English annual reports in Europe

Transcription

1 English annual reports in Europe A study on the identifiation and reeption of genre harateristis in multimodal annual reports originating in the Netherlands and in the United Kingdom

2 Published by LOT phone: Janskerkhof 13 fax: BL Utreht The Netherlands Cover illustration: Meta Niemeijer ISBN NUR 616 Copyright 2008: Elizabeth B. de Groot all rights reserved.

3 English annual reports in Europe A study on the identifiation and reeption of genre harateristis in multimodal annual reports originating in the Netherlands and in the United Kingdom een wetenshappelijke proeve op het gebied van de Letteren Proefshrift ter verkrijging van de graad van dotor aan de Radboud Universiteit Nijmegen op gezag van de retor magnifius prof. mr. S.C.J.J. Kortmann, volgens besluit van het College van Deanen in het openbaar te verdedigen op woensdag 11 juni 2008 om uur preies door Elizabeth Bouhien de Groot geboren op 23 september 1976 te Delfzijl

4 Promotores: Copromotor: Prof. dr. M. Gerritsen Prof. dr. C. Nikerson (Swiss Graduate Shool of Management Studies, Bangalore, India) Dr. H.P.L.M. Korzilius Manusriptommissie: Prof. dr. V.K. Bhatia (City University of Hong Kong) Prof. dr. P. Gillaerts (Lessius Hogeshool, Antwerpen) Prof. dr. A.M. Smelik

5 Foreword Dear readers, Based on the researh results of the present genre study, it would seem to be appropriate to start this Duth-English foreword with a salutation to those who have opened the book. In this foreword, I would like to aknowledge all the people that have made my PhD projet a joyful and valuable experiene. First, I would like to thank Marinel Gerritsen, Hubert Korzilius and Catherine Nikerson for their exellent supervision, in terms of researh ativities and thesis write-up but also in terms of personal motivation. I believe I was part of a wonderful team inluding members with omplementary expertise. Marinel, you always kept a general yet very lear view on the main issues in the researh projet and taught me what researh planning is all about. Hubert, your lessons in statistis helped me to establish a link between qualitative analyses and quantitative researh tehniques. And Catherine, even at a distane your insights in genre theory were enlightening and enabled me to refine my theoretial and analytial frameworks. Thank you all for sharing your knowledge and enthusiasm in this projet. Next, I would like to thank the people who have ontributed to different phases in the study, either by their atual partiipation or by their failitation of researh ativities. I am most grateful to Marle Smit, Christian Burgers, Tammy Costamallo and Sanne Heymann for thinking along and assisting in the interoder reliability tests. Likewise, I spent many interesting hours in disussion with the Duth and British informants that were involved in the in-depth interviews about annual reports. Although I annot mention their names here, I am very thankful for their time and information. I would also like to thank the finanial analysts in London who either ompleted my questionnaires in the response analysis or who provided me with a ontrolled researh setting for this analysis. Aordingly, I would like to express my partiular gratitude to Reinier de Jong, who helped me set up the response analysis in the exiting area of Canary Wharf. Throughout the past four years, my working environment formed a major impulse for my researh ativities. I am very grateful to my olleagues within the department of Business Communiation Studies for their support and interest in the researh projet. In partiular, I would like to thank Hans Hoeken and Carel Jansen for their ontinuous onfidene and enouragement, and for allowing me some extra time in ompletion of my study. Jos Hornikx, Suzanne de Bakker and Renske van Enshot- Van Dijk, I greatly appreiate your feedbak about onduting, writing and publishing a PhD study. At times, it was a relief to know we had shared experienes in our lives as PhD students. Furthermore, I would like to thank Berna Hendriks and Brigitte Planken for their support and insights into professional disourse in English. Brigitte, thank you also for lending me your room at the beginning of my researh projet; in addition to the helpfulness of olleagues, this made it even easier for me to transform from an outsider into an insider in our department.

6 Other people have ontributed in their own ways to the ompletion of this study. Speial thanks go to Gisela Redeker, who informed me about the researh proposal on English annual reports in 2003 and who enouraged me to further explore my researh skills. Moreover, I would like express my gratitude to Ken Hyland, Vijay Bhatia and John Swales for sharing their expertise in the fields of metadisourse, genre analysis and disourse ommunities, respetively. Their wisdom formed a rih soure of information for the present study. Finally, I would like to thank my friends and family. During my time as a PhD student they remained nearby to share pleasurable moments, to put things in perspetive and to motivate me into making the researh projet a suess. In partiular, I would like to thank Greetje Gompelman and Patriia Langhout for their patiene and persistene with regard to our friendship. I am proud to have them as my paranimfen. Similarly, I would like to thank Erik-Jan Bennema, who kept enouraging me during our numerous offee breaks in the university library. As for family and family-in-law, I would like to thank them for their unonditional support and understanding throughout the past years. Even in times when I was preoupied with work and hardly approahable, they kept providing amiable and omfortable homes. Two people in partiular ontributed to the realization of this dissertation: they are my mom and dad, who granted me a valuable eduation and taught me to work hard for the things I wanted to ahieve in life. Lieve pap en mam, jullie hebben eigenlijk de basis gelegd voor dit proefshrift. Ik ben bijzonder dankbaar voor het feit dat jullie me hebben laten studeren en voor de onvoorwaardelijk steun die je me hebt geshonken bij de keuzes die ik daarin en daarna maakte. Als je je best maar doet, was jullie motto. Dat heb ik de afgelopen jaren gedaan, met dit boekwerk als resultaat. Ik ben er trots op jullie als ouders te hebben! The last but ertainly not the least person I would like to aknowledge in this foreword is Mathieu Niemeijer, the love of my life and my best friend. Mathieu, I am very grateful for your presene and ontinuous support during the past four years, and for your interesting views on my researh topi. Sometimes it seemed as if we were hardly together, when you pluked me from behind my desk at the middle of the night or dropped me off at the university library on Saturday and Sunday mornings. But we were and we will be. Zwolle, maart 2008 Elizabeth de Groot

7 Contents 1. Introdution: English in European business English in ross-ultural and interultural business disourse The status of English in the Duth business ommunity The urrent annual report as a business ommuniation tool Comparing English annual reports by Duth and British multinationals Aims and researh questions Overview of this genre study Theoretial sope: text as rhetorial genre Soial onstrutionism Community and disourse A definition of professional genre Genres as typified soial ations The rhetorial purpose of genres Strategi intention in genres Content, move struture and lexio-grammar of genres Networks of genres The annual report as a rhetorial genre Conlusion Researh design for the ross-ultural study of English annual reports Generi haraterization Approah to multimodal text analysis Seletion of annual reports and speifi texts Cross-ultural data analysis: qualitative and quantitative desriptions Interoder reliability tests for ross-ultural analyses of enre features Charaterization of the genre ontext of annual reports in the Netherlands and in the United Kingdom A working definition of genre ontext: disourse ommunity Contextualizing annual reports in the Netherlands and the UK Design of the qualitative survey in Duth orporations Design of the in-depth interviews in Duth and British orporations Results English language poliy for Duth-based annual report Results ommuniative ontext Duth-based and UK-based annual reports General ommuniative purposes of the annual report Reader audiene of the annual report Guidelines for the ontents of the annual report Communiative objetives per setion in the annual report English language-poliy for the (statutory) annual report Guidelines for and purposes of photographs in the annual report Prodution proess of the annual report Conlusion and disussion Cross-ultural omparison of ontextual elements Impliations for genre theory and genre analysis 89

8 5. Charaterization of the genre ontent of Duth-English and British-English annual reports A working definition of genre ontent: oneptual themes Coneptual themes in written texts Coneptual themes in photographs Analyzing themes in Duth-English and British annual report texts Design for the identifiation of text themes Design for the identifiation of photo themes List of text themes and photo themes in annual reports Interoder reliability test for text themes and photographi themes Themes in management statements Themes in the running texts of the management statements Themes in photos and text fragments of the management statements Themes in orporate profiles Themes in the running texts of the orporate profiles Themes in photos and text fragments of the orporate profiles Themes in operational reviews Themes in the running texts of the operational reviews Themes in the photos and text fragments of the operational reviews Conlusion and disussion Cross-ultural omparison of ontent elements Impliations for genre theory and genre analysis Charaterization of the rhetorial genre struture of Duth-English and British-English annual reports A working definition of genre struture: moves and strategies Steps versus strategies The differene between moves and themes in rhetorial genre struture The demaration of moves and strategies in written texts Photographs in rhetorial genre struture Analyzing moves and strategies in Duth-English and British annual report texts Design for the identifiation of textual moves and strategies Design for the identifiation of visual moves and strategies List of textual and visual moves and strategies in annual reports Interoder reliability test for textual and visual moves and strategies Moves and strategies in management statements Moves and strategies in the running texts of the management statements Moves and strategies in photos and text fragments of the management statements Moves and strategies in orporate profiles Moves and strategies in the running texts of the orporate profiles Moves and strategies in photos and text fragements of the orporate profiles 152

9 6.5 Moves and strategies in operational reviews Moves and strategies in the running texts of the operational reviews Moves and strategies in photos and text fragments of the operational reviews Conlusion and disussion Cross-ultural omparison of strutural elements Impliations for genre theory and genre analysis Charaterization of the generi lexio-grammar of Duth-English and British-English annual reports A working definition of generi lexio-grammar: metadisourse The (on)text-speifi realization of metadisourse Photographs in the analysis of metadisourse Analyzing metadisourse in Duth-English and British annual report texts Design for the identifiation of textual metadisourse Design for the identifiation of visual metadisourse Integrated model of textual and visual metadisourse for annual reports Interoder reliability tests for metadisourse in texts and photographs Metadisourse in management statements Metadisourse in the running texts of the management statements Metadisourse in photos and text fragments of the management statements Metadisourse in orporate profiles Metadisourse in the running texts of the orporate profiles Metadisourse in photos and text fragments of the orporate profiles Metadisourse in operational reviews Metadisourse in the running texts of the operational reviews Metadisourse in photos and text fragments of the operational reviews Conlusion and disussion Cross-ultural omparison of metadisourse elements Impliations for genre theory and genre analysis Conlusion ross-ultural genre analyses Genre ontext Genre ontent Genre struture Generi lexio-grammar The English annual report as a global genre set Impliations for genre theory Impliations for genre analysis Limitations of the ross-ultural genre study and suggestions for further researh 245

10 9. Reader response to Duth-English and British-English annual reports Genre studies and text reeption Hypotheses for the interultural response analysis Method interultural reader response analysis Partiipants Materials Dependent variables Pretest of the questionnaire Design and proedure Statistial analyses Results of the experimental reader response analysis H1: effets on pereptions of orporate image H2-4: effets on pereptions of persuasiveness of the annual report text H5: effets on general preferenes for written texts and photographs RQ: the use of annual report texts Conlusion and disussion Interultural effetiveness of genre features in English annual reports Impliations for genre theory, genre analysis and genre pratie Limitations of the interultural response study and suggestions for further researh Closing remarks 283 Referenes 285 Appendix I: text ontent management statements 303 Appendix II: text ontent orporate profiles 305 Appendix III: text ontent operational reviews 307 Appendix IV: text struture management statements 309 Appendix V: text struture orporate profiles 310 Appendix VI: text struture operational reviews 311 Appendix VII: lexio-grammar text management statements 312 Appendix VIII: lexio-grammar text orporate profiles 313 Appendix IX: lexio-grammar text operational reviews 314 Appendix X: questionnaire experiment 315 Samenvatting 325 Curriulum Vitae 338

11 Chapter 1 Introdution: English in European business Nowadays, profiieny in English would seem to be inherent in the notion of good ommerial prospets aross the borders of European nations. Both outside and within Europe, the signifiane of English as an international business language has grown ever sine the seond world war, after whih the teahing of English inreased exponentially (Brumfit, 2001) and the language aquired eonoultural funtions, meaning that it beame the produt of the development of a world market and global developments in the fields of siene, tehnology, ulture and the media (Seidlhofer, 2003: 9). As for Europe in partiular, the interest in English experiened a boost during and after the 1990s, most likely due to the establishment of the European Union and thus to the eonomi integration of European ountries (Berns, 1995; Phillipson, 2003; Seidlhofer, Breiteneder & Pitzl, 2006). Despite the emphasis on language diversity and demorati language poliies by the Language Poliy Division of the European Counil (Language Poliy Division, 2007), English has gained a dominant position as an international language of operation in ommerial and politial ontexts in Europe. Consequently, European multinationals now fous on English in their international advertising ampaigns or international negotiations and EU-institutions suh as the European Central Bank have taitly adopted English as the main means of ommuniation (e.g. Gerritsen, et al., 2007; Seidlhofer et al., 2006). The prominent status of English in general has been the topi of muh debate. Some fear that the spread of English will ause a language problem, threatening to engulf and replae indigenous European languages (Rogerson- Revell, 2007: 106). From this perspetive, an English-only language situation in Europe seems inevitable (Van Els, 2000). Others ritiize the idea of English-as-alanguage problem and suggest that it is more realisti to expet an Englishgenerally language situation in Europe. They argue that Europeans will not need to hoose between their own native language and English; people an use their native language within their national or language ommunities and English as their European language (Van Els, 2000; Rogerson-Revell, 2007). Whatever perspetive is taken, however, the inreasingly important status of English within Europe is believed to be an irreversible fat. Interestingly, the prominene of English as a foreign or international language is not pereived to be the same throughout Europe. In the Netherlands and in the Sandinavian ountries, for example, English language developments seem to have been partiularly powerful. The 2005 Eurobarometer on languages in Europe (2006) showed that more than 85% of the partiipants in the Netherlands, Sweden and Denmark laimed that they spoke English well enough to hold a onversation. Booij (2001) and Graddol (2000), moreover, argue that suh researh results indiate a transition from English as a foreign language to English as a seond language in the Netherlands and in Sandinavia. Presumably, the high profile of English in these ountries is owing to their relatively small number of loal language speakers in ombination with their historial dependene on international trade and ooperation (e.g. Graddol, 2006; Hoffmann, 2000). The relevane of English as an international language is espeially evident in European business settings. The growing number of publiations on English in European ompanies suggests that English has beome a major international

12 business language aross Europe (e.g. Gerritsen et al., 2007; Rogerson-Revell, 2007). The spread of English throughout Europe and its strong position in international business an be explained in terms of either linguisti imperialism or pragmatism. Aording to Phillipson (1998), linguisti imperialism is a hegemoni phenomenon. It is a legay of the British olonial period, as it onerns the establishment and ontinuous reonstitution of strutural and ultural inequalities between English and other languages (Phillipson, 1998: 104). Linguisti imperialism implies that the native-like use of business English is more benefiial than other language variants, beause it gives aess to material and assoiative supremay, e.g. finanial alloations and influene or images of professionalism and modernism. Conversely, the pragmati motivation for the widespread interest in English fouses on the effiieny, relevane and eonomy in the use of the international language. Here, the funtionality of English is regarded as utilitarian, i.e. important for international business, and idealisti, i.e. potentially allowing for further ross-border ommuniation and mutual understanding (Seidlhofer, 2003). From this view, the use of business English in Europe primarily aims to failitate international interations and does not involve a drive for native profiieny in the foreign language. In reent years, the pragmati perspetive has gained onsiderable support in the field of English for professional international ommuniation. As far as internal orporate ontats are onerned, for example, many European multinationals have deided to apply English as an international working language not tied to any national base. This is illustrated by the German ompany Siemens or the German-Frenh ompany Aventis, whih introdued English as the offiial orporate language, and by the Duth ompany ING, whih presently refers to English as the offiial medium of ommuniation within their Duth headquarters (Daelemans, 2005; The Eonomist, 2004; Truhot, 2002). In these ases, English enables business people to leverage international allianes and it failitates the ommuniation or working proesses between affiliates from potentially different language bakgrounds. Additionally, ases in whih English has been ruial to the external ommuniation of European multinationals are the Finnish forest industry group Metsä-Serla hanging its name to M-real or the Duth onglomerate Philips inluding English-language slogans in their international advertisements, i.e. Sense and simpliity (Taavitsainen & Pahta, 2003; Philips, 2004). Suh strategi deisions are generally aimed at expanding and maintaining the orporation s foreign markets, at emphasizing the global presene of the business and at establishing a orporate identity that is omprehensible for international ustomers (Taavitsainen & Pahta, 2003; Truhot, 2002). Notwithstanding the inreasing popularity of the pragmati use of international English, there are other instanes, however, in whih the symboli meaning of English implied in Phillipson s linguisti imperialism still seems to prevail as the most important reason for its use. As an illustration, several studies have shown that Spaniards tend to assoiate the English language with (inter)national soial and professional prestige, presumably as a result of their politial history in whih English has long been a suppressed foreign language. Smith (1997: 22) remarks that Spanish businessmen feel a need to demonstrate their mastery of ommerial English to position themselves as respetable, internationally mobile professionals. Furthermore, investigations by Gerritsen, et al. (2007) and Piller (2001) indiate that marketers promoting (inter)national produts in the Netherlands and in Germany use English to improve the pereived status of these produts. More speifially, Gerritsen et al. (2007) and Piller (2001) find that

13 English in TV-ommerials and print ads tends to be assoiated with symboli onsumer harateristis suh as modernity, sophistiation, dynamism and globalization. In multinational organizations or negotiations, the pragmati or imperialisti use of English may involve both native and non-native variants of English, i.e. professionals may depend on English as their mother tongue (EMT) or English as their lingua frana (ELF). ELF intermediates between non-native speakers of English who do not have a shared language bakground; it enables these people to onnet based on ommon interests and onerns aross languages and ommunities (Seidlhofer et al., 2006). Sine transations throughout orporate Europe an in priniple involve both EMT and ELF speakers, it may be more orret to refer to English as an international language (EIL) in these situations. MKay explains that EIL or International English is used by native speakers of English and bilingual users of English for ross-ultural ommuniation. International English an be used both in a loal sense between speakers of diverse ultures and languages within one ountry and in a global sense between speakers from different ountries (2002, 132). The present study will fous on the use of English in an area where it has aquired a high profile as an international (business) language: English produed in the Netherlands. More speifially, it will explore Duth-English annual report texts originating in large Duth multinationals and will ompare those to native English annual report texts produed in a different European nation, i.e. written in British orporations. Before elaborating on the researh aims and questions of this investigation in the last setion of this hapter (1.5), the following subsetions will first introdue the study s prinipal motivations: the relevane of a ombination of ross-ultural and interultural researh (1.1), the status of English in Duth orporate life (1.2), the position of the annual report as a business ommuniation instrument (1.3) and the importane of omparing the Duth and British business ultures (1.4). 1.1 English in ross-ultural and interultural business disourse The first basi motivation of this study lies in the belief that a multiultural approah to English business disourse is best served by a ombination of ross-ultural and interultural perspetives. Here, business disourse is defined as talk and writing between individuals whose main work ativities and interests are in the domain of business and who ome together for the purpose of doing business (Bargiela- Chiappini & Nikerson, 1999: 2). The spread of English as the dominant language for European business relations implies a omplex ommuniation proess, where Europeans who are part of fundamentally different national or language ultures ome together in many situations where they need to reah a mutual understanding of the aims of their interation in a language that may not be their first language. Aordingly, the multiultural (European) setting in whih English is used as a native or non-native language may well influene the outome of the business disourse. This is illustrated by Shaw, et al. (2004), who ondut an international study on ustomer omplaint dialogues in English. They show that in omparison to north-west Europeans from Belgium, Denmark, Sweden and the UK, Italians have substantially different pereptions of what is preferable and aeptable problemsolving disourse. For instane, north-west Europeans showed a signifiantly

14 stronger preferene for straightforward but sinere and relational dialogues than Italians, whereas the Italians did not reveal a lear preferene for one type of dialogue but displayed an equal appreiation for all types of dialogues that were non-minimalist, i.e. relatively long and polite. Shaw et al. (2004) onlude that there may be national ulture norms that influene business disourse in terms of omplaining, apologizing and omplaint handling, and the fat that texts are in English does not override these norms. Indeed, this outome suggests that both native and non-native speakers of English tend to interat in aordane with the soio-ultural norms whih govern the use of their own first language (Vandermeeren, 1999: 275). With regard to disourse in orporate ontexts, the oneptualization of terms suh as ulture or ultural norms is omprehensive and ontinues to be a topi of multidisiplinary debate (Harris & Bargiela-Chiappini, 1997; Bargiela- Chiappini & Nikerson, 2003). Business disourse an be regarded from various ultural dimensions that eah affet the typial disursive praties in a speifi business setting. Generally, a distintion is made between national or regional ulture, and orporate or organizational ulture (e.g. Bargiela-Chiappini & Nikerson, 2002; Suhan & Dulek, 1998; Ulijn, O'Hair, Weggeman, Ledlow & Hall, 2000). National ulture is a maro-level onept overing general disourse onventions typial to a ompany s region of operation; it omprises the inherited olletive programming that distinguishes one group or ategory of individuals sharing a partiular nationality, e.g. values, attitudes and behaviors, from another group (Claes & Gerritsen, 2007; Hofstede, 2001; Morden, 1995, 1999). A ompany s orporate ulture, on the other hand, influenes the prodution and interpretation of business disourse on a more organization-speifi level and enompasses fators suh as orporate size, orporate ativity and struture, (ommuniation) tehnology and methods of ontrol (Bargiela-Chiappini & Nikerson, 2002; Suhan & Dulek, 1998). In business ommuniation studies, national ulture and orporate ulture are not regarded as stati onepts, but are believed to interplay: values and norms inherent to the national ulture may influene or determine the ulture of a partiular orporation (Bargiela-Chiappini & Nikerson, 2002; Ulijn et al., 2000). Fators like organizational struture or ontrolling methods are often tied to the ontemporary harateristis of the national ulture in whih the ompany is situated. With respet to orporate struture, for example, national governane regulations require a two-tier system in the Netherlands, where exeutives and non-exeutives are part of two separate boards (i.e. the Exeutive Board and the Supervisory Board). Compared to the Anglo-Saxon one-tier system, in whih exeutives and non-exeutives reside in one management board (i.e. the Board of Diretors), the Duth organizational struture has a distint impat on the way in whih the internal deision-making proesses operate, e.g. the appointment of non-exeutive diretors (Hooghiemstra & Van Manen, 2004). In this example, features of the national ulture oinide with orporate norms, all of whih affet the disursive praties adopted in the deision-making proess. Varner (2000) proposes an additional ultural level interfaing the two onepts of national ulture and orporate ulture : the national business ulture. She explains that while national ulture reflets the values and onventions inherent to a given soiety (e.g. involving preferene levels for individualism or power distane), business ulture exists within this larger ulture and only omprises the soial norms relevant for business relations. As suh, it inludes the ommonly preferred or aeptable business objetives, strutures and poliies (e.g. with regard to human resoure or finanial management). Finally, orporate ulture ensues from

15 the national business ulture in the sense that it onerns the onnetion between generally preferred business behaviors and the development of ompany-speifi strategies, information flows and deision-making ativities (Varner 2000, 2001). Again referring to the above example on two-tier and one-tier organizations, it ould be argued that variations in governane struture are owing to differenes in the Duth and Anglo-Saxon business ultures, whih in turn may be based on higherorder values in the Duth and British soieties. Sine the present study seeks to explore the general use of English in annual reports produed by diverse Duth and British orporations (see also setion 1.5), it fouses on the onepts of national ulture and general business ulture. As suh, it aounts for the fat that the disourse analyzed is developed in different ommerial ommunities that eah belong to a larger, distint national system. Following Varner (2000, 2001), inluding both onepts in the investigation of annual reports in priniple would seem to offer the fullest view on the fators possibly shaping the disourse in both nations; omitting either national or business ulture as a researh objet ould limit the number of aspets potentially relevant in determining the impat of the ultural ontext on text. Researh involving multiple ultures and their speifi use of (business) disourse an be onduted in two ways: from a ross-ultural approah or from an interultural perspetive. A ross-ultural view on disourse ommonly omprises omparative researh into the similarities and differenes in ommuniative habits aross distint ultures. It involves the separate examination of ulture-speifi tendenies and, subsequently, a ontrastive study to determine to what degree these tendenies differ (e.g. Claes & Gerritsen, 2002; Gudykunst, 2003; Harris & Bargiela-Chiappini, 1997). Cross-ultural omparisons in general have been the topi of many business ommuniation studies. Examples of ross-ultural analyses of English business disourse an be found in Ponini (2004), who examines and ompares linguisti features in English promotional texts for Italian, Australian and Amerian wineries or wine regions and in Upton and Connor (2001), who explore and ontrast the rhetorial struture and politeness strategies in English appliation letters by Amerians, Finns and Belgians. Interultural studies generally fous on the disourse of people from different ultures who engage in ommuniation. In partiular, the interultural approah to business disourse onerns studying how two or more individuals (or groups of individuals) from different ultures interat in a partiular work-related enounter (Harris & Bargiela-Chiappini, 1997: 6). Thus, interultural researh exposes the effets of possible similarities and variations in ommuniative behavior between ultures on the interation between agents from these different ultures. In that sense, it is distint from the ross-ultural perspetive on disourse, whih exlusively involves the identifiation of suh similarities and variations per ulture. Relatively few studies have onentrated on interultural business ommuniation researh let alone interultural researh on European English business disourse. As suggested by Bargiela-Chiappini and Nikerson (2003) and Harris and Bargiela- Chiappini (2003), this is perhaps aused by the fat that it is still a relatively new and evolving aademi field. Interultural disourse studies have been onduted, for instane, by Louhiala-Salminen, Charles and Kankaanranta (2005), who examine the interultural pereptions and English disourse in meetings and s within a Swedish-Finnish merger and by Planken (2005), who analyzes faework in English sales negotiations among native speakers of English, Duth, Frenh, Italian, Japanese and German. In these situations, where all the members of the multinational group ommuniate in English, interultural researh seems

16 partiularly relevant. As Rogerson-Revell explains: While people may well need to speak the same language in suh multilingual ontexts, they may not neessarily speak the same way (2007: 118). Interultural researh enables analysts to pinpoint if and how different ways of using the same language atually ause diffiulties in soial or ommuniative relationships. Gudykunst (2003) argues that ross-ultural ommuniation studies should preede (i.e. are a prerequisite for) interultural analyses. The above desriptions of both researh perspetives support this perspetive: before being able to evaluate the interation between partiipants from different ultural bakgrounds, analysts need to know in what manner their bakgrounds may have shaped the interational habits of eah of these partiipants. As will be referred to later in this hapter, the present researh projet inludes both ross-ultural analyses i.e. ulture-speifi text analyses and an interultural investigation i.e. the exhange of Duth-English texts with potential UK-based readers. It starts with the ross-ultural omparison of annual report texts and then moves on to an interultural reader response analysis of these texts. These analyses are onduted to reveal whether the Duth-English and British texts enlose a similar way of using language and, subsequently, what the effets of any variations in language use are on the UK-based audiene of these texts. 1.2 The status of English in the Duth business ommunity Apart from the importane of ross-ultural and interultural views on business disourse, another reason for the present study relates to the growing status of English in the Duth business ontext. In general, the use of English has beome a ruial element in disourse praties within the Duth professional and eduational spheres. In line with the general trend, business English has gained momentum both for national ommerial purposes, to ommuniate with Duth lients, and for international objetives, where English funtions as an international language to ommuniate with foreign olleagues or lients. As for the use of English on a national sale, for instane, it is prominent in billboard advertisements, in printed magazine ads and TV ommerials for produt promotion and in newspaper and online job advertisements (e.g. Gerritsen & Nikerson, 2004; Gerritsen et al., 2007; Van Meurs, Korzilius, Planken & Fairley, 2007). Furthermore, the use of Duth-based English as an international language in ross-border interation has gained importane as the result of the inreasing internationalization of traditionally Duth orporations. Several strategi proesses in partiular have ontributed to the growing signifiane of Duth-English as a lingua frana. For example, a number of ompanies of Duth origin, e.g. Koninklijke Olie (Shell), Margarine Unie (Unilever), Elsevier (Reed Elsevier), Koninklijke Hoogovens (Corus) and CMG (LogiaCMG), have entered into mergers with Anglo-Saxon multinationals and have adopted English as the primary medium of ommuniation in order to failitate internal working relations (e.g. Nikerson, 2000). In addition, Duth multinationals suh as AEGON, AHOLD, ING and Wolters Kluwer have strengthened their international position by employing board members or even CEOs from outside the Netherlands. These orporate situations have also aused a greater need for English as an internal international language, at least in the exeutive offies where the multinational board is seated (e.g. Daelemans, 2005). Finally, several Duth ompanies with headquarters in the

17 Netherlands rely on English as the main external language for the expansion and retention of their European or global markets. For instane, the teleom ompany KPN and the animal nutrition ompany Nutreo Holding maintain English-only websites for their non-duth stakeholders. And Ten Cate, a large materials manufaturer, positions itself worldwide with an English-language slogan promising Materials that make a differene (e.g. Nikerson & De Groot, 2005). Another external ommuniation tool that has been affeted by the emergene of English as the main international business language for Duth orporations, is the annual general report. In his 2001 study on languages of aount in Europe, Parker suggests that multinational ompanies based in ountries haraterized by a relatively small language ommunity (e.g. Duth) may deide to keep aounting reords in English [ ] to ommuniate to the rest of the world and he ontinues to explain that in the Netherlands foreign ompanies may file finanial statements in Duth, English, Frenh or German (2001: 140, 144, emphasis EBdG). Five years on, the English language situation for annual reports originating in Duth multinationals would seem to have beome quite different, suh that over the past four years in partiular, the English language has beome of ruial importane in the finanial ommuniation disourse pratied by traditionally Duth multinationals. From 2003 to 2006, major listed orporations like Philips, AEGON, AHOLD, KPN, Heineken, ING, Rodamo, TPG/TNT, Van der Moolen and Wolters Kluwer gained substantial shareholder support for the introdution of an offiial English-language annual general report, i.e. the version of the report that is delivered to the Duth Chamber of Commere and is legally binding. These developments learly mark a shift in Duth orporations English language poliy on annual reports (Nikerson & De Groot, 2005). As suh, they drew ritial media attention when the new poliies were first implemented, as testified to by headlines like Philips outlaws mother tongue for annual report (de Volkskrant, 2003) or Protests against AEGON s abolition of Duth annual report (Communiatie Online, 2004). Today, English-oriented shifts in language poliies for annual reports of Duth multinationals are publily aepted as simply a neessity, sine English is believed to failitate partiipation on the international markets. However, the ommuniative onsequenes of these poliy hanges have reeived hardly any onsideration, neither from a textual viewpoint nor from a response perspetive. Therefore, the present investigation seeks to gain further insight in the design and effetiveness of the English annual reports originating in Duth listed multinationals. It will provide text desriptions of Duth-English annual reports, and will ompare these to desriptions of native British annual reports. Subsequently, this study will explore the effet of the Duth-English and British annual report texts from the perspetive of UK-based stakeholders in partiular. 1.3 The urrent annual report as a business ommuniation tool The previous subsetions referred to the relevane of ross-ultural and interultural views on business disourse (setion 1.1) and the importane of English for Duth orporations (setion 1.2) as primary reasons for the present researh projet. Another main reason involves the omplex and evolving nature of the annual report as a multimodal business ommuniation tool. With regard to their ommuniative funtion, first of all, orporate annual reports are omprehensive douments that

18 have often formed ontroversial researh materials (Thomas, 1997). They generally onsist of a series of individual omponent texts of different lengths and on various topis, and exist in a network of omplementary instruments like preliminary announements and analyst presentations (Garzone, 2005; Rutherford, 2005). As suh, some have haraterized the annual report as a text type in itself, inluding several sub-text types like the CEO s letter to shareholders (e.g. Garzone, 2004, 2005). Others have identified the annual report as a broader aounting narration text type; this means that it resembles text types like interim reports or finanial press releases and ontains more speifi text types like the operating and finanial review (Rutherford, 2005). Despite the ontroversy surrounding the speifi haraterization of annual report texts, analysts have repeatedly tried to establish the annual report as a text type in its entirety. Conerning audiene, for example, studies on annual reports by European-based and Amerian-based ompanies have reurrently suggested that the primary reader group of these reports omprises (institutional) shareholders, whereas the seondary target groups inlude readers suh as finanial analysts, journalists, bankers or reditors, orporate employees and staff (Anderson & Imperia, 1992; Hyland, 1998b; Jameson, 2000; David, 2001; Garzone, 2004). As for the main ommuniative purposes of the annual report, these studies have argued that it is to provide information about the ompany s finanial or operational performane and to support the ompany s impression management by onveying its personality or philosophy and expressing onfidene in its future prospets (Anderson & Imperia, 1992; Garzone, 2004; Hyland, 1998; Malavasi, 2005). In reent years, worldwide aounting inidents e.g. involving Enron or AHOLD have evoked debates about the redibility of the finanial statements in annual reports. As a result, finanial aounting authorities have introdued more stringent politial regulations that restrit freedom in aounting figures but add to the transpareny of the annual report texts in relation to the finanial aounts (e.g. Rutherford, 2005). Within Europe, for instane, the European Union Aounting Regulation determined that all orporations listed in a European Seurities market were to adopt the International Finanial Reporting Standards (IFRS) from 1 January 2005 onwards. These standards equate the design of European orporations finanial statements and fous on profit quality rather than profit height (IASplus, 2005). For the annual reports of European multinationals the onsequene of the IFRS is threefold. First, they failitate international omparisons of finanial performane and enlarge the hane of deteting any irregularities (IASplus, 2005). Seond, the IFRS stimulate the use of English for the drawing up of the orporate annual report. Beause the IFRS were originally published in English for international distribution, they have provided support for the fat that English is the most important international finanial language and is thus most onvenient for the international use and evaluation of the annual report (e.g. ING, 2006; Van der Moolen, 2006). And third, sine these standards prevent ompetitive positioning on the basis of finanial aounts, European orporations will be inreasingly reliant on the non-finanial texts within the annual reports, as a means of distinguishing themselves from their ompetitors (Nikerson & De Groot, 2005: 328). From a disursive viewpoint, the last two effets have had or are likely to have a onsiderable impat on the prodution of the annual report texts and their ommuniative funtions. As an example, several listed Duth multinationals have prioritized English as the base language for their annual report, partially in order to failitate ompliane with the IFRS and display an orientation towards the international finanial markets (e.g. ING, 2006; Wolters Kluwer, 2006).

19 Another harateristi of ontemporary annual reports onerns their multimodal nature (Kress & Van Leeuwen, 2001). While the annual report was initially used as a dislosure instrument for fatual finanial figures, today it has evolved into an elaborate investor relations tool whih ontains both written texts and visual illustrations (Bexley & Hynes, 2003). With respet to the multimodality of reent annual reports, the textual mode is often ombined with and omplemented by the photographi visual mode, as opposed to the shemati or diagrammi visual mode; many of the non-finanial texts in annual reports, suh as over pages, management forewords, operational reviews or diretors biographies, are now supported by photographs. It has been argued that the inreased use of photographi support in orporate publiations stems from the need to establish a tangible representation of the ompany, beause visible presene has funtioned as a traditionally aepted prerequisite for authentiity a quality widely onsidered ruial for orporate legitimay (Guthey & Jakson, 2005: 1058). Hene, photographs in annual reports are not only expeted to failitate the reading proess but are also believed to enhane ways of knowing the truth of the ompany, both in relation to the textual details and as individual representations of the ompany s soial ideologies (Benshop & Meihuizen, 2002; David, 2001; Graves, Flesher & Jordan, 1996; Preston, Wright & Young, 1996). The reprodution of orporate truth in annual reports is generally established through the integration of photographs of real things in natural settings or doumentary photographs whih onveniently oneal some of the tehniques the reators use to represent the truth (David, 2001: 209). For instane, an annual report of a shopping mall developer may ontain images of buildings in spotlights, photographed at night. Although these realisti pitures aentuate the beauty and serenity of the malls for the readers of the annual report, they are hardly onsistent with the ongestion suh enters attrat in daily life (David, 2001). In other words, the rhetorial value of photographs in annual general reports lies in their simulated transpareny, whih supposedly onvines the readers of the natural, ideologial truth that resides in the operational and finanial aounts of the ompany (Graves et al., 1996; Guthey & Jakson, 2005). The above onsiderations about the ontemporary annual report indiate that this doument has been aknowledged as an international and versatile business ommuniation instrument. In general, it is believed to serve informational as well as promotional purposes for a range of disourse partiipants. These purposes, moreover, are ommonly realized through both the written texts and the photographs in the annual report. It was also disussed that the annual report either as a speifi text type or a broader aounting narrative is ommonly identified as a oherent series of texts that onstitute the full report and math its aims and audienes. The ombination of these general features suggests that the annual report an be defined as a multimodal genre of organizational ommuniation (Yates & Orlikowski, 1992). It is this text perspetive that will underpin all the desriptive and response analyses aounted for in the present study. The genre onept will return later in setion 1.5 and in partiular in Chapter 2. Here, it will be further defined as a primary starting-point for the investigation of multimodal, non-finanial Duth- English and British-English annual report texts.

20 1.4 Comparing English annual reports by Duth and British multinationals So far, this study has been motivated by the relevane of ross-ultural and interultural disourse analysis, the relevane of English for Duth multinationals and the relevane of the annual report as a multimodal genre of organizational ommuniation. Its final motivation lies in the assumption that a omparison of Duth-English and British annual reports provides a relevant researh example, in geographial terms but also with respet to the speifi genre analyzed. First, within the European Union registered ompanies in the Duth business ommunity seem to math partiularly well with British ompanies quoted on the London stok exhange, beause both the Duth and UK eonomies inlude large and mid-size multinationals operating in similar markets or industries and maintaining similar finanial positions (Forbes, 2003; Nikerson & De Groot, 2005). Sometimes this has led to fiere international ompetition, as is illustrated by a reent press release of TNT: due to ontrat wins with large British ompanies suh as Centria, BT and Sainsbury s, TNT is now Royal Mail s main ompetitor in the UK (TNT, 2007). In other ases, similar ommerial goals have stimulated Duth and British orporations to engage in global trading relationships or major allianes (e.g. Shell, Unilever, et.; see setion 1.2). Here, reasons of shared entrepreneurial heritage, workable ultural differenes and high levels of English flueny by the Duth were believed to ditate ommerial and strutural suess (Lesquillier, 2007; Pfanner; 2006). Geographially, the Duth and British business ommunities an thus be onsidered as ompetitive and omparable ontexts in whih English is applied as the dominant international business language. With regard to the international annual report, Duth and British orporations prinipally rely on a generially similar doument written in the same language (i.e. English) to ommuniate with their foreign stakeholders. Sine different national and business ultures are involved here, Duth and British ompanies may draw on distint disourse onventions and then impose these upon English as the international business language in the annual report. Similar ultural influenes on professional English disourse have been pointed out and evidened in previous ross-ultural studies on English appliation letters by Amerians, Belgians and Finns (Upton & Connor, 2001), on English internal s by British and Duth professionals (Nikerson, 2000), or on student-professor s by students of 34 nationalities (Bjørge, 2007). The idea that Duth and British ultural influenes ould be at play in the prodution and interpretation of the English annual report is further strengthened by past experienes in and theorizations about Anglo-Duth business enounters. In suh English-language interations, professionals have sometimes met with ulturespeifi ommuniation norms that resulted in misunderstandings, mutual bias and, ultimately, unfavorable business onsequenes. A former Strategy diretor at Koninklijke Hoogovens (i.e. the Duth partiipant in the Corus merger), for instane, partially blames Corus organizational turmoil on the dominant, authoritarian and vague ommuniation of British managers overruling their Duth olleagues (Lesquillier, 2004). On the other hand, a former Shell onsultant explains that the demorati but diret approah to orporate ommuniations by the Duth tends to ause frition when used with the British, who seem to have a greater preferene for ahievement-orientation and diplomay (Lesquillier, 2004). As several studies have indiated, suh misommuniation between Duth and British

21 professionals may be aused by fundamental differenes in ultural values. Whereas the Duth and British national ultures are generally similar with regard to the preferred levels of soial hierarhy and harmony (i.e. power distane), the extent to whih feelings are shown (i.e. neutrality vs. emotion) and the role of the individual (i.e. individualism), they vary substantially in their oneptions of the relevane of ontextual information (i.e. high-ontext vs. low-ontext), the degree to whih unertainty is aepted in life (i.e. unertainty avoidane) and the preferred degree of soial or prestigious behavior (i.e. femininity vs. masulinity). With regard to these ross-ultural variations, the Netherlands is marked by a lower ontext, a higher avoidane of unertainty and a higher level of femininity than the UK, where people tend to prefer higher-ontext ommuniation, avoid risks to a lesser degree and are more masuline (Claes & Gerritsen, 2007; Hofstede, 2001; Morden, 1995). Culture-speifi preferenes ould also be refleted in the English annual reports published by Duth and British orporations ompeting or ooperating on the European market. And aordingly, any ultural variations in the realization of these reports may affet the international reader response to either the Duth-English or British-English variant (e.g. Shaw et al., 2004). Based on these onsiderations, the present study suessively fouses on ross-ultural text desriptions of the Duth- English and British annual reports and on the response to any ross-ultural text variations observed. With regard to the ompetitive relationships between Duth and British multinationals, for example, a ontrastive study of Duth-English and British annual reports ould potentially reveal ulture-speifi variations in the multimodal design and effet of ompetitors texts used to inform or reruit international stakeholders (e.g. KPN versus BT). In the ase of Anglo-Duth mergers or partnerships, ross-ultural and interultural analyses of English annual reports might raise awareness of ulture-based differenes in the realization and interpretation of finanial disourse that may be detrimental to the realization of an Anglo-Duth version of the annual report (e.g. annual reports printed by Reed Elsevier and Unilever). 1.5 Aims and researh questions The previous paragraphs provided an overview of key onepts motivating the present researh. As was outlined, the analytial framework of this study speifially fouses on the ross-ultural and interultural pratie of English business disourse within the non-finanial texts inluded in the multimodal genre of the annual report produed by Duth and British multinationals. It was argued that a ombination of ross-ultural and interultural analyses is essential for a full and aurate view on ultural differenes in disourse and their ommuniative impat (Gudykunst, 2003). Also, it was emphasized that the inreasing status of English as an international business language for European multinationals together with the growing importane of non-finanial annual report texts as a promotional underurrent form topial and thus relevant issues of investigation (e.g. Seidlhofer et al., 2006; Hyland, 1998). Moreover, it was proposed that these English texts are part of a professional genre, as the linguisti and graphi design of the annual report is reurrently used to fulfill partiular orporate purposes for a broad audiene (e.g. Yates and Orlikowski, 1992; Garzone, 2004). Finally, refletions on the many ooperative and ompetitive relations that urrently exist between Duth and UK ompanies (some of whih have been mentioned throughout this hapter) indiated

22 that Duth-English and British annual reports would form relevant researh materials for the present ross-ultural and interultural genre analysis. In more theoretial detail, the identifiation of the annual report as a professional genre implies that it is a onventionalized text type whih is shaped by and shapes its speifi ommuniative ontext. From a genre perspetive, the ommuniative ontext of the annual report, inluding its purposes and partiipants, binds this text to typial substantive and formal multimodal harateristis, i.e. ontent, struture and lexio-grammatial elements in texts and images (Miller, 1994; Swales, 1990; Bhatia, 1993). At the same time, its reurrent use also redefines the annual report as a professional text type that is used in a given ommuniative situation to address partiular objetives and readers (Nikerson, 2000; Rutherford, 2005; Yates & Orlikowski, 1992). Assuming, moreover, that the annual report onsists of different underlying texts, it should here be noted that the above genre desription may also apply to eah of these individual texts. Garzone (2004), for instane, indiates that the CEO s letter to shareholders generally provides a personal summary of results and is therefore featured by an interpersonal style of language use that annot be found elsewhere in the annual report. The existene of subordinate text types intertextually related to and embedded in an overarhing text type has been dealt with in various theoretial terms, e.g. subgenres, genre sets or genre systems. Together with the general genre onept, the onept of subordinate genres will be further explained in the theoretial frame of this study (Chapter 2). The pratial or analytial relevane of these onepts within the present investigation of Duth-English and British annual report texts will be aounted for in the methodology setion (Chapter 3). Prior researh taking a similar genre approah to annual reports is omparatively sparse. This seems rather surprising, sine the annual report makes a ruial and ative ontribution to the establishment and maintenane of relationships with orporate stakeholders on whih ompanies rely for their finanial and operational ontinuity. In addition, the ontemporary annual general report is still evolving both as a national and international orporate ommuniation tool and therefore ontinues to offer opportunities for advaned investigations in applied linguistis. Some of the studies referred to in this hapter have foused on the ommon features of English annual report texts from one or various ountries, but have omitted the ross-ultural omparison of business disourse. Hyland (1998b), for instane, examines the generi disourse generally applied in CEO s letters and diretors reports of multinationals based in the UK, USA and Hong Kong. Additionally, Malavasi s trans-ultural genre study of annual reports (2005) omprises the management statements, business desriptions and orporate governane reviews produed by Italian, Frenh, Swiss, German and British banks. Another illustration is Rutherford s (2005) generi approah towards the ontent of the operating and finanial reviews in reports of UK ompanies. Conversely, Nikerson and De Groot (2005) inorporate the ross-ultural in their analysis of Duth-English and British-English management statements, assuming that the annual report genre may reflet distint disourse strategies in business ommunities embedded in various national ultures. This assumption is also at the basis of Garzone (2004), who ontrasts the use of personal pronouns in CEO s letters of Italian-based annual reports to the pronouns used in CEO s letters of foreign annual reports. In none of these reent studies, nor in any other genre study known to the present researher, has the ross-ultural investigation of non-finanial annual report texts been expanded by the inorporation of interultural response analyses. A small

23 number of genre-based investigations did use existing information on ontext and text features to test the mono-ultural and interultural impat of other professional text types, e.g. Lagerwerf and Bossers (2002) on Duth business proposals, Vaughan and Dillon (2006) on Amerian digital newspapers or Shaw et al. (2004) on European and Japanese omplaint handling dialogues. But questions remain about the effets of different ulture-speifi genre features in the interultural use of annual reports in partiular. It should be noted that reader response to annual reports has been the topi of several experimental investigations, suh as Milne and Patten s (2002) study about the effet of soial dislosures on investment-deision-making or Stanton, Stanton and Pires s (2004) study on the response to partiular topis in nonfinanial texts on orporate impressions. However, these studies are primarily experiment-driven and do not onsider genre-speifi issues suh as the relationship between ontext and text design. Drawing on Gerritsen and Nikerson (2004) and Nikerson and De Groot (2005), the present study will ontend that eah ulture involved, i.e. Duth ulture and British ulture, may impose its own national, (pragma)linguisti and organizational norms upon the English disourse used in the non-finanial annual report texts analyzed. Aordingly, there is a possibility that disourse strategies or genre features in the annual report differ ross-ulturally and perhaps affet the interultural pereption of the orporations involved. Based on these onsiderations, one of the main purposes of this study onerns the doumentation of any ulturespeifi preferenes in the prodution and interpretation of non-finanial English texts inluded in Duth-based and UK-based annual reports. That is, it aims to detet the general similarities and differenes between Duth-English and British-English non-finanial texts, explain the ourrene of substantial ross-ultural variations and measure whether the Duth-English texts are as effetive as the native English texts from the viewpoint of international UK-based stakeholders. Furthermore, this investigation seeks to ontribute to a further theoretial and analytial development of the generi approah to ross-ultural and interultural researh of multimodal annual report texts. More speifially, it will test and, if relevant, refine genre theory as a researh tool for both qualitative (i.e. desriptive) and quantitative (i.e. statistial) analyses of texts and photographs and as a preparatory instrument for reader response experiments. Considering these general objetives, the urrent study revolves around the following two main researh questions: I. What are the genre harateristis of English, non-finanial texts in multimodal annual reports produed by Duth orporations and in what way do these differ from the genre harateristis of English, non-finanial texts in multimodal annual reports produed by British orporations? II. How do the typial genre harateristis of the English, non-finanial texts in multimodal annual reports produed by Duth orporations affet the UK-based reader s pereptions of the orporate image and the persuasiveness of the text, and how does this ompare to the typial genre harateristis used in British- English annual reports? The multimodal researh materials used to answer these questions omprise annual reports whih report on the 2003 finanial year and whih were originally realized in English.

24 To answer the first primary researh question i.e. to establish and ontrast the genre features of the Duth-English and British annual reports the present investigation inludes two analytial phases based on oneptualizations in the genre theory (e.g. Swales, 1990; Bhatia, 1993; see also Chapter 2). These two phases are in line with the generi notion that text types are marked by partiular ontextual features on the one hand and substantive, strutural and lexio-grammatial text features on the other. To begin with, the present study proeeds from a set of researh aims and researh questions that speifially fous on exposing the ommuniative ontext of the Duth-English and British annual reports (see RQ 1ae below). Subsequently, it ontains a set of objetives and questions that are to shed light on the multimodal text harateristis of the Duth-English and British annual reports (see RQ 2a-d below). The answer to the seond entral question of this study i.e. to analyze the impat of Duth-English and British-English genre features on UK-based readers pereptions of orporate image and persuasiveness of the text is sought through a researh phase dediated to partiular reader response-related objetives and questions. It should be noted that this final researh phase is to originate from the answers to the first main researh question for whih ross-ultural text distintions will be investigated; suh ross-ultural differenes will form the researh variables in the reader response analysis (see RQ 3a- below). Charaterization of the genre ontext of (English) annual reports originating in the Netherlands and in the UK, from the perspetive of their produers The aims of this first researh phase were (i) to desribe the English language poliy for annual reports originating in Duth orporations, (ii) to desribe the separate ommuniative ontexts in whih the Duth-English and British annual reports are realized, and (iii) to ompare and ontrast the ontextual harateristis for the exposure of any ross-ultural differenes. These aims resulted in four speifi researh questions: RQ 1a. Aording to text produers in Duth multinationals, whih English language poliy is maintained for the Duth-English annual general report? RQ 1b. Aording to text produers in Duth multinationals, what is the typial ommuniative ontext of the Duth-English annual general report? RQ 1. Aording to text produers in British multinationals, what is the typial ommuniative ontext of the British-English annual general report? RQ 1d. To what extent do the typial ommuniative ontexts of the Duth-English and British annual general reports orrespond and differ? RQs 1a-1d will be answered in Chapter 4. Charaterization of the textual and photographi disourse in the Duth- English and British annual reports The aims of this seond researh phase were (i) to desribe the ontents, rhetorial struture and lexio-grammatial features of the texts and the photographs in the Duth-English and British annual reports, (ii) to ompare and ontrast these elements for the exposure of any ross-ultural differenes and (iii) to give a ulturerelated explanation of the omparative results. The following speifi researh questions ensued from these objetives: RQ 2a. What are the typial ontent, struture and lexio-grammar of the texts and photos in the English annual general report produed by Duth multinationals? RQ 2b. What are the typial ontent, struture and lexio-grammar of the texts and photos in the English annual general report produed by British multinationals?

25 RQ 2. To what extent do the typial ontent, struture and lexio-grammar of the texts and photos in the Duth-English and British annual general reports orrespond and differ? RQ 2d. How might the existene of these textual similarities and differenes be explained? RQs 2a-2d will be answered in eah of the Chapters 5 to 7. Response of UK-based readers to the typial generi harateristis of the Duth-English and British annual report The aims of this third researh phase were (i) to independently measure the effets of the typial generi features of the multimodal Duth-English and British annual report texts, from the perspetive of UK-based readers, (ii) to ompare these effets and (iii) to determine the effetiveness of the genre features of the Duth-English annual report texts in relation to the generi features of the British annual report texts. These aims led to three speifi questions: RQ 3a. What is the UK-based readers response to the typial written and photographi disourse in the non-finanial texts of English annual general reports produed by Duth multinationals? RQ 3b. What is the UK-based readers response to the typial written and photographi disourse in the non-finanial texts of English annual general reports produed by British multinationals? RQ 3. Aording to the UK-based reader, is there a differene in effetiveness between the multimodal generi features of the Duth-English annual report texts and the British annual report texts? RQs 3a-3 will be answered in Chapter Overview of this genre study The struture of this dissertation is largely based on the different researh phases desribed in setion 1.5. Chapter 2 establishes the theoretial basis of the study. It explains in more detail the generi approah that underpins all analytial steps taken, i.e. by referring to the interrelated onepts of soial onstrutionism, language use ommunity and rhetorial genre. Next, Chapter 3 offers an integral overview of the methodology applied to answer the first main researh question (onerning the ross-ultural genre investigation). It disusses the proedural approah to the multimodal text analyses, the sampling of researh materials, the researh steps shared aross the different textual and photographi analyses and the design of the interoder tests onduted to measure the reliability of the results per multimodal text analysis. The first researh phase is overed by Chapter 4, whih presents the analytial frame and outome of the ontext analysis of Duth-English and British annual reports (RQs 1a-d). It desribes a survey and interviews with ommuniation and investor relations offiers in Duth and British multinationals; the survey and interviews were onduted to map the generi ontext of the English annual reports. Chapter 5 disusses the analysis of genre ontent or substane in Duth-English and British annual report texts (RQs 2a-2d). It offers definitions of textual and photographi themes as elements of genre substane and gives an overview of results for the themati analysis of texts and images. Chapters 6 and 7 are organized in a similar way. In Chapter 6, the researh results for the analysis of rhetorial genre struture in the texts and pitures of the Duth-English and British annual

26 report texts are preeded by a more detailed explanation of rhetorial moves and strategies as elements of this struture (RQs 2a-2d). Chapter 7 ontains an elaborate desription of metadisourse items as onstituents of the generi lexio-grammar and then presents the final results of the metadisourse analysis of multimodal Duth-English and British annual report texts (RQs 2a-2d). Chapter 8 briefly summarizes the researh results of the ross-ultural omparisons of ontext and text features desribed in Chapters 4 to 7. It also draws onlusions in relation to the main researh objetives and the first main researh question of this study. The most typial results and impliations of the (on)text analyses were assimilated in a reader response experiment. The experiment was arried out to find an answer to the seond main researh question and is disussed in Chapter 9. This hapter omprises information about the theoretial frame, the respondents and the design of the reeption analysis and desribes the outome of the experiment, whih partiularly fouses on the omparative effet of Duth-English and British-English disourse (RQs 3a-3). It also puts forward several theoretial and pratial impliations of the ombination of ross-ultural and interultural genre studies.

27 Chapter 2 Theoretial sope: text as rhetorial genre In setion 1.1 of the previous hapter, the English annual general report was assoiated with the onept of business disourse, whih means that the doument is applied in professional ontexts to fulfill partiular business-related objetives. As suh, the annual report was assumed to be a negotiated textualization, onstruted by soial ators as they go about their daily ativities in pursuit of organizational and personal goals (Bargiela-Chiappini & Nikerson, 2002: 274). In more detail, it was noted that this textualization is reurrently used for the purposes of onveying performane details and establishing a positive orporate image among a potentially broad range of soial ators (e.g. Garzone, 2004; Hyland, 1998). Furthermore, sine the present study involves text produers operating in Duth and British orporations, it was suggested that the language applied in the English annual report might be subjet to ulture-speifi preferenes in text realization and text interpretation (e.g. Nikerson & De Groot, 2005). The idea that language is used to aomplish rhetorial ations bound to partiular soial settings is fundamental to general theories on disourse. The term disourse has motivated a variety of researh perspetives in multiple disiplines, e.g. soiology, ethnomethodology, psyhology, philosophy and linguistis. It ommonly refers to written, spoken or visual language as a form of soial pratie that is onditioned by and signifies the ontext in whih the language is applied. Therefore, disourse an be defined as situated ation: it not only overs the transmission of texts with speifi funtions in a given time and plae but also enompasses the enatment of affiliations with ertain institutions, soial groups or ultures (Bhatia, 1993; Fairlough, 1995; Gee, 1999; Titsher, Meyer, Wodak & Vetter, 2000). Within the area of applied linguistis in partiular, the analysis of disourse has generally foused on understanding the struture and funtion of language use to ommuniate meaning (Bhatia, 1993: 3). Here, disourse analysis onerns language beyond the sentene level, often as a series of interrelated utteranes. It has been approahed from various theoretial and analytial angles, depending on the importane assigned to the soial fators in text-ontext relationships. The early approahes to disourse studies in applied linguistis were primarily onerned with the analysis of language use varieties on the basis of lexio-grammatial text haraterizations; these studies are highly notational and offer little or no explanation as to the rhetorial funtions of linguisti forms or the speifi soial situations in whih suh funtional forms are used, e.g. register analysis (Halliday, 1978) and pragmati grammatial-rhetorial analysis (Selinker, Lakstrom & Trimble, 1973). Although similar linguisti studies still ontribute to the field, the need for explanatory text interpretations has also generated disourse approahes in whih the presene or absene of organizational and lexio-grammatial text features is aounted for by a seletion of soial fators, e.g. interational analysis evaluating text from sender/reeiver perspetives (Renkema, 2001; Sinlair, 2004) and ritial disourse analysis relating text to soial power (Van Dijk, 1993; Fairlough, 1995). In response to the limitations pereived in the above disourse approahes i.e. restrited information on the ontextual rationale underlying language or text varieties as well as minimal onsideration of onventional regularities in these varieties another, more integral perspetive was introdued, i.e. the generi approah to disourse. Genre analysis fouses on the desription of

28 substantive and formal text features and seeks to explain how these features have been influened by the soio-ultural situation in whih the text is used (Bhatia, 1993; Swales, 1990). Regarding the ontology and analytial objetives of these diverse approahes to disourse analysis, it appeared that the genre perspetive in partiular would best serve the first entral researh aim of this study: the exposure of any ulture-speifi preferenes in the reation and interpretation of Duth-English and British-English annual reports. As was desribed in setion 1.5, the present investigation seeks to doument and ompare the general features of the written texts and photographs inluded in the non-finanial setions of the Duth-English and British reports. Moreover, it aims to reveal ulture-dependent reasons why any similarities and espeially differenes between the two ultures might exist. This means first of all that the urrent study should be desriptive and exploratory by nature, rather than presriptive or evaluative (e.g. the interational Correspondene-Consisteny- Corretness-model by Renkema, 2001). Seond, the urrent researh requires an analytial perspetive that allows for a full desription of multimodal text features, ranging from the written and visual ontent to the lexio-grammar in texts and pitures. Therefore, approahes suh as Halliday s register analysis (1964), whih remain foused on the level of words or word strings, do not seem to provide suffiient information. Sine relatively little is known about the impat of the Duth and British ultures on the prodution of the annual report, this study should moreover omprise a broad survey of the ultural fators potentially affeting the textual and visual harateristis of the non-finanial setions. Aordingly, it should not be limited to a partiular seletion of soial influenes, like the (re)establishment of soial dominane through disourse in Van Dijk (1993). Finally, the present study onsiders the onventional use of the annual report. Despite the fat that hanges in sender and reeiver needs as well as new regulations have resulted in subtle adjustments in its ommuniative purposes, ontent and design, its reurrent use has aused the annual report to evolve into a fairly stable type of doument over time. Of all the approahes mentioned previously, the generi perspetive (e.g. Swales, 1990) is the only one expliitly aounting for stabilization in disourse. It ombines an open perspetive on ognitive and soio-ultural aspets of text typifiation with linguisti insights (Bhatia, 1993). In sum, then, the integral view on disourse offered by genre analysis allows for a relevant theoretial sope in the present investigation of Duth-English and British annual reports. The following subsetions will provide a further theoretial definition of the ore onepts from whih the analysis of professional genres arises, i.e. soial onstrutionism (setion 2.1), disourse-based ommunities (setion 2.2) and texts as typified rhetorial ations (setion 2.3). These onepts together provide a oneptual frame whih embraes the first entral researh question onerning the ontrastive identifiation of disourse in Duth-English and British annual report texts. Therefore, their relevane for the present study will be disussed in terms of their appliability to the analysis of professional text types in general and of annual reports in speifi. Although genre theory in priniple overs multimodal disourse (Bhatia, 2004), it should be noted that it has traditionally been applied to written texts in partiular. Therefore, the urrent disussion of genre models primarily fouses on professional texts, as opposed to texts and visuals. At the end of this hapter (in setion 2.4), the role of visual illustrations in genre theory will be referred to. Furthermore, as the soial onstrutionist view has often inspired operationalisations of ommunities of professional language use and as studies on

29 genres have often pointed at the influene of soial onstrutionism and the professional ommunity, several ontributions will be ited in more than one subsetion (e.g. Berkenkotter and Hukin, 1995; Miller, 1994; Swales, 1990; Yates & Orlikowski, 1992). 2.1 Soial onstrutionism Genre analysis has its origin in soial onstrutionism, whih was inorporated into applied linguistis from the field of philosophy. Aording to soial onstrutionists, the use of language omprises soial ations that depend on shared ognitive definitions of the disourse situation. As suh, text prodution and text interpretation onstitute soial proesses. From a soial onstrutionist perspetive, real life is built up of soial relations motivated by ommunal needs, suh as ommunity membership, interation and position within the group (Grabe & Kaplan, 1996; Freedman & Medway, 1994). There is no suh thing as a universal foundation for onstruting reality, or as Bruffee assumes in his overview of soial onstrutionist literature there is only ( ) a onsensus arrived at for the time being by ommunities of knowledgeable peers. Conepts, ideas, theories, the world, reality and fats are all language onstruts generated by knowledge ommunities and used by them to maintain ommunity oherene (1986: 777). Soial onstrutionism basially supposes that reality is psyhologially onstrued by human beings within the boundaries of their ultural and historial bakgrounds. Burr ontinues that, although values and beliefs of soial life are ulturally and historially loated and produed, human beings individually evaluate the appropriateness of these values in ertain soial situations (1998: 24). Burr s explanation is supported by Bruffee, who laims that the interpretation of soial life is not grounded on the fatual or objetive interpretation of physial reality but on our beliefs about the physial reality that shoves us around (1986: 777). This implies that a ombination of soial and ognitive dimensions are involved in the soial onstrutionist view on the (re)prodution and identifiation of soial ations: soial ations are repeatedly performed on the basis of (un)onsious hoie-making about what is preferred within the soial ontext in whih the ation ours. Therefore, reurrent enatment of soial ativity depends on an inter-subjetive proess of meaning-making, where soial ators reah a mutual understanding of the situation and the ations it requires. Moreover, Berkenkotter and Hukin (1995) propose that the onnetion between soial and ognitive proesses is established within situated ognition, whih ontinually evolves on the basis of what is needed and pratied in everyday soial situations and whih thus enables partiipation in soial ativities. Norms for performing soial ations in ertain ontexts are stored in, retrieved from or adapted within situated ognition. Aording to Berkenkotter and Hukin, as soial ators, we onstantly monitor our ations and reognize the available patterns through whih we might at at any given moment, yet we are apable of modifying those patterns to aommodate our reading of the rhetorial moment (1995: 24). Bazerman (1997) further explains that prior experiene or knowledge learned about partiular soial situations raises awareness of possible effets of soial ations in these situations and hene of the responses these ations require.

30 As Bruffee (1986) points out, soial onstrutionism as suh holds a ruial role for disourse-as-situated language use, in the sense that language and soial situations are part of a mutually onstrutive relationship. Language does not only provide a desription of the world as it is pereived in a partiular time and spae, but it also onstruts this world and may even alter it. On the one hand, the use of language is a soial ation realizing soial proesses, e.g. presenting an environmental statement to lients or exhanging internal memos with olleagues. On the other hand, the soial proesses enated through language use help to sustain the knowledge needed to manage these proesses against the bakground of partiular historial-ultural onventions, e.g. preserving the environmental statement or the internal memo as speifi means of ommuniation. Additionally, by linguistially expressing shifts in the knowledge underlying the enatment of soial proesses, hanges in soial knowledge or meaning are established, e.g. as is the ase in adjustments in the aeptable amount and type of information distributed by the environmental statement or the eletroni memo (e.g. Bazerman et al., 2003; Burr, 1995, 2003; Yates and Orlikowski, 1992, 1994). As suh, language-based soial ations respond to and (re)onstrut the soial world. For interatants, the use of language thus omprises the ontinual ativation of mental representations of the soial world in whih their texts are to be embedded. These mental representations help interatants determine what language use is relevant, for instane, in addressing ertain audienes, purposes or themes (Rubin, 1988). The onept of disourse as soial ation also implies that mental onstruts of the disursive situation are generally based on onsensual values whih have been established and reinfored through a onventionalized use of language. Aordingly, modifiations in language use onventions generally reflet developments in soial meaning, i.e. hanges in the organization of soial ativities or soial relations (e.g. Bazerman, 1997). The soial onstrutionist approah to disourse has proven to be a useful tool in explaining language use in professional settings. Soial onstrutionism predominates, for instane, in disourse analyses by Pogner (1999) and Wegner (2004). Both researhers fous on desribing the onstrution of partiular texts along the lines of speifi ontextual dimensions. Pogner (1999) investigates the omposition of an energy onept by Danish onsulting engineers and a German muniipality. His study reveals that differenes in national business ulture and organizational ulture aused the Danish writers and German lients to have ontrasting pereptions of the status and funtion of the energy doument. This led to a ommuniative onflit: while Danish onsultants viewed the energy onept as an expert doument failitating the onsultany ativity with lients, German authorities regarded this text as the final onsultany produt refleting their deision-making proess. Pogner onludes that the realization of the energy onept is indeed a soial ativity, where ultural expetations towards the doument and the tehnial planning task are onstruted ( ) in interation (1999: 117). Wegner s study (2004) on the onstrution of a natural areas management plan in a muniipal ommunity projets a similar disourse situation. She analyzes what proesses are involved in developing a rather unommon text that annot be assoiated with a text type already existent in the soial ommunity. Her findings reveal that partiipants in the omposition of the plan drew on soial knowledge needed to draft familiar doument types, e.g. general management reports, as well as on soial knowledge needed to understand the general situation in whih the natural areas plan funtions. Aording to Wegner, the onstrution of the plan omprised a

31 learning proess: Partiipants resoures inluded their substantial shared knowledge of the soio-ultural, historial and institutional ontext that supplies the relevant soial languages and praties for their enterprise (2004: 446). A third researh example whih emphasizes the importane of soial onstrutionism in organizational disourse is Bargiela-Chiappini (1999). She onduts a ross-ultural omparison of the rhetorial design, themati ontent and lexial hoies in Italian and British human resoure management (HRM) magazines. Her analysis indiates that, while Italian HRM magazines applied values of ontrol and diretorship, the British HRM magazines inorporated more human values. Aording to Bargiela- Chiappini, any differenes between the Italian and British magazines stemmed from ontemporary ultural differenes onerning the field of HRM: both reader and writer ontribute by drawing from their experiential baggage, thus making the at of interpretation not only historially and soially situated, but also possible in the first plae (1999: 148). In the present study of annual reports produed in the Netherlands and in the United Kingdom, soial onstrutionism is of equal importane, i.e. for defining and motivating the English disourse in these texts. In the Duth and British business ontexts as in the entire western world the annual report an generally be marked as a purposive soial ation. Considering the gradual evolution of the annual report from a written fat sheet into a multimodal investor relations instrument (Hynes & Bexley, 2003), it is not an isolated, stati text. Instead, it has ontinually been shaped by and has repeatedly reprodued and modified the ommuniative situation in whih it is used. Depending on the needs of the soial group within whih the annual report was written, read or evaluated, the text has been subjet to adjustments in rhetorial funtion, stakeholder audiene and subjet matter. For example, in the past deade Duth and British orporations have been required to extend their orporate governane and orporate soial responsibility (CSR) aounts in their annual reports in order to meet new regulations and gain goodwill with speial interest groups (e.g. Hooghiemstra, 2003; Kolk, 2003). Furthermore, for the ontrastive analysis of English language use in the Duth- English and British annual reports, the soial onstrutionist perspetive allows disourse identifiation within a wide ultural frame. Choies about what the annual report should do and look like are grounded in an awareness of soial onventions in text prodution. Sine rules for text design may differ per ulture, regarding the annual report as a produt of its soio-ultural bakground ould aount for any ross-ultural differenes in text design. More speifially, the onept of the annual report as a soial onstrut helps to larify why and under what onventional onditions Duth and British text produers and stakeholders apply and aept partiular text features in the annual report and reognize them as germane (Bargiela-Chiappini & Nikerson, 1999). Although soial onstrutionism in priniple allows for disourse interpretations at all ontextual levels, ranging from the individual to the global, it should here be noted that the present study fouses on situated larifiations drawn from the national business ultures of the Netherlands and the UK. 2.2 Community and disourse Engagement in disourse does not generally involve just any random group of individuals. Instead, the use of language is tied to and shaped by a group or multiple

32 groups of peers. Theory in applied linguistis has referred to suh groups in terms of ommunities that form the soure of both knowledge and soial ations. Although there is general onsensus about ommunal harateristis suh as oherene through shared values and truth norms, disourse onventions and ooperation, different oneptualizations of ommunities for language use have been proposed. Four of those have been of substantial influene on disourse analysis: speeh ommunity, interpretative ommunity, disourse ommunity and ommunity of pratie (e.g. Beaufort, 1997; Borg, 2003; Pogner, 2005). The oldest of these onepts, i.e. speeh ommunity, has its origin in soiolinguistis (Hymes, 1974). It omprises a variant of language use that is distint from the language spoken or written in other ommunities. Members of a speeh ommunity are bound involuntarily (i.e. by birth or aident) to a ertain spae whih holds linguisti norms and are therefore part of onrete groupings that an be identified rather easily, e.g. the Afrian-Amerian speeh ommunity (Borg, 2003; Harris, 1989; Swales, 1990). Whereas the speeh ommunity onept aounts for the linguisti oherene in a ommunity, the interpretative ommunity fouses on its soio-ognitive bonding. The interpretative ommunity was introdued by Fish (1980) and reflets a ommon state of mind among ommunity members that is reahed and onstrained by shared values and assumptions about what is to be expeted and aeptable in soial enounters. It overs groupings that are not physially identifiable, onsisting of individuals who do not neessarily engage in ahieving interests and objetives, e.g. the ommunity of law (e.g. Borg, 2003; Harris, 1989; Pogner, 2005). To establish a link between the use of language and ommunity-speifi norms, the linguisti onept of disourse ommunity was initiated (Swales 1990). Disourse ommunities are soio-rhetorial networks of speial interest groups that produe texts with ommon objetives. In disourse ommunities, the use of language is a demonstration of (desired) membership and thus of the awareness of ommunal onventions. The dependene on soial purpose as the main argument for ommunal oherene implies that the disourse ommunity onerns groupings whose boundaries may be vague and whose texts may alter in order to fulfill this purpose, e.g. a ommunity of engineers drafting an energy onept (Bargiela-Chiappini & Nikerson, 1999; Borg, 2003; Pogner, 2005). The last ommunity onept relevant to applied linguistis involves the soioultural ommunity of pratie (Wenger & Snyder, 2000). Its emergene is a ritial reation to the strong emphasis on ommuniative purpose and text prodution in the disourse ommunity (Borg, 2003). Communities of pratie fous on ommon expertise and joint enterprise as the main reasons for ohesion and not on any speifi type of soial ation. Members of a ommunity of pratie belong to relatively manifest groupings and are self-seleted pratitioners, e.g. a projet management ommunity (Pogner, 2005; Wenger & Snyder, 2000). Within genre-based researh of professional or business texts, the onept of disourse ommunity has been of onsiderable interest. Perhaps this is owing to the fat that the majority of generi studies revolve around the design of partiular types of purposive written or spoken text and thus fous on a speifi language-related soial ation loated and reated in a ertain time, spae and membership. In these studies, the exhange of experienes or information through any additional organizational disourse is often of minor importane. As suh, the onept of disourse ommunity has been used to interpret the ommuniative ontext of various professional text types, e.g. researh artiles (Hyland, 1998), business proposals (Lagerwerf & Bossers, 2002), internal business (Nikerson, 2000),

33 energy onepts (Pogner, 1999), sientifi onferene presentations (Rowley-Jolivet, 2004), For-Your-Information letters (Vergaro, 2005) or sales letters and business faxes (Zhu, 2005). Some studies have also referred to the disourse ommunity onept to ontextualize the annual general report in partiular (Rutherford, 2005; Solbjørg Skulstad, 1996, 2005). In more detail, a disourse ommunity an be identified by a set of ommuniative purposes, interommuniation mehanisms or forums among its members, a olletion of text types used for information exhange, a typial lexion or terminology and altering memberships (Swales, 1990). These features indiate that a disourse ommunity aims to establish a ontinuous ommunality of publi interest, a set of disourse onventions used to produe and disseminate knowledge and voluntary partiipation (Swales, 1990; Pogner, 2005). With regard to the professional or business domain, Sollon, Bhatia, Li and Yung explain that disourse ommunities are related to soial identities that are an essential element of one s professional or workplae identifiation (1999: 23). Members of a given business or disiplinary ommunity at in aordane with the soial and disourse norms maintained in that ommunity; aordingly, their use of language does not merely reveal what they wish to ommuniate but also who they laim to be (Planken, Van Hooft & Korzilius, 2004; Sollon et al, 1999). Moreover, several investigations have foused on a ulture-based definition of the professional disourse ommunity, i.e. they have aounted for the role of national and professional ultures in the interpretation of the disourse ommunity of a text. These studies have pointed out, for instane, that professional disourse ommunities an exist both within and aross national ultures, e.g. the US-based federal government ommunity (Beaufort, 1997) versus the ommunity in whih both British and Italian For Your Information letters are produed (Vergaro 2005). Additionally, these studies have shown that disourse ommunities may arise at different ultural levels; disourse ommunities an exist within a orporate institution suh as an Italian sportswear ompany (Ponini, 2002), within a professional expertise group suh as the ommunity of independent marketers (Ponini, 2002) and within geographially determined institutions suh as the loal ity government or the national Amerian eduational department (Beaufort, 1997). As for the present study on annual reports originating in Duth and British ompanies, the onept of disourse ommunity is only partly relevant. Similar to what was observed in previous studies on the onstrut of the annual report (e.g. Rutherford, 2005; Solbjørg Skulstad, 1996, 2005), it appears that the ontext within whih annual reports are generally produed orresponds largely with Swales (1990) haraterization of a disourse ommunity. From the perspetive of text prodution, annual reports are realized by self-seleted members of a speial interest group who repeatedly asribe informational and promotional purposes to the text and who provide it with a reognizable design (e.g. Garzone, 2004, 2005). The only ommunal feature that seems to be problemati in the prodution ontext of the annual report onerns the presene of an interommuniation mehanism between text produers: these professionals usually fous on ativities within their own work area and hardly engage in interation with professionals in other ompanies (see the Café Owner Problem in Swales, 1990: 25). Sine the annual report has evolved into an established publi doument, however, it would seem that the absene of a regular forum among annual report produers does not neessarily eliminate the expert ommunity. This is likely to be the ase for reators of other types of business texts also (e.g. orporate websites or produt brohures).

34 The ulture-based aounts of the disourse ommunity onept (Beaufort, 1997; Ponini, 2002; Vergaro, 2005) are partiularly useful in determining and motivating the ultural fous in the prodution ontext of the urrent Duth-English and British annual reports. In the light of the ulture-speifi desription and rossultural omparison of Duth-English and annual report texts, the present view on disourse ommunity onerns both a national and a multi-national fous. In speifi, it involves multiple national ultures and the business ultures enlosed therein. The present fous on nation-level disourse ommunities is intentional: sine one of the main researh aims of this study is to make valid generalizations of text prodution and text interpretation aross the Netherlands and the UK, the effets of industry-related or ompany-related issues were to be redued as far as possible. In priniple, the onept of national business ultures allows for a nation-speifi analysis of a variety of organizations in diverse industrial setors, without atually highlighting these setor and orporate differenes. Consequently, the identifiation of disourse ommunity on the level of national business ulture enabled the exlusive analysis of the genre features typially inluded in the Duth-English and British annual reports and for the omission of potential influenes of industry or orporate identity. Although the disourse ommunity onept was appliable to the prodution side of the Duth-English and British annual report, and hene to their ross-ultural omparison, the onept does not seem to over the ontext in whih readers reeive these annual reports. Some (genre) analysts have suggested the inorporation of text reeivers in the onept of disourse ommunity, laiming for instane that the onept relates to the relatively stable overall funtioning of ommunities of readers and writers in whih genres arry out the soial goals of the ommunity (Beaufort, 2000: 188). Suh referenes to reeivers of the text, however, often imply an imagined readership; they do not atually omprise real readers but writers mental representations of these readers and their reader requirements (e.g. Hyland, 2007). Therefore, most reader referenes in the onept of disourse ommunity ontinue to reflet a view of writing, i.e. reader referenes are used to aount for reader orientations expressed in the text. Genre-based studies that make mention of existing reeiver audienes ommonly do so in terms of ommunities of readers or simply readership (e.g. Hyland, 2007; Hyland & Tse, 2004; Vergaro, 2002).In the ase of business texts suh as the annual report, moreover, the non-expert reader audiene does not seem to be inluded in the disourse ommunity of expert text produers. As Bhatia explains, these readers belong to a larger set of people outside the ommunity, one of whih generally is ordinary publi and the other(s) may be disourse ommunities other than those to whih [the ommunity members] belong (2002: 9). Presumably, the abstrat position of non-expert readers in the identifiation of disourse ommunities is a result of the disiplinary ultures from whih the disourse ommunity onept emerged. Originally, this onept was established for the explanation of speial interest groups within aademi disiplines, where both produers and reeivers of the texts (i.e. researh artiles) are experts in the field (Hyland, 1998a, 2000; Swales, 1990). Aordingly, the onept of disourse ommunity is less relevant with regard to the seond main researh question of the present study, i.e. onerning the interultural response to Duth-English and British annual report texts. Clearly, a different oneptualization is needed to aount for the international reader audiene of these texts. A survey of the other theoretial views on ommunities of language use does not offer a solution; they are either narrow-angle views on language or

35 soio-ognitive groupings (i.e. speeh ommunity or interpretative ommunity) or they remain foused on interation within an expert ommunity (i.e. the ommunity of pratie). Considering the potential worldwide reah of the English annual reports by Duth and British orporations, a reader-related onept like the international or global ommunity of readers would seem to be more appropriate here. Mahin and Van Leeuwen explain that members of this ommunity need not share a language [ ] but they do share an involvement with the same modalities and genres (2003: 493). Its international dispersal and linguisti heterogeneity suggest that this reader ommunity is muh more a global speifi interest group than a group united by a ommon language (Mahin & Van Leeuwen, 2003). As for the various international audienes of the Duth-English and British annual reports, they are also unified primarily beause of their shared interest in the ompany under review and their mutual reognition of the annual report as an answer to ompany-related interests. Moreover, it is the use of these annual reports in partiular that brings them into (indiret) ontat with the produers of these texts. In other words, interation between the global reader ommunity and the expert disourse ommunity is established through the ommuniative links set out by the annual report; through the annual report, different ommunities interset and thus form a ommon point of referene. 2.3 A definition of professional genre The assumption that disourse is purposive soial ation whose repeated enatment onstitutes and maintains the soial system of a ommunity, is fundamental to the theory of genre. It aounts for the fat that language use an evolve into text types: the reurrent expression of partiular utteranes to ahieve the same ommuniative purpose in a speifi soial situation may eventually result in the evolution of standardized, distintive texts. These texts or text types reveal how a ommunity typially operates as a soial group and, as suh, they are generally defined as genres (e.g. Bakhtin, 1986, Bargiela-Chiappini & Nikerson, 1999; Nikerson, 2000; Orlikowski & Yates, 1994). During the past three deades, theorizations about the genre onept have been entral to three traditions, i.e. Australian systemi funtionalism, North Amerian New Rhetori and English for Speifi Purposes (ESP) or Languages for Speifi Purposes (LSP, a term introdued to over the non- English studies of applied linguistis) (Freedman & Medway, 1994; Hyon, 1996; Swales, 2000; Yunik, 1997). The oneptualization of genre in these traditions differs on motivational and funtional grounds. For the Australian systemi funtional shool, disourse analysis omprises the exposure of language use onventions for the purpose of empowering minority groups. Its ideology-based desription of texts primarily fouses on formal language features and indiretly refers to their funtion in soial disourse settings. Aording to Halliday (1978: 145), genre has an enabling funtion in the realization of texts: genre omprises the soial value of a text, whih determines what soial ation ( field ) or what partiipant roles ( tenor ) should be performed through the use of partiular language features ( register ). Here, genre plays only a subordinate role in the identifiation of text types (Freedman & Medway, 1994; Hyon, 1996). In ontrast with the systemi-funtionalist approah to genre, the North Amerian New Rhetori shool has no higher ritial goal and plaes the genre onept at the enter of disourse lassifiation. Rhetoriians laim that genres allow individuals to fulfill

36 soially identified needs in a manner that is reognizable for other members within the same ommunity. Here, the emphasis is more on the rhetorial ation to be aomplished through the genre than on the ontents and form of the generi disourse. Additionally, the rhetorial genre is onsidered to be dynami, i.e. it evolves, sustains for a period of time and then dissolves again (e.g. Bakhtin, 1986; Bazerman, 1997, 1994, 2003; Berkenkotter & Hukin, 1995; Miller, 1994). Lastly, the non-ideologial Languages for Speifi Purposes tradition appears to omprise ideas of both the Australian and the New Rhetori shools. Like the Australian shool, LSP aims at typifying texts on the basis of formal features. And like the New Rhetori shool, it relates these features to the dynami soial setting in whih the texts are used. In the LSP tradition, genre is fundamentally assoiated with ommuniative purpose, whih explains why a text ontains typial formal elements and speifi ontent features. Moreover, it assumes that genres are not simply independent entities but belong to networks of texts whih interatants draw upon depending on the disourse situation (Bhatia, 1993; Freedman & Medway, 1994; Hyon, 1996; Swales, 1990, 2000; Yunik, 1997). The above overview suggests that LSP and New Rhetori oneptualizations of situated genre in partiular refer to soial onstrutionist ideas (disussed in setion 2.1.). LSP studies on texts used in aademi or business settings have often been based on the soio-historial work of rhetoriians, presumably beause a ombined approah provides insight in the proedures, praties, and onventions that make the text possible and relevant to a partiular soio-rhetorial ontext (Bhatia, 2002: 5). The LSP genre tradition fouses on ontextualized language desriptions and on an interdisiplinary researh approah involving (applied) linguistis, soiology and psyhology (e.g. Bhatia, 1993). As suh, it is valuable for the present study, whih seeks to establish a link between multimodal features of English annual reports and their Duth or British ultural bakgrounds. Based on the fundamental genre theorizations of Miller (1994), Fairlough (1992a, 1992b, 1996), Swales (1990) and Bhatia (1993), LSP genres in professional settings an be defined as follows: Genres are situated soial ations haraterized by partiular ommuniative or rhetorial purposes reognized and mutually understood by the members of the professional ommunity in whih the ations reour, often as part of a network of interrelated events. The rhetorial purposes of these ations form the rationale for the genres, i.e. they onstrain and explain the audienes of the genres as well as their prototypial ontent, disourse or move struture and lexio-grammar. These onstraints are often exploited by the experiened members of the disourse ommunity in order to ahieve sender-speifi strategi intentions within the sope of the soially reognized genre purposes. This genre definition will be dealt with in more detail in the next subsetions. After a further theoretial view on the ontext of the professional genre i.e. involving the onepts of soial ations (setion 2.3.1), rhetorial purpose (setion 2.3.2) and strategi intentions (setion 2.3.3) an overview will be given of the ore elements of the genre i.e. its prototypial ontent, move struture and lexio-grammar (setion 2.3.4). Finally, an explanation will be given of the intertextual networks to whih the professional genre often belongs (setion 2.3.5). Eah theoretial onept will be linked to the identifiation of annual reports in general. Setion finalizes the disussion on professional genres by relating the theoretial onepts to the present Duth-English and British texts in partiular.

37 2.3.1 Genres as typified soial ations First of all, genres are typified soial ations beause they reurrently emerge in a given rhetorial or soial setting. Assuming that soial onstrutionism is inherent in language use, genres repeatedly aquire their meaning from a given rhetorial situation as well as from the larger soial or ommunal ontext in whih this situation arose (Miller, 1994). For instane, Chinese orporations stimulating longterm produt sales through the promotional letter genre generally refer to Chinese ultural norms of fae-keeping and harmony (Zhu, 2005). The rhetorial setting of a genre does not involve repetitive enounters with a material situation; rather it omprises an inter-subjetive, soially defined onfiguration, a mutual understanding of situations as somehow omparable, similar, or analogous to other situations (Miller, 1994: 29). These soial onstruts determine what intentions, soial relations and rhetorial praties are ommonly deemed aeptable and relevant within a ertain language use ontext. They are stored in interatants stok of soial knowledge, whih in turn is the result of enulturation through learning and prior experiene (e.g. Bazerman, 1994; Berkenkotter & Hukin, 1995; Miller, 1994; Yates & Orlikowski, 1992). The idea that the generi text as soial ation is a produt of the rhetorial situation in whih it reours, has two onstrutionist impliations. It indiates firstly that genres are apable of organizing and lassifying human behavior within ommunities of language use. Sine genres are repeated responses to ertain disourse ontexts, they express situation-speifi onventions in language use and, hene, say something about the typial interational oherene within the ommunity. As Bazerman (2003) points out, genres ontain typifiations of soial ativity and of relations that help people make mutual sense of the life world (456). Additionally, genres as text-based soial ations ensure the endurane of a ommunity. The reurrent realization of partiular texts in typial rhetorial situations oinides with the reprodution and reinforement of language use onventions and thus of the soial grouping itself. The reiproity between ommunity members disourse habits and the establishment of soial systems is known as the struturation proess (e.g. Berkenkotter & Hukin, 1995; Giddens, 1979, 1984 in Yates & Orlikowski, 1992). Struturation an take two forms: maintenane and transformation. Maintenane ours when the (re)prodution of genres is used to onfirm the interational orientation in the disourse as a soial fat. It refers to the ontinuous exeution of ongoing disursive and therefore soial praties. Genre transformation takes plae when historial developments form new soial strutures and text types are extended or modified to adapt to these strutural hanges. In turn, the enatment of these evolved text types generates and establishes the new soial strutures, thus sustaining the entire soial system (Berkenkotter & Hukin, 1995; Nikerson, 2000; Yates & Orlikowski, 1992). The interplay between generi maintenane and transformation is illustrated in Akar and Louhiala- Salminen (1999), whose study on the business fax shows that it was an evolving genre at the time of researh, i.e. moving from a formal business text to a more informal and onise text type. Akar and Louhiala-Salminen onlude that although the rhetorial purpose of the fax has remained the same, it is affeted by new tehnology and new business praties ( ) As a result, a new set of onventions for onduting business is emerging that is mediated by the partiular features of the message system (1999: 220, 221).

38 With regard to the annual report as a soial ation, it has been noted that members of the investment ommunity expet the annual report to serve ommon soial objetives and to meet legal, traditional and ethial requirements (Jameson, 2000). Also, the annual report is believed to be subjet to ontinual soio-historial developments in national and business ultures. Before and after it is maintained for a partiular period of time, the annual report is usually transformed as a result of shifts in, for instane, stakeholders reading needs, medium use, soial attitudes towards finanial reporting, reporting praties, and reporting regulations (Garzone, 2004; Jameson, 2000; Nikerson & De Groot, 2005; Rutherford, 2005). As an illustration, annual reports have generally evolved from a highly numerial text into a multimodal text with many qualitative performane dislosures. And within the UK in partiular, the Aounting Standards Board introdued more stringent guidelines for the design of the Operational and Finanial Review after aounting inidents in the US business arena (Rutherford, 2003) The rhetorial purpose of genres An additional harateristi of genres as soial ations is their typial rhetorial purpose or soial motive, whih determines the existene and design of partiular text types. The harater of the soial motive is twofold: rhetorial situations in whih genres are used imply the need for the performane of ertain ommuniative ations (i.e. motive as situational need), whih at the same time are motivated and goal-oriented (i.e. motive as text purpose). The relationship between rhetorial situation and soial motive is referred to by Miller (1994) in terms of exigene or objetified soial need. The objetified soial need enompasses the assumptions and expetations shared among members of a soial group about the disursive ativities that are needed for rhetorial partiipation under partiular situational onditions. Aording to Miller, it is an understanding of the soial need in whih I know how to take an interest, in whih one an intend to partiipate (1994: 30, 31). Thus, the exigene overs a set of ommunity-internal patterns determining for members whih objets, events and interests reflet whih soial need and, subsequently, whih soial ations should be taken to meet this need. This an be illustrated by Devitt (1991), who douments that tax aounting firms produe and exhange the genre of the tax protest as a response to the need to address tax authorities with an argument for the lient s tax position, after previous negotiations with a tax agent have been ineffetive. While the rhetorial situation gives rise to soial needs, these needs in turn beome motives for the soial ations performed in reation to these needs. Or, the exigene embedded in the rhetorial ontext provides the text with a soially reognized ommuniative purpose (Miller, 1994). Initially, Swales regarded ommuniative purpose as a privileged riterion for genre identifiation: sine it influenes the ontent, form and audiene of a genre, purpose operates to keep the sope of a genre ( ) narrowly foused on omparable rhetorial ation (1990: 58). This first operationalization of generi purpose suggested that texts reflet unique and evident purposes, prinipally allowing for an inontrovertible taxonomy of text types. Although this idea of ommuniative purpose proved relevant in many genre studies, it was ritiized for its stati text-orientation, i.e. for the fat that it did not aount for subtleties in text purpose within the boundaries of the generi norms (Askehave & Swales, 2001; Bhatia, 1993). Communiative purpose as suh was found to be neessary but not suffiient in the (ross-ultural) identifiation of

39 genres with multiple or layered objetives (e.g. Askehave, 1998, 1999). Although purpose is still seen as the main devie for distinguishing between text types, urrent oneptualizations of generi purpose are more ontext-driven. To obtain an aurate identifiation of a given text type, genre studies have inreasingly based their text interpretations on ommunity-speifi values, onditions, expetations and interational repertoires (Askehave & Swales, 2001; Bhatia, 2002, 2004; Zhu, 2005). The ontextual approah to genre analysis proved to be relevant, for example, in Askehave s (1999) study of ompany brohures: these texts are generally believed to promote a orporation but they may also inlude underlying strategi purposes suh as portraying the organization as a qualified partner in order to enhane long-term lient relationships. This set of primary and seondary purposes is reognizable only by moving beyond the obvious and taking a broader view on what disursive spae is allowed within the rhetorial situation of the ompany brohure. In line with these onsiderations, annual reports have always been marked by partiular onventional ommuniative or rhetorial purposes. At the most abstrat level, annual reports are presently regarded as disursive praties through whih orporations institutionalise the systems of ideas, beliefs, meanings and onepts underlying their ideologies (Fox, 2006: 358). Thus, similar to other orporate genres, the annual report is a ommuniation tool used for the expression of the organizational philosophy. More onretely, in reent years it has been observed that annual reports generally fous on informing a broad range of orporate stakeholders about the ompany s performane and supporting the ompany s impression management. Aordingly, the annual report aims to exhibit both objetive figures and biased evaluations reassuring stakeholders of the health, sustainability and trustworthiness of the organization (e.g. Garzone, 2004; Hyland, 1998b; Malavasi, 2005; Rogers, 2000). Furthermore, Garzone (2004) and Hyland (1998b) indiate that the degree to whih the informative and affetive goals are relevant in the annual report differ per omponent, e.g. the management letter to stakeholders is more evaluative than the Diretors report whih ontains a more fatual reord of (board) ativities and poliies Strategi intention in genres The hermeneuti approah to genre purpose suggests that genres may be realized through both standardized and strategi intentions in language use. On the one hand, the overarhing, universally reognized purpose of a text is expressed through language that has traditionally been assoiated with the ahievement of that partiular purpose and whih distinguishes one genre from another, e.g. the language that has onventionally been used to identify the omprehensive orporate promotion -purpose in the ompany brohure. On the other hand, these soiohistorial language onstraints may be exploited by strategi onsiderations exhibiting institutional dispositions, writer reativity, originality or even ultural authentiity. Here, ognitive onsiderations are at the basis of sender-speifi language use whih, although still interpretable for members of the wider ommunity, results in the subtle individualization of the genre, e.g. the language that is potentially used in ompany brohures to express the underurrent aim of positioning the orporation and sustaining trading allianes (e.g. Askehave & Swales, 2001; Bhatia, 1993).

40 Sender-speifi hoies embedded in the realization of a rhetorial genre an be regarded from two perspetives: individual exploitation and ultural interpretation. The expression of an individual s private intentions within the frame of a genre s soially reognized objetives was addressed by Miller, who laims that the objetified nature of the genre still provides an oasion, and thus a form, for making publi our private versions of things (1994: 30). She explains that the definition of a generi motive and thus of a genre emerges from an individual s own reonstrution and ategorization of a partiular rhetorial situation, and that this indues him or her to determine how to engage in it in a soially reognizable and interpretable way (1994: 31). Miller s argument is supported by Bhatia s (1993) work on professional genres, whih assumes that expert members of a disourse ommunity have the knowledge and skill to make innovative use of genre norms. Bhatia exemplifies this idea by referring to experiened newspaper reporters who often sueed in imposing desired perspetives on otherwise objetive news reports (1993: 15). In addition to a writer s individual realizations, strategi hoies for genre disourse an onern ulture-dependent onsiderations: in different ultures, the ommuniative purpose of the same genre may be affeted by distint disursive preferenes, leading to subtle ultural variations in the omposition of a genre. In her analyses of disourse strategies aross Italian and English business letters, for example, Vergaro (2002, 2005) finds that genre purpose as the ommon denominator for global text typifiation does not rule out the impat of ultural variability in terms of linguisti utteranes and the positioning of similar elements in text struture. Consequently, her studies indiate that the loalization of senderspeifi hoies in ultural bakground an ontribute to a full pragmati desription of seemingly idential genres pratied internationally. Corresponding ideas an be found in Mauranen s (1993) analysis of Finnish-English and Amerian eonomi texts. Her study is based on the premise that these douments are built on global resoures governed by the genre and persuasive tatis direted by national ultural preferenes. Aording to Mauranen (1993), writers rhetorial hoies to enhane text redibility depend on what is pereived to be onvining disourse; suh hoies are limited by the value and belief systems prevailing in the disourse ommunity whih onstitute the soial ontext for the text (Mauranen, 1993: 5). If this ommunity represents multiple soial systems e.g. Finnish and Amerian sholars eonomi texts are driven by global genre features and ulture-driven features, where the latter tend to hold aross disiplines and genres but be typial of a national ulture (Mauranen, 1993: 5). Despite the urrently stabilized purposes of the annual report and its globally reognized text features, ross-ultural text analyses in partiular have also revealed sender-speifi strategies in the textual realization of these purposes. For example, in their investigation of the introdutory management forewords in Duth- English and British annual reports published in 2001, Nikerson and De Groot (2005) find that these texts vary onsiderably in the inlusion of salutations, operational data, future objetives and board details. Moreover, Lampi s (1992) pilot study of Chairman s statements in Finnish and British annual reports shows rossultural variations in assumptions about text struture, linguisti style and writerreader relations (in Yli-Jokipii, 1998). And in her analysis of Italian and non-italian CEO s Letters, Garzone (2004) observes that the Italian-based orpus is marked more by impersonal ompany referenes and less by first person plural pronouns referring diretly to the sender of the message. In all these studies, ulture-speifi onsiderations aount for the differenes in strategi hoies made to ahieve the rhetorial objetives of the annual report texts.

41 2.3.4 Content, move struture and lexio-grammar of genres Apart from the situational and motivational onditions that ontextualize a genre, it is typified by substantive and formal text features. These text features, inluding ontent, struture and lexio-grammar, are influened (in)diretly by the ontextual irumstanes. Context reflets the rationale underlying the usage of partiular text features in a given text type. Contrary to the genre purpose embedded in the rhetorial ontext, the substantive, strutural and linguisti properties of a genre are not absolutes but prototypial harateristis; their realizations allow for some variation aross instanes of the same text type. Therefore, they suggest that genre typifiation rests on the identifiation of family resemblane instead of definitional resemblane (e.g. Connor & Wagner, 1998; Lok & Lokhart, 1998; Swales, 1990). The ontent of a genre involves the information or proposition refleted in the text. In priniple, it is a ognitive onept; it is what Miller has alled genre substane or the semanti value of disourse, [whih] onstitutes the aspets of ommon experiene that are being symbolized in the text (1994: 32). In their analysis of aademi texts, Berkenkotter and Hukin explain that genre ontent omprises mental proesses in whih shared experiene within a partiular ommunity, e.g. norms, bakground knowledge, opinion or soial situation, is translated into informational value that is appropriate within the frame of the purposive text (Berkenkotter & Hukin, 1995; Hukin, 2002). Similarly, in her investigation of tax aounting genres, Devitt (1991) argues that the typial rhetorial situation embodied in eah tax aounting genre gives this genre its expliit substantive meaning, thereby handling one omponent of the tax aountant s work situation. As for the identifiation of typial text ontent, genre studies have sometimes referred to an informational landsape omprising a set of interrelated details on objets, ators and events in a time-spae setting (e.g. Bazerman, et al., 2003). Some have also foused on the atual subjet matters in the genre, suh as Berkenkotter and Hukin (1995) who look at topis in onvention abstrats. Others have observed the partiular message orientation represented in the subjet matters, e.g. informational and relational genre substane in CEO presentations (Rogers, 2000, 2001) or internally and externally oriented genre ontent in eletroni memos (Yates and Orlikowski, 1992). Sine all these interpretations of genre substane assume that it is haraterized by situationspeifi expressions of ommon experiene or knowledge, it an be onluded that there may be variations between soial or professional groupings in what they believe this ommon ground ontains, i.e. they may onsider a different ontent to be appropriate in order to aomplish the same rhetorial ation (Nikerson, 2000: 41). This has been illustrated by Sims and Guie (1992), who arried out a omparative study of native and non-native English letters of inquiry onerning a US university. Their results show that non-native MA and PhD andidates from Europe, Asia and the Middle-East provided more unneessary professional and personal information than native Amerian andidates. The prototypial rhetorial organization of texts is referred to in genre theory as the move struture, in whih text fragments are distinguished as marostrutural onstituents with distintive rhetorial funtions that eah ontribute to the realization of the general purpose of the full text (e.g. Bhatia, 1993; Henry & Roseberry, 2001; Nikerson, 2000; Swales, 1990; Upton & Connor, 2001). The move struture is not merely a reader s individual shemati onstrut of a text; rather, it represents a soio-ognitive pattern that members of a partiular disourse

42 or professional ommunity typially use for the onstrution and understanding of that genre to ahieve speifi ommuniative purposes (Bhatia, 1993: 21). The individual s identifiation of genre struture thus depends on the reognition of onventionalized regularities in disourse organization inside a given ommuniative ontext. Considering that genre struture emerges from soioognitive onventions, it an also be argued that ultural differenes may exist in the typial rhetorial organization of a text type (e.g. Connor, 2000; Nikerson & De Groot, 2005). Vergaro (2002, 2005) onfirms this idea in her ross-ultural analyses of Italian and English business letters; in these studies, eah national ulture used different strutural elements to realize the same genre purpose. Similar results an be found in Zhu (2000), who douments ultural distintions in the organization of native English and Chinese sales letters. The rhetorial onstrution of a text is based on the ognitive onepts of moves, and individual tatis within these moves. Sine moves are onventional elements in the organization of a genre, they determine its prototypial struture, i.e. their presene in the text struture determines the reognizability and therefore integrity of a given genre. With regard to the advertising letter, for example, the moves establishing redentials, introduing the offer and soliiting response are ruial to the typifiation of this genre, whih aims to aumulate profit for orporate purposes (Bhatia, 2004: 95). In addition to the prototypial moves, the rhetorial struture of a text also ontains allowable elements used within these moves that are available to an author for reative or innovative genre onstrution (Bhatia, 1993: 32). These allowable elements are strategies by means of whih an author exeutes moves and by means of whih a genre thus beomes personalized; they annot atually alter the ommuniative objetive of the text. Distint from moves, individual tatis are therefore non-disriminative elements in the identifiation or reognition of a genre. The fat that the allowable tati indiating value of the offer is sometimes absent in the introduing the offer -move of a promotional letter does not neessarily eliminate the move and ertainly does not threaten the integrity of the entire genre. As for generi distintions between groups or ultures, Bhatia adds that these variations are likely to our at the level of individual strategies. He notes: loal ultural onstraints are unlikely to effet substantially the essential move-struture of a speifi genre; however, it is very likely that they will have signifiantly interesting impliations for the realization of ertain moves (1993: 38). In addition to a prototypial ontent and struture, genres also ontain a harateristi lexio-grammar. The generi analysis of linguisti features was first introdued by Halliday (1978), who expliitly foused on linguisti variation to distinguish between text types. Bhatia argues that suh language analyses prinipally offer a surfae-level text desription; he believes that they reveal very little ( ) about the way soial purposes are aomplished in and through [genres] in settings in whih they are used (Bhatia, 1993: 18). However, the investigation of lexiogrammatial elements remains a relevant aspet within the holisti approah to texts in the LSP genre tradition. That is, the empirial evidene gained through lexiogrammatial researh enables genre analysts to reah the final and most onrete stage of formal text typifiation, whih as suh is omplementary to the ontextual, substantive and strutural identifiation of a genre. Genre theorists argue that the lexio-grammatial nature of a text type is onstrained by the soial situation, inluding rhetorial funtion or ommuniative purpose (e.g. Bhatia, 1993, 2004; Swales, 1990). This an be illustrated by Solbjørg Skulstad s (2005) study of orporate environmental reports. Her findings show that

43 ation markers and deiti referenes are partiularly useful in guiding readers in their proessing of the informative aim of the report, e.g. as in This booklet gives readers some bakground fats about the main environmental poliies (Solbjørg Skulstad, 2005: 77). Additionally, as Hyland (2000) notes in his study on aademi abstrats, the rhetorial aim of a genre presents authors with linguisti possibilities whih an be rhetorially exploited. For example, he shows that in abstrats aimed at influening readers deisions about whether the enlosed artile is worth reading positive adjetives suh as great and important are inluded to laim signifiane of the artile topi. Aording to Hyland (2000), these lexiogrammatial items beome disursive markings that add to the persuasiveness of the text and thus to the realization of the ommuniative purpose of the text. The notion that the lexio-grammar of a genre is influened by the rhetorial ontext suggests that it is also subjet to ommunity-speifi or ulture-speifi preferenes. Several studies have onfirmed that variation in the relative frequeny of linguisti elements an result from language onstraints posed by disiplinary ulture (e.g. Dahl, 2004; Hyland, 2000) and national ulture (e.g. Nikerson, 2000; Vergaro, 2002, 2005). Studies aross national ultures have also shown that not only substantially distint ultures vary in their linguisti genre preferenes, but also those ultures whih have had frequent ontats and whose languages and ultures are superfiially relatively similar to one another, e.g. reognizable differenes between European ultures (Mauranen, 1993: 6). Several studies have speifially foused on the substantive, organizational or lexio-grammatial elements used to ahieve the onventional and strategi intentions of annual report texts. As for genre ontent, Rutherford (2005) looks at the themati ontent of the operating and finanial reviews in annual reports originating in UK orporations. He uses word frequenies to establish the atual themes and analyzes harged words to determine the ompanies evaluations of these themes. Furthermore, studies by Garzone (2004) and Nikerson and De Groot (2005) revolve around the identifiation of text strutures in management letters to stakeholders. They establish the rhetorial organization of the text by dividing it into onventional and optional fragments that support the informative and interpersonal aims of the text. Finally, the lexio-grammatial features of the annual report genre have been the topi of researh in Malavasi (2005), Rutherford (2003) and Thomas (1997). Malavasi (2005) examines evaluative adjetives, nouns and verbs that reveal how European banks enfore reader admiration for the organizational values, results and prospets refleted in the annual report. Rutherford (2003), moreover, onentrates on syntatial omplexity in the operating and finanial reviews of British annual reports in order to determine the degree of obfusation of poor performane. And Thomas (1997) investigates the use of verb onstrutions and agents in positive and negative news inluded in management forewords of annual reports by a US orporation Networks of genres Finally, a genre an be defined by its intertextual onnetions within a network of text types. Intertextuality plays a role in the development of new genres as well as in inidental or stabilized inter-generi referenes (Bargiela-Chiappini & Nikerson, 1999). As for intertextuality in genre evolution, Miller explains that the new is made familiar through the reognition of relevant similarities (1984: 29). In her view, when new situations are enountered whih do not math prior experiene or

44 existing knowledge for a partiular genre, language users ompose a new text type based on a seletion of referenes to texts typially used under onditions that are also reognized in the new rhetorial situation. Miller adds that if a new typifiation proves ontinually useful for mastering states of affairs, it enters the stok of knowledge and its appliation beomes routine (Miller, 1994: 29). For example, the virtual team room genre in organizational settings was inspired by existing text onventions for fae-to-fae meetings and ontats (Yates & Orlikowski, 2002). Similar to Miller (1994), Bakhtin (1986) believes that language users presuppose the ourrene of preeding utteranes. He assumes that their urrent utterane enters into one kind of relation or another (builds on them, polemiizes with them, or simply presumes that they are already known to the listener). Any utterane is a link in a very omplexly organized hain of other utteranes (Bakhtin, 1986: 69). As suh, Bakhtin s ideas aount for the inidental or onventional intertextual links between established genres, i.e. speakers who intentionally blend existent genres or instanes of a genre to plae their own aent on the text used. Through this proess of re-aentuation one genre is onstituted through another; or, (part of) the rhetorial format of one genre is embedded in another genre for its realization (Bakhtin, 1986: 80). This is the ase, for instane, when researh artile introdutions refer to prior aademi reports on ongruent researh topis (Hyland, 2000; Swales, 1990). Drawing on Miller (1994) and Bakhtin (1986), several lassifiations of generi intertextuality have been introdued more reently. In her study on disourse in a tax aounting environment, Devitt distinguishes three kinds of intertextuality: (1) referential intertextuality or linguisti referenes aross texts, e.g. linguisti referene to a lient s tax return in a letter to tax authorities, (2) funtional intertextuality or rhetorial oherene between texts, e.g. a lient s information request reating the need for a firm s response letter and (3) generi intertextuality or the full olletion of text types reurrently available for professionals disourse situations, e.g. all genres used by tax aountants (Bargiela-Chiappini & Nikerson, 1999; Devitt, 1991). Another yet overlapping lassifiation of intertextuality is provided by Fairlough (1992a, 1992b), who proposes a distintion between manifest and onstitutive intertextuality among genres: (1) manifest intertextuality inludes the (in)diret presentation of one or more text(s) in another text, e.g. Sir Edward said yesterday: and (2) onstitutive intertextuality involves the multigenre omposition of a situation, e.g. an eduational setting allowing for a mix of offiial lassroom genres and playground genres (Fairlough, 1992b). There are three onepts that are ommonly referred to for a further explanation of the atual networks established through intertextual relationships between genres. First, there is Devitt s (1991) onept of genre set, whih is pereived to ontain a olletion of different text types used by ommunity members to fulfill a speifi rhetorial need in their professional ontext. The genres inluded in the set reflet the ativities and soial relations inherent to the members professional funtion, e.g. the olletion of oral and written texts that tax aountants use to ompose a formal report on the firm s position on tax questions (Devitt, 1991; Spinuzzi, 2004). A seond onept overing the intertextual oherene between genres is Bazerman s (1994) genre system, whih seems to onern the full series of genres relied upon by all parties engaged in disourse aross ommunities or aross groups within a given ommunity. Genre systems arise from interdependent text types that are enated in sequene and whose funtions and forms are interloked, e.g. the olletions of texts that aountants produe and reeive in their ontats from and to lients, from and to authorities, and from and to

45 olleagues (Bazerman, 1994; Yates & Orlikowski, 2002). Thirdly, intertextual relations among genres are desribed in terms of genre repertoire. Aording to Orlikowski and Yates (1994), genre repertoire reflets all ommuniative praties available within a group, i.e. it ontains the multiple, various and interating text types that ommunity members have routinely enated over time. It appears that this onept overs both the funtional genre sets and the interative genre systems a ommunity an rely on: it would seem to omprise both the texts individual members use to fulfill their own ommunal funtion and the text these members use in interation with ommunity-internal or external ommuniators (Spinuzzi, 2004). Therefore, genre repertoire is pereived to inlude all texts members design and reeive within the frame of their ommunity-speifi position, e.g. the use of internal eletroni memos, dialogues, proposals, ballot questionnaires, ballot responses and ballot results within the ommunity of designers of an artifiial intelligene language (Orlikowski & Yates, 1994). Studies on annual reports have suggested that these texts are not independent genres, either; rather, they seem to be involved in different intertextual networks of related text types. Jameson (2000), for instane, points out that the annual report itself onsists of multiple sub-genres, suh as the letters to stakeholders, finanial highlights or diretors biographies. She finds that eah subgenre is a self-ontained entity involving unique narrators, purposes, readers and internal oherene; however, she also emphasizes that eah subgenre has ( ) not neessarily unbreakable links to what preedes and what follows and as suh ontributes to the onstitution of the full report (Jameson, 2000: 32). Aordingly, she defines the overarhing text type the annual report as a genre set. In her analysis of the proess of fundraising for aademi researh, Tardy (2003) plaes annual reports in a different intertextual ontext, relating them to the series of texts used in the interation between priniple investigators, funding agenies and the investigators aademi institutions. Here, the annual report is part of a genre system in whih it funtions as one of the funding ageny s works of referene for the investigator seeking finanial support. Yet another view on the intertextual nature of the annual report genre is refleted in both Rogers (2000) and Fox (2006). Rogers (2000), who fouses on the annual report as a genre of finanial reporting, argues that this text type is part of an extensive olletion of texts ommuniation or investor relations managers an rely on for finanial dislosures, e.g. analyst meetings, investor meetings, interim reports, press releases, funding plans. Although in Fox s study (2006) the annual report is defined more generally as a orporate publi disourse genre, she shows that as suh it also belongs to a large olletion of interrelated texts by means of whih ommuniation managers an profile the ompany, e.g. mission statements, media advertisements, business guidelines. Presumably, these studies asribe the annual report to two distintive genre repertoires, i.e. one more finane-related and one more marketing-related The annual report as a rhetorial genre Based on the theoretial definition of professional genres and of the annual report in partiular, a learer piture an now be drawn of the annual report genre as it is pereived in the present study. Generally, the soio-historial situation in whih the Duth-English and British-English annual reports are produed and onsumed is determined by two nation-level fators: the overarhing national ulture and the general business ulture. These ultures in turn influene the values and praties of

46 the ommunities in whih the annual report is of partiular use as a means of dislosure. With regard to the present Duth-English and British-English texts (and for many other business texts), the omposition of these ommunities is quite omplex: they need to enlose the expert produers of the annual report on the one hand and the variety of readers of the annual report on the other hand. For the Duth-English texts in speifi, these ommunities omprise writers related to a Duth ompany and a broad international reader audiene with different soioultural bakgrounds. And for the British-English texts, these ommunities ontain annual report produers assoiated with UK orporations and readers who belong to either national or international target groups. Within the writer and reader ommunities, the development of the Duth-English or British annual report is bound to rhetorial needs and rules for its existene. That is, the English annual report is produed and onsumed as a response to the soial purposes Duth and British orporations and their (inter)national stakeholders wish to ahieve. On the prodution side or in the disourse ommunity, writers of the Duth-English and British annual reports translate the rhetorial purposes of these reports into a olletion of interdependent texts that eah omprise speifi ommuniative purposes and reognizable ontents, strutures and linguisti harateristis. As suh, both the Duth-English and the British-English annual report would seem to represent a genre set. This is shematially illustrated in Figure 2.1. In light of the present attempt to offer a general desription of the annual reports originating in Duth and British orporations, it should also be observed that these reports may ontain genre features that respond to or overlap with other genres produed by ommuniation or investor relations managers in Duth and British ompanies. Figure 2.2 depits this form of intertextuality. It shows that the English annual report is likely to o-exist with other finanial ommuniation tools, e.g. press releases, interim reports, shareholder meetings or finanial websites but that it an also be used as a omplementary tool in a mix of marketing ommuniation instruments, suh as produt folders, orporate brohures, radio and TV interviews or sustainability reports. In both ases, the annual report as a genre set is a potential element in a orporation s genre repertoires. Inside suh networks, ongruent texts might share subordinate objetives, substantive or formal text features and even ontextual denominators (e.g. text produers, national values), yet eah text is unique, espeially onerning its disourse situation.

47 Soial ontext = national ulture Soial ontext = general business ulture Disourse ommunity = text produers Global reader ommunity = (inter)national readers Disourse situation = rhetorial need for AR 2.4 Conlusion Genre (set) = AR text purpose ontent struture lexio-grammar Figure 2.1: Shemati representation of annual report (AR) genre-in-ontext Genre (set) = AR: text purpose, ontent, struture, lexio-grammar e.g. Genre = Letter to Shareholders: purpose, ontent, struture, lexiogrammar e.g. Genre = Operational review: purpose, ontent, struture, lexio-grammar Genre = Shareholder meeting Genre = Print advertisement Genre = Finanial press release Genre = Interim report Genre = Corporate brohures Genre = Sustainability review Genre repertoire = Genre repertoire = Finanial ommuniation tools Marketing ommuniation tools Figure 2.2: Shemati representation of annual report (AR) in genre repertoires

48 2.4 Conlusion The previous setions provided an overview of theoretial oneptualizations of texts in relation to their soial surroundings. The onepts disussed i.e. soial onstrutionism, ommunity and rhetorial genre formed the sope for text identifiation that is entral to the present study. The general relevane of the generi approah to texts in professional environments was already implied by some of the studies referred to (e.g. Bazerman, 1994; Bhatia, 1993; Yates & Orlikowski, 1992): knowing the motivational, substantive and formal requirements of a given text type in a partiular professional situation enables language users to interpret the meaning of what is said or written in that setting. Familiarity with a professional genre thus implies the ability to behave appropriately in a speifi rhetorial ontext and thus failitates the ommuniation proess among users of professional genres (e.g. the writers and readers of Duth-English and British annual reports). The definition of professional texts-as-rhetorial-genres indiates that the generi approah to texts provides more than just a theoretial framework for disourse analysis. Apart from offering theoretial onepts on whih suh an analysis an be built, genre theory appears to determine what the analysis should fous on in the identifiation of text types. This ambiguous role of the genre perspetive is illustrated in Luzón s (2005) review of genre researh in tehnial ommuniation, whih relates genre to the onepts of lines of researh, methods of analysis and topis of analysis. He onludes that genre is a pratial analytial ategory whih provides the appropriate framework for researhing and understanding tehnial ommuniation issues (2005: 292). Similar ideas are presented in Yli-Jokipii (1998) and Charles (1998), who suggest that genre theory supplies the terminology and methodology for studies on business disourse. As Yli- Jokipii indiates, the genre approah involves the reation of oneptual frameworks by whih business writing problems an be investigated (1998: 97). Nikerson (2000) is one of the few to have expliitly distinguished the ontologial and analytial relevane of the LSP genre model. She refers to soial onstrutionism, genre as typified soial ation and struturation as the fundamental theoretial onepts for the analysis of internal business orrespondene, and fouses on the generi ontext, ontent, lay-out, medium, rhetorial struture and lexio-grammar as the main items of analysis. Regarding LSP genre theory as a method of analysis with a theoretial ground, it an be argued that it is muh more a tool of inquiry than a theoretial framework. Drawing on Gee s desription of soial disourses, genres an be defined as thinking devies that guide us to ask ertain sorts of questions (1999: 37). With respet to the present study, the genre perspetive thus direts the researh steps needed to investigate the international business English in annual reports, while handing a theoretial explanation of the basi and important features of the disourse. This will be elaborated upon in Chapter 3, whih presents the researh design for the rossultural omparison of Duth-English and British annual report texts. As a means of inquiry, the LSP genre model has both strengths and limitations, the most important of whih generally onern its thik, situated desription of text types on the one hand, and the risk of lapsing into atomisti or presriptive typologies of texts types on the other (e.g. Badger & White, 2000; Bhatia, 1993). The main potential weakness of text presription an be overome by a ontinual awareness of the dynami soial situation in whih texts our and the impat this has on all text levels (Bhatia, 1993); this study therefore seeks to remain

49 onsious of the influene of genre ontext. There is yet another ruial diffiulty in the genre model that has been the topi of urrent debates on disourse analysis: its relevane for multimodal texts. As the LSP genre analysis originates in an ationrelated, purpose-driven and holisti approah to texts, it prinipally overs all reognizable types of texts, inluding purely verbal texts, verbal texts with graphs or still images, verbal utteranes pronouned in moving images, hypertext, et. (Askehave & Nielsen, 2005; Bateman, Delin & Henshel, 2006; Bhatia, 2004). Until reently, however, genre studies had a mono-modal harater, they exlusively foused on spoken or written language in professional texts. As a result of this onsistent one-dimensional view on disourse, the LSP model was found not to be fully equipped for the growing variety of multimodal texts in whih analysts presently plae an interest. In order to apture the design of multimodal language use, the genre model was therefore adjusted (e.g. Askehave & Nielsen, 2005) or further defined in terms of suitable levels of text desription (e.g. Bateman et al., 2006). In partiular, questions seem to remain about the position of visuals in the generi move struture and the labeling of visual lexio-grammatial elements. Nevertheless, several genre-based studies have suessfully explored the relationship between the verbal and the visual, offering important initial insights in the onept of the multimodal genre. For instane, some have analyzed the visual ontent based on the objets observed and then determined the funtional meaning of this ontent within the multimodal genre, e.g. Rowley-Jolivet s analysis of tables, graphs and pitures in sientifi onferene papers (2002), Jameson s study of photographs used in annual reports (2000) or Thurlow and Jaworski s (2003) examination of globalization-related pitures in in-flight magazines published by different airlines. Others have linked visual ontent to the strutural or linguisti design of the genre, e.g. the assoiation between photographs and rhetorial tatis in Askehave and Nielsen s analysis of orporate homepages (2005) or the onnetion between photographs and deiti referenes in Fuertes-Olivera, Velaso- Saristán, Arribas-Baño and Samaniego-Fernández s (2001) investigation of print advertisements. The present study builds on these examples and seeks to ontribute to the further extension of the LSP genre model as a means of inquiry for multimodal disourse. Having defined the oneptual and analytial frame of this study in the previous subsetions, the next hapter will fous on its appliability to the present researh design. Following Nikerson (2000), Chapter 3 will relate the generi onepts of soial ontext, ontent, struture and lexio-grammar to the analytial steps taken in order to find an answer to the first main researh question introdued in setion 1.5.

50

51 Chapter 3 Researh design for the ross-ultural study of English annual reports As was indiated at the end of the previous hapter, the LSP genre-model and the onepts assoiated with it are ruial to the first and seond researh phases of the present study (see setion 1.5). In general, these onepts are relevant for the methodology used to find an answer to the first main researh question: What are the genre harateristis of English, non-finanial texts in multimodal annual reports produed by Duth orporations and in what way do these differ from the genre harateristis of English, non-finanial texts in multimodal annual reports produed by British orporations? More speifially, the genre onepts are at the basis of the researh designs needed to answer RQs 1a-1d and 2a-2d, whih respetively fous on the similarities and differenes between (1) the ommuniative ontext and (2) the multimodal text features of Duth-English and British annual reports. Sine LSP genre theory provides a useful means of inquiry for disourse analysis, i.e. it determines the fous for the identifiation of texts in a partiular disourse ontext, it is used here as an analytial instrument for the ross-ultural investigation of English annual reports originating in Duth and British multinationals. Consequently, the idea that underlies the methods disussed in this hapter is that texts an be identified as situated ations with a onventional rhetorial purpose and a prototypial ontent, struture and lexio-grammar (e.g. Bhatia, 1993; Miller, 1984; Swales, 1990). The following setions elaborate the researh steps taken for the ontrastive ross-ultural analysis of Duth-English and British annual report texts; they suessively desribe the methodologial approah taken in the generi haraterization of English annual reports (setion 3.1), the multimodal nature of the analysis (setion 3.2), the sampling of annual report texts (setion 3.3), the registration of ourrenes of text features and the statistial rossultural omparison of these features (setion 3.4), and the interoder tests arried out to determine the reliability of the researh results (setion 3.5). It should be noted that this hapter does not inlude an explanation of the tehniques applied to answer the seond main researh question of this study, i.e. onerning the UK-based reader s response to the multimodal genre features in Duth-English and British annual report texts (setion 1.5). This seond question involved a different, interultural researh setting whih required a different methodology than the ross-ultural text analyses that are disussed below. Whereas the ross-ultural investigation of annual reports was primarily writer-oriented and desriptive, the interultural study of these texts foused on the readers and their reeption or evaluation of the texts. In generi terms, the ross-ultural analyses were intended to reveal the rhetorial onventions in the Duth-English and British annual report texts, while the interultural analysis was designed to investigate and ompare the rhetorial impat of these Duth-English and British text onventions on their (inter)national readers in the UK (e.g. Ahuvia, 1998; Paul, Charney & Kendall, 2001). Beause Chapters 4-8 all refer to ross-ultural text analyses and their impliations, the researh design for the interultural reader response study is presented prior to its findings in Chapter 9.

52 3.1 Generi haraterization Assuming that the annual report is a rhetorial genre within the professional domain, the methodology for the present ross-ultural analysis of Duth-English and British annual report texts enompassed four levels of text features. These levels are: ontext, ontent, struture and lexio-grammar. In setion 2.3, an overview of LSP genre studies showed that rhetorial text types are generally marked by (1) a partiular soial ontext, inluding the typial ommuniative purpose of the text, (2) a speifi substane or informational landsape that indiates what the text is about, (3) a speifi text struture, onsisting of rhetorially funtional text fragments and (4) a pattern in the use of ertain lexio-grammatial elements (e.g. Bhatia, 1993, 2004; Miller, 1984; Swales, 1990; Yates & Orlikowski, 1992). In order to establish a thik desription of the annual report as a genre in the Duth and British national business ultures, all four levels of analysis need thus be inorporated in the urrent researh design (Bhatia, 1993). Within genre studies, the position of the ontextual, ontent, strutural and lexio-grammatial levels of analysis may differ, depending on whether the transript (i.e. text) or the soial system in whih the transript ours is taken as the starting-point for investigation. Aording to Askehave and Swales (2001), there are two proedures that determine the staging of a generi text analysis. First, they introdue the text-based or linguisti proedure, whih starts with the identifiation of struture, style, ontent and text purpose and moves upwards towards a desription of the text in its soio-historial ontext. And seond, they propose the ontext-based or ethnographi proedure, moving downwards from the examination of soial or ommunity-speifi onditions, values and norms towards the purpose, substane and form of partiular text types (Askehave & Swales, 2001). Some genre analysts have primarily referred to one rather than the other of these proedures in their desription of a text type. Lassen (2006), for instane, relies on Swales and Askehave s text-based proedure. She investigates formal text features suh as onjuntions, negations and rhetorial struture to determine the ommuniative purposes of press releases in biotehnology. Askehave and Swales, however, suggest that opting for either one of the two proedures may not ause optimal results in the identifiation of text types. They argue that genre analysis should ombine the two in order to aount for the evolving and dynami nature of genres: genre ategorization [ ] is more a matter of text-in-ontext inquiry than straightforward textual [ ] srutiny on the one hand, or introspetive system building, on the other (2001: 209). An example of suh a text-in-ontext based study is Rowley-Jolivet and Carter-Thomas (2005), who analyze the link between the ontext of onferene presentation introdutions and both their move strutures and subjet pronouns. Additionally, Nikerson (2000) onduts a survey of the orporate ontext for written and eletroni business texts in an Anglo-Duth ompany and speifially relates this ontext to the struture and the linguisti style of messages. Considering that the present genre study aims to explore English annual reports about whih relatively little is known to date, it seemed reasonable to start from a general inventory of situational and funtional onstraints and then ontinue with a set of ontextualized disourse analyses of the annual report texts (i.e. analyses of ontent, struture and lexio-grammar). Therefore, this study follows the text-in-ontext-based proedure used in Rowley-Jolivet and Carter- Thomas (2005) and Nikerson (2000).

53 Having deided on the researh proedure for the urrent investigation of professional texts, the LSP genre model provides a range of analytial tehniques that an be used to exeute this proedure. This is partiularly true for the tehniques available for the analysis of the rhetorial ontext, ontent, and lexio-grammar of a genre. Whereas the analysis of rhetorial text struture in priniple is bound to a psyholinguisti or ognitive method, i.e. the observation of moves and allowable tatis, the tehniques for the study of ontext, ontent and lexio-grammar an be drawn from various disiplines, suh as soiology, psyhology, soio-linguistis or marketing (Bhatia, 1993; Charles, 1998; Swales, 1990). Apart form the psyholinguisti tehnique applied for the identifiation of text struture in Duth- English and British annual reports, the present study employed three additional tehniques to analyze text ontext, ontent and lexio-grammar: (1) ethnography Duth and British informants views on the purpose, audiene and prodution proess of the English annual reports, (2) ognitive mapping an analysis of onept-based themes in these reports and (3) linguistis an investigation of lexio-grammatial markers of text organization and interation inluded in the annual reports. Table 3.1 further illustrates all the distintive tehniques used at eah level of analysis. It shows the ohesion between the speifi researh questions provided in setion 1.5 and the orresponding researh tehniques, as well as the reasons for their implementation. Detailed motivations and desriptions of the researh methods an be found in the hapters separately overing eah level of analysis (Chapters, 4, 5, 6 and 7). Table 3.1: Methods for the ross-ultural analyses of the English annual report texts. Researh questions (Ch.1) Contextual identifiation and omparison of the Duth- English and British annual reports (RQs 1a 1d). Level of analysis Context Method/tehnique Motivation Chapter Qualitative survey among listed Duth orporations. In-depth interviews with Duth and British listed orporations. Registration of behavioral pattern (i.e. English language poliy) in relatively many ases. Detailed insights in informants soial knowledge of purpose, readers and prodution proess of annual report. Ch. 4 Identifiation and omparison of the multimodal disourse in the Duth- English and British annual reports (RQs 2a 2d). Content Struture Corpus analysis of oneptual themes, based on ognitive-psyhologial proess of mental mapping. Corpus analysis of moves and strategies, based on ognitive organization of the text. - Insight in ognitive + semanti nature of genre ontent. - Coneptual approah allows for situated, multimodal themes. Explains how text fragments help realizing ommuniative purpose of the full text. Ch. 5 Ch. 6 Lexiogrammar Corpus analysis of metadisourse items, based on rhetorial interpretation of lexion and grammar. - Interpretational frame for the language features of genre. - Link between lexiogrammar and rhetorial purpose of text. Ch. 7

54 It should be noted that the present genre investigation omprised both a multimethod and a multidisiplinary approah. Most LSP-based genre studies have foused on qualitative i.e. detailed and ontextualized - text identifiation; this is beause of the thik desription of texts that underpins the genre philosophy (Bhatia, 1993; based on Geertz, 1973). Some genre analysts, however, have also emphasized the importane of quantitative methods. They suggest for instane, that quantitative data point at salient genre harateristis (Swales & Rogers, 1995). Or they endorse the idea that quantitative researh methods are empirially grounded and validated, and thus more ompelling (Rutherford, 2005: 354). The relevane of this approah in LSP genre studies is illustrated in, for instane, Rutherford s (2005) orpus analysis of harged themes in UK annual reports and Herring and Paolillo s (2006) multivariate analysis of gender-related language style in aademi web logs. Furthermore, the multidisiplinary nature of the urrent genre analysis entails partnership researh, whih inludes an epistemologial and methodologial merging of related disiplines. As Bargiela-Chiappini and Nikerson explain, a multidisiplinary approah offers the opportunity to work towards a ertain degree of methodologial and theoretial onsolidation or onvergene (2002: 276). An example of a multidisiplinary genre study is Swales and Rogers (1995), where both linguisti and management insights are referred to in order to identify different orporate mission statements. 3.2 Approah to multimodal text analysis The fat that ontemporary annual reports generally ontain both written texts and visual illustrations (graphs, diagrams, tables and photographs, see also setion 1.3) means that a genre study of suh reports needs to inorporate both textual and visual disourse. This is also onfirmed by Jameson s (2000) study of verbo-pitorial relations in the subgenres of annual reports published by US-based organizations. She finds that eah narrative element struture, theme, style, narration and subgenre mix has both verbal and visual manifestations that interat with one another (2000: 33). Jameson observes, for instane, that the struture in annual report texts is realized through typographial marks, bulleting or headings, that textual information is highlighted in graphs or harts, and that the reation of narrators and sender-reader interation ours by means of management portraits. In the present study on Duth-English and British annual reports, the oherene between verbal text and visual illustrations was analyzed on the basis of Kress and Van Leeuwen s (2001, 2006) oneptualization of multimodal disourse analysis. This subsetion explains how their approah to multimodal text analysis has influened the present investigations of genre ontext, ontent, struture and lexiogrammar. The full explanation is summarized shematially in Figure 3.1. Aording to Kress and Van Leeuwen, the analysis of multimodal disourse omprises the examination of omposite texts that aquire their meaning from multiple modes, where a mode onerns any material resoure whih is used in reognizably stable ways as a means of artiulating disourse (2001: 25). The resoures inluded in a multimodal text an vary from general modes e.g. writingas-mode or images-as-mode to speifi modes e.g. objets-as-mode, olor-asmode, sene-as-mode or events-as-mode in photographs. Whether something is a mode depends on the text type in whih it is used. For instane, the earth olors of furniture portrayed on a front page of a home magazine or the soft golden-brown

55 olors surrounding the romanti ouple in a offee advertisement may artiulate the immaterial produt values of nostalgia and bliss, thus ausing the olor-as-mode to be both prominent and signifiant. However, the olor-as-mode need not have the same important disursive funtion in all over pages or ads. In ads fousing on the material produt itself, for example, the objet-as-mode is supreme (Kress & Van Leeuwen, 2001, 2006). Moreover, Kress and Van Leeuwen argue that the multiple modes in a text should not be analyzed as disrete parts adding to the sum of meaning in the text; rather, the parts should be looked upon as interating with and affeting one another (2006: 177). Again referring to the offee ad, they explain that the brief produt-related text and the relatively large portrait of a man and woman snuggling together have integrated meanings as the photograph expresses and emphasizes the promise of what might be when using the produt desribed in the adjoining text fragment, i.e. romanti intimay (Kress & Van Leeuwen, 2006). The above examples show how the different modes interat in multimodal texts: they may reinfore eah other by arrying the same message in different ways, they may fulfill omplementary roles by providing extra information about the same thing or they may be hierarhially ordered, meaning that one mode is dominant over the other in the interpretation of a text (Kress & Van Leeuwen, 2001). As was already outlined in setion 1.3, the multimodal harater of the annual report genre is generally intended to failitate the reading proess and to open up ways of presenting a orporate truth. Here, photographi illustrations in partiular draw attention to ertain text details and also allow for a more tangible expression of the ultural notions underlying the textual desription of the ompany, the industry it operates in, and the business praties it adheres to (e.g. David, 2001; Jameson, 2000; Preston et al., 1996). The reprodution of orporate truth in annual reports is ommonly established through the inlusion of pitures of real-life things, i.e. doumentary pitures. This is beause realisti photographs (...) telling an extended story onstitute a doumentary style, one that an suggest the standards of auray and honesty to whih reports are held (David, 2001: 21). Sine doumentary photographs seem to apture reality, they form a believable way of telling the ompany s story. Although they may atually distort or omit aspets of reality in order to onstitute a new ompany-speifi reality, the natural settings depited in doumentary photographs presumably establish an atmosphere of orporate authentiity and, hene, reader aeptability (David, 2001; Guthey & Jakson, 2005; Preston et al., 1996). Moreover, David argues that the gravity of the message enhanes the persuasiveness of photos in annual reports: the ethos of the reports as onveyors of finanial data adds redibility to the promotional setions [where] the elaborate visuals reate a positive, emotional ontext for interpreting the fatual data presented in the text (David, 2001: 217). As an illustration, in his qualitative survey of annual reports of the Burton Group, MKinstry (1996) finds that the ombination of doumentary photographs of shop interiors, employees and ustomers with numbers of stores in adjoining text exerpts not only ontributes to the informative or fatual dislosure on operational performane; it also ontributes to the reation of a solid, safe image of Burton as one of the largest fashion retailers in the UK. Considering the rhetorial importane of doumentary photography in the informative and promotional purposes of the annual report genre (e.g. David, 2001; Graves et al. 1996), the present multimodal analysis entered on the relationship between texts and doumentary photographs in Duth-English and British annual report texts. Furthermore, this study of the multimodal annual report genre revolved around a seletion of meaningful modes embedded in the texts and doumentary

56 photographs analyzed. With respet to the texts in the English annual reports, the analysis foused on the mode of written language, or the writing-as-mode. This general mode enlosed both the running texts or narratives in the reports as well as the text fragments adjoining the realisti photos. Apart from telling its own story, the writing-as-mode funtions as a ontextual deixis or an interpretative frame for the oexisting photographs; it is used to set the tone, to indiate how the other modes and the objets and strutures whih they artiulate are to be read and understood (Kress & Van Leeuwen, 2001: 31). In other words, the written texts provide a partiular pointing for the meaning-making of images (Kress & Van Leeuwen, 2001: 32). As for the investigation of doumentary pitures, a more speifi hoie in modes seemed to be appropriate, sine eah situated piture in priniple yields an assemblage of signifying visual modes (Kress & Van Leeuwen, 2001). It was determined that the present multimodal study should fous on the oherene between writing-as-mode on the one hand and the photographi objet-as-mode, ativity-as-mode, sene-as-mode, emotion-as-mode and ompositional framing-asmode on the other hand. This seletion of modes may not have overed or fore grounded the full olletion of visual modes represented in the Duth-English and British researh materials. However, it allowed for the interpretation of all pitures inluded in the researh materials and failitated the identifiation of what was showased in the photos (e.g. David, 2001). In other words, sine the seleted visual modes omprised fairly straightforward, reognizable pitorial elements, this would seem to result in a set of relatively unambiguous, onrete image interpretations. Additionally, it was antiipated that these four speifi modes would provide suffiient insight in visual meaning so as to be able to identify any ross-ultural differenes in photography. To identify exatly whih pitorial elements would have to be related to the different photographi modes, Smelik s (1999) definition of image omponents was referred to. In her framework for visual analysis, she suggests that the manner of representation in images is determined by image omponents suh as mise-en-sène and amera perspetive. Aording to Smelik (1999), mise-en-sène omprises the omplex set of framed objets, subjets and ativities. It enompasses prodution design in terms of props, settings, ostumes, faial expressions, behavior and plaement of props. Smelik (1999) ategorizes both size of frame and amera angle as elements of amera perspetive. The size of frame involves the depth of field reated by the amera, i.e. it determines viewer-portrait intimay through lose-ups, medium shots and long shots. The angle of the amera suggests whih perspetive the viewer should take towards the visual ontent; bird s views, for instane, highlight the powerlessness of the portrayed item, whereas frog s views emphasize the impressiveness of the portrayed item. Table 3.2 illustrates how Kress and Van Leeuwen s modes of disourse and Smelik s image omponents were ombined to establish the analytial fous in the photographs. Aordingly, the multimodal meaning of photos in annual reports was defined by analyzing the persons, faial expressions, buildings, equipment, professional or natural environment, ativities, professional or soial events and framing tehniques refleted in the images. The messages in the surrounding running texts and adjoining text fragments were used to aquire the full meaning of the doumentary photographs.

57 Table 3.2: Researh fous for analysis visual metadisourse visual modes and omponents, based on Kress and Van Leeuwen (2001, 2006) and Smelik (1999). Mise-en-sène Objet-as-mode - Main people/objets: gender, skin olor, hair olor, physial appearane objet. - Costume: lothes. Ativity-as-mode - human behavior, movement. Sene-as-mode - Environment: physial ontextual elements. - Position people/objets: plae of person/objet relative to surroundings. Emotion-as-mode - faial expression, posture. Camera perspetive Framing-as-mode - Frame size: lose-up, medium-shot, long-shot - Camera angle: eye-level view, low angle, high angle. As will be indiated in Chapters 5-7, the visual objet-as-mode, ativity-as-mode and sene-as-mode were partiularly useful in the more abstrat analyses of oneptual themes (Chapter 5) and strutural elements (Chapter 6). The visual emotion-as-mode and framing-as-mode were also relevant in the onrete analysis of lexio-grammatial elements (Chapter 7). In the present study, the different textual and visual modes were used to identify the written and photographi genre features within Duth-English and British annual reports. This is shown in Figure 3.1. First, the ontextual investigation onsidered the ommuniative aims, target groups and design proess of the written texts as well as of the photographi ompositions in the annual reports. Next, the analyses of text features depended on the language in the writing-as-mode and the photographi elements in the five speifi visual modes to suessively determine the themes, moves/tatis and metadisourse in the multimodal annual report texts. Multimodal genre analysis Text: writing-as-mode Doumentary photography: visual objet-as-mode, ativity-as-mode, sene-as-mode, emotion-as-mode, framing-as-mode Context: purpose, audiene, design (Ch.4) Content: themes (Ch.5) Struture: moves/ tatis (Ch.6) Lexio-grammar: metadisourse (Ch.7) Figure 3.1: Multimodal genre analysis of annual reports.

58 3.3 Seletion of annual reports and speifi texts As was disussed in the first hapter, the present researh projet fouses on nonfinanial texts that form the persuasive underurrent of Duth-English and British- English annual reports. It was noted that sine the introdution of the International Finanial Reporting Standards whih were proposed in 2003 and formally standardized the finanial aounts from 2005 onwards European orporations are inreasingly dependent on the non-finanial setions of their (English) annual reports as tools of impression management (e.g. IASPlus, 2005; Nikerson & De Groot, 2005). Several studies both within Europe and beyond have indiated whih setions of the annual report are likely to ontribute to its promotional potential. Hyland (1998b), Nikerson and De Groot (2005) and Thomas, (1997), for instane, demonstrate that the managerial letter to stakeholders aims to influene reader pereptions of the ompany and to establish reader ommitment to that same ompany. In addition, Rutherford (2005) argues that the operational and finanial review is used to establish a positive orporate image. Brammer and Pavelin (2006) and O Donovan (2002) stress that the environmental dislosure setion is useful for enhaning the orporate reputation and for laiming legitimay. With regard to the speifi persuasive funtion of photographs in the non-finanial annual report texts, moreover, Benshop and Meihuizen (2002) and Anderson and Imperia (2002) show that images an reflet the soial or ideologial onstrution of the orporate identity. Graves et al. (1996) add that pitures in annual reports may also ontribute to a ompany s reputation management by depiting it as a powerful or redible organization with high-quality produts. Studies suh as the above have indiated that different setions and aspets in the annual report ould be part of the researh materials for the present multimodal genre investigation. In order to reate a sample of researh texts that would math the disursive situation of both the Duth-English and British annual reports, the sampling proedure was primarily based on the interview results in the ontextual analysis (whih will be fully desribed in Chapter 4). During the interviews, Duth and British informants were asked to desribe separately eah annual report setion, produing a list of texts with various purposes and readers. They pointed out that four setions in the Duth-English and British annual reports in partiular add to the ompany s impression management: (1) the Chairman s statement in the British annual reports; (2) the Chief Exeutive Offier s statement in the Duth-English and British annual reports; (3) the orporate profile desriptions in the Duth-English and British annual reports; (4) the operational review in the Duth-English and British annual reports. It was suggested that these setions ontribute to the general promotional potential of the annual report, as their multimodal design is bound to relatively few regulations. Also, the informants laimed that these texts did not aim at any speifi soial or politial interest group (whih would have narrowed down their promotional sope). Consequently, the English management statements, orporate profiles and operational reviews were seleted as researh materials in the present genre study. Although the study by Nikerson and De Groot (2005) revealed some rhetorial differenes aross the Duth-English statements by the Chief Exeutive Offier (CEO), the British Chairman s Statements and the British CEO s Statements, they were all inluded as management statements in the urrent

59 researh soures. The ombination of statement types is supported by the finding that the Duth CEO s Statements over muh of the rhetorial ground realized in the British Chairman s and CEO s Statements ombined (Nikerson & De Groot, 2005: 341). The four text types that formed the fous of the present researh projet were derived from Duth-English and British annual reports reording the orporate ativities and results in the 2003 finanial year. Sine the finanial alendars in the United Kingdom and the Netherlands appeared to be asynhronous in some ases, the study inluded annual reports with losing dates ranging from 1 July June At the time of the sampling proess, all orporations whose annual reports were eligible for seletion were either registered on the Amsterdam Stok Exhange or on the London Stok Exhange. The sampling methods used for the seletion of English annual reports, resembled those applied in prior researh on ultural influenes in finanial reporting (e.g. Hussein, 1996). They inluded a mix of purposive sampling and luster sampling (e.g. Neuendorf, 2002): after olleting units of analysis that met the seletion requirements within the wider population of Duth-English and British- English annual reports, these units were divided among relatively homogenous lusters. First, the sampling of annual reports was based on predetermined riteria involving market value or market apitalization (the number of issued shares multiplied by their market prie), English language poliy and Anglo-Duth merger history. These topis were used as starting points for the purposive sampling proedure beause (1) a orporation s market value impats the legal requirements for finanial reporting, the use of the annual report on international markets and therefore the pratie of English as an international business language, and (2) a orporation s Anglo-Duth merger history often results in British-English language bases for the annual report, thus obsuring the influene of Duth-English in its prodution proess (this has happened, for instane, in Unilever, Reed Elsevier and Shell). Considering these issues, the exat riteria for the purposive sampling of Duth-English and British annual reports were as follows: - the annual reports were published by traditional Duth or British orporations, i.e. ompanies whih had their original headquarters in the Netherlands and whih still had their headquarters in the ountry of origin at the time of text sampling. Therefore, ompanies suh as Wolters Kluwer or Philips were inluded in this study, while Anglo-Duth ompanies suh as Reed Elsevier or Shell were exluded; - the annual reports were published by Duth and British orporations with a large to mid-size apital base; - the Duth and British orporations realized an original English version of the 2003 annual report, i.e. English was used as the base language for the annual report. The purposive sampling proedure was sueeded by the industry-related lustering of English annual reports by Duth ompanies of different apital size. Sine this study aims to provide a general omparative overview of genre harateristis in Duth-English and British annual reports, moreover, all industries represented in the sample of Duth-English texts were mathed with idential industries represented in the British-English texts. These mathes did not always omprise orrespondene in market value; as there was no industry-speifi or ountry-speifi onsisteny in disrepanies for this riterion, it was antiipated that market value would not have a onsiderable effet on the present researh results. Apart from allowing a generalization of results aross diverse setors, the luster sampling proedure also

60 ontributed to the omparability of researh materials. Table 3.3 shows whih industries and orporations were inluded in the present omparative study. The industry labels were derived from The Forbes Global 2000 (Forbes, 2003). Table 3.3: Duth and British orporations inluded in the researh sample, per industry. Industry Duth orporations British orporations Bank ABN AMRO HSBC Insurane AEGON Aviva Diversified finanials ING Old Mutual Chemials AKZO NOBEL ICI Teleom servies KPN Versatel BT Vodafone Media VNU Pearson Wolters Kluwer WPP Food and drug retail Numio Boots Ahold Sommerfield Transportation TPG Exel Business servies and supplies Randstad Rentokil Initial Vedior Hays IT, Software and servies ASML Invensys Getronis Sage Group ASM Inernational Maroni Trading ompanies/retailing Hagemeijer Van der Moolen Wolseley Inhape Property investment Rodamo Europe British land Conglomerates Philips Smiths Group Speialty retail Hunter Douglas KingFisher Constrution IHC Caland Balfour Beatty As Table 3.3 indiates, both sampling proedures resulted in a seletion of English annual general reports published by 44 ompanies in total, i.e. 22 Duth ompanies in 15 different industries and 22 British equivalents. For the analyses of genre ontent, struture and lexio-grammar, all Duth-English and British management statements, profiles and operational reviews were extrated from these 44 annual reports. Table 3.4 presents the outome of this final seletion proedure, projeting the number of eligible texts by English language variant. It should be observed that the number of texts differs per disourse mode analyzed. Sine not all 137 annual report texts involved in the investigation of the running written texts ontained photographs, the number of materials for the pitorial analysis was smaller (i.e. 97). In turn, not all 97 texts holding photographs inluded text fragments adjoining these images. Therefore, the total number of units for the analysis of photo-speifi text fragments was 73.

61 Table 3.4: The number of texts for the study of running texts, photos and photo-related text fragments. Chairman s Statements CEO s Statements Corporate profiles Running texts Duth-English British-English Photographs Duth-English British-English Texts with photo Duth-English British-English Operational reviews Total Cross-ultural data analysis: qualitative and quantitative desriptions Whereas the first part of the ross-ultural desription of Duth-English and British annual report texts i.e. the ontextual investigation (Chapter 4) inluded a primarily qualitative interpretation of data, the identifiation of themes, move struture and metadisourse (Chapters 5-7) predominantly involved a quantitative proessing of data. After a brief disussion of the qualitative data analysis applied in the haraterization of genre ontext, this subsetion explains the quantitative perspetive maintained throughout the generi analyses of substantive and formal text features. Assuming that disourse situations are soio-ognitive onstruts, the analysis of genre ontext annot solely depend on the tangible information available in the text itself (Flowerdew, 2005). Therefore, the present study foused on qualitative methods that would allow insight into the mental onstrution of soial reality within the disourse ommunity reognizing and produing the genre of the English annual report. A quantitative inventory of the disourse situation of a text generally redues this situation to fixed or isolated data that hardly explain the (ognitive) ausal relationship between the soial ontext and disursive hoies in the text (Wendt, 2002). Aordingly, a quantitative approah to the analysis of disourse ontext would not really reveal the rationale underlying the onrete text features in the researh materials of the present LSP-based genre study. By ontrast, qualitative tehniques tend to offer a more exploratory view on the interpretational patterns ommunity members ommonly rely on when establishing the soial signifiane of a partiular text (Wendt, 2002). In the ase of a ontextual investigation of annual reports, a qualitative approah to data proessing thus permits a omprehensive exploration or desription of olletive systems. In line with Wendt s (2002) suggestions for the methodology in foreign language studies and Bhatia s (2002) omments about analytial perspetives on applied genre researh, the present study relied on a qualitative survey and semi-strutured interviews to obtain information about the ommuniative ontext and writing praties of expert members in the disourse ommunity within whih the Duth- English and British annual reports are realized. A full aount of the qualitative ontext analysis is provided in Chapter 4.

62 The identifiation of genre-speifi text features took plae by means of orpus analyses that primarily omprised quantitative methods of data proessing. LSP genre theorists initially ritiized quantitative orpus-based analyses, as they believed it revolved around rigid text oding that did not aount for the dynami nature of texts within their soio-historial ontext (e.g. Swales, 1990). To date, however, genre studies in professional disourse have repeatedly inluded a quantitative approah to orpora in order to underpin the reliability and exatness of results (Upton, 2002; Henry & Roseberry, 2001). Flowerdew (2005) adds that quantitative orpus analyses in priniple allow for disourse-level investigations of a large number of omplete texts, within the frame of the ethnographi or soioultural dimension of text usage. For example, Basturkman (1999) looks at key word onordanes to detet the topis disussed in strutural elements of book blurbs for literature on English Language Teahing (ELT) and, hene, tries to establish what the ontemporary issues were in ELT pedagogy. Upton (2002) and Flowerdew and Dudley-Evans (2002), moreover, present absolute frequenies of ourrene of move struture elements to show how text struture helps to realize the ommuniative purpose of diret mail letters and editorial letters. As a final illustration, Dahl (2004) studies the frequenies of lexial text organizers and their onordanes to reveal the relative refletion of writer and reader roles in researh artiles produed in multiple languages and aademi disiplines. Suh genre studies inlude relatively large orpora but maintain the viewpoint of text-in-rhetorial ontext, in partiular through analyzing and tagging word strings or single linguisti elements whih are rhetorially funtional within the struture of the text (Flowerdew, 2005). Similar to the above examples, the present genre study also ontained frequeny ounts to determine whih substantive and formal features were typial to the multimodal orpora of Duth-English and British management statements, profiles and operational reviews. Eah orpus analysis started with a manual identifiation of generi items observed in the written texts and in the photographs of the various annual report setions. The items were labeled and their frequenies of ourrene were registered. For eah Duth-English and British text type, frequeny ounts depended on the presene or absene of an item within an instane of the text type, irrespetive of how many times a ontent item, strutural item or lexio-grammatial item appeared in that partiular instane. For example, the data registration showed the number of Duth-English CEO s statements that ontained a given text theme, but not the number of times this theme ourred within one of the Duth-English CEO s statements analyzed. In other words, the orpus analyses were based on doument frequeny ounts, expressing how many texts belonging to a partiular Duth-English or British annual report setion ontained the item in question. Suh frequeny ounts were relevant in the urrent analysis of general genre features, beause it needed to explore the ontent, strutural and lexiogrammatial items that were typially (non-)existent in a ertain text type (e.g. Lee & Myaeng, 2002). The results of the manual data registration were inluded in omputerized (SPSS) databases, whih were reated for the quantitative omparison of frequenies of ourrene in the Duth-English and British annual reports. Per annual report setion, separate SPSS files were realized for genre features pereived in the running written texts, in the doumentary photographs and in the text fragments adjoining these photographs. Eah file ontained a ountry/ompany(row)-by-genre item(olumn) matrix, e.g. the file for genre features in the written texts of the English management statements inluded ountry odes/ompany names in the rows

63 and the full list of themes diserned within the researh materials in the olumns. Following genre theory (e.g. Swales, 1990; Bhatia, 1993), the ross-ultural omparison foused first of all on the text items that were most typial or persistent in the annual report setions. More speifially, it initially observed the ontent items, strutural items and lexio-grammatial items that ourred in 50% or more of the total sample of texts for a setion type in one business ulture or in both ultures. The researh findings indiated that a fous on majority ounts was espeially useful for the genre items in the running written texts, as opposed to the photographs and the written exerpts adjaent to the photos. Sine genre items were generally sparser in the limited number of photographs and in the brief exerpts, often these items did not our in 50% or more of the text instanes for a partiular annual report setion. Even so, the omparative analysis of photographs and exerpts yielded a number of substantial ross-ultural variations, either in relation to an item ourring in a majority or in a minority of the multimodal annual report setions. Therefore, the presentations of results in the following hapters will desribe the 50% ounts for the running written texts in the English annual reports and all frequeny ounts for the doumentary images and adjoining text fragments in these reports. In order to test the differenes between the Duth-English and British orpora per setion of the annual report, Chi-Square tests were onduted (Kilgarriff, 2001). Sine Chi-Square is a statisti that measures whether the samples analyzed are distint enough to allow for signifiant variations aross the populations from whih these samples have been drawn, it was espeially useful for the present study, whih aimed to identify typial ross-ultural distintions in rules for text design. For eah annual report setion, one test aross the Duth-English and British-English text types was arried out to establish the value of the adjusted standardized residual and to determine whether the frequeny of a genre item was lower or higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96). Chi-Square tests have been used in several other (genre-based) studies of disourse aross ultures, e.g. in Shaw et al. (2004) who analyze signifiant international differenes in the aeptane of disourse strategies in ustomer omplaint dialogues and Bjørge (2007) who investigates signifiant international variations in the expression of power distane in student s. Finally, it should be noted that Fisher s Exat tests were applied to supplement the Chi-Square tests; they were arried out for metadisursive elements with low expeted frequeny sores in one or more text types, i.e. where > 20% of ells had an expeted ount less than 5. Fisher s Exat tests only produe a signifiane value and no formal test statisti (Fleiss, 1981; Kilgarriff, 2001). They have also been used as omplementary measures for ross-ultural differenes in other disourse studies, e.g. in Hinkel s ontrastive analysis (1994) of English text interpretations by Asian and Amerian students. 3.5 Interoder reliability tests for ross-ultural analyses of genre features If genres are defined on the basis of the soio-ognitive reognition of onventions in language use within a partiular ommunity (see Chapter 2), their substantive and formal identifiation would seem to depend on subjetive judgments or on the personal realm of psyhologial reality whih is onditioned by the way we think

64 of things in the world (Paltridge, 1994: 296). In priniple, the subjetive potential that is intrinsi to the genre onept ould affet the reliability and empirial validity of generi text analyses, as different individuals may have distintive pereptions of the nature or demaration of generi text features. Nevertheless, not many genre studies have atually verified the reliability of the identifiation and labeling of these features. Referring to the sope of the present researh, some studies have reported on the quantitative results of inter-oder tests to evaluate the reliability of the oding system for moves. Crookes (1986), for example, uses student raters to test the reliability of existing moves from Swales 1981 study of genre struture in introdutions of researh artiles. And Kanoksilapatham (2005) employs a speialist to determine the reliability of both new and existing moves in all setions of biohemistry researh artiles. A few other studies have inluded multiple oders evaluations to determine the reliability of their labeling of metadisursive items. For instane, Hyland (1998a, 1998b) uses lists of predetermined metadisourse items in the interoder reliability tests for rhetorial investigations of aademi researh artiles and business letters to shareholders. Assuming that genre analyses in general gain in solidity and redibility by the exeution of interoder reliability tests, the urrent study omprised interoder tests for the interpretation of ontent features, strutural features and lexio-grammatial features in the written texts and photographs of eah annual report setion. Interoder reliability tests measure the extent to whih two or more oders agree on a partiular rating system and, more speifially, they analyze to what degree similar values an be expeted to emerge when the measurement is arried out on a seond oasion (Rietveld & Van Hout, 1993: 190). In the present study, eah orpus analysis omprised an interoder test performed by two oders. In all ases, the author was the first oder in the test; the seond oder was a (graduate) student in either Business Communiation Studies or Cultural Studies, with a nearnative profiieny in the English language. The seond oders were asked to register textual and photographi genre items on the basis of the same researh steps exeuted by the first oder. Consistent with Neuendorf (2002), the interoder reliability tests omprised 10% of eah sample of available texts per setion of the Duth-English and British annual report. Consequently, the two oders investigated 14 texts in every interoder reliability test: 2 Duth-English CEO s statements, 2 British CEO s statements, 2 British Chairman s statements, 2 Duth-English profiles, 2 British profiles, 2 Duth-English operational reviews and 2 British operational reviews. Table 3.5 shows how many pages were analyzed per interoder reliability test (and per orpus analysis). For most of the levels of analysis, the two oders rated both the written texts and the photographs inluded in these pages. It must be noted that the texts used in eah interoder analysis were also inluded in the full-sample investigations of genre features. The reasons for allowing the admission of these texts was that all genre features defined during the interoder reliability tests were also traed in the full sample of texts (i.e. no new text-related or photo-related features needed to be added to or subtrated from the existing lists of features). Furthermore, the oders involved in the reliability analysis exlusively analyzed the running written texts and photographs in the test texts, i.e. they did not examine the text fragments adjaent to the photographs. Sine the interoder analysis was intended to test the reliability of the interpretation of genre features in the written and visual modes, it seemed plausible to restrit this analysis to the larger written texts and the doumentary images.

65 Table 3.5: Interoder reliability tests total number of pages analyzed in Duth-English and British annual reports. Text feature analyzed Number of pages Content: oneptual themes 96 Struture: moves/tatis 53 Lexio-grammar: metadisourse a Total 249 Note a : Sine a different seond oder was involved in the reliability test for metadisourse in photos, a new sample of texts had to be used. This sample was based on the 10% riterion for the multimodal annual report texts only. As was the ase for the interoder tests in whih written texts and photos were analyzed simultaneously, not all pages inluded visual illustrations. Eah time before onduting the formal interoder reliability test, the two oders partiipated in several training sessions during whih they disussed the interpretation and appliability of preliminary lists of textual and visual themes, strutural elements or metadisursive elements. These preliminary lists had been reated by the first oder in an exploratory researh phase, where readings of prior studies and observations of the given annual report setions had resulted in a first overview of text features. Following the brief introdutory meetings about the purpose and method of the interoder reliability test, the two oders engaged in a number of sessions in whih they onduted small pilot studies to pretest the preliminary oding lists, and in whih they established orrespondene about the required researh steps for the different analyses of text harateristis. The pilot studies usually aused debates about the reognition and demaration of genre features and led to improvements in the oding lists. The annual report texts used for these pilot studies were not inluded in the random sample prepared for the offiial interoder reliability tests. Similar to the regular text analyses and the pilot studies in the interoder analysis, the offiial interoder tests ontained both manual data olletions and statistial analyses. Manual data registrations were inserted in SPSS files, whih failitated the atual omparison of setion-speifi results aross oders in terms of interoder agreement perentages and interoder agreement rates. Interoder agreement rates were established on the basis of mean Kappa alulations. The qualifiation of test results for Kappa was based on Rietveld en Van Hout (1993). The final results of eah interoder reliability test are presented separately in setions 5.2.4, and The methodology introdued in this hapter will be further explained in Chapters 4-7, where it will be speified aording to the type of (on)text analysis involved. Chapter 4 will first of all proeed with a disussion of the ontext analysis, whih was onduted to establish the ommuniative situation in whih the Duth-English and British-English annual reports are ommonly produed.

66

67 Chapter 4 Charaterization of the genre ontext of annual reports in the Netherlands and in the United Kingdom Having outlined the theoretial and methodologial framework for the ross-ultural analyses (in Chapters 2 and 3 respetively), this hapter omprises the first researh phase in the identifiation of the Duth-English and British annual report texts. In line with the text-in-ontext approah to disourse (Askehave & Swales, 2001), this first researh phase fouses on the analysis of the rhetorial situation in whih these texts are used. It relates to the first main researh question and strives to answer the speifi RQs 1a to 1d, whih refer to the desription and ross-ultural omparison of the ommuniative ontext of annual reports by Duth and British multinationals (see setion 1.5). In more detail, the following subsetions will outline and ontrast ultural preferenes with regard to the English language poliy, ommuniative aims, target groups and prodution proess of the full annual report genre. The analysis of these ontextual aspets was grounded on the generi assumption that the soial ommunity that owns the annual report sets the requirements for the multimodal design of its ontent, struture and linguisti style (e.g. Swales, 1990; Jameson, 2000). The perspetive taken in the present ontext analysis is writer-oriented: only Duth and British ommunity members who had been on the prodution side of the 2003 annual reports were involved in the investigation. The reason for this one-sided approah is that the produers of the Duth-English and British-English annual reports atually determined the onstrut of these texts and were thus responsible for the multimodal text features observed in the urrent researh materials (these will be disussed in Chapters 5-7). Therefore, to be able to understand the reasons for the ourrene of these text features, an insight needed to be gained into the underlying priniples followed by the text produers. As was indiated in setion 2.2, the soial grouping to whih the produers of business texts belong, is generally referred to as the disourse ommunity. Drawing on Vergaro (2005), the present ross-ultural investigation of genre ontext foused on the situational preferenes of two different ultural groups that are part of the same expert disourse ommunity, i.e. the annual report produers in Duth orporations and the annual report produers in British orporations. Sine this investigation aims to map the disourse onventions applied in eah individual business ulture, it does not aount for any interultural relations. It should be noted that ontext-based preferenes typial to the reeivers of the annual report will be dealt with in the interultural reader response analysis presented in Chapter 9 of this dissertation. Setion 4.1 provides a more speifi working definition for genre ontext as it is oneived of in the present study. Setion 4.2 omprises the ethnographi frameworks used to ondut the ontext analysis: it further defines the ontextual aspets under investigation and elaborates on the exat tehniques applied in the qualitative survey and in-depth interviews among Duth and British informants. Next, setions 4.3 and 4.4 disuss the omparative researh results for the survey and interviews respetively. These setions will be sueeded by an overview of main onlusions and a disussion in setion 4.5.

68 4.1 A working definition of genre ontext: disourse ommunity Following Miller (1994), it an be assumed that rhetorial ontext is the fundamental prerequisite for establishing a generi produt; as suh, rhetorial ontext is essential to the desriptive analysis of a partiular text type (e.g. Connor, 2004). This subsetion will provide a further operationalization of the onept of genre ontext and will develop a working definition that is relevant for the present study of annual reports (see Table 4.1). Aording to the LSP genre theory explained in setion 2.3, the reurrent reognition of a ertain disourse situation not only eliits a typial text design to fulfill a rhetorial need, but it also helps to organize soial behavior in the disourse ommunity and to sustain this ommunity by expressing its values and beliefs system (e.g. Berkenkotter & Hukin, 1995; Miller, 1994; Yates & Orlikowski, 1992). Given the importane of genre ontext as a basis for typified soial ations, several genre analysts have tried to define the onept of professional ommunity in terms of standard or ommon types of elements, i.e. they have suggested what types of elements generally onstitute the ommunity. Bhatia (1993) first of all notes that ontextual analyses in the identifiation of unfamiliar genres need to raise awareness of why ommunity members write the way they do. Aording to Bhatia (1993), situational explorations potentially address an array of ommunal aspets, e.g. the soial relations between the produers and the audiene of the text, the soiohistorial positioning of the ommunity, linguisti traditions and textual networks and the soial funtion of the text. Askehave and Swales (2001) argue that the soial or ommuniative text funtion in partiular is a deisive fator in genre identifiation, but they also emphasize that it is framed by its soial environment. Consequently, they believe that ontextual genre inquiry also involves the values and goals of the disourse ommunity in general, the material onditions of this ommunity, its work rhythms and horizons of expetations and its textual networks and etiquettes (Askehave & Swales, 2001). To failitate a systemati analysis of genre ontext, Yates and Orlikowski (2002) offer a more ategorized overview of these and other situational onstituents. Genres and genre systems, they propose, are onditioned by six ommuniative dimensions: (1) why, i.e. the soially reognized purpose of a text, (2) what, i.e. expetations about the substantive text design, (3) who, i.e. the soial ties between the initiator and addressee of the text, (4) how, i.e. expetations about the medium and formal design of the text, (5) when, i.e. temporal expetations or deadlines for the text and (6) where, i.e. expetations about the loation of the text (Yates & Orlikowski, 2002). Finally, in her attempt to desribe the soial situation of a text, Beaufort (1997) expliitly relates the ontext of the written genre to the professional disourse ommunity. She speifies that the oneptual hallenge in developing a working definition of disourse ommunity is to isolate those partiular aspets of a soial group that bear upon the ats of writing (Beaufort, 1997: 489). Hene, she suggests that genre ontext is embedded within the disourse ommunity. Drawing on anthropologial, rhetorial and soiolinguisti resoures, Beaufort (1997) puts forward six ommunal elements that identify the interplay between soial ontext and generi patterns; hene, these elements onstitute a working definition of genre ontext that is relevant for the urrent investigation of business texts. As is shown in Table 4.1, these elements onern either the ommuniative ativities or the soial harater of the disourse ommunity. The table also indiates that the ontextual elements referred to in Askehave and Swales (2001), Bhatia (1993) and Yates and Orlikowski

69 (2002) an be inorporated in Beaufort s oneptualizations. For instane, elements suh as text purpose, norms for genre design, writer roles and material onditions are also inherent in the overviews of Askehave and Swales, Bhatia, and Yates and Orlikowski. Table 4.1: Working definition of genre ontext, based on Beaufort (1997: 489, 490). Genre ontext ommuniative ativities in disourse ommunity 1. Modes for ommuniation, networks of related texts, soially reognized purposes of texts. 2. Norms for text or genre features (ommunity-speifi or shared aross ommunities), linguisti traditions or etiquette, expetations about the substantive and formal design of texts. 3. Roles for writers, situation-speifi writing tasks, soial relationships between the writer/initiator and reader/reeiver of text. Genre ontext soial properties of disourse ommunity 1. Community-speifi values and goals that influene text prodution, soio-historial or oupational plaement of the ommunity, work habits, prospets. 2. Material onditions, available ommuniation tools, media, temporal and spatial expetations for texts. 3. Input by the individual soial ator (histories, goals, skills). Features of genre ontext suh as the ones displayed in Table 4.1 have formed the basis of a number of studies on business texts. In their investigation of English genres in the Brazilian business ommunity, for instane, Barbara, Celani, Collins and Sott (1996) look at organizational profile, text purpose, language poliy, frequeny of use and medium to establish whih text types are most ommonly used in English and by what kind of ompany. The findings of their study show that the status of English in Brazilian ompanies was inreasing at the time of researh, partiularly for proposals, reports and projets used in large industrial organizations (Barbara et al., 1996). An aount of genre ontext within a Finnish professional setting is provided in Louhiala-Salminen s examination (1997) of the English business fax. Before analyzing the form of this genre used by a Finnish export ompany, she skethes the soial situation in whih the faxes our in terms of orporate profile, language of the texts, (inter)national ommuniation partners, layout of the message and ommuniative purpose of the faxes. Louhiala-Salminen onludes that the realization of business texts (i.e. genres) is a dynami proess, where the situation with all of its elements (the interatants and their roles, medium, power relations, language used, native/non-native status, to name a few) defines the norms to be followed in the disourse of these texts (1997: 331). Nikerson (2000), moreover, defines the ontext of internal business genres in terms of orporate ulture and orporate ativity. In her study, orporate ulture omprised aspets suh as autonomy, frequeny and type of information supply or languages, while orporate ativity inluded operational ativities, ode of disourse and the internal position of departments. Nikerson s (2000) results indiate that language and disourse patterns in written texts are strongly related to both elements of the organisational ontext. With regard to the rhetorial ontext of a multimodal business genre, Rowley-Jolivet and Carter-Thomas (2005) distinguish several ontextual features that govern the spoken and visual text in onferene presentations. Their study

70 fouses on the position of the onferene presentation (CP) within the network of related texts, the ommuniative funtion of the onferene presentation, the media applied in it, its speaker-listener relations and the time limits to whih it is bound. The results of Rowley-Jolivet and Carter-Thomas investigation reveal that the means of persuasion ( ) are losely linked to the ontextual onstraints and purposes of the CP genre. Speakers are aware of the diffiulty for their audiene of proessing in real time a lengthy spoken monologue with a heavy information load, and adapt both their argument and the alloation of roles aordingly (2005: 64). A final relevant example of the identifiation of genre ontext is derived from Zhu (2005), who examines and ontrasts multiple national ultures to be able to explain the differenes between English business texts from Australia/New Zealand (NZ) on the one hand and Chinese business texts on the other hand. She observes how historial developments in soial values, market eonomy and tehnology have affeted the design of sales letters, sales invitations and faxes in eah ulture. In her onlusion, Zhu notes that ontext-based differenes refleted in the text are partiularly likely to influene the interultural effetiveness of that text: NZ and Chinese managers seemed to follow different riteria for effetive persuasion. What was onsidered appropriate by Chinese managers might not be onsidered as suh by the NZ managers (2005: 183). The following subsetion will further elaborate on the ontextual identifiation of a different business genre originating in various national business ultures, i.e. the Duth-based and UK-based annual report. 4.2 Contextualizing annual reports in the Netherlands and the UK Drawing on the idea that a need still exists for the deeper analyses of proesses, ontexts and purposes of disourse (Connor, 2004: 292), the remainder of this hapter will be dediated to expanding the understanding of genre ontext with respet to the (English) annual reports produed in the Netherlands and in the UK. Several previous studies have speulated about the interdependene between soial ontext and annual report dislosures by Duth and British orporations, e.g. Santema, Hoekert, Van de Rijt and Van Oijen (2005) who relate Duth orporate governane preferenes and ultural values to levels of strategy dislosure in Duth annual reports, or Stittle (2002) who relates the regulatory framework established by UK finanial authorities to levels of ethial reporting in British annual reports. However, little researh has reently been onduted to speify this ontext on the basis of empirial, ethnographi information. As this study suggests, in order to exeute a full genre analysis of Duth-English and British annual reports, an aurate oneptualization of their ommuniative setting was needed. Based on the working definition of genre ontext and based on the fat that a native English and a non-native English ulture were involved, nine ontextual features of Duth(- English) and British annual reports were studied by means of a qualitative survey and in-depth interviews with Duth and British informants. Eah feature an be related to one or more of the elements in Beaufort (1997, see Table 4.1): a) Duth and English language poliies in annual reports of Duth orporations linguisti traditions; b) general ommuniative purpose(s) of the texts in the full annual report soially reognized purpose;

71 ) speifi ommuniative purpose(s) per textual setion of the annual report soially reognized purpose; d) general ommuniative purpose(s) of the photographs in the full annual report soially reognized purpose; e) general reader group(s) of the full annual report writer-reader relationship; f) speifi reader group(s) per setion of the annual report writer-reader relationship; g) guidelines for the multimodal design of the annual report norms for genre design/ ommunity-speifi values; h) writing proess underlying the realization of the full annual report, inluding identifiation of authors involved norms for genre design/ writing tasks/ material onditions/ temporal expetations; i) prodution proess for the multimodal annual report norms for genre design/ writing tasks/ material onditions/ temporal expetations. Sine the present study revolved around general disourse onventions within ultural groups (that were part of the same international disourse ommunity), it seemed plausible to fous first of all on their olletive preferenes. Informants in Duth ompanies partiipated in the qualitative survey as well as in the in-depth interviews, whereas the British informants were only engaged in the indepth interviews. This is beause the qualitative survey aimed to map the orporate language poliies speifially related to English as a non-native international language for the 2003 annual reports of Duth ompanies; the in-depth interviews were intended to reveal the ommuniative purposes, reader groups and realization proess of the Duth(-English) and British annual reports. During the interviews, attention was paid to the importane of both the texts and the photographs inluded in these reports. Partiipants in the qualitative survey and in the interviews were employed at large and mid-size orporations quoted on the Amsterdam stok exhange and on the London stok exhange. These orporations were onsidered to be rih and relevant soures of information, as they are legally obliged to publish annual reports and they depend on these texts as important ommuniation tools for international stakeholders. While organizing the qualitative survey and the in-depth interviews, it beame apparent that it would be impossible to ondut the survey and the interviews with individual informants who were fully aountable for the annual report of a partiular ompany. Usually, a team of staff members was responsible for the reation of this report. Hene, both the qualitative survey and interviews took plae with informants who had been losely involved in oordinating the realization proess of the annual report or who had been atively involved in writing up the 2003 report. All the people partiipating in either the survey or the interviews were offiers working in the ommuniations or investor relations department Design of the qualitative survey in Duth orporations In order to find an answer to RQ 1a, i.e. mapping the English language poliy for annual reports originating in the Duth business ulture, Duth orporations were invited to partiipate in a qualitative survey. At the time of researh, the Amsterdam stok exhange omprised a total of 49 large (AEX) and mid-size (Midap) orporations in diverse industries. With regard to the analysis of English and Duth language poliies in annual reports produed in the Netherlands, all orporations on

72 the Amsterdam stok exhange were inluded in the qualitative survey. Information about language poliies by all registered Duth ompanies was ruial, as an exploratory pre-analysis during the first year of the projet (i.e. 2004) had suggested that Duth ompanies had taken various approahes to writing and translating the Duth and English versions of their 2003 annual reports. The qualitative survey findings were intended to larify the language harater of the Duth-English materials. And as suh, they were used to deide whether Duth-English translations or original English versions of the annual reports should be inluded in the text analyses (desribed in Chapters 5-7). A qualitative survey fouses on theorization by means of omparing a relatively large number of ases and aims at re-onstruing meanings, assumptions and behavioral patterns (Wester & Peters, 2004: 40). Here, the survey was an effiient method for olleting information that would enhane existing knowledge of the English language poliies generally adopted for annual reports by Duth organizations. The 49 Duth ompanies were asked three questions through the e- mail, onerning (1) the language of the original version of the annual report, (2) the first language of the writers and translators of the (English) annual report and (3) any expeted hanges in the language poliy adopted for the report. It must be noted that these questions were not posed in a qualitative survey of all 49 ases. A total of 11 ompanies quoted on the Amsterdam stok exhange were exluded from the survey, beause they had already agreed to partiipate in the in-depth interviews. To avoid having to ontat the same ompanies twie, it was deided that the three questions initially planned for the survey, would be inorporated in the in-depth interviews in Duth ompanies. Consequently, 38 large or mid-size orporations in the Netherlands reeived the qualitative survey by e- mail. After five months and follow-up requests for partiipation in the survey, 37 ompanies had answered the survey questions either by or telephone. Adding the 11 interview responses to the 37 survey responses resulted in a response rate of 98%, i.e. 48 out of 49 ompanies. The results of the qualitative survey are desribed in setion Design of the in-depth interviews in Duth and British orporations The answers to RQs 1b to 1d, i.e. desribing and omparing the ontextual harateristis of annual reports in the Netherlands and in the United Kingdom, were obtained by means of a series of in-depth interviews with Duth and British informants. Determining the role played by the annual report in the Netherlands and in the UK required a large number of questions that would fous on the pereived role, objetives, audiene and development of the annual report in general. Sine the detailed responses to these questions ould not be inorporated in a survey, the funtion of the annual report was mapped by semi-standardized open interviews during whih all topis of interest ould be investigated in detail (e.g. Flik, 1998). In semi-standardized interviews theory-driven [ ] questions are asked for the purpose of making the interviewee s impliit knowledge more expliit (Flik, 1998: 84). In this ase, a topi list was used to organize and ondut the semistandardized interviews. The first interview was used to pilot this list; a revised version of the list was applied in the suessive interviews. While a total of 49 large and mid-size ompanies in diverse industries were quoted on the Amsterdam stok exhange, a muh larger number of equivalent

73 ompanies were quoted on the London stok exhange. Beause in-depth interviews with informants of all registered Duth and British orporations would not have been feasible and would probably have yielded redundant information in the end, a sample was drawn from the total population of annual reports. To form a relatively representative sample of Duth and British interview partiipants from different types of large and mid-size orporations, a maximum variation sampling strategy was applied (e.g. Lindlof, 1995; Flik, 1998). Maximum variation sampling is a purposeful sampling proedure, whih enloses a range of differentiating qualities aross ases. Therefore, it uts aross variation in the sample and enables the researher to explain a situation through ore harateristis that are shared among members in this sample. As suh, the sample seleted here allowed for the study of general features of the soial ontext of annual reports, thus avoiding the influene of any possible industry-speifi or size-related features. In the present study, ompanies in the Netherlands providing informants for the in-depth interviews differed in size and were operating in eight industries: diversified finane, hemials, media, business servies and supplies, food, onglomerates, IT/software and servies and onstrution and infrastruture (see Table 4.2). Although the maximum variation sampling proedures were also used in seleting British interview partiipants, here this proedure was preeded by onveniene sampling, whih relies on the seletion of readily available units (Neuendorf, 2002: 87). Instead of drawing a variation sample from the total population of British ompanies that had sent their annual reports, only the Londonbased ompanies were admitted to the researh population. Sine it would be more pratial to ondut several in-depth interviews in a partiular period of time, it was deided that all British informants be seleted from orporations situated in one region in the UK. As onveniene sampling is primarily expediene-based, it is generally onsidered to be the least aeptable sampling method (Lindlof, 1995; Neuendorf, 2002). In the present study, however, the methodologial defiit of the onveniene sampling method was moderated by the sueeding variation sampling proedures. British ompanies whose staff members partiipated in the interviews also differed in size and were operating in five industries: media, teleom, diversified finane, onglomerates and insurane (see Table 4.2). The diversity in orporate aspets overed by this sample allowed for a broad view on genre ontext that surpassed the levels of size and industry. Although the Amsterdam and London stok exhanges enompassed ompanies in many additional industries, the findings showed that a point of informative saturation had been reahed after onduting eleven interviews in the Netherlands and five interviews in the UK. At that point, the analysis of information provided started to display reurrent patterns in question-answering, in the Netherlands as well as in the United Kingdom. Table 4.2 offers an overview of the ompanies that engaged in the in-depth interviews (under the strit requirement of anonymity).

74 Table 4.2: Duth and British orporations partiipating in the in-depth interviews. Nation Industry Size Frequeny Netherlands Diversified finane Large 2 a Business servies and supplies Mid-size 2 a Conglomerates Large 1 IT/Software and Servies Mid-size 1 Chemials Large 1 Media Large 1 Food Large 1 Constrution and Infrastruture Mid-size 2 a United Kingdom Media Large 1 Teleom Large 1 Diversified finane Large 1 Conglomerates Mid-size 1 Insurane Large 1 Note a. Companies lassified in the same industrial ategory were marked by different types of orporate ativities: (1) the large Duth ompanies in the diversified finanial industry varied in emphasis on banking and insurane, (2) the mid-size Duth ompanies in the business servies industry onerned employment servies and doument systems, and (3) the mid-size Duth ompanies in the onstrution and infrastruture business varied in terms of maritime or land tehniques. The analysis of the interview data was onduted aording to a themati oding proedure. In themati oding, the researh issue is the soial distribution of perspetives on a phenomenon or a proess (Flik, 1998: 187). The proedure begins with a first interpretation of ases, or individual responses to interview questions. After finalizing the first interpretation phase, the themati oding proedure requires a deepening analysis of ases, resulting in a more aurate system of response ategories per question. For eah informant, the response ategories per question are ross-heked aross all interviews; this should eventually lead to a themati struture whih underlies the analysis of ases, in order to inrease their omparability (Flik, 1998: 188). For the present study, the interview data were initially ategorized on the basis of the topi list used in the interviews. During subsequent oding phases this lassifiation system was adapted and redued aross interviews, generating seven main ategories that ould be ompared aross informants: 1) general ommuniative purposes of the annual report; 2) reader audiene of the annual report; 3) guidelines for the ontents of the annual report; 4) ommuniative objetives per setion in the annual report; 5) English language-poliy for the (statutory) annual report; 6) guidelines for and purposes of photographs in the annual report; 7) prodution proess of the annual report. Setion 4.4 presents a omparative overview of the data obtained in the interviews with Duth and British orporations.

75 4.3 Results English language poliy for Duth-based annual report The answers given to the first question in the survey are dislosed in Table 4.3: it presents details of the original language for the 2003 annual reports of the 48 large and mid-size Duth orporations that partiipated in either the survey or the interviews. As an be seen in Table 4.3, over 58% (i.e. 28) of the Duth orporations laimed that they had used English as the base language for the 2003 annual report. Only 39.6% (i.e. 19) of these orporations had still applied their first language for the realization of the original version of the report. Moreover, it was observed that the large ompanies in partiular had adopted English as the original language for the finanial report. Whereas 19 of the 23 large ompanies (82.6%) indiated that they had used English as the language for the original draft of the report, only 9 of the 25 mid-size ompanies (36.0%) had used English as the base language. One respondent from a large Duth organization explained that the Duth and English versions of the annual report are realized simultaneously. Table 4.3: Base language for annual reports by Duth ompanies of different size. Company size Large ompany Mid-size ompany Total % Original language annual report Duth English Duth+English Total Subsequently, the survey foused on the native languages (L1) of the writers and translators of the annual reports. The findings on the L1 of writers of the original drafts of the reports are displayed in Table 4.4. Respondents stated that the original Duth version of their annual reports had been written by native speakers of the Duth language. With respet to the original English versions published by orporations in the Netherlands (58.3%, see Table 4.3), however, it was found that half of these versions had been written by native speakers of English (14, i.e. 50%). Furthermore, 13 out of 28 original English reports (i.e. 46.4%) were laimed to have been written by a ombination of native English speakers and native Duth speakers, whih implies that in several ompanies the development of the annual report involved a ollaborative writing proess in whih a team of staff members was responsible for drafting the English report. In only one survey ase, the respondent replied that the original English version of the annual report had been written by a native Duth speaker with an aademi degree in English. Table 4.4: L1 of writers for the original versions of annual reports by Duth ompanies. Original language annual report Duth English Duth + English Total % L1 writer original annual report Duth English Duth+English Total

76 The answers to the survey question about the translators first language (L1) are provided in Table 4.5. This question aused five missing values, sine five respondents did not know whether native Duth speakers or native English speakers or both, had been responsible for the English translations of their original Duth annual report. They explained that ompanies in the Netherlands often hire external translation agenies without getting to know the atual translators or their ultural bakgrounds. Where the original language of the 2003 annual report was given as Duth, an English translation had been realized and vie versa. Translations from Duth into English had predominantly been exeuted by native speakers of the English language. i.e. in 15 out of 16 orporations (93.8%). Again, respondents pointed out that either British or Amerian writers had been employed to reah a native level of written English. In one orporation, a native Duth speaker with an aademi English degree had been assigned the task of realizing a Duth-English translation. While the majority of the Duth translations of English originals were laimed to have been realized by native speakers of Duth (19 out of 26, or 73.1%), a small number of ompanies had employed a team of native Duth and English speakers to reate these Duth translations. A team of translators had also been involved in the simultaneous establishment of Duth and English versions of the annual report. As Table 4.5 indiates, moreover, there were five of the 43 partiipating Duth ompanies (11.6%) that did not translate their original English annual report into a Duth report in 2003; for these five ompanies, the L1 of the translator was thus irrelevant. Table 4.5: L1 of translators for the translated versions of annual reports by Duth ompanies. Original language annual report Duth English Duth + English Total % L1 translator annual report Duth English Duth+ English irrelevant Total The final question in the survey onerned future perspetives on the language poliies for annual reports. In Table 4.6, expetations about any short-term alterations in the urrent language poliy are related to the language used for the original version of the report. Exept for four respondents, none of the interviewees antiipated any offiial hanges in the language poliies that had been used for the 2003 annual reports. Two ompanies that had been writing original Duth versions and English translations up to 2003, did report plans to hange their language poliies: they expressed their desire to introdue English as the original language for the 2004 annual report, provided that their proposal was supported by shareholders at the next annual general meeting. Another responding orporation that had published English originals and Duth translations of its annual report in 2003, laimed it would stop produing any further Duth versions of the report from 2004 onwards. And finally, one ompany that had been exeuting an English-only poliy throughout the past years was planning on re-introduing a Duth translation of its annual report in 2004, i.e. to meet the needs of Duth stakeholders with insuffiient English profiieny.

77 Table 4.6: Speulations about hanges in language poliy, related to the base language of the reports. Original language annual report Duth English Duth + English Total % Change language poliy No Yes Total While reent language trends had already shown the growing soial and professional importane of English in the Netherlands (e.g. Graddol, 2000; Gerritsen & Nikerson, 2004), the qualitative survey results desribed in this subsetion revealed that these trends have had a major impat on Duth orporations annual reports in partiular. The results suggested that the professional status of English within these reports ompetes with or even exeeds the Duth language status. 4.4 Results ommuniative ontext Duth-based and UK-based annual reports A total of 16 Duth and British orporations partiipated in in-depth interviews to explain the ommuniative ontext in whih the annual report was generally produed. In line with genre theory (e.g. Bhatia, 1993), this ontext was assumed to have had an effet on the substantive and formal design of the annual report (e.g. see Davison, 2002). The following subsetions doument and ontrast the information obtained during the interviews in the two different national business ultures. They suessively deal with the Duth and British responses to the seven primary topis aquired through the themati oding proedure: general ommuniative purposes of the annual report (setion 4.4.1), reader audiene of the annual report (setion 4.4.2), guidelines for the ontents of the annual report (setion 4.4.3), ommuniative objetives per setion in the annual report (setion 4.4.4), English language-poliy for the annual report (setion 4.4.5), guidelines for and purposes of photographs in the annual report (setion 4.4.6), and prodution proess of the annual report (setion 4.4.7) General ommuniative purposes of the annual report In general, Duth and British interviewees regarded the annual report as a rather stati doument, ontaining a defined orporate profile and a historial reord of results. Aording to the informants, the annual report in itself annot have real news value and is merely supportive of other more dynami ommuniation instruments that publish updated information about the ompany s identity and performane. They noted that the annual report as suh does not have an absolute rhetorial funtion but that it is part of a network of orporate texts that are funtionally related. Both Duth and British partiipants tended to assoiate the annual report with a variety of finanial ommuniation tools in partiular; they mentioned its ties with preliminary announements (finanial douments that appear before the annual report and ontain updated finanial details, e.g. per quarter), road

78 shows (stakeholder meetings with CEOs or Investor Relations managers that are based on the ontent of the preliminary announements and annual report), press releases, web pages on investor relations issues, interim reports and annual general meetings. The annual report was believed to be a useful means for rereading or verifying what is ommonly ommuniated through other orporate hannels. Furthermore, it was onsidered to be a relevant onstituent of some of these other hannels (e.g. the report as a preparatory or referene work for the annual general meeting). While the Duth partiipants in the interviews onfined themselves to the definition of the annual report as a finanial ommuniation instrument, the British partiipants indiated that it may also fulfil a relevant funtion within the mix of marketing ommuniation tools. British informants argued that the annual report an be regarded as a omprehensive orporate brohure, sine it offers an elaborate information pakage about the identity, strategy, produts or servies and market position of the ompany. They added that the annual report may therefore prove to be an interesting doument for stakeholders who do not have a diret performanebased interest in the ompany but who wish to know more about its organization, divisions and produts. When the Duth and British informants were asked to larify the objetives of the full annual report, several agreed that the priniple objetive of the annual report is to fulfil the legal obligation of publishing an annual report, or to meet the legal requirements set for orporate finanial reporting. Both Duth and British informants put forward the following two ommuniation purposes: a.) First, the annual report is intended to offer transparent, informative desriptions of the orporate profile and performane. It needs to ommuniate fatual and truthful details about the orporate identity, strategi priorities, management issues, operational and finanial results and orporate prospets. Furthermore, it offers insight in the organizational ondut, in terms of orporate soial responsibility, orporate governane, human resoure management, et. b.) Seond, the annual report is intended to affet reader attitudes; it needs to establish a positive orporate image by emphasizing managerial effiieny and orporate reliability. The annual report aims to eliit general goodwill and investment interests in the ompany. Compared to the regulated annual aounts, the annual review is more suitable for this ommuniative objetive, as it enables the ompany to inlude simplified, visualized messages in whih they are free to emphasize orporate suesses and fous on a onsistent orporate marketing style. The Duth and British informants agreed that the primary objetive of the annual general report is dislosing information, partly as a onsequene of the legal requirements that have formalized the report. Several interviewees added that the extent to whih readers of the annual report are informed or persuaded may differ, depending on the text part involved (i.e. finanial aounts or annual review). The tehnial and detailed annual aounts are highly informative and therefore ruial to finanial readers, who are mainly interested in objetive finanial data. The onise and illustrated annual review, however, was onsidered to be less formal and more strategi; it was expeted to be partiularly relevant for positively influening the attitudes of non-finanial stakeholders. As suh, it was regarded by the British interviewees as a useful ontribution to the marketing ommuniation mix.

79 Informants in the Netherlands and in the UK also believed that the annual report does not serve the purpose of ativating stakeholder behaviour. Although the doument was referred to as a relevant tool for informing readers and influening reader attitudes, informants laimed that it is too stati and tehnial to be able to atually influene or urge (investment) behaviour. Rather, the annual report should support attempts to ativate behaviour, preferably in ombination with fae-to-fae ontats between orporate staff and orporate stakeholders. Interviewees suggested that analyst meetings and shareholder presentations in partiular are more effetive for the enouragement of desired stakeholder ations Reader audiene of the annual report The annual report serves multiple target groups in the Duth and British business ultures. As was already implied in their desriptions of the ommuniative purposes of this report, informants distinguished two general types of readers: finanial readers and non-finanial readers. Apart from this rough haraterization of target groups for the annual report, Duth and British informants further speified its reader audiene in several subgroups, eah of whih has a unique interest in the ompany. A ombination of Duth and British interview responses yielded seven reurrent subgroups: - (prospetive) institutional and private investors; - finanial analysts and stok brokers; - (finanial) media; - (prospetive) lients and suppliers; - soial interest groups/non-governmental organisations (NGOs); - (regulatory) authorities; - (prospetive) employees and managers. Informants in the Netherlands and in the UK aknowledged that these different reader groups do not have idential information needs and added that ompanies usually employ different ommuniation strategies to make the annual report relevant and appealing for them. More speifially, Duth and British ompanies tended to take notie of themes, amount of information provided, voabulary, tone of voie and text struture. Duth and British informants offered several overlapping examples of the reader-speifi rhetori inorporated in the annual report. For instane, setions on share information inluding details on share apital, share listings, shareholder analyses, dividend or share prie are inluded to serve (potential) investors. And sustainability themes are disussed to satisfy NGOs. Also, finanial professionals are generally addressed in tehnial jargon, explaining fatual data on results, finanial figures and prospets on future performane. Some Duth orporations inluded glossaries in their annual reports to help non-finanial readers understand the finanial data in the written texts. Although informants were aware that suh rhetorial strategies annot make the annual report of equal interest to all readers, e.g. politial and soial stakeholders still need to derive extra details about business ondut from other orporate soures like business priniples or sustainability douments, they assumed that these strategies do ontribute to the overall reader-friendliness of the text. After speifying the various reader groups involved in the annual reports, interviewees were asked to plae these groups in order of priority. A few informants advoated the idea that no reader group be prioritized over others. Most partiipating orporations, however, did have a lear opinion about the relative importane of the

80 speifi reader audienes. In eah national business ulture, informants generally established a top three of the most relevant target groups. But as is shown in Table 4.7, there were ross-ultural differenes in the struture of the top three rankings. Whereas the British informants were very deisive about the relevane of the shareholders as the primary finanial audiene of the annual report, Duth informants evaluated these readers as equally important as the other finanial target groups (i.e. analysts, brokers and journalists). Compared to British orporations, moreover, Duth orporations asribed more value to readers who work within their own organization; internal staff belong to the broad audiene of the British annual reports, but are addressed as a separate important reader group in Duth-based annual reports. Table 4.7: Top-three ranking reader groups for the annual report. Audiene annual report in the Netherlands 1. Finanial groups: (prospetive) institutional and private investors, stokbrokers, finanial analysts, (finanial) media; 2. Internal groups: (prospetive) employees and managers; 3. Broad audiene: authorities, (prospetive) lients, suppliers, soial interest groups/ngos. Audiene annual report in the UK 1. Shareholders: (a) institutional investors, (b) private investors; 2. Other finanial groups: finanial analysts, stokbrokers, (finanial) media; 3. Broad audiene: authorities, (prospetive) lients, suppliers, (prospetive) employees and managers, soial interest groups/ngos. Duth informants explained that finanial and internal reader groups are given high priority, as they influene the finanial and operational effiieny of the ompany. To establish goodwill in these groups, the Duth annual report needs to provide them with details on orporate results and orporate prospets. Conversely, British interviewees onsistently argued that the investor audiene is of partiular importane beause the annual report is one of the few written douments shareholders an refer to for a full aount of the orporate results. Although British informants did onsider the (prospetive) employees to be fairly relevant for the annual general report, the fat that relatively few employees regularly request the report, has led the informants to believe that the internal workfore does not form a ruial audiene. Both in the Netherlands and in the UK, the remaining readers in the ategory of broad audiene were deemed as being least influential with regard to finanial and operational performane and were given low reader priority for the annual report Guidelines for the ontents of the annual report A pre-study of Duth-English and British annual reports overing 2003 results revealed that there was a general variability in the setions and topis inluded in these texts. Apparently, some items in these annual reports were ompulsory (e.g. finanial review) and others were not (e.g. profile, management foreword). From the perspetive of written disourse, this distintion between obligatory and optional items may have influened the degree of freedom with whih these items were treated in the annual reports of Duth and British orporations. To determine what might have aused the inlusion or omission of partiular items in the annual report, Duth and British partiipants in the in-depth interviews were asked to explain their

81 guidelines in the write-up of the report. Their responses are shown in Table 4.8. Both Duth and British informants reognized the major influene of legal requirements but also suggested that these legislations still allow for individual ontributions in dislosures, espeially with regard to the visual design of the texts. They indiated that several non-legal riteria were generally referred to in the prodution of the annual report ontents. Despite the fat that the Duth and British responses orresponded for the legal, soio-politial, peer-related and historial guidelines, the table also reveals that there were several ross-ultural distintions in the hoie of additional guidelines. Here, the Duth partiipating ompanies seemed to employ a slightly wider variety of riteria for the design of the annual report than the British partiipants. It must be noted that the riteria mentioned by the informants partiularly onerned the annual reports published up to the 2004 finanial year. After 2004, several (inter)national developments in aountany and soiety have taken plae that might have affeted the riteria but not the table of ontents for the sueeding annual reports, e.g. new standards for the operating and finanial review in the UK, issued by the British Aounting Standards Board in May 2005 (Finanial Reporting Counil, 2005). The various guidelines informants laimed to use for the ontent of the annual report in the Netherlands and in the UK ould be divided into three ategories: obligatory (based on legal standards), best pratie (based on general soio-politial issues) and optional (based on riteria set by the own ompany). This is illustrated in Table 4.8. Sine many legal aountany standards and non-finanial guidelines, suh as ethial issues or historial traditions, have had a prominent impat on the design of the annual report over the past years, it was diffiult for both Duth and British interviewees to make an exat distintion between setions or topis that were legally required and those that were desirable from a soial or traditional viewpoint. Consequently, it was impossible for them to determine the preise amount of authorial freedom for these items within the annual report. Table 4.8: Guidelines for the write-up of items in the annual report. Type of guideline 1. Obligation = law item must omply with legal requirements. 2. Best pratie = soio-politial issues item should inlude issues labeled as ethial and soially responsible. 3. Option = ompetitors annual report item should be in line with reports of peers. Relevant setion Duth annual report Report of the exeutive board (inl. operational and finanial review) *, report of the supervisory board *, independent auditors report, finanial statements. Corporate soial responsibility (CSR) review, remuneration report. Corporate governane report. Relevant setion British annual report Partiular topis in the operating and finanial review (e.g. finanial + operational results, researh and development, human resoure management), diretors report *, remuneration report, finanial statements. Chairman s statement *, CEO s statement *, orporate soial responsibility (CSR) report, ompany profile. Operating and finanial review.

82 Table 4.8 ontinued. Type of guideline 4. Option = historial traditions item should math orporatespeifi onventions in reporting. 5. Option = reader needs item should meet reader needs. 6. Option = orporate poliy item should reflet the philosophy/poliies of the ompany. 7. Option = benhmarking item should strive for high external ranking. 8. Option = management approah to finanial reporting item should meet the management approah. Relevant setion Duth annual report Table of ontents of the annual report. Shareholder information setion (inl. share pries, distribution of shareholders). Design of the divisional review of operations (e.g. by produt group or region). Disussion of topis that ensure high evaluation sores. Relevant setion British annual report Table of ontents of the annual report. Plain design versus expressive design (inl. suessful ase studies, visual illustrations). Note * : UK-based annual reports ontain statements of an exeutive and a non-exeutive representative, as well as a report of the full board of diretors. Duth-based annual reports ontain one statement of an exeutive representative, one report of the board of exeutive diretors and one report of the board of supervisors (see setion 4.4.4) Communiative objetives per setion in the annual report Both in the Netherlands and in the UK, the annual reports appeared to onsist of a reurrent variety of setions with distintive funtions and designs. Consequently, Duth and British informants were asked to desribe several of the (multimodal) texts inluded in the annual report of their ompany, partiularly in terms of ommuniative objetives, (reader) audiene and propositions. All informants agreed that the Duth and British annual reports are generally onstituted of a pathwork of texts that reflet various narrative voies, purposes and target groups and that deal with different topis. Some informants added that the ommuniative style throughout these different annual report setions is usually kept onsistent, so as to suggest a unified type of text. This is illustrated in Table 4.9, whih presents the Duth and British desriptions of a number of popular non-finanial narratives in the Duth-English and British annual reports and of the finanial aounts that are inluded in eah English annual report. Other figure-based or non-narrative texts suh as finanial highlights, five-year summaries and diretors biographies, are not aounted for in the table, sine they fell outside the sope of the present study (whih foused on the narrative texts preeding the finanial aounts). Table 4.9 indiates that Duth and British partiipants sometimes had diverse ideas about the rhetorial value of speifi annual report setions.

83 Table 4.9: Pereived topis, ommuniative aims and reader groups of popular non-finanial texts in Duth-based and UK-based annual reports. Setion report Duth annual report British annual report Corporate profile Topis: identity, mission, values (fundamental views), markets (ore ativities), strategi priorities, orporate strengths. Aims: informing readers with limited knowledge of ompany, onvining readers of unique business harateristis + solid strategi basis for the future, ontextualize results of past year + future. Readers: broad audiene (finanial + non-finanial readers). lients). Chairman s statement CEO s statement Operational review Topis: identity, mission, values (fundamental views), markets (ore ativities), strategi priorities, orporate strengths. Aims: informing readers with limited knowledge of ompany, onvining readers of unique business harateristis + solid strategi basis for the future, ontextualize results of past year + future. Readers: broad audiene (emphasis on shareholders, Topis: personal opinion about (finanial) results, management situation, future outlook. Aims: offering an informative and parental top-line overview of results, ontextualizing information in sueeding setions, providing the ompany with a personal fae, establishing reader-writer relationship. Readers: broad audiene, but fous on shareholders. Topis: personal view on orporate profile, operational/finanial results, strategy, market position, future. Aims: offering an informative and informal summary of results, ontextualizing information in sueeding setions, onvining readers of reliability of the ompany, providing the ompany with a personal fae, establishing reader-writer relationship. Readers: broad audiene. Topis: detailed fats about identities, strategies, operational results, explanations of results, future prospets and ase studies per business unit/ produt group. Aims: offering biased but detailed information about segmental orporate results, onvining readers of suess/ uniqueness/reliability of the ompany. Readers: (potential) shareholders, lients, staff in segments. Topis: strategi overview of operations, suesses, problems, solutions, reasons underlying the results. Aims: offering biased but detailed information about orporate results, ontextualizing information in (segmental) operating review. Readers: shareholders. Topis: detailed fats about identities, strategies, operational results, explanations of results, future prospets and ase studies per business unit/ produt group. Aims: offering biased but detailed information about segmental orporate results, onvining readers of suess/ uniqueness/reliability of the ompany. Readers: private shareholders.

84 Table 4.9 ontinued. Setion report Duth annual report British annual report Finanial review Topis: fats about finanial results in past year(s), inluding profits/losses, debts, ash flow, et. Aims: informing readers about finanial position of the ompany. Readers: shareholders, analysts, brokers, journalists (finanial readers). Topis: fats about finanial results in past year(s), inluding profits/losses, debts, ash flow, et. Aims: informing readers about finanial impliations of results disussed in CEO s statement + operational review. Readers: institutional shareholders, analysts, brokers, Corporate governane/ remuneration report Corporate soial responsibility (CSR)/ human resoure management (HRM) report Topis: orporate struture, orporate management poliies/proedures, payment poliies/shemes for orporate diretors. Aims: informing readers in a transparent manner, onvining readers of reliability of the ompany. Readers: private shareholders, analysts, journalists, authorities. Topis: poliies/ativities supporting the ompany s soial/environmental/ eduational ommitment, poliies/ management of personnel. Aims: informing readers in a transparent manner, onvining readers of reliability and humane harater of the ompany. Readers: authorities (CSR), NGOs (CSR), (potential) employees (HRM), shareholders journalists (finanial readers). Topis: orporate struture, orporate management poliies/proedures, payment poliies/shemes for orporate diretors. Aims: informing readers in a transparent manner, onvining readers of reliability of the ompany. Readers: shareholders, journalists, authorities. Topis: poliies/ativities supporting the ompany s soial/environmental/ eduational ommitment, poliies/ management of personnel. Aims: informing readers in a transparent manner, onvining readers of reliability and humane harater of the ompany. Readers: broad audiene (emphasis on NGOs). (HRM). Diretors report Topis: fats about issues involving the exeutive and nonexeutive diretors. Aims: formally informing readers. Readers: readers interested in management/governane topis. Supervisory board report (nonexeutive diretors) Topis: supervisory board s approval of finanial aounts, roles/proedures exeutive and supervisory diretors, monitoring ativities supervisory board. Aims: informing readers about evaluation exeutive board s ativities. Readers: non-finanial readers interested in internal management praties.

85 Table 4.9 ontinued. Setion report Duth annual report British annual report Finanial aounts Topis: tehnial fats and figures about the ompany s finanial performane. Aims: informing readers about the finanial position + prospets of the ompany. Readers: institutional shareholders, analysts, journalists (finanial readers with intermediary role ompanypubli). Topis: tehnial fats and figures about the ompany s finanial performane. Aims: informing readers about the finanial position + prospets of the ompany. Readers: institutional shareholders, analysts, journalists, authorities. As Table 4.9 indiates, it seems that the most obvious ross-ultural differenes in annual report setions involved the managerial forewords, the reports of diretors and the operational review. First, informants desriptions revealed that Duth annual reports only have one management statement written by the CEO, whereas British annual reports have two management statements, i.e. one introdutory statement signed by the supervising Chairman and one sueeding statement signed by the CEO. Presumably, this differene is the result of the fat that Duth orporations have a ontinental two-tier management system and the British orporations have an Anglo-Saxon one-tier system (e.g. Hooghiemstra & Van Manen, 2004). In a two-tier system, the board of exeutives and the supervisory board (omprising non-exeutives) are autonomous organs that work independently from one another. Hene, a statement written by the management of a Duth orporation needs to be signed by the CEO, who is the exeutive leader of the management board. A one-tier system ontains one management board in whih both exeutive and non-exeutive diretors are inluded. Aordingly, Anglo-Saxon annual reports generally ontain management statements of both representatives of the board: the CEO, who is running the daily ativities of the ompany, and the Chairman, who is responsible for supervising and advising the exeutive board. It seems that the ross-ultural variation in types of management statements has affeted the purposes and topis overed by these texts. As an be seen in Table 4.9, the Duth CEO s statement has a hybrid format ompared to the two British statement types. The Duth text was identified as a personalized message about the orporate identity, results and market position; therefore, Duth informants believed it is interational as well as strategi. British informants, however, onsidered the Chairman s statement to be interational in partiular, while they defined the more elaborate CEO s statement as primarily informative and strategi. Thus, the Duth CEO s statement seems to fulfill rhetorial objetives that are distributed aross the Chairman s and CEO s statements in the British annual report. In addition, there was also a ross-ultural distintion in the reader groups aimed at through the managerial forewords. While the Duth informants explained that the CEO s statement should reah a broad (non-)finanial audiene, British informants foused on the shareholders as the most important readers for both the Chairman s and CEO s statements. These reader preferenes are in line with informants opinions about the general reader audiene of the full annual report: British informants expliitly asribed more importane to the separate shareholder group than Duth informants, who regarded shareholders as one of several relevant reader groups (setion 4.4.2).

86 Moreover, the differene between the two-tier system in the Netherlands and the one-tier system in the UK seemed to have influened the design of the diretors reports in the Duth and British annual reports. Aording to the Duth informants, the Duth annual report usually inludes a report by the exeutive board (to whih the operational and finanial reviews belong) and a report by the supervisory board. The inlusion of these two different diretors reports is onsistent with the two-tier system of Duth organizations, whih would seem to aount for the fat that both boards submit a separate review about past ahievements. Similarly, British informants explanation of the diretors report as a text about issues involving exeutives as well as non-exeutives orresponds with the one-tier system in UK ompanies. As UK ompanies have one entral board of management, their annual reports only inlude one report of diretors. A third ross-ultural differene in Table 4.9 onerns the pereived reader audiene of the operational review in Duth and British annual reports. Duth responses indiated that the operational review is relevant for multiple reader groups, i.e. shareholders, lients and employees. Conversely, British responses emphasized the importane of the shareholders as the primary reader group for the operational review. As was suggested before, this ross-ultural differene may have been aused by the overall British preferene for the shareholder ommunity as the main audiene for the full annual report. Apart from the differenes aross national business ultures, the desription of annual report setions also revealed several ruial ross-ultural similarities. For instane, Duth and British responses showed that the rhetorial aims of the various annual report setions all ontribute to the realization of the two general ommuniative purposes of the annual report (setion 4.1.1). Informants explained that throughout the annual report, texts inorporate either one of the general objetives i.e. informative or persuasive or inorporate both of these objetives informative-persuasive. In the light of the present genre study, whih intends to onentrate on non-finanial annual report texts with a promotional potential, Table 4.9 displays five types of setions that speifially fulfill this riterion: the orporate profile, the management statements, the operational review, the orporate governane/remuneration report and the CSR/HRM report. Also, informants indiated that the general reader groups reognized for the full annual report (setion 4.4.2) are addressed in various setions of the report. Eah setion should add to the fulfillment of the information needs assumed to be harateristi of the omprehensive audiene of the annual report. For example, it appears that the orporate governane/remuneration report and the CSR/HRM report in partiular aim at speial interest groups, e.g. authorities or NGOs. Within the urrent investigation, it was deided that these speial interest setions would be omitted from further analyses and that the fous should be on the non-finanial texts reporting about the general profile and performane of the ompany (i.e. the orporate profile, the management statements and the operational review, see also setion 3.3). A final onlusion that an be drawn from the results in Table 4.9 is that none of the setions in the annual report is idential to another setion. Although some overlap may exist in purposes or readers, eah text in the annual report serves a unique set of aims and topis. In other words, the annual report is omposed of different setions that are intertextually related to onstitute the purposes of the omplete doument.

87 4.4.5 English language-poliy for the (statutory) annual report As was indiated in setion 1.2, prior researh has pointed out that the status of English as an international business language is inreasing, both within the Netherlands and beyond (e.g. Daelemans, 2005; Rogerson-Revell, 2007). Empirial data obtained through the qualitative survey among Duth listed ompanies (setion 4.3) speifially onfirmed this trend for annual reports originating in Duth business ulture: a majority of these ompanies already used English as the base language for their 2003 annual reports. Between 2003 and 2007, several Duth orporations ontinued with this language poliy for the statutory version of the annual report. Large orporations suh as Philips, KPN and Wolters Kluwer gained overwhelming shareholder support for the proposed introdution of an English statutory report; at their annual general meetings, more than 98% of total shareholder votes favored English over Duth (Philips, 2003; KPN, 2004; Wolters Kluwer, 2006). Any disomfort with the proposal that English should be the statutory language for the annual report was redued by the affirmation that Duth summaries of the English report would be made available for Duth stakeholders. Similarly, it seemed that listed ompanies in the UK predominantly depend on English as an international finanial language. Judging from the British annual reports eligible for seletion in the present study, British ompanies almost exlusively use their native language to ommuniate with international stakeholders about their finanial position. Their 2003 annual reports only sporadially referred to foreign language publiations. To determine the ideas underlying the adopted language poliies for the annual report, the Duth and British interviewees were asked to explain their language hoies. The informants in the Netherlands argued that there were legal, ultural and pratial reasons for the maintenane of an L1 doument, either in the form of a full Duth annual report or a Duth summary of results. First of all, they stated that Duth is inorporated as a primary language beause their ompany is quoted on the Amsterdam Stok Exhange and needs to omply with Duth legal requirements for aounting and finanial reporting in annual douments and annual general meetings. They stressed that in the ase of onflits or other ritial irumstanes, it is ruial to have a report that an be aurately interpreted by the Duth authorities, e.g. the Court or Chamber of Commere. Moreover, it was argued that Duth is maintained to express an alliane with the ultural heritage of the Duth home market. Aording to a majority of the informants, a third reason for the ontinued use of Duth in annual reports involves Duth stakeholders poor skills in foreign languages. Informants feared that both internal and external target groups, i.e. employees, staff, (elderly) shareholders or journalists, do not have suffiient profiieny in English to fully and orretly understand the ontents of an Englishonly report. To avoid the exlusion of Duth reader groups, some orporations ontinued to produe a Duth version of their annual report. For omparable reasons, some of the partiipating Duth ompanies had preserved foreign languages suh as German or Frenh as media of ommuniation with international stakeholders. In many ases, however, the prodution of these foreign language reports had again been stopped owing to a lak of stakeholder demand. The limited fous on foreign languages that had initially been pereived in the British annual reports, was endorsed by the British interview partiipants. Apart from a few ompanies publishing additional performane summaries in Japanese or Frenh, all partiipating ompanies only produed a full annual report in English.

88 Informants explained that the different styles of English employed in the nonfinanial texts (the annual review) and in the finanial texts of the annual report justify their English-only poliy. While the non-finanial texts are usually written in informal tabloid English, the finanial or tehnial texts are generally drawn up in fairly formal Queen s English. The British informants expeted international readers of the annual report to at least understand the simplified English language used in the non-finanial texts, as these normally ontain little tehnial detail and are tailored to a broader audiene. The expliit reasons Duth and British interviewees provided for the strong English-language poliy in international finanial reporting are displayed in Table It shows that these reasons are mainly pratial and ommerial in nature. It also indiates that Duth informants put forward a few more internally-oriented explanations for the use of English in their annual reports. To be able to ommuniate the orporate ativities to foreign business partners or business units, some Duth ompanies have introdued English as the primary language for internal ontats. This inludes the prodution and distribution of the annual report. Furthermore, many informants represented ompanies managed by international boards, whose non-duth diretors are unlikely to have any profiieny in Duth but are usually familiar with English. Here, English is prioritized for the annual report, so as to reate a doument that an be understood by managers who need to formally agree with its ontent. Table 4.10: Reasons for using English in annual reports by Duth and British ompanies. Reason for use of English Duth annual report British annual report 1. International struture of the ompany X (joint ventures, subsidiaries, departments). 2. Management of the group ompany X inludes international diretors. 3. Commerial operations in international X X markets (international marketing). 4. International distribution of shareholders. X X 5. Registration on the London/New York X X stok exhange (English report is obligatory). 6. Global profiieny in English as a finanial/tehnial language. X X Overall, the reasons presented in Table 4.10 reveal the assumption that English is an international business language whih is properly understood by people at various levels in the professional domain, whether they are native speakers or non-native speakers of the English language. Neither the Duth nor the British orporations appeared to atively onsider the atual effet of their English annual report on international readers, e.g. in terms of the text omprehensibility, text redibility, orporate attrativeness or reliability of the ompany Guidelines for and purposes of photographs in the annual report Nearly all 2003 annual reports produed by Duth and British orporations ontained non-finanial texts that were supported by photographi illustrations of real-life objets. As was explained in setions 1.3 and 3.2, these doumentary

89 pitures have distintive rhetorial funtions in annual report texts: they tend to doument the truth of the ompany represented in the annual report and failitate reader understanding and appreiation of orporate information (e.g. David, 2001; MKinstry, 1996; Preston, et al., 1996). In order to establish the speifi rhetorial value of doumentary images in annual reports of Duth and British ompanies, the in-depth interviews foused on the pragmati riteria for and ommuniative objetives of photographs in annual report texts. Duth and British informants suggested that (doumentary) images are prevalent in the non-finanial texts preeding the fatual finanial aounts. Many also agreed that suh images are inluded to primarily fulfil the needs of nonfinanial readers, as they make texts more appealing and omprehensible for those who have relatively little tehnial knowledge about business administration. Aordingly, informants explained that they used visuals as a means of allowing different types of readers aess to the same fundamental message in the annual report. In partiular, they laimed that doumentary images enable the ompany to ommuniate fatual and thus identifiable visual meanings that require limited ognitive proessing and therefore aid the delivery of the message. The Duth and British interviewees added that the seletion of pitures in their annual reports is often based on strit riteria, sine ompanies usually have a lear view of the orporate truth they wish to ommuniate. Table 4.11 offers an overview of these riteria for eah national business ulture. It reveals that Duth and British orporations partly set the same guidelines for determining the suitability of photographs. Whereas Duth orporations tended to use an additional riterion related to emotional appeal, British orporations applied two extra riteria involving the illustration of historial and textual information. Table 4.11: Seletion riteria for doumentary photos in Duth-based and UK-based reports. Criteria for inlusion of a given image 1. Consisteny in orporate ommuniations images should math the orporate ommuniations style refleted in other publiations. 2. Logial relation to orporate philosophy images should represent the orporate identity, e.g. values, struture, employees, lients, ativities/produts. 3. Clear and purposeful meaning images should have an identifiable ommuniative value that is related to other texts/images in the report. 4. Supportive of (entral) theme in report images should underline the main theme of the report and the speifi theme of an annual report setion. 5. Appeal to reader pereptions images should be attrative and relevant for readers, readers should be able to identify with the visual ontent. 6. Refletion of (historial) developments images should depit the up-to-date position of the ompany. 7. Explanation of textual information images should failitate interpretation of textual details. Duth annual British annual report report X X X X X X X X X X X Complementary to the disussion on guidelines for admission or omission of partiular images, Duth and British informants were asked to explain what

90 ommuniative purposes images are likely to have one they have been aepted for the annual report. Informants replies indiated that photographs are inorporated in the annual report for a number of text-external and text-internal reasons, most of whih onerned orporate identity, the expliation of textual ontent and reader attitudes. Table 4.12 displays all purposes that were typially asribed to pitures in the annual report. It points out that the Duth and British partiipants largely agreed on these purposes; the British interviewees ontributed one extra purpose related to the visual struturing of the text (i.e. to add to the readability of the text). Table 4.12: Communiative objetives of photos in Duth-based and UK-based reports. Rhetorial purpose of images in the annual report 1. to visually illustrate the orporate identity portraits of diretors, lients, produts, markets, et. 2. to provide the ompany with a personal fae portraits of diretors, employees. 3. to emphasize the added value of produts/servies portraits of usage, manufaturing or end produt of highly tehnial/abstrat produt. 4. to emphasize high apaity/professionalism portraits of suessful ases, speialist witnesses, impressive produt qualities. 5. to support the ompany s marketing ommuniations portraits of produts/servies that are in line with orporate branding poliy. 6. to establish a entral theme in the report portraits of items that are topially related. 7. to appeal to reader attitudes portraits expressing a desirable atmosphere of openness, integrity or warmth (related to ompany or produts/servies). 8. to add to the readability of the text portraits breaking up textual information or enhaning liveliness of the text. Duth annual report X X X X X X X British annual report X X X X X X X X After elaborating on the ommuniative objetives of photographs in annual report texts, the Duth and British informants were invited to outline the pratial use of these photographs. They were shown several images extrated from the 2003 annual report of the ompany they worked for and were asked to desribe the rhetorial value of these images in the multimodal doument. Their answers revealed that orporations generally take different approahes to visual rhetori in annual reports: orporations may assign a variety of rhetorial funtions to images with different or layered meanings. This is illustrated in Table 4.13 whih provides two examples derived from interviews in a Duth and in a British onglomerate, i.e. a orporation that onsists of heterogeneous business divisions.

91 Table 4.13: Informants explanations of the rhetorial value of two photos in annual reports. Industry/business ulture Conglomerate the Netherlands Conglomerate the United Kingdom Photograph (Piture on over annual report) (Piture in operational review) Explanation photograph It represents one of the major regions in whih the ompany operates. Instead of fousing on Western people in its photography, the organization hose to fous on Asian people in order to position an important growth market that is less well-known than the Western markets. Furthermore, the projetion of a human being is to express the fat that the ompany tries to be a people-business. Thus, the piture visualizes an internal and external (regional) target group and emphasizes orporate poliy. It pitures an objet in whih the ompany s produt is implemented to make the objet operable. As suh, it displays the end produt and, sine this end produt is impressive and relevant to everyday situations, the photo also implies the added value of the ompany s produt. Moreover, the fat that the portrayed end produt will be used by Departments of Defense, i.e. the end users of the ompany s produt, indiates that the ompany manufatures high quality produts. The effet of the photograph is strengthened by its textual explanation: The Joint Strike Fighter, whih will be the highest volume fighter jet in prodution, inorporates Smiths systems ranging from power management, to atuation, to mirowave able Prodution proess of the annual report In order to gain further insight in the prodution-related ontext of the annual report, Duth and British interviewees were invited to desribe the partiipants and proesses involved in the realization of the full multimodal doument. Their explanations of the design proess overed both the written and the visual disourse in the report. Informants in the Netherlands and in the UK agreed that the annual report is ommonly established through ollaborative proedures in whih one projet manager or oordinator-writer orhestrates all ativities. The projet manager, who generally leads a team of staff members or a speial annual report

92 ommittee, maintains a entral organizational role in the prodution of the annual report. He or she is responsible for planning the prodution proess, assigning tasks, olleting information, drafting texts, seleting visual materials and ooperating with external agenies. Staff members primarily onsult the projet manager about the themati ontent or language use in the annual report and evaluate proposals for its graphi design. Duth and British informants noted that the projet manager is usually employed at the ommuniations or investor relations department. Furthermore, Duth and British responses largely orresponded with regard to the suessive prodution phases in whih the projet manager and staff are ommonly engaged. These phases ould therefore be integrated in one model for the design of multimodal annual reports in the Netherlands and in the UK (see Figure 4.1). Proposal theme annual report + planning + theme and aims of annual report to external designers Colleting information (reports divisions, interviews, desk researh press releases/news/ reports) First textual draft of annual report by projet manager + team of staff members First textual draft to initial authors (divisions, CEO/ Chairman) for revision and approval Final multimodal Final textual Suessive Suessive annual report to draft+multimodal textual draft + textual draft of exeutive board+ design of annual multimodal annual report by audit ommittee/ report by projet designer onept projet manager auditors manager+staff to exeutive + staff members (+supervisory members+ board for board) for external revision and 8 7 approval designers approval 6 Figure 4.1: Prodution proess of the annual report in the Netherlands and the UK. Both Duth and British informants explained that the prodution of the annual report ommenes during late summer, in the year prior to the final publiation date. The first step is to propose a entral theme and a time-table, whih are also distributed to external designer agenies involved in the multimodal onstrut of the annual report (step 1 in Figure 4.1). Next, the projet manager gathers information for the different setions in the annual report (step 2 in Figure 4.1). Suh information is mostly aquired through preliminary reports by divisional and departmental diretors, through interviews with these diretors and/or through desk researh with internal and external messages issued by the ompany during the past year. Staff involved in the initial gathering of information generally inlude the Chairman, several members of the ommuniation department, divisional diretors and the Chief Finanial Offier (CFO). The introdutory statement of the annual report is usually delivered in full format, by the CEO in the Netherlands and the Chairman in the UK. All relevant data are integrated in one text-based onept (step 3 in Figure 4.1), whih is then returned to the initial soures for evaluation (step 4 in Figure 4.1). At this stage, the projet managers involved in the reation of English reports for Duth ompanies are often assisted by native English editors who assess the orretness of

93 the English used and insert any neessary adjustments. The first text onept generally moves bak and forth several times between the projet manager and the initial authors before it is aepted (steps 4 and 5 in Figure 4.1). After the divisional diretors and (non-)exeutives have individually approved of the early textual draft of the annual report, it is presented to the exeutive board, in ombination with a onept for its multimodal design (step 6 in Figure 4.1). The multimodal design onept ommonly results from a ooperative relationship between the projet manager and external professionals. Based on briefing sessions with the projet manager, external designers propose ommuniation strategies for the annual report that math both the orporate identity and the rhetorial funtion of the multimodal text, e.g. in terms of visual style, key messages, report struture and tone of voie. The multimodal design of the annual report usually ontains images that are either derived from a ompany s standard stok of photographs or that are taken anew by an external photographer who is also involved in the initial briefings. The graphi designers pre-selet images from the final photo olletion available and integrate these in the multimodal onept(s) for the annual report. The exeutive diretors need to formally aept the multimodal onept of the annual report, but they often initiate another elaborate revision phase before they give their final onsent (steps 6 and 7 in Figure 4.1). At the end of this phase, the offiial multimodal draft of the report is again presented to the exeutive board, and to the independent auditors and/or audit ommittee. In the Netherlands, it is also handed to the members of the supervisory board (step 8 in Figure 4.1). One these parties have approved the draft (and any translated versions of this draft), it is prepared for publiation. In its published form, the annual report is often used again as a soure of information for next year s report (arrow step 8 step 2 in Figure 4.1). 4.5 Conlusion and disussion The analyses outlined in this hapter foused on the registration and omparison of the ommuniative ontext whih haraterizes the annual report genre in the Duth and British business ultures. As suh, they provided an answer to eah of the RQs 1a-1d. The findings of the qualitative survey and in-depth interviews showed that the genre ontext for the Duth-based and UK-based annual reports was similar but not idential. Considering that the rhetorial value of a genre is one of the most important riteria for its integrity (e.g. Askehave & Swales, 2001; Swales, 1990), there were several ruial ross-ultural similarities between the ontexts of these reports: they orresponded with regard to general ommuniative purposes, types (albeit not priority) of readers and ommuniative funtions of photos. This would seem to explain why the Duth-based and UK-texts are regarded by both Duth and British informants as instanes of the same genre (i.e. the annual report). The results of the ontext analysis also revealed several ross-ultural variations between the Duth-based and UK-based annual report, most of whih were aused by different pereptions of readership and different national onventions in organizational strutures (see Figure 4.2). In sum, these results support Zhu (2005), who argues that different ultural bakgrounds may affet the prodution of a partiular genre, but these ultural differenes need not reate fundamental generi distintions. Setion will summarize the findings related to the marked topis in Figure 4.2. In line with the seond main researh aim of this study, moreover, this onlusion will end with several impliations for genre theory and genre analysis (setion 4.5.2).

94 Duth (English) -overlap- British - Status of English: English as base language for the annual report - Priority target groups: 1. Institutional investors, private investors, analysts, journalists 2. Employees, managers 3. Authorities, lients, suppliers, NGOs - English language poliy: English as an international business language improves hanes to operational and ommerial benefits - Referene book in the finanial ommuniation mix - Priority ommuniative objetives: 1. To inform readers about profile, poliies and performane 2. To influene reader attitude, to reate a positive orporate image - Target groups: institutional investors, private investors, analysts, journalists, authorities, lients, suppliers, employees, managers, staff, NGOs - Brohure in the orporate ommuniation mix - Priority target groups: 1a. Institutional investors b. Private investors 2. Analysts, journalists 3. Authorities, lients, suppliers, employees, management, NGOs - Unique riteria for table of ontent: 1. Reader needs for themes 2. Corporate poliy 3. Benhmarking - Unique text types in table of ontent: 1. CEO s Statement a 2. Exeutive Board Report (title) b 3. Supervisory Board Report - Evaluation in prodution proess: major role supervisory board - Criteria for table of ontent: 1. Legal dislosure requirements 2. Soio-politial/Best pratie issues 3. Competitors annual reports 4. Corporate reporting onventions - Similar text types in table of ontent: 1. Managerial foreword a 2. Corporate profile 3. Operational review b 4. Finanial review b 5. Compulsory note on exeutive and supervisory proedures in the board 5. Corporate governane report 6. Corporate Soial Responsibility report 7. Finanial statements - Communiative funtion photographs: 1. To illustrate the orporate identity 2. To provide the ompany with a personal fae 3. To illustrate the added value of produts 4. To emphasize high apaity and professionalism 5. To support the marketing ommuniation poliy 6. To establish a entral theme 7. To appeal to reader emotions - Prodution proess: ollaborative proess between projet manager, staff, diretors and external designer agenies. - Unique riteria for table of ontent: management approah to finanial reporting (plain vs expressive) - Unique text types in table of ontent: 1. Chairman s Statement a 2. CEO s Statement a 3. Operating and Finanial review (title) b 4. Diretors Report - Communiative funtion photographs: to improve readability of the text by visualizing (abstrat) information - Evaluation in prodution proess: major role exeutive ommittee Figure 4.2: Cross-ultural overlap and variation in the genre ontext of the annual report. Note a : The unique texts in the Duth and British annual reports belong to the higher-order ategory of managerial forewords. Note b : The unique texts in the Duth and British annual reports (i.e. they arry unique titles) omprise an operational review and a finanial review. Note : The unique texts in the Duth and British annual reports onern ompulsory notes on management proedures.

95 4.5.1 Cross-ultural omparison of ontextual elements Assuming that a genre is onstruted and reonstruted by the disourse ommunity [ ] in whih members [ ] reate and share its meanings and writing onventions (Zhu, 2005: 178), the present ontext study relied on the expertise and praties of annual report produers in the Netherlands and in the UK. Sine the inventory of ulture-speifi ontext features (RQs 1a-1) was inherent in the omparison of features aross ultures, the summaries below fous on the ontrastive results (RQ 1d). The first ontextual element disussed, i.e. the status of English in the annual report, was not studied aross national business ultures. Sine insights in this element were partiularly relevant for the final seletion of researh materials drawn from Duth ompanies (see setion 3.3), it was not analyzed within the British business ulture. To finalize the ross-ultural analysis of situational elements, this subsetion will briefly point at the omparability of genre ontexts for Duth-based and UK-based reports. Status of English for Duth-based annual reports As for the status of English, the survey results showed that most of the listed Duth ompanies already produe an original English annual report, whih is often written with the help of native English speakers. For the 2003 annual report, five Duth orporations did not publish a Duth annual report at all; they only ommuniated with (inter)national stakeholders through the English language. The survey results also indiated that Duth AEX orporations in partiular tend to apply English as the base language for their annual reports. This ould be explained by the fat that these large ompanies operate extensively on international markets. By antiipating language poliy hanges in favour of the English language, some respondents implied that the status of English as an international language for finanial reporting is likely to improve in the next few years. Comparison of English language poliy Aording to the Duth and British partiipants in the in-depth interviews, the strong English language poliy for the annual report is primarily attributable to the ommerial effiienies of and legal obligations for English in international marketing. This holds partiularly for international markets where orporations have headquarters and where they obtain revenues through operations and shareholdings. Another reason for Duth ompanies to produe an English annual report involves the fat that a growing number of Duth multinationals employ foreign board members who an only rely on English-language texts for a full understanding of the (performane-related) information they need to formally approve of. All arguments given by the Duth and British interviewees suggested that native English and nonnative English readers of the annual report are believed to be able to understand the English disourse in this report, beause of the inreasing global distribution of English as a finanial or tehnial medium of ommuniation. Comparison of role annual reports in orporate ommuniation mix Interviewees from the Netherlands and the United Kingdom expressed different ideas about the role of the annual report genre in the orporation s mix of ommuniation instruments. Whereas both the Duth and British informants positioned the annual report within the finanial ommuniation mix, only the British informants expliitly related it to the marketing ommuniation mix as well.

96 Duth and British informants proposed that the annual report omprises an arhival work of referene that is based on or omplements other, more dynami finanial ommuniation tools. As suh, it belongs to the network of finanial publiations in the Duth and British business ultures. However, the interview results in the UK also showed that British ompanies see broader ommuniation opportunities for the annual report. British informants believed that, as the annual report ontains a omprehensive pakage of ompany-related bakground information, it is also useful as a orporate brohure for stakeholders with a general (as opposed to a diret finanial) interest in the ompany. Comparison of the general ommuniative objetives of the annual report Despite the different views on the position of the annual report as a orporate ommuniation tool, Duth and British interviewees agreed on the overall ommuniative purposes of the report in its entirety. They laimed that the primary objetive of the annual report is to inform reader groups about a variety of ompanyspeifi issues, while its seondary objetive is to influene reader attitudes. Apart from providing objetive information, the annual report is intended to establish a positive orporate image among readers so as to eliit goodwill and a (finanial) interest in the ompany. All informants believed that the non-finanial texts preeding the finanial aounts are most useful in trying to ahieve this seond goal. Furthermore, Duth and British informants emphasized that the annual report does not in itself atively influene investment behaviour; they argued that the annual report may help to urge this behaviour in ombination with other, interpersonal investor relations ativities, e.g. meetings with the (institutional) investor ommunity. Comparison of the general ommuniative target groups of the annual report With respet to the overall audienes of the full annual report, Duth and British interviewees listed idential reader groups haraterized by unique information needs that were usually addressed by speifi rhetorial strategies in the report. Despite the itation of equivalent reader types, however, there were ross-ultural differenes in the priority assigned to eah of these groups. The Duth informants stated that the omprehensive finanial reader group (i.e. institutional and private investors, analysts and journalists) is the most important audiene for the annual report, followed by the internal target groups (employees and managers) and the broad audiene (lients, speial interest groups and poliy-holders). Conversely, British informants onsidered the shareholders as the ruial audiene for the annual report, with partiular emphasis on the institutional shareholders. They regarded the remaining finanial professionals as the seond important reader audiene, and plaed employees in the least important target group, among readers in the general publi. Therefore, the interview results revealed that in spite of its dual ontribution to the orporate ommuniation mix, the British report is a shareholder doument most of all. In Duth business ulture, the annual report aims at a less speified professional audiene with a diret finanial interest in the ompany. Comparison of the riteria for the table of ontents in the annual report In informants desriptions of the riteria for the ontent of the annual general report, there was a high degree of ross-ultural similarity in terms of guidelines determined by law, soio-politial issues, ompetitors ontent hoies and ompany-speifi reporting onventions. The small variation that did exist aross the ontent riteria for Duth-based and UK-based annual reports exlusively

97 involved optional guidelines, e.g. benhmarking or management s approah to finanial reporting. Presumably owing to the large variety of legal, soial and ethial developments that have reently impated the finanial reporting proess, Duth and British informants ould not provide a unanimous and lear-ut distintion between annual report setions that are either ompulsory or non-ompulsory in the Netherlands and in the United Kingdom, respetively. Comparison of the text types inluded in the annual report Although several Duth and British interviewees laimed that they do not onsiously onsider the existene of different setions or text types while working on the annual report, their explanations of the ommuniative value of various annual report texts revealed that they do reognize rhetorial distintions between these texts. Both Duth and British informants observed that the non-finanial annual report texts in partiular may differ with respet to ommuniative purpose, target group and, hene themati ontent. A ross-ultural omparison of responses, moreover, showed that annual reports produed in Duth orporations are generally onstituted by texts that are subtly different from the texts in the British annual reports. This was illustrated espeially in referene to the managerial forewords. Most likely as a onsequene of the distint Duth two-tier management system and the British one-tier management system, the Duth annual reports only ontain a CEO s statement, whereas the British annual reports inlude a Chairman s statement and a CEO s statement. Judging from the informants responses, the Duth CEO s statement is marked by ontextual features that an be traed in both the British Chairman s statement and the British CEO s statement. Like the British CEO s statement, it generally ontains biased and relatively detailed information about the performane of the group ompany; like the British Chairman s statement, it is interational and needs to provide the ompany with a personal fae. It should be noted that in all their text desriptions, the British informants tended to emphasize the importane of the shareholders as a ruial reader group, whih is in line with the general notion that shareholders form the number one reader audiene for the full British annual report. Finally, Duth and British interviewees indiated that, although the annual report onsists of setions that an be defined as independent text types, the ombination of these setions is required for the onstitution of the omplete annual general report. Comparison of the ommuniative funtion of photographs in annual reports Duth and British informants largely agreed on the riteria for and ommuniative purposes of photographs in the annual report. Aording to the informants, a piture is generally onsidered appropriate for the annual report if it is onsistent with the orporate ommuniation style, if it properly reflets the orporate philosophy, if it serves an obvious purpose within the annual report and if it relates to the (entral) theme of the report. They also shared the opinion that the ommuniative objetives of photographs in the annual report primarily omprise the illustration of the orporate identity, the personifiation of the ompany, the emphasis on added value and high quality of orporate ativities, the refletion of the orporate marketing poliy, the realization of (entral) themes and the appeal to readers emotions. Despite the ioni or fatual nature of doumentary images in the annual report, Duth and British responses indiated that these images may still demand quite a lot of ognitive proessing by readers. Suh images often have omplex rhetorial funtions that require the interpretation of multiple or layered visual meanings.

98 Comparison of the prodution proess of the annual report Finally, interview results onerning the realization proess of the annual report showed that in both the Duth and British business ultures, the annual report is based on information or texts derived from different people inside and outside the organization. In the Netherlands as well as in the UK, the information providers usually engage in a ollaborative design proess, whih is headed by one internal projet manager who orhestrates all proedures for the prodution of the annual report and is largely responsible for writing and editing all textual and graphi onepts. From the moment the first multimodal drafts for the annual report have been realized, the exeutive boards of Duth and British ompanies are atively involved in the revision and approval of the text. Despite the many ross-ultural similarities in the ollaborative prodution proess of the annual report, a fundamental differene seemed to onern the role of non-exeutive board members in the offiial (dis)approval of the multimodal text onept. While the Duth informants explained that non-exeutive diretors, i.e. members of the supervisory board, are losely involved in judging the auray of the final draft of the multimodal annual report, the British informants suggested that besides the nonexeutive Chairman only exeutive diretors are involved in evaluating this draft. Genre ontext and the Duth-based and UK-based annual reports as a genre set Consistent with RQs 1a to 1d, the above summary of survey and interview results highlighted and ontrasted the typial ontextual features of annual reports originating in the Netherlands and in the UK. Considering that RQ 1d fouses on the extent to whih the ommuniative ontext of Duth-based and UK-based annual reports differed, it would seem that RQ 1d in partiular refers to the extent in whih genre ontext marks these reports as omparable instanes of the same genre. Although annual reports are reognized as suh on a global sale, ross-ultural researh on these business texts requires that they be subjeted to an empirial analysis of their atual ontextual omparability. Here, the omparability of texts involves the detetion of similarities in the situational interpretation of annual reports in varied national business ultures (e.g. Miller, 1994). As the researh results suggest, the genre set onept forms a means of explaining in what way the annual reports produed in the Duth and British business ultures are alike. The informants in the in-depth interviews speifially indiated that the full annual reports of Duth and British orporations onsist of series of omponent setions or omponent text types. They suggested that the different text types within ompanies annual reports are interrelated in terms of purpose, audiene and topi of the message. These interrelated text types were believed to generate the full annual report: informants explained that the report atually derives its ommuniative value from the ombination of texts it ontains. As suh, the results of the present ontext analysis appear to onfirm the observations of previous genre-based studies of annual reports, where the report setions were found to reflet multiple narrators, different target groups or various topis (Jameson, 2000) and different degrees of informative or affetive objetives (Garzone, 2005). Based on the study of ontextual elements, it an thus be onluded that both the Duth-based and UK-based annual reports omprise a genre set. As was outlined in setion 2.3.5, Devitt (1991) defines a genre set as a olletion of diverse texts or genres produed and used within a partiular oupational field. She argues that genre sets emerge as a onsequene of the existene and stabilizing funtion of (...) intertextuality, where intertextuality involves referenes to similar rhetorial audiene, purpose or subjet within and aross genres (Devitt, 1991: 341).

99 Combining the results of the urrent qualitative survey and in-depth interviews, it an also be observed that the full annual reports of Duth and British ompanies omprise similar, although not idential sets of genres. They are omparable examples of the same genre set in the sense that they have a large number of ontextual features in ommon, suh as general and speifi ommuniative purposes, reader groups and the riteria or proesses involved in the multimodal design of the annual report. Following Miller (1994), this sharing of ontextual elements auses the Duth-based and UK-based annual reports to have a omparable exigene, i.e. they are solutions to the same soial or rhetorial problems in a given professional setting. Having determined that the genre ontexts of annual reports originating in the Duth and British business ultures are suffiiently equivalent to identify the reports as omparable instanes of the same genre set, it seems reasonable to inlude these reports in a further ross-ultural study of multimodal genre features (see Chapters 5-7). Given the high status of English as an international finanial language in the two business ultures, moreover, it seems plausible to speifially fous on the ross-ultural omparison of genre features in Duth-English and British-English annual reports. Finally, the interview results indiated that annual reports produed in the Netherlands and in the UK do not funtion as independent business text. As a genre set within the finanial ommuniation mix of Duth and British orporations, for example, the annual report supports and is omplemented by a range of different but oherent texts or genres, suh as press releases, interim reports and shareholder meetings. Orlikowski and Yates (1994) have referred to suh an intertextual network as a orporation s genre repertoire (see also setion 2.3.5). Apart from the omparable use of the Duth-based and UK-based annual report as a genre set within the repertoire of finanial text types, the researh findings revealed that British annual reports are also employed in the repertoire of marketing genres. The assignment of the annual report to diverse genre repertoires is in line with findings by Rogers (2000) and Fox (2006), who laim that the annual report o-exists with genres in finanial reporting and orporate publi disourse genres, respetively Impliations for genre theory and genre analysis In response to the seond main researh question of the present study, this subsetion will put forward several impliations that an be drawn from the ontext investigation presented in this hapter. These impliations may help refine both genre theory and genre analysis in general. The theoretial importane of disourse ommunity in the identifiation of genre ontext As was explained at the beginning of this hapter, the relationship between genre ontext and the disourse ommunity in whih the genre is employed, implies an identifiation of situational elements that is strongly onditioned by text prodution onventions (e.g. Hyland, 2007; Swales, 1990). The disourse ommunity onept omprises the soial or ommuniative issues that influene the writing environment. In priniple, it allows insight in the soial status or funtion of texts, in the onventions for text design, in the writers soial roles, in the general writing behaviours of these writers, et. (Beaufort, 2000). Therefore, a fous on disourse ommunity may be fruitful in establishing the rhetorial praties and onventions that underlie the omposition of professional genres, whih generally have well-

100 defined aims and audienes. As the present researh results indiate, the expertise of members belonging to the disourse ommunity of annual report produers ontains a rih soure of information; these expert members were apable of drawing a omprehensive overview of the rules, proesses and issues that generally determine the multimodal design of the annual report in the Netherlands and in the UK in partiular. This overview, moreover, would seem to be a relevant resoure in further analyses of speifi genre features, where the existene of partiular substantive or formal items should be regarded as the produt of the genre s rhetorial setting (in Chapters 5-7). In spite of the usefulness of text produers expertise in the identifiation of the professional genre ontext, it must be noted that the onept of disourse ommunity provides a limited overview of ontextual elements. Sine it fouses primarily on the expert ommunity involved in the realization of a given professional text type, it does not entail a detailed onsideration of the genre ontext as it is pereived by the reeivers of this text type. Although a diret and insightful relation an be established between the text produers ideas within the expert disourse ommunity and the atual multimodal omposition of their texts, these ideas should not be regarded as naturally representative of the ontextual pereptions of the reader ommunities, who may have a shared interest in the texts but who are unlikely to share the same expert bakground (see also setion 2.2). The theoretial importane of ross-ultural variations in genre ontext Notwithstanding the fundamental ontextual similarities between the annual reports of Duth and British ompanies, the urrent in-depth interviews also exhibited several subtle ross-ultural differenes. In partiular, these differenes onerned the pereived position of the annual report in the orporate ommuniation mix, the prioritization of reader groups and the rhetorial value of and partiipants in individual annual report texts. Based on these observations, it an be argued that orporations from different business ultures are allowed a degree of freedom in the manifestation of genre ontext, while maintaining the integrity of the genre involved. It appears that as long as there is suffiient overlap between the disourse situations or genre ontexts within whih texts are used, instanes of business texts suh as the Duth-based or UK-based annual report are reognized as examples of the same professional genre. In other words, ross-ultural variations that do not touh the ore harateristis of a genre (e.g. ommuniative purpose, see Askehave & Swales, 2001) are unlikely to ause fundamental generi distintions between texts originating in different national business ultures. Although the individual or ulture-speifi exploitation of ontextual norms for partiular text types is hardly addressed in genre theory, the present researh results indiate that the flexibility of ontextual boundaries aross ultures is as muh an issue in the generi disourse approah as the flexibility of strutural or lexio-grammatial genre rules aross ultures (e.g. Mauranen, 1993; Vergaro, 2002, 2005). The analytial relevane of the qualitative survey and in-depth interviews In general, the qualitative survey and the in-depth interviews allowed for a detailed and pratial insight in the ommuniative ontext in whih the annual reports of Duth and British ompanies are typially realized. Consistent with Wester and Peters (2004), the qualitative survey appeared to be an effiient devie to apture the norms in English and Duth language poliies for the Duth-based annual report of the 2003 finanial year and beyond. With regard to the in-depth interviews, moreover, the semi-standardized design allowed for a strutured but open onversation with annual report produers and, aordingly, resulted in a relevant

101 and rih desription of genre ontext (e.g. Flik, 1998). In ombination, the findings of the qualitative survey and the in-depth interviews provided a potential basis for rationalizing the substantive and formal features of Duth-English and British annual report texts (see Chapters 5-7). The analytial relevane of language poliy issues The results of the qualitative survey showed that English-language annual reports of Duth orporations are regularly produed with the help of native English speakers. Considering the intervention of native English writers or translators, it ould be suggested that the impat of Duth language onventions has been redued in the original English versions of the Duth-based reports. Consequently, the relevane of the omparison between Duth-English and native British-English annual report texts might be questioned. In the urrent study, however, the fous was on ulturespeifi, as opposed to language-speifi, influenes on text prodution. Assuming that ultural identity goes beyond language use in the sense that ulture is expressed through language (e.g. Claes & Gerritsen, 2007), it is argued here that the ontrastive analysis of English annual reports originating in traditionally Duth and traditionally British orporations remains a relevant researh example. The analytial relevane of genre ontext for reader response analyses To date, no study known of has expliitly related the onept of genre ontext to the analysis of reader response. More speifially, and within the sope of the present study, it remains to be investigated (1) if the ontextual features determined by the text produers atually apply to the readers of the genre and (2) if ontextual elements that are typial to a partiular genre in one ulture are appreiated or understood as suh by genre reeivers in another ulture. Again referring to the results of the in-depth interviews, it would be interesting to find out whether the English annual report texts really reah their intended international readers and whether the English texts ahieve the ommuniative purposes defined for these readers. For example, a reader effet analysis ould fous on the relevane of Duth- English and British-English management statements, as pereived by the investor. As an initial step in the generi identifiation of Duth-English and British-English annual reports, this hapter foused on the ommuniative situation in whih these reports are produed. The next hapter ontains the first in a series of three textbased analyses, in whih the textual and visual features within the management statements, orporate profiles and operational reviews are identified and ompared aross ultures. Chapter 5 will present a ross-ultural study of the ontents of these annual report texts.

102

103 Chapter 5 Charaterization of the genre ontent of Duth-English and British-English annual reports Having identified the genre ontext of the annual reports originating in Duth and British orporations, the present hapter omprises a first analysis of text-related features that are generally defined by this genre ontext. It inludes an investigation of the multimodal genre substane (Miller, 1994) or multimodal text ontent of Duth-English and British-English annual report texts. Aordingly, this hapter aims to find a partial answer to speifi RQs 2a to 2d, whih fous on the rossultural identifiation, omparison and ontextual explanation of the typial textual and visual disourse features in the Duth-English and British annual reports (see setion 1.5). In general, it has been suggested that the information ontent of orporate annual reports is onstrained by and rafted on the basis of the soial or aounting ontext in whih these reports are produed (e.g. Rutherford, 2005; Smith & Taffler, 2000). This suggestion is underpinned by genre theory, whih assumes that an understanding of what the text is about is losely related to knowledge about the ommuniative situation in whih this text re-ours. Genre analysts have argued that onsiderations of soial situation, disourse partiipants and ommuniative purpose in partiular form the rationale for the kind of information onveyed through a partiular text type (Berkenkotter & Hukin, 1995; Swales, 1990). Furthermore, the ontent of a text is often onsidered to be the most basi feature by whih to distinguish a partiular genre, sine it influenes the formal strutural and lexio-grammatial genre harateristis that symbolize this ontent (e.g. Berkenkotter & Hukin, 1995; Nikerson, 2000). Based on these onsiderations and onsistent with the multiultural disourse ommunity perspetive taken towards the analysis of genre ontext in Chapter 4 (e.g. Hyland, 2000; Vergaro, 2005), the investigation presented below foused on ontent-related text elements employed by expert reators of the annual report in the Duth and British business ultures. Genre ontent aross ultures was analyzed to determine to what extent Duth-based and UK-based produers of the English annual report realize these douments in a generially (dis)similar way. Setion 5.1 further explains the theoretial assumptions that guided the ross-ultural ontent analysis. It suessively disusses the nature of textual ontent and visual ontent in multimodal disourse and, as suh, provides a working definition of the onept of genre ontent. Setion 5.2 elaborates on the methods used to analyze the speifi ontents of Duth-English and British annual reports. Additionally, it presents the lists of text themes and photo themes that were relevant for the present ontent study. Setion 5.2 ends with a report on the results of an interoder test that was onduted to measure the reliability of these methods and lists. Setions show the ross-ultural results for typial ontent features in different setions of the annual report, i.e. in the management statements, profiles and operational reviews. A onlusion and disussion with regard to the full ontent analysis an be found in setion 5.6. In line with the seond main researh objetive (setion 1.5), this setion will also provide several impliations for genre theory and genre identifiation in general.

104 5.1 A working definition of genre ontent: oneptual themes As was indiated in the theoretial sope of this dissertation (setion 2.3.4), studies on the ontent of business genres in general and annual reports in partiular have adopted various tehniques to map the propositions in these texts. For instane, Berkenkotter and Hukin (1995) identify genre ontent in terms of topis in the text, while Rogers (2000, 2001) looks at reader-orientation in the message of the text. Also, it appears that genre ontent an be analyzed by quantitative measures, suh as Rutherford s word ounts (2005), or by qualitative measures, suh as Nikerson s message desriptions (2000). An overview of generi ontent studies also reveals that different disiplinary or national ultures may have different pereptions of what is aeptable in the onstrution of the informational landsape in a given text type, i.e. they may have different ideas about what the ontent of a genre should be omprised of. Sims and Guie (1992), for instane, found differenes between the ontents of English letters of inquiry by US, Asian and European students. Despite the important analytial impliations of these diverse approahes to genre ontent, several questions remained as to their relevane for the present analysis of the multimodal ontent of annual reports originating in different national business ultures. First, it appears that Miller s (1994) definition of genre ontent as a ognitive-semanti onept has reeived relatively little attention in genre analysis. Aording to Miller, genre ontent i.e. genre substane onerns the semanti value of disourse whih holds aspets of ommon experiene that are symbolized and made reognizable or interpretable through the formal features of a text (1994: 32; see also setion 2.3.4). Most ontent-based genre studies have offered a general qualitative or quantitative representation of the ontent items observed but have hardly explained what ognitive proess was involved in identifying these items within the text. Hene, it was neessary to expand the analytial frame for the analysis of genre ontent in terms of ognitive interpretations. Seond, analyses of genre ontent seem to refer primarily to texts used within one type of (ultural) ommunity; few genre analysts have addressed distintions in text ontent of similar genres aross ultures. Assuming that genre ontent omprises the symbolization of aspets of ommon experiene in a partiular disourse ommunity, different ultures may employ different ontents in the same situations involving the same partiipants in order to ahieve the same rhetorial ation (Nikerson, 2000: 41). Although genre ontent may not in itself define a partiular text type in a given soial ontext, it offers a basi starting point for the present genre study of annual reports produed in the Duth and British business ultures. As will be shown in Chapters 6 and 7, the examination of genre ontent is omplemented by the analyses of the strutural and lexio-grammatial onstruts of annual report texts. Finally, it should be noted that any analysis of genre ontent so far has mainly been restrited to the written text. Genre ontent expressed through other disursive modes inorporated in a text have only sporadially been referred to, e.g. the analysis of visual omposition in onferene presentations by Rowley-Jolivet (2004). The present study posits, however, that the multimodal view on ontent analysis will ontribute to a more omplete understanding of the annual report genre as it is produed by Duth and British ompanies (see also Kress & Van Leeuwen, 2001).

105 In an attempt to provide answers to the above questions, this study foused on oneptual themes as the onstituents of genre ontent. Aording to researh in grounded theory, a theme generally represents an integrated set of disrete onepts that are (non)verbally expressed and that identify or label a soial phenomenon (e.g. routine ativities, happenings). In other words, a theme is disovered when onepts are ompared one against another and appear to pertain to a similar phenomenon (Strauss & Corbin, 1990: 61). Ryan and Bernard add that themes as onept-based lassifiations for phenomena in soial reality also apply to multimodal resoures; they are abstrat [ ] onstruts that link not only expressions found in texts but also expressions found in images, sounds and objets (Ryan & Bernard, 2003: 87). With regard to the investigation of text ontent, the grounded theory approah is related to a seond approah, i.e. the ognitive mapanalyti perspetive. Text analysis in grounded theory fouses on the researher s intuitive observation of onepts that explain general soial odes; onversely, text analysis in ognitive mapping ombines the researher s qualitative data interpretation with the quantitative identifiation of oneptual themes expressed in the onrete words of the text (Ryan & Bernard, 2000). As Carley points out, map analysis ompares texts in terms of both onepts and the relationships between them [ ] Differenes in the distribution of onepts and the relationships among them aross texts provide insight into the similarities and differenes in the ontent [ ] of the texts (1993: 92). Carley (1993) and Carley and Palmquist (1992) further explain that ognitive mapping implies situated onepts, the meaning of whih depends on multiple ontextual fators suh as the link between ognitive models and language, the relation between words and meaning and the soial knowledge required for an aurate understanding of the oneptual networks. Although the definitions of oneptual themes in grounded theory and ognitive map-analysis are epistemologially different, they may well be omplementary within the analysis of text ontent. Following Ryan and Bernard (2000; 2003), it an be argued that ontent analysis would benefit from the suessive exeution of open oding praties, used to ategorize the themes pereived in text fragments or in other modes of disourse (as in grounded theory), and oneptual networking, where key words in the texts or key elements in other modes of disourse are then assoiated with eah theme (as in ognitive mapping). These proedures would allow ontent analysts to use key words to searh for additional ourrenes of the theme in large orpuses of text (Ryan & Bernard, 2000: 792). With regard to the present genre study, the ognitive map-analyti approah to text themes would seem to over all relevant aspets of genre ontent: (1) it depits the informational landsape within the text, (2) it aounts for the ognitive proesses involved in the establishment of the semanti value of the text, (3) it allows for the integration of soio-ultural ontexts that potentially influene the genre ontent and (4) it enompasses the information refleted in both verbal and visual languages. In addition, the analysis of oneptual themes allows for the implementation of qualitative as well as quantitative data proessing, thus failitating the identifiation of large orpora of researh materials. Overall, the identifiation of themes as ognitive oneptual networks in partiular appears to be an appropriate means by whih to obtain insight in the ognitive-semanti nature of genre ontent in multimodal annual report texts originating within the Duth and British business ultures. Compared to prior genre studies that have foused on text ontent (e.g. Berkenkotter & Hukin, 1995; Rutherford, 2005), this multi-method study of onept-bases themes would seem to be able to generate a more omprehensive or onrete idea of what the instanes of a genre are about. In line

106 with the present researh fous, setions and further disuss the refletion of oneptual themes within the ontents of written texts and doumentary photographs Coneptual themes in written texts Within written texts, themes an be traed in subordinate onepts expressed through a single word suh as friend or writing or a phrase suh as works well with others (Carley & Palmquist, 1992: 607). In this sense, language-based onepts or keywords underlying the text theme remain individual ideas without meaning as long as they are not interonneted by a theme that relates them to the same phenomenon. The relationships between thematially oherent onepts an be shematized in ognitive mapping models. The (inomplete) network of onepts shown in Figure 5.1 is based on Carley and Palmquist s (1992) extration and generalization of onepts ommuniated in an interviewer-student onversation about how to do researh. One of the generalized onepts is books/literature, whih an be regarded as an abstrat theme diretly inferred from verbalized onepts in the spoken text. Text fragment interview Interviewer: How would you find out Students: Magazines, books, enylopedias Interviewer: Anything else. Student: Nothing offhand, I mean, there s things, If you had resoures you would know something offhand. Generalized onept = theme Books/literature Extrated onepts = words/phrases magazines books enylopedias resoures Figure 5.1: Cognitive map of oneptual theme traed in verbal text (based on Carley & Palmquist, 1992: 607). Defining themes as ognitive maps or networks of semantially related onepts suggests that the identifiation of themes depends on analyzing oneptual lues inorporated within the text. Ryan and Bernard (2003) propose eight interrelated interpretational tehniques, or srutiny tehniques by whih verbal expressions an be linked to a higher-order theme in the mind of the interpreter. As is shown in Table 5.1, these primarily onern the observation of (1) repetitions of linguisti expressions with similar oneptual meanings, (2) indigenous typologies, (3) metaphors and analogies, (4) transitions in oneptual themes, (5) similarities and differenes in oneptual meaning, (6) linguisti onnetors between keywords for the oneptual theme, (7) missing data referring to implied themes and (8) theoryrelated onepts enountered in previous studies. It appears that these srutiny tehniques involve the study of individual verbal items and then a higher-level omparison of themati patterns to enable the identifiation of separate themes. With regard to the present study of written douments, Ryan and Bernard s tehniques

107 offer lues about what to look for in a text when identifying the themati ontent of a genre. Sine a pratial explanation of the identifiation of themes failitates the reprodution of and therefore enhanes the reliability of a ontent analysis, Table 5.1 further outlines the srutiny tehniques as they were used in the present study. To emphasize the appliability of these tehniques, the examples in Table 5.1 were extrated from the Duth-English and British-English researh materials analyzed. Table 5.1: Detailed desription of Ryan and Bernard s (2003) srutiny tehniques identifiation of text themes. Srutiny tehnique Explanation + example English annual report 1) Repetitions Reurrent linguisti expressions referring to the same onept. Example: suessfully aquired, expanding, divest, a 40% stake in, disposals, aquisitions and equity interest for the investments/divestments theme. 2) Indigenous Terms speifi to a ommunal (or ultural) domain. typologies Example: linguisti referenes to the ost management theme, e.g. ost savings, ost ontrol and operating expenses, in Duth-English 3) Metaphors and analogies operational reviews. Ambiguous linguisti expressions referring to the same onept. Example: The past ouple of years have been tough ( ) and some of us have felt the hill for the business ontext theme. 4) Transitions Linguisti markers indiating a shift in themes. Example: Retail profit grew by 12.8% to 372 million ( ) B&Q also inreased its share of the UK s market, where also marked a transition from the finanial performane theme to the operational performane theme. 5) Similarities and differenes in oneptual meaning 6) Linguisti onnetors 7) Missing data referring to implied themes 8) Theory-related onepts Abstrat interpretation of similarities or differenes between two or more sets of words, phrases or texts that eah omprise a partiular themati onept. Example: the answer to the question how does this fragment on orporate governane vary from preeding/sueeding fragments? Linguisti expressions that either onnet keywords belonging to a partiular theme or disonnet them. Example: The opportunity for UK Retail to improve the returns on its urrent portfolio in 2004 is lear. However, we are also extremely well plaed to ontinue playing an ative role in the onsolidation within the industry over the oming years, where However still indiates the ontinuation of the onfidene in the future theme. Conealed information whih is not made expliit in the text but whih atually refers to an underlying theme. Example: the real operational impat of bad eonomi or politial or environmental onditions is left impliit, despite its ruial role in the overview of last year s results (e.g. the tone of the themes business ontext and operational results is onsistently optimisti, despite diffiult times). Referenes to relevant text themes observed in previous studies and to relevant reasons for any distintions between theory and the urrent material. Example: drawing on text themes found by Osborne at al. (2001) in presidents letters to stokholders in US annual reports. For instane, their theme marketing with keywords division, produt, sales and market and their theme R&D with keywords new, failities, plant and researh were also relevant for the present study.

108 As for the present investigation of genre ontent, the advantage of these ognitive tehniques for the interpretation of oneptual themes lies in their multimodal appliability: whether the data ome in the format of video, audio, or written douments, handling them is always helpful for finding themes (Ryan & Bernard, 2003: 89). This will be illustrated in the next subsetion, whih elaborates on the identifiation of oneptual themes in the visuals of multimodal texts Coneptual themes in photographs Themati analyses based on ognitive mapping are not restrited to verbal texts, but an also be used for multimodal texts. The relevane of oneptual themes in images is shown by Kress and Van Leeuwen s semioti interpretation of multimodal texts in popular magazines, where different types of visualizations signify the disursive ontent of photographs and artiulate how these photographs, and any additional text, should be pereived and understood (Kress & Van Leeuwen, 2001). In their investigation of pitures of kithens inluded in English home magazines, they repeatedly observe photographi meanings in whih the themes of tradition, heritage and aesthetis are brought together to reate a mood of nostalgia (Kress & Van Leeuwen, 2001: 31). These photographi themes are reognized as suh by the interrelation of visualized onepts, or what Kress and Van Leeuwen (2001) refer to as a set of semioti signs, aross several visual modes. With respet to the objetsas-mode, for instane, the main themes are established by the ombined and situated meaning of portrayed objets suh as a table and hairs with turned legs, soft furnishings and enamel jugs. The same ognitive proess also applies to the oloras-mode, where the mapping of pereived olors, like warm reds, bright yellow and blue-and-white heks, evokes the themes related to nostalgia. Returning to the notion of theme as a ognitive lassifiation of key onepts, the photographi realization of themes through the objets-as-mode introdued by Kress and Van Leeuwen (2001) an also be shematially depited as a ognitive network (see Figure 5.2).

109 Generalized onept = theme Nostalgia (tradition, heritage, aesthetis) Extrated onepts [piture: table + [piture: soft [piture: enamel = visual signs hairs with furnishings] jug] turned legs] Figure 5.2: Cognitive map of theme traed in objets-as-mode in photograph (based on Kress & Van Leeuwen, 2001: 30, 31). Although this has not yet been expliated in previous work, most of the srutiny tehniques disussed in Ryan and Bernard (2003, in setion 5.1.1) also appear to be relevant for a repliable and thus reliable identifiation of onept-based themes in visuals. Tehniques 1-5 and 7-8 in partiular an be translated onto the analysis of themati ontent in visual illustrations, e.g. photographs. Tehnique 6 (involving linguisti onnetors), however, seems to be appropriate for the analysis of verbal text themes exlusively. An alteration of this tehnique into visual onnetors within doumentary photographs would not add value, sine the objets portrayed in the piture are already logially onneted by the piture frame, e.g. the link between visual onepts related to the nostalgia theme in English kithen images is inherent to the fat that they are plaed together inside a piture frame. To illustrate the usefulness of srutiny tehniques 1-5 and 7-8 for the identifiation of photographi themes in annual reports, Table 5.2 shows an adapted version of Ryan and Bernard s model. For eah tehnique, it speifially refers to riteria that helped define the visual themes in these reports. Again, the examples inluded in the table were taken from the urrent researh texts. It must be noted that the observation of photographi themes partly depends on the themes diserned in the written texts with whih they were juxtaposed in the annual report. These written texts offered an interpretative ontext, i.e. they direted the identifiation of visual themes (Kress & Van Leeuwen, 2001).

110 Table 5.2: Detailed desription of Ryan and Bernard s (2003) srutiny tehniques identifiation of photo themes. Srutiny tehnique Explanation + example English annual report 1) Repetitions Reurrently visualized objets referring to the same onept. Example: pitures of book + produt name + logo subsidiary, website + produt name, searh system in use (plaed with lient testimony) for the orporate produt theme. 2) Indigenous Visualized objets speifi to a ultural domain. typologies Example: pitures of man/woman at work, lothing with ompany logo, man at onstrution site, woman in front of omputer for the 3) Metaphors and analogies employee theme in British CEO s statements. Ambiguous visual expressions referring to the same onept. Example: the theme identity of the group ompany is metaphorially visualized as a globe in the orporate profile, thus signaling the international orporate harater. 4) Transitions Visual markers (i.e. visualized objets) indiating a shift in themes. Example: within the orporate profile desription, a piture ontaining a smiling woman in formal dress, sitting at a desk fouses on a theme that differs from the theme refleted by an adjaent piture ontaining the front of a building and the ompany logo on the building. Here, there is a shift from the employee theme to the store theme. 5) Similarities and differenes in oneptual meaning 7) Missing data referring to implied theme 8) Theory-related onepts Abstrat interpretation of similarities or differenes between two or more visualized objets or photographs that eah omprise a partiular themati onept. Example: the answer to the question how does the refletion of the management team theme in this image vary from the preeding or sueeding images? Conealed information whih is not made expliit in the photographs but whih atually refers to an underlying theme. Example: a hemial ompany s inlusion of images related to the CSR theme, but its exlusion of images related to problemati waste disposals during the year. Referenes to relevant photo themes observed in previous studies and relevant reasons for any distintions between theory and the urrent material. Example: drawing on photo themes found by Preston and Young (2000) in annual reports of multinational orporations. For instane, their theme globalization with key visualizations of globes was also relevant for the present study. 5.2 Analyzing themes in Duth-English and British annual report texts Assuming that eah text refers expliitly to its typial subjet, thereby reating a pattern of referene that reveals expliitly the text s professional purpose (Devitt, 1991: 343), the following subsetions will fous on the analysis of the typial genre ontent in multimodal Duth-English and British-English annual reports. To date, the information ontents inluded in Duth-based and UK-based reports in partiular have only been investigated inidentally. For instane, Santema et al. (2005) analyze the degree in whih annual reports from the Netherlands (and from other European ountries) ommuniate about the orporate strategy theme, whih is defined by key onepts suh as business profile, qualitative and quantitative targets and divisional

111 strategies. And in his historial study of annual reports by UK fashion retailer Burton pl, MKinstry (1996) notes that photographs ontain different ontents (e.g. fashion models, typial ustomers, fany stores) depending on the soio-eonomi situation of the ompany. Studies suh as these suggest that a fundamental overview of prototypial ontent elements in international annual reports of Duth and British orporations remains to be established. Therefore, the ontent study in this hapter will revolve around the desription and omparison of these prototypial elements in various Duth-English and British annual report texts (the sampling of texts was already aounted for in setion 3.3). As was pointed out in setion 5.1, genre ontent will here be defined in terms of oneptual themes ommonly inorporated in written texts and photographs. In general, the present ontent analysis indutively derived oneptual themes from running written texts, photographs and text omments adjaent to the photographs. As an exploratory researh approah was taken, moreover, the themes were initially registered by hand. An exploratory, non-automated registration of themes usually allows for a detailed and dynami examination of thematially related onepts, providing the investigator with a base voabulary that an be expanded as additional texts are read (Carley & Palmquist, 1993: 610). Within the urrent study, full readings of the researh texts enabled the identifiation of reurrent and new onepts that ould either be related to an existing theme (label) or that ould ontribute to the emergene of a new theme (label). Both textual and visual themes were identified and olleted on the basis of Ryan and Bernard s (2003) srutiny tehniques, after whih they were inorporated in SPSS files (see setion 3.4 for more details on data proessing). Subsetions and will suessively desribe the analytial steps taken to identify the textual and photographi themes in the Duth-English and British annual report texts. Subsetion will provide an overview of all the themes that were diserned in the written texts and in the doumentary photographs of the different annual report setions. And subsetion will present the results of the interoder test onduted to establish the reliability of these themes and their identifiation Design for the identifiation of text themes Drawing on theorizations about onept-based themes, the themati ontent of the present Duth-English and British researh materials was mapped by grouping together the meaning of several key words or word strings into a higher-order themati onept (e.g. Carley & Palmquist, 1992). The text theme future performane, for example, ould be defined through the words or phrases: prospets, (out)look at, we expet, forward, 2004, reovery, growth, ontinue, Company X will be investing. Keywords oneptually onstituting the future theme all referred to orporate objetives and/or speulations for the short- to mid-term. The key-phrases and keywords in italis in Figure 5.3 show how this future performane theme ould be diserned in the atual researh texts: the ombination of referenes to the next finanial year (in 2004), modal verbs (will), referenes to future happenings (growth) and expetations for the future (we expet) reveals the existene of this theme.

112 Looking at 2004 and espeially at our home markets, the United States and Brazil will experiene healthy eonomi growth while Europe in general, and the Netherlands in partiular, will expand more moderately. We expet our operating result and net profit to grow further in 2004, albeit more slowly. Figure 5.3: Example theme-in-text future performane theme (ABN AMRO annual report, 2003: 5). Keywords were not neessarily restrited to a single theme and ould be interhangeable aross several themati shemata. The keyword reovery, for instane, ould also be asribed to the business ontext theme, whih in addition omprised words and phrases suh as during, seond half of the year, eonomi environment, the developed world. The possibility to interhange keywords oneptualizing different themes depended on the textual ontext in whih these words ourred. While the keyword reovery in the future performane theme was surrounded by words referring to the future and thus implied eonomi or orporate reovery-in-the-future, the keyword reovery in the business ontext theme was surrounded by words referring to the reporting year and therefore suggested eonomi or orporate reovery-in-the-past. Table 5.3 displays five additional examples of onept-based text themes that were registered during the analysis. A list of all text themes enountered in the Duth-English and British management statements, profile desriptions and operational reviews is presented in setion Table 5.3: Examples of text themes and their onstitutive onepts in the written texts. Text theme Shareholder onfidene Marketing HRM/People General finanial results Mission statement Constitutive onepts (keywords/phrases) a The path to reovery, I am onfident that, mindful of our responsibilities, well-positioned, progress, I have no doubt, potential. Brand strategy, positioning, we operate aross 26 markets, expanding our market boundaries, marketplae, market-driven, distribution hannels. Attrating, employees, motivating, developing, I thank management and employees, We have asked our people to dig deep, they have delivered, efforts, dediation, ommitment to the human resoures funtion. Finanial results, profits, losses, in summary, results, finanial performane, results summary, operating profit, group profit. To be a leading provider of, mission, vision, in all that ompany X does, it strives to, our aim is. Note. a The lists of onstitutive elements are not exhaustive, but give insight in words and key phrases that ould typially be related to a partiular text theme Design for the identifiation of photo themes In the present study, visual themes were part of doumentary photographs that were intended to illustrate or emphasize the ompany s identity and its ativities (see setion 4.4.6). The analysis of photographi themes was based on the multimodal disourse approah of Kress and Van Leeuwen (2001), whih means that the fous was on visual modes or visual resoures of expression and the themati relationship between the visual signs or onepts embedded in these modes. More speifially,

113 the analysis of photographi themes primarily omprised the assoiation between objets, ativities and senes that were displayed in the image and, if relevant, the relationship between these visual items and the surrounding written texts. Thus, the establishment of photographi ontent involved the visual objet-as-mode, ativityas-mode and sene-as-mode on the hand and the writing-as-mode on the other hand (see setion 3.2 on multimodal disourse). An example of the identifiation of onept-based photo themes an be drawn from a 2003 annual report published by a supermarket hain. Here, a piture showing isles with shelves of food produts, shopping arts and adult females in asual lothes ould be ategorized under the theme servie area. Figure 5.4 projets an image for whih this theme was registered. Figure 5.4: Example theme-in-photograph servie area theme (Somerfield annual report, 2003:11). Similar to the text themes, photographi themes were definable by elements that ould also be onstitutive of other photographi themes. A photograph zooming in on shelves of food produts, for instane, ould also ontribute to the visual produt theme. Table 5.4 provides four additional examples of photographi themes that were diserned in the Duth-English and British researh texts. A full list of all photographi themes in the managerial forewords, orporate profiles and operational reviews is inluded in setion Table 5.4: Examples of photographi themes and their onstitutive onepts in the doumentary images. Photographi theme Work area CEO Employee (male) Client (female) Constitutive elements (signs in objets-as-mode/written languageas-mode) a Steel onstrution for building, steel beams, muddy ground, trator ontainer, metal fene, small projetion adult male, yellow work oat, helmet. Adult male, desk + paperwork + eletroni planner, suit + neat shirt + tie, headline above piture: From the CEO. Head + part of shoulders adult male, neat shirt, phone in hand, headline above piture: X s enhaned prospets as a foused Speialist Reruitment Group. Adult Asian female, woman s suit, smiling, holding produt of ompany X, text next to piture: Name: Dr. Ying, Industry: Healthare, Customer sine: Note. a The lists of onstitutive elements are not exhaustive, but give insight in pitorial signs that ould typially be related to a partiular photographi theme.

114 5.2.3 List of text themes and photo themes in annual reports Based on preliminary observations of the ontents of the written texts and the photographs in the Duth-English and British-English annual reports, two lists of themes were omposed that would represent the textual ontents and the visual ontents of all sampled researh materials. During the interoder reliability test (to be disussed in setion 5.2.4), these lists of themes were verified and adapted to inrease their relevane for the multimodal annual report texts. The final versions of the lists omprised a total of 189 themes for the written texts and a total of 82 themes for the pitures in the management statements, orporate profiles and operational reviews (see Tables 5.5 and 5.6, respetively). As was suggested in the desription of srutiny tehniques (subsetions and 5.1.2), the textual themes and photo themes were established by referene to both prior analyses on annual reports and the urrent researh texts. Studies that helped to establish the textual themes in the Duth-English and British annual reports inluded Osborne, Stubbart and Ramaprasad (2001), Santema and Van de Rijt (2001) and Smith and Taffler (2000). Moreover, studies that ontributed to the identifiation of themes in doumentary images were MKinstry (1996), Preston and Young (2000) and Preston et al. (1996). Sine eah of these studies had speifi researh foi, they ould not fully over the updated ontents of the present Duth- English and British annual report texts. Therefore, textual themes suh as ost redution/management or IFRS and photo themes suh as logistis or ommuniations/promotions needed to be introdued to be able to analyze the full informational landsape of these texts. Furthermore, it should be noted that the lists of textual and visual themes in Tables 5.5 and 5.6 are not bound to any speifi text type in the annual report. A disrete distintion between themes per annual report setion was diffiult, beause the researh data pointed out that several themes ourred more or less often in multiple setions. The textual theme orporate strategy, for instane, was traed in a majority of the management statements and orporate profiles. And the visual theme workplae was relatively prominent in both the profiles and operational reviews. Finally, it must be added that the text themes and photo themes in the seleted annual report texts did not reour in a partiular order of appearane. Sine they ourred in a random position in different instanes of similar or various text types, they are listed here in alphabetial order.

115 Table 5.5: Themes identified for written texts in Duth-English and British-English annual reports. Text themes apoli ommitbo atboard ommitsu ativfou ommrisk atsuper ommunit adshare ompat advie omparis afterdat omplod agm onflit appresul onssegm artasso onssupp attendan onsult auditors ontat awrank ontrat baklog oopnetw balane orpgov basisaud ostred boarddet ountris busimpli sr apital urreny ases urrrisk ashflow debt hannel debtrisk laims direxter Colophon dirrespo dividend donation earnshar eduatio environm estimati evaluati exepite faility finalen finhighl fininstr finpoli finresul finsegm fintarge forwlook founddet founddr foundpri foundsha funstaf futsegm future geomarke glossary goodwill govode heasafe history hrm/peop identity idmarket idsegm ifrs inentiv inome independ index indexwor indusseg Indus/ont inflatio instrris insuran insurris intontr intelpro interest interris invdiv invrelat itrdsegm itresdev legalris liqrisk liquidit loans markdist markrisk marksegm marposit matontr membfoun memboard memsuper mission monunion nonopite numbempl obligati Offbalan opompan opost operrisk ophighli opinaudi opresult opsegm ordright partner pensbene pension perfgrap positseg press prinipl prodmat proposal purhase ref22f refurr refusa registra regulati relpartt remboard remhang remom remnonex rempak rempoli remsuper remunera repaudit repnomin reportin respdiau riskom riskfat riskman roi roleboar rolesupe sales shaanaly shaapit shafund shaindex shainter shaoptio shaprie shareon sharehol sharelis shareval shaservi sperigh strategy stratseg strusha strutur supplpay supwork taxation tradrisk training treasury usp workond workoun Table 5.6: Themes identified for photographs in Duth-English and British-English annual reports. Photo themes atommu awarrank atedu balane atenvir building addvalpr buspeopl air alulat allboard eo allexebo eremony allsupbo hairman animal hild applprod lothes Awardwin ompass ompromo srtool urtain endprodu equipm famlien flhmem flfempl flmempl female femhil femlien femempl fieldpro globe grdptboa grolien grouempl hand indivboa indptboa justie light liquor loommu loenvir logisti mafeempl mainboar malhil mallien male malefema malempl manufpro mlifemp mlimemp money pailfem pailien pailmem peoommu peoedu photo plant produt rawmater resommu resedu resenvir resobje serviare servie stairs store strsign teens toys water workarea Interoder reliability test for text themes and photographi themes An interoder reliability test was arried out to assess the reproduibility of the themati oding proedures used in the Duth-English and British annual report texts. Two oders partiipated in this test, i.e. the author of the present study and a graduate student in Business Communiation studies with a near-native profiieny in English. The two oders analyzed the themati ontent of both the written disourse and the photographs in the texts randomly sampled for the reliability test. A more detailed desription of the test proess was already provided in setion 3.5. Table 5.7 gives an overview of the test results, inluding range of Kappa values, standard deviations (SDs) and interoder agreement perentages.

116 Table 5.7: Evaluation of interoder agreement for the text themes and photographi themes. Setion/ text type Qualifiation Mean SD Min. Agreement Kappa max. % Management Substantial statements/text Management Almost perfet statements/photo Profiles/ TEXT Almost perfet Profiles/PHOTO a Operational Moderate reviews/text Operational reviews/photo Almost perfet 0.00 min./max.= Note. a Sine the sample of profile desriptions for the interoder reliability test did not ontain any pitures, Kappa ould not be alulated for the registration of photographi themes in these texts. Aording to the rule of thumb set for Cohen s Kappa, qualifiations for interoder reliability ranged from moderate (.59) to almost perfet () with respet to the textual and photographi themes inluded in the statements, profiles and reviews (Rietveld & Van Hout, 1993). Consequently, the applied oding system for the exploratory identifiation of textual and photographi themes in Duth-English and British annual reports an be regarded as a reliable instrument. 5.3 Themes in management statements The following subsetions report on the textual and photo themes that were diserned in the Duth-English and British management statements. The results for the text themes will be disussed in setion The results for the photo-related themes, inluding those embedded in the text fragments adjoining the images, will be presented in setion Themes in the running texts of the management statements The full list of text themes diserned in the Duth-English and British CEO s statements and in the British Chairman s statements is inluded in Appendix I. It indiates that these statements are thematially omplex douments whih over a range of orporate issues, suh as finane, operations, orporate strategy, marketing, human resoure management (HRM), Corporate Soial Responsibility (CSR), share management, et. A total of 15 themes appeared to be typial of the realization of the ontents in the management statements, i.e. they ourred in more than 50% of one, two or all three statement types. Table 5.8 shows the frequeny ounts for the most popular themes, together with any signifiant ross-ultural differenes in themati ontent.

117 Table 5.8: Typial text themes ourring in 50% of the management statements. Theme Du-CEO Br-CEO Br-Chair n=20 n=18 n=16 Count % Count % Count % Total χ 2 (df p % = 2) Board details a b <.001 Confidene CSR 2 a b <.05 Corp.govern. 11 b a <.05 Corp.strategy b a <.001 Corp.industry Dividend 4 a a b <.001 Finan.results Future HRM/People 20 b a <.05 Invest.divest Marketing Oper.results Op.result unit Position Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. The ross-ultural omparison of themes first of all showed statistially signifiant differenes between the British CEO s statements and the British Chairman s statements, in partiular with respet to the orporate strategy and the board details themes. Whereas the British CEO s statements inluded signifiantly more details on the orporate strategy theme, the British Chairman s statements omprised signifiantly more instanes of the board details theme. Seondly, the analysis of text themes indiated signifiant variations between the Duth-English and the British CEO s statements, where the former revealed signifiantly higher frequeny sores for both the HRM/people theme and the orporate governane theme. A third signifiant differene was observed between the Duth-English CEO s statements and the British Chairman s statements: the British Chairman s statements dislosed signifiantly more information on the CSR theme than the Duth CEO s statements. Finally, the orpus analysis suggested that the three types of statements differ signifiantly with regard to the dividend theme. Compared to the Duth CEO s statements and the British CEO s statements, the British Chairman s statements inluded signifiantly more details on this theme Themes in photos and text fragments of the management statements Table 5.9 inludes all photo themes observed in the multimodal management statements. It shows that the pitorial ontents of Duth-English and British CEO s statements and British Chairman s statements entered around speifi ompanyrelated items: management members, workplae and employees. As a relatively large number of the speifi themes registered for the managerial forewords were

118 related to these three items, they were lustered for the desription of themati ontent. Clustering avoided the fragmentation of results and allowed for a more omprehensive and meaningful explanation of photographi themes. Themes that had an ambiguous meaning and would ause ategorial mixing, e.g. referring to employees + lients, and themes that were single representatives of a ertain ategory, were plaed in the luster other themes. Table 5.9 also highlights the signifiant differenes in these themes aross Duth-English and British statements. Furthermore, Table 5.10 displays the frequeny sores per themati luster in order to determine whether there were any signifiant differenes aross statements for groups of oherent themes. Clustered results for the photo themes are presented in a separate table, beause some individual annual reports ontained several speifi photo themes that were part of the same themati luster. For the doument frequeny ounts used in the present investigation (see setion 3.4), simply adding the ounts for the speifi photo themes would thus lead to inaurate ourrene alulations. Table 5.9: Frequeny sores photo themes in management statements. Theme Du-CEO Br-CEO Br-Chair n=17 n=16 n=14 Count % Count % Count % Total % Members of management All ex. Board 6 b χ 2 (df = 2) <.05 CEO b a <.001 Chairman 0 a a b <.001 Indiv.exe.dir Main board dir Employees Female employee b <.05 Group employees Male employee 0 a b <.05 Workplae Building Equipment Logistis Servie area Store Work area Other themes Communiation Female Fmlient+fmempl Fmlient+mlempl Male Mllient+fmempl Male+female Pair of lients Produt Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. p

119 A omparison aross statements indiated that the Duth-English CEO s statements ontained a signifiantly high sore for photographs displaying all exeutive diretors in the ompany. Next, pitures of female employees were signifiantly overrepresented in the British CEO s statements. Table 5.9 also shows that the British CEO s statements and the British Chairman s statements vary signifiantly with respet to the CEO theme of inluded pitures; not surprisingly, pitures of the CEO ourred exlusively in the former type of statement. Another statistially signifiant differene in photographi themes was observed between the Duth- English and the British CEO s statements, where photographs of male employees were not inorporated in any Duth-English CEO s statement but appeared in a signifiantly higher number of British CEO s statements. A final signifiant distintion onerning the Chairman theme ould be observed aross all three statement types. Photographs portraying the non-exeutive Chairman, quite logially, only appeared within the text boundaries of the British Chairman s statements. It must be noted that the signifiant results for the speifi themes (as opposed to the lustered themes) in management-related photographs are of partiular importane from the perspetive of genre identifiation. Sine the individual and grouped objets portrayed in these photos refer diretly to the sender(s) of the message, they are highly speifi to the management statements in whih they are inorporated. Hene, a mere generalization of these speifi themes into the lustered theme members of management would not do justie to the visual ontents of the management statements. Table 5.10: Frequeny sores lustered photo themes in management statements. Theme Du-CEO Br-CEO Br-Chair n=17 n=16 n=14 Count % Count % Count % Total χ 2 (df = % 2) Members of management Employees 0 a b Workplae b p <.05 <.05 Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. Fisher s Exat tests for the remaining lustered themes showed two additional ross-ultural differenes in photographi ontent (see Table 5.10). First, there was a statistially signifiant distintion between the Duth-English CEO s statements and the British CEO s statements for the ategory representing employee-related themes. Here, the set of British CEO s statements inluded signifiantly more texts portraying the ompany s employees. Seond, the lustered themes revealed a statistially signifiant result for the British CEO s statements only: these ontained a signifiant overrepresentation of texts referring to the workplae theme. In most management statements that ontained visual illustrations, the photographs were supported by textual explanations of the visual ontent. Sine these photographs predominantly portrayed the (non-)exeutive sender(s) of the message, the adjaent text fragments typially desribed the identity of the manager sending

120 the message, i.e. in terms of name and/or professional funtion. Table 5.11 displays all text themes diserned in the photo-related text omments. Table 5.11: Frequeny sores text themes in photo-related omments in management statements. Theme Du-CEO Br-CEO Br-Chair n=12 n=13 n=9 Co % Count % Count % Total unt % Confidene Consumers Consumers unit Corp.strategy Corp.strengths Faility Future HRM/People Identity market Identity unit Industry Manager unit Marketing Marketing unit Oper.highlights Oper.results Oper.result unit Produt Prodution Note.. Fisher s Exat Test. χ 2 (df = 2) p Fisher s Exat tests indiated that there was no statistially signifiant differenes in the distribution of text themes in photo-related exerpts within multimodal management statements produed in the Netherlands and in the UK. 5.4 Themes in orporate profiles While setion disusses the typial themes observed in the running texts of the Duth-English and British profiles, setion presents the oneptual themes in the doumentary images and in the text fragments supporting these images Themes in the running texts of the orporate profiles Appendix II ontains the set of text themes that was diserned in the Duth-English and British-English orporate profiles. It reveals that the orporate profiles were thematially speified. Profile desriptions primarily foused on text themes that refer to the ompany s identity and the ommerial environment it operates in, e.g. ore values, produts/servies, markets, strategi fous points and orporate strengths. Although some orporate profiles inluded information about general operational results of the group ompany, themes referring to orporate performane

121 were learly of minor importane in these texts. As is displayed in Table 5.12, only five themes appeared in more than 50% of the Duth-English and/or British profiles, i.e. they were most typial to the Duth-English and/or British texts. The table shows the frequeny of ourrene for these themes and displays any signifiant differenes between main themes aross the profiles. Table 5.12: Typial text themes ourring in 50% of the orporate profiles. Theme Du-Profile Br-Profile n=21 n=20 Count % Count % Total % χ 2 (df =1) p Corp.strategy Corp.strengths Identity Identity unit Identity market Mission Note.. Fisher s Exat Test. The analysis of text themes in the orporate profiles indiated that the Duth-English and British-English texts predominantly ommuniate the same information: Chi- Square tests and Fisher s Exat tests did not exhibit any statistially signifiant differenes between the main text themes aross the profile desriptions Themes in photos and text fragments of the orporate profiles The ontent analysis of photos in the orporate profiles showed that the Duth- English and British-English texts were primarily illustrated by pitures onerning the workplae, the employees, the ompany s ustomers and its produts. Sine most of the photographs that were observed ould be distributed among the four broader visual ategories mentioned in setion 5.3.2, they were again lustered to avoid fragmentation of interrelated visual themes. To aount for different frequeny alulations, moreover, Tables 5.13 and 5.14 separately present the sores for speifi themes and the frequeny sores by luster. Signifiant ross-ultural distintions are also inorporated in these tables. Ambiguous photographs that were oneptually related to multiple themes (e.g. female employee + female lients) or single themes were inluded in the other themes ategory.

122 Table 5.13: Frequeny sores photo themes in orporate profiles. Theme Du-Profile Br-Profile n=8 n=10 Count % Count % Total % χ 2 (df = 1) p Customers Child Group of lients Male lient Pair of lients Employees Female employee Group employees Male employee 0 a b <.05 Workplae Building Equipment Servie area Work area Produt Added value produt End produt Produt Raw materials Other themes Balane Ceremony Communiations Female Globe Hand Male Mal.lient+mal.empl Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. For the speifi photo themes in the profiles, Fisher s Exat test only exhibited a statistially signifiant distintion aross ultures for pitures refleting the male employee theme: these ourred in a signifiantly larger number of the British profile desriptions than the Duth-English profile desriptions. In line with this finding, Fisher s Exat test revealed a statistially signifiant result for the general themati luster labeled employees. Visual themes related to the workfore (either female or male) ourred signifiantly more often in the British than in the Duth- English profiles. This is illustrated in Table 5.14.

123 Table 5.14: Frequeny sores lustered photo themes in orporate profiles. Theme Du-Profile Br-Profile n=8 n=10 Count % Count % Total % χ 2 (df =1) p Customers Employees 0 a b <.05 Workplae Produt Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. With regard to the text fragments adjoining the photographs in the Duth-English and British profiles, no typial themes were diserned. The theme with the highest sore in both business ultures was the Produt/servie desription theme, whih ourred in 33.3% of the Duth-English texts and in 37.5% of the British-English texts. This text theme usually aompanied images of produts/servies in pratie or on display. As is indiated by Table 5.15, no signifiant ross-ultural differenes were found for the themes in the text fragments explaining the images. Table 5.15: Frequeny sores text themes in photo-related omments in orporate profiles. Theme Du-Profile Br-Profile n=3 n=8 Count % Count % Total % χ 2 (df = 1) p Communiations Consumers unit Corp.strengths Finan.highlights Identity Id.market unit Identity unit Marketing Oper.highlights Oper.results Oper.results unit Produt Note.. Fisher s Exat Test. 5.5 Themes in operational reviews This setion will first desribe and ompare the themes found in the running texts of the Duth-English and British-English operational reviews (subsetion 5.5.1). It will then ontinue with a disussion of the themes refleted in the images and inorporated in the text exerpts adjaent to these images (subsetion 5.5.2).

124 5.5.1 Themes in the running texts of the operational reviews As is indiated in Appendix III, the extensive operational reviews published by Duth and British orporations inluded a number of topis similar to the management statements. However, the full list of text themes for the reviews also shows that their informational value was less variable and more speifi, as they predominantly dislosed issues related to the past and future operations of orporate divisions on ommerial markets. Although text themes overing general finanial results by division were also enountered relatively often, these themes ommonly funtioned as mere starting-points for the disussion of themes about operational performane. Table 5.16 presents frequeny sores for the 16 main text themes within the Duth-English and British-English operational reviews and shows any statistially signifiant differenes in themati ontent aross the two national business ultures. Table 5.16: Typial text themes ourring in 50% of the operational reviews. Theme Du-OperRev Br-OperRev n=22 n=20 Count % Count % Total % χ 2 (df = 1) p Award/Ranking Consumer unit Context unit Corp. strengths Cost managem. 11 b a <.05 Fin. highlights Fin.result unit Future division Identity market 14 b a <.05 Identity division Invest.divest Marketing unit Op.result unit Position unit Sales Strategy unit Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. First, the Chi-Square test results showed that Duth ompanies tended to inlude the textual ost management theme signifiantly more often than British ompanies. And seond, it appeared that a signifiantly larger number of Duth ompanies, as ompared to British ompanies, had inorporated the identity of the market theme in their texts Themes in the photos and text fragments of the operational reviews Compared to the management statements and the profile desriptions, the Duth- English and British operational reviews ontained many more visual illustrations

125 and, aordingly, a larger total of photographi themes. On the one hand, these pitures refleted the internal orporate or divisional identity by means of photographs related to management members, employees and the workplae; on the other hand, they portrayed what the ompany offers people outside the organization by means of pitures of produts/servies and lients. Table 5.17 offers a rossultural overview of the speifi, unambiguous themes that ourred most frequently in the photos of the Duth-English and British reviews. Table 5.18 again gives a separate aount of the frequeny sores per luster of themes that were semantially related. Table 5.17: Frequeny sores photo themes in operational reviews. Theme Du-OperRev Br-OperRev n=15 n=17 Count % Count % Total % χ 2 (df = 1) p Customers Child Family-lients Female lient Fem.lient + hild Male lient Male lient + hild Pair of lients Employees Female employee Group employees Male employee Mal.empl.+fem.empl Workplae Building Equipment Servie area Store Work area Produt Added value produt Appliation produt End produt Produt Prod.manufaturing Raw materials Servie Members of management CEO Indiv.dpt. diretors Group dept. diretors

126 Table 5.17 ontinued. Theme Du-OperRev Br-OperRev n=15 n=17 Count % Count % Total % χ 2 (df = 1) p Other themes Awards/ranking Ativity ommunity Ceremony Communiations Female Hand Light Male Male + female Mal.lient+fem.empl Mal.lient+mal.empl Pair lient+mal.empl People in ommunity Note.. Fisher s Exat Test. Tables 5.17 and 5.18 indiate that Chi-Square tests and Fisher s Exat tests did not reveal any signifiant ross-ultural differenes in the themati onstrut of photographs within Duth-English and British operational reviews. Table 5.18: Frequeny sores lustered photo themes in operational reviews. Theme Du-OperRev Br-OperRev n=15 n=17 Count % Count % Total % χ 2 (df = 1) p Customers Employees Workplae Produt Members of management Note.. Fisher s Exat Test..71 The themati analysis of text exerpts plaed with the doumentary photographs in the reviews showed a reurring use of the produt/servie theme in both the Duth- English and British texts. Considering that a majority of the English operational reviews were strutured in aordane with the divisional organization of the ompany, many orporations appeared to use images and adjaent texts to speify the produts/servies manufatured or delivered by a partiular division (see Table 5.19). Other text themes, e.g. related to desriptions of divisional performane, also re-ourred relatively often in the text fragments.

127 Table 5.19: Frequeny sores text themes in photo-related omments in operational reviews. Theme Du-OperRev Br-OperRev n=13 n=15 Count % Count % Total % χ 2 (df = 1) p Business ontext Bus.ontext unit Communiation Confidene Consumers Consumers unit Corp.strategy Corp.strengths CSR Faility Finanial results Finan.results unit Future Future unit HRM/People Identity market Id.market unit Identity unit Marketing Member board Op.highlight unit Operat.results Oper.result unit Produt Shareholders Strategy segm Note.. Fisher s Exat Test. A ross-ultural omparison based on Chi-Square tests and Fisher s Exat tests revealed that there was no statistially signifiant variation in the text themes Duth and British orporations used for the texts illustrating the images in their English reviews. 5.6 Conlusion and disussion The ontent analyses desribed in this hapter aimed to explore the ross-ultural use of oneptual themes in multimodal annual reports originating in the Netherlands and in the UK. As suh, they were intended to provide a further response to the primary researh objetive of this study (see setion 1.5). In partiular, these analyses aimed to address the speifi RQs 2a-2 by identifying and omparing the ontent features ourring within the written texts and doumentary images of the Duth-English and British-English researh materials. It an be onluded that Duth and British orporations largely relied on similar themes in the onstrut of their English annual report texts. For eah type of text analyzed, moreover, they used a series of themes that differed from the sets of themes in other

128 annual report setions. Cross-ultural variations in genre ontent primarily ourred for the management statements and for the themes inorporated in the written running texts. The emergene of substantive differenes aross Duth-English and British annual report texts underpins the results of studies suh as Sims and Guie (1992), whih suggests that text produers in different ultures may have different pereptions of the appropriate or effetive ontent of a genre. Setion will briefly reount the main ross-ultural findings for the onept-based themes, and it will provide ontextual explanations for these findings if possible (see RQ 2d, setion 1.5). It will also elaborate on the degree to whih the Duth-English and British annual report texts orrespond or differ in genre ontent. In line with the seond general researh aim of this study, setion will put forward several impliations that may ontribute to the refinement of genre theory and genre analysis in general Cross-ultural omparison of ontent elements Assuming Devitt s laim that eah text refers expliitly to its typial subjet, thereby reating a pattern of referene (1991: 343), the themati analyses of texts and photographs in annual reports proved to be relevant in distinguishing ulturespeifi preferenes in the substantive onstrut of this genre. The themati similarities and distintions (RQs 2a-2) observed aross the Duth and British business ultures are summarized and ontextualized (RQ 2d) below. Also, the omparability of Duth-English and British annual report texts will be disussed at the end of this subsetion (i.e. a fous on RQ 2). Comparison of themati ontent in management statements The analysis of genre ontent revealed many similarities in the text themes aross the Duth-English CEO s statements, the British CEO s statements and the British Chairman s statements. The three types of statements primarily reported on grouplevel issues: the main text themes in all statements involved orporate performane as well as orporate poliies and were foused on assuring readers of the future wellbeing of the ompany. With regard to the photo themes, the researh results showed no signifiant ross-ultural differenes for pitures referring to orporate failities and orporate equipment. Also, the Duth-English and British management statements showed similar use of the speifi photo themes related to groups of employees and to individual members of the board. Finally, both the Duth-English and British management statements typially relied on the theme related to the identifiation of board members within the text fragments adjoining the photographs. Despite these substantive similarities, several signifiant distintions were also observed for the text themes and photo themes in the Duth-English and British management statements. Cross-ultural variations in text ontent were observed for themes onerning the ompany s ondut in terms of its soial responsibility, HRM poliy, governane poliy, management hanges, strategi priorities and dividend. For instane, the number of Duth-English CEO s statements fousing on the textual CSR theme was signifiantly smaller than the number of British Chairman s statements onsidering this theme. Furthermore, the textual dividend theme omprising aounts of dividend, dividend poliies or payments was inluded in a signifiantly larger number of the British Chairman s statements than the Duth- English and British CEO s statements. Several signifiant results were also

129 doumented for the photo ontent of these statements, in partiular with respet to photographs portraying orporate employees, full management teams and orporate workplaes. Pitures portraying all members of the exeutive board, for example, were signifiantly over-represented in the Duth-English CEO s statements. Here, the sender of the message was portrayed as part of a olletive responsible for the ompany s daily management. Also, a relatively large number of British CEO s statements ontained visual themes related to diverse aspets of the orporate work area. There were no signifiant ross-ultural differenes with regard to typial themes in the text fragments adjaent to the doumentary images. The general onlusion that an be drawn from these results is that the Duth-English and British-English management statements are not idential as far as genre ontent is onerned. In spite of some overlap in verbal-pitorial ontent, the signifiant variations in both text themes and photo themes suggest that (1) the Duth-English and the British managerial forewords differ generially and that (2) the Duth-English CEO s statement is a hybrid doument that is generially similar to both the British CEO s statement and the British Chairman s statement (whih are themselves not generi equivalents). As for the distintions in textual ontent, muh of the differene observed seems to stem from the fat that the supposed writers of the texts, the CEOs and Chairmen, fulfill distint funtions within Duth and British organizations and from the fat that their statements have distint positions and purposes within the annual report. British Chairmen have a supervisory funtion and represent the needs of shareholders, while Duth and British CEOs exeute the ompany s daily management and thus represent orporate interests first of all. But whereas the British CEO s statement appears after the Chairman s statement in British reports and is more narrative in nature, both the British Chairman s statement and the Duth-English CEO s statement are introdutory, interational douments in the annual report (see 4.4.4; Hooghiemstra & Van Manen, 2004). As an illustration, the professional mandate of the British Chairman may aount for the fat that a onsiderably larger number of Chairman s statements than Duth-English and British CEO s statements ontained the textual dividend theme. With respet to the signifiant distintions in photographi substane, it appears that the Duth-English CEO s statements and British Chairman s statements primarily ontained images of the sender of the message; the British CEO s statements inluded both senderrelated photographs and ompany-related photographs (i.e. showing employees or workplae). These distintions may have been aused by the fat that the British CEO s statement is generally not an introdutory text and tends to be more narrative and often more extensive; it provides a relatively large amount of textual information that an be illustrated by different types of doumentary photographs (e.g. pitures of employees at work, during one of the operational ativities desribed in the text). Comparison of themati ontent in orporate profiles An analysis of similarities in the text ontent of the Duth-English and British- English orporate profiles revealed that these two text types share a themati fous on issues overing the general orporate identity: text themes primarily onerned the ore viewpoints and ompetenies of the ompany responsible for the annual report. Sine no signifiant variation was observed for the speifi text themes that ourred in 50% or more of the Duth-English or British profiles, it an be onluded that Duth-English and British orporations relied on equivalent textual ontents for the realization of these douments. Cross-ultural orrespondene was also found in the themes onstituting the photo ontent and the ontent of the photo-

130 related text fragments in the profiles. No substantial differenes were observed, for instane, in the number of Duth-English and British profile desriptions ontaining pitures of orporate lients, produts or workplaes. Thus, orporations in both business ultures showed little variation in text themes and photo themes inorporated to introdue the ompany to its audiene. Signifiant distintions between the Duth-English and British orporate profile desriptions only ourred for two interrelated photo themes. First, the speifi male employee theme was observed in a signifiantly larger number of British profiles than Duth-English profiles. Congruously, the general employee theme also appeared in images inluded in signifiantly more British than Duth- English profiles. Only onsidering the text themes inorporated in the substane of Duth- English and British-English profile desriptions might wrongly imply that these douments are generially idential. Although they are predominantly the same in terms of textual ontent, the analysis of themes inorporated in photographs revealed that they annot be regarded as real equivalents; the analysis emphasized that Duth and British orporations tended to reate a slightly different informational landsape when photographially identifying themselves in the English profile desriptions. Duth ompanies mostly inluded photographs of produt-related, workplae-related and ustomer-related items to support the orporate identity explained in the text. None of these ompanies portrayed its employees for this purpose. Conversely, the British orporations did not inorporate any ustomer-related photos but a substantially large number of them did inlude pitures of the workfore. Obviously, the British ompanies, in ontrast to the Duth ompanies, preferred to portray the people who are responsible for the daily operations of the organization as a means of profiling themselves. Comparison of themati ontent in operational reviews The investigation of substane in Duth-English and British-English operational reviews suggested that these douments ontained a large number of themati similarities in the written text and photographs. The text ontent of operational reviews published by Duth and British orporations predominantly addressed themes dealing with divisional ahievements, i.e. desriptions of diverse divisions, finanial and operational results per division and issues that had impated divisional operations. Also, it appeared that Duth and British orporations used the same broad range of typial photo themes to omplement the information provided in the written texts of their operational reviews. In both national business ultures, a relatively similar number of reviews ontained pitures of different types of ustomers and employees, of workplae-related objets, of produt-related items and of exeutive diretors. With regard to the themes in the photo-related text fragments, Duth and British orporations predominantly relied on the same typial subjet matter, i.e. the produts theme. As for the signifiant distintions in genre ontext aross the Duth- English and British-English operational reviews, there were only two text themes that highlighted a signifiant result. The Duth-English reviews and the British reviews showed signifiant ross-ultural differene for the ost management theme and the identity of the market theme. In the Duth-English reviews, information about operational osts or ost redutions and details about the market in whih a division operates were signifiantly more important than in the British operational reviews.

131 The Duth-English and British operational reviews had muh in ommon. Signifiant ross-ultural variation was only found for two text themes and was not observed for photo themes at all. While the two signifiant results for the textual substane of the reviews might have been aused by ontemporary soio-eonomi trends or normative ommuniative hoies, these auses did not form the basis of any more fundamental differenes between the Duth-English and British operational reviews. From a generi viewpoint, it an therefore be argued that the operational reviews in the 2003 annual reports of Duth and British ompanies were substantively similar to a very large extent, in terms of both textual and photographi ontent. They were marked by only a small number of ross-ultural distintions that did not seem to affet the generi integrity of the operational review. Genre ontent and the Duth-English and British-English annual reports as a genre set After reviewing the ross-ultural orrespondene and variation in oneptual themes inluded in the multimodal annual report texts, the question remains as to the degree in whih the Duth-English and British-English texts are omparable. In generi terms, omparability refers to similarities in the ontent-related interpretation of annual reports used for the same purposes in different national business ultures (e.g. Miller, 1994). In line with Devitt s observation in tax aounting genres, it was found that eah Duth-English and British annual report text partially derived its meaning from other texts, those that onstitute its subjet matter (1991: 343). Aordingly, the researh results again indiated that the annual report in the two national business ultures omprised a genre set (Devitt, 1991; see also setion 4.5.1). Within the Duth-English and British-English report, eah setion or text type ontained a partiular olletion of primary text themes and photo themes that did not math the olletion of main themes inluded in any other text type in the annual report. For instane, the Duth-English and British operational reviews were marked by a variety of reurrent text themes onerning divisional performane, while the English-language profile desriptions repeatedly foused on a set of main text themes referring to the orporate identity. However, the researh findings simultaneously indiated that some themes reourred within several different Duth-English and British text types. The textual orporate strategy theme, for example, appeared in orporate profiles as well as in operational reviews. And the textual Operational results per division theme was observed in both the management statements and the operational reviews. These findings thus onfirmed the idea that the different types of annual report texts analyzed here i.e. foreword, profile and review were haraterized by ontent-based intertextual oherene. From the perspetive of genre ontent, they would seem to be independent text types, eah with a unique themati onstrut, that o-exist through intertextual ties. Together, these genres build the annual report-as-a-genre set. Having determined that both Duth-English and British annual reports onsist of different text types featured by themati intertextuality, the researh results onfirmed that the Duth-English and British annual reports are omparable examples of the same genre set. The ontent study of the management statements, orporate profiles and operational reviews showed that the English-language annual reports produed in the Duth and British business ultures ontained text types that were largely similar in terms of textual and photographi themes. Consistent with the observations made in setion 4.5.1, it was found that most themati variations

132 existed in the three types of management statements. Here, the Duth-English CEO s statements orresponded thematially with both the British Chairman s statements (e.g. for the textual HRM/people theme or the photographi workplae theme) and the British CEO s statements (e.g. for the textual orporate strategy theme or the photographi hairman theme). At the same time, however, these statement types also revealed an overall agreement in the use of partiular text themes (e.g. General finanial results, Investments and divestments and Future prospets ). Despite some signifiant differenes in text themes and photo themes aross text types, the overall ross-ultural similarities in genre ontent thus suggested that the Duth-English and British annual report setions analyzed in the present study an be typified as omparable text types or genres within omparable instanes of the annual report genre Impliations for genre theory and genre analysis The seond main researh aim of the present study was to ontribute to a further development of the generi approah to multimodal disourse in (professional) texts. In the light of this objetive, the ontent analysis yielded several impliations for genre theory in general and genre analysis in partiular. The theoretial importane of oneptual themes in genre ontent The themati approah taken towards the investigation of genre ontent allowed for a ognitive and situated perspetive on multimodal information ontent. Within the ognitive map-analytial tradition, ontent is established by mental networks of ontextualized linguisti onepts or visual onepts that are semantially related. Here, eah network represents a oneptual theme (Carley & Palmquist, 1992; Ryan & Bernard, 2000). Whereas the running texts and text fragments in the annual report setions provided keywords as oneptual onstituents of textual themes, the pitures inluded in the texts offered visualized objets and events that signaled partiular photographi themes. As suh, the themati investigation of multimodal text ontent seems to be in aordane with fundamental notions of substantive genre identifiation. Drawing on Miller (1994) first of all, an investigation of textual and photographi themes enables researhers to look at the semanti value of disourse, in whih ommon experiene among disourse partiipants is symbolized through language (i.e. keywords) and images (i.e. visualized objets). In addition, ognitive mapping theory assumes that the lassifiation of thematially related onepts is based on shared soial knowledge whih helps readers determine whih onepts are likely to be attributable to whih theme in a partiular part of the text. Therefore, the present analyses are also onsistent with Berkenkotter and Hukin (1995), who argue that ommunity-speifi experiene is transferred to the informational value (i.e. ontent) of a partiular type of text. Moreover, the present study of textual and visual themes supports Nikerson s suggestion (2000) that different ommunities may prefer to express various aspets of ommon experiene (i.e. informational value) in their texts; apart from the ross-ultural overlap in textual and photographi genre ontent, the results of this theme-oriented study also revealed subtle ross-ultural distintions (see also Sims & Guie, 1992). The theoretial importane of ross-ultural variations in genre ontent Although the Duth-English and British text types analyzed here ould be defined as similar genres in omparable examples of the same genre set (i.e. the annual report),

133 partiular attention should still be paid to the signifiant ross-ultural differenes that were diserned. These differenes subtle as they may be indeed indiated that the two business ommunities have slightly different pereptions of the ontent required for the annual report genre, i.e. they have different ideas about what ontent is appropriate to a partiular purpose in a partiular situation at a partiular point in time (Berkenkotter & Hukin, 1995: 13). Culture-speifi ontent preferenes were predominantly exhibited for the English-language management statements, where the Duth-English CEO s statements were nevertheless hybrid and thus omparable genres in relation to both British versions of the management statement. Culture-speifi preferenes for text themes or photo themes were less prominent in the English profile desriptions and in the English operational reviews. Consistent with Miller s (1994) and Swales (1990) assumption that genre ontent is shaped by the rhetorial or disourse situation in whih the genre is used, it may be argued that the ross-ultural differenes in themes were aused by distint genre onditions. With regard to the management statements, for example, many of these differenes ould well be attributable to ommuniation norms and soio-eonomi irumstanes. In partiular, the frequent use of the interpersonal HRM/people theme and the orporate governane theme in the Duth-English CEO s statements might be the result of the typial interational nature of the message, on the one hand, and the introdution of a new Duth orporate governane ode at the time of publiation, on the other hand (e.g. Hooghiemstra et al., 2004). In addition, the overrepresentation of the visual all exeutive board members theme in the Duth- English management statements may be explained by different soial phenomena, e.g. by the fat that the Netherlands is a more feminine and therefore more ooperation-minded ulture than the British ulture (e.g. Claes & Gerritsen, 2007) or by the fat that suh olletive pitures add to the redibility of annual reports originating in a small ountry where multinationals are wathed losely in the ontext of a vulnerable, open eonomy (Kolk, 2003: 286). The theoretial importane of photographs in genre ontent A third theoretial impliation involves the role of photographi themes in the identifiation of the genre ontent of professional texts. As was disussed above, the ognitive mapping approah to the study of themes allowed an insight into genre substane that was omprised of ontextualized onepts as well as semanti relations whih were not bound to written language alone (Ryan & Bernard, 2003). In the urrent researh projet, this approah proved fruitful in the ross-ultural omparison of informational landsapes in the multimodal texts of international annual reports. With regard to the images in the Duth-English and British annual reports, the themati ontents were reognized by plaing and onneting the meaning of visualized key-objets within the rhetorial ontext of the text (e.g. Kress & Van Leeuwen, 2001). Although pitures have reeived relatively little attention in analyses of genre ontent, in the present study they were revealed as ruial in the identifiation of some genres inluded in the Duth-English and British annual reports: had it not been for the investigation of photo themes, the Duth-English and British profile desriptions would inorretly have been typified as idential instanes of the same genre. Thus, the present ontent study onfirms the ontention that it is indeed preferable to inlude pitures in generi analyses, as they add value to the thik desription of partiular text types within and aross ultures.

134 The analytial relevane of qualitative and quantitative methods in the identifiation of genre ontent The investigation of genre ontent in Duth-English and British annual reports started with the manual olletion of multimodal data and then moved on to the SPSS appliation in order to alulate signifiant ross-ultural differenes in text themes and photo themes. In other words, it omprised both interpretative or qualitative and numerial or quantitative analyses (Ryan & Bernard, 2000). This multi-method approah to the identifiation of themes was partiularly useful in the present study, whih sought to explore the prototypial ontent features in large Duth-English and British-English orpora. Here, the manual or qualitative proedures allowed for the ognitive interpretation and labeling of oneptual networks or themes-in-ontext, while the quantitative proedures enabled rossultural omparisons of genre ontent with a large number of researh texts. The analytial relevane of interoder reliability tests in the identifiation of genre ontent Apart from measuring the reliability of identifiation and labeling proedures in the present ontent investigation, the interoder reliability tests helped to establish more aurate lists of text themes and photo themes. Therefore, the findings of these tests not only onfirmed that interoder reliability analyses enhane the reproduibility or redibility of the analysis of oneptual themes, but they also suggested that suh analyses may improve the validity of the atual researh steps and researh results. As Ryan and Bernard point out, researhers will be better able to desribe themes on the basis of the systematial observation of interoder agreements and disagreements for key onepts or disrete oneptual themes (2000: 792). The analytial relevane of genre ontent for reader response analyses A final analytial impliation onerns the relevane of ontent-related results for reader response analyses in whih both the Duth-English and British-English annual report texts are involved. The themati analysis presented in this hapter has indiated several signifiant differenes between the texts extrated from the Duth- English annual reports and the texts taken from the British-English annual reports. These signifiant differenes in partiular are interesting in a researh setting aimed at measuring the effetiveness and appreiation of English-language texts, from the perspetive of their (inter)national reader groups. For example, it would be relevant to test whether international investors do or do not prefer to read details about dividend in the introdutory statement of the annual. In addition, it would be useful to investigate if UK-based stakeholders think less positively of a Duth ompany when pitures of employees are fully omitted from the orporate profile (sine in the UK this text often ontains suh pitures). Aordingly, the themati differenes that were observed in the previous setions ould be used as independent variables in interultural effet studies that inlude English annual reports produed in both national business ultures. The ontent analysis disussed in this hapter provided a first answer to the speifi researh questions in the seond analytial phase, i.e. onerning the desription and omparison of multimodal disourse features in Duth-English and British annual report texts (see setion 1.5). Chapter 6 will expand the answer to these speifi researh questions by fousing on the strutural elements typially used in the Duth-English and British management statements, orporate profiles and operational reviews.

135 Chapter 6 Charaterization of the rhetorial genre struture of Duth-English and British-English annual reports The previous hapter showed that there were several signifiant ross-ultural differenes with respet to the textual and visual themes inorporated in the Duth- English and British annual report texts. Aordingly, it indiated that the Duth and British ompanies maintained slightly different pereptions about the neessary genre ontent of their multimodal annual reports. The analysis desribed in the urrent hapter is a further extension of the seond researh phase, whih fouses on the identifiation of textual and photographi disourse in Duth-English and British-English annual report texts (see setion 1.5). More speifially, this hapter seeks to reveal the rhetorial strutures of the Duth-English and British annual report texts (RQs 2.a-b), and to ompare these strutures in order to determine to what extent they orrespond or differ aross business ultures (RQ 2.). If relevant, the omparative struture-related results will be explained by the ommuniative ontext in whih the Duth-English and British reports were realized (RQ 2.d). The onept of a rhetorial genre struture ensues from the observation that speialist writers seem to be fairly onsistent in the way they organize their overall message in a partiular genre (Bhatia, 1993: 29). This has also been pointed out in prior researh on annual reports, i.e. both with respet to the overall organization of annual report setions (i.e. Jameson, 2000) and with respet to the text struture of individual setions (e.g. Garzone, 2004; Solbjørg Skulstad, 1996). In general, the rhetorial struture of a genre omprises its prototypial and thus reognizable internal organization, whih is onstituted by standard and optional strutural elements that in turn ontribute to the realization of the genre purpose. As suh, the rhetorial struture allows for a purposeful organization of the ontent of a genre (Swales, 1990). Within genre theory, moreover, it is regarded as a soio-ognitive onept: it enompasses regularities in the strutural interpretation of texts and as suh reflets aumulated and onventionalized soial knowledge available to a partiular disourse or professional ommunity (Bhatia, 1993: 21). The rhetorial struture of a genre is thus losely related to the soial ontext in whih it is used; it enompasses soial rules that reveal how members of a language use ommunity typially organize their disourse. Based on the generi interpretation of rhetorial text struture, several studies have indiated that different ultures may have different notions about the strutural elements that are required to build a partiular genre (e.g. Vergaro, 2005; Zhu, 2005). Consistent with the analyses of genre ontext and genre ontent, the analysis of prototypial genre strutures in Duth-English and British annual report texts revolves around ulture-speifi preferenes expressed by expert members in the multiultural disourse ommunity of annual report produers. This hapter will first a more detailed working definition of the onept of rhetorial struture, as it is used in the present genre study (setion 6.1). In setion 6.2, the hapter ontinues with an overview of the proedures for the investigation of strutural elements in the texts and photos of the Duth-English and British annual reports. This setion also inludes the lists of ompulsory and allowable strutural elements that were generally identified within the English annual reports and it desribes the results of an interoder reliability test. The findings of the rossultural analyses of strutural elements in management statements, profiles and

136 operational reviews are disussed in setions, 6.3, 6.4 and 6.5 respetively. The presentations of results will be sueeded by a onlusion in setion A working definition of genre struture: moves and strategies Before disussing the relevane of the rhetorial text struture to the present researh materials, this paragraph will first provide a further theoretial explanation of its omponents. A summary of this explanation is displayed in Figure 6.1. As was indiated in setion 2.3.4, genre theory assumes that the rhetorial struture of a genre is based on onventional and allowable omponents, eah of whih serves a typial ommuniative intention whih is always subservient to the overall ommuniative purpose of the genre (Bhatia, 1993: 30). The onventional omponents or moves determine the prototypial marostruture of a genre: they are rhetorially funtional text fragments that are reurrently applied within a given ommunity to ahieve a partiular text purpose. Their presene or absene in a text tends to vary the nature of the genre signifiantly, often introduing new or additional onsiderations in the ommuniative purpose of the text (Bhatia, 1993: p.21). Therefore, moves are disriminative elements or neessary onstituents of the struture of a genre and, as suh, they distinguish between genres within and aross national or disiplinary ultures. Connor (2000), for instane, examines variation in moves to determine whether any generi differenes exist between US grant proposals in the humanities and sienes. While she finds that these proposals are generially similar in terms of some of the moves, i.e. the territory, gap, goal and means moves, she also observes that the siene proposals onsistently ontain the researh question move but the humanities proposals do not. Bhatia (1993) argues that genre struture does not neessarily inlude a tight sequene of a fixed set of moves, suh as the three-move struture Swales (1990) assigned to researh artile introdutions (i.e. 1. establishing a researh territory 2. establishing a researh nihe 3. oupying the nihe ). He suggests that it is the onsistent presene or absene of partiular moves that typifies a genre most of all, rather than, for instane, the ontent or length of moves and their position within the genre struture. Additionally, several studies of professional genres have shown that a rhetorial genre struture an omprise (a) ompulsory and optional moves, (b) separate moves and intertwined moves, and () a linear and a flexible organization of moves (e.g. Bhatia, 1993; Connor, 2000; Garzone, 2005). Both the ompulsory-optional distintion and the mixing of moves are illustrated by Garzone s analysis of CEO s letters (2005). For instane, she explains that moves providing a performane narrative are obligatory, while moves on orporate governane are absent in some texts. Furthermore, she finds that moves on operational performane and future priorities may be realized separately in some letters but an also be intertwined in others, meaning that the performane move may ontain information about the ompany s plans for the future and the future move may omprise details about operations in the past year. With regard to the flexibility of move strutures, Bhatia (1993) notes that moves in sales promotion letters an our in various positions. He shows, for example, that the offering inentives move an be positioned diretly after the produt detailing move but an also be part of the opening move in promotion letters.

137 Moves may onsist of optional or allowable rhetorial tatis. These tatis reveal how individual text produers have hosen to realize partiular moves (Henry & Roseberry, 2001). In other words, the allowable tatis exploit the onventional rules within the rhetorial boundaries of the move, often for the purpose of greater effetiveness in a very speifi soio-ultural ontext, originality or very speial reader onsiderations (Bhatia, 1993: 20). Sine the rhetorial value of allowable tatis is framed by the rhetorial funtion of the move, they have no diret influene on the ommuniative purpose of a genre (this is different from the ompulsory moves, whih do affet the generi purpose). As these tatis are of minor importane to the strutural typifiation of a genre, they are defined as nondisriminative genre elements (Bhatia, 1993). In his analysis of introdutions to researh artiles, Swales (1990) refers to the allowable tatis as steps, while Bhatia (1993) uses the term strategies in his study of business and legal texts. Both approahes have gained support in genre investigations. Drawing on Swales (1990), Flowerdew and Dudley-Evans (2002: 472, 473) explain that the move preparing the reader for the deision in editorial letters inludes one or more of the following steps in a fairly preditable sequene : referring to the submission apologizing for a delay in responding interpreting reviewers reports. Following Bhatia (1993), Henry and Roseberry (2001) note that the move promoting the andidate within letters of appliation an be realized by a number of strategies in random order, e.g. listing relevant skills, stating how skills were obtained or naming present job. Although steps and strategies are rhetorially similar (i.e. both are the writer s individual ontributions to the establishment of a move), their possible positions within the move would seem to differ. Aording to Henry and Roseberry, if the move an only be realized by a series of strategies in a partiular order then, following Swales (1990), strategies an be thought of as steps (2001: 154). Hene, if a move is typially realized by tatis that do not generally our in a partiular sequene, then, following Bhatia (1993), they an be typified as strategies. These notions imply that Bhatia s strategies an be applied to both well-strutured moves and to moves that have a less-defined organization. Several genre analyses have endorsed the idea that generi texts may ontain strategies that do not always appear in a standard number and/or order. In his explanation of the rhetorial struture of book blurbs, for instane, Bhatia (2004) indiates that the appraising the book move may be onstituted by different series of the same strategies, i.e. (a) previewing the book indiating the value of the book desribing the book or (b) previewing the book desribing the book indiating the value of the book. The sheme in Figure 6.1 provides a working definition of genre struture; it illustrates the theoretial onepts of rhetorial moves, steps and strategies that are entral to the identifiation of rhetorial genre struture.

138 Rhetorial struture = ognitive pattern onstrued for the prodution and interpretation of a genre in a partiular disourse situation. Moves = prototypial text entities with a partiular rhetorial funtion that ontribute to the realization of a genre s ommuniative purpose. Steps = individual tatis used by a writer to realize a move; they our in a fixed sequene. Strategies = individual tatis used by a writer to realize a move; they may our in random order. Figure 6.1: Working definition of rhetorial genre struture onstitutive elements (Bhatia, 1993; Swales, 1990). As was indiated by the theoretial survey in this subsetion, the elements of genre struture have been the fous of attention in many genre studies. However, few of these studies have aounted for the hoie between steps and strategies or for the way strutural elements were identified. Furthermore, only a small number of studies have applied the onept of the rhetorial move struture to texts ontaining multiple modes of disourse, e.g. texts and graphs or pitures. In order to demonstrate the appliability of the rhetorial genre struture to the present study of annual report texts, the following subsetions will elaborate on the demaration of strutural omponents and on the role of multimodality in strutural analyses. More speifially, they will aount for the urrent fous on rhetorial moves and strategies (setion 6.1.1), for the differene between moves and themes in strutural genre analyses (setion 6.1.2), for the linguisti demaration of moves and strategies (setion 6.1.3) and for the role of photographs in the identifiation of moves and strategies (setion 6.1.4) Steps versus strategies Swales steps and Bhatia s strategies basially differ in two aspets, the first and most obvious of whih was already mentioned in setion 6.1. Clearly, these oneptualizations vary in the positioning of tatis within moves: steps usually our in a linear sequene, while strategies may our in various positions inside a move. In Henry and Roseberry s strutural analysis of job appliation letters, this distintion led to the onlusion that where more than one strategy was used to realize a move none appeared to be part of a series in a speifi order and so none was onsidered to be steps (2001: 159). A seond, more subtle distintion between steps and strategies is introdued by Bhatia (1993), who indiates that there is a differene with regard to the rhetorial explanation of the terms. On the one hand, it seems that Swales steps are primarily referred to as a text produer s options for the realization of a move; on the other hand, strategies are generally defined as the options a text produer has to onstrut a move in a way that he or she believes to be useful for an effetive and suessful aomplishment of the

139 ommuniative purpose of the genre (Bhatia, 1993: 32). Aordingly, Bhatia implies that the onept of genre strategies fouses more on the rhetorial value individual text produers add to the organization of the genre. Both the flexibility in allowable tatis and the rhetorial value of the tatis are addressed in Nikerson and De Groot s (2005) study of Duth-English and British managerial forewords in 2001 annual reports. Based on their finding that the managerial forewords ontained moves whih were onstituted by ompulsory and optional tatis in varying positions, Nikerson and De Groot use Bhatia s (1993) rhetorial strategies to oneptualize these tatis. Furthermore, they argue that the strategy-onept is most relevant for the type of disourse analyzed, where the writer is alled upon to use the genre reatively to persuade the reader that the orporation onerned is worth (re)investing in (Nikerson & De Groot, 2005: 335). Following Bhatia (1993), Henry and Roseberry (2001) and Nikerson and De Groot (2005), and onsidering the limited data available on the rhetorial struture of annual report texts (in partiular for orporate profiles and operational reviews), the present study assumes the existene of strategies-in-moves throughout the disussion of genre struture The differene between moves and themes in rhetorial genre struture In defining move struture as the ognitive organization of the message in the text, the distintion between moves and strategies and themes deserves further attention. Although these strutural elements and ontent elements may be related, they do not neessarily oinide. Whereas moves and strategies-in-moves are marked by boundaries in rhetorial funtion (Bhatia, 1993), themes transend these boundaries and may our either within or aross them. Two fragments extrated from the urrent researh materials and labeled aording to the lists of moves and strategies in setion illustrate this point. For instane, in Figure 6.2 the orporate profile inludes the identifying the ompany -move, whih overs multiple themes (identified in the ontent analysis of orporate profiles disussed in setion 5.4.1). [1. Every WPP ompany is a distintive brand in its own right; all with their own identities and own areas of expertise.] [2. That is their strength. What they have in ommon is in harnessing intelligene, talent and experiene to bring ompetitive advantage to their lients.] Figure 6.2: Text example two strategies-multiple themes (WPP annual report, 2003: 8). The move in Figure 6.2 is realized by two strategies: (1) offering a general profile of the ompany and (2) fous on orporate strengths and redibility. Looking at themati ontent, it was found that the first strategy omprises the textual orporate identity -theme, while the seond strategy reflets the orporate strengths -theme. A ontrasting example is displayed in Figure 6.3, whih ontains a text fragment drawn from an English operational review. Here, one theme (identified for operational reviews in setion 5.5.1) is aounted for aross the boundaries of multiple strategies in the move reporting on operational performane.

140 [1. The doubling of our total R&D effort paid off with a number of innovations reahing the market in Europe.] [2. Examples are Sikkens Autolear WB, the first workable waterborne learoat in the ollision repair industry, and Sikkens Autolear III, our fastest growing two pak urethane learoat.] Figure 6.3: Text example two strategies-one theme (Akzo Nobel annual report, 2003: 67). Again, two strategies onstitute the move in Figure 6.3: (1) providing speifi operational performane details and (2) desription of the produt. Contrary to the strategies in Figure 6.2, however, both strategies in this text fragment are dediated to the text theme researh and development per division. The way in whih the strutural elements and the theme o-ourred in the written text presented in Figure 6.3 is similar to the manner in whih strutural form and themati ontent were ombined within the doumentary photographs in the annual reports. Both the struture and ontent analyses of photographs foused on objets, ativities and senes displayed in the images (see also setions 3.2 and 5.2.2); therefore, the investigations of visual themes, moves and strategies were bound to and overed the same visualized items within the frame of partiular photographs. This interrelation an be larified by the photograph in Figure 6.4. Within an English operational review, a portrait of online relatives in front of a laptop omputer gave rise to the identifiation of one move targeting the market and one onstitutive strategy desription of the ustomer (see setion 6.2.3) on the basis of one theme that was diserned for this photograph in the analysis of genre ontent, i.e. family lients (see setion 5.5.2). Figure 6.4: Photo example one strategy-one theme (British Teleom annual review, 2003: 6) The demaration of moves and strategies in written texts The disussion of themes and the identifiation of strutural elements is a ruial element in the explanation of the demaration of moves and strategies in the written text. In line with the generi notion that textual boundaries between strutural elements are based primarily on reognizable shifts in the rhetorial funtion of the text ontent (e.g. Bhatia, 1993, 2004; Swales, 1990), the examples in setion demonstrate that themes may help define the boundaries of these elements. Paltridge points out that it is a ognitive rather than linguisti sense that guides our pereption of textual division and further explains that the typifiation of strutural elements in a genre depends more on onvention, appropriateness and ontent than on formal signals suh as lexial ohesion or lay-out of the text (1994:

141 295). Genre analyses in both the ESP and Australian traditions (see setion 2.3) have indiated that formal patterns annot always be reliable indiators of textual boundaries. In his omparison of sales promotion letters, for example, Bhatia (1993) finds that distintive pronominal referenes may represent idential rhetorial values in a partiular move. Aording to Bhatia, both text fragments in Figure 6.5 omprise the move establishing redentials, but distint language items are used to realize this move. While example (1) fouses on ompany-related strengths by means of using pronominal self-referenes, example (2) refers muh more to the needs of the ustomer by means of using seond-person referenes. (1) We are an established ourier inorporated loally sine 1971 when ourier servie was virtually unknown. We have ome a long way and today we an boast that we have established ourselves as a speialist in the field of ourier servie, espeially to destinations within Indonesia, i.e. Jakarta, Medan, Surabaya et. (2) Have you ever wished there were one omputer ourse providing you with on-site onsultany to assist your staff in solving problems or implementing appliation? Figure 6.5: Example pronouns in move, from Bhatia (1993: 86, 87). Furthermore, Paltridge (1994) shows that lexial ohesion, although initially onsidered to be a boundary-marker by Hasan (1989), atually extends aross the boundaries of moves and annot fully aount for distintions between strutural elements. Paltridge exemplifies his point with part of a servie enounter taken from Hasan (1989). The example in Table 6.1 shows that the lexial hain ontinues aross strutural elements with different rhetorial funtions and, as suh, lexial ohesion annot be relied upon for the demaration of these elements. Table 6.1: Example lexial ohesion and moves, from Paltridge (1994: 290). Strutural element Role partiipant Text Lexial hain Sales request Customer: Can I have ten oranges and a kilo of bananas, please? Sales ompliane Vendor: Customer: Yes, anything else? No thanks. Sale Purhase Purhase losure Vendor: Customer: Vendor: That ll be dollar forty. Two dollars. Sixty, eighty, two dollars. Thank you. dollar dollars dollars In addition, the theme-related examples drawn from the present researh (disussed in setion 6.1.2) revealed that genre ontent annot stritly define boundaries of moves or strategies either: themes may indiate boundaries of moves or strategies but may also over both moves and strategies. Paltridge s laim (1994) that the identifiation of boundaries between elements of genre struture is primarily ontent-based an therefore be modified: although ontent may be an indiator of suh boundaries, it is no more regulative than formal text features, e.g. pronominal referenes or lexial hains. The findings presented here support the idea that the

142 main riterion for the demaration of moves and strategies is not formal or ontentrelated but indeed pragmati or rhetorial in nature. Consequently, the typifiation of boundaries between moves and strategies annot be based on a fixed list of formal or ontent-related riteria; rather, it is grounded in the reognition of shifts in rhetorial value aross different text fragments. Despite the limitations of formal and ontent-related riteria for the demaration of strutural elements and the importane of rhetorial funtion, ontent and formal features still remain helpful devies in determining where the rhetorial funtion of one text fragment ommenes and another one ends. The above text examples taken from the Duth-English and British-English annual reports illustrate this point with regard to ontent features. Depending on whether moves or strategies are the units of analysis, shifts in themes may ontribute to the definition of beginnings and ends for these strutural elements, i.e. the theme researh and development per division for the move reporting on operational performane and the two themes orporate identity and orporate strengths for the strategy offering a general profile of the ompany and the strategy fous on orporate strengths and redibility, respetively. As for the demarating role of lexiogrammatial features, an example an be found in Hyland s study of researh artile abstrats (2000). He explains that for the purpose move and the sueeding produt/main findings move the hoie of presentation verb often distinguished the two moves, items suh as disuss, desribe, explore and address marking intentions, and show, demonstrate, find and establish signaling results (Hyland, 2000: 69). Although suh ontent-based and form-based riteria for the demaration of elements in the written text may not hold for eah move or strategy in a partiular genre at all times (e.g. Bhatia s example in Figure 6.5, in whih distint pronominal expressions still reflet the move labeled establishing redentials ), it will here be argued that they are nevertheless relevant tools for a reliable analysis of genre strutures: these riteria enable genre analysts to reprodue how they reognized boundaries between text fragments with different rhetorial funtions Photographs in rhetorial genre struture In spite of the growing importane of multimodality in the onstrut of professional or aademi texts (e.g. Kress & Van Leeuwen, 2001, 2006; Rowley-Jolivet, 2004), the role of photographs within the rhetorial struture of a given text type has been investigated only sporadially. Within the frame of multimodality analysis, for instane, Mahin and Van Leeuwen (2004) refer to photographs in a skin are advertisement that may highlight the result or added value of the produt promoted (by portraying a person with beautiful skin) or images that may desribe the instrumental solution to a skin problem (by portraying the produt promoted). As for ESP/LSP genre analysts in partiular, Bhatia (2004) desribes the strutural funtion of pitures in a restaurant advertisement by taking into onsideration the visualized items in the photo, the themati ontent refleted in these portrayals, the position of the photo within the ad and the ommuniative purpose of the entire ad. He notes that a piture of people dining, positioned at the top of the ad and diretly beneath the headline, omprises the move reader attration. He asribes the move produt details to a piture of food at the bottom of the ad. Similar examples an be found in Askehave and Nielsen s genre analysis of a European industrial ompany s digital homepage (2005). They identify visual moves and strategies on the basis of the items portrayed, the themes expressed

143 through these visualized items, the position of the photo, the ontent of adjoining text exerpts and the aim of the full web-page. Askehave and Nielsen (2005) show that the move attrating attention is established by both visual and textual strategies, i.e. through sequentially flashing images of produts, prodution proesses and onsumers whih are aompanied by verbal one-liners highlighting unique selling points. For instane, their definition of strategies in the attrating attention move depended on the rhetorial assoiation between a piture of a man lying on a patio working on a laptop with a up of offee beside him and an adjoining text fragment saying We are proud of adding flavour to your day (Askehave & Nielsen, 2005: 136). Although Askehave and Nielsen do not refer to the pitorial and verbal strategies with speifi labels, they ould atually be given names, e.g. (a) desribing the produt, desribing the manufature proess or identifying the ustomers for the images of end produts, laboratory researhers at work or ustomers enjoying the end produt and (b) desribing value of the produt for the photo-related text omments. Also, Askehave and Nielsen (2005) find that the move identifying web-owner is realized strategially by images of the ompany, its buildings or its staff, thus by visual strategies that ould be labeled presenting orporate failities and giving fae to the ompany. Sine the ontext analysis (Chapter 4) and the ontent analysis (Chapter 5) already indiated that doumentary photographs hold relevant ommuniative purposes and themes in relation to the written texts of the Duth-English and British-English annual reports, it was deided that these photographs should also be inluded in the present analysis of rhetorial genre struture. The visual analyses by Bhatia (2004) and Askehave and Nielsen (2005) were referred to for an indiation of how to ondut a photo-based investigation of genre struture. In setion 6.2 both the textual and visual moves and strategies identified in the English managerial statements, profiles and operational reviews are further desribed. 6.2 Analyzing moves and strategies in Duth-English and British annual report texts Setion 6.1 emphasized that the urrent identifiation of moves and strategies was driven mainly by the theoretial notions of Swales (1990) and Bhatia (1993), who propose that eah strutural element should ontribute to the realization of the ommuniative purpose of the genre. In line with genre theory, it was noted that the most important riterion for assigning rhetorial value to elements of disourse struture was thus funtional rather than formal. However, it was also argued that ontent-based features or lexio-grammatial features in written texts and photographs an be useful indiators for the definition of the boundaries between funtional units in the text. Previous studies on the rhetorial struture of annual reports have adopted varied perspetives: some have foused on the identifiation of moves and strategies in management letters to stakeholders (e.g. Garzone, 2005; Gillaerts, 1996; Nikerson & De Groot, 2005; Solbjørg Skulstad, 1996), some have looked at information strutures within and aross annual report setions (e.g. Jameson, 2000), and some have investigated the (antithetial) struturing of the entral theme in texts and photographs of annual reports (e.g. Davison, 2002). It would seem, however, that there is still a need for the general analysis of rhetorial strutures in different setions of annual reports originating in different national business ultures. Consequently, the present study enters on the identifiation of all

144 moves and strategies embedded in the texts and doumentary photographs of management statements, orporate profiles and operational reviews of Duth- English and British-English annual reports. Similar to the proedures of the ontent analysis desribed in setion 5.2, the strutural data for texts and images were first olleted by hand and were then registered in SPSS files. The following subsetions will further speify the tehniques applied to establish the moves and strategies in the present researh materials. Setions and provide an overview of analytial frames used for the reognition of moves and strategies in written texts and photos. The appliation of these frames resulted in lists of reurrently reognized moves and strategies that were used in the full analysis of strutural elements in texts and images. The final versions of these lists are disussed in setion Setion reports on the interoder test that was onduted to measure the reliability of the moves and strategies identified Design for the identifiation of textual moves and strategies While ontinually onsidering the ommuniative purpose of the full text in the identifiation of strutural elements, design-related, ontent-related and languagerelated indiators were studied as a means to demarate moves and strategies in the written texts of the Duth-English and British annual reports. Although shifts in layout, themes and linguisti realizations were not always equally relevant for the demaration of all text-based moves and strategies (see setion 6.1.3), in various ways they proved to be helpful devies in signaling where one funtional entity ended and another began. Drawing on Hasan (1989), Martin (1992), Connor and Mauranen (1999) and Vergaro (2005), for instane, the design-related, ontentrelated and language-related lues that were used for strutural demarations in the present study omprised: extra spae or headings (e.g. Strategy versus Market expansion), semanti ohesion/lexial referene (e.g. profit-loss-interest-taxurreny), lexial ohesion (e.g. profit-earnings-inome), marked themes (e.g. the returns are likely to improve; So far, all ontrats have been extended), oherene markers (e.g. onsequently, in addition), illoution markers (e.g. I am pleased to report a landmark year ) or tense and modality (e.g. In early 2004, we sold our Brazilian retail hain ; We will also ontinue to ooperate fully ). The example in Figure 6.6 is extrated from a British operational review and illustrates how various substantive and formal features ontributed to the demaration of moves and strategies in the written texts. [A. (a)the expeted ost of our redundany programme is 45.5m. (b) Cost savings related to those departures amount to 31.4m in total, of whih we invested 5.8m in reating 290 new jobs in our stores.] [B. () By the time you read this, most of these jobs will already have gone, and we will be looking ahead to a further round of job uts in the urrent year, (d) with the aim of bringing our total head offie team down to around 1,500 people.] Note: [A., B.] represent moves, (a)-(d) represent strategies Figure 6.6: Example role purpose, design, ontent, language in demaration text (Boots annual review, 2003: 7). In Figure 6.6, the extra white spae suggests that there may be a distintion in informational value between the two text fragments. It appears that the first

145 fragment is primarily onerned with reporting on finanial performane, as an be inferred from semanti and lexial relations (e.g. ost, ost savings,, invested). This report on finanial results, moreover, is realized through two sentenes with different rhetorial funtions: one antiipating the future osts of the redundany program that was exeuted (lexial hoie expeted, ost, 45.5m) and one disussing the investments arried out in the past year as a result of ost savings (lexial hoie ost savings, 31.4m, invested, 5.8m and past tense invested). The seond fragment, on the other hand, seems to fous on speulating about the ompany s future performane in relation to the results of the redundany program, as an again be inferred from semanti and lexial relations (e.g. by the time, jobs, already, ahead, further, job uts, urrent year, aim, bring down, people). Again, these future speulations are expressed through two phrases that have rhetorially different values: one introduing the reader to future prospets (e.g. expetations) on job-utting (lexial hoie by the time, jobs, gone, further round of job uts, urrent year = 2004 and future tense will have gone, will be looking) and one reporting on objetives set for the up-oming year (lexial hoie aim, bring down). Within the British operational review, whih generally aims to provide an optimisti and strategi overview of divisional results (see setion 4.4.4), these two fragments ontain two moves (A) reporting on finanial performane and (B) speulating on future performane. Eah move onsists of two strategies (a) future prospets and (b) finanial performane details for move (A) and () future prospets and (d) future objetives for the short-term for move (B) Design for the identifiation of visual moves and strategies A similar interpretational approah was taken towards the analysis of strutural elements in the photographs and in their adjoining text fragments. The main riterion for the identifiation of photo-related moves and strategies in Duth- English and British annual reports was the relationship between the rhetorial funtion of depited items and the ommuniative purpose of the full text. Aordingly, the genre-based analysis of the strutural role of photographs in annual report texts resembled Bhatia s (2004) examination of photographs in produt ads and Askehave and Nielsen s (2005) study of images in the orporate homepage. As was indiated in setion 3.2, the strutural investigation of annual report photos speifially foused on visual items in the objets-as-mode, ativity-as-mode and sene-as-mode (Kress & Van Leeuwen, 2006). This meant that visualized objets, ativities and senes were taken into onsideration in the demaration and identifiation of rhetorial entities within photographs. The semanti oherene between the visualized items within the boundaries of the piture frame was also regarded as an important visual lue for the strutural interpretation of images. The moves and strategies refleted in the adjaent text fragments were simultaneously analyzed with the strutural elements in the photographs. The analysis of these text fragments was based on the formal linguisti lues disussed in setion An explanation of the two examples in Figure 6.7 illustrates how rhetorial funtion, ontent and formal visual indiators add to the demaration and labeling of moves and strategies in photographs inluded in the urrent researh texts.

146 (1) (2) Nany MKinstry ore Selfridges, Birmingham Haden Young ompleted eletrial and mehanial and fit-out servies at the new Selfridges store in Birmingham. Figure 6.7: Two examples role purpose, ontent, form in demaration images (CEO Wolters Kluwer annual report, 2003: 7; Building Balfour Beatty annual report, 2003: 22). The identifiation of strutural elements in example (1) of Figure 6.7 is relatively unompliated. Example (1) shows a piture of Nany MKinstry, CEO of Wolters Kluwer in the ompany s 2003 annual report. The piture was inluded in the CEO s statement, whih is the introdutory text in Duth-English reports and serves to introdue the readers to the group ompany s performane throughout the past year. It also aims to establish a personal relationship with the reader (see setion 4.4.4). Referring to the ommuniative objetives of the CEO s statement and to the fat that the piture is positioned next to the running written text, it an be assumed that the atual sender of the message is portrayed here. This assumption is endorsed by the relationship between Nany MKinstry s signature at the end of the statement and the objet (i.e. an individual woman in formal dress) visualized in the photograph. The assumption is further expliated by the lexial realization Nany MKinstry, whih is integrated with the image and whih orresponds with both the information at the end of the running text and the visualized objet. Through the photograph of the CEO in her own introdutory statement, the identity of the Wolters Kluwers ompany beomes personalized: the ompany obtains a human fae by whih the ompany an be reognized. As suh, the photograph reflets a move with the rhetorial funtion identifying the ompany, whih is realized by referene to a partiular item of identifiation, i.e. the CEO s fae. Therefore, the move is onstituted by the rhetorial strategy giving fae to the ompany. The same move and strategy an also be assigned to the adjoining text, whih omprises a diret referene to the rhetorial value of the photograph. Although the rhetori of the image in example (2) of Figure 6.7 involves a similar interpretation proess, its position in the Balfour Beatty operational review of 2003 and the written texts referring to the image ause this proess to be more omplex than in the CEO example. Together with several other pitures of buildings and adjoining text fragments, this photograph was plaed at the bottom of the operational review pages, whih in British annual reports disloses biased and strategi information on operations by orporate division (see setion 4.4.4). Drawing on suh ontextual features, the piture of the Selfridges store in example (2) would seem to be related to the 2003 operational results of one of Balfour Beatty s divisions (i.e. Building, Building Management and Servies). This observation is first of all supported by the rhetorial oherene between the portrayed objet (a modern-style building) in the divisional review and an exerpt in the running text saying that Balfour Beatty s subsidiary Haden Young ompleted major works for the Meteorologial Offie in Exeter, King s College Hospital in London and the new Selfridges store in Birmingham... (Balfour Beatty, 2003: 23). Furthermore, the written text adjoining the piture at the bottom of the page further refines the rhetorial interpretation of the image. It states that Haden Young

147 ompleted ore eletrial and mehanial and fit-out servies at the new Selfridges store in Birmingham. Thus, the adjoining text fragment suggests that the building in the piture is not manufatured by the Balfour Beatty ompany alone, but that it is the end produt realized with help from the tehnial servies the ompany provided throughout the past year(s). The portrayed objet, its textual explanations and the implied instrumental value added by the Balfour Beatty produt (proessed as a photo theme in setion 5.5.2), make it plausible to asribe the following move and strategy to the image: the move labeled detailing the produt, as it refers to the onstrution ativities of the ompany, and the strategy ommuniating value of the produt, as the ompany s ativities have ontributed to the funtioning of an impressive new building. The move and strategy for the adjaent text fragment an be based on its lexion (e.g. Haden Young, eletrial and mehanial and fit-out servies, store), ation verb (to omplete) and past tense (ompleted). These lues reveal that the text fragment reflets the move reporting on operational performane realized by the strategy operational performane details. The examples in Figure 6.7 projet photographs that an be marked by a single rhetorial move established by a single strategy. However, some pitures in the present researh materials inorporated a two-fold ontribution to the rhetorial struture of the full annual report text. In the Duth-English operational review of Philips 2003 annual report, for instane, a lose-up piture of a young Afro- Caribbean woman holding a reently developed mp3-player (see Figure 6.8), was assigned two overlapping moves: (A) the move labeled detailing the produt with the strategy desription of the produt and (B) the move targeting the market with the strategy desription of the ustomer. While the running written text and the adjoining text fragment only highlight the ation of produt development and the features of the produt, the piture fouses on the new produt as well as on the target group for whih it has been reated. Key Ring In 2003 Philips introdued the innovative Key Ring range that allows onsumers to wear their MP3 audio, digital images and information. These players provide a highly portable platform for listening while on the move, with quik and easy PC uploading and simultaneous reharging thanks to USB onnetivity. The Camera Key Ring even allows onsumers to reate their own instant memories, yet is small enough to be added to a key ring for easy arrying and wearing. Figure 6.8: Example multiple moves in one photograph (Philips annual report, 2003: 24) List of textual and visual moves and strategies in annual reports Preliminary interpretations of strutural elements in the written texts and in the photographs of the urrent researh materials, resulted in lists of moves and strategies that were relevant for the full investigation of all 137 English management statements, profiles and operational reviews. The final versions of these lists were omposed during an interoder reliability test, whih is further disussed in setion

148 The use of predetermined lists in the analysis of moves and strategies is relatively ommon in genre-based researh, e.g. Swales s move-step approah for aademi artiles formed the basis for strutural analyses of hairman s statements (Solbjørg Skulstad, 1996) and onferene proposals (Hallek & Connor, 2006) and Bhatia s move-strategy approah for promotional letters was used for strutural analyses of job appliation letters (Henry & Roseberry, 2001) and unsoliited promotional s (Barron, 2006). The disussion below outlines the rhetorial funtions of the 17 moves and 34 strategies that were identified and used in the strutural analyses of Duth-English and British-English annual report texts. The funtional desriptions of rhetorial moves and strategies for the urrent researh materials were either new, i.e. introdued as a result of strutural elements that were unique for the Duth-English and British-English texts, or were derived from previous studies on rhetorial strutures in professional genres. The urrent moves and strategies referring to the orporate situation, the orporate strategy, the finanial and operational results, the future ourse of the ompany, the management situation, the expressions of sympathy and the addressing of the reader were (in)indiretly drawn from genre studies of management letters onduted by Garzone (2005), Gillaerts (1996) and Nikerson and De Groot (2005). Moreover, the desriptions for moves and strategies assoiated with detailing the produt, targeting the market, establishing redentials and ommuniating the mission statement were derived from Bhatia s work (1993, 2004) on advertising letters, philanthropi fundraising letters and aademi introdutions. Finally, there were two kinds of strutural elements that proved to be relevant for the present annual report texts but that had not been identified in prior researh. First, moves and strategies refleting the overall profile of the organization provided a ontext whih briefly introdued stakeholders to the type of ompany they would be reading about in sueeding setions of the annual report and whih would perhaps help readers understand why partiular deisions had been made during the past year(s). Seond, this study revealed that moves and strategies in the form of orientating headlines, headings and intertextual referenes were ating as rhetorial devies to antiipate reader orientation to the text, i.e. to indiate what readers ould expet from given information and to failitate the reading proess. Before presenting the two final lists of moves and strategies (Tables 6.2 and 6.3), several omments need to be added to aount for their usefulness in the present study of rhetorial strutures in annual report texts: - the lists do not assoiate partiular moves and strategies with ertain setions in the Duth-English and British annual reports; the observation that several moves and strategies ourred aross different annual report setions onflited with an initial attempt to establish a strit ategorization of typial strutural elements per text type. For instane, the textual move ommuniating orporate strategy was prominent in the Duth-English CEO s statements and the operational reviews as well as in the British Chairman s statements, the CEO s statements and the operational reviews; - the lists do not assoiate partiular strategies with ertain moves: the analysis revealed that some strategies were not bound to one move only and had thus been used to establish different moves in various setions of the annual report. This is illustrated by the observation that the strategy strategi auses/onsequenes ourred within three moves: reporting on finanial performane, reporting on operational performane and ommuniating orporate strategy ;

149 - the lists do not present the identified moves and strategies in a partiular order of appearane. A linear, full-text struture in the appliation of moves and strategies ould rarely be diserned in any of the texts examined within the Duth-English and British reports. Most strutural elements ould our in varying positions within different instanes of an annual report setion. The move identifying the orporate situation, for instane, was found in various positions in the Duth-English and British operational reviews, suh as in the beginning as a way of providing a general ontext for the operational results or in the body of the text as a means of providing a speifi ontext for partiular operational ahievements; - the lists do not distinguish between written and visual modes sine many moves and strategies were relevant for text as well as for images. For example, the move detailing the produt/servie with either of the strategies desription of the produt/servie or value of the produt/servie ould be found in both written texts and photographs within the Duth-English and British operational reviews. It must be emphasized that these general observations did not hold for all moves and strategies in all instanes. Some ombinations of textual moves and strategies or visual moves and strategies inluded in the list were more ommon in partiular annual report setions than others. With regard to the written texts, for example, the move expressing sympathy with the strategy words of appreiation ourred most frequently in the British hairman s statements and in the Duth-English CEO s statements. Also, the move presenting the mission statement/ore values with strategy mission statement/ore values almost exlusively appeared in the Duth-English and British-English orporate profile desriptions. With regard to the doumentary photographs, for instane, it was found that the move identifying the ompany with strategy giving fae to the ompany ourred in all multimodal Duth-English and British management statements but in a smaller number of profiles and operational reviews. Tables 6.2 and 6.3 display the full lists of multimodal moves and strategies respetively. Sine rhetorial strategies onstitute rhetorial moves i.e. they are the author s individual tools for the realization or exeution of moves (Bhatia, 1993) only Table 6.3 inludes strategi examples taken from the urrent researh materials. Within the annual report texts, these strategies were used to establish the rhetorial moves displayed in Table 6.2.

150 Table 6.2: Rhetorial moves identified for texts and photographs in Duth-English and British annual reports. Move Desription 1. Identifying the ompany Offers a general fatual summary of the orporate profile, onisely states what kind of ompany is represented in the annual report. 2. Presenting the mission Motivates why the ompany is here and refers to the ore statement/ore values values that are at the basis of the ompany s performane. 3. Communiating orporate strategy Identifies at a fundamental level. Clarifies whih strategi ourse the ompany has been and will be taking on the long term, from an operational and finanial perspetive. 4. Detailing the Elaborates upon the pratial and added value of the produt/servie produt/servie that is offered by the ompany. 5. Targeting the market Desribes the ompany s markets in terms of geography, industry or type of ustomer and the way in whih it is atually targeting these markets. 6. Considering ustomer needs 7. Showing ommitment to proper business ondut 8. Identifying the orporate situation 9. Reporting on the leadership situation 10. Reporting on finanial performane 11. Reporting on operational performane Highlights speifi intrinsi ustomer needs/wishes to whih the offered produt is onsidered to be the right solution. Emphasizes the ompany s involvement in the realization of proper business ondut, e.g. priniples and ativities in HRM, governane, CSR. Depits the past or urrent situation of the ompany, by referring to external fators (e.g. eonomy, ompetitors) or general performane outome (positive/negative). Aounts for hanges in the exeutive/supervisory board and for the professional or biographial bakground of management members. Reflets the ompany s finanial performane and gives insight in its finanial position, e.g. in terms of profits, sales, dividend, ash flow. Presents the operational, non-finanial ahievements of the ompany in the past year (s), e.g. in terms of aquisitions, researh & development. 12. Speulating on future Previews the future ourse of the ompany, its performane performane and its markets. 13. Establishing redentials Fouses on orporate strengths, expertise, awards and the ability to reognize and fulfill reader (=stakeholder) needs. 14. Expressing sympathy Conveys the writer s feelings towards events/ahievements that have been realized by or that have affeted internal or external stakeholders. 15. Addressing the reader Aims to involve readers in the text, e.g. by salutations, selfintrodutions and soliiting response. 16. Antiipating reader orientation 17. Indiating speial share ownership Informs readers about the details that are to be ommuniated in a next text fragment or that have been disussed in previous setions. Aounts for speial shareholder groups/interests that have or have had a relatively large influene on orporate ativities.

151 Table 6.3: Rhetorial strategies identified for texts and photographs in Duth-English and British annual reports. Strategy 1. General profile of the ompany 2. Giving fae to the ompany 3. Mission statement/ore values 4. Strategi fous points long term 5. Strategi auses/onsequenes 6. Desription of the produt/servie 7. Value of the produt/servie 8. Desription of target market(s) 9. Desription of the ustomer(s) 10. Customer needs Desription + examples Briefly introdues the ompany to the readers. Hunter Douglas is the world leader in window overings and a major manufaturer of arhitetural produts. Hunter Douglas has its Head Offie in Rotterdam, the Netherlands, and a Management Offie in Luerne, Switzerland. Hunter Douglas employs about 15,900 people. Projets the people who work within the ompany. [Piture:] ¾ shot of a smiling woman in her work outfit with logo, who holds a leaning mahine. Desribes the ompany s fundamental views and the raison d être. ING s mission is to be a leading, global, lient-foused, innovative and low-ost provider of finanial servies through the distribution hannels of the lient s preferene in markets where ING an reate value. Presents the ompany s strategi priorities for the past/oming years. Our priority is to fous on investments in attrative growth markets where we an ahieve or maintain leadership positions. (...) Where we annot meet our own strit eonomi riteria, we will urtail our investments or look at reverting to an export model. Explains the strategi auses/onsequenes related to performane/poliy. The added apaity will enable us to meet inreased demand from our ustomers. Depits instrumental/pratial harateristis of the produt/servie. [Piture:] Close-up of wine bottles in a supermarket rak. Depits the added value of the implementation/use of the produt/servie. When risis omes, our solutions support the medial are you reeive and ensure the presriptions you need are waiting for you and your family. Speifies the markets/industries the ompany operates or fouses on. It has interests in plumbing and heating distribution in Ireland, Italy, Denmark, The Netherlands, Hungary and the Czeh Republi. Speifies the type(s) of ustomers the ompany aims at. Our ustomers are drawn from aross the industry: selfemployed ontrators, large onstrution businesses and governments. Establishes expliit ustomer wishes and demands. People on the move need onvenient and seure aess to up-to-date information, appliations and deision-making tools.

152 Table 6.3 ontinued. Strategy 11. Involvement in HRM/CSR, et. 12. Market onditions 13. Corporate performane ontext 14. Management details 15. Finanial performane details 16. Operational performane details 17. Future prospets 18. Future objetives short term 19. Confidene in the future 20. Awareness of reader needs Desription + examples Refers to poliy-making and ativities related to sustainability/integrity. We atively promote a ulture of aring for the environment. We respet the rights and dignity of our employees and ustomers, and we treat both with high regard. Our partnership with World Food Programme demonstrates a ompany-wide ommitment to working for a better world. Sets the external sene that is at the basis of the orporate performane. Exiting 2002 the Company was onfronted with a further deline in the Information and Communiation Tehnology (ICT) market in all major geographies. Indiates the nature/diretion of the annual results in general. The past 12 months have been dominated by hange. Even as reently as January this year, Invensys faed an unertain future; but what a differene a few months and a lot of hard work have made in tabilizing the Group. Presents hanges in or professional/biographial data of management. I am delighted that Mihael Welms joined the board on 29 January Mihael brings with him a wealth of retail experiene. Desribes finanial results on group-level or division-level. For the year ended 31 July 2003, Smiths generated operating profit from ontinuing ativities of 372m, exeeding last year s result by 2%, in spite of the translation effet of an unfavourable US Dollar exhange rate. Desribes operational results/ativities on group-level or division-level. we ahieved a number of other groundbreaking wins this year, inluding a six-year ontrat with ITV to reate a new multimedia network, and a five-year deal with Alliane & Leiester. Presents expetations and speulations for the future performane/ontext. We expet our industry to grow as those savings and investments aumulate. Presents expliit orporate aims for the short term (1-2 years). During 2004 Castorama plans to invest 48 million in refurbishing or reloating 10 stores. Expresses management s onfidene in the future ourse of the ompany. We fae the future with the onfidene that omes from knowing that we have an exellent set of produts, a loyal and satisfied ustomer base. Reognizes pratial needs of stakeholders and their reading interests. Your ompany believes this is not in the interests of shareholders, ustomers or employees.

153 Table 6.3 ontinued. Strategy 21. Fous on strengths and redibility 22. Words of appreiation 23. Words of ompassion 24. Self-introdution of the sender 25. Salutation to the reader 26. Polite ending 27. Indiretly soliiting response 28. Desription of old/new information 29. Evaluation of information in text 30. Emphasis on strategi fous points 31. Introdution of shareholders 32. Marketing ommuniation means Desription + examples Emphasizes exellene and trustworthiness (awards, rankings, et.) In 2003 Medial Systems strong performane reeived reognition in the form of four Frost and Sullivan Awards, inluding that of Medial Imaging Company of the Year. Expresses appreiation for ahievements/onfidene by stakeholders. We want to speifially mention the outstanding ontribution made by our management and staff throughout the Group (...) we would like to thank them for their efforts. Expresses ompassion for stakeholder involvement in harmful inidents. I know I speak on behalf of everyone at HSBC in offering my deepest sympathies to the families of those who lost their lives and those who were injured. Speifies the identity (e.g. name, professional role) of the sender. On behalf of the Managing Board, I am writing to you about the performane of your bank in Comprises salutation of sender seeking ontat with reader. Dear fellow shareholder,... Comprises sender s polite ending of the text. Best regards,... Enourages readers to reat (indiretly) to the message of the sender. The diretors reommend that you vote for all of the items of business at the AGM. Presents given or up-oming information (headings, intertextuality). Rik Haythornthwaite will provide more detail on our struture and individual businesses in his Chief Exeutive s review. Adds sender s value to the up-oming information (headings). [Heading] Strong exeution against produt roadmaps Highlights strategi fous entral to the up-oming information (headings). [Heading] Organi Growth: A Top Priority Presents (major) shareholders or their speial interests in the ompany. At the end of 2003, the State held 19.3% of our outstanding shares. Desribes marketing ommuniation instruments or hannels. Examples of the orporate advertising ampaign to launh the new brand in broadsheet and speialist publiations. [+ pitures of ads]

154 Table 6.3 ontinued. Strategy 33. Company history 34. Corporate equipment/faility Desription + examples Refers to the historial development of the ompany. Versatel was inorporated under the laws of The Netherlands on Otober 10, 1995, as a private ompany with limited liability. Versatel onverted its legal struture from a B.V. to a publi ompany with limited liability on Otober 15, Desribes the tools and failities used for the ompany s daily operations. [Piture:] Full shot of HSBC s new Vitorian-style head offie in St James Street, London Interoder reliability test for textual and visual moves and strategies In the present study an interoder reliability test was onduted for moves and strategies in the texts and photographs inluded in Duth-English and British annual reports. The author was the first oder in the test; again, the seond oder was a graduate in Business Communiation studies with near-native profiieny in English. Based on the findings of a pilot test in the interoder analysis, the preliminary lists of moves and strategies were adapted for use in the formal interoder reliability test and in the full-sample analysis of annual report texts (see setion 3.5 for a further disussion of the proedures for the interoder analysis). Table 6.4 shows the results of the formal interoder reliability test, i.e. it presents the mean Kappa values and the mean agreement perentages for the observation of ombinations of moves and strategies in written texts and photographs. Table 6.4: Evaluation of interoder agreement for ombinations of moves+strategies in the written texts and ombinations of moves+strategies in the photographs. Setion/ text type Qualifiation Mean Kappa SD Min. max. Agreement % Management Moderate statements/text Management Almost statements/photo perfet Profiles/ TEXT Substantial Profiles/PHOTO Slight Operational Substantial reviews/text Operational reviews/photo Fair Based on the rule of thumb set for Cohen s Kappa, the results of the interoder reliability test show that agreement in qualifiations between oders ranged from negative (-.25) to almost perfet (). The overall mean Kappa value was substantial (.62) (Rietveld & Van Hout, 1993). Table 6.4 indiates that the applied oding system for the exploratory identifiation of textual moves and strategies in English annual report texts was a more reliable researh tool than the oding system

155 for the analysis of photographi moves and strategies, in partiular with regard to the profiles and operational reviews. 6.3 Moves and strategies in management statements The following subsetions reveal the moves and strategies in the written texts and in the photos of the Duth-English and British-English management statements. Setion desribes and ompares the typial strutural elements in the written texts of these English statements, while setion desribes and ompares the strutural elements in the photographs Moves and strategies in the running texts of the management statements Appendix IV indiates that Duth and British orporations tended to use a variety of textual moves and strategies to realize the ommuniative purpose of their Englishlanguage management statements. Several of the rhetorial moves and strategies were typial of both the Duth-English and the British management statements; therefore, these elements defined the genre struture of the annual report in both national business ultures. Cross-ultural similarity partiularly emerged for moves and strategies reporting on the orporate strategy, the orporate performane situation, the finanial and future results and the orporate strengths. It was also observed for overt presentations of the presumed writers of the texts and for headings highlighting the informative value of the sueeding text fragments. Table 6.5 shows whih textual moves and strategies ourred in 50% or more of the Duth-English and/or British management statements. Table 6.5: Typial moves and strategies in running text of 50% of management statements. Moves/Strategies Du-CEO n=20 Br-CEO n=18 Br-Chair n=16 Count % Count % Count % Tot.% χ 2 (df p = 2) Moves 3. Corp. strategy Corp. ondut Corp. situation Leadership a b < Finan.perform Operat.perform a < Future perform Sympathy a < Address.reader Read.orientation 11 a b <.05

156 Table 6.5 ontinued. Moves/Strategies Du-CEO n=20 Br-CEO n=18 Br-Chair n=16 Count % Count % Count % Tot.% χ 2 (df p = 2) Strategies 1. General profile Strategi fous Str.onsequene Desr.produt a < Involvement Ext. onditions Perform.ontext Board details a b < Detail fin.perf Detail opr.perf Prospets Objetives Confidene fut Reader needs Strengths Appreiation 20 b a b < Self-intro Salutation 12 b a < Desr. Info Emph.str.fous Combinations d Move 3 + Strategy 4 + Strategy 5 + Strategy 16 + Strategy 17 Move 7 + Strategy 11 Move 8 + Strategy 12 + Strategy 13 Move 9 + Strategy 14 + Strategy 21 + Strategy 22 Move 10 + Strategy 12 + Strategy 15 + Strategy 16 Move 11 + Strategy 4 + Strategy 5 + Strategy 12 + Strategy 15 + Strategy 16 + Strategy 17 + Strategy 21 n= n= n= n= n= n= n= n= n= n= n= n= n= n= n= n= n= n= a <

157 Table 6.5 ontinued. Moves/Strategies Combinations d Move 12 + Strategy 12 + Strategy 16 + Strategy 17 + Strategy 18 + Strategy 19 Move 14 + Strategy 22 Move 15 + Strategy 24 + Strategy 25 Move 16 + Strategy 28 + Strategy 29 + Strategy 30 Du-CEO n=20 Br-CEO n=18 Br-Chair n=16 Count % Count % Count % Tot.% χ 2 (df = 2) n= b 7 n= n= b 63.2 n= n= n= n= a 0.0 n= n= n= n= n= p < < Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. d Combination of move ourring in 50% of the Duth-English and/or British management statements + strategies ourring in 50% of this move. As an be seen in Table 6.5, there was statistially signifiant variation between the Duth-English CEO s statements and the British CEO s statements for three elements. Expliit referenes to the informational development of the text through the move antiipating reader orientation were signifiantly more prominent in the British CEO s statements than in the Duth-English CEO s statements. Conversely, the salutation to the reader strategy as a rhetorial devie to greet the reader appeared relatively more often in the Duth-English CEO s statements than in the British CEO s statements. This signifiant result also applied to the ombination of move 15 addressing the reader and strategy 25 salutation to the reader, whih was observed in signifiantly more Duth-English CEO s statements than British CEO s statements. Furthermore, Fisher Exat test revealed a signifiant differene aross all three types of statements. It showed that the strategy with rhetorial funtion words of appreiation ourred in a signifiantly larger number of Duth- English CEO s statements and British Chairman s statements than British CEO s statements. As for the British Chairman s statements alone, statistially signifiant differenes ould be asribed to the reporting on operational performane move, the desription of the produt/servie strategy and the ombination of move 11 reporting on operational performane and strategy 15 finanial performane details. For all these strutural elements, the British Chairman s statements showed relatively low frequeny sores. A statistially signifiant under-representation was also found for the British CEO s statements, in whih the move with rhetorial funtion expressing sympathy was signifiantly under-represented. Next, a signifiant result ould be diserned for the Duth-English CEO s statements. It appeared that in these statements the ombination of move 12 speulating on future performane and strategy 18 future objetives for the short term was signifiantly

158 over-represented, i.e. it was a relatively prominent rhetorial devie for informing readers about the future ourse of the ompany. Finally, two statistially signifiant differenes were observed between the British CEO s statements and the British Chairman s statements. The move reporting on leadership situation ourred signifiantly more often in the British Chairman s statements than in de British CEO s statements. And a relatively larger number of British Chairman s statements ontained the management details strategy, as ompared to the British CEO s statements Moves and strategies in photos and text fragments of the management statements This subsetion presents the ross-ultural researh results for all moves and strategies in the photographs and in the adjaent text fragments within the Duth- English and British management statements. For both the Duth-based and UKbased management statements, the rhetorial funtion of moves and strategies in the photographs foused on the personalized identifiation of the ompany, i.e. by portraying board diretors (the CEO or the Chairman) who laimed to be responsible for a given text. As for the Duth-English and British CEO s statements in partiular, they also orresponded in the distribution of visual moves and strategies related to the desriptive illustration of orporate produts or servies. Table 6.6 provides a omparative overview of the strutural elements that were observed in the doumentary images of Duth-English and British managerial forewords. It shows that only the visual moves and strategies related to the personalized identifiation of the ompany ourred in 50% or more of the English management statements. Table 6.6: Frequeny sores moves and strategies in photographs of management statements. Moves/Strategies Du-CEO n=17 Br-CEO n=16 Br-Chair n=14 Count % Count % Count % Tot.% χ 2 (df = 2) Moves 1. Id.ompany Detail.produt Target.market p Strategies 2. Giving fae Desr. produt Customers Communiat Faility Combinations d Move 1 + Strategy 2 n= n= n= Note.. Fisher s Exat Test. d Combination of move ourring in 50% of the Duth-English and/or British management statements + strategies ourring in 50% of this move.

159 With regard to the moves and strategies that were refleted in the photographs of the management statements (displayed in Table 6.6), Fisher s Exat tests did not reveal any statistially signifiant variations aross the Duth and British business ultures. In line with the rhetorial value of moves and strategies in the photographs of the Duth-English and British management statements, the adjoining text fragments often ontained strutural elements that aimed to larify the identity of the diretor portrayed and thus to provide the orporation with a personal fae. Also, both the Duth-English and the British-English CEO s statements inluded adjaent fragments with headings that highlighted the relevane of the information in the sueeding sentenes and, simultaneously, of the ontent of the photos. Table 6.7 shows the ross-ultural results of the analysis of photo-related omments. Table 6.7: Frequeny sores moves and strategies in photo-related omments of management statements. Moves/Strategies Du-CEO N=12 Br-CEO n=13 Br-Chair n=9 Count % Count % Count % Tot.% χ 2 (df = 2) Moves 1. Id.ompany Corp.strategy Detail.produt Target.market Busin.ondut Corp.situation Oper.perform Establ.redent Adress.reader Read.expetat Strategies 1. Corp.profile Giving fae Strategi fous Strat.onseq Desr.produt Value produt Desr.market Desr.ustomer Involvement Eon.sene Perf.ontext Op.perf.detail Fut.prospets Corp.strengths Self-intro Desr.info Evaluat.info Emph.strat.inf Faility p

160 Table 6.7 ontinued. Moves/Strategies Combinations d Move 1 + Strategy 2 Du-CEO N=12 Br-CEO n=13 Br-Chair n=9 Count % Count % Count % Tot.% χ 2 (df = 2) n= n= n= p.63 Note.. Fisher s Exat Test. d Combination of move ourring in 50% of the Duth-English and/or British management statements + strategies ourring in 50% of this move. As is displayed in Table 6.7, there were no statistially signifiant differenes in the use of rhetorial moves and strategies within photo-related omments inluded in the Duth-English and British management statements. 6.4 Moves and strategies in orporate profiles Subsetions and disuss and ompare the rhetorial moves and strategies that were found in the written texts and photographs of the Duth-English and British orporate profiles Moves and strategies in the running texts of the orporate profiles The list of all strutural elements that were observed in the running (written) texts of the Duth-English and British-English orporate profiles are inluded in Appendix V. The frequeny distributions for the textual moves and strategies in these profiles revealed a fous on strutural elements with a speifi rhetorial funtion. Table 6.8 indiates that both the Duth-English and the British profiles typially onsisted of textual moves and strategies that onerned the basi presentation or profiling of the group ompany, i.e. the desription of ore harateristis, produts or servies and orporate strengths. The English profiles also ontained rhetorial elements that enhaned reader orientation in the texts (i.e. headings pointing out the informational value of the sueeding paragraph).

161 Table 6.8: Typial moves and strategies in running text of 50% of orporate profiles. Moves/Strategies Du-Profile Br-Profile n=21 n=20 Count % Count % Total % χ 2 (df p =1) Moves 1. Identify ompany Presenting mission Corporate strategy 18 b a < Detailing produt Orientating reader Strategies 1. General profile Mission/ore values Strategi fous Strat.onsequene Desription produt Value produt Desription market Desript.ustomer Operat.perf.details Corp. strengths Desription info Combinations d Move 1 + Strategy 1 Move 2 + Strategy 3 Move 3 + Strategy 4 + Strategy 5 Move 4 + Strategy 6 + Strategy 7 + Strategy 21 Move 16 + Strategy 28 n= n= n= n= n= n= n= n= n= n= Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. d Combination of move ourring in 50% of the Duth-English and/or British management statements + strategies ourring in 50% of this move. A Chi-Square test revealed that the move with rhetorial funtion ommuniating orporate strategy aused signifiant variation aross the orporate profiles (see Table 6.8). Compared to the British-English profiles, this textual move ourred in a signifiantly larger number of Duth-English orporate profile desriptions.

162 6.4.2 Moves and strategies in photos and text fragements of the orporate profiles As the Duth-English and British orporate profiles ontained a smaller number of photographs and photo-related text fragments than the management statements and the operational reviews, the rhetorial-strutural role of images and omments in the profiles was relatively limited. Table 6.9 displays all visual moves and strategies that were observed in the photographs in the Duth-English and British orporate profiles. In line with the textual moves and strategies, the strutural elements in the doumentary images mainly involved the identifiation of the ompany, in partiular through metaphorial profiling (e.g. piture of a globe), illustrations of produts and produt values and presentations of orporate ommuniations and orporate failities. Table 6.9 shows that the English profiles ontained few typial strutural elements, i.e. moves or strategies ourring in 50% or more of the Duth-English and/or British profiles. Table 6.9: Frequeny sores moves and strategies in photographs of orporate profiles. Moves/strategies Du-Profile Br-Profile n=8 n=10 Count % Count % Tot.% χ 2 Moves 1. Identify ompany Mission Detailing produt Targeting market Business ondut Oper.performane (df=1) p Strategies 1. General profile Giving fae Mission Desript. produt Value produt Desription market Desript. ustomer Involvement Oper.perf.details Communiations Faility

163 Table 6.9 ontinued. Moves/strategies Du-Profile Br-Profile n=8 n=10 Count % Count % Tot.% χ 2 Combinations d Move 1 + Strategy 1 + Strategy 2 + Strategy 34 Move 4 + Strategy 6 + Strategy 7 n= n= n= n= (df=1) p Note.. Fisher s Exat Test. d Combination of move ourring in 50% of the Duth-English and/or British management statements + strategies ourring in 50% of this move. Fisher s Exat tests ould not detet signifiant variation aross the visual moves and strategies in the Duth-English and British orporate profiles. This implies that there were no substantial ross-ultural differenes in the strutural funtions attributed to photos in these texts. The moves and strategies that were found in the text fragments adjaent to the doumentary photographs in the English profiles are presented in Table Compared to the moves and strategies diserned in the images, the moves and strategies in the photo-related omments show less rhetorial variety: they primarily fous on the general identifiation of the ompany.

164 Table 6.10: Frequeny sores moves and strategies in photo-related omments of orporate profiles. Moves/strategies Du-Profile n=3 Br-Profile n=8 Count % Count % Tot.% χ 2 Moves 1. Identifying ompany Corporate strategy Detailing produt Finan.perform Operat.perform Establ. redentials Reader orientation (df=1) p Strategies 1. General profile Desript. produt Desription market Desript. ustomer Fin.perf.details Oper.perf.details Future objetives Corp. strengths Desription info Evaluation info Communiations Combinations d Move 1 + Strategy 1 + Strategy 32 n= n= Note.. Fisher s Exat Test. d Combination of move ourring in 50% of the Duth-English and/or British management statements + strategies ourring in 50% of this move. Similar to the analysis of visual moves and strategies in the English profiles, the investigation of moves and strategies in the photo-related text fragments did not reflet any statistially signifiant differenes aross the Duth and British national business ultures. 6.5 Moves and strategies in operational reviews The following subsetions will desribe and ontrast the frequenies of ourrene for the typial moves and strategies in the running texts (setion 6.5.1) and in the photos and adjaent omments (setion 6.5.2) of the Duth-English and British- English operational reviews.

165 6.5.1 Moves and strategies in the running texts of the operational reviews The variety of textual moves and strategies listed in Appendix VI suggests that the rhetorial genre strutures of the Duth-English and British operational reviews required text onstituents with diverse disursive funtions. Table 6.11 shows the moves and strategies that ourred in the running written texts of 50% or more of the Duth-English and/or British operational reviews. It indiates that these texts were typially onstituted by strutural elements that larify the influenes on and results of past and future ativities: more speifially, textual moves and strategies mainly foused on orporate strategy, the pratial effet of this strategy, finanial and operational performane details, market onditions and onsiderations for the future. In addition, both the Duth-English and British reviews ontained many headings orientating readers in the text. In general, the funtion of strutural elements in Duth-English and British operational reviews differed from the rhetorial funtion of elements in the management statements and orporate profiles, whih onentrated on summarizing overall ompany performane and on identifying the group ompany, respetively. Table 6.11: Typial moves and strategies in the running text of 50% of the operational reviews. Moves/Strategies Du-OperRev Br- OperRev n=22 n=20 Count % Count % Total % χ 2 (df = p 1) Moves 1. Identify ompany Corpor. strategy Detailing produt Targeting market Identify situation Finan.perform Operat.perform Future perform Reader orientat Strategies 1. General profile 22 b a < Strategi fous Strategi onseq Desript. produt Value produt Desript. market Desript.ustomer Customer needs Involvement Extern.onditions Perform.ontext Fin.perf.details Oper.perf.details

166 Table 6.11 ontinued. Moves/Strategies Du-OperRev Br- OperRev n=22 n=20 Count % Count % Total % χ 2 (df = p 1) Strategies 17. Future prospets Future objetive Future onfiden Corp. strengths Desription info Evaluation info Combinations d Move 1 + Strategy 1 + Strategy 6 Move 3 + Strategy 4 + Strategy 5 + Strategy 12 + Strategy 16 + Strategy 17 Move 4 + Strategy 1 + Strategy 4 + Strategy 6 + Strategy 16 Move 5 + Strategy 1 + Strategy 6 + Strategy 8 + Strategy 9 Move 8 + Strategy 4 + Strategy 12 + Strategy 13 + Strategy 15 + Strategy 16 + Strategy 17 Move 10 + Strategy 1 + Strategy 4 + Strategy 12 + Strategy 13 + Strategy 15 + Strategy 16 + Strategy 17 n= n= n= b 12 8 n= b 7 8 n= n= n= n= n=9 2 0 a 9 4 n=6 3 0 a 3 5 n= n= < <

167 Table 6.11 ontinued. Moves/Strategies Du-OperRev Br- OperRev n=22 n=20 Count % Count % Total % χ 2 (df = 1) Combinations d Move 11 + Strategy 1 + Strategy 4 + Strategy 5 + Strategy 6 + Strategy 7 + Strategy 9 + Strategy 10 + Strategy 12 + Strategy 13 + Strategy 15 + Strategy 16 + Strategy 17 + Strategy 18 + Strategy 21 Move 12 + Strategy 4 + Strategy 12 + Strategy 16 + Strategy 17 + Strategy 18 + Strategy 19 Move 16 + Strategy 28 + Strategy 29 n= b n= a 6 n= n= a n= b 8 n= p < < Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. d Combination of move ourring in 50% of the Duth-English and/or British management statements + strategies ourring in 50% of this move. Whereas no substantial variation ould be diserned for the most prominent moves in the written texts of the Duth-English and British operational reviews, the rhetorial strategies and the ombined moves and strategies in partiular did reflet statistially signifiant differenes aross the two national business ultures. First, the Duth-English and British operational reviews differed in terms of the strategy general profile of the ompany, whih was used to identify either the group ompany or its divisions; this strategy ourred in a signifiantly larger number of Duth-English texts than British-English texts. Similar signifiant results applied to the ombination of move 4 detailing the produt with strategy 4 strategi fous points, the ombination of move 5 targeting the market with strategy 6 desription of the produt, and the ombination of move 11 reporting on operational performane with strategy 10 ustomer needs ; these three ombinations were found more often in the Duth-English reviews than in the British-English reviews. Finally, there was a statistially signifiant distintion for the ombination of move 12 speulating on future performane with strategy 18

168 future objetives for the short term. This ombination was found in signifiantly more British than Duth-English operational reviews Moves and strategies in photos and text fragments of the operational reviews The Duth-English and British operational reviews were observed to ontain the largest variety of moves and strategies in the photographs and in the photo-related omments. This may be owing to the fat that the operational review is generally one of the most extensive and narrative non-finanial setions in the Duth-English and British annual reports. Table 6.12 presents the moves and strategies that appeared in the photographs of the sampled reviews per national business ulture. The table indiates that the typial visual elements predominantly ommuniated about the general identity of a orporate division, the people operating the division (i.e. staff or employees), the equipment used to realize divisional ativities, and the instrumental features of division-speifi produts. Table 6.12: Frequeny sores moves and strategies in photographs of operational reviews. Moves/Strategies Du-OperRev Br-OperRev n=15 n=17 Count % Count % Tot.% χ 2 p (df=1) Moves 1. Identifying ompany Corporate strategy Detailing produt Targeting market Business ondut Operat.performane Establish redentials Strategies 1. General profile Giving fae Strategi fous Desription produt Value produt Desription market Desription ustomer Involvement Operat.perf.details Corporate strengths Communiations Faility

169 Table 6.12 ontinued. Moves/Strategies Du-OperRev Br-OperRev n=15 n=17 Count % Count % Tot.% χ 2 Combinations d Move 1 + Strategy 2 + Strategy 34 Move 4 + Strategy 6 Move 5 + Strategy 8 + Strategy 9 n= n= n= n= n= n= (df=1).00 p Note.. Fisher s Exat Test. d Combination of move ourring in 50% of the Duth-English and/or British management statements + strategies ourring in 50% of this move. As Table 6.12 shows, there were no statistially signifiant differenes in the strutural importane of photographs within the Duth-English and British-English operational reviews. In other words, the rhetorial genre struture of the multimodal reviews produed in the Netherlands and in the UK ontained a similar distribution of visual moves and strategies. The strutural analysis of text fragments adjoining the doumentary images in the operational reviews revealed that these fragments had diverse rhetorial funtions in relation to the rhetorial value of the photographs. This is illustrated by the extensive list of strutural elements in Table Both in the Duth-English and in the British-English operational reviews, high frequenies of ourrene were registered for moves and strategies identifying the ompany, desribing the orporate produts or servies and reporting on operational results.

170 Table 6.13: Frequeny sores moves and strategies in photo-related omments of operational reviews Moves/Strategies Du-OperRe Br-OperRev n=13 n=15 Count % Count % Tot.% χ 2 p (df=1) Moves 1. Identifying ompany Mission/ore values Corporate strategy Detailing produt Targeting the market Customer needs Business ondut Corporate situation Leadership Finan. performane Oper. performane Future performane Establish redentials Addressing reader Reader orientation Strategies 1. General profile Giving fae Mission Strategi fous Strategi onsequene Desription produt Value produt Desription market Desription ustomer Customer needs Involvement External onditions Performane ontext Finan.perf.details Operat.perf.details Future prospets Future objetives Confidene in future Reader needs Corporate strengths < Self-intro Desription info Evaluation info Emphasis str. fous Communiations Faility

171 Table 6.13 ontinued. Moves/Strategies Du-OperRe Br-OperRev n=13 n=15 Count % Count % Tot.% χ 2 Combinations d Move 1 + Strategy 1 + Strategy 2 Move 4 + Strategy 6 + Strategy 7 Move 11 + Strategy 16 Move 16 + Strategy 28 + Strategy 29 n=7 5 5 n=9 8 3 n= n= n=7 3 3 n= n= n= (df=1) - p Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. d Combination of move ourring in 50% of the Duth-English and/or British management statements + strategies ourring in 50% of this move. Whereas the Duth and British orporations did not differ signifiantly in their use of doumentary photographs as rhetorial devies for the realization of genre purpose, they did differ slightly with regard to the rhetori applied to desribe these photos. Table 6.13 shows that the strategy fous on strengths and redibility was used for photo desriptions in a signifiantly larger number of the British-English operational reviews than the Duth-English operational reviews. 6.6 Conlusion and disussion The overall fous of this hapter was on the textual and visual entities that ontributed to the rhetorial genre struture and thus to the realization of the ommuniative purposes of Duth-English and British annual report texts. More speifially, the analysis of rhetorial genre struture aimed to provide insight in the typial moves and strategies adopted in the English annual report texts and the degree to whih these texts differ struturally aross the two business ultures. As suh, this analysis was related to RQs 2a-2b, whih onerned the desription of multimodal disourse features per ulture, and to RQ 2, whih involved the rossultural omparison of disourse features (see setion 1.5). The researh results showed that ompanies in the Duth and UK business ultures used a speifi series of textual and visual moves and strategies in eah annual report setion. For instane, the strutural elements in the English management statements mainly reported on the orporate strategy or results and added to the personifiation of the orporate identity, whereas the strutural elements in the English profiles presented the ompany s harater or produts and ontributed to its metaphorial identifiation. With regard to the ross-ultural omparison of rhetorial genre struture (RQ 2), moreover, it was observed that variations in strutural elements

172 espeially ourred aross the Duth-English and British management statements and that ross-ultural variations in general emerged for the textual moves and strategies in partiular. Aordingly, it an be onluded that the present researh results are onsistent with Bhatia (1993), Vergaro (2005) and Zhu (2005), who found subtle ross-differenes in the rhetorial genre struture of various business letters. In addition to a further disussion of ross-ultural similarities and differenes in rhetorial genre struture, subsetion will also onsider ontextbased explanations for these ross-ultural researh results, speifially related to speifi RQ 2d. In line with the seond main researh aim of this genre study, i.e. refining the generi perspetive on professional disourse (setion 1.5), this hapter will end with a subsetion about the impliations of this strutural analysis for genre theory and genre analysis Cross-ultural omparison of strutural elements Assuming that the strutural organization of a professional genre reflets preferred ways of ommuniating intention in speifi areas of inquiry (Bhatia, 1993: 29), this subsetion summarizes and explains the strutural preferenes expressed in the Duth-English and British management statements, profiles and operational reviews. Thus, it offers a more elaborate answer to RQs 2a-2d. Sine the ulture-speifi registration of textual and visual moves and strategies was inherent in the rossultural omparison of these strutural elements, the summaries below enter on ontrastive results. To highlight the researh results for the different types of strutural elements, these elements have been typographially marked. Comparison of rhetorial genre struture in management statements Several similarities were diserned for the textual and photo-related moves, strategies and ombinations of moves and strategies in the Duth-English and British management statements. As for the moves in the running written text, many English CEO s statements and British Chairman s statements used text fragments reporting on the group ompany s strategy, finanial and future performane and on the identity of the sender of the message (i.e. the CEO or Chairman). Popular strategies were those ommuniating about the ompany s identity and strategy, soial responsibility, performane ontext, overall performane, future prospets and objetives, and orporate strengths. Another prominent strategy involved the rhetorial orientation of the reader through the text. With regard to the ombinations of textual moves and strategies, there was high ross-ultural agreement for elements explaining the orporate strategy, business ondut, performane onditions and finanial performane, and for elements announing the informational value of preeding or sueeding text setions. The moves and strategies observed for the photographs and adjaent text fragments were idential throughout the Duth-English and British management statements: photo-related moves, strategies, and ombinations of moves and strategies primarily onerned the personalized identifiation of the ompany (i.e. by management portraits). The strutural investigation of Duth-English and British management statements also exhibited a number of ross-ultural differenes. These differenes exlusively involved the moves, strategies and ombinations in the running written texts of the English statements. The Duth-English and British management statements showed variation in textual moves related to the ompany s leadership situation, operational ahievements, sympathy towards stakeholders and reader-

173 orientation in the text. Here, the leadership-related and sympathy-related moves ourred least often in the British CEO s statements, the move on operational results ourred least often in the British Chairman s statements and the reader-orientation move ourred least often in the Duth-English CEO s statements. Strategi variation aross the two national business ultures was observed for the textual strategies involving produt desriptions, details on management (hanges), words of appreiation and text-initial salutations. Whereas the strategy desription of the produt appeared least often in the British Chairman s statements, the strategies related to management details, expressions of appreiation and salutations were least ommon in the British CEO s statements. As for the strategies related to management details, this finding seems to be onsistent with the fat that relatively few British CEO s statements (and relatively many British statements by the supervisory Chairman) ontained the move labeled reporting on leadership situation. Finally, some ross-ultural distintions emerged with respet to partiular ombinations of textual moves and strategies. These distintions onerned the ombination of the move reporting on operational performane with strategy finanial performane details, the ombination of the move speulating on future performane with strategy future objetives, and the ombination of the move addressing the reader with strategy salutation to the reader. The first ombination was least prominent in the British Chairman s statements, the seond ombination ourred most often in the Duth-English CEO s statements and the third ombination was least ommon in the British CEO s statements. This last observation was in line with the signifiant ross-ultural results for the individual salutation strategy in the written texts of the English management statements. The strutural analysis of management statements indiated that the Duth- English and British-English texts omprised rhetorial strutures that differed signifiantly in terms of textual moves, strategies and ombinations of moves and strategies. It should be noted that ross-ultural variations involved a number of differenes between the Duth-English CEO s statements and British Chairman s statements on the one hand and several differenes between the Duth-English CEO s statements and the British CEO s statements on the other hand. Here, the Duth-English CEO s statement thus formed a hybrid genre, whih both orresponded with and varied from the two British statement types. Many of the ross-ultural differenes seemed to have originated in the diverse professional roles of the presumed writers of the text, in the diverse positions of the texts within the annual report and, hene, in their ommuniative purpose. Similar observations also emerged in the investigations of genre ontext and genre ontent (see setions and 5.3, respetively). Both the English statements by the Duth CEO and the English statements by the British supervisory Chairmen were introdutory or interational texts in the annual report and, as suh, shared strutural elements that added to the realization of ommonalities in text purpose. The expression of sympatheti feelings towards employees and shareholders, for instane, was present in a signifiantly small number of British CEO s statements but in a relatively large number of Duth-English CEOs statements and British Chairman s statements. It ould be argued that the sympathy move was more popular in the Duth-English CEOs statements and British Chairman s statements beause these were onsidered to be personal annual report texts that needed to establish reader-writer relationships. Conversely, Duth and British orporations implied that the Duth-English and British CEO s statements were similar in their more elaborate and biased narrative on group operations (see setion 4.4.4); it was perhaps for that reason that the disussion of the leadership situation, for example, was typially inluded in the

174 genre struture of the supervisory Chairman s statement, but rarely in the CEO s statement. Thus, the Duth-English and British CEO s statements also shared strutural units that helped establish the overlapping aspets of their rhetorial aims. Comparison of rhetorial genre struture in orporate profiles The analysis of moves and strategies in the written text and pitures of the Duth- English and British profiles mainly revealed ross-ultural similarities. In the realization of text struture that would support the ommuniative aim of the full annual report text, ompanies in the Netherlands and in the United Kingdom typially relied on textual moves, strategies, and ombinations of moves and strategies that foused on the general orporate profile, ore values, produt desriptions and evaluations, market desriptions and orporate strengths. Additionally, both Duth-English and British profiles ontained brief information (p)reviews by means of whih readers were guided through the text. A few rossultural similarities were also found for the strutural elements in the photographs and adjoining text fragments of the English orporate profiles. Within the atual photographs, the move with rhetorial funtion identifying the ompany and the strategy with rhetorial funtion giving fae to the ompany were partiularly prominent among both the Duth-English and British profiles. The analysis of text fragments adjaent to the photos again showed a shared preferene for strutural elements that foused on the general identifiation of the ompany. It seemed that these elements highlighted the importane of the photographs for the profiling of the Duth or British ompany. Apart from the many similarities in the rhetorial struture of the Duth- English and British profile desriptions, the strutural analysis only revealed one ross-ultural differene with regard to the moves in the running written texts. Although the textual move ommuniating the orporate strategy ourred in a majority of the English-language profiles in both business ultures, it appeared in a signifiantly larger number of Duth-English texts than British texts. This suggests that the Duth ompanies in the present sample, as ompared to the British ompanies, typially asribed more rhetorial value to strategi information in the (international) identifiation of their ompany. Overall, the results of this strutural analysis of English orporate profiles show that these douments are marked by similar rhetorial elements aross the Duth and British business ultures. Exept for the signifiant ross-ultural differene in the appliation of moves reporting on orporate strategy, Duth and British ompanies typially built their orporate profiles on similar numbers of orresponding moves and strategies. This leads to the onlusion that the Duth- English and British orporate profiles ould not be distinguished as two different genres. The assumption of omparability between profiles was further endorsed by the findings of the ontextual analysis (setion 4.4.4) and the ontent analysis (setion 5.4), whih revealed that the Duth-English and British profiles refleted similar purposes and themes in similar disourse situations. It would seem that these annual report texts an therefore be lassified in the same genre ategory, with only slight distintions in the extent to whih partiular themes and moves were manifested as prototypial features of the genre. Comparison of rhetorial genre struture in operational reviews The rhetorial struture of the extensive Duth-English and British operational reviews also showed a high degree of similarity. Typial textual moves aounting for the strategi identity of the ompany, its produts/ servies, markets,

175 performane ontext, performane details and future prospets were distributed more or less equally in the running written texts of the Duth-English and British reviews. In addition, the move antiipating reader orientation, as a means of previewing or realling information for the reader, ourred in all operational reviews. With respet to the rhetorial strategies in the texts of these reviews, rossultural similarity ourred for elements ommuniating the ompany s (divisional) profile, its strategi fous points, the presentation of its produt features, its marketing priorities, performane onditions, finanial and operational performane, future speulations and unique selling points. Furthermore, both the Duth-English and British operational reviews inluded strategies that direted readers through the text on the basis of desriptive referenes to past or up-oming information. Finally, the Duth-English and British reviews showed similar frequeny distributions for ombinations of moves and strategies related to the identifiation of the ompany, the presentation of strategi information, the desription of the market situation, the presentation of finanial results and the orientation of the reader in the text. There was also high ross-ultural agreement on the use of moves and strategies in the photographs and adjaent text fragments of the operational reviews. Both the Duth- English and British reviews ontained photos with moves that explained the orporate identity and the qualities of the orporate produt. Popular strategies for the pitures in these reviews foused on giving fae to the ompany and desription of the produt. Common ombinations in the photographs of the English reviews were the move identifying the ompany together with the rhetorial strategies giving fae to the ompany and orporate equipment and the move detailing the produt with the strategy desription of the produt. The text fragments adjoining the photos in the operational reviews typially omprised moves reporting on produt features and strategies providing produt desriptions and operational performane details. A ombination that was prominent in the photospeifi omments involved the move detailing the produt with the strategy desription of the produt/servie. Next, the analysis revealed several signifiant differenes between the rhetorial strutures of the Duth-English and British operational reviews. In the written text, these differenes only emerged for the optional strategies and for the ombination of partiular moves and strategies. As for the textual strategies, a signifiantly larger number of Duth-English operational reviews than British reviews inluded the allowable tati of profiling a business (unit), e.g. to ontextualize divisional results or to support market desriptions. With regard to the ombination of textual moves and strategies, the strutural analysis showed signifiant ross-ultural variations for the move detailing the produt ombined with the strategy strategi fous points, the move targeting the market with the strategy desription of the produt, the move reporting on operational performane with the strategy ustomer needs and the move speulating on future performane with the strategy future objetives for the short term. While the first three ombinations appeared most often in the Duth-English reviews, the fourth ombination was most ommon in the British operational reviews. Finally, it was observed that the strategy labelled fous on strengths and redibility was signifiantly more prominent in the photo-speifi text fragments of the British reviews than of the Duth-English reviews. The British orporations in the present sample typially used these fragments to exploit the orporate suesses related to the objets or events portrayed in the photograph. Considering that the moves in partiular reflet the ommuniative purpose of a text and therefore determine the integrity of a genre, it may be onluded that

176 the Duth-English and British operational reviews belong to the same genre ategory. These texts showed no signifiant distintions in the typiality of individual moves, they differed only with regard to the extent to whih the writers individual tatis were ommonly used to exploit the onventional norms for these moves. Again, this onlusion seems to be supported by the ontextual and themati analyses of the Duth-English and British operational reviews (see setions and 5.4, respetively). Moreover, the observation that these operational reviews serve similar ommuniative purposes i.e. to provide information on divisional performane and to establish a positive impression of the ompany (see setion 4.4.4) would seem to explain the high degree of strutural equivalene between the Duth-English and British reviews: in order to establish these aims, the Duth and British ompanies in the present sample employed moves (and many strategies) with similar rhetorial funtions. Contextual explanations for the subtle ross-ultural differenes in the rhetorial genre struture of the Duth-English and British operational reviews remain more speulative and thus lead to tentative onlusions. It might be argued, for instane, that Duth orporations preferene for additional produt desriptions in the targeting the market move or onsiderations of ustomer needs in the reporting on operational performane move is attributable to the fat that Duth-English operational reviews aim to reah an audiene of lients, shareholders or employees while the British reviews mainly address an audiene of shareholders (see setion 4.4.4). Genre struture and the Duth-English and British-English annual reports as a genre set The summary and ontextualization of results provided in the paragraphs above suggest that the identifiation of the present Duth-English and British researh materials as examples of international annual reports is supported by the many ross-ultural similarities in rhetorial genre struture. From a strutural perspetive, the Duth-English and British-English annual reports an be regarded as omparable instanes of the same genre set. As in the previous genre analyses presented in Chapters 4 and 5, omparability refers to the degree to whih similar genre features (here: moves and strategies) are relied upon to solve the same rhetorial questions in a partiular ommuniative situation (e.g. Miller, 1994). The analysis of rhetorial strutures revealed that eah text type within the Duth-English and British annual reports was marked by a unique olletion of onventional and allowable elements. None of the text types or annual report setions analyzed was based on strutural elements that were equally typial to another text type in the English annual report. In addition, it was found that many individual rhetorial elements did not only our in one text type, but were used aross different text types inside the Duth-English and British-English annual report. For example, the reporting on operational performane move and the strategi auses/onsequenes strategy, although most typial to the written text in the operational reviews, were present in all three text types within the urrent Duth- English and British researh materials. Eah text type was haraterized by a unique set of moves and strategies, some of whih were used aross different text types (i.e. intertextually, see Devitt, 1991) within the same annual report. From a strutural perspetive, it ould hene be onluded that the Duth-English and British annual reports were not merely individual professional genres; instead, they appeared to ontain a ompilation of different genres that had been integrated into and thus allowed for the existene of larger texts (i.e. annual reports). The assumption that rhetorial struture ontributed to the emergene of Duth-English and British

177 annual reports as genre sets is similar to Jameson s (2000) suggestion that eah setion in the (US) annual report has a struture of its own whih simultaneously ontributes to the non-linear hyper-struture of the full report. In addition, the strutural analysis of Duth-English and British management statements, profiles and operational reviews showed that eah of the genres within the annual reports ontained onventional and allowable elements that were idential aross the two business ultures. For instane, the move identifying the orporate situation was almost equally typial to the marostruture of the running (written) text of the Duth-English and British management statements. This was also the ase for the ombination of the identifying the ompany move with the strategy orporate equipment/ faility refleted in photographs within the Duth-English and British operational reviews. It must be noted that ross-ultural similarities were most frequent and most expliit in the rhetorial struture of the Duth-English and British orporate profiles and operational reviews. Sine the Duth-English CEO s statements, the British Chairman s statements and the British CEO statements eah represented different kinds of managers with distintive professional roles in the organization and sine they appeared at different plaes within the annual report, they showed the largest number of signifiant variations in moves and strategies aross ultures. Despite the strutural differenes among the Duth-English and British management statements, however, the results of the ontext, ontent and struture analyses did not reveal different management-related genres aross ultures. Beause of the hybrid format of the Duth-English CEO s statement, it seemed to be generially ompatible with both the British Chairman s statement and the British CEO s statement whih may indeed be reognized as different genres within the British annual report (see e.g. Nikerson & De Groot, 2005). In general, the analysis of moves and strategies thus implied that the sampled Duth-English and British management statements, profiles and operational reviews an be reognized as omparable instanes of the same text types. Or, in Swales terms (1990), the strutural omparison of annual report texts aross ultures showed suffiient family resemblane among these texts to be able to identify them as examples of the same setions in the genre set of the annual report Impliations for genre theory and genre analysis Referring to the seond main researh aim of this study, the analysis of moves and strategies disussed in this hapter suggested several fundamental ideas about the generi approah to professional disourse and about the identifiation of professional disourse in partiular. Several relevant researh impliations are provided in the following paragraphs. The theoretial importane of moves and strategies in genre struture The strutural analysis of English annual report texts originating in the Duth and British business ultures onfirmed the relevane of Bhatia s (1993) onept of rhetorial moves and strategies, as opposed to Swales (1990) onept of moves and steps. It showed that the moves inluded in the present management statements, profiles and operational reviews were based on flexible series of allowable tatis. Consistent with the researh results of, for example, Bhatia (1993) and Henry and Roseberry (2001) the allowable tatis did not our in a fixed sequene within partiular moves, but were either applied in different moves at different positions or not applied at all. Moreover, the strutural analysis highlighted the disursive

178 importane of these tatis within the professional annual report texts. Following Nikerson and De Groot (2005), it would seem as if they reflet subtle (and often also ulture-speifi) rhetorial hoies aimed at a more effetive realization of the text purpose. This an be illustrated by the strategy salutation to the reader, whih was a typial onstituent of the opening move in the introdutory Duth-English CEO s statements. Referring to the genre ontext of Duth-based annual reports (setion 4.4.4), it ould be argued that Duth text produers in partiular prefer this rhetorial strategy as a means to enhane the persuasiveness of the professional text. The theoretial importane of ross-ultural variations in genre struture Although the Duth-English and British annual reports generally represented omparable genre sets, at the level of the individual text types or annual report setions the urrent analysis also revealed signifiant ross-ultural distintions in the appliation of moves and strategies. Thus, ompanies in the two national business ultures showed ulture-speifi preferenes for the establishment of the rhetorial struture of a partiular text type. The ombination of the move speulating on future performane with strategy short-term future objetives, for instane, was partiularly prominent in the Duth-English CEO s statements and less so in the two British statement types. Consistent with Vergaro (2005) and Zhu (2000), there was some ross-ultural variation in the appliation of individual moves, espeially for the management statements. However, ulture-speifi preferenes for partiular strutural elements in the annual reports mainly related to the use of rhetorial strategies. Aordingly, this finding is in line with Bhatia s suggestion (1993) that ultural differenes between genres are most likely to arise in the allowable strategi devies used to onstrue and exploit the moves, and are less likely to appear in these onventionalized moves. Some of the ross-ultural differenes in moves and strategies ould be aounted for by ontextual aspets analyzed in the investigation of genre ontext (disussed in Chapter 4). As an illustration, the high frequeny of the reporting on leadership move within the British Chairman s statements ould be explained by the professional mandate of the Chairman, who is a representative of the shareholders and supervises the ativities of the exeutive board (inluding the CEOs). Furthermore, the frequent ourrene of the expressing sympathy move in the Duth-English CEO s statements and the British Chairman s statements ould be explained by the ommuniative purpose of these statements, whih involves the reation of a reader-writer relationship. For other ross-ultural variations, only tentative explanations ould be offered on the basis of the ulture-speifi data olleted among Duth and British informants (see setion 4.4). The theoretial importane of photographs in genre struture The addition of photographs to the traditionally text-based analysis of genre organization resulted in a more omplete view on the rhetorial struture of professional texts. In urrent business pratie, orporations heavily rely on multimodal genres when ommuniating with their internal and external stakeholders. Considering the omplementary roles of images in the onstrution of suh texts or of any multimodal text (Kress & Van Leeuwen, 2001), it would seem to be insuffiient to regard the rhetorial onstrut of a genre like the annual report exlusively from a written text perspetive. This study has shown that it is possible to inlude both text and photography in the analysis of rhetorial moves and strategies. Therefore, the traditional method for analyzing genre struture also appears to be relevant for the strutural identifiation of multimodal disourse.

179 Moreover, Bhatia s remark that pitures in a promotional genre are often used to highlight moves whih have traditionally been realized in terms of lexiogrammatial resoures (2004: 65) applied to the urrent researh materials. Rhetorial funtions of (written) text units were often intensified by doumentary images of ompany-speifi objets and senes with ongruent rhetorial values. Notwithstanding the general importane of photographs in the strutural interpretation of multimodal business genres, the results of the present analysis indiated that variation in moves and strategies aross national business ultures ourred primarily in the written text. With respet to the rhetorial elements inorporated in the photographs or in their adjoining text fragments, only one signifiant distintion ould be doumented. This finding implies that, as far as the rhetorial struture is onerned, doumentary photographs do not express ulturespeifi preferenes in the onstrution of the annual report as a genre set. In their 2003 annual reports, the Duth and British orporations tended to rely on photographs with similar strutural roles in relation to the ommuniative purposes of the diverse annual report setions. This is different from the results of the themati analysis (presented in Chapter 5), where photographi themes did mark ulture-speifi preferenes, e.g. the employee-related images that were typial to the British orporate profiles. The analytial relevane of formal riteria for the demaration of moves or strategies The interpretation of moves and strategies on the basis of both funtional and formal riteria of demaration helped to solidify the present study of genre struture (see setion 6.2.1). It allowed for a more expliit and therefore repliable researh proess whih proved to be useful in the speifi study of rhetorial strutures in Duth-English and British annual report texts. Shifts in themes, lay-out and linguisti realizations turned out to be relevant signals for the identifiation of textual and visual units with different rhetorial funtions. Aordingly, the urrent analysis of genre struture supported the findings of both Paltridge (1994) and Connor and Mauranen (1999). On the one hand, it suggests that the determination of textual boundaries depends on ognitively pereived divisions in ontent and that, as a result, the genre analyst needs to move away from the physial aspets of language [...] to how the text, as a whole, is onditioned by [text-] external onsiderations (Paltridge, 1994: 296). On the other hand, the analysis supported the notion that there are a number of linguisti lues that texts generally use to indiate internal boundaries, whih together with more obvious boundary markers suh as typographi ones, an be utilized in analysis (Connor & Mauranen, 1999: 51). The analytial relevane of interoder reliability tests in the identifiation of genre struture Drawing on prior studies by Crookes (1986) or Kanoksilapatham (2005), the urrent analysis of rhetorial genre struture was supported by an interoder reliability test. This test overed the identifiation of both textual and visual moves and strategies. Interestingly, the test results showed that the interoder agreement sores were higher for the text-based moves and strategies than for the photo-based elements. It appeared that the identifiation of the rhetorial funtion of photographs allowed for more ambiguity than the identifiation of rhetorial elements in the written text. This seemed to be true for the doumentary images in the orporate profiles and in the operational reviews in partiular. The smaller degree of interoder orrespondene for the labeling of visual moves and strategies in these annual report setions is

180 likely to have been aused by the omplexity of the images and the subjetive interpretation of visual meaning. Whereas the single rhetorial value of management portraits in the managerial forewords was unambiguous from the perspetive of the two oders (i.e. the identifying the ompany move ombined with the strategy giving fae to the ompany ), the larger number of pitures with often multiple rhetorial funtions in the genre strutures of orporate profiles and operational reviews allowed for more interpretational freedom. For example, one oder onsistently asribed the move identifying the ompany to photographs of produts in a manufaturer s orporate profile, while the seond oder onsistently referred to the move detailing the produt. Despite the fat that both oders used the same labels for the visual strategies, the interoder agreement sore for the images in this partiular profile was very low due to a different interpretation of visual moves. The analytial relevane of distinguishing omplex and simple texts The strutural analysis showed that none of the Duth-English and British text types examined ontained a strit sequene of moves or a fixed order of strategies-inmoves. Exept for a single element, suh as the move identifying the ompany at the beginning of the orporate profile or the move addressing the reader with strategy salutation to the reader at the beginning of Duth-English CEO s statements, no moves or strategies in any of the Duth-English and British texts ould be bound to fixed positions in whih they onsistently reourred. Moreover, the Duth-English and British management statements, profiles and operational reviews were extensive douments whih usually omprised a large number of text fragments with different but interrelated rhetorial funtions and whih often ontained rhetorial elements that related to elements in other text types within the annual report. Aordingly, the Duth-English and British texts were haraterized by an intriate web of moves and strategies. The ourse of ation followed in both the interoder test and the full-sample analysis indiated that the investigation of rhetorial genre struture is a very intensive researh method for texts that are marked by flexible and omplex sets of moves and strategies. As suh, the present Duth-English and British genre sets were different from, for instane, the ompat job appliation letters analyzed by Henry and Roseberry (2001) or the brief For- Your-Information letters studied by Vergaro (2005). These letters inluded a limited number of sequential moves and optional strategies with distint rhetorial funtions that ould be more easily reognized and defined. Consequently, studies of move strutures in suh onise texts provide a relatively manageable and onise piture of the soio-ognitive patterns that underlie the typial organization of professional disourse. Assuming the relevane of moves and strategies as possible indiators of these soio-ognitive regularities in disourse strutures (Bhatia, 1993), areful readings of the present Duth-English and British annual report texts resulted in lists of re-traeable moves and strategies that overed all strutural elements in the management statements, profiles and operational reviews. It should be noted, however, that the genre-based method for analyzing text struture is perhaps more onvenient and less intensive for the identifiation of moves and strategies in more onise and well-strutured texts. The analytial relevane of genre struture for reader response analyses Finally, it an be argued that the present analysis of genre features again revealed several ross-ultural differenes that ould be highly relevant in a reader response analysis of annual reports originating in different national business ultures. The signifiant ross-ultural distintions in typial moves and strategies that were

181 observed in the management statements, profiles and operational reviews ould be used as independent variables in the interultural evaluation of annual report effetiveness. For example, it would be interesting to determine if openness about the orporate strategy (i.e. found to be more prominent in the Duth-English than in the British profiles) influenes the international redibility of the orporate profiles. From a rhetorial view, moreover, it ould be fruitful to evaluate whether international stakeholders prefer or rejet a personal salutation as the opening sentene of the introdutory text in the English annual report. Having disussed the rhetorial genre struture of the Duth-English and British annual report texts in this hapter, the next hapter will report the final analysis onduted to provide an answer to RQs 2a-d (onerning the identifiation and omparison of multimodal disourse features in these texts). Chapter 7 ontains an investigation of the typial lexio-grammatial harateristis refleted in the written texts and photographs inluded in the Duth-English and British management statements, profiles and operational reviews.

182

183 Chapter 7 Charaterization of the generi lexio-grammar of Duth-English and British-English annual reports So far, the analyses of the ommuniative situation, ontent and struture of Duth- English and British annual report texts have revealed a relatively large degree of ross-ultural overlap as well as subtle ross-ultural variations for the preferred onstrut of these texts. The previous investigations also indiated that some rossultural differenes ould be explained by ontextual data (see setion 4.4, or Claes & Gerritsen, 2007). This hapter reports on a multimodal text analysis that finalizes the study of genre features in Duth-English and British annual report texts (i.e. the seond researh phase in setion 1.5). The study presented below revolves around the prototypial lexio-grammatial features or style (Swales, 1990) in the written and visual disourse of these texts. To answer the speifi researh questions, it offers ulture-based desriptions of the textual and visual lexio-grammar in the annual report texts (RQs 2a-2b), and ompares the lexio-grammar aross ultures to determine to what extent the Duth-English and British texts rely on different lexio-grammatial text features (RQ 2). The results of the lexio-grammatial analysis will again be related to relevant aspets of genre ontext (RQ 2.d.). A relatively small number of prior studies have foused on the lexiogrammatial haraterization of annual reports in partiular, e.g. some have analyzed linguisti features in letters to shareholders (Thomas, 1997) and some have investigated linguisti features in multiple non-finanial setions of the annual report (Hyland, 1998; Malavasi, 2005). In general, early researh on lexio-grammatial text features were primarily notational: they provided empirial evidene for the high or low frequeny of partiular linguisti elements in a text (e.g. Halliday, 1978). Within ESP genre theory, suh studies were ritiized for the absene of reasons as to why linguisti features were seleted and distributed in a ertain way within a given text type (Bhatia, 1993; Nikerson, 2000). Aordingly, genre-based investigations of lexio-grammatial style aim to speify the way ommunity members typially assign restrited values to various aspets of language use (they may be features of lexis, syntax or even disourse) when operating in a partiular genre (1993: 26). The present identifiation of lexio-grammatial elements in Duth-English and British annual reports will also fous on the atual disursive funtions of these elements within the rhetorial genre struture of the reports. As suh, it differs from the disussion in setion 6.2.1, where lexio-grammatial elements were referred to as means of demarating moves and strategies. In line with the analyses of genre ontext, genre ontent and genre struture (Chapters 4-6), the lexio-grammatial analysis of annual report texts examines and ompares ulture-speifi preferenes in lexio-grammatial style, employed by Duth-based and UK-based members of the expert disourse ommunity of report produers. Setion 7.1 will provide a working definition of generi lexio-grammar that is partiularly relevant for the urrent researh materials. Setion 7.2 disusses the methodology applied in the analysis of text-based and photo-based lexiogrammar; it offers details about the analytial steps taken, about the model that was used to exeute the analysis and about the interoder test that was onduted to measure the reliability of this model. The results of the lexio-grammatial analysis for the management statements, profiles and operational reviews will be presented in setions Finally, setion 7.6 ontains a onlusion and disussion.

184 7.1 A working definition of generi lexio-grammar: metadisourse Genre analysts have looked at various types of elements in the analysis of the lexiogrammatial style of partiular text types. Bhatia (1992), for instane, analyzes the pragmati use of nominals in the desription and positive evaluation of promoted goods in the produt advertising genre. And in a later study, Bhatia (1993) refers to nominals, propositional phrases and syntati disontinuities as lexio-grammatial markers of legislative provisions. In her analysis of the eletroni workplae memo, moreover, Forey (2004) fouses on the theme-rheme onstrutions in this genre to aount for differenes in meaning-making by different reader groups. A fourth example is Upton and Connor s (2001) study of the genre of the appliation letter; they examine modals and formulai expressions in order to doument the exeution of politeness strategies in this genre. Suh linguisti genre analyses have often referred to the rhetorial value of individual lexio-grammatial items in professional genres, i.e. they have often omitted a basi model systematizing and underpinning the analysis of lexio-grammatial elements with partiular rhetorial funtions. In general, this need not pose any analytial problems within relatively small-sale and/or qualitative studies of seleted linguisti forms. However, in largesale, quantitative investigations of genres, there is a need for a lear set of (rhetorial) ategories in whih the lexio-grammatial data an be organized. For the urrent analysis of lexio-grammatial style in Duth-English and British annual report texts, a relevant ategorial model was found in onept of metadisourse. Hyland and Tse (2004) explain that metadisourse is typially used as an umbrella term to inlude a heterogeneous array of ohesive and interpersonal features whih help relate a text to its ontext by assisting readers to onnet, organize, and interpret material in a way preferred by the writer and with regard to the understandings and values of a partiular disourse ommunity (2004: 157). The onept originates in Halliday s systemi-funtional definition of text types, whih distinguishes three semanti metafuntions of a text, i.e. the ideational funtion or text ontent, the interpersonal funtion or the partiipatory text potential and the textual funtion or the text-forming potential (1978: 112). Coneptualizations of metadisourse are inspired in partiular by Halliday s interpersonal and textual funtions whih refer to non-propositional, rhetorial aspets of a given text. Aordingly, metadisourse is a pragmati onept whih enters on the way in whih writers use lexio-grammatial features to onvey their personality, redibility or sensibility towards the reader, and stane towards the topi and the message in the text. It reflets how writers seek to influene the reader s understanding of the text ontent and it exposes the writer s attitude or intentions towards the text and its audiene (e.g. Crismore, Markkanen & Steffensen, 1993; Hyland, 1998a, 2000). In pratial terms, metadisourse omprises language analysis in whih words, phrases, main lauses and also puntuation and typographial marks are examined as lexio-grammatial devies potentially used to help the reader organize the disourse and evaluate the message provided by the writer (Crismore et al., 1993). The definition of metadisourse implies that its intent is two-fold: to larify text organization and establish reader-writer interation. Initially, these purposes were aounted for in metadisourse models with two major ategories labeled textual metadisourse and interpersonal metadisourse. In all models, these ategories were further divided into speifi funtions, suh as onnetives for textual metadisourse and attitude markers for interpersonal metadisourse (e.g.

185 Crismore et al., 1993; Hyland, 1998a, 2000; Vande Kopple, 1985). Although the metadisourse model has remained prinipally the same sine its introdution, its more reent version by Hyland and Tse (2004) ontains two important modifiations. First, Hyland and Tse inlude a more aurate distintion between the main rhetorial ategories. Sine the interpersonal dimension of language use is inherent to the metadisourse onept and an draw on both organizational and evaluative features (2004: 168), Hyland and Tse avoid the textual-interpersonal ontrast. Instead, they put forward a set of interative or organizational ategories and a set of interational ategories. Here, the term interative is applied to refer to the writer s management of the information flow to guide readers through the text [ ] and interational to refer to his or her expliit interventions to omment on and evaluate material (Hyland & Tse, 2004: 168). Seond, Hyland and Tse s model presents an adapted taxonomy of distintive and unambiguous sub funtions, based on prior researh of business texts and aademi texts in partiular (e.g. Hyland, 1998b, 2000). Table 7.1 presents the metadisourse model proposed by Hyland and Tse (2004). It shows that metadisursive ategories with partiular rhetorial funtions an be realized through different types of lexio-grammatial items. For instane, transitions may be realized by onjuntions or sentene-initial phrases and hedges may take the form of modal verbs or mitigations. Table 7.1:Hyland and Tse s model of metadisourse (2004: 169). Category Funtion Examples Interative resoures = Help to guide reader through the text Transitions Frame markers Endophori markers Evidentials Code glosses Express semanti relation between main lauses (e.g. additive, ontrastive). Refer to disourse ats, text stages/sequenes, topi shifts. Refer to information in other parts of the text, make additional material available. Refer to the soure of (quoted) information in other texts. Help readers grasp meaning/funtions of text ontent through explanation/speifiation. in addition, but, thus, and finally, in sum, I argue noted above, see page, in setion aording to X, (X, 1998) namely, suh as, in other words Interational resoures = Involve the reader in the argument Hedges Withhold writer s full ommitment to information in the text. Boosters Emphasize fore or writer s ertainty towards text ontent. Attitude markers Express writer s attitude towards information in the text. Engagement Expliitly refer to or build a relationship markers with the reader. Self-mentions Expliitly refer to the author of the text. might, would, perhaps in fat, definitely, it is lear unfortunately, important note that, you an see I, we, my, our Several versions of the metadisourse model have been applied in the ross-ultural analyses of professional genres. These investigations have shown that metadisourse is a useful tool for omparing lexio-grammar in genres aross ultures, yielding interesting details about ulture-speifi preferenes in linguisti style. Mauranen (1993), for instane, fouses on interative resoures suh as transitions, frame

186 markers and evidentials in her analysis of Finnish-English and native English eonomi journal artiles. She finds that Anglo-Amerian writers rely muh more on these metadisursive resoures than their Finnish olleagues. Nikerson s analysis (2000) of Duth-English and British messages within an Anglo-Duth multinational furthermore reveals a differene aross ultures with respet to partiular interational resoures. Her results indiate, for example, that the Duth- English texts ontain more boosters and olletive self-mentions than the British- English texts. A third example in whih the metadisourse model was used for the ross-ultural identifiation of a genre onerns Vergaro s investigation (2005) of interative and interational resoures in Italian and English For-Your-Information letters. Vergaro reveals that individual self-mentions and hedges in the form of modals our more often in the English texts, whereas frame markers in the form of sentene-initial illoutionary ats appear more frequently in the Italian texts. Although the ategories in Hyland and Tse s (2004) metadisourse model were onsidered useful for large-sale, ross-ultural text analyses, it was observed that his model required further larifiation with respet to the metadisursive study of multimodal professional genres in general, and of multimodal annual reports in partiular. To be speifi, details needed to be added about the (on)text-dependent nature of the identifiation of metadisourse items in a ertain text type and about the role of metadisourse in photographi illustrations inluded in a given text type. These issues are ruial for the investigation of metadisursive elements embedded within written texts and images of genres in partiular disiplines or ultures. Therefore, subsetions and elaborate on (on)text-speifi realizations of metadisourse and on metadisourse in photographs, respetively The (on)text-speifi realization of metadisourse The first issue that needs further elaboration onerns the realization of metadisursive items in genres ommonly used in or bound to partiular disiplines or ultures. The ontext-based lassifiation of lexio-grammatial items is not a new issue in the analysis of metadisourse; it is often left impliit in the analytial framework of metadisourse studies. Sine similar lexio-grammatial realizations may not have equal metadisursive value aross genres or ultures, however, the influene of ommuniative situation deserves to be highlighted. Assuming that an understanding of the pragmatis of metadisourse must be loated in the settings whih determine its use and give it meaning (Hyland, 1998a: 438), metadisourse omprises ontext-speifi or text-speifi preferenes, not only in terms of the frequeny of ourrene of ertain metadisursive items but also in the atual expression of suh items. What may be a rhetorially funtional lexio-grammatial element in one type of text may not have any suh metadisursive relevane in another text type. This point an be illustrated by examples derived from Swales study of aademi texts (1990) and from the urrent researh materials (i.e. Duth- English and British annual report texts). Swales notes that the establishment of the main researh topi at the beginning of a researh artile introdution an be realized by the phrases in Table 7.2.

187 Table 7.2: Example realization metadisourse (based on Hyland, 2000; Swales, 1990). Funtion Realizations = Evidentials in the form of unassigned ativity in the field Claiming entrality of the researh topi - The effet of has been studied extensively in reent years - Many investigators have reently turned to - Reently, there has been wide interest in In Table 7.2, the reurrent lexio-grammatial item reent(ly) is a ruial onstituent of the general evidential that is represented in eah phrase. As a disursive marker, it needs to draw reader attention to the lively, signifiant or well-established researh area that is reported on in the artile (Swales, 1990: 144). Within the Duth-English and British annual report texts analyzed here, the same lexio-grammatial item ourred relatively frequently, primarily in the rhetorial text ontext presented in Table 7.3. Table 7.3: Example realization metadisourse Duth-English and British annual reports. Funtion Realizations metadisourse Communiating operational performane details - Old Mutual and Nedor have also reently entered into a formal relationship greement (Operational review Old Mutual, 2003) - Having ompleted the imaging portfolio with the reent aquisitions, the fous (CEO s statement Philips, 2003) The instanes of reent(ly) in the annual report texts show that they, unlike the item reent(ly) in Swales example, do not atually ontribute to the rhetorial funtion of the text fragment. Contrary to the ourrene of the item in the artile introdutions, in the annual report texts these instanes are not disursive markings that need to emphasize the newness of the ativities desribed; they merely signal text-external temporal relations. Thus, reent(ly) is not a potential (part of a) metadisourse feature in these business texts, while it is essential to the realization of the metadisursive evidential in the aademi texts analyzed by Swales (1990). The lexio-grammatial item reent(ly) would be a keyword in the metadisourse study of researh artile introdutions but not in the analysis of annual report texts. The above examples show that, in addition to ommonly shared realizations of metadisourse elements (e.g. and or but for transitions and would or about for hedges), different genres may be marked by numerous highly (on)text-speifi realizations of metadisourse. This would result in metadisourse studies of an extensive size. As a result, many genre analysts have restrited their investigations of the (relative) frequenies and distributions of metadisourse items, perhaps to avoid having to examine all the typial realizations of items in the metadisursive ategories. They have restrited the sope of their researh, either by seleting a limited number of researh texts (e.g. Mauranen, 1993; Valero-Garés, 1996), by referring to existing inventories of metadisourse items (e.g. Coniam, 2004; Hyland, 1998b) and/or by hoosing partiular metadisourse ategories or funtions as the researh fous (e.g. Bondi, 2002; Dahl, 2004; Mauranen, 1993; Solbjørg Skulstad, 2005; Vergaro, 2005). With regard to the present study of metadisourse in annual report texts, it was deided that suh limitations should in priniple be exluded. Sine one of the main objetives of this study was to establish and ompare full exploratory overviews of the typial lexio-grammatial features in Duth-English

188 and British business texts, all ategories in the metadisourse model (i.e. Hyland & Tse, 2004) were used as starting points for the ross-ultural inventories. As will be disussed in setion 7.2.1, indeed the wide metadisursive sope of this study resulted in a voluminous researh design for the identifiation and labeling of instanes of metadisursive items Photographs in the analysis of metadisourse As the present genre analysis strives to aount for both text and photographs, the seond issue that is onsidered here is the role of photographs in the analysis of metadisourse. So far, this hapter has foused on a text-based desription of metadisourse, beause the metadisourse model was designed in priniple for written texts (e.g. Crismore et al., 1993; Hyland and Tse, 2004). Aordingly, the vast majority of generi analyses have only looked at the verbal realizations of metadisourse items. In his attempt to define visual metadisourse, however, Kumpf (2000) emphasizes that authorial influene extends beyond the written text and also determines the visual elements that onstitute the text. He argues that we must inquire how metadisourse affets the visual deisions and subsequent visual ontrol ( ) writers exert on their douments (Kumpf, 2000: 402). Although Kumpf s initiative to adjust the ategorization of metadisourse to multimodal disourse is important for the present purposes, his ategories of visual metadisourse e.g. attration, hunking, onsisteny are only based on textual metadisourse (i.e. Hyland and Tse s interative metadisourse) and they remain rather general; they do not yet allow for a speifi labeling of individual visual elements with partiular organizational and interational funtions. In his work on aademi writing, Hyland (2000: 119) touhes briefly upon the subjet of visual metadisourse. Aknowledging that the sienes [ ] reonfigure available linguisti resoures for making meaning through the rhetorial ombination of images and texts, he believes it is through endophori metadisourse that readers of sientifi texts an swith between written and visualized (i.e. diagram) messages. A similar relationship between the metadisourse onept and images in a professional text an be found in Fuertes-Olivera et al. (2001). Their analysis of English in multimodal produt advertisements is based on the idea that photographs involve the audiene in onstruting a range of messages for themselves (Fuertes-Olivera et al., 2001: 1303). To aount for the rhetorial signifiane of visual elements, the analysis shows that the rhetorial relation between slogans or headlines and pitures is primarily established through endophori markers. In an ad portraying a fashion model in ombination with a L Oréal make-up item, for instane, the text Milla Jovovih is wearing Quik Stik in Sand Beige refers to the woman as well as to the produt in the piture (Fuertes- Olivera et al, 2001: 1303). As suh, it puts the aent on the oherene between the qualities of the beautiful model and the added value of the produt. Even though the approah adopted by Hyland (2000) and Fuertes-Olivera et al. (2001) would appear to be highly relevant from the urrent genre perspetive, at least one inadequay an still be observed in relation to the present researh objetives. While Hyland and Fuertes-Olivera et al. fous on the rhetorial link between propositions represented in the text and in the photographi images, the present investigation also seeks to identify whether eah of these propositions in themselves ontain metadisursive elements, regardless of the rhetorial verbalpitorial link in the multimodal text. With respet to metadisourse refleted in the

189 visual propositions in partiular, a relevant analytial framework was found on the basis of Kress and Van Leeuwen (2006). Their work on meaning-making in visual images does not expliitly refer to the onept of metadisourse, but it builds a tool for visual analysis around Halliday s (1978) three metafuntional interpretations of a text (see setion 7.1 for an introdution to these metafuntions with regard to metadisourse in written texts). Kress and Van Leeuwen distinguish (1) the ideational or representational metafuntion, whih omprises the people, plaes and objets depited in the image (i.e. the partiipants ), (2) the interpersonal or interational funtion, whih inludes the reader-writer engagements, and the distane and point-of-view suggested towards the portrayed partiipants and (3) the textual or ompositional funtion, whih enompasses the informational value of the tehnial organization of partiipants (e.g. plaement, framing, reality value) (Higgins, 2003; Kress & Van Leeuwen, 2006). Comparing this taxonomy for visual images to the metadisourse onept for texts, it an be onluded that the interpersonal and ompositional metafuntions resemble the interational and interative resoures in Hyland and Tse s (2004) model. Therefore, the present investigation of visual metadisourse inorporates Kress and Van Leeuwen s interational and ompositional ategories. Consistent with the lexio-grammatial approah taken in this hapter, visual metadisourse will be studied in terms of the visual lexis and grammar of the photographs. Drawing on Kress and Van Leeuwen, this means that the fous will not only be on the individually visualized objets or subjets i.e. the lexis but also on the way in whih these elements are ombined into meaningful wholes (Kress & Van Leeuwen, 2006: 1) i.e. the grammar. Kress and Van Leeuwen argue that suh a lexio-grammatial approah to visual metadisourse gives insight in the meanings of regularities in the way image elements are used ( ) in expliit or systemati ways (2006: 1). To be able to desribe the meaningful lexiogrammatial onstrut of images, they rely on theorizations of soial semioti sign systems. For their oneptualization of the semioti sign as a devie through whih metafuntions are realized, they refer to De Saussure s linguisti definition of a sign onsisting of a signifier (i.e. the potentially meaningful sensory form) and an arbitrarily related signified (i.e. the atual meaning generated by this form). However, they also modify De Saussure s definition with referene to the philosophial notions of Peire, who laims that not all semioti signs are merely arbitrary and that some are motivated and ontext-based onstituents of meaning whih are interonneted inside the interpreter s mind (Kress & Van Leeuwen, 2006: 8). They ontinue to ombine the two semioti models by integrating Peire s triadi sheme of ions, indies and symbols together with De Saussure s system of signifier and signified. The ombination of these two semioti perspetives is shematially illustrated in Figure 7.1 (a further explanation of the Peirean signs is provided later in this subsetion).

190 Signified (meaning) Signified Signified resemblane fatual/logial onvention = ion ontiguity = index = symbol Signifier (form) Signifier Signifier Figure 7.1: Semioti models ombined (based on Kress & Van Leeuwen, 2006: 8; Lister & Wells, 2006: 72, 79). Desriptions of visual images in Kress and Van Leeuwen largely remain on the level of signifiers and signifieds, without a further labeling of the interpretative relationships between these two semioti omponents. For instane, in their analysis of the metafuntion of a realisti piture on the Ford Mondeo (automobile) website, they desribe the observed forms, e.g. a ar parked in front of a ouple, and amera work, e.g. a low-angle long shot with the ar in the foreground, before explaining the interational meaning of the image: viewers are made to look up to [the partiipants], and they are made to see them as if they notie the ar and the stylish ouple from aross the street, with envy (Kress & Van Leeuwen, 2006: 132). The theorizations on the lexio-grammar of visual images by Kress and Van Leeuwen support the urrent analysis of photographs in annual reports in two fundamental ways: (1) they offer main ategories for metafuntions in visual disourse and (2) they provide insight in meaningful visual elements on the basis of whih these metafuntions an be identified in images. Despite these basi keys to the analysis of visual metadisourse, two ruial diffiulties remain with respet to the model of visual analysis that is required for the present study of lexio-grammar in business genres. First, the major ategories of interpersonal and ompositional resoures are not further divided into rhetorial sub-funtions, whih allows only for a general and omprehensive inventory of metadisursive items in photos. And seond, it seems that the semioti model that was predominant in Kress and Van Leeuwen (2006), i.e. De Saussure s sign system of signifier-signified, will not yield a suffiiently rih and organized set of data by means of whih suh rhetorial subfuntions of visual elements an be fully explained. De Saussure s arbitrary ode of signs does not allow for a systemati and ognitive motivation of the relationship between tangible form and (rhetorial) signifiane; in metadisursive terms, it an desribe if a partiular visual element is an instane of interpersonal or ompositional/interative metadisourse of some kind, but it annot further speify or label the lexio-grammatial realization of a metadisursive element at any level of deliay. Based on De Saussure s system, for instane, the ouple in the image on the Ford website ould be identified as a visual explanation of the Mondeo lients, whih would make this ouple a ompositional/interative resoure and a visual ode gloss in partiular. In this example, however, De Saussure s system does not allow for a further speifiation of the formal realization of the visual ode gloss (suh as the label adverb for the textual ode gloss namely ). In an attempt to establish a model of visual disourse that would be as aurate as Hyland and Tse s (2004) model of text-based metadisourse, it was deided that the image-based model would be omposed of Peirean sign theory (Wollon, 1998), the model for textual metadisourse (Hyland & Tse, 2004) and the photo-related findings in the ontext analysis of Duth-English and British annual reports (setion 4.4.6).

191 Model of visual metadisourse for annual report texts In order to propose a model of visual metadisourse that overs photographs in annual reports, it is first of all neessary to return to the photo-related results of the ontext analysis in setion These results showed that Duth and British informants generally asribed eight general ommuniative purposes to the use of photographs in their English-language annual reports. Table 7.4 summarizes the pereived purposes of photographs. In general, it appears that photos were intended to take up both the readers ompetene in text interpretation and the ompany s attitude towards its own profile, produts and ahievements. In other words, they were primarily inluded to enhane the readability of the full text and to onvey a positive orporate image. Drawing on Kress and Van Leeuwen (2006) and Hyland and Tse (2004), it an thus be onluded that the ommuniative objetives of photographs in Duth-English and British annual reports onerned (a) the ompositional or interative metafuntion as well as (b) the interpersonal or interational metafuntion. Regarding the photographi purposes proposed in the interviews as the funtional explanations of metadisursive ategories, they an be used to form a link between Kress and Van Leeuwen s image-based metafuntions and Hyland and Tse s rhetorial resoures. This is illustrated in Table 7.4. Table 7.4: Overview of ommuniative funtions of photos in annual reports (based on setion 4.4.6), in relation to general ategories of metadisourse (derived from Hyland & Tse, 2004; Kress & Van Leeuwen, 2006). Metadisursive funtion Compositional/interative Readability Readability Readability Interpersonal/interational Image Image Interative+interational Readability/Image Readability/Image Readability/Image Communiative funtion To support the marketing ommuniation poliy To establish a entral theme To improve readability of the text by visualizing (abstrat) information To appeal to reader emotions To provide the ompany with a personal fae To illustrate the orporate identity To illustrate the added value of produts To emphasize high apaity and professionalism The variety of photo-related objetives informants referred to in the analysis of genre ontext (disussed in setion 4.4.6) implies that the photographs ontained diverse rhetorial sub-funtions, e.g. interative transition markers realized in the form of image omponents that supported the entral annual report theme. Table 7.5 shows a more detailed overview of the speifi metadisursive ategories or subfuntions that an be established on the basis of the pereived purposes of images in Duth-English and British annual reports. In other words, it presents a model of visual metadisourse that is grounded in theory (i.e. Hyland & Tse, 2004; Kress & Van Leeuwen, 2006) but that is partiularly relevant for the researh materials in the present genre study. Compared to the model of textual metadisourse (Hyland & Tse, 2004), the model of visual metadisourse in Table 7.5 exludes the ategories

192 of interative frame markers and endophori markers, and interational hedges. This is owing to the fat that frame markers and hedges were not represented in the photographs of the Duth-English and British annual reports. The endophori markers were left out beause (a) the ontent of one image did not expliitly refer to the ontent of another image and (b) the endophori oherene between written texts and photographs was believed to be intrinsi to the (intentional) use of images in Duth-English and British annual reports. Therefore, the explanation of the other ategories of visual metadisourse depended on the assumption that the meaning of photographs was always related to the ontents of the written texts. Table 7.5: Model of visual metadisourse, based on ommuniative funtions (setion 4.4.6) and metadisourse model of Hyland and Tse (2004). Category Funtion Examples Compositional (Interative) resoures = Help reader to interpret photographi ontent Transitions Express semanti relationship: realize Consistent portrayal of a entral theme throughout the report. employees, value produt. Evidentials Code glosses Refer to extra information soure: prove the ompany s high apaity/ professionalism (referred to in text). Exemplify/speify ideational material: - Illustrate fatual internal/external identity of the ompany represented in text. - Improve readability of the text by visualizing (abstrat) written information. - Illustrate instrumental value of produts/ servies of the ompany (referred to in text). - Support marketing ommuniation poliy (referred to in text). Portrait of award(s). Portrait of employee(s), portrait of headquarter, fatual portrait of produt, portrait of advertisement, portrait of lients. Interpersonal (Interational) resoures = Involve the reader in the photographi ontent Boosters Emphasize fore in proposition: fous Frog-view portrait of on the ompany s high apaity/ produt produed by professionalism (referred to in text). ompany. Attitude markers Engagement markers Self-mentions Express sender s attitude to proposition: - Illustrate the assoiative internal/external identity of the ompany represented in text. - Illustrate the assoiative added value produts/servies of ompany (referred to in text). Build relationship with readers: appeal to the readers emotions or attitudes. Refer to the author(s) of the text: to provide ompany with a personal fae. Portrait of employee in formal dress, portrait of smiling lient using orporate produt, frog-view of CEO/ headquarters. Close-up of manager, portrait of baby lient. Portrait of manager, portrait of employee(s).

193 Tables 7.4 and 7.5 imply that the analysis of metadisourse in photographs is more omplex than the analysis of text-based metadisourse. Contrary to metadisursive funtions or items in written texts, some metadisursive funtions or items in images may belong to multiple metadisursive ategories. This is due to the fat that different layers of meaning an be represented by the elements in an image. As is illustrated in Figure 7.2, a piture of a smiling employee in uniform whih is inluded in a orporate profile, not only visualizes the orporate identity in a fatual way (i.e. interative ode gloss), but also embodies the assoiative orporate qualities of optimism and friendliness (i.e. interational attitude marker) and the we mentioned in the written text (i.e. interational self-mention). The urrent study foused on the full metadisursive potential of photographs in annual report texts, sine it aimed to explore the visual metadisourse generally embedded in these multimodal business genres. Consequently, it essentially overed all metadisursive ategories, funtions and items represented in a partiular photograph (see Figure 7.3 for a further exemplary analysis of the piture of the Rentokil-employee). Interative ode gloss Interational attitude marker Interational self-mention Figure 7.2: Visual metadisourse ategories in photo of employee, based on Table 5 (Rentokil profile, 2003). As was suggested above, a further explanation of the model of visual metadisourse an be realized through the Peirean sign system. Peire s signs are phenomenologial in nature, i.e. they are anhored in the real world and over signifiation beyond the language boundaries (Mik, 1986). Aordingly, Peire s triadi distintion between ions, indies and symbols (see Figure 1) not only allows for a substantial or organized view on meaning-making, but it also enables a ontext-based and multimodal interpretation of disourse within and aross ultures. This supports the idea that the Peirean sign system is a helpful devie in the identifiation of visual metadisourse in annual reports originating in the Duth and British business ultures. Within the ontext of visual analysis, Peire s ion an be defined as a sign in whih the physial form observed (e.g. objets, subjets, ativities, senes) shows resemblane to what it represents or means (Lister & Wells, 2001; Wollen, 1998). In onrete terms, the ion is ontained within the image ( ) The represented has no other referent than its own representation: the image of the hair does not refer to any other hair but the hair of whih it is the image (Mitry, 2000: 91). The index is a more omplex sign, as it involves an existential onnetion between the portrayed form and its meaning, i.e. a logial ausal relationship or fatual reasoning. As suh, the interpretation of an indexial sign depends more on

194 the ontext in whih it is pereived than the signifiation of an ioni sign, whih an be meaningful in itself (Lister & Wells, 2001; Mitry, 2000; Wollen, 1998). In pitures of drought migrants, for instane, the feeling of distress an be refleted by indies in the form of faial expressions eyes ast down, blank stare and bodily position sitting with knees raised (Lister & Wells, 2001: 79). Likewise, if a amera shot shows a doll lying on stone steps in a orner of a garden, this doll is an index of the presene of an unknown little girl (Mitry, 2000: 103). Peire s third sign, the symbol, is not related to its tangible form by similarity or fatual logi, but by (ultural) onvention or law. The meaning of a symbol is assumed and highly ontextual (Mitry, 2000; Wollen, 1998). Visual symbols may be reurrent and ommon, suh as the shell-logo representing the Shell oil ompany in any kind of orporate ommuniation (Nöth, 1975). But visual symbols may also have arbitrary temporary signifiane in a reognizable instane of metaphorial meaning-making. This is the ase, for instane, when a lose-up of a ladle slowly filling and falling to the bottom of a buket is used to symbolize the refusal, misery and lethargy felt by the main harater (Mitry, 2000: 188). Or when a lose-up of a doll, preeded by a shot of a little girl taken away and a shot of searhing parents, beomes a symbol for the missing girl (Mitry, 2000). It should be noted that ions, indies and symbols are not mutually exlusive; all three aspets frequently ( ) overlap and are opresent (Wollen, 1998: 84). Obviously, neither indies nor symbols take shape without underlying observations on the ioni level. And sometimes symbols annot seem to exist without the prior exeution of indexial relationships either, for instane with regard to the symboli piture of the doll whih only makes sense if the previous shots have raised awareness of the fat that a girl has gone missing and that onsequently her parents have started looking for her or for something that might lead them to her. This line of argumentation would also seem to hold true for photographs enlosed in multimodal texts suh as the urrent annual reports. These photographs an be analyzed as ompilations of interrelated visual signs ions, indies and symbols. Again referring to the portrait of an employee in Figure 7.2, for example, it an be argued that the employee symbolizes the Rentokil-ompany, based on the notions that (a) she is related to the ompany profiled in the text and (b) that she atually works for this ompany, judging from her uniform, the tool in her hands and the aompanying text fragment. Here, ioni observations have been put together to form indexial meanings whih in turn have added to a symboli interpretation of the photo within the ontext of the text. Drawing on the semioti theory of Peire, the basi version of the model of visual metadisourse (shown in Table 7.5) an now be finalized. Peire s sign system suggests that, similar to the language-based realizations of metadisursive resoures, the realizations of visual metadisursive ategories may emerge in different forms. While the language-based self-mention, for instane, an our in the form of the personal pronoun we, the photographi self-mention may take the shape of a Peirean pitorial index. This is illustrated by Figure 7.3, whih demonstrates what the visual metadisourse model may look like with the inlusion of relevant semioti signs. The figure speifially fouses on the metadisursive explanation of the employee portrait in the Rentokil profile. It sets out the semioti interpretation of this portrait as well as the metadisursive resoures and ategories the semioti signs help to establish. Figure 7.3 indiates that this semioti interpretation involves the identifiation of visual omponents that are losely related to the writing-as-mode (text fragment adjoining the image) and to the photographi objet-as-mode (person, lothes), ativity-as-mode (working) and

195 emotion-as-mode (faial expression) (Kress & Van Leeuwen, 2001; Smelik, 1999, see setion 3.2). It should be noted that the metadisursive sheme in Figure 7.3 does not omprise the omplete model, simply beause the Rentokil image does not represent all possible ategories. Furthermore, it may be antiipated that visual metadisourse in annual report texts omprises a limited number of ioni signs: the omplexity of the funtions and meanings of photographs inside these multi-purpose and multimodal business texts is likely to give rise to a variety of metadisursive indies and symbols. Sign Ion Index Symbol Signifiation Realisti fous on Cauasian woman, smiling, head tilted, faing amera, wearing shirt + sign, handling mahine. Western woman is working, she is employed by ompany mentioned in text and named in logo on shirt, she works for its Failities Management department named in adjoining text, she is a happy/open/friendly employee. Working woman is personifiation of Rentokil ompany (in logo), Rentokil is a professional, optimisti and friendly organization. Category Funtion Examples Compositional (Interative) resoures = Help reader to interpret photographi ontent Code glosses Illustrate fatual internal/external identity of ompany presented in text. employee exemplifies orporate identity desribed in profile text = symbol Interpersonal (Interational) resoures = Involve the reader in the photographi ontent Attitude markers Illustrate the assoiative internal/external identity of the ompany represented in text. - employee enjoys her job at the ompany = index - employee represents positive qualities of orporate identity (happy, friendly) = symbol Selfmentions Expliitly refer to the author(s) or sender(s) of the text = provide ompany with a personal fae. employee is not neessarily writer, but she represents we behind ompany desribed in profile = index Figure 7.3: Model of visual metadisourse, based on ommuniative funtions (setion 4.4.7), metadisourse model of Hyland and Tse (2004) and Peirean sign system (e.g. Wollon, 1998). Setion 7.2 below elaborates on the steps taken to distinguish the formal realizations of both text-related and photo-related instanes of metadisursive funtions within the urrent researh materials. All analytial steps were based on the models of textual and visual metadisourse introdued in the previous sub-setions.

196 7.2 Analyzing metadisourse in Duth-English and British annual report texts As was determined in setion 7.1, the present analysis of lexio-grammatial style in Duth-English and British annual reports enters on the onepts of textual and visual metadisourse, where metadisourse refers to the lexio-grammatial elements that diret readers to how they should understand, evaluate and respond to propositional information (Hyland, 1998: 50). A theoretial survey indiated that the traditionally text-based model of metadisourse ould be adapted to over photographi illustrations in addition written texts. Aordingly, a model of textual metadisourse and a model of visual metadisourse were introdued that were expeted to be of partiular relevane for the study of written texts and photographs in Duth-English and British management statements, orporate profiles and operational reviews. To date, little researh has been onduted on the use of metadisursive items in annual reports in partiular. Hyland (1998b), for instane, investigates instanes of all metadisursive ategories in CEO s letters and diretors reports of UK, US and Hong Kong ompanies, while Solbjørg Skulstad (2005) examines the ourrene of a seletion of metadisursive ategories in British Chairman s statements. Although the present study builds on these exploratory analyses, its sope is extended to inlude both multimodal disourse and a larger number of annual report setions. In line with the researh proedures in the previous analyses of genre ontent and genre struture (Chapters 5 and 6), the fullsample analysis of textual and visual metadisourse omprised the manual identifiation of single items that were then inserted as data in SPSS files (see also setion 3.4). Setions and expand on the researh proedures for the identifiation of metadisursive elements in the written texts and in the photos of the Duth-English and British-English annual reports. Setion presents the full model of multimodal metadisourse that was used for the investigation of the present researh texts, i.e. it presents a model in whih both the model of textual metadisourse and the model of visual metadisourse are integrated. The integral model also ontains examples extrated from the English annual reports. Finally, setion disusses the results of the interoder test that was arried out to evaluate the reliability of the interpretation of metadisourse in the texts and photos Design for the identifiation of textual metadisourse Setion 7.1 pointed out that the analysis of textual metadisourse in the Duth- English and British annual reports was based exlusively on Hyland and Tse (2004). Their model was inluded in Table 7.1 and is integrated in the model of multimodal metadisourse disussed in setion Hyland and Tse s model was used to establish ulture-speifi overviews of metadisursive items typially ourring in the written texts of the English management statements, profiles and operational reviews. Subsequently, these overviews were ontrasted aross ultures to determine if the Duth and British business ultures relied on different metadisursive ategories or different lexio-grammatial realizations of metadisursive ategories in their onstrut of English annual report texts. The relevane of ross-ultural omparisons of textual metadisourse in professional genres is onfirmed by studies suh as Nikerson (2000), whih reveals variations in interational metadisourse aross Duth-English and British business s, or Dahl (2004), whih shows

197 differenes in interative metadisourse aross Norwegian, English and Frenh researh artiles. Neither the text-based model by Hyland and Tse (2004) nor the model of multimodal metadisourse in setion 7.2.3, inorporates speifiations of formal realizations of metadisursive ategories. As both the Duth-English and British annual reports ontained reourring words and phrases to realize partiular metadisursive ategories, however, Table 7.6 gives an impression of the types of formal text-based realizations of metadisourse that were generally observed in these texts. Table 7.6: Common formal realizations of textual metadisourse in annual report texts. Metadisursive funtion/ategory Interative transitions Interative frame markers Interative endophori markers Interative evidentials Interative ode glosses Interational hedges Interational boosters Interational attitude markers Interational engagement markers Interational self-mentions Textual realization onjuntions (thus), sentene-initial phrases (in addition). phrases or words staging text (first), illoutionary ats (we introdue). words/phrases of referene (analyzed further in). words/phrases of attribution (awards), quotations (says CEO). explanatory or speifying words/phrases (for example). passives (was ompleted), modal verbs/verbs (would, feel), mitigating adjetives/adverbs (possibly), mitigating nouns (possibility). verbs of ommitment (we believe), emphati adjetives/adverbs (substantially), emphati nouns (exellene). verbs of stane or evaluation (I respet), evaluative adjetives/adverbs (important), evaluative nouns (suess). words or phrases of engagement (you, your, take healthare ). singular pronouns (I), plural pronouns (we), group-related nouns (the Group), management-related nouns (the Board), funtion-related nouns (Chairman), nobility-related nouns (lord, sir), orporate names (Numio). To be able to gain a general overview of the textual metadisourse typially used in the Duth-English and British annual reports, the linguisti analysis in priniple was designed to enompass all metadisursive funtions, ategories and formal realizations (see Tables 7.1 and 7.6). The appliation of Wordsmith Tools 4.0 (Sott, 2006), in ombination with several qualitative researh ativities, allowed for the full overage of the metadisourse elements. First, WordSmith Tools 4.0 was used to reate wordlists for eah Duth- English and British text type involved in the present analysis (i.e. wordlists for Duth-English CEO s statements, profiles and reviews and British CEO s statements, Chairman s statements, profiles and reviews). To enable proessing in the WordSmith Tools Wordlist utility, all texts were onverted from downloaded PDF-files or sanned Word-files to simple ASCI-files. The wordlists ontained a series of words running from the most frequently used item to the item ourring least often in a partiular type of Duth-English or British annual report text. Beause the wordlists ontained an extensive number of lexio-grammatial items for eah text type (e.g. up to around 6,000 words in eah of the two lists for the operational reviews), it was deided to draw a minimum frequeny boundary for the initial detetion of possible metadisourse elements. Setting a minimum frequeny

198 boundary inluded only the most frequent and thus most ommon and relevant items in a partiular Duth-English or British text type. The boundary in the wordlist for eah text type was set at the minimum frequeny perentage of 0.01%, meaning that words ourring in less than 0.01% of the texts inorporated in a wordlist for a given text type were exluded from the analysis. Below this line, WordSmith Tools 4.0 did not offer any further frequeny perentages, beause the ourrene of items in the texts was too low. This boundary allowed for a manageable yet suffiiently long list of words. Plaing this boundary at a higher frequeny perentage ould have yielded shorter wordlists that would not have provided representative overviews of existing metadisursive elements in the annual report setions. Up to the boundary of 0.01% in the wordlists, any lexial and/or grammatial element that ould be (part of) a metadisurive element in the written text was highlighted for further examination. It was deided that a small number of metadisursive realizations would be exluded from the present analysis: endophori markers in the form of ontextualizing phrases (e.g. with these objetives in mind ), engagement markers in the form of asides (e.g. as you will agree ) and typographi elements (e.g. exlamation marks or bullets). These types of metadisursive realizations were often demarated by ommas or had been erased during the proess of text onversion. Therefore, they were extremely diffiult to trae on the basis of the WordSmith Tools wordlists. A manual hek of these formal realizations of textual metadisourse revealed that they ourred in relatively few management statements, profiles and operational reviews. Sine the metadisursive ategories to whih these rare types of formal realizations belong were still represented in the analysis and sine these types of realizations were onsistently omitted from the given ategories, their absene is unlikely to have had a serious impat on the final researh results. Further examination of possible metadisourse items took plae by separately inserting the items highlighted in the wordlists in the WordSmith Conordane appliation. Similar to many other genre studies on metadisourse (e.g. Bondi, 2002; Coniam, 2004; Nikerson, 2000), onordaning was inluded to determine if an instane of a given lexio-grammatial item was either a metadisursive or a syntatial element. In situations where the highlighted item involved a rhetorially funtional element, the onordane was used to settle its metadisursive label and realization type (e.g. hedge modal verb, mitigation, or passive). The labeling of metadisursive features was based on Hyland s list of ommonly applied features (2000), a few studies related to orporate reports (Hyland, 1998b; Solbjørg Skulstad, 2005) and the present researh materials. To obtain a olletion of metadisursive features that would be as large as possible within the limitations of this study, eah highlighted item that ould be extended linguistially was provided with an asterisk to generate the onordanes of all its linguisti forms ourring in the text type onerned. For instane, the highlighted items believe and suess were inserted in the Conordane appliation as *believ* and *suess* to get onordane results for believebelieves-believable-unbelievably and suess-suessful-suessfullyunsuessful. This approah led to an overview of onordanes for words that would either re-appear later in the wordlist (as less frequent items) or that would be part of the set of least frequent words that were not adopted during the initial identifiation of researh items. The ommon lexio-grammatial elements that were found to have a rhetorial funtion in ases of a partiular text type were ategorized aording to the text-based metadisursive funtions of Hyland and Tse s model

199 (2004). The verb believe, for instane, was plaed in the ategory hedge verb of non-ommitment or in the ategory booster verb of ommitment, depending on the text ontext in the onordane. The adjetive suessful was generally ategorized as attitude marker evaluative adjetive Design for the identifiation of visual metadisourse It was argued in setion that a ombination of theorizations by Hyland and Tse (2004) and Kress and Van Leeuwen (2006) on the one hand and empirial data on genre ontext (setion 4.4.6) on the other, would yield a lassifiation system by means of whih the rhetorial meaning of the observed visual lexio-grammar ould be defined. The model of visual metadisourse was illustrated in Table 7.5 and is integrated in the model of multimodal metadisourse in setion Similar to the establishment of metadisursive elements in the written texts of the annual reports, it might be expeted that orporations in the Duth and British business ultures maintain different ways of realizing visual metadisourse. This idea is supported by studies suh as Paek (2005), who finds that signifiantly more Korean than US newspaper ads inlude elebrity endorsers to provide evidene for (i.e. as interative evidentials for) the quality of orporate produts. Therefore, the present study fouses on the desription and omparison of metadisursive ategories and lexiogrammatial realizations of metadisourse in Duth-based and UK-based annual reports. In order to omplement the model of visual metadisourse introdued in setion and integrated in the model of multimodal metadisourse in setion 7.2.3, Table 7.7 shows formal visual realizations of metadisourse that re-appeared relatively often in the urrent management statements, profiles and operational reviews. The labeling of these formal visual realizations are grounded in Peire s triadi sign system (e.g. Wollen, 1998; see setion 7.1.2). Table 7.7: Common formal realizations of visual metadisourse in annual report texts. Metadisursive funtion/ategory Interative transitions Interative evidentials Interative ode glosses Interational boosters Interational attitude markers Interational engagement markers Interational self-mentions Visual realization Portraits of employees (symbols), photos of produts (symbol). Photo of award (index related to external soure), portrait of lient next to ase study (index related to external soure). Portrait of employee (symbol), photo of headquarter (index), photo of promotional material (index), CEO portrait (symbol). Photo of produt-in-frog-view (symbol), photo of award (symbol related to ompany). Portrait of happy lient (index related to produt; symbol related to ompany) a, photo of use produt (index related to produt; symbol related to ompany). a Portrait of manager-in-lose-up (symbol), photo of baby (symbol). Portrait of divisional manager (index). Note. a Example photo of offee mahine, end produt of Invensys: indexial meaning = end produt failitates the reation of a offee break that is enjoyed by professional ustomers (indies = happy ustomer is result of relevane produt); symboli meaning = Invensys suessfully fulfils the offee needs in the daily life of professional ustomers (symbols = ustomer represents the happy-making apaity of the ompany and produt represents the usefulness of the existing ompany).

200 In order to aquire a broad and representative view on the rhetorial lexio-grammar in the doumentary images of Duth-English and British annual reports, the analysis of visual metadisourse needed to over as many metadisursive instanes and realizations as possible. This was ahieved by inluding visualized details in the identifiation of metadisursive items represented in all the seleted photographi modes of disourse (see setion 3.2). Drawing on Kress and Van Leeuwen s (2006) visual modes of disourse and on Smelik s (1999) image omponents, the semiotisbased investigation of visual metadisourse foused on (a) persons, lothes, buildings or equipment to disern metadisursive items in the objets-as-mode, (b) suggested human behavior or movement to observe metadisursive items in the ativity-as-mode, () interrelated objets onstituting the setting to find metadisursive items in the sene-as-mode, (d) faial expressions or posture to disern metadisursive items in the emotion-as-mode and (e) lose-ups shots or lowangle views to trae metadisursive items in the framing-as-mode. Peire s triadi sign system of ions, indies and symbols enabled an organized and onerete analysis of the metadisursive potential embedded in the various image omponents, starting at the most manifest meaning-level in the piture and building up to its most arbitrary interpretation. In the searh for visual metadisourse, the initial step was to doument per image omponent the fatual appearanes of phenomena in the piture, i.e. their ioni refletions. Next, the ioni signs were interrelated and related to the text-ontext surrounding the piture to determine per omponent whether it refleted a ontiguous or indexial meaning that evolved logially from this interonnetion of visual and/or textual signs. Finally, both ioni and indexial interpretations were used to see if they provided the basis for the highly ontextual meanings with long-term or temporary arbitrary value, i.e. visual symbols. The degree of diffiulty aused by these interrelations varied per photograph. In ase higher-level meanings ould be based on straightforward ioni representations (e.g. an employee symbolizing the orporate identity), this would require less ognitive effort than when these meanings had to be built on ioni representations that were part of visual metaphors (e.g. an arobat symbolizing the orporate identity). Figure 7.4 offers an expliit overview of details obtained through the semioti analysis of image omponents in the photograph of a Duth orporate exeutive.

201 Category omponent Main people/objet Position photo Emotions Ativity Clothing Environment Frame size/ amera view Ions in photograph Male, almost bald, glasses. One-page photograph, man in entral/left position. Slightly squinted eyes, smile slightly upward on right, lightly rossed hands. Sitting still behind desk, looking straight at amera, paperwork and planning put aside. Light shirt, dark tie, dark suit, rimless glasses, large wath. Desk, paper + pen, eletroni planner, fruit bowl, painting with minimalist reation by Immenee. Medium shot, low angle (just below eyeheight). Ioni meaning Male adult, individual. Heavy fous on man. Man smiles and sits almly. Man does administrative work, addresses the amera lens. Man wears formal, professional lothing, with a few modern aspets. Workplae, deorated with homely and modern objets = personal touh. Realisti, relatively lose look from lower position. Category omponent Main people/objet Position photo Emotions Ativity Clothing Environment Frame size/ amera view Indies in photograph Male adult, individual, CEO s statement in TPG annual report. Fous on man, CEO s statement in TPG annual report. Man is alm, smiling with a light smirk, CEO s statement in TPG annual report. Man does offie work, addresses amera lens, CEO s statement in TPG annual report. Formal, professional lothing with some modern elements, CEO s statement in TPG annual report. Professional workplae, personalized by art with soial nature and homely aessories, CEO s statement in TPG annual report. Realisti, relatively lose view up to the waist, from low angle, CEO s statement in TPG annual report. Indexial meaning Man is Chief Exeutive Offier of ompany publishing annual report, he is sender of text. As CEO, man plays a major role within TPG ompany. CEO is optimisti, satisfied. He expresses onfidene and reassurane towards viewer/reader. As ative manager of the ompany, CEO expresses transpareny and sinerity towards the viewer/reader. CEO is professional and onventional, but with a sense of what is modern/fashionable. The CEO s offie is marked by professionalism and modernism as well as oziness and tenderness. Viewer/reader is fored to look up to the CEO in a natural way, from other side of desk.

202 Category omponent Main people/objet Position photo Emotions Ativity Clothing Environment Frame size/ amera view Symbols in photograph Funtion CEO, sender of the text, CEO s statement in TPG annual report, TPG. Major funtion CEO, sender of the text, CEO s statement in TPG annual report, TPG. Funtion CEO, CEO is happy/ satisfied and reflets onfidene/reassurane, CEO s statement in TPG report, TPG. Funtion CEO, CEO is open/honest to audiene, CEO s statement in TPG annual report, TPG. Funtion CEO, onventional but fashionable lothes, CEO s statement in TPG report, TPG. Funtion CEO, offie is business-like and artisti but ozy and tender, CEO s statement in TPG annual report, TPG. Funtion CEO, looking up to the CEO from the other side of the desk, CEO s statement in TPG annual report, TPG. Symboli meaning CEO is individual personifiation of TPG ompany. CEO is most essential personifiation of TPG. TPG is an optimisti ompany with onfidene in the future. TPG is a transparent and open-minded ompany. TPG is a traditional ompany with an eye for modern trends. TPG is professional business whih at the same time has a modern and soiable nature. Like its CEO, TPG is a powerful ompany that ompels respet and admiration. Figure 7.4: Visual analysis of CEO portrait, based on visual omponents and semioti signs (TPG CEO s statement, 2003). The researh results shematized in Figure 7.4 show that a semioti analysis of seven image omponents aused a omplex network of meaning-making. They also reveal that signifiation emerged gradually, starting at the ioni level (i.e. what was pereived) and possibly evolving into a sign at the symboli level (i.e. what did these pereptions tell the reader about the ompany). It is this gradual harater of meaning-making that allowed for a subtle study of different realizations of metadisursive funtions and ategories in photos. Sine a detailed desription of the ognitive interpretation of images (as in Figure 7.4) would require enormous spae, it was deided that results setions should only inlude the final outome of the semiotis-based study of visual metadisourse; these setions only indiate whether the metadisursive realization onerned a visual ion, index or symbol. The semioti signs and their meaning, whih were doumented for the different image omponents, helped to define the metadisourse in the doumentary photographs, i.e. they generated and onstituted the metadisursive elements in these photos. This was already demonstrated in the identifiation of visual metadisourse in the Rentokil photo in Figure 7.3 (setion 7.1.2). It an again be explained with regard to the CEO portrait analyzed in Figure 7.4. The links between semioti signs and metadisursive ategories or funtions within this portrait are expliated in Table 7.8.

203 Table 7.8: Semioti realization of visual metadisourse in CEO portrait. Category metadisourse Interative Code gloss Interational Attitude marker Interational Self-mention Funtion Illustrate the fatual internal/external identity of the ompany represented in text = Exemplifiation of the ompany that is the topi of disussion in the text. Illustrate the assoiative internal/external identity of the ompany represented in text = Evaluative qualifiation of the people behind the organization and the ompany itself. To provide the ompany with a personal fae = Diret referene to the writer of the text (the I ). Examples + semioti realization CEO is ompany = symbol - CEO enjoys job at ompany = index - CEO represents positive qualities of ompany (optimisti, onfident, open, professional, modern, soiable, and powerful = symbols CEO is writer = index The weaving together of the semioti or lexio-grammatial onstrut and the metadisursive value of a photograph was partiularly useful with regard to the ross-ultural omparison of annual reports. Sine it provided a detailed view on how an image ontributed to the organization of and the writer s stane in the full multimodal annual report text, it also enabled the detetion of subtle variations in the lexio-grammatial realization of partiular metadisursive funtions. This an be illustrated by a final referene to the pitures of the Rentokil-employee (in Figure 7.2) and the TPG s CEO (in Figure 7.4). Although they are single images extrated from different text types (i.e. a British profile and a Duth-English management statement), a omparison of these images yields findings that ould also be enountered in ross-ultural ontrasts. In partiular, a omparison shows how the same metadisursive funtions and ategories an be realized by different semioti onstruts. Both pitures omprise a symboli ode gloss foused on the identifiation of the ompany and an indexial self-mention referring to the orporate people behind the text. While these metadisursive items are realized through a visualization of the ative workfore in one piture, however, they are established through a formal management portrait in another Integrated model of textual and visual metadisourse for annual reports Although it was argued that the metadisursive ategorization in the model of Hyland and Tse (2004) an be onsidered as appropriate for the analysis of lexiogrammatial style in written texts and photographs (see setions 7.1 and 7.1.2), readings of prior genre studies and a preliminary examination of the urrent English researh texts indiated that the funtions and examples in this model ould be further defined. An initial attempt to speify the funtions and examples in the metadisourse model was further improved during the interoder reliability test (to be disussed in setion 7.2.4). The modifiations implemented in Hyland and Tse s (2004) original version of the model primarily aimed to enhane the relevane of the metadisursive funtions as riteria for the identifiation of metadisursive elements in the written texts and images of the annual reports. In other words, the adapted

204 funtions and examples were intended to diret the present researh fous better, i.e. they were to be more helpful in determining what to look for in the multimodal researh materials. With regard to text-based metadisourse, adjustments in funtions and examples were inspired mainly by Barton (1995), Crismore and Farnsworth (1990), Crismore, et al.(1993), Dahl (2004), Hyland (1998b, 2000, 2007), Mauranen (1993), Nikerson (2000), and Solbjørg Skulstad (2005). Some modifiations in examples were also derived from the Duth-English and British annual report texts themselves. As was outlined in setion 7.1.2, photo-based adaptations to the originally text-based model were realized with referene to theorizations by Kress and Van Leeuwen (2006) and to researh results of the analysis of genre ontext (see setion 4.4.6). Furthermore, the urrent model of visual metadisourse was inspired by prior analyses of photographs in annual reports (e.g. David, 2001; Jameson, 2001; MKinstry, 1996; Preston et al., 1996). Table 7.9 shows a ompilation of the revised version of the text-based model of metadisourse by Hyland and Tse (2004) and the photo-based model of visual metadisourse introdued in setion Aordingly, Table 7.9 presents an integrated model of multimodal metadisourse. Table 7.9: Revised model of metadisourse for Duth-English and British annual report texts. Interative resoures: help to guide reader through the text/ help reader to interpret photographi ontent Category/label Funtion/Identifiation Examples TEXT Examples PHOTO Transitions Transitions signal relations in the writer s thinking, helping readers to interpret links between different ideas of the writer. TEXT: Transitions express a semanti relation between two main lauses. A summing-up in the form of noun phrases (renewals and settlements), -ing phrases (strengthening and developing) or infinitive phrases (to build and to foster) is grammatial and not metadisursive. Transitions an omprise onjuntions (and, but), sentential adverbs (nevertheless, onsequently) or prepositional phrases (due to, despite). PHOTO: Establish a entral theme throughout the annual report. In addition, but, thus, and, while, however, also, to that end, so that, as, although, as a result, as (onnetion between main lauses). Consistent portrayal of employees, onsistent portrayal of added value produt, onsistent portrayal of ompany s loal ativities.

205 Table 7.9 ontinued. Interative resoures: help to guide reader through the text/ help reader to interpret photographi ontent Category/label Funtion/Identifiation Examples TEXT Examples PHOTO Frame markers Frame markers refer to disourse/ illoutionary ats and sequenes/stages in the text. Frame markers indiate whih rhetorial at the writer is performing through the text (I argue here) and guide readers in interpreting the text struture (firstly, in summary). My purpose here is to, I am pleased to report, I would like to pay tribute to, I would like to thank, I will now disuss, In onlusion, finally. Endophori markers Endophori markers omprise loational referenes to information in other parts of the text (as we said above), in other texts inside the base doument (as outlined in the CEO s review) or in texts other than the base doument (as outlined in the CEO s strategy presentation). Endophori markers do not always ontain expliit referenes to other text fragments; they may also establish an interpretative or ideational ontext and as suh prepare the reader for the information that is to follow (When we take unertainty into aount, the results hange dramatially ). noted above, set out in the finanial review, more info may be found on, see Fig., in setion 2, - Refleting the opportunities outlined before, - From a finanial perspetive, Evidentials Evidentials omprise external soures referred to for authority. TEXT: Evidentials PHOTO: refer to the soure of Emphasize the information drawn ompany s from another text high apaity/ (Einstein laimed professionalism that ). To inrease the (referred to in fore of the argument, text). writers sometimes assign authority to themselves (as a smoker, I am offended by ). Aording to X, Z states, (Y, 1990), said Arne Svendsen, a Produtionoordinator at Arla (lient), Gerard Kleisterlee (CEO) noted Portrait of lient(s) next to a asestudy, portrait of award(s).

206 Table 7.9 ontinued. Interative resoures: help to guide reader through the text/ help reader to interpret photographi ontent Category/label Funtion/Identifiation Examples TEXT Examples PHOTO Code glosses Code glosses introdue the elaboration of what has been said, in order to ensure that the reader grasps the writer s intended meaning. Code glosses omprise the reformulation or exemplifiation of information provided. TEXT: Reformulation aims to speify the information from a different perspetive and to reinfore the writer s intention (in other words, in partiular, known as, indiating that). Exemplifiation aims to larify the information by referring to an example or typial ase (for example, like, suh as). If linguisti items refer to expliit onsequenes or onlusions from what has been said, they do not have a metadisursive funtion but a syntatial funtion (In 2003 we made progress with our program to ut osts, exeeding our ambitious target ). PHOTO: - Illustrate the fatual internal/ external identity of the ompany presented in text. - Improve readability of the text by visualizing (abstrat) written information. - Illustrate the instrumental value of produts/ servies of the ompany (referred to in text). - Support the marketing ommuniation poliy (referred to in text) ( ), - -, i.e., espeially, in partiular, speifially, or, namely, e.g., for instane, say, inluding. Portrait of employee(s), portrait of headquarter, fatual portrait of produt, portrait of advertisement, fatual portrait of operational ativity in 2003.

207 Table 7.9 ontinued. Interational resoures: involve the reader in the argument/ in the photographi ontent Category/label Funtion/Identifiation Examples TEXT Examples PHOTO Hedges Hedges enable writers to withhold full ommitment to the truth-value of statements/ propositions. They allow the writer to retain redibility by avoiding diret responsibility for the information in the text (might, perhaps). A speifi type of hedge by means of whih the writer an distane both himself and the reader from the text is the passive. Passives mitigate fae threats = defoalization (Beause delivery time was delayed again Beause you delayed ). would, it is possible, it is likely, ould, there is good reason to believe, we now see sope for, a ontrat was obtained from, the merger has been ompleted, the sale was announed, generally, some, a number of, slightly. Boosters Boosters amplify the message in the text, i.e. they emphasize degree instead of quality. TEXT: Boosters enable writers to express full ommitment to the truthvalue of statements/ propositions. They emphasize fore or the writer s ertainty in the textual information and, as suh, allow the writer to express onfidene and deisiveness. They an often be replaed by really or a great deal (We made good progress ). PHOTO: Emphasize the ompany s high apaity/ professionalism (referred to in text). In fat, definitely, it is lear that, obviously, we are onfident that, we expet that, we believe (we are in a great business), I know, signifant(ly), partiularly (important), more than, demonstrably, we are determined, strive for, we are wellplaed to, will (promise). Frog-view of building onstruted by ompany, frog-view of stroller produed by ompany s hemials.

208 Table 7.9 ontinued. Interational resoures: involve the reader in the argument/ in the photographi ontent Category/label Funtion/Identifiation Examples TEXT Examples PHOTO Attitude markers Attitude markers express the writer s affetive values toward the information in the text/photo. TEXT: Attitude markers ommuniate the writer s feelings of surprise, importane, obligation, agreement, et. They do not reflet the writer s ommitment to propositions or the general publi opinion. Attitude markers an omprise attitude verbs (disagree), modals (must), sentene adverbs (interestingly), typography (!) and adjetives (areful). PHOTO: - Illustrate the assoiative internal/external identity of the ompany presented in text. - Illustrate the assoiative added value produts/ servies of the ompany (referred to in text). should be, unfortunately, surprisingly, disappointing (deision), I hope, I regret, hallenging (sales performane), healthy (levels of profitability), satisfatory, suessful, notable. Portrait of employee in formal dress, portrait of happy lient using orporate produt, portrait of end produt that is ruial to daily life of ustomer. Engagement markers Engagement markers expliitly refer to or aim to establish a relationship with the reader. Readers are expliitly addressed, either to fous their attention or inlude them as disourse partiipants. TEXT: Engagement markers inlude diret addresses (you), imperatives/ diretives (reall), real questions answered later in the text, rhetorial questions, tag questions and asides (as you will agree). PHOTO: Appeal to the readers emotions or attitudes. note that, onsider, you an see that, dear shareholder, to our shareholders, let me now go through. Close-up of manager (writer of text), portrait of baby lient. Self-mentions Self-mentions expliitly refer to the atual writer of the text. TEXT: Writers may use PHOTO: pronouns, referenes to the To provide (type of) organization the he/she is part of, referenes ompany to his/her professional with a funtion or orporate personal names. fae. I, we, my, our, us, mine, the ompany, it, its, on behalf of my olleagues on the board I, HSBC. Portrait of manager, portrait of employee(s).

209 7.2.4 Interoder reliability tests for metadisourse in texts and photographs In line with the proedures of the preeding text analyses (Chapters 5 and 6), an interoder test was onduted to measure the reliability of the identifiation and labeling of metadisursive elements. In previous hapters, the text-related and photo-related items were investigated simultaneously by the same oders. Sine the analyses of metadisourse in written texts and photographs eah rested on fundamentally different designs and disiplinary insights, however, it was deided that two reliability tests with two different pairs of oders would be required here. Therefore, the results of the interoder tests for textual metadisourse and for visual metadisourse are disussed separately. Interoder results for metadisourse in texts The interoder analysis for metadisourse in texts was again onduted by two oders. The first oder was the author of this thesis, while the seond oder was a graduate in Business Communiation studies with a high-level profiieny in English (and who had not been involved in the ontent and struture analyses). Table 7.10 douments the extent to whih interoder orrespondene was reahed for textual metadisourse, i.e. in terms of Kappa values and mean interoder agreement perentages. Table 7.10: Evaluation of interoder agreement for metadisursive elements in the written texts. Setion/ text type Qualifiation Mean SD Min. Agreement Kappa max. % Managem.statements Almost perfet Profiles Substantial Operational reviews Substantial Cohen s Kappa values in Table 7.10 indiate that interoder orrespondene for the qualifiation of textual metadisourse ranged from substantial (.77) to almost perfet (.82). The mean Kappa value was substantial (.79) (Rietveld & Van Hout, 1993). These values suggest that the oding system for metadisursive items in the written texts of the Duth-English and British annual reports was a reliable instrument. Interoder results for metadisourse in photographs The interoder reliability analysis for the visual metadisourse was more laborious than any of the other interoder tests in the present study beause of the more subjetive meaning-making proesses involved. This interoder analysis partiularly foused on the lexio-grammatial realizations or semioti signs ontaining visual metadisourse. Sine the identifiation of elements of visual metadisourse primarily depended on the interpretation of these signs, the interoder reliability test evaluated the labeling of ions, indies and symbols rather than the oding of metadisursive funtions (whih were inherent in the semioti meaning of signs). Two oders were involved in the interoder reliability test for metadisourse in doumentary photos within annual reports. The author of this study was the first oder; the seond oder was a Cultural Studies student who had taken ourses in visual analysis and had a high level of English profiieny.

210 Coder 1: 1) Main people/objet 2) Position in photo Coder 2: 1) Main people/objet Ioni meaning Indexial meaning Symboli meaning Group of three men, 2 white and 1 blak. Fous on multiultural group of three men. Men in group are employees of Old Mutual-South Afria. Multiultural group of Old mutual employees they work together on the basis equality. Group of employees is personifiation of Old Mutual-South Afria. The multiultural, equal and ooperative workfore typifies Old Mutual. Ioni meaning Indexial meaning Symboli meaning Three men, 2 white and 1 blak. Three men are employees of Old Mutual-South Afria. 2) Position in photo Fous on three men. Colleagues work together as equals. Three employees are a symbol of Old Mutual-South Afria. Old Mutual values equality and ooperation. It is an international ompany and employs people from diverse ethni bakgrounds. Results omparison: - People-omponent: full agreement for the ioni, indexial and symboli meaning of signs; - Position-omponent: disagreement for the multiultural quality in the ioni and indexial meaning of signs + disagreement for the international quality in the symboli meaning of signs. Agreement for the ioni fous on the male persons, the indexial qualities of equality and ooperation for the workfore and the symboli qualities of equality and ooperation for Old Mutual. Figure 7.5: Comparison of interoder results for visual analysis. In the formal interoder reliability test, the oders filled in shemes ontaining the ioni, indexial and symboli interpretations of eah image omponent inluded in Figure 7.4. The definitive omparison of shemes was based on keywords in the desriptions of ioni, indexial and symboli meaning. For eah oder, these keywords were marked. Next, they were heked for interoder agreement or disagreement and missing ases (where one oder presented an extra semioti value). Figure 7.5 illustrates this proess of marking and omparing. The results for interoder agreement at all three levels of semioti signifiation are presented in Table It shows Kappa values and mean interoder agreement perentages.

211 Table 7.11: Evaluation of interoder agreement for metadisursive elements in the photographs. Setion/ text type Qualifiation Mean SD Min. Kappa max. Manag.statements: Ion Index Symbol Profiles: Ion Moderate Index Symbol Moderate Operation. reviews: Ion Agreement % Index Symbol Table 7.11 reveals that interoder agreement sores for the qualifiation of visual metadisourse in the annual reports were low. For most types of signs, Kappa values ould not be alulated or showed negative sores. Kappa values that did emerge from the interoder test were moderate (i.e..42 and.47) (Rietveld & Van Hout, 1993). In general, the interoder analysis indiated that the semiotis-based oding system was the least reliable of all oding systems used so far in the identifiation of multimodal genre features. This suggests that the photo-related results in setions were relatively less stable or reliable. 7.3 Metadisourse in management statements This first results setion desribes and ompares the metadisursive elements that ourred in the written texts (setions 7.3.1) and in the photographs and adjaent text fragments (setion 7.3.2) of the Duth-English and British-English management statements Metadisourse in the running texts of the management statements Based on Appendix VII, whih ontains all the frequeny sores for textual metadisourse in the management statements, a list ould be made of the metadisourse elements that ourred in the written texts of 50% or more of these statements. This list is shown in Table As for the interative resoures, the management statements typially inluded lexio-grammatial elements onneting ideas, announing the disursive goals of the text and referring to additional information inside the text itself or in other texts within the annual report. Table 7.12 also indiates that interational resoures were partiularly popular in the written texts within the Duth-English and British management forewords. These texts primarily ontained lexio-grammatial elements defoalizing and mitigating the message, enforing the message, evaluating the message and establishing readersender interation.

212 Table 7.12: Typial metadisourse elements statements. in running text of 50% of management Metadisourse Du-E CEO n=20 Br-E CEO n=18 Br-E Chair n=16 Count % Count % Count % Total % χ 2 (df = 2) Interative: Transition Conjuntion Sentene-initial Frame marker Text stager 11 b <.05 Illoutionary at Endoph.marker Code gloss Interational: Hedge Passive Mitigation (modal) verb Booster Verb of ommitm. Emphati Attitude Evaluat. adj/adv Evaluat. noun Evaluat. verb Engage.marker a <.05 Self-mention Corp.name pers.sing.pron pers.plur. pron Funtion Management b <.05 Group-rel.nouns Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. Aording to the results in Table 7.12, Chi-Square tests revealed three statistially signifiant results for the Duth-English and British management statements. First, there was a signifiant over-representation in the number of Duth-English CEO s statements that inluded frame markers in the form of text staging signals, suh as firstly or finally. The Duth-English statements showed a relatively high preferene for the expliit organization of textual information. Seond, interational engagement markers were signifiantly under-represented in British CEO s statements. Thus, relatively few British CEO s statements inluded lexiogrammatial elements (in)diretly addressing the readers of the text. Finally, a signifiantly large number of British Chairman s statements ontained selfreferenes in the form of management-related nouns; the British Chairman s p

213 statements often inorporated a narrator perspetive representing the group ompany s board of diretors Metadisourse in photos and text fragments of the management statements Sine the photos in the management statements generally inluded a limited number of metadisursive realizations, all results of the visual analysis have been integrated in Table This table shows that the metadisursive elements observed in the Duth-English and British management statements typially onerned visual ode glosses in the form of indexial and symboli illustrations of the ompany or its ativities. Portraits of managers presumably representing the writer(s) of the statement and therefore personifying the ompany were partiularly popular. In addition, the table indiates that visual metadisourse often involved attitude markers in the form of indies and symbols highlighting the positive values of the ompany or its produts, in partiular within the British CEO s statements. These visual attitude markers espeially involved values of optimism and professionalism, primarily based on photos of smiling managers in formal lothing. Table 7.13: Frequeny sores metadisourse elements in photographs of management statements. Metadisourse Du-CEO Br-CEO Br-Chair n=17 n=16 n=14 Count % Count % Count % Tot. % Interative: Transition onjuntion Consistent theme Code gloss illustrate id/ produt/ market Fous CEO/ 10 a Chairman (symbol) 0 Board members (symbol) Fous CEO + board 6 b members (symbol) Corporate produt in use (index) Corporate workplae inside/ outside (index) Corporate 0 a b employees (symbol) Corporate market/ b marketing (index) Aommodation (symbol) Corporate operations (ion) χ 2 (df=2) p.06 <.05 < <.001 <

214 Table 7.13 ontinued. Metadisourse Du-CEO Br-CEO Br-Chair n=17 n=16 n=14 Count % Count % Count % Tot. χ 2 p % (df=2) Interational: Booster emphasize high apaity Impressiveness produt/servie (symbol) Attitude evaluate id or produt Optimism, onfidene, enthusiasm (symbol) Sinerity, openness 14 b a <.05 (symbol) Business-like/ reserved nature (symbol) Professionalism, seriousness (symbol) Eye for style/ modernism (symbol) Friendliness, soiability (symbol) Involvement, ommitment to tasks (symbol) Powerfulness, respetability (symbol) Multi-gender nature b <.05 (symbol) Male-oriented nature (symbol) Multi-ultural b <.05 nature (symbol) Global nature b <.05 (symbol) Grandeur, wellestablished nature (symbol) At servie of stakeholders (symbol) Added signifiane, quality produt/ servie (index) Produt auses positive emotional state (index)

215 Table 7.13 ontinued. Metadisourse Du-CEO Br-CEO Br-Chair n=17 n=16 n=14 Count % Count % Count % Tot. % Interational: Self-mention sender Fous CEO/ 10 a Chairman (index) 0 Board members (index) Fous CEO + board 6 b members (index) Engagement marker appeal Intimay with ompany (symbol) χ 2 (df=2) p <.05 <.05 5 a b <.05 Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. Compared to the British management statements, there was a statistially signifiant under-representation in the number of Duth-English statements with visual ode glosses in the form of individual CEO portraits symbolizing the orporate identity. Not surprisingly, this signifiant ross-ultural distintion was also found for indexial self-mentions in the form of pitures of individual CEOs. Perhaps these signifiant results were due to two other ross-ultural differenes, whih omprised a signifiant over-representation in the number of Duth-English statements ontaining symboli ode glosses and indexial self-mentions in the form of images of omplete management teams. Relatively many Duth-English CEO s statements showed a preferene for orporate identifiation and writer identifiation on the basis of portrayals of the management olletive. Next, it was observed that ertain metadisursive realizations were relatively prominent in the photos of the British CEO s statements. These statements showed statistially signifiant over-representations for visual ode glosses identifying the orporate markets or marketing ativities and for visual attitude markers symbolizing the ompany s multi-gender nature (i.e. photos with both sexes), its multiultural nature (i.e. photos of people from multiple ethni bakgrounds) and its global nature (i.e. photos of employees/lients from multiple ultures). Also, signifiantly more British CEO s statements than Duth-English CEO s statements inorporated images of employees in the symboli illustration of (i.e. as visual ode glosses for) the orporate identity. Finally, the signifiane tests revealed two ross-ultural distintions between the Duth-English CEO s statements and the British Chairman s statements. On the one hand, it was found that a signifiantly larger number of Duth-English CEO s than British Chairman s texts inluded visual attitude markers in the form of orporate symbols of sinerity. It was signifiantly less ommon in Chairman s statements to portray managers looking straight into the amera lens, and thus at the reader of their texts. On the other hand, the tests showed that signifiantly more British Chairman s statements than Duth-English CEO s statements ontained visual attitude markers in the form of orporate symbols of

216 intimay. Intimate lose-up shots of the orporate managers were signifiantly less ommon in the Duth-English CEO s statements. These are interesting results: while the first result seemed to diminish the level of viewer-ompany intimay in the British Chairman s statements, the seond result appeared to ompensate for this. In quite a few ases, photographs in the management statements were aompanied by written text fragments explaining the pitorial ontent or adding to it. All metadisursive elements observed in these fragments are displayed in Table The table shows that adjaent omments in the British CEO s statements inluded the largest variety of metadisursive realizations and that the British Chairman s statements only refleted interational self-mentions in the form of nobility titles. Additionally, it indiates that interative frame markers direting the readers viewing proess for the photos and interational self-mentions in the form of orporate names mainly ourred in the text fragments of both the Duth-English and British CEO s statements. Table 7.14: Frequeny sores metadisourse elements in photo-related omments of management statements. Metadisourse Du-CEO Br-CEO Br-Chair n=12 n=13 n=9 Count % Count % Count % Tot.% χ 2 (df=2) p Interative: Transition Conjuntion b <.05 Frame marker Viewing stages Endoph.marker Code gloss Interational: Hedge Mitigation Passive b Mitigating (modal) verb Booster Emphat. adj/adv Emphati noun Verb of ommit Attitude Evaluat. adj/adv b Evaluative noun Evaluative verb Self-mention Corp. names pers.plur.pron Group-rel.nouns Nobility titles b < < <.05 Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test.

217 As Table 7.14 points out, there was a statistially signifiant over-representation in the number of British CEO s statements with photo-related omments that ontained transitions based on onjuntions, hedges in the form of passives and attitude markers in the form of evaluative adjetives or adverbs. These metadisursive items were non-existent in the photo-related texts of the Duth-English statements and the British Chairman s statements. In addition, Table 7.14 shows that a signifiantly over-represented number of British Chairman s statements with photo-related texts ontaining self-mentions referring to the nobility status of the sender of the text. 7.4 Metadisourse in orporate profiles Setion outlines and ontrasts the metadisursive items typially observed in the written texts of the orporate profiles; setion deals with the results for the photos and adjaent text fragments in these profiles Metadisourse in the running texts of the orporate profiles Appendix VIII ontains all metadisourse elements ourring in the written texts of the English profiles. The elements that were diserned in 50% or more of the Duth- English and/or British profiles are inluded in Table The Duth-based and UKbased profiles shared many typial metadisursive elements in the running written texts. With regard to the interative resoures, both predominantly omprised transitions in the form of onjuntions and explanatory ode glosses. As for the interational resoures, they typially ontained hedges in the form of passives and mitigating expressions, emphati and evaluative referenes to orporate qualities, and self-mentions related to the ompany or the orporate olletive (i.e. the Group ompany).

218 Table 7.15: Typial metadisourse elements in running text of 50% orporate profiles. Metadisourse Du-Profile Br-Profile n=21 n=20 Count % Count % Total % χ 2 (df = 1) p Interative: Transition Conjuntion Code gloss Interational: Hedge Passive Mitigation Booster Verb of ommitm Emphati adj/adv 21 b a <.05 Attitude Evaluative adj/adv Self-mention.48 Corporate name pers.plur.pron. 14 a b <.05 Group-rel.nouns Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. First, Fisher s Exat tests showed that a signifiantly larger number of Duth- English profiles than British profiles ontained textual boosters realized by adjetives and adverbs emphasizing ompany-related information. Seond, it was observed that signifiantly more British profiles than Duth-English profiles inluded olletive pronouns (i.e. referenes to the people behind the organization) as a means of establishing self-mentions. Self-mentions in the Duth-English profiles refleted a stronger preferene for neutral group-related referenes Metadisourse in photos and text fragments of the orporate profiles Due to the restrited number of realizations of visual metadisourse within the profiles, all findings for the photo-analysis were integrated in Table It shows that photographs in both the Duth-English and British profiles foused on indexial exemplifiations of orporate produts and marketing as visual ode glosses and on indexial illustrations of positive produt values as visual attitude markers. The images in the Duth-English profiles also revealed multiple instanes of visual attitude markers in the form of symbols referring to orporate qualities (e.g. global harater) and indies referring to (emotional) produt values. Moreover, the images in the British-English profiles revealed additional preferenes for visual ode glosses realized through employee-related symbols of the ompany, visual boosters in the form of symbols referring to the orporate grandeur, visual attitude markers in the form of various symbols of orporate qualities and visual self-mentions established through employee-related indies. In sum, the British profiles tended to ontain a

219 larger variety of metadisursive realizations in photographs than the Duth-English profiles. Table 7.16: Frequeny sores metadisourse elements in photographs of orporate profiles. Metadisourse Du-Profile Br-Profile n=8 n=10 Count % Count % Tot.% χ 2 (df=1) p Interative: Transition onjuntion Consistent theme Code gloss illustrate id, produt or market Corporate employee/-s (symbol) 0 a b <.05 Corporate produt in use (index) Corporate workplae inside/outside (index) Corporate market/ marketing (index) Aommodation (symbol) Corporate operations (ion) Interational: Booster emphasize high apaity Impressiveness produt/servie (symbol) Attitude evaluate id or produt Optimism, onfidene, enthusiasm (symbol) Sinerity, openness (symbol) Professionalism, seriousness (symbol) Friendliness, soiability (symbol) Involvement, ommitment to tasks (symbol) Powerfulness, respetability (symbol) Multi-gender nature (symbol) Male-oriented nature (symbol) Multi-ultural nature (symbol) Global nature (symbol) Added signifiane, high-quality of produt/servie (index) Produt auses positive emotional state (index) Coherene/balane, synergy (symbol) Self-mention sender Coporate employee/-s (index) 0 a b Engagement marker appeal Initimay with ompany (symbol) <.05 Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. Table 7.16 shows that signifiantly more British profiles than Duth profiles inorporated visual ode glosses realized by employee portraits symbolially identifying the ompany. In aordane with this finding, a signifiantly larger number of British-English orporations used pitures of employees as indexial realizations of self-mentions, i.e. they depited people that represented the we or the Group that was profiled in the written text. This seond result follows logially from the first signifiant differene, as these results refer to the same lexiogrammatial elements in the photographs. The illustrative funtion of employee

220 pitures automatially makes the portrayed employees the embodiment of the ompany that introdues itself in the orporate profile. All metadisursive elements that were found in the photo-related text fragments of the profiles are displayed in Table Partiularly within the Duth-English profiles, these fragments showed minor metadisursive funtionality. As for the photo-related fragments in the British-English profiles, there was a general preferene for interational emphati and evaluative lexis and self-mentions in the form of ompany-based referenes. Table 7.17: Frequeny sores metadisourse elements in photo-related omments of orporate profiles. Metadisourse Du-Profile Br-Profile n=3 n=8 Count % Count % Total % χ 2 (df = 1) p Interative: Transition Conjuntion Sentene-initial Endoph.marker Code gloss Interational: Hedge Passive Mitigation Booster Emphati adj/adv Attitude Evaluat. adj/adv Evaluative noun Self-mention Corporate name pers.plur.pron Group-rel.nouns Note.. Fisher s Exat Test Signifiane tests did not reveal any substantial ross-ultural differenes in the use of metadisourse in the photo-related text fragments of the Duth-English and British profiles. 7.5 Metadisourse in operational reviews The metadisourse in the running written texts of the Duth-English and British operational reviews will be presented and ompared in setion Subsequently, the metadisursive items in the photos and in the adjaent omments will be reported separately in setion

221 7.5.1 Metadisourse in the running texts of the operational reviews Appendix IX douments the omplete list of all metadisursive realizations observed in the written running texts of the Duth-English and British operational reviews. Table 7.18 shows the realizations of textual metadisourse that ourred in 50% or more of the reviews analyzed. From the perspetive of interative metadisourse, transitions in the form of onjuntions and phrases relating ideas, and ode glosses further larifying textual information, were partiularly popular in both the Duth- English and British reviews. Furthermore, majority sores were registered for lexio-grammatial features highlighting disursive ats, pointing at intertextual relations and disussing awards and rankings as proof of good performane. In terms of interational metadisourse, the prinipal devies used in the Duth-English and in the British reviews involved a variety of lexio-grammatial items reating a distane between writer and message (i.e. hedges), emphasizing operational and strategi information (i.e. boosters) and expressing the writer s attitude towards the textual information (i.e. attitude markers). Majority sores were also traed for interational referenes to the orporate olletive as the assumed sender. Table 7.18: Typial metadisourse elements in running text of 50% operational reviews. Metadisourse Du-OperRev Br-OperRev n=22 n=20 Count % Count % Total % Interative: Transition χ 2 (df = 1) Conjuntion Sentene-initial Frame marker Text stager 11 b a <.05 Illoutionary at Endoph. marker Evidential Reognized result Code gloss p

222 Table 7.18 ontinued. Metadisourse Du-OperRev Br-OperRev n=22 n=20 Count % Count % Total % χ 2 (df = 1) Interational: Hedge Passive Mitigation (modal) verb Booster Verb of ommit Emphati adj/adv Noun Attitude Evaluat. adj/adv Evaluat. noun Self-mention Corp.name pers.plur. pron Group-rel.nouns Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test. Chi-Square tests revealed one statistially signifiant distintion between the metadisourse elements in the running texts of the Duth-English and British reviews. A signifiantly larger number of Duth-English than British reviews ontained frame markers indiating stages in the text, e.g. by expliit text struturing signals suh as first, seondly or finally Metadisourse in photos and text fragments of the operational reviews Aording to the results in Table 7.19, the photos in the Duth-English and British reviews showed a relatively large olletion of interative ode gloss realizations. Here, images inorporating produt-related indies, faility-related indies and market-/ ustomer-related indies were espeially popular as a means of explaining or speifying the ompany, strategy or operational ativities disussed in the text. Additionally, the Duth-English and British reviews onveyed a preferene for the visual realization of interational attitude markers, where both types of reviews made substantial use of photographs depiting smiling employees (optimismsymbol), formal lothing (professionalism-symbol) and added value inherent to the produt itself (added signifiane-index). Another metadisursive element that was applied relatively frequently in the English reviews was the visual booster, i.e. in the form of lose-ups, bird-views and frog-views symbolially emphasizing the apaity, quality or omplexity of the portrayed produt. p

223 Table 7.19: Frequeny sores metadisourse elements in photographs of operational reviews. Metadisourse Du-OperRev Br-OperRev n=15 n=17 Count % Count % Tot.% χ 2 (df=1) p Interative: Transition onjuntion Consistent theme (symbol) Code gloss illustrate id, strategy, produt or market Fous CEO/Chairman (symbol) Fous divisional diretor (symbol) Team of divisional diretors (symbol) Corporate employee/-s (symbol) Fous on belongings employees (symbol) View on wide landsape (symbol) Extraurriular/reative ativity (symbol) Corporate produt in use (index) Raw materials produt (index) Corporate workplae/faility inside/outside (index) Corporate headquarters (index) Corporate market/ marketing (index) Corporate soial responsibility projet (index) Corporate operations (ion) Evidential reognition positive performane Fous on satisfied lient (symbol) Fous on elebrities-in-ation (symbol) Fous on award (symbol) Interational: Booster emphasize high apaity Impressiveness produt/servie (symbol) Impressiveness workplae/faility (symbol) Impressiveness market (symbol) Attitude evaluate id or produt Optimism, onfidene, enthusiasm (symbol) Sinerity, openness (symbol) Business-like/reserved nature (symbol) Professionalism, seriousness (symbol) Eye for style/modernism (symbol) Friendliness, soiability (symbol) Involvement, ommitment to tasks (symbol) Powerfulness, respetability (symbol)

224 Table 7.19 ontinued. Metadisourse Du-OperRev Br-OperRev n=15 n=17 Count % Count % Tot.% χ 2 (df=1) p Interational: Attitude evaluate id or produt Multi-gender nature (symbol) Multi-ultural nature (symbol) Global nature (symbol) Grandeur, well-established nature (symbol) At servie of stakeholders (symbol) Care, tenderness for ustomers (symbol) Contributes to daily life (symbol) Contributes to professional life (symbol) Contributes to life of young and old (symbol) Contributes to life of small/large businesses (symbol) Helps dreams ome true (symbol) Responsibility, soial awareness (symbol) Extensive employee-related awareness (symbol) Creativity, onstrutiveness (symbol) Added signifiane, high-quality of produt/servie (index) Produt auses positive emotional state (index) Coherene/balane, synergy (symbol) Clear-sighted, foused nature (symbol) Self-mention sender Fous CEO/Chairman (index) Fous divisional diretor (index) Team of divisional diretors (index) Coporate employee/-s (index) Engagement marker appeal Initimay with ompany (symbol) Note.. Fisher s Exat Test. Signifiane tests did not reveal any substantial ross-ultural differenes, implying that the produers of the Duth-English and British reviews relied on similar photographi signs or lexio-grammar in their visual rhetori. Within the multimodal operational reviews, photos were often larified by aompanying text fragments referring either diretly or indiretly to the meaning of the image. All instanes of metadisourse in these fragments are displayed in Table The table shows that photo-related omments in the Duth-English and British reviews typially inluded hedges (passives), boosters (emphati lexis), attitude markers (evaluative lexis) and self-mentions (orporate names, first person plural pronouns). While the photo-related omments in the Duth-English reviews also

225 inluded interative onjuntions, the British reviews ommonly ontained interative ode glosses. Table 7.20: Frequeny sores metadisourse elements in photo-related omments of operational reviews. Metadisourse Du-OperRev Br-OperRev n=13 n=15 Count % Count % Total % χ 2 (df = 1) p Interative: Transition Conjuntion Sentene-initial Frame marker Text stage Illoutionary at Endoph Marker Evidential Reogized result Quotations Code gloss Interational: Hedge Passive Mitigation (modal) verb Booster Verb of ommit Emphati adj/adv Attitude Evaluat. adj/adv Evaluat. Noun Evaluat. Verb Engage.marker Self-mention Corp.name pers.plur. pron Management Group-rel.nouns Note.. Fisher s Exat Test. As Table 7.20 points out, the results of Chi-Square and Fisher s Exat tests did not exhibit any ross-ultural differenes in the use of metadisursive features within the photo-related text fragments of the Duth-English and British operational reviews. 7.6 Conlusion and disussion Consistent with the first main researh aim of this genre study (see setion 1.5), the analysis of multimodal metadisourse foused on the doumentation of ultural

226 preferenes in the lexio-grammatial design of English annual report texts. In response to the speifi researh questions, the metadisourse analysis omprised ulture-speifi inventories of text features (RQs 2a-b) and ross-ultural omparisons of these text features to determine to what extent the Duth-English and British reports differed at the lexio-grammatial level (RQ 2). In general, the results in this hapter indiated that previous answers to RQs 2a-b (in Chapters 5 and 6) were onfirmed by the investigation of textual and visual metadisourse. The identifiation of metadisursive elements in the Duth-English and British management statements, profiles and operational reviews showed that eah annual report setion in the two national business ultures omprised a set of multimodal lexio-grammatial elements that is different from the set of features represented in another setion. Comparing the management statements and the orporate profiles, for example, it appeared that the running texts in the statements inluded a larger variety of typial metadisursive items than the running texts in the profiles. With referene to RQ 2, it ould be observed that there were ross-ultural variations for a number of metadisursive realizations, most of whih ourred aross the Duth- English and British management statements. Furthermore, it was found that the majority of ross-ultural differenes onerned textual or visual realizations of interational metadisourse. Following Hyland, it an thus be onluded that the ontrastive researh results in setion unovered some generi variations as well as [ ] distintions in ommunity praties (2000: 129). In setion of this onlusion, the main ross-ultural similarities and differenes observed in the metadisourse analysis will be summarized and, if relevant, they will be related to ontextual features of the Duth-English and British annual reports (RQ2d). Setion will fous on the impliations of the metadisourse investigation for genre theory and genre analysis in general. As suh, it also onsiders the seond main researh goal of the present projet, i.e. the further development or refinement of genre-based disourse analysis Cross-ultural omparison of metadisourse elements The ross-ultural analysis of lexio-grammar in multimodal business texts emerged from the idea that metadisourse fouses our attention on the way writers projet themselves into their work to signal their ommuniative intentions (Hyland, 1998a: 437). The summary below highlights the similarities and differenes in metadisourse between the Duth-English and British annual report texts (RQs 2a- 2); it also refers briefly to the ross-ultural omparability of these texts (RQs 2). Next, it will indiate that generi variations may be owing to ulture-dependent disourse praties (RQ 2d). Comparison of metadisourse in management statements Aross the Duth-English and British management statements similarities in the use of typial metadisursive elements ould be observed in the running texts, the photographs and the photo-related omments. In the running written texts, a great deal of ross-ultural orrespondene was found in the use of interative (senteneinitial) onjuntions, illoutionary ats, intertextuality markers and explanatory ode glosses. Cross-ultural similarities in textual metadisourse was also found in the use of interational distaning devies, verbal or lexial emphatis, evaluative expressions and group-related or writer-related self-mentions. The agreement in photographi metadisourse aross English management statements involved an

227 even distribution of interative onjuntions related to the entral theme of the report, ode glosses identifying diverse orporate aspets, emphati and evaluative illustrations of produt values and board-related self-mentions. Finally, rossultural orrespondene was also observed for the interative and interational resoures embedded in the photo-related omments. Results showed almost equal frequenies of ourrene for phrases framing the image interpretations, endophori expressions, ode glosses, mitigation devies, emphati and evaluative nouns and verbs and self-mentions refleting the orporate olletive. The present analysis also suggested several signifiant differenes in the use of partiular lexio-grammatial realizations of metadisourse. Aross business ultures, it was determined that a onsiderably large number of Duth-English CEO s statements (as opposed to British statements) ontained frame markers suh as first or in short to sequene information in the written texts. Additionally, it was observed that a signifiantly small number of British CEO s statements inluded engaging pronouns like you(r) or salutations like dear shareholder and that a signifiantly large number of British Chairman s statements ontained management-related nouns like the Board s as a means of referring to the ompany responsible for the annual report. Moreover, many signifiant distintions arose aross the Duth-English and British management statements with regard to metadisourse in photographs. Compared to the British management statements, a substantial number of Duth-English CEO s statements showed visual ode glosses and self-mentions established through portraits of management teams personifying either the ompany or the writer of the text. Conversely, a signifiantly small number of Duth-English CEO s statements inluded ode-glosses and selfmentions based on portraits of individual managers embodying the orporation or the author. Next, onsiderably more Duth-English CEO s statements than British Chairman s statements ontained attitude markers expressing orporate sinerity through frontal views of employees. In ontrast, signifiantly more British Chairman s statements than Duth-English CEO s statements inluded attitude markers expressing intimay with the ompany through lose-ups of employees. Apparently, both introdutory douments of the annual report use photos to seek intense ontat with the reader, but the manner in whih this is done differs per national business ulture. Furthermore, the visual analysis pointed out that a signifiantly large number of British CEO s statements omprised ode glosses in the form of indies related to the orporate marketing strategy, and attitude markers onsisting of signs symbolizing the multi-gender, multi-ultural and global nature of the orporation. Finally, it appeared that signifiantly more British than Duth- English CEO s statements inluded ode glosses in the form of employee portraits symbolizing the ompany. As for the photo-related omments, signifiant distintions mainly onerned the British CEO s statements: a substantial number of photo-related omments in these statements showed transitions expressed through onjuntions, hedges in the form of passives, and attitude markers expressed by qualifying lexis. The fourth signifiant result in the texts adjoining photographs involved the over-representation of the number of British Chairman s statements displaying self-mentions in the form of nobility titles suh as lord or sir. Considering the metadisursive similarities and differenes between the Duth-English and British management statements, it an be argued that these statements generally orrespond aross ultures but that they also ontain subtle ulture-speifi realizations of textual and visual metadisourse. Consistent with the results of the previous analyses on ontext, ontent and struture (Chapters 4-6), the metadisourse analysis showed that the Duth-English statements were linguistially

228 similar to both the British-English Chairman s and CEO s statements. In this sense, there was ross-ultural overlap for the use of metadisursive elements in written texts and photographs (this does not mean, however, that there were no ross-textual distintions between the statements within the British annual reports). While the running texts in the Duth-English and British CEO s statements inluded few selfmentions in the form of management-related nouns, for instane, the photos in the Duth-English CEO s statements and British Chairman s statements did not omprise any ode glosses in the form of indexial illustrations of the market. Moreover, there was also ross-ultural overlap with regard to all three statement types, e.g. the similar distribution of onjuntions or ode glosses in the written texts. With regard to the variations aross national business ultures, it seems that some ross-ultural differenes an be explained by the prodution ontext of the English management statements (setion 4.4.4). For example, the fat that British CEO s statements ontained relatively few engagement markers in the written texts but many attitude markers in the form of visual symbols of the ompany, may be explained by the assumption that these statements are highly narrative and need not establish reader-writer relationships (like the Duth-English CEO s statements and the British Chairman s statements). They appear to require a more neutral form of address and seem to allow for greater reativity in visual metadisourse. Other ross-ultural distintions are likely to be related to national ultural values that have influened the multimodal disourse in management statements. For instane, the findings that a relatively large number of Duth-English CEO s statements ontained text stagers, visual ode glosses in the form of portraits of the management team, and visual attitude markers in the form of employees views in the amera lens may be explained by the low-ontext and feminine ulture of the Netherlands, whih emphasizes ooperation and diretness (as opposed to the UK ulture, whih is less low-ontext and masuline in nature, see e.g. Claes & Gerritsen, 2007; Hofstede, 2001). Comparison of metadisourse in orporate profiles There was a high degree of ross-ultural similarity for the written texts, photographs and photo-related omments in the Duth-English and British profiles. With respet to the typial metadisursive features in the written texts of these profiles, similarities were found for interative onjuntions logially relating textual propositions and for phrases and words exemplifying or speifying textual information. Also, the written texts in the profiles showed an even distribution of defoalizing passives and mitigating lexis, emphati verbs, qualifying adjetives or adverbs and self-mentions in the form of orporate names or group-related nouns. Cross-ultural similarities in photo-based metadisourse were observed for the use of interative onjuntions visually onstituting the entral theme of the annual report and visual ode glosses illustrating the ompany s produt, its failities, ustomer or markets, headquarters and operations. Furthermore, orrespondene in visual metadisourse emerged for visual signs boosting and evaluating the produt or ompany qualities and for amera-related signs establishing viewer-sender intimay or engagement. There was full ross-ultural orrespondene in metadisourse for the photo-related text fragments of the Duth-English and British profiles; these fragments ontained onjuntions and sentene-initial transitions, endophori markers in phrases and website referenes, hedges expressed through passives and mitigating lexis, emphati and evaluative lexis and self-mentions in the form of orporate names, first person plural pronouns and group-related nouns.

229 The analysis revealed several ross-ultural differenes in the lexiogrammatial realizations of metadisourse within the written texts and the photographs inluded in the Duth-English and British orporate profiles. As for the running written texts, it was determined first of all that boosters in the form of emphatis (e.g. more than or major ) ourred in a signifiantly larger number of Duth-English profiles than British profiles. With regard to the textual selfmentions, it was found that substantially more British profiles than Duth-English profiles ontained first person plural pronouns like we or our to refer to the sender or the ompany behind the message. Sores for metadisourse in the photographs of Duth-English and British profiles revealed that a signifiantly larger number of British than Duth-English profiles ontained visual ode glosses realized by employee portraits symbolizing the orporation, and visual self-mentions in the form of employee portraits illustrating the narrator in the text. The results of the metadisourse analysis aross ultures indiated that the Duth-English and British profiles largely rely on the same rhetorial lexiogrammar. Exept for a few ross-ultural differenes in textual and visual realizations of metadisourse, no substantial variation was enountered. Similar to the profile-related results in the ontent analysis (Chapter 5) and struture analysis (Chapter 6), the high degree of metadisursive orrespondene may be owing to shared ontextual features, i.e. the Duth-English and British profiles largely serve the same purposes and readers (see setion 4.4.4). Given the ross-ultural orrespondene in prodution ontext, the ross-ultural distintions in textual and visual metadisourse are diffiult to explain. Referring to work suh as Dow (1999; in Nikerson, 2000), it seems that differenes in the use of emphatis may reflet ulture-speifi preferenes in language use. Comparison of metadisourse in operational reviews The Duth-English and British operational reviews showed many ross-ultural similarities in the use of textual and visual metadisourse. The typial metadisursive elements in the written texts of the reviews refleted orrespondene for interative onjuntions and sentene-initial transitions, phrases indiating illoutionary ats, intertextuality markers, evidentials referring to awards and rankings and ode glosses explaining textual information. In addition, ross-ultural similarities were observed for interational passives and mitigations, lexion boosting or evaluating information about the ompany and self-mentions in the form of first person plural pronouns and group-related nouns. There was full orrespondene with regard to the diverse metadisursive realizations in the photographs. The English reviews showed a similar distribution of visual transitions supporting the entral theme of the report, visual ode glosses established through illustrations of orporate people, produts, markets, failities or operations and evidentials in the form of portraits of lients, elebrities or awards. Moreover, similarities were observed for visual boosters and attitude markers highlighting the orporation and its people, produts, markets or failities, self-mentions referring to the individual or team of people behind the text and engagement markers establishing reader-writer intimay. The metadisursive items found in the photorelated omments of the reviews also refleted only ross-ultural similarities. There was orrespondene for interative transitions, disursive frame markers, intertextuality markers, evidentials for reognized results and explanatory ode glosses. Furthermore, orrespondene was shown for interational passives and mitigations, emphati and evaluative verbs or lexis, reader-oriented engagement

230 markers and self-mentions in the form of plural pronouns, ompany-related or funtion-related nouns. Signifiant ross-ultural differenes in the use of metadisourse ould only be traed in the written texts of the Duth-English and British operational reviews. Compared to the British operational reviews, a signifiantly larger number of Duth-English reviews inluded frame markers in the form of text stagers organizing the information in the text. Generally, it an be onluded that there was very little variation in the use of textual and visual metadisourse inorporated in the Duth-English and British operational reviews. Of all the annual report setions analyzed, the English operational reviews showed the least ross-ultural differenes in the lexiogrammatial realizations of metadisourse. With regard to the ross-ultural distintion in text stagers, this differene may well be attributable to national ultural norms. As was suggested in the disussion of the metadisourse analysis of management statements, the relative prominene of text stagers in the Duth-English texts may be explained by the low-ontext ulture of the Netherlands, whih values diret and expliit ommuniation (e.g. Claes & Gerritsen, 2007). Metadisourse and the Duth-English and British-English annual reports as a genre set The results of the ontrastive metadisourse analysis, summarized in the previous paragraphs, underpin the identifiation of Duth-English and British annual reports as internationally reognized business texts. Judging from the large degree of rossultural similarity in textual and visual metadisourse, it seems that the two English variants of these texts mainly use similar metadisursive elements to realize similar text purposes. In answer to RQ 2 in partiular, it may thus be onluded that the Duth-English and British annual report texts are marked by omparable metadisursive onstruts (e.g. Miller, 1994). Comparisons of the textual and visual metadisourse on the level of text types first of all supported the idea that within the Duth-English and British annual reports, none of these setions ontained idential multimodal features. In eah business ulture, the management statements, profiles and operational reviews refleted a unique series of typial metadisursive elements. With regard to the British annual report douments, for example, engagement markers were most typial of the written texts in the British Chairman s statements, and evaluative photographs highlighting the ompany s ontribution to the daily life of ustomers only appeared reurrently in the operational reviews. Hene, suh findings lend support to the onlusion made in previous hapters that the Duth-English and British annual reports really onsist of a olletion of different text types, or genres. In the urrent metadisourse analysis this onlusion was further strengthened by the observed inlusion of intertextual endophori markers, whih expliitly indiated that different kinds of texts were integrated and oherent within one umbrella doument. As suh, the metadisursive design of the Duth-English and British annual reports appeared to be ongruent with Devitt s (1991) general oneptualization of the genre set, and in partiular with Jameson s (2000) laim that the annual report is a set of interrelated (sub)genres. More speifially, the fat that the Duth-English and British management statements, profiles and operational reviews shared numerous metadisursive features also suggested a degree of lexio-grammatial omparability aross ultures for the text types inorporated in the international annual report as a genre set. For example, the running texts in both the Duth-English and the British operational

231 reviews showed relatively high frequenies for the use of interative evidentials in the form of referenes to awards or rankings. As in previous hapters, the idea that Duth-English and British annual reports onsist of a set of (at least three) omparable instanes of the same genres was thus onfirmed Impliations for genre theory and genre analysis At the end of this hapter, several impliations an be drawn from the results of the metadisourse analysis in the written texts and photographs of Duth-English and British annual reports. These will be used to provide a further answer to the seond main researh objetive of this genre study, i.e. the theoretial and analytial development of the generi approah to disourse. The theoretial importane of metadisourse in generi lexio-grammar The investigation of the textual and visual lexio-grammar of management statements, profiles and operational reviews was onduted on the basis of the metadisourse onept, primarily derived from Hyland and Tse (2004). In brief, this onept involves linguisti elements that help realize the rational, redible and affetive appeals that ontribute to the persuasiveness of a text (Hyland, 1998b: 226). As suh, it proved to be a useful means of analysis, both for the study of lexio-grammar in texts and for the lexio-grammatial examination of photographs in Duth-English and British annual reports. More speifially, the analysis of metadisourse provided an insight into the ulture-speifi lexion and grammar Duth and British orporations employed to ahieve the ommuniative objetives of these multimodal douments (see also Solbjørg Skulstad, 2005). In line with Hyland s (1998b) previous study on CEO s Letters and Diretors reports, the metadisursive realizations observed in the urrent analysis showed that the produers of the Duth-English and British annual report texts put partiular effort into reating a logial and persuasive message to gain support for their proposition. This finding involved textual metadisourse as muh as visual metadisourse, meaning for example that the illustration and evaluation of ompany-related information (i.e. interative ode glosses and interational attitude markers) were reurrent elements in the researh results for both the written texts and the photographs in all annual report setions. Furthermore, the model of metadisourse proposed by Hyland and Tse (2004) was found to be partiularly useful for a ategorial perspetive on generi lexio-grammar. It enabled the lassifiation of lexio-grammatial elements in both texts and photographs on the basis of their rhetorial funtionality within the ontext of the text. Considering that genre theory in priniple fouses on rhetorial text identifiation that moves from the level of disourse ontext to the level of lexio-grammatial text features (e.g. Swales, 1990; Bhatia, 1993) and onsidering the present intent to establish a systematial overview of rhetorially funtional lexio-grammar, Hyland and Tse s (2004) oneptualization of metadisourse proved to be a relevant and fruitful devie for a lexio-grammatial genre analysis. The theoretial importane of ross-ultural variations in generi lexio-gramar Apart from the observed ross-ultural omparability of annual report setions, the metadisourse analysis also allowed for the identifiation of ross-ultural variations in the use of metadisursive elements. As suh, it was a useful tool for the ontextdependent definition of the business texts. The present study learly supported the

232 onlusions in Mauranen (1993), Nikerson (2000) and Vergaro (2005), whih indiated that ertain realizations of metadisourse in genre onstrution may be more or less preferable in one ulture than in another. An example taken from the present researh material illustrates this: ompared to the British management statements and the operational reviews, the Duth-English management statements and reviews refleted a substantially higher preferene for frame markers in the form of text-staging referenes suh as finally or in short. Unlike other ross-ultural differenes, suh as the small number of engagement markers in the written text of the British CEO s statements, variations suh as the one related to text stagers ould not be linked to the annual report s prodution ontext in a given business ulture. These variations seemed to be related to more general ultural onventions in (business) ommuniations. Aordingly, the ross-ultural distintion for text stagers may have been affeted by the fat that the Netherlands has a stronger lowontext ulture than the UK, implying that Duth ommuniators generally have a higher preferene for expliit and linear disourse than British ommuniators (e.g. Claes & Gerritsen, 2007; Ulijn & StAmant, 2000). A similar argument would seem to apply to the observations that (1) a signifiantly large number of Duth-English CEO s statements inluded visual ode glosses and self-mentions realized by photos of the management team, and that (2) the British Chairman s statements showed an over-representation of nobility titles in the photo-related omments. Both observations ould be asribed to the higher degree of masulinity in UK ulture, as ompared to Duth ulture; masuline ultures generally fous more on individual aknowledgement or prestige and less so on ooperation and ompromise (e.g. Claes & Gerritsen, 2002; Hofstede, 2001). It should be noted that these are tentative ideas; this study did not provide diret evidene for suh ulture-based explanations. Relating the urrent researh results to the results of other studies on metadisourse in Duth-English and British business texts, it is interesting to note that some of these results overlap and some are ontraditory. For instane, in her study on internal ommuniation inside an Anglo-Duth ompany Nikerson (2000) finds that Duth-English writers make muh more use of emphatis than their British olleagues; this finding is in line with the signifiantly higher frequenies of emphati adjetives and adverbs that were diserned in the present Duth-English orporate profiles. Therefore, the metadisourse analysis of annual report texts seemed to support Dow s (1999) suggestion that Germani speakers tend to upgrade propositional information, while English speakers prefer to use down-graders (also in Nikerson, 2000). On the ontrary, Nikerson s observation that Duth-English writers showed a higher preferene for the appliation of the pronoun we (exluding the reeiver of the text) seems to onflit with the present result that selfmentions in the form of first person plural pronouns ourred in a signifiantly larger number of British orporate profiles. Based on this example, it would seem that metadisursive items whih are ulture-speifi in one genre may not be so in another genre. The theoretial importane of photo-based metadisourse Furthermore, the urrent lexio-grammatial analysis suggested that the investigation of doumentary photographs ontributes to a full understanding of metadisourse in multimodal genres. As was illustrated in the results setions of this hapter, the semiotis-based interpretation of photographs helped to establish metadisursive differenes aross text types or genres and aross national business ultures. Also, the reliane on Peire s system of signs proved helpful in establishing the rhetorial hoies taken within the visual lexio-grammar, yielding logi-based

233 and detailed data at the most elementary level of meaning in photographs (i.e. similar to the lexio-grammatial language level). Finally, the semioti analysis allowed for a further investigation of the assumed endophori relationship between written texts and images in the annual reports; onsequently, it offered insights into the metadisursive elements embedded within the images. These insights, then, omprise omplementary information to the ideas put forward by Hyland (2000) and Fuertes-Olivera et al. (2001), who exlusively analyzed the endophori link between the ontent of texts and the ontent of photographs but who did not look at the metadisursive elements inorporated in the photos. The analytial relevane of orpus size in the study of textual metadisourse The present metadisourse study also put forward some methodologial impliations with respet to the design of the textual metadisourse analysis. As was already implied in setion 7.2.1, the identifiation and labeling of metadisursive items in the written texts turned out to be an intriate task. Despite the limitations set for the WordSmith wordlists on whih the analysis was based, the tasks of finding and registering metadisursive elements, realizing omparable frequeny tables and alulating signifiant ross-ultural results still required the investment of onsiderable time and effort. Apparently, the quantifiable investigation of metadisourse in a large sample of 137 annual report texts ranging from one to over 20 pages demands a long-term fous. Hene, it seems advisable that any metadisourse analyses to be onduted in a limited period of time should either ontain a restrited orpus-size or should fous only on a seletion of metadisursive realizations (e.g. as in genre studies of Bondi, 2002; Mauranen, 1993; Valero- Garés, 1996). Nonetheless, in the light of the present researh aims, the approah taken towards the design for the analysis of textual metadisourse did result in a omprehensive and exploratory overview of metadisursive elements generally onstituting the (rhetorial) linguisti style in the Duth-English and British annual reports. The analytial relevane of interoder reliability tests in the identifiation of metadisourse As was the ase in the previous analyses of prodution ontext, themati ontent and rhetorial struture (Chapters 4-6), interoder tests were onduted to measure the reliability of the identifiation of metadisursive elements in annual report texts. More speifially, the interoder analyses measured the possibility of repeated measurement and thus underpinned the oding system used in the present study to identify metadisursive items. The interoder test for textual metadisourse aused few interpretation problems between the oders, as was evidened by the high interoder agreement sores. More omplexities arose during the interoder test for the elements of visual metadisourse. It ould be argued that the main reason for the low interoder agreement in the reliability test for visual metadisourse onerns the poly-interpretable potential of photographs. The interoder test showed that different interpretations may have various auses. During the interoder analysis, partiipants sometimes ativated different ommon knowledge in the establishment of a text-photo relationship. For instane, one of the oders was not aware that a piture of London s skyline atually zoomed in on the City -setion of London. Here, the image of the dynami finanial enter of London was losely related to the identifiation of the finanial servie provider publishing the annual report. In addition, the outome of the interoder test was affeted by interoder disrepanies in the positioning of

234 idential semioti meanings within the analyti shemes: for several photographs, oders atually introdued the same interpretations, but these interpretations were either asribed to different photographi omponents or disussed at different semioti levels. Finally, the low interoder agreement sores were owing to the enforement of a given semioti meaning at all sign levels. This ourred, for example, with regard to a lose-up image of a light-olored folder with Old Mutual s logo and a produt name ( Personal Finanial Assets ). While this image merely showed the physial form of one of Old Mutual s produts, a meaning that was established at the indexial level by both oders, one oder added that this folder was a symboli representation of the ompany s professionalism and ommitment to ustomers, without further referene to any pitorial lues that might have led to this symboli interpretation. This illustrates that the signifiation of photographs in (multimodal) texts is impated by subjetive meaning-making that an lead to different (priorities in) interpretations of visuals. From an analytial perspetive, it an be argued that higher interoder agreement sores may have been ahieved by using higher-order ategories of meaning-making (e.g. remaining on the level of interative-interational funtions). However, this would be at the expense of a detailed study of visual realizations of metadisourse. For the produers of international annual reports, these interoder results would seem to be of ruial importane, for they emphasize the ommuniative risk of doumentary photographs; without areful onsideration for publiation, suh photos might end up onveying a message among readers that was not intended by the ompany. The analytial relevane of metadisourse for reader response analyses A final remark onerns the relevane of the urrent findings on metadisourse for the exeution of a reader response analysis. The investigation of metadisursive items in texts and photos revealed several substantial differenes in the lexiogrammatial onstrut of genres within the Duth-English and British annual reports. These ross-ultural distintions in partiular are interesting in a researh setting aimed at measuring the interultural effetiveness and appreiation of different English-language versions of the same business genre. For example, it would be interesting to test UK-based readers reation to the many text-staging frame markers in the Duth-English annual report texts. With regard to the introdutory douments of the annual report, it would moreover be relevant to find out what these readers reations would be to pitures of managers faing the audiene, as opposed to portraits of managers looking away from the amera. Aordingly, the ross-ultural metadisursive distintions that were observed in this hapter are likely to be useful as independent variables within an experimental study of the reeption of international English annual reports. Chapters 4-7 have provided answers to speifi RQs 1a-d and RQs 2a-2d, onerning the ulture-speifi desription and ross-ultural omparison of ontextual, substantive and formal genre features of Duth-English and British- English annual report texts. Sine the multimodal analyses presented in Chapters 4-7 overed all types of genre harateristis and resulted in a thik desription of the business genre involved (e.g. Bhatia, 1993), Chapter 8 will disuss several overarhing onlusions and impliations with regard to these (on)text analyses.

235 Chapter 8 Conlusion ross-ultural genre analyses In aordane with the assumption within genre theory that a text type reours in a partiular soio-ultural ontext and ontains a prototypial ontent, struture and lexio-grammar (e.g. Nikerson, 2000; Swales, 1990; Yates & Orlikowski, 1992), the general exploration of the Duth-English and British annual reports in Chapters 4 to 7 foused on the prodution setting of these reports, as well as on the themes, strutural moves and strategies, and metadisourse inorporated in the reports. Eah of the analyses of (on)text features provided further insight in the generi onstrut of multimodal annual report texts produed in the Duth and British business ultures. Following the genre-in-ontext approah to disourse in partiular (Askehave & Swales, 2001; Nikerson, 2000; see setion 3.1), the thik desription of the annual report texts arose from the investigation of the disourse ommunity in whih these texts had been produed (Chapter 4) and was finalized with the identifiation and situated explanation of textual and visual metadisourse (Chapter 7). After returning to the main researh objetives and researh questions underlying the analyses of genre features in the multimodal annual reports, this hapter will provide an overall summary of the results presented in Chapters 4 to 7. Moreover, it will refer to the general impliations for genre theory and genre analysis that were generated by the present researh results. As was indiated in Chapter 1, this genre study was based on the premise that the English multimodal annual report is an influential business text for multinational ompanies. Globalizing markets and the implementation of international standards for finanial reporting have strengthened the position of English as a global finanial language and have inreased the persuasive relevane of multimodal non-finanial texts in the annual report in partiular (e.g. Hyland, 1998; IASPlus, 2005; ING, 2006; Van der Moolen, 2006). Regarding these eonomi, soio-politial and linguisti developments, and onsidering the fat that relatively few studies have desribed the English annual report as a form of international business disourse (e.g. Hyland, 1998), it was deided that the present researh should seek to establish a rhetorial understanding of the international English annual report, speifially those reports produed in the Duth and British business ultures. In the light of these onsiderations, the primary researh aim of this study was to doument any ulture-speifi preferenes in the use of English-language annual reports originating in Duth and British orporations (see setion 1.5). Chapters 4 to 7 demonstrated that an initial step in the ahievement of this researh objetive was to identify and ompare the disursive features text produers had inorporated in the written texts and photographs of the Duth-English and British- English annual reports. Given the fat that these annual reports an be seen to represent an internationally reognized text type serving partiular purposes and readers in a speifi business ommuniation ontext, it was deided that this omparison should be arried out within the frame of genre theory (Bhatia, 1993; Yates & Orlikowski, 1992). To date, genre-based investigations of annual reports have predominantly foused on qualitative desriptions of (small) orpora of written texts (e.g. Garzone, 2004, 2005; Gillaerts 1996; Jameson, 2001; Nikerson & De Groot, 2005). Aordingly, the seond main researh aim of the urrent analysis was to ontribute to the further theoretial and analytial development of a generi approah to multimodal disourse in international business texts, i.e. in international

236 annual reports in partiular. In other words, this study was intended to expand the relevane of genre theory as an instrument for qualitative and quantitative analyses of written texts and images in business genres originating in different national business ultures (see also setion 1.5). It was shown in setion 1.5 that these general researh objetives resulted in two main researh questions, the first of whih is repeated below. The seond main researh question does not diretly relate to the ontext and text analyses presented in Chapters 4-7 and will be addressed separately in Chapter 9, whih disusses the reader effet of ulture-speifi genre features in English annual report texts. I. What are the genre features of English, non-finanial texts in multimodal annual reports produed by Duth orporations and in what way do these differ from the genre features of English, non-finanial texts in multimodal annual reports produed by British orporations? Consistent with the generi approah to disourse analysis (Swales, 1990), this first researh question was divided into speifi researh questions addressing the ommuniative ontext assoiated with the Duth-English and British annual reports on the one hand, and the textual and visual ontent, struture and lexio-grammar of these reports on the other hand. Table 8.1 reapitulates the speifi researh questions related to the identifiation of genre ontext and multimodal genre elements. Chapter 4 provided detailed answers to RQs 1a-1d and Chapters 5-7 aounted for the answers to RQs 2a-2d. Table 8.1: Speifi RQs related to the first main researh question of the present genre study (see setion 1.5). RQs genre ontext RQ 1a. Aording to text produers in Duth multinationals, whih English language poliy is maintained for the Duth- English annual general report? RQ 1b. Aording to text produers in Duth multinationals, what is the typial ommuniative ontext of the Duth- English annual general report? RQ 1. Aording to text produers in British multinationals, what is the typial ommuniative ontext of the British- English annual general report? RQ 1d. To what extent do the typial ommuniative ontexts of the Duth- English and British annual general reports orrespond and differ? RQs genre ontent, rhetorial genre struture and generi lexio-grammar RQ 2a. What are the typial ontent, struture and lexio-grammar of the texts and photos in the English annual general report produed by Duth multinationals? RQ 2b. What are the typial ontent, struture and lexio-grammar of the texts and photos in the English annual general report produed by British multinationals? RQ 2. To what extent do the typial ontent, struture and lexio-grammar of the texts and photos in the Duth-English and British annual general reports orrespond and differ? RQ 2d. How might the existene of these textual similarities and differenes be explained? The researh design (in Chapter 3) argued that the speifi RQs in Table 8.1 ould best be answered on the basis of a multidisiplinary analytial framework. While genre theory did provide for units of observation in the analysis of rhetorial genre struture, i.e. in terms of moves and strategies (Bhatia, 1993), it did not speify suffiiently how genre ontext, genre ontent and generi lexio-grammar should be

237 analyzed. In the present study, the genre-based framework was therefore extended by models derived from various disiplines, i.e. ethnography (disourse ommunity Beaufort, 1997), ognitive mapping (oneptual themes Carley & Palmquist, 1992) and linguistis (metadisourse Hyland & Tse, 2004). The appliation of this multidisiplinary framework aross the Duth and British business ultures revealed that the Duth-English and British-English annual reports were haraterized by many generi similarities and several ross-ultural differenes. Thus, it showed that a single internationally reognized business genre may still ontain ulture-speifi realizations. Table 8.2 offers an overview of the ross-ultural differenes that were observed for the present researh materials. As suh, it presents the partial answers to RQs 1d and 2, whih ould not have been solved without answering RQs 1a-1 and RQs 2a-2b first. The answers to RQs 1a-1 and 2a-2b were inorporated into the presentations of results in Chapters 4 to 7. Table 8.2 displays the ross-ultural variations for management statements, profiles and operational reviews per series of RQs, i.e. RQs 1a-1d for genre ontext and RQs 2a-2 for genre ontent, genre struture and generi lexio-grammar. Sine the results in Table 8.2 do not show the ontextual explanation of ross-ultural differenes in genre features, speifi RQ 2d is not represented here. It should furthermore be noted that the ontextual differenes aross the two national business ultures onern qualitative findings, while the ross-ultural differenes for disourse features involve quantitative results (obtained on the basis of the researh methods disussed in Chapter 3). The signifiant results in Table 8.2 indiate whether an item was observed in a signifiantly higher or lower number of annual report texts, or whether it was observed in a signifiantly over-represented (high) or under-represented (low) number of annual report texts. Table 8.2: Overview of ross-ultural differenes in genre features, per level of analysis (see Chapters 4-7). Speifi RQs Level of analysis (Genres in) the annual reports Contextual haraterization and omparison of the Duth- Priority reader groups Finanial readers English and (investors, media, British annual analysts) + employees reports (RQs Relevane of English - 1a 1d). Context Du-Eng annual report Br-Eng annual report Corp.ommuniation mix Finanial mix Finanial + marketing mix Criteria table of ontents Unique texts in ontents International organization struture, international board members Reader needs, orporate poliy, benhmarking CEO s statement, supervisory board report Institutional investors, private investors + analysts, media Management approah to finanial reporting Chairman s statement, CEO s statement, diretors report, general fous on investors Major role exeutive Diretors in prod.proess Major role supervisory board board Unique funtion photos - To add to readability of the text

238 Table 8.2 ontinued. Speifi RQs Level of analysis (Genres in) the annual reports Charaterization and omparison of the typial multi-modal disourse in the Duth-English and British annual reports texts (RQs 2a 2). Content text Du-Eng CEO Br-Eng CEO Br-Eng Chair Corp. Strategy higher lower Board details lower higher HRM/People higher lower Corp.governane higher lower CSR lower higher Dividend lower lower higher Content photo All exeut.board members high CEO higher lower Chairman lower lower higher Employees lower higher Workplae high Content photo Du-Eng Profile Br-Eng Profile Employees lower higher Content text Du-Eng Oper.review Br-Eng Oper.review Cost management higher lower Identity market higher lower Struture text a Du-Eng CEO Br-Eng CEO Br-Eng Chair M: Leadership lower higher M: Operat. performane low M: Sympathy low M: Reader orientation lower higher S: Desription produt low S: Management details lower higher S: Appreiation higher lower higher S: Salutation higher lower M+S:Oper.perf.+fin.detail low M+S:Fut.perf.+fut.objet. high M+S: Address.read+salut. higher lower Struture text a Du-Eng Profile Br-Eng Profile M: Corporate strategy higher lower Struture text a Du-Eng Oper.review Br-Eng Oper.review S: General profile higher lower M+S:Corp.strategy+strat.fous higher lower M+S: Market+desry.produt higher lower M+S: Oper.perf.+ust.needs higher lower M+S: Fut.perf.+fut.objet. lower higher Struture photo omment a S: Corporate strengths lower higher

239 Table 8.2 ontinued. Speifi RQs Level of analysis (Genres in) the annual reports Charaterization Lexio-gram. text Du-Eng CEO Br-Eng CEO Br-Eng Chair and omparison Text stager high of the typial Engagement marker low multi-modal Self-mention: high disourse in the management Duth-English and British Lexio-gram. photo Code gloss: fous low annual reports CEO/Chair texts (RQs 2a Code gloss: fous high 2). CEO+board Code gloss: employees lower higher Code gloss: market(ing) high Attitude: sinerity higher lower Attitude: multi-gender high Attitude: multi-ultural high Attitude: global nature high Self-ment.: fous low CEO/Chair Self.ment.: fous high CEO+board Engagement: intimay low high Lexio-gram. photo omment Conjuntion high Hedge: passive high Attitude: evaluative high adv/adj Self-mention: nobility title high Lexio-gram. text Du-Eng Profile Br-Eng Profile Booster: emphati higher lower adv/adj Self-ment.: 1 st lower higher pers.plur.pron. Lexio-gram. photo Code gloss: employees lower higher Self-ment.: employees lower higher Lexio-gram. text Du-Eng Oper.review Br-Eng Oper.review Text stager higher lower Note. a. M = move, S = strategy, M+S = move + strategy. The results in Table 8.2 allow for an overarhing response to RQs 1d and 2. Judging from these results, most ross-ultural differenes emerged at the lexiogrammatial level of analysis and with regard to the Duth-English and British management statements in partiular. Aordingly, differenes aross the two national business ultures mainly ourred for the most onrete and detailed units of observation (e.g. Bhatia, 1993). The large number of ross-ultural variations for the management statements is likely to be aused by the fat that the British-English annual reports inluded two statement types, while the Duth-English annual reports ontained only one statement type whih resembled both the British Chairman s statement and the British CEO s statement. Moreover, ross-ultural differenes ourred espeially within the writing-as-mode, whih means that the Duth-English and British annual report setions were most different with respet to the rhetorial

240 devies used within the written running texts. It appeared that the generi design of the Duth-English and British annual reports was partiularly similar in terms of ommuniative ontext and ontent-related text features. The subtle ross-ultural variations and the high degree of overlap that were observed in the ontextual features of the annual reports would seem to form one of the main reasons why the Duth-English and British reports are generally reognized as instanes of the same genre set (e.g. see Askehave & Swales, 2001; Swales, 1990). The relatively small number of ross-ultural differenes in themati ontents in partiular seem to onfirm the assumption in setion 5.1 that genre ontent in itself may not define a partiular text type in a given rhetorial setting. As an be seen in Table 8.2, the singular analysis of genre ontent would have ignored additional ross-ultural variations in genre features at different levels of analysis, i.e. at the levels of rhetorial struture and lexio-grammar. Finally, Table 8.2 indiates that the fewest differenes emerged for the Duth-English and British orporate profiles and for the photographi modes in general. The overall orrespondene in multimodal rhetori aross the Duth-English and British profiles may be underpinned by the rossultural sharing of basi ontext features in these setions (e.g. with regard to purposes, readers, see setion 4.4.4). The relatively small ross-ultural variation in visual genre features implies that Duth and British orporations largely relied on the same rhetorial elements in the onstrut of doumentary photographs for the annual report. Complementary to the ulture-speifi items in Table 8.2, subsetions will provide a more elaborate summary of the ross-ultural results found for the rhetorial ontext, ontent, struture and lexio-grammar of the Duth-English and British-English annual report texts. As suh, they expand and speify the answers to the speifi RQs 1a-d and 2a-2, for eah level of analysis. In order to address RQ 2d, the following subsetions will also onsider the ontext-based explanations that were proposed for the ross-ultural results in multimodal text features (see Chapters 5 to 7). It should again be noted that the summaries below do not expliitly refer to the analyses of genre features per national business ulture: sine these analyses are inherent in the ross-ultural omparison of genre features in the Duth- English and British annual report texts, setions will primarily disuss the ontrastive results. Consistent with the seond main aim of this study, setions 8.6 and 8.7 outline what impliations the present researh results have for genre theory and genre analysis in general. Finally, this hapter will identify several limitations of the urrent study and will propose suggestions for further genre-based investigations (setion 8.8). 8.1 Genre ontext The ross-ultural desription and omparison of genre ontext were based on a qualitative survey, whih mapped the language poliies for the English annual reports of Duth orporations only, and in-depth interviews, whih analyzed the prodution ontext for the Duth-based and UK-based annual reports. After giving an answer to RQ 1a, onerning the English language poliies, this subsetion will fous on the similarities and differenes in the prodution ontext for the Duth- English and British annual reports (RQ 1d). The answers to RQs 1b-1, i.e. the ulture-speifi desriptions of this ontext, were already provided in the tables of results in Chapter 4.

241 As for the language poliies in annual reports (RQ 1a), a majority of the Duth listed ompanies already laimed to use English as the base language for these texts. They explained that original English annual reports are usually established through a proess in whih preliminary English reviews from Duthbased and overseas divisions are ombined in a unified doument produed by either native English speakers or by a group of non-native and native speakers of English. Although it ould be argued that the involvement of native English speakers in the prodution proess of Duth-English annual reports might have redued Duth language influenes, the present fous on ulture-speifi (rather than languagespeifi) influenes on texts used by traditional Duth orporations still validates the ontrastive analysis of English annual reports originating in the Netherlands and in a native English business ulture (i.e. British). The ulture-speifi analysis of genre ontext (RQs 1b-1) and the subsequent omparison of genre ontext aross the two national business ultures (RQ 1d) revealed that the Duth-English and British annual reports were realized in similar but not idential ommuniative situations. With regard to the ross-ultural similarities in ontextual features of the English annual reports, both Duth and British ompanies believed that English is an established global finanial language whih enhanes ommerial suess and guarantees legal ompliane on international markets. They also agreed that the annual report (1) aims to inform stakeholders about the (strategi) identity and performane of the ompany, and (2) serves the purpose of establishing a positive orporate image. It was suggested that the English annual reports should fulfill these ommuniative objetives with a similar range of finanial and non-finanial reader groups. Furthermore, the tables of ontents in the Duth-English and British annual reports were laimed to be based on largely the same legal, soio-politial, historial and peer-related riteria. Contiguously, the tables of ontents in these reports inluded similar text types, e.g. orporate profiles, operational and finanial reviews or orporate governane reports. Aording to Duth and British ompanies, the photographs in the annual reports need to be purposeful and ompany-related, i.e. they need to appeal to reader emotions and refer to the orporate identity, the value of the orporate produt and the themes in orporate marketing ativities. Finally, it was observed that the prodution proess of both the Duth-English and British annual reports is ommonly orhestrated by a oordinator-writer who ooperates with graphi designers, departmental offiers and board members to draft and finalize the multimodal annual report. The ross-ultural differenes in ontextual features of the English annual reports were already displayed in Table 8.2. It indiates that only Duth ompanies tended to apply English in their annual reports to meet the information needs of non- Duth partiipants in the internal organizational struture. Additionally, the Duth ompanies foused on the annual report as a finanial ommuniation tool, while the British ompanies tended to use it as both a finanial and a marketing ommuniation instrument. Another distintion involved the priority in reader groups, where the Duth ompanies prioritized all finanial and group-internal readers and the British orporations primarily foused on shareholders. There were also ross-ultural variations in the use of guidelines for the tables of ontents of the English annual reports and in the atual omposition of these tables of ontents. For example, annual reports of two-tier orporations in the Netherlands ontained a CEO s statement representing the exeutive board and a separate report of the supervisory board; onversely, annual reports of one-tier ompanies in the UK inluded a CEO s statement refleting the opinion of the exeutive board members,

242 a Chairman s statement representing the non-exeutive board members and a diretors report by the whole board of diretors. With regard to the rhetorial value of images, only British ompanies added that they use photographs to improve the readability of the annual report texts. And finally, Duth ompanies as opposed to UK ompanies laimed that the supervisory or non-exeutive diretors are usually involved in the final approval of the annual report before it is published. In line with the generi text-in-ontext perspetive (Askehave & Swales, 2001; Bhatia, 1993; Nikerson, 2000; Swales, 1990), the results of the situational analysis ould be related to the findings in the analyses of multimodal text features (summarized in setions ). As was pointed out at the beginning of this hapter, the present researh results suggested that orrespondene in the ontextual features of the Duth-English and British annual reports was refleted in rossultural similarities for the textual and visual themes, moves or strategies and metadisourse. With referene to the Duth-English and British orporate profiles, for instane, all substantive and formal elements underpinned the shared ommuniative purposes of the text, i.e. to highlight orporate strengths and to identify the ompany for new readers and for the ontextualization of results in the sueeding annual report setions (see setion 4.4.4). Several of the situational differenes aross the Duth-English and British annual report texts also seemed to offer a diret explanation for ross-ultural variations in multimodal text features. As an illustration, the high frequenies of ourrene for the orporate strategy theme and the operational performane move within the Duth and British CEO s statements were more than likely aused by the fat that these statements (as opposed to the British Chairman s statements) should provide a strategi summary of finanial and operational results. On the other hand, there were ross-ultural variations in textual or visual features that only seemed to have an impliit or tentative link with the ontextual aspets examined in the present study. With respet to the Duth-English operational reviews, for example, it was implied in setion that the prominene of onsiderations of ustomer needs within the operational performane move was owing to the fat that these reviews aim to ommuniate with lients, shareholders or employees, while the British reviews mainly address an audiene of shareholders (see setion 4.4.4). It should also be noted that some ross-ultural differenes in multimodal text harateristis ould ertainly not be explained by the urrent ontext-related results. The frequent ourrene of text stagers in the Duth-English management statements and operational reviews, for instane, ould not be asribed to any of the aspets analyzed in the prodution ontext of the Duth-English and British annual reports. Instead, this ross-ultural differene in textual metadisourse was aounted for by national ultural norms desribed in previous researh (i.e. with regard to lowontext ultures, see Claes & Gerritsen, 2007). 8.2 Genre ontent The identifiation and omparison of genre ontent in the Duth-English and British annual reports were based on the analysis of textual and visual themes. This subsetion disusses the main similarities and differenes that were observed for these themes within the management statements, orporate profiles and operational reviews (RQ 2). The ulture-speifi desriptions of oneptual themes (RQs 2a- 2b) were already displayed in the tables of results in Chapter 5. At the end of this

243 subsetion, the findings of the ontrastive ontent analysis will be plaed within the broader rhetorial ontext of the annual report texts (RQ 2d). The analysis of multimodal genre ontent revealed that speifi setions of the Duth-English and British annual reports ontained both idential and distintive themes. As for the ross-ultural similarities, the Duth-English and British management statements showed orrespondene in the use of performane-related, market-related and onfidene-related text themes and in the use of general photo themes related to members of management. Sine many of the Duth-English and British management statements inluded images of diretors, themes in the text fragments adjaent to these images mainly involved the identity of the manager(s) depited. The written texts of the English orporate profiles were thematially speified in both business ultures, i.e. they predominantly reported on themes related to the orporate strategy and identity. As the images in these profiles often refleted visual themes related to the orporate workplae and the (appliation of) orporate produts or servies in partiular, the adjoining text fragments mainly ontained the theme orporate produt. Both the Duth-English and British operational reviews inluded a variety of written text themes, ranging from divisional results to divisional strategies and strengths. These reviews also showed similar photographi ontents, where pitures refleted visual themes related to ustomers, employees, workplae failities and produts. Congruous with the visual themes, the photo-related omments primarily ontained text themes desribing the divisional operations and produts illustrated in the image. Table 8.2 shows that there were also several ross-ultural differenes in the themati onstrut of the Duth-English and British annual report texts. With regard to the management statements, ross-ultural differenes onerned textual themes overing the orporate strategy, dividend, governane, sustainability and stakeholder appreiation. For example, the board details and dividend themes were typial of the British Chairman s statements. The analysis of photo themes showed that produers of the Duth-English statements preferred portrayals of the olletive management teams, while produers of the British CEO s statements preferred photo themes related to individual managers, employees and orporate workplae failities. Only one ross-ultural differene was found for the themati ontent of the Duth-English and British orporate profiles: there was again a typial British preferene for employee-related themes in the photographs inorporated in these texts. Finally, the analysis of text themes in the operational reviews revealed that the ost management and the market identity themes were more prominent in the Duth-English than in the British texts. Considering RQ 2d, the ross-ultural results for textual and visual themes in the speifi annual report setions ould be ontextualized, with respet to both their role in the full annual report and their relationship to the ommuniative setting. First, it was observed that several different text types in the British annual report (i.e. the CEO s statements and the profiles) typially ontained photographs refleting the employee theme. Sine this visual theme predominated in several British text types, it would seem that the theme was not merely bound to individual annual report setions originating in the British business ulture, but that it was typial of the British annual report or of British business disourse in general. Seond, the ontent analysis of photos indiated that doumentary images ontributed both to the themati distintion between text types and to the rossultural distintion between themes in the same text type. For instane, the diversity of visual themes was larger in the operational reviews than in the management statements of the Duth-English and British annual reports. Also, the ross-ultural

244 variation aross Duth-English and British profiles in partiular was determined by the photographi employee theme. A third observation onerns the situational explanation of ross-ultural differenes in genre ontent. Many of these differenes ould be assoiated with (1) the ommuniative purpose of the annual report setion (see RQs 1a-1d), e.g. the prominene of the appreiative HRM/People theme in the interational Duth-English CEO s statements and British Chairman s statements but not in the narrative British CEO s statements, (2) the professional mandate of the writer, e.g. the prominene of the dividend theme in the statements of the British supervisory Chairman but not in the Duth-English CEO s statements, (3) soio-politial developments, e.g. the introdution of the Duth Tabaksblat ode in early 2004 whih presumably aused the prominene of the orporate governane theme in the Duth-English CEO s statements but not in the British statements and (4) national soial values, e.g. the prominene of photo themes related to the full board in Duth-English management statements that were produed within a ulture marked by a more feminine and ollaborative nature (e.g. Claes & Gerritsen, 2007; Hofstede, 2001). 8.3 Genre struture The rhetorial strutures of the Duth-English and British annual report texts were mapped and ontrasted in terms of textual and visual moves and strategies. The urrent subsetion fouses on the omparative findings of the strutural analysis of Duth-English and British management statements, profiles and operational reviews (RQ 2). These findings were based on the ulture-speifi identifiation of moves and strategies (RQs 2a-2b) that was already dealt with in the results setions of Chapter 6. In the disussion below, the ross-ultural observations for moves and strategies will be assoiated with ontextual aspets (RQ 2d). The study of strutural elements revealed both ross-ultural overlap and ross-ultural variation in the use of textual and visual moves and strategies. Crossultural similarities in the rhetorial struture of the Duth-English and British management statements onerned textual moves and strategies that reported on the (finanial) performane, strategy or strengths of the ompany, that evaluated the performane ontext or the future ourse of the ompany and that established a reader-writer relationship. The moves and strategies in the photos and in the adjoining text fragments of the managerial statements primarily ontributed to the text purpose by identifying the ompany and providing it with a personal fae. The written texts in the Duth-English and British orporate profiles shared moves and strategies identifying the ore harateristis, produts and strengths of the ompany. Moves and strategies in the images and adjaent omments within the profiles mainly foused on the general or personal identifiation of the ompany represented in the text and on the desription or valuation of the orporate produt. Both the Duth-based and UK-based operational reviews were onstituted by a range of textual moves and strategies, e.g. reporting on past performane of divisions, explaining results against a strategi bakground, evaluating the performane ontext, speulating on divisional ahievements in the future and improving reader-friendliness of the text. Visual moves and strategies mainly involved the (personal) identifiation of the ompany and the illustration of orporate produts mentioned in the running text. The photo-related text fragments

245 in the English reviews ommonly inluded moves and strategies that presented the ompany s identity, produts and operational ations. As an be seen in Table 8.2, ross-ultural differenes in the rhetorial struture of speifi annual report setions only onerned textual elements. Aross the Duth-English and British management statements, there was signifiant rossultural variation for textual moves and strategies desribing board hanges, future objetives and operational performane (based on finanial details) as well as for textual moves and strategies realizing appreiation for stakeholders, reader friendliness of the text and reader-writer interation. The addressing the reader move ombined with the salutation strategy, for instane, was more typial of the Duth-English than the British management statements. The Duth-English and British orporate profiles only varied substantially in the use of textual moves reporting on orporate strategy; these moves were enountered in onsiderably more Duth-English than British profiles. In the Duth-English and British operational reviews, ross-ultural differenes were found for textual moves and strategies profiling orporate divisions, highlighting ustomer needs underlying operational results, outlining strategi foi, desribing market-produt relations and setting short-term future objetives. For example, the general profile strategy ourred more often in the Duth-English than in the British operational reviews. With respet to the photo-related omments in the reviews, there were signifiantly more British than Duth-English reviews that typially inluded strategies stressing the orporate strengths in these fragments. To find an answer to RQ 2d, the results of the strutural investigation were plaed in-ontext, i.e. in the text-internal and in the text-external ontext of the full annual report. First, it appeared that ross-ultural distintions in moves and strategies were onfined to speifi types of texts within the annual report. The analysis revealed a variety of differenes between the Duth-English and British annual report texts, but none of those seemed to reflet ulture-speifi patterns maintained throughout the annual report. Some signifiant findings were even ontraditory aross text types: the move speulating on future performane with the strategy short-term objetives ourred more often in the Duth-English than in the British management statements, while it was more popular in the British than in the Duth-English operational reviews. Furthermore, the researh results suggested that doumentary photographs played a minor role in the strutural definition of text types or of ross-ultural differenes in text types inluded in the annual report. There was regular overlap in the rhetorial funtion of visual moves and strategies aross text types; different Duth-English and British annual report setions ontained photographs with similar strutural labels, e.g. the visual move detailing the produt together with the strategy desription of the produt ourred in all Duth-English and British annual report setions. Additionally, no signifiant rossultural distintions were found for the moves and strategies embedded in the photos. This suggests that Duth and British orporations asribed similar strutural funtions to the images in their English annual reports. Finally, it was noted that some strutural differenes aross the two national business ultures ould be explained by situational fators. These ross-ultural variations might have been produed by (1) the ommuniative purpose of the annual report setion (see RQs 1a-1d), e.g. the prominene of the reporting on operational performane move in the Duth-English CEO s statements (whih aim to inform about operations) but not in the British Chairman s statements (whih fous on a summary of finanial results), (2) the professional mandate of the writer, e.g. the prominene of the reporting on leadership move in the texts by the British supervisory Chairman but

246 not in the Duth-English CEO s statements, or (3) national soial values, e.g. the prominene of the move addressing the reader with strategy salutation in introdutory management statements originating within the more feminine Duth business ulture (e.g. Claes & Gerritsen, 2007; Hofstede, 2001). 8.4 Generi lexio-grammar A model of metadisourse was used to desribe and ompare the lexio-grammatial features of the multimodal disourse in the Duth-English and British annual reports. This subsetion enters on the ross-ultural similarities and differenes that were diserned for the textual and visual metadisourse in these texts (RQ 2). The ulture-speifi desriptions of metadisourse that formed the basis of these rossultural results (RQs 2a-2b) were inorporated in the results tables of Chapter 7. To address RQ 2d, this subsetion will also inlude a ontextualization of results. The analysis of metadisursive items in the written texts and photographs showed that there were ross-ultural similarities as well as many ross-ultural differenes in the lexio-grammatial onstrution of the Duth-English and British annual report texts. With regard to the ross-ultural similarities, the Duth-English and British management statements ontained a relatively equal distribution of textual metadisourse in the form of transitions, disourse ats, information (p)reviews, ontent speifiations, hedges, emphatis, evaluative expressions and (pro)nominal self-referenes. The metadisourse in the photos of these statements mainly onerned produt-related or workplae-related ode glosses and attitude markers related to professionalism or ommitment. It was observed that the photorelated omments in the English management statements orresponded for metadisursive items staging the reader s viewing of the image and relating the visual information to the orporate name. Both the Duth-English and British orporate profiles inluded text-based onjuntions and ode glosses, and a relatively large variety of interational items suh as hedges, empathi verbs, attitude markers and nominal self-mentions. Cross-ultural orrespondene in visual metadisourse involved produt-related, work area-related or market-related ode glosses, boosters emphasizing the impressiveness of the produt and attitude markers related to the professionalism, modernity or added value of the produt. The photo-related text omments in the profiles primarily ontained endophori markers, evaluative expressions and self mentions in the form of orporate names. Textual metadisourse that was prominent in the Duth-English and British operational reviews omprised transitions, illoutionary ats, endophori markers, award-related evidentials, ode glosses, hedges, boosters, positive attitude markers and (pro)nominal self-mentions. The images in these reviews refleted a range of visual metadisursive items, e.g. produt-related or faility-related ode glosses, devies boosting the impressiveness of produts or work plaes, attitude markers depiting the ompany or its produts as modern, professional and adding to onsumers lives, self-mentions referring to divisional diretors and indiators of reader-ompany intimay. Photo-related omments in the English reviews revealed a similar use of interative and interational resoures (e.g. transitions, frame markers, evidentials, boosters, engagement markers). The relatively large number of ross-ultural differenes in textual and visual metadisourse were highlighted in Table 8.2 at the beginning of this hapter. Textual metadisourse within the Duth-English and British management statements

247 differed only in terms of items staging the text, addressing the reader and referring to the author as management. Text stagers, for instane, were more ommon in the Duth-English than in the British management statements. Cross-ultural variations in visual metadisourse onerned ode glosses identifying the ompany by its board members, employee-related or market(ing)-related ode glosses, attitude markers stressing orporate sinerity and modernity, self-referenes in the form of management portraits and engagement markers reating intimay with the ompany. Visual ode glosses illustrating the ompany by its management team, for example, ourred more often in the Duth-English than in the British management statements. The photo-related omments varied aross ultures for textual onjuntions, passives, evaluative expressions and self-mentions in the form of nobility titles. Nobility titles, for instane, were most typial of the British Chairman s statements. In the running texts of the orporate profiles, ross-ultural differenes were found for boosting lexis, whih ourred more in the Duth- English profiles, and for self-mentions in the form of first person plural pronouns, whih were more prominent in the British profiles. There were ross-ultural variations for visual ode glosses identifying the ompany by its employees and employee-related self-mentions, both of whih ourred more often in the British- English profiles. The operational reviews revealed only one ross-ultural distintion in textual metadisourse: text stagers were more typial of the Duth- English than the British reviews. To answer the question as to how the metadisursive similarities and differenes aross ultures an be explained (RQ 2d), the results of the metadisourse analysis were further ontextualized. It was noted first of all that several ulture-speifi preferenes in metadisourse reourred aross different types of annual report texts. As an illustration, frame markers in the form of text stagers were most ommon in the written texts of the Duth-English management statements and the Duth-English operational reviews. In addition, visual ode glosses identifying the orporation by its employees were found most often in the photographs of the British CEO s statements and the British profiles. These findings may suggest that ross-ultural differenes in textual or visual metadisourse our beyond the level of speifi text type and involve ulture-speifi preferenes for multimodal disourse in the full annual report or in business disourse in general. Congruous with the first observation, the metadisursive funtion of doumentary images was found to ontribute to the identifiation of different text types within the Duth-English and British annual reports and to the identifiation of ross-ultural variations aross these different text types. For example, the variety of visual metadisursive items in the English operational reviews was muh greater than the variety of items in the photographs within the English management statements and English profiles, eah of whih ontained their own unique sets of visual metadisursive features. With regard to the visual metadisourse in the management statements, moreover, a ross-ultural distintion was observed for visual ode glosses identifying the ompany by its management olletive: these ode glosses were more typial of the Duth management statements than of the British statements, whih foused on visual ode glosses identifying the ompany by its individual diretors. Finally, the present researh results implied a possible relationship between the ulture-speifi use of metadisourse and the prodution ontext of the annual report. It seemed that some results may be explained by (1) the ommuniative purpose of the annual report setion (see RQs 1a-1d), e.g. the prominene of engagement markers in the interational Duth-English CEO s and British Chairman s statements but not in the narrative British CEO s statement, (2)

248 the ommuniative purpose of the annual report (RQs 1a-1d), e.g. the relatively large variety in realizations of interational resoures as opposed to interative resoures to support the general aim of affeting reader attitudes, or (3) national soial values, e.g. the prominene of text stagers within management statements and operational reviews originating in the Netherlands, whih is a low-ontext ulture (lower than the low-ontext ulture in the UK) that is generally marked by a preferene for expliit language use (Claes & Gerritsen, 2007). 8.5 The English annual report as a global genre set The identifiation, omparison and ontextual explanation of genre features in the Duth-English and British annual reports (RQs 1a-1d, RQs 2a-2d) indiated that the generi design of the management statements, profiles and operational reviews in these reports was similar (albeit not idential). The results of the (on)text analyses showed that there was a relatively high degree of ross-ultural orrespondene in the prodution setting, in the textual and visual themes, in the textual and visual rhetorial struture, and in the textual and visual metadisourse of the speifi annual report setions. This suggests that produers of the English annual reports in Duth and British orporations had overlapping pereptions of the typial onstrut of this business text. Moreover, it would seem that suh overlap in (on)text harateristis further onfirmed the idea that Duth-based and UK-based produers of the annual report applied language use onventions that are universally reognized. That is, the genre analyses showed that these text produers adopted similar disourse standards by means of whih they realized a doument that they believed would be identifiable internationally as a orporate annual general report. Aordingly, it may be argued that the ross-ultural orrespondene found aross the Duth-English and British annual report texts underpins the notion that annual reports are universally onventionalized or global genres with similar rhetorial funtions and patterns of language use aross different national business ultures (e.g. Bhatia, 1993; Nikerson, 2002). Despite the large degree of ross-ultural similarity in genre features, the present investigation also revealed subtle differenes in the rhetorial ontext, ontent, struture and lexio-grammar of the Duth-English and British annual reports (see also Table 2). There was ross-ultural variation for both the textual and the visual elements in the annual report texts, thus implying that text produers in Duth and British orporations expressed ulture-dependent i.e. loal preferenes - in the multimodal onstrut of the English annual report (e.g. Nikerson, 2002). In general, signifiant generi differenes aross national business ultures existed in relation to (1) a partiular text type in the annual report, e.g. a substantially larger number of Duth-English than British operational reviews ontained the ost management theme, (2) a partiular level of analysis, e.g. frame markers in the form of text stagers ourred in a larger number of Duth-English than British management statements and operational reviews, and (3) a partiular business ulture, e.g. employee-related themes and metadisourse were more typial to the photographs in the British than in the Duth-English annual report texts. While the English annual reports originating in Duth and British orporations may be instanes of a global genre (i.e. a genre that is internationally reognized), these ross-ultural variations indiate that within suh a genre ompanies have some freedom in the manifestation of ontextual, substantive and formal disourse

249 features. These ulture-speifi approahes to business genres thus provide some insight into the relationship between individual ompanies disourse ativities and the soio-ultural system within whih these ompanies are situated (Mayes, 2005; Vergaro, 2002, 2005; Zhu, 2005). As was pointed out in the interim onlusions of Chapters 4 to 7, the overview of ross-ultural similarities and differenes in (on)text features also suggested that both Duth-based and UK-based produers of the English annual report regarded the annual report as a olletion of varied but oherent text types or genres that onstitute a uniform genre set. Drawing on Devitt (1991), the onept of genre set refers to a set of texts interating to aomplish the rhetorial funtion of the full annual report; the reurrent use of these texts have helped to define and stabilize the annual report as a genre set. Following Jameson s aount of US annual reports (2000), the setions or text types embedded within both the Duth-English and the British annual reports were genres in their own right; although onneted by intertextual relations, they often established a non-linear organization of these reports, allowing readers to enter, leave, and reenter the text at will, reating, in essene, their own struture of the reports (Jameson, 2000: 8). Considering the many ross-ultural similarities and subtle ross-ultural differenes in genre features, moreover, it an be onluded that both the Duth-English and British- English annual reports onsisted of omparable instanes of different genres that had been unified within the annual report as a genre set. In other words, the Duth- English and British annual reports omprised genre sets that were omparable aross ultures, i.e. global genre sets. Here, the omparability of instanes of genres involves the reognition of reurrent situations that require a partiular rhetorial response; although realizations of a genre may differ, when multiple texts serve to fulfill idential soial needs in a ertain rhetorial situation they are likely to be omparable instanes of the same genre (Miller, 1994). Or, in line with Mayes notion about Japanese and Amerian professional disourse, the Duth-English and British annual reports are omparable genre sets beause they address the same exigene, a ommon soial motive that is revealed through regular patterns of situated ation whih may be influened by loal preferenes in language use (2005: 342). 8.6 Impliations for genre theory Both the analytial frameworks and the findings of the genre investigations yielded several theoretial impliations that add to the development of genre theory as a disourse approah to (international) multimodal (business) texts. Hene, these impliations relate to the seond main researh purpose disussed at the beginning of this hapter (and in setion 1.5). A first theoretial impliation an be drawn from the oneptualizations of disourse ommunity in setions 2.2 and 4.1. Within genre theory, the disourse ommunity onept is primarily writer-oriented or expert-oriented (e.g. Hyland, 1998, 2007; Swales, 1990). Owing to its origins in the aademi disiplines, where experts usually ommuniate with other experts, this onept is problemati in analyses of business texts that are produed by expert ommuniators but that are distributed among a varied group of (non)expert readers. This was illustrated by the ontextual investigation of Duth-English and British annual reports (setion 4.4), whih showed that ommuniations or investor relations offiers need to reah a

250 range of reader audienes that are unlikely to share the offiers expertise and that have nothing in ommon but their interest in orporate ahievements, e.g. investors, employees, lients, non-governmental organizations or authorities. In order to aount for all the various disourse partiipants engaged in the use of a partiular business genre, it would therefore be relevant to further expand the definition of the ommunity onept. While the onept of disourse ommunity may be maintained for the exlusive identifiation of expert produers of (international) professional texts, it an be usefully omplemented by the onept of (global) reader ommunity to allow for the identifiation of the many potential reeivers of these texts (Mahin & Van Leeuwen, 2003). Seondly, in setion 2.4 it was suggested that the ESP/LSP genre theory would be a relevant means of inquiry for the analysis of business disourse suh as the international annual general report. Based on genre analysts omments (e.g. Luzón, 2005; Yli-Jokipii, 1998) and prior studies on business genres (e.g. Nikerson, 2000), it was antiipated in the present study that the genre perspetive would provide an analytial fous, i.e. that it would point to the written and visual elements by means of whih the Duth-English and British annual report texts ould be identified and ontrasted. Indeed, the methodology desription in setion 3.1 outlined that ESP/LSP genre theory provides four different levels of multimodal text analysis, i.e. the ommuniative ontext, the ontent, the rhetorial struture and the lexio-grammar refleted in the written texts and photographs of the annual reports. These four levels of analysis not only failitate a systemati investigation of various multimodal text elements, but also allow for the thik desription and omparison of internationally established business texts (i.e. annual reports) as produts of their soio-ultural ontexts (e.g. Bhatia, 1993). Following Nikerson, genre theory thus enables omparisons to be made aross [ ] genres used in different ultures [and with] similarities in terms of the situations and partiipants involved, and the ommuniative ations aomplished (2000: 187). Even though it was argued here that genre theory in priniple offers foused starting points for the omprehensive analysis of written texts and photographs in international annual reports, it was not very expliit with regard to the analytial frames and researh steps required for the full realization of this analysis. Although genre theory inorporates a psyholinguisti method for the analysis of rhetorial text struture (i.e. the identifiation of moves and strategies), it does not set riteria for the demaration of strutural elements or for the examination of genre ontext, ontent and lexio-grammar. In the present study of multimodal annual report texts, there was thus a need to develop an interdisiplinary frame inluding different disiplinary traditions in the ontextualized analysis of disourse (Nikerson, 2000: 187). As was explained in setion 3.1, the urrent interdisiplinary frame was based on a ombination of insights derived from ethnography, ognitive mapping, psyholinguistis and linguistis to establish the generi ontext, ontent, struture and lexio-grammar of annual report texts, respetively. Chapters 4-7 showed that for eah level of analysis, these disiplines were used to provide an analytial model relevant for the investigation of textual and visual realizations of a partiular genre feature. Judging from the results of the multimodal text analyses, genre theory as a means of inquiry was made more omplete and more pratiable by the integration of different disiplinary approahes. Moreover, the interdisiplinary relevane of these approahes was emphasized by the observation that researh results aquired at one level of analysis ould be related to results obtained at a different level of analysis, e.g. the oherene between data from the ethnographi ontext analysis and data from the analysis of oneptual themes, moves and strategies, and

251 metadisourse. As suh, the multidisiplinary framework proved to be an important devie for the thik desription of professional texts (e.g. see also Askehave & Swales, 2001; Bhatia, 1993; Yli-Jokipii, 1998). A fourth theoretial impliation involves the analysis of photographs as onstituents of the multimodal business text. Exept for a few investigations (e.g. Askehave & Nielsen, 2005; Fuertes-Olivera, et al., 2001), the verbal-pitorial rhetori in professional disourse has reeived relatively little attention in genrebased researh. Considering that visuals in multimodal douments are often intended to supplement written texts (Kress & Van Leeuwen, 2001), that the design of visuals may differ aross ultures (Kress & Van Leeuwen, 2006) and that visual illustrations have well-defined purposes in Duth-English and British annual reports (see setion 4.4.6), it would seem to be a logial hoie to inlude photographi images in the genre study of annual reports or in studies of other multimodal business genres. Compared to prior researh, the ontextual, substantive and formal analysis of doumentary photographs resulted in a more omplete identifiation of annual report texts produed in two different national business ultures. Additionally, it aused this identifiation to be more aurate. In eah national business ulture, the analysis of photos helped to unover differenes between text types within the annual report (e.g. the management-related photo themes in management statements versus the produt-related photo themes in operational reviews). Furthermore, it helped to reveal ross-ultural variations for a speifi text type within the annual report (e.g. the photo theme all exeutive board members in the Duth-English management statements versus the photo theme individual CEO/Chairman in the British management statements). From a ross-ultural perspetive, the role of doumentary images was most distintive in the analysis of oneptual themes and metadisursive items, e.g. the visual employee theme was the only signifiant distintion between the Duth-English and British orporate profiles; it was least distintive in the analysis of moves and strategies, where the analysis of photographs did not yield any ross-ultural variations. Finally, the study of multimodal annual reports suggested that the typiality of partiular kinds of genre features may depend on the ultures and text types analyzed. This impliation follows on from the fat that several rossultural differenes observed in the present investigation did not orrespond with findings in previous studies on professional genres. Based on his analyses of promotional and legal texts in Anglo-Saxon and Asian ultures, for example, Bhatia (1993) antiipates that ultural onstraints are unlikely to have a substantial influene on the move struture of a genre, but that it is very likely that suh onstraints affet the strategi realizations of moves. Sine the strutural analysis presented in Chapter 6 revealed that ross-ultural differenes onerned moves as muh as rhetorial strategies (see also Table 8.2 above), the urrent researh results suggest that Bhatia s laim does not hold for ultural variations aross English annual reports of Duth and British orporations. A similar argument applies to Nikerson (2000), who finds that Duth text produers prefer first person plural pronouns in their orporate s whereas British writers prefer the use of first person singular pronouns. These results are in ontrast with the present researh results for textual metadisourse in orporate profiles (setion 7.4), whih showed that a signifiantly larger number of British than Duth-English texts ontained first person plural pronouns. Suh onfliting findings aross studies support the notion that eah text type should indeed be regarded as a produt of its own design ontext in a given soio-ultural situation (Swales, 1990). For instane, the multimodal features that are typial of the onstrut of Duth-English annual report texts may

252 not be typial of the design of annual report texts or other professional genres originating in other national business ultures. Likewise, the multimodal features that are typial of the onstrut of Duth-English annual report texts may not be typial of the design of other business texts within the Duth business ulture. 8.7 Impliations for genre analysis In addition to the impliations that add to the onept of genre as a theory, the present study put forward several impliations for genre as an analytial approah. The impliations for genre analysis primarily emerged from the multimodal genre investigations disussed in Chapters 4-7. Together with the theoretial impliations, the researh impliations realize the seond main researh purpose, i.e. the development of genre theory as a means of inquiry for international and multimodal annual report texts. First, it should be noted that the researh models and the identifiation of genre features grew more omplex as the investigation shifted to more onrete levels of analysis. In general, genre analysts have aknowledged the omplexities involved in the generi approah to disourse. Bhatia (2002), for instane, refers to the implementation of both disourse onventions and individual exploitations in the realization of genre purpose or genre struture. Swales and Feak (2000), moreover, point out that long and omprehensive texts usually ontain large series of metadisursive items to redue the ognitive effort required for reader interpretations. However, none of these analysts seem to have foused on the omparative omplexity of the models and researh steps needed to establish the different levels of genre analysis, i.e. for the identifiation of genre ontext, ontent, struture and lexio-grammar. In partiular, the urrent study indiated that the analytial framework and the design that were required in the investigation of themati ontent were less elaborate than the researh frames and designs that were required in the identifiation of moves or strategies, and metadisourse. These differenes in omplexity would seem to depend on the size of the units of observation, the demaration of the units of observation and the ategorization of the units of observation. Whereas the present analysis of oneptual themes involved a relatively simple ategorization of abstrat units, the investigations of strutural elements and metadisursive items onerned the demaration and stratified ategorization of more onrete units of analysis. In the full sample study of 137 multimodal annual report texts, the themati analysis yielded a data set that was less omplex and more manageable than the data sets aquired in the strutural and lexio-grammatial analyses. In addition, the present study implies that these omplexities may result in laborious researh models and designs in the ase of omprehensive analyses of large orpora of lengthy multimodal business texts. The omplexities in arrying out a omplete genre researh ould in general be irumvented by small-sale text analyses of onise genres or genre analyses that revolve around predetermined ategories of text features (e.g. Bhatia, 1993; Bondi, 2002; Dahl, 2005; Hyland, 1998a, Zhu, 2000); A seond analytial impliation onerns the interation between multimodal genre features at different levels of analysis. In his introdutory work on ESP genre theory, Swales (1990) already emphasizes the relationship between ommuniative ontext and genre design; he laims that the rhetorial situation in whih a genre is used shapes the shemati struture of the disourse and

253 influenes and onstrains hoie of ontent and [linguisti] style (Swales, 1990: 58). Similar ideas an be found in Bhatia (1993, 2004), Nikerson (2000) and Zhu (2005). The genre analyses presented in Chapters 4-7 revealed orresponding textontext assoiations, e.g. the prominene of text fragments identifying the ompany in the Duth-English and British orporate profiles or the prominene of text stagers in annual reports originating in the low-ontext business ulture of the Netherlands. Moreover, these analyses suggested an interdependene between various text-based genre features. Interrelations between text features primarily arose when a genre element at one level of the analysis helped to demarate or determine a genre element at another level of the analysis (and thus aused an overlap of results). For example, the board details theme in the written texts of the English management statements was fundamental to the definition of the textual move reporting on the leadership situation in these statements. Similarly, the photo themes all exeutive board members and employees in the management statements oinided with the visual (symboli) ode glosses fous on CEO and board members and orporate employees, respetively. Aordingly, it would seem that genre explanations do not only benefit from the identifiation of ontext-text relations but may also gain from observations of the interation between text-based features aross different levels of analysis. As the interrelations between disourse ontext and text features on the one hand, and between different types of text features on the other, add to the thik desription of a genre i.e. they provide a funtional and grounded desription of language use (Bhatia, 1993: 5) they should be inherent in genre analyses in general. Another aspet of the present study that adds to the refinement of omparative genre analysis in general involves the statistial approah to data proessing. With the exeption of a few statistis-based genre publiations (e.g. Herring & Paolillo, 2006; Rutherford, 2005), quantitative inventories of text features aross (instanes of) genres have mainly been restrited to absolute ounts or perentages of frequeny of ourrene, in both large-sale and small-sale examinations (e.g. Flowerdew & Dudley-Evans, 202; Hyland, 1998b, 2004; Nikerson, 2000; Swales & Rogers, 1995). Sine the urrent genre investigation foused on testing the relationship between national business ulture and genre features in annual report texts, it inorporated statistial tests. These were used in the ross-ultural analyses of genre items to help determine to what extent any variations aross the orpora of Duth-English and British-English texts might be expeted to arise aross the populations of all Duth-English and British annual report texts. The signifiane of patterns in language use annot be tested by means of basi univariate analyses (e.g. frequeny ounts or frequeny perentages) alone; therefore, the inlusion of statistial tests (here: bivariate Chi-Square tests) tends to strengthen the researh findings of genre analyses aross text types or aross ultures. It must be added, moreover, that statistis-based findings need not onflit with genre assumptions of the ontextualization of disourse. Although the intersetion between the quantitative and qualitative data in the present study were not analyzed statistially, the results for speifi genre features ould first of all be embedded within qualitative situational onsiderations. For example, the introdution of the Tabaksblat ode supposedly triggered the signifiant prominene of the orporate governane theme in the written texts of the Duth-English management statements. Seondly, the quantitative results in the text analyses did not emerge from the examination of single (i.e. superfiial) lexio-grammatial items but were based on onsiderations of text ontext, e.g. the identifiation of oneptual themes by assoiating ongruent key words/visual elements or the

254 labeling of metadisursive items by examining their position within a set of onordanes. In sum, the appliation of statistial tests in ontrastive genre investigations yields more solid or reliable results that support a generi approah to disourse. With regard to the analysis of photographs in the annual report as a multimodal genre set, the present study also showed that it was possible to assign the visual and written text elements to the same ategories in the genre-based researh models. This ould be aomplished by integrating oneptualizations of visual rhetori in traditionally text-based researh frameworks. For instane, the strutural role of photographs ould be defined in terms of Bhatia s (1993) moves and strategies and the metadisursive relevane of photographs ould be expressed in terms of Hyland and Tse s (2004) resoures and funtions. Here, the strutural ontribution of visuals was based on the rhetorial funtion of photographs in relation to the full text purpose (Askehave & Nielsen, 2005) and the metadisursive ontribution of visuals was based on the semioti signs onstituting the photographs (Kress & Van Leeuwen, 2006; Wollon, 1998). Moreover, the investigation of photographi elements suggested that visual features found in the multimodal genres might be assoiated with the role of pitures within the prodution ontext of these genres (whih was desribed within the frame of the disourse ommunity onept by Beaufort, 1997). For example, themes and strutural elements refleted in portraits of diretors or employees ould be explained by the fat that photos in Duth-based and UK-based annual reports serve the purpose of providing the ompany with a personal fae. In addition, the ontextual role of photographs formed the basis for the establishment of a model for multimodal metadisourse in Chapter 7. Aordingly, there was onsisteny in the models for the analysis of the prodution ontext of multimodal annual report texts and for the investigation of multimodal ontent, struture and lexio-grammar in these texts. The onsisteny in researh models at eah level of analysis allowed for the aquisition of textual and visual data of a similar texture, therefore enabling the integration of photographs as a ontextual, substantial and formal supplement to written texts in professional genres originating in different national business ultures. As suh, these observations endorse the usefulness and appliability of visual analyses in genre identifiation, and they underpin the idea that ontext-text relations involve visual disourse as muh as written disourse. A final impliation for genre analysis relates to the exeution of interoder tests as a means of establishing the reliability of researh results. It was already explained in setion 3.5 that very few genre studies have verified the extent to whih the identifiation and labeling of genre features in the researh texts were atually reliable (e.g. Crookes, 1986; Hyland, 1998b; Kanoksilapatham, 2005; Upton & Connor, 2001). Assuming that the generi approah to disourse is liable to soioognitive and thus subjetive interpretations of text features (Paltridge, 1994), it was deided that the present study should inlude interoder tests for eah level of text analysis in order to enhane the objetivity and thus redibility of the full-sample investigations (this exludes the qualitative analysis of genre ontext). The interoder reliability tests learly indiated what analytial problems were inherent in the urrent identifiation of different types of genre features. More speifially, these tests revealed that the demaration of textual moves and strategies, and the labeling of semioti signs as items of visual metadisourse required strit ategorizations in order to reate a workable researh situation for multiple oders. With regard to the moves and strategies, an expliit lassifiation system resulted in plausible Kappa values for textual elements in partiular. For the items of visual

255 metadisourse, however, the reation of a semioti lassifiation system still did not guarantee an aeptable degree of interoder orrespondene. In the latter situation, the oders often agreed on the overall metadisursive meaning of an image but they showed different qualifiations for the ioni, indexial or symboli onstituents of this meaning. These findings might explain why interoder reliability tests have reeived relatively little attention in analyses of rhetorial text strutures and in semiotis-based analyses of images in partiular. As suh analyses depend on omplex ognitive proesses of meaning-making, it will be diffiult to reah a suffiient degree of interoder agreement for the identifiation of the units of observation. Given these diffiulties, however, the exeution of interoder tests remains a relevant means for substantiating the final outome of the study. 8.8 Limitations of the ross-ultural genre study and suggestions for further researh Although the ross-ultural analyses of (on)text features desribed in Chapters 4 to 7 were arried out on the basis of deliberate hoies of materials, national business ultures and theoretial and analytial frameworks, the exploratory nature of these analyses yielded some limitations that reflet suggestions for further researh. These will be disussed below. Fous of the ross-ultural genre study The seletion of Duth-English and British-English annual report texts as the primary genres under investigation brought about a relevant but limited fous within the present genre study. Other types of written or oral disourse and other national business ultures were exluded from this study, thus rendering researh results that only apply to the sampled researh materials and the Duth and British business ultures. Considering this limitation, it would be useful to expand the investigation and look at different business texts, different ultures or even different languages. With regard to the generi orrespondene between the English annual reports and other English-language business texts, it was already indiated that the typiality of genre features may be influened by both the seleted text types and the seleted ultures (see setion 8.6). Nevertheless, the question remains as to whether the ross-ultural results found for the Duth-English and British annual reports an also be traed in other related text types inluded within the genre repertoires to whih the annual reports belong (e.g. in interim reports, shareholder meetings, press releases). That is, it would be interesting to analyze if these related text types are haraterized by intertextual similarities in their generi onstruts. The internationalization of ompanies within and outside Europe, moreover, inreases the need for further researh of English annual reports originating in different national business ultures. It would be fruitful, for example, to ompare the genre ontext, ontent, struture and/or lexio-grammar of international English annual reports published by Frenh multinationals (suh as BNP Paribas) with the genre features in the Duth-English and British annual reports analyzed here. Comparisons of English annual reports aross additional national business ultures ould larify the ulture-related questions. Despite the assumption that the international English annual report is affeted by the ultural bakground in whih it is produed, it may finally be of interest to determine to what extent the non-native English annual report atually reflets ulture-speifi features that are also inherent in the native

256 language of its ulture of origin. With referene to the present ase, it would be insightful to determine to what degree the Duth-English annual report texts reflet harateristis that would also be inorporated in a native Duth annual report. Analytial framework in the ross-ultural genre study Beause the researh models in the urrent genre analyses were developed for the identifiation of multimodal genre features in English annual reports in partiular, there is a need for further testing of these models in relation to annual reports or other professional genres aross a variety of different (business) ultures. Although the present models were based on previous studies on professional genres, it is likely that the present versions of the researh models will have to be adjusted or extended in order to make them fully appliable to other text types besides the annual report. As an illustration, sine the lists of oneptual themes and visual metadisursive items are speifially attuned to the ommuniative situation of the multimodal annual report, the urrent labels of these genre elements may not be appropriate to other professional text types. Nonetheless, the fundamental usefulness of the researh models may be measured by suessive genre studies that follow the same basi theoretial and analytial steps in establishing and exeuting analyses of textspeifi features. The fruitfulness of translating a researh model from one genre to another has already been shown in numerous analyses of moves and strategies in professional genres (e.g. Bhatia, 1993, 2004; Henry & Roseberry, 2001; Nikerson, 2000), but it may just as well be appliable to the researh model for oneptual themes and visual metadisourse. For instane, the model of visual metadisourse ould be adapted in aordane with the ommuniative funtion text produers asribe to photographs in different business genres. Contextual dimensions in the ross-ultural genre study The analysis of the ommuniative situation of the annual report in Chapter 4 entered on the onventions and proesses that influene the prodution of this multimodal business text. Consequently, the ross-ultural inventory of ontextual genre features (1) ontained only a limited number of ontextual dimensions and (2) was primarily onditioned by respondents expertise in text prodution. In other words, this inventory did not offer insight into higher-order soial values or norms that typially influene the English annual report within the Duth and British business ultures, e.g. the degree in whih these ultures are feminine or masuline. Sine the present study did not provide ulture-speifi data on the existene of suh soial rules within the design ontext of the Duth-English and British annual reports, their potential impat ould only be explained tentatively by means of theoretial referenes to prior studies on ross-ultural or interultural ommuniation (e.g. Claes & Gerritsen, 2007; Hofstede, 2001). Therefore, later investigations should explore extra ontextual dimensions in order to establish a broader (i.e. national-soial) empirial understanding of the ultural effets on Duth-English and British-English annual reports. Subsequently, these investigations ould also ontribute to the further ultural investigation of other kinds of Duth(-English) and British business texts. An additional ontextual fator that also needs to be reognized in suessive studies on the annual report is the dynami nature of this genre. The rhetorial funtion of the annual report in general has hanged several times sine its first publiations, i.e. from a finanial fat sheet into an investor relations tool (Hynes & Bexley, 2003). Furthermore, the introdution of international finanial reporting standards have influened the ontents or language poliies for annual

257 reports (e.g. IASPlus, 2005; ING, 2006). And the inreasing importane of the web has aused new interests in PDF-versions or HTML-versions of the originally printed annual report. Although the status of the PDF-file may largely orrespond with the status of the printed annual report, the HTML-version of the annual report whih allows for links with other types of web texts is likely to alter the nature of the (printed) annual report as a genre set due to the harateristis of the medium (Askehave & Nielsen, 2005). With the rapid soio-politial and tehnologial innovations in finanial reporting, it would seem that the urrent synhroni approah to the annual report does not do justie to the real nature of this genre. For a more omplete understanding of the annual report genre, more longitudinal studies should be onduted; these studies ought to explain in what way the annual report is evolving as a professional multimodal genre. Methodologial issues in the ross-ultural genre study From a methodologial perspetive, there are three limitations that might lead to an extension of the present study: (1) the exlusion of the orporate harateristis industry and apital value in the sampling of Duth-English and British annual reports, (2) the inlusion of British Chairman s and CEO s statements as independent text types and (3) the use of doument frequeny ounts. Although the urrent genre analyses did not indiate that the industry and apital value of partiular orporations diretly affeted the final researh results, it ould first of all be fruitful to initiate analyses in whih the effet of these orporate aspets are measured formally. This is partiularly relevant in studies that aim to ompare the genre onventions for annual reports in one orporate setor to the genre onventions in another setor. Seondly, it would be advisable to determine the interdependene between the British Chairman s statements and the British CEO s statements, whih were regarded as individual genres in the present study. Beause the British Chairman s and CEO s statements typially appeared in sueeding order within one annual report, they may not be as independent as they were here assumed to be. Hene, the investigation of British management statements should be expanded to inlude the intertextual relationships between these two types of texts within UK-based annual reports exlusively. A third methodologial reommendation for further researh onerns the need for relative frequenies of ourrene, as opposed to doument frequenies, within the proessing of substantive, strutural and lexio-grammatial data. Even though the registration of doument frequenies was aeptable from the exploratory perspetive maintained in the present study, for a further refinement of this study it would be relevant to look at relative frequeny ounts that reveal how often a genre item or genre feature ours in a given instane of a partiular text type. The sores for relative frequeny ounts will reveal to what degree the urrent findings are onfirmed (i.e. enlarged or redued) in a more sophistiated researh situation. Reader response and the ross-ultural genre study Finally, the results of the genre analyses disussed in Chapters 4-7 yielded interesting results, not only from a text-desriptive or ross-ultural point of view but also from the perspetive of future response studies. It was emphasized that the ross-ultural variations observed for the ontext, ontent, struture and lexiogrammar of the English annual reports ould well be implemented as independent variables within an effet study. Here, the signifiant results of the statistial analyses ould form the basis for hypotheses to be tested among the readers of the annual report. For example, the fat that metadisursive text stagers were

258 signifiantly less ommon in the British annual report texts than in the Duth- English texts might lead to the hypothesis that British shareholders prefer annual report texts without text stagers over annual report texts with text stagers (based on the similarity-attration paradigm by Byrne, 1971). Therefore, the next step would be to analyze the response of international stakeholders to the signifiant rossultural differenes in the onstrut of the English annual report. This ontrastive response analysis would be a feasible step with regard to the present researh materials, sine it was onluded that the annual report as a genre set onsisted of omparable text types aross the Duth and British business ultures, distinguishable from eah other on the basis of a set of subtle ross-ultural differenes (see setion 8.5). Response studies would seem to be of partiular importane to the produers of the English annual reports, as these studies may allow insight into the effetiveness of the ommuniation strategies used. Chapter 9 will present a first attempt in translating signifiant ross-ultural differenes in genre features into a response analysis among international readers that are believed to be the primary audienes of the annual reports of Duth and British orporations.

259 Chapter 9 Reader response to Duth-English and British-English annual reports The previous hapters presented ontrastive (on)text analyses of multimodal English annual reports originating in Duth and British business ultures. These analyses have resulted in ross-ultural overviews of prototypial genre features in the written texts and photographs of Duth-English and British management statements, orporate profiles and operational reviews. For eah of the two national business ultures, the ross-ultural investigations foused on the prodution ontext of the annual report texts, as well as on the oneptual themes, the rhetorial struture and the metadisourse inluded in these texts. As suh, Chapters 4 to 7 provided answers to the first main researh question and to RQs 1a-1d and 2a-2d, all of whih revolved around the doumentation of ulture-speifi preferenes in the prodution of multimodal disourse in Duth-based and UK-based annual reports (see setion 1.5). The results setions in Chapters 4 to 7 showed that Duth and British orporations draw on several distint genre onventions in the multimodal realization of English-language texts for the annual report. As was pointed out at the end of Chapter 8, it would be relevant from a text prodution perspetive to expand the urrent genre study and determine the international reader response to signifiant ross-ultural differenes in the multimodal disourse of Duth-English and British annual reports. Similar ideas an be found in studies suh as Paul, et al. (2001), Rogers (2000) and Vaughan and Dillon (2006), whih suggest that genre researh is more omplete when it inorporates both the prodution and reeption of texts. Therefore, this final hapter disusses an interultural reader response analysis whih was onduted to measure the atual impat of ross-ultural genre variations, i.e. it measured the effetiveness of the Duth-English annual report texts as ompared to the British-English texts. More speifially, the response analysis in this hapter was intended to answer the seond main researh question of the present genre study (introdued in Chapter 1): II. How do the typial genre harateristis of the English, non-finanial texts in multimodal annual reports produed by Duth orporations affet the UK-based reader s pereptions of the orporate image and the persuasiveness of the text, and how does this ompare to the typial genre harateristis used in British-English annual reports? This seond question was further divided into several speifi researh questions whih failitated the interultural reader response analysis. They are inluded in Table 9.1. These RQs also helped to realize part of the first main researh objetive, whih not only foused on apturing ulture-dependent preferenes in the prodution of English annual reports but whih also foused on reording ulturedependent preferenes in the interpretation of these texts.

260 Table 9.1: Speifi RQs related to the seond main researh question of the present genre study (see setion 1.5). RQs reader response to genre RQ 3.a. What is the UK-based readers response to the typial written and photographi disourse in the nonfinanial texts of English annual general reports produed by Duth multinationals? RQ 3.b. What is the UK-based readers response to the typial written and photographi disourse in the nonfinanial texts of English annual general reports produed by British multinationals? RQ 3.. Aording to the UK-based reader, is there a differene in effetiveness between the multimodal generi features of the Duth-English annual report texts and the British annual report texts? The response analysis that was onduted to find an answer to RQs 3a-3, was based on the ross-ultural variations observed for the genre ontext of the Duth-English and British annual reports (Chapter 4) and for the substantive, strutural and lexiogrammatial genre features in these reports (Chapters 5-7). As will be explained below, these ross-ultural variations were used as independent variables or researh topis in the present effet study of speifi Duth-English and British-English annual report texts. Setion 9.1 sets out the role of response analysis in ontemporary genre researh. Setion 9.2 aounts for the relevane of researh on interultural ontexts in genre reeption and also inludes the hypotheses appliable to the present response analysis. Setion 9.3 disusses the methods used in this analysis, onsequently overing its partiipants, stimuli, measures and design plus proedures. Setion 9.4 provides an overview of the researh results, and it is sueeded by a onlusion and disussion in setion Genre studies and text reeption Considering that genre theory refers to texts as onventionalized disursive events that are initiated and interpreted to fulfill ertain ommuniative needs within a language use ommunity (e.g. Miller, 1994; Swales, 1990), there is a fundamental relationship between the sender and reeiver of a genre. First, members of a ommunity negotiate meaning based on their norms and expetations for the design of a partiular text (i.e. genre) and for the way in whih this text should be understood. Genre design thus indiates if and how a genre is sustained among ommunity members and implies whih soial strutures and relationships are onfirmed through its use (Bhatia, 2002; Vaughan & Dillon, 2006). In other words, the text or genre produt is the sender s physial manifestation of soio-ultural disourse onventions, while the reeiver s interpretation of the genre reflets the soio-ultural signifiane of the text in a given situation. Seond, the notion that genres are primarily typified by a partiular soial need or exigene whih is translated into a typial text purpose suggests that genres are dialogial: they not only need to ahieve something from the perspetive of the writer or the speaker but they also need to affet the reader or the listener (e.g. Solbjørg Skulstad, 2005; Yates & Orlikowski, 1992). Or, as Rogers puts it, genres draw attention to how disourse

261 atually funtions in the experiene of the individuals who interatively reate and interpret it (2000: 430, 431). She believes that the pratial impliations of generi purposes annot be fully aounted for without an understanding of how these purposes are interpreted by their audiene, i.e. without an understanding of whether these purposes ontribute to the fulfillment of the audiene s needs and expetations. In their study on sientifi texts, Paul et al. add that genre analyses should not only fous on the typial features of the text produt but should also provide insight in the rhetorial effet or ommunal aeptane of these features (2001: 388). Based on the haraterization of genres as interative manifestations of a soial system and as dialogial texts, it an be argued that both text prodution and text reeption are inherent in the genre onept. Aordingly, any genre study should ideally omprise the aspet of atual reader reation and response to the texts analysed (Solbjørg Skulstad, 2005: 349). Despite the above arguments for the integration of reeption analyses in genre researh, a gap remains in user-based genre studies within the professional domain (as was observed by Berkenkotter & Hukin in 1995, and by Askehave & Swales in 2001). To date, relatively few studies have onsidered genres as utteranes reeived. Genre studies known to have inorporated reader reations to examples of a given text type generally omprise experiments testing the effet of genre features or textual responses ommenting on or reviewing the genre under investigation. Some of the earliest experimental analyses of genre reeption were onduted by Loker (1999), who foused on the influene of different rhetorial strategies in negative letters refusing students in-store onsumer redit and in negative letters denying them admission to graduate shool. In her genre-based experiment among MBA and aounting students, moreover, Rogers (2000) analyzed the pereived ommuniative profile of CEO presentations in finanial analyst meetings. Additionally, Lagerwerf and Bossers (2002) experimentally investigated lients responses to the struture and linguisti style of onventional and unonventional examples of the business proposal. For the interultural use of genres in partiular, Shaw et al. (2004) tested the response to various versions of English ustomeromplaint dialogues among Business students from six different nations (and language bakgrounds). The aim of this study was to show how presumed ulturespeifi politeness strategies affet the pereived preferene, aeptability and familiarity of the ustomer-omplaint text. Furthermore, Van Meurs, Korzilius and Hollander (2006) foused on the effet of English in Duth-based job advertisements among Duth students in eonomis and management. They speifially analyzed the attitude towards the reruiting ompany, the attrativeness and omprehensibility of these ads. A final illustration of experiments involving genre reeption is Vaughan and Dillon (2006), whih tested undergraduate students appreiation of genre-onforming and genre-violating strutures in web-based newspapers. With regard to the textual response studies of professional genres, these an be exemplified by Solbjørg Skulstad s (2005) qualitative examination of student teahers replies to omputer-mediated text analyses of fellow students. In partiular, Solbjørg Skulstad explored the rhetorial strategies respondents use to express their identity as both a friendly student and professional teaher in the new genre of the digital response omment. Another example is James, Sholfield and Ypsiladis (1994), who analyze native and non-native English writers reations to the persuasive strategies used in non-native English versions of the letter of appliation.

262 9.2 Hypotheses for the interultural response analysis User-based genre studies suh as the ones mentioned in the previous setion predominantly report on genre reeption in mono-ultural situations. With the exeption of studies like Shaw et al. (2004), hardly any refer to a ommuniative ontext in whih the sender and/or respondents of a genre have different ultural i.e. professional, national or language bakgrounds. This is rather surprising, as the many genre studies that have been onduted aross ultures (e.g. Connor, 2000; Dahl, 2004; Mauranen, 1993; Nikerson, 2000; Nikerson & De Groot, 2005; Valero-Garés, 1996; Vergaro, 2005; Zhu, 2000) have generally yielded interesting materials for the further analysis of the interultural reeption of partiular professional genres. Based on Dahl s (2004) observation of different linguisti onventions for sientifi artiles in English, Norwegian and Frenh, for instane, it would be relevant to investigate the evaluation of English artiles by a Norwegian or Frenh audiene. Similarly, Vergaro s findings (2005) of distint strutural and linguisti rules for Italian and English For-Your-Information letters would allow for an interesting response study in whih the English letter is judged by international Italian lients. And with respet to the use of non-native English in professional interultural ontexts, a good starting-point for response analyses is provided by Mauranen s (1993) study of Finnish-English and US eonomi texts. Sine these texts show ross-ultural variations in linguisti onventions, it would be fruitful to determine if international sientifi experts interpret native US texts differently from Finnish-English eonomi texts. Following Hoeken and Korzilius, the results of these ross-ultural genre studies reveal that doument designers from different ultural bakgrounds differ in their opinions on what makes a doument effetive (2003: 288). Partiularly for ommerial organizations operating internationally, this is a ruial fat, possibly with far-reahing onsequenes. In situations where a ompany is involved in international mergers, global marketing, global investments or internationalization of management or staff, its suess partly depends on the effetiveness of its oral and written ommuniation with people from different ultural bakgrounds. If orporations have no awareness of the different expetations foreign stakeholders may have with regard to the design or language used in the orporate texts these stakeholders reeive, they ould be in for an unexpeted and irreoverable loss of business (e.g. Bargiela-Chiappini & Nikerson, 2003; Vandermeeren, 1999). As Riks (1999) illustrates in his overview of international business blunders, this may inlude disappointing sales figures, negative publiity, extreme marketing osts or the disontinuation of joint ventures. Presumably, suh dramati responses to interultural orporate ommuniations might have been avoided or redued if the orporation had had a better understanding of the possible ulture-speifi impat of its texts. Here, genre-based studies suh as Shaw et al. (2004), or any other interultural response analysis of business texts (e.g. Ulijn & St. Amant, 2000) would have been useful soures of information: these studies present orporations with empirial data that indiate how their texts are likely to be pereived by a partiular international audiene. The English annual reports analyzed in the present genre investigation also play a signifiant role in the interultural ommuniation of Duth and British orporations. The interview results in setions and showed that the Duth-English and British annual reports generally serve several purposes in relation

263 to a variety of international readers, on whih the orporation s finanial ontinuity and general publi impression more or less depend. Duth-English and British annual reports aim to inform international readers about the orporate results in the past finanial year and they are intended to eliit positive assoiations with the ompany. Both Duth and British ompanies believed that their English annual report is espeially relevant for foreign readers with (in)diret finanial interests in the organization, i.e. institutional and private investors, finanial analysts and finanial journalists. In line with the genre studies disussed above (Dahl, 2004; Mauranen, 1993; Vergaro, 2005), the (on)text analyses of the speifi English management statements, profiles and operational reviews (setion and Chapters 5-7) furthermore revealed some ross-ultural differenes in disourse onventions that may influene reader response in international ommuniation settings. Although there was ross-ultural variation in the prodution ontext, themes, struture and metadisourse of all text types analyzed in the English annual reports, the largest number of differenes existed between the Duth-English and British management forewords. For example, the Duth-English CEO s statement was interational as well as strategi in nature, while the British Chairman s statement was mostly interational and the British CEO s statement predominantly provided a strategi insight in the ompany s performane. Compared to the Duth-English CEO s statements, moreover, the British management statements more often inluded the Corporate Soial Responsibility theme and lose-up pitures of managers, but less frequently ontained salutations at the beginning of the text and pitures of the full management team. To the extent that the English annual report texts reflet generi variations between the Duth and British national business ultures, it ould be predited that in an interultural response analysis with UK-based readers the British-English texts will be most suessful. Drawing on the similarity-attration paradigm (Byrne, 1971; Shadid, 2000) it may be antiipated that the persuasiveness of a text is enhaned by similarities between the disourse onventions applied in the text and the disourse onventions with whih the reeiver of the text is familiar. This means that if a finanial advisor on the British market reads both Duth-English and British annual report texts, he or she is likely to have a preferene for the ompanies or the texts that onform most to the norms of disourse prevailing in the British business ulture. Sine the Duth-English annual report texts are based on different genre rules for both written disourse and photographs, it is possible that they will not meet the expetations of readers who are most familiar with the British-English disourse onventions. Therefore, the Duth-English multimodal texts ould be expeted to produe more measurable negative effets with UK-based respondents. These refletions are in line with the theoretial onept of genre, whih assumes that the substantive and formal text features typially employed in a given disourse situation reflet what members of a ommunity ommonly antiipate to be appropriate or effetive rhetorial behavior (e.g. Berkenkotter & Hukin, 1995; Swales, 1990). Assuming that the Duth-English and British annual reports in general serve informative and promotional purposes and onsidering the ross-ultural variation in multimodal text features, the present response study fouses on the effet of ross-ultural differenes in disourse onventions on ratings of text persuasiveness and orporate image. In partiular, it will test this effet from the perspetive of UK-based finanial professionals (see setion 9.3.1). Based on the similarity-attration paradigm, preditions about the outome of this study with

264 finanial readers in the UK market led to five hypotheses. The first hypothesis diretly ensued from the promotional purpose of the Duth-English and British annual reports and involved the impat of the typial Duth-English and British- English genre features on the pereived image of the ompany represented in the annual report texts. Here, orporate image refers to readers interpretation of or attitude towards ompany-related attributes, i.e. readers assessments of the ompany on those attributes (Fombrun, 1996; Van Riel & Fombrun, 2007). H1: For finanial professionals in the UK, annual report setions designed in aordane with British-English genre onventions eliit a more positive orporate image than annual report setions designed aording to Duth-English genre onventions. In relation to the informative objetive of the annual report in partiular, the urrent response analysis examined interultural effets of ross-ultural variations in genre features on the pereived persuasiveness of the Duth-English and British annual report texts. To be able to test these effets, the response analysis inorporated three aspets of persuasiveness: general attitude towards the texts, redibility of the texts and omprehensibility of the texts. Whereas attitude toward the text and text redibility reveal the extent to whih readers find the information in the texts attrative and believable (Makenzie & Lutz, 1989), text omprehensibility refers to the aessibility of the information in the texts (Maes, Ummelen & Hoeken, 1996). Therefore, the evaluation of these three aspets of persuasiveness will give an indiation of how likely it is that the informative details in the texts will atually be absorbed by the UK-based professionals. The seletion of aspets of text persuasiveness resulted in three additional hypotheses: H2: For finanial professionals in the UK, annual report setions designed in aordane with British-English genre onventions eliit a more positive attitude towards the text than annual report setions designed aording to Duth-English genre onventions. H3: For finanial professionals in the UK, annual report setions designed in aordane with British-English genre onventions are pereived as more redible than annual report setions designed aording to Duth-English genre onventions. H4: For finanial professionals in the UK, annual report setions designed in aordane with British-English genre onventions are pereived as more omprehensible than annual report setions designed aording to Duth-English genre onventions. To further analyze the persuasiveness of the Duth-English and British annual report texts, the reader response analysis inluded general reader preferenes. Preferene was measured in two ways: (1) as the preferene expressed by UKbased professionals for a full Duth-English or British text, and (2) as a preferene expressed by UK-based professionals for speifi genre features that were typial of the written texts and photographs in the Duth-English or the British-English annual reports. Sine these preferene measurements were not expliitly assoiated with either the informative or the promotional purposes of the English annual reports, they were relevant to the extent that they provided insight into the general

265 appreiation for Duth-English and British multimodal disourse onventions. Again referring to the ultural imprint of genre rules and the attration of familiar language use (Byrne, 1971), a fifth hypothesis was proposed: H5: For finanial professionals in the UK, annual report setions designed in aordane with British-English genre onventions are preferred over annual report setions designed aording to Duth- English genre onventions. While hypotheses 1-5 primarily overed the response to ross-ultural differenes in multimodal genre features (i.e. themes, moves/strategies and metadisourse), the ontext analysis in Chapter 4 yielded several ross-ultural distintions that were also relevant in the interultural effet study of Duth-English and British annual reports. Cross-ultural variations that were of partiular interest onerned the prioritization of target groups for the annual report and of the ommuniative purpose of speifi annual reports setions (e.g. the management statements). Considering the urrent fous on the interultural effetiveness of Duth-English and British annual reports, it would be fruitful to explore if Duth and British ompanies pereptions of the readers and objetives of the English annual report texts atually orrespond with the use of these texts in pratie. Aordingly, the final issue in this reader response study was addressed in the form of an exploratory researh question: RQ: Do UK-based finanial professionals use the English annual report texts in a way that is ulturally ongruent with British ompanies pereptions of the purposes and partiipants of these texts? The details of the measures referred to in the above hypotheses and researh question will be elaborated upon in setion The answers to both the hypotheses and the researh question are intended to establish the effetiveness of the Duth-English annual report, as ompared to the impat of the Britishl report. 9.3 Method interultural reader response analysis The present study measured the effetiveness of multimodal texts in Duth-English and British annual reports on the basis of a reader response experiment. The following subsetions desribe the partiipants in the urrent experiment (9.3.1), the materials and independent variables (9.3.2), the dependent variables (9.3.3), and the design and proedures (9.3.4) Partiipants The interultural response analysis inluded finanial readers residing and working in the United Kingdom. Sine the ontext analysis suggested that institutional investors in partiular belong to the primary audiene of the annual report in both the Netherlands and in the UK (setion 4.4.2), it was initially deided that the partiipants in the experiment should be working for institutional investment organizations in the UK (e.g. insurane ompanies, banks, pension funds; see Van

266 Riel & Fombrun, 2007). It appeared, however, that the employees of suh organizations hardly ever use the annual report as an investor relations tool, beause the portfolios of investment produts and servies are generally ompiled on the basis of ompany evaluations onduted by finanial advisors suh as analysts or traders (stok brokers). Consequently, the fous of the experiment shifted from institutional investors to finanial analysts and stok brokers. This resulted in a sample of respondents employed by at least seven different investment banks and two other finanial analysis institutions loated in London, i.e. in the City and at Canary Wharf. Although this response study was originally based on the premise that the evaluation of Duth-English texts by native British readers would represent a realisti situation, an on-site exploration of the researh population added relevant information. First, the population of finanial analysts and brokers in London was found to ontain many professionals who are native speakers of British-English but who rarely fous on British orporations. For instane, it inludes British analysts who are speialized in the evaluation of Italian or Ukrainian ompanies, in whih ase they did not really fit the desription of readers of British-English annual reports. Seond, only a selet group of native British analysts and brokers seemed to read British annual reports for investment-related purposes. This is explained by the observation that the London-based population of analysts and brokers in fat overs multiple nationalities. A large number of professionals were born and raised outside the United Kingdom and are non-native speakers of English, e.g. they speak Frenh, German or Italian as their first language. Nonetheless, many of these nonnative professionals regularly analyze UK orporations and use British texts as a soure of information. As suh, they have familiarized themselves with the way in whih British orporations design their annual reports. Despite their non-british bakground or their non-native ommand of the English language, in priniple these professionals mathed the desription of readers of British annual reports. Sine a fous on one of these groups British advisors on non-uk ompanies, British advisors on UK ompanies, and non-british advisors on UK ompanies would have affeted the validity of the experiment and would have reated the risk that the researh sample would be of insuffiient size, the general professional profile of the partiipants was defined as follows: partiipants are UKbased finanial analysts or stok brokers who regularly fous on traditional British orporations and who have a native or near-native profiieny in English. Although initial ontats with potential respondents were established randomly, atual partiipation in the experiment depended on a person being either a native British- English speaker or an international UK-based professional with experiene in the evaluation of British orporations. Table 9.2 provides a detailed overview of the full sample profile. As Table 9.2 shows, the partiipants in the reader response experiment were primarily male (82.9%) and of relatively young age (a mean age of 27.5 years). Overall, they had just under four years of work experiene as finanial professionals. Consistent with the preliminary observations, the sample ontained an almost equal distribution of native British and non-native British professionals. Independent-samples t-tests revealed no signifiant differenes between the experimental results of respondents with a native British bakground and a nonnative British bakground. The total number of 35 partiipants for the experiment was established at a statistial power of.80, a medium effet size (d =.50) and an alpha of.05 (e.g. Aron & Aron, 1997).

267 Table 9.2: Sample profile partiipants reader response experiment English annual reports Sample harateristis Profile N=35 Professional funtion Analyst Broker/trader 85.7% 14.3% (30) (5) Professional experiene (in years) Gender Male Female Age Country of origin UK Non-UK Mean 3.90 SD % (29) 17.1% (6) Mean SD % (18) 48.6% (17) Materials The researh materials for the experiment were based on the results of the rossultural genre analyses disussed in Chapters 4-7. As was mentioned previously, the generi study of the Duth-English and British management statements in partiular showed many ross-ultural text variations that might be influential in situations involving an interultural readership. In addition, these statements often overed only one page in the English annual report, therefore enabling an experiment to be designed using authenti text formats. As the Duth-English and British profiles and operational reviews were either too similar or too extensive, the present reader response analysis speifially foused on testing the effetiveness of the Duth- English managerial forewords, as ompared to the British-English management statements. This was also an appropriate hoie from the perspetive of genre ontext: both Duth and British orporations believed that the introdutory statements in the annual report appeal to a broad audiene, with a slightly stronger emphasis on shareholders among the British orporations. Considering the wide appeal of the Duth-English and British management statements and the apparent importane of finanial analysts and stok brokers as readers of the annual report in the UK market (see setion 9.3.1), it seemed plausible to investigate how relevant the genre of the management statement really is to this target audiene. Inherent in the seletion of English management forewords as researh materials, was the diffiulty of having to test three types of texts, i.e. the Duth- English CEO s statements, the British-English Chairman s statements and the British-English CEO s statements. Although it was suggested that all statements types be inorporated to ensure full eologial validity, subjeting respondents to a omprehensive questionnaire with one Duth-English statement and two British statements did not appear be a feasible option, given the length of time it would take to omplete suh a questionnaire. As previous ontats with Duth and British finanial professionals had indiated that potential partiipants generally work a heti shedule, the ompilation of experimental texts was based on the signifiant ross-ultural results found in the Duth-English CEO s statements on the one hand,

268 and on the merged sets of signifiant results observed in the British CEO s and Chairman s statements on the other. This allowed for a questionnaire ontaining one Duth-English version and one British-English version of the managerial foreword. It should be noted that speial attention was given to the ultural variations aross the Duth-English CEO s statements and the British Chairman s statements, sine both are the first forewords in the annual report of Duth and British ompanies. The omparative effet of genre features in the Duth-English CEO s statement and the British Chairman s statement ould have impliations for the design of introdutory texts in English annual reports published by Duth multinationals. Furthermore, as there were no available fats on the respondents awareness of the two possible senders of the British text (i.e. the Chairman or the CEO) and interpretational influenes related to the sender s professional funtion (i.e. supervisory versus operational) needed to be redued, the experimental texts were referred to as Letters to the Shareholders. The title Letter to the Shareholders suggested one ommuniative purpose for both the Duth-English and the British texts and the avoidane of funtional referenes in this title was intended to plae a stronger fous on their generi text features. Referenes to the professional role of the sender of the statement were also omitted from the losures of the Letters. The Letters to the Shareholders used in the experiment were onstruted in order to resemble authenti Duth-English and British management statements as losely as possible (see Appendix X). In priniple, the Duth-English Letter was based on the CEO s statement of a Duth finanial servies ompany, while the British-English Letter was based on the Chairman s and CEO s statements of a British finanial servies ompany. The Duth-English and British annual reports from whih these statements were drawn, involved the 2003 finanial alendar year. Therefore, both had been inluded in the generi text analyses desribed in Chapters 5-7. The authenti statements of the Duth and British finanial servies ompanies were relevant primarily beause they ontained many of the ulture-speifi genre features that were disernable in the Duth-English and British management statements. As suh, they inluded most of the independent variables that needed to be tested in the response experiment. If a required variable was absent in the seleted statements, appropriate text fragments were extrated from English management statements in 2003 annual reports of Duth and British orporations in the same industrial setor. These reports had also been part of the prior text analyses. Although the two test Letters for the experiment needed to suggest that they were representing two distint finanial servies orporations, the authenti versions of the Duth-English and British texts from whih they had been generated were too different to be able to realize a omparable set of Letters. To prevent pereptual effets as a result of disrepanies in performane details or market information, the Duth-English and British Letters were onstruted to tell similar orporate stories, e.g. with regard to profit data, operational ativities, eonomi onditions, soial trends, external rankings and the name of the sender of the text. The uniformity of Letters was further enhaned by referene to the genre features that were found to our aross all Duth-English and British management statements. More speifially, the themes, moves and strategies, and metadisursive items typially existing in a similar (high) perentage of all statement types were inluded to establish generi equivalene between the Duth-English and British Letters in the experiment. For example, the theme Shareholder onfidene, the rhetorial strategy Displaying orporate strengths and attitude markers in the form of evaluative adjetives were observed in more than 70% of all original English management statements and were thus integrated in the two test Letters. The

269 projetion of omparable orporate messages and the maintenane of ross-ultural similarities observed in the text analyses (Chapters 5-7) supposedly redued performane-related distrations and distrations related to distintive but irrelevant disourse items during the partiipants evaluations of the texts. Appendix X shows how omplete examples and fragments of the Duth-English and British Letters to the Shareholder were inorporated in the questionnaire. In total, the test Letters for the experiment ontained 17 out of the 38 rossultural differenes observed in the written texts and photographs of the authenti Duth-English and British statements. These 17 ulture-speifi genre harateristis are shown in Table 9.3; they represent the independent variables used in the response analysis. That is, the full examples of the test Letters were manipulated on the basis of ulture-speifi textual elements, while the fragments of these letters were manipulated on the basis of both ulture-speifi textual and visual elements (Appendix X). Sine the effet of most of the ross-ultural variations in genre features would already be measured by the experimental analysis of the running texts and photos in the management statements, it was deided that rossultural differenes related to the text omments adjoining the images be exluded from the experiment. Table 9.3: Independent variables reader response experiment, based on ross-ultural differenes in themes, moves and strategies, and metadisourse inluded in the Duth-English and British management statements (see results setions in Chapters 5-7). Genre feature Du- CEO Br- CEO Br- Chair. Text themes (setion 5.3) % % % Total χ 2 p % (df = 2) Corporate strategy b 43.8 a 77.8 <.001 Board details a 87.5 b <.001 CSR 10.0 a b <.05 Dividend 20.0 a 16.7 a 81.3 b <.001 Photo themes All exeutive board members 35.3 b <.05 Text moves (setion 6.3) Information leadership a 87.5 b <.001 Reader orientation 55.0 a 94.4 b <.05 Text move + strategy Operat.perf. + finanial details a <.05 Future perf. + future objetives 66.7 b <.05 Addressing reader + salutation 63.2 b 0.0 a <.001 Text metadisourse (setion 7.3) Frame marker: Text stagers 55.0 b <.05 Photo metadisourse Code gloss: Fous CEO/Chairman 58.8 a <.05 Code gloss: Fous CEO+board 35.3 b <.05 Attitude: sinerity/openness 82.4 b a <.05 Self-mention:Fous 58.8 a <.05 CEO/Chairman Self-mention: Fous CEO + board 35.3 b <.05 Engag.marker: Intimay ompany 29.4 a b <.05 Note. a,b. The frequeny of a theme was a lower or b higher than might be expeted on the basis of row and olumn totals (i.e. adjusted standardized residuals < or > 1.96).. Fisher s Exat Test.

270 To preserve a high degree of eologial validity and pragmati value in this study, three riteria were applied in the seletion of the 17 independent variables displayed in Table 9.3. Beause there were no existing guidelines for the seletion of genre features in reader response analyses, these riteria were explorative and pragmati in nature. They primarily onerned salient or large ross-ultural differenes between the Duth-English and British management statements, and between the Duth- English CEO s statements and the British Chairman s statements: 1) Signifiant results (in Table 9.3) were used as variables in the experiment when the text analyses had shown that they were either signifiantly overrepresented or under-represented in the management statements of one of the two national business ultures. Compared to both the British Chairman s and CEO s statements, for instane, the Duth-English statements displayed an over-representation of frame markers in the form of text stagers. But the Duth-English texts revealed an underrepresentation of pitures ontaining an individual manager as a means of identifying the ompany. Therefore, texts stagers were inluded in the Duth-English Letter and portraits of individual managers were inluded to represent the British-English Letter. 2) Signifiant results (in Table 9.3) were used as variables in the experiment when the text analyses had shown that there was at least a 30% differene between the Duth-English management statements on the one hand and both types of British management statements on the other hand. For example, a larger number of Duth-English statements (82.4%) projeted the sinerity of the orporation through a piture of a manager staring towards the amera lens, with at least a 30% differene ompared to the highest frequeny sore for the British-English statements (50%). In ontrast, the move Orientating readers through the text ourred more often in the two types of British statements, with at least a 30% differene between the lowest frequeny sore for the British statements (87.5%) and the sore for the Duth-English statements (55.0%). Consequently, the experiment overed images of managers looking into the amera for the Duth-English Letter and the move Orientating readers for the British Letter. 3) Signifiant results (in Table 9.3) were used as variables in the experiment when the text analyses had shown that there was at least a 45% differene between the Duth-English CEO s statements and the British Chairman s statements. This was the ase, for instane, with regard to the Dividend theme, the move Reporting on leadership situation and the engaging lose-ups of orporate managers. Suh textual or visual elements were all inorporated to onstitute the British test Letters in partiular. It should be noted that these riteria atually applied to the photographi CEO theme and the photographi Chairman theme as well. Consistent with the professional mandate of the sender of the text, images of the CEO ourred signifiantly more often in the Duth-English CEO s statements and pitures of the Chairman appeared more frequently in the British Chairman s statements. However, these ross-ultural results are not inorporated in Table 9.3 beause professional funtion was not inorporated into the reader response analysis (see above). The seleted independent variables presented in Table 9.3 were entered in the researh materials in four different ways: (1) as elements in omplete Letters to

271 the Shareholders, (2) as elements in statements about the Letter to the Shareholders, (3) as elements in text fragments within the Letter to the Shareholders and (4) as elements representing the visual design of the Letter to the Shareholders. First, the 10 text-based variables were inorporated in omplete versions of a Duth-English and British Letter to the Shareholders. These were ulturally ongruent texts in whih all typial textual features were plaed in-ontext. The final versions of the Duth-English and British Letters eah ontained lose to 500 words (see setions A + B in Appendix X). The omplete Letters allowed for a natural presentation of variables (i.e. within a text of realisti length and sope) and for the analysis of fulltext effetiveness. The effet of several text-based genre features was also investigated by means of evaluative statements (see setion E in Appendix X) and text fragments (see setion F in Appendix X) highlighting typial substantive, strutural and metadisursive features of the Duth-English or British Letter to the Shareholders. Both devies permitted a fous on isolated genre features and provided a quik way for respondents to express their preferene for partiular ulture-bound text elements. Finally, the photo-related variables were inorporated into individual images typial of the visual design of a Duth-English or British Letter (see setion G in Appendix X). Considering that the present experiment aimed to aount for the effetiveness of ulture-speifi visualizations of orporate managers, it examined the impat of photo-based variables that had been separated from their textual environment. In other words, respondents were onfronted with single pitures of diretors, that had previously been found to have a typial substantive or formal funtion in either the Duth-English or British-English management statements Dependent variables In the omparative response analysis of Duth-English and British-English Letters to the Shareholders, measures for the dependent variables were primarily adopted from prior effet studies. Although some of the existent measures drawn upon ontained seven-point or eleven-point sales, all measures in the present reader response analysis were designed to inlude five-point sales. This hoie of sales was based on the researh objetives of the urrent response analysis and on the riteria used for the questionnaire (e.g. Shmalensee, 2003): while the aim of the experiment was to monitor the impat of Duth-English and British texts in a way that was meaningful to both aademis and pratitioners in business ommuniation, it also needed to aommodate the busy shedules of the partiipants in the experiment. To realize a valid questionnaire that would minimize distrations by onsisteny in sales and failitation of quik answers, it was deided that all sales with more than five values would be replaed by five-point sales. The reliability of these five-point sales was alulated on the basis of Cronbah s Alpha values per measure. The qualifiations of Cronbah s Alpha were derived from Van Wijk (2000). Additionally, the questionnaire inluded two multiple response items and one openended item to hek whether the UK-based partiipants use of the management statement was onsistent with what ould be expeted based on the ontextual genre analyses of the Duth-English and British annual reports (setion 4.4). The following paragraphs will further disuss the measures for the independent variables. It should be noted that the response experiment omprised a within-group design in whih partiipants needed to read a Duth-English as well as a British Letter to the Shareholders. After reading eah Letter, partiipants had to

272 respond to the image-related measures and the measures related to text persuasiveness (see setion 9.3.5). The measures for general reader preferenes and the measures for the general use of annual report texts were inorporated one in eah questionnaire. Corporate image Corporate image was assessed with regard to the full English Letters to Shareholders (see setions A + B, Appendix X). After reading eah text, subjets were asked to express their opinion about orporate attributes refleted in the written disourse. A measurement sale was seleted that would relate to the reader needs of the finanial analysts and brokers partiipating in the experiment. Several studies on investor or finanial relations have indiated that finanial professionals evaluation of orporate dislosures depends on attributes suh as the ability to gain earnings, the quality of orporate management, the produt quality, the return on investments, the strategi future position of the ompany, the transpareny of information and the pratie of sustainability (e.g. De la Fuente Sabaté & De Quevedo Puente, 2003; De Chernatony & Harris, 2000; Pava & Krausz, 1996). Compared to other imagerelated measures evaluating orporate behavior, personality or redibility (e.g. Berens & Van Riel, 2004), these attributes are best represented in the Reputation sale used in the annual Fortune Survey of Most Admired Companies (Fombrun & Shanley, 1990). Therefore, this sale was used for the measurement of orporate image refleted in the urrent researh materials. The fat that the Fortune ratings are alled a reputation measurement does not prinipally prelude its use as a orporate image measurement. As Van Riel and Fombrun laim, reputations evolve from orporate images in the sense that the onept of orporate reputation aptures the effets that [ ] images have on the overall evaluations whih stakeholders make of ompanies (2007: 40). They ontinue that reputation an therefore serve a useful funtion by gauging the overall estimation in whih the organization is held by its onstituents and so measure the effetiveness of the organization s ommuniations with those stakeholders (Van Riel & Fombrun, 2007: 40, 41). In other words, the Fortune Reputation sale is a devie by whih the orporate image pereived by the reeiver of business texts an be established. The Fortune ratings were originally introdued to establish the general orporate exellene of US firms among key audienes whose main interest was in the finanial position of these ompanies, i.e. senior exeutives, outside diretors and finanial analysts (Fombrun & Shanley, 1990; Fryxell & Wang, 1994). They have sine been referred to in various reputation studies, e.g. related to finanial performane (Hammond & Sloum, 1996; Roberts & Dowling, 2002), layoffs (Flanagan & O Shaughnessy, 2005) or media exposure (Wartik, 1992). The Fortune survey omprises an eight-item, eleven-point sale (0 = poor/ 10 = exellent). Partiipants are asked to rate US firms on eah of the following eight attributes: (1) quality of management, (2) quality of produts or servies, (3) long-term investment value, (4) innovativeness, (5) finanial soundness, (6) ability to attrat, develop and keep talented people, (7) ommunity and environmental responsibility and (8) use of orporate assets (Fombrun & Shanley, 1990). The items of long-term investment value, finanial soundness and use of orporate assets over finanial ends, while the other items over apabilities and strategi means through whih the finanial ends an be ahieved (Chun, 2005). Several studies have suggested that the eight items in the Reputation sale are highly interdependent, together onstituting an overall impression of the ompany. Fombrun and Shanley (1990) and Fryxell and Wang (1994), for instane, show that the ratings for these

273 seemingly distint attributes strongly orrelate and that these ratings are diretly affeted by the subjet s pereptions of the finanial potential of the ompany. Aordingly, they argue that the Reputation sale of the Fortune survey should be reognized as a one-dimensional measure whih does not allow for the separate analysis of the eight attributes. The present study employed the Fortune sale to alulate the mean sores for orporate image refleted in eah English Letter to the Shareholders. The original version of the Fortune Reputation sale was replaed by a version with a five-point sale, ranging from 1 = poor to 5 = exellent. For both the Duth-English and the British-English Letters in the present experiment, the reliability of the eight-item, five-point sale was adequate: Cronbah s α =.80 for the Duth-English text and Cronbah s α =.72 for the British-English text. Attitude towards the text and pereived text redibility In order to investigate the effetiveness of the written disourse in the Duth-English and British management statements, the experiment first of all foused on attitude towards the text and redibility of the text (see setions A + B, Appendix X). Similar to the measurement of orporate image, these dependent variables were analyzed with regard to the omplete English Letters to the Shareholders. Attitude towards the Letters and pereived text redibility were examined on the basis of MaKenzie and Lutz s (1989) semanti differential sales. Apart from allowing insight into the reader s general appreiation of the Duth-English and British annual report texts, MaKenzie and Lutz s (1989) sales permitted a fous on the observed trustworthiness of the information provided in the texts. Taking into onsideration that one of the main riteria for finanial readers evaluation of orporate dislosures is the redibility of texts (Merer, 2004, 2005), this aspet was espeially relevant in the urrent study. For annual reports in partiular, it would seem that the effetiveness or appreiation of the written disourse is losely related to the pereived authority of the texts. Although MaKenzie and Lutz (1989) originally applied their attitude and redibility measures to printed produt advertisements, response studies in marketing and advertising have demonstrated that these measures are also relevant in establishing the effetiveness of other persuasive business texts. For instane, Thornsteinson and Highhouse (2003) used them for printed job advertisements, Gong and Maddox (2003) for web advertisements, Orth and Holanova (2004) for gender portrayals in print ads and Morimoto and Chang (2006) for diret marketing hannels. Thus, it would seem that these measures are also appropriate for the testing of reader response to the informative and promotional Duth-English and British annual report texts. For eah dependent variable, MaKenzie and Lutz (1989) introdue a threeitem, seven-point rating system. Attitude towards the text is measured using the semanti differential sales (1) good/bad, (2) pleasant/unpleasant and (3) favorable/unfavorable. Text redibility is analyzed using the semanti differentials (1) onvining/unonvining, (2) believable/unbelievable and (3) unbiased/biased. The urrent response study inorporated all three items for Attitude and Credibility. It inludes mean sores for attitude towards the Letters on the one hand and pereived redibility of these Letters on the other hand. As is shown in setions A + B of Appendix X, they were ombined in the questionnaire to represent the pereived persuasiveness of the full texts (together with the measure of text omprehensibility). The number of values per sale were redued to five, ranging from 1 = bad/unonvining to 5 = good/onvining. The reliability of the three-item, five-point sale for attitude towards the text was adequate for the Duth-English

274 Letter (Cronbah s α =.79) and moderate for the British Letters (Cronbah s α =.62). Reliability tests for the text redibility sale were insuffiient for the Duth- English Letter (Cronbah s α =.50) and adequate for the British Letter (Cronbah s α =.71). Pereived text omprehensibility The effetiveness of the written disourse in the Duth-English and British-English management statements was also analyzed in terms of text omprehensibility (see setions A + B, Appendix X). Again, this dependent variable was studied in relation to eah of the omplete Letters to the Shareholders in the experiment. The sales used to measure the effet of ulture-speifi genre features on pereptions of text omprehensibility were derived from Maes et al. (1996). The term text omprehensibility has been used by sholars to refer to a variety of different text harateristis or riteria, inluding a oherent struture, a ertain degree of logi or preditability, ommon ground referenes or lear ommuniation of goals (Duffy et al., 1989). A detailed disussion of this omprehensive onept falls beyond the sope of the present study. However, the rating system provided by Maes et al. (1996) seemed to be a relevant tool for apturing the impat of this onept with regard to the urrent English business texts, as has also been indiated in several reent user-based studies of other professional genres, e.g. job advertisements (Van Meurs, Korzilius & Hermans, 2004) and personal agent reports in publi affairs websites (Nieuwboer, Maes & Swanepoel, 2005). Pereptions about text omprehensibility in Maes et al. (1996) is measured by a six-item, seven-point balaned sale of semanti differentials. The differentials are: (1) diffiult/easy, (2) simple/omplex, (3) unlear/lear, (4) poorly organized/well organized, (5) logial/illogial and (6) onise/lengthy. Exept for the replaement of the seven values by five values (ranging from 1 = diffiult to 5 = easy), the text omprehensibility measure was retained in the urrent response experiment (see setions A + B in Appendix X, where it is ombined with the sales by MaKenzie & Lutz, 1989). An overall insight in the omprehensibility of the Letters was established by mean sores, as opposed to item-related sores. The reliability of the text omprehension sales was good for the Duth-English Letter to the Shareholders (Cronbah s α =.83) and adequate for the British-English Letter (Cronbah s α =.77). Investment intentions In addition to the analysis of image-related pereptions eliited by the Letters to the Shareholders, the reeption study also overed the influene of the full text versions on readers behavioral intentions (see setions A + B, Appendix X). By means of mapping subjets investment intentions per English Letter, it aimed to test the investment worthiness of the Duth and British ompanies responsible for these Letters. Here, sales for produt purhase intentions used in onsumer response studies offered relevant effetiveness measures to antiipate response behavior. Smith (1993), for instane, rates purhase intentions by asking onsumers how likely they would be to buy a partiular ola brand promoted in a print ad and how likely they would be to reommend it to their friends. A similar rating system an also be found in more reent onsumer response studies on other genres, e.g. the potential purhase effet of nutritional details on food produt pakages (Garretson & Burton, 2000) or the produt purhase intentions generated by soial responsibility details in a newspaper artile about the manufaturer (Klein & Dawar,

275 2004). Although the purhase intention sales were originally introdued to analyze the subjets willingness to buy a given onsumer good, the investment-related role of finanial advisors (i.e. analysts and brokers) meant that these sales were an appropriate measure for the present study. Smith s (1993) brand measure ontains a two-item, seven-point likelihood sale. The two items relate to (1) purhase intention and (2) intention to reommend purhase to others; eah item is tested by a sale with endpoints labeled zero likelihood to ertain. To make Smith s rating system appliable to the partiipants of the urrent response analysis, the items were first of all designed to refer to (1) investment intention and (2) intention to reommend investment to a professional olleague. The likelihood of investment behavior was indiated on a five-point sale, ranging from 1 = zero likelihood to 5 = ertain; the general intention to invest was established on the basis of mean sores for the two items in this measure. The reliability of the five-point sales was good for both the Duth-English and the British Letter to the Shareholders (Cronbah s α =.90 and Cronbah s α =.85, respetively). Text preferene Besides analyzing reader pereptions and intentions eliited by the Letters to the Shareholders, the user-based investigation also inluded several one-item, five-point sales to expliitly measure reader preferenes with regard to the typially Duth- English and British-English texts. Preferene was measured for both the omplete versions of the English Letters and for speifi multimodal elements ommon in either the Duth-English or British Letter (see setions C, E, F, G in Appendix X). After reading the Duth-English and British Letters to the Shareholders and answering the subsequent questions about orporate image, text persuasiveness and investment intentions, the partiipants were asked to indiate whether they had a preferene for one of the two Letters. Partiipants answered on a five-point sale with extremes labeled Letter CorpA and Letter CorpB, where the underlying referenes to the ultural origins of the letters and orporations Duth or British depended on the versions of the questionnaire (see setion 8.3.4). If no lear or slight preferene ould be expressed, partiipants were asked to mark the middle value of the sale. To hek if reader preferene was related to manipulations of ulture-speifi genre features in the texts, partiipants were asked to add a brief motivation for their response (setion C, Appendix X). Next, reader preferene was analyzed on the basis of a series of statements that referred to individual substantive and formal elements typifying the texts of either the Duth-English or British-English management statements. Eah statement started with the phrase The Letter to the Shareholders should inlude details about, and ended with various genre features for whih differenes had been found between the Duth and British national business ultures (setion E, Appendix X). For example, the statement The Letter to the Shareholders should inlude details about the management situation (e.g. board hanges) overed the rossultural differenes observed for the text theme board details and the move providing information about the leadership situation. The partiipants were asked to indiate on a five-point sale to what extent they agreed or disagreed with eah statement. As for the statement about the management situation, reader (dis)agreement suggested whether partiipants had a preferene for Letters to the Shareholders either inluding or exluding information about board members or management hanges.

276 Textual genre harateristis for whih ross-ultural variation had been observed were also analyzed by means of pairs of exerpts drawn from authenti Duth-English and British management statements. For eah distint genre feature, the questionnaire ontained one exerpt with and one exerpt without this feature, polarizing the exerpts through a five-point sale. Response to the move/strategy Addressing the reader/salutation, for instane, was measured using a text fragment beginning Dear shareholders on one end of the sale and a seond text fragment exluding the salutation on the other end of the sale. Exept for the salutation, these text fragments were idential. This is illustrated in setion F, Appendix X. Similar to the sales for the propositions, the rating system for the text fragments required partiipants to express to what degree they preferred either the exerpt with or the exerpt without a ulture-speifi genre item. If no preferene ould be expressed, they needed to mark the middle value of the sale. Fourthly, a one-item, five-point sale was used to measure the impat of several visual features for whih a ross-ultural differene had been diserned. This sale was applied to three pairs of management photographs whih were based on images enountered in authenti Duth-English and British annual reports. To avoid person-dependent assoiations and to be able to apture responses to the atual portrayal of the persons, real-life images of managing diretors were replaed by staged photos of models (taken by a professional photographer). These models enated the same poses as those in the authenti photographs, resulting in professional opies of the original versions of the images. Based on the advie of business ommuniation experts and based on the results of the pretests, one team image and a series of portraits of one individual were seleted from the initial olletion of photographs for inlusion in the final draft of the questionnaire. Two pairs of photos in the questionnaire showed the same individual portrayed in different poses and from different angles. One pair of pitures omprised an image of the individual model and an image of a management team inluding the individual model (setion G, Appendix X). Like the rating system for the text exerpts, eah portrait in a set of images was plaed at the extreme end of the fivepoint sale. As an illustration, the ross-ultural variations related to the visual theme all exeutive board members and the visual ode glosses fous on CEO/Chairman and fous on CEO + board members were aounted for by a sale polarizing an image of a management olletive and an image of an individual manager. For eah instane of the sale, partiipants were asked to express to what degree they preferred the visual design of either one of the ulture-speifi portraits. Partiipants were asked to mark the middle value of the sale if they had no lear or slight preferene for one of the images. Text usage Finally, the questionnaire ontained two multiple response items, two one-item fivepoint sales and an open-ended item for the analysis of partiipants general use of the annual report and of the Letters to the Shareholders (setion D in Appendix X). These measures were inluded to hek several of the observations in the investigation of genre ontext (see Chapter 4). As was outlined in setion 4.4, Duth and British ompanies had slightly different standards for the purposes and partiipants of their annual report and of their managerial forewords in partiular. To determine whether these ross-ultural distintions were refleted in the way UKbased readers tended to use annual report texts, they were asked to give their opinion about the relevane, objetives and sender of these texts. Any (British) ultural influenes would have important impliations for the international use of the Duth-

277 English annual report texts, sine the effetiveness of suh texts depends partly on the degree to whih foreign readers reognize and aept their rhetorial onventions. The multiple response items were inluded to identify what rhetorial funtions UK-based finanial professionals ommonly asribe to the annual report and to the management statement inside this report. The response ategories in these multiple hoie questions were based on the information aquired during the ontextual genre analyses of the Duth-English and British annual reports: eah of the two questions ontained a ompilation of ommuniative purposes that Duth and British orporations had assigned to either the annual report or the managerial statement. For example, partiipants were asked why they read the Letters to the Shareholder and were asked to mark one or more of the following options: (a) to get a quik informative view on the ompany s general performane, (b) to evaluate the general performane of the ompany, () to get an informative impression of the ompany s personality or harater, (d) to evaluate the ompany s personality or harater or (e) other. Two additional one-item five-point sales were inluded in the response study to reveal the importane of the annual report and the management statement in the professional lives of the partiipants. As for the annual report, partiipants were expliitly asked how important this doument is in their daily ativities. They responded using a five-point sale ranging from 1 not important at all to 5 very important. With respet to the importane of the managerial foreword, partiipants were asked how often they read this text when onsulting the annual report. For this item, they answered using a five-point sale ranging from 1 hardly ever to 5 almost always. These questions allowed a omparison between the orporations pereived importane of the texts and the interest that their finanial stakeholders atually show in these texts in pratie. The third type of test-item related to the use of the management statement in partiular involved an open-ended question about the presumed sender of this text. Setion showed that the introdutory Letter to the Shareholders in Duth and British annual reports is signed by different people, i.e. the CEO in the Netherlands versus the Chairman in the UK. Sine the questionnaire aimed to hek whether UK-based readers were aware of the fat that a real British Letter to the Shareholders is written by the supervising diretor and not by the managing diretor (like the suessive CEO s statement in British reports) partiipants were asked to indiate who they thought the produer of this text is Pretest of the questionnaire Before the questionnaire (i.e. the materials and measuring instruments) was presented to the partiipants, it was pre-tested among several finanial and ommuniation experts. The pretest was onduted with the help of four Duth finanial aademis (who were either PhDs or professors in Aounting and Finane), one British leturer in aademi writing for eonomi students and one Professor in Business Communiation studies. The finanial aademis were asked to judge the authentiity and interpretational effet of the figures and ativities desribed in the Letters and fragments; they also evaluated the possible impat of desribing the 2003 finanial year, as opposed to a more reent alendar year. Furthermore, the finanial respondents were asked to omplete the form and to onsider if they ould identify with the response ategories in the sales and in the

278 multiple response items. The native British leturer was asked to look at the omparability and larity of the English used in the two Letters and text fragments. Beause the fous needed to remain on the seleted rhetorial items or variables, any additional language-related effets had to be avoided as muh as possible. Moreover, the British leturer was asked to examine the English used in the response ategories, as some of the items in the questionnaire had been translated from Duth into English (e.g. the sale for text omprehensibility by Maes et al., 1996). The Business Communiations professor speifially looked at the ounterbalaning effet in the questionnaire as well as at the ambiguity, experimental relevane and user-friendliness of the response ategories. The results of the pretests indiated that a slight ross-textual differene in profit information was needed in the Letters, in order to imply the involvement of two distint orporations. If idential figures had been used, finanial readers would probably have imagined the same orporate sender. This might have aused the transferene of meaning-making aross the two Letters, e.g. if dividend details in the British text influened the interpretation of the Duth-English letter, whih might in turn have inluded idential profit figures but no expliit referene to the Dividend theme (as the text analyses revealed that it does not typially ontain this theme). The pretest respondents believed that the figures provided in the preliminary drafts of the Letters were realisti and that the small ross-textual differene in figures did not distrat reader attention away from the rhetorial design of the texts. They did not think that referenes to the 2003 finanial year would affet reader interpretations, sine both texts reported on the same year. Furthermore, the pretest results suggested that the English used in the Duth-English and British Letters and text fragments was lear, omprehensible and omparable. Some of the finanial aademis proposed several adjustments onerning the equivalene of English information sequenes, tone of voie in the English text and missing jargon ommon in international finanial English. With regard to the manipulation of the Letters, pretest partiipants laimed to be aware of the similar orporate stories refleted in the two texts, but aknowledged that they had different opinions about the two orporate senders beause of the way in whih a given story was told. Overall, the time it took to omplete the questionnaire appeared to be aeptable and the response ategories were of interest to their target group. The feedbak aquired in the pretest was used to realize the final version of the Duth-English and British Letters to the Shareholders Design and proedure A within-subjets design was used to measure the response to the Duth-English and British-English management statements among UK-based finanial professionals. Eah of the 35 respondents read and evaluated the Duth-English as well as the British-English version of the Letter to the Shareholders. The within-subjets design, whih repeatedly tests the same variables on the same subjet under different treatment onditions, is generally ritiized for its exposure to ontext effets that may alter the impat of a given treatment, e.g. sequene or arry-over effets and sensitization to the experimental hypotheses (Greenwald, 1976; Maes et al., 1996). However, in the present experiment its benefits made it a useful tool for the analysis of reader response to English annual reports: (1) the within-group design ontrolled for individual differenes, minimizing the effet of variability among individuals and highlighting the effet of variability in treatment, and (2) the within-group

279 design allowed for an experimental setting that was eologially valid, where finanial professionals ommonly ontrasting information soures of peer ompanies needed to read and evaluate texts of two different finanial orporations. To ounterbalane the ontext effets, moreover, both the order of the omplete English Letters and the order of the six English text fragments were rotated on the basis of a Latin-square design. This resulted in four sequenes of questionnaire items, whih were evenly distributed among the finanial professionals partiipating in the within-subjets experiment. Table 9.4 offers a onise overview of these four versions of the questionnaire. Although the number of partiipants in eah version was relatively small, a GLM was performed to test whether text version influened the sores on the dependent variables. In one ase (i.e. text redibility), there appeared to be an interation between the sores for Duth-English and British items and text version. As will be shown in setion 9.4.2, however, the separate effet of text redibility was not statistially signifiant in the experiment. Therefore, the effet of text version was limited. Table 9.4: Four versions of the questionnaire for UK-based analysts and brokers. Order English Letters to Shareholders Order of omparisons of English text fragments struture/metadisourse 1. British-English Duth-English + Headings Text stagers Salutation 2. British-English Duth-English + Text stagers Salutation Headings 3. Duth-English British-English + Headings Text stagers Salutation 4. Duth-English British-English + Text stagers Salutation Headings The English-language questionnaires were ompleted by analysts and brokers in the finanial enters of London (i.e. the City and Canary Wharf areas), between June They partiipated in the experiment during their work breaks. All partiipants worked individually, filling in their own questionnaire at a loation inside or outside their workplae. The time needed to omplete the questionnaire was minutes. While emphasizing the fat that the experimental materials were based on authenti annual report texts, partiipants were instruted to arefully read and evaluate the information presented as if they were atually judging a ompany s redibility and investment worthiness. To help avoid arry-over effets even further, they were asked not to leaf bakwards (or forwards) through the pages of the questionnaire; prior evaluative answers were to have as little effet as possible on the sueeding answers. After partiipants finished reading one of the full English Letters, they had to answer a series of questions omprising sales on whih they ould indiate their response to the orporate image refleted in the text and the persuasiveness of the text (i.e. in terms of attitude towards the text, text redibility and text omprehensibility). This proedure was repeated for the seond Letter in the questionnaire. Following the evaluation of the two omplete Letters to the Shareholders, partiipants had to answer several questions about the importane and purposes they generally asribe to the annual report and to the management statement in partiular. These questions were plaed between the full-text evaluations and the evaluations of Letter fragments, beause this would fore partiipants to distane themselves from their responses to the full-text questions before expressing their opinion about isolated instanes of typial text features. The questionnaire thus ontinued with several questions about speifi themes, moves or strategies, metadisursive elements and photographs of the Duth-English and British Letters to the Shareholders. A omplete example of version 1 of the

280 questionnaire is inluded in Appendix X. One the questionnaires were ompleted, they were handed bak to the researher, who thanked the partiipants and asked them if any additional questions had arisen while working through the experimental materials. All partiipants were offered a small Duth token of appreiation. The proedure for the within-subjets experiment was idential for all partiipants, and was arried out within a ontrolled environment for 20 out of the 35 respondents. The ontrolled environment omprised either a onferene room within an investment bank or a publi plae in the City and Canary Wharf areas where professionals took their offee or lunh. In all ontrolled ases, the researher remained in the viinity of the partiipants, to answer any intermediate questions or to redue distrations by external stimuli. Beause relatively few finanial professionals ould spare a break for partiipation in the response analysis and beause the ontrolled ases revealed no diffiulties, it was deided that those professionals willing to partiipate in their own time would also be inluded in the experiment. Consequently, 15 out of the 35 subjets read the texts and filled in the questionnaire in an unontrolled researh setting. These subjets were given the questionnaire and the instrutions on a ertain day and either personally returned the ompleted form a few days later or sent the form bak to the researher s work address through regular mail servies. Considering the findings in the pretests and in partiular the trouble-free ontrolled situations, these unontrolled ases were also inorporated in the final researh results. Independent-samples t tests showed no signifiant differenes between the experimental results of respondents within a ontrolled and respondents within an unontrolled researh situation Statistial analyses In the response experiment, all measures were designed to inlude omposite sales (see also setion 9.3.3). In line with Van Wijk (2000), it was deided that sales with low Cronbah s Alpha values (e.g. the text redibility sale for the Duth-English Letter to the Shareholders) would be preserved in the questionnaire. The experimental data for the Duth-English and British texts were analyzed and ompared using SPSS (version 14). Paired-sampled t-tests were performed to analyze the results for orporate image, attitude towards the text, text redibility, text omprehensibility and investment intentions. With regard to the full-letter preferenes, propositions, text fragments and photographs, one-sampled t-tests were used, with a theoretial mean or test value of 3. Test value 3 was assumed in order to measure the extent to whih the sample mean deviated from the most neutral reader evaluation in the sales (in all sales, sore 3 represented no preferene or no opinion ). Hypothesizing a higher appreiation for typially British genre features among the UK-based partiipants (H1-5, setion 8.2), all statistial analyses inluded one-sided testing. Multiple response frequenies were alulated to study subjets motivation for the full-letter preferenes, the pereived funtion of the annual report in general and the pereived funtion of the management statement in partiular. Frequenies were used to establish the pratial importane of both the annual report and the management statement and to determine who subjets believed to be the sender of this statement. Multiple responses and frequenies were intended to reveal whether the use of annual report texts by the UK-based partiipants atually orresponded to British orporations ideas about the purpose and partiipants of these texts (RQ, setion 9.2).

281 9.4 Results of the experimental reader response analysis The findings of the experiment are presented in the order of the hypotheses and researh question listed in setion 9.2. Therefore, the following subsetions suessively disuss the results for pereptions of orporate image, text persuasiveness, text preferene and the pratial use of annual report texts H1: effets on pereptions of orporate image The sores for partiipants ratings on the two image-related sales (i.e. orporate reputation and investment intentions) are shown in Table 9.5. A t-test indiated that the different variants of English Letters to the Shareholders signifiantly affeted pereptions of orporate image: the reputation of the ompany represented in the Duth-English Letter was rated as less attrative than the orporate reputation represented in the British Letter (t(34) = 2.52, p <.01, one-tailed). Although the ratings for investment intentions were again higher for the British-English Letter to the Shareholders, the variable English language variant did not have a statistially signifiant effet on the willingness to invest in either the Duth or British ompany (t(34) = 1.15, p =.13, one-tailed). Table 9.5: Pereptions of orporate image, tested with paired-sampled t-tests (1 = poor/zero likelihood and 5 = exellent/ertain). Pereived orporate image Mean SD t-value (df = 34) Corporate reputation/image Duth-English Letter British-English Letter Investment intentions Duth-English Letter British-English Letters p (onetailed) < H2-4: effets on pereptions of persuasiveness of the annual report text Table 9.6 presents partiipants sores on questions about the persuasiveness of the two English Letters to the Shareholders. These sores reveal that were no statistially signifiant differenes in the general attitude towards the texts (t(34) = 1.31, p =.10, one-tailed) and in the assessment of the redibility of the texts (t(34) = 0.84, p =.21, one-tailed). However, partiipants did have signifiantly different views on the omprehensibility of the two English Letters. Their ratings showed that they evaluated the Duth-English Letter as less omprehensible than the British- English Letter (t(34) = 1.88, p <.05, one-tailed).

282 Table 9.6: Pereptions of text persuasiveness, tested with paired-sampled t-test (1 = negative and 5 = postive). Pereived persuasiveness Mean SD t-value (df = 34) Attitude towards the text Duth-English Letter British-English Letter Text redibility Duth-English Letter British-English Letters Text omprehensibility Duth-English Letter British-English Letters p (one-tailed) < H5: effets on general preferenes for written texts and photographs General multimodal disourse preferenes were mapped in relation to an evaluation of the omplete written Letters, and to judgments of propositions on textual items, text-based exerpts and single management photographs. First of all, the sores for partiipants full-text preferenes are inluded in Table 9.7. Assuming a theoretial mean of 3 ( no preferene ), the response analysis revealed that UK-based professionals had a signifiantly higher preferene for the full-text version of the British-English Letter than for the omplete Duth-English Letter to the Shareholders (t(34) = 2.08, p <.05, one-tailed). Table 9.7: Full-text preferene, tested with one-sampled t-test and test value 3 (1 = Duth- English and 5 = British-English). Full-text preferene Mean SD t-value p (one-tailed) (df = 34) Letter to the Shareholders (written text) <.05 Table 9.8 presents an overview of the top three aspets that were repeatedly ited by the partiipants to motivate their preferene for either one of the English Letters to the Shareholders. Interestingly, many of these fators seem to be onsistent with the ulture-speifi genre features typial to the preferred text. For instane, partiipants laim that the Duth-English Letter was most attrative beause of its detailed desription of operations seems to be onsistent with the fat that the move/strategy Presenting operational performane/disussing finanial details was inorporated in the Duth-English Letter but not in the British-English Letter. Following the design of the authenti management statements, the former inluded more information on operational ativities where the latter inluded intertextual referenes ( previews ) to other texts in the annual report. Furthermore, Table 9.8 shows that the attrativeness of the British-English Letter was often asribed to its lear, well-strutured organization. Again, this explanation is in line with the typial use of text fragments with headings in British-English management texts and the general absene of these strutural devies in Duth-English Letters. It should be noted that some of the main reasons for preferring the British Letter over the Duth- English Letter are rather abstrat; aspets suh as ompleteness and professionalism

283 of the text or quality of English annot be diretly assoiated with any of the genre features dealt with here. Table 9.8: Motivation for preferene Duth-English or British Letter ( top three, per Letter variant). Motivation for preferene Duth-English or British Letter Duth-English Letter to the Shareholders (n=13) 1. Clear struture text stagers 2. Details on orporate strategy 3. Muh detail on operational ativities British Letter to the Shareholders (n=21) 1. Clear struture headings 2. Finanial details dividend 2. Conise overview of ativities 2. Complete overview of ativities 3. Professional text 3. Better English Perentage (ount) 38.5 (5) 23.1 (3) 15.4 (2) 71.4 (15) 19.0 (4) 19.0 (4) 19.0 (4) 14.3 (3) 14.3 (3) Next, partiipants multimodal disourse preferenes expressed through the statements, text exerpts and photographs are displayed in Table 9.9. Based on test value 3, the results for the propositions reveal that the UK-based partiipants had a signifiantly higher preferene for Letters inluding the following typially British themes: management details (t(34) = 6.83, p <.001, one-tailed), CSR (t(34) = 1.87, p <.05, one-tailed), and dividend (t(34) = 8.62, p <.001, one-tailed). Conversely, Table 9.9 indiates that UK-based professionals expressed a signifiantly higher preferene for Letters inluding the typially Duth-English orporate strategy theme (t(34) = 10.53, p <.001, one-tailed) and the strategy providing future objetives (t(34) = 7.04, p <.001, one-tailed). The findings for the text exerpts show that the UK-based partiipants rated British Letter fragments with headings and informative previews as more preferable than Duth-English Letter fragments without headings and with additional operational information (t(34) = 4.43, p <.001, one-tailed). There was no statistially signifiant differene for the assessment of fragments either inluding or exluding the Duth-English text stagers (t(34) = 0.52, p =.30, one-tailed) and salutations (t(34) = 0.16, p =.44, one-tailed). Ratings for the management portraits led to signifiant results for all visual genre features tested. Partiipants had a signifiantly higher preferene for Letters ontaining typially Duth portrayals of the management team as opposed to the individual manager t(34) = 1.93, p <.05, one-tailed) and of the management in medium shot as opposed to lose-up (t(34) = 5.37, p <.001, one-tailed). However, they rated typially British pitures of managers looking away from the amera as opposed to looking into the amera as signifiantly more preferable (t(34) = 2.08, p <.05, one-tailed).

284 Table 9.9: Preferene for various textual and visual elements of the English Letters to the Shareholders, tested with one-sampled t-test and test value 3 (1 = as in Duth-English and 5 = as in British-English). Preferene textual and visual elements Mean SD t-value (df = 34) Propositions Letter should inlude Theme: management situation Theme: CSR Theme: dividend details Theme: orporate strategy Struture: short-term future objetive Text fragments Preferene for Move: (no) headings/preview Metadisourse: (no) text stagers Move: (no) salutation Photographs Preferene for Code gloss/self-mention: team indiv Attitude: view into/away from amera Engagem.marker: medium/lose-up shot p (onetailed) <.001 <.05 <.001 <.001 <.001 < <.05 <.05 < RQ: the use of annual report texts Finally, the experiment foused on the professionals general appliation of annual reports and of management statements in partiular, i.e. in terms of the pereived usefulness, purpose and sender of these texts. Table 9.10 presents details about the pratial relevane partiipants asribed to the annual report on the one hand and the Letter to the Shareholders on the other. With regard to the importane of the annual report in their professional lives, the mean sore of 3.71 suggests that the full doument is at least of some interest and thus a relevant soure of information for orporate evaluations. As is shown in Table 9.10, a majority of the urrent sample of partiipants onsidered the annual report to be important (22.9%) to very important (37.1%). Judging from the present results on pereived importane of the annual report, it seems that these are in line with the notion shared by both Duth and British ompanies that the full report is intended to ommuniate with finanial stakeholders in partiular (see setion 4.4.2). A different result emerged for the pereived relevane of the speifi management statement. The mean sore for reading frequeny of the management statement (2.74) and the relatively large number of partiipants that marked the answering ategories of oasionally (25.7%) and hardly ever (25.7%) indiate that the text is read only sporadially by this finanial target group. Several partiipants explained that the management letter usually omprises a trifling amount of finanial information, thus offering insuffiient detail for orporate qualifiations. They believed that other annual report texts, suh as the Operating or Finanial reviews, provide more omprehensive performane desriptions and are therefore more important to the exeution of their professional funtion.

285 Table 9.10: Pereived pratial relevane of annual report and management statement (1 = not important at all/hardly ever and 5 = very important/almost always). Pratial relevane annual report (text) Frequeny % (ount) Importane annual report in professional life Not important at all 2.9 (1) Not very important 20.0 (7) Of some importane 17.1 (6) Important 22.9 (8) Very important 37.1 (13) Reading frequeny management statement Hardly ever 25.7 (9) Oasionally 25.7 (9) Half of the time 14.3 (5) Most of the time 17.1 (6) Almost always 17.1 (6) Mean SD The findings in Table 9.11 reveal the UK-based professionals pereptions about the ommuniative objetives of the annual report and of the management statement. Their sores suggest that the omplete annual report serves informative and evaluative purposes primarily. As for the informative value of the annual report, the partiipants in the urrent experiment tended to fous on details about orporate results (18.4%) and orporate strategy (17.6%). Regarding the evaluative aim of the annual report, these partiipants laimed that they used the report to assess the orporate profitability (15.4%) and future prospets (14.0%) in partiular. Notwithstanding the audiene-speifi fous on partiular (finanial) issues in the annual report, the balane between informative and evaluative objetives was onsistent with the genre ontext depited by the Duth and British orporations (see setion 4.4.1). Both the English annual reports produed in the Netherlands and in the UK were assigned a primary informative purpose and a seondary promotional purpose, where the promotional value of the report did not omprise the atual influening of investment behavior (see setion 4.4.1). Table 9.11 also inludes partiipants sores on a multiple response question about the ommuniative purposes of the management statement in the annual report. It points out that this text is espeially useful as a soure of performane highlights. Almost half of the partiipants responses (43.1%) indiate that the managerial foreword is usually read to get a quik informative outline of the orporate results. Additionally, 19.6% of these responses showed that management statements should inform the reader about the personal nature of the ompany. Although orporate identifiation played a minor role in the ommuniative aims of the report in its entirety, several partiipants seemed to think that the introdutory management letter speifially should be designed to leave a personalized impression of the ompany. In omparison to the pereived aims of the management letter to the purposes set by the Duth and British orporations (in setion 4.4.4), it an be observed that partiipants responses primarily reflet genre rules that were attributed to the Duth-English CEO s statement and the British Chairman s statement. In line with ompanies assumptions about the Duth-English CEO s statement and the British Chairman s statement, UK-based professionals laimed that the Letter to the Shareholders should provide an informative summary of results, a personalized view on the orporate identity and a sense of goodwill onerning the orporate identity.

286 Table 9.11: Pereived purposes of the annual report + pereived purposes of the management statement (most popular purposes, > 3% ourrene). Text type + ommuniative purposes/sender of the text Annual general report ommuniative purposes (N = 35, 136 responses): 1. To get informed about ompany s operational/finanial results 2. To get informed about the ompany s strategy 3. To evaluate the ompany s profitability 4. To evaluate the ompany s future prospets 5. To get informed about the ompany s identity 5. To evaluate the ompany s effiieny 6. To failitate investment deision-making 7. To diretly ativate investment behavior Management statement ommuniative purposes (N = 35, 51 responses): 1. To get a quik informative view on ompany s performane 2. To get an informative impression of the ompany s personality 3. To evaluate the ompany s personality 4. To evaluate the ompany s general performane 5. To get a feel for the leadership quality 5. To evaluate management relationship with shareholders 5. To get a feel for the orporate strategy Perentage (ount) 18.4 (25) 17.6 (24) 15.4 (21) 14.0 (19) 9.6 (13) 9.6 (13) 6.6 (9) 3.7 (5) 43.1 (22) 19.6 (10) 11.8 (6) 9.8 (5) 3.9 (2) 3.9 (2) 3.9 (2) The final ontextual item that was inorporated in the experiment onerned the presumed writer of the Letter to the Shareholders, i.e. the person signing the text. The UK-based professionals provided a variety of possibilities, the most popular of whih are presented in Table It shows that 27.3% of the partiipants identified the Chairman of the ompany as the general writer or signer of the introdutory foreword in the annual report; this is indeed the ase for introdutory management statements in British annual reports (see setion 4.4.4). As many as 30.3% of the partiipants, however, responded that the CEO or operational exeutive is the typial sender of the management letter. This is onsistent with the Duth-English genre rules for the management introdution. However, it is not in line with the British- English genre onventions in whih CEOs fulfill the role of protagonist in follow-up but not in introdutory statements (setion 4.4.4). A small group of partiipants (21.2%) indiated that the writer of the management foreword ould either be the Chairman or the CEO. Table 9.12: Pereived sender of the management statement (most popular senders, > 3% ourrene). Text type + ommuniative sender of the text Management statement sender of the message (N = 33): 1. CEO 2. Chairman 3. Chairman/CEO 4. Communiations/Investor Relations offiers 5. Do not know Perentage (ount) 30.3 (10) 27.3 (9) 21.2 (7) 9.1 (3) 6.1 (2)

287 9.5 Conlusion and disussion The reader response experiment presented in this hapter revolved around the interultural effetiveness of multimodal texts in non-native English annual reports. More speifially, it aimed to analyze how typial textual and visual features of Duth-English and British-English annual reports influene UK-based readers pereptions of orporate image and the persuasiveness of the text (see RQ II at the beginning of this hapter). After mapping partiipants separate assessments of ulture-speifi genre features in omplete management statements and in fragments of suh statements (RQ 3a-b), the experiment foused on the omparison of assessment sores to determine if there was a differene in effetiveness between the multimodal features of the Duth-English and British annual report text (RQ 3). In general, the findings indiated that UK-based finanial professionals had a higher appreiation for the omposition of those genre features that are typial of the British-English management statement. However, the results also revealed patterns of appreiation that did not differ signifiantly aross the English text variants, and patterns of appreiation in whih the Duth-English genre features showed signifiantly better sores. More speifially, UK-based finanial professionals expressed a more positive opinion about the textual design of the British-English than of the Duth-English Letter to the Shareholders, but overall voied a more positive attitude towards the visual design of the Duth-English than of the British managerial foreword. As suh, these observations are in line with prior researh outomes. Some analyses of the reeption of professional genres have shown that texts designed in aordane with genre onventions ommon in the subjets ommunity are more effetive than texts violating these ommunityspeifi rules; they have shown that onventional texts eliit greater reader appreiation (Lagerwerf & Bossers, 2002) or higher performane advantages in text usage (Vaughan & Dillon, 2006). Experiments inluding multiple ommunities or ultures have exhibited a more subtle relationship between the use of partiular genre features and genre effetiveness. Shaw et al. (2004), for instane, indiate that subjets from different nations may have ulture-speifi preferenes for a given generi style but may aept other styles that are typial to foreign national or business ultures. They argue that this ould be owing to the gloalization proess, with loal preferenes surviving against a bakground of international homogenization (Shaw et al., 2004: 396). The following subsetions will draw further onlusions on the effetiveness of Duth-English and British annual report texts. Subsetion provides an answer to the hypotheses and additional researh question introdued in setion 9.2. In subsetion 9.5.2, a brief overview will be given of the general impliations for genre theory, genre analysis and interultural genre use. Therefore, this subsetion relates to the seond main researh aim of this genre study. Subsetion inludes researh limitations and impliations for further studies. Finally, subsetion ontains several losing remarks Interultural effetiveness of genre features in English annual reports This setion summarizes the experimental results and outlines whether or not the expetations in hypotheses 1 to 5 (H1-H5) were onsistent with UK-based professionals pereptions of disourse in the Duth-English and British annual

288 report texts. It also answers the additional researh question (RQ1) with regard to the pratial relevane of annual reports and management statements in the lives of finanial analysts and brokers. H1: more positive orporate image refleted in British annual report text With regard to the orporate image refleted in the textual onstrut of omplete Letters to the Shareholders, there was one statistially signifiant differene between the partiipants evaluation of the Duth-English and British texts. In terms of pereived orporate reputation, the British Letter was evaluated as more positive than the Duth-English Letter. However, the results also indiated that the use of Duth-English or British text features in the full Letters did not affet partiipants investment intentions. Apparently, ulture-speifi genre harateristis an influene pereptions about the general orporate image, but they do not have a signifiant impat on the finanial reader s reported willingness to invest in either one of the ompanies represented in the texts. Perhaps this is owing to the fat that the intention to atually purhase shares, depends on more ruial fators, e.g. a ompany s eonomi history. Therefore, H1 was partly supported. H2-4: greater persuasiveness of the British annual report text The findings for the pereived persuasiveness of the full English Letters to the Shareholders revealed that ulture-speifi text ontent and text struture did not have a statistially signifiant effet on subjets appreiation of (H2 attitude) and onfidene in (H3 redibility) the text. Conversely, UK-based professionals aquainted with British-English texts rated the typial British-English Letter as signifiantly more omprehensible (H4 omprehensibility) than the onstrut of the typial Duth-English Letter. In other words, the partiipants did not like or trust the British-English text signifiantly more than the Duth-English text, but they did appreiate the British-English text more for its level of omprehensibility. Aordingly, although the results did not provide empirial support for H2 and H3, they did onfirm H4. It should be noted that the results for the pereived redibility of the text are reliable to a limited degree, owing to the low Cronbah s Alpha sores for the redibility sale (see setion 9.3.3). H5: greater preferene for the British annual report text The experiment revealed a statistially signifiant differene between partiipants general evaluation of the full Duth-English and British Letters. In addition, it pointed to mixed results for partiipants general assessment of speifi text-based and photo-based genre features in the Duth-English and British Letters. As for the full-text evaluations, UK-based professionals regarded the textual design of the British-English Letter to the Shareholder as more preferable than the Duth-English text. Interestingly, this preferene largely depended on those aspets of the text for whih the British-English Letter also showed a signifiantly higher sore: omprehensibility of the organization or oherene of the text. With regard to professionals evaluations of isolated textual harateristis, it was observed that there was a signifiantly higher preferene for British-English genre onventions related to the inlusion (as opposed to the exlusion) of management details, CSR details, dividend details and headings or intertextual previews. Nevertheless, the UK-based partiipants passed more favorable judgments on management letters inluding (as opposed to exluding) typial Duth-English elements suh as orporate strategy and future objetives. Their evaluation of the Duth-English and British texts did not show a signifiant preferene for the presene or absene of text

289 stagers and salutations. With respet to partiipants ratings for isolated visual features, it was found that the photographi design of the Duth-English Letter to the Shareholder in partiular was preferred. In line with Duth-English genre onventions, partiipants preferred images of the management olletive (as opposed to individual managers) and medium-shots (as opposed to intimate loseups) of managers as a means of introduing the orporate sender of the message. They only expressed a greater preferene for the British way of pituring managers with their view away from the amera lens. In short, the researh results suggested that there was partial empirial support for H5. These observations are an indiation that the Duth-English genre features that had been embedded in the omplete Letter and for whih hypothesis five was not supported in isolated examples (i.e. the orporate strategy and future objetives themes, text stagers) were not deisive in partiipants full-text evaluations. Presumably, the importane of typial British genre harateristis suh as the dividend theme or the strutural headings outweighed the relevane of these Duth-English features in their assessment of the full text. RQ: the use of annual report texts by UK-based finanial professionals The experimental findings indiated that British ompanies pereptions of the purposes and partiipants of annual report texts (see setion 4.4) was onsistent with partiipants use of the annual report but was only partially onsistent with their use of the management statements. Partiipants pereptions about the pragmati relevane of the annual report indiated that their use of the full text mathes the text usage antiipated by both British and Duth orporations: UK-based professionals indiated that the annual report fulfils an important to very important funtion in their lives. Conversely, the sores for reading frequeny of the management statement implied that the finanial professionals should not be seen as a major audiene for this partiular text. A majority of the partiipants tended to read the managerial foreword oasionally or hardly ever. This would seem to have little onsequene for the Duth-English version of the text, whih is designed to reah all types of international stakeholders, but it might have larger impliations for the British-English version of the statement, for whih the investors are a main reader audiene. With respet to the ommuniative funtion of the annual report in its entirety, the UK-based partiipants marked purposes that were onsistent with orporate requirements for the rhetorial value of the Duth-English and British annual reports, i.e. (1) to be informed about the orporate results and prospets, and (2) to be persuaded of the finanial health of the ompany (see setion 4.4.4). This onlusion does not quite hold for the pereived ommuniative funtion of the management statement. While Duth and British orporations mainly haraterized the management Letter as a personal introdution to the ompany and its results, the partiipants mainly used it for the purpose of obtaining highlights about the orporate results. The subsequent suggestion that the introdutory Letter might not meet all finanial reader needs was onfirmed by some of the professionals, who pointed out that this text is uninteresting beause of its lak of detail. It ould be argued, however, that other text types inluded in the Duth-English and British annual reports (e.g. the British CEO s statement or the Duth-English operational and finanial reviews) might make up for readers pereived defiienies in the atual management introdution. Finally, it was observed that the UK-based professionals had different ideas about the professional funtion of the writer of the Letter to the Shareholders. Thus,

290 they had different notions about the frame of referene within whih the Letter is ommonly produed. Their varied entries might be attributable to the fat that they had sometimes gained experiene with annual reports of both Anglo-Saxon and non- Anglo-Saxon orporations. Unlike the British orporations, many non-anglo-saxon ompanies suh as the Duth have annual reports ontaining one introdutory management statement signed by the CEO (see also Garzone, 2004). It ould be argued that in the interultural use of the Duth-English CEO s foreword, for instane, readers assuming a hairman-as-writer will miss supervisory details, e.g about dividend. Although this information is offered in the Duth-English Report of the Supervisory Board, its omission in the managerial foreword might influene the effetiveness of this partiular text. Considering the experimental results in general, it would seem plausible to suppose that the typial British-English genre features have a stronger positive effet on the UK-based finanial professionals than the typial Duth-English features. In terms of effetiveness, the Duth-English Letter to the Shareholders therefore appears to be slightly less suessful than the British-English Letter. Aordingly, it may be onluded that these observations largely support the theoretial assumption that disourse onventions and expetations imprinted in the ulture or ommunity to whih the reader aspires or belongs have a basis in the reader s text evaluation (e.g. Byrne, 1971; Shadid, 2000). Established members of a language use ommunity possess a speial familiarity with text types that onform to the shared values and norms. In the ase of genre reeption, suh text types will be easily reognized and are likely to eliit a more positive response than non-onforming text types (Berkenkotter & Hukin, 1995; Lagerwerf & Bossers, 2002; Swales, 1990; Vaughan & Dillon, 2006). It should be noted that these onlusion do not hold for all disourse features analyzed. As was the ase in Shaw et al. (2004), the present findings also indiated that UK-based partiipants had a higher preferenes for speifi genre features that were ulturally inongruent. Considering their appreiation for partiular textual and visual elements typial of the Duth-English annual report, this might indiate a onvergene or harmonization of national norms in finanial disourse Impliations for genre theory, genre analysis and genre pratie The results of the reader response experiment put forward several impliations for the theorization and identifiation of professional genres, as well as for the appliation of professional genres in interultural settings. This subsetion will elaborate on these impliations and, as suh, adds to the realization of the seond main researh purpose whih fouses on the development of genre theory as a researh tool for multi-method analyses (see setion 1.5). The theoretial relevane of interultural genre reeption The present experiment was based on the assumption that genres, as a produt of living in a soial ontext whih requires the use of texts to ommuniate meaning, inorporate both the prodution and reeption of texts (see also Solbjørg Skulstad, 2005; Vaughan & Dillon, 2006). The relationship between the ontext-dependent reation of a genre and its ontext-dependent interpretation would seem to be of partiular importane in the interultural use of genres. This was illustrated in the present response study. Whereas the preeding genre analyses of (on)text features

291 indiated that Duth and British orporations rely on different ultural expetations with regard to the English-language supply of orporate information in their annual reports, the experiment suggested that the reeivers of these international business texts also interpret these texts on the basis of their own ultural expetations. As Devitt puts it, readers rereate texts in their own ontext (2000: 710). In general, the results indiated that the genre onventions applied in the full British-English annual report text were more in line with UK-based partiipants pereptions of appropriate finanial disourse than the genre onventions used in the Duth-English annual report text. Nevertheless, all partiipants were under the impression that they were reading instanes of the same genre, irrespetive of the differenes in genre features. The present researh outome thus suggests that (professional) genres, although produed in various ultures, an survive internationally owing to readers reognition of similarities in (on)text features. However, it adds to this onlusion that the ategorization of a genre differs from the reeption of the genre; international readers may express greater appreiation for ulture-speifi (substantive or formal) realizations of genre instanes. In sum, the present reader response experiment has shown that it is insightful to take genre theory beyond the lassifiation or desription of partiular text types and to embrae the onept of genre reeption as a means of disussing the effet(iveness) of partiular genres. Following Askehave and Swales, it an be argued that a fous on genre reeption deepens and widens our understanding of the role of disourse in ontemporary soiety (2001: 196). The analytial relevane of genre ategorization in the study of genre reeption The urrent study has shown that ross-ultural genre investigations an be used as guiding priniples for studies on the interultural response to genres. More speifially, it has shown how different variants of English as an international finanial language an have a sinere impat on the effetiveness of multimodal annual reports exhanged in Western European business relations. As suh, this study has highlighted the possible expansion of genre researh in general. To date, genre researh has mainly omprised text identifiation on the basis of qualitative desriptions or orpus analyses (e.g. Zhu, 2005; Upton & Connor, 2001) or to a lesser extent reeption analysis on the basis of experiments or textual response studies (e.g. Lagerwerf & Bossers, 2002; Solbjørg Skulstad, 2005). As was suggested in the present investigation, however, these different approahes an be ombined to establish a multi-method and omprehensive examination of (professional) text types that are produed in one partiular ulture but used aross the globe. It also onfirms the idea that the results of prior ross-ultural studies on genre realizations (e.g. Dahl, 2004; Mauranen, 1993, see setion 9.2) an be used in investigations that measure the international reader effet of any ross-ultural variations observed. The pratial relevane of the study of interultural genre reeption In spite of the regular involvement of native English editors in the prodution of Duth-English annual reports (see setion 4.3), the genre analyses in Chapters 5 to 7 revealed the existene of multimodal text features that are typial of the Duth- English disourse, as opposed to the British-English disourse. Subsequently, the response experiment in the present hapter showed that these ulture-speifi features yield an interpretational effet among members of an important international reader audiene. For ruial onepts suh as orporate image and text omprehensibility, the impat of Duth-English disourse onventions was more

292 negative than the impat of British disourse onventions. Aordingly, the experimental results suggested that the use of English in Duth-based annual reports may not always render the desired rhetorial effet. In ontrast with Duth ompanies pereptions about the funtionality of English as an international finanial language (see setion 4.4.5), the use of Duth-English in their annual reports may not always failitate interultural ommuniation. In ontrast with more general pereptions about the value of international business English (e.g. Gerritsen et al., 2007; Piller, 2001), moreover, the use of Duth-English in annual reports may not always enhane the pereived image of the ompany responsible for the text. Within the frame of the present study, this ould have serious onsequenes for Duth ompanies depending on stakeholders in the British market in partiular (as the partiipants in the experiment were working and residing in the UK). Further researh should be onduted with regard to the effet of Duth-English annual report features on reader pereptions in different foreign markets. In broader terms, the present experiment underpins the notion that suessful interultural ommuniation an be seen as a [ ] proess of ahieving optimal ompetene aross ultures (Zhu, 2005: 185). It implies that the (further) aommodation of English in international business texts to standards of English ommon in the reeiving ulture will enhane the hane of positive responses to the texts and thus of realizing their ommuniative purposes Limitations of the interultural response study and suggestions for further researh Despite the steps taken to maximize the eologial validity of this study, some artifiialities might have influened the experimental situation. These will be disussed below. Parameters for the seletion of partiipants The present study assumed that three parameters were ruial in the seletion of partiipants: (1) general professional funtion, i.e. investment advisor, (2) resideny/ workplae in the United Kingdom and (3) regular evaluation of British orporations. Parameters suh as gender, years of experiene, advisory fous or native ountry were not taken into aount. For the purposes of this exploratory experiment, the level of deliay for the sampling riteria seemed to be suffiient. Moreover, these riteria enabled the realization of a sample of respondents who are not easily approahable and whose work shedules do not easily aommodate partiipation in the experiment. Even so, a further refinement of the subjet profile ould yield additional details. The importane of paying attention to advisory fous, for example, was underlined by one finanial professional who suggested that the genre of the management statement generally is more relevant to advisors working loser to the orporate market (e.g. brokers). Involvement in the researh materials The responses onerning the usefulness of the management statement in partiular implied that subjets were probably not very involved in the experimental texts. A majority of the professionals noted that the introdutory Letter is sporadially used as a soure of information; in a normal reading ontext, many subjets would have deided to avoid this text and move on to another annual report text. Therefore, their ratings of orporate image, persuasiveness of the text and text preferenes might not

293 always have been grounded on areful judgments of the English texts. This makes the urrent experimental situation similar to the onditions in Hoeken and Renkema (1998), whih measures the orporate image-related effet of orretions in newspaper artiles. One way to get around these issues in a subsequent study would be to use different annual report setions for researh materials. Several of the subjets pointed out, for instane, that they tend to grant more redibility to the operating and finanial reviews, whih ontain additional information they an atually use for their orporate evaluations. And following Hoeken and Renkema, as interest inreases, the probability that they form a well-based judgment will inrease as well (1998: 59). Time pressure during partiipation A ongruent onern is the redued involvement of subjets due to time pressure. In the experiment, subjets were asked to interpret and evaluate two full texts, six text fragments, six photographs, five propositions and a few multiple hoie ategories. Although the questionnaire was designed to be as onise as possible, this was a omprehensive task for a professional reader audiene that is under ontinuous time pressure. Aordingly, some subjets might have sanned through the researh materials, ausing somewhat superfiial judgments of the English text variants. Notwithstanding the external validity of this situation (with regard to the daily ativities of the finanial professionals), the impat of time limits might have been mitigated by presenting subjets with one text version, i.e. in a between-subjets experiment. This would probably require muh effort in terms of reruiting a suffiient number of finanial professionals, but it would be interesting to test the urrent hypotheses for international partiipants who have read only the Duth- English or British-English management statement. A fous on isolated photographs Finally, the eologial validity might have been affeted by the restrited number of ontrolled fators in the analysis of photographs. This experiment ontrolled for the gender and appearane of the models. As the models needed to resemble authenti images of real-life managers in their oupational surroundings, they were seleted on the basis of appropriate gender, age, posture and faial expressions. Moreover, the fat that the same male was used in the individual portraits and in the portrait of the management team ontrolled for bias with respet to the portrayed persons (e.g. Orth & Holanova, 2004). Conversely, the response experiment did not ontrol for another important aspet, i.e. the positioning of the photographs in the multimodal text. Beause the subjets fous needed to remain on the manner of portrayal, it was deided to rule out any environmental influenes and submit the images in isolation. Sine photos in annual reports ommonly appear in a multimodal spae in whih texts and images are omplementary (Kress & Van Leeuwen, 2001, 2006), however, reader evaluations of pitures-in-ontext ould produe different pereptual results. Therefore, further reeption researh should be arried out in whih a ombination of written texts and photographs are presented to the partiipants Closing remarks Drawing on the text-in-ontext perspetive within ESP/LSP genre theory (e.g. Askehave & Swales, 2001; Bhatia, 1993, 2004), this study inluded a ombination of ross-ultural and interultural disourse analyses. More speifially, it foused

294 on the realization and reeption of textual and visual features in English annual reports originating in two European business ultures. While the ross-ultural investigation showed that a partiular professional genre may ontain different multimodal realizations aross ultures, the interultural analysis indiated that different realizations of the same professional genre may influene international reader response. Both types of analyses yielded impliations for the further refinement of theoretial onepts and for the extension of methods for genre researh. Sine this study explored several innovative aspets of the analysis of Duth-English and British annual reports in partiular, it has left sope for further researh on multimodal genres that are produed in different ultures and used in interultural business relations.

295 Referenes Aounting Standards Board. (2005). ASB Issues Reporting Standard on the Operating and Finanial Review (OFR). Press noties. Retrieved 31 July 2007, from After Babel, a new ommon tongue: The European Union. (2004, August 7). The Eonomist (US). Ahuvia, A.C. (1998). Soial ritiism of advertising: on the role of literary theory and the use of data. Journal of Advertising, 1, Akar, D. & Louhiala-Salminen, L. (1999). Towards a new genre: a omparative study of business faxes. In F. Bargiela-Chiappini & C. Nikerson (Eds.), Writing business: genres, media and disourses (pp ). Essex, UK: Pearson Eduation Ltd. Ammelrooy, P. v. (2003, April 2 ). Philips doet jaarverslag in moerstaal in de ban (Philips outlaws mother tongue for annual report). De Volkskrant. Anderson, C. J. & Imperia, G. (1992). The Corporate Annual Report: A Photo Analysis of Male and Female Portrayals. Journal of Business Communiation, 29(2), Aron, A. & Aron, E. (1997). Statistis for the Behavioral and Soial Sienes: A Brief Course. Upper Saddle River, NJ: Prentie Hall. Askehave, I. (1998). A funtional interpretation of orporate brohures: from ontext to text. Unpublished PhD thesis, The Aarhus Shool of Business. Askehave, I. (1999). Communiative purpose as genre determinant. Hermes - Journal of Linguistis, 23, Askehave, I. & Nielsen, A. E. (2005). Digital genres: a hallenge to traditional genre theory. Information Tehnology & People, 18(2), Askehave, I. & Swales, J. M. (2001). Genre identifiation and ommuniative purpose: a problem and a possible solution. Applied Linguistis 22(2), Badger, R. & White, G. (2000). A proess genre approah to teahing writing. ELT Journal, 54(2), Bakhtin, M. (1986). The problem of speeh genres. In M. Bakhtin (Ed.), The problem of speeh genres and other late essays (pp ). Austin: University of Texas Press. Barbara, L., Celani, M.A.A., Collins, H. & Sott, M. (1996). A survey of ommuniation patterns in the Brazilian business ontext. English for Speifi Purposes, 15(1), Bargiela-Chiappini, F. (1999). Meaning reation and genre aross ultures: human resoure management magazines in Britain and Italy. In F. Bargiela- Chiappini & C. Nikerson (Eds.), Writing business: Genres, media and disourses (pp ). London and New York: Longman. Bargiela-Chiappini, F. & Harris, S. J. (1997). Managing Language: The disourse of orporate meetings. Amsterdam, Philadelphia: John Benjamins Publishing Company. Bargiela-Chiappini, F. & Nikerson, C. (1999). Business Writing as Soial Ation. In F. Bargiela-Chiappini & C. Nikerson (Eds.), Writing business: Genres, media and disourses (pp. 1-32). London and New York: Longman. Bargiela-Chiappini, F. & Nikerson, C. (2002). Business disourse: old onepts, new horizons. International Review of Applied Linguistis in Language Teahing, Speial Issue on Business Disourse, 40(4),

296 Bargiela-Chiappini, F. & Nikerson, C. (2003). Interultural Business Communiation: a rih field of studies. Journal of Interultural Studies, 24(1), Barron, A. (2006). Understanding Spam: a maro-textual analysis. Journal of Pragmatis, 38(6), Barton, E. L. (1995). Contrastive and Non-Contrastive Connetives. Written Communiation, 12(2), Basturkman, H. (1999). A Content Analysis of Elt Textbook Blurbs: Refletions of Theory-in-Use. RELC Journal, 30(1), Bateman, J., Delin, J. & Henshel, R. (2006). Mapping the multimodal genres of traditional andeletroni newspapers. In T. Roye & W. Bowher (Eds.), New Diretions in the Analysisof Multimodal Disourse (pp ). New Jersey: Lawrene Erlbaum Assoiates. Bazerman, C. (1994). Systems of genres and the enatment of soial intentions. In A. Freedman & P. Medway (Eds.), Genre and the New Rhetori (pp ). London: Taylor and Franis. Bazerman, C. (1997). Disursively strutured ativities. Mind, Culture, and Ativity, 4(4), Bazerman, C., Little, J. & Chavkin, T. (2003). The Prodution of Information for Genred Ativity Spaes: Informational Motives and Consequenes of the Environmental Impat Statement. Written Communiation, 20(4), Beaufort, A. (1997). Operationalizing the onept of disourse ommunity: a asestudy of one institutional site of omposing. Researh in the Teahing of English, 31(4), Beaufort, A. (2000). Learning the Trade: A Soial Apprentieship Model for Gaining. Written Communiation, 17(2), Benshop, Y. & Meihuizen, H. E. (2002). Keeping up gendered appearanes: representations of gender in finanial annual reports. Aounting, Organizations and Soiety, 27(7), Berens, G. & Riel, C. B. M. v. (2004). Corporate assoiations in the aademi literature: three main streams of thought in the reputation measurement literature. Corporate Reputation Review, 7(2), Berkenkotter, C. & Hukin, T. (1995). Genre knowledge in disiplinary ommuniation: ognition/ulture/power. Hillsdale, NJ: Lawrene Erlbaum Assoiates. Berns, M. (1995). English in the European Union. English Today, 11(3), Bhatia, V. K. (1992). Pragmatis of the Use of Nominals in Aademi and Professional Genres. Pragmatis and Language Learning, 3, Bhatia, V. K. (1993). Analysing genre: Language use in professional settings. Essex: Longman Group Ltd. Bhatia, V. K. (2002). Applied genre analysis: a multi-perspetive model. Iberia, 4, Bhatia, V. K. (2004). Worlds of written disourse: a genre-based view. London: Continuum. Bjorge, A. K. (2007). Power distane in English lingua frana ommuniation. International Journal of Applied Linguistis, 17(1), Bondi, M. (2002). Attitude and Episteme in Aademi Disourse: Adverbials of Stane Aross Genres and Moves. Textus, 15(2), Booij, G. (2001). English as the lingua frana of Europe: a Duth perspetive. Lingua e Stile, 36(2),

297 Borg, E. (2003). Key onepts in ELT: disourse ommunity. ELT Journal, 57(4), Brammer, S. J. & Pavelin, S. (2006). Corporate Reputation and Soial Performane: The Importane of Fit. Journal of Management Studies, 43(3), Bruffee, K. A. (1986). Soial onstrution, language and the authority of knowledge: a bibliographial essay. College English, 48, Brumfit, C. J. (2001). Individual Freedom in Language Teahing: Helping Learners to Develop a Dialet of their Own. Oxford: Oxford University Press. Burr, V. (1995). An introdution to soial onstrutionism. London: Routledge. Burr, V. (1998). Overview: Realism, relativism, soial onstrutionism and disourse. In I. Parker (Ed.), Soial onstrutionism, disourse and realism (pp ). London: Sage Publiations. Burr, V. (2003). Soial Construtionism New York: Routledge. Byrne, D. (1971). The attration paradigm. New York: Aademi Press. Carley, K. (1993). Coding Choies for Textual Analysis: A Comparison of Content Analysis and Map Analysis. Soiologial Methodology, 23, Carley, K. & Palmquist, M. (1992). Extrating, Representing, and Analyzing Mental Models. Soial Fores, 70(3), Charles, M. (1998). Europe: Oral business ommuniation. Business Communiation Quarterly, 61, Chernatony, L. & Harris, F. (2000). Developing Corporate Brands Through Considering Internal and External Stakeholders. Corporate Reputation Review, 3(3), Claes, M. T. & Gerritsen, M. (2007). Culturele waarden en ommuniatie in internationaal perspetief [Cultural values and ommuniation from an international perspetive]. Bussum, the Netherlands: Coutinho. Coniam, D. (2004). Conordaning oneself: Construting individual textual profiles. International Journal of Corpus Linguistis, 9(2), Connor, U. (2000). Variation in rhetorial moves in grant proposals of US humanists and sientists. Text, 20(1), Connor, U. (2004). Interultural rhetori researh: beyond texts. Journal of English for Aademi Purposes 3(4), Connor, U. & Mauranen, A. (1999). Linguisti analysis of grant proposals: European Union researh grants. English for Speifi Purposes, 18(1), Connor, U. & Wagner, L. (1998). Language use in grant proposals by nonprofits: Spanish and English. New Diretions for Philanthropi Fundraising, 22, Corbin, J. M., & Strauss, A. (1990). Basis of qualitative researh. grounded theory proedures and tehniques. Newbury Park, California: Sage. Crismore, A. & Farnsworth, R. (1990). Metadisourse in Popular and ProfessionalSiene Disourse. In W. Nash (Ed.), The Writing Sholar. Studies in Aademi Disourse (pp ). London: Sage. Crismore, A., Markkanen, R. & Steffensen, M. (1993). Metadisourse in persuasive writing: a study of texts written bij Amerian and Finnish university students. Written Communiation, 10(1), Crookes, G. (1986). Towards a validated analysis of sientifi text struture. Applied Linguistis, 7, Daelemans, T. (2005). Nederlands in de vergaderzaal is een luxe. Taalshrift: Tijdshrift voor Taal en Taalbeleid. Retrieved July 2005, from

298 Dahl, T. (2004). Textual metadisourse in researh artiles: a marker of national ulture or of aademi disipline? Journal of Pragmatis 36(10), Davenport, E. (2001). Knowledge management issues for online organisations: 'ommunity of pratie' as an exploratory framework. Journal of doumentation, 57(1), David, C. (2001). Mythmaking in annual reports. Journal of Business and Tehnial Communiation, 15(2), Davison, J. (2002). Communiation and antithesis in orporate annual reports: a researh note. Aounting, Auditing & Aountability Journal, 15(4), Devitt, A. J. (1991). Intertextuality in tax aounting: generi, referential and funtional In C. Bazerman & J. Paradis (Eds.), Textual dynamis of the professions: historial and ontemporary studies of writing in professional ommunities (pp ). Madison: University of Wisonsin Press. Dijk, T. A. v. (1993). Priniples of ritial disourse analysis. Disourse & Soiety, 4(2), Duffy, T. M., Higgins, L., Mehlenbaher, B., Cohran, C., Wallae, D., Hill, C., Haugen, D., MCaffrey, M., Burnett, R., Sloane, S. & Smith, S. (1989). Models for the Design of Instrutional Text. Reading Researh Quarterly, 24(4), Els, T. J. M. v. (2000). De Europese Unie, haar Instituties en haar Talen. Enkele taalpolitieke beshouwingen [The European Union, its Institutions and its Languages. Some Language Politial Observations]. Nijmegen: Katholieke Universiteit Nijmegen. European Commission. (2006). Eurobarometer 2006: Europeans and their Languages (No. Speial Eurobarometer 243 / Wave 64.3): Diretorate General for Eduation and Culture. Fairlough, N. (1992a). Disourse and text: linguisti and intertextual analysis within disourse analysis. Disourse & Soiety, 3(2), Fairlough, N. (1992b). Intertextuality in ritial disourse analysis. Linguistis and Eduation, 4, Fairlough, N. (1995). Media Disourse. London: Edward Arnold. Fish, S. (1980). Is There a Text in This Class? Cambridge: Harvard University Press. Flanagan, D. J. & O'Shaughnessy, K. C. (2005). The effet of lay-offs on firm reputation. Journal of management, 31(3), Fleiss, J. L., Levin, B.A. & Paik, M. C. (1981). Statistial Methods for Rates and Proportions. New York: Wiley-Intersiene. Flik, U. (1998). An introdution to qualitative researh. London: Sage Publiations. Flowerdew, J. & Dudley-Evans, T. (2002). Genre analysis of editorial letters to international journal ontributors. Applied Linguistis, 23(4), Flowerdew, L. (2005). An integration of orpus-based and genre-based approahes to text analysis in EAP/ESP: ountering ritiisms against orpus-based methodologies. English for Speifi Purposes, 24(3), Fombrun, C. & Shanley, M. (1990). What's in a name? Reputation building and orporate strategy. Aademy of Management Journal, 33(2), Fombrun, C. J. (1996). Reputation: Realizing Value from the Corporate Image. Boston MA: Harvard Business Shool Press.

299 Forbes. (2003). International: The Global Retrieved May 2004, from Forey, G. (2004). Workplae texts: do they mean the same for teahers and business people? English for Speifi Purposes, 23(4), Fox, R. (2006). Corporations' ideologies: a new subfield of study of orporate ommuniation. Corporate Communiations: An International Journal, 11(4), Freedman, A. & Medway, P. (1994). Loating genre studies: anteedents and prospets. In A. Freedman & P. Medway (Eds.), Genre and the new rhetori (pp. 1-20). London: Taylor and Franis. Fryxell, G. E. & Wang, J. (1994). The Fortune Corporate 'Reputation' Index: Reputation for What? Journal of Management, 20(1), Fuente Sabaté, J. M. d. & Quevedo Puente, E. d. (2003). The Conept and Measurement of Corporate Reputation: An Appliation to Spanish Finanial Intermediaries. Corporate Reputation Review, 5 (4), Fuertes-Olivera, P. A., Velaso-Saristán, M., Arribas-Baño, A. & Samaniego- Fernández, E. (2001). Persuasion and advertising English: meatdisourse in slogans and headlines. Journal of Pragmatis, 33(8), Garretson, J. A. & Burton, S. (2000). Effets of Nutrition Fats Panel Values, Nutrition Claims, and Health Claims on Consumer Attitudes, Pereptions of Disease-Related Risks, and Trust. Journal of Publi Poliy & Marketing, 19(2), Garzone, G. (2004). Annual Company Reports and CEOs' Letters: Disoursal Features and Cultural Markedness. In C. N. Candlin & M. Gotti (Eds.), Interultural Aspets of Speialized Communiation (pp ). Bern: Peter Lang. Garzone, G. (2005). Letters to the shareholders and hairman's statements: textual variability and generi integrity. In P. Gillaerts & M. Gotti (Eds.), Genre variation in business letters (pp ). Bern, et.: Peter Lang. Gee, J. P. (1999). An introdution to disourse analysis: Theory and method. London: Routledge. Gerritsen, M., Korzilius, H., Meurs, F. v. & Gijsbers, I. (2000). English in Duth ommerials: Not understood and not appreiated. Journal of advertising researh, 40(4), Gerritsen, M. & Nikerson, C. (2004). Fat or Fallay? English as an L2 in the Duth Business Context. In C. Candlin & M. Gotti (Eds.), Interultural Aspets of Speialized Communiation (pp ). Bern Peter Lang. Gerritsen, M., Nikerson, C., Van Hooft, A., Van Meurs, F., Nederstigt, U., Starren, M. & Crijns, R. (2007). English in produt advertisements in Belgium, Frane, Germany, the Netherlands and Spain. World Englishes, 26(3), Gillaerts, P. (1996). The adress to the shareholders in annual reports: a genologial approah. In 1996 European Writing Conferenes. Barelona: Institut de Cienies de l'eduaio. Gong, W. & Maddox, L. M. (2003 ). Measuring Web Advertising Effetiveness in China. Journal of Advertising Researh, 43(1), Grabe, W. & Kaplan, R. B. (1996). Theory and pratie of writing: an applied linguisti perspetive. London/New York: Longman. Graddol, D. (2000). The Future of English? A guide to foreasting the popularity of the English language in the 21st entury. London: The British Counil/ The English Company (UK) Ltd.

300 Graddol, D. (2006). English Next. London: British Counil. Graves, O. F., Flesher, D.L. & Jordan, R.E. (1996). Pitures and the bottom line: the television epistemology of U.S. annual reports. Aounting, Organizations and Soiety, 21(1), Greenwald, A. G. (1976). Within-Subjets Designs: To Use or Not To Use? Psyhologial Bulletin, 83(2), Gudykunst, W. B. (2003). Foreword. In W. B. Gudykunst (Ed.), Cross-Cultural and Interultural Communiation (pp. vii-ix). Thousand Oaks: Sage. Guthey, E. & Jakson, B. (2005). CEO portraits and the authentiity paradox. Journal of Management Studies, 42(5), Hallek, G. B. & Connor, U. M. (2006). Rhetorial moves in TESOL onferene proposals. Journal of English for Aademi Purposes, 5, Halliday, M. A. K. (1978). Language as a Soial Semioti. London: Edward Arnold. Hammond, S. A. & Sloum, J. W. (1996). The impat of prior firm finanial performane on subsequent orporate reputation. Journal of Business Ethis, 15(2), Harris, J. (1989). The Idea of Community in the Study of Writing. College Composition and Communiation, 40(1), Harris, S. & Bargiela-Chiappini, F. (1997). The languages of business: Introdution and overview. In S. Harris & F. Bargiela-Chiappini (Eds.), The languages of business. Edinburgh: Edinburgh University Press. Harris, S. & Bargiela-Chiappini, F. (2003). Business as a site of language ontat. Annual Review of Applied Linguistis, 23, Hasan, R. (1989). The struture of a text. In M. A. K. Halliday & R. Hasan (Eds.), Language, ontext, and text: aspets of language in a soial-semioti perspetive. Oxford: Oxford University Press. Henry, A. & Roseberry, R. L. (2001). A narrow-angled orpus analysis of moves and strategies of the genre 'Letter of Appliation'. English for Speifi Purposes, 20(2), Herring, S. C. & Paolillo, J.C. (2006). Gender and genre variation in weblogs. Journal of Soiolinguistis, 10(4), Higgins, J. M. D. (2003). International Conferene: Visual Images in the Press: The Case of the Okinawa G8 Summit. Paper presented at the Knowledge & Disourse: Speulating on Disiplinary Futures, Hong Kong. Hinkel, E. (1994). Native and Nonnative Speakers' Pragmati Interpretations of English Texts. TESOL Quarterly, 28(2), Hoeken, H. & Korzilius, H. (2003). Conduting experiments on ultural aspets of doument design: why and how? Communiations, 28, Hoeken, H. & Renkema, J. (1998). Can orretions repair the damage to a orporate image aused by negative publiity? Corporate reputation review, 2(1), Hoffmann, C. (2000). The Spread of English and the Growth of Multilingualismwith English in Europe. In J. Cenoz & U. Jessner (Eds.), English in Europe. The Aquisition of a Third Language (pp. 1-21). Clevedon: Multilingual Matters. Hofstede, G. (2001). Cultures onsequenes: Comparing values, behaviors, institutions and organizations aross nations (2nd ed.). Thousand Oaks, CA: Sage. Hooghiemstra, R. (2003). The Constrution of Reality: Cultural Differenes in Selfserving Behaviour in Aounting Narratives. Erasmus University Rotterdam, Rotterdam.

301 Hooghiemstra, R. & Manen, J. v. (2004). The independene paradox: (im)possibilities faing non-exeutive diretors in the Netherlands. Corporate Governane: an international review, 12(3), Hukin, T. N. (2002). Textual silene and the disourse of homelessness. Disourse & Soiety, 13(3), Hussein, M. E. (1996). A omparative study of ultural influenes on finanial reporting in the US and the Netherlands. International Journal of Aounting, 31(1), Hyland, K. (1998a). Persuasion and ontext: The pragmatis of aademi metadisourse. Journal of Pragmatis, 30(4), Hyland, K. (1998b). Exploring orporate rhetori: metadisourse in the CEO's letter. Journal of Business Communiation, 35(2), Hyland, K. (1998). Hedging in sientifi researh artiles. Amsterdam: John Benjamins. Hyland, K. (2000). Disiplinary disourse: soial interations in aademi writing. Harlow et.: Longman. Hyland, K. (2007). Applying a gloss: exemplifying and reformulating in aademi disourse. Applied Linguistis, 28(2), Hyland, K. & Tse, P. (2004). Metadisourse in Aademi Writing: A Reappraisal. Applied Linguistis 25(2), Hymes, D. (1974). Foundations in soiolinguistis: an ethnographi approah. Philadelphia: University of Pennsylvania Press. Hynes, G.E. & Bexley, J.B. (2003). Understandability of bank's annual reports. Paper presented at the 2003 Assoiation for Business Communiation Annual Convention, Albuquerque, New Mexio. Hyon, S. (1996). Genre in three traditions: impliations for ESL. TESOL Quarterly, 30(4), IAS Plus. (2005). International Finanial Reporting Standards in Europe: Reent hronology of IFRSs in Europe. Retrieved 31 July 2007, from ING. (2006). Annual General Meeting of Shareholders of ING Groep N.V. (minutes 25 April 2006). Amsterdam: Author. James, C., Sholfield, P. & Ypsiladis, G. (1994). Cross-ultural orrespondene: Letters of appliation. Oasional Paper. Dublin: Centre for Language and Communiation Studies, Trinity College. Jameson, D. A. (2000). Telling the investment story: a narrative analysis of shareholder reports. Journal of Business Communiation, 37(1), Kanoksilapatham, B. (2005). Rhetorial struture of biohemistry researh artiles. English for Speifi Purposes, 24(3), Kilgarriff, A. (2001). Comparing Corpora. International Journal of Corpus Linguistis, 6(1), Klein, J. & Dawar, N. (2004). Corporate soial responsibility and onsumers' attributions and brand evaluations in a produt harm risis. International Journal of Researh in Marketing, 21(3), Kolk, A. (2003). Trends in sustainability reporting by the Fortune Global 250. Business strategy and the environment, 12(5), KPN. (2004). Notulen Algemene Vergadering van Aandeelhouders van Koninklijke KPN N.V. (minutes Annual General Meeting, 15 April 2004). Amsterdam: Author. Kress, G. & Leeuwen, T. v. (2001). Multimodal Disourse: the modes and media of ontemporary ommuniation. London: Arnold.

302 Kress, G. & Leeuwen, T. v. (2006). Reading Images: the grammar of visual design. London, New York: Routledge. Kumpf, E. P. (2000). Visual Metadisourse: Designing the Considerate Text. Tehnial Communiation Quarterly, 9 (4), Lagerwerf, L. & Bossers, E. (2002). Assessing Business Proposals: Genre Conventions and Audiene Response in Doument Design. Journal of Business Communiation, 39(4), Language Poliy Division. (2007). Counil of Europe language eduation poliy. Retrieved 31 July 2007, from Lassen, I. (2006). Is the press release a genre? A study of form and ontent. Disourse Studies, 8(4), Lee, Y.-B. & Myaeng, S. H. (2002). Text genre lassifiation with genre-revealing and subjet-revealing features. Paper presented at the the 25th annual international ACM SIGIR onferene on Researh and development in information retrieval, Tampere, Finland. Lesquillier, S. (2007). Hels Huwelijk (a devil of a marriage) [Eletroni Version]. Transformatieblog, Retrieved 31 July Lindlof, T. R. (1995). Qualitative ommuniation researh methods. Thousand Oaks CA: Sage. Lister, M. & Wells, L. (2001). Seeing beyond belief: Cultural Studies as an approah to analysing the visual. In T. v. Leeuwen & C. Jewitt (Eds.), Handbook of Visual Analysis (pp ). London: Sage. Lok, G. & Lokhart, C. (1998). Genres in an Aademi Writing Class. Hong Kong Journal of Applied Linguistis, 3(2), Loker, K. (1999). Fators in reader responses to negative letters: experimental evidene for hanging what we teah. Journal of Business and Tehnial Communiation, 13(5), Louhiala-Salminen, L. (1997). Investigating the genre of a business fax: a Finnish ase study. Journal of Business Communiation, 34(3), Louhiala-Salminen, L., Charles, M. & Kankaanranta, A. (2005). English as a lingua frana in Nordi orporate mergers: Two ase ompanies. English for Speifi Purposes, 24(4), Luzon, M. J. (2005). Genre analysis in tehnial ommuniation. IEEE Transations on Professional Communiation, 48(3), Mahin, D. & Leeuwen, T. v. (2003). Global shemes and loal disourses in 'Cosmopolitan'. Journal of Soiolinguistis, 7(4), Mahin, D. & Leeuwen, T. v. (2004). Global media: generi homogeneity and disursive diversity. Continuum: Journal of Media & Cultural Studies, 18(1), MaKenzie, S. B. & Lutz, R. J. (1989). An empirial examination of the strutural anteedents of attitude toward the ad in an advertising pretesting ontext. Journal of marketing, 53(2), Maes, F., Ummelen, N. & Hoeken, H. (1996). Instrutieve teksten. Analyse, ontwerp en evaluatie. Bussum: Coutinho. Malavasi, D. (2005). Banks' annual reports: an analysis of the linguisti means used to express evaluation. Paper presented at the Assoiation for Business Communiation 7th European onvention, Copenhagen. Martin, J. R. (1992). English Text: System and Struture. Philadelphia, Amsterdam: Benjamins.

303 Mauranen, A. (1993). Contrastive ESP rhetori: metatext in Finnish English eonomis texts. English for Speifi Purposes, 12(1), Mayes, P. (2005). Linking miro and maro soial struture through genre analysis. Researh on language and soial interation, 38(3), MKay, S. (2002). Teahing English as an International Language. Oxford: Oxford University Press. MKinstry, S. (1996). Designing the annual reports of Burton PLC from 1930 to Aounting, Organizations and Soiety, 21(1), Merer, M. (2004). How do investors assess the redibility of management dislosures? Aounting Horizons, 18(3), Meurs, F. v., Korzilius, H., Planken, B. & Fairley, S. (2007). The effet of English job titles in job advertisements on Duth respondents. World Englishes, 26(2), Meurs, F. v., Korzilius, H.P.L.M. & Hermans, J.J. (2004). The influene of the use of English in Duth job advertisements: An experimental study into the effets on text evaluation, on attitudes towards the organisation and the job, and on omprehension. ESP Aross Cultures, 1, Meurs, F. v., Korzilius, H.P.L.M. & Hollander, A. den. (2006). The use of English in job advertisements on the Duth job site Monsterboard.nl and fators on whih it depends. ESP Aross Cultures, 3, Mik, D. G. (1986). Consumer Researh and Semiotis: Exploring the Morphology of Signs, Symbols, and Signifiane. The Journal of Consumer Researh, 13(2), Miller, C. (1994). Genre as soial ation. In A. Freedman & P. Medway (Eds.), Genre and the New Rhetori (pp ). London Taylor & Franis. Milne, M. J. & Patten, D. M. (2002). Seuring organizational legitimay: an experimental deision ase examining the impat of environmental dislosures. Aounting, Auditing and Aountability Journal, 15(3), Mitry, J. (2000). Semiotis and the Analysis of Film. Bloomington: Indiana University Press. Morden, T. (1995). International ulture and management. Management Deision, 33(2), Morden, T. (1999). Models of National Culture: A Management Review. Cross Cultural Managament, 6(1), Morimoto, M. & Chang, S. (2006). Consumers Attitudes toward Unsoliited Commerial and Postal Diret Mail Marketing Methods: Intrusiveness, Pereived Loss of Control, and Irritation. Journal of Interative Advertising, 7(1), Neuendorf, K. A. (2002). The ontent analysis guidebook. Thousand Oaks, CA: Sage. Nikerson, C. (2000). Playing the orporate language game: an investigation of the genres and disourse strategies in English used by Duth writers working in multinational orporations. Amsterdam NL, Atlanta GA: Editions Rodopi BV. Nikerson, N. & Groot, E. B. d. (2005). Dear Shareholder, Dear Stokholder, Dear Stakeholder: The Business Letter Genre in the Annual General Report In P. Gillaerts & M. Gotti (Eds.), Genre Variation in Business Letters (pp ). Bern: Peter Lang.

304 Nieuwboer, I., Maes, A.A. & Swanepoel, P. (2005). The persuasive power of peer guides in Web sites that promote HIV/AIDS voluntary ounselling and testing. Paper presented at the International Professional Communiation Conferene, IPCC Nöth, W. (1975). Semiotik. Eine Einführung mit Beispielen für Reklameanalysen. Tübingen Niemeyer. O'Donovan, G. (2002). Environmental dislosures in the annual report: Extending the appliability and preditive power of legitimay theory. Aounting, Auditing & Aountability Journal, 15(3), Orlikowski, W. J. & Yates, J. (1994). Genre repertoire: the struturing of ommuniative praties in organizations. Administrative Siene Quarterly, 4, Orth, U. R. & Holanova, D. (2004). Consumer response to sex role portrayals in advertisements: effets of inongruity and prejudies on emotions and attitudes. Journal of Advertising, 32(4), Osborne, J. D., Stubbart, C.I. & Ramaprasad, A. (2001). Strategi groups and ompetitive enatment: a study of dynami relationships between mental models and performane. Strategi Management Journal, 22(5), Paek, H. J. (2005). Understanding Celebrity Endorsers in Cross-ultural Contexts: A Content Analysis of South Korean and US Newspaper Advertising. Asian Journal of Communiation, 15(2), Paltridge, B. (1994). Genre analysis and the identifiation of textual boundaries. Applied Linguistis, 15(3), Parker, R. H. (2001). Aounting in Europe: European languages of aount. The European Aounting Review, 10(1), Paul, D., Charney, D. & Kendall, A. (2001). Moving beyond the moment: reeption studies in the rhetori of siene. Journal of Business and Tehnial Communiation 15(3), Pava, M. L. & Krausz, J. (1996). The assoiation between orporate soialresponsibility and finanial performane: The paradox of soial ost. Journal of Business Ethis, 15(3), Pfanner, E. (2006). Dual itizens, one bottom line: in a hanging Europe, the British- Duth model yields new benefits - and drawbaks. International Herald Tribune. Retrieved February 2007 from Philips. (2003). Algemene Vergadering van aandeelhouders van Koninklijke Philips Eletronis NV (minutes Annual General Meeting 27 Marh 2003). Amsterdam. Philips. (2004). Press release: Philips realigns around new brand promise. Retrieved 31 July 2007, from Phillipson, R. (1998). Globalizing English: Are linguisti human rights an alternative to linguisti imperialism? Language Sienes, 20(1), Phillipson, R. (2003). English-only Europe? Challenging language poliy. London: Routledge. Piller, I. (2001). Identity onstrutions in multilingual advertising. Language in Soiety, 30, Planken, B. (2005). Managing rapport in lingua frana sales negotiations: A omparison of professional and aspiring negotiators. English for Speifi Purposes, 24(4),

305 Planken, B., Hooft, A. van & Korzilius, H. (2004). Promoting Interultural Communiative Competene through Foreign Language Courses. Business Communiation Quarterly, 67, Pogner, K. H. (1999). Disourse ommunity, ulture, and interation: on writing by onsulting engineers. In F. Bargiela-Chiappini & C. Nikerson (Eds.), Writing business: genres, media and disourses (pp ). New York, London: Longman. Pogner, K. H. (2005). Disourse Communities and Communities of Pratie On the soial ontext of text and knowledge prodution in the workplae. Paper presented at the 21st EGOS Colloquium, Berlin. Ponini, G. (2002). Investigating disourse at business meetings with multiultural partiipation. International review of Applied Linguistis 40, Ponini, G. (2004). Disursive Strategies in Multiultural Business Meetings. Bern: Peter Lang. Ponini, G. (2005). Construting an International Event in the Wine Industry: An Investigation of s in English and Italian. In P. Gillaerts & M. Gotti (Eds.), Genre Variation in Business Letters (pp ). Bern, et.: Peter Lang. Preston, A. M., Wright, C. & Young, J.J. (1996). Imag[in]ing annual reports. Aounting, Organizations and Soiety, 21(1), Preston, A. M. & Young, J. J. (2000). Construting the global orporation and orporate onstrutions of the global: a piture essay. Aounting, Organizations and Soiety, 25(4-5), Renkema, J. (2001). Underover researh into text quality as a tool for ommuniation management: the ase of the Duth tax department. In D. J. R. Neutelings (Ed.), Reading and Writing Publi Douments (pp ). Amsterdam: Benjamins. Riks, D. A. (1999). Blunders in international business. Oxford UK: Blakwell. Riel, C. B. M. v. & Fombrun, C. J. (2007). Essentials of Corporate Communiation: implementing praties for effetive reputation management. London, New York: Routledge. Rietveld, T. & Hout, R. v. (1993). Statistial tehniques for the study of language and language behavior. Berlin et.: Mouton de Gruyter. Roberts, P. W. & Dowling, G. R. (2002). Corporate reputation and sustained superior finanial performane. Strategi Management Journal, 23(12), Rogers, P. (2001). A ommuniative perspetive for investor relations: tapping the interfae between orporate onerns and user needs. Unpublished Working paper. University of Mihigan Business Shool. Rogers, P. S. (2000). CEO Presentations in Conjuntion with Earnings Announements. Management Communiation Quarterly, 13(3), Rogerson-Revell, P. (2007). Using English for International Business: A European ase study. English for Speifi Purposes, 26(1), Rowley-Jolivet, E. (2002). Visual disourse in sientifi onferene papers: A genre-based study. English for Speifi Purposes, 21( 1), Rowley-Jolivet, E. (2004). Different visions, different visuals: a soial semioti analysis of field-speifi visual omposition in sientifi onferene presentations. Visual Communiation, 3(2), Rowley-Jolivet, E. & Carter-Thomas, S. (2005). The rhetori of onferene presentation introdutions: ontext, argument and interation. International Journal of Applied Linguistis, 15(1),

306 Rubin, D. L. (1988). Four dimensions of soial onstrution in written ommuniation. In B. A. Rafoth & D. L. Rubin (Eds.), The soial onstrution of written ommuniation. Norwood, NJ: Ablex Publishing. Rutherford, B. (2003). Obfusation, Textual Complexity and the Role of Regulated Narrative Aounting Dislosure in Corporate Governane. Journal of Management and Governane 7(2), Rutherford, B. A. (2005). Genre analysis of orporate annual report narratives: a orpus linguistis-based approah. Journal of Business Communiation, 42(4), Ryan, G. W. & Bernard, H. R. (2000). Data management and analysis methods. In Y. Linoln (Ed.), Handbook of Qualitative Researh (pp ). Thousand Oaks: Sage. Ryan, G. W. & Bernard, H. R. (2003). Tehniques to identify themes. Field Methods, 15(1), Santema, S., Hoekert, M. & Rijt, J. van de, Oijen, A. van. (2005). Strategy dislosure in annual reports aross Europe: a study on differenes between five ountries. European Business Review, 17(4), Santema, S. & Rijt, J. v. d. (2001). Strategy dislosure in Duth annual reports. European Management Journal 19(1), Shmalensee, D. H. (2003). The perfet sale. Marketing Researh, 15(3), Sollon, R., Bhatia, V., Li, D. & Yung, V.. (1999). Blurred genres and fuzzy identities in Hong Kong publi disourse: foundational ethnographi issues in the study of reading. Applied Linguistis, 20(1), Sott, M. (2006). WordSmith Tools. Oxford: Oxford University Press. Seidlhofer, B. (2003). A onept of International English and related issues: from 'real English' to 'realisti English'? Strasbourg: Language Poliy Division, DG IV Diretorate of Shool, Out-of-Shool and Higher Eduation, Counil of Europe. Seidlhofer, B., Breiteneder, A. & Pitzl, M. L. (2006). English as a lingua frana in Europe: hallenges for applied linguistis. Annual Review of Applied Linguistis, 26, Selinker, L., Lakstrom, J. & Trimble, L. (1973). Tehnial rhetorial priniples and grammatial hoie. TESOL Quarterly, 7(2), Shadid, W. (2000). Communiatieve ompetentie in interulturele ommuniatie. P&M, 90, Shaw, P., Gillaerts, P., Jaobs, E., Palermo, O., Shinohara, M. & Verkens, J.P. (2004). Genres aross ultures: types of aeptability variation. World Englishes, 23(3), Sims, B. R. & Guie, S. (1992). Differenes between business letters from native and non-native speakers of English. Journal of Business Communiation, 29, Smelik, A. (1999). Effetief beeldvormen: theorie, analyse en praktijk van beeldvormingsproessen. Assen: Van Gorum & Comp. B.V. Smith, M. & Taffler, R.J. (2000). The Chairman's statement: a ontent analysis of disretionary narrative dislosures. Aounting, Auditing & Aountability Journal, 13(5), Smith, R. (1997). English in European Spanish. English Today, 13, Smith, R. E. (1993). Integrating information from advertising and trial: proesses and effets on onsumer response to produt information. Journal of marketing researh, 30,

307 Solbjørg Skulstad, A. (1996). Rhetorial organization in hairmen's statements. International Journal of Applied Linguistis, 6(1), Solbjørg Skulstad, A. (2005). Competing roles: student teahers using asynhronous forums. International Journal of Applied Linguistis, 15(3), Solbjørg Skulstad, A. (2005). The use of metadisourse in introdutory setions of a new genre. International Journal of Applied Linguistis, 15(1), Spinuzzi, C. (2004). Desribing Assemblages: Genre Sets, Systems, Repertoires, and Eologies. Austin: The University of Texas. Stanton, P. Stanton, J. & Pires, G. (2004). Impressions of an annual report: an experimental study. Corporate Communiations, 9(1), Stittle, J. (2002). UK Corporate Ethial Reporting A Failure to Inform: Some Evidene from Company Annual Reports. Business and Soiety Review, 107(3), Suhan, J. & Dulek, R. (1998). From Text to Context: An Open Systems Approah to Researh in Written Business Communiation. Journal of Business Communiation, 35(1), Swales, J. M. (1990). Genre analysis: English in aademi and researh settings. Cambridge, New York, et.: Cambridge University Press. Swales, J. M. (2000). Languages for speifi purposes. Annual Review of Applied Linguistis, 20, Swales, J. M. & Rogers, P. S. (1995). Disourse and the projetion of orporate ulture: the mission statement. Disourse & Soiety, 6, Taavitsainen, I. & Pahta, P. (2003). English in Finland: globalisation, language awareness and questions of identity. English Today, 19(4), Tardy, C. (2003). A Genre System View of the Funding of Aademi Researh. Written Communiation, 20(1), Thomas, J. (1997). Disourse in the marketplae: the making of meaning in annual reports. Journal of Business Communiation, 34(1), Thornsteinson, T. J. & Highhouse, S. (2003). Effets of goal framing in job advertisements on organizational attrativeness. Journal of Applied Soial Psyhology, 33(11), Thurlow, C. & Jaworksi, A. (2003). Communiating a global reah: Inflight magazines as a globalizing genre in tourism. Journal of Soiolinguistis, 7(4), Titsher, S., Meyer, M., Wodak, R. & Vetter, E.. (2000). Methods of Text and Disourse Analysis. London: Sage Publiations. TNT. (2007). TNT Post onfirms Centria ontrat win. Retrieved 31 July 2007, from 07_TNT_Post_onfirms_Centria_ontrat_win.html Truhot, C. (2002). Key aspets of the use of English in Europe. Strasbourg: Language Poliy Division, DG IV Diretorate of Shool, Out-of-Shool and Higher Eduation, Counil of Europe. Ulijn, J., O'Hair, D., Weggeman, M., Ledlow, G. & Hall, H.T. (2000). Innovation, Corporate Strategy, and Cultural Context: What is the mission for international business ommuniation. Journal of Business Communiation, 37(3), Ulijn, J. M. & StAmant, K. (2000). Mutual interultural pereption: How does it affet tehnial ommuniation? Some Data from China, the Netherlands, Germany, Frane, and Italy. Tehnial Communiation, 47(2), Upton, T. A. (2002). Understanding Diret Mail Letters as a genre. International Journal of Corpus Linguistis, 7(1),

308 Upton, T. A. & Connor, U. (2001). Using Computerized Analysis to Investigate the Text Linguisti Disourse Moves of a Genre. English for Speifi Purposes, 20(4), Valero-Garés, C. (1996). Contrastive ESP rhetori: Metatext in Spanish-English eonomis texts. English for Speifi Purposes, 15 (4), Vande Koppele, W. J. (1985). Some exploratory disourse on metadisourse. College Composition and Communiation, 36, Vandermeeren, S. (1999). English as a lingua frana in orporate writing. In F. Bargiela-Chiappini & C. Nikerson (Eds.), Writing business: Genres, media and disourses (pp ). London and New York: Longman. Van der Moolen. (2006). Notulen van de Algemene Vergadering van Aandeelhouders van Van der Moolen Holding N.V. (minutes Annual General Meeting 5 April 2006). Amsterdam: Author. Varner, I. I. (2000). The Theoretial Foundation for interuitural Business Communiation: A Coneptual Model. TheJournal of Business Communiation, 37(1), Varner, I. I. (2001). Teahing Interultural Management Communiation: Where Are we? Where Do We Go? Business Communiation Quarterly, 64(1), Vaughan, M. W. & Dillon, A. (2006). Why struture and genre matter for users of digital information: a longitudinal experiment with readers of a web-based newspaper. International Journal of Human-Computer Studies, 64, Vergaro, C. (2002). Dear Sirs, what would you do if you were in our position? Disourse strategies in Italian and English money-hasing letters. Journal of Pragmatis, 34(9), Vergaro, C. (2005). 'Dear sirs, I hope you will find this information useful': disourse strategies in Italian and English For-Your-Information (FYI) letters. Disourse studies, 7(1), Wartik, S. L. (1992). The Relationship between Intense Media Exposure and Change in Corporate Reputation. Business & Soiety, 31, Wegner, D. (2004). The Collaborative Constrution of a Management Report in a Muniipal Community of Pratie. Journal of Business and Tehnial Communiation, 18(4), Wendt, M. (2002). Context, Culture and Constrution: Researh Impliations of Theory Formation in Foreign Language Methodology. Language, Culture and Curriulum, 15(3), Wenger, E. & Snyder, W. M. (2000). Communities of pratie: The organizational frontier. Harvard Business Review, 78(1), Wester, F. & Peters, V. (2004). Kwalitatieve Analyse. Uitgangspunten en Proedures (Qualitative analysis. Starting-points and proedures) Bussum: Coutinho. Wijk, C. v. (2000). Toetsende statistiek: basistehnieken. Een praktijkgerihte inleiding voor onderzoekers van taal, gedrag en ommuniatie. Bussum: Coutinho. Wollen, P. (1998). Signs and Meaning in the Cinema London: Seker and Warburg. Wolters-Kluwer. (2006). General Meeting of Shareholders of Wolters Kluwer N.V. (minutes April 26, 2006). Amsterdam. Yates, J. & Orlikowski, W. (1992). Genres of organizational ommuniation: a struturational approah to studying ommuniation and media. Aademy of Management Review, 17,

309 Yates, J. & Orlikowski, W. (2002). Genre Systems: Struturing Interation through Communiative Norms. Journal of Business Communiation, 39(1), Yli-Jokipii, H. M. (1998). Europe: Business Writing. Business Communiation Quarterly, 61, Yunik, S. (1997). Genres, registers and soiolinguistis. World Englishes, 16( ). Zanoli, C. (2004, 20 April 2004). Protest tegen afshaffen Nederlands jaarverslag Aegon (Protests against AEGON s abolition of Duth annual report). Zibb Finane. Retrieved 23 September 2007, from Zhu, Y. (2000). Strutural Moves Refleted in English and Chinese Sales Letters. Disourse Studies, 2(4), Zhu, Y. (2005). Written ommuniation aross ultures: a soioognitive perspetive on business genres. Amsterdam/Philadelphia: John Benjamins Publising Company.

310 Referenes annual reports Annual reports Duth orporations ABN AMRO. (2004). Annual Report Amsterdam: Author. AEGON. (2004). Annual Report Den Haag: Author. AHOLD. (2004). Annual Report Zaandam: Author. Akzo Nobel. (2004). Annual Report Arnhem: Author. ASM International. (2004). Annual Report Bilthoven: Author. ASML. (2004). Annual Report Veldhoven: Author. Getronis. (2004). Annual Report Amsterdam: Author. Hagemeijer. (2004). Annual Report Naarden: Author. Hunter Douglas. (2004). Annual Report Rotterdam: Author. IHC Caland. (2004). Annual Report Shiedam: Author. ING. (2004). Annual Report Amsterdam: Author. KPN. (2004). Annual Review The Hague: Author. KPN. (2004). Annual report and Form 20-F The Hague: Author. Numio. (2004). Annual Report Zoetermeer: Author. Philips. (2004). Management Report Eindhoven: Author. Philips. (2004). Finanial Statements and Analysis Eindhoven: Author. Randstad. (2004). Annual Report Diemen: Author. Rodamo. (2004). Annual Report 2003 Book I. Rotterdam: Author. TPG. (2004). Annual Overview Amsterdam: Author. TPG. (2004). Annual report 2003/Form 20-F. Amsterdam: Author. Van der Moolen. (2004). Annual Report Amsterdam: Author. Vedior. (2004). Annual Report Amsterdam: Author. Versatel. (2004). Annual Report Amsterdam: Author. VNU. (2004). Annual Report Haarlem: Author. Wolters Kluwer. (2004). Annual Report Amsterdam: Author. Annual reports British orporations Aviva. (2004) Annual Review. London: Author. Aviva. (2004) Annual Report and Aounts. London: Author. Balfour Beatty. (2004). Annual Report and Aounts London: Author. Boots. (2004). Annual Review Nottingham: Author. Boots. (2004). Annual Report and Aounts Nottingham: Author. British Land. (2004). Annual Report and Aounts London: Author. BT. (2004). Annual Review London: Author. BT. (2004). Annual Report and Form 20-F London: Author. Exel. (2004). Annual Report and Aounts Berkshire: Author. Hays. (2004). Annual Report and Aounts London: Author. HSBC. (2004) Annual Review. London: Author. HSBC. (2004) Annual Report and Aounts. London: Author. ICI. (2003). Annual Review London: Author. ICI. (2004). Annual Report and Aounts London: Author. Inhape. (2004). Annual Report and Aounts London: Author.

311 Invensys. (2004). Annual Report and Aounts London: Author. Kingfisher. (2004). Annual Review and Summary Finanial Statements London: Author. Kingfisher. (2004). Annual Report and Aounts London: Author. Maroni. (2004). Annual Report and Aounts London: Author. Old Mutual. (2004). Annual Report and Aounts London: Author. Pearson. (2004). Annual Report & Aounts London: Author. Rentokil Initial. (2004). Annual Report East Grinstead: Author. Sage Group. (2004). Annual Report and Aounts Newastle upon Tyne: Author. Smiths Group. (2004). Annual Review London: Author. Smiths Group. (2004). Report and Aounts London: Author. Somerfield. (2004). Annual Report and Aounts 2003/04. Bristol: Author. Vodafone. (2004). Annual Review Berkshire: Author. Vodafone. (2004). Annual Report for the year ended 31 Marh Berkshire: Author. Wolseley. (2004). Report and Aounts Reading: Author. WPP. (2004) Annual Report and Aounts. London: Author.

312

313 Appendix I: text ontent management statements List of themes (ounts) in running written texts of Duth-English and British- English management statements. Du-CEO n=20 Br-CEO n=18 Br-Chairman =16 atboard atsuper agm awrank balane boarddet apital ashflow hannel laims ommunit omparis omplod onssegm onssupp ontat oopnetw orpgov ostred sr urreny debt dividend earnshar eduatio environm evaluati exepite faility finhighl fininstr finpoli finresul finsegm fintarge foundsha funstaf futsegm future geomarke goodwill govode heasafe history hrm identity idmarket idsegm ifrs 1 0 0

314 Table Appendix I ontinued. Du-CEO n=20 Br-CEO n=18 Br-Chairman =16 inentiv inome indusseg industry intontr intelpro interest invdiv invrelat itrdsegm itresdev liquidit markdist marksegm marposit memboard mission numbempl opost ophighli opresult opsegm partner perfgrap positseg prinipl prodmat purhase refurr regulati remunera reportin roi roleboar sales shafund shaprie shareon shareval sperigh strategy stratseg strutur supwork taxation training usp workond 0 2 0

315 Appendix II: text ontent orporate profiles List of themes (ounts) in running written texts of Duth-English and British- English orporate profiles. Du-profile n=21 Br-profile n=20 awrank 1 1 busimpli 0 1 hannel 3 1 laims 1 1 onssegm 2 2 onssupp 6 9 onsult 0 1 ontat 1 3 oopnetw 1 1 orpgov 0 1 ostred 0 1 sr 2 3 environm 0 1 exepite 0 1 faility 2 1 finhighl 1 7 fininstr 0 2 finpoli 0 1 finresul 2 2 finsegm 3 2 fintarge 0 1 forwlook 1 0 futsegm 1 1 future 3 5 geomarke 3 5 heasafe 0 1 history 5 3 hrm 2 2 identity idmarket 5 10 idsegm indusseg 0 2 industry 3 7 instrris 0 1 intelpro 1 1 invdiv 0 3 itresdev 1 1 markdist 7 5 marksegm 4 2 marposit 4 5 mission numbempl 0 6 opompan 2 2 ophighli 1 4 opresult 3 7 opsegm 1 3 partner 1 1 perfgrap 0 1 positseg 2 4 prinipl 1 2 prodmat 3 1 regulati 1 1 reportin 2 2

316 Table Appendix II ontinued. Du-profile n=21 Br-profile n=20 shareon 1 1 sharehol 2 0 sharelis 4 3 strategy stratseg 4 5 strutur 4 0 usp workond 0 1

317 Appendix III: text ontent operational reviews List of themes (ounts) in running written texts of Duth-English and British- English operational reviews. Du-OperRev n=22 Br-OperRev n=20 apoli 2 0 afterdat 1 0 awrank 6 11 baklog 1 0 balane 3 0 boarddet 2 4 apital 2 3 ashflow 0 2 hannel 1 1 laims 2 1 ommunit 3 1 omparis 0 1 omplod 1 0 onssegm 9 11 onssupp 3 4 onsult 1 2 oopnetw 6 1 orpgov 2 0 ostred 11 3 ountris 1 0 sr 4 4 urreny 4 4 urrrisk 1 0 debt 1 1 dividend 0 1 earnshar 0 1 eduatio 0 1 environm 1 3 evaluati 2 2 exepite 0 2 faility 7 5 finhighl 6 11 fininstr 1 0 finpoli 1 0 finresul 6 3 finsegm fintarge 2 1 forwlook 1 0 funstaf 2 0 futsegm future 9 7 geomarke 4 3 goodwill 4 0 heasafe 1 2 history 2 1 hrm 4 4 identity 3 2 idmarket 14 5 idsegm ifrs 1 1 inentiv 0 1 inome 10 1 indexwor 1 0

318 Table Appendix III ontinued. Du-OperRev n=22 Br-OperRev n=20 indusseg industry 8 6 insuran 1 0 intontr 3 0 intelpro 2 0 interest 0 1 invdiv itrdsegm 10 5 itresdev 3 4 legalris 1 0 liqrisk 1 0 markdist 7 5 marksegm marposit 3 3 numbempl 3 2 opompan 3 1 opost 9 3 operrisk 1 0 ophighli 8 6 opresult 9 5 opsegm partner 5 8 pensbene 0 1 pension 2 0 positseg prinipl 2 3 prodmat 7 5 regulati 5 2 relpartt 1 0 reportin 2 4 riskfat 1 0 riskman 2 0 roi 0 2 sales 12 9 shaapit 1 0 shareon 6 6 strategy 9 5 stratseg strutur 9 4 taxation 2 1 training 0 1 usp workond 0 1

319 Appendix IV: text struture management statements List of moves and strategies (ounts) in running written texts of Duth-English and British-English management statements. Du-CEO n=20 Br-CEO n=18 Br-Chairman =16 MOVES 1. Id.ompany Mission Corp.strategy Detail produt Target market Customer need Corp.ondut Corp.situation Leadership Finan.perform Oper.perform Futur.perform Credentials Sympathy Address.reader Read.orient STRATEGIES 1. General profile Mission/values Strategi fous Strateg.onseq Desr.produt Value produt Desript.market Desr.ustomer Custom.need Involvement Eonom.sene Perf.ontext Board details Finan.perf.det Oper.perf.det Fut.prospets Fut.objetives Confidene Reader needs Corp.strengths Appreiation Compassion Self-intro Salutation Polite ending Soli.response Desript.info Evaluat.info Emph.str.fous History 1 1 0

320 Appendix V: text struture orporate profiles List of moves and strategies (ounts) in running written texts of Duth-English and British-English orporate profiles. Du-profile n=21 Br-profile n=20 MOVES 1. Id.ompany Mission Corp.strategy Detail produt Target market Customer need Corp.ondut Corp.situation Leadership Fin.perform Oper.perform Futur.perform Credentials Read.orient Share owner 2 0 STRATEGIES 1.General profile Mission/values Strategi fous Strateg.onseq Desr.produt Value produt Desrip.market Desr.ustom Custom.need Involvement Eon.sene Perf.ontext Board details Fin.perf.det Oper.perf.det Fut.prospets Fut.objetives Confidene Reader needs Corp.strength Appreiation Desript.info Evaluat.info Emp.str.fous Shareholders Communiat History Faility 1 0

321 Appendix VI: text struture operational reviews List of moves and strategies (ounts) in running written texts of Duth-English and British-English operational reviews. Du-OperRev n=22 Br-OperRev n=20 MOVES 1. Id.ompany Mission Corp.strategy Detail produt Target market Customer need Corp.ondut Corp.situation Leadership Fin.perform Oper.perform Futur.perform Credentials Sympathy Addres.reader Read.orient Share owner. 0 1 STRATEGIES 1.General profile Mission/values Strategi fous Strateg.onseq Desr.produt Value produt Desrip.market Desr.ustom Custom.need Involvement Eon.sene Perf.ontext Board details Fin.perf.det Oper.perf.det Fut.prospets Fut.objetives Confidene Reader needs Corp.strength Appreiation Self-intro Polite ending Soli.respons Desript.info Evaluat.info Emp.str.fous Shareholders Communiat History Faility 7 4

322 Appendix VII: lexio-grammar text management statements List of metadisursive items (ounts) in running written texts of Duth-English and British-English management statements. Du-CEO n=20 Br-CEO n=18 Br-Chairman =16 Interative Transition Conjuntion Sentene-initial Frame marker Text stager Illoutionary at Endoph. marker Evidential Code gloss Interational: Hedge Passive Mitigation adj/adv Mitigation noun (modal) verb Booster Verb of ommitm Emphati adj/adv Emphati noun Attitude Evaluat. adj/adv Evaluat. noun Evaluat. verb Sentene-intial Engage.marker Self-mention Corp.name pers.sing.pron pers.plur. pron Funtion Management Group-rel.nouns

323 Appendix VIII: lexio-grammar text orporate profiles List of metadisursive items (ounts) in running written texts of Duth-English and British-English orporate profiles. Du-profile n=21 Br-profile n=20 Interative Transition Conjuntion Sentene-initial 10 5 Frame marker Text stager 2 2 Illoutionary at 8 0 Endoph. marker 7 5 Evidential 3 2 Code gloss Interational: Hedge Passive Mitigation adj/adv Booster Verb of ommitm Emphati adj/adv Attitude Evaluat. adj/adv Evaluat. noun 8 6 Evaluat. verb 5 1 Engage.marker 1 4 Self-mention Corp.name pers.plur. pron Group-rel.nouns 19 15

324 Appendix IX: lexio-grammar text operational reviews List of metadisursive items (ounts) in running written texts of Duth-English and British-English operational reviews. Du-OperRev n=22 Br-OperRev n=20 Interative Transition Conjuntion Sentene-initial Frame marker Text stager 11 4 Illoutionary at Endoph. marker Evidential Reognition result Quotation 3 2 Code gloss Interational: Hedge Passive Mitigation adj/adv Mitigation noun 2 2 (modal) verb Booster Verb of ommitm Emphati adj/adv Emphati noun 11 9 Attitude Evaluat. adj/adv Evaluat. noun Evaluat. verb 6 6 Engage.marker 9 5 Self-mention Corp.name pers.plur. pron Funtion 2 6 Management 2 5 Group-rel.nouns 18 19

325 Appendix X: questionnaire experiment Version 1 of the questionnaire, distributed among UK-based finanial professionals. Setion A: Setion B: Setion C: Setion D: Setion E: Setion F: Setion G: British-English Letter to the Shareholders + questions orporate image, persuasiveness of the text and investment intentions. Typial British-English disourse features are highlighted (based on results setions 5.3, 6.3 and 7.3). Duth-English Letter to the Shareholders + questions orporate image, persuasiveness of the text and investment intentions. Typial Duth-English disourse features are highlighted (based on results setions 5.3, 6.3 and 7.3). General preferene for the omplete Duth-English Letter or the omplete British-English Letter. Use of annual report and the Letter to the Shareholder in general (based on results setions 4.4.1, and 4.4.4). Statements about the ontent and struture of the Letter to the Shareholders (based on results setions 5.3, 6.3 and 7.3). Preferene for strutural elements and metadisursive items in the written text of the Letter to the Shareholders (based on results setions 5.3, 6.3 and 7.3). Preferene for themati, strutural and metadisursive elements in the photographs of the Letter to the Shareholders (based on results setions 5.3, 6.3 and 7.3).

326 Setion A: Letter CorpA Please read the text below and answer the questions on the next page.

327 Questions Letter CorpA Please mark the answer you find most appropriate. 1. After reading this Letter to the Shareholders, how would you rate the ompany publishing the text on eah of the following attributes: Poor Exellent a) Quality of management b) Quality of produts/servies ) Long-term investment value d) Innovativeness e) Finanial soundness f) Ability to attrat/develop/keep talented people g) Community and environmental responsibility h) Use of orporate assets 2. After reading this Letter to the Shareholders, how would you rate the text on eah of the following attributes: a) Good Bad b) Unpleasant Pleasant ) Diffiult Easy d) Simple Complex e) Favourable Unfavourable f) Unonvining Convining g) Clear Unlear h) Poorly organized Well organized i) Believable Unbelievable j) Unbiased Biased k) Logial Illogial l) Conise Lengthy 3. Based on the information in the above Letter to the Shareholders, zero likelihood ertain a) how likely would you be to invest money in CorpA? b) how likely would you be to reommend CorpA to a professional olleague?

328 Setion B: Letter CorpB Please read the text below and answer the questions on the next page.

329 Questions Letter CorpB Please mark the answer you find most appropriate. 1. After reading this Letter to the Shareholders, how would you rate the ompany publishing the text on eah of the following attributes: Poor Exellent a) Quality of management b) Quality of produts/servies ) Long-term investment value d) Innovativeness e) Finanial soundness f) Ability to attrat/develop/keep talented people g) Community and environmental responsibility h) Use of orporate assets 2. After reading this Letter to the Shareholders, how would you rate the text on eah of the following attributes: a) Good Bad b) Unpleasant Pleasant ) Diffiult Easy d) Simple Complex e) Favourable Unfavourable f) Unonvining Convining g) Clear Unlear h) Poorly organized Well organized i) Believable Unbelievable j) Unbiased Biased k) Logial Illogial l) Conise Lengthy 3. Based on the information in the above Letter to the Shareholders, zero likelihood ertain a) how likely would you be to invest money in CorpB? b) how likely would you be to reommend CorpB to a professional olleague?

330 Setion C: Text preferene Please indiate to what degree you prefer either the Letter to the Shareholders of CorpA (Setion A) or the Letter of CorpB (Setion B). The middle irle indiates no preferene. We would like you to briefly motivate your answer. Letter CorpA. Letter CorpB. Motivation answer: Setion D: General information Please mark the answer you prefer or provide your own answer. 1. In your profession, how do you use the annual report? (more than one answer possible) To get informed about a ompany s: identity. strategy. operational & finanial results. To evaluate a ompany s: effiieny. profitability. future prospets. To failitate: the deision-making proess in investments. To diretly ativate: investment behaviour. Other: 2. How important is a orporation s annual report in your professional life? Very important Not important at all 3. When you read the annual report, how often do you read the Letter to the Shareholders? Almost always Hardly ever

331 4. Why do you read the Letter to the Shareholders? (more than one answer possible) To get a quik informative view on the ompany s general performane. To evaluate the general performane of the ompany. To get an informative impression of the ompany s personality or harater. To evaluate the ompany s personality or harater. Other: 5. Who do you think writes (i.e. signs) the Letter to the Shareholders? Setion E: General ontent Letter to the Shareholders Please indiate to what extent you agree or disagree with eah statement. The Letter to the Shareholders should inlude details about Absolutely agree Absolutely disagree the management situation (e.g. board hanges). orporate soial responsibility poliies and ativities. the dividend paid to shareholders. the long-term orporate strategy. short-term future objetives (for the next finanial year).

332 Setion F: Struture Letter to the Shareholders You will now be presented with three pairs of exerpts drawn from different Letters to the Shareholders. For eah pair, please indiate to what degree you prefer either one of the two exerpts. The middle irle indiates no preferene. Our business Your ompany ontinues to make progress by innovating in our traditional markets and by growing revenues in all the new wave markets in whih we operate. More information about our individual divisions is set out in the Operating review. Returns to shareholders We are reommending a full year dividend of 8.5 pene per share. The full year dividend is 31% up on last year, and over four times higher than two years ago. Strategi progress and outlook The strong growth in turnover and the strong loyalty of our ustomers show that our strategy is working. Your ompany ontinues to make progress by innovating in our traditional markets and by growing revenues in all the new wave markets in whih we operate. General insurane sales were higher than in 2002 and our asset management business also produed a good result. Our pension business benefited from a strong ontribution from reent aquisitions. We are reommending a full year dividend of 8.5 pene per share. The full year dividend is 31% up on last year, and over four times higher than two years ago. The strong growth in turnover and the strong loyalty of our ustomers show that our strategy is working. First, signifiant progress was made during the year in expanding our multi-hannel distribution systems in the Unites States and Taiwan. Seondly, we ompleted the aquisition and integration of a number of independent finanial advisor firms in Spain. Finally, the partnership with Savings, one of the largest savings banks in India, was an important development in Overall, we will ontinue to develop partnerships in those markets where we expet growth During the year we made signifiant progress in expanding our multi-hannel distribution systems in the Unites States and Taiwan. We ompleted the aquisition and integration of a number of independent finanial advisor firms in Spain. In 2003, another important development was our partnership with Savings, one of the largest savings banks in India. We will ontinue to develop partnerships in those markets where we expet growth. Overall, 2003 was a good year for CompanyX. Our reord results show the diversity of our business against a bakdrop of improvement in most of the world s major eonomies. Our performane also benefited from a strong ontribution from reent aquisitions and our ontinuing investment in new produts and servies D ear shareholders, Overall, 2003 was a good year for CompanyX. Our reord results show the diversity of our business against a bakdrop of improvement in most of the world s major eonomies. Our performane also benefited from a strong ontribution from reent aquisitions and our ontinuing investment in new produts and servies

333 Setion G: Photographs You will now be presented with three pairs of photographs. For eah pair, please indiate to what degree you prefer either one of the two pitures as an illustration to the Letter to the Shareholders in general. The middle irle indiates no preferene.

334 Setion H: Professional bakground Please desribe or mark your answer. 1. Could you briefly desribe your professional funtion? 2. How many years of experiene have you had as a finanial professional?. years. 3. You are Male. Female. 4. Did you grow up in the United Kingdom? Yes. No. 5. Did you grow up speaking more than one language? Yes, namely: No. 6. What is your age? Please use this spae if you wish to add any omments: If you wish to reeive a summary of the researh results, please fill in your address: Thank you very muh for your partiipation!

335 Samenvatting Aanleiding Het belang van Engels als internationale bedrijfstaal is in Europa vooral toegenomen in de jaren negentig van de 20 ste eeuw. Vanaf die periode werd het Engels veelvuldig gebruikt bij de eonomishe integratie van Europese landen en verkreeg het ten opzihte van andere Europese talen een dominante positie in internationale politieke en ommeriële ontexten (Berns, 1995; Seidlhofer, Breiteneder & Pitzl, 2006). Tegenwoordig vervult het Engels in het Europese bedrijfsleven doorgaans twee funties: (1) een pragmatishe funtie, waarbij het Engels bijdraagt aan een effiiënte internationale ommuniatie, bijv. bij het Duitse bedrijf Siemens dat wereldwijd het Engels als offiiële interne taal hanteert, en (2) een symbolishe funtie, waarbij het Engels samenhangt met een prestigieus imago, bijv. in Europese landen waar Engelstalige tv-ommerials worden gemaakt om het gebruik van een produt te assoiëren met een moderne en dynamishe levensstijl. In beide gevallen wordt het Engels als moedertaal of vreemde taal gebruikt om over vershillende nationale bedrijfsulturen heen de professionele ommuniatie zo effetief mogelijk te laten verlopen (Gerritsen et al., 2007; Philipson, 1998; Piller, 2001; Seidlhofer, 2003; Varner, 2000, 2001). Ook voor veel Nederlandse bedrijven speelt het Engels een prominente rol in het internationale handelsverkeer: de taal staat entraal bij internationale fusies (bijv. Shell), bij managementteams met niet-nederlandse leden (bijv. ING), op internationale orporate websites (bijv. KPN) en in internationaal gebruikte slogans (bijv. Sense and Simpliity van Philips). Een ander ommuniatie-instrument dat de laatste jaren is beïnvloed door de groeiende status van het Engels is het jaarverslag. Het besluit van op de Nederlandse beurs genoteerde organisaties als Aegon, Rodamo en Wolters Kluwer om over te gaan tot een statutaire ofwel leidende versie van het jaarverslag in het Engels, werd reentelijk door het merendeel van hun aandeelhouders gesteund. Dit zou kunnen betekenen dat bedrijven de professionele relevantie van Engelse jaarverslagen hoger waarderen dan die van de Nederlandse. Naast de internationalisering van de markten is de invoering van de IFRS (International Finanial Reporting Standards) een van de voornaamste redenen voor Nederlandse multinationals om een offiieel Engelstalig jaarverslag te publieren. De IFRS zijn internationale regels voor de finaniële verslaglegging, die werden geïntrodueerd na enkele boekhoudinidenten (bijv. Ahold). Bij de IFRS staat het ontwerp van de jaarrekening internationaal vast en ligt de nadruk op de kwaliteit in plaats van de hoogte van de winst. Sinds 2005 dienen alle beursgenoteerde bedrijven in Europa te voldoen aan de IFRS (IAS-plus, 2005). Volgens vershillende Nederlandse ondernemingen benadrukken de Engelstalige verslagleggingregels het belang van Engels als internationale (finaniële) taal. Als gevolg daarvan zien zij het Engelse jaarverslag als een effiiënt middel om hun ativiteiten aan het buitenlandse publiek te rapporteren (ING, 2006; Van der Moolen, 2006). Ondanks de waarde die wordt geheht aan het gebruik van Engels in jaarverslagen van Nederlandse bedrijven, is er nog weinig bekend over de effeten ervan op het ontwerp en de ontvangst van deze verslagen. Dit is opmerkelijk, omdat de teksten en illustraties in het Engelse jaarverslag een brede internationale lezersgroep moeten informeren over en overtuigen van de kwaliteiten van het bedrijf of het bedrijfsresultaat (Garzone, 2004; Hyland, 1998; Jameson, 2000).

336 De bovengenoemde eonomishe, politieke en linguïstishe ontwikkelingen vormden de aanleiding van dit onderzoek, dat zih riht op een groter begrip van de effeten van het Engelse jaarverslag als internationaal ommuniatiemiddel. Het onderzoek betreft in het bijzonder Engelse jaarverslagen die geprodueerd zijn in Nederlandse en Britse bedrijfsulturen en die rapporteren over het jaar Onderzoeksvragen Hoofdstuk 1 presenteert de doelen van het onderzoek en de onderzoeksvragen. In deze studie wordt het jaarverslag geïnterpreteerd als een professioneel genre, wat betekent dat het wordt gezien als een onventioneel teksttype waarmee in een speifieke professionele situatie telkens hetzelfde ommuniatiedoel en dezelfde doelgroepen worden gediend (Bhatia, 1993; Swales, 1990). Van het jaarverslag zijn uitsluitend de Engelstalige, niet-finaniële teksten geanalyseerd. De niet-finaniële onderdelen gaan vooraf aan de jaarrekening en zijn veelal multimodaal van aard (Kress & Van Leeuwen, 2001): ze bevatten vaak geshreven teksten en beelden die de bedrijfsinformatie toegankelijk maken voor zowel finaniële als niet-finaniële lezers (David, 2001). Door de introdutie van de IFRS, die voorkomen dat Europese bedrijven zih op grond van hun jaarrekening nog kunnen ondersheiden, is de persuasieve funtie van de niet-finaniële teksten belangrijker geworden (Nikerson & De Groot, 2005). Omdat er in de Nederlandse en Britse eonomieën sprake is van onurrerende multinationals die op dezelfde markten opereren en het Engelse jaarverslag inzetten om internationaal stakeholders te bereiken, leverden met name de Nederlands-Engelse en Brits-Engelse jaarverslagen relevant vergelijkingsmateriaal op (Forbes, 2003; Nikerson & De Groot, 2005). Er is relatief weinig onderzoek gedaan naar het Engelse jaarverslag als internationaal genre. Sommige studies hebben op basis van de genretheorie een globale beshrijving gegeven van kenmerken van delen van Aziatishe, Amerikaanse en Europese jaarverslagen (bijv. Hyland, 1998; Malavasi, 2005); andere hebben speifieke rossulturele vershillen in kaart gebraht door jaarverslagen uit vershillende Europese landen te vergelijken (bijv. Garzone, 2004; Nikerson & De Groot, 2005). Geen van deze genreanalyses heeft ehter geleid tot een effetstudie waarin de internationale response op typishe kenmerken van een Engels jaarverslag uit een bepaald land werd gemeten. Uit bovenstaande uitgangspunten vloeiden twee onderzoeksdoelen voort. Het eerste doel betrof het inventariseren van ultuurgebonden voorkeuren in het ontwerp van en in de waardering van Nederlands-Engelse en Britse jaarverslagen. Ten tweede beoogde dit onderzoek bij te dragen aan de theoretishe en methodologishe ontwikkeling van genreanalyse als instrument voor het bestuderen van de effetiviteit van multimodale teksten in de internationale bedrijfsommuniatie. De twee hoofddoelen leidden tot twee entrale onderzoeksvragen: I) Wat zijn de genrekenmerken van Engelse, niet-finaniële teksten in multimodale jaarverslagen geprodueerd door Nederlandse bedrijven, en hoe vershillen deze van de genrekenmerken van Engelse, nietfinaniële teksten in multimodale jaarverslagen geprodueerd door Britse bedrijven?

337 II) Hoe beïnvloeden de typishe genrekenmerken van de Engelse, nietfinaniële teksten in multimodale jaarverslagen geprodueerd door Nederlandse bedrijven de perepties die lezers in Groot-Brittannië hebben ten aanzien van het orporate imago en de overtuigende kraht van de tekst, en wat zijn de perepties van deze lezers in geval van genrekenmerken die typerend zijn voor de Britse jaarverslagen? Voor de beantwoording van deze onderzoeksvragen is eerst het genreonept nader gedefinieerd. Volgens deze definitie, die is beshreven in hoofdstuk 2, zijn genres retorishe handelingen ofwel teksten die binnen een soiale groep herhaaldelijk worden gebruikt om een zekere ommuniatieve behoefte te vervullen. Doordat bepaalde teksten steeds opnieuw worden ingezet om een zelfde doel te bereiken (d.w.z. een behoefte te vervullen), zijn ze binnen de soiale groep in de loop der tijd herkenbaar geworden als voorbeelden van een teksttype dat geshikt is voor speifieke ommuniatiesituaties (Miller, 1994). De genretheorie gaat er eveneens vanuit dat de onventionele doelen van deze teksttypen of genres het publiek, de inhoud, de strutuur en de taal van teksten bepalen. Als zodanig geven de doelen een verklaring voor het inhoudelijke en formele ontwerp van de teksten (Swales, 1990). In een verkoopbrief, bijvoorbeeld, ombineert men het produtaanbod vaak met sterke produt- of bedrijfseigenshappen om het doel van de tekst koopgedrag stimuleren te helpen realiseren (Bhatia, 1993). Om de entrale onderzoeksvragen te kunnen beantwoorden zijn in totaal vijf analyses uitgevoerd. De eerste vier analyses gaven samen antwoord op de eerste hoofdvraag en betroffen rossulturele tekstvergelijkingen, de vijfde analyse is uitgevoerd om de tweede hoofdvraag over lezerseffeten te beantwoorden: 1) een analyse van de ommuniatieve ontext van Nederlands-Engelse en Britse jaarverslagen; 2) een inhoudsanalyse van de teksten en de foto s in 3) een struturele analyse niet-finaniële onderdelen uit de 4) een linguïstishe analyse Nederlands-Engelse en Britse jaarverslagen; 5) een internationale responsanalyse van typishe tekstuele en visuele kenmerken in niet-finaniële onderdelen uit Nederlands-Engelse en Britse jaarverslagen. De methoden die zijn gebruikt voor de rossulturele ontext- en tekstanalyses, worden toegeliht in hoofdstuk 3. Hoewel genrestudies tot op heden vooral kwalitatieve beshrijvingen van geshreven tekst bevatten, wordt in hoofdstuk 3 uitgelegd dat binnen de huidige studie ook kwantitatieve onderzoekstehnieken zijn toegepast. Bij de rossulturele analyses zijn de data verzameld door kwalitatieve interpretaties van geshreven tekst en foto s. Vervolgens is getoetst of de gevonden vershillen ook statistish signifiant waren. Voorafgaand aan iedere tekstanalyse is bovendien een test voor de interodeur betrouwbaarheid uitgevoerd om de betrouwbaarheid van de resultaten te meten. In de volgende paragrafen worden de speifieke opzet en de resultaten van de vershillende analyses samengevat. Bij de resultatenbeshrijving van de tekstanalyses ligt de nadruk op de signifiante vershillen die werden gevonden in de rossulturele vergelijkingen. Contextanalyse jaarverslagen Om inziht te krijgen in de ommuniatieve ontext waarin het jaarverslag wordt gebruikt, is in hoofdstuk 4 onderzoht welke taalkeuzes, funties, doelgroepen, ontwerponventies en soiale waarden ten grondslag liggen aan de produtie van het Nederlands-Engelse en Britse jaarverslag (bijv. Beaufort, 1997). Hiervoor zijn een

338 kwalitatieve survey en diepte-interviews gehouden met ommuniatiemanagers en Investor Relations managers in Nederlandse en Britse beursgenoteerde bedrijven. De kwalitatieve survey vond plaats onder 48 Nederlandse bedrijven en belihtte het Engelse taalbeleid voor het jaarverslag in Nederland. De resultaten van de survey toonden aan dat een meerderheid van de beursgenoteerde organisaties in Nederland het Engels als basistaal had gebruikt voor het 2003 jaarverslag. Respondenten verklaarden daarnaast dat originele Engelse jaarverslagen tot stand komen op basis van een proes waarbij vaak Engelstalige rapporten van (inter)nationale divisies in één doument worden geïntegreerd door redateurs met Engels als moedertaal of door een groep moedertaalsprekers van het Engels en het Nederlands. Dit proes kan de linguïstishe invloed van het Nederlands op de Nederlands-Engelse jaarverslag over 2003 hebben verminderd. Omdat dit onderzoek zih rihtte op ultuurspeifieke (in plaats van taalspeifieke) kenmerken in teksten die herhaaldelijk worden gepublieerd door bedrijven die vanaf hun oprihting vanuit de Nederlandse ultuur opereren, lijkt het ehter onwaarshijnlijk dat alle ultuurgebonden ommuniatieonventies in deze teksten zouden zijn verdwenen met de inshakeling van Engelssprekende redateurs. Bij de diepte-interviews waren 11 Nederlandse en vijf Britse bedrijven betrokken. Informanten beantwoordden vragen over de rol van het Engels voor het internationale verslag, de doelen en lezers van het jaarverslag, de rihtlijnen voor het ontwerp van het jaarverslag en de funtie van foto s in het jaarverslag. Zowel de Nederlandse als de Britse informanten gaven aan dat het Engelse verslag vooral relevant is voor het behalen van internationale ommeriële suessen. Ook waren Nederlandse en Britse informanten het grotendeels eens over de ommuniatiedoelen (informeren, een positief imago realiseren), doelgroepen (investeerders, interne lezers, speial interest groepen) en inhoudsriteria (wettelijk, politiek, historish) voor het jaarverslag. In beide nationale bedrijfsulturen bleek het verslag bovendien uit gelijksoortige teksten te bestaan, waarbij volgens de informanten vooral de bedrijfsprofielen, de managementvoorwoorden en de overzihten van operationele ontwikkelingen 1 bijdragen aan de overtuigingskraht van het jaarverslag. Dit komt onder meer door de foto s in deze teksten, die doorgaans de bedrijfsidentiteit, de produten en de orporate marketing illustreren. De diepte-interviews onthulden ook rossulturele vershillen in de retorishe ontext van het jaarverslag. Enerzijds hehtten Nederlandse informanten veel waarde aan het verslag als onderdeel van de finaniële ommuniatiemix en als middel om intern te ommunieren met buitenlandse medewerkers. Anderzijds beshouwden de Britse informanten het jaarverslag als een onderdeel van de finaniële én de marketing ommuniatiemix en vonden zij de investeerders de belangrijkste lezers van het verslag. Het idee dat investeerders een ruiale doelgroep vormen voor het Britse jaarverslag werd onderstreept in de beshrijvingen van de vershillende onderdelen van het jaarverslag: vergeleken met de Nederlandse informanten hadden de Britse informanten meer aandaht voor deze doelgroep in teksten zoals het managementvoorwoord en het overziht van operationele ontwikkelingen. Verder suggereerden de informanten dat vershillen in nationale organisatiestruturen ten grondslag liggen aan de vershillende typen managementteksten in de Nederlands-Engelse en Britse verslagen. In Nederlandse 1 De overzihten van operationele ontwikkelingen zijn in jaarverslagen uit Nederland doorgaans onderdeel van het Verslag van de Raad van Bestuur, terwijl deze overzihten in Britse verslagen veelal onderdeel zijn van de Operating and Finanial Review ( Operational review ).

339 bedrijven zijn de Raad van Bestuur en de Raad van Commissarissen autonome organen, wat binnen het jaarverslag resulteert in een voorwoord van de bestuursvoorzitter (CEO) en een rapport van de Raad van Commissarissen. In Britse bedrijven behoren de direteuren en de ommissarissen tot één managementteam, wat binnen het jaarverslag resulteert in een voorwoord van de bestuursvoorzitter (CEO), een voorwoord van de voorzitterommissaris (Chairman) en een geïntegreerd managementrapport (Report of Diretors). Op grond van de resultaten van de ontextanalyse konden de onderzoeksmaterialen voor de tekstanalyses worden gespeifieerd. Wat betreft de uit Nederland afkomstige materialen, zijn alleen Engelse jaarverslagen geseleteerd die door traditionele Nederlandse bedrijven waren gemaakt en waarvoor het Engels als basistaal was gebruikt. Dit leverde een verzameling op van 22 Nederlands- Engelse verslagen van bedrijven uit 15 vershillende setoren. Vervolgens zijn 22 Engelse jaarverslagen verzameld van traditionele Britse bedrijven uit equivalente setoren. Van de Engelse jaarverslagen zijn uitsluitend de managementvoorwoorden (de Nederlands-Engelse CEO s statement, Britse CEO s statement en de Britse Chairman s statement), de bedrijfsprofielen en de overzihten van operationele ontwikkelingen in de tekstanalyses bestudeerd. Een volledige opsomming van de betrokken setoren en bedrijven is gepresenteerd in Tabel 3 in hoofdstuk 3. Inhoudsanalyse: tekstthema s en fotothema s De typishe inhoudelijke kenmerken van het Nederlands-Engelse en Britse jaarverslag zijn geïdentifieerd en vergeleken door middel van thema-analyses (gebaseerd op Carley & Palmquist, 1992; Kress & Van Leeuwen, 2001). Deze staan beshreven in hoofdstuk 5. Enerzijds werden hierbij sleutelwoorden zoals brand strategy, positioning, markets en distribution hannels geïnterpreteerd als elementen van een oneptueel tekstthema ( marketing ). Anderzijds werden in een foto items als stalen steiger, nieuwbouw, trator en metalen omheining geïnterpreteerd als een oneptueel fotothema ( werkterrein ). De meest voorkomende tekst- en fotothema s in de managementvoorwoorden, bedrijfsprofielen en operationele overzihten zijn opgenomen in Tabellen 8-19 in hoofdstuk 5. De tabellen laten zien dat er een aantal thematishe overeenkomsten waren tussen de Nederlands-Engelse en Britse varianten van deze drie typen teksten. Tekstthema s over het bedrijfsresultaat, de marktbenadering en het vertrouwen in de toekomst kwamen bijvoorbeeld vaak voor in zowel de Nederlands-Engelse als de Britse managementvoorwoorden. In de Nederlands-Engelse en Britse overzihten van operationele ontwikkelingen kwamen veel fotothema s voor met als fous de klanten, de werkvloer en bedrijfsproduten. De resultaten van de inhoudsanalyse toonden wat betreft de managementvoorwoorden vershillen aan tussen ulturen en tussen teksten binnen een ultuur. Terwijl het tekstthema bedrijfsstrategie vaker voorkwam in de Britse CEO s dan in de Britse Chairman s statements, kwam het tekstthema managementdetails vaker voor in de Britse Chairman s dan in de Britse CEO s statements. De tekstthema s Human Resoure Management en orporate governane werden vaker gevonden in de Nederlands-Engelse dan in de Britse CEO s statements. De analyse liet zien dat het tekstthema Corporate Soial Responsibility vaker werd gebruikt in de Britse Chairman s statements dan in de Nederlands-Engelse CEO s statements, en dat het tekstthema dividend vaker werd gebruikt in de Britse Chairman s statements dan in zowel de Nederlands-Engelse als Britse CEO s statements. Daarnaast gaven de onderzoeksresultaten voor de visuele inhoud aan dat het fotothema alle leden van het managementteam het meest

340 voorkwam in de Nederlands-Engelse CEO s statements, en dat de fotothema s werknemers en werkvloer het meest frequent was in de Britse CEO s statements. De inhoudsanalyse van de bedrijfsprofielen demonstreerde alleen een Britse voorkeur voor illustraties met het fotothema medewerkers. Bij de Nederlands- Engelse en Britse operationele overzihten werden alleen rossulturele vershillen gevonden voor de tekstthema s kostenbeleid en marktidentiteit : beide kwamen meer voor in de Nederlands-Engelse dan in de Britse teksten. Strutuuranalyse: moves en strategieën in tekst en foto s De retorishe opbouw van de Nederlands-Engelse en Britse managementvoorwoorden, profielen en operationele reviews is in kaart gebraht door een move-analyse (Bhatia, 1993, 2004; Swales, 1990). Hierover wordt gerapporteerd in hoofdstuk 6. Bij deze analyse worden teksten opgedeeld in moves ofwel tekstfragmenten die een speifieke retorishe funtie vervullen ten aanzien van het tekstdoel. Vervolgens worden binnen de fragmenten de strategieën geïdentifieerd die gebruikt zijn om de moves vorm te geven. Een voorbeeld uit het materiaal van dit onderzoek: in de persoonlijke voorwoorden van het management waren tekstuele uitdrukkingen als Dear shareholders (groet) of the diretors reommend that you (diret verzoek) vershillende strategieën om de move het aanspreken van de lezer te realiseren. Hoewel de analyse van moves en strategieën tot op heden vooral is toegepast in studies naar de retorishe strutuur van geshreven tekst, is ze in het huidige onderzoek ook gebruikt voor het bestuderen van foto s. De analyse van moves en strategieën bleek een geshikte methode te zijn voor het bepalen van de struturele funtie van foto s. Op basis van deze analyse werd een foto van de CEO in een managementvoorwoord bijvoorbeeld gekenmerkt door de move identifieren van het bedrijf en de strategie het bedrijf een persoonlijk geziht geven. Tabellen 5-13 in hoofdstuk 6 geven weer welke moves en strategieën het meest voorkwamen in de teksten en foto s van de managementvoorwoorden, de bedrijfsprofielen en de overzihten van operationele ontwikkelingen. De tabellen tonen aan dat er diverse struturele overeenkomsten werden gevonden tussen de Nederlands-Engelse en Britse teksten. Zo kwamen tekstuele moves en strategieën waarin de bedrijfsomgeving werd geëvalueerd in gelijke mate voor in de Nederlands-Engelse en Britse voorwoorden. Bij de bedrijfsprofielen werden in de foto s rossulturele overeenkomsten gevonden voor moves en strategieën geriht op de algemene of persoonlijke identifiatie van het bedrijf en op de instrumentele of toegevoegde waarde van het produt. De resultaten van de move-analyse lieten alleen vershillen zien voor de geshreven tekst. Bij de managementvoorwoorden was sprake van rossulturele en rosstekstuele vershillen. De tekstuele move rapporteren over leidershap en de tekstuele strategie presenteren van managementdetails kwamen vaker voor in de Britse Chairman s dan in de Britse CEO s statements. De tekstuele move rapporteren over operationele resultaten in ombinatie met de strategie presenteren van operationele details en de strategie beshrijven van het produt werden alle het minst vaak gebruikt in de Britse Chairman s statements. De tekstuele move sympathiseren met stakeholders werd het minst vaak gevonden in de Britse CEO s statements. Verder bleek de tekstuele move speuleren over de toekomst in ombinatie met de strategie korte termijndoelen stellen het meest aanwezig te zijn in de Nederlands-Engelse CEO s statements. Terwijl de tekstuele move de lezer wegwijs maken in de tekst vaker voorkwam in de Britse dan in de Nederlands-Engelse CEO s statements, kwam de move het aanspreken van de lezer in ombinatie met de strategie groeten vaker voor in de Nederlands-Engelse

341 dan in de Britse CEO s statements. Ten slotte werd de tekstuele strategie waardering uitspreken voor stakeholders vaker gevonden in de Nederlands- Engelse CEO s statements en de Britse Chairman s statements dan in de Britse CEO s statements. Bij de bedrijfsprofielen toonden de onderzoeksresultaten alleen aan dat in de Nederlands-Engelse profielen een grotere voorkeur bestond voor de move het ommunieren van de bedrijfsstrategie dan in de Britse profielen. Bij de operationele overzihten bleken de tekstuele strategie introdueren van het bedrijfsprofiel, de move detailleren van het produt geombineerd met de strategie presenteren van de strategishe fous, de move bespreken van de marktbenadering geombineerd met de strategie beshrijven van het produt en de move rapporteren over operationele resultaten geombineerd met de strategie noemen van klantenbehoeften typisher te zijn voor de Nederlands-Engelse dan de Britse teksten. De tekstuele move speuleren over de toekomst geombineerd met de strategie korte termijndoelen stellen kwam vaker voor in de Britse dan in de Nederlands-Engelse overzihten. Lexio-grammatiale analyse: metadisourse in tekst en foto s Als laatste is een tekstanalyse uitgevoerd waarmee de typishe lexio-grammatiale eigenshappen van Nederlands-Engelse en Britse jaarverslagen werden bepaald en vergeleken. Deze studie, die is beshreven in hoofdstuk 7, is gebaseerd op het metadisourse-model (Hyland & Tse, 2004). Hierbij wordt in de tekst gekeken naar woorden en zinsdelen die bijdragen aan de organisatie van een tekst ( interatief ) of aan het realiseren van ontat tussen zender en ontvanger ( interationeel ). Bijvoorbeeld, in het onderzoeksmateriaal werden uitdrukkingen als therefore of we introdue gebruikt om duidelijk te maken wat er in de daaropvolgende tekstfragmenten ging gebeuren en werden uitdrukkingen als extremely diffiult of you door de shrijver gebruikt om zijn/haar houding ten opzihte van de boodshap of de zender duidelijk te maken. Ondanks het feit dat het metadisoursemodel is ontworpen voor de analyse van retoriek op woord- en zinsniveau, is in de huidige genrestudie een relatie gelegd tussen dit model, de theorie van multimodale disourse en de theorie van de visuele semiotiek (Kress & Van Leeuwen, 2006). Deze ombinatie van theoretishe inzihten heeft uiteindelijk geleid tot een model waarmee behalve de tekstuele metadisourse ook de visuele metadisourse in foto s kon worden bestudeerd. Zo konden foto s van managers of hoofdgebouwen worden geïnterpreteerd als illustraties van het bedrijf in de tekst ( interatief ) en werden afbeeldingen van lahende klanten gebruikt om de veronderstelde kwaliteit van het bedrijfsprodut te benadrukken ( interationeel ). Tabellen in hoofdstuk 7 presenteren de vormen van tekstuele en visuele metadisourse die het meest voorkwamen in de Nederlands-Engelse en Britse managementvoorwoorden, bedrijfsprofielen en overzihten van operationele ontwikkelingen. Ook de resultaten van deze tekstanalyse duidden op een aantal rossulturele overeenkomsten voor de drie teksttypen. Bij de Nederlands-Engelse en Britse managementvoorwoorden werd bijvoorbeeld evenveel gebruik gemaakt van interatieve metadisourse in de vorm van woorden of zinsdelen die betekenisrelaties weergeven, die een taalhandeling bevatten, die naar informatie elders in de tekst verwijzen en die de inhoud van de tekst verder speifieren. De profielen in beide ulturen bevatten vooral visuele metadisourse in de vorm van interatieve speifiaties van produten, werkterreinen of markten en interationele markeerders van de professionaliteit, de moderniteit of de toegevoegde waarde van het produt. De analyse van lexio-grammatiale kenmerken in de managementvoorwoorden liet ook vershillen zien tussen ulturen en teksten. In de

342 geshreven tekst werden vershillen gevonden voor interationele markeerders van shrijver-lezerrelaties ( you, note that ), die het minst vaak voorkwamen in de Britse CEO s statements, en verwijzingen naar de shrijver als het management ( the board ), die het vaakst voorkwamen in de Britse Chairman s statements. Daarnaast toonden de resultaten aan dat interatieve strutuuraanduiders ( first, finally ) het meest werden gebruikt in de Nederlands-Engelse CEO s statements. Uit de resultaten van de fotoanalyse bleek dat er in de Britse CEO s statements een signifiante voorkeur bestond voor interatieve speifiaties van het marketingbeleid in de tekst (klanten) en voor interationele illustraties van het bedrijf als geëmanipeerd (vrouwelijke en mannelijke medewerkers), multiultureel (medewerkers met vershillende etnishe ahtergrond) en internationaal (medewerkers van internationale divisies). Interatieve speifiaties van het bedrijf in de tekst en interationele verwijzingen naar de shrijver in de vorm van foto s van het managementteam werden vaakst gevonden in de Nederlands-Engelse CEO s statements. Daarentegen kwamen illustraties van het bedrijf en auteurverwijzingen in de vorm van foto s van individuele managers het minst vaak voor in deze teksten. Foto s van medewerkers als interatieve speifiaties van het bedrijf kwamen vaker voor in de Britse dan in de Nederlands-Engelse CEO s statements. Interationele illustraties van het bedrijf als open en opreht (blik van managers/medewerkers rihting amera) werden vaker gevonden in de Nederlands-Engelse CEO s statements dan in de Britse Chairman s statements, terwijl interationele markeerders van intimiteit met het bedrijf (lose-up van managers) vaker werden gevonden in de Britse Chairman s statements dan in de Nederlands-Engelse CEO s statements. Bij de bedrijfsprofielen werden in de Nederlands-Engelse teksten meer interationele markeerders van nadruk ( obviously, signifiant ) gebruikt dan in de Britse teksten. De Britse profielen bevatten meer verwijzingen naar de shrijver in de vorm van de eerste persoon meervoud ( we, our ). De analyse van de visuele metadisourse toonde aan dat interatieve illustraties van het bedrijf in de tekst en interationele verwijzingen naar de shrijver in de vorm van foto s van medewerkers vaker voorkwamen in de Britse dan in de Nederlands-Engelse profielen. Wat de operationele overzihten betreft, werd alleen aangetoond dat interatieve strutuuraanduiders ( first, finally ) het meest frequent waren in de Nederlands- Engelse dan in de Britse teksten. Conlusie rossulturele tekstanalyses Het eerste onderzoeksdoel en de eerste onderzoeksvraag rihtten zih op het identifieren en vergelijken van multimodale genrekenmerken in niet-finaniële teksten uit Nederlands-Engelse en Britse jaarverslagen. Ter beantwoording van de eerste vraag zijn bovengenoemde ontext- en tekstanalyses uitgevoerd, waarbij rossulturele vergelijkingen hebben geleid tot overzihten van overeenkomsten en vershillen in de geshreven tekst en de foto s. Het overziht van de rossulturele genrevershillen, die ook hierboven entraal stonden, is gepresenteerd in Tabel 2 van hoofdstuk 8. Over het algemeen toonden de onderzoeksresultaten dat Nederlands-Engelse en Britse jaarverslagen bestaan uit vershillende teksttypen die elk in beide nationale bedrijfsulturen veel dezelfde ontextuele en tekstuele eigenshappen bevatten, maar die ook ultuurgebonden kenmerken hebben. Het grote aantal rossulturele overeenkomsten in ommuniatieve ontext, tekstuele en visuele elementen lijkt erop te wijzen dat de Nederlands-Engelse en Britse jaarverslagen behoren tot dezelfde internationale genreset (Devitt, 1991). Dit betekent dat het Engelse jaarverslag in de twee ulturen herkend en gebruikt wordt als een vergelijkbare ompilatie van samenhangende teksten met elk een eigen doel,

343 lezersgroep, inhoud en vorm. Daarnaast maakt de ombinatie van de vershillende teksten het bestaan van het overkoepelende genre jaarverslag mogelijk (Jameson, 2000). Verder lijken de huidige onderzoekresultaten te bevestigen dat het retorishe doel van een tekst een bepalende fator is bij het identifieren van een genre: ondanks de ultuurspeifieke kenmerken in de drie typen onderzoeksteksten, lijkt met name de rossulturele eenheid in ommuniatiedoel voor het jaarverslag als geheel én voor de aparte teksten te hebben gezorgd voor overeenkomsten in taalgebruik en voor de omshrijving van de Nederlands-Engelse en Britse jaarverslagen als vergelijkbare voorbeelden van dezelfde genreset (Miller, 1994). Een overziht van de signifiante rossulturele vershillen in genrekenmerken (zie Tabel 2 in hoofdstuk 8) biedt een algemene blik op de aard van de vershillen. Ten eerste blijkt dat vershillen tussen bedrijfsulturen vooral gevonden zijn op het lexio-grammatiale analyseniveau. Met andere woorden, ulturele voorkeuren waren vooral zihtbaar bij de meest gedetailleerde en onrete analyse-items. Bovendien tonen de resultaten aan dat er met name sprake was van vershillen tussen de Nederlands-Engelse en Britse managementvoorwoorden, dus bij teksttypen waar in Nederland maar één variant voor bestaat (de CEO s statement) en in Groot-Brittannië twee varianten (de CEO s statement en de Chairman s statement). Ten derde blijkt dat de meeste vershillen in tekstkenmerken voorkwamen in de geshreven onderdelen van de Nederlands-Engelse en Britse jaarverslagen; kennelijk is er tussen de twee bedrijfsulturen meer overeenstemming over de typishe visuele ondersteuning van deze geshreven teksten. Niettemin lijken visuele kenmerken een belangrijke rol te spelen bij de identifiatie van genres. Enerzijds kunnen ze bijdragen aan het ondersheid tussen diverse typen nietfinaniële teksten (bijv. de diretiegerelateerde fotothema s in de managementvoorwoorden versus de klant-, produt- en werkvloergerelateerde fotothema s in de operationele overzihten). Anderzijds kunnen foto s een indiatie zijn voor ultuurspeifieke voorkeuren binnen een bepaald teksttype (bijv. het managementteam thema in de Nederlands-Engelse voorwoorden). Ten vierde kan nog worden gesuggereerd dat de rossulturele vershillen in het huidige onderzoek niet alleen gebonden zijn aan bepaalde teksttypen in het jaarverslag maar mogelijk ook aan ulturele onventies in de (finaniële) bedrijfsommuniatie. Sommige tekstkenmerken waren immers typish voor een bepaalde tekst in een bepaalde bedrijfsultuur (bijv. het tekstthema dividend in de Britse Chairman s statements) en andere kenmerken waren typish voor vershillende teksttypen in een bepaalde bedrijfsultuur (bijv. het fotothema medewerkers in de Britse CEO s statements en de Britse bedrijfsprofielen). In navolging van de genretheorie (Swales, 1990) is in dit onderzoek geprobeerd een relatie te leggen tussen de ontext waarin een tekst tot stand komt en de typishe multimodale kenmerken van de tekst. De resultaten van de tekstanalyses lijken erop te wijzen dat een aantal van die kenmerken het direte resultaat zijn van ommuniatieve aspeten die in de ontextanalyse zijn bekeken, andere zijn waarshijnlijk het gevolg van nationale voorkeuren. Over het algemeen kunnen de vershillen tussen de Nederlands-Engelse en Britse teksten verklaard worden door: 1) het ommuniatiedoel van het jaarverslag, zoals de beïnvloeding van lezerattitudes door het veelvuldig gebruik van interationele metadisourse, 2) het ommuniatiedoel van een niet-finaniële tekst, zoals het presenteren van een operationele samenvatting met relatief veel operationele details in de Britse CEO s statements, 3) de professionele funtie van de shrijver, zoals het rapporteren over de managementsituatie in het voorwoord van de Britse toezihthoudende Chairman, 4) soiaal-politieke ontwikkelingen, zoals het veelvuldig gebruik van het tekstthema

344 orporate governane in Nederlands-Engelse managementvoorwoorden na de introdutie van een nieuwe governane ode, en 5) nationale soiale waarden, zoals het gebruik van foto s van het managementteam (in plaats van foto s van individuele managers) in voorwoorden die zijn geprodueerd in het feminiene Nederland. Theoretishe en methodologishe impliaties van de rossulturele tekstanalyses Het tweede doel dat werd gesteld bij de eerste onderzoeksvraag betrof de verdere theoretishe en methodologishe ontwikkeling van genreanalyse. Vanuit theoretish perspetief kan ten eerste worden geonludeerd dat het genreonept niet alleen een kader shetst voor de interpretatie van teksten, maar ook voor de identifiatie van een tekst. Anders gezegd, in het huidige onderzoek heeft het genreonept rihting gegeven aan de bestudering van teksten en foto s in het Engelse jaarverslag; het maakte duidelijk welke tekstniveaus of welke tekstaspeten moeten worden beliht om inziht te krijgen in het totale ontwerp van het jaarverslag. Bovendien bleek dat het genreonept ruimte biedt voor een volledige en veelzijdige tekstinterpretatie door de mogelijke integratie van methoden uit vershillende wetenshappelijke disiplines (bijv. interviews etnografie, metadisourse linguïstiek). Ten tweede heeft dit onderzoek aangetoond dat de visuele retoriek in teksten een plaats heeft binnen de genreanalyse. Hoewel genrestudies tot op heden vooral geshreven teksten betroffen, demonstreren de huidige fotoanalyses dat visuele kenmerken opgenomen kunnen worden in vershillende onderzoeksmodellen die dan leiden tot een ompletere interpretatie van multimodale ommuniatie. Vanuit methodologish perspetief kan worden gesuggereerd dat een ombinatie van kwalitatieve en kwantitatieve tehnieken de resultaten van een genreanalyse meer solide maakt. Voorafgaand aan iedere tekstanalyse werden interoder testen uitgevoerd. Deze testen ondersteunden niet alleen de betrouwbaarheid van de onderzoeksresultaten, maar gaven ook aan welke problemen kunnen optreden bij de analyse van bepaalde genrekenmerken. Zo bleek tijdens de interoder betrouwbaarheidstesten dat het afbakenen van tekstuele moves en strategieën en de analyse van visuele metadisourse zeer bewerkelijke proessen zijn. In de formele tekstanalyses werden de data die door kwalitatieve tekstinterpretatie waren verkregen, verwerkt in een statistish programma om de signifiantie van eventuele rossulturele vershillen te kunnen berekenen. Door het gebruik van statistishe toetsen kon de betekenis van de onderzoeksresultaten worden versterkt en konden waargenomen patronen in taalgebruik vervolgens worden gerelateerd aan de ulturele of ommuniatieve ontext van de betreffende teksten. Hoewel nog weinig genrestudies zih hebben geriht op het effet van typishe genrekenmerken in een bepaalde tekst of ultuur, bieden de resultaten van de huidige tekstanalyses aanleiding voor verder onderzoek naar genrerespons. De signifiante vershillen tussen de geshreven teksten en de foto s in de Nederlands- Engelse en Britse jaarverslagen zouden in een responsstudie kunnen worden gebruikt als onafhankelijke variabelen op basis waarvan lezervoorkeuren in kaart worden gebraht. Hierover gaat het laatste hoofdstuk in dit proefshrift. Responsanalyse Nederlands-Engelse en Britse managementvoorwoorden Hoofdstuk 9 staat in het teken van de tweede onderzoeksvraag die foust op de pereptie van lezers in Groot-Brittannië ten aanzien van de multimodale genrekenmerken in het Nederlands-Engelse en Britse jaarverslag. Ter beantwoording van deze vraag is een experiment uitgevoerd waarbij finaniële

345 analisten in Londen Nederlands-Engelse en Britse managementvoorwoorden hebben geëvalueerd in termen van waargenomen orporate imago en overtuigingskraht van de tekst. Op basis van de assumptie dat taalgebruikonventies waarmee men bekend is aantrekkelijker zijn dan onventies waarmee men onbekend is (Byrne, 1971), kon de onderzoeksvraag verder worden uitgewerkt in vijf hypothesen en één subvraag. In de hypothesen werd ahtereenvolgend verondersteld dat de analisten uit Groot- Brittannië een positiever imago zouden toeshrijven aan het bedrijf ahter de Britse tekst (H1), en dat zij de Britse tekst aantrekkelijker (H2), geloofwaardiger (H3) en begrijpelijker (H4) zouden vinden. Daarnaast werd verondersteld dat de finaniële analisten een algemene voorkeur voor de Britse tekst zouden hebben (H5). Voor de beantwoording van de subvraag werd nagegaan of de doelen en doelgroepen van het Britse jaarverslag, zoals die in de ontextanalyse werden geformuleerd door de Britse informanten, daadwerkelijk aansluiten bij de lezerbehoeften van de finaniële analisten. In totaal namen 35 finaniële analisten in Londen deel aan het experiment. Door middel van een vragenlijst gaven zij hun mening over twee volledige managementvoorwoorden en over enkele vragen, stellingen, tekstfragmenten en foto s gerelateerd aan het ontwerp van de voorwoorden. Een voorbeeld van de stimuli en de bijbehorende vragenlijst is opgenomen in Appendix X. De omplete voorwoorden waren zodanig gemanipuleerd dat ze typishe Nederlands-Engelse of typishe Britse tekstkenmerken bevatten. Ieder voorwoord werd door de respondenten geëvalueerd met behulp van meetinstrumenten voor orporate reputatie en investeringsintenties (samen orporate imago), aantrekkelijkheid van de tekst, geloofwaardigheid van de tekst en begrijpelijkheid van de tekst. Voor de andere items in de vragenlijst werden shalen gebruikt waarbij de tegenpolen representatief waren voor een voorkeur voor ofwel de Nederlands-Engelse genrekenmerken ofwel de Britse kenmerken. Onder de 35 respondenten werden vier versies van de vragenlijst verspreid, waarbij de volgorde van de items steeds vershillend was. Respondenten vulden de vragenlijst op loatie in tijdens hun werkpauzes. De resultaten van het experiment toonden aan dat de finaniële analisten in Londen een positievere orporate reputatie toekennen aan het bedrijf ahter het Britse voorwoord dan aan het bedrijf ahter het Nederlands-Engelse voorwoord. Hun beoordeling liet ehter geen vershil zien voor de intentie om te investeren in deze bedrijven. Bij de evaluatie van de managementvoorwoorden bleek dat de analisten de Britse tekst begrijpelijker vonden. Wat betreft de aantrekkelijkheid en geloofwaardigheid van de Nederlands-Engelse en Britse teksten was er geen vershil. Toh werd over het algemeen de voorkeur gegeven aan het Britse voorwoord. De belangrijkste reden voor deze voorkeur was de heldere strutuur in de Britse tekst, met name gerealiseerd door kopjes tussen paragrafen. De reaties op de stellingen en de tekstfragmenten gaven aan dat finaniële analisten deels de Nederlands-Engelse kenmerken in managementvoorwoorden prefereerden (voor de thema s bedrijfsstrategie en toekomstdoelen ) en deels de Britse kenmerken (voor de thema s managementsituatie, maatshappelijk verantwoord ondernemen en dividend en voor de struturele elementen kopjes ). Bij de evaluatie van de foto s bestond vooral een voorkeur voor de visuele retoriek in de Nederlands-Engelse voorwoorden (voor portretten van het managementteam en voor mediumshot opnamen van managers) en enigszins voor de visuele retoriek in de Britse voorwoorden (voor portretten van managers waarin zij wegkijken van de amera). Op de vraag hoe relevant het jaarverslag is, antwoordden de finaniële analisten dat het verslag (heel) belangrijk is. Maar op de vraag hoe vaak zij het

346 managementvoorwoord lezen, was het antwoord een enkele keer. Finaniële analisten gaven aan dat het jaarverslag voornamelijk wordt gebruikt om informatie te verkrijgen over de bedrijfsresultaten en de bedrijfsstrategie. Het managementvoorwoord wordt meestal doorgenomen om een snelle indruk te krijgen van het bedrijfsresultaat en de bedrijfsidentiteit. Volgens de meeste finaniële analisten wordt het managementvoorwoord ondertekend door de CEO of de toezihthoudende Chairman. Hoewel de Britse tekstvariant op enkele ruiale punten (reputatie, begrijpelijkheid, algemene voorkeur) beter soorde dan de Nederlands-Engelse variant, liggen de antwoorden op de hypothesen en de subvraag genuaneerder. De resultaten laten immers zien dat op andere punten (tekstthema s en visuele retoriek) de Nederlands-Engelse genrekenmerken beter soorden of dat er geen vershil in waardering was. Op grond van de resultaten kan worden geonludeerd dat er gedeeltelijke empirishe ondersteuning was voor H1 (orporate imago) en H5 (algemene voorkeur), dat H2 (aantrekkelijkheid) en H3 (geloofwaardigheid) niet werden ondersteund, maar dat H4 (begrijpelijkheid) wel werd bevestigd. Een andere onlusie is dat de finaniële analisten uit Groot-Brittannië het jaarverslag als geheel gebruiken op een manier die de Britse bedrijven volgens de resultaten in de ontextanalyse (zie hoofdstuk 4) ook verwahten, maar dat het managementvoorwoord in de praktijk van minder belang is dan verwaht. Aldus lijken de resultaten van het experiment tot op zekere hoogte de assumptie te ondershrijven dat taalgebruikers een hogere waardering hebben voor tekstonventies waarmee zij binnen een soiale groep het meest bekend zijn. Teksten waarin deze onventies worden toegepast zijn herkenbaarder en daardoor vaak aantrekkelijker dan teksten waarin niet aan de onventies wordt voldaan (Byrne, 1971; Vaughan & Dillon, 2006). Theoretishe en methodologishe impliaties van de responsanalyse De experimentele resultaten implieren dat responsonderzoek naar genrekenmerken bijdraagt aan de verdere ontwikkeling van het genreonept. Ze ondersteunen daarmee ook het tweede entrale onderzoeksdoel. Ten eerste wijst het experiment erop dat genretheorie zih niet alleen zou moeten beperken tot de lassifiatie van teksttypen, maar dat het ook de reeptie van teksttypen zou moeten behelzen. Teksten worden niet alleen geprodueerd vanuit een bepaalde soiale ahtergrond, ze worden door de ontvangers ook geïnterpreteerd vanuit eigen soiale verwahtingen. Vooral in internationale ommuniatieontexten kan een genre pas effetief zijn als zender en ontvanger overeenkomstige ideeën hebben over de typishe genrekenmerken. Een breder begrip van de effetiviteit van een genre als interatieve tekst is dus gebaat bij theoretishe inzihten in zowel het ontwerp van de tekst als de respons op de tekst. Ten tweede blijkt uit de responsanalyse dat genreonderzoek niet noodzakelijk gebonden is aan desriptieve methoden maar dat het evengoed een basis biedt voor experimentele methoden. Methodologish gezien kan genreonderzoek dus op logishe wijze worden uitgebreid met een ombinatie van tehnieken die een volledig beeld geven van het (inter)nationale gebruik en effet van (professionele) teksttypen. Als laatste kan vanuit praktish oogpunt worden opgemerkt dat responsanalyses van professionele genres interessant zijn voor bedrijven die internationaal afhankelijk zijn van hun stakeholders. De huidige resultaten laten bijvoorbeeld zien dat finaniële lezers in Groot-Brittannië in bepaalde opzihten een voorkeur hebben voor het Britse jaarverslag en dat het Nederlands-Engelse jaarverslag daardoor aan effetiviteit inboet. Voor Nederlandse

347 bedrijven die zih via hun jaarverslag willen rihten op deze lezers is dit een relevant gegeven. Tot slot In dit proefshrift is gerapporteerd over een onderzoek naar de aard van en de respons op tekstuele en visuele kenmerken van Engelse jaarverslagen die tot stand zijn gekomen in twee Europese bedrijfsulturen. De rossulturele vergelijkingen toonden aan dat een professioneel genre vershillende multimodale kenmerken kan hebben in vershillende ulturen. De responsanalyse liet zien dat deze vershillende kenmerken de effetiviteit van het genre kunnen beïnvloeden bij internationale doelgroepen. Beide typen analyses bieden suggesties voor verder onderzoek naar 1) de verdere ontwikkeling van het genreonept, 2) het ultuurgebonden en multimodale ontwerp van professionele genres en 3) het gebruik van deze genres in de internationale bedrijfsommuniatie.

348 Curriulum Vitae Elizabeth de Groot was born on 23 September 1976 in Delfzijl, the Netherlands. After she reeived her Athenaeum diploma from the Fivel College in Delfzijl, she went to the United Stated and reeived a high shool diploma. She returned to the Netherlands to study Communiation Studies and English Language and Culture at the University of Groningen. In 2001 she graduated in Communiation Studies from the University of Groningen, on the basis of a Master thesis about the visualization of ultural aspets in Duth TV-ommerials. In 2003 she graduated in English Language and Culture from the University of Groningen, on the basis of a Master thesis about the oneptualization of word meaning. In she worked as a junior leturer in the Departments of English and Applied Linguistis, University of Groningen. During that same period, she worked as a part-time researher at the University of Groningen, onduting several opinion polls within the Duth business ontext. In 2004 she joined the Department of Business Communiation Studies at the Radboud University Nijmegen, where she worked as a junior researher until In 2008 she was appointed leturer in this department. She teahes a variety of ourses and supervises a number of Bahelor and Master theses into aspets of (English) Business Communiation.