STEPS IN LANGUAGE DOCUMENTATION AND REVITALIZATION JACK MARTIN NICK THIEBERGER
Steps in Documentation Steps in Community Relations
STEPS IN LANGUAGE DOCUMENTATION Ideally: as rich as possible a set of recordings of as many people as possible, from a range of ages and both women and men Text, audio, video, photographs Useful records Accessible records
QUIZ How many languages are there in the world? How much is recorded for each of them?
Size of icon 0-20 points 1-5 points for each of four categories: Grammar; Lexicon; Texts; Media corpus white = 0; red = 1-5; purple = 6-10; yellow = 11-15; green =16-20 http://bit.ly/docindex
WE WANT To create as rich a record as possible for as many languages as possible How do we get there?
LINGUISTS AND SPEAKERS Linguists will be creating the most detailed records for most languages Have a responsibility to use the best possible methods Have changed a lot in the past generation Speakers create records and need help to make records last
METHODS AND TOOLS New ways of: Recording easier and cheaper than before Transcribing linked to media Tracking what we have done metadata tools Crowdsourcing new records, annotating old ones
Tiawi itraus traus traus traus, natrauswen ga itaos nlag. Itrausi pan kaipa. Me komam uta laap kin uto mau, a? Malen umat, inom. The old man can talk and talk and talk, his story is like the wind. He tells it and it is gone. But there aren't many of us left. When we die, it will be finished. Kalfapũn Mailei, 1998, Erakor Village, Efate, Vanuatu (NT1-98009-A, 1932.36 1942.16)
BEFORE FIELDWORK Funding for your fieldwork (course on grant writing) Write consent forms and translate them into a relevant language Planning what to take, how to use equipment, test it before you leave (course on audio 1 & 2; video 1 & 2) Plan backup for records made in the field (course on data management and archiving) Learn what software will help with fieldwork (many courses at CoLang) Talk to an archive about how they want records to be delivered (course on data management and archiving)
COMPUTERS Software tools for organising field materials Databases for keeping track of what you have recorded and what files contain (database course) Transcription (Elan course) Annotation of transcriptions (Fieldworks course) Creation of a lexical database (Fieldworks course)
Playable media, produced in Elan Text, interlinearised in Fieldworks
QUALITY OF RECORDS YOU CREATE OF CONTENT AND OF FORM Content culturally relevant, range of genres, old and young speakers, men and women. Elicitation, using various stimuli. Form recordings are clear and useful and have descriptions that allow others to identify what is in the recordings. Text is structured.
CONTENT OF RECORDS Record narratives, songs, conversations, greetings etc. Observations, fieldnotes Elicitation, using various stimuli Relevant word lists (course: Lexicography; Introduction to Rapid Word Collection) Plant and animal identification books (course: ethnobotany; ethnobiology) Astronomical charts, software Stimulus videos (http://fieldmanuals.mpi.nl/) Recording song and performance (course: song documentation) etc
TOO MANY TOPICS FOR A LINGUIST TO DEAL WITH ON THEIR OWN Star Trails showing Polaris at the center. (photo courtesy of Jerry Schad)
FORM OF RECORDS Learn how to use your equipment Pay attention to, e.g. background noise for recordings and bright or dim light for photography Know how to use your microphone Use a tripod to keep video steady Keep track of your recordings in fieldnotes Use most suitable software (courses in audio, video, transcription, Elan, Miromaa, )
PREPARATION Getting to know speakers Establish ground rules for your work, shared understanding of your goals and those of the speakers signed consent forms that allow you to use the material recorded (and to archive it for others to use) (Community Language Archives course) Developing a spelling system that is elegant and acceptable to the speakers (orthography course)
ORGANISING YOUR RECORDS Keeping track of your files Using metadata entry tools SayMore (http://saymore.palaso.org/) Arbil (tla.mpi.nl/tools/tla-tools/arbil/) ExSite9 (http://www.intersect.org.au/exsite9) Spreadsheets or databases
ORGANISING YOUR RECORDS Know how to convert text from the output of one tool to the input format of another tool \id 061:005 \aud NT1-98002-A 316.43 320.317 \tx Preg nafkal skot namẽr \mr preg nafkal skot namẽr nig Emlakul. nig Emlakul \mg make war with people of p.name \fg To fight the people of Malakula.
ORGANISING YOUR RECORDS \id 061:005 \aud NT1-98002-A 316.43 320.317 \tx Preg nafkal skot namẽr \mr preg nafkal skot namẽr nig Emlakul. nig Emlakul \mg make war with people of p.name \fg To fight the people of Malakula.
ORGANISING HISTORICAL RECORDS In 1904, 500 copies of a blank wordlist questionnaire for Aboriginal languages of Western Australia Around 220 filled out 1647 word prompts 193 sentence prompts Data management and archiving course
IMAGE= 51-038T-XII2D17 ID= 0392
REGULAR EXPRESSIONS Convert large files of text globally (course on data management and archiving) Convert lexical files into formats required for online dictionaries or phone apps
CONVERT FROM THIS TO THIS Somali dictionary, 26,000 entries
CREATE IN FORMATS THAT CAN BE REUSED Lexical files can be used in: online dictionaries books (dictionaries for children, language learners, special domains) lots of dictionary apps for mobile devices
MA IWAIDJA!
MOBILE APPS FOR RECORDING Aikuma System to record stories and then to record a translation
MOBILE APPS FOR RECORDING Aikuma http://lp20.org/aikuma/pilot_project.html
Me, selwan ag kupi emãe, tiawi itraus traus traus traus, natrauswen ga itaos nlag. Itrausi pan kaipa. Me komam uta laap kin uto mau, a? Malen umat, inom. But when you are far away the old man can talk and talk and talk, his story is like the wind. He tells it and it is gone. But there aren't many of us left. When we die, it will be finished. Recording made Recording in the field. Named Kalfapũn archived with Mailei, 1998, Erakor Village, and described Efate, in Vanuatu persistent fieldnotes. Transcribed identification by hand in the field. (NT1-98009-A, 1932.36 1942.16) Playable media, Transcribed in Elan. Timecodes Included in texts glossed in Fieldworks
GET ADVICE http://www.nflrc.hawaii.edu/ldc/
GET ADVICE http://rnld.org
LOTS OF EXCITING OPPORTUNITIES Assessing tools for their usefulness in doing what we want to do Creating better records Putting them in digital archives Building ways of finding, viewing and working with them Learning about new methods Enjoy CoLang!