Socio-Economic and Demographic Differentials between Hindus and Muslims in India

Transcription

1 Socio-Economic and Demographic Differentials between Hindus and Muslims in India Abusaleh Shariff This paper reviews the socio-economic and demographic data according to religion available from various censuses, National Sample Survey and academic publications since the independence of India. Indicators such as, the structure and levels of employment, of living and of education according to religion arc discussed. The fertility and mortality indicators, distribution and growth of population are also presented. The paper emphasises the need to strengthen the data based which would allow a study of ethnic and religious differentials in socio-economic and educational achievements. I Introduction INDIA is a multi-cultural, multi-ethnic and multi-lingual country. People belonging to many religions, such as Hinduism, Buddhism, Jainism, Sikhism, Islam and Christianity live in this country since time immemorial. However, there appears to be substantial differentials in the socioeconomic and demographic profiles of major religious communities in India, mainly emerging from socio-cultural and historical reasons. This paper presents various types of social, economic and demographic data gathered from various censuses and National Sample Survey publications and discusses differentials according to religion. Il is hoped that this analysis would provide a lead in understanding the contemporary Indian society in a better perspective. The main purpose here, however, is to generate interest among the readers to pursue further research on this issue, both at the micro-disaggregated and macro levels. At the outset, it must be emphasised that various religious groups in India, especially Hindus, Muslims and Christians are not homogeneous populations. Each one is divided and subdivided into innumerable castes, sects and cultural groups. For example, the scheduled castes and the scheduled tribes, besides a large number of socially backward castes are a part of the Hindu community. Similarly, the Sunnis, Shias, Bohras and many other sects or cultural groups constitute the Muslims. It is my view that studying socio-economic differentials in India at the level of religious aggregation is simplistic and not advisable. However academic research has to reflect the socio-economic and cultural realities of its age. The contemporary politicisation of religion is such that a new focus which can be called as 'political economy and political demography of religions' is the need of the hour. It is in this context, even at the cost of being simplistic and seemingly un-academic perspective, that this paper is conceived and written. It is necessary that relevant facts and figures found in the official and academic records and publications are put together for an objective assessment of reality. Section 11 of this paper presents the growth and distribution of population according to religion in India and in its major states. Section III discusses data on employment, occupation, levels of living and education according to various religions. The following section presents some fertility and mortality indicators according to religious affiliation and fifth section presents a discussion and some conclusions. II Population Profile According to Religious Affiliation Now let us take a look at the growth of population since independence. The Indian government as a routine exercise conducts complete enumeration of population every 10 years, which is known as decennial census. The last census was conducted in 1991 whose Economic and Political Weekly November 18,

2 preliminary results are flow becoming available However, for many types of information the latest sources still arc the 1981 Census publications. Census gives population figures by major religions for all India and slates and union territories. The religious affiliation as reported by the head of the household during the census operations is presented in broad categories. For example, all Hindu castes including the scheduled castes and the scheduled tribes are identified as Hindus. Similarly, many Muslim sects and sub-sects have been clubbed as Muslims. The figures are presented in Table 1. India got independence in 1947 which also led to partition of the country. A large number of Muslims moved over to the then Pakistan. In 1951 (the first census after independence) there were 35 million Muslims living in India, forming the largest minority. According to this census there were 304 million Hindus and about 8.3 million Christians in India. The Hindu population (including the scheduled castes and scheduled tribes) grew at the rate of 20.3,24.7 and 24.2 per cent, respectively, during the past three decades and in 1981 they were 550 millions or about 82.6 per cent of India's population. The Muslims population grew at the rate of about and 30.6 per cent during the same period and by 1981 there were 76 million Muslim living in India. The growth of Christians was and 17.4 during the corresponding periods. The decadal growth rate during was substantially lower than that of the previous decades, for the total population as a whole. This means the growth rates of its components, such as by various religions would also be lower compared with respective rates during A fact to note, however, is that while Hindus were the significant majority, Muslims constituted 11.4 per cent and Christians constituted about 2.4 per cent of the total population in India. Table 2 provides the distribution of population according to religious affiliation for 15 major states in India. Excepting in Jammu and Kashmir and Punjab, the Hindu population is in majority in all the remaining 13 states considered. However, the Hindu population in Kerala and West Bengal was 58 and 77 per cent respectively which is less than the all-india average of 82.6 per cent. It can be seen that 64 per cent of population in Jammu and Kashmir are Muslims, This is the only state where Muslims are in the majority. The next largest proportion, about 21 per cent each, is found in West Bengal and Kerala. Muslims constitute about 16 per cent in Uttar Pradesh, 14 per cent in Bihar, 11 percent in Karnataka and 9 per cent in Maharashtra. The stales Where Muslim population is about 5 per cent or less are Tamil Nadu Orissa, Madhya Pradesh, Haryana and Punjab. Christians are mostly concentrated in Kerala, about 21 per cent, and in Tamil Nadu, about 6 per cent of the state population. The other areas where Christians are found in significant percentages are some stales in the northeastern parts of India. Overall, about 52 per cent of Muslims in India reside in three states, namely Uttar Pradesh, Bihar and West Bengal. Another interesting feature of Muslims is that, with a few noteworthy exceptions, on the whole a larger proportion of Muslims live in urban areas, when compared with members of other religions (Table 3). For example, in all the south Indian states the percentage of Muslims living in urban areas is more than double that of the Hindu population, excepting in Kerala. It is likely that the proportion of Muslims in urban areas may have increased over the past four decades. The other states where the proportion of urban Muslims is substantially higher than the stale average are Rajasthan, Gujarat and Uttar Pradesh. Conversely, in West Bengal, Bihar, Kerala and Jammu and Kashmir where a sizeable number of Muslims live, the proportion in urban areas is either lower than or about the same as the stale average. At the all-india level, however, while only 22 per cent of the Hindus live in urban areas, about 29 per cent of Christians and 34 percent of Muslims do so. Yet a point to be noted is that 66 per cent of Muslims and 71 per cent of Christians still live in Indian villages. Ill Socio-Economic Differentials Among Populations of Various Religions Socio-economic conditions of a population can be assessed by studying indicators, such as. land ownership, occupation, worker population ratio, literacy and school continuation rates. Such data, for the first time, have been made available by the National Samples Survey Organisation (NSSO) through its 43rd round survey conducted during NSSO is probably the most scientific and dependable source of data on various types of social and economic information for both the states and national levels. These surveys are conducted on a sample of households round the year and every year, often with a special focus. The 43rd round surveyed about 1.3 lakh households spread over about 8,400 sample villages and 4,600 urban sample blocks in the whole country. Publications emerging out of 43rd round survey contains some cross-tabulations by major religions, which are discussed in this paper. Although information can be made available separately for various slates the published statistics are now available only at the level of all-india. 2948

3 OCCUPATION AND WORK PARTICIPATION Some rare data on occupation and work participation rates both for rural and urban is presented in Table 4. Accordingly, in rural India, about 44 per cent of Hindus (includes scheduled castes and scheduled tribes) undertake agricultural occupations as opposed to only 36 per cent of Muslims. However, 22 per cent of rural Muslims arc engaged in non-agricultural and non-labour occupations as opposed to only 12 per cent Hindus. Nevertheless, over one-third of both religious populations work as agricultural/ manual labourers in rural areas. In urban areas, whereas 47 per cent of Hindus (includes the scheduled castes and scheduled tribes) are employed in regular wage/salaried occupations in organised sector (Table 5) only 29 per cent of Muslims are thus employed. Majority of Muslims are self-employed, lor example, 53 per cent of them as opposed to 36 per cent of Hindus are self-employed in urban areas. However, per cent of all the religious categories work as casual labourers in urban areas. Information on type of employment and sex specific work participation rates are difficult to find. This information in three broad categories of employment, namely, selfemployed, regular workers and casual workers was made available in the NSS 43rd round publications. 'Self-employed', are those who earn their livelihood by working on their own traditional occupations and enterprises including own agriculture. 'Regular workers' are identified as those who work on long-term salaried jobs. All the modern public and private service sector jobs are included in this category casual workers are those who work for wages on a day-to-day basis. Table 6 presents structure of employment and 'work participation rates' (WPR) for various religious groups, by sex and residence (i e. whether living in rural and urban areas). Note that the WPRs for males in both rural and urban areas for all religions are fairly high The male WPR is marginally higher for all religious groups in rural areas. The differentrals between religious groups for males is also marginal. The female WPRs arc substantially lower in both rural and urban areas. Besides, the disparities in WPRs between religious groups are substantial. For example the WPR among urban Muslim females is as low as 11 per cent while this rate is 16 per cent for Hindus and 24 per cent for Christians. Similarly the rural WPR for females is only 20 per cent for Muslims, while it is 34 per cent and 37 per cent for Hindus and Christians respectively. Apart from the lack of work opportunities for females, the cultural factors such as practice of purdha and female seclusion might have affected the female WPRs in India. When we take a look at the distribution of employed persons by category of work it is interesting to note that over 53 and 60 per cent of male and female Muslims, respectively, in urban areas and 59 and 68 per cent of males and females, respectively, in rural areas are 'self-employed', This percentage is substantially lower for both Hindu and Christian males and females in urban areas. On the contrary Hindus' and Christians' share in 'regular-workers' category is substantially higher for both males and females in urban areas. Muslims' employment in regular wage/salaried jobs is marginal. This presents a very dismal picture especially for Muslim females since their WPRs are very low to start with and of them only an insignificant proportion is in regularly paid jobs. One possible reason for this may be the lack of secular education among the Indian Muslim females (see Tables and a discussion below). In the 'casual workers' category Muslims' share of employment is comparable to other religious categories. OWNERSHIP OF LAND Distribution of land ownership describes the intensity of resource base especially of a rural population dependent upon agriculture. Table 7 presents land ownership data for the all-india rural population. Excepting the 'other religions category (which is a relatively smaller proportion of the total population) relative landlessness is more among Muslims. As opposed to only 28 per cent of Hindus 35 per cent Muslims are landless. Hindus also have a relative advantage in owning land in large size class. For example, while one-fifth of Hindus own five acres or more of land, only one-tenth of Muslims belong to this category. RELATIVE LEVELS OF LIVING The NSSO is known to provide data relating to the monthly per capita expenditures (MPCE) for both rural and urban populations of India. The MPCE is considered a better index of levels of living than the monthly per capita income GDP figures. Distribution of population according to MPCE presents a reasonable picture of the relative levels of living. Such information is given in Tables 8 and 9 for rural and urban Source: NSS 43 Round, (Scheduled-10). areas, respectively. There are substantial differences in levels of living between rural and urban areas and among the different TABLE 7: DISTRIBUTION OF PERSONS BY SIZE CLASS OF CULTIVATED LAND AND RELIGION : RURAL INDIA Economic and Political Weekly November 18,

4 religions in urban India. The relative difference between the Hindus and Muslims is marginal in rural areas, but the difference is considerable in urban areas. For example, while 53 per cent of Muslims fall into the category of less than Rs 160 MPCE, this percentage is only 36 among Hindus. At the higher per capita expenditure classes the proportion of Hindus is considerably larger. It may be recalled that most of the Muslims in urban areas are self-employed as opposed to a majority of Hindus who are employed in the organised/salaried sector. It is likely that the levels of income generated by selfemployment is substantially lower than the average income originating from the organised sector. The Christian population both in rural and urban areas has a relatively better distribution of the MPCE. RELATIVE LEVELS OF EDUCATION Levels of literacy and percentage of younger population continuing in educational institutions are considered good indicators of social development. Tables 10 and 11 present information on levels of education achieved by males and females according to religious categories. It is revealing that in rural areas Christians are by far the most educated. While illiteracy among Muslim men is 58 per cent, it is 51 per cent among Hindus and only 34 per cent among Christians. For females, respective percentages of illiteracy are 76, 75 and 43. Christians retain their lead even in higher education categories, for example, 9 percent of Christian males and 8 per cent of females are secondary educated, whereas these figures for the Hindus (includes the SCs and STs) and the Muslims respectively are 5.7 and 1.7 per cent, and 3.4 and 0.8 per cent. Further, while non-negligible proportion of Muslim females and only 0.6 per cent of Muslim males are found in graduate and above category, these proportions are 0.2 and 1.2 per cent for Hindus, and 1.5 and 1.8 per cent for Christians. Relative to rural areas, in urban areas the literate constitute a fair proportion among all the religious categories. The male illiterates in urban areas are only 19 per cent. 25 per cent and 42 per cent for Christians. Hindus and Muslims, respectively, and female illiterates are 23 per cent, 42 per cent and 59 per cent, respectively. Literacy rates at the level of secondary education in the same order are as follows: 20 per cent. 17 per cent and 8 per cent for males, and for females, 21 per cent. 11 per cent and only 4 per cent. Higher educated proportions are relatively better among Christian and Hindu males. 8 per cent each, but only 2.3 per cent of Muslim males are reported to be graduates and above. This achievement among females is 5.5 per cent among Christians, 4.2 percent for Hindus and only 0.8 percent for Muslims. Tables 12 and 13 present data on continuing education by religious categories for rural and urban areas. Education continuation rates both in rural and urban areas are better among Christian males as well as females. The enrolment and continuation rates are also better in urban than in rural areas for all religious categories. However, at each level and in both rural and urban areas Muslim rates are substantially lower than Hindus Economic and Political Weekly November 18, 1995

5 For example, in age group in urban areas, 56 per cent Hindu boys are continuing education as opposed to only 35 per cent among Muslims. These percentages are 42 and 19 for females. For age groups 20 years and over, the respective percentages are 323 and 12 for Hindus and 10 and 15 for Muslims. In rural areas the scenario is similar but at much lower levels for all religious categories. Thus far we know that in terms of all the socio-economic parameters discussed above the Muslims are relatively worse-off than the majority Hindus which includes the SCs and STs and the Christian population in both rural and urban areas in India. In the following some evidences on the demographic aspects according to major religions in India are presented. Demographic transition theories have informed that the socio-economic factors influence the demographic parameters to a large extent. Accordingly socioeconomic improvement of a group or a region would depress fertility, reduce mortality leading to an overall improvement in the quality of life. Kerala is the lowest. On the other hand the q2 for rural Muslims in almost all states has been fairly lower than for Hindus and in many states lower than for Christians, In urban areas as well, with the noteworthy exceptions of Uttar Pradesh, Bihar and West Bengal, the q2 for Muslims is lower than the Hindus. A combination of a positive growth of population and a low infant mortality rate (also low death rate) might have enabled a growth rate of population as a whole. Nevertheless, one wonders as to what could be the possible reasons for low child mortality among Muslims while the fertility rates are supposedly higher, and that they are also socio-economically worse-off. It may be possible to provide valid reasons for such a paradox through a three way classification and also through multivariate analytical methods. Such data arc not yet made available by the censuses. However, some relevant data of this son is published by the Operations Research Group, Baroda who have been undertaking National Family IV Demographic Differentials and Future Growth of Population FERTILITY AND MORALITY DIFFERENTIALS A comparative picture of the household size for three religious populations according to place of residence is presented in Table 14. Overall the household size for Muslims both in rural and urban areas is higher than that of the Christians and Hindus for all states and for all India. A larger household si/e among the Muslims suggests; (a) a relatively higher practice of living in joint family situations, and (b) that actual family si/e per couple is also relatively high among Muslims. Various census publications contain data which present differentials in fertility and mortality of religion. Tables 15 and 16 contain this data for 14 major states in India. General marital fertility rate (GMFR) presents the number of children born in a year per thousand married women in the age group 15-49, The total marital fertility rate ('I'M FR) presents the total number of children expected to be born to a married woman during her whole reproductive period starting at age 15 and ending at around 50 years. The child mortality rate (q2) presents number of childen who are born alive but dying before their second birthday per thousand live born children. Since the child mortality depend upon the socio-economic development of a group, it is considered a composite and summary index of development and standard of living. The data presented on the above indices suggest that both in the urban and rural areas in almost all states, the GMFR and TMFR are consistently higher among the Muslims when compared with the Hindus and the Christians. Fertility among the Christians of Economic and Political Weekly November 18,

6 Planning Surveys at intervals of about a decade since Distribution of the mean number of live births (MNLBs) and proportion surviving according to religion, education and income are presented in Table 17. These figures are the national average from a survey of about households spread overall the states and union territories found in the mainland India. Mean number of live births for the country as a whole is Overall the Hindus have about 13 percent lower MNLBs than the Muslims. Women's education has a negative effect on fertility among all the religious categories. However the fertility differential between the Hindus and Muslims still remain. A fact to be noted, however, is that while the fertility among the well educated (secondary level and over) Hindus was 31 per cent lesser than the illiterates, among the Muslims also this differential was as much over 27 per cent. Besides, with an increase in education of women, fertility differences between the Hindus and the Muslims decreases considerably. This suggests thai educating women will depress fertility irrespective of religious affiliation. TABLE 18: CURRENT CONTRACEPTIVE USERS ACCORDING TO RELIGIOUS CATEGORIES ALI INDIA, (Millions ) Similarly an increase in levels of family income depress fertility albeit of lower magnitude among all the religious categories (Panel ii, Table 17). Nevertheless, the effect of education decreasing fertility levels both among Muslims and Hindus is substantially larger than the depressing effect on fertility of increase in levels of income. The survival status of children is relatively better among the Muslims (Panel iii, Table 17), and this relative advantage is higher in lower income levels. An increase in the level of incomes will lead to improved survival among the Hindus. However, an apparent 4 per cent advantage in survival among the Muslims seems to be marginal. Use of Modern Contraceptives The use of modern contraceptives is marginally lower among the Muslims when compared with the Hindus. However, the usership rates are substantial than what one is given to understand or generally talked about. For example, while the proportion of the Hindus is 85.2 per cent of sample population (Table 18), their percentage in contraceptive users is 86.1 Whereas, the Muslims constitute 10.6 per cent of the sample population, their share in the eontraceptors is only 8 per cent. Further, the regular users of contraceptives are 50.3 of all eligible women among the Hindus and this percentage is 40.7 percent for Muslims. Although the usership rates are relatively low, Muslims are not averse to acceptance of family planning as it is normally portrayed in the print media. Panel (i) of Table 19 presents change in the family planning usership during according to religious categories. From a developmental point of view it is interesting to note that the family planning use which was only 14 per cent and 9 per cent for the Hindus and Muslims, respectively, during 1970 has increased to 46 per cent and 34 per cent, respectively, in 1988, The percentage change during was 22 and 14 percentage points for the Hindus and the Muslims, respectively, but during this percentage change was 10 and 11 respectively, relatively more among the Muslims. Panel (ii) of Table 19 presents change in the inter-spousal communication of family planning among the religious groups. It is noteworthy that this change has been highest among the Muslims, recording a 20 percentage point higher than in 1980 which is more than an increase among the Hindus which is only 16 percentage points. However, as far as Muslims are concerned, the trend in acceptance of family planning is encouraging, indeed the percentage change is relatively larger than the Hindus between This evidence allays a common allegation that Muslims are against the use of family planning. Indeed the evidence is to the contrary. Future Growth of Muslim Population The available evidence suggests that in spite of declining fertility and mortality for the population as a whole and also lor all its religious components, the relative fertility is higher and relative child mortality is lower among the Muslims compared to other religious populations in India. Does this mean that over a period of time the Muslim population would exceed the Hindu population in India? An answer to this can be found in simulation exercises undertaken in a resarch conducted at the Gokhale Institute of Politics and Economics, Pune [Mistry 1990]. Table 20 reproduces the results of this simulation exercise. It can be seen that if the base population of minority community is 10 per cent, and if the differential in the exponential growth rate is (rate of growth of total population- r 1 minus rate of growth of minority population - r 2 ) then the percentage of minority population will be 12 per cent in 50 years duration, 14.2 per cent in 100 years, 16,8 percent in 150years Economic and Political Weekly November 18, 1995

7 19.8 in 200 years and so on. Note that these percentages are achieved only if the growth rate differential remains unchanged over a 200-year period. But normally the growth rate differentials do not remain static and the tendencies are towards a secular equalisation of growth rates among all the groups of human populations. However, the present Hindu-Muslim fertility and population growth rate differentials are not large and are likely to narrow down in future [Mistry 1990], Besides, any minority can become a majority only when three conditions are jointly met; that the minority constitutes a significant proportion of the total population, the population growth rate diffential is large and the duration is substantially long. Since Muslims constitute only about 12 per cent of the total population and that the growth rate differential is about 0.005, at the national level it is not possible that the Muslims can outnumber the Hindus in India for a considerable period of time to come. V Conclusions This paper reviews some socio-economic and demographic characteristics of population according to religious affiliation in India. The Hindus are the significant majority in all states excepting in Jammu and Kashmir and Punjab, where Muslims and Sikhs respectively outnumber them. Among the 14 major states, the Christians are found in substantial percentage only in Kerala. It is found that the growth of Muslims is higher than that of the other religious populations. Muslims are found to be dwelling relatively more in urban areas than the other sets of people. Muslims are mostly self-employed and their share in regular paid jobs is low. Hindu population is relatively belter employed in regular employment in urban areas. The work participation of Muslim females is externely low. The land holding is relatively better among the Hindus than the Muslims, and Muslims work on non-agricultural occupations in substantial proportions in rural parts of India. Muslims by far are least educated when compared with Hindu and Christian populations in India. The scenario is same if one studies the student population both in rural and urban areas as well. Muslim women are the least educated of all the possible categories discussed in this paper. Although the MPCE compares well between the Hindus and Muslims in rural areas, it is highly lopsided favouring Hindus in the urban areas. On the whole the Muslims are socio-economically worse off in all parts of the country. The levels of fertility as measured by GMFR and TMFR are higher for Muslims in both rural and urban areas. But the child mortality among both the urban and rural Muslims is comparatively low. A combination of a positive growth of population and a low child mortality seems to have enabled a marginally higher growth of Muslim population in India. More recently, however, there has been a noteworthy increase in the acceptance of family planning methods among the Muslims. In fact the percentage change in the family planning use is marginally higher than for the Hindus. The inter-personal communication regarding the family planning use has also increased substantially among the Muslims in India. There, however, seems to be a paradox. Muslims in India are relatively poorer than the other populations, they have recorded relatively low levels of MPCEs, do not work as regular wage earners, are mostly illiterates, have relatively high fertility but contrary to the expectation they maintain a low child mortality. A question arises as to what could be the possible reasons for low child mortality among Muslims while the fertility rates are supposedly higher. Social scientists and demographers need to present a reasonable answer to this contradictory demographic phenomenon. A simulation exercise reproduced in this paper suggests that the Muslim population will not overtake the Hindu population in India during a foreseeable future. The recent type of the communal and anti-secular minded opinion being expressed that Muslims soon will outnumber Hindus in DISCUSSION CANDOUR has a way of cropping up in the most unexpected places. In a Hindustan Lever ad, for instance, that I saw in the EPW a few years ago, the message was something like this; we do not sell products, we sell feelings (cosmetic feelings, the ad might have said, but didn't) like beauty, self-confidence, freshness, vigour, etc. More recently there was, once again in the EPW, an article by Paul Brass (September 9) in which, while making a case for increased access to sensitive areas for American academics. Brass referred to the "nearly two centuries long political drive by upper class elites to establish a 'national culture' for India based on their own values and interpretation of Indian history". The sub-title of the article, 'Virtue Unrewarded', nicely sums up one of its themes, and nothing further needs to be said in this matter, except that il would be interesting to know whether the sub-title was inserted by the editorial staff of the weekly, or Brass had himself put it there, 'up front'. But the case for liberalised access was not based solely on virtue. Other things figured as well, like commerce, and the need to free academics from the 'Four Faces of Indian Bureaucracy', Such a proliferation of India does not stand the lest of the scientific and academic enquiry. This paper, however, concludes by emphasising the need to strengthen the data base which would allow a study of ethnic and religious differentials in socio-economic and educational achievements, The Census of India can take a lead in this direction by publishing the cross tabulations of literacy by religion and distribution of workers by religion and industry at the level of India, states and various districts. Further, there is a need to undertake focused micro-studies to clearly understand the nature, type and magnitude of socio-economic problems faced by various communities in rural and urban India. References Mistry. M B (1990) 'Hindu-Muslim Fertility Differentials in India'. M Phil Dissertation, Gnkhale Institute of Politics and Economics. University of Poona. National Sample Survey Organisation (1990): Sarvekshana: Results of the Fourth Quinquennial Survey on Employment and Unemployments NSSO, Ministry of Planning, Government of India. New Delhi Operations Research Group (1990): Family Planning Practices in India: Third All India Survev Operations Research Group. Baroda. Candour about the Twice-Born Sudhanshu Ranade arguments is, of course, only to be expected. As John Lewis put il once, once you have decided, for some reason, to support a cause like, say, human rights in China, it makes sense to broad-base yourself by adducing all sorts of other reasons as well The example is mine: Lewis's reference at that time was to poverty alleviation in the days when the World Bank, under the stewardship of former secretary of state McNamara, was espousing this cause of poverty alleviation 'directly' rather than 'indirectly' as at present. But let us return to the point about 'upper classelites' attempting to establish 'anational culture in their own image'. The quoie on this was in the article, immediately preceded by a statement that those studying the glory of India's past were being allowed tree access because their output 'feeds into' the hegemonic enterprise. A sound move, from the point of obtaining additional support from liberal circles in India for lifting restrictions on access. And, in an apparent attempt at leaving no stone unturned while reaching out to the already anguished hearts in this lobby, Brass immediately followed up the 'national culture' formulation with the warning that "time may yet be coming... Economic and Political Weekly November 18,