SPECIAL ARTICLEEconomic & Political Weekly EPW octoBER 11, 200849Understanding the Paradox of Changes among Dalits in PunjabParamjit S Judge, Gurpreet Bal A study of dalits in two districts of Punjab reveals that ghettoisation remains common (including in urban areas). While the traditional caste occupational structure has changed, this is less so among dalits in rural areas. Caste endogamy remains the norm. The study shows that casteism is powerfully embedded in the collective consciousness. The caste system is oppressive due to discrimination, exclusion, exploitation and untouchability, but at the same time it situates the dalits in the system as a collective identity. Their desire is to end all kinds of discrimination, oppression and exclusion, but not the caste system in its entirety. This is an instance of a deep-rooted internalisation of a world view. This paper is based on interviews of 1,600 dalit respondents randomly selected in Amritsar and Jalandhar districts of Punjab.Since independence, the forces that have been unleashed to ameliorate the conditions of the dalits in India include con-scious efforts on the part of the government through con-stitutional provisions, reservation policies and special develop-ment programmes. Besides, there have been numerous instances of the efforts by the dalits through political mobilisation towards improving their conditions. All these efforts towards changing the socio-economic conditions of the dalits for the better are des-tined to have a bearing on the caste system as such. This paper is an attempt, based on empirical investigation, to understand the paradox involved in the changes that are taking place among the Punjabi dalits. The main objective of this paper istoexamine whether at the empirical level all these forces have made an impact with regard to the conditions of the dalits in Punjab. Equally important is to see whether the caste system has been eroded at different levels of the lives of the dalits. Ghurye (1969) identifies the following features of the caste system: segmental division of society, hierarchy, restriction on feeding and social intercourse, civil and religious disabilities and privileges of the different sections, lack of unrestricted choice of occupation and restrictions on marriage. In certain respects, the present study is linked to the identification and examination of certain dimen-sions of the caste system that are expected to change.In the specific context of Punjab, with dalits constituting 28.85 per cent of the population of the state, attempts have been made to identify certain key dimensions of change in the socio-economic conditions of the dalits [Jodhka 2002 and Judge 2004]. Based on the empirical research some dimensions have been examined by these two studies, namely, social ecology, religious places and economic conditions.1 Equal attention is needed for the investi-gation into the issues that are linked with the way the dalits construct their orientation towards change with regard to caste endogamy and caste identity. It may be added that the caste sys-tem and casteism are equally important in impeding and faci-litating certain basic changes in the system. The caste system is marked by political and economic asymmetry and cultural plu-rality, whereas casteism is the system of ideas and principles that justify the system. As a system of justificatory ideas, casteism is embedded in the mind and the world view of the people and maypersist longer than the changes in the caste system through education, occupation and social ecology. Keeping in view the central features of caste, the following aspects have been exam-ined here to map the changing (and not so changing) contours of Punjabi dalits: social ecology, hereditary occupations, religious places, caste endogamy and caste identity. The present paper is a part of the report submitted to the University Grants Commission that funded the project ‘Education, Empowerment, Emigration and Entrepreneurship: A Sociological Study of Mobility among the Dalits in Punjab’. The authors wish to acknowledge the assistance provided by Bhupinder Thakur, Gurpreet Singh and Manoj Birdi as research staff. Critical comments of M Rajivlochan are gratefully acknowledged.We are grateful to the referee of this journal for giving useful comments as a result of which we have been able to improve the paper tremendously.Paramjit S Judge (paramjitjudge@yahoo.co.uk) and Gurpreet Bal (gbal.judge@gmail.com) are at the Department of Sociology, Guru Nanak Dev University, Amritsar.

SPECIAL ARTICLEoctoBER 11, 2008 EPW Economic & Political Weekly50Before proceeding to a discussion of the above-mentioned as-pects of the changing and perpetuating conditions of the dalits, it is important to briefly state the methodology of the study. The study is based on the interview of 1,600 dalit respondents randomly selected in two districts of Punjab, namely, Amritsar and Jalandhar. An equal number of the sample (400) was drawn from each of the four units that were identified as social types, namely, Amritsar city, Amritsar villages, Jalandhar city and Jalandhar villages.21 SocialEcology The social organisation of physical space is one of the most im-portant dimensions of understanding change in a society. The notion of the ghetto basically emerged in the Jewish context in which it was understood that the Jewish population was confined to overcrowded localities. There are historical reasons for the emergence of the phenomenon and in the post-Holocaust period, the situation has somewhat undergone a change with regard to the Jews. The ghet-tos, however, have not disappeared. It has been found that ethnic communities and blacks live in ghettos. Two reasons are as-cribed to the process, namely, a low in-come level and sociocultural/racial dis-tinction from the dominant population of the community. The latter is so strong that it tends to become the only reason for ghettoisation. Aninformalnetwork develops in which the dominant com-munity creates a system of checks to ensure the denial of entry to the “others” as residents in the locality. In other words,the social organisation of space occurs in such a manner that the economic dimension in the determina-tion of physical space gets marginalised.3 It is important to mention that few attempts have been made to understand the social organisation of the urban physical space with regard to the dalits. One of the ways of understanding the social organisation of space is to know the caste background of the neighbours. It is important to begin our discussion on the neighbourhood patterns. Table 1 shows the caste background of the two neighbours of the respondents. It may be pointed out that the respondents were asked to reveal the caste background of the neighbours residing on both sides of their houses.It is evident from the data that the respondents have been pre-dominantly living in a neighbourhood where the people of their own caste reside. The fact that draws immediate attention is that even the cities are not free from caste-based neighbourhoods.4 The social ecology of the urban areas acquires importance be-cause of the general understanding that cities are the modern centres where caste distinctions do not operate in the market sit-uation. Possession of capital and goods for exchange becomes the most crucial basis of interaction. So far as the market is con-cerned, similar principles may operate, though the western experiences of immigrants reveal that even the market situation could create exclusionary practices on the basis of birth-ascribed features. Influenced by the city life in the western countries and repulsed by the Indian village life, B R Ambedkar did not only create a critique of the village republic the way it was constructed by the British administrators [Moon 1989], but also asked the dalits to leave the villages and go to the cities [Virdi 2004]. How-ever, as Judge and Bal (2005) have argued, the migration of the Punjabi dalits neither followed any such perception that their conditions would undergo change nor was there any increase in migration rate. Even now the proportion of the dalits in urban Punjab is less than that of villages. As a matter of fact the expand-ing cities have enveloped the villages that made many of them city-dwellers. The city was to them a place where the market situ-ation could be clearly separated from the residential space. In the case of residential spaces the concentration of dalits in certain localities is obvious. It is evident from the data in Table 2 that in the case of 18 wards in Amritsar and 21 wards in Jalandhar the percentages of the scheduled castes (SCs) are above their population per-centage in the state ranging from 29 to 85 per cent.There is another fact that comes to the fore in the case of all the localities in both the cities. In Amritsar, balmikis in Gowal Mandi, mazabis in Guru Ki Wadali and meghs in Haripura locali-ties are in an overwhelming majority. In Jalandhar ad-dharmis/chamars in Boota Mandi and Mithapur, balmikis in Ali Mohalla and meghs in Bhargo Camp are living in their own caste localities. In such a situation in which social ecology corresponds with the caste background, the classical notion of the city is seriously challenged. The political economy perspective would lead us to the understanding that such a situation could have emerged due to the overlapping of class and caste positions. It may be partially true. But it was seen that even in the case of eco-nomic mobility, as is illustrated by the ad-dharmis of Boota Mandi,socialecology is marginally altered. The prosperous ad-dharmi respondents had then started living in the adjoining GuruTeghBahadur Nagar in palatial houses, but the locality was still mainly inhabited by the ad-dharmis. Sharma (2003) in his study of a small town in Rajasthan has shown how the social organisation of space is closely associated with the caste. There is a likelihood of a small town showing the ecological segregation of castes. The interesting aspect is that all the big cities of Punjab like Amritsar, Ludhiana and Jalandhar have the same kind of pattern. Such a pattern organises the social space in such a way that the perpetuation of caste identities seems highly probable. At the same time, such aggregation of people of one caste at one place facilitates the mobilisation process. The villages hardly show any changed scenario with regard to the settlement pattern. As is evident from Table 1, the dalits continue to have segregated houses. In the villages of Amritsar Table 1: Caste Background of Two NeighboursCaste Status Urban Rural TotalSame caste 736 (92.00) 766 (95.75) 1,502 (93.88)Both higher 7 (0.87) 15 (1.88) 22 (1.37)One same, one higher 48 (6.00) 18 (2.25) 66 (4.12)One high, one lower 9 (1.13) 1 (0.12) 10 (0.63)Total 800 (100) 800 (100) 1,600 (100)Figures in brackets are columnwise percentages.Table 2: Number of Wards according to the Percentage of the Scheduled Castes in Amritsar and Jalandhar% of the Amritsar JalandharScheduled Castes Number % Number % of Wards of Wards of Wards of Wards0-5 16 25.81 6 10.915-10 1117.7447.2710-20 1320.971832.7320-29 4 6.456 10.9129-50 1625.811323.6450-85 2 3.228 14.55Total 6210055100Source: Census of India, 2001.

SPECIAL ARTICLEEconomic & Political Weekly EPW octoBER 11, 200851district the dalit locality is called ‘thathi’, whereas in the Jaland-har district it is called ‘chamarhli’ [Judge 2004]. It is already argued by many scholars that the location of the dalit colony is towards the southern-western direction of the village.2 HereditaryOccupations Occupational mobility is an indicator of both economic and social mobility. Sociologists take occupation, education and income as the determining variables of the class position of the individuals and families. Occupations do not change in a makeshift manner, as there is a definite relationship between economic change and the emergence of new occupations that require certain skills. Capitalism/industrialisation is credited with creating occupational diversification that broke the traditional here-ditarycharacter. In India, caste and occupation had close proximity to the extent that even the caste names could reveal the nature of hereditary occupation. For example, chamar, mochi, bhangi, lohar, nai are some of the caste names that are the names of the occupations as well. In the existential world of village India, the caste name invariably invoked the occupation. For example, jat meant peasant, saini meant vegetable grower, julaha meant weaver, and so on. Obviously, the arrival of capitalism with new occu-pations that required education was destined to break the corres-pondence between caste and occupation. There has been a lot of diversifica-tion of occupations among the dalits, which is clear from the fact that the dalit respondents are involved in as many as 46 occupations in both the ur-ban and rural areas. However, every-thing has not changed. As is evident from the data in Table 3, 28.75 per cent in the cities and 49.5 per cent of the dalits respondents in villages are still involved in the low status occupations. In the case of male respondents, sweep-ing and agricultural labour, and in the case of female respondents and male respondents’ wives, to be a maidservant or domestic help is the predominant occupation. It was also noticed that many of them tried to conceal their occupa-tion. Table 3 also shows that among the respondents 17.12 per cent in urban areas and 9.66 per cent in rural areas do not have any occupation. A probe into the matter reveals that they had actually concealed the fact that they were either sweepers or maidservants. Understanding the change and persistence in the occupations of the dalits depends upon how we look into the matter with re-gard to the occupational characteristics of the respondents. One way of understanding is that there has been a definite change in occupations as there is some percentage (about 6 per cent) of re-spondents occupying prestigious occupations as they are indus-trialists, big traders, large landowners, etc, whereas the other would be that some of the respondents are still located on the lowest occupational rungs. It is also plausible that there is a class formation among the dalits. Another way of looking at the empirical reality is that change is taking place in a gradual manner. It may be pointed out that dalit castes have also been at the bottom of the class structure. They were involved in polluting occupations. There is a difference between caste and class mobil-ity. One is related with status enhancement, whereas the other is economic. In the case of dalits the two were combined and it could be anticipated that if the dalits stopped doing defiling work their status could change. Such a change has occurred in alimited fashion though the nexus between caste and occupa-tion has been fractured. 3 ReligiousPlaces All the respondents, irrespective of their rural or urban backgrounds, were assertive with regard to one query, namely, whether they were allowed to enter the temple/gurdwara belon-ging to upper dominant castes. What was that assertion? There is no restriction on their entering any temple/gurdwara. Rather many balmiki respondents in both Amritsar and Jalandhar cities laughed at the query and declared that “nobody could dare stop us from entering a temple”. However, the dalits of Punjab are struggling against the upper dominant castes, particularly the jat Sikhs, on issues concerning the religious places. The issue of reli-gious places has historically evolved in two phases. First, it was related with the right to worship in the temple. Ambedkar led dalits to make a forced entry in the tem-ples in the initial stages of his struggle for equality of worship [Zelliot 1970]. The second stage began when the dalits either had their own religious places or they started “reclaiming” their religious places. So far as the issue of entry into religious places is concerned, it has almost been settled. Jodhka (2002) and Judge (2004) have noted the practice of not allowing dalits to enter the temple in Patiala district. But, such instances in Punjab are rare.The issue of religious place is an im-portant dimension of intercaste interaction and caste system. One dimension of religious place is what Jodhka (2002) has brought out in his study, wherein he noted that in 51 villages the dalits had their own gurdwaras though the number of such villages could be much more. It has been pointed out that there are 10,000 dalit gurdwaras in Punjab [Webster 2007]. The implication of this finding could be very interesting. One of the interpretations of this empirical fact is that the dalits are not allowed to enter the gurdwara and in protest they have constructed their gurdwaras [Jodhka 2002 and Bal 2007]. The practiceofuntouchabilityat religious places among the Sikhs is almost absent in the Doaba region of Jalandhar, Hoshiarpur, KapurthalaandNawanShahar districts of Punjab.5 There is a widely held belief that there is caste hierarchy among the Sikhs though the caste system lacks rigidity that is prevalent in other parts of the country. The existence of the practice of untouchability is invariably denied.There is another way of understanding the issue of the sepa-rate gurdwaras. Let us look into the historical and anthro-pological dimensions of the religious places and their caste-linked Table 3: Categorywise Occupational Levels of the Respondents Occupational Prestige Category* Urban Rural TotalI 25 (3.12) 16 (2.00) 41 (2.56)II 46 (5.75) 9 (1.12) 55 (3.44)III 8 (1.00) 1(0.12) 9 (0.56)IV 58 (7.25) 20 (2.20) 78 (4.73)V 171 (21.38) 147 (18.38) 318 (19.69)VI 124 (15.50) 133 (16.62) 257 (16.06)VII 231 (28.88) 396 (49.50) 627 (39.18)VIII/NA 137 (17.12) 78 (9.66) 215 (13.43)Total 800 (100) 800 (100) 1,600 (100)*These occupational categories have been constructed on the basis of D’souza (1985). These categories have been arranged in descending order in terms of prestige of the occupation.Figures in brackets are columnwise percentages.

SPECIAL ARTICLEoctoBER 11, 2008 EPW Economic & Political Weekly52existence in Punjab, which, metaphorically, could be stated as “a religious place of one’s own”. Historically, the social structure of a Punjab village typically followed the all-India pattern in which the entire social organisation revolved around agriculture. In the present Punjab, prominent castes of landowners are jats, sainis, lobanas, kambohs, mahatons, khatris and brahmins. The numeri-cally preponderant and politically dominant caste among them in Punjab is that of the jats. Since more information is available on the jats, it is appropriate to select them for an illustration of the entire issue. The jat landowner traditionally had a large number of service castes that were related to him under the “Jajmani” system. As the entire village community depended on agrarian production for its existence and livelihood, the land-owner occupied the central position so far as the economic life of the village was concerned [Judge 2002]. All other castes were limited in number except for the agricultural workers who invari-ably belonged to the dalit castes. This fact of the existence ofthe service castes and their proportion in the population of village has been noticed by Dirks (2002: 76) in India in general, “No vil-lage contained more than a few of these families…”. The con-struction of a religious place in such a situation could only be done under the patronage of the landowning caste. Most of the villages in the Doaba region in particular are with-out any temple despite the presence of the Hindus, whereas all the jat Sikh villages have gurdwaras. Over a period of time, the number of gurdwaras in the villages has begun to rise. There is a general impression that there is a caste basis for the gurdwaras. Each caste tends to have its own gurdwara. Such an observation is partially true. To understand the reality underneath, one may raise a question as to why the jats have more gurdwaras if the logic of caste basis is taken into consideration. As a matter of fact there is a gurdwara for each locality called ‘patti’ in Punjab. The dalits having their separate gurdwara is a recent phenomenon. The green revolution process brought changes in the Punjab economy and society. The arrival of migrant labour pushed the dalits out of the agrarian relations, as they were displaced as agricultural labourers due to which they had to go to the cities to look for work in the non-agricultural sector. The conditions of the dalits relatively improved and they diverted their attention to the affairs of god. The balmikis constructed their temples, ad-dharmis their Ravidas temples and ramdasias and mazabis their gurdwaras. It should be pointed out that so long as a caste com-munityispoverty-strickenand its members are finding it diffi-cult to make both ends meet,itmay not be able to construct its templeorgurdwarainthevillage. In the case of the Sikhs it is quite probable that during the period of the Singh Sabha move-ment the construction of gurdwaras according to the locality and caste was discouraged due to its claim of equality among all the Sikhs. The conversion of jats to Sikhism in the Doaba region at the turn of the 20th century due to the proselytisation activities of the Singh Sabha movement leads us to raise a pertinent ques-tion, that is, what did happen to their religious places of worship? Since the jats as the landowning class had a Hindu past, it was expected that they might have constructed some temples. We know that so far as ancestor worship is concerned those places are still intact. However, the question of temples relating the jats to their Hindu past has not been settled. There are two pos-sibilities. One, there were no temples of any sort in the villages of Punjab. Second, the jats after their conversion to Sikhism appropriated the sacred placeaswell.So far as the Doaba region of Punjab is concerned, the above argument holds well, but the Majha region in which Amritsar district falls presents a different case. During data collection the mazabi respondents of Amritsar district reported caste-based exclusion of religious practices. Many of them said that the upper caste Sikhs did not allow them to carry the sacred book to their residence for purpose of performing various rituals/ceremonies on the pretext that their houses were dirty. Similar information was given by the mazabis of Guru Ki Wadali – an erstwhile vil-lage that has become a locality of Amritsar city. No such thing was reported in Jalandhar district. The respondents while asked about the religious places said that their villages had common gurdwara(s). It clearly shows that a certain degree of religious exclusion is practised in some parts of Punjab.It may be reiterated that a separate gurdwara does not imply that its existence is a consequence of exclusion. There are three kinds of gurdwaras of the Sikhs. First, there are historical gurd-waras. These gurdwaras are linked with the Sikh gurus or cer-tain events in the Sikh history. The Golden Temple in Amritsar is one of the examples of historical gurdwaras. The second type of a gurdwara is what have now become very well-known reli-gious places, namely the ‘deras’.6 The deras are a part of both the Hindu and the Sikh religious traditions of Punjab; and if we take the Sufi tradition into consideration, then it is possible to put even the places of the Sufi ‘fakirs’ as a part of the same tra-dition. In the early medieval period, there were deras of ‘Nath yogis’. Sufi saints had their own deras. After the emergence of Sikhism, the institution of deras could be found among them. The ‘nirmalas’ and the ‘udasis’ had their own deras. These deras have also come into conflict with the mainstream religious establishments. Recently, deras have also become places of contestation among different castes. The third kind of gurdwaras is called the social gurdwaras. It is here that the issue of untouchability or exclusion has emerged. In the 10 villages studied in Amritsar and Jalandhar districts, there was no historical gurdwara. A question was asked about the number of religious places and their responses are given in Table 4. It is interesting to note that Table 4: Number of Religious Places in Each VillageVillage Jat/KambohCommonMazabiHinduDeraSufiChurchBalmikiRavidas GurdwaraGurdwaraGurdwaraTemplePir’sTombTempleMandirGehri 1 1 2 1 1 Galowali 1 1 1 1 Bhoma 1 3 1 2 1 Kotli Dhole Shah 1 1 1 1 Nawan Pind 1 1 1 1 1 Udhowal 2 8 1 Adharman 1 1 2 1 Meham 2 1 2 1 1Bath 1116 Mehtpur 2 8 1 1Total 7 7 6 6 3 39 3 7 2(First five villages are from Amritsar district, whereas the rest five are from the Jalandhar district.)

SPECIAL ARTICLEEconomic & Political Weekly EPW octoBER 11, 200853there was a clear-cut distinction between the Amritsar and Jalandhar villages. In the case of Amritsar villages, the predomi-nant trend was to identify the gurdwara with the caste, whereas the villages in Jalandhar district had the predominance of com-mon gurdwaras. Notably, the mazabis and balmikis abound in terms of religious places though the largest number of religious places in villages was that of Muslim Sufi saints. The Muslims of the eastern Punjab migrated to the western part of Punjab after Partition. However, it may be noted that while mosques in villages have disappeared, the tombs of Sufi saints are well-maintained. It implies that the composite religious tradition of Punjab has not been completely eliminated by the efforts at constructing a single religious tradition.Finally, the existence of a distinct religious space is a different dimension from that of a separate gurdwara if the issue impli-cates the religious practices of the dalits. Comparing to the ideal typical constructions of the deprivation of dalits, the existence of a place of worship is a major transfor-mation in the conditions of the dalits. The issue of the gurdwara is complex because some of the respondents re-ported discrimination that comes close to untouchability. However, the con-struction of the gurdwara on the part of the Sikh dalits could be considered in terms of their ability to do so. Caste was always present and strong even earlier. It was not what Gupta (2007) thinks about the rising tide of caste consciousness, rather the change is with regard to the rising awareness of the dalit castes and its manifestation in the public sphere. The political economy of the religious place indicates that there is some improvement in the economic conditions of the dalits. The existence of separate gurdwaras for different castes presents an illustration of untouchability as well as of dalit mobility however limited it may be.74 Endogamy and Intercaste MarriagesCaste endogamy is regarded as one of the major citadels on which the caste system rests. The notions and practice of hypergamy and hypogamy existed in the Indian tradition, but these are more a part of mythological stories than reporting of actual cases. The combination of caste endogamy and the institution of arranged marriage have made caste almost a permanent system. Caste en-dogamy is also a result of a principle that has transformed into a mentality. Vovelle (1990: 12) defines the examination of mental-ity “as the study of the mediations and of the dialectical relation-ship between the objective conditions of human life and the ways in which people narrate it, and even live it”. In a way the notion of mentality implicates the combination of objective and subjec-tive conditions with a difference from traditional understand-ing of this dichotomy. The subjective is understood in termsof how it is narrated/told. It seems natural to think, expect and anticipate that the people belonging to the lower castes whose stake is in the demolition and destruction of the system would support intercaste marriages. Caste identity presupposes endogamy in spite of the fact that in a patriarchal system the identity of the male is more important than that of the female. The respondents were asked to give their views on the significance of caste endogamy if they thought that it was desirable. Table 5 indicates the views of the respondents with regard to caste endogamy.It is obvious that the dalit respondents living in villages over-whelmingly consider caste endogamy significant in comparison to the urban dalits. Less than 3 per cent of rural respondents stated that caste did not matter, whereas 29.12 per cent of the respondents in the urban areas viewed endogamy as insigni-ficant. In both the settings there was comprehensive support for caste endogamy. Tradition is a vague word and the caste system could be plausibly regarded as a part of Indian tradition. Since the rural respondents are supporting caste endogamy, it is clear that this view cuts across caste lines. On the other hand, a consid-erable proportion of the urban respondents regarded caste endo-gamy as not significant. Keeping in view this fact, it is proper to see the castewise responses of the urban respondents. The conjecture is that the ad-dharmis and meghs would be more tilted towards the insigni-ficance of caste endogamy than those of mazabis and balmikis. The data presented in Table 6 show clearly that among those who empha-sised traditions as the reason for supporting caste endogamy, the majority belonged to the mazabi, ad-dharmi/chamar and sansi castes. At the same time, among the ad-dharmi respondents one-fourth of them conside-red the issue unimportant. If we take a closer look at this table and see how each caste is distributed in terms of its emphasis on caste endogamy, then only meghs could be considered to be more amenable to marginalise caste endogamy. In that sense, our conjecture is partially substantiated. The second main caste that does not attach a crucial significance to caste endogamy is that of the balmikis. It may be concluded that the dalit castes consider caste endogamy significant as a part of the tradition.Related with the issue of endogamy is the event of intercaste marriage. The respondents were asked whether intercaste mar-riages could destroy the caste system. The responses are given in Table 7 (p 54 ). Again there are rural-urban differences in the sense that 96 per cent of the urban respondents are of the view that it is possible to end the caste system if intercaste marriages Table 5: Significance of Caste EndogamySignificance UrbanRuralTotalTradition 518 (64.75) 775 (96.88) 1,293 (80.82)Better understanding 20 (2.50) 1 (0.12) 21 (1.31)Family support 24 (3.00) 1 (0.12) 25 (1.56)Does not matter 233 (29.12) 23 (2.88) 256 (16.00)Do not know 5 (0.63) -5 (0.31)Total 800 (100) 800 (100) 1,600 (100)Figures in brackets are columnwise percentages.Table 6: Castewise Responses: Why Caste Endogamy Is Significant(Urban)Responses Caste Tradition Better Family Does Not No Total UnderstandingSupportMatterResponse between Spouses Ad-dharmi/chamar 150 (71.77) 2 (0.95) 4 (1.91) 53 (25.36) - 209 (100)Megh/kabir panthi 141 (57.08) 4 (1.62) 9 (3.64) 92 (37.25) 1 (0.40) 247(100) Balmikis 109 (56.48) 7 (3.63) 5 (2.59) 71 (36.78) 1 (0.52) 193 (100)Mazabi 108 (80.59) 7 (5.22) 5 (3.73) 13 (9.70) 1 (0.75) 134 (100)Sansis 3 (100) ----3 (100)Dumnas 7 (50.00) -1 (7.14) 4 (28.57) 2 (14.29) 14 (100)Total 518(64.75) 20 (2.50) 24 (3.00) 233 (29.12) 5 (0.62) 800 (100)Figures in brackets are rowwise percentages.

SPECIAL ARTICLEoctoBER 11, 2008 EPW Economic & Political Weekly54come into practice. Interestingly, most of the rural respondents have showed uncertainty with regard to intercaste marriages. They remain undecided about the issue. The urban respondents are more articulate with regard to the issue than the rural respondents. One of the reasons is that all the urban respondents have personal knowledge of such marriages due to which many of them think that such marriages never succeed or caste is deep-rooted.Anattempt was also made to further examine the urban respondents. Table 8 presents castewise data on whether in-tercaste marriages could destroy the caste system. Predict-ably, the ad-dharmis, meghs, sansis and dumnas constituted the major proportion of those respondents who were of the view that it was possible to end the caste system through intercaste marriages. Balmikis and mazabis were not sure of the proposition. In terms of education, as shown in Table 9, the responses are predictable. It is widely understood that education creates aware-ness. Obviously, educated respondents in the urban areas con-sider the proposition as a possibility. It may be further pointed outthat even the educated respondents are not wholeheartedly supporting the view. The issue of intercaste marriage has wider ramifications. In a society in which modern institutions provide space for interaction among people belonging to different castes, there is always a possibility of certain frequency of intercaste mar-riages. The issueis also related to how the other caste groups understand it. The following may not be a common occurrence,butsuchcases appear once in a while: Despitefilingawritin the Punjab and Haryana High Court for security, a couple – Ashwani Kumar and Baljeet Kaur – were murdered by the girl’s relatives – “punishment” for marrying against her family’s wishes.Ashwani, a Balmiki, and Baljeet, a jat, belonged to Rauli village falling under Mehtpur police station (The Sunday Tribune, July 15, 2007, p 1).The case refers to the area in Jalandhar district from where the sample of villages was drawn for the present study. In contrast to Amritsar district, the dalits of Jalandhar district, as we have seen, have done well in different walks of life. How-ever, the upper caste dominant jats do not think so. There is still another issue with regard to intercaste marriages. Can we look at the possibility of intercaste marriages among the dalits? The same rule applies here also. The caste hierarchy among the dalits prevents any possibility of breaking caste barriers through intercaste marriages[Judge2002].However, it is equally important to know the actual number of intercaste marriages in which at least one of the couple is from a dalit caste. Table 10 gives the data on the cases of inter-caste marriages in which dalits are involved. The data could be interpreted in various ways of which three seem plausible. Let us first look into the dominant pattern. A majority of cases (56.35 per cent) reporting intercaste marriages are that of between a dalit boy and a non-dalit girl. One of the reasons could be that since the dalit boys are socially more mobile than dalit girls, owing to the factor of patriarchy, the mobile and educated dalits can marry from the upper castes. An illustration of this could be given of the family of Steven Kaler inBootaMandi, Jalandhar who is a factory owner. The castes of the wives of his three brothers are brahmin, upper caste Jain and khatri. Secondly, in some cases the girls belong to the backward classes like nai and jheer, where the differences in the positions in the caste hierarchy could not have evoked a hostile response. Thirdly, a case of a non-dalit girl marrying a dalit boy is invari-ably controversial and the family of the girl opposes it. The conflicttends to keep the case alive in memory. On the other hand, a dalit girl marrying a non-dalit boy makes the girl invisible as she leaves the locality and there is not much reac-tion to such cases. After some time the case is forgotten. How-ever, it is clear that despite all preferences for cast endogamy and reactions against intercaste marriages, change is also taking Table 7: Whether Intercaste Marriages Can Destroy Caste Systems Response UrbanRuralTotalPossible 384 (96.00) 16 (4.00) 400 (100)Desirable 107 (52.45) 97 (47.55) 204 (100)Caste is deep- rooted 140 (92.72) 11 (7.28) 151 (100)Such marriages do not succeed 36 (92.31) 3 (7.69) 39 (100)Undecided 133 (16.50) 673 (83.50) 806 (100)Total 800 (50.00) 800 (50.00) 1,600 (100)Figures in brackets are rowwise percentages.Table 8: Castewise Responses to Whether Intercaste Marriages Can End Caste System(Urban)Responses Caste Possible Desirable Caste Is Such Marriages No Total Deep-rooted DoNot Response SucceedAd-dharmi/ chamar 125 (59.81) 25 (11.96) 17 (8.13) 9 (4.31) 33 (15.79) 209 (100)Megh/ kabir panthi 138 (55.87) 38 (15.38) 35 (14.17) 2 (0.81) 34 (13.76) 247 (100)Balmikis 71 (36.79) 28 (14.51) 52 (26.94) 9 (4.66) 33 (17.10) 193 (100)Mazabi 40 (29.85) 16 (11.94) 32 (23.88) 15 (11.19) 31 (23.13) 134 (100)Sansis 3 (100) ----3 (100)Dumnas 7 (50.00) -4 (28.57) 1 (7.14) 2 (14.28) 14 (100)Total 384 (48.00) 107 (13.38) 140 (17.50) 36(4.50) 133 (16.62) 800 (100)Figures in brackets are rowwise percentages.Table 9: Intermarriages as a Means to End the Caste System according to the Education of the Respondents (Urban)Responses Caste Possible Desirable Caste Is Such Marriages No Total Deep-rooted DoNot Response SucceedIlliterate 28 (7.29) 11 (10.28) 33 (23.57)11 (30.56) 30 (22.56) 113 (14.12)Literate 11 (2.86) 2 (1.87) 8 (5.71) 3 (8.33) 4 (3.01) 28 (3.50)Primary 55 (14.32) 21 (19.63) 24 (17.14) 5 (13.89) 25 (18.80)130 (16.25)Middle 59 (15.36) 16 (14.95) 36 (25.71) 4 (11.11) 21 (15.78) 136 (17.00)Matriculation 99 (25.78) 35 (32.72) 28 (20.00) 7 (19.44) 37 (27.82) 206 (25.75)+2 53 (13.81) 12 (11.21) 4 (2.86) 3 (8.33) 9 (6.77) 80 (10.00)College/graduate 53 (13.81) 7 (6.54) 5 (3.57) -5 (3.76) 71 (8.88)Postgraduate 7 (1.82) 2 (1.87) --2 (1.50) 11 (1.38)Diploma/ professionaldegree 19 (4.95) 1 (0.93) 2 (1.43) 3 (8.33) - 25 (3.12)Total 384 (100) 107 (100) 140 (100) 36 (100) 133 (100) 800 (100)Figures in brackets are columnwise percentages.Table 10: Incidence of Intercaste Marriages Involving DalitsCaste Status of the Couple Amritsar Amritsar District Jalandhar Jalandhar District TotalBoy dalit, girl dalit 13 (27.66) 3 (25.00) 8 (17.78) 3 (13.63) 27 (21.43)Boy dalit, girl non-dalit 20 (42.55) 7 (58.33) 32 (71.11) 12 (54.55) 71 (56.35)Boy non-dalit, girl dalit 14 (29.79) 2 (16.67) 5 (11.11) 7 (31.82) 28 (22.22)Total 47 (37.30) 12 (9.53) 45 (35.71) 22 (17.46) 126 (100) (100)(100)(100)(100)Figures in brackets are columnwise percentages.

SPECIAL ARTICLEEconomic & Political Weekly EPW octoBER 11, 200855place. However, the general response of the dalit respondents regarding intercaste marriage is negative.5 DalitIdentity The project of ending the caste system seems to have lost direc-tion in the face of tradition and resilience of caste. The defining features of caste clearly have showed that the attack on caste has to take place from all fronts if it is to be destroyed. If the caste differences do not disappear or wither away with all round ef-forts, then the next logical question is what is to be done? Is it possible to end a system that has been in existence for centuries with the stroke of a pen or micro-level efforts? One of the preconditions for the destruction of the caste system is the end of all caste-based identities. One of the numerous castes of Punjab is that of the jats. They are also the dominant caste in the villages. They are proud of their jat identity. Similarly, the graffiti written on four-wheelers like trucks in particular would show the arrogantjatdisplaying himself. The most popular graffiti is ‘putt jattande’(sonsofjats). No other caste group would find such mention. Recently, a similar trend could be found among the chamars of Punjab, as is evident from the following graffiti: ‘putt chamaran de’ (sons of chamars), then “chamar power”. However, the following graffiti behind a truck is notable for its identity assertion: ‘Awen bharam ne mutiaran de, marhi neet rakhade munde chamaran de’ (Young women are mistaken in thinking that the chamar boys have evil intentions). Such examples may be rare, but present before us a different trend in the caste issue. There has been a rise in caste assertion and it is not from the upper castes where it was always present. The dalits have begun to assert their caste iden-tity. The contemporary political scenario is also favourable to such assertion. Mayawati’s coming to power in Uttar Pradesh is an im-portant aspect of this assertion. Whatever might be her political strategy, she essentially represents dalits. The present is marked by the perpetuation of caste in which the politics of numbers has led to a situation where it is in the interest of the dalits to articulate and manifest and maintain their caste identities.6 ConclusionsThe findings of the empirical study point towards the paradox of change among dalits. At the level of the caste system, inequalities and exclusion continue to show their existence, the evidence of which could be ascertained on the basis of data on social ecology, occupation and access to religious places. Secondly, at the level of casteism as a world view the dalits are unable to construct the plausibility of a society without caste. The relevance and indis-pensability of caste in marriage has been largely accepted by them though there are apprehensions that without ending caste endog-amy caste may not disappear. Articulation of dalit identity for po-litical gains in the electoral process is again an important issue of concern. From the perspective of the dalits, the end of the caste system is the most radical change that could be thought of. The traditions, however, do not allow the dalits to emerge as a homo-geneous category. Casteism is powerfully embedded in the collec-tive consciousness of the dalits as a result of which their orienta-tion towards caste system is ambivalent. For them the caste sys-tem is oppressive due to discrimination, exclusion, exploitation and untouchability, but, at the same time, it situates them in the system as a collective identity. Their desire is to end all kinds of discrimination, oppression and exclusion, but not the caste sys-tem. It is an instance of deep-rooted internalisation of a world view – a process of reification in which like the Marxian market and relations emerging from that caste becomes a natural, abso-lute and ahistorical fact for members of the society. While the need is to think of Indian society without caste,thepossibilityof caste differences without inequality seems real. Notes 1 Jodhka (2002) has examined the issue of untouch-ability by further probing into the issues like access to village streets, sharing of drinking water sources, unclean occupation, practice of untouchability in modern institutions, police behaviour, etc. 2 Each unit has distinct social characteristics in terms of caste. For example, Amritsar city consists of balmikis, mazabis and meghs, Amritsar village mainly the mazabis, Jalandhar city balmikis, ad-dharmis and meghs and Jalandhar villages were comprised ad-dharmis and balmikis. It may be informed that the dalits are 28.78 per cent in the Amritsar district and 37.69 per cent in Jalandhar district. 3 There could be an exception to the general rule in the sense that in the class society some rich blacks, such as Oprah Winfrey, may be living in the posh and white-dominated neighbourhood. 4 Judge and Bal (2005) have shown how the organi-sation of the social space in Amritsar and Jalandhar cities in the case of the dalits exhibits strong association with caste. 5 The following empirical observation may be impor-tant in understanding how untouchability has de-clined in the Doaba region: In village Ajram in Hoshiapur district there are three gurdwaras out of which one belongs to the jats and the other non-dalit castes, and two to the ad-dharmis. When the authors attended a ‘bhog’ ceremony in the gurdwara meant for the jats and other non-dalits, it was interesting to find that the priest belongedtothejulahacaste. This priest was a stopgap arrangement,becausethe gurdwara priest was unwell and he belonged to the mazabi caste. The person who prepared the ‘langar’ (community food) belonged to the ad-dhar-micaste. On top of that, the women who were cooking chapattis belonged to the bazigar caste. 6 Recently, clashes between ‘Sikhs’ and the followers of Sacha Sauda sect have made the word ‘dera’ common. 7 The issue of religious places provides for varying interpretation largely due to the complexity in-volved in the situation. Having a separate gurdwara thus could also be understood as “assertion for autonomy” as stated by Jodhka (2002) though re-ligious communities tend to construct horizontal solidarity unless there are internal sectarian fissions. ReferencesBal, Gian Singh (2007): ‘Viaktigat Suntarata, Dera Sacha Sauda Vivad and Dalit-Kisan Samasya’, Nawan Zamana, July 22 (Punjabi). Dirks, Nicholas B (2002): Castes of Mind: Colonialism and the Making of Modern India, Permanent Black, New Delhi.D’souza, Victor S (1985): Economic Development, Social Structure and Population Growth, Sage, New Delhi.Ghurye, G S (1969): Caste and Race in India, Popular Prakashan, Bombay.Gupta, Dipankar (2007): ‘The Threat from Within’, Hindustan Times, July 28. Jodhka, Surinder S (2002): ‘Caste and Untouchability in Rural Punjab’,Economic & Political Weekly, May 11, pp 1813-23.Judge, Paramjit S (2002): ‘Punjabis in England: The Ad-Dharmi Experience’,The Economic & Political Weekly, August 3, pp 3244-50.– (2004): ‘Interrogating Changing Status of the Dalits of Punjab’ in Harish K Puri (ed),Dalits in Regional Context, Rawat, Jaipur, pp 100-31.Judge,ParamjitSand Gurpreet Bal (2005): ‘Dalits and Urban Social Space’, Indian SocialScience Review, 7(2), July-December, pp 98-120.Moon, Vasant (1989): Dr Babasaheb Ambedkar: Writings and Speeches,Vol 5, Government of Maharashtra, Bombay.Sharma, K L (2003): ‘The Social Organisation of Urban Space: A Case Study of Chanderi– A Small Town in Central India’, Contributionsto Indian Sociology, Vol 37, No 3, September-December.Virdi, S L (2004):Bahujan Manifesto, Guru Ravi Dass Society of Calgary, Calgary.Vovelle, Michel (1990):Ideologies and Mentalities, Polity Press, Cambridge.Webster, John C B (2007): ‘The Dalit Sikhs: A History?’ in Tony Ballantyne (ed),Textures of the Sikh Past: New Historical Perspectives, Oxford University Press, New Delhi, pp 132-54.Zelliot, Eleanor (1970): ‘Learning to Use Political Means: The Mahars of Maharashtra’ in Rajini Kothari (ed), Caste in Indian Politics, Orient Longman, New Delhi, pp 29-69.