On the face of it, the expression ‘Dalit Muslims’ would appear to be a contradiction in terms. It is commonly held that Islam is an egalitarian religion and there are no status differences among Muslims. As such, there is no question of the prevalence of untouchability among them and a category called ‘Dalit Muslims’ cannot be said to exist. This is the standard line that is handed down whenever any reference to ‘Dalit Muslims’ is made.
This would be a perfectly understandable position to take were it not for the fact that considerable evidence exists to suggest that a category called “Dalit Muslims’ does exist in India. Ghaus Ansari argued on the basis of evidence from the decennial censuses that Muslims in India were divided into three broad categories that he called the ashraf (noble born), ajlaf (mean and lowly) and arzal (excluded). Each of these categories was further divided into a number of groups which, following the practice of the decennial censuses, he chose to designate as castes. Since Ansari was relying on the evidence supplied by the decennial censuses, he could not examine the process of mutual interaction among these castes. He generally suggested that the three broad categories he had identified constituted a hierarchy in which the castes were ranked in an order of social precedence. How this hierarchy was constituted and what was the basis on which the rank order was settled were questions that Ansari could not discuss on account of the limitations of the data he used.
More focused research on social stratification among Muslims in the early seventies and subsequently relied upon empirical methods, painstakingly collecting information on actual, day-to-day interactions among the Muslim communities. This research succeeded in providing a more grounded picture of the situation of the groups whom Ansari had called arzal. It demonstrated that in terms of day-to-day social interactions the arzal existed on the margins of society. Even so, the range of dimensions of interaction that this research explored was restricted to areas of commensality, endogamy and sociality. It showed that the arzal engaged in the lowly occupation of scavenging, confined their marriages within the group and were excluded in the villages as well as the towns into separate residential quarters in which members of the other categories did not live. This research also noted the existence among the arzal communities of a system of internal government and social control with a hereditary official who regulated the life of group members and punished any transgressions of group norms besides settling domestic or intra-group disputes. Since much of this early research was focused on local communities, villages and towns, and covered groups falling into what Ansari had designated as arzal and ajlaf, the range of information on the arzal communities does not go beyond this limited range. For example, it is silent on the exclusion of the arzal communities in the ritual and religious spheres as well as on whether the religious specialists who cater to the ashraf and ajlaf communities also minister to them.
One question raised by this research was how the presence of groups whom Ansari had called arzal be explained. Should they be seen in strictly occupational terms as practitioners of a distinct occupation that in their case happened to be lowly and demeaning without status connotations? Or, should it be seen as arising from more fundamental and intrinsic considerations requiring evaluation of groups into a ranked social order? Opinions on this significant point were substantially determined by how one viewed the position of Islam in relation to social stratification. Those who took the position that Islam was against any social stratification and posited the inherent equality of all human beings tended to represent the presence of arzal communities as merely an occupational division without any status implications.  From their point of view, the disabilities and exclusion characterizing the arzal communities applied to individuals and were relevant only in the occupational realm. Once their members move outside the occupational realm they are on par with everyone else. Others did not flatly take the position that Islam was against social stratification. They viewed the existence of arzal communities as reflecting a system in which groups were ranked as superior and inferior and individuals carried the burden of their group status through having to suffer disabilities and exclusion as members of groups.
Since the most elaborate expression of a system of social stratification wherein groups are ranked as superior or inferior and individuals are forced to carry the burden of their group status is found in the caste system, one way of characterizing the presence of arzal communities could be in terms of caste. However, since Islam in popular imagination is seen as the harbinger of social equality, such characterization is open to contestation on ideological grounds. This has precisely been happening in sociological research on the arzal communities. At the behavioural level, sociologists are willing to concede that there are elements of caste in Indo-Muslim society. However, as soon as the discussion shifts from behaviour to ideology they recoil form their position, seeking to add caveats or hedge around the issue by admitting unabashedly that when they apply the term in the context of a Muslim group they are using it in a loose sense. Two recent writings by Husnain and Nazir exemplify this tendency most eloquently.
Husnain locates his discussion in the context of the question whether the concept of caste can be applied to the system of social stratification of a community professing a faith other than Hinduism. His conclusion is bald and simple: ‘It is true that the egalitarian social order of Islam stands in sharp contrast with the ideology of caste yet the ‘Indian Islam’ and ‘Hindu Caste System’ have been able to achieve a substantial compatibility’. He then goes on to offer a host of explanations for why this should be the case. He writes:
Hutton sounds convincing when he says that when Muslims and Christians came to India, the caste was in the air and the followers of even these egalitarian ideologies could not escape the infection of caste. Moreover, the overwhelming majority of Indian Muslim population comes form the lower Hindu castes who have been coming into the fold of Islam to escape from social persecution and the oppressive socio-economic disabilities. They were also attracted and lured by the social egalitarianism of Islam but the search for equality proved a mirage. In many cases there were improvements in their socio-economic condition yet the goal of social equality remained illusive. Moreover, in most of the cases the people embracing Islam gave up their religious faith but not the caste that was brought forward even to a new socio-religious milieu. Thus, it would be apt to say that while Islam may not be having castes or caste-like groupings, the Indian Muslims do have.
No sooner that he has made this sociological formulation, Husnain becomes uncomfortable. As if fearing that he might have committed an almost sacrilegious act by declaring that there is caste among Indian Muslims, he wishes to recoil from it. Cryptically, he adds:
But in the present paper an attempt is being made to stay clear of the issue whether the model of social stratification among the Indian Muslims is the replica of the Hindu caste system or not. The author, in this paper, shall be using the term caste and caste system among the Indian Muslims in a conveniently loose manner. It is undisputed that there are groups of people among the Muslims who are organised more or less like the Hindu castes but this is also true that many of them are less rigid because Islam, theoretically at least, permits marriage between different classes of believers.
Not only that. He looks for crutches that would enable him to perform this summersault. He finds one in the following statement of Nazir, which he quotes approvingly:
. . . . It is necessary to make a distinction between a caste system and caste labels: the former refers to a local system of hierarchically ordered corporate groupings involving division of labour, occupational specialisation, unequal dependence, and recruitment by birth only; the latter refers to a set of non-local, non-corporate named groups which provide a ranking hierarchy, and which do not involve occupational specialisation, unequal dependence, and recruitment by birth only.
‘Perhaps,’ concludes Husnain, ‘the “caste system’ and “caste like groupings” among the Indian Muslims with all its fluidity may be better analysed and better understood through this observation’. 
This assumes that Hindus live under ‘the caste system’. Muslims only use castes labels. Several theoretical and empirical questions are raised by this assumption. First, how is this assumption made? Is it made on the basis of a piece of empirical research? Or, is it made on entirely a priori grounds. As far as I am aware, there has to date been no empirical research which can be said to have established beyond the shadow of a doubt that Muslims do not live under a caste system and only use caste labels. Indeed, if such empirical research existed, the dilemma these authors (and others) face over how to characterise Muslim social stratification in India would not exist. It exists because available empirical research has demonstrated that social stratification of Muslim communities in India and beyond is marked by features of the caste system. It is, therefore, clear that the assumption is made on a priori grounds. As believing Muslims committed to upholding the widely proclaimed Islamic egalitarianism as axiomatic, they cannot face up to the behavioural reality that Muslims live under a caste system. They not only assume the distinction between ‘the caste system’ and ‘caste labels’ but go on to suggest that it constitutes a viable framework for analysing and understanding Muslim social stratification in India. It is used as a smokescreen to avoid facing the harsh behavioural reality of caste among Muslims in India.
Second, is there an empirical basis to the assertion that Muslim social organisation in India is ‘a set of non-local, nor-corporate named groups which provide a ranking hierarchy, and which do not involve occupational specialisation, unequal dependence, and recruitment by birth only’?  Nazir does not make explicit the level at which he is talking. Is he talking about the categorisation of Muslims into the broad categories of ashraf, ajlaf and arzal. If that is his point of reference, then his characterisation of Muslim social organisation as a set of non-local, non-corporate groups can be said to have some validity. However, it would invalidate the distinction between ‘the caste system’ and ‘caste labels’ since similar broad division exists in the form of varna categories in ‘the caste system’. Ansari used the three broad categories of ashraf, ajlaf and arzal in the collective sense but clearly recognised that they were divided into smaller named groups that were distinguished from one another by occupation, endogamy and sociability. Thus, if Nazir’s reference is to the groups at this level, then his description of Muslim groups is wholly erroneous. Let us look closely at the empirical evidence in order to determine whether the distinction he posits between ‘the caste system’ and ‘caste labels’, and by implication between Hindu and Muslim modes of social organisation, is confirmed by available studies.
Sociological research on Muslims in India as opposed to lay and impressionistic writings continues to be thin. Evidence brought together by Ahmad (1973) and subsequent research demonstrates that Muslim groups which are the point of reference here, for which words biradari and zat are commonly used, are local and corporate entities. Even biradaris or zata such as Saiyyid, Sheikh and Ansaris, which are dispersed widely and found in different parts of a district, state or the county, are identified by their affiliation to a particular territory and restrict their marriages to members within that territory. Of course, how that territory is distinguished varies widely. For Sayyids, Shiekhs and Pathans, which resent being characterised as biradaris and prefer to be described as zats, the association to territory is expressed through appending the name of the territory to its name. Thus, one hears of Sayyids of Satrikh, Sheikhs of Allahabad, Kidwais of Baragaon or Kasauli and Pathans of Malihabad. In the case of biradaris that have an internal organisation of government and social control (called biradari or zat panchayat) this territorial association is defined by the jurisdiction of the biradari panchayat. The Ansaris in Rasulpur, where I carried out fieldwork, were divided into concentric circles of three and thirteen villages. They confined their marriages to thirteen villages though Ansaris existed in neighbouring areas as well.
This is not all. Considerable evidence exists to show that the biradaris or zats are associated with particular occupations, are inter-dependent (tied into patron-client relationships of the jajmani type), and are endogamous. This does not mean that all members of a biradari or zat necessarily practice the occupation with which their group is traditionally associated. There has been much variation throughout history among biradaris and zats, as indeed there has been within castes, in the extent to which their members remain tied to the practice of their traditional occupation. Biradaris and zats higher up in the social hierarchy did not usually have a traditional occupation and there was no close association between biradari or zat and traditional occupation. On the other hand, biradaris and zats further down the social ladder had traditional occupations and their association with occupation was strong. This was not significantly different from the picture of groups in what Nazir would characterise as ‘the caste system’. Risley’s following observation makes this explicit:
In theory each caste has a distinctive occupation, but it does not follow that this traditional occupation is practised by its members . . . . The traditional occupation of the Brahmans is the priesthood, but in practice they follow all manner of pursuits. Many are clerks or cooks, while some are soldiers, lawyers, shop-keeprs and even day-labourers, but they remain Brahmans all the same. The Chamars of Bihar are workers in skin, but in Orissa they are toddy-drawers. In Orissa and the south of Gaya the Dhobi is often a hewer of splitter of wood. In Bihar and Bengal the Dom is a scavenger or basket maker, but in the Orissa states he is a drummer or basket maker and has nothing to do with the removal of nightsoil: in Chittagong and Assam he is a fisherman, in Cashmere a cultivator and in Kumaon a stone mason.
The argument that Muslim groups, biradaris and zats, are not based on recruitment by birth only is equally fallacious. Like the groups in what Nazir would call ‘the caste system’, Muslim biradaris and zats are based on recruitment by birth only. There is no process by which one can become a Saiyid, Shiekh or Julaha except that of birth. It is for this reason that when someone marries into another biradari or zat, he is not integrated into another biradari or zat but retains his or her original biradari or zat association. There exists a possibility in the case of biradaris and zats to attempt social mobility and end up becoming a Sayid, Shiekh or Pathan in course of time through inventing a rationale and a genealogy. Where such social mobility occurs, the basis of recruitment to the biradari or zat does not change. The biradari or zat just ends up becoming another biradari or zat, and comes to be known by another name, to which recruitment continues to be based on the principle of birth. This is again not significantly different from the situation in ‘the caste system’ where castes have the possibility of changing their antecedents and name through the process of social mobility. Thus, the point that both biradaris and zats are ‘less rigid, because Islam, theoretically at least, permits marriage between different classes of believers’  is not empirically established. It is commonly asserted without a substantial basis in any empirical research.
This raises fundamental questions. Why Husnain and Nazir as well as a host of other researchers who have worked on the sensitive question of the existence of caste among Muslims are so strongly persuaded to posit that there are significant differences between ‘the caste system’ and the system of biradaris and zats? Is it that these differences actually exist but empirical research has so far failed to unearth them? Or, is it that they are persuaded into asserting these differences contrary to empirical evidence out of extraneous considerations? Is it that they are prone to emphasising these differences because as believing Muslims they are familiar with the Islamic discourse that asserts that Islam preaches social equality and are afraid to take a contrary position? Or, is it that asserting these differences is a defence mechanism whereby they can simultaneously adhere to their disciplinary obligation as social scientists as well as their religious obligation to uphold what is commonly considered the Islamic view on social stratification? My own view has been that the tendency to emphasise differences between ‘the caste system’ and the system of biradaris and zats arises from some such considerations, but I would refrain from making any such point here. I would like, instead to explore whether their starting point that Islam is an egalitarian religion and preaches social equality theologically and sociologically valid. This is central to understanding their standpoint.
There is need to ask three different questions of the Islamic text if we are to understand Islam’s position with respect to social stratification and social equality. First, whether Islam is opposed to social stratification as such or is merely opposed to social inequality. Second, what is truly the Islamic attitude towards social inequality that existed in the society in which Islam evolved and took root. Finally, whether the social equality that it proclaims, and to which reference is always made when it is suggested that Islam is an egalitarian religion, is a description of an existing state of affairs in society or is merely an ideal that is given to mankind as a direction in which it should strive. It is necessary to ask these questions in order to understand the nature of the emphasis on egalitarianism and social equality in Islam. Basic to these questions is the sociological dictum that no society beyond the most primitive in the sense of lacking any kind of economic surplus can be truly egalitarian. This was the point at the heart of Veblen’s Theory of the Leisured Class wherein he argued that as societies generated economic surplus there almost always developed some form of social stratification. Of course, Veblen’s concern was an analysis of the lifestyle and consumption pattern of the class that controlled the economic surplus and the symbolic and behavioural expressions of its privileged position. Even so, the substantive theoretical point of his analysis was that once a society starts generating economic surplus some form of social stratification is bound to emerge. Pitirim A. Sorokin articulated this point as a general statement:
Any organised social group is always a stratified social body. There has not been and does not exist any permanent social group which is “flat”, and in which all members are equal. Unstratified society, with a real equality of its members, is a myth which has never been realised in the history of mankind. This statement may sound paradoxical and yet it is accurate. The forms and proportions of stratification vary, but its essence is permanent, as far as any permanent and organised social group is concerned.
On even the most casual reading of the Islamic scriptural text one is struck that quite irrespective of the emphasis it places on equality of human beings Islam’s orientation is remarkably hierarchical. Its hierarchical orientation comes in a wide variety of fields. First, the relationship of the believers with non-believers is conceived in strictly hierarchical terms with the believer, the dhimmi and the kafir constituting a clear hierarchy. Second, the relationship of Allah to the believer is conceived in hierarchical terms. It is a relationship of subordination and subservience so much so that the individual believer must prostrate before Allah in daily prayers and must at the same time see himself as utterly powerless in relation to Him. Any number of passages exist in the Islamic scriptural text that endorse the relatively lowly standing of the believers, whether as individuals or as a collective entity, in relation to Allah. Second, the relationship of the wife to her husband is clearly conceived in hierarchical terms even if the text does not distinguish between them in terms of the religious duties enjoined upon them. This is sometimes cited by Muslim feminists and Muslim modernists to argue that Islam guarantees equality of gender and does not place a Muslim woman in any inferior position to a man. However, in reality a woman is subordinate to a man and the relationship between them is seen as constituting a hierarchy wherein the woman stands in relation to a man in the same position as the individual stands in relation to the community and the community stands in relation to Allah. Fatima Mernissi characterises this orientation of Islam in relation to women by the concept of nusuz, which implies an unequal relationship. Islam makes a distinction between the wives of the Prophet and other women and the responsibilities placed on them are also distinctly varied. Indeed, the Quranic verse that orthodoxy used at a later stage in the development of Islam to impose the custom of veiling for Muslim women originally related to the wives of the Prophet. Finally, the relationship between the master and slave is conceived in clearly hierarchical terms even if the master is called upon to deal with the slave with kindness and merit is assigned to those who would free their slaves. Thus, it is clear that the framework of Islamic thinking is deeply imbued with the notion of hierarchy and social stratification.
It is true that the Arab society in which Islam evolved did not possess great differences of wealth, but economic differentiation between ordinary Bedouins and the trading classes did exist. One can easily imagine that they would have differed with respect to their wealth, material possessions and lifestyles and Islam could not have brushed them under the carpet. It would have been required to deal with them, as they would have been reflected in their behaviour and mutual attitudes. As far as the Islamic scriptural text is concerned, it clearly recognises such distinctions in society and prescribes appropriate forms of behaviour for each. It asks those deprived in social and economic terms to be content and to live according to their means. It is repeatedly said in the text that Allah is All-seeing and would reward the poor for their poverty on the day of judgement. At the same time, the wealthy and rich, while they are allowed to live in their riches and to spend according to their economic standing, are warned not to be too proud of their material possessions. Moreover, they are asked to show kindness to those who are deprived and poor and to part with a portion of their wealth and income for the poor. Even the poor are conceived in hierarchical terms: first come the near ones followed by orphans and then the destitute and the deprived. If some kind of social stratification had not existed in society, Islamic scriptural text would neither have referred to those differences, nor indicated appropriate forms of behaviour for them. It would also not have sought to device an economic framework for the redistribution of wealth in a manner that the poor and able to meet both ends meet. It is, thus, clear that the emphasis that Islamic scriptural text places on social equality does not describe an existing state of affairs.
If the worldview of the Islamic scriptural text is hierarchical and it admits of social and economic differences in society, then how should we interpret its emphasis upon social equality? One way to interpret this can be to ignore that Islamic orientation is hierarchical and to argue that it stands for egalitarianism as an absolute value. I would argue that those who maintain that Islam contemplates no social stratification are interpreting Islam in precisely this way. Even when they encounter social differentiation and stratification, they glibly ignore it and flash the proclaimed egalitarianism of Islam as a social reality. The other way of interpretation can be to recognise a fundamental difference. This is the difference between the society as it exists and as it ought to exist and to maintain that the Islamic proclamation in favour of social equality is more in the nature of an ideal for the future than a description of an existing state of affairs. My own position is that drawing this distinction is important in any consideration of the question of the presence or absence of caste and caste-based social stratification in Indo-Muslim society. It enables us to see that a distinction has to be made between the society which exists, and where caste- or class-based distinctions may exist, and a future state of society where they are expected to disappear and give rise to an egalitarian society. This distinction applies to Islam as much as to any other ideological system that proclaims social equality as an ideal.
This distinction should not be entirely unfamiliar to us in India. As is well known, Indian society has been the most unequal society, the social inequality being institutionalised in the caste system. India’s constitution went on to declare India to be a casteless and classless society. In so doing, the constitution was not proclaiming that social inequalities of the past had entirely disappeared and the society was egalitarian from the time it was promulgated. The only sensible way would be to recognise that, while social inequalities persist, the ideal that the Constitution provides is that of egalitarianism. This is also true of Islam. It proclaims social equality to be an ideal, but recognises social inequalities existing in society. By this token, there is no contradiction between Islamic support for an egalitarian society as a future goal and presence of caste or class differences as a social reality.
Social realities have a way of prevailing over sociological and theological formulations. Contrary to the argument of some sociologists and most theologians that caste does not exist among Muslims and untouchability is disallowed in Islam, the expression ‘Dalit Muslims’ has been finding increasing mention in the discourse of traditionally backward Muslim communities in recent years. However, there does not yet exist any clear understanding of what this expression actually means or which castes or groups it is supposed to denote. On the one hand, it has been used to denote a whole range of Muslim castes which are currently included in the category of the Other Backward Classes. On the other hand, it has been used to denote those Muslim castes or groups which converted from the ‘untouchable’ Hindu castes or are so severely stigmatised and are subjected to such extreme forms of social exclusion that would render them comparable to the Scheduled Castes.
The Mandal Commission compounded and reinforced this confusion. As is already well-known, the Commission’s task was to identify Other Backward Castes and to determine whether they should be eligible for reservation along the lines of the Scheduled Castes and Scheduled Tribes. There was no difficulty in this with respect to Hindu castes because administrative policy clearly recognised a distinction between Scheduled Castes and Other Backward Castes. Because the presidential order of 1950 clearly and arbitrarily laid down that ‘No person wo professes a religion different from the Hindu religion shall be deemed to be a member of a Scheduled Caste’, the Mandal Commission lumped the severely stigmatised and extremely excluded among the Muslims with Muslim Other Backward Castes for purposes of affirmative action. Therefore, when the urge for equality and social justice seized the imagination of the lowest social groups in other religious traditions and the word Dalit came to be seen as a short-cut carrier of that aspiration, the expression ‘Dalit Muslims’ came to be used for a wide variety of groups other than those severely stigmatised and excluded and on that ground comparable to Hindu ‘ex-untouchable’ castes for whom the term ‘Scheduled Castes’ was reserved.
One can arrive at an assessment of the extent of confusion that prevails at present with respect to the expression ‘Dalit Muslims’ by reading between the lines in the statements of those claiming to speak on their behalf as well as by considering the castes that they have been tempted to include under that category. N. Jamal Ansari writes: ‘. . . it is an established fact that Indian Muslim community is divided into castes and has a large deprived section. . . . Before discussing constitutional provisions in respect of Dalits and exclusion of all ‘Dalit Muslims’ from those provisions, I think we must define “Dalit Muslims”. Dalit means downtrodden, oppressed, suppressed and backward. Also, Dalit stands for untouchable and depressed classes. The term “Dalit” applies to members of those menial castes that have been graded lowly which they have inherited by accident of birth’. Likewise, Ali Anwar  uses the words ‘pasmanda’ (meaning downtrodden and backward) and Dalit interchangeably and includes under ‘Dalit Muslims’ castes like Bhatiyara, Tikyafarosh, Itafarosh, Halalkhor, Khakrob, Mogalzada and Chirimar only some of which can be said to be severely stigmatized and excluded. In all such statements and lists, as their reading suggests, the expression ‘Dalit Muslims’ has been used as a generic term to denote all Muslim castes that are educationally and socially backward.
Clearly, there is need to define Dalit Muslims in more precise terms. Ansari suggested in his early work that the relationships between the ashraf and ajlaf on the one hand and the arzal on the other were shaped by considerations of social distance taking on the characteristics of untouchability. He mentioned that the members of the category called arzal were excluded both physically and socially. From a physical point of view, they tended to inhabit excluded localities and did not mix with the members of the other two categories. When it came to social intercourse, their relationship was characterized by strict maintenance of social distance and deference so that the members of the arzal communities had minimal and limited interaction with the members of the other communities. The expression should be restricted to refer to these castes alone. Since the expression ‘Dalit’ has come to acquire pejorative connotations, though it was originally used by the Dalit Panthers Movement as a short-hand way of referring to the extremely deprived and excluded castes, many Muslims may not like that expression to be used in the context of Muslim castes. Even so, there is need to recognise that the castes to whom we have here referred to as Dalit Muslims do form a class separate from the other categories of Muslim castes and need to be distinguished on account of the extreme degree of stigmatisation and exclusion suffered by them.
There is need for rich and focussed ethnographic research on such castes. This research should seek to understand the attitudes of the non-arzal castes and groups toward the members of the arzal category and to gauge the extent and intensity of discrimination suffered by them today. It is possible that with the introduction of sanitary toilets and other technological changes the arzal castes no longer engage in the demeaning and defiling occupation of scavenging but social distance from them continues to be maintained. It is also possible that the forms of discrimination and stigmatisation practised against the arzal castes have changed, but they may have taken other forms. Only focussed social research can indicate the contemporary situation of the arzal castes in contemporary Muslim society.
 Gaus Ansari, Muslim Castes in Uttar Pradesh, Lucknow, Ethnographic and Folk Culture Society, 1959.
 Imtiaz Ahmad, `Introduction’, in Imtiaz Ahmad (ed.), Caste and Social Stratificaion among Muslims, Delhi, Manohar Book Service, 1973.
 From my observations of growing up in a Muslim family I am able to recall a number of instances of both open and silent discrimination practiced against these castes. We had a Lalbegi woman come to clean the toilets in our house. She was on the best of terms with my mother and would sit for hours together gossiping with my mother. Whenever my mother would offer her pan, she would wrap her hand with her dupatta to receive it. My mother used to drop the pan in her hand, making sure that her hand did not touch the Lalbegi woman’s hand. On occasions of marriage the family would come and sit in a corner and wait until all guests had eaten and left. It would then be given food in vessels they brought with them. They did not eat the food there, but instead took it with them to be eaten at home. On sacrificial eid the family was not given any portion of the meat. It was given the intestines which were kept aside for them. It is possible that some of these forms of discrimination have changed, but there is no evidence to show that they have disappeared.
Some evidence exists to show that there is discrimination against these Muslim castes in the religious spheres. I found during fieldwork in eastern Uttar Pradesh that members of these castes did not go to the mosque for prayers and if they went they had to stand in the back rows. It has been mentioned by many observers that such groups often have their own mosques. N. Jamal Ansari notes that ‘in certain areas of Uttar Pradesh and Bihar there are separate mosques and burial grounds’ for these castes (Paper presented at the seminar on Dalit Muslims organized by Deshkal Society, New Delhi, 2004). Establishment of own mosque would call for a level of prosperity for the groups as a whole. Whether they have attained such levels of prosperity is something on which very little information exists.
 They were admitted to the house, but only for the performance of their services, in doing which they were required to ensure that they did not touch members of the household or other things that carried the risk of becoming polluted or rendered impure by their touch. They were asked to be present at social functions and festivals on social occasions such as marriages or life cycle rituals, but this was not by virtue of any social right. It was by way of recompense for the service they rendered throughout the year. On all such occasions they sat some distance away from the main entrance for the guests and waited until everyone had been entertained to be given food that they were not supposed to eat then and there but carry it in the containers they brought with them to be eaten at home.
 Charles Lindholm has argued that many of the features found in Muslim society are similar to those found among Muslims in other parts of South Asia and on that basis has argued that the Muslim social stratification found in India is an extension of the system found elsewhere (see Charles Lindholm, ‘Paradigms of Society: A Critique of Theories of Caste among Indian Muslims’, European Journal of Sociology, 1965, pp. 131-140). Many Muslims are themselves inclined to take a similar line of argument. This argument would have been tenable if Islamic scriptural sources had provided a blue-print of an Islamic social stratification system. This not being the case, the argument fails to sustain itself. It is plausible that Islam did modify certain social practices including that of caste. Whatever practices were not sanctified by Islam but existed in India were attenuated. Whatever practices existing in India were in conformity with the Islamic ethos became more rigid. Thus, purdah practices, which already existed even in India, were rendered more rigid and strict and caste principles were relaxed or made less restrictive.
 Nazir, 2898.
 Husnain, p. 2
 Husnainm pp. 207-08.
 Husnain, pp. 207-08.
 Nazir, 1993: 2898
 pp. 208.
 Nazir, pp. 2898.
 Census of India, 1902, pp. 350-51.
 Husnain, pp. 208.
 Pitirim A. Sorokin, Social and Cultural Mobility, The Free Press, New York, 1959, pp. 13-14.
[17Mernissi, Fatima (1986). Femininity as subversion: Reflection on the Muslim concept of nushuz. In Devaki Jain and Diana L. Eck(eds.), Speaking of faith: Crosscultural perspectives on women, religion, and social change. New Delhi: Kali for Women. 464-476.
Anwar, Ali (2005). Masawat Ki Jung (in English), New Delhi, Indian Social Instittue.
(Prof. Imtiaz Ahmad could be reached at firstname.lastname@example.org)