ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Class, Caste or Race: Veils over Social Oppression in Pakistan

Pakistani society prefers to silence any discussion about caste-based discrimination and oppression in the public domain even though such marginalisation is widespread in many parts of the country.

Letter from South Asia

Class, Caste or Race: Veils over Social Oppressionin Pakistan

Pakistani society prefers to silence any discussion about caste-based discrimination and oppression in the public domain even though such marginalisation is widespread in many parts

of the country.


hen asked recently to think about “social marginalisation” in Pakistan, and to carry out fieldwork on the issue in rural and urban areas of the country, an experienced team of researchers knew they were up against some of the usual sources of resistance. It is easy enough to think, speak and write about “economic” poverty in Pakistan – the government’s point men notwithstanding. Societal causes of deprivation and marginalisation such as caste, religion and ethnicity find few takers and many detractors. But as our intrepid researchers got about their work, new demons that proved to be familiar, and yet quite unnerving, reared their heads.

I have enjoyed the privilege, over the last many years, of working with a gifted group of field researchers who relish the opportunity of challenging their own imaginations and asking difficult questions.1 We have worked together on a number of social policy issues including rural poverty, urban governance, school performance, bonded labour and migration. Our last assignment took us on a whirlwind tour along the length of the country – from Khyber to Karachi – and across its breadth from Badin to Gwadar. Part of the brief was to understand and document diverse processes of social marginalisation. Or the systematic marginalisation of individuals, families, and groups due to their “social” attributes such as caste, traditional occupation, kinship, ethnicity, religion and lifestyles.

The fieldwork took us to diverse urban neighbourhoods, irregular settlements (‘katchi abadis’), villages, hamlets, migrants’ campsites and beggar colonies in many districts: Peshawar in North-Western Frontier Province (NWFP), the Islamabad Capital Territory, Faisalabad, Lahore, and Toba Tek Singh in central Punjab, Dera Ghazi Khan and Rajanpur in southern Punjab, Quetta, Kech and Gwadar in Balochistan, and Sanghar, Badin and Karachi in Sindh. We visited a number of places that were quite familiar and also some places that were new.

There was a prior list of groups or “communities” that were of primary interest. This included groups that were known to face caste-based oppression, such as the ‘kammis’ (service castes) and “menial” Muslim shaikhs in Punjab and the NWFP; groups such as the ‘masihi’ Christians in Punjab, and the bheels, kolhis and other scheduled caste Hindus in Sindh who, in addition to caste-based oppression, were also vulnerable to religious discrimination.2 Traditional beggars, minstrels and groups known to follow a peripatetic lifestyle represented another distinct process of marginalisation from the mainstream economy and society. In some areas, notably Makran in southern Balochistan with a history of African slave trade, race was expected to be an important dimension of social marginalisation.

Caste Is Another Country, Another Time

We expected this study to raise the usual hackles in Pakistan. There is little tolerance in the public domain of any serious discussion about caste and caste-based oppression, social hierarchies, and discrimination. The Right silences such talk by shouts of “we are all Muslims” and “caste is another country” – obvious which country that might be. In fact, the denunciation of “the evil caste system” is a standard hymn in the rightist intellectual’s repertoire on India, Hindus and the Two-Nation Theory. For the Left in Pakistan, when there was one, it was all about class. Caste or other societal sources of inequality and oppression were seen as unnecessary diversions from class struggle. The Left, of course, never succeeded in large-scale class mobilisation either, and it is tempting to put this failure down to sterilised and mechanical ideas about class that did not address societal conditions.

There is yet another popular response to the question of caste hierarchy that can be seen as a synthesis of the cruder early rightist and leftist approaches. Those educated enough to have read a colonial gazetteer – that fountainhead of societal knowledge – know that caste is not another country, and that caste was class in what would be Pakistan. It is hard to argue with something written in correct English by a real Englishman, or a properly trained Indian Civil Services officer who was the next best thing. The answer is that caste is the past – that was then, now is different. For evidence, some offer the standard anecdote of the “low-caste” person from their own village who went to Dubai and did well, while others wax lyrical about Pakistan having an urban “feel” to it.

The public silencing on caste contrasts with an obsession with it in private dealings and transactions. I was once heckled at a policy forum in Islamabad while presenting findings of research that highlighted the significance of traditional caste and occupational hierarchies in the working of labour markets. It was quite bizarre to be at the receiving end of a leftist “class-not-caste” harangue from a Jamaat-e-Islami trade unionist – reassuringly, he later conformed to type and accused our research of besmirching the image of Islam and Pakistan before an international gathering. Even more ironic was the fact that a dominant theme of coffee-break chat

Economic and Political Weekly January 13, 2007

among transfer-hungry bureaucrats assembled at that meeting was the favourable attitude of a high official in the Punjab provincial government towards applications emanating from rajput sources. Such encounters are common enough for the few of us who do this type of work, and I frequently find myself swapping “war stories” with my friend Ali Cheema who teaches economics in Lahore and is in many ways closer to the frontlines.

In fact, the kinship group, known variously as ‘zaat’, ‘biraderi’ and ‘quom’ in different parts of the country, remains a key – perhaps the key – dimension of economic, social and political interaction. It is a paradox that Pakistani society outside of Karachi, but also quite substantially within that metropolis, is united in being divided into robust kinship groups. The common practice of cousin-marriage – the small Hindu and Christian communities are exceptions in this regard – contributes to the strength of extended patriarchies. Families and extended social networks based upon kinship ties are important for Pakistanis and they acknowledge and share these values across ethnic and sectarian boundaries. Everyone instinctively knows that market transactions and political coalitions will be mediated through kinship group ties.

Perhaps the public silencing on biraderi and quom is born out of embarrassment with the continuity of “traditional” social forms in the face of modernist aspirations of the Islamist and cosmopolitan types respectively. If the most common understanding of quom is the kinship group, then whither the great nation-building projects?3 Pakistanis appear, at one level, to be obsessed with patriarchal notions of honour, with all of the violent misogyny that goes with them. But the robustness of wide social networks based on kinship groups also accounts for feelings of solidarity, group-based collective action, relative autonomy from state and market, and a culture of hospitality.4

How Do You Treat Your Kammis?

The trouble is that the biraderis and quoms are not all equal, and public silencing of the issue is very much about perpetuating existing hierarchies. The inequality is so severe and deeply embedded in parts of the country that it is hardly even noticed. An urbane politician from Lahore did not intend any offence when, outraged at the state’s desecration of the remains of a Baloch tribal leader killed in a military operation, he blurted out, “we don’t treat even kammis like that!” No offence was taken, of course, because there was no-one to speak up for the kammi biraderi, or to ask our Lahori friend how he treated his kammis.

It was not surprising to me when in the course of our fieldwork on social marginalisation we ran into a senior official who admonished me for purposively selecting sites in Punjab with a large concentration of ‘chuhras’. It was easy enough for this person to use what would be regarded as politically incorrect language about an oppressed community – the “untouchable” sweepers who are mostly Christian – in a private conversation with a virtual stranger. In fact the great Punjabi village joke is about caste, and at its receiving end usually stands a chuhra, a ‘mussali’ (a pejorative term for Muslim shaikh menial workers) or a kammi. Luckily for our research the fieldwork was not stopped and we went on to meet many chuhras, mussalis and kammis in Punjab, their counterpart ‘neech zaat’ (low caste), ‘badnasal’ (bad lineage) in NWFP, and ‘ghulams’ (slaves) in Balochistan.

Verbal abuse is just one part of it. In Peshawar we heard about how children from a Christian “sweeper” colony were barred from local government schools – schools that encouraged the enrolment of children of Afghan refugees who were not even Pakistani citizens. In brick kilns within the Islamabad Capital Territory, Muslim shaikh workers were held in conditions of virtual bondage. Youths who did manage to venture out to Rawalpindi’s Raja Bazar for recreation had the local police pounce on them and extract bribes for no reason other than their “mussali appearance”. Bonded labour is also rife in Toba Tek Singh villages where mussali farm servants remain at the beck and call of Araeen farmers, as do bheel tenants of Sindhi landlords in Sanghar. Verbal abuse leads to physical violence in many cases, and “light” slapping is not even reported as “a beating”.

We documented cases across the country – in Peshawar, Faisalabad, Quetta and Sanghar – of rapes perpetrated against “low-caste” women from chuhra, mussali, lachhi and scheduled caste Hindu communities respectively. The perpetrators were all well known and there was a feeling that they committed these crimes because they could get away with it, knowing full well that the victims were socially and politically weak. In fact, these rapes were only the most extreme instances of sexual violation suffered by the marginalised groups. In the language of the dominant groups the “low castes” had no honour, and certainly no honour that could be defended. The Khans in Peshawar, who regarded themselves as the racially pure descendents of 11th century Pashtun invader tribes from Afghanistan thought that the women of their ‘hamsayas’ (literally neighbours, but used as a euphemism for dependent service castes) such as the Toorkhail (literally “black lineage”) and ‘kisabgars’ (menials) were of lax social morals. In any case the hamsaya men, unlike the “pure” Pashtuns, would not/could not protest openly if their women did contract illicit liaisons with other men.

Everyday forms of exclusion revolve around taboos regulating eating and drinking together and sharing utensils. Bheels, kolhis, bagris, lachhis and other scheduled caste Hindus that we met in Sindh and Balochistan are served in plates and bowls that are kept separate from the rest in local eateries. They have to wash these utensils themselves after use. The same is true of the Christian chuhras in Punjab and the NWFP, who are also denied the ‘huqqa’ (traditional smoking pipe) which otherwise circulates freely in the Punjabi village.

The reasons offered by the Muslims for these food taboos – “they eat ‘haram’ (food forbidden in Islam) such as carrion, pork, and reptiles, so it is forbidden for us to eat with them” – are curiously reminiscent of brahmanical notions of ritual purity. Hardly anyone bothers to substantiate claims about the consumption of haram food on the part of these marginalised groups, or how, for that matter, eating or sharing utensils with

Economic and Political Weekly

available from:

Churchgate Book Stall

Churchgate Station Opp Indian Merchants Chamber Churchgate Mumbai - 400 020

Economic and Political Weekly January 13, 2007 people who may have had partaken of food forbidden to Muslims might be offensive to Islam. In fact, there is little problem in sharing meals with “upper caste” and upper class Christians and Hindus. And conversely, even second-generation Muslim converts from the marginalised “low-caste” groups are subjected to the same food taboos – they “appear” like the carrion-eaters, so they must be the same. In fact, it is quite common to hear allegations that even Muslims of long-standing such as the Muslim shaikhs (mussalis), changars, lohris, shahikhels and others are not “proper” Muslims.

There are many stories of change too, in some cases dramatic ones, across the country. Individuals from “low castes” have broken out of their subject status and entire groups have acquired social equality and political power to an extent. The Khaskhelis of Sindh, for example, have memories from just a couple of generations ago of having been referred to as “slaves” by their former masters, the Talpur Mirs. While still mostly poor, they are no longer in a position of dependence, and are represented in positions of influence in politics and in public life. In Makran in southern Balochistan the Darzadas or descendents of African slaves have experienced economic and social mobility. They insist on being referred to as Baloch, and the “pure Baloch” acknowledge their rise and point to important political leaders in the Baloch nationalist movement who are from a Darzada background. Migration to Muscat and mass recruitment in the 1970s and 1980s in the Royal Oman Army, in which all men from Makran were classified simply as “Baloch” is cited as a key channel of mobility.

Zaat Is Race

The story of the Makran Darzadas is clearly about race – the distinct racial origins of the former slaves and masters are clearly visible, and the history of enslavement, trafficking, and emancipation a relatively recent one. But then how different, really, is the story elsewhere in Pakistan (and northern India for that matter). The public silencing of “caste” in Pakistan implies that unlike India there is not much political capital vested in that term. If we can go from some sanitised notion of “class” to “caste”, then what prevents us from pulling away yet another veil covering social oppression? “Caste” after all implies some legitimised “system” of the division of labour. The Oxford English dictionary calls it “any of the hereditary Hindu social classes; any exclusive social class”. The common translation in Pakistani languages is zaat, which is not about social class but related to ‘nasal’ (lineage), or quite literally, race.

The colonial gazetteer of Punjab based on the first population census of the province proposed that the “vagrant, menial, and artisan castes”, which made up over a quarter of the population including all of the groups such as kammis, chuhras and mussalis mentioned here, “…include the great mass of such aboriginal element still to be found in the Panjab”.5 Its author Ibbetson then went on to grade these various groups in terms of their “habits” such as eating carrion and vermin, and lax social mores, while acknowledging that these groups were responsible for much of the industry and the farm labour in the province. Ibbetson and his peers, unlike contemporary Pakistanis, felt free to write about race, culture, caste and class, as complementary features of social inequality, perhaps because they were unperturbed by their own racism. But one and a quarter of a century hence the public silencing and the private obsession with zaat means that the village joke is still on us.




1 Azmat Ali Budhani and Hussain Bux Mallah are co-veterans of many such campaigns.

2 “Scheduled Caste Hindu” is an official census category in Pakistan. Affirmative action regulations do exist, but are largely ignored by government. Many of the castes and tribes included in the “schedule” would be referred to as adivasi or dalit in India.

3 Quom is also the Urdu word for nation.

4 To get a very human sense of this dialectic I would strongly recommend Mukhtar Mai’s autobiography – this ultimate victim and resister of the extended patriarchy remains deeply attached to her family, village and social networks. In terms of integrity, candour, warmth, humility, and insight, In the Name of Honour published by Virago Books stands at the far end of the spectrum from that other Pakistani autobiography published in 2006.

5 Ibbetson, Denzil (1986), Panjab Castes, reprint of the chapter on ‘The Races, Castes and Tribes of the Panjab’, 1883, Government Printing Press, Lahore, p 266.

Economic and Political Weekly January 13, 2007

To read the full text Login

Get instant access

New 3 Month Subscription
to Digital Archives at

₹826for India

$50for overseas users


(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top