Which diversity?

Kumkum Sangari

back to issue

I cannot presume to make a well-informed intervention in the multiculturalism debate, or speak in more than an impressionistic way about the variety of multiculturalisms. There seem to be two major tendencies: those that recognise diversity only to contain or repress it, and those that wish to mobilise diversity for a radical notion of democracy.

Both tendencies result from and respond to the social, economic and political conditions of minorities and dominated groups within a conjuncture marked by the collapse of colonial empires, the weakening of ‘multi-ethnic’ nations, a delegitimation of the state, and an apparent erosion of the hegemony of dominant metropolitan centres through the reorganisation of capital on a global scale. The expansion in capitalist commodity production and infotech has been accompanied by the reformation of ‘culture’ as an idiom of resistance against a variety of ‘centralisms’ and as a domain of self-creation: a concept of culture that seems poised to bypass material social constraints and, paradoxically, to become a source or locus of collective rights to self determination.1 Further, as the integrative capacity of class mobilisation declines, ethnic identities and mobilisation have increasingly become the language of social action.

The multiculturalisms usually grouped under the first tendency – conservative, ‘colonial’, liberal, ‘modernist’ – have been critiqued for their white racist and Eurocentric assumptions, implicit cultural hierarchies, levelling and homogenising agendas, and for their situation within the cultural logic of multinational capitalism. Their politics has been shown to revolve around ‘incommensurable’ identities, reified ethnicities/cultures/nationalisms, ‘harmonious’ assimilation, and market-friendly consumption, and has in fact been opposed by the multiculturalisms grouped under the second tendency – critical, insurgent, postmodernist. These denounce racism, Eurocentrism, hegemonic nationalism, and the liberal pluralist emphasis on the unity and boundedness of cultures; they are cautious about exalting or essentialising group particularisms, and committed to a fluid, dialogic, liminal heterogeneity.2

Liberal multiculturalism sets out to manage cultural diversity within the nation state while postmodernist multiculturalism positions itself outside or across national boundaries. In practice, critical multiculturalisms not only differ in the degree of attention they give to structures of power, but also about whether multiculturalism is to be added on to a democratic project or to stand in for the democratic project. It seems to me that neither type of multiculturalism can come to grips with the diversity of the Indian subcontinent which does not quite conform to a ‘multi-ethnic’ state. Nor does it lend itself to a contest between homogeneity and heterogeneity, provide a ready ground for differential rights, or compress itself into many of the other agendas implicit in multiculturalisms.

The history of subcontinental diversity cannot be confined to the two moments – colonialism and its aftermath – which constitute the multicultural debate in Europe and North America. The mechanisms for social diversification, the apparatus for channeling or repressing it, the ideological ground for confusing cultural diversity with systemic inequality, not only preceded colonisation but overdetermined it. What is more, these recombined with colonial ideologies in many ways as cultural diversity was pulled into competing ideological locales – the British administrative rationale for rule over ‘chaos’, a pluralist Hinduism fashioned to swallow all but Semitic religions, and the secular nationalist compact enshrined in the Constitution.



The numerous individuals or even groups who do not belong to a single ‘ethnic’ identity in which region, religion and language coincide, and the equally numerous overlaps in religious and linguistic identities in India cannot be subsumed within the liberal schema of ‘ethnic’ multiculturalism. In fact many of the differences in the constitution of ethnicity followed on the nature of colonisation. The imposition of colonial rule on the tribal mode of production, on those that were feudal or tributary, the subsequent migration of colonised groups to imperialising countries, and the demography of white settler regimes have produced at least four distinct registers of ethnicity, and each generated many specific subsets or constellations.

India had complex and varied patterns of demographic settlement and migration over the centuries, and, unlike many colonised countries, a non-settler colonisation without a substantial influx of migrants or a wholesale decimation of native populations with its accompanying logics of guilt and reparation. The liberal multicultural notion of discrete ethnic groups suggests marginality or subordination to some dominant group, and seeks a peaceful management of settled differences and identities. In precolonial India, however, influential groups were formed from both immigrants and prior residents, while no single religious group had a history of only victimage.

The static ascriptions of ethnicity (and corollary multicultural policies) have largely been consolidated in relation to immigrants in Europe and America; they often derive from earlier colonial conflations of origin, race, and culture and suppress the dynamic, multifaceted constitution of each so-named ‘race’. Though multiculturalism can function as an attempt to break with the model of hierarchical assimilation in Euro-American countries where immigrants are at the bottom of the ladder,3 it offers only a thin crust of ‘cultural autonomy’ since processes of economic assimilation and homogenisation continue beneath it.



Religious minorities in India cannot, however, be similarly identified as underclasses or victims of forcible transplantation by the capitalist labour market and have been historically subject to both processes of inclusion and exclusion, assimilation and othering. Liberal multiculturalism sets out to deal with the influx of ‘other’ cultures into Europe and America and to ‘integrate’ them, while what we need today in India is precisely to resist the isolation of such pre-given ‘others’ and their preemptive ‘assimilation’ into a Hindu rashtra. In this sense, liberal multiculturalism concedes ‘separate’ and/or incommensurable religions/cultures/communities too glibly and too quickly and can complement communal agendas.

In contrast, postmodernist multiculturalism advocates a relational multiculturalism committed to changing power relations and offers both sympathy and an epistemological advantage to the oppressed. This professedly critical and polycentric multiculturalism rejects a unitary West and absolutised ethnic identities; it celebrates a transgressive, border crossing hybridity envisaged as a protest against or a reversal of colonial violence, and, with the end of formal colonial rule, sets up national boundaries – seen as a systematic principle of differentiation – as the major remaining obstacle to be overcome. Solutions are sought in an autonomy for restructuring intercommunity relations within and beyond the nation state according to the internal and partially overlapping imperatives of diverse communities.4

However, in this decentred model, all types of mobility – of the individual or group, of labour or culture – seem to occupy a level plane of equivalence; material structuration is overshadowed by a spatial concurrence of diversities that simulate the synchronicity of the marketplace. It leaves little room to ask questions about how transnational intercommunity alliances will resist economic imperialism and the finitude imposed by economic exploitation or to formulate a practical and ethical horizon against which to pose issues of inequality and common rights.



It seems to me that if mapping the specific relationalities and antagonisms, the determinate material coherences and disparities within which cultural diversity has been constituted is a necessary first step, then the cognitive grids of both a normative liberal and a liminal postmodernist multiculturalism can be pre-emptive. Can we jump to either an ‘ordered’ or an interstitial multiculturalism before ‘descriptions’ or even approximations are in place? Religion, patriarchy and culture still await disentangling, given the combination of different religions with similar patriarchal arrangements, different regional cultures with common religions, and different religions with shared regional cultures in India.



In fact it is from such diverse ensembles that ‘ethnic’ groups abroad have been culled by liberal multiculturalism in a series of conflationary moves. What is more, culture is an ideologically laden and debatable term, especially for feminists. ‘Culture’ may sustain women but it also curtails their freedom. Should culture be defined as a contested field of values and meanings that determines practices or itself as a set of practices, a way of life, a synonym for society. And in the latter case whose description of culture will be privileged?

An equally pressing concern among those struggling against oppressive structures is how to align democracy with a cultural diversity that was also embedded in and shaped by numerous inequalities. In this respect too it is difficult to square a liberal communitarian pluralism that suggests differential rights and/or laws on religious lines with an egalitarian and feminist commitment to a secular, multireligious society. This liberal communitarianism does not take into account either the patriarchal resistance by the state and religious groups to a universalisation of rights for women, or the pragmatism of a Hindu right that can take selective recourse to both homogenisation and religio-cultural differentialism.

Liberal multiculturalism is based on a distinction between resident majority and incoming ‘minority’ cultures, and has been preoccupied with their religions and patriarchies. In fact a sizeable chunk of the multiculturalism debate has hinged not on ‘culture’ but on the spectacular emblematisation of diverse religions and patriarchal practices from veiling to polygamy.5 Ironically, many of the patriarchal privileges that are now being coded as ‘cultural’ rights of immigrants in certain types of Euro-American multiculturalism have been defended from the colonial period onwards in India as ‘religious’ rights. And both these redefinitions have helped to efface the specificity and distinctions between legal, patriarchal, and religious plurality as well as those historical processes through which patriarchies were often jointly constituted – in the elisions of 19th century liberalism – across the colonial divide.



Homogeneity and heterogeneity, then, cannot be taken on board either singly or wholesale. They cannot, like particularism and universalism, be taken by themselves. We need to talk concretely about types of heterogeneity and homogeneity, about types of universalism and particularism, as well as the class and other discriminatory social relations that underly particular forms of diversity – on qualitative, ethical and egalitarian grounds. Subcontinental cultural plurality is not open to an unproblematic celebration – whether as civilizational splendour or as a sign of Hindu tolerance. If it was rich in religious and other diversities it also had patriarchal and casteist features. Thus, the struggle against patriarchies may be pitted against forms of homogenisation or against forms of heterogeneity, or both, depending on their nature in a specific conjuncture.

Take the case of the personal laws that govern religious groups and are upheld as a sign of religious pluralism. These laws have in fact helped to homogenise five fixed religious identities – Hindu, Islamic, Christian, Judaic and Zoroastrian – from a welter of sects and customs, prised apart the relational history of religions, separated them into bounded units, patrolled borders and inhibited conversion.

Hindu personal law forecloses even a theoretical space for agnostics, partial converts, atheists, or sects who do not wish to be defined as ‘Hindu.’6 In all this, it bears a striking affinity to Hindutva that represents all past religious differences and dissent as differences and dissent from within – tribals as incomplete Hindus, Buddhists and Sikhs as breakaway Hindus who can be intermittently incorporated – and invokes Hinduism both as a supreme assimilative power and a practice of ingesting diverse traditions and sects. This same over-inclusive and over-elastic legal definition of Hindus has also been central to defining scheduled castes.7



Or, take the Hindu right’s demand for a uniform civil code. This too has to be seen in relation to the fact that the Hindu right can not only appropriate both universalism and particularism but also that both strategies are put to work towards the same goals. If one strategy of the Hindu right is to aggregate, stereotype and demonise Muslims and Christians as monolithic transnational communities, then another strategy, sustained by a colonial legacy of classification on lines of pure origin, impure racial inter-mixture and native converts, is to differentiate between Muslims and Christians on particularist lines of descent as aliens and natives for purposes of ‘reconversion’.

There is a case, then, not only for reconnecting issues of religious identity with those of class, caste and gender, but also rethinking sharp and soft definitions of religious identity in ways that go beyond the present purview of multiculturalisms. This would involve rejecting a dehistoricised definition of religions as immutable or exclusive, a definition that coincides with homogenising community claims, and is amenable to exploitative political instrumentalities which encash ascriptive constituencies.



In the Indian context a critique of the essentialist or hegemonising drives of the nation state cannot usefully be made by invoking essentialised religious particularity and communities. Nor can it be made by jeopardising women’s rights – rights that the state itself is reluctant to universalise. It is only by seeking a guarantee for the inalienable rights of all women as citizens that women can be empowered to actively challenge and redefine the present political arena. As I have suggested elsewhere, if new laws are devised they would have to deal with the multiplicity of patriarchies and move, not towards a universal principle that picks up their common elements and flattens them into a ‘core’, but towards imagining a concretization of the universal that can take into account both similarities and differences.8

In sum, I believe there has to be, at once, a case for certain types of homogenisation in the arena of rights, and a case for encouraging certain forms of plurality. Religious plurality, mired as it is in patriarchal practices and the caste order, is unthinkable without a transformative political agenda and has to be affirmed within an egalitarian commitment to social justice.



The scattered points I have made can perhaps be pulled together by looking at some of the contradictions and potentials that inhere in religious conversion. Conversion can occupy the same coercive material and patriarchal structures of power as institutionalised and politically constituted religions do, or it can signify acts of rejection, dissent, choice and freedom of movement within a multireligious formation. It can carry the intent of homogenisation and work as a mechanism of pluralisation. It can be used to strengthen the illusion of a non-proselytic and accommodative Hinduism or, in the shape of Dalit mass conversion, it can jeopardise the majoritarian pretensions of Hindutva from ‘below’. Conversion, thus, can be fed into the demographic paranoia of the Hindu right or it can further the interests of a secularism that rests on a recognition of the conditions and mechanisms of religious plurality.9

Conversion has usually been confined to a semantic of instant and dramatic change of faith that lends itself to sensational dyads of force and illumination, pragmatism and belief. It can, however, be given an active role in weaving the fabric of religious plurality, and shown to carry the weight of numerous other temporal frames, pathways, agencies, shapes, transactions, and levels of self-consciousness in the partial or the substantial shifts from one belief system to another.

Historically, these transactions have been multilayered, deep or shallow, combative or dialogic; and within them curiosity, contiguity and theological compatibilities have been more decisive than conquest or clientage. They have transgressed boundaries but at times also effaced them. They have ranged from sharp repudiation to long drawn out, ongoing processes of gradual assimilation, absorption, permeation, diffusion described by D.D. Kosambi, Richard Eaton and other historians. (These were not always unilateral: the ‘assimilation’ of tribals into the caste order was often accompanied by the persistence of many non-Sanskritic beliefs among tribals and their incorporation by upper castes.)



These processes, that did not always entail moving from one self-contained religion to another, but, as it were, from one ‘blend’ to another, were in turn determined by a number of material factors. For instance, shifts in the mode of production, as well as caste mobility, class processes and occupational diversification worked beneath the assimilative logic of Sanskritisation, Brahminisation, Islamicisation, Sikhisation and Christianisation.

Ironically, even though these logics have been increasingly tethered to a separation of the ‘signs’ of religious identity among loose, unclassifiable or converted groups, each wave of such desyncretising, purifying and homogenising campaigns, each entry of low castes, untouchables and tribals into Hindu, Christian or Islamic frameworks, has often led to new intermixtures. Even the Vishwa Hindu Parishad’s brand of Hinduisation is creating new double or mixed identities. For instance, some Christian adivasis keep photos of Jesus and Hanuman and describe themselves as Hindu-Christians in Dangs.10

Both temporalities of conversion – sudden and gradual – thus have produced ‘partial conversions’ and created linked or overlapping religious networks by altering the relationships between belief systems, and by propelling reconfigurations within belief systems through accretion, overlayering, reinterpretation, substitution or fusion of gods, beliefs and practices. This relative lack of thoroughgoing transformation, this frequently noted asynchrony between change in belief and change in the totality of social usages and religious practices, has some major implications.



First, the very fact that different degrees and types of change have occurred demands a requestioning not only of the conflation of religion with culture but also the inflation of religious primordiality. The mechanisms of conversion may often have worked also as mechanisms for splitting religious belief from ‘culture’ in the broader sense, since the purview of a particular set of beliefs could be far narrower than the provenance of culture. The more pragmatic variants of conversion too can signal the priority of other affiliations over and above religious belonging, or display religion (as in some instances of inter-religious marriage) either as less important than or as an obstacle to affective personal bonds. And, significantly, conversions based on an explicit rejection of caste or patriarchy can drag religion quite consciously from birth-bound and primordialist into willed definitions and agential terrains.

Second, it bears on the relation between religious diversity and ‘inauthenticity’. Precisely because conversion did not imply thoroughgoing transformation (even of religious belief), involved both change and continuity, it has been, ironically enough, a mechanism for the production of religious diversity within a multi-religious formation – in which the historical crystallisation and enactment of religious identities did not conform to a model of five monolithic religions.

Over time, incomplete and partial conversions produced alterations within and between belief systems, divergences and carry-overs that were in fact a principle of cohering diversities. It is this contingent character of conversion, as producer of a cohering diversity which does not or cannot absolutise difference, that is labelled ‘inauthenticity’. And it is the consequent overlaps and contingent reconfigurations of religions that have come under the combined pressure of colonial Hinduisation and Islamicisation, the inclusive ‘legal’ definition of a Hindu in personal law, the anti-conversion agenda and hegemonic aspirations of Hindutva. Partial conversions, that blur the distinctions between ‘minority’ and ‘majority’ have become a more vulnerable target of competing claims than those ‘complete’ (or sufficiently ‘complete’) conversions that can be slotted under ‘separate’ cultures.



Yet if inauthenticity is a condition of genuine religious plurality, a principle and potential for cohering diversities, it should be open neither to the Hindu right’s designs – whether as object for ‘inclusion’ or reconversion – nor to state legislation.

A secular valuation of composite culture would place continued usages along with rituals and customs that differ from designated religious identities into a framework of ‘partial conversion’ within a multireligious formation. It would see conversion as not just a change of religion by individuals or groups but imbricated in changes in religions, in their genesis and mutation, in the production of discrete ambitions and overlapping realities.11 Conversion has created interfaces, parallelisms or meeting grounds between religions, a low level interweaving or overlayering of practices and beliefs, composite cults, selective syncretisms, and ambiguous identities.



A secular position would argue for more flexible religious identities rather than the ‘completion’ of unfinished or partial conversions. The Hindu right, however, pulls all such continuity and ‘incompleteness’ into an absorptive and supposedly pluralist Hinduism, then presses them into the single teleology of ‘reconversion’. Moreover, reconversion is conceived as a demographic onslaught on Muslims and Christians by separating the descendants from the converts, the outsiders from the insiders.

At the same time, the exclusivism embedded in laws has to be interrogated. We need a conception of law that does not routinely seek or depend on declarations of a finished religious identity, preserves the potentials for healthy fluidity but does not humiliate minorities. This can only come about through a perspective on law built on the understanding that wherever religious beliefs and communities are at stake in the near political future there will be – as there has been in the near political past – a contest between the processes and agencies of both fixation and change. In these circumstances laws should under rather than overdefine religious identity, cease to be preoccupied with fixity of religious definitions or communities, honour the individual’s right to choose where to belong, recognise existing diversity, processes of change within and between denominations, and maximise certain types of flexibility. Can we afford a legal definition of religion that precludes religious intermixture or change and which can be used to legitimise ‘separate cultures’?



Third, a recognition of partial religious transformation may assist in formulating a notion of the secular that is not eternally fixated on a majority but can leave space for its disaggregation and recombination, and for the emergence of new religious identities. Present definitions in fact obstruct the multiplication of sects seeking to be independent of Hinduism both in practice and in legal self-designation. Do we want merely a ‘prefabricated’ and regimented religious pluralism that has stemmed from the combination of the longue duree and deep temporality of precapitalist belief systems with the short-term rationalities of a capitalist political economy and electoral political logics? Or do we want a secularism that acknowledges and honours the mechanisms of religious pluralisation that have developed from the historical specificities of a multireligious social formation, and is quite distinct from pluralist Hinduism in which ‘flexibility’ is a majoritarian device?

Do we want to read our history as inevitably tending towards separate religions (and therefore cultures) or as one that had, and may still have, other possibilities which the Hindu right is trying so desperately to annex or foreclose – as is evident in its repeated and violent attack on syncretic cults? For the Hindu right, partial conversion is at once a threatening and a handy rationale for purification and reconversion.

However, partial conversion can provide secularism a way for uncovering the relationality of religions, for confronting the facts that no subcontinental religion has a single set of customs or cultural practices that adhere to it, and that both the presence of an active multireligious formation and the movement between belief systems can often loosen the grip of primordiality or entail a reappraisal of religion. This would enable secularism to go beyond the protection of hermetically sealed enclaves as well as beyond laws that preside over unalterable, standardised or normative identities and merely stipulate procedural principles for ordering a (static) multireligious society.



It is especially important for Indian secularism, constitutionally committed to democratic principles, reformist intervention and a multireligious society to re-evaluate conversion. The fact that conversion has been a mechanism for diversification and has taken place, whether as process or conjunctural event, more often than not, within the parameters of hierarchical power relations – state, class, institutions, caste, patriarchies – suggests that it is these rather than conversion that must be isolated as objects of critique.

In a context where religious plurality and structures of oppression coexist, and entail protection of the one alongside critique of the other, finer distinctions need to be made from firmer ground. This involves ensuring that the mechanisms for religious diversification are not stifled (legally or otherwise), that exploitative particularisms which close the avenues for pluralisation are resisted, and that religions are singled out only in areas where they function directly as guarantors of patriarchal, caste and communal oppression. A freer and more flexible formation of religious groups and identities will in the long run assist feminist and democratic struggles for substantive and consciously shared concepts of inalienable rights across religious differences.

In conclusion, multiculturalism cannot simply be grafted on Indian secularism while a national secular, in so far as it is still a reclaimable ideal, can still work along the grain of certain types of religious pluralisation. The case for plurality and inauthenticity requires not so much a leverage from critical, insurgent, or postmodernist multiculturalism as a refashioning of the egalitarian secularism we have inherited by tethering it to the historical specificity of the subcontinent. This in turn may reposition the emergent multivocal identities of the global diaspora (the primary constituency of liberal and postmodernist multiculturalism) in relation to the complex identities and religious pluralisation on the subcontinent, and resituate the interplay of old and new inequalities.

What is more, the way the dilution of immigrant ‘ethnic’ identities is ignored by official liberal multiculturalists abroad and resisted by the Hindu right may not be entirely discrete phenomena. The management of religious and cultural diversity, the recreation of minorities and majorities, may at some level be connected transnational processes. Consider that a minority can simultaneously be reconstituted as a majority in the era of globalisation: if ‘Hindu’ minorities are subject to racism and the objects of ameliorative multiculturalism abroad they are also being offered an originary and permanent ‘majoritarian’ status in India by the Vishwa Hindu Parishad.



1. On some of these issues see, Terence Turner, ‘Anthropology and Multiculturalism’, in David Theo Goldberg (ed.), Multiculturalism: A Critical Reader, Oxford, Blackwell, 1994, pp. 407, 419-20.

2. See Slavoj Zizek, ‘Multiculturalisms, or, the Cultural Logic of Multinational Capitalism’, New Left Review 225; the essays in Multiculturalism, edited by Goldburg, especially Henry E. Giroux, ‘Insurgent Multiculturalism and the Promise of Pedagogy’, pp. 326-27, 336.

3. See Kevin Macdonald, ‘Identity Politics’, Arena, June-July 1994, pp. 19-20.

4. See for instance Robert Stam and Ella Shohat in Goldberg, op. cit; Robert Stam, ‘Eurocentrism, Polycentrism, and Multicultural Pedagogy’ and Ella Shohat, ‘The Struggle over Representation’, in Roman de la Campa and Michael Sprinker (eds.), Late Imperial Culture, Verso, London, 1995.

5. As is evident, for instance, in the ‘Feminism and Multiculturalism Debate’ in Boston Review, October-November 1997.

6. See Kumkum Sangari, ‘Politics of Diversity: Religious Communities and Multiple Patriarchies’, Economic and Political Weekly 30(51 and 52), 23 and 30 December 1995, pp. 3287-3310, 3381-3389.

7. See Marc Galanter, Competing Equalities, Oxford University Press, Delhi, 1991, p. 144.

8. Sangari, ‘Politics of Diversity’, op. cit.

9. Some of these themes and the argument that follows have been elaborated in Kumkum Sangari, ‘Gender Lines: Personal Law, Common Law, Conversion’, Social Scientist 27 (5-6), May-June 1999.

10. Reported by Ghanshyam Shah, ‘Conversion, Reconversion and the State: Recent Events in Dangs’, Economic and Political Weekly 34(6), 6 February 1999, p. 315.

11. See Kumkum Sangari, ‘Tracing Akbar: Hagiographies, Popular Narrative Traditions, and the Subject of Conversion’, in Neera Chandoke (ed.), Essays in Honour of Prof. Ravindar Kumar, Tulika, Delhi, forthcoming.