Caste Identity Versus Class Solidarity: Some Speculative Notes

January 26, 2013

P. K. Vijayan


Much has been discussed and written about caste politics, and about caste and politics. My purpose here is not to revisit and rehearse these debates, even if I do refer to some of them in passing. Nor the aim of article is to analyze the multiple alliances and conflicts that take shape through mobilizations and consolidations along caste lines. These have occurred historically, or continue to occur contemporarily. I wish to focus here on an issue that is increasingly taken for granted, in these debates as well as in the processes around caste politics. The issue is the understanding of caste as a political category. Specifically, I want to focus on how caste is often treated as political primarily – often solely – in terms of the political formations that arise out of individual caste identities. I seek to open out the politics of this treatment. While the validity and importance of such an understanding of the politics of caste are undeniable, my contention is that it is not sufficient to understand the politics of caste in these terms. The obvious question then is, how else can we, or should we, approach this issue?

Caste Politics and Politics of Caste
But even before we address that question, it is perhaps essential to distinguish between two seemingly identical, apparently interchangeable terms: ‘caste politics’ and ‘the politics of caste’. The first term, ‘caste politics’, arguably refers to political processes, conjunctions and dynamics to which caste is central. These processes focus on, engage with and/or emphasize issues of caste. So for instance, the formation of organizations like trade unions, social service groups, special interest groups, political parties, etc, along the lines of caste identities, would constitute ‘caste politics’. Conversely, the second term, ‘the politics of caste’, refers to the terms on which caste in general becomes a politically deployed category, within a general field of similarly deployed categories, such as class, gender, religion, region, language, ethnicity, etc. The important point here is that this deployment need not be only in the field of organized politics. Unlike the ambit of ‘caste politics’, ‘the politics of caste’ can be played out wherever, whenever and however the issue of caste may play a determining role – marriage, tenancy, employment, associational affiliation, etc. So, for instance, when Andre Beteille writes of ‘the peculiar tenacity of caste’ (EPW, 31 March 2012), he is implying that the longevity of caste as an issue is itself politically motivated, and is therefore referring not (just) to caste politics but to its underlying ‘politics of caste’ – i.e. to the political disposition that promotes caste politics. (Indeed, Beteille’s argument is that caste as an everyday experience is weakening, but caste as a vehicle of political mobilization is growing in strength – but I will return to this argument later.) Here, now, we can discern the specific form of the distinction between the two terms: while ‘caste politics’ is necessarily a part of any ‘politics of caste’, the latter is much more than the former, it encompasses and addresses the political aspects of the phenomenon of caste in all its dimensions.

Caste and Class
This distinction is crucial to understanding the politics of caste. In fact, the persistent collapse of this distinction has generated some fundamental controversies regarding the political nature of the phenomenon of caste. Among some left political formations, the very validity and viability of caste as a political category has been questioned, on the grounds that it distracts from the actual political dynamic in society. Caste – like gender, race, ethnicity, etc – will disappear once the class bases of these opiates are countered. Such an understanding is a direct product of the collapse of the distinction noted above, namely, the politics of caste is reduced to the identity-form of caste politics, wherein political mobilization is solely on the basis of caste identity, leaving no room for class (or gender, ethnicity, region, etc) to feature as a determining element in the constitution of the mass identity. Caste here is understood as an obstacle to the emergence and concretization of class solidarities [1] (a very crucial distinction is maintained here between solidarity and identity, which I will return to shortly). The legitimacy of caste as a viable basis for political mobilization is thus challenged, or at least critiqued as being insufficient and/or abortive.

A fundamental difficulty here lies in the perception of caste as actually and conceptually structured like class. To the extent that caste is frequently associated with profession (the association is more intractable the lower one goes in the caste hierarchy), and consequently, with the income range associated with those professions, it bears some class-like attributes. Further, in the cases of both caste and class, professions involving directly physical labour are generally located lower in the hierarchy, while those involving less physical labour are located higher. This has only served to reinforce the assumed equivalence between caste and class. However, it is necessary to remind ourselves that caste relations (inter-caste and intra-caste) are constituted and operate in ways that are very different from class relations. While both class and caste may be inherited, may persist over generations, may appear unchangeable, there are some crucial differences. Caste itself is (in sociological terms) an ascribed identity, while class is an achieved condition. In this sense they are not only categorically different, they are ontologically different. It is for this reason that one’s class status can change with greater ease than one’s caste identity. One’s class status is a function of the larger set of economic relations in which one exists, as well as of the economic conditions carved out by those relations. One’s class status can change (for better or for worse) without there necessarily being a change in those larger economic relations and/or conditions. For example, if my income suddenly increases or decreases, and I am shifted out of my current class position, that alone does not change or even affect the overall class dynamics around me. One’s caste identity, however, is neither so relative nor so independent of the larger social relations within which caste identities are constituted. They cannot be changed at the individual level (even if they may be disavowed or disowned), but historically, have changed only for entire communities, over long periods of time, in the course of which the perceptions of the caste-status of a particular caste have adjusted for material changes in the social relations between caste communities. That is, extended changes in the material relations between caste communities have brought about changes in the social, or caste, relations between them. [2] A case in point here might be the changing caste status of the Nair communities in Kerala from ‘shudra’ to ‘kshatriya’. [3]

However, even this clumsy possibility for mobility within the caste order is limited to the middle orders, and is absent at the two (defining) ends of the caste hierarchy. These are marked by presumptions of extreme purity (Brahmins) and extreme pollution (Dalits/ outcastes). These presumptions at the extreme ends of the caste order are inflexibly maintained. There is no possibility for Brahmins (as a community) to lose high-caste status, just as there is no possibility for Dalits (as a community) to lose outcaste status. [4] They cannot change precisely because they are definitive – they define the limits of the caste system. There are several additional factors of some consequence here: (a) if we understand ‘caste’ in general as referring to the larger ‘varna’ system, then each varna has a number of jatis within it; (b) there are regional variations within the same caste, and these variations are often concentrated in specific regions (whether or not in the form of specific jatis); (c) the degree of caste homogeneity increases, and the variety of intra-caste jatis decreases, as we move up the caste hierarchy, and vice versa (i.e., there is a very large number of jatis jostling at the bottom of the caste hierarchy, which are often region specific, while there are much fewer at the top, and these are often pan-Indian); (d) the extent to which caste is associated with physical labour tends to increase as we go down the caste hierarchy; and (e) although caste is denied by other socio-religious communities (Muslims, Christians, Sikhs, etc), the fact is that it persists and prevails in them almost as much as in the ‘Hindu’ community. Taken together, these points suggest that, at the top of the caste pyramid, labour and/or profession is more generalizable, because it is less associated with specific physical (and specifically physical) labours or economic functions. Therefore, these (upper) castes and their economic roles tend to be more applicable across regions, spread out, rather than tied to particular regions. [5] Furthermore, because of this ability of the upper castes to participate economically in a cross-regional way, they also exercise a cross-regional (pan-Indian) control over the economy – and consequently over the administrative, bureaucratic and judicial processes (in short, over the Indian state).

A very crucial consequence of this pan-Indian presence and control of the upper castes has been the emergence of nationalism as essentially an upper caste identity. As a result, the lexicon (or vocabulary) of the discourse of national politics is, by default, an upper caste one. Consequently, any oppositional political discourse has been forced to adopt either an alternative upper caste nationalism – e.g. the Jan Sangh or the Bharatiya Janata Party – or to mobilize around an alternative, lower caste identity – e.g. the Janata Party in the seventies or the Bahujan Samaj Party much later. This was the only way available to articulate their politics from within this discourse of identity politics. The politics of representation that formed the basis of this identity politics, was first introduced in the nineteenth century, under British rule; the Indian National Congress (or the Congress Party as it is commonly called today) was the first to exploit that form of identity politics to present upper-caste hegemony as representative of ‘Indian’ identity. The Congress has subsequently continued to maintain a powerful hold on this upper-caste hegemony, although from the nineteen eighties it has had to compete with the Bharatiya Janata Party and other Hindu right groups. The latter have recently enjoyed as much (or as little) success on the national stage, broadly speaking, as the Congress, in this endeavor. Lower caste identity based political formations, very significantly, have enjoyed only regional success, rather than national (even if, in the process, they have managed to make some impact on the national political stage). [6] However, no lower caste platform has shown any possibility of emerging as a national force, primarily because (as we observed above) there is no national lower caste presence in quite the same economically enabled way as the national upper caste presence. On a different tack, political agendas that have sought to introduce an alternative discourse – that of class, equity, equitability, secularism, democratic rights, etc. (in whatever form) – have gone one of two ways. Either they been subsumed by the regional-caste formations noted above (e.g., the Janata Party); or they have been confined to states where the alternative vocabulary was already widely present and accepted, and therefore could be articulated and elaborated politically (e.g., the mainstream Communist parties, especially in Kerala and West Bengal). In this sense – in terms of the impact of (upper) caste identity in shaping the national political discourse – a politics of caste has been at work, without being noticed or acknowledged as a politics of caste, for several decades – perhaps even going back to the decades before independence. That is, under the overt projection of a secular, democratic, liberal political framework, there has been a persistent and powerful upper caste bias that has resisted and prevented the possibility of an alternative political framework emerging. Through its implicit and sustained promotion of (caste) identity as the basis of political mobilization and consolidation, it has either been unwilling to, unable to or has expressly not wanted to, promote class solidarity (except whenever that solidarity coincided with (upper) caste solidarity). Objectors to ‘caste politics’, such as Andre Beteille, have thus failed to take into account the larger and much longer-duration ‘politics of caste’ that has been the central theme of all politics in the country since well before independence.

Solidarity vs. Identity
The crucial point here is what I had mentioned in passing earlier – the distinction between the ideas of solidarity and identity. Identities – especially shared identities – are not only ascribed (precisely because they are shared), as noted earlier, but they assume the characteristics of a property – a set of defining or definitive attributes that are owned by the holder of the identity, indeed that define the identity. These attributes need not necessarily be manifested physically, as symbols, or marks: they may be beliefs, values, norms, even languages or dialects. However, once this is understood, we can observe something quite unusual: the conventional way in which we think of ownership gets inverted. The attributes that define the identity make their possessors part of a larger community of others who share the same attributes; consequently, they are not the owners of those attributes as much as that they become owned through those attributes. [7] That is, these attributes mark the individual who owns them as belonging to the community that shares them: the individual is rendered object, the property of the community. The totality of the shared attributes that define the individuals in the community then constitutes an imaginary or notional ‘body’ of the community at large – the notional (or discursive or ideological [8]) ‘body’ that owns its individual members. This notional body claims the actual body of the individual by ‘inhabiting’ – literally, possessing – the actual body of the individual with its community rituals and practices. The affect-based rituals and practices of the community that the individual participates in, and shares with others, synchronize the many real bodies of the larger notional body. [9] In other words, the notional body is constituted of the affective (emotions, feelings, psycho-social impulses and affiliations) practices and rituals that individual members of the community come to share and perpetuate. Caste-based identities, specifically, in this sense, are constituted around the notion of the particular caste, the ‘caste-body’, as it were, and its specific caste rituals (apart from the caste profession/occupation – which in any case generates a set of bodily practices specific to that caste or profession).

The caste system in general is fairly totalizing – in the sense that one either possesses a caste identity or one is out-caste, ‘mleccha’. Hence, attempts to disown one’s caste identity by repudiating parts or all of the notional body – the caste-body – do not necessarily take one outside the system of caste in general. The caste body continues to claim – to own – the individual even in such a case of repudiation, by virtue of a primary attribute, birth. And this is true even if that ownership claim is not exercised by the ‘owner’ (the caste body) or acknowledged by the ‘owned’ (the individual). [10] Most significantly, from the point of view of political mobilization, the affective hold that caste maintains and retains over the individuals of the community actively and passively, implicitly and explicitly, overtly and covertly, interferes with the possibilities and processes of building political solidarities not based on identities. These affective holds are evident not just in the deciding of marriages, for instance, but in customs governing cuisine and eating: vegetarian/non-vegetarian; with or without garlic and onions; extents of non-vegetarianism (ranging from egg-eating to fish-eating to chicken-and-mutton eating to pork-but-not-beef eating to beef-but-not-pork eating, and other such permutations bordering on the absurd); who cooks meat, where it will be cooked, where served, how served, and very importantly, with whom consumed – and so on. These caste specific customs and conventions are crucial in deciding the extent of inter-caste interactions. And when a particular caste-based community is – as is often the case in the lower rungs of the caste pyramid – specific to a region, it would become particularly difficult to articulate cross-regional solidarities. In this sense, Beteille’s argument, noted earlier, that caste as an everyday experience is weakening, while it is growing as a political consolidation, makes an untenable distinction between the two: simply put, a weakening of the affective hold of caste – which would constitute the basis of its everyday experience – would automatically permit the emergence of voluntary class-based solidarities – which unfortunately, have happened few and far between, thereby giving the lie to Beteille’s argument.

In such circumstances, class-based solidarities tend to emerge in a fragile and tenuous way, and tend to be on a much smaller scale than the identity-based consolidations noted earlier. This is because the defining characteristic of all solidarities is that, unlike identity-based consolidations, they are substantially voluntary. They require individual members to actively choose to be part of the solidarity. Even when such solidarities are the effects of common economic need, or commonality of economic circumstances, they remain primarily voluntary, so long as there is no bond or relation between these individuals other than the economic need that requires collective action from them (or the recognition of the political need to mobilize on and around questions of economic needs – which amounts to the same thing). As such, unless there is a sustained sense of economic need that is widespread, such solidarities are likely to remain confined to being generated for, and addressing only, the specific economic need in question. This returns us to the point that they tend to be on a much smaller scale than identity-based consolidations. Capitalism of any and all hues would thrive on identity politics, as well as thrive in a scenario dominated by identity politics, because of the following reasons:

(a) This form of politics dismisses the economic disparities that all forms of capitalism ceaselessly generate. This kind of politics may occasionally address such economic disparities from and through caste-based collective action, but they are generally dismissed in favour of maintaining the cohesion and coherence of the caste-based notional collective body. [11] This in turn also corresponds with and serves the organizational structuring of all forms of capitalism, which are generally task-and-specialization based. It thereby facilitates the operations of capitalism rather than challenging them.

(b) The transformation of the individual into a property – or more accurately, of the relations between the individual and the collective into a relation of mutual ownership – also harmonizes well with the commodifying character of all capitalist systems, rendering the individual as well as his/her community into commodities at the service of the operations of capitalism. (At the very least, there is a familiarity with the sense of being property – of belonging to someone or something – that can be most advantageous for any form of capitalism to exploit.)

(c) Class-based solidarities would prioritize and address specifically the economic disparities generated in and through capitalism, thereby posing serious challenges to the continued functioning of the capitalist system. Identity based politics evidently prevents – or at least hampers – the formation, consolidation and widespread activation of such solidarities. Such solidarities cannot take shape precisely because of the demands of affiliation and allegiance to the community that are integral to identity politics. Consequently, there is an inevitable resistance to the voluntary formation of solidarities by transcending the identity-community. And perhaps the biggest challenge is – as is so often the case – when class and caste overlap to the extent of appearing almost identical, or at least inextricably related. Then, caste identity is mistaken for class solidarity, and a politics of caste is mobilized as if it were – or in place of – a genuinely revolutionary politics, that aims for systemic transformation.

My arguments so far suggest that one cannot ignore the power of caste bonds. They cannot be dismissed as false consciousness, or even as ideological constructs more or less independent of the economic circumstances they exist in, however persuasive this might appear to be. The caste system operates on a logic of ownership that transforms the individual (indeed the community itself) into an object. That is, the individual is an object, a belonging of the larger identity-community as well as a subject belonging to it. In the process, it shackles and stifles the ability of the individual to think beyond the immediate hold of the community. As such, caste works in favour of capitalism, as a ready and already regulated system of ownership of bodies and therefore of labour. This is particularly true of the lower rungs of the caste pyramid, where caste and class positions overlap significantly; as we move up the pyramid, and the correspondence between caste and class is less straightforward, the correspondence between caste and occupation or profession also weakens. That is, as we noted earlier, for a certain range of professions (especially those entailing relatively less physical labour), a degree of caste flexibility is increasingly to be found. These professions therefore permit a greater degree of voluntary adoption by the individual, rather than being ascribed through caste affiliation. That is, the profession may be chosen by the individual, rather than ascribed to him/her by his/her caste. But this voluntariness does not necessarily mean that resistant class solidarities can emerge in these more flexible levels of the caste pyramid (although implicitly complicit class solidarities can and often do). This is because of the voluntary basis of the involvement in the profession. [12] That is, the fact of choosing the profession voluntarily makes the possibility of resistance to (the administration of) that profession more unlikely.

The central paradox that emerges here is that, on the one hand, an essential prerequisite for the formation of resistant class solidarities is voluntariness – in the sense of an active willingness to be part of a larger solidarity not based on affective claims of identity; on the other hand, the conditions that make possible such voluntariness do not encourage the formation of solidarities. They open the possibility of voluntariness precisely so the volunteering individual can participate in those professions – i.e., participate productively in that political economy. The facilitation of such voluntarism is not towards some lofty end of producing the ‘autonomous’ or ‘liberated’ individual (all claims to the contrary of capitalism notwithstanding), but towards the greater efficiency of functioning of this mode of production. As such, this voluntariness is only seemingly so, is voluntary only in relation to the professional and occupational predispositions of the caste system (which it thereby appears to be breaking from). It is itself predisposed towards the service of the conditions that have facilitated such apparent voluntarism, rather than against them. In this sense, it is predisposed by the very conditions of its own production to militate against any program for the voluntary construction of resistant class solidarity. We have already observed that the middle layers of the caste pyramid permit a degree of flexibility, while the two ends are more or less the defining and absolute – and therefore inflexible – forms of caste. It follows from this that even the degree of apparent voluntarism permitted in the middle has its caste limitations as much as its class or economic limitations. As such – and this is absolutely crucial to understanding the dynamics at work here – the caste system helps to reproduce and perpetuate inequities and inequalities that are themselves perpetuated – if not actively generated – by the capitalist political economy. This is true even when the system appears to be in opposition to it, as for instance through caste-based mobilizations against social and economic injustices, or for social and economic rights. To the extent that these are ‘owned’ by the caste community, within a larger system of caste-defined operational limitations within the caste system, such rights and redressals remain confined to the caste identity. They are of consequence only within the terms of that identity-system, not in terms of a larger, more pervasive and more fundamental process of guaranteeing rights and liberties that would entail the dismantling of the capitalist political economy itself. In short, it is clear that the caste system is deeply integrated, in both ideological and material ways, with the political economy of capitalism, however one may understand that term. [13]

The obvious question that arises, in the wake of this admittedly rather bleak understanding of class and caste relations, is the classic one: “What is to be done?” I do not presume to know the answer. However, by way of making a tiny contribution to a gargantuan debate (not to speak of a gargantuan project), I would like to spell out some of the questions and possibilities that arise from the above discussion. Perhaps the first need is for a full and clear recognition of the place, function, operation and significance of caste, in the formation of political communities and mobilizations – towards which this paper hopes to contribute. To attempt class-based mobilizations while ignoring, dismissing or down-playing caste is to embark on a doomed project. [14] Beyond this, there is also perhaps a need to examine how the paradox of ‘voluntarism’ noted above can be cultivated into a full-fledged systemic contradiction: this is by way of exploring political strategies beyond those of the politics of caste. But perhaps the task of greatest urgency is to address and dismantle the logic of ownership that operates in both capitalism and the caste system – to examine and critique how it might be similar in both cases, how different, what are the points of connection and mutual reinforcement, and what might perhaps be contradictory between them, etc. It is also necessary to examine and critique the principles and procedures by which this logic is replicated, reproduced and sustained in both cases. Intuitively, there is a strong possibility that at the heart of this logic of ownership in the caste system as well as in capitalism, lie the principles of inheritance and patrilineage. Because these are profoundly affective – and therefore ideologically loaded – even as they are profoundly material – and therefore possess a concrete presence and force – questions of inheritance and patrilineage are extemely difficult to challenge. And yet they are most likely to prove central to deciding the course of any genuinely democratic political change. Tackling this will entail examining a range of much wider questions, and will call for the deployment of important insights into these areas generated in feminist debates, for instance. But perhaps most significantly, tackling this will entail the promotion and propagation of the possibility of alternative, collective systems of ownership: doing away with inheritance altogether, for instance, for if all ownership is collective, then nothing can be inherited or endowed – but these are larger questions, well outside the scope of this paper, which will have achieved enough if it provokes some thought, debate, strategy and action on these and related matters.

1. Just as gender, religion, ethnicity, etc, might also be perceived, in other contexts, to be similar obstacles.

2. In certain understandings of caste, an individual within these caste communities can change his/her caste status. A Brahmin can lose his/her status through a variety of pollutions, while a Dalit can raise his/her status through a variety of provisions for purification. However, these provisions almost always apply only to individuals and are not available to communities as a whole. They are dependent on the individual’s actions and have no reference to the community’s understanding of them. And finally, they are not available to very many communities – for instance, in the case of lower castes in Muslim or Christian communities.

3. For a general account of this, see

4. This argument was first made by F C Bailey in his book, Caste and the Economic Frontier (Manchester, 1957), and has since been disputed for specific cases of Brahmin or Dalit castes, but remains convincing to me as a general argument.

5. The case of migrant labour – for instance, Bihari labour in Punjab – is apparently counter-intuitive to this argument, but in fact, reinforces it, because, even if this indicates the possibility of the migrancy of labour at the lower end of the caste pyramid, the nature of the labour does not change, and consequently, neither does the caste – migration is horizontal mobility, not vertical.

6. This reinforces my contention that lower castes’ effectivity (by which one means a combination of extent of presence and of agency, or power of intervention in their environments) has largely remained bound by the region of the specific caste presence, because of the emphasis on caste identity as the basis of mobilization.

7. Speculatively, these attributes may have some connection to the profession/vocation of the community in question, in the case of caste, or to the theology/faith in question, in the case of religion – and so on, for ethnicity, region, language, race, etc. Interestingly, gender-difference does not create quite the same sense of community, although alternative sexuality (or sexualities) does – but that is a discussion for another paper.

8. I am using these terms not interchangeably or as synonyms (being well aware of the distinction between them), but to suggest that all three terms could apply to the conception of the community ‘body’ that I am referring to.

9. ‘Notional’ here meaning, not superficial or insubstantial but as ‘notion’ – an imagined phenomenon. These general remarks are necessarily preliminary theoretical observations that will require a more detailed elaboration; for now, they must be read as an outline or sketch of that future elaboration, presented here primarily to frame the specific discussion on caste.

10. It is in this sense that caste remains a powerful social-psychological force, even amongst individuals who, on the face of it, may have disowned and disavowed many or even all of the trappings of their caste attributes: there are psychological dimensions to property relations that transfer onto the sense of owning and being owned by the community, perhaps of becoming and being property oneself. This is of course specifically applicable to caste, but is also true of race, ethnicity, even to gender and language or linguistic identity, to some extent; however, it does not hold for religious and national identities – a very telling divergence, but one which must – along with all the other speculations here – be explored elsewhere.

11. These remarks on identity politics are made here with specific reference to caste politics, but may be applicable to other identity categories such as gender, race, ethnicity, sexuality, etc.

12. Again, this does not mean that resistant class solidarities cannot emerge here either, only that the need for them is unlikely to be felt uniformly, or with the same intensity as at the lower levels of the hierarchy, given the voluntary dimension of the economic activity.

13. Because, as I have tried to show, the integration is at the very fundamental level of the principle of ownership of labour. This deep integration of the caste system with capitalism in our context, has also led to controversies in several quarters regarding the character of the political economy, i.e., as to whether it is feudal, or semi-feudal or capitalist – because, obviously, the caste system has integrated with capitalism to varying degrees and in diverse ways in different regions, leading to confusions regarding the exact character of this political economy.

14. And possibly also invite (very legitimate) suspicions about possible (upper caste) discomfort in dealing with issues of caste in general, as has happened so often in many left debates in the past.

No Comments »

Leave a comment