Accommodation and conflict


back to issue

WHAT are the most important changes to occur in India since Independence? For this writer, two things stand out: the emergence of a democracy with deep roots in society; and the decline in the power of caste hierarchies across most of rural India.

The latter change is not as widely recognized as it should be, but abundant evidence from diverse regions plainly indicates that it has been occurring – unevenly, but widely enough to be a national trend.1 As Dipankar Gupta has stressed, ‘caste’ tends increasingly to denote ‘difference’ more than ‘hierarchy’.2 The very foundations of Indian society (and politics) are shifting.

But what does this imply for social and political dynamics? What sorts of accommodations and conflict do we see, and which of these themes predominates? These preliminary notes from my field research in Andhra Pradesh and Karnataka open a discussion of these issues.3 


It is important to stress that ‘caste’ can mean three different things: ‘varna’, ‘jati’ and ‘jati-cluster’. Varnas – the four large traditional divisions of Hindu society, which exclude Dalits – have less importance in South India than elsewhere because there are no indigenous Kshatriyas and Vaisyas in the South,4 and Brahmins there constitute less than 4% of the population, so the Sudra category contains upwards of 80%. Jatis are much smaller endogamous caste groups – within which people marry their children – and jati remains strong, even though hierarchies among jati’s are waning. Jati clusters have emerged in recent decades, as similar jatis align with one another to enhance their numerical strength (and thus their influence) within the political system. Karnataka’s Vokkaligas, for example, are a cluster of six jatis.

What does the declining power of caste hierarchies over the thinking and actions of ordinary rural folk mean for social and political interactions at the village level, and for politics at higher levels? Conventionally, we might expect more violence as formerly dominant groups seek to maintain the old hierarchies by force. This field research has found that savage, murderous incidents have indeed occurred in some localities, and less serious but still outrageous clashes and abuses often happen. But surprisingly, in these two states (Andhra and Karnataka), those trends are outweighed by the emergence of uneasy but resilient accommodations between castes – not least between formerly dominant castes and Dalits – even in villages located quite close to localities that have experienced spasms of lethal violence.


Let us look at what has – and has not – changed. We must exercise restraint in describing changes that have occurred in caste systems,5 and in interactions between castes. Ambiguities abound here, as with all other things Indian. A term like ‘transformation’ is still too strong to be an accurate generalization – although it is appropriate across much of rural India. Three words of caution are in order here. First, the overwhelming majority of rural dwellers still marry their children to others from their own jati, and jati still has an immense impact on social life and interactions. It is still crucial as a material reality, enhancing or undermining people’s opportunities and capacities. Hierarchies have waned, but the institution of jati remains potent.6 

Second, at election time, people still pay attention to caste – to their jati, and/or their jati cluster – as they decide how to vote. But caste does not count for as much as the media suggests. A lot of other considerations loom large – indeed, often larger.7 So, while the declining power of hierarchies does not imply a declining importance of caste in voting behaviour, and thus in higher level politics, we should not overstate that importance.

Third, hierarchies have declined unevenly in different areas. We quickly discover this when we move between sub-regions of Indian states, or even between districts within sub-regions. For example, hierarchies have lost much more of their former power in the southern Karnataka sub-region than further north in the sub-region formerly ruled by the Nizam of Hyderabad. Land was and is more equitably distributed in the former than anywhere else in India and as a result, caste inequalities have long been less marked than elsewhere. But in the latter sub-region, inequalities in landholding – and in society, and in power relations between castes – were and remain much greater.

But even within these sub-regions, we find significant variations. In southern Karnataka, when we move from districts near Mysore city where hierarchies have declined and new accommodations have emerged (see below) to Mandya district (located between Bangalore and Mysore), we find far less change. That is because Vokkaligas in Mandya constitute roughly half of the district population – vastly in excess of their numbers elsewhere. That enables them to sustain many old hierarchical practices, even though Dalits and OBCs there no longer accept them.


To get a sense of changes in interactions between castes in these states, let us consider (all too briefly) four episodes which exemplify the patterns that predominate in village life today.8 Across much, probably most, of rural Andhra Pradesh and Karnataka, the landowning castes which once dominated village society no longer do so. They cannot dictate to others and get their way. Perhaps the most vivid depiction of this change is in a study of one southern Karnataka village where a small ‘festival’ – in which unpaid labour was performed by Dalits and OBCs – no longer happens. In former times, the dominant Vokkaligas would name a day when grain from their land was ready to be threshed and bagged for market, and others in the village cooperated out of deference. The word ‘Vokkaliga’ means ‘those who thresh’, but the actual work of threshing was left to other, lesser castes. Today, the old deference has evaporated, and those other castes ignore requests to cooperate free of charge.9 The Vokkaligas have now stopped demanding and started paying.


That example is about something which has stopped happening. But as caste hierarchies have lost their potency, other things which were unthinkable not so long ago have started occurring. In another Karnataka village, an informal council of elders, in which landowning Lingayats predominate, still exercises considerable influence over social life. (It should not be confused with the formal, elected panchayat.) On a given date each year, the Lingayats take out a procession in connection with a celebration particular to them. It customarily passes through the Dalit section of the village, and as a courtesy, Dalits have long tidied the street in their quarter along which it proceeds. When an especially insensitive Lingayat became leader of the informal council, he decided that the procession would bypass the Dalit quarter – but he did not bother to inform the Dalits. They were offended, and took a decision which could not have occurred until recently – to impose a total boycott of the Lingayats until that leader was replaced.


It swiftly revealed how interdependent the village had become, and how little the formerly dominant caste could now do to persuade Dalits to abandon the boycott. Negotiations ensued, and the Lingayats eventually agreed to remove the offending leader – to the astonishment of many. This incident revealed to everyone – not least to the Dalits – that new power dynamics now prevailed, and that negotiation and accommodation were required to manage local affairs.10 

Sometimes, however, severe violence is unleashed by the formerly dominant castes against Dalits. The next two cases focus less on those savage incidents in individual localities than on the social dynamics which emerged within nearby villages in the wake of these outrages. The first thing to note is that in both cases, and in every other such instance in these two states, violent acts in single villages never spread to neighbouring areas. They remain confined to individual localities. But do they poison caste interactions outside, but close to, those localities? The answer is ‘no’.

The first case comes from coastal Andhra Pradesh. In one village, a ghastly massacre of Dalits by members of the formerly dominant Kamma caste occurred after a perceived affront. Several years afterward, people in nearby areas acknowledged that social interactions in that village – where the perpetrators of violence had been convicted and jailed – remained extremely tense and bitter. But they stressed that in their localities, quite close by, somewhat uneasy but resilient understanding had been negotiated (and periodically renegotiated), so that severe tension and violence were avoided.

An illustration of such understanding emerged from Guntur district in that sub-region. Three Dalit youths had teased a Kamma boy, and when news of this reached Kamma elders, they met to discuss it. Some argued that they should give the three youths a thrashing. But the predominant view was that ‘we cannot do such things anymore.’ Instead, the incident would be discussed with Dalit elders. The talks led to an agreement that the three young Dalits would apologize to the boy whom they had teased. The apology was then made, and accepted, and peaceable normality was restored.11 This kind of dialogue and accommodation represents the prevailing pattern, even across sub-regions where lethal atrocities have occasionally taken place.


A very similar picture emerges from contiguous areas of Chamarajanagar and Mysore districts which are located near another grotesque outburst of violence. In one village, Lingayats engaged in a massacre of Dalits after Dalit youths in a rural cinema put their feet up on the backs of chairs near where Lingayat girls were sitting. (Note that the youths had enough income to afford to sit in the most expensive part of the cinema hall – where the options are ‘floor’, ‘bench’ and ‘chair’.) This sparked a violent reaction from Lingayat men which spiralled into a savage killing spree against Dalits – and (again, as is now the norm) led to convictions and imprisonment for the murderers. Social relations in that village remain toxic years later, but in localities quite close by, the reaction (then and since) has been similar to that seen in coastal Andhra.


Lingayats and Dalits, sitting together, recently spoke to this writer about that massacre in a remarkably cool, matter of fact, manner. Their descriptions of social dynamics in their own villages nearby, even became jocular at times. They laughed at and among themselves as they referred to each group’s frustrations with the new social dynamics. (That sort of humour is seldom seen elsewhere in these two states because both elites and Dalits are more uneasy.)

Their comments indicated that new, workable relationships had been negotiated. The old social hierarchies had waned, and the Lingayats were unhappy about that. New, mainly economic, hierarchies and modes of exploitation had evolved in which caste connections helped some Lingayats to flourish, albeit alongside lesser numbers of Dalits and OBCs. Dalits were not content with that, but they welcomed the economic gains that many Dalits and OBCs had made by commuting by rail into booming Mysore city for well-paid day labour on construction sites. Those gains gave them greater autonomy, and the confidence and skills to assert themselves back in their villages. A far more complex, ambiguous, but less inequitable set of dynamics had emerged.

Such reassuring scenes, and the social accommodations that make them possible, would not have occurred two decades ago. But they should not distract us from another depressing reality that has also followed from the waning of the old hierarchies. On those quite rare occasions when severe violence has occurred in these states, it has been more savage than in former times. In those days, it usually entailed non-lethal punishment for individuals who had deviated from what most members of society – including many from disadvantaged groups – saw as accepted social norms. Now, it entails murderous violence against groups by local elites infuriated by their inability to dominate, and by the widespread rejection of the old norms.

The predominant trend, however, is the self-restraint now exercised by formerly dominant castes in both states. It is the result of a change of mind, not a change of heart12 – of pragmatic calculations rather than empathy with Dalits. The old elites face changes both within and outside the village which they cannot stem (see below), so most of them reckon that the best strategy is to accept (grudgingly) their loss of influence, and to concede as little as possible. They resort to an uneasy political correctness which minimizes risks and damage from forces outside the village, and which frees them to pursue mainly economic opportunities within and beyond it.


Only a brief survey of these risks and opportunities – and thus of multiple causes of change – internal and external to villages is possible here, but they merit attention.13 Let us begin with externalities. The importance of education and government social programmes in promoting occupational differentiation and eroding hierarchies is often rightly noted, but the impact of the latter appears to be outweighed by other official interventions. The rather effective implementation in these two states of the 1989 Atrocity Act14 looms large. Many poorer Dalits are not fully aware of it,15 but formerly dominant castes certainly are, and it inspires restraint among them.16 Many worry less about convictions on atrocity charges than about prolonged court cases which can prevent them from engaging in other advantageous activities.


The application of this act, and other government efforts to protect or support Dalits, are reinforced by three things that have gained importance in recent years: (i) the competition between major political parties for badly needed Dalit votes in these states; (ii) the greatly increased penetration of rural areas by Dalit organizations,17 and (iii) the somewhat increased media coverage of such matters in rural areas. In former times, perpetrators of violence against Dalits often concluded (correctly) that they would not be found out, or that the state was on their side or would look the other way. Now abuses are likely to become known, and the state is more likely to intervene on the side of victims.


Another set of externalities consists of opportunities for what Surinder Jodhka has called ‘exits’ from villages, which ease tensions by separating potential or real adversaries. To his ‘exits’, I would add external ‘distractions’ – a second ‘pull’ factor which inspires less elite involvement in contested, frustrating village affairs. In these states, ‘exits’ are widely available to members of the formerly dominant castes who are exasperated by ‘push’ factors: the waning of deference; by the need to grapple with assertive Dalits and OBCs in contentious local politics, and to seek votes in panchayat elections from them; by diminishing returns from agriculture as inheritance subdivides lands – and in Andhra Pradesh, by Naxalites.

‘Exits’ imply migration – permanent or temporary – to urban centres or (in Andhra Pradesh far more than Karnataka) overseas. Emigration overseas from coastal Andhra is quite significant, and from Telangana, massive.18 Many Dalits and OBCs also exit villages, usually temporarily, either for extended periods to work in urban centres or – if they live near booming cities – as commuting labourers. ‘Distractions’ for members of formerly dominant castes consist mainly of opportunities for investment and leisure in large, booming cities – an especially pronounced trend in Andhra Pradesh among Rayalaseema Reddys.

Of course, internalities also include occupational differentiation which has made Dalits less dependent on elites for their livelihoods. Ironically, much of this has been triggered by decisions by the formerly dominant castes to abandon old paternalistic arrangements in pursuit of economic advantage. That erodes the old hierarchies and deference, and now frustrates them, but they have contributed to it.

That erosion is also the result of long experience of local democracy among Dalits and OBCs – in Karnataka’s strong panchayats, and in fierce local party and factional competition in Andhra Pradesh where successive state governments have kept panchayats weak. That experience has greatly enhanced disadvantaged groups’ ‘political capacity’ – their political awareness, confidence, skills and connections to allies. Political capacity helps Dalits to make better use of their often considerable numerical strength – they are the largest (or nearly the largest) voting blocs in many state assembly constituencies, and in many localities. In parts of Andhra Pradesh, their numerical weight has increased somewhat as a result of ‘exits’ by members of the formerly dominant groups. One last internal element is an important generational change: a greater acceptance among younger members of formerly dominant castes of greater roles for Dalits and OBCs, and of the need for more equitable social relations.


These various trends help to explain the paradoxical picture that emerges in rural Andhra Pradesh and Karnataka. Very occasional spasms of grotesque violence occur – and they are more severe than two decades ago – as the formerly dominant castes lash out amid the waning of hierarchies. But it is far more common to encounter the negotiation (and periodic renegotiation) of calculated, uneasy but resilient accommodations. They prevent violence from occurring on anything like the scale, or with anything like the severity, that might be expected when something as fundamental as caste hierarchy loses its power over the thinking and actions of rural dwellers.



1. See for example, G.K. Karanth, ‘Caste in Contemporary Rural India’, in M.N. Srinivas (ed.), Caste: Its Twentieth Century Avatar. Penguin, New Delhi, 1996, p. 106; A. Mayer, ‘Caste in an Indian Village: Change and Continuity’, in C.J. Fuller (ed.), Caste Today. Oxford University Press, Delhi, 1997, pp. 32-64; G.K. Karanth, Change and Continuity in Agrarian Relations. Concept, New Delhi, 1995; S.R. Charsley and G.K. Karanth (eds.), Challenging Untouchability: Dalit Initiative and Experience From Karnataka. Altamira Press, London, 1998; D. Gupta, Caste in Question: Identity or Hierarchy? Sage, New Delhi, London and Thousand Oaks, 2004; S.S. Jodhka, ‘Caste and Untouchability in Rural Punjab’, Economic and Political Weekly, 11 May 2002, pp. 1813-23; and S.S. Jodhka and P. Louis, ‘Caste Tensions in Punjab: Talhan and Beyond’, Economic and Political Weekly, 12 July 2003, pp. 2923-36.

2. Dipankar Gupta, Interrogating Caste: Understanding Hierarchy and Difference in Indian Society. Penguin, New Delhi, 2000.

3. I am grateful to the Harry Frank Guggenheim Foundation for a grant to support this research which also focuses on several other states.

4. I owe this formulation to discussions with M.N. Srinivas.

5. The use of the plural is crucial here since, roughly speaking, each linguistic region of India has its own distinctive caste system – which differs somewhat or markedly from those in other regions.

6. See the discussion of the materiality of caste in J. Manor, ‘Prologue’, in R. Kothari (ed.), Caste in Indian Politics (second edition). Orient Blackswan, New Delhi, 2010, pp. xxii-xxiv.

7. See the section on caste and voting behaviour, based on data from Lokniti and the National Election Study, ibid., pp. xi-lxi.

8. Enough similar cases have emerged to indicate that these examples are fairly typical.

9. G.K. Karanth, V. Ramaswamy and R. Hogger, ‘The Threshing Floor Disappears: Rural Livelihood Systems in Transition’, in R. Baumgartner and R. Hogger (eds.), In Search of Sustainable Livelihood Systems: Managing Resources and Change. Sage, New Delhi, Thousand Oaks and London, 2006, pp. 265-74.

10. I am grateful to Kripa Anantapur, who has studied this episode, for a detailed account of it.

11. I am grateful to K.C. Suri for an explanation of this incident.

12. I owe the insights here and in the previous paragraph to K.C. Suri. They proved to be valid in field investigations in most parts of both of these (and some other) states. See also, for further ideas, P. Price, ‘Changing Meanings of Authority in Contemporary Rural India’, Qualitative Sociology 29, 2006, pp. 301-16; and P. Price, ‘Honor and Morality in Contemporary Rural India’, in L. Joseph, M. Mahler and J. Auyero (eds.), New Perspectives in Political Ethnography. Springer, New York, 2007, pp. 88-109.

13. Discussions in late December 2011 with V. Ramaswamy in Bangalore and K.C. Suri in Hyderabad – and field work in rural areas in 2010 and 2011 – has informed the discussion of the themes set out in this section.

14. Its full title is ‘The Scheduled Castes and the Scheduled Tribes (Prevention of) Atrocities Act, 1989’.

15. Interviews with a senior police officer, Hyderabad, 30 December 2011; and with two civil servants, Bangalore, 5 January 2012.

16. I encountered palpable anxiety about the act not just in these states but in under-developed areas of rural Madhya Pradesh where hierarchies are strong, during Digvijay Singh’s time as chief minister (1993-2003).

17. This penetration has occurred despite internal conflicts which have affected Dalit organizations in these and many other states.

18. I am grateful to Anil Kumar Vaddiraju for stressing this. Migrants from rural Rayalaseema tend to go to Hyderabad rather than overseas.