¹ The influence is also that of Derrida, as from Writing and difference or Positions (tr. Bass, A.; London, 1978 and 1981)Google Scholar, but specifically is from Foucault's opposition between power and the ‘great fantasy of a social body constituted of the universality of wills’; Power/knowledge: selected interviews and other writings (ed. Gordon, C.; Brighton, 1980)Google Scholar, 55. A somewhat fuller discussion of these points will be published separately, and was also offered when a version of parts of this essay, centred on the trial of Maulvi Ahmadullah in 1865, was presented to the eleventh European Conference of Modern South Asian Studies in Amsterdam in 1990; I am grateful to Peter van der Veen for that opportunity, to Jürgen Lütt as discussant, and to other panel members for their comments.

² Bourdieu, Pierre, Equisse d'une théorie de la pratique… (Paris, 1972), 180CrossRefGoogle Scholar.

³ A nationalist conscience: M. A. Ansari, the Congress and the Raj (New Delhi, Manohar, 1987; pp. xvii, 277, Rs. 200)Google Scholar. Detailed references to this book will not be given; note that an appendix at pp. 137–44 reprints the National Pact of 1924 (mentioned below), with one page out of order.

⁴ On Nationalism and communal politics in India 1916–1928 (Delhi, 1979)Google Scholar and Mohamed Alt: ideology and politics (Delhi, 1982)Google Scholar; see also ‘In search of integration and identity: Indian Muslims since independence’, Economic and Political Weekly (hereafter EPW), Special Number, November 1988.

⁵ See Government of India, Home Department Political B Proceedings, nos. 44–5, March 1913, National Archives of India, New Delhi. The point is discussed briefly in Robb, P. G., The emergence of British policy towards Indian politics (New Delhi, forthcoming), ch. iiiGoogle Scholar.

⁶ Report on Indian Constitutional Reforms, Cd. 9109 (1918), especially pp. 185–8. The Report's paragraph headings on communal electorates are instructive, as follows: ‘They are opposed to the teaching of history’, ‘They perpetuate class divisions’ (that js, the ‘partisan’ and not the ‘citizen’), and ‘They stereotype existing relations’.

⁷ See for example Francis Robinson, ‘Professional politicians in Muslim politics 1911–1923’, in Pandey, B N. (ed.), Leadership in South Asia (New Delhi, 1977)Google Scholar.

⁸ On such mid-riot changes see Siddiqi, Majid H., ‘History and society in a popular rebellion: Mewat 1920–1933’, Comparative Studies in Society and History, 28, 3, 1986CrossRefGoogle Scholar, and Das, Suranjan, ‘Towards an understanding of communal violence in twentieth century Bengal’, EPW, 27 August 1989Google Scholar.

⁹ See Pandey, Gyanendra, The ascendancy of the Congress in Uttar Pradesh, 1926–1934; a study in imperfect mobilisation (Delhi, 1978)Google Scholar, and Sarkar, Sumit, ‘The conditions and nature of subaltern militancy’, in Guha, Ranajit (ed.), Subaltern studies, III (Delhi, 1984)Google Scholar.

¹⁰ Bipan Chandra similarly, in a valuable but extremely instrumentalist essay, leaves some tantalizing contradictions in his main argument: that communalism was not a popular movement except in riots, and that it had ‘hardly any religion in it’ though ‘communal consciousness’ existed and religion was ‘a large part of the life of a pre-capitalist people’; Communalism in Modern India (Delhi, 1981)Google Scholar. Such an approach was challenged by Gyan Pandey in a review article, ‘Liberalism and the study of Indian history’, EPW, 15 October 1983, and an alternative was already on offer in Hardy, Peter, The Muslims of British India (Cambridge, 1972)CrossRefGoogle Scholar.

¹¹ Anderson, Benedict, Imagined communities: reflections on the origin and spread of nationalism (London, 1983)Google Scholar.

¹² See not so much Said, Edward W., Orientalism (London, 1978)Google Scholar, as Inden, Ronald, ‘Orientalist constructions of India’, Modem Asian Studies, 20, 3, 1986CrossRefGoogle Scholar. See also Pandey, Gyanendra, ‘The colonial construction of “communalism”’, in Guha, Subaltern studies, vi (Delhi, 1989)Google Scholar.

¹³ Bayly, C. A., ‘The pre-history of “communalism”? Religious conflict in India 1700–1860’, Modern Asian Studies, 19, 1, 1985CrossRefGoogle Scholar.

¹⁴ There is a large literature on these subjects; particularly relevant to the present paper are Lelyveld, David, Aligarh's first generation: Muslim solidarity in British India (Princeton, 1978)Google Scholar, Metcalf, Barbara Daly, Islamic revival in British India: Deoband 1860–1900 (Princeton, 1982)CrossRefGoogle Scholar, Robinson, F. C. R., Separatism among Indian Muslims: the politics of the United Provinces’ Muslims, 1860–1923 (Cambridge, 1974)Google Scholar, Ahmed, Rafiuddin, Bengal Muslims, 1871–1906: a quest of identity (Delhi, 1981)Google Scholar, Minault, Gail, The Khilafat movement: religious symbolism and political mobilization in India (Delhi, 1982)Google Scholar, and Hardy, , Muslims. A useful bibliography will be found in Hasan, Mushirul (ed.), Communal and pan-Islamic trends in colonial India (New Delhi, 2nd ed. 1985)Google Scholar.

¹⁵ In addition to the references cited in the preceding note, see Shah, Mohammed, ‘The emergence of a Muslim “middle class” in Bengal: attitudes and rhetoric of communalism’, University of London Ph.D. 1990Google Scholar.

¹⁶ See Seal, Anil, ‘Imperialism and nationalism in India’ and other essays in Modern Asian Studies, 1, 3, 1973 (Locality, Province and Nation).Google Scholar

¹⁷ Jinnah's memorandum of March 1929 referred to ‘adequate safeguards for the protection and promotion of Muslim education, languages, religion, personal law, and Muslim charitable institutions’. See Philips, C. H (ed.), The evolution of India and Pakistan 1858 to 1947: select documents (London, 1962), 228–37Google Scholar.

¹⁸ Iqbal proposed a self-governing Muslim state in the north-west of India, arguing that ‘The life of Islam as a cultural force in this living country very largely depends on its centralization in a specified territory’; the relevant passages are reproduced in The evolution of India and Pakistan, 239.

¹⁹ ‘The Lingua Franca of India’, in Comrade, 22 July 1912, argued that Urdu was not ‘essentially’ Muslim, but the ‘vernacular’ of Muslims, containing in its theological borrowings the ‘consolation’ of religion.

²⁰ For a discussion of some of this literature see Beckett's, J. V. introduction and other essays in Clyve, Jones (ed.), Britain in the first age of party 1860–1750: essays presented to Geoffrey Holmes (London, 1987)Google Scholar. See also, on ‘social cohesion’, Christie, Ian R., Stress and stability in late eighteenth century Britain: reflections on the British avoidance of revolution (Oxford, 1984)Google Scholar.

²¹ See, for example, Nehru, Jawaharlal, An autobiography (London, 1936; New Delhi 1962), chs. xix and Ivi: ‘The real struggle to-day in India is not between Hindu culture and Muslim culture, but between these two and the conquering scientific culture of modern civilization’ (pp. 136–7)—and Gopal, S. (ed.), Selected works of Jawaharlal Nehru, vols. 3 and 6, 8 and 10 (New Delhi, 1972–1977)Google Scholar: note, in Tribune, 27 November 1933, that to ‘lay stress on communal problems rather than national ones…is obviously anti-national’; in the Bombay Chronicle, 11 December 1933, that ‘Today in India there is absolutely no cultural or racial difference between the Muslim and Hindu masses’; and in the Hindustan Times, 12 January 1937, that ‘The realities of today are poverty and hunger and unemployment and the conflict between British imperialism and Indian nationalism. How are these to be considered communally?’ and so on. It will be seen that in seeking a ‘modern’ antiimperialism he equated nationalism, social progress and the Congress, while associating reactionaries, elitist self-interest and the Muslim League. More important, perhaps, he interpreted culture and identity from a material point of view, and nationalism from a geographical one; he was pleased to admit that he was ‘totally unable to think along…communal lines’. Compare Chatterjee, Partha, Nationalist thought and the colonial world: a derivative discourse? (Delhi, 1986), ch. vGoogle Scholar.

²² The nature of Congress support is one of the most active questions of current debates, with contributions ranging from those of Chandra, Bipan and his colleagues, as in India's struggle for independence (New Delhi, 1988)Google Scholar, to those in Ranajit, Guha (ed.), Subaltern studies, I–VI (Delhi, 1982–1989)Google Scholar; see Mukherjee, Mridula, ‘Peasant resistance and peasant consciousness in colonial India: “subalterns” and beyond’, EPW, 8 and 15 October 1988Google Scholar. On business support, contrast Mukherjee, Aditya, ‘The Indian capitalist class’, in Sabyasachi, Bhattacharya and Romila, Thapar (ed). Situating Indian history (Delhi, 1986)Google Scholar, with Markovits, Claude, Indian business and nationalist politics 1931–39: the indigenous capitalist class and the rise of the Congress Party (Cambridge, 1985)CrossRefGoogle Scholar. The tendency seems to be to think either that the alliance behind Congress was wholly predictable (being among nationalists) and beneficial, or that it was self-interested on the part of élites (quite ready to back the British if necessary) and thus responsible for retarding an Indian social and political revolution from below. The truth may lie in between. Issues of continuity are raised (with some exaggeration) by Brown, Judith M., Modern India: the origins of an Asian democracy (Oxford, 1985)Google Scholar, and Potter, David C., India's political administrators, 1919–1983 (Oxford, 1986)Google Scholar.

²³ A modification of Herschmann's concept of choices between ‘exit’ and ‘voice’ may be helpful here: Muslims who ‘voiced’ the collectivity of religion were more likely (but not bound) to ‘exit’ from other possible allegiances.

²⁴ Gandhi too envisaged improvements—not from ‘modernization’, which he rejected—but from advances in society and human nature, and this led him to mount no more than a partial attack on, and indeed a qualified defence of, such institutions and beliefs as caste, the duty of women to men, and so on. In 1921 he claimed that untouchability made ‘swaraj impossible of attainment’, but also that ‘the spirit of kindness…is slowly but steadily gaining ground in the hearts of the masses’. Effectively, he called for tolerance (or personal ‘truth’) in place of social reform. See The collected works of Mahatma Gandhi, passim; the quotation is from vol. xix (Ahmedabad, 1966), 571–2Google Scholar. Thus in the 1920s and 1930s his emphasis on spinning and the superior or softer moral qualities of women (for example, the view that ‘women are more suitable than men’ for work in khadi or temperance) may be seen to be directed at modifying the conduct of men and rather at perpetuating than at improving conditions for women; while his emphasis on the Harijans similarly seemed addressed at caste Hindus rather than at the practical suffering of untouchables (which any caste attitudes clearly increased), for all Gandhi's comments on equality of respect, bread labour and so on. A convenient presentation of these points—and evidence of their advocacy by Gandhians—is in Prabhu, R. K. and Rao, U. R. (ed.), The mind of Mahatma Gandhi (Ahmedabad, 1945; 3rd ed. 1967), 107–11, 198–205 and 398–402Google Scholar; the above quotation on women is from a speech at Dandi, 13 April 1930, in Phatak, N. R., Source material for a history of the freedom movement in India, vol. III, part 3: Mahatma Gandhi, 1929–1931 (Bombay, 1969), 35Google Scholar. Similarly Gandhi's many speeches and writings about Muslims often seem directed at Hindus rather than at the felt worries among Muslims; he, supported their outrage over the Khilafat issue rather than expressing them as his own. Note his Congress Presidential Address at Belgaum in 1924, arguing that communal disturbances were fomented by interested persons, that communities could live together only as friends and not under the coercion of the state, and that swaraj depended on unity and social justice; ibid. vol. II, part 2; 1922–1929, 485–502.

²⁵ This point is discussed in Robb, , Emergence, ch. iiiGoogle Scholar. The one ‘class’ the British might seem to have accepted was that of ‘untouchables’; but they were defined not in economic but in ‘caste’ or ‘non-caste’ terms, as a community. For the foregoing see also Robb, Peter, ‘Ideas in agrarian history; some observations on the British and nineteenth-century Bihar’, Journal of the Royal Asiatic Society, 1, 1990Google Scholar.

²⁶ One need look no further than the various movements described in the early chapters of Hardy, Muslims, but see also Freitag, Sandria B., Collective action and community: public arenas and the emergence of communalism in North India (Berkeley, 1989), ch. iiGoogle Scholar. For insights into the complexity of communal identity, however, see Kumar, Nita, The artisans of Banaras: popular culture and identity 1880–1986 (Princeton, 1988)Google Scholar.

²⁷ See above, n. 12; and also Robb, , ‘Ideas in agrarian history’Google Scholar.

²⁸ See Daly Metcalf, Barbara (ed.), Moral conduct and authority: the place of adab in South Asian Islam (Berkeley, 1984)Google Scholar. See also Chaudhuri, K. N., Asia before Europe: economy and civilisation of the Indian Ocean region from the rise of Islam to 1750 (Cambridge, 1990), chs. ii and iiiGoogle Scholar; ideas from chs. i and viii–x on long-term change are also taken up briefly below; my thanks to the author for letting me read this before publication. On race see also Lévi-Strauss, Claude. The view from afar (tr. Neugroschel, J. and Hoss, P.; Oxford, 1985)Google Scholar, preface and ch. i, and From honey to ashes (tr. J., and Weightman, D.; New York, 1973), ch. xviiiGoogle Scholar.

²⁹ For example, Attlee's speech of 4 June 1935, Parliamentary Debates, H.C., vol. 32, col. 1824–1928Google Scholar, on the necessary role of the Congress in the creation of ‘modern India’, and on political division along economic lines, its being ‘in the modern world the natural development’.

³⁰ Sir Lyall, Alfred, The rise and expansion of the British dominion in India (London, 1894; fifth edition 1910; rpr. 1919), 2–3 and 41–7Google Scholar. The European dimension of views of India is currently being reasserted by Majeed, Javed, as in his ‘Orientalism, Utilitarianism and British India’, read at SOAS on 7 June 1990Google Scholar.

³¹ Chirol, Sir Valentine, India (London, 1925; 4th impression 1930), 1–55Google Scholar.

³² Thomson, S. J., The silent India: being tales and sketches of the masses (Edinburgh and London, 1913) v–viii and 1–34Google Scholar.

³³ Craddock, Sir Reginald, The dilemma in India (London, 1929), 1–17Google Scholar and. passim. These ideas were predictable for many reasons: the experience of Italian and German unification or Greek, Irish and Slav nationalism; the claims of social equity, by the later nineteenth century often thought achievable through collective restrictions on self-interest and not through Utilitarian individualism or the beneficent selfishness of the market, that is, through ‘community’, which raised anew the role of state interference and the relationship between state and people; and hence of course in a range of nineteenth-century thought, repeatedly discussed, not least in India. Bankimchandra Chattopadhyay noted, for example, that J. S. Mill disagreed with Comte on community and nationalism (see Raychaudhuri, Tapan, Europe reconsidered: perceptions of the west in nineteenthcentury Bengal (Delhi, 1988), especially pp. 84–5, 132–55 and 168–71)Google Scholar; it would be rewarding to know still more of the dialogue between Western and Indian ideas on these and other matters, which produced the kinds of view represented by Craddock. Moreover, the debate continues. For a recent argument in favour of having a state divorced from ethnicity, but arising out of civil society, not ordering it, see Sheth, D. L., ‘State, nation and ethnicity’, EPW, 25 March 1989Google Scholar.

³⁴ For this reason Ansari's case parallels the point made by Hardy that the Muslim League and the Jamiat al-Ulama-i-Hind, though divided over Pakistan, were equally pro-Muslim (Muslims, 239–55).

³⁵ Raychaudhuri, , Europe reconsidered, 1–13 and passimGoogle Scholar.

³⁶ On such links and parallels, see Bayly, Susan, Saints, goddesses and kings: Muslims and Christians in South Indian Society, 1700–1900 (Cambridge, 1989), 41, 85–6, 94–6, 123Google Scholar, and more broadly ch. iii (on a South Indian Islam in a Hindu context). This book raises a number of important general points—for example, about different chronologies of conversion, and the appeal of ‘sacred energy’ rather than egalitarian teaching, both of which link religion with politics. Note, at p. 463, that ‘“high” religion never acquired a dominance over the autonomy of local belief and workship in the south’ even in the later nineteenth and twentieth centuries, despite the influence of the colonial state. Moreover orthodoxy was less insisted on than orthopraxis—any increase in the former seems to be attributed to leadership—and (p. 459) ‘At no time in the immediate pre-colonial period was there a clear and unambiguous process at work by which boundaries between different south Indian groups or “communities” were being irrevocably hardened…. All the same, significant differences in identity did exist’ (emphasis added). On palaeontology see Gould, Stephen Jay, Wonderful life: the Burgess Shale and the nature of history (London, 1990)Google Scholar in comparison with, say, Bonner, John Tyler, The evolution of complexity by means of natural selection (Princeton, 1988)Google Scholar. This reference does not mean to suggest that Gould's final conclusion (if I understand it), that the forerunner of modern vertebrates survived by chance, can be applied to the evolution of Muslim separatism.

³⁷ Robinson, , Separatism, 33Google Scholar. Compare above n. 10.

³⁸ I owe to Maureen Perrie and Madhavan Palat this example of a word capable of meaning ‘community’, ‘world’ and ‘peace’.

³⁹ It is this emphasis on relative pervasiveness which makes the argument of this paper complementary to rather than contradictory of the very valuable studies appearing on peasant and low-caste consciousness, as by Amin, Shahid, ‘Gandhi as Mahatma’ and Gautam Bhadra, ‘The mentality of subalternity’, in Guha, Subaltern studies, III and VIGoogle Scholar, or Prakash, Gyan, ‘Reproducing inequality: spirit cults and labor relations in colonial eastern India’, Modern Asian Studies, 20, 2, (1986)CrossRefGoogle Scholar.

⁴⁰ This essay does not take up the appropriateness of secularity for South Asia, which nowadays is questioned—see Mitra, Subrata, ‘The limits of accommodation: Nehru, religion and the state in India’, South Asia Research, 9, 2, 1989CrossRefGoogle Scholar—and if it had it would have concluded that secularism may be like democracy, the worst possibility except for all the others.

⁴¹ Alexander, Dru (ed. and tr.) The letters of Jacob Burckhardt (London, 1955)Google Scholar, 191 and passim; and also Burckhardt, , Judgements on history and historians (tr. Zohn, Harry with an introduction by Trevor-Roper, H. R.; London, 1959), for example pp. 61–6Google Scholar; and see Antoni, Carlo, From history to sociology (tr. White, Hayden V.; London, 1962), ch. vGoogle Scholar.

⁴² Despite the acceptance of government and institutional influence in politics (above, n. 16), this aligns with Guha, Ranajit (‘Dominance without hegemony and its historiography’, Subaltern studies, VI)Google Scholar in some of his criticism of Anil Seal and David Washbrook. But the two positions seem partly compatible. There is common culture, whatever the vitality of common elements within groups, only in a limited sense: ideas, as collective memory, form part of an environment in which all identity and behaviour are contingent, and even protest against such norms is often forced into the frame of what it opposes, thus reflecting or even perpetuating them: there is no anti-caste movement without caste, no counter-culture without dominant culture. But tracing an Indian (Hindu?) idiom—authority, subservience and protest and danda, dharma, bhakti—should not imply that forms or their content remained constant over time; they are continually reinterpreted or reinvented, and no single version (dominant, subordinate, Indian, British) should be regarded as particularly ‘real’. Identities too have more than one origin and path of development, and are influenced by institutions as well as ideas. Guha admits (p. 255) that the recognition of belonging to ‘one's own country’ (a specialized identity) has to be achieved; it follows that anti-colonialism too can be understood in comparison with colonialism, whose forms and structures it may commandeer. Thus, as Guha suggests, Washbrook offers a false choice (opposed in the very concept of this paper, as in my original review of The emergence of provincial politics; Cambridge, 1976)Google Scholar when he claims that, in political history, ‘contemporary processes’ are more interesting than ‘ideational antecedents’ (even supposing they can be distinguished, or the former can exclude ideas). But thus Guha also makes a false dichotomy: between the Indian and the British history of India. There are more than two positions or schools. One may argue for characteristic or persistent ‘Indian’ views, but here Guha's position seems excessively essentialist and ahistorical. British and Indian history are at once distinct and continuous, and the elements of thought each culture has borrowed from the other are too complex merely to be cited as evidence of oppression. Guha rejects the ‘appropriation’ of Indian by British history (ever since James Mill), but, at least since Mill, demonstrably a part of each is a ‘portion’ of the other. This does not deny the value of Indian history for and by Indians—or Bengali(s), not at all the same!—see Guha, Ranajit, An Indian historiography of India: a nineteenth-century agenda and its implications (Calcutta, 1988)Google Scholar.