4
All civilizations have language but societies do not put this universal implement to the same use. Historically, language can hardly be treated as a homogeneous entity. It can be divided in many ways—into its various strata, its distinct elements, the differing types of competences which are gathered up into the general notion of a language. The social functions accredited to different strata of language in Indian civilization (though the situation in South India is different in some important respects) appear to be interestingly different from the European case; in this essay I shall try to analyse how language contributes to the formation, and rupturing, of social identities, and I do this by focusing primarily on the Bengali speech community.1 This does not imply any claim to the exemplary or precedental quality of this particular case. Though the story of the relation between language and politics is bound to be different in the different language zones of South Asia's exceedingly diverse culture, these might reveal some similarity of processes, though not an identity of the exact line of events. In the last part of the essay, however, I shall speak more generally about the linguistic processes at work in the Indian nation state.
Language not only unites people, it also as effectively divides them. Another way of putting this would be to say that language is, socially, not merely a means of communication but also of deliberate incommunication. It causes not only feelings of identity but also of enmity: often the most indestructible barriers among people are ‘walls of words’. By this I do not mean merely that the process of formation of one linguistic identity generates a sharper sense of differentiation from others: for instance, the more the people of a particular region become ‘Bengalis’, the more their sense of separateness from surrounding languages like Oriya or Hindi. This is evidently true; but this is not the only sense in which language creates incommunication. Language divides ‘internally’ as well, and not to pay attention to this process often distorts historical accounts about linguistic identities.
People ‘having’ the same language do not have it in the same way. Socially, linguistic competence confers on people capacities, and their absence correspondingly takes them away. Being Bengali is an identity coming out of a person's having the Bengali language; but clearly, all Bengalis do not have this language in the same way or to the same extent. Thus, they enjoy the political ‘rights’ of Bengaliness to a patently unequal degree: for some rights stemming from Bengaliness must be indivisible, but others are unequal and stratified. The bhadralok of Calcutta speak the Bengali standard language, one which has resemblances on one side with the ‘high’ language in which Tagore wrote his poetry, but also, on the other side of the cultural spectrum, with the language spoken in the bazaar by the fisherman, the maid in the babu household, or by criminals on the margins of urban Calcutta. And these are not rightly separated orbits sufficient in themselves but a complex of words pulled in different directions by the internal logic of each social practice. The historical existence of the Bengali language is a complex fact in which all these sublanguages (or linguistic subpractices) must find adequate and properly judged representation. Language as it is socially used thus has to be broken down into various subparts—high and low language, literary and common language, the guru (high, of greater merit) and the chalita (conversational) language (a special distinction of twentieth-century Bengali), the literates’ and the illiterates’ language. Such differences are not merely aural or cultural, but also political. Being able to use a language in certain ways enables a person to do certain things socially; others who do not possess such linguistic competence simply cannot perform them. Often, these people are reduced to varying states of dependence on those who are more skilled, and their access to the whole of the social universe is mediated by this latter group who can consequently control this tenuous access.2
The use of writing by moneylenders, the scourge of the indebted peasantry, is one example. And the peasants cannot be blamed if they consider writing not as a means of enlightenment but of oppressive mystification. Peasant revolts, historians have argued, show a particular intensity of anger against written records because they relate to the linguistic practice of writing differently. The complexity of the story of language and identities cannot be tackled without a sufficiently complex conception of the gradations of competence in language and its political effect. I shall call this the ‘internal economy’ of language.
The manner in which this internal economy was structured in traditional India seems to have been peculiar, and interestingly different from the European case. Jacques Derrida's work has emphasized the primacy in European culture of the grammatological, of the written over the spoken part of language.3 This is asserted strongly, although in many European cultures the term for language comes from the Latin lingua or tongue, which might lead one to infer the primacy of the spoken inscribed in the etymology or language about language itself. But there appears to exist a more fundamental difference. The idea of logos in the Greek philosophical tradition clearly draws attention to language as a means of reasoning, of ratiocinative activity, giving this exceptional significance over other things that can be done by means of language. This is related, though this is not the place to try to show the connection in detail, to Gadamer's argument about the deep-seated privileging of the epistemological in Western cultural tradition.4 The first difference is that in the classical Indian tradition though a great deal of attention is given to nada5— the originary form of sound, and therefore, the source of both language and music—and though it is occasionally equated with God himself, it seems to lack a clear equivalent to the typically European concept of the logos— speech, writing, reasoning, episteme, science. By contrast, the Indian philosophic and aesthetic curiosity appears directed towards vak— the irruption of the utterance which universalizes language rather than writing. Writing—inscribing something on a receiving medium—is as much to create meaning as to stain virgin space. The Indian tradition's way of treating writing is, in interesting ways, ambiguous. Writing is fixing, giving an idea a kind of material immortality, something that can be done only to the rarest, at least to the most significant, of thoughts, ideas of extraordinary importance in some sense, an exalted fate that common speech does not deserve. It is however possible to argue just the reverse of this case: material things are destructible, to consign ideas to material existence is thus to subject them to the law of decay and destruction. Towards writing the Indian tradition shows a strangely complex combination of reverence and mistrust.
There is a second analytical difficulty. The way Indian culture structures the internal practices of language cannot be adequately captured by a standard distinction between oral and literate cultures. Indian culture arranges the institutional transaction between literate-oral in very complex ways; here the distinction between literate and oral is not homologous to the one between educated and illiterate. Even the educated have their own traditions and institutions of oral performance. It is not that this culture does not know writing; rather, in spite of knowing writing from a very early stage of its history it clearly uses writing quite sparingly. This peculiar configuration of knowing the gift of writing, yet abjuring its use in social transactions, seems to indicate that in traditional Indian culture linguistic practice is governed by a ‘theory’ of distribution of functions between speaking and writing. Some features of this functional distribution must be noticed before we come to a discussion of more modern history. First, the intimate connection so common in European culture between institutions, i.e. the extension in the scale of social practices in time and space, and consigning things to the fixity of writing, does not seem to obtain here.6 Enormous and essential structures of social exchange and communication are entrusted to oral continuity rather than written codification. Although writing and speaking can both create continuity, their manner of doing so are different. In Indian society, except for the vast numbers of the directly productive classes of people, the upper strata depended on literacy and related means of social control. The upper strata required use of literacy for different functional requirements specific to the social practices they engaged in. Trading practices on any large scale required literacy and commercial records which could not be handled without literacy and numeracy among commercial groups. Wielders of political power depended on some amount of minimal administrative documentation: occasionally, there was evidence of impressive elaborate accounting of revenue resources of their realms.7 Brahmins provided some of these services. But they were notoriously selfish repositories of the society's skills of literacy and learning of its practical applications, and they performed two types of essential functions. They lent their knowledge of literacy and its more specialized uses to the Kshatriyas or those who controlled political authority. A two-way traffic results from this caste monopoly of skills and cultural assets. Administrative ordering, deft use of the technical apparatus of legal principles, financial accounting, and bookkeeping require the use of these relatively rare and jealously monopolized attainments. Those who had formal political power depended on the Brahmins for these essential functional necessities as much as for more general moral legitimation. Caste barriers also constituted effective prevention against possible diffusion of these scarce resources and their increased availability to other groups in society. Secondly, Brahmins also performed the exceedingly important activity of rituals which ensured the imaginary continuance of the social order by keeping it supplied with its essential symbolic collective representations, and vital concepts for the construction of the collective self.
To be able to perform these materially and symbolically essential functions the Brahmins had to ensure that the relevant skills were reproduced in undiminished quality within their own caste. Relatively little of this function of institution-maintenance was given over to writing or writing-based forms of training. Although Brahmins had to be literate, a surprisingly large part of this training came to be not ratiocinative and written but passive and mnemonic.8 Memory is never an unfailing implement for perfect or reliable reproduction. Dependence on oral continuity gave to these practices a peculiar character which may be read wrongly by social scientists. They confer on these continuing institutions the solidity and relative immobility common to similar institutions in written cultures. Here, because the internal mechanism was oral, and because there were rarely standardized written institutional histories, institutional continuity could be compatible with great flexibility. The eternal religion (sanatana dharma) can therefore keep changing constantly. Though this may not have ensured continuity the way writing would have done, it did have a compensating advantage. Reliance on memory fulfilled the historical need for flexibility within a formally rigid structure in a subtle and partly covert fashion. In reasoning activities no mnemonic reproduction could be perfect, a reproduction without variation, without slippages caused by simple incompetence, inadvertence, interpretation, or deliberate intent. Indeed, quite often the key to understanding the unchanging structures is the astonishing amount of unannounced change which these sanction. Rigid formal rules of social conduct were meant to ensure a remarkable degree of conformist continuity. Since these rules were uncodified and worked through an oral tradition, they afforded those who directed society considerable space for informal amendments. Sometimes a reputation for rigidity accomplished things in subtle ways which an actual immobility would not have done.9 No wonder the record of Indian history is marked by a paradox—a rhetoric of immutability (as for instance in sanatana dharma, an eternal religious order invented in the nineteenth century) cut through by historical evidence of silent and surreptitious change. A good example is provided by the fascinatingly mobile history of the supposedly immutable order of castes.10
Traditional Indian society thus had a highly literate culture, but inside it literacy was guarded with great jealousy through institutional arrangements which strictly prevented its extension, so that it was always a strong sellers’ market in literacy. Ordinary unlettered people carried on their daily existence through spoken vernacular dialects. In the nature of things, these vernaculars varied a great deal, and in most cases did not have standardized written scripts of their own before the tenth century. Above these productive and oral classes presided a bilingual elite which commanded the use of an esoteric language which could not be mastered by others due to caste prohibitions. The elite could carry on their own internal discourse (which could be either theological or political) in a language which had the strange quality of a partial publicity, with the dual quality of being public and secret at the same time.11 It was public to insiders, but closed, esoteric, secret to outsiders who did not have the requisite skills. This had significant political results; because it made the scale of possible collective action or consultation asymmetric between the elites and the subaltern social groups. While elite discourse could range across the entire subcontinent, the discourse of subordinate groups necessarily remained trapped in the close boundaries of their vernacular dialects. Thus, while conservatism and reaction could be subcontinental in spread, dissent was condemned to be mostly local. Only those reform initiatives were likely to succeed which used the implements of the elite discourse itself against its ideological structures.
Fig. 1. Alterations in linguistic structures during medieval times
Social and political changes in medieval times brought some alterations in the structure of language and literacy, but these were not so fundamental as to introduce changes in the linguistic economy. Two types of change occur. Due to the political power of Muslim rulers—a reality brahminical Hindu society could not ignore—Arabic and Persian came to slowly occupy a part of the exalted position of the earlier esoteric language—Sanskrit. But it was more a situation of power than of authority. As Muslim power became entrenched in North India, especially those social groups and specialized castes which traditionally used literacy in the administrative service of the government, now extended their skills to the new language of power.12 It must be recognized, however, that this does not seem to have threatened the privilege of Sanskrit. Hindu society responded to the reality of Muslim rule partly through an attempt to marginalize the power of the state and eject it out of the circle of Hindu social practices, a move which succeeded in part due to the pragmatism of the rulers themselves. Although this could not wholly succeed, it did result in a sharper hiatus between the political power of the state, and social dominance inside these communities.13 Power and prestige inside these communities remained intimately related to real or supposed competence in Sanskrit; officiating Brahmins may have had a tenuous grip on the intricacies of Sanskrit grammar, but this hardly mattered since the effectiveness of these practices was symbolic, not grammatical. Indeed, the stable political power of Muslim rulers and the process of conversion to Islam of those who used this to escape the repressive rigidity of the caste system set in motion two contrary responses in Hindu society. On one side this gave rise to a more frantic traditionalism, but on the other also to types of exchange with Islamic culture. A tendency towards utilitarian secular exchanges was discernible in the practices of castes like the Kayasthas who were given to government employment traditionally, or the trading groups who tended to treat all men equally as long as this served the objective of commercial expansion. But there were also the more intellectually grounded proposals for religious exchanges advocated by bhakti doctrines—at the same time more ambitious and vulnerable.
Religious developments have an intimate relation to our story of languages. Throughout this period there occurred a slow development of vernacular languages through the gradual separation of their emerging literatures from the high Sanskrit tradition. Comparatively little research has been done on this process; but the lines of development and their basic interconnections are clear. And it is instructive to dwell on this process because the radicalism of the reform religions displays some peculiar features. The emergence of a vernacular literature does not indicate an immediate change in the narrative or imaginative content of literary culture, or a sharp, dramatic, conscious rupture with the classical tradition, however striking the disjunction may appear in retrospect. They arise haltingly, always making reverential genuflections in the direction of the high tradition and its texts, which they were eventually to undermine. These are not seen or begun as revolts against the logic of exclusion of the common people from aesthetic and religious seriousness built into the classical Hindu tradition. Rather, their first and most impressive texts are attempts to stretch the riches of this high culture towards the lower, culturally deprived orders. Their implicit justification would have been that, if religiosity and aesthetics were significant and valuable for all human beings, those without the use of Sanskrit should not be deprived of these values. As a consequence, these literatures assume a consciously subaltern relation between themselves and the high classical texts. The narratives around which these literatures grew were in most part old esteemed narratives from the classical tradition—the Ramayana, the Mahabharata, or similarly modelled stories of what in Bengal were called the mangalkavyas, which, though exalting unknown and upstart deities, do it in a style and form too easily recognizable as derived.14
One of the great texts on this translation of religiosity into a popular register was the Ramcharitmanas of Tulsidas.15 Many of the complex and ambivalent properties of this new vernacular literature could be found in this, its acknowledged masterpiece. In neighbouring Bengali regions, too, similar texts were composed: the Ramayana by Krttivas and the Mahabharata of Kashiramdas were written out of similar cultural impulses, though in artistic quality they do not compare with Tulsidas's text. But Tulsidas himself, like a number of other lesser poets, was no mean composer of Sanskrit lyrics, and vernacular writers often had to legitimate their claim to serious attention by writing in the more exalted language.16
Despite this, vernacular literature's poetic traditions began an undeclared revolution. Within the formal terms of continuity with classical traditions in terms of narratives, forms, and texts, these ‘translations’ in the vernaculars were hardly passive cultural creations; and they gradually produced an alternative literature which told the same stories with subtle alternative emphases to alternative audiences. It founded a cultural strand that Namwar Singh has elegantly called a dusri parampara, a second tradition.17 It has been pointed out quite often that the Ramcharitmanas was in several respects a deeply innovative text. Its ambition of taking the classical narrative to a common audience forced some alterations in the narrative and characterological structures. The humaneness of the idol is reinterpreted from a serenity that befits the almighty to an intimacy of a more accessible kind. As the epic is rewritten into the bhakti register, this gives a softness and sentimentality to the characterization which was lacking in the classical original. Modern interpreters of these texts sometimes make an anachronistic mistake by conceiving them in a style common to modern heterodoxies. Modern rebellions announce themselves even before they are wholly successful: revolutions in traditional cultures tended to hide the fact of their being revolts. They pretend to respect a continuity which they do not in fact practise. By declarations of continuity, however, they circumvent the censorship; precisely by their submission, they create a space for themselves in which the narrative axis, patterns of denouement, and delineation of characters all undergo a subtle transformation. Krishna, the incomparable warrior of the high tradition, becomes Krishna the incomparable lover. Ideas about union with god and his availability to his devotee go through the most radical change.18 All the while this is accompanied by formal declarations of continuity, apparently testified to indubitably by the sameness of the narratives, the figures, the tropes, and the symbols. Yet, at the same time, in a slow, undeniable open process these religions and their observance become public spectacles of a kind of defiance and heterodoxy.
These new texts could do this because they were practising an activity which was tolerated by the Hindu canonical tradition. The ascription of texts to mythical all-seeing sages was clearly a sanctioned device of collective writing, of having open-ended texts which went through constant proposals of embellishment and extension. This must happen of course in oral retelling of stories: but it does happen in written cultures like India in peculiar ways. Nameless poets tried to smuggle their favourite works into great texts to savour the ironical taste of an unnamed immortality.19 Oral incantation made this possible to an even greater extent, creating something like a latent market of interpolations of which only those favoured by a wide public eventually made it into the text itself. Accumulations of this kind sometimes changed the narrative structure, rasa structure, flavour, or even the meaning of the work. Hinduism had a tradition of such kumbhilakas, of this strange form of plagiarism in reverse, of people claiming not someone else's work as their own, but their own as someone else's. Bhakti literature in its celebrated translations used this general sanction of free retelling to interpret a new religion into existence.
Gods and goddesses of the classical tradition retain their audience at the cost of this alteration of their character. The stateliness of the central divine characters is replaced by their accessibility and humaneness.20 This principle of accessibility extends to the formal poetics of the compositions. The verses are composed in a simple and dignified language, in the most accessible of metres—in Bengali the payar for instance, the language and rhymes most suited for oral incantation by common people. Literary creations of the Vaishnava cult around Chaitanya displayed similar attributes and converted the figure of Krishna into a figure of love. Krishna's physical form, the idol itself, is transformed: the warrior form with four arms displaying the shankha, chakra, and gada somewhat softened by the residual padma, is replaced by the anthropomorphic symbol of the child, Gopal, or the adolescent Krishna of pleasurable and excusable transgressions. The predominant rasa is changed from aishwarya to the one of vatsalya or shringara—both in their different ways vehicles of nearness and indistinction. The language considered adequate for this religiosity is the vernacular. Like Buddhism much earlier, the bhakti movements favoured the lower language; but since these were touched by religiosity, this gave these tongues a new dignity. It will be historically misleading, however, to exaggerate this point; the extrication from Sanskrit remains incomplete in several significant ways. The Bengali Vaishnava padavali literature shows this ambivalence with remarkable clarity. Its poetic compositions hover between the clear folkishness of the language of Chandidas or Jnanadas on one side, and the comparatively sanskritized Bengali used by Govindadas. Both are equally indispensable parts of a single literary trend. Most interestingly, this linguistic tension is also reflected in the ways in which Chaitanya's life story is told by his biographers. Vrindavandas's Chaitanyabhagavat chronicles it in anthropomorphic terms, the extraordinariness of his acts marked by his miraculous humaneness. By contrast, Krishnadas Kaviraj's Chaitanyacharitamrita emphasizes the divinity of this existence, and sees in every act a metaphor of the order of the universe. It is hardly accidental that Vrindavandas writes in a limpid Bengali, while Krishnadas's work is only formally written in Bengali: in its more sombre moments it reminds us that we are in the presence of God himself by lapsing into appropriately complex Sanskrit, heightened by the erudite compositions in the more complex metre, mandakranta. Although Chaitanya is an avatar of Krishna he is saved from the heaviness of the classical type by the infusion into his image of Radha (because he is, according to Vaishnava theology, radhabhava krishnaswarupa). The figure of the feminine in the Indian tradition is often the symbol of the approachable, the popular, the proximate, the unthreatening and accessible. Because Chaitanya combines in himself the features of Krishna with the lovability of the feminine, he is an androgynous god, the deity of a popular religion which can capture the imagination of relatively subaltern peasant elements in the Hindu society of eastern India.
This indicates a further interesting fact about how people thought of themselves, and conferred self-descriptions on their relevant community. It is not only that the disengagement from Sanskrit is yet incomplete, and the continued popularity of the work of Jayadeva demonstrates its unassailed position as the norm for Vaishnava poetic compositions. More importantly for my central hypothesis, the poetry of Vidyapati from neighbouring Mithila is considered an inalienable part of the Vaishnava heritage. Given the later linguistic identity of this region as Bengali, both Jayadeva and Vidyapati would have raised problems of subsumption, as the first was an inhabitant of Orissa and the second of Mithila. If anything, this shows that despite the unmistakable beginnings of a distinctive vernacular literature, peoples’ identity must have been primarily determined by their belonging to a religious sect rather than to a common speech group.
This suggests the following hypothesis. Between the two great layers of language, the upper layer of Sanskrit, and the lower layer of dialects in medieval times, a slow process of differentiation had begun. But the vernaculars were still held together in linguistic intelligibility within a linguistic structure marked by the commonness of Sanskrit at the high end and the easy neighbourly intelligibility of the dialects below.
Gramsci said a language contains a certain conception of the world. To put it another way, there is usually a strong historical connection between a natural and a conceptual language. Seen this way, the origin of vernacular languages appears to be intimately linked to an internal conceptual rebellion within classical brahminical Hinduism. Along with the emergence of a new conception of religiosity, based on a more intimate relation between man and his deity, this movement required a new type of language. To fulfil its functions, it had to be a language of transparency and nearness rather than of inaccessibility and distance. If the purpose of the act of worship was to underline the inequality between God and man, His infinitude against the vulnerable finiteness of all human capacities, the inaccessible sonority of Sanskrit achieved that perfectly. Bhakti Hinduism, like strands of European Protestantism, sought to destroy the brokerage of the Brahmins between the devotee and his god. The philosophical argument behind this was, as is well known, that the act of worship, the most significant act of attunement of man to God's universe, must be transparent and meaningful, and against the earlier semiotics of forbidding distanciation bhakti religion brought in a whole new semiotic of nearness and informality. All this could be accomplished only in a language which, although still distanced by the question of literacy, was closer to actual common speech in two senses. It was, first, a literature in the vernacular. Secondly, it was a vernacular written, though with an ineffable elegance, like the common spoken word. The idea was, in its ideal, utopian, perfect phases, like a pedagogy of the oppressed: poetry spoke a language of the people, expecting that over the long term popular language of worship and sensibility would be exalted by this poetry.
Later, when vernacular literatures were fully formed in colonial times, some of them, like Bengali, would often seek to confer upon themselves a history with a suitably impressive antiquity. In constructing these images of their pasts, they constantly gerrymander historical frontiers. Histories of Bengali literature, a required segment of the formation of ‘cultured’ Bengalis, standardly claim Vidyapati (from Mithila) and Jayadeva (from Orissa) as parts of their own history. Unlike what Gellner says about the histories nations construct of themselves, there is nothing really fraudulent about this;21 for these were truly part of the genetic process of the formation of these cultures. But this also shows something equally and vitally true: that there was no Bengal in the modern sense. There was already perhaps an identity of the language, but no linguistic identity. Linguistic identity is not formed by the simple objective fact of some people having a common language; it lies in a more deliberate choice to see this fact as the essential criterion of their identity. The primary reason for this mild paradox—why the language existed without a linguistic identity of its people, I shall suggest—was due to the fact that it was a different kind of social world altogether. My argument is not that this social world ‘lacked’ some of the features that the modern world has; but that it was a world of a different kind, which we must try to understand by historical inference.
Elsewhere I have tried to approach this problem through a distinction between a fuzzy and an enumerated (in the sense of being counted) social world.22 Traditional society is made up of a structure of groups which are, in some crucial cases though not in all, fuzzily conceived,23 of fuzzily conceived space,24 and fuzzily sensed and imaged time.25 It is impossible to develop a full argument here about the diverse ways in which the social world, social space, and social time are fuzzy conceptions, and further how these conceptions are interlocked. Indian society, as anthropologists have argued, is of course segmented. But these segments of the lived world do not seem to have frontiers as in any organization of modern social space. Villages do of course begin and end, and therefore have their boundaries, though here, too, in the most immediate form of lived space, there is a certain ‘approximateness’ and ‘indeterminability’. Boundaries of villages, it has been pointed out for many regions of the precolonial world, were not drawn the way states would be represented in maps. It is important to stress that this traditional world is a world without maps, and is therefore devoid of a spatial mentality, inextricable from modernity, which thinks in terms of mapped spaces, and its corollary, measured distances. Inhabitants of the traditional world may thus be said to have a conception of farness rather than of calibrated distance.26 Beyond the village, the boundedness of regions, subregions, languages, kingdoms is undetermined or indeterminable with modern precision. ‘Boundaries’ do exist, things, spaces, groups, do begin and end. But they tend to shade off, merge, graduate. It is a different way of organizing difference from the modern one with which we are familiar.
Language illustrates this principle of organization of difference very clearly. In the area lying between, say, Benaras and Puri, a traditional observer would have heard dialects slowly and imperceptibly changing, such that, with historical clairvoyance these could be ascribed to what would eventually become three distinct vernaculars—Hindi, Bengali, and Oriya. But it would be impossible, within that world, to determine where Bengali ended and Oriya began; in any case, the change would be decisively different from the standard modern organization of linguistic difference as this is fixed on to maps. Differences would shade off the way distinctly different colours are arranged in a spectrum. It is a world, to put it dramatically, of transitions rather than of boundaries.
Some other ontological features of this social world are of serious consequence for any understanding of politics. First, not merely are the boundaries hazy or spectrum-like, people and objects inhabiting this world are radically uncounted. It is also an unenumerated world. In this world, again, Hindus and Muslims would have perceived their difference (though I doubt they would do it in the modern way); Bengalis and Oriyas do so as well, just as do Shaivas and Vaishnavas. But they do not live in a world in which knowledge of how many Hindus or Muslims or Vaishnavas or Bengalis there are is a part of the commonsensical social knowledge determining patterns of social action. This has a crucial political corollary. Self-identifications are in some senses fuzzy and uncounted: thus, Vaishnavas do not know how many of them are there in the world, and they are crucially incapable of considering what they can do to force the structure of their social world to their collective benefit, if all who are like themselves act together.27 Distinct languages can thus exist in this society without uniting or dividing people in the manner in which we familiarly see these processes in modern contexts.
British colonialism introduced decisive and irreversible changes in the structure of traditional society in this respect. The establishment of colonial power created a different structure of culture by a combination of deliberate policy and unintended consequences. Colonialism had to force social behaviour into recognizable institutional forms, which could be done only by introducing a new type of discourse. Colonialism did not allow to subject Indians any option in this matter; they were obliged, differentially according to their social status, to respond to its demands, identifications, and actions. Colonial laws and rules restructured Indian society in fundamental ways. Behind these laws there was a more fundamental conception of what laws were, how they related to the facts of the social world, a primarily bourgeois-rationalist discourse of legality. To understand and practically respond to the colonial-legal it was imperative for all Indians to understand this discourse; but it follows that the need was more intense according to their intimacy and involvement with colonial power. Sub-Brahmin groups seeking quick upward mobility through the pathways opened by colonial administration, newly-endowed landlords under the Permanent Settlement, and those armed with a university degree who could get professional placements in the new society had more to do with this system, and therefore showed understandable eagerness to enter schools, pass public examinations, read European history, and get a grip on these conceptual assumptions upon which the colonial institutional system was founded. For understandable reasons, they themselves, and to a lesser extent their colonial patrons, acclaimed this as their acquisition of the difficult, yet potentially universal, principles of rationalism. By contrast, other groups which, by choice or circumstances, had less to do with colonial institutions acquired less of this conceptual language, and eventually came to be derided by the babus for their lack of rational inclinations. Interestingly, it is not only the poorer and culturally deprived social groups which fell into this ignominy, but also those who, on account of cultural pride or other such non-maximizing grounds, remained within the institutions and practices of traditional culture. However, in general, colonialism imposed on society a radically new, unfamiliar discourse, a conceptual grid, an alphabet, without which these institutions were uninhabitable, unworkable, and unintelligible. Institutions are after all the external superstructure of practices, and by altering the logic of institutional functioning colonial rule also forcibly altered the structure of traditional practices and the social understandings that went with them.
Though early colonial administrations by and large followed a policy of non-interference in cultural affairs, there were strict limits to this tolerance. Colonial administrations could hardly dispense with one essential prerequisite of effective rule: intelligibility of this world to the rulers themselves. As a result of the unanswerable power that attached to the moves of the colonial administration in its days of dominance, even though sometimes these moves were misidentifications, they had real consequences. Gradually, colonialism introduced into this social world entirely unfamiliar processes and institutions drawn from the enormous cognitive apparatus that rationalism had by this time created in the West, by which alone the colonizers could make this world cognitively and practically tractable. Surely, colonialism was an enterprise of introducing modernity only in truncated forms.28 However, one particular aspect of modernity the colonial state did introduce with effectiveness—the modern imperative of setting up social connections on an unprecedented scale. The extension of scale of social action brought in, as sociological research has amply shown, pressures towards standardization in varying forms.29 European missionaries of the Srirampur (Serampore) College sought to fashion printed alphabets for Bengali, and by 1800 religious tracts meant to popularize Christianity were being printed at their press. These foreign pioneers of printing chose from among various styles of calligraphy, keeping in mind technical constraints, but once this contingent choice was made it came to have decisive standardizing effects on even Bengali writing. Ironically, the way generations of modern Bengalis would read and write their own tongue was fatefully decided by choices made by a peculiarly skilled Englishman.
The most significant standardization, however, occurred in the areas of spoken and literary language; and this begins the mysterious process by which a high literature which they would never be able to read nonetheless confers an identity on the illiterate. Calcutta, an unknown village before the advent of the British, came to acquire as the colonial capital unprecedented economic, administrative, and commercial eminence. The prime beneficiaries of the colonial social transformation either came from the society in Calcutta or journeyed there to remain close to the source of all colonial beneficence. The social elite resident in that city thus came to enjoy an eminence unknown in earlier times, and indeed, one which was structurally impossible. In a matter of less than a hundred years the Calcutta babu came to acquire an unequalled capacity to do social and cultural norm-setting, growing out of his opulence, and to screen demands made on the colonial administration. Slowly, the language of the Calcutta bhadralok, with occasional skilful mixtures from areas which had a reputation for particularly mellifluous accents, came to be regarded as the norm language for bhadralok Bengalis in all regions of this linguistic area. Of course, this norm-setting process could not have been completed without the print media. One of the reasons for lack of standardization in earlier language was the lack of scale-extension of any particular speech form. The arrival of printing altered this fact by exerting two types of pressures on its users. First, language had to be standardized not only at the material level of letters to be used in print; but also, this standard Bengali could then be dedicated to modern uses as the vehicle of high literature, of science, of serious intellectual instruction. Secondly, once this standard or norm language became established, elites or aspirants to elite status from other, more outlying, regions began to emulate its accents and written idioms. The gentry of Medinipur, presumably content in earlier times with speaking a dialect which kept them in close touch with their local lower-order brethren, and more significantly, in intelligible proximity with Oriyas, now began to read and speak like journals written in the norm language of Calcutta. Any claim to linguistic distinction in the traditional language-economy rested on the command of esoteric languages. Dialect Bengali would be spoken the same way by all social groups. Now the marks of distinction had to be entered into the vernacular itself. Traditionally, though the two languages, Sanskrit and Bengali, may have been placed in a relation of hierarchy, each language was internally more equal. Two changes now occurred in this structure. English quickly displaced Sanskrit from the status of esoteric language: Sanskrit was relegated to the position of an archaic tongue. More significantly, the internal economy of the Bengali language itself became distinctly more hierarchical. As an aspiring high language, Bengali, in its increasingly sophisticated form, came to have interestingly complex relations with Sanskrit and English, the two languages from whose tyranny it was supposed to emancipate its cultivated speakers. The resources of the colloquial language of the street—its early literary masters like Vidyasagar and Bankimchandra felt—were too meagre and insubstantial to carry the burden of such tasks. The more Bengali would replace Sanskrit as a high language, therefore, the more it would ‘become’ like Sanskrit in a sense, both in the direct sense of borrowing from its vocabulary, and being less accessible. Indeed, the new ‘high’ Bengali took on board a whole range of attributes from both Sanskrit and English, the two languages with which it existed in a relation of cultural contestation. Further, its internal economy would open up greater possibilities of hierarchy through the infinite refinement of ‘literary’ writing styles, and of ‘cultured’ pronunciation, an unfailing marker of increasing social differentiation.30
Sometimes this emergence of a standard Bengali and a culture of transaction of high functions in it rather than in Sanskrit or English is treated absentmindedly as a democratization of the linguistic field, a judgement to be taken with some caution. Within its incontestably democratic trends were lodged sharper inequalities of a new kind. The peasantry, the Hashim Sheikhs and Rama Kaivartas of Bankim's famous essay, stood no chance of comprehending the argument in which they figured, and which was made on their behalf—for no other reason but that they could hardly understand its Sanskritic grace.31 Ironically, a speech community certainly grew up in nineteenth-century Bengal, as Bankim's Vande Mataram illustrated, with seven crores of Bengalis, a vast majority of them illiterate, prepared to worship their linguistic motherland, extending to her a form of reverence earlier reserved only for scriptural deities. Yet, though this produced a sense of political community around language or speech, it was anything but a community of the same speech.
This historical sketch, though obviously crude and minimal, helps us put to rest some prejudices that social science analyses of identity formation have uncritically picked up from later nationalist discourse. Both in mature nationalism and in social science, it is customary to seek systemic differences between traditional and modern identities. A vulgar Weberianism which almost always subtly informs nationalist thinking on this question often makes it out as if linguistic identities are primordial, while the national identity of India is modern.32 This view is false, it appears, on two different counts. (i) The identities of region and nation are both products of the same historical-cultural processes which produced a mapped world out of the earlier fuzzy one. Political identities based on language are therefore equally modern though the languages on which they are based have a distinct historical existence from much earlier times. (ii) This historical imperative—i.e. both these are modern and historically intertwined, making it difficult to play one against the other—makes for a peculiar configuration of a two-layered identity for individuals and groups. A regional identity is subsumed in a larger national one—a fundamental historic fact that, later, more anxious and simple-minded forms of nationalist ideology have often sought to deny.33
Within Bengali fiction, Bankimchandra's work is fascinating, I have argued elsewhere, because he shows quite precisely how this consciousness is formed; how some crucial cultural choices are made, how within the newly formed intense consciousness of being Bengali critical weaknesses are perceived.34 The sufferings of political humiliation and rightlessness are felt not to be a peculiar problem of the Bengalis, but common to all those who lived under British rule. Besides, if a credible political coalition was to be built up, which could effectively menace British power, a regional identity of Bengalis alone was unlikely to impress them. Authors like Bankim displayed what I have termed an anticolonial consciousness before they chose their nation.35 After considerable hesitation, and a few false starts, Bengali creative writing, by the middle of the nineteenth century, accomplishes a substantial gerrymandering of the boundaries of its imagined community. Initially, it appears that the ‘we’ who should oppose the British are to be Bengalis; but after considering proposals about making Hindus their nation, they eventually choose to be Indians. Reflecting this process graphically, Bankim breaks the boundaries of his initially restricted Bengali regional identity and begins to represent Rajputs, Sikhs, and Marathas in their saying of ‘we’, in their collective self-description. As parallel processes go on in other parts of the country, this contributes, by the beginning of the twentieth century, to the making of the familiar contours of Indian nationalism. Still, under this nationalistic configuration of self-consciousness, other identities of language, occasionally linguistically perceived ones of religion or subregional cultures, are always present in an indistinct and politically inactive state.36 Much of current Indian politics revolves around how nationalism decides to deal with them—to attack and destroy them as competing attractions or to give them a place within its own internal architecture. Precedents of European nationalism, through which Indian nationalists initially sought to understand and think through their own world, presented an immediate problem. Successful European nations created, in most cases, culturally homogeneous states based crucially on the unity of one language. Indian national identity seemed to ‘lack’ one of the main prerequisites of a modern nation, a feature which helped them to weather periods of political adversity.37 Indian nationalism and its state inherited from its inception a ‘language problem’: i.e. the problematicity of its cultural formation in the absence of a single unifying language. Two ways appeared to be open to it and the state it aspired to establish after Independence was achieved. One arguable, but unlikely, road was to pursue a single-minded strategy of cultural homogenization based on the primacy (others would call it domination) of a ‘majority’ language. This would have implied trying to solve the cultural problem by a Soviet model: the only drawback was that the conditions for its practical realization did not exist. Hindi did not have a statistical absolute majority in India, nor was its political supremacy firmly established by the cultural processes of nationalist politics. It was therefore the second solution which was taken up by the national movement—a strategy of accepting the legitimacy of linguistic self-identifications of people in their regions, giving it a place in a second order identity of Indian nationalism and using a critically significant political diglossia.
The nationalist movement used two cultural devices to solve this difficulty of according legitimacy to regional identities but still requiring a second order self-identification of people as Indians. In its mature stages, it generated a plausible and powerfully articulated narrative of India's immemorial past, a logic of cultural unification lodged in the essentials or the depths of Indian history, into which, through the creation of a ‘composite culture’, Muslims were integrated in medieval times. Nehru's Discovery of India was of course the classic of this narrative construction which saw the imagining of India as having been accomplished in the past.38 As part of an anti-imperialist ideology, it had an understandably powerful appeal, though in my judgement the exclusive reliance on this narrative for solving all problems after Independence has had a deleterious effect on political thinking. This logic of creating unity within plural strands, the ecumenical tolerance which gradually absorbed others into an extended notion of the self, was felt to have created this imagination and its practical institutions already in the past; British colonialism disturbed its continuity only temporarily. This implied that once the national state was achieved, this buried cultural imagination of India would again come into its own. This became a deeply held and widespread belief that undergirded much nationalist practice of mobilization and state-building. Nehru's text, written at the moment of its imminent triumph, was a systematic and passionate statement of this narrative of culture; and I think Nehru's curious inattentiveness to cultural construction cannot be explained without understanding how seriously his generation believed in the story they had told themselves. As I said earlier, as an empowering narrative for the nationalist movement this picture of India's past was immensely powerful. For rational reconstruction of the state, in the less romantic period after freedom, this narrative was more a hindrance to realistic policy than a help. It put the process of the cultural construction of the nation in the past, rather than in the future; it saw that as an accomplished task rather than as a requirement to be fulfilled if India was to stay together as a nation state. It encouraged forgetfulness and negligence about the cultural reproduction process in all its forms—in everyday life, social practices, and education, its most significant institutional form.39
But this narrative structure was supported by a deeper cultural arrangement that was more directly linguistic, which determined and mediated crucial political relations, and decided who could, literally, speak to whom. To rejoin my earlier argument, the mixed elite drawn from the educated, bilingual intelligentsia of different regions found themselves within a triangle of cultural or discursive exchange—exchanges with the lower orders of their regional people, with peoples of other regions through their leaders who were culturally and politically similarly placed, and exchange with the British, apart from the more general and exalted need to receive information about scientific developments in the Western world. The most effective and economical means of dealing with this set of discursive demands was a diglossia or cultural bilingualism.40 Through political experimentation, the national movement came to settle on this device of political culture quite fundamentally. Common training in English education, with a common syllabus, common cultural preferences and tastes, and common biases provided the preconditions for this situation, but it was cemented by the evident functionality of this arrangement for the growth of an Indian political movement.
Some consequences of this invisible cultural fact deserve some mention. It ensured in an oblique and unobtrusive way a kind of elite domination of the higher levels of nationalist mobilization, even though subaltern dissent against colonialism was more extreme and visceral and middle-class defiance more careful and circumspect. The linguistic economy partly guaranteed that despite this, it was only the middle-class elements who would provide the all-India leadership. Vernacular speakers could storm into the leadership of linguistically homogeneous areas, but there their political stars stopped climbing. Compulsions of this kind, working not through the explicit and usually resented logic of social class or status, but the subliminal agencies of language and communication, introduced an elitist counterweight into the inherent populism of the nationalist movement.
This diglossia was also politically rational in a narrower sense. Indian culture, it is often remarked, is characterized by an easy heteroglossia. After the entry of English education on a large scale, the situation stabilized into a fairly common structure of diglossia among the educated. A pure Indian bilingualism (i.e. using two Indian languages equally fluently for serious intellectual activity) was not very common among the elite: bilingualism rather meant the ability to use a vernacular and English. This meant that a bilingual person was not thickly aware of cultural or political development in only two areas; under this arrangement, he was of course thickly aware of his own vernacular-based regional culture, but also thinly of all others. The success of nationalism was made possible by political conceptual co-ordination between the various vernacular regions. However, often this bilingualism was internally uneven; those who could use these languages were unequally fluent in them. This yields a structure similar to the flower-shaped arrangement suggested by the Dutch sociologist of language Abram de Swann, with the inconvenience of having widely irregularsized petals.41 The petal representing Hindi would account for about 40 per cent of India's population next to those of comparatively small languages like Kashmiri or Manipuri. The recent admission of the political significance of some languages which cannot he ascribed specific regional or spatial location, like Nepali or Sindhi, would make the structure of actual linguistic exchanges still more complex. Additionally, such bilingualism is bound to be in individual cases asymmetric: a speaker would not have equal competence in both, and since competence confers and abridges rights and capacities, each would thus support the cause of the language he is more proficient in. For individuals, thus, we should write not only that person X has E+V, but also indicate a certain slope or precedence: E/V or V/E, because these would, in political action, make a material difference.
Fig. 2
Language thus acted as a necessary process of filtration, or ‘gate-keeping’; it would filter out inconvenient, extreme, radical, intransigent demands from subaltern social groups from reaching higher bodies. This was no small factor in enabling the higher decision-making bodies to maintain their immaculate middle-class ambience of restraint, and polite gentility. This gives rise to a certain paradox of participation. Certainly, the English-knowing bilingual elite represented the largest number of people from all vernacular regions; but this implied an inverse relation between the extent or width of representation and its intensity or intimacy. Conversely, seen from the point of view of ordinary people living in the dialect or vernacular spaces of the pyramid of speech, the further their demands were carried into the political world by their representatives, the less the control they could exercise over them.42 The Indian people became, like democratic people elsewhere, the generalized reason, the universal justification for things done in their name which they would have found largely incomprehensible. The greatest example of this paradox was the constitution. It was apparently part of their tryst with destiny to give themselves the longest, the best, and the most intricate constitution in the world: there is a minor irony considering the fact that over 70 per cent of the country was illiterate and had no clear idea of what they had been given. Making a constitution of this kind was a strange but historically inevitable mixture of giving people unprecedented rights and also keeping them securely out of their reach.
Fig. 3
What is remarkable in this story of language and the conferment of identity is a paradox that emerges in a comparison between pre-Independence years and the time after Independence. Abstractly, it appears that the cultural problems of nationalism can be handled better when the ideological force of nationalism is assisted by the material power of the modern state. Yet, on an historical view, it seems that the Indian national movement, when it had to contend with the enmity and obstructiveness of the colonial state, tackled some cultural processes better than the nation-state. In certain ways, the guiding logic of the complex network of culture, language, identities, and discourses seems gradually to slip out of the grasp of the state, leading to a further paradox. The state is becoming increasingly powerful in technical and material terms and at the same time abjectly ineffective in regulating, let alone solving, cultural conflict.
The Nehruvian elite, which inherited effective control after Independence, faced a complex initial situation. Indeed, the issue of an official or ‘national’ language occasioned some of the most acrimonious debates in the Constituent Assembly. Its proceedings also showed another ironic fact: how deep the influence of European precedents was on the minds of the Indian intelligentsia. On the language issue, the assembly saw a division between moderates and Hindi extremists centred on three major issues: (i) whether India should have a national language, and if it did whether it would be the broad ecumenical form of Hindustani, or a Sanskritized Hindi, (ii) what would be the status of English, how quickly this mark of foreign subjugation was to be phased out and replaced by the national language; and (iii) the largely symbolic problem about the system of numerals to be adopted. A nation, it was argued by those who favoured Hindi, must have a national language; and this could only be an Indian language. Of the various Indian languages, since there was, ‘unfortunately’, no clear majority language, this place should go to Hindustani or Hindi as the language spoken by the largest plurality. Strong and insistent demands were also made in the Constituent Assembly for having a Hindi version of the constitution adopted as the original in place of the English one. Interestingly, even moderates on the language issue conceded the idea of having a single language being a precondition of firm, unassailable nationalism. Eventually, however, the Assembly adopted a more complex, if less decisive, line in favour of what was known as the Munshi-Ayyangar formula, which pragmatically reflected the structure of discursive exchange which underlay the national mobilization.43
The influence of European examples was so deep that it determined the manner in which the constitution-makers framed their question. A single language, one of the prime features of European nationalism, was missing in the Indian case: the nation-state lacked a common language. Pragmatically, with their comparative knowledge and historical sense, however, the Nehruvian leadership realized that the cultural form of Indian nationalism must be different from the European norm. Nehru's government pursued a policy of gentle exhortation on the significance of Hindi as an official language for the union government and communication between states. The actual use of languages in government, however, diverged substantially from its declared policy. The introduction of large-scale economic planning, and the related emphasis on science, technology, and the processing of standardized information naturally increased the subtle forms of power exercised by the high bureaucracy simply on account of their inabilities and preferences. They could converse more easily in English and carry on these intricate activities more conveniently without the acute problems bound to be raised by translations many times over between the vernaculars. English became more entrenched in the bureaucratic and private managerial echelons after Independence because the earlier political compulsion of using vernaculars to underline self-respect had less relevance. To those outside the charmed circle of English education, this gave grounds for the complaint that India was reconquered for English after Independence.44 The language of India's nation-building was, unfortunately, English.
The first wave of regional disturbances that the new state faced was not of its own making: it may have contributed to their occurrence only by forgetting the promises the Congress had made. Congress acknowledged during the freedom movement that administrative boundaries instituted by British rule were arbitrary, and had to be reordered after Independence. After 1920 the Congress organization decided to work on the basis of linguistic provinces rather than of administrative boundaries, implicitly reinforcing the commitment. Events at the time of Partition and the experience of constitution-framing may have made the ruling group more cautious in this regard. However, the original organization of states was, by the mid-1950s, facing serious disaffection in some parts of India. The structure of this disaffection is interesting. Initially, regional discontent did not arise in all parts of the country simultaneously, but only in those parts in which some communities had, under the colonial dispensation, enjoyed sub-colonial advantage over others. Early opportunities of colonial employment had imparted a peculiar class structure to Bengali society, since Bengalis had supplied the entire demand for middle-class bilingual professionals not only for the Bengali-speaking areas but for the entire northern and eastern parts of the empire. In the Madras Presidency Tamil enjoyed largely similar superiority for similar reasons. Linguistic groups who were disadvantaged by this naturally hoped that with the end of colonialism such privileges would also disappear. Strong regionalist sentiments arose over Congress tardiness and hesitation. Although this caused temporary embarrassment to Nehru's government, some peculiarities in the situation made the eventual transition to linguistic states remarkably easy and smooth. The actual grievance was against the regionally superordinate cultures rather than the central government. The moral strength of the case of these regions was matched by the weakness of their adversaries. Besides, constitutional rights, unlike material resources, were not scarce or restricted things, and thus the achievements of the rights of one group, instead of threatening, furthered the claims of others.
Interestingly, although the Nehru government conceded these demands, and gave it a generalized form by instituting the states reorganization commission, Nehru himself expressed some forebodings. First, he thought living in mixed, multilingual administrative units would provide education in common living and reduce linguistic chauvinism.45 Secondly, he anticipated difficulties in administratively applying the principle: areas like the north-east have in fact proved chronically difficult to settle linguistically.46 Yet Nehru agreed to linguistic reorganization because of the undeniable force of the argument of democracy—that use of the vernacular in administration would bring government closer to the people, because that was the language they were illiterate in. In the case of the demand for a Punjabi suba, he persistently refused to grant it because he thought this was merely a communal divide under a thin subterfuge of language.
The reorganization of states on the basis of language still left major difficulties. In some areas, like Assam, linguistic groups were too mixed to allow for their clear administrative disentanglement without making the units too small. Even such reordering has not, much afterwards, solved the question of enclaves within majority languages. Hindi remained anomalously placed inside the constitutional structure, since it was stretched from Bihar to some parts of Punjab. This might have encouraged some kind of interest coalition among its speakers, and since their resources taken together would have been quite substantial, this may have renewed the conflicts of the Constituent Assembly once more. A potentially difficult situation was avoided precisely due to the internal divisiveness of the Hindi language area. Hindi was historically less standardized than other vernaculars,47 and several of its subregional forms could claim to be much more than mere dialects of some general norm language,48 though khadi boli from early on made known its pretensions to that status.49 Apart from the problem of the standard, Hindi was riven by the acrimonious dispute between a Sanskritized Hindi and a Persianized Urdu tearing apart the spontaneous historical form of Hindustani, the language in which most people actually spoke in North India. This was also the form more easily intelligible to neighbouring vernacular speakers, and in which much of the high literature was written. Internal dissensions within Hindi led to a circumstance which turned ironically to be beneficial to the new state. The potential mobilization for Hindi as a ‘national’ language which would compulsorily replace English, and make India look more like a unilingual European nation-state, failed to gather momentum because it was hard to decide which kind of Hindi would enjoy these privileges of universal aggrandisement. After 1956 the union government has faced infrequent difficulty from this direction, except for some areas like the north-east, where linguistic determination of state boundaries has remained intellectually unconvincing and politically contested.
One of the major features of Nehru's development policy that has escaped serious critical attention was its economism: a superstitious belief in the powers of economic growth to dispel all evils like rays of light in the enveloping darkness of traditionalism. Thus, although individual members of that early governmental elite had impressively high cultural attainments, their dedicated economism made them deeply negligent about processes of culture. While economic processes were minutely analysed, alternative policies assessed and debated, they did not see cultural processes or institutions as serious objects of either historical reflection or deliberate policy, except in terms of budgetary allocations. Unattended cultural reproduction processes, left to the violence and malignancy of both a vicious market and un-principled state manipulation, have now produced something like a massive counter-revolution of culture. In the rest of this essay I shall try to present a bare outline of this process and its relation to the crisis of the state.
If states, to be durable, have to depend on something like a common consciousness, the Indian national state faced an obvious difficulty, not so much because of the multiplicity of its languages as due to the fact that there were two rather different types of ‘common sense’ about the world that were lodged in the English and the vernacular segments of its culture. True, serious transactions had begun between these two orbits of common sense in the nineteenth century and during the freedom movement. But such exchanges remained unfinished and irregular. As India did not have a natural coincidence of a common culture and a common language, the quotidian self-recognition of its citizens depended crucially on an argumentative and ratiocinative process which could be fashioned by education and deliberate cultural policies. Such a self-recognition of Indianness is primarily a conceptual construction and is to be produced and maintained by means of cultural arguments quite different from the easy recognition of a collective self given in the mere utterance of a natural language. It has to be an indirect, second-order process, quite distinct from the process of linguistic self-recognition, of creation and stabilization of a political common sense. Identity in India must have more than one layer. From this point of view, Nehru's government and its successors have consistently misconstrued the historical function of education. Due to their deeply economistic reasoning, they considered education an input in the creation of skills necessary for a modern economy, and entirely missed its more fundamental common-sense-forming role. These roles are performed by different levels of the educational structure. Technological education was given nearly exclusive priority, including science education as its base, and naturally also higher education. This led to a state of affairs in which, while a large number of middle-class graduates were trained to be engineers, medical practitioners, or planners, the large majority were not trained in being Indians. Such training, sorely missed in recent years, has been left to the canned patriotism provided abundantly and free of charge with the television news.
The later evolution of Indian politics has shown such superstitious belief in the enlightening powers of economic processes to be fallacious. Economically upwardly mobile farmers have not become more kindly towards other rural classes—their traditional neighbours in the daily life of the countryside.50 Professional middle classes have not shown increased comprehension of others’ justified demands on the resources of the state.51 Under the primitive capitalist conditions that exist in India, greater economic prosperity does not seem to increase a group's understanding of the overall requirements of distributive justice: rather an insensitivity to distributive questions may produce the drive for strong commercial enterprise. In general, economic growth simply failed to radiate the kind of illumination and understanding expected of it by the development model.
Cultural defaults of the nation-state, begun in the absentmindedness of Nehru's government, were intensified by successor regimes. While all governments paid ceremonial attention to high culture and art, made larger budgetary allocations to higher education, they considered their responsibility completed by such exertions and showed little interest in organizing social reflection on the relation between democracy, poverty, and culture. These delicate processes were thus left to an increasingly chaotic and iniquitous market of cultural skills and goods whose spontaneous trends were ‘corrected’ by an increasingly arbitrary state which specialized in correcting injustices by other injustices rather than the unaccustomed intellectual practice of following a judicious policy over a long term.52 Indian cultural life saw the emergence of a progressive intelligentsia whose only vocation in life, after pursuing individual self-interest, was to lead the Indian people, especially the indigent, to liberation—but it had forgotten to pick up their language. Those who spoke those languages often had political horizons too narrow to consider national interests, and were mired by pulls of regional or subregional selfishness. The contradiction between the two common senses, lodged largely in English and the vernaculars, has now assumed close-to-crisis intensity, bringing the conceptual basis of the Indian state to great strain. The process of the imagining of India has fallen a victim to quietly effective processes of bifurcation and heterogeneity. The neglect of education—leaving it to the market—meant a chasm between high and low culture; incompatible with the logic of democratic politics, this has persisted. Entrant elites at the state levels of the Indian federation, drawn from farmer groups, perceived this system, with some justification, as one of cultural or linguistic untouchability, an English-based caste system. On the other hand the urban elite, systematically withdrawing from the diglossia into exclusive use of a coarsening functional English, reacted negatively, often showing their paradoxical belief that only those who could not speak in any Indian language could be safely called Indian nationalists; others were assigned derogatory ethnicities like Punjabi, Bengali, Bihari, Tamil, etc. The unplanned and unorganized spread of vernacular education, which is often distinctly inferior in quality, accentuated this process. The diglossia that had formed the cultural basis of the national movement has tended to fall apart. Educational differentiation has led to a situation in which a monolingual English-using elite is faced by an equally monolingual vernacular aspirant group contesting its power with increasing urgency. Both languages suffer coarsening in this friction, though it is the slow slide of English into functional inelegance which is most observable.53 Thus the discursive structure of nationalism is being broken down in two ways. First, the English-using managerial elites at the top of the pyramid are getting distanced from the vernacular-using lower slices, failing to perform the task of creating and disseminating a nationalist common sense which must not be an ossified collection of ideas from past nationalist leaders, but an active configuration of ideas which all negotiate. Secondly, since in our model the major part of the exchanges between the vernacular segments did not occur laterally, but through the bilingual top, these vernacular sections are coming into increasing friction.54
By and large, till recently, the Indian state has faced relatively infrequent difficulty from linguistic separatism. Common peoples’ quotidian linguistic identities seem to lie quiescent in a pleasant, inoffensive, taken-for-granted kind of way, until they become vehicles for other types of grievance through a malignant elective affinity.55 Though both the Assam and the Punjab problems show pronounced linguistic aspects, in these disturbances the linguistic demand exists in interestingly complex relation with others.
One can also glimpse the emergence of a different sort of problem in Indian politics which might revive the question of Hindi chauvinism in a new form. Since Indira Gandhi's victory in the general elections of 1971, directly populist rhetoric had been rather faint in Indian political discussion. Populism has in recent times seen a great revival, forcing democratic issues on the political agenda of the nation, mixed in a distinctly Tocquevillesque fashion, with dangerous disregard for norms of institutional circumspection. Mrs Gandhi spoke in the name of the great majority of the poor; others now speak in the names of the equally great majorities of the ‘backward castes’, of the neglected Hindus, or the insulted, disprivileged speakers of Hindi, the language of the majority of Indians. We saw earlier how the question of what to do with Hindi was alleviated by the other one of what was Hindi. Since Independence, the situation has changed in several ways. The first is the noticeable retreat of Urdu into the cultivated middle-class parlour, under the assault of a state-patronized and Sanskritized Hindi. Because of the evident disruption of the nationalist diglossia it has become possible now to claim that the division of the rich and the poor, of the urban exploiters and the rural sufferers, the upper castes and lower, are all translatable into the symbolic divide between English, the language of the undeservingly privileged, and Hindi, the speech of unmerited suffering. A significant recent development has been the declaration of four chief ministers of northern states that Hindi should displace English in official administrative functions. They explicitly demanded that it should be used in inter-state communication, and as the exclusive language for entrance examination of various types of government and semi-government bureaucracy. Its major supporting argument is that insistence on English artificially and unjustly wrongfoots speakers of Indian languages—an entirely understandable idea up to this point. Yet their demand is not that linguistic iniquity should be ended, but that it should be turned in favour of native Hindi speakers. They should enjoy the privilege that belonged to English users because they constitute the ironic majority of 40 per cent of the Indian population. Clearly, this would not be perceived as equitous by those who do not speak Hindi, only a proposal to disadvantage others. English, they would claim, is more equitous because it places all Indians in a position of equal disadvantage.
Undoubtedly, in the past forty years India has seen a process of uneven but also unprecedented enfranchisement of the common people. Language rights constitute a vital part of this process—the right to protest becomes attenuated if it is not a right to protest in one's own language. Through the processes of democracy, identities have undergone rapid reconfiguration, and political groups are naturally trying to split and recoalesce identities in ways preferable to themselves. Communal elements have put forward the considerable appeal of a Hindu majoritarianism. In response, the Janata Dal advanced a consolidation, also majoritarian, of backward castes, splitting the potential Hindu coalition on caste lines. The proposals for Hindi majoritarianism, though statistically on somewhat weaker grounds, bear interesting connections with Hindu communal sentiment. Historically, the trend most dangerous for the survival of Indian democratic government will be a process of mutual attraction, a fatal elective affinity, between various forms of majoritarianism, a coalition struck between differently grounded majoritarian demands. Indeed, as some people have argued at different points in recent history, the majority demand of a Hindu state, instead of the present one based on ‘Western’ secularism and the majority demand for a Hindi-dominant state, can coalesce because the two groups of beneficiaries of such privilege would overlap to a large degree. ‘Being Indian’ has involved a multi-layered identity in which the upper and more general identification subsumes, but does not nullify, the less general and particular ones. The new majoritarian trends threaten this historical form and the specific equilibrium that the nation-state has tried to give it. In a major irony of history, although most of these forces and their political leaders see themselves as being uncompromisingly anti-English and anti-Western, they wish, despite their strident indigenous rhetoric, to replay a European paradigm of nationalism in which being Indian must find confirmation in speaking Indian and writing Indian. If such a convex majoritarianism gradually takes shape, convex because this puts the majoritarianism of religion on top of that of language, compensating the statistical inadequacy of Hindi by the overwhelming dominance of the Hindus, this can really fatally threaten the cultural presuppositions of the Indian nation-state. That would illustrate the disruption of a democratic state by the unqualified and uncomprehending application of an equation between democracy and majority rule. For democracy cannot exist without the rule of the majority; but equally, it cannot exist as only majority rule, unmodified by other subtler, juster, latent, equitable principles which make it not only the most acceptable, but also the most delicate and perishable among forms of modern governance.
Austin, Granville. 1966. The Indian Constitution: Cornerstone of a Nation. Oxford: Clarendon Press.
Chattopadhyay, Bankimchandra. 1964 (rpnt.). Bangadesher Krishak. In Bankim Rachanavali, vol. ii. Calcutta: Sahitya Samsad.
Cohn, Bernard S. 1987. An Anthropologist among the Historians and Other Essays. Delhi: Oxford University Press.
Dasgupta, Jyotirindra. 1970. Language Conflict and National Development. Berkeley: University of California Press.
de Swann, Abram. 1988. In Care of the State. Cambridge: Polity Press.
———. 1990. Political and Linguistic Integration in India: Monopolistic Mediation versus Language Integration. Draft paper presented at seminar on Changing Relations between State and Society in India and Trends Towards Emerging European States. New Delhi: 5-9 March.
Derrida, Jacques. 1974. Of Grammatology (translated by Gayatri Chakravorty Spivak). Baltimore: John Hopkins University Press.
Douglas, W. and P. Hill. Trans. 1971. The Holy Lake of the Acts of Rama. Bombay. Oxford University Press.
Fox, Richard G. Ed. 1977. Realm and Region in Traditional India. Delhi: Vikas.
Gadamer, Hans Georg. 1975. Truth and Method. London: Sheed and Ward.
Gellner, Ernst. 1983. Nations and Nationalism. Oxford: Blackwell.
Gopal, S. 1979. Jawaharlal Nehru, vol. 2. Delhi: Oxford University Press.
Kaviraj, S. 1991a. The Imaginary Institution of India. In Partha Chatterjee and Gyanendra Pandey, eds, Subaltern Studies VII. New Delhi: Oxford University Press.
———. 1991b. On the Construction of Colonial Power. In Shula Marks and Dagmar Engels, eds, Foundations of Imperial Hegemony. Oxford: Oxford University Press.
———. 1991c. State, Society and Discourse in India. In James Manor, ed., Rethinking Third World Politics. London: Longman.
———. 1995. The Unhappy Consciousness: Bankimchandra Chattopadhyay and the Formation of Nationalist Discourse in India. Delhi: Oxford University Press.
Khubchandani, Lachman M. 1991. Language, Culture and Nation-building. Shimla: Indian Institute of Advanced Study and Manohar Publishers.
Prasad, R.C. 1989. Ed. and trans. Ramacharitamanas. Delhi: Motilal Banarsidass.
Saberwal, Satish. 1989. India: The Roots of Crisis. Delhi: Oxford University Press.
Sen, Sukumar. 1965. Banglar Sahitya Itihas (in Bengali). New Delhi: Sahitya Akademi.
Singh, Namwar. 1975. Dusri Parampara ki Khoj (in Hindi). Delhi: Rajkamal Prakashan.
First published in Nationalstaat und Sprachkonflikte in Süd- und Südostasien, ed. Dagmar Hellmann-Rajanayagam and Dietmar Rothermund (Stuttgart: Steiner, 1992), pp. 25–68.
1 A standard history of the Bengali language and its literature is Sen 1965.
2 Bernard Cohn has done pioneering work on these questions: see especially chapters 6 and 10 of Cohn 1987. An excellent discussion of language and political identity can be found in Dasgupta 1970.
3 Derrida 1974: Chapters 1 and 2, argues this against earlier linguistic theories.
4 Gadamer 1975.
5 There are references often in traditional Indian thinking to nadabrahma. Nada is the originary meaningful sound which is given great ceremonial significance in two of its cultural forms, the intelligent power of words and the sound of music, its aesthetic often begins with general considerations of this kind.
6 Some internal diversities in the tradition must be noticed. In the Buddhist tradition there was greater reliance on writing in order to fix the meaning of something especially valuable. The preachings of the Buddha are written down by disciples; but the Hindu tradition by contrast seems to be interested in inscribing significant ideas on the more quizzical tablet of memory.
7 Both the Arthashastra and the Ain-i-Akbari, central texts for two different periods, point to habitual and widespread use of administrative records.
8 The difference is between the ability to think through a problem and the ability to reproduce from memory standard recipes. Even current Indian educational practice shows over-reliance on memory-oriented skills.
9 The difference between the two forms of the caste system—of varna and jati—might have something to do with this.
10 The sociological controversy about caste is instructive in this respect. Without denying Dumont's general claim that what is essential to caste is the ideological form of hierarchy, one can still admit the evidence of historical change through the process known after Srinivas as Sanskritization.
11 Considerations of such things get unnecessarily embroiled in controversies about nationalist pride, accusations of orientalism, and such other judgemental disputes. Few social scientists in India have, for instance, pursued the question of the exact relation between what is indubitably common and what is conceptually public. Existence of commonness does not mean that this will be conceptualized as public space or public interest or public institutions. The work of Satish Saberwal raises this question (though not in this language): cf. Saberwal 1989.
12 This linguistic configuration continued undisturbed down to the time of Ram Mohan Ray who was proficient in both Sanskrit and Arabic-Persian, besides his native Bengali and colonial English. It is remarkable how quickly this heteroglossia is destroyed by the cultural processes of colonialism. By the time of Bankimchandra and Tagore, proficiency in Arabic-Persian and familiarity with Islamic culture are not required as marks of a cultured Bengali. The Bengali bhadralok elite had decided to give themselves a resolutely Hindu past.
13 I have suggested elsewhere that the precolonial state was marked by the twin features of being spectacular and marginal: Kaviraj 1991c.
14 Mangalkavyas were composed in honour of nonclassical Hindu deities like Manasa, the snake goddess, Chandi, and one of the best-known texts in this tradition is Mukundaram Chakrabarti's Chandimangal.
15 Several English translations of the Ramcharitmanas are available: for instance, Prasad 1989 and Douglas and Hill 1971.
16 A good example from Bengal, though in a later period, is Bharatchandra, who wrote some wonderfully innovative metric poetry apart from his larger works, Annadamangal and Vidyasundarkavya; but he, too, established his claim to poetic recognition by insignificant versification in Sanskrit.
17 Namwar Singh 1975, in which the author's argument partly follows the critical work of Hazari Prasad Dwivedi, specially his reading of Kabir.
18 I have discussed this with reference to Krishna and Radha in Kaviraj 1995.
19 The Sanskrit word for such writers is kumbhilaka.
20 Though both in the Valmiki Ramayana and in the Uttararamacharita of Bhavabhuti, Rama is seen moved to tears, a demeanour not considered incompatible with his status as a hero or an avatara of Narayana.
21 Gellner 1983.
22 Kaviraj 1991b.
23 Social groups have fuzzy ends or edges in a world in which differences are organized as spectrums rather than divided by boundaries. Arguments of this kind, though conceptualized differently, are found in Fox 1977, especially the papers by Burton Stein, Bernard Cohn, and Ainslie T. Embree.
24 Spatial perception could be similarly rendered into a traditional idea of farness against the modern sense of distance.
25 Anthropological work provides sufficient evidence about a sense of time that is linear in relatively short spans, but whose edges become vague before and after a few generations.
26 What I mean is that they may think of place X as ‘far’ Y as ‘farther’. They would not have the means, much of the time, to say that X is 500 miles, and Y is 5000.
27 Kaviraj 1991b.
28 The economic aspect of this argument—that colonialism did not introduce a capitalist economy of the Western type, but a peculiarly distorted form—is too well known to require special mention.
29 Cohn 1987. See the essay entitled ‘Census, Social Structure and Objectification in South Asia’.
30 Subsequently, Bengali showed an interesting line of development on this count. Through the radio and the great popularity of the Bengali film, the ‘cultured’ way of speaking Bengali has become since the 1960s very widespread. While this may have extended the ‘civilizing’ process in a certain sense, the diction and accent have become distinctly less significant social markers.
31 Chattopadhyay 1964: 288.
32 Nationalism in India has had a particularly strong connection with modernist developmentalism. As it develops, nationalist discourse decisively sides with modernity, and sees everything modern with approval and things ‘traditional’ with suspicion if not with straightforward hostility. Additionally, it often tends to consign all premodern forms to primordiality (in the derogatory sense of an aboriginal primitiveness). But it is useful to distinguish between two distinctions widely used in the literature. Primordial/modern is in one sense a chronological historical distinction. In a related but quite distinct sense, this may also mean a taking/making of identities: primordial identities are those people have to take, modern ones are those they make on their own, or so they would like to believe.
33 This is a point that has been made consistently by radical left parties against the grain of mainstream nationalism.
34 Kaviraj 1995: ch. 4.
35 What I mean is that their decision to oppose British colonialism comes chronologically before they decide who this ‘we’ was—Bengalis, Hindus, Indians?
36 The connection of Urdu with Islam, and of Punjabi more recently with Sikhism. Yet, it is interesting to note, with Khubchandani, that Urdu, especially Hindustani, is the language of a culture rather than of a religion. Similarly, those who equate Punjabi too closely with the Sikh religion forget that there is a Punjabi-speaking people in Pakistan, and there is a Punjabi ‘diaspora’. Khubchandani 1991.
37 Curiously, any historical or social feature that Indian society has that is different from standard forms in European history is seen as a feature it does not have; its presence is seen as a lack and theorized consistently in this fashion.
38 For a critical discussion of this narrative structure, see Kaviraj 1991a.
39 Though there is no space for outlining this here, my argument requires as a complement a critical discussion about educational structures, practices, and policies.
40 Any observant student of Indian culture would notice the common prevalence of diglossia and heteroglossia. After the coming of colonialism, the typical structure of Indian bilingualism is English-vernacular rather than vernacular 1 and vernacular 2.
41 The floral model is used in de Swann 1988: ch. 3, and used in a discussion on India in idem 1990.
42 In democratic theory a distinction is commonly made between delegation and representation. Something like this also exists in the Indian national movement, though this is seen as two pragmatic forms rather than two theoretical positions engaged in explicit debate.
43 For an excellent account of the debates on the language issue in the Constituent Assembly, see Austin 1966: ch. 12.
44 One of the telling examples of this was the alteration in the structure of school education. A strong emphasis on symmetrical bilingual education in the period from the 1930s to the 1960s produced a fairly distinctive school system in the areas that had enjoyed the differential advantages of colonial education. This was gradually replaced by a flourishing privileged private sector of English-medium schools and an underprivileged state sector of vernacular schools. Because of the strong correlation between the acquisition of English language and chances of good middle class employment, there has been a constant demand for rapid expansion of these ‘public schools’.
45 Occasionally, such ideas went to an unpractical extreme—as in the case of the intensely unpopular suggestion for the merger of Bengal and Bihar. For Nehru's thinking on linguistic states, see Gopal 1979: ch. 12.
46 Professional linguists sometimes argue that the idea that there should be sharply defined territorial boundaries not only for nation states but also for linguistic units inside them was unsuited to the Indian situation and bound to cause problems. Cf. Khubchandani 1991.
47 For a discussion of the issues involved in the Hindi-Urdu-Hindustani debate, Khubchandani 1991, and an earlier account in Dasgupta 1970, chs 2 and 6.
48 Maithili has caused endless debates on this count—whether it was a dialect of Hindi or a separate but closely related language.
49 Even this has to be qualified, because during the early development of modern Hindi literature eminent figures like Bharatendu Harishchandra had suggested khadi boli as the form of prose and brajabhasha as the form for poetry.
50 The demands of rich farmers show this insensitivity quite dramatically.
51 The middle class's hostility to any suggestion of redistributive moves was reflected in its violent reaction to the declaration that the recommendations of the Mandal Commission would be partially implemented.
52 This is reflected in the irrationalities of the schools system. Public schools, run on lines of merciless private enterprise, treat these institutions as business propositions with high fees, impossible entry criteria, and a quality of training which is hardly commensurate with the expenses or the advertisement. Government schools, starved of basic resources and plagued by bureaucratic controls are not allowed to stabilize into any reasonable pattern by the appallingly arbitrary way in which changes are made in their educational schemes.
53 I would argue against some enthusiastic linguists that while the Indian way of using English must be legitimately different, the question of linguistic elegance should not be confused with this. Elegance or distinction is achieved when language rises above mere functionality, and can express complexity and subtlety of ideas. From this angle one can find a clearly perceptible coarsening of the linguistic culture. And it is not the inadequately bilingual speakers who are responsible but the parasitic, culturally sterile, and imitative elites whose only language is this vulgarized, subliterary English.
54 It must be seen, however, that this model is meant to be an ideal-typical construction. Exchanges took place bilaterally between regional cultures: e.g. Bengali literature (especially the novels of Bankimchandra Chattopadhyay, Rabindranath Tagore, and Saratchandra Chatterjee) were widely translated in other languages. My point is that the primary political exchange happened systematically through bilingualism of the elite.
55 Khubchandani 1991 makes this point about Punjab. Despite the creation of a Bengali-speaking Bangladesh, there was no noticeable urge on either side for a sentimental merger.
