ISSN (Online) - 2349-8846
-A A +A

Three Planes of Space

Examining Regions Theoretically in India

Sudipta Kaviraj ( teaches Indian Politics and Intellectual History at Columbia University.

There are three ways of looking at the question of regionality in India: “generalisation,” “fragmentation” and “composition.” Generalisation means gathering primary information from a determinate region, and assuming that all parts of India possess identical or similar characteristics. Fragmentation means believing that, since the regions are so fundamentally real, nothing existing beyond the regional level has any serious, compelling historical reality. Composition asserts that regions are historical and remain bound together within one single frame of some kind: political, economic or cultural.

Just like the concept of the dog that does not bark, the concept of a region does not exist on the material ground of space. It is a concept and exists historically and, like any other social concept, is subject to the basic rules of historicity. It is always legitimate to ask when and why it came into existence, and what people were doing with it and to themselves in devising a concept exactly like that.

To think fundamentally about something is a paradoxical act—simple in some ways, hard in others—of thinking of our categories, even the most elementary among them, as provisional, fallible and corrigible. These three characterisations are separate types of activities. To treat a concept as provisional—the most profound task of historicisation—is to acknowledge, as an a priori rule, that ideas, which we think through to describe, analyse and evaluate social objects, are historical and may not have existed at all times. This implies that they appeared at a particular time, and will disappear presumably. But this also imposes a responsibility on us to try and determine when and why a concept emerged. Concepts can be fallible in many ways. It could be that academics use a concept unproblematically, which ordinary people do not either use, or use in a different way. Corrigibility naturally stems out of this prior task. If we find that either our concepts do not do their referential work, or actors use others, there is a demand for correcting them. Any historical study of social science or historical thinking will discover quickly that much of the history of social science consists precisely in this—how the categories through which analysts see and examine their social objects have changed in history, and with what consequences. Here, we are concerned mainly with two concepts, namely India and its regions. I suspect that India is in some ways the much more problematic concept, and therefore more interesting to examine; but, historically speaking, regions are no less challenging.

Concepts like “nation” and region are historical in two clear senses. In the first sense, we can, the moment we expand the temporal span of attention, see that these concepts are referentially unstable/dynamic and have changed through historical processes like state formation, economic change, cultural configurations, etc. These are also historical in the second sense, in that these are formed by contingent epistemic processes, that is, through highly specific ways in which intellectuals or social agents have epistemically formed these concepts by actual, determinable epistemic/cognitive moves that we can study. To think about them critically is to suspend the easy belief that these are self-evidently clear, but ask what these are like: Are these real in the sense that they point to some definable reality, and real in what ways? I shall start with the reality of regions and in the end I shall add a few disorienting/critical remarks about the always contentious idea of India.

How regions are formed, and how these formations changed in history are among the central questions for understanding Indian social reality, which, strangely, remains underexplored in social science. Diversity is generally celebrated, or more recently with the rise of Narendra Modi’s Bharatiya Janata Party (BJP), deplored, but rarely studied with attentiveness. Yet, one of the most fundamental truths about India is that it is, in a manner of speaking, unlike itself. Its capacity for confounding generalisations is infinite, and the primary reason is that whatever we say about “India,” it is always possible to find a counterexample to contest it.

What we designate as India is such a large space that it is reasonable from the start to expect that what it contains inside would not be uniform whatever dimension we consider. We should presume lack of uniformity or difference/diversity. As there are several vectors of difference—religious identity, natural geographic variety, language—each of these can in principle yield a one-dimensional map of diversity. India is so difficult to conceive precisely because of this diversity of diversities, the different kinds of difference that constitute it. Each of these elements of diversity—religion, language, caste—yields a spatial map, and the idea of India at any time is a convex product of such maps of single-element diversity superimposed on one another. I think this form of profound complexity should be the taken-for-granted starting point for any examination of Indian politics, large or small. It is important to remind ourselves that the effects of this structure of complexity have to be read not merely when we study subjects that span across India, like capitalism, or the state. It has to be deciphered even in the smallest localities in the smallest pieces in which capitalism and state exist in their infinitesimal incarnations which are combined and composed to produce their large forms.

A Leftist Parable

Let me start with a leftist parable, an incident of leftist history that first pushed me to think of this as a problem. In the late 1940s, the Communist Party of India (CPI) decided, in retrospect somewhat unwisely, to declare that the political independence was false and, consequently, to challenge Jawaharlal Nehru’s new government through military insurgency. That policy went through two clear phases, separated by an interesting line of difference. The communists did not undertake any serious initiative without serious theoretical examination. After theoretically deciding that the Nehru government was a mere arm of not only the indigenous bourgeoisie, but also of Western imperialism, they examined the social forces on which their impending revolution was to be based. Initially, under the leadership of B T Ranadive, a trade union leader based in the great city of Bombay, they decided that the revolution was to follow the model of the Russian Revolution. The proletariat—based in urban and industrial centres like Bombay and Calcutta, and along the network of Indian Railways—were the most revolutionary class in Indian society, and a strike at the metropolitan cities would paralyse the country’s capitalist economy and install a revolutionary government. After nearly six months of strenuous efforts to trigger an urban revolution, catalysed by a massive railway strike, they accepted the failure of this policy and decided that, since India was primarily an agricultural society, they should pin their hopes on the radical peasantry as the leading revolutionary class, and attempt a Chinese-style revolutionary uprising based in the most remote agrarian regions. From their already existing base among the radicalised peasantry in Telangana in Andhra Pradesh, they wanted to launch a peasant revolution that would gradually encircle the cities, and capture national power.

If we examine their theoretical analyses of Indian society and its contradictions, we find a paradoxical picture. In effect, there was a strange disconnect between their theory of society, and the theory of revolution. Diversity was so overwhelmingly apparent to any observer that analysts of Indian society could not escape registering this undeniable fact. But, Marxists had an advantage in being armed with a “theory” of diversity, with some distinctive characteristics. Registration of evident difference in Indian thought tended to be a static spatial picture, reflected in the nationalist idea that had become commonplace: “unity in diversity.” This idea was also capable of internal complexities and it could simply mean the adjacency of diversities in the vast space of the subcontinent. But, the nationalist idea emerging from the late 18th century was more sophisticated.

I have suggested elsewhere that the reading of François Guizot’s history of the Western civilisation, which in turn drew upon Hegel, but provided it with a special twist, played a significant part in this enhancement. Guizot sought to reveal a logic of history running through the confusing diversity of events of European history. Bhudev Mukhopadhyay in Bengal strove in his sociological essays to discern a similar “logic of history” in the long Indian past. He saw a tendency of outsiders to approach India, drawn by her wealth and comfortable agricultural plenty, which periodically brought in elements discordant to the prior culture. But the “essence” of Indian history, its peculiar genius, was to find a new coherence which made these additional elements congruent again to a more complex and richer form of civilisational unity. Not just spatial diversity, this was a deeper “logic” of uniting the elements without suppressing their individual character.

The deeply idealistic historical tone of this narrative was not suited to the materially inclined communists. They seized upon a tougher economically inflected notion of uneven and combined development: Leon Trotsky’s concept drawn from Karl Marx’s earlier reflections on making sense of the confusing patterns of economic change in Europe under the impact of capitalist growth. The theory of “uneven and combined development” dispensed with two common presuppositions in thinking about diversity in temporal terms: linearity and uniformity. First, it stressed that the impulse of capitalist development transformed societies powerfully, but unevenly in spatial terms: some parts changed more rapidly than others, thus producing regional inequality where none existed before. But, it also emphasised that the picture of inequality at any point of time did not develop in a linear way by simply replicating and accentuating the states of danced and backward economic conditions in regions. There could be explosive spurts of growth in areas that were backward in an earlier stage. To use more colourful Hegelian tropes, even economic history could develop through its wrong side. Clearly, this constituted a remarkably sophisticated and forceful initial theory to start thinking about the question of regional diversity in colonial/postcolonial India.

Yet, in their actual theory of revolution on how to mobilise ordinary people’s discontent to produce deep social transformation, they threw away this powerful analytical tool. For the period that Ranadive’s group was in the leadership, they insisted that the proletarian revolution was “the correct line” for the Indian revolution. When that proved impossible, they were equally convinced that this line was wrong, and they went for a similar singular correct line modelled on the Chinese Revolution. In retrospect, the search for a single correct line contradicted the profound intuition in the idea of combined and uneven development. Even after this experience, the search for a correct line did not cease. The party continued to search for a singular correct line for the Indian revolution instead of recognising that if a political party sought justice and liberation for ordinary people in India, this would have meant justice against and liberation from quite dissimilar structures of social oppression in different parts of the country. The “correct line for West Bengal could not be the correct line for neighbouring Bihar.” A revolution in India must be a coalition of quite different revolutions in different regions.

This is a parable, but it reveals the difficulty and the significance of coming to grips with the fundamental question of regionality in thinking about India. For those like me who try to understand politics and the state, this difficulty is acute. An economist can add the per capita incomes of two adjoining states and divide that sum by two. We cannot follow a procedure that would even remotely be reasonable. I shall focus primarily on forms of political regionality and its connection to the making and unmaking of regions of other kinds—economic and cultural regions.

Dual Ontology of Space

There are three ways, I shall suggest, of looking at the question of regionality in India. These are three distinct analytical operations, which could be named generalisation, fragmentation and composition. Generalisation means gathering primary information from a determinate region, and assuming that all parts of India possess identical or similar characteristics. Conventionally, Bengali historians were generally, and in part justifiably, accused of such procedures like taking too seriously the idea that what Bengal thinks today India thinks tomorrow, assuming presumably that their descriptions of India, though not immediately accurate, will become true with a time lag.

Fragmentation means believing that, since the regions are so fundamentally real, nothing existing beyond the regional level has any serious, compelling historical reality. To turn this into a forceful slogan: there is no India, only its regions/India does not exist, only its regions/India is not real, only its regions.

As I do not find either of these ways of thinking persuasive, so I am forced into a position in thinking about this question, which I shall call composition and which does not deny the fundamental reality of the region, but, instead of denying the reality of India, it adds two further arguments. First, it asserts that regions are historical; they are formed historically, and thus get unformed or transformed in time. Second, the fact that these regions remain bound together within one single frame of some kind—political, economic or cultural—imparts some serious qualities to the regions and how they function. I shall offer an example from contemporary democratic politics. We shall return to these ways of conceptualising space and diversity later on.


Regionality is created, or regions are formed by three separate kinds of forces—political, economic and cultural—and regions are also consequently of three kinds. These three kinds of regionalities are produced by distinct kinds of logics, and their boundaries are maintained in different ways. One form of regionality also affects others, and the power of the state, in particular, affects the two other types of region formation. All three forces are communicative in the sense that they produce a certain kind of currency exchanged between individuals and groups, and the circulation of these objects—of political power, material goods and cultural artefacts—create zones of common experience and intelligibility, which have the historical effect of producing what we call regions. These three kinds of regionalities might overlap, for understandable reasons.

Ernest Gellner’s analysis of the formation of nationalism is an illustration of this kind of argument. States are sometimes produced by partially fortuitous circumstances—like victory in wars—but once they stabilise, they strive to create a form of cultural uniformity through state-driven processes of cultural production, like common educational policies, common syllabi, etc. Conversely, in some cases, if other historical circumstances produce a sense of cultural singularity among a group of people, this kind of cultural identity then becomes a basis for a demand for a separate state especially if fragments of such a nation are distributed among a cluster of states. The cases of Italy and Germany were so powerful imaginatively, precisely because they endorsed a fable of a pre-existing nation demanding and eventually realising a state of their own. Historically, the interaction and intersection between these three processes of region formation is a fascinating field of study.

We can illustrate this with the example of West Bengal. Bengal was a historically recognisable region in ancient India, designated by a regional name in Sanskrit, Banga; though it is important to retain a clear distinction between the principles of coherence in spatial entities found in premodern and modern history. Usually, in premodern cases, territorial entities were defined by some kind of centre, defined culturally or politically with lines radiating from that centre and weakening as they spread outwards; in modern times territories are typically bounded by spatially recognisable boundary lines. Thus, there were no modern-style linear boundaries between neighbouring regions like Anga, Kalinga, and Kamarupa, though their differences were marked by clear frontiers and while passing them one moved from one defined region to another. But, it is not clear what held these territorialities together in medieval times, before the rise of a definable regional vernacular. In later medieval times, for instance, a powerful cultural formation like the religious constellation of Gaudiya Vaishnavism emerged, which has an interesting spatial history. It developed from an undoubtedly Bengali centre in Navadvip, Chaitanya’s birthplace and Bengal’s erstwhile capital, but as it spread irresistibly in the next decades and centuries, it spread over a much larger spatial area covering the whole of present-day West Bengal and Bangladesh, Odisha, Manipur, and parts of Assam and Bihar, with a powerful satellite in distant Vrindavan. But, the interesting fact about this is the disconnection between the religious and linguistic identity. From the history of the Chaitanya movement, it is clear that by this time there was existence of a linguistic identity of the Bengali language, and Chaitanya himself and most of his major intellectual followers were Bengalis and composed many of their celebrated texts in Bengali. At the same time, the orbit of Chaitanya’s religion encompassed several linguistic regions, with followers from regions which spoke Bengali, Odia, Hindi, Manipuri, etc.

It is only with the rise of modern cultural institutions and processes like printing, standardisation of scripts, and the harnessing of linguistic identity to state processes, that this linguistic identity comes to have a political form and that Bengaliness becomes a politically mobilisable force. From the mid-19th century to the time of partition, Bengal developed a forceful cultural identity around its emerging literature, but the lines of religious division could be seen clearly in internal debates about the predominantly Hindu nature of this cultivation, and disputations among Muslim intelligentsia about how they could place themselves inside this cultural constellation. Such implicit divisions are mobilised periodically, initially by the British proposal for the partition of Bengal in 1905, and more successfully during the creation of Pakistan. After partition, a cultural identity linking the western and eastern parts of Bengal still continues. But, it is clear that under conditions of political modernity, states become the most causally dominant force of regionality. Despite a vestigial and weakening presence of a literary Bengali culture, the division of the linguistic community into two separate and conflicting states overrides this identity. The state acts as a container for economic processes, cutting the dependence of the jute-producing eastern part of the province on the western concentration of mills around Kolkata.

The fates of the pre-existing regionalities of Bengal and Punjab demonstrate the dominance of the processes of political region formation. Quite similarly, cultural processes have marked the literary spheres of Punjabi and Urdu. Political and economic boundaries shape, form and determine the nature of social experience in such a profound fashion that literary and cultural expression is easily subordinated to those boundaries. Literary expression in the two parts of Bengal continues to draw upon the prior artistic heritage, but the experience it gives voice to is decisively separated by the divergent experience of the two countries.

Premodern Regionality

As I study political power historically through the discipline of historical sociology, the long-term processes of premodern regionality is crucial for my understanding of power. A persistent difficulty of the study of modernity is its unworried one-sidedness. Modernists simply assume that modernity is a fundamental transformation of the social world, which I believe to be true, but inadequate for the real understanding of historical change. An understanding of change—in a literal sense—can never be successfully one-sided. However much we know the consequences of a transformation, if we do not know the character of the state of affairs from which the change occurred, we cannot say we understand the nature of that change. To state that Bengali literature changed after the late 18th century is a correct statement, but it is historically underdetermined. Not because Bengali literature did not change fundamentally, because it did. It is underdetermined because, if we do not understand what Bengali literature was really like before this change, we can have a true abstract picture that a change occurred, but not really a historically accurate picture of how it changed, because that, minimally, requires a two-sided picture of what the conditions were like before the change happened, and after.

It is in this sense that much of our historical picture of modernity and its change is cognitively unsatisfactory and inadequate. That is why a correct—even thumb line—picture of the premodern is not an embellishment, an optional addition, but a constituent part of our understanding of the modern. Of course, there are serious difficulties in this cognitive enterprise: historians of modernity cannot simultaneously become historians of the premodern. But, it is reasonable to demand a short-term historical picture or a stylised structural model of the premodern configuration for our understanding of modernity to be credible. I do not study political power beyond the late Mughal period, but this period is rich in chronicles and historical evidence, which, however, are not always easy to read historically. In the chronicles, it is easy to see these territorialising or spatialising processes clearly.

Our argument might become clearer if we artificially use three terms in a strict conceptual separation: territory, space and place. By “territory” we shall designate the primordial, most material form of existence, where it is mere extension without boundaries, and without conceptual markers of use. We shall use the term “space” to indicate some form of bounded territorial entity in a third-person language, that is, in the sense that most ordinary users of language will refer to that territorial configuration by the generally accepted common name. Awadh, or Bengal would be a marker like that, irrespective of whether it is politically or culturally seen as unified by some criterion or not. So, it would not be an incoherent statement to say that after the Vijayanagara empire fell, the Deccan remained disunited, because this states two things about that territory: that there is a commonly intelligible territory—space—that is generally known as the Deccan, or its vaguer Sanskrit equivalent, dakshinatya; and, additionally, that this space does not have any political form which covers all of it. Similar statements can be made of other spaces like Bengal, Kalinga, Awadh, Sindh, or even Hindustan or India. We can then reserve the term “place” for a more intense experiential sense of a place invested/throbbing with affect.

Several kinds of spatialities overlap in the Mughal domain. In the Islamic world, the Mughal empire is clearly seen as part of a cultural Persianate cosmopolis—with intense circulation of individuals and groups, especially with military, intellectual and cultural skills. The historical decline of the Persian empire, or sectarian hardening of the state immediately led to a steady flow of major figures of Persian letters into North India. But, despite this vast cosmopolitan constellation, which was culturally defined, the Mughal empire shows interesting politically defined subregional forms. In the last stages of the empire, several ethnic groups drawn from definable regions in the surrounding world—Mughals, Afghans, Turks—interact, at times as individual nobles and at times as clan groups who can depend on clan support in a world of insecurity. The empire works through a complex organisation of four identifiable northern regions—territories bordering on Iran, the central territory around Delhi and Agra, and two other parts of the empire which expand and contract depending on military fortunes—the eastern part stretching into Bengal, Assam and Odisha, and the southern empire in the Deccan where Aurangzeb spent his last years fighting Shivaji and other southern sultanates.

Every time the imperial centre is weakened through uncertainties of succession or ineffectual rule, these regions tend towards informal or formal separation into independent kingdoms. The eventual destruction of the Mughal empire occurs through an intensification of this process. Although at times of succession these distinct and more politically homogeneous and administratively manageable regions tended to develop pretensions of autonomy, rulers like Shah Jahan or Aurangzeb were able to restore the stability of the centre. After Aurangzeb, this process of reclamation of a revitalised centre failed precipitiously and, in the story of Mughal decline, we can see the revival of these long-term durable political regions submerged by the tides of imperial power. Some commentators would trace these regionalities back to the times of the Mahabharata, which also shows, if read through relentless political reduction, the same identifiable logic of durable regions and fragile empires. In the case of the Mughals, after the death of Aurangzeb, two regions decidedly drifted away through the formation of the Nawabi of Bengal through Alivardi Khan and Murshid Quli Khan, and the almost instant secession of the Deccan segment, the dominions of the Nizam-ul-Mulk.

Colonial Regionality

After the hiatus of half a century, British colonial power gradually restored the imperial centre, though its spatial formation was unusual, because the British were a maritime power. The authority of the imperial centre was restored after two sweeping movements from coastal bases in the east and the south. A firm establishment of British power came after the final defeat of the Marathas, the Sikhs and other serious contenders for political dominion cancelled the separation and fragmentation of political regions, and colonial rule brought in newer dynamics of region formation. From economic research on Mughal and British economies, it seems likely that before colonial rule, and the emergence of an entirely new kind of imperial economy, Indian society contained a range of common premodern economic formations. Since economic transactions were primarily local, a great variety of production economies existed on a local scale, which exchanged goods through various levels of trade. The colonial economy introduced an entirely new dimension to economic life by first linking parts of the Indian economy to the British imperial structure and indirectly to the world economy, which in a real modern sense developed in the 19th century.

Parts of the Indian economy, particularly in the coastal regions, reorientated their production responding to impulses from the English centre of the colonial empire. Internally, however, colonial power in the broader sense effected a fundamental transformation of economic structure, which has to work as the starting point for an analysis of regionalities that existed at the time of independence. Establishment of manufacturing industries in the colonial economy introduced a new kind of regional differentiation based on relationships of inequality and interdependence. Dependence of the coastal economies on their agricultural hinterland for raw materials and labour force created a new kind of unequal diversity with sharp inequalities of incomes, opportunities and collective regional power.

This kind of inequality between regions does not seem to have existed in premodern times, which were characterised by more equal forms of economic/productive diversity. Forms of regionality which had deep political consequences stemmed from this growth of regionally unequal industrialisation in the colonial period, though the “colonial” period, of course, varied a great deal between different parts of India. The entire region of northern India came to be dominated by a vast Bengali elite who had originated from western Bengal’s relatively narrow base, but established their sub-imperial dominance over an immense space from Shillong to Lahore. Subsequently, after independence, these regions saw intense dissatisfaction against Bengalis, sometimes explicitly in the form of regional movements.

The gradual retraction of this sub-imperialist Bengali middle class back into West Bengal partly contributed to the crisis of the West Bengal economy after independence. But, in a long historical view, it appears that premodern regionalities were not always entirely erased by modernity. At times, these were overwritten by new regional formations in all our three spheres—economy, politics and culture—probably because colonial power slowly brought in processes of the modern state, which alter the relation between states and their populations, and, most significantly, turn the boundaries of states into containers of all processes in ways that were not possible in earlier stages of history.

Postcolonial Regionality

Nehru’s thinking was deeply influenced by some aspects of Marxist economic analysis, and if we read the planning documents to search for its economic ideal, the legacy of colonial regionality looms very large. Regional inequality is a condition that is directly targeted by locating new heavy industries precisely in the regions that are regarded as the most backward. Development is equated rather uncritically with industrial and urban development. To exaggerate, every tree felled and every village swallowed by an expanding neighbouring city is a step in the direction of this industrial image of development.

It is not entirely surprising that Nehru’s first serious political challenge—which surprisingly questioned the electoral sovereignty of his Congress party—came from a historical effect of the regional inequalities produced in the colonial era. The first dramatic illustration of the force of this political sentiment generated by unequal economic growth was, of course, the agitation for Andhra Pradesh. It is clear that Nehru was caught by surprise by the occurrence of these protests and their intensity. The protests showed that there were deeper forces in the political world that could not be translated easily into the terms of electoral victory and defeat. Remarkably, this redrew the map not merely of the Madras state, but of the entire country. Interestingly, Nehru was averse to an early conversion to a linguistic reorganisation of Indian federalism, although this was a kind of implicit promise of the Congress mobilisation. By organising its segments along linguistic lines in the 1920s, the Congress delegitimised the administrative spatial distribution of British India. Behind his aversion lay an anxiety about accentuation of linguistic boundaries and differences in the peoples’ political imagination. Despite Nehru’s initial disinclination to concede linguistic states, he had the statesmanlike vision to realise that this demanded a general and not a local resolution; that unequal development was not a problem specific to Madras, and should be resolved nationally.

State-led Industrialisation

Reducing regional inequalities produced and exacerbated by colonial economic processes was one of the major objectives of Nehruvian planning. But, its successes were uneven, with large unintended consequences. The Nehru government used deliberate policy to locate large-scale industries in areas that were mainly agricultural—like siting giant public sector steel plants in Bihar, Orissa and Madhya Pradesh—with the expectation that these will benefit from the conventional spread effects of industrialisation. Some small-scale effects were certainly produced for the local economy, but, at least in the medium term, these industries did not have a dramatic effect on the region in which they were located or immediately surrounding it. Migration by the local poor to low-income occupations in metropolitan cities did not cease. On a smaller scale, it produced an effect quite similar to colonial industrialisation and professionalisation. Industrialisation in Bihar produced an enormous opportunity for employment of the large professional elites; for example, those with engineering degrees, in West Bengal who would otherwise have faced a severe contraction of their prospects. These industries also attracted highly skilled professionals from other regions as well, particularly from Tamil Nadu and the South which began to feel the effects of the initial wave of caste-based reservation in the 1950s.

Instead of producing an industry-fuelled development of a whole region that was given a new potential centre, this produced a crossflow of migration, of poorer labourers to West Bengal and of higher-income professionals to these centres. Location of single immense plants in the middle of agrarian hinterlands did not lead to the expected results.

Disjunction between Political and Economic Regions

After independence, formation of regions did not cease with the linguistic reorganisation of the federal structure in the mid-1950s. It appears that the container effect of the state is far more forceful at the level of the state itself rather than its internal federal constituents. Any trade or economic exchange between neighbouring nation states must go through the elaborate legal mechanism of international trade. Economic processes inside India are not similarly subject to a containerising effect of federal state boundaries. But, boundaries of economic regions do not always coincide with political state boundaries in two different ways. First, the growth of the capitalist economy did not respect state boundaries, and in some cases clearly gave rise to new economic regionalities. One of the most obvious instances of such transformation is the creation of a relatively industrialised zone in the western part of Uttar Pradesh (UP) in the 1970s, probably partly fuelled by agricultural prosperity associated with the green revolution wheat production.

The increasing distinctness of these regions created, by implication, another large primarily agrarian region consisting of eastern UP and Bihar. In subsequent decades, the development of industries in parts of Haryana adjacent to Delhi has in effect created a zone of stratified industrialisation that stretches from western UP to the eastern parts of Haryana across the capital region of Delhi, which, in its peripheries, now houses a vast sector of lower-level and low-technology industrial production. What is remarkable is that this effectively economically unified region—with a network of common interests of corporations, business groups and working people—has not demanded a translation of this economic regionality into a political form.

This is in sharp contrast to the relative success of two kinds of breakages of unified political units on the argument that there were internal economic enclaves that were consistently neglected by the states of which they were a part. Several cases have occurred where a section of a state has been carved out and established as a separate political unit, on the basis of ethnic identity. The breakup of Bombay into Maharashtra and Gujarat was primary on the grounds of linguistic regionality, and, subsequently, states like Mizoram and Arunachal Pradesh were created by using the same logic. But, a combination of economic and ethic demands have always been more effective. While the structure of the Indian federation acquired an appearance of permanence after the 1960s, successful secessions into new states became possible in the case of Jharkhand, Uttarakhand and Chhattisgarh on the grounds of economic backwardness. This is also true of the creation of Telangana, although by the time the separation happened it was ironical to treat the areas around Hyderabad as economically backward regions.

We must notice, however, a clear opposition between these cases of creation of new states—a kind of political acknowledgement of economic–cultural regionality—and the earlier example of the industrial region straddling the three states of Haryana, Delhi and western UP, which has not led to any political demand. This suggests that purely economic regionality might not be politically effective. These interests might be too indirect for political agents to discern, or this might show that economic commonality of interest by itself is not sufficient to work the creation of political regionality that requires an additional force of cultural or linguistic cement of some kind to turn this region in-itself into a region for-itself. We had seen that early colonial modernity introduced a new logic of producing regionalities through the disequalising character of industrial economic change. Regions which were even or equal before were turned newly unequal by the establishment of modern industries and, therefore, the forcible conversion of their agrarian surroundings into raw-material and labour-supplying hinterlands. It appears that the two major engines of economic change—the state and market—have continued to drive the force of disequalisation, with the market drawing unequal resources into some developmental nodes, and the state allowing the absorption and use of its revenues disproportionately on some regions, creating new centres and, correspondingly, new peripheries. This process primarily elicits a political response that demands the secession of the backwards areas to form their own states.

Liberalisation seems to have affected the template of regionalities in some unprecedented ways. Policies of liberalisation were preceded from the early 1980s in any case by a persistent alteration in the character of planning. Clearly, after the Third Five Year Plan, the central government abandoned its ambition of setting serious macroeconomic objectives for the entire economy. Planning turned increasingly into a statistical, stocktaking exercise after the fact. After liberalisation policies were adopted in 1991, the state actively withdrew from some areas of economic decision-making where it was dominantly effective earlier. As a consequence, state governments were allowed a new kind of policy freedom through which they would compete among themselves to attract foreign direct investment (FDI) and other kinds of investment by offering tax incentives and other forms of inducements by infrastructure development. Some states, recently freed of central supervision, performed more successfully—like Gujarat under Narendra Modi or Andhra Pradesh under Chandrababu Naidu. The cumulative effect of these developments can be a new accentuation of regional inequality, since at least private capital investment is likely to be drawn towards the new hubs.

It is possible to discern two speeds of development in different clusters of states, which are misleadingly averaged out for the entire country. But, the two speeds can be observed over two groups of states that are not regionally clustered. There is a clear division in the pace and nature of development between these two groups. But, the spaces are too large, and too internally diverse in other terms to view this as a dynamic producing new regionalities, and this clearly raises a question that is parallel to the classical Marxist debates about the two notions of class: class in itself and class for itself. In a different language, we can call these categories that are analytical and agentive. When analysts can see similarities or a unifying logic unifying a space, that yields an analytical purchase. But, when inhabitants of a particular space come to view themselves as not merely a collectivity with common interests, that self-identification becomes agentive. Analytical and agentive descriptions are quite different, simply because analytical descriptions need not be agentive. How real are these analytical and agentive categories of regions? It is rather pointless to debate which one of them is real, because they are both real in different ways.

Principles and Regionality

In concluding this discussion of the historical formation and unformation of regionalities, several principles need to be remembered. First, regions are formed by different kinds of causal processes, and these regions may not have overlapping boundaries. In understanding regions—even with the narrower focus of regional political economy—it is important to remember the region-forming powers of cultural and, particularly, political forces. To study the unnamed region of incipient industrialisation that straddles UP, Delhi and Haryana, we must understand that an economic region has a reality, with real effects. But it lacks a cultural self-description, and its economic processes are mediated by the bureaucratic decision-making structures of three state governments, and these too are subject to the accidental qualities of electoral politics. But, even economic exchange is helped by the fact that some version of Hindi can be understood across the entire region.

Second, many of these regions have long histories which continue to have effects in the present. Cultural and linguistic regions, in particular, have long histories that create long chains of memory. Economic regionalities have histories and memories of a different kind, usually starting with decisive events from the colonial period. Some economic initiatives of the state—like Nehru’s industrialisation plans—sought to counteract the legacies of these histories. But, in other cases, where economic change is primarily driven by markets, investment and growth tend to increase around areas that are already developed, because markets tend to take advantage of facilities that already exist rather than invest in new infrastructure. But, the future of region formation cannot be just an inert extrapolation from pre-existing or present trends.

Third, it is because of this kind of complexity that the only accurate way of thinking about the relation between these regions and the idea of a supervening India is not through generalisation or fragmentation, but by way of what I have called composition. However, we have to add another methodological dimension to this compositional analysis, which is the peculiar sense in which a real India exists.

Is There a Real India?

It is a legitimate question, then, to ask. Where is India in this picture? Clearly, we have two options of conceiving India if we give a prior assent to the reality of regions in this fundamental sense. Some analysts are misled by the legitimate emphasis on the reality of regions in this picture to conclude that, if the regions are real in this sense, it does not leave any room for India to exist, or to be real in any significant sense. At best, India is a kind of abstract container of these regions without any real characteristics which require either separate description or analysis. But, I do not agree with this argument, which I have called fragmentation. I want to propose a third kind of space, which we can only call India or Indian because it is impossible to seriously locate these features elsewhere. In that case, our second option, if we reject the idea of a mere container, is to think of a second-order or a second-level reality called India.

I shall offer this argument through a political illustration, a thought experiment. If we imagine that, for some reason, the individual states of the Indian union become all independent states, what would be their political character? I think it could be surmised safely that a major number of these states would succumb to irresistible pressures/forces of authoritarian rule. Even in real terms, it could be stated safely that the quality of democratic government was more palpable at the level of the central government rather than the states. Many states in India are dominated by either single parties that are organised around individual leaders, or social groups that dominated political life either through political parties or other organisations in a way that their domination could not be assailed. But, such political leaders who often in reality governed their states in distinctly undemocratic ways turned into stout defenders of a strictly democratic treatment when they dealt with the central government. In some historic instances, immensely powerful political leaders of various states found it in their interest to nominate a weak incumbent for the position of the Prime Minister rather than one among themselves, for fear that this individual would establish unassailable dominance.

This can be presented more abstractly in rational choice terms. If there are several roughly equal political agents who are totally dominant in their own regions, it would be rational for them as political actors, if they have to coexist within the framework of a common state structure, to demand that the structure should be entirely democratic. This does not require a change in their political disposition, but simply a lucid appreciation of what lies in their rational interest. This is what I indicate as second-order reality. This shows that the fact that these rulers are part of the Indian federal state structure has an overwhelming shaping influence on their decision-making and political behaviour, but on a plane that is quite distinct from the plane of state politics. The frame that is called India is extremely effective and influential, but not at the level of their actions in state politics. It would be false to think of India here as a mere container of these regions. The only adequate description of the reality of India would be to conceive of it as a separate second-order plane. It is in this sense that India is a second-order reality, but emphatically a crucial part of the structure of reality of our political world. Just as, if we ignore the reality of regions, we get a skewed and misleading picture of Indian politics, if we do not grasp this second-order reality of India, we get a picture that is skewed from the other direction.

India as a Second-order Space

Let me now turn to the original conceptual model I advanced about conceiving or picturing regionality, and restate my argument about the three levels of space or, more fashionably, the first, second and third space. The first stratum or plane must be a brute fact, material, unconceptualised space that is unmarked and nameless. I am tempted to use allegorically Walter Benjamin’s idea of an empty, homogeneous space. I must confess that I find the idea of an empty, homogeneous time of modernity, so admired by his followers, both striking and inscrutable. (It is not inscrutable in itself. Of course there are calculations which routinely use “empty homogenous” time; it is the assignation of modernity that makes it inscrutable. It is hard to see why modern people either would or necessary do use such a conception of time, while premodern ideas were imaginatively diverse.)

However, an empty, homogeneous conception of space is clearly conceivable and analytically useful. If I drive five miles through the insane traffic of Delhi, and drive five miles through the enchanting mountains of the Kumaons, in a sense, these are comparable only as empty homogeneous space. But, my experience of driving and several other experiential attributes of these two identical stretches of distance would be vastly divergent. The first space, or the first level of space must be territoriality in this sense; for our purposes, this space need not be topographically featureless, that is, empty, but as long as it is unmarked, unnamed and unbounded, that represents in a sense a zero degree of space. Conceptions of regions are a second-level conceptualisation of this space as something: as Ayodhya, or Awadh, or UP. The primary point is that any human activity with and relating to space will require some kind of conceptualisation of this kind, including characterisation as regions. But, if our earlier argument is correct, then we also usually have a third space, or a third level of space conceptualised as India. That level is the repository of the causal efficacy of specific processes, which cannot be accurately grasped by simply a serial addition of regional spaces.

That is why India is not an illusion, or a fantasy, but a reality of a second-order. What happens there is a distinctive order of happenings. These cannot be reduced to what happens in the regions.

Updated On : 17th Nov, 2017


(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top