Version classiqueVersion mobile

Essays on population and space in India

Christophe Z. Guilmoto
Alain Vaguet

Part III. Features of Indian urbanization

7. Indian City, Hindu City ?

Factors and Processes of Spatial Segregation

Odette Louiset-Vaguet

Texte intégral

  • 1 Visaria and Visaria (1995).

1In the world list of achievements in the matter of urbanization, India occupies a prominent position, both in terms of number (240 million in 1995),1 and regarding the existence of cities which are counted among the largest in the world. To this incontestable importance of the urban phenomenon is often contrasted a moderate rate of urbanization : almost two out of every three Indians still live in rural zones. Far from opposing each other, town and countryside form a continuum, not only because of improved means of communication, but also because of the spreading of urbanized zones. Diffuse and physical urbanization proceeds, but this weakens neither the capacity of the countryside to retain the country dwellers, nor the scarcely urbanizing tendencies of a caste society which seems to have a phobia about the unknown. Above all, these two spaces recognize the same values proper to Hinduism and, because of this, present analogies in their spatial organization.

2The Indian urbanitas is disconcerting, exhibiting the invariants of the universally urban, as well as of Indian sociocultural specificity. This combination exhorts us to derive benefit from the parallels established with other urban territories and forms of evolution, whether in the developing or in the industrialized countries. The examination of urban forms and issues is in no way advanced by a compartmentalized analysis.

  • 2 Godard (1994).

3However, it is at the same time known that the results depend on the methods of approach. If at the outset the Indian city seems to sustain an inductive approach which accords primacy in particular to the large scale, the deductive method makes it possible to draw back and place the Indian city in the framework of the large urban types and processes : a slow urbanization notwithstanding its extreme characteristics (Calcutta), many sided towns struck by crisis, as in all countries of the world, but a specifically communal internal structuring. It therefore appears to us to be essential to follow a comparative method in order to distinguish the “universal tendencies of processes proper to local cultures” to “make good tools of knowledge and avoid hasty transfers of models of urban management from one region to another.”2 It is from this perspective that we propose to contribute by way of a few examples to the debate on the Indian city.

The Indian urban world contained in the universal urban

Indian urbanization on statistical trial

  • 3 The Census of India divides towns into 6 classes on the basis of the number of inhabitants. Class I (...)

4India would seem to be impervious to a statistical analysis of the urban phenomenon ; while the relative figures place India in the bottom group, absolute values put it at the head of the list. At a rate of urbanization of 26 per cent, the urban population of the country is more numerous than the total population of the overwhelming majority of the world’s countries. By the same token, the growth rate of the urban population establishes no records whatsoever among the developing countries : the cities with more than 100,000 inhabitants3 experiencing the most sustained growth recorded an annual rate of 3.09 per cent during the 1980s. Rates twice as high are currently to be observed in tropical Africa ; and, above all, Indian urban growth appears to have fallen off in comparison to the preceding decades. However, this was expressed in an intercensal growth (1981-1991) of the 300 cities in Class I of 44.2 million persons, which represents a greater population than that of the 107 cities in India with more than 100,000 inhabitants, in 1961. All classes taken together, the cities in the country counted at the last census, in 1991, 72 million city dwellers more than in 1981. Thus, in the Indian context, relatively moderate rates represent a genuine challenge in terms of housing, education, employment, etc. In India, and no doubt more than elsewhere, the quantitative alone smoothes over the realities.

  • 4 At least 5,000 inhabitants, a minimum density of 400 inhabitants/, at least 75 per cent of wo (...)

5The problem of adequate definitions/demarcations arises at every turn. The criteria retained by the Census of India are common,4 but induce notions of threshold. What is more, the passage of an agglomeration from one class to another does not always obey the criteria cited. Thus, the admission of a city to the category of agglomerations with several million inhabitants leads to greater investments by the state. This explains in part why the urban limits at times assiduously (and mysteriously) skirt around certain villages (which are, in fact, urbanized) to avoid the fateful threshold of a million inhabitants. Nevertheless, once the pseudo-count has been published, it is used by all without compunction.

6A comparison of two Asian demographic giants confirms the inadequacy of the statistical approach. Indian and Chinese data, both relative and absolute, present similarities, whereas the situation of urban spread, the evolution of cities, their place in the economy and their relation to rural zones, differ profoundly. Detailed or thematic approaches must be made in order to discern more clearly the facets of Indian cities, which are both many-sided and deeply marked by their specifically Indian features.

Million-plus cities, “stars ” of growth

  • 5 Moriconi-Ebrard (1993), p. 91.

7The presence of numerous large cities constitutes a further element in the Indian paradox : approximately one third of the city dwellers lived in 30 cities of over a million inhabitants in 1995 (Visaria and Visaria, 1995). There were only 13 such cities in 1981 and the 1991 Census enumerated four metropolises exceeding 5 million inhabitants. The Indian cities with populations of millions assert themselves as the “stars” of growth. This phenomenon obeys the classic schema of developing countries, but also the world-wide tendency during the years 1950-1990 : in the poor countries, secondary and tertiary activities were concentrated in a limited number of cities situated on the upper rungs of the urban hierarchy ; on the global level, there is a veritable “inflation” in the size of cities, at the same time as a generalization of the phenomenon.5

8The paradox of these great agglomerations stems from their essential role in urban growth, while the potential of the countryside to retain its population and the decline in the migratory flux to the cities are confirmed. Large cities, whose very contrasting image sustains the Indian subconscious traversed by Gandhian ideas : they are at the same time isles of light, key economic centres, and places where dangers loom, very often embodied by Calcutta and its tutelary goddess, Kali. With Calcutta, the theme of the dying city is privileged.

9Million-plus cities “are capable” of the best, as of the worst. They constitute favourable places of exchange and innovation, but much greater demands are made of them than they can fulfil, particularly in terms of employment and housing. On the other hand, they have the highest literacy rates and the lowest child mortality rates. These favourable indicators are certainly influenced by the presence of a prosperous elite, but the most disadvantaged nevertheless benefit there from better conditions than in the villages, even if they share among themselves only the “crumbs.”

  • 6 Regarding this point, we do not entirely share the point of view of F. Moriconi-Ebrard when he writ (...)

10Officially, the very large city does not have full approval and the Indian state encourages notably the growth of small and medium-sized towns, which, however, have many difficulties in gaining ground when they are in competition with the large centres. Urban macrocephaly is not absent from the Indian Union and the country suffers from an unsound equilibrium.6 In effect, macrocephaly is experienced on the state level. Very often, the capital concentrates economic, political and cultural powers and reaps both the beneficial effects of growth, and the price to be paid for it.

Cities resembling other cities in the Third World

11As in other countries of the Third World, the growth of cities accompanies a veritable crisis, which affects above all the most destitute, for whom the “right to the city” is a daily struggle. The shortage of housing is the most visible syndrome, expressed in particular by the existence of slums which shelter between one third and 40 per cent of the Indian city dwellers. Subsidized housing, only little developed because of the lack of means, is in any case hardly allocated to the most impoverished, who then opt for the illegal occupation of urban lands and construct their own dwellings. Nevertheless, the lack of housing does not affect alone the most destitute. Private companies primarily build lucrative upper-class residences for the most affluent. Access to urban land and land speculation constitute one of the most acute problems, and the laws voted in the mid-1970s to reduce inequalities were often circumvented, not to forget the perverse effects which further accentuated speculative tendencies. If the high cost of housing represents a dissuasive factor in terms of integration, confining the villagers to rural zones near large centres, it also constrains the poor to envisage “informal” strategies and contributes to the elaboration of an entire system (occupation, construction, settlement) on the margins of the laws and planning (O. Vaguet, 1986). Of course, the questions of housing and the unequal access of city dwellers to land are only the visible portion of the iceberg of under-development and socio-economic disparities. Unemployment, under-employment, illiteracy, precarious diet and health serve to complete the picture of a city termed in crisis. Lacking a sufficiently developed industrial sector, the tertiary sector is overcrowded and primarily composed of the informal sector, the relations of which to the formal sector are close, but which offers its workers only meagre wages and no security whatsoever.

12From the legal to the illegal town—without strict borders—, there are a great many possible declensions. The Indian city presents a countenance characteristic of cities in the developing countries : incompleteness, unattractive, and ill-assorted buildings. Intra-urban contrasts are pronounced, between the central business district (CBD) exhibiting an American-style urbanism and the historical centres which are heavily populated and often slum-ridden. Traces of colonization remain and are inscribed in an often bi-of polynuclear morphology gradually complicated over the course of time (Bourcier et al., 1996). Intra-urban vacant spaces become scarce and new urban peripheral forms emerge, illustrating individual strategies or policies. Indian cities are becoming dense and spreading. In the image of what is to be observed in Europe, the agglomeration strengthens on the small scale by extending its territory ; on the large scale, the urban appears to be diluted in the rural environment which it transforms, and its limits become more blurred. Peri-urbanization is going on and causes profound changes, in the nearby rural zone as well as in the practice and management of space by city dwellers.

13Until now, questions very similar to those encountered elsewhere stand out. However, the city does not constitute an autonomous space, but is an element of the society which it accommodates, and which accommodates it. The specificity of India is the role played by caste in the structuring and functioning of society (Dumont, 1966). The socio-religious hierarchy proper to Hinduism does not constitute a foundational element of urbanitas in India, but rather declines Indian urbanitas. Louis Dumont, whom one reproached for having ascribed too great an importance to caste in his works on India, responded that caste was neither more nor less central, nothing more (Deliege, 1993).

Indianness in the city (caste and urban territoriality)

Caste, foundation of spatial segregation ?

  • 7 The information concerning these two types of quarters were collected by us in the course of househ (...)
  • 8 Slums : under-integrated, and by extension, poor quarters, shantytowns. Those mentioned here are sl (...)

14Urbanization is generally presented as going together with “social disorganization”, that is, with a weakening, or indeed disappearance, of the organizing principles of traditional societies. This conception inherited from the European 19th century tends to establish a correlation between massive urbanization and industrialization, these two processes inexorably entailing a change in social behaviour to culminate in the “stereotyped pattern of a uniform urban personality and that of a linear evolution” (Roncayolo, 1990). Thus, Bettelheim (1962) wrote that social distance was the most reduced in the large Indian cities, and one perceives the author’s conviction that there is an obligatory passage from traditional caste society, which would first take place in these agglomerations as a consequence of the presence of the industrial sector and the new human relationships thereby engendered. This question appears to be central in studies bearing on the actual socio-spatial configuration of Indian cities. Does caste remain the segregating and organizing principle, or is it demolished by new principles of the social division of space ? To respond to this question, we shall call to mind the organizing principles of the traditional city, and then give examples of urban quarters which reveal new tendencies : the housing developments of the middle and upper classes7 on the one hand, and slums8 on the other.

15The anarchic “aggregation” of populations in the Indian city is often contrasted with “the organization of the colonial city (Conté, 1996)." reality, the configuration of the Indian city, and here more precisely of the Hindu city, obeys very strict principles, whether in the determination of a site, town-planning norms, or socio-spatial organization.

  • 9 Sekhar (1985), pp. 80-100.

16The Vastu Vidya, or “secret science of architecture”, elaborated during the Vedic period, is anchored in brahmanic ideology. The latter is projected in space with its concern for order and preservation of the degree of purity of the “upper” communities. The Vastu Vidya, over a long period transmitted by the architects-priests from father to son, only became the subject of written treatises during the medieval period (AD 750-1200). Town, village, temple, all obey a single principle : their configuration is an image of the cosmos. The latter is symbolized by the mandala, a magic diagram composed of a circle encompassed by a square. Every structure must uphold the cosmic Order and “represent a renewed victory over the danger of dislocation” (Volwahsen, 1968). Simply stated, the town is the image of the cosmos, architecture and urbanism are closely linked to the sacred. Nevertheless, the town is not only that which contains, it is also “contained”, and the principles of the Hindu religion concern the functioning of society by structuring it in hierarchical castes. In the matter of spatial organization, the caste system prescribes segregation due to Hinduism’s taboos on contact. The Vastu Vidya also indicates the position of caste respective of the cardinal points, the orientation of openings, the form of the structure and the number of storeys authorized. The priests decide upon the place where each caste is settled, a location also determined in respect to the other communities (Doshi, 1994). To the highest caste falls the immediate proximity to the centre, which is defined by the temple in temple-towns or villages, and by the palace of the king or raja in political capitals. Thus, the Brahmins live in the agraharam, the quarter reserved for them very near the place of worship. From this point of view, the temple-towns in south India present the plan most nearly approaching the mandala. Their urban structure spreads out in concentric rings around the cella representing the Brahman (absolute being) around which the cosmos is organized. The sacred character of the place diminishes as the distance from the centre increases. The first Vedic towns are initially presented as mono-caste, then as reserved for a few groups occupying political, administrative, commercial or religious functions, who alone lived in the fort (some sort of intramural town). Every morning, the “little people” waited at the city gates to enter and work for the “important society.”9

Figure 7.1 : Localization of scheduled castes in Amaravathi (Andhra Pradesh)

Figure 7.1 : Localization of scheduled castes in Amaravathi (Andhra Pradesh)

17This model of urban residential segregation was nevertheless influenced from outside. Non-Hindu sovereignties created or besieged towns. Then, beginning in the 13th century, the Muslims strengthened the administrative, political and military function of towns, where they established themselves and implemented a division of space based on an ethnic criterion : original Muslims (as opposed to Indian converts), and more particularly Arabs, occupied the top of the hierarchy. Despite everything, the traditional principles were still applied, as witnesses Jaipur, built in the 18th century according to the norms of Hindu urbanism, and more specifically the spatial distribution of castes in conformity with the model dictated by the Vastu Vidya (Doshi, 1994). Beginning in the 18th century, British presence modified not only the localization of towns by favouring coastal settlements, but it also instituted a new ethnic segregation by isolating the European town from the indigenous town ; this differentiation is still perceptible today. The new quarters built during the colonial period respond to foreign social and town-planning norms. Thus, the urban model has been from the outset progressively modified.

  • 10 This correlation was noted, and reflects what Bettelheim (1962) designates as the historical advant (...)

18Since that period, caste seems to have lost ground to the advantage of socio-economic determinants. The multi-caste composition of recent quarters seems to be the unanswerable argument of those who believe in a “normalization” of Indian society. However, does the apparent easing up of segregation prescriptions mean that caste has become inoperative in the distribution of socio-religious groups in space ? The examples presented here bear witness to the complexity of the phenomenon, on the one hand as a result of official choices by Indian democracy, which abolished caste (1950 Constitution), and, on the other, inherited relations between traditional ritual status and socio-economic status.10

Figure 7.2 : Localization of agricultural castes (Naidus and Yadawas) in Amaravathi

Figure 7.2 : Localization of agricultural castes (Naidus and Yadawas) in Amaravathi

19A recent study conducted in a small town of 10,000 inhabitants in Andhra Pradesh (Hamon, 1997) attempts to identify the socio-religious changes at work, in both professional and spatial domains. The appearance of new professions does not confine the members of a caste to the exercise of a traditional occupation. Thus, the Vadderas (stone breakers, unskilled road workers) in the town of Amaravathi are divided into two groups : those who continue their traditional occupations, and those who have cycle-rickshaws. It is here seen that, although the collective strategy of incursion into a new profession indeed constitutes a change, this cannot be construed as a transformation. The social promotion of Dalits assumes great importance in this context and the integration of some of them in public office indicates the distance covered, a noteworthy distance when the ritual status of untouchability is considered.

Figure 7.3 : Localization of upper castes in Amaravathi

Figure 7.3 : Localization of upper castes in Amaravathi

20Nevertheless, few among the scheduled castes are those who benefit from such social advancement resulting of a government quota policy. Moreover, the new professions requiring the highest qualifications most often fall to the upper castes. Finally, some occupations resist urbanization, as they are not justified by “modern” economic logic. This is the case for activities closely associated in terms of taboos respected by Hindus, who leave to cobblers, washermen, barbers, etc., the “charge” of being in contact with “polluting materials.” From the strictly spatial point of view (Figures 7.1-7.3), and more precisely from the perspective of segregation, taken as a whole the families always make up homogeneous blocks according to caste and, above all, the Dalits are separated from the rest of the town by a space left unoccupied, the “playground”, a boundary which few Dalits succeed in crossing. Those who do have recently arrived in Amaravathi as transferred government employees. This signifies that occupational mobility can accompany or be accompanied by residential mobility.

21Finally, two facts are prominent in understanding the great complexity of the situation. In 1987, a new caste quarter reserved for the Naidus was created. In effect, the municipal authorities decided to rehouse the community, although its members are far from counting among the poorest in the town. They represent one quarter of the electorate and are easily mobilized by the municipal power in the hands of another peasant caste. Invalidating the preceding event, in 1989, the “Lecturer Colony” housing estate was built, a quarter intended to accommodate teachers from the nearby university and which is, therefore, likely to be multi-caste. In reality, this neighbourhood has very few residents from either extreme of the hierarchy.

Figure 7.4 : Spatial distribution of castes in the Tatabad quarter (Coimbatore, Tamil Nadu)

Figure 7.4 : Spatial distribution of castes in the Tatabad quarter (Coimbatore, Tamil Nadu)

22This study demonstrates the complex web of factors explaining the socio-spatial organization of urban space. Striking, however, is the infrequency of individual mobility. Most often, cases of collective mobility are witnessed, whether in the professional or residential spheres. This seems to confirm the impact of caste as well as its overlapping with socio-economic factors which, moreover, have certainly never been absent from Indian society. Change takes place, but often appears to be “contained” by fundamental principles of Hindu civilization.

The middle classes in the caste town

23In the context of growing pressure on urban land and speculation, urbanized areas progress with the production of new constructed spaces. The middle and upper castes are the producers of part of these new spaces. In Coimbatore, a city with over a million inhabitants and ranked second behind Chennai in the urban hierarchy of Tamil Nadu, the peri-urban countryside is particularly shaped by these categories, the numerical importance of which is increasing, in a context of economic growth and, in particular, of industrial diversification (Vaguet and Vaguet, 1992).

24For the middle categories and the most affluent, central spaces represent inconveniences. They prove to be inaccessible to the former because of the high cost, notably lor whoever would want to own his accommodation. Households consider the acquisition of accommodation to be a priority, a duty vis-à-vis the children to whom the inheritance will eventually fall. For the latter, the more well-to-do, centrality becomes synonymous with congestion and pollution, thus with inconvenience, while the new quarters are vaunted for the quality of life they would offer. Distance or, more exactly, settling in a peri-urban zone, does little to the isolate families who avail of a means of transport, whether scooter, motorcycle or automobile (private or “company car”).

25The peripheral housing developments are very diverse ; they are distinguished by the size of the plots, the quality of the habitat and the attention given to the landscape (more or less wooded). The location is also of importance ; land prices vary, of course, according to this criterion.

  • 11 Family surveys conducted by us in Coimbatore, January 1992, in fifteen middle- and upper-class quar (...)
  • 12 The sociologist Grafmeyer (1994) underscores the crucial character of “the question of scale of obs (...)
  • 13 As opposed to 15 to 20 per cent of the Indian population (Pouchepadass, 1984).

26The great majority of quarters visited11 are multi-caste and, in this, differ from differentiated in quarters of traditional petty bourgeoisie in the old parts of the city, which appear as juxtaposed mono-caste cores. Nevertheless, this “mix” of recent quarters must be nuanced, notably at a more detailed level of observation.12 In effect, the families (a majority of whom are Hindus) counted more than one third Brahmins, one fourth peasant castes (Naidus and Gounders, dominant and rival communities in Coimbatore) and 10 per cent Kshatryas and Vaishyas. Taken as a whole, therefore, 45 per cent of the families are “twice-born”13 and the remaining 55 per cent are divided among agricultural and service castes of “respectable” or medium level. Scheduled castes and tribes are nearly absent. A monographic study carried out in the same agglomeration (Eliot, 1993, see Figure 7.4) underscored the dominance of upper castes in a residential estate developed by the Tata firm in the 1970s : nearly 40 per cent of the Hindus are Brahmins, while the other castes rank in the middle of the hierarchy. In the old part of the city, in Sukkrawarpet, the findings are similar, but for the presence of two communities traditionally found in this area (jewellers and weavers), in addition to the Naidus and Brahmins. This over-representation of categories of high status can be seen as a indication of socio-spatial segregation.

27A tendency was noted in all neighbourhoods to bring together castes in blocks or parts of a street such that the immediate neighbours belong to the same community. This formation of mono-caste cores reveals the combination of two systems of values, with the overlapping of the ritual hierarchy and that of income levels. The Brahmins remain particularly sensitive to the question of contact with other castes. The frequency of classified advertisements is newspapers of the type “flat to let to strict vegetarian” bears witness to the demarcation of a level of “purity”, beneath which one is not prepared to accept living together, This goes beyond a mere overlap of caste and class likely to disappear at the end of an “economic recovery” staged by the lower castes.

Figure 7.5 : Distribution of households in the N’Kunta slum in Hyderabad (Andhra Pradesh)

Figure 7.5 : Distribution of households in the N’Kunta slum in Hyderabad (Andhra Pradesh)

The slum, territory of untouchability

28The slum has the outer appearance of an impoverished hamlet, isolated from the main village. This isolation traditionally affects Dalits and tribals because of the already noted taboos. Based on household surveys conducted in different slums in Hyderabad, the socio-economic and socio-religious profile of these slums was drawn up. Other than the precariousness and poverty, which are of no surprise, the characteristic of these quarters is the clear over-representation of Dalits. In one of the slums, N. Kunta, nearly two thirds of the Hindus (who are here in a large majority) belong to the scheduled castes (see Figures 7.5 and 7.6). The other families consist of either scheduled tribes or service castes who have one point in common : their particularly low status in the Hindu hierarchy (launderers, barbers, cobblers, fisherfolk, etc.). Figure 7.7 illustrates relative grouping according to jati (sub-caste) : barbers in the north-east, cobblers in the east, tailors in the north-west. To an area in the south dominated by Dalits (Malas and Madigas) is opposed in the north a more composite area.

Figure 7.6 : Distribution of scheduled caste households, N’Kunta, Hyderabad

Figure 7.6 : Distribution of scheduled caste households, N’Kunta, Hyderabad
  • 14 Dalits account for 16 per cent of the total population.

29The situation in the other slums visited is comparable. Municipal departments undertook a survey in the 1980s, and estimated that three quarters of the inhabitants of shantytowns in Hyderabad were Dalits, as opposed to 9.5 per cent of the city dwellers (Census of India, 1981).14 Like the Dalits’ hamlet in the village, the slum is not a crossing point but, on the contrary, a place to be avoided.

30Outside the slum, as within, caste remains a spatial reality. Communities have a tendency to group together and to form juxtaposed homogeneous blocks. The multi-caste character of the shantytown exhibits but a weak evolution ; the castes that tolerate the proximity of the scheduled castes themselves belong to lower-status groups. Their own impurity perhaps makes them less susceptible to the risk of pollution. An almost “horizontal” grouping of castes can be seen, wherein the originality and identity proper to the castes are preserved, but the common denominator is the low socio-religious status. In Coimbatore, the sociologist C. Sekhar (1985) notes the same tendency : the relegation of Dalits to 6 (out of 30) municipal wards in which 60 per cent of them are concentrated, while they comprise less than 8 per cent of the city’s inhabitants.

Figure 7.7 : Distribution of service castes, N’kunta, Hyderabad

Figure 7.7 : Distribution of service castes, N’kunta, Hyderabad

31The innovation of the city would reside in the fact that it generates some amount of inter-caste proximity, if not promiscuousness, at a level not found in the village, as if urbanization would have entailed a mitigation of the segregation process, but within well-defined limits. For the majority of castes, the impurity syndrome is still very much alive. In effect, if launderers and scheduled castes are neighbours in a slum, members of upper castes who find themselves in the greatest destitution refrain from occupying a hut there. The Indian shantytown is not only a synonym of poverty, but is also synonymous with impurity, all the more so as the under-integrated quarters occupy urban zones abandoned by the conventional settlement, or peripheral spaces deprived of amenities. Thus, the residents of shantytowns accept to live in a malodorous or dangerous environment, which is, in the eyes of middle and upper castes, outside the norm.

32The mechanisms involved in the integration/marginalization of shantytowns in the Indian metropolis are complex. The under-integrated quarter, here as elsewhere, does not elude a negative moral image in which poverty and delinquency or deviance are associated. The slum is viewed as a core of social pathology, not without similarity to the working classes/dangerous classes of the European 19th century and, more recently, to the low-income housing colonies in French suburbs. Of great interest, but eminently complex from the perspective of the Western researcher, caste, which institutionalizes inequality in Indian society, seems to play a contradictory role. The slums occupy a place equivalent to that of the untouchables’ hamlets, and find a place in the city, as the lower castes find their place in society ; and, at the same time, the image of these quarters is made negative by their ritual status, adding to the conspicuous poverty the suspicion of social pathology.


33It is because the confrontation of the universal and the particular does not appear to us to be a contradictory process, but is rather seen to enhance our considerations, that we focus on urban phenomena and processes observable throughout the world, as well as on the expression of the profound originality of the Indian world. These two aspects, far from existing separately, converge to produce the Indian city.

34The major question concerns the role played today by caste. It is not exclusively anthropological in nature, for castes are also spatialized social formations. Caste appears to continue to intervene in the composition of neighbourhoods, although it only partially intersects the actual socio-economic order, and although the logic of Indian cities differs but little from that of other agglomerations in the world. The given examples are far from providing a categorical response, and it would be as wrong to think that caste is no longer actively at work in terms of spatial segregation, as it would be wrong to assert that it remains immune to all outside influence.

35Multi-caste areas are often perceived as a sign of the system’s weakening. However, the over-representation of upper and middle castes in the well-to-do housing estates and the over-representation of lower castes in the slums confirm the survival of segregation according to socio-religious criteria. The neighbourhoods and the process of drawing closer together observed in recently established quarters certainly confers on them a greater heterogeneity than that of the former cores, but this proximity is contained and appears between communities of comparable or compatible levels.

36The relay of caste by “neo-castes”, as suggested by G. Heuzé (1982, p. 204), enables the continuance of Hinduism’s fundamental taboos : the economically upgraded Dalits but rarely lives beside upper castes ; by the same token, the poor person belonging to a ritually high status avoids at any price the proximity of “polluting” castes in disadvantaged quarters. These observations are confirmed by matrimonial behaviour, which is still in the majority of cases dictated by socio-religious prescriptions.


1 Visaria and Visaria (1995).

2 Godard (1994).

3 The Census of India divides towns into 6 classes on the basis of the number of inhabitants. Class I includes “large towns”, that is, those having 100,000 or more inhabitants.

4 At least 5,000 inhabitants, a minimum density of 400 inhabitants/, at least 75 per cent of working males engaged in non-agricultural activities, the existence of specifically urban characteristics.” See Bose (1980)

5 Moriconi-Ebrard (1993), p. 91.

6 Regarding this point, we do not entirely share the point of view of F. Moriconi-Ebrard when he writes that India is exempt from urban macrocephaly.

7 The information concerning these two types of quarters were collected by us in the course of household surveys. The middle- and upper-class quarters are located in Coimbatore, (Tamil Nadu). The notion of middle and upper-classes would merit a lengthy exposition. Very succinctly, one witnesses an inflation of intermediate categories between the poor and the affluent. The plural appears to us to be appropriate because, as much from the point of view of the standard of living, as from the position in professional life, this notion refers to very diverse situations. In our opinion, the middles classes are distinguished from the poor by their capacity to consume, going beyond (although to very different degrees) “subsistence economy" and, above all. by their capacity to project themselves into the future, in this instance, to invest in the future of their children. In India, one currently speaks of the lower middle, middle and upper classes.

8 Slums : under-integrated, and by extension, poor quarters, shantytowns. Those mentioned here are slums in the capital of Andhra Pradesh, Hyderabad.

9 Sekhar (1985), pp. 80-100.

10 This correlation was noted, and reflects what Bettelheim (1962) designates as the historical advantage of the upper castes and the for a long time exclusive access ol Brahmins to knowledge ; the concept of dominant castes advanced by Srinivas puts however this correlation into perspective (Deliège, 1993). Some groups hold economic and political power without a similar position in the ritual hierarchy.

11 Family surveys conducted by us in Coimbatore, January 1992, in fifteen middle- and upper-class quarters.

12 The sociologist Grafmeyer (1994) underscores the crucial character of “the question of scale of observation which could appear to be of a purely technical order, (but) in fact involves that of the very meaning which one accords to the idea of segregation”.

13 As opposed to 15 to 20 per cent of the Indian population (Pouchepadass, 1984).

14 Dalits account for 16 per cent of the total population.

Table des illustrations

Titre Figure 7.1 : Localization of scheduled castes in Amaravathi (Andhra Pradesh)
Fichier image/jpeg, 59k
Titre Figure 7.2 : Localization of agricultural castes (Naidus and Yadawas) in Amaravathi
Fichier image/jpeg, 62k
Titre Figure 7.3 : Localization of upper castes in Amaravathi
Fichier image/jpeg, 53k
Titre Figure 7.4 : Spatial distribution of castes in the Tatabad quarter (Coimbatore, Tamil Nadu)
Fichier image/jpeg, 62k
Titre Figure 7.5 : Distribution of households in the N’Kunta slum in Hyderabad (Andhra Pradesh)
Fichier image/jpeg, 58k
Titre Figure 7.6 : Distribution of scheduled caste households, N’Kunta, Hyderabad
Fichier image/jpeg, 60k
Titre Figure 7.7 : Distribution of service castes, N’kunta, Hyderabad
Fichier image/jpeg, 72k


Universite de Rouen, France.

© Institut Français de Pondichéry, 2000

Conditions d’utilisation :

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search