The power of space in a traditional Hindu city

Robert l. Levy

Each society offers up its own peculiar space...as an object for analysis and overall theoretical explication. ~Heuri Lefebvre

INTRODUCTION: NATURALIZED AND SACRED SPACE

In the early 1970s I began anthropological studies in Bhaktapur, a city of forty thousand people in Nepal's Kathmandu Valley. Bhaktapur was identified by its inhabitants as a 'Hindu city' and had been shaped from its founding in the twelfth century as the center of a royal city-state by ~astric and Brahmao.ical ideals for such a royal city. The material, social, and cultural structures that resulted shaped Bhaktapur's life and people for many centuries and still dominated the life of the city at the time of my study.l I will focus here on one aspect of Bhaktapur, its 'sacralized' urban space, and by noting some contrasts of the city's mode of organization with some quite different communities try to suggest some of the 'peculiar' features of that space and of the Hindu civic order into which it is woven. I will, necessarily, make use of some materials and arguments I have published elsewhere (Levy 1990). I approached my study of Bhaktapur not as an Indologist, but as an anthropologist, and I perceived and questioned its Hindu order in some traditional anthropological ways. How did that order seem to 'work' for (and against) the community and its inhabitants? How does the community appear when it is compared and contrasted with other k/nds of communities? I had just completed studies in Piri, a small, comparatively isolated Polynesian village (Levy 1973). In comparison with Piri, what was greatly striking to me about Bhaktapur was

International Journal of Hindu Studies 1, 1 (April 1997): 55-71. © 1997 by the World Heritage Press Inc.

56 / Robert I. Levy the great number of its elaborate and dramatic 'religious' material forms and actions, prominent aspects of the city's urban life. The striking contrast with low-key, undramatic, and workaday Piri made this density of 'symbols' noteworthy--and problematic. Piri's life certainly was full of a quiet sort of 'symbolism'--having one's house placed in a certain place in the village was a 'symbol' of one's relative status; the way men and women worked, stood, moved, and gazed were 'symbols' of the village construction of gender. But most of Piri's 'symbolism' was part of, in a sense hidden in, the ordinariness of everyday life, leading villagers to imagine that locally constructed forms were, in fact, 'the natural way that things are.' Locally constructed meanings were embedded in, hidden in, the ordinary. They were naturalized, made to seem inevitable in the way that the heaviness of a rock is part of its "nature.' This naturalization of local village cultural assumptions gave them much of their stable force in the ordering of everyday life. But Bhaktapur, alongside its own naturalized locally constructed meanings (for example, the complex of feelings, understandings, and actions connected with 'pollution'), had quite another kind of symbolic life, an enormous set of objects, forms, and activities that were clearly differentiated from the ordinary by being marked in various ways to show that they belonged to an extraordinary realm, a realm (to use unavoidable but problematic terms) of the 'religious' and the 'sacred.' In Bhaktapur sacralized2 symbols were extensively put to work to organize much of the time, space, status, economic and psychological life of the city, helping, in short, to shape it into a community. They were doing something for Bhaktapur, presumably, that Piri was able, for the most part, to do without. Let us consider those aspects of Bhaktapur's space which are both (i) sacralized and (ii) at the level of the city-as-city. Sacralized space contrasts with mundane, ordinary space and is (in company with sacralized images, times, status divisions, and actions) clearly marked through various devices as being extraordinary, powerful, and participant in a transcendent world, a world elaborately defined by means of the extensive meaning-giving resources of Hinduism. Alongside the city's sacralized space are the mundane but still, of course, meaningful spaces of the city--responses to its topography, material witnesses to a variety of historical projects and happenings, responses to and indexes of the economic, utilitarian, and communicative needs of the city. Sometimes these other spaces are related to sacred ones, sometimes they are quite independent and secular. Bhaktapur's sacralized space is not only at the level of the city-as-city. There are also other levels and kinds of sacred space, relating to substructures of the city, on the one hand, and its distant outer environment, on the other. (The symbolic order of the city's immediate or neighboring environment--in contrast

The power of space in a traditional Hindu city / 57

to the more distant environment--is, as we shall see, an integral and essential part of the organized meaning of the city and is, in contrast to the distant outside, extensively defined by civic symbols and festivals.) Subcivic sacralized space includes the highly differentiated and significant spaces of the body, the minutely ordered horizontal and vertical space of the house, the spaces of the multitude of kin and craft and caste organizations, and the inner spaces of the city's temples and shrines and its royal palace-complex--whose position in the city (in contrast to their inner structures) allows them to act as important markers in the organization of city space. At a level larger than the city are the sacralized places in the Kathmandu Valley, greater Nepal, and India, the Hindu outside in which the city is embedded. These are spaces whose ritual representation is in pilgrimage and great religious fairs (and whose secular order is in the political and economic sphere). It is likely, by the way, that these subcivic and extracivic orders will survive the destruction of the increasingly fragile civic order of Bhaktapur and its replacement by a more modern city, as household and individual religion persist within secular, modern Indian cities, and pilgrimages and religious fairs persist outside of them. Between these sacralized smaller and larger spaces are the spaces that belong to the city in itself, a major ground of its citizens' identity, and the whole in relation to which so many forms--gods, hierarchical social units, callendrical timv---derive their intelligibility and produce their effects.

THE CITY OF ORTHOGENETIC TRANSFORMATION

The people who inhabited the Kathmandu Valley prior to its conquest in the late eighteenth century by the Gorkhi armies from the western hills of present day Nepal and the resulting incorporation of the Valley into the larger multicultural area and polity that was to become Nepal were called Newars. They had their origin in Tibeto-Burman speaking peoples who had settled in the Kathrnandu Valley in, perhaps, the seventh century BCE. By the first or second century CE the local Tibeto-Burman rulers had been replaced by a dynasty of north Indian rulers who called themselves Licchavis and who were presumably related to the Liechavi rulers of Vaigali in northern India. Over the centuries an integrated Newar synthesis of Indian and Tibeto-Burman forms developed. The early Sanskritic court culture and the Himalayan culture of the people influenced each other, but the Sanskritic forms and ideals remained and flourished as dominating religious, social-structural, and political influences. From early times various currents of Indian religion were introduced into the Valley and, operating

58 / Robert I. Levy alongside the Brahmat).ical religion of the ruling courts, slowly became interwoven into the religious life of the Newars and their developing urban centers. After the fall of the Licchavi dynasty in the ninth century a new dynasty, a local one, eventually arose, which modem Newars consider their founding rulers, the builders of what they take to be their society and culture as Newars. These were the Mallas under whom wealthy and elaborate urban centers, Kathmandu, Patan, and Bhaktapur----each surrounded by a dependent and bounded hinterland of agricultural lands and secondary towns--developed. These cities and towns became centers of literate High Culture, with elegant art--painting and architecture, wood, stone and metal sculpture, and music--as well as the efflorescence and development of, for the most part, South Asian religious ideas, forms, and practices. This sociocultural efflorescence was made possible by the wealth and leisure, for some, derived from intensive year-round agriculture enabled by the Valley's rich alluvial soil, watered by extensive irrigation systems which had been begun in Licchavi days. The Valley authorities were also able to exploit their position on trade routes between India and Lhasa in Tibet to collect tolls from and sell services to Indian traders, as well as to provide services and goods directly to Lhasa itself. It was in this rich, stimulating, and expanding context that Bhaktapur was formally established (on a base of earlier settlements) as a royal center in the twelfth century. Bhaktapur eventually became 'the metropolis of the Malla dynasty and the nerve center of its culture and civilization' (Hasrat 1970: xxxix). The Valley hegemony of Bhaktapur waxed and waned, and eventually the Valley and its Newar society split into three small kingdoms, each with its royal central town, complete with king and court, and its dependent hinterland, a division which facilitated the Newar cities' eventual fall to covetous invaders from beyond the Valley. The Newar chronicles and the historical memory of the people of Bhaktapur credit a particular man, the Rija Jayasthiti Malla, with, in the mid-fourteenth century, restoring a traditional Hindu urban order in Bhaktapur, an order which had 'fallen into disuse.' According to one of the chronicles, with the assistance of his five pandits...[he divided] the people into castes and... [made] regulations for them. Each caste [in Bhaktapur from now on] followed its own customs. To the low castes dwellings, dress and ornaments were assigned, according to certain rules. No sleeves were allowed to the coats of Kasais [butchers]. No caps, coats, shoes, nor gold ornaments were permitted to Podhyas [untouchables]. Kasais, Podhyas and Kulus [low-status drum makers] were not allowed to have houses roofed with tiles, and they were obliged to show proper respect to the people of castes higher than their own

The power of space in a traditional Hindu city / 59

(Wright 1972 [1877]: 183, 187, 182f.). Jayasthiti, the chronicles assert, determined where people should live in the city, controlled marriage and sexual liaisons, determined the proper rites of passage for each social group and the proper extent of purification for different groups, and specified punishments for those who followed the profession of another (for citations of relevant chronicles, see Levy 1990: 42ff.). He assigned special cremation grounds to various groups, and he required that the 'four varn. as' followed the traditional ancient rules for building the houses that would be appropriate to their status. Jayasthiti is thus credited by chronicles and by present day Newars with rationalizing an order out of previous fragments, an order making use of a multitude of South Asian forms and ideas in an attempt to force a particular type of coherence-justified and represented through Hindu traditions, forms, and sacred force--on a city and its inhabitants. Bhaktapur, that is, the city-in-itself, was the central reference in relation to which time, social structure, hereditary occupation, and a pantheon of urban deities were organized. They became part of an organization which framed the experience of the city's people in their everyday activities and, most dramatically and inescapably, in the course of a multitude of festivals which were set against and above everyday life. The ordering that is attributed to Jayasthiti was real to the chroniclers of Bhaktapur, it is what they were trying to explain retrospectively. But the attribution is a narrative simplification. As the art historian Mary Slusser summed it up for the social order (and certainly equally relevantly for other aspects of the Hindu ordering of Bhaktapur attributed to him): [During Jayasthiti's reign] new concepts of administration, nascent in the early Malla years, became clearly established .... But he cannot be credited with introducing the caste system into Nepal, nor with single-handedly infusing hierarchy into Nepalese society, two deeds on which his fame popularly rests. The Indian caste system was in effect in the Nepal Valley from at least the beginning of the Licchavi Period, as inscriptions attest. Similarly, the complex system of subcastes that ordain Valley social behavior must be viewed as the product of centuries of gradual accretion, not a sudden imposition by law. Significantly, [Jayasthiti'sl...own annals make no mention of these undertakings .... Nonetheless, [Jayasthiti]...may well have codified the particular social patterns that had developed by his time, and thus given established local custom the force of law (1982: 59). The order he is supposed to have fathered was to characterize all the major

60 / Robert I. Levy Newar cities and (although modified by their smaller scale) many of the Valley's small towns and was to endure in Bhaktapur for some six hundred years. What kind of order is it in the comparative perspective--this time not of small villages like Piri, but of 'cities' ? It is not that of the multicultural, heterogeneous secular conglomerations that we usually think of as cities, places whose multicultural complexity set them against their rural contexts and made them into transformative historical forces. It rather resembles the order of the ancient places that the anthropologists Robert Redfield and Milton Singer in a seminal essay on 'the part played by cities in the development, decline, or transformation of culture' call (in a neologistic phrase that emphasizes their essential contrast to multicultural cities) 'cities of orthogenetic transformation' (1954: 53, 59). According to Redfield and Singer, in the kinds of cities most of us now live in (and are generally familiar with in history) one or both of the following things are true: (i) the prevailing relationships of people and the prevailing common understandings have to do with the technical not the moral order, with administrative regulation, business and technical convenience; (i0 these cities are populated by people of diverse cultural origins removed from the indigenous seats of their cultures. They are cities in which new states of mind, following from these characteristics, are developed and become prominent. The new states of mind are indifferent to or inconsistent with, or supersede or overcome, states of mind associated with local cultures and ancient civilizations (1954: 57). The 'cities of orthogenetic transformation,' in stark contrast, were places 'which carry forward, develop, elaborate a long-established local culture or civilization. They are cities that convert the folk culture into its civilized [that is, civic] dimension' (Redfield and Singer 1954: 57). It is this conversion of long-established local culture----of whatever characterized the nascent Hinduism of South Asian ancient village and small town practice--that Jayasthiti and, presumably, the designers of ancient Indian cities were trying to bring about? Redfield and Singer conceived the city of orthogenetic transformation as arising someho~v naturally out of the modes of integration and cultural assumptions of the simpler local communities preceding and surrounding it, with whatever transformations were necessary for its new conditions of urban life. Those smallscale precursors were, plausibly, organized something like Piri, that is, by means of the construction of a shared community of interacting understandings, and were communities in which religious forms had comparatively limited functions.4 Village-like devices of face-to-face order work for the internal ordering of the component units of a city like Bhaktapur. But something else is necessary for

The power of space in a traditional Hindu city / 61 the tying together of the relations of such units into a large urban whole. Bhaktapur suggests that one thing that is at hand for this integration of small component units into a complex city is the expansion and differentiation of dramatic, attention grabbing, deeply interesting and compelling religious forms. Hinduism in its most developed urban forms might be conceived of as a sort of symbolic machine which has the power to organize space, time, society, and people's private mental worlds on the large scale of the communal order of a 'unicultural' city. It is a kind of religion--in scale and in the uses to which it is put--which differs significantly, on the one hand, from that of simpler communities and, on the other, from that of the still larger imperial and economic orders which were to dissolve Hinduism's kinds of cities. It is the kind of religion necessary to a city of orthogenetic transformation.

BHAKTAPUR'S SACRALIZED SPACE

The sacralized space of the city as a whole is conceived and portrayed by Bhaktapur's literati and experienced by its ordinary people as a yantra, a complex symmetrical pattern of spaces, boundaries, and points, of divisions and unities, of conflicts and resolutions, a bounded device concentrating some sort of extraordinary power within its outer circular 'man..dalic' boundary in a way that makes the organized space of the city fundamentally different from the wild, dangerous, yet vital, and generative space of its exterior environing space. The city's encircling boundary follows what was once the city wall and is protected by a ring of Tantric 'mother-goddesses,' mat.rk~s, whose mien and stories (recounted in many ways in the city) make them semantically appropriate to this task--as the appearance and mythicand legendary accounts of all of Bhakiapur's space-marking deities help make them semantically appropriate to the particular kind of space they inhabit, protect, mark, or represent. These goddesses are located at the eight principal compass points and placed in a sequence derived from their stories in canonical texts (for the most part from the Dev[ MdJ~tmya). The events and arrangements that occur outside the boundaries marked by the ring of goddesses are set in the 'neighboring outside' of the city where significant statements defining the inside of city through its contrast and transactions with the neighboring outside are made. Here, just beyond the boundary in the inauspicious south, the absolutely impure untouchables must live, signaling the basic, if significantly graded, purity of all those who dwell within the city. Here, in the neighboring outside, are located the shrines of the ances~al deities of the many 'clans' (thars) that are the essential components of the city's status

62 / Robert I. Levy and occupational systems. The external position of these shrines are reminders to city dwellers that their historic origins were in various parts of the Valley and of South Asia beyond the city--for the shrines are thought by some to have been placed where they are to commemorate the first immigrants' last pause before entering and becoming defined by the city. It is a reminder that the city itself highlights and defines a life which began elsewhere before birth into the city and which will after death be again transformed to some other ground. And thus, as everywhere in South Asia, cremations take place outside of the city and, in Bhaktapur (as elsewhere, when possible), on a river bank, where the water will carry the corpse's ashes still further away from the city. The Tanlric goddesses of the boundaries protect the city against those disorders which are beyond the powers of the quite different gods (the benign deities-Vi.s.nu, Laks.mi, Siva, P~xvati, Gao.e~a, and the like) who, placed for the most part in temples and shrines within the city, protect the moral order of the interior of the city, protect it by insuring and providing models for proper relations among the city's people. The Tantric goddesses (and their subordinate male consorts) protect against 'extra-moral' problems---invasion, war, earthquakes, disease, inauspicious astrological portents--and, above all, struggle with and channel the forces that might disturb climate and soil, and thus disrupt the growth of grain in the fields, which are also just beyond the boundaries of the city. There are nine such matrkas, eight at the periphery and one in the center of the city. As well as protecting the city as a whole, they preside over and protect eight radial and one central segment of the city. Through the focal point of the 'houses,' which, in distinction to their boundary shrines, each one has within the city (structures where their processional images are kept), the bounding power of each goddess is brought into an octant of the city, while the centrally located goddess protects her central circular area. There is suggestive evidence that in the past the central goddess may--in accordance with the concept of the yantra--have been conceived to have concentrated the energies of the peripheral goddesses at the city center (Levy 1990: 167f.) There are other centers of the city----e,ach one a point in which some balance of urban divisions is achieved and in which some transcending power is concentrated. (The relation of these various central points to each other and their presumed higher unification is the sort of thing that the esoteric knowledge of the city, known to people with proper initiations, is supposed to encompass.) One center is the site of the royal palace and its adjacent temple-complex of the city's Tantric political deity, the patron goddess of the MaUa kings, who, although long supplanted by the Gorkh~ dynasty, have a persisting and significant ritual presence in Bhaktapur. This palace and its temple-complex were constructed

The power of space in a traditional Hindu city / 63 towards the top of the rise on which the city is built and are the center of a series of city-encompassing concentric circles of residence of different social units, whose status falls progressively towards the periphery of the city, hitting bottom with the residence of the butchers just within the city boundaries and that of the untouchables just outside the city boundaries. A third center of balance of division (and in this case of antagonistic opposition) unites the city's two halves, a type of division represented in the sacralized space of most other old Newar cities and towns. The antagonism, struggle, and eventual unification of the two halves is enacted at length during the course of a sometimes bloody nine day long solar new year festival. The struggle comes to a conclusion and resolution at a point along the line separating the two city halves, when with unified effort a great garlanded tree trunk is laboriously set upright to signal the start of the solar new year. In addition to these large city divisions, Bhaktapur is also divided into twentyfour village-like neighborhoods which function as the face-to-face moral units of the city in which households and larger kinship groupings are embedded. These each cluster around neighborhood shrines of Ga0.ega, whose function as the giver of siddhi, the essential initiatory power on which the success of all ritual undertakings depend, suits him to be the focal god of the neighborhood. Each neighborhood is also guarded by one or more dangerous space-protecting deities. These urban spaces and their resident deities are tied together in 'saeralized action' for the most part through the annual lunar and solar festival cycles, which contain seventy-nine festivals, some of several days duration. These festivals are very varied in their actions and symbolic resources and through the course of the year develop the definition, meanings, and larger relations of each one of the city's sacralized spaces through complex narrative devices. The sacralized urban spaces are also relevant to and orienting for many of the smaller ritual events in the city, particularly household rites of passage and acts of worship for various special, contingent purposesMbut it is the annual festivals that involve the city's total population and speak of the city as a whole as the space in which these local acts take place and have their meaning. Many of the annual festivals follow the city's major festival route, which is arranged so that it winds through most of the major divisions of the city--and thus acts as a physical connector of the divisions. The city also makes use of the symbolic power of points which signify the collapse of order, namely, crossroads which can lead from and to anywhere, places where, as one man put it, 'who knows whom you may meet?' The major crossroads are sites of uncanny and dangerous deities and spirits; they are places where polluting materials can be placed to be absorbed by the chthonic deities who reside in stones half or entirely buried there.

64 / Robert I. Levy

The people of Bhaktapur encounter and invest and reinvest with meaning these spaces which have to do with the construction of the city on its own level during the ordinary movements and activities of their lives and during a large number of expensive and time consuming festivals and family and individual rites. These festivals and rites compete with the demands on capital and lime required for the modernization of Nepal--and have, consequently, been enormously reduced in quantity and scale in the once similar, but now multicultural and modernized city of Kathmandu. But they have held out longer in Bhaktapur, still doing much of their old work. In Bhaktapur, and places that once resembled it, the resource of sacralized space, deities, ritual action, time, status, myth, and legend are woven into compeUing messages and instruction which are repeatedly delivered to the city's people. In the course of the annual [festival] cycle Bhaktapur moves each of its citizens into each of the arenas that are the essential units and varieties of their social experience and of phases of their selves. Positioned successively in each arena, citizens find the appropriate forms of the city's symbolic world rotating around them, engaging them in contemplation and action (Levy 1990: 597-98). The city's symbolic order seems to do much more than simply reflect and mystify some more 'real' order, it seems to be an active resource, a peculiar resource, for creating that order, s

THE POWER OF BHAKTAPUR'S SPACE

There is something about the powerful, sacralized, symbolic world of Bhaktapur which seems problematic when seen from the perspective of the 'modern Western world,' with its ideals (if not reality) of building communal order through politics, rationally balanced interest, legislation, and utilitarian and practical reason. Bhaktapur's sacralized space, and other aspects of its order, evade these putatively reasonable principles of order in two quite different ways. I have been emphasizing one peculiarity of Bhaktapur's space, its sacralized order. Before turning to that space's other 'non-rational' power, a comment is needed on what is entailed in adherence to or 'faith' in an extraordinary supernatural order. Such orders are almost by definition known in a different way than the way in which

The power of space in a traditional Hindu city / 65 the 'common sense' and 'mundane' world is known; that is what makes them extraordinary. The assent to a sacralized proposition has to be achieved by mature individuals and supported by the community in a different way than assent to the proposition that, say, fire will burn one's hand. As an example of one pressure and channel for the formation of 'faith' in Bhaktapur, I have argued at length that for many (perhaps most) people the 'psychological force' of the city's ubiquitous blood sacrifice, with which people are involved in many festivals, in rites of passage and in Tantric worship, helps to motivate and shape assent to aspects of the city's religious system which are open to doubt when people are thinking with ordinary reason and judging 'religious' phenomena from the basis of their everyday experience (Levy 1990: 332ff., 571ff., passim). Faith in Bhaktapur's sacralized world, whatever the various mechanisms that generate and sustain it, is a matter of something other than the pragmatic and sensuously based logic of everyday life. There is another aspect of Bhaktapur's sacralized space that conveys meanings in a way that resists reason. This is the 'experiential' semantic force of the elements out of which the sacralized spatial forms are constructed. These elements are lines, bounded areas, central points, and their compounds--similarly sized and thus equivalent units, larger areas subsuming smaller areas which subsume still smaller ones, circular outer boundaries, axes of division belonging to neither (or all) of the units they divide, insides and outsides, and, adding a third dimension, up and down, higher and lower. These spatial elements contribute an aspect of directly experienced meaning, sensed as significant form, to the sacralized constructions. These meanings--boundaries which prohibit and divide, points which center and define a periphery, high places which require work and achievement to seize and which afford more comprehensive views (and which define by contrast a set of meanings of low places), safe insides and adventurous but perhaps dangerous outsides--are all meanings related directly to experience, everyone's experience, rather than references to something else. (In a similar way Hindu icons of gods are to a considerable degree directly meaningful. They constitute, in a sense and in part, the god's meaning rather than referring to a transcendent elsewhere beyond the image. This is why they were perceived as 'idols' by monotheists.) These elemental meanings are, then, not referential in the way that most symbolic forms are. They are more like the danger directly implied by a burning building, that is danger itself, than a sign or diagram referring to danger. The city's sacralized space conflates the intuitive understanding of the meanings of the elements of spatial order with a sacralizing superstructure, anchored in and enabled by these spatial elements, of powerful religious symbolism. Both these levels are resistant to critical discussion and 'rationality'--and out of their alliance are constructed the urban realities of

66 / Robert I. Levy hierarchy, equality, similarity, difference, inclusion, and exclusion. This kind of space is, in Henri Lefebvre's term, a community's 'representational space,' 'space as directly lived through its associated images and symbols, and is the space of "inhabitants" and "users"' (1991: 39). In their encounters with representational space a community's 'members obey social norms without knowing it--that is to say, without recognizing those norms as such, Rather they live them spatially: they are not ignorant of them, they do not misapprehend them, but they experience them immediately' (Lefebvre 1991: 230). Lefebvre notes that this is the space also of 'some artists...and a few writers and philosophers, who describe and aspire to do no more than describe' (1991: 39). Redfield and Singer (1954: 57) suggest the same idea in claiming that the 'intellectuals,' that is, the writers and philosophers of the city of orthogenetic transformation, were literati, while those of the heterogeneous cities were, rather, intelligentsia. 'Intelligentsia' are those who struggle against the tyrannies, as they feel them to be, of representations. Their imagined homeland is Classical Greece.

OTHER ORDERS: CLASSICAL GREECE AND ISLAM

Assertions about the peculiarity of Hindu space (and other aspects of its communal organization) and the particular peculiarities that are chosen to characterize it are generated by a 'Western' discomfort, for some, about its implications. The ideal defining historical contrast, the mirror opposite, is in what we have chosen to emphasize in the Classical Greeks, who are conventionally thought to have pushed beyond the anti-individualistic tyrannies of sacralized civic organization, opening the way for new freedoms--and new kinds of tyrannies. As Bernard Williams writes of the ancient Greeks, we do not study the Greeks as anthropologists study other societies to learn about human diversity, for the Greeks are among our cultural ancestors, and our view of them is intimately connected with our view of ourselves .... To learn about the Greeks is... [in contrast to what is usefully learned from studying our contrasts and similarities with non-ancestors] more immediately part of self understanding. It will continue to be so even though the modern world stretches round the earth and draws into itself other traditions as well... [for] the Greek past is specially the past of modernity .... The process by which modernity takes in other traditions will not undo the fact that the modem world was a European creation presided over by the Greek past (1993: 3).

The power of space in a traditional Hindu city / 67

In this organizing perspective, Hinduism with its literati, its virtuosi of representational form, is decidedly not one of 'our' cultural ancestors. 6 Commentators on the development of ancient Greek cities write of the development of an urban planning which, in contrast to Near Eastern settlements and to earlier archaic Greek towns, emphasized their secular order and geometric regularity. These plans have suggested, at least to some observers, places whose purely geometrical control of space might be considered the architectural corollary of the growth of philosophical speculation .... The grid...offered a wholly abstract method whereby the city could be planned with only minimal regard for topography (Scully 1962: 187) and, one might add, with minimal regard for Bhaktapur's kind of imaginative mythic construction. Even if the plans of the Greek towns and cities suggested a development of secular and geometric emphases at the expense of 'mythic' and 'religious' spatial forms, sacralized forms and activities--temples, gods, processions, and festivals --were, of course, profusely present. Yet even within the sacralized sphere itself there are suggestions of some provocative contrasts with Hindu practice. For example, in Bhaktapur's religious processions most commonly an image of the god is brought out of its temple or shrine and carded around the main festival route, or through some particular segment of the city, contributing to the meanings and sacredness of the urban spaces through which it is carded. But in Classical Greece this movement of deities was exceptional. According to Walter Burkert, 'processions with images of gods--which play a major role in the ancient Near East--are [in Greece] an exception.7...Such a moving of the immovable is an uncanny breaking up of order' (1984: 92). While in Bhaktapur gods invade the city and animate its order, in Greece they were, it would seem, beginning to be kept in their proper and constrained place out of the city's secular space. While in Classical Greek cities, in contrast to Bhaktapur, for the most part gods were kept in their shrines and out of processions, the human membership of those processions had other significant contrasts with Bhaktapur's. Thus, in many Greek and Hellenistic festivals individuals were represented who had achieved their statuses through competition, including the victors of athletic contests (Fox 1986:80)3 Bhaktapur's Hindu festivals have none of this--it is the immemorial, unquestionable statuses themselves that are repetitively represented. These fragments of difference in the relations of the secular and the sacralized in urban space are to be remarked because they are consonant with a view of Greek cities that they were breaking loose from the tyranny of mythic forms by

68 / Robert I. Levy means of techniques of abstraction and comparison. As Lefebvre wrote of the archetypical Greek theorist Aristotle, his categories of the understanding of space and time can be understood as 'generalities in some way superior to the evidence supplied by the body's sensory organs' (1991: 1), that is, as an extrication from the directly sensed implications of constructed urban space and, at the same time, the 'irrational' tyrannies of its sacralized superstructure. It all took place, the story goes, as part of a new situation which put great pressure on these old modes of community organization. 'The late sixth to the early fifth century,' writes Martin West in an essay on early Greek philosophy, was a time when the Greeks were developing a particular interest in the beliefs and customs of other nations. As Xenophanes' arguments illustrate, the effect was to make them aware how much their own beliefs and customs were based on mere convention, which might profitably be challenged (1988:111). Looking at Bhaktapur's space from the perspective of these ideals, it is seen as binding in its particularistic significant forms, forms which can only be fully grasped and maneuvered through long residence, perhaps even by birth in a city so concretely, complexly, and locally marked, constructed so as to be resistant to abstract deconstructive analysis, and thus resistant to the transformative idea of 'mere convention.' It is a kind of space which is designed to be inhibiting to people on the move such as the intelligentsia of Athens or to the carriers of Islam into South Asia. Islam suggests one more relevant refraction of the peculiarity of Hindu urban spaces like Bhaktapur's. The ideal urban space of the Muslims who conquered the Indian royal cities (and who forayed into Bhaktapur in the fourteenth century) was of a radically different sort than the urban spaces they encountered in South Asia's cities of orthogenetic transformation. Islam--and its contesting divisions --needed cities and city space which were adequate to the tasks and governance of empire. Islam's responses to such South Asian cities was an abhorrence of their symbolic forms, their 'idols.' The 'idols' that they destroyed...[helped to] anchor...each South Asian royal city at the center of the universe--thus generating innumerable local universes. In irreconcilable contrast, however, 'the basic objective in the expansion of Islam was to acquire political control over an area and to set up the symbols of Islamic sovereignty' (Inalcik 1984: 10). Within this expanding and universal Islam, the Islamic world view 'determined the physical and social landscape of the city. The city was supposed to become a space where the prescriptions of the Islamic religion could be performed fully and appropriately' (Inalcik 1984:

The power of space in a traditional Hindu city / 69

9). Instead of the city being a center transcending secular history and geography in order to center itself in a mythic history and space, it became an off-center marker in a universal and presumptively objective and real grid of mundane space and history whose presiding god was incorrigibly transcendent. The tasks of symbolsmand thus their form and nature--in such a transformed city are altered. They become thinned out, more universal, easier to read by the various kinds of people who were to belong more to a universal Islam than to a particular city. Law, standardized Islamic law interpreted and enforced by a bureaucracy, became central to the regulation of such cities .... Local symbolic forms-including local spatial constructions--had to be made to represent an abstract, rational, and universal political, social, and ethical order. As far as possible, local...[worlds like Bhaktapur's] had to be dissolved (Levy 1990: 619). These imperial Islamized citiesmalthough they eventually developed, to some degree, secondary local forms of religion, particularly in their rural hinterlands and among the 'lower classes' of the cities themselves (Gellner 1994" 18f.)--were in their dominating ordering of a radically different kind than the South Asian ones they encountered. They were at the service of a different political order. In these Islamic cities, as in well laid out modern ones, we move from a city like Bhaktapur, locally known and 'felt' by its long rooted citizens, significantly different in the layout of its space and of its festivals from each of the other Valley cities, to a more generally theorized interchangeable kind of city, a member of a set of Islamic cities, units in an empire. You don't have to be born into such a space to know it, the small and intimately known space of the hedgehog is being transformed into the larger, more superficial and abstract space of the generalizing, powerful, and dangerous fox. The Muslims' effect on Bhakiapur was transient; their depredations into that city, paradoxically, may well have made the Hinduizing reforms of Jayasthiti Malla easier, as he replaced ruined physical structures and, perhaps, the shaken morale of the city's people. The ordering of Bhaktapur through the power of Hindu ideas and images was restored, elaborated and strengthened and developed into a 'climax community' of Hinduism, where that system of sacralized symbols, interpretations, and codes could demonstrate its full potential for a certain and (for Greeks, Muslims, and us) peculiar kind of communal ordering. It was to work well after its own peculiar fashion for another five hundred years.

Notes 1. In this paper I use the present tense to characterize Bhaktapur as it was at the time

70 / Robert I. L e v y of my studies. 2. I use 'sacralized' rather than 'sacred' in this paper to suggest a process (rather than a state), a process which results in different degrees and kinds of 'sacredness' being attributed to various structures and acts. 3. It is important to note that this 'level' of order is that which has to do with the strategies and techniques for trying to further overall urban integration, the integration of the city-in-itself.' The Newar city---and its citizens---contain many kinds of South Asian religious and social ideas and practices which the larger urban ordering variously uses, struggles with, or ignores. 4. Residues of assumptions about the evolution of 'man' and 'culture' once made it seem plausible that earlier, more 'primitive,' communities depended more on religion for social and psychological uses, and thus anthropologists and historians focused on the religious dimensions of such communities' organization. I suggest that communal religion, at least, is an organizational resource exploited in its fullest form by large and complex 'archaic' communities. 5. Such a claim has to be argued in two ways, first from the top down, as I did in Mesocosm, through the detailed interpretation of the structure and meaning of the symbolic order, and, also, from the bottom up, in an analysis of the effects of the symbolic order as revealed in individuals' accounts of their life. I have treated some of the bottom up material in Mesocosm and in Levy (1996) and hope to present it in more detail elsewhere. 6. This same sense of the ambiguous relation of Hinduism to the 'modern' is expressed in academic uncertainty as to whether India has undergone our ancestral 'transformations of the axial age' (see Levy 1990: 22f.). 7. There were a very few exceptions to this generalization. An important one was during the festival of Dionysos Eleuthereus in Athens, whose statue was moved within the city. This seems to reflect the peculiar meanings of Dionysos as a deity. 8. Fox is writing specifically of Hellenistic festivals, but his remarks are equally true of Classical ones.

References cited Burkert, Walter. 1984. Greek religion: Archaic and classical (trans. John Raffan). Oxford: Basil Blackwell. Fox, Robin Lane. 1986. Pagans and Christians. New York: Viking. Gellner, Ernest. 1994. Conditions of liberty: Civil society and its rivals. London: Hamish Hamilton. Hasrat, Bikrama Jit. 1970. History of Nepal, as told by its own and contemporary chroniclers. Hoshiarpur: V. V. Research Institute Press. Inalcik, Halil. 1984. Istanbul as an Islamic city. Paper presented at a Conference on Meanings of the City, Wisconsin, 25-27 October. Lefebvre, Henri. 1991. The production of space (trans. David Nieholson-Smith). Cam-

The power of space in a traditional Hindu city / 71

bridge, MA: Blackwell. Levy, Robert. 1973. Tahitians: Mind and experience in the Society Islands. Chicago: University of Chicago Press. Levy, Robert. (written in collaboration with Kedar R~ij Rijop~dhyiya). 1990. Mesocosm: Hinduism and the organization of a traditional Newar city in Nepal. Berkeley: University of California Press. Levy, Robert. 1996. Essential contrasts; differences in parental ideas about learners and teaching in Tahiti and Nepal. In Sara Harkness and Charles Super, eds., Parents' cultural belief systems, 123-42. New York: Guilford. Redfield, Robert and Milton Singer. 1954. The cultural role of cities. Economic development and cultural change 2, 1: 53-73. Scully, Vincent. 1962. The Earth, the temple, and the gods. New Haven: Yale University Press. Slusser, Mary Shepherd. 1982. Nepal mandala: A cultural study of the Kathmandu Valley. Princeton: Princeton University Press. West, Martin. 1988. Early Greek philosophy. In Joan Boardman, Jasper Griffin, and Oswyn Murray, eds., Greece and the Hellenistic worM, 126--41. Oxford: Oxford University Press. Williams, Bernard 1993. Shame and necessity. Berkeley: University of Chicago Press. Wright, Daniel. 1972 [1877]. History of Nepal. Kathmandu: Nepal Antiquated Book Publishers.

ROBERT I. LEVY is Professor Emeritus of Anthropology at the University of California, San Diego and, at present, Research Professor of Anthropology at Duke University.