Ritual movement in the city of Lalitpur.

Pickett, Mark A.

This article intends to explore the ways that the symbolic organization of space in the Newar city of Lalitpur (P.atan) is renewed through ritualized movement in the many processions that take place throughout the year. Niels Gutschow (1982: 190-3) has done much to delineate the various processions and has constructed a typology of four different forms. (1) In my analysis I build on this work and propose a somewhat more sophisticated typology of processions that differentiates the types still further.

Sets of Deities

Space is defined by reference to several sets of deities that are positioned at significant locations in and around the city. Gutschow (ibid.: 165, Map 182) shows a set of four Bhimsen shrines and four Narayana shrines which are all located within the city. Both sets of four shrines encircle the central palace area. Gutschow (ibid.: 65, Map 183) also shows the locations of eight Ganesh shrines that are divided into two sets of four. One of these groups describes a polygon that, like those of the shrines of Bhimsen and Narayana, encircles the palace area.

It is not clear what significance, if any, that these particular sets of deities have for the life of the city. They do not give rise to any specific festival, nor are they visited, as far as I am aware in any consecutive manner. Other sets of deities have very clear meaning, however, and it is to these that 1 shall now turn to first more on the boundaries which give the city much of its character.

City Boundaries

As Slusser points out (1998 [I]: 92-3, and [II]: plates 95; 96) the old Malla cities were walled settlements punctuated at many points by gates. This is attested by contemporary and later accounts such as those of Kunu Sharma (1961) and Oldfield (1981: 1, 95-6, 102-3, 111) (2). It is possible, as Slusser did herself to trace the precise location of the old wall even though there is no extant substantial evidence. Local residents know the exact point at which they are within (dune) or outside (pine) the old walls. Place names such as Kwalkhu (Kwalakhu) and Ikha Lakhu tell of former gates and boundaries going back, in the latter case, to pre-Malla times (Map 1).

[ILLUSTRATION OMITTED]

Local residents tell of twelve or more gates that used to perforate the walls of Malla era Lalitpur. When asked where exactly the gates were, however, they are often at a loss to give the precise locations of more than at the top, had the names of only two gates: tah dhvaka on the west and jhval dhvaka on east.

The easiest way to tell where several of the gates were placed, however, is by an analysis of the locations of some of the low castes. Although most residents of Lalitpur were not intentionally situated at any particular distance from the city centre, two low caste groups were clearly located on, or relocated to the boundaries of the city at some time in the distant past. The Dyahla. Untouchables were placed immediately outside the old Malla city gates, whereas the (unclean but touchable) Khadgi Butchers were, for the most part, placed immediately inside. (3) Since these castes still live largely in their traditional localities, the simple matter of distinguishing one caste from the other can lead to, at particular locations, precise mapping of old city boundaries (Map 2). The Dyahla. were to be located not simply outside the city wall but more particularly outside the gates. In this way their polluted presence would protect the city from marauders or evil spirits (bhut/pret) which, like the city's inhabitants, would also find them repulsive.

[ILLUSTRATION OMITTED]

Further evidence for this emphasis on the importance of the division between the 'inside' and 'outside' of the city is provided by an analysis of Lineage Deity sites. Lineage Deities are always sited outside the settlement. The reason for this is not clear, although their resemblance as natural stones (Prakrit), to the aniconic pithas may have something to do with it. Perhaps also the relation of the cult of the Lineage Deity with the worship of dead ancestors (pitri) led to such an association. Many lineages retain an oral history of the establishment or relocation of their Lineage Deity. It is possible, then, by comparing these oral historical accounts with the actual cult and location of these shrines to make some tentative deductions about the time in which a certain lineage migrated, or how the lineage in question was established in the first place (Locke, 1985:517; Gellner 1992: 239-40).

Why this preoccupation with boundaries? The nineteenth century Sanskritist Max Muller drew attention to the fact that

 Hindu towns seem to bear out [the] notion [that] a town is a region
 that is, as it were. walled off from the sourrounding country. Such
 land as was inhabited by an organised community had to be set off
 from the country, which was unstructured, uncultured, not
 'urbanized'. (Quoted in Gutschow and Kolver 1975: 20)

Hence the Sanskrit pur means both 'wall' and 'town'. The wall therefore acted as a boundary between the structured, urbanised inside and the unstructured, wild outside.

Such a concern with boundaries goes back at least to Vedic times. Michael Witzel explains that the Vedas made a distinction between grama, 'settlement', and aranya, 'wilderness' (1997: 519). The aranya was dangerous and lull of threats. It was where the barbaric and uncivilised aborigines and dreaded demons lived. By demarcating the boundaries, therefore, the Aryan settlement became a sacred space resembling the Vedic offering ground. In this it seems the ancient Aryans were not alone. Jameson points out that the Newar city's concern with boundaries, particularly those between order and chaos, offers striking analogies with those of that other ancient Indo-European civilisation, Greece (Jameson 1997: 487).

Ashoka Mounds

Gutschow and Kolver suggest that Lalitpur was subject to planning from a very early stage and that this became necessary in order to delimitate the new foundation on the fusion of the earlier separate settlements (1975: 20). The four mounds (slupas or thurs) that ring the city seem to have been set up to give, on the one hand, an expression of unity to the space contained within, and on the other hand some sort of orientation to that space. Their location, though notionally at the cardinal points of the compass are up to 24[degrees] off the true geographic axes. In this the mounds correspond to the major trade routes described above (Map 1: cf. Gutschow 1982: 154). Most of the streets of Lalitpur are laid out in a grid pattern roughly parallel to these, trade routes. It would seem that the city grew up self-consciously oriented by the four mounds which themselves were built to sanctify the space within and lend some legitimacy, if it was needed, to the development of the city through its trade. Could it also be significant that, located, as they are at the four entry points of what was, after all, a crossroads the stupas were intended to counteract the inauspiciousness of this node? (5)

Power-places

In and around the city of Lalitpur there are thousands of Power-places (pitha or pigandyah) each consisting of a simple unhewn stone. These stones are often several feet below ground level with steps leading down for the convenience of worshipper. Rarely are these 'hyperethral shrines', as Slusser calls them, covered although, like the Bhavani Power-place in the courtyard of 'House' of Managh., they may be surmounted by brass naga serpents and related paraphernalia and surrounded by a stone border. Many, perhaps, all of these shrines have a Sanskrit name but it is clear that most, if not all, represent local cults that were later Sanskritised. Locals usually interpret these shrines as sites where pieces of Parvati's decomposing body fell to earth when, according to the myth, the grieving and distraught Shiva carried it on his shoulder.

Many residents make a distinction between 'true' Power-places and those that are, as it were, interlopers. The number of true Power-places, they say, is twelve, but a heated argument will often ensue when trying to define which of the plethora of possibilities are rightly included in the group. Although many Power-places are located within the old city boundaries, all twelve of the True Power-places are located outside the city (Map 3). The twelve consist of the Eight Mother Goddesses (Ashtamgtrika) with the addition of a shrine each belonging to Bhairava, Kumar, Ganesh and Siddhilakshmi (Table 1).

[ILLUSTRATION OMITTED]

Power-place worship (pith puja) is an essential part of the autumnal festivals of yenyah and Mohani, and the equivalent vernal Mohani festival of Paha Cahre. During Mohani, procession to the Power-places begins on Kayashtami (Bhadra- Shukla/Yalathwa 8) and goes on for a full lunar month. Most Newars visit the Power-Places in the mornings playing as they go the damaru (dabu dabu, small, one-handed double-headed drum)

(cf. Gutschow, 1982: 173, fig. 194). Maharjans, on the other hand, visit the Power-places in the evening playing the dhimay (large, double-headed drum) and bhusayah. (accompanying large cymbals).

The Eight Mother Goddesses (Ashtamatrika) are worshipped everywhere around the Kathmandu valley as manifestations of Devi, the Goddess, the city of Bhaktapur can be neatly divided into nine sectors corresponding to the Eight Mother Goddesses plus a ninth, central goddess, that are situated at Power-Places around its perimeter (cf. Gutschow's map in Levy, 1990:155).

In spite of Toffin's claim to the contrary (1991: 484) it is not at all easy to trace an eight-fold division in Lalitpur corresponding to the Eight Mother Goddesses. Rather, as Barre et al. (1981: 117) have pointed out for Panauti. there seems to be only two, or perhaps three, of the Mother Goddesses that have a significant cult--Mahalakshmi, Bal Kumari and Vishnu Devi. If there is, in fact, an eight-fold division it is not at all as well developed as it is in Bhaktapur. Furthermore, there are no 'god houses' (dyah che) within the city corresponding to Power-places outside as there are in Bhaktapur (Levy 1990: 23; Vergati 1995: 39).

Apart from the significance of the Eight Moither Goddesses for the festivals outlined above, Lalitpurians also do Power-place worship during the course of various life-cycle rituals. It is instructive to map the Power-place worship of various lineages. An analysis of Power-Place Worship of those Pengu Dab (Tamvah, Marikahmi. Sikahmi, and Lwahakahmi) who still live within the confines of the old city, for instance, demonstrates that the attachment to a particular Power-place is determined by two factors: descent and location. Ultimately location wins over descent as, when a group moves on a permanent basis, they take up the worship of the Power-place that is related to that locality. Their traditional Power-place cult is not necessarily abandoned, however, and at important life-cycle events a puja will be offered to both.

The question arises as to the purpose of these Mother Goddess shrines. It would seem that we are dealing here with the systematising of a number of existing deities into a set in order to give some form to them and raise their significance for the city as a whole. Gutschow and Kolver (1975:21) suggest that the system of the Eight Mother Goddesses was probably meant to raise the status of what had become a royal settlement.

It seems to me, however, that the system does more than this. Each of the Mother Goddesses is situated outside the city, in the wild, disordered land where the demons lurk and bandits may attack. In the same way as the Untouchable Dyahla were positioned outside the gates of the city, presumably during Malla times, the Eight Mother Goddesses were neatly situated to ward off the danger of that other world. The reason that fierce gods and goddesses are better than benign deities as policemen of boundaries is because people are afraid of them (Gellner 1997: 552). By making them into a set and giving them Sanskrit names, medieval Lalitpur was formalising their protective role. A set of eight also speaks about unity in diversity. The medieval city was a disparate one with lineages and castes all pulling away from increasing central control. By making the eight into a set, the city was saying something about the unity of the city itself; a unity against outsiders that revolved around the royal centre from which processions to the outside would begin and to which they would return.

There is a further group of goddesses that we must examine here. These are the Ten Great Knowledges (Dashmahavidya). Again, the exact complement of the Dashamahavidya is contested (Table 2). Not all of the Ten Great Knowledges are of equal importance. The cult of Bagalamukhi is important for all the city's inhabitants. Her propitiation is seen as important for the prevention of cholera (Slusser 1998: 322). As with the twelve 'true' Power-places, the cult of the Dashmahavidya is especially important for Kayashtami (Bhadra- shukla/Yalathwa 8) when Dashmahavidya and the Ashtamatrika are not so much in their number as in their location. With the exceptioin of Vajrayogini (Tara) of Puco, each of the Great Knowledges, accompanied, in its vicinity, by an image of Bhairava, is located within the confines of the city (Map 4). (9) They are not, however, evenly distributed throughout as one might expect; none are located further east than Tripurasundari in Caka Bahi. Dil Mohan Tamrakar is of the opinion that the Dashmahavidya represents an older group of protective deities that were once situated outside the boundaries of the settlement. Indeed, he reported a tradition that the image of Bal Kumari at Kwache is a copy of the one at Caka Bahi, which is located adjacent to that of Tripurasundari. (10) It seems highly probable, then, that the Dashmahavidya represented the protective deities, analogous to the twelve True Power-places, of the Licchavi city of Yupagramadranga. (11)

[ILLUSTRATION OMITTED]

Cremation Sites

Related to the locations of the Eight Mother Goddesses are the cremation sites (mashan) of the city. Three of the crematioin grounds-Shankhamul mashan, Manohara Manimati mashan and Yappa mashan--are located adjacent to Mother Goddess shrines--those of Camundg, Bal Kumgri and Vishnu Devi respectively. (12) The majority of Lalitpurians cremate their dead at Shankhamul, which was refurbished in the 19th century under Rana patronage. (13) Each of these three cremation grounds is located near a river. The Yappa mashan, however, located at the Vaishnavi shrine at Nakhu, is surprisingly at some distance from the river itself. Could this be because the river has meandered a different course since the cremation ground was established? The Shankhamul cremation ground is at the riverside itself and consists of a number of bathing and cremation ghats. It is similar in appearance though not in size to those at Banaras and Pashupati. It is also at some distance from its accompanying Power-place, that of Camunda. One could surmise that the river had meandered away from the shrines since they were established. In the case of Shankhamul the importance of the cremation ground enforced its relocation with the retreat of the riverbank. Lalitpurians, in fact, make a distinction between the mashan at the riverside and the actual cremation platform (depah or dip) which is usually up on the bank and, as we have seen, sometimes at some distance from the river itself. At Vishnu Devi there are three cremation platforms--one for Vajracaryas and Shakyas, another, nearest the shrine, belonging to Maharjans and other high and middle castes, and a third at some distance from the two for the Khadgi.

The cremation ground at Shankhamul has four cremation platforms each with its own shelter (phalca). High castes and Maharjans may use the largest. Tandukars, Khadgis, and Vyanjankars have one each. The Untouchable Dyahla do not have a cremation platform at all but burn their dead separately down by the river (Gellner and Pradhan 1995: 167). All castes throw the ashes of the dead in the river upon completion of the cremation.

The Central Hub: the Palace Square

At the centre of the city of Lalitpur, lies the palace area (layaku) referred to by locals as Mangah and marked on maps as 'Darbar Square'. The present palace buildings were built, for the most part, in the seventeenth and early eighteenth centuries in place of previous palace buildings. The site now occupied by the palace seems to have been the seat of the monarchy from Licchavi times. The adjacent water fountain, the Mangahhiti, was built around 400 AD. and is the oldest known structure in the locality. Local tradition relates that the palace was moved to its present site from a mound at Patuko, one block north and west of Mangah, during Licchavi times. According to this tradition, the Patuko mound was built by the Kirati kings (Shrestha and Malla 1971: 36; Gellner 1996: 129; Landon 1876: 209). The myths of origin of the nearby Kwa Bahah refer to oppression by the Kiratis who were forced from their palace by hundreds of bees and chased to the locality of Cyasah to the northeast where eight hundred (cya sah) of them died. Lalitpurians regard the Vyanjankar caste, which inhabits Cyasah, as the descendants of those who survived.

The relocation of the palace to its present site, precisely at the intersection of the two major trade routes seems to have been an attempt to express spatially what was already a reality politically that the king was the centre of the life of the city. We will look at this more later on when we consider the movement of gods around city space.

The legitimation of the central, role of kingship in Lalitpur was accomplished first and foremost by the construction of various temples within and in front of the new palace. The present day arrangement of monuments in Lalitpur's palace square demonstrates a very carefully thought out strategy to sanctify the central, royal space by establishing a number of seats of gods. The intended effect of this sanctification was no doubt to lend legitimacy to the king's reign. The amazing density of concentration of major temples of brick and stone in such a small area bears testimony to careful design. Although the buildings are of very divergent designs and built with various materials the overall visual effect is not crowded. Nineteenth and early twentieth century foreign observers marvelled at the beauty of the square as does many a tourist today. Since its designation as a UNESCO World Heritage Site in 1979 many of the monuments have undergone extensive restoration. The architect Sekler (1987: 66-7) describes the design of the space thus:

 By having related heights, horizontal lines were carried from one
 building to the next and all this helped to unify the overall
 impression. Were it not for the unification through these means
 and through the underlying ordering principles of number and
 measure, the total impact might have been too restless, even
 disturbing, owing to the strong contrasts of shapes, textures and
 colours. As it is, unity and diversity balance each other in a most
 successful manner.

Why was it so important to design the square with such an approach? The architects were without a doubt expressing certain values by doing so.

 The strong visual order imposed on the buildings of the Darbar
 square in all likelihood was based on an equally powerful and
 complex belief-system. This order, together with the outstanding
 artistic craftsmanship of the masters of the past and with the
 originally impeccable-now, alas, broken-hierarchy of scale, made
 Patan Darbar Square one of the great historic urban spaces of the
 world (ibid.: 67).

So the overall effect of the location and design of the Palace Square is to demonstrate the unity of the city around and under the rulership of her king. This centrality of kingship, though, since the conquest of Lalitpur by Prithvi Narayan Shah in 1768 greatly weakened politically, is still strongly expressed ritually. We will see how this works below.

Ritual Movement through Space

There are several procession routes (Skt. pradakshinapatha) in the city of Lalitpur. Processions following these routes vary greatly in significance for the city. The great festival of Mataya, for instance, attracts many thousands of. devotees, whereas the Narasimaha Jatra, which follows a different route, is essentially the province of only one caste.

It is helpful at this point to tease out a typology of the procession routes for the city and to try to discern the significance of each one within the overall context of the annual march of the calendar.

One could take as the most basic division between the processions, those that have the worship of a deity as their main raison d'etre and those that do not. The only one that falls in the latter category, however, is the funeral procession, or 'Way of the Dead'. Another basic division, at least for the vast majority of those processions that are for the purpose of worship, is between those in which the devotees visit fixed images of the deities, and those in which movable images are transported through city space.

Niels Gutschow (1982:190-3) has differentiated four types of procession route: centrifugal, linear, centripetal, and convoluted. The last two seem to be variations of the same category. Moreover, Gutschow counts Pitha Puja as a centrifugal procession. Although it does include an excursion from the city I include it as a separate type of centripetal procession because it is, nevertheless, focused on the centre, as I will demonstrate. I have, therefore, constructed a new typology that attempts to take these factors into consideration (Table 3).

Centrifugal Processions

Certain processions are centrifugal, that is, they start at the centre, or at least on the inside, of the city and work their way outside (ibid.: 193, fig. 233). The classic festival in which centrifugal procession plays a significant part is that of Gatha Mugah which takes place in the middle of the monsoon. Effigies of the demon Gatha Mugah are constructed by children during the day and in the evening carried out of the city on a clearly defined route to a place where it is burned.

The other procession to take a centrifugal character is that of the funeral (Map 5). (14) On the death of a resident, pallbearers, other members of the death guthi, family members, and friends gather at the house of the dead. In a short time, the corpse is brought out, a procession forms and then moves at, what seems to a European, a hurried pace through the streets of the dead person's locality (twah) and on through the city to the cremation ground outside. The route, the 'way of the dead' (silapu), is always the same. Everyone knows the route that he will take on his last journey from his home, should he have the fortune of breathing his last breath at the place, for men at least, that he more than likely breathed his first. The route followed by each locality is, in most cases, an apparently rational one to the outsider, though not always. In some places a diversion seems to be made so that the route is not the most direct. The most glaring detour is that followed by the residents of the localities south of the royal palace.

[ILLUSTRATION OMITTED]

A funeral procession from Haugah or Tangah, for instance, makes its way down the main road towards the palace at Mangah but then, instead of going straight down the road in front of the palace and the great temples of the palace square, turns one block west to Mahapa. Only then does it resume its northward journey before heading east again at Konti to exit the city and finally arrive at Shankhamul. Two phenomena are noteworthy here. Firstly, the funeral procession does not exit the city at the nearest gate but heads, instead, across the city. Secondly, the procession makes an inefficient detour around the centre instead of going straight through it. These two phenomena lead me to conclude that the normal and ideal centrifugal pattern is modified by the importance of the site of cremation. This importance is almost certainly a measure of its antiquity. In the context of Bhaktapur, Gutschow and Kolver (1975:27) suggest that the ways of the dead doubtless belong to the oldest materials that can be found. If this is so, as one would expect, then the later superimposition of the palace on this pattern led to the need to modify the pattern in order to avoid that place of supreme purity. When I asked why the procession must avoid the palace area, informants told me that the king loved his subjects so much that he mourned whenever he saw a funeral procession. This was how Dipak Lal Tamrakar expressed it:

 The reason why the way of the dead is around Mangah is so that
 the king would not see the procession from out of his window. If
 he did then he would grieve for the dead as for one of his family.
 Why give him grief and make him eat only one meal that day?

From the popular perspective, therefore, the processions took a detour to avoid giving the king unnecessary suffering. Such a story fits in with the devotion with which the Nepalis, in general, honour their monarch. (15)

The main significance we must understand about centrifugal processions, however, is this need to remove impurity and vileness from the city. Both corpses and effigies of demons, representing the demons themselves must be taken out of the city. The city must be kept pure in order to function as a sacred space. All impurity must be removed. It can be said that the mundane daily task of the city's sweepers, a group still consisting largely of the Untouchable Dyahla is also centrifugal entering the city as they do only to clear the rubbishes of urban life. In Malla times, in fact, they were only allowed into the city at daybreak and had to be out again before it got dark. Even today, one can enter one of the localities of the Dyahla and be hit by the overpowering stench of the putrefying carcasses of dead dogs, that have been removed from the confines of the old city and simply left in front of the dwellings of the sweepers, who seem to be oblivious to the foetid atmosphere.

Linear Processions

The movement of deities through city space is affected, first of all, by whether those deities normally dwell within or without that space. Linear processions are those that, against along prescribed routes, enter from the outside, continue through city space and exit again to return to the deity's home (ibid). Several deities are carried through the city in this way. All are resident outside the city. Of these, three are of the Mother Goddesses, Mahalakshmi, Bal Kumari, and Vishnu Devi (ibid.: 173, Fig. 195). (16) Another is of the powerful, Tantric image of Ganesh called Jala Vinayaka or Kwenadyah. Though Kwenadyah has its own major temple at Co Bahah, south of the city, where the Bagmati River flows through a gorge on its way out of the Valley, the large brass mask-like processional image is kept for most of the year at another temple at Puco. It is this heavy image that is mounted on a palanquin and carried through the southwestern sector of the city on its annual jatra. The great annual chariot festival Bugadyah also describes a linear route, as one would expect of visiting deity (Map 6).

[ILLUSTRATION OMITTED]

Centripetal Processions

Centripetal processions are those that emphasise the integration of city space. By processing around city space--clockwise keeping the centre on ones respectful right--the festival integrates the diverse communities that make up the localities of the city. There are three types of centripetal procession: royal, convoluted, and excursive.

Royal centripetal processions: The majority of processions of Lalitpur follow a centripetal or integrative pattern. That is, by circumambulating the centre of the city they effect the unification of city space (Map 7). (17) The main procession route of the city, this integrative pradakshinapatha, follows a route that wends its way around the centre of city space taking in as it does ten localities (twah). Such centripetal processions are those that take place in the festivals of Sa Paru, Nyaku Jatra, Mataya Ganesh Puja, Narasimha Jatra, Krishna Jatra, and Bhimsen Jatra. All centripetal processions follow a clockwise route--that is the centre of the city is always kept on the right of the processing devotees. There is an important exception here, however. At the point immediately to the north of the royal palace, the procession does not circumambulate the city centre and the palace by progressing to the east and through the localities of Valakhu and Nyauta. Instead, the procession route takes a direct line to the centre taking the devotees immediately in front of the palace and thence east to Saugah and Nugah.

[ILLUSTRATION OMITTED]

This integrative or centripetal pradakshinapatha has been designed so that it winds directly in front of the palace. Why is this? One thing we can be sure of is that this pradakshinapatha does not describe an older city boundary as Barre et al. suggested for Panauti (198: 41) and Gutsehow himself suggested for Bhaktapur (Gutschow and Kolver 1975: 21; cf. Toffin 1991). The route does not follow any known boundary that has been discovered by historians. Furthermore, there is no concept of inside and outside as there would be if the route were considered a boundary such as a wall. Rather, such a pradakshinaptha binds the city together as Gutschow later realised (1982: 190). But furthermore, and this is what makes this procession route different from the other kind of centripetal route (the convoluted one to which I will turn shortly) this route is a royal route. The event consciously includes the king in a way that convoluted processions do not, and hence I call this royal centripetal pradakshinapatha. (18) Interestingly, in addition to using the designation thya for those who are invited to a feast or similar event, Lalitpurians also use in to refer to those who are allowed to take part in the Mataya and Nyaku Jatra, in a sense similar to that reported by Ishii for Satungal (1978). (19) Certain localities (the ten through which the pradakshinapatha traverses) are thya--in, and certain are ma thya--out. The boundaries are very clearly and precisely understood. Haugah, Caka Bahi, Saugah, and Ikha Lakhu, for instance are thya but Yanamugah, Kuti Saugah. Gasah, and Tangah are ma thya. It is those twah which are considered thya that today still have the rotating responsibility for Nyaku Jatra. Some residents are of the opinion that Tangah, Kuti Saugah, Yanamugah, etc. were not part of the city when the festival was inaugurated.

All the festivals that make use of the royal centripetal pradakshinapatha take place during the dark half of the month of Gula. (20) What is the significance of this? Tradition has it that Buddha himself chose this month for his worship because no other gods would. There is more to it, though, as most of the jatras that take place at this time, have nothing to do with Buddha at all. Sa Paru, Narasimha Jatra, Krishna Jatra, and Bhimsen Jatra in fact are unashamedly non-Buddhist. The explanation must be sought not in ideology but in structure. This lunar fortnight follows hard on the heels of the first lunar festival of the year--Gatha Mugah Cahre. In that festival we see a symbolic enactment of the removal of all that is base and evil from the city. As such it is a negative festival. It is there to undo the worst effects of that period of chaos, which resulted in the collapse of the urban order during the first and heaviest half of the monsoon. With the festivals of Gulaga (Bhadra-krishna paksha) we have the beginning of the positive reconstruction of that order. It is as if the people were saying that urban unity is of utmost importance.

I suspect that the royal centripetal jatras of Gulaga (Bhadra-krishna paksha) are no older than the beginning of the Malla period, unless older jatras changed their route to conform. It seems highly likely to me that King Siddhi Narasimha Malla established the royal centripetal pradakshinapatha in the seventeenth century.

Large urban festivals both express and create cohesion in the city. As Clifford Geertz (1973: 93) remarks:

 Culture patterns have an intrinsic double aspect; they give
 meaning, that is objective conceptual form, to social and
 psychological reality both by shaping themselves to it and by
 shaping it to themselves.

Or, as Levy (1990:199) puts it for Bhaktapur, the town's spatial divisions both give meaning to an take meaning from their special divinities, symbolic enactments and their associated legends and myths.

Convoluted centripetal processions: Gutschow's fourth type of procession is, for want of a better term, a 'convoluted' one but one that is nevertheless centripetal. This kind of procession is exemplified by Lalitpur's Mataya festival (Gutschow 1982:190, fig. 226). Such a procession is integrative in that it circumambulates city space but not royal in the sense that Sa Paru is, as it pays no attention to the central space itself. Mataya is, in fact, the archetype of such processions as its modern analogues to other deities demonstrate. The purpose of such festivals, which includes the older Bahah. Puja as well as the modern Krishna Puja and Bhimsen Puja, is for each of the participants to visit each and every shrine of a particular deity within city space. In the case of Mataya this involves thousands of devotees walking, often barefooted, around the city on an extremely convoluted, but for all that ordered, route to each of the Buddhist votive shrines called caityas (or in more colloquial Newar, cibhahs) (cf. Gutschow 1982:170, fig. 190). During the course of the procession the devotees also circumambulate the tour mounds (thur or stupa) that are positioned at the four cardinal directions around the city. To reach three of these mounds it is necessary to make a detour, as it were, outside the old city boundaries as they were never encompassed by the old city's urban encroachment.

The procession of Bahah Puja, which takes place the day before Mataya, visits each of the dozens of Buddhist monasteries around the town. In doing so, the line of devotees is forced to cross the centripetal pradakshinapatha at Keshav Narayana Cok, the northernmost courtyard of the royal palace. This is a most interesting and telling phenomenon as, at precisely the same time, the self-consciously shivamargi procession of Sa Paru is making its way around the city on the centripetal pradakshinapatha.

The line of Buddhist devotees then has to cross the procession of Sa Paru. The reason for this is historical. In order to allow for the northward extension of the palace compound (with the building of Keshav Narayana Cok) the Malla kings of the early seventeenth century were compelled to uproot the monastery of Hah Bahah from its ancient site in Mangah. It was then transplanted 400m westward at its present site in Ga Bahah. Not to be intimidated by such an act of royal aggression, the devotees at Bahah Puja continue to offer puja to Hah Bahah at its older site by placing fruit and other offerings on the step of the courtyard's golden door (Plate 2). It is this continued act of defiance that leads to the two processions to continue to cross each other's path. What was truly amazing to this writer was to observe the two processions going on in this way, without the "slightest interest in, or even, acknowledgement of, the other's existence.

Mataya, though not following the route of the royal centripetal pradakshinapatha, nevertheless is related to it in three ways: firstly, the procession is organised by a committee, on a rotational basis, from one of the ten localities (twah) on that route; secondly, the Nyaku Jatra--the procession of musicians that plays on the same day as Mataya--does circumambulate the city on the centripetal route; and thirdly, it is also a clockwise integration of city space. What then is the relationship between the two routes? It is apparent that the royal centripetal pradakshinapatha is an invention of the Malla kings, and that Mataya was too well established, and too closely followed to allow the kings to change it. They did, however, bring it under some sort of central control by delegating the responsibility for its organization to the ten localities along the integrative route and ordering the Nyaku Jatra to play along that route simultaneously.

Excursive centripetal processions: Power-place worship, which involves procession to each of the Twelve True Power-places in turn, is also a kind of centripetal procession. (21) The difference with this procession, however, is in the fact that this procession takes place, for the most part, outside the city. Each day a procession forms at the city centre, Mangah, and makes an excursion out of the city to the particular shrine of the day.

Having done the puja, the procession retraces its steps to the palace (cf. Gutschow 1982:173, fig. 194). The integrative character of the procession is in the set as a whole. This is reemphasized on the ninth day of Mohani, (Syako tyako) when all twelve Power-places are visited at one go.

Sanctifying city space

The location of deities in and around the city, and the movement of deities or, alternatively, the movement of their devotees through city space all have a symbolic significance. In relation to the Navadurga of Bhaktapur, Gutschow and Basukala (1987:147) suggest that the gods enact an elaborate set of processions which aim, in the broadest sense, at ritual taking possession of space, of a realm sacred to and sanctified by these gods:

 The appearance of the Navadurga in the streets and lanes not only
 serves to substantiate the actual presence of the gods in town. It
 is in fact much more: the gods come to each and every quarter as
 if to prove that they form the component parts of the whole, the
 town of Bhaktapur with its heterogeneous spatial and social
 structure. Thus the appearance of the gods confirms and reaffirms
 the special quality of an urban as opposed to the rural
 environment. The gods represent the essence of the urban
 environment (ibid.: 155).

Conclusion

Although the bounded nature of settlements seems to have been thoroughly established in ancient times this does not seem to he true of their centredness to any great extent. The Malla kings, as masterful politicians, however, took full advantage of the sacred texts that sanctified the city as a mandala. Although the centrality of the palace was already an accepted notion they took it to a new plane. The value of centredness became equivalent to that of kingship. All roads in ancient Europe may have led to Rome, but in medieval Lalitpur they led, ideally at least, to the palace. Ritual movement through sacred city space reinforced such conceptions.

References

Barre, V., P. Berger, L. Feveile and G. Toffin. 1981. Panauti: une ville au Nepal. (Collection Architectures.) Paris: Berger-Levrault. (French).

Geertz, C. 1973. The Interpretation of Cultures. New York: Basic Books.

Gellner, D.N. 1984. "Cities and Mandalas (Reviev of Barre et al.)" CNS 12: 115-126.

--1992. Monk, Householder and Tantric Priest, Newar Buddhism and its Hierarchy of Ritual. Cambridge: CUP.

--1996. "A Sketch of the History of Lalitpur (Patan) with Special Reference to Buddhism". CNS 23: 1.

--1997. "Does symbolism construct an urban mesocosm?": Robert Levy's Mesocosm and the question of value consensus in Bhaktapur. IJHS 1 (3): 541-64.

Gellner, D.N. and R.P. Pradhan. 1995. "Urban Peasants: The Maharjans (Jyapu) of Kathmandu and Lalitpur". In D.N. Gellner and D. Quigley (eds.) Contested Hierarchies: A Collaborative Ethnography of Caste among the Newars of the Kathmandu Valley, Nepal. Oxford: Clarendon Press, pp. 158-185.

Gutschow, N. 1982. Stadtraum und Ritual der Newarischen stadte im Kathmandu-Tal: Eine architekture-anthropologische Untersuchung. Stuttgart: Kohlhammer. (German).

Gutschow, N. and G.M. Basukala. 1987. "The Navadurga of Bhaktapur--Spatial Implicatioins of an Urban Ritual." In Gutschow and Michaels (eds.), Heritage of the Kathmandu Valley: Proceedings of an International Conference in Lubeck. Sankt Augustin, GFR: VGH Wissenschaftsverlag.

Gutschow, N. and B. Kolver. 1975. Ordered Space, Concepts and Functioins in a Town in Nepal. Wiesbaden: Franz Steiner.

Ishii, H. 1978. "Agricultural Labour Recruitment among the Newars and other groups in the Sub-Himalaya Areas". In Charles Romble and Martin Brauen (eds.), Anthropology of Tibet and the Himalaya. Zurich: Ethnological Musemu of the University of Zurich.

Jameson, M.H. 1997. "Sacred space and the city: Greece and Bhaktapur". IJHS 1(3): 485-99.

Landon, P. 1976 [1928]. Nepal. (Bibliotheca Himalayica 16.) Kathmandu: Ratna Pustak Bhandar.

Levy, R.I. 1990. Mesocosm: Hinduism and the Organisation of a Traditional Newar City in Nepal Berkeley and Los Angeles: University of California Press.

Locke, J.K. 1985. The Buddhist Monasteries of Nepal: A Survey of the Bahas and Bahis of the Kathmandu Valley. Kathmandu: Sahayogi.

Oldfield, H.A. 1981. Sketches from Nepal. 2 vols. Delhi: Cosmo Publications. (First published 1880.)

Sekler, E.F. 1987. "Urban Design at Patan Durbar Square: A Prelimainary Inquiry" In Gutschow and Michaels (eds.).

Sharma, Kunu. 1961. Kirtipataka Yogi Naraharinath (tr.). Lalitpur: Jagadamba Prakashan. (VS 2018) (Sanskrit, Nepali.)

Shrestha, C.B. and U.M. Malla. 1971. "Urban Centres of the Kathmandu Valley." The Himalayan Review, 4: 33-39.

Slusser, M.S. 1998 [1982]. Nepal Mandala: A Cultural Study of the Kathmandu Valley. 2 vols. Kathmandu: Mandala Book Point.

Toffin, G. 1991. "Urban Space and Religion: Observations on Newar Urbanism". In Toffin (ed.) 1991. Man and his House in the Himalayas: Ecology of Nepal. New Delhi: Sterling. (First Published in French 1981.)

Vajracharya, Dhanavajra. 1964. "Mallakalama desharakshako vyavastha ra tyasaprati prajako kartavya". Purnima, 1 (2): 20-30. (VS 2021) (Nepali)

Vergati, A. 1995. Gods, Men and Territory: Society and Culture in Kathmandu Valley. New Delhi: Manohar Publishers and Distributors.

Witzel, M. 1997. "Macrocosm, Mesocosm, and Microcosm: The persistent nature of 'Hindu' beliefs and symbolic forms". IJHS 1 (3): 501-39.

Notes

(1.) Gutschow's work is in German but the maps are self-explanatory so the reader without German is not at such a disadvantage as would first appear.

(2.) cf. Dhanavajra Vajracarya (1964).

(3.) The fact that some Khadgis also lived outside the old wall indicates that perhaps the distinction between Untouchable and merely Water-unacceptable was at times hazy.

(4.) The data for this map is partially gleaned from Gutschow (1982:162, Map 176) and from the map prepared by Nutan Sharma and published in Gellner and Pradhan (1995: 275, Fig 9.2). I have interpolated the position of the wall a little differently. See also Gutschow and Kolver (1975:49) and Gutschow's maps in Levy (1990:164 & 179) for a similar mapping of Bhaktapur.

(5.) Crossroads are usually considered inauspicious in South Asia. They are the abode of evil spirits such as the reverse-footed kicikini.

(6.) This list came out of a hard won consensus after a long discussion involving Nem Krishna Tamrakar and a group of elderly Maharjan men at Haugah.

(7.) Gutschow (1982:163, Map 177) shows the locations of Lalitpur's Eight Mother Goddesses.

(8.) I am grateful to Nutan Sharma, Dil Mohan Tamrakar and Nem Krishna Tamrakar for helping me prepare this list.

(9.) Some of the names and positions of the Dashmahavidya in Gutschow's map are different (1982:165, Map 183):

(10.) The proximity of these two deities leads to the common but mistaken belief that the Mahavidya is Bal Kumari.

(11.) This also fits with the existence of the Lineage Deity of the Pahma Sikahmi (and Pahma Shrestha) at Thapahiti just along the road east of Caka Bahi, which must have been outside the city at the time it was established.

(12.) A fourth cremation ground is used exclusively by the Dyahla of Thati and is located about 400m southwest of their settlement.

(13.) Olfield's report that the majority of Lalitpur's cremations took place at Bal Kumari seems unlikely (1981: 125).

(14.) Gutschow's map is inaccurate (ibid.: 163, Figure 177). He draws strict boundaries so that the inhabitants of Kuti Saugah, for instance, go to Bal Kumari. In fact, though Maharjans of that locality cremate their dead at Bal Kumari, the Sikahmi go to Shankhamul via (Cyasah on a circuitous route (Map 4). Likewise, the Barahi of Yanamugah.

(15.) This devotion was clearly demonstrated after the palace massacre of 2001 when thousands of men shaved their heads as if in mourning for their father.

(16.) Gutschow was apparently unaware of the Vishnu Devi Jatra. The two jatras of Mahalakshmi (Gutschow's map only shows one) have, since Gutschow wrote, become defunct. Gutschow suggests that each of these Jatras traces a route around the territory corresponding to that ashtamatrika. An analysis of the corresponding localities in terms of the cult, however, reveals that this is, in fact, not the case. I am not aware of an analogous procession for any of the other of the Eight Mother Goddesses.

(17.) Gutschow's list of stopping places on the pradakshinapatha is not wholly accurate (ibid.: 168, fig. 188 caption).

(18.) For many residents this is the pradakshinapatha.

(19.) The term as it is used in Lalitpur, however, seems not to have the connotation of chronological precedence that Ishii reports for Satungal.

(20.) There is an exception to this in that small groups of men also use the route on Swaya Punhi which, being a modern festival in the Valley, does not follow traditional rules.

(21.) Sometimes Pitha Puja may be more informal with no fixed number of Power Places visited.

Table 1: The Twelve 'true' Power-places of Lalitpur (6)

Batuk Bhairava, Lagankhel
Mahalakshmi (*), Tasi (S)
Kumar Raja, Kusunti
Vishnu Devi (Vaishnavi) (*), Yappa (SW)
Brahmayani (*), Neku
Maheshvari (Rudrayani) (*), Palinka
Indrayani (*), Lwahagalah (NW)
Ganesh, Hasapatah
Dhuma Varahi (*), Dhantila (NW)
Camunda (Sikali)(*), Sikabahi, Shankhamul (N)
Bal Kumari (*), Kwache (E)
Siddhilakshmi (Siddhi Caran) Bangi

(* denotes Mother Goddess, matrika) (7)

Table: 2 The Ten Great Knowledges (Dashmmahavidya) of Lalitpur (8)

Tripurasundari, Bal Kumari, Caka Bahi
Swet Kali, Ta Bahah
Bhairavi, Bahalukha
Kamala, Purnacandi
Vajrayogini (Tara), Puco
Bhuvaneshvari, Nakah Bahi
Bagalamukhi, Kumbheshvara
Dhumavati, Dhum Bahah
Matangi, Swatha
Chinamasta, Mamadu Galli

Table 3: A typology of Processions

Centrifugal Ways of the dead (Silapu)
 Gatha Mugah
Linear Mahalakshmi Jatra
 Bal Kumari Jatra
 Vishnu Devi Jatra
 Bugadyah Jatra
 Kwenadyah Jatra

Centripetal Royal Sa Faru
 Nyaku Jatra
 Mataya Ganesh Puja
 Narasimha Jatra
 Krishna Jatra
 Bhimsen Jatra

 Convoluted Bahah Puja
 Mataya
 Krishna Puja
 Bhimsem Puja
 Dashmahavicya Puja
 Vasundhara Puja
 Excursive Pitha Puja