Chapter Twelve
The Civic Ballet: Annual Time and the Festival Cycles

Introduction

To bring Bhaktapur's symbolic order to coordinated life, a temporal system and a patterning of events within the tempos of that system are necessary. There are in Bhaktapur two large classes of symbolic enactments that tie together public space, individuals, social units, deities, and time into a larger assemblage. In these enactments, the various matters that we have dealt with in earlier chapters become elements in a civic performance . Social roles, significant space, the complexes of meaning represented by deities, priests, and modes of worship emerge and become realized m these performances, which, like their constituent aspects, follow traditional patterns. The first class of these civically significant enactments are the rites of passage, the samskara s (app. 6), whose sequence is determined—often with considerable leeway in their exact timing—by the stages of the life cycle, fine-tuned through astrological considerations. In Bhaktapur these rites entail references beyond the individual and household to the patrilineal extended family, the phuki , and beyond that to some larger civic units (primarily the twa : and the mandalic^[*] area pitha ). But the samskara s’ central importance is in defining the individual in relation to household, extended family, and, with marriage, to an allied kin group within his or her status level. Their relation to the larger city is secondary, and for the most part simply emphasizes the phuki ’s relation to neighborhood and Mandalic^[*] Segment, as all phuki worship does. It is the second class of temporally

coordinated ritual enactments that are the true civic enactments. That group is primarily constituted by those activities that take place in the city following the dictates of the calendar. These include practices that are appropriate—in many cases necessary—for particular days of the week, of the lunar fortnights or of the solar month. However, at the center of our concern in this study are those events whose occurrence depends on the annual cycle itself. This particular temporal set, the yearly events, has special urban emphases and uses in contrast to the smaller and larger cycles.^[1]

Many of these annual events are associated with feasts encompassing one or another social unit, and are designated as nakha cakha (Kathmandu Newari, nakha: cakha: ), literally "to feed and associated activities."^[2] "Nakha cakha " may be glossed with appropriateness by "festival," with that word's connections to "feast" and "festive." The term "festival" also works well for some of the other major and public events of the year, particularly if "festival" is extended to include some restrained, minor, or routine "celebrations." However, there are other significant annual events—a day for the protective rubbing of bodies with oil, a day on which the moon should not be viewed, and so on—for which the term is inappropriate. Hindu calendrically determined events include two sorts of events that, although distinguished by classical terminology, are blurred in actual usage. These are vrata events and utsava events. "Vrata " implies a "religious duty," and is used often in Bhaktapur in the strong sense of "religious or ascetic observance taken upon oneself, austerity, vow, rule, holy work such as fasting and continence" (Macdonell 1974, 304). The other term, "utsava, " indicates, traditionally, "festival or holiday." Gnanambal, in a report on surveys of Indian "festivals," notes that the term "vrata " has a wider usage in many parts of India to include "festivals," especially where fasting is a necessary element (Gnanambal 1967). There is, nevertheless, in usage, he adds, a "faint distinction" of vrata and utsava as evidenced by the presence of the two terms in many parts of India. Kane (1968-1977, vol. V, p. 57) also remarks on the difficulty of using the termino-logical distinction for discriminating actual events.

The annual events we will consider contain among them elements of vrata and utsava and sometimes of neither, and we will use various glosses for them, all meaning no more than "calendrically determined event of general civic importance." During the course of the year in Bhaktapur there are some seventy-nine of these, and as on some days there are more than one and as, in contrast, a few last for two days,

there are a total of seventy-four days each year during which some part or all of the city is involved in one or another such event. This is in addition to weekly and monthly activities during the course of the year, as well as the pilgrimages and mela s or "fairs," taking place elsewhere in the Valley (or beyond) every year or every few years in which people in Bhaktapur may participate.^[3] These gross enumerations are, however, misleading as our discussions in later chapters and a more refined enumeration in chapter 16 will indicate. Later chapters will provide a clearer view of the types and density of calendrical events in Bhaktapur.

The calendrical events, derived from South Asian tradition and the Kathmandu Valley's long history, are, like the supernatural members of the city's pantheon, of interest to us primarily not as a collection of witnesses to that history but as aspects of an ongoing, meaningful contemporary urban life. We will say something, if only in passing, about each calendrical event, but we will treat certain events at much greater length, and among these we will be most particularly concerned with what we call the "Devi cycle" and its climaxes in Mohani, the Autumn Harvest festival sequence, and in the related performances throughout nine months of the year of the Nine Durgas troupe. The events we emphasize are, evidently, those we take to be of particular integrative importance.

The Calendar

Bhaktapur, typically of South Asia, has both a solar annual calendar and a lunar one.^[4] While the great majority of festivals are determined by the lunar calendar, there is one major festival sequence (Biska:, the solar New Year sequence) and one other annual event which are determined by the solar cycle. The lunar year is normally divided into twelve lunar months. The lunar month begins on the day following the new or dark moon, which ends the previous month.^[5] The Newar lunar month is divided into a first "bright" half, corresponding to the waxing of the moon, and ending with the full moon; and a second "dark" half, corresponding to the waning phase of the moon and ending with the new moon. The bright fortnight is called tumla ; the dark fortnight, khimla .^[6] To designate individual lunar fortnights, terms for the dark half and light half of the lunar month—ga and thwa , respectively—are added to the name for the particular month. The name of the month is itself a compound including the morpheme la , meaning lunar month. Thus the

first lunar fortnight of the lunar year, Kachalathwa, means the bright or waxing fortnight (thwa ) of the lunar month (la ) of Kacha, which is followed by Kachalalaga, the dark fortnight (ga ) of the month of Kacha. Table 2 lists Bhaktapur's lunar fortnights, the Sanskritic equivalent months, and the approximate correspondences of the fortnights to the Western year. The table does not include the intercalary period, which has to be added every third year to adjust the lunar to the solar year, and which does not usually affect the ritual calendar.

The full-moon day, which ends the bright lunar fortnight (the first half of the month), is called punhi , and the dark or new-moon day, which ends the dark fortnight is called amai^[7] The other days of each lunar fortnight are given ordinal numeric Sanskritic names, with the

exception of the fourteenth day of the dark fortnight, whose Sanskrit name, caturdasi , is usually replaced by the Newari term "ca:re ". The lunar days are called tithi . Although they vary in length (from about 21.5 to 26 hours), they are, for most purposes, made to correspond to ordinary solar days.

The solar year, which is of very much lesser importance for the ritual cycle, contains twelve months. The names of these months are the same as for the Lunar months,^[8] although there is, perhaps, a tendency to use the Sanskritic forms more for them. The days of the solar year, gate , begin at sunrise. They are arranged in a seven-day week deriving, as the solar calendar in general does, from the same sources thus sharing some of the same astral references as the Western days of the week. The occasional necessity of relating the solar and lunar years requires a complicated set of rectifying conventions that are not relevant here.

Approaches to Meaning

In the next three chapters (chaps. 13 to 15) we will describe the annual calendrical events. We will look for aspects of form and thematic content, and for similarities and contrasts that contribute to the meaning-fulness of clusters of and subcycles of calendrical events, as well as of the entire annual collection. We will introduce here some of the issues and approaches that will concern us in our detailed presentation of the festival year.

Cycles

There are many ways of sorting Bhaktapur's calendrical events. Our first rough sorting has been into events of the solar year and of the lunar year, and then a further division of the lunar year into one set that constitutes a clearly interrelated and extremely important group, the "Devi cycle," and another, residual, lunar group. The solar calendar has only one important festival sequence, Biska:. The lunar cycle as a whole (as opposed to the clearly integrated Devi cycle) seems on the surface at least to be a mixture of miscellaneous events. We will, however, be concerned with its deeper patterns insofar as they may—or may not—exist.

Marc Gaborieau (1982) has proposed a "structure" for the Indo-Nepalese calendar that, although problematic for the Newars at least,

provides a useful point of departure for a discussion of the meaning of the overall lunar cycle. He notes that some writings on Hindu time (he cites Mus [1932-1934] and Zimmer [1951]) represent it as having the following properties: (1) each of the different divisions of time (days, years, cosmic periods, etc.) is arranged in a cycle; (2) these cycles are formally similar; (3) each cycle has a beginning and an end; (4) each cycle has a movement from order toward disorder and culminates in a state of chaos that precedes the regeneration, which will mark the beginning of a new cycle; and (5) the stages of chaos and regeneration are not considered parts of the temporal cycle, but temporary escapes from time. They represent the "axis which communicates with eternity."

Gaborieau argues that this schema is reflected in the Indo-Nepalese festival cycle. For the Indo-Nepalis, he writes, the four-month period beginning with the summer solstice—the period of the monsoon and the major work prior to the rice harvest—are considered inauspicious months, but it is also the period for the majority of the year's festivals. The period begins with the festival of Hari Sayani, the time when Visnu^[*] goes to sleep for a period of four months "leaving the earth to the demons." In Gaborieau's speculation, those months are out-of-ordinary secular time. The first two months correspond to the period of disorder, the second to the period of "regeneration." The festivals during this period "manifest radical disorders and reversals followed by profound restorations of order" (1982, 16). In contrast, he argues, the eight-month period beginning with the first winter month (Marga in November/December) is an auspicious period where life follows its normal course, and household ceremonies for good luck, and prosperity, and the like, take place as do lineage ceremonies. He further argues that those inauspicious events, the disorders and reversals, that do take place during this period primarily concern the lower castes. The "cyclical mystery" of the year as expressed in its festivals is the privileged experience of the upper, twice born castes.

How far this schema is adequate for the sorting of the Indo-Nepalese festival calendar is for others to judge; the involution of festival practices, the relative secularization of most of the Indo-Nepalese groups, will make an anthropological critique difficult. In its details, this schema does not work for Bhaktapur's calendar, but for certain aspects of that calendar and with different timings, it will provide (in chap. 16) a useful point of departure for an analysis of the possible implications of the arrangement of all the components of the annual cycle within that cycle.

Whatever the internal structure of the overall annual calendar of events may be, there is the partially related question of how the meanings of the events might be affected by external cyclical events of a different order. Is there any relation of the meanings of festival events to the phases of the moon and to the sun's course and seasons beyond their clock-like uses in coordinating the cycle? The kinds of data we deal with show only scanty echoes of these cosmic events. But when we consider still another external cycle (in turn, dependent on the seasons), the rice agricultural cycle, particularly in its relation to the Devi cycle, the connections are of central importance.

Selection from the Hindu Set of Festivals

Although most of the elements of Newar symbolic life are taken from the inventions and developments of South Asian history—supplemented by some significant bur quantitatively minor Newar and Himalayan forms—there is, as we have seen in relation to the urban pantheon, a necessary selection among these elements. There are quantitative considerations—only a certain number of elements can be understood and put to use in the civic system. There are also considerations of propriety; some forms do not fit in, or are redundant, have their places filled, as it were, by other symbolic elements. As is the case with all inventories of South Asian possibilities, the list of calendrically anchored events noted in the classical literature and religious texts is very large. Kane has what is presumably an almost exhaustive list of calendrical vratas and utsava containing well over one thousand events throughout India's vast history and extent (1968-1977, vol. V, pp. 253-462). These vary in their general importance and occurrence through out historic time, space, or class of devotees.^[9]

Bhaktapur's calendar selects and rejects from this group, and invents—or often pieces together from existing fragments—its own festivals. The most salient contrast of the Bhaktapur calendar for Bhaktapur's citizens is with other Newar calendars and with Indo-Nepalese calendars. Not only does the presence or absence of calendrical events in Bhaktapur reflect an active selective in relation to other calendars, but so, and often more significantly, does their local importance. Thus a festival of general South Asian importance may be present in Bhaktapur, but only in some residual and unimportant manifestation. We will consider the questions of selection and emphasis in the following chapters.

Aspects of the Analysis of Calendrical Events

When we narrow our horizon to look at Bhaktapur's various calendrical events in themselves and in their similarities and contrasts to each other, we must seek appropriate and relevant aspects for analysis and contrast. We must attend to the social and spatial units involved. What are the static and dynamic uses of those units? Which deities are made use of? Is the deity or deities moved; do people move? Where? To what purpose? Who are the human actors and audiences? What are the different sorts of actions? What are the themes and narratives portrayed and recounted? Are there narrative "plots," with conflicts, tensions, climaxes, resolutions? What kinds of symbolic forms and rhetoric are used to contribute to meaning? What kinds of themes are there in various events? What problems seem to be dealt with? How do participants seem not only to act in but also to respond to particular calendrical events? How are various city units tied together—through "parallel" devices (with various units doing the same sorts of things at the same time) or through "serial" or "interactive" devices, with some sort of meaningful movements and encounters systematically interrelating different kinds of actors and social units in the course of the event?

Such questions are in the background of our considerations of calendrical events, but we have not dealt with these issues explicitly in relation to all the calendrical events noted in the following chapters, for many minor festivals many of them are irrelevant. These various elements of festival meaning become fully relevant only in the more developed festivals, those that are more important to Bhaktapur by various criteria, which we will present in the following chapters.

A catalogue of their potential resources for generating and expressing order and meaning, in fact, is liable to make the annual events seem more exhaustively integrative and constitutive of the city's symbolic system than they really are. That task falls on selected ones. The question of which potential resources are, in fact, used or neglected by particular individual events and throughout the course of the annual cycle is an empirical one.

We will see that the events vary in importance from "trivial" to what we call "focal" events, events of central importance to the city,^[10] and we will make an estimate of the relative importance of the various calendrical events as being of minor, moderate, or major importance. In the next three chapters we will lose ourselves among the trees of the annual cycle. In chapter 16 we will return to the view of the forest, and

the consideration of its differential contributions to urban order. What do these annual events do for Bhaktapur and its people? How do they do whatever it is that they do?

The Inclusion and Sequential Numbering of Calendrically Determined Events

Some arbitrary decisions regarding two issues were necessary. First, when is a day of the year that is given a particular name really to be considered an event of some civic importance and thus to be listed? In some cases, for example, a particular full-moon day, the day may be given a special differentiating name, but the activities characterizing it are no different than for other such days, the day being an unexceptional member of the, for the most part, internally undifferentiated set of twelve full-moon days. A particular day may have some special differentiating feature of very minor present importance, or be of interest in only a personal and optional way to some few individuals or households, or to some particular thar . It has been optional in some few cases whether to include or exclude an event; we have usually (but not consistently) excluded it unless there is some clear suggestion of civic importance according to our criteria. The minor processions of certain deities that we include and list among the "minor" festivals are also of very marginal importance, in this case because they have little or no present following or attendance—but they are clearly public urban events dictated by the calendar, echoes sometimes of events of some past importance and possible nuclei, in some interesting cases, for some future efflorescence.

A second problem is how to distinguish separate events for enumeration within some larger festival sequential complex, such as the autumn Mohani festival or the solar Biska:. We have dealt with this in somewhat different ways in those two cases, in large part following local conventions.

In should be noted that the sequential numbers that we use to designate festivals in the following chapters are derived from the sequence of the entire group of annual festivals starting with the lunar New Year's Day. Thus the numbering of the events in the three individual cycles indicate their position in the overall annual collection of events. (The annual festival calendar is given in summary form in app. 5.)