Navigation – Plan du site

AccueilIssues57TranslationFestivals, time and space: the st...


Festivals, time and space: the structure of the Indo-Nepalese version of the Hindu calendar

Marc Gaborieau
Traduction de Matthew Cunningham


This article provides an interpretation of the annual cycle of festivals among the Hindus of the hills of Nepal. After recalling the main technical data of the Hindu calendar, the author develops his approach in two stages. Through an ethnographical inventory of festivals, he finds out the main divisions of the annual cycle. Using then Indian theoretical texts, he reconstructs deductively the conception of time which underlies this cycle of festivals. This sacred time is analogous to sacred space: the ‘Four Months’ of the Rainy Season (July-October), during which all important festivals take place, are like a third vertical dimension, the dimension of eternity, which brings men into contact with gods at the end of each annual cycle.

Haut de page

Entrées d’index

Haut de page

Notes de la rédaction

This is a translation of an article published in 1982 as ‘Les Fêtes, le temps et l’espace : structure du calendrier hindou dans sa version indo-népalaise’ in L’Homme 22 (3): 11–29. The original version is accessible on the open-access Persée portal and we are grateful to them for allowing us to publish this translation. We remind the reader that we have respected the style of the original, except for references that needed to be reformatted and corrected or where the author has provided minor modifications.

Texte intégral

The sacred times and spaces in which rites are performed and myths created are qualified to receive them. The spaces are true temples. The times are festivals.
Marcel Mauss, Œuvres I: 30.

1The ethnologist is ceaselessly confronted with festivals, which offer an often spectacular illustration of social structures. Each of them forms a whole that could be the subject of a monograph. But is there an order presiding over their sequence on the annual calendar? In other words, do they constitute a cycle? I will consider this question for South Asia, by more particularly examining the Hindu calendar.

2Two types of source are used. First, ethnographic observation. Hindus (and Buddhists to a large extent), have a common basic calendar; but every region, and often even every ethnic group, has developed its own variant over the centuries. Here we will not be getting lost in the analysis of these multiple versions. To keep the structure simple, we will limit ourselves to one region, the middle mountains of Nepal; and within that region, a single ethnic group, the Indo-Nepalese (Gaborieau 1978a: 213–233), to the exclusion of the Newar (Toffin 1979). I will base my analysis on work by Western and Nepalese colleagues, as well as on my own observations on Nepal, drawing comparisons with various North Indian regions if necessary (Stevenson 1971: 262–342; Tripāṭhī 1971). The second type of source used: ancient books containing theoretical data on the organisation of time. Even though these come from fields as diverse as astrology, astronomy, rituals, architecture and iconography, one finds in them a remarkable convergence of views.

3My study will be divided into three parts: an overview of the broad features of the Hindu calendar, necessary for understanding my argument; an inventory of the festivals that take place in the course of a year; finally, a summary that aims to bring out the structure of the annual cycle.

1. The Hindu calendar

4The Hindu calendar is lunisolar (cf Bīrūnī 1964, I: 319–388; Renou and Filliozat 1947).

1.1. Solar computation

  • 2 To facilitate possible comparison with other regions of the Indian subcontinent, I have most often (...)

5The major divisions of the calendar follow the course of the sun. The year, varṣa2, which serves as a reference, is a solar year of three hundred and sixty-five and one quarter days; it is divided into twelve solar months, māsa, of thirty days or more, their beginning corresponding with the sun’s theoretical entrance into one of the zodiacal signs. This first day of each month is called saṅkrānti.

6The year includes two intermediary divisions. It is first divided into two equal periods, anaya, of six months each. The one in which the sun, starting from its lowest point, moves towards its highest point, towards the north, uttar, is called uttarāyaṇa; it begins on the winter solstice, whose date, for lack of corrections, falls around 13 January: this is the rising period. The descending period, corresponding to the sun’s displacement from its highest to its lowest point, towards the south, dakṣiṇa, is called dakṣiṇāyana; it begins on the summer solstice around 13 July. These two periods mark the major beats of the annual rhythm.

7Next comes a finer division into seasons, ṛtu, of which there have been six since the Early Middle Ages, that is, three for each of the rising and descending periods. Each season therefore lasts two months.

8Among the festivals, only those marking the beginning of a solar month—in particular the two solstices—have their date set based on the solar computation.

9Table 1 sums up the divisions of the solar year.

Table 1: The Hindu calendar

Table 1: The Hindu calendar

The circle representing the year is divided into equal parts successively indicating the two periods (at the centre), the six seasons (middle ring) and the twelve months (outer ring). Each subdivision shows the Sanskrit term (generally used here), the Nepali term (if it exists), the English translation, and finally reference points on our own calendar.

1.2. Lunar computation

10The date of most festivals depends on a lunar computation that is combined with the solar calendar. The year is divided into twelve lunar months, each corresponding to an observed lunation; they have the same names as the solar months. The lunar year is shorter than the solar year by nearly ten days (355 ½ days compared with 365 ¼ days); the shift is corrected by adding an intercalary month every three years that contains no festivals. The Hindu lunar year thus follows the rhythm of the solar year and stays in harmony with that of the seasons (unlike the Muslim calendar). Table 1 therefore also applies to the lunar calendar.

11Every lunar month is divided into two periods, the two fortnights, pakṣa, corresponding to the lunar phases: the dark fortnight, krṣṇa pakṣa, from the day after the full moon to the new moon, amavasya (the ‘no-moon day’) and the bright fortnight, śukla pakṣa, from the new moon to the full moon, pūrṇimā. As the vocabulary indicates, the dark fortnight is of ill omen, and is dedicated to the worship of demons and the dead, while the bright fortnight is dedicated to divinities of good omen. The fortnight chosen to mark the beginning of the month varies depending on the region and ethnic group; the Indo-Nepalese currently start with the dark fortnight. Regardless of that choice, the days of each fortnight are numbered from one to fifteen; the dating of festivals according to the lunar computation are as follows: nth day of the dark (or bright) fortnight of such-and-such month.

12In its essence, this lunisolar calendar is not intended to set the rhythm of the civil year: depending on the region, ethnic group and dynasty, the new year can be placed in any season (cf Bīrūnī 1964, II: 8–9). Neither does it revolve around agricultural work. It is above all a religious calendar.

2. Inventory of festivals

13To determine the major divisions on this religious calendar, I will proceed by induction, establishing an inventory of the festivals that punctuate it. Since it is not known where the beginning of the year is located, nor even whether or not it has a defined beginning and end, I will arbitrarily start with the descending period.

  • 3 For more details, see the excellent calendar recently published (Bouillier 1979: 58–80) and the ar (...)

14This review will be brief for several reasons. From the exhaustive inventory, I will only retain festivals that seem relevant to the subject at hand. I have left out those whose law of distribution over the course of the year does not clearly appear: the fairs, melā, that attract crowds of pilgrims to sacred sites on a specific date, as well as the birthdays of numerous Hindu divinities. Also disregarded are the innumerable fasts, vrata, that are performed especially by women, and have been studied by V. Bouillier (Bouillier 1982). A mass of important festivals remains; for lack of space, I will limit myself to reconstructing their sequence and categorising the themes they illustrate3.

2.1. The descending period

15The descending period that begins at the summer solstice comprises two subdivisions of unequal duration: the tradition sets apart the first four months, caturmāsa (Skt.) or caumās (Nep.), which will henceforth be called the ‘Four Months’; they are followed by the two months of Winter.

2.1.1. The Four Months

16These Four Months correspond to the time of the monsoon and of major agricultural works. Hindus consider them inauspicious months of ill omen. This character already appears in the rites marking its beginning. In the lunar computation, it is the eleventh day of the bright fortnight of āsāḍha: one celebrates hari śayanī, the day when Hari, that is to say Viṣṇu, withdraws to sleep in hell accompanied by other gods, leaving the earth to the demons, particularly Bali. In the solar computation, it is the summer solstice, celebrated out of fear at nightfall through evil-expulsion rites; married daughters must visit their parents one last time before the monsoon, which could kill them. This inauspicious aspect is emphasised by taboos: family ceremonies of good omen—initiations and marriages—as well as ceremonies of prosperity and protection of the village and of the lineage cannot take place during the Four Months period; during the first two, aside from major purification rites, one must not bathe in the rivers, which are considered impure and dangerous at that time.

17However, most festivals occur during those Four Months. Two take place during the bright fortnight of śrāvaṇa: the fifth day, nāga pañcamī, is dedicated to the serpents that are the masters of the soil and rain; the fifteenth day, the full moon, named śrāvaṇa purṇimā, ṛṣi tarpanī and rakṣā bandhana (Skt.) or janai purṇimā (Nep.), respectively sees a partial return of the gods to earth, a purification of initiated men from the three upper classes of twice-born who renew their sacred thread, and protection rites (Macdonald 1962, 1972; Jest 1966). The third and fifth days of the bright fortnight of bhādra, tīj and ṛṣi pañcamī, are marked by the women’s major annual festival (Bista 1969; Bouillier 1982; Tripāṭhī 1971: 151–170). In āśvina, the first ten days of the bright fortnight are dedicated to the festival of the greater dasaī (Nep.), also called daśaharā or durgā puja (Skt. and H.): under the aegis of the three great goddesses, Dasai marks the victory of the gods over the demons, and the re-establishment of political order (Gaborieau 1978b: 41–43; Toffin 1981). Finally, in kārtika, from the thirteenth day of the dark fortnight to the second day of the bright fortnight, the festival of divālī (H.) or tihār (Nep.) permanently repels the spectre of the demons and death, and ensures the prosperity of the family, business, or farm (Gaborieau and Helffer 1968–1969).

18These festivals are also the largest of the year. They manifest the most radical disorders and reversals, followed by the deepest restorations of order. These two phases are present in both of them; let us systematically place them in groups. The reversals, firstly, are found at all levels. At the cosmic level, the monsoon is conceived as a disintegration of the world, and one sees demons—whose invading presence is extensively evoked during Dasai and Divali—gain the upper hand over the absent gods. Then, the boundaries between the worlds of the living and dead disappear: it is believed that the dead are particularly close during the months of Asvina and Kartika; the dark fortnight of Asvina, under the name of either pitṛ pakṣa (Skr.) or sorā śrāddha (Nep.), is entirely reserved for their worship (Stevenson 1971: 326–329); likewise, Divali, or under its other name yama pañcaka, is partly dedicated to the worship of the dead and to their god Yama. Finally, the disorders reach into human society itself: during the month of Sravana, the lower castes turn arrogant and demand additional remuneration from their upper-caste bosses (Gaborieau 1977: 29); in the west of the country, untouchable jesters, bhāṛ, can even insult them in order extort money from them; the following month, during Tij, women challenge the authority of their husband and mother-in-law. The disorder attains its paroxysm during the two months of autumn; the thirty-two days that elapse from the beginning of Dasai to the end of Divali constitute a kind of interregnum: royal authority is challenged during those two festivals; no activities are undertaken (other than urgent agricultural work); people play games reserved for this period of the year: kites and swings (both linked to worship of the dead) and gambling; children, certain renouncers, and untouchable minstrels beg under religious pretexts.

19Each of these festivals then contributes to re-establishing order. At the cosmic level, the gods are gradually called back: at the Sravana full moon during Dasai, as well as during Divali, which sees Vishnu’s victory over Bali and the end of the demons’ reign; Vishnu ‘awakens’ to come back and rule over the earth on the eleventh day of the bright fortnight of Kartika, celebrated under the name hari bodhinī; at the end of the festival, the dead and their god Yama are totally cast out. Human society rebuilds itself little by little: during Tij, after having eaten well, sung, danced and been purified, women resume their role as submissive wives; the previous month, on the occasion of Janaipurnima, the three upper ‘classes’ of twice-born—by purifying themselves and renewing the sacred thread emblematic of their status—recovered their superiority over the low castes that had dared lift their heads. The last two festivals go even further: during Dasai, authority is reasserted in the family, lineage and kingdom; this festival places at the forefront political power and the Ksatriya who traditionally wield it. With Divali, the social and political dimension is implicitly recalled, emphasis being placed on family prosperity, along with farms and businesses, the Vaiśya’s traditional field of activity.

2.1.2. The two months of Early Winter

20Mārgaśira and pauṣa include no festival, strictly speaking. However, the atmosphere has changed radically. All ordinary activities have resumed. No one has the least suspicion of ill omen; it is an auspicious time. The celebration of initiations and marriages is starting again. In Margasira, collective festivals of prosperity and protection also take place. These are of two types: the first type are celebrated by the whole village community, which worships the gods of the soil and the goddesses of illness, to ensure the fertility of the earth and protect the health of people and animals (Gaborieau 1978b : 43–44); the others are a matter for each lineage, which honours its clan divinities, kuldevatā, and the god of livestock, Gaiṛu, in order to obtain abundant, healthy progeny, prosperous livestock, and wealth (Bista 1972).

2.2. The rising period

21The winter solstice seems to be the symmetrical opposite of the summer solstice. It is celebrated joyfully, no longer out of fear. After a bath and a salutation of the rising sun, which is beginning its ascending course, each house-headman throws a banquet; he invites his married daughters, to whom he gives gifts. But the symmetry stops there: the summer solstice marked a break between an auspicious period of normal activity and the beginning of the inauspicious Four Months; the winter solstice does not mark any break, and the whole of the rising period will have the same characteristics as the two preceding Early Winter months. It is an auspicious period, of good omen, that sees the uninterrupted pursuit of ordinary agricultural, commercial, administrative and political activities; one continues to celebrate initiations and marriages (except for during the final month of Asadha, which marks the transition towards the Four Months), as well as ceremonies of prosperity and protection (village ceremonies are spread over six months); lineage worship preferably takes place in vaiśākha or in jyeṣṭha, at the full moon.

22During all of this time, hardly three festivals take place, only the second of which has some importance, in North India more than in Nepal. These are: vasanta pañcamī, the fifth day of the bright fortnight of the month of māgha (Tripāṭhī 1971: 305 sq.); holī (Nep. and H.), the spring festival, from the eight to the fifteenth day of the bright fortnight of phālguna (Macdonald 1969); caite dasaī or little Dasai (Nep.), the eighth day of the bright fortnight of caitra. Can these festivals be compared to those of the Four Months?

23The first two should be grouped together, as indicated by the most detailed data concerning North India (Tripāṭhī 1971: 305–307), where there exists no discontinuity from one to the other; they take place during the season called Late Winter, yet they are Spring celebrations, as shown by the term vasanta and especially by the content of the celebrations. During Vasanta pancami, one venerates Sarasvati, goddess of knowledge—this is the most important rite for the Nepalese; one also worships the gods of love, Kāma and Rati. In North India, this date marks the beginning of preparations for the festival of Holi, which is a carnival; after having lit the new fire and offered the first fruits of the winter harvest, one gives free reign to disorders and reversals: the rules of decency are suspended in chants and dances; social hierarchies are reversed: women insult men and members of the lower castes fight those of the upper. In Nepal, rites involving excesses have disappeared: in the west of the country, men still chant and dance around fires; elsewhere, young people settle for insulting passers-by by sprinkling red powder on them.

24Little Dasai then marks the restoration of order. It is only a pale replica of autumn’s big Dasai. Only the eighth day is celebrated, and at a very small scale. In autumn, it was the entire edifice of the kingdom that was rebuilt; in the spring, official celebrations are limited to the royal palace and the seats of large administrations; it is the time often chosen for the beginning of the civil year (Caitra in North India, Vaisakha in Nepal). Little Dasai is above all a family festival during which sharecropping and service contracts are renewed with untouchable specialised castes (Gaborieau 1974: 327–329; 1977: 28).

25The festivals of the rising period are therefore not very remarkable, and even less so in Nepal than in North India. Disorders and reversals mainly affect the lower strata of society: it is traditionally said that Holi is the festival of the Śūdra. The re-establishment of order is superficial and primarily intended for the lower casts and the administration. Neither phase displays the depth and dimensions of the festivals of the Four Months period.

26This brief inventory leads to specific conclusions. The split into two equal periods—rising and descending—does not enable us to understand the division of the cycle of festivals. The real division is of another type; the only common point is the summer solstice. This solstice, in which the rising period gives way to the descending period, also marks the beginning of a short division of the year, the Four Months, an inauspicious time when ordinary activities are interrupted; but it is the quintessential time of festivals, characterised by radical reversals followed by a total re-establishment of order. Next comes a long period of eight months that, beginning with the Early Winter season, overlaps the winter solstice and continues until the summer solstice.

3. Structure of the annual cycle

27How can this cycle be conceived of, in order to bring out an overall structure in the Hindu calendar? In particular, where should the beginning and end of the religious year be placed?

3.1. Current conceptions

28Informants spontaneously supply various responses often confirmed in the literature and not always reconcilable. We will look at two of them.

29The first response is based on the traditional division of society into four classes, assigning a specific festival to each of them. One therefore isolates four festivals considered most important; they are enumerated beginning at the summer solstice, and allocated to the classes in descending hierarchical order. The three festivals attributed to the three upper classes take place in the short division of the year, the Four Months: Rsi tarpani during the Sravana full moon, with the renewal of the sacred thread and the worship of inspired sages, is the festival of the Brahmans, priests and scholars; Dasai, in Asvina, with the total overhaul of the political edifice, is the festival of the Ksatriya, kings and warriors; Divali in Kartika, with the worship of money, account books and livestock, is the festival of the Vaisya, merchants and farmers. As for the Sudra, the lowest class dedicated to the service of the three others, they are excluded from the short period and consigned to the long division, in which they are allocated the spring festival of Holi. This classification, often overlooked by specialists, has considerable depth. In fact, the top three classes, those made up of the twice-born who have access to Vedic initiation, knowledge and rituals, have their festival during the Four Months, which seem to be the sacred period of the year when men and gods communicate. The Sudra, who have no access to superior rituals and knowledge, have no festivals during this sacred period.

30The second response divides the year in three equal parts, each placed under one of the three great gods of Hinduism. The year starts at the end of autumn with the awakening of Vishnu, who reigns for the four-month duration of the seasons of Early Winter and Late Winter, regenerating and ordering the universe. The months of Spring and Summer fall under Brahma. The Four Months, which cover the Rains and Autumn, fall—positively—under Śiva, god of creation and destruction; negatively, they are plunged into grief by the absence of Vishnu, who has gone to sleep in hell.

31These two responses agree with the conclusions of the preceding section, making the Four Months into a specific time of year. It is then that the three upper classes have the place of honour; the absence of Vishnu makes possible the disorders and reversals that characterise the major festivals. Once again, the points of division are clear: the year divides into two unequal periods of four and eight months. But does the sacred division of the Four Months constitute the beginning or end of the religious year?

3.2. Time and space

32Speculations on religious time, in both Indian thought (Mus 1932–1934, passim; Zimmer 1951: 11–19) and in comparative studies (Mauss 1968: 3–39; Eliade 1968: 310–343), agree that a religious year must have a well-defined beginning, middle and end, forming an ordered succession, a cycle giving onto a new beginning, onto another cycle identical to the previous one. From these schools of thought, we can extract the following principles of interpretation:

(a) Religious time has the following properties:

  • the different divisions of time (days, years, cosmic periods…) each form a cycle;
  • these cycles are homologous;
  • each cycle has a beginning and end;
  • each cycle represents a devolution from order towards disorder to the point of chaos, before the regeneration that will mark the beginning of a new cycle;
  • the end of the cycle, that is to say the phase of chaos and regeneration, is not considered part of the temporal cycle; it represents the axis that connects with eternity.

(b) Religious space has the following properties:

  • the different religious spaces (cosmos, town, temple, human body as microcosm…) form a whole;
  • these spaces are homologous;
  • their path is directed: it includes a point of departure and a point of arrival;
  • each of these paths leads from order to disorder culminating in chaos, then to regeneration, which begins a new identical path;
  • the end of the journey is not situated in horizontal terrestrial space, but follows a vertical axis that connects earth with hell and heaven.

(c) These two series of propositions are completed by a last, very important proposition: the temporal cycles and the spaces are homologous. A term-to-term correspondence can be established between the moments in the temporal cycle and the orientation points in space; travelling through a given space corresponds to a time cycle.

33This conception of time in its relationship with space provides the key to interpreting the Hindu calendar. Table 1, a simple provisional reference, offers a spatial representation of the annual cycle, but it is like a wheel without beginning or end; it is too homogeneous since all parts of time are equal in it; it is too flat since it lacks the third vertical dimension corresponding to the other worlds and to eternity; finally, it is not oriented in relation to the cardinal points.

34It is possible to give a much more adequate interpretation of Hindu time. Paul Mus (1932–1934, 33: 847) has masterfully paved the way: ‘the orientation point was given over to the four winds, in the course of one year, […] and this schema was completed by a celestial orientation point matching the point where time starts, and where it is destined to enter back into itself’. This is clear and impeccable in the abstract. But reading Mus’s book leaves one unsatisfied: nowhere does he indicate how these principles can be concretely applied to marking the divisions of time, the paths of space, and the spatiotemporal correspondences.

3.3. The annual cycle

35The present article’s contribution is specifically to show this, based on more limited material. I will make use of a treatise on Buddhist iconography written in Sanskrit over the course of the first millennium AD, entitled the Sādhana mālā; it is accessible in a nearly complete English translation (Bhattacharya 1958). I will no doubt be criticised for making reference to a Buddhist tradition that is somewhat removed from the Indo-Nepalese population studied here. But calendrical conceptions are shared, and furthermore, in the Hindu books consulted, I found no text that illustrates our subject as clearly and succinctly.

36Of the many correspondences established in the Sādhana mālā, we will only retain those that are absolutely necessary for understanding the annual cycle. For the temporal cycles, our points of reference will be both a period shorter than the year, the day, ahorātra, and a longer period, caturyuga or mahāyuga, the whole of the four ages, a cosmic period spanning from a re-creation to a destruction. In Indian thought, both Hindu and Buddhist, starting with each re-creation, the Universe passes through four ages, yuga, of increasing degeneracy—religious order, dharma, diminishing by one quarter in each of them. The first, satya yuga, is the perfect time when dharma is complete; in the second, tretā yuga, only three-quarters remain; and so on…; in the fourth age, kali yuga—the one in which we now live—there remains only one quarter, which will gradually disappear from now until the end of the time ahead. As for space, we will only retain the cardinal points without dwelling on the microcosms that are altars, temples, towns and the human body. Finally, the most important correspondences will be established between those three homologous temporal cycles on the one hand, and space on the other.

Table 2: Correspondence of temporal cycles between themselves and with the orientation points of space

Order no.* Year Day Cosmic period
Season Month Moments Ages Cardinal points
1 Early Winter mārgaśira dawn satya north-east = centre
2 Late Winter māgha noon tretā east
3 Spring caitra mid-afternoon dwāpara south
4 Summer jyeṣṭa dusk kali west
5 Rains śrāvaṇa midnight the end of time north
6 Autumn āśvina o o o
* Cf Table 3

37The results of the examination of those texts are presented in Table 2, and they are very enlightening. It is obvious that the annual cycle has a beginning that is well-defined and can be clearly pinpointed. Just as the path through a religious space, pradakṣina, always begins in the north-east, identified with the centre, every day begins at dawn, and every cosmic period begins with the satya yuga; likewise, the religious year begins with the Early Winter season, which inaugurates the long division of the year, after the end of the Four Months.

38This long division of eight months leading to the summer solstice presents a certain simplicity: it is the continuous unfolding of time, analogous to that of the moments of the day or of the four ages of a cosmic period. Confirming my earlier conclusions, the winter solstice, corresponding both to midday (or the second age of the cosmic period) and to the east, marks no break or point of departure; it is just one reference point among others in the unfolding of the long division. Projected into space, this corresponds with the horizontal ritual path clockwise, always keeping the centre to its right and starting from north-east identified with the centre, in order to successively pass the east, south and west to reach the north. In all respects, these eight months represent the ordinary course of time in its auspicious large division that nothing comes along to disturb. However, a remarkable feature should be noted in contrast to our calendar: the winter solstice, which for us is the bad season and marks the sun’s lowest point on the horizon, corresponds to the Hindu midday and to the east, the brightest and most valued points. Conversely, the summer solstice—which for the Hindus marks a break (the beginning of the short division of the year)—inaugurates the dark and dangerous period of the Four Months. That which for us is the beautiful season when the sun is highest in the sky, is for Hindus midnight, the end of time, the end of the ritual path.

39Those Four Months present some interesting problems on Table 2. They include two seasons. Only the first, that of the Rains, has spatiotemporal correspondences: this season begins in the north, at midnight, and inaugurates the end of time; it must logically correspond to the end of the cycle, to the disintegration that prepares the beginning of a new cycle; in space, it is equivalent to the finale loop of the path that leads back from the north towards the centre. A surprise awaits us in the sixth season, Autumn: the table is empty; there is no correspondence in time or space. Yet it is the moment of the largest festivals, Dasai and Divali, which are separated by a long interregnum. Are the correspondences established by tradition unsound, the six-season system unable to be connected with the other divisions of time and space? This is very unlikely; in those types of traditions, nothing is left to chance and everything is carefully calculated.

40It is therefore the moment to introduce the third dimension, representing verticality and eternity, which was lacking in our overly flat view of the calendar. Those two months of the final season, whose festivals revive the battles of the gods and demons, and bring the dead back among the living, are in reality outside of time. As I have said, the end of the cycle is not in time, but represents the axis that connects with eternity; the Indian texts say that in Autumn, the faithful are in this interval, this remainder, śeṣa, where there is nothing; everything is annihilated and regeneration is being prepared; it is in this primordial moment, in these primordial waters, that gods, demons and men, the dead and living communicate; the boundaries and roles are not yet set, or rather they are challenged and then reinstated before the universe resumes its normal course. Hence the numerous reversals one sees during that interregnum.

41From the spatial point of view, the same applies. The end of the path is not located in horizontal space but in the third vertical dimension, the world axis that connects the earth with heaven and hell. The whole of space must therefore be considered in its three dimensions. The displacement undergone during the season of the Rains towards the centre was also a descent into hell. In my interpretation, Autumn’s path must also go according to the vertical axis at its centre outside of horizontally deployed space, from the bottom up, to reach the top of the axis that will serve as a point of departure for a new path.

42From this perspective, the year will be spatially represented not as a flat circle, but as a sort of helix that starts from the top of a vertical central axis, describes an arc on a horizontal plane, and finally descends back towards the base of the axis (cf Table 3).

Table 3: Spatial representation of the annual cycle of festivals

Table 3: Spatial representation of the annual cycle of festivals

The cardinal points are arranged as on our compass rose. The oblique line linking the centre to the north-east represents the vertical axis. The year’s path begins at the top of that axis, identified with the north-east, and goes clockwise; the numbers on this path mark the beginning of the seasons and refer to the spatiotemporal correspondences indicated on Table 2. On the outside are the calendrical reference points: seasons, lunar months subdivided into fortnights; a double line marks the separation between two months at the full moon; a single line, the separation between the fortnights, dark and then bright at the new moon. The date of festivals, depending on whether they last one or several days, is indicted by a dot or bold line on the inner side of the helix.

43Thus we have a clear view of the annual cycle. The path that the arc follows from the north-east to the north represents the long division of the year, from the beginning of Early Winter to the summer solstice; it is the auspicious period when life follows its normal course with family ceremonies of good omen, festivals of prosperity and protection, and lineage worship; the rare festivals in the strict sense are rather flat. During the short division of the Four Months, one returns from the circumference towards the centre and follows the path of the vertical axis; normal life is abandoned to begin the end of the cycle, the inauspicious period from which ceremonies of good omen and ceremonies of prosperity are excluded; it is the time of disorders, reversals and restorations of order expressed in this period of the year when festivals are most numerous and largest, particularly during the two last months that relate to the path of the vertical axis, to the identification of time and eternity. This season of Autumn is that of the great restoration of order on the cosmic, political, social and family levels.

44Does such a conception, developed based on a particular model—that of the Indo-Nepalese of Nepal—lend itself to generalisation? It definitely should for the whole Indian domain, Indo-Nepalese culture being only one of the subcontinent’s multiple cultural variants. For example, I have already shown the deep relationship of the festival of Divali in Nepal and North India with that of Onam in South India (Gaborieau and Helffer 1968–1969). Why would it not be the same elsewhere for other festivals?

45These reflections lead to more general considerations of religious anthropology. Thinkers as diverse as Mauss, Eliade and Leach have highlighted the same properties of religious time: it is divided into periods of different natures. On the one hand are those—not very pronounced, flat so to speak—that represent ordinary secular time; on the other hand are those that, affected by deep upheavals, manifest a break with ordinary life, generating disorders and reversals; they enable connection with the sacred; they are festivals. Under this conception, the annual cycle is cut in several places.

46All of this applies to the Hindu calendar, with the exception of one detail: in India, the only cut is that of the Four Months, located at the end of the cycle; this is when the genuine festivals take place, the quintessential festivals, in which reversals and disorders reach their paroxysm while the radical restoration of order heralds the renewal of the cycle. The other festivals, which take place during the long division of eight months, are simply worship or carnivals that do not reach the dimension of the sacred; in particular, spring celebrations marking the new agricultural and civil year remain secondary; they do not disrupt the course of everyday life. Primarily concerning the lower class of the Sudra and the administrations, they do not have the depth of the end-of-cycle festivals that give the three upper classes of twice-born special access to the sacred.

Haut de page


Anderson, M. 1971. The Festivals of Nepal. London: George Allen & Unwin.

Bhattacharya, B. 1958. The Indian Buddhist Iconography. Calcutta: Mukhopadhyay.

Bīrūnī, A R. 1964. Alberuni’s India: an account of the religion, philosophy, literature, geography, chronology, astronomy, customs, laws and astrology of India about A.D.1030, edited, with notes and indices, by E C Sachau (second edition). 2 vols. Delhi: S. Chand (Trübner’s oriental series).

Bista, K B. 1969. ‘Tīj ou la fête des femmes’. Objets et Mondes 9(1): 7–18.

Bista, K B. 1972. Le culte du kuldevata au Népal, en particulier chez certains Ksatri de la vallée de Kathmandou. Paris: Éditions du CNRS (Cahiers népalais).

Bouillier, V. 1979. Naître renonçant. Une caste de Sannyāsi villageois au Népal central. Nanterre/Paris: Laboratoire d’Ethnologie/Klincksieck (Recherches sur la Haute Asie 5).

Bouillier, V. 1982. ‘Si les femmes faisaient la fête... A propos des fêtes féminines dans les hautes castes indo-népalaises’. L’Homme 22(3): 91–117.

Crooke, W. 1923. ‘The diwali, the Lamp Festival of the Hindus’. Folklore 34: 267–292.

Crooke, W. 1968. The Popular Religion and Folklore of Northern India. 2 vols. Delhi: Munshiram Manoharlal.

Eliade, M. 1968. Traité d'histoire des religions. Paris: Payot.

Gaborieau, M. 1974. ‘Classification des récits chantés. La littérature orale des hindous de l'Himalaya’. Poétique 19: 313–332.

Gaborieau, M. 1977. ‘Systèmes traditionnels des échanges de services spécialisés contre rémunérations dans une localité du Népal central’. Purusārtha 3: 3–70.

Gaborieau, M. 1978a. Le Népal et ses populations. Bruxelles/Paris: Éditions Complexe /Presses Universitaires de France.

Gaborieau, M. 1978b. ‘Le partage du pouvoir entre les lignages dans une localité du Népal central’. L’Homme 18(1-2): 37–67.

Gaborieau, M and Helffer M. 1968-1969. ‘Problèmes posés par un chant de Tihār’. L’Ethnographie 62/63: 69–89.

Helffer, M. 1977. ‘Une caste de chanteurs-musiciens : les Gaine du Népal‘. L’Ethnographie 73: 45–75.

Jest, C. 1966. ‘La fête du Janaipurnimā à Patan’. Objets et Mondes 6(2): 143–152.

Jośi, S M. 1956 (V S 2014). Hamro lok-saṃskṛti (Our folk culture). Kathmandu: Ratna Pustak Bhandar.

Kandanvā, X. 1963 (V S 2020). Nepāli jan sāhitya (Nepali folk literature). Kathmandu: Royal Nepal Academy.

Leach, E R. 1968. ‘Deux essais concernant la représentation symbolique du temps’. In Critique de l'anthropologie, pp210–230. Paris: Presses Universitaires de France.

Leonard, R G. 1965. Nepal and the Gurkhas. London: Her Majesty's Stationary Office.

Lévi, S. 1905-1908. Le Népal. Étude historique d'un royaume hindou. 3 vols. Paris: Annales du musée Guimet.

Macdonald, A W. 1962. ‘Notes préliminaires sur quelques jhākri du Muglan’. Journal Asiatique 1: 107–139.

Macdonald, A W. 1969. ‘Note sur deux fêtes chez les Tharu de Dang’. Objets et Mondes 9(1): 69–88.

Macdonald, A W. 1972. ‘Le janaipurnimā et le gosaīkunda’. In Langues et techniques. Nature et société. Volume II: Approche ethnologique, approche naturaliste, edited by J M C Thomas and L Bernot, pp73–80. Paris: Klincksieck.

Monier-Williams, M. 1970. A Sanskrit-English Dictionary. Oxford: The Clarendon Press.

Nepāl rājkīya pancānga (Nepal government calendar), Varanasi-Kathmandu. (We consulted in particular the following years: V S 2023 [1966–1967], 2026 [1969–1970] and 2032 [1975–1976]).

Mauss, M. 1968. ‘Introduction à l'analyse de quelques phénomènes religieux’. In Œuvres I: Les fonctions sociales du sacré, pp3–39. Paris: Éditions de Minuit.

Mus, P. 1932-1934. ‘Barabudur. Les origines du stupa et la transmigration, essai d’archéologie religieuse comparée’. Bulletin de l’École française d’Extrême-Orient 32: 269–439; 33: 577–980; 34: 175–400.

Pignède, B. 1966. Les Gurungs. Une population himalayenne du Népal. Paris/La Haye: Mouton.

Platts, J T. 1884. A Dictionary of Urdu, Classical Hindi and English. London: Oxford University Press.

Renou, L. and Filliozat J. 1947. L’Inde classique. 2 vols. Paris: Payot.

Śarmā, B C. 1962 (V S 2019). Nepālī śabdakoś (Nepali dictionnary). Kathmandu: Royal Nepal Academy.

Stevenson, S. 1971. The Rites of the Twice-Born. Delhi: Oriental Reprints Corporation. London: Oxford University Press.

Thomas, P. 1961. Epics, Myths and Legends of India (12th edition). Bombay: D B Taraporevala Sons & Co.

Toffin, G. 1979. ‘Aspects religieux de la royauté néwar au Népal’. Archives de Sciences sociales des Religions 48(1): 53–82.

Toffin, G. 1981. ‘Culte des déesses et fête du Dasaï chez les Néwar (Népal)’. Purusārtha 5: 53–82.

Tripāṭhī, R P. 1971. Hinduõ ke vrata, parva aur tyauhār (Hindu's fasts, feasts and festivals) (9th edition). Allahabad: Lokbharati Prakashan.

Turner, L. 1931. A Comparative and Etymological Dictionary of the Nepali Language. London: Routledge/Kegan Paul.

Vajrācaryā, P R. 1963 (V S 2020). Hāmro saṃskṛti (Our culture). Kathmandu.

Zimmer, H. 1951. Mythes et symboles dans l'art et la civilisation de l’Inde. Paris: Payot.

Haut de page


2 To facilitate possible comparison with other regions of the Indian subcontinent, I have most often used the Sanskrit terminology without providing the multiple equivalents in modern vernacular languages. It is this terminology that is in use on the religious calendars of various regions of the subcontinent. The origins of certain terms is indicated by the following abbreviations: (Skt.) = Sanskrit, (H.) = Hindi, (Nep.) = Nepali.

3 For more details, see the excellent calendar recently published (Bouillier 1979: 58–80) and the articles and monographs listed in the bibliography.

Haut de page

Table des illustrations

Titre Table 1: The Hindu calendar
Légende The circle representing the year is divided into equal parts successively indicating the two periods (at the centre), the six seasons (middle ring) and the twelve months (outer ring). Each subdivision shows the Sanskrit term (generally used here), the Nepali term (if it exists), the English translation, and finally reference points on our own calendar.
Fichier image/jpeg, 229k
Titre Table 3: Spatial representation of the annual cycle of festivals
Légende The cardinal points are arranged as on our compass rose. The oblique line linking the centre to the north-east represents the vertical axis. The year’s path begins at the top of that axis, identified with the north-east, and goes clockwise; the numbers on this path mark the beginning of the seasons and refer to the spatiotemporal correspondences indicated on Table 2. On the outside are the calendrical reference points: seasons, lunar months subdivided into fortnights; a double line marks the separation between two months at the full moon; a single line, the separation between the fortnights, dark and then bright at the new moon. The date of festivals, depending on whether they last one or several days, is indicted by a dot or bold line on the inner side of the helix.
Fichier image/jpeg, 101k
Haut de page

Pour citer cet article

Référence électronique

Marc Gaborieau, « Festivals, time and space: the structure of the Indo-Nepalese version of the Hindu calendar »European Bulletin of Himalayan Research [En ligne], 57 | 2021, mis en ligne le 15 décembre 2021, consulté le 07 décembre 2023. URL : ; DOI :

Haut de page


Marc Gaborieau

Chercheur associé au CEIAS

Directeur d'études retraité à l'EHESS

Directeur de recherche retraité au CNRS

Haut de page

Droits d’auteur


Le texte seul est utilisable sous licence CC BY 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search