Zoroastrian festivals fall into two broad categories. There are the seven feasts of obligation, that is, No Rōz (Nowrūz) and the six gāhānbārs (gāhāmbār; q.v.), which formed the framework of the religious year, and which it was a sin not to keep; and others, which it was a merit, not a duty, to observe. This second category can be subdivided into major and minor feasts. The former, kept generally throughout the community, were in honor of great yazatas (benign divinities) of the Zoroastrian pantheon, with the exception of Frawardīgān (q.v.) and Sada, which is a winter fire-festival. Of the minor ones, of which not much is recorded, a number were in honor of locally venerated yazatas, with whose cults fairs often grew up. Abū Rayḥān Bīrūnī, writing around 1000 C.E., said of them: “We cannot fix them, as little as we can the watercourses of a torrent, it being impossible to count them” (Āṯār, p. 230, tr. Sachau, p. 217). One which he did record, Waḵš-angām, was celebrated in Chorasmia in honor of Waḵš (Waḵš-angām), divinity of the river Oxus (ibid., p. 237, tr. p. 225). There were also annual festivals at appointed times at pilgrim shrines (Boyce, Stronghold, pp. 241-70), and each Parsi Ātaš Bahrām (see ĀTAŠ) holds an annual festival to commemorate the day of its sacred fire’s enthronement. Individuals or communities have sometimes established yearly festivals to celebrate some local event, and sometimes such a festival evolved in spontaneous thanksgiving after a remarkable happening, as with the Sasanian feast of Ābrīzagān (q.v.; Bīrūnī, Āṯār, pp. 228-29, tr. Sachau, pp. 215-16). In addition to annual festivals, small occasional ones may be celebrated at any time by a family or a local community, in thanksgiving or worship, and these, too, are called jašans, which is the generic Zoroastrian term for a festival.

Jašan or jašn are the Middle and New Persian forms of Avestan yasna, meaning “an act of worship,” and a religious service is an essential part of every festival. Such services range in solemnity according to the importance of the festival, from the long and elaborate Visperad, celebrated (with other acts of worship) at each of the seven obligatory feasts, to the Yasna itself, and down to a simple Āfrīnagān (q.v.) for minor observances. Since every jašan is essentially a holy occasion, those taking part should be in a state of physical and ritual cleanliness, for dirt and pollution belong to the evil creation and prevent prayers and worship reaching the divine beings. They should also seek to banish from their thoughts any “demons” of anger, grief, resentment, or the like, and try rather to entertain contentment, cheerfulness, and charitable feelings towards all, such as are pleasing to the yazatas.

On the obligatory festivals all but necessary work was forbidden, and the other great feast-days were also generally kept as holidays. Preparations were made in advance, houses were scrupulously swept and cleaned, and people wore their best clothes. In traditional usage everyone first attended at least part of the religious services, saying their own prayers while these were solemnized by the priests; or if unable to do this, they at least took part by sharing in the food then blessed (Modi, 1937, pp. 424-25; Boyce, Stronghold, pp. 34-35, 39-40, 232-33, 234). Thereafter it was a pleasant duty to be as merry as possible, since in Zoroastrian doctrine joyfulness is a positive virtue, a weapon to defeat sorrow and care. Feasting, the friendly and enjoyable sharing of food and drink, forms a prominent part of Zoroastrian festivals, with the feasts sometimes being provided by an individual (either through a charitable endowment or by a single act of hospitality), sometimes, especially in the case of the gāhānbārs, being communal banquets, with everyone contributing.

There were traditional festal foods, and some dishes were associated with particular festivals (Bīrūnī, Āṯār, p. 235, tr. Sachau, p. 221; see also individual feast days); wine was regularly drunk, with toasts being given. There seems also always to have been music, either at the feast, or afterward, or both. Male and female musicians figure, with banqueters, in Sogdian wall paintings (e.g., Azarpay, Pls. 28-30; Belenizki, pp. 107, 209); and the Sogdian divinity Artimpasa, identified as Avestan Aši (q.v.; Grenet, pp. 41-45), is represented on a gold plate of the 4th century B.C.E. enthroned among musicians and merry revelers, who are presumably celebrating her feast (Bessonova, p. 101, fig. 25). In the same century, in the far west of the Zoroastrian world, the Greek poet Diogenes wrote of Bactrian girls in Anatolia worshipping the yazata Anāhīt (q.v.) “in her laurel-shaded grove the while they, ‘mid plucking of triangles and pectides, thrum the magadis in responsive twanging, where also the flute, in Persian fashion, joins its welcome concord to the chorus” (Athenaeus, Deipnosophistae 14.636; Boyce and Grenet, Zoroastrianism, p. 204). When in the first century B.C.E. Antiochus I of Commagene founded annual festivals to celebrate his own birthday and coronation day, he endowed lavish banquets, at which wine was to be served unstintingly and music played as long as those present wished. The musicians were women permanently attached to the sanctuary which was the center of the cult (see Boyce and Grenet, Zoroastrianism, p. 341, with references). At Damascus (q.v.) in the third century B.C.E. the sound of revelry at “a Persian festival” rose up from the countryside round about so loudly that it reached the city walls (Polyaenus, 4.15). Still in modern times Zoroastrians enjoy music and dancing at their festivals, and down into the second half of the 20th century at traditional Zoroastrian centers in Persia worshippers would sing and dance on festal days at shrines themselves (Boyce, Stronghold, pp. 91-92; idem, 1975, pp. 114-15). Other pleasant sociable diversions were also entered into, such as story-telling, play-acting, and mime, with yet others which varied according to place and time. In Zoroaster’s own distant day chariot-racing appears to have been deeply enjoyed. In Parthian Iran racing of horses is mentioned at Nowrūz (Faḵr-al-Dīn Gorgānī, p. 35, l.47, tr. p. 20), whereas athletic contests took place at Anāhīt’s festivals in Lydia under Roman rule (Boyce and Grenet, Zoroastrianism, pp. 239-41). Particular customs and observances were associated with the major feasts, and these in time attracted speculations about their origin, and so about the origin of the festivals themselves (such as were gathered by Bīrūnī, Āṯār, pp. 215-18, 220-21, 222-23, for Nowrūz, Tīragān and Mehragān).

In fact, since Zoroastrianism is an ancient faith, there is no record of the founding of any of its major festivals. It seems a reasonable surmise that Nowrūz, the holiest of them all, with deep doctrinal significance, was founded by Zoroaster himself, while the six gāhānbārs, whose Young Avestan names show that they were originally pastoral and farming festivals, were probably adopted and rededicated to the faith by his followers during the “Young Avestan” period, that is about 1200-800 B.C.E. (Boyce, 1993, p. 105). Between sunset of the day of the 6th gāhānbār and sunrise of Nowrūz was celebrated Hamaspaθmaēdaya (later known, in its extended form, as Frawardīgān), and this and the gāhānbārs are the only festivals named in the surviving Avesta. The observance of Mehragān is known from Achaemenid times, and it and Nowrūz, Tīragān, and Sada are alluded to for the Parthian period. References to the keeping of festivals is fuller for the Sasanian epoch, and most abundant from the Middle Ages down to the present day (for details see individual festivals). The observances which are best known are those of Persians, and their religious calendar tends to be considered as “the” Zoroastrian one, since it appears to have been adopted throughout the Sasanian empire, and so has been followed by the Parsis and Persian Zoroastrians alike; but thanks to Bīrūnī (Āṯār, pp. 234-40) there has long been some knowledge also of festivals kept locally by the Sogdians and Chorasmians, and this has been added to this century through archeological discoveries. There are materials also for Zoroastrian observances as kept in Armenia (see armenia iii, Religion).

In Zoroaster’s time his people appear to have had a calendar of 360 days, divided into twelve months of thirty days each, with probably a thirteenth month added every six years or so to keep it in accord with the seasons. During the “Young Avestan” period a specifically Zoroastrian form of this calendar was created by dedicating each of the thirty days in a month to one of the yazatas (for the list see above, IV, p. 661), with at this stage the 8th, 15th and 23rd days devoted, it seems, to Apam Nąpāt, Haoma (qq.v.), and Dahma Āfriti (see DAHM YAZAD; Boyce, 1993, pp. 108-10). The thirty dedications are set together in Yasna 16, presumably for mnemonic purposes. The dedications of the twelve months to twelve out of the same thirty yazatas were probably made later in the Achaemenid period, for there is no listing of them in the Avesta, where the names of only seven of them occur, incidentally and in late contexts (Bartholomae, Air. Wb., col. 1171). Thereafter whenever a day- and month-name had the same dedication (such as day of Mithra, month of Mithra), the day was celebrated as a festival for that yazata. New feast-days were thus created for Ahura Mazdā (q.v.), to whom the first day of the month was devoted, and the six great Aməša Spəntas, the Fravašis (qq.v.), Tištrya/Tīri, Mithra, Āpas (see ĀBĀN) and Ātar (see ĀDUR). Moreover, possibly under Zurvanite influence (Nyberg, pp. 132-34), the 8th, 15th and 23rd days were now rededicated to the “Creator (Dadvah) Ahura Mazdā.” (These are in fact the only attested dedications of the four days, the hypothesis proposed being that appropriate alterations were made at this time in Yasna 16). In consequence there were from then on four festivals for Ahura Mazdā (q.v.) in the month Dadvah (Mid. Pers. Dai), a midwinter month when the creator’s power and the worship and joyfulness of his creatures were especially needed to combat the evils of cold and darkness. Bīrūnī (Āṯār, p. 235, tr. Sachau, p. 222) noted that in his time the Sogdians held fairs (in conjunction evidently with religious observances) on the three “Creator” days.

In the Achaemenid period the Zoroastrian calendar was further reformed, in the interests of more accurate time-measurement, by the addition of the epagomenae, that is, five extra days set at the year’s end (see Marshak, de Blois). Their introduction evidently caused deep bewilderment, with many people, priests as well as lay, suspicious of the validity of the celebration of holy days on the “new” dates. Accordingly they henceforth kept the festivals on both the “new” dates (doubtless at first under strong official compulsion) and the “old” ones, five days later, so that, for example, Nowrūz was now observed on the first and sixth of Frawardīn. The latter, as the supposed “old” date, was called the “Great” Nowrūz. The same duplication affected the other six obligatory feasts, and a number of the major festivals also, while great confusion, whose effects are still felt today, developed with regard to the festival of the fravašis (q.v.). This, instead of being the observance of the one night between sunset of 30 Spendarmad and sunrise of 1 Frawardīn, was extended perforce through the five days of the epagomenae which now lay between, and for the majority—because of bewilderment still existing in the second year of reform—through the last five days of Spendarmad also. The extent of the confusion which was caused could hardly have been foreseen by the reformers. A Middle Persian name for the epagomenae was andargāh, the “between-time,” and no doubt it was intended that they should simply stand between the old year and the new, with minimal dislocation of customary usages; but other names given to them, the “robbed” or “stolen” days, (rōz ī truftag/duzīdag; Bundahišn, tr. Anklesaria, p. 28, 1.21) shows the hostility their introduction evoked. The controversy attending it is still clearly attested in the Pahlavi books, which contain texts both for and against the reform; and this led to a wrong conclusion, namely that the reform itself took place in the Sasanian period (Boyce, 1970); but it appears that these Middle Persian texts must in this respect preserve texts and ideas which had been handed down orally through the intervening centuries. The reason why they continued in transmission was doubtless that the controversy itself continued to be a living one, with championing of the relative importance of the “greater” and “lesser” feast-days.

According to Bīrūnī (Āṯār, p. 224, tr., p. Sachau, 209), this lingering problem was partly resolved early in the Sasanian period (i.e., during the high priesthood of Kirdēr), when Hormoz I (r. 272-73 C.E.), joined the “greater” and “lesser” days of Nowrūz and Mehragān into single-day festivals. This set a pattern for other major festivals, and six-day observances became standard, though at some later time they were reduced to five days, probably under the influence of the epagomenae, with the fifth day then being regarded as the “greater” one. Hostility to the epagomenae, the source of all the trouble, must have vanished much earlier, and they had come, indeed, to be regarded by Persian Zoroastrians as especially holy days, for their priestly authorities had dedicated each of the five days to one of Zoroaster’s five Gāθās, and recitation of the appropriate Gāθā on each day was declared to be a highly meritorious observance. This usage was not universally adopted throughout the Zoroastrian community. Thus there is evidence that in Sogdia and Chorasmia the five days were simply called by the same names as the first five days of the month (Bīrūnī, Āṯār, 47-48, tr. pp. 57-58; Henning, p. 251, n. 58). Yet the epagomenae, even though still being termed the “stolen days” in some Sasanian texts, were necessarily treated everywhere as holy, because they largely coincided with the sixth gāhānbār (observed presumably on 30 Spendarmad before the Achaemenid reform, but eventually from 30 Spendarmad to the 5th Gāθā day), and were embedded in the cluster of festivals (sixth gāhānbār, Frawardīgān, Nowrūz) which merged to form the greatest festival season of the Zoroastrian year, with seventeen days of unbroken observances, from 25 Spendarmad to 6 Frawardīn. Naturally, only the rich, the leisured, and priests could set aside so much time all at once for religious festivals, but there was a special atmosphere during the whole period, and all took part some of the time, with the whole community joining in the holiest celebrations, namely the sixth gāhānbār (coinciding with the five Gāθā days and the last five days of Frawardīgān), the “farewell to the fravašis” on the last night of the epagomenae, and the Lesser and Greater Nowrūz.

Other calendar reforms eventually followed the Achaemenid one, because this created a 365-day calendar which slipped back slowly against the natural year, losing a month every 120 years. In the early 6th century C.E. it seems that a bold measure was taken to bring the seven obligatory feasts back into proper relation with the seasons. In 507-511 C.E. the spring equinox coincided with 1 Ādur (Āḏar); and apparently in one of these years it was decreed within the Sasanian empire that Nowrūz should be celebrated on 1 Āḏar instead of 1 Frawardīn, with the six gāhānbārs also moving to keep their fixed positions in relation to it, together with Frawardīgān, the epagomenae now being set at the end of the preceding month, Ābān. The other great non-obligatory festivals were not shifted, since they were tied to month- and day-names. Nowrūz was accordingly celebrated for the next half a millennium at the beginning of Āḏar month, even though this in its turn slipped backwards through the natural year. Piety and traditionalism ensured, however, that some celebrations continued also on 1 Frawardīn, which was still reckoned for civil purposes to be the beginning of the year. In due course, in 1006 C.E., 1 Frawardīn came to coincide with the spring equinox, and was once more celebrated as the religious Nowrūz (for details see de Blois), with the epagomenae being moved to stand again before it, at the end of Spendarmad month, and the gāhānbārs and Frawardīgān also returning to their former places in the calendar year. But because Nowrūz had by then kept on 1 Āḏar for five hundred years piety forbade that day’s abandonment, and religious services, and no doubt for a long time festivities also, continued for generations to be held on it as well.

The last reform of the 365-day calendar was made by Parsis, probably in or around 1125 C.E., when in order to bring 1 Frawardīn back to the spring equinox they repeated the preceding month, as Second Spendarmad, putting the epagomenae after it, so that, in festal terms, Frawardīgān and the sixth gāhānbār were celebrated then. This is the only certain instance of the theory being put into practice of the 365-day calendar being kept in harmony with the seasons by the intercalation of a thirteenth month every 120 years. It inevitably caused confusion, with puzzled people doggedly celebrating Nowrūz that year on the sixth day of Second Spendarmad, which, without this reform, would have been the first day of Frawardīn. Thereafter, right down into the second half of the 20th century—some 700 years later—6 Spendarmad has been kept as a festival day among Parsis, being known (among other names) as the “abandoned No Rōz,” Sōḍī Nahrōj (Khareghat).

This is the last instance of a recurring phenomenon, that every reform of the 365-day calendar led to some duplication of festivals, which was then piously maintained for generations, when not in perpetuity. The reason for this is clear: calendar changes are notoriously confusing in practice, and most people evidently did not understand the theory; and they were deeply concerned not to neglect days which had once been kept holy, in order not to fail in their duty to the yazatas. Evidence for perplexity arising at each attested calendar reform, and the keeping of the religious Nowrūz at 1 Āḏar for some sixteen generations, make it in the highest degree doubtful that there ever were regular intercalations of a month every 120 years in the 365-day calendar (for skepticism about this on other grounds, see Bickerman). The Parsis’ quite possibly unique intercalation of a month meant that thereafter their calendar, which came to be known as the Šenšāī (popular variant Šāhānšāhī) was a month behind that of the Persian Zoroastrians, which came to be called the Qadīmī (Parsi Kadmī). In the early 20th century a group of Parsis, inspired by Kharshedji Rustamji Cama (q.v.), adopted the Gregorian calendar, known as the Faṣlī, with a fixed Nowrūz on 21 March, and a day intercalated every fourth year after the epagomenae. There are accordingly now three Zoroastrian calendars in use, by which the same festal days are celebrated, but on different dates. However, with urbanization of the community, and Zoroastrians generally entering into the mainstream of life in the various lands in which they now live, many festivals have been abandoned and others shortened and simplified, with the loss of old observances. Traditionalists still, however, maintain the chief festal days, and even the skeptical keep Nowrūz, though tending to celebrate this (as Shiʿite Muslims do) as a secular feast.


Bibliography (for cited works not given in detail, see “Short References”):

G. Azarpay, Sogdian Painting, Berkeley and Los Angeles, 1981.

A. M. Belenizki, Mittelasien, Kunst der Sogden, Leipzig, 1980.

S. Bessonova, Religioznye predstavleniya Skifov (Religious concepts of the Scythians), Kiev, 1903.

E. J. Bickerman, “The ‘Zoroastrian’ calendar,” Archiv Orientálnî 35/2, 1967, pp. 197-207.

M. Boyce, “On the Calendar of Zoroastrian Feasts,” BSO(A)S 30/3, 1970, pp. 513-39.

Idem, “Mihragān among the Irani Zoroastrians,” in Mithraic Studies, ed. J. R. Hinnells, 2 vols., Frome and London, 1975, I, pp. 106-18.

Idem, Zoroastrianism, Its Antiquity and Constant Vigour, Columbia Lectures on Iranian Studies 7, Costa Mesa, 1993.

F. de Blois, “The Persian Calendar,” Iran 34, 1996, pp. 39-54.

Faḵr-al-Dīn Asʿad Gorgānī, Vīs o Rāmīn, M. A. Todua and A. A. Gwakharia, Tehran, 1349 Š./1970; tr. G. Morrison as Vis and Ramin, New York, 1972.

F. Grenet, “L’Athéna de Dil’berdžin,” in Cultes et monuments religieux dans l’Asie Centrale préislamique, ed. F. Grenet, Paris, 1987, pp. 41-45.

W. B. Henning, “A Sogdian God,” BSO(A)S 28/2, 1965, pp. 242-54; repr. in idem, Selected Papers II, Acta Iranica15, 1977, pp. 617-29.

M. P. Khareghat, “The Daryāī Nōrōz,” in J. J. Modi Memorial Volume, Bombay, 1930, pp. 118-30.

B. I. Marshak, “Istoriko-kul’turnoe znachenie sogdiyskogo kalendarya,” in T.B. Knyazevskaya, ed., Mirovaya kul’tura: Tradicii i sovremennost, Moscow, 1991, pp. 183-97; tr. W. J. Pittard as “The Historico-Cultural Significance of the Sogdian Calendar,” Iran 30, 1992, pp. 145-54.

J. J. Modi, The Religious Ceremonies and Customs of the Parsees, 2nd ed., Bombay, 1937; repr. 1986.

H. S. Nyberg, “Questions de cosmogonie et de cosmologie mazdéennes II,” JA, 1931, pp. 1-134.

(Mary Boyce)

Originally Published: December 15, 1999

Last Updated: January 26, 2012

This article is available in print.
Vol. IX, Fasc. 5, pp. 543-546