The Circle of Ancient Iranian Studies
Iranian Religions: Zoroastrianism
By: Mark Garrison
Fire altar is a term adopted by modern researchers to designate the stand upon which sacred fire was placed. Strictly speaking, the designation "fire altar" is incorrect, since the structure was not used to receive a sacrifice, but simply to hold the fire for the purposes of veneration, probably contained within a metal or clay bowl. Mary Boyce has suggested the appellation "fire-holder"; although more accurate, this designation has not won wide acceptance (Boyce, Zoroastrianism II, p. 52; Houtkamp, p. 24). Because the term fire altar has such immediate and direct connotations of Zoroastrian worship, it should probably also be restricted to those structures which have a clear Zoroastrian religious context. Scholarship concerning fire altars has rarely been this precise, and the term has in general been used to characterize any altar-like structure which may have held sacred fire (see EIr. III, pp. 7-9, for terms designating the fire altar).
The question of the identification of fire altars is further complicated by our lack of understanding of the exact role of sacred fire in ancient Iran (and how that role may have changed over time), how and when the structures identified as fire altars were used in ancient ritual, and whether the appearance of a fire altar indicates the presence of Zoroastrian ritual or some other, unknown religion (Genito, 1987; Herrenschmidt; in general see Boyce, Zoroastrianism I and, II; Boyce and Grenet, Zoroastrianism, for criteria for identifying the presence of Zoroastrian religion).
Classical literary sources and Achaemenid artistic evidence (e.g., seals, tomb reliefs, etc.) suggest that sacred fire assumed a central role in the religion of the Achaemenid kings, although its exact nature is lost to us (Houtkamp, for the pictorial evidence; Boyce, Zoroastrianism II, passim, esp. pp. 50-62, 112-16; Yamamoto, 1979, pp. 26-36); the question is intimately tied to the continuing debate on Achaemenid state religion and the role that Zoroastrianism played (cf. Boyce, Zoroastrianism I and II; Herzfeld; Beneviste; Duchesne-Guillemin, 1953; idem 1958; Gerschevitch, pp. 8-22; Kellens, ed.; for evidence provided by the Persepolis Fortification tablets see Koch, 1977 and idem, 1991; cf. Boyce, Zoroastrianism II, pp. 132-37, commenting on Koch). Most scholars have sought the origins of Zoroastrian and Achaemenid fire worship in veneration of the household hearth fire (Boyce, Zoroastrianism I, pp. 154-55; Genito, 1982, passim, esp. pp. 231-32). How much earlier we can project back the worship of fire is unclear (for an extreme view see Khlopin, who traces the fire altar back to the fourth millennium B.C.E.).
J. Houtkamp and Yumiko Yamamoto (1979) have collected the archaeological and pictorial evidence for fire altars in the Achaemenid, Seleucid, and Parthian periods (see also Schippmann, 1971, pp. 473-85; Stronach, 1985; Genito, 1987). Of the two types of altar which the pictorial evidence preserves, the distinctive Iranian altar with a stepped top and base has been quite frequently identified in the archaeological record (Moorey; Boyce, Zoroastrianism II, pp. 145-148; Yamamoto, 1979, pp. 30-36, identifies three types of altar). On the other hand, the second form of altar, the crenellated tower altar, is almost entirely absent in the Iranian archaeological record (but cf. the rock-cut altar in Fârs at Bâgh-e Bodra [Stronach, 1966, pp. 223-24, fig. 7], perhaps Achaemenid in date, which conceivably represents an example of the tower altar). Various other types of structures have also been characterized as fire altars.
Predating the Achaemenid period is the altar found in the central temple at Tepe Nûš-e Jân (near Hamadân), dating to about 750-600 B.C.E. (PLATE I; Stronach, 1973; Houtkamp, p. 34; for other references see Boyce, 1975, pp. 456-57). This plastered mud brick altar had a straight-sided socle rising directly from the floor, surmounted by four steps. In the middle of the square upper surface a shallow, hemispherical fire bowl showed extensive traces of burning (Stronach, 1984, pp. 479-80). It is clear from the shallowness of the bowl, however, that it could never have sustained ever-burning fire. The striking stepped form of this altar suggests that it could have been a prototype (pre-Zoroastrian) for the later stepped stone altars of the Achaemenid period (Stronach, personal communication).
The two well-known square limestone plinths in the sacred precinct at Pasargadae (Fârs) may have some connection to a fire ritual (PLATE II; Stronach, 1978, pp. 138-45, figs 70-71, 74; Yamamoto, 1979, pp. 28-29, for a survey of interpretations). The southern one has a stepped top and stone steps. The northern one may also have had a stepped top. How exactly these structures were used is unknown, although the southern one is identified as a podium for viewing fire displayed on the northern one. Construction technique and historical circumstance of the site of Pasargadae suggest a date at the time of Cyrus. Pasargadae has also produced a fragment of a three-stepped altar top with a deep bowl and two three-stepped altar bases (one of which may belong to the altar top; see Stronach, 1978, pp. 141-42, fig. 72, pl. 107b; Houtkamp, p. 37; PLATE III). Boyce (Zoroastrianism II, pp. 51-52) takes these fragments to be the earliest known fire altar.
At Altin Tepe-10 (q.v.; south of Turkmenistan), in one of the corner rooms in Structure II, there was found "a heap of three-stepped unbaked clay pyramids covered with white gypsum," which have been interpreted as remains of a fire altar (Sarianidi, p. 102, figs. 49-50; Houtkamp, p. 34). The suggested date is late 5th century B.C.E.
At Dahan-e Gholâmân (q.v.; Sîstân) Building 3 (QN3), a large square courtyard surrounded by porticoes and rooms (towers) in the corners, was described by Scerrato as a "holy building" (Scerrato, 1979, p. 712). In its earlier Phase A the building contained some forty-seven fire "chests" (ibid., p. 716). In phase B, the complex was completely re-modeled, and the porticos housed "large ovens," "stoves" and "fire-places." In the middle of the courtyard there were three rectangular altars, each standing on a wide base and having small stairways (perhaps present in Phase A as well; Scerrato, 1979, p. 724). Scerrato (1979, p. 719) suggested that a fire burned inside each altar. In a slightly later Phase B1, ten to twelve low rectangular platforms, each having a hemispherical depression in its center to contain fire, were installed in the north portico. Scerrato (1979, pp. 725, 731-33) suggested that building QN3 provided the earliest evidence for the worship of three deities and the "Indo-Iranian doctrine of the three fires" (cf., however, Boyce, Zoroastrianism II, pp. 128-31, articulating the un-Zoroastrian character of the complex). Building QN3 is usually dated in the Achaemenid period (cf. Houtkamp, p. 16; Schippmann, 1971, pp. 50-57, figs. 6-7). Dahan-e Gholâmân has also yielded the stepped base to a "fire altar" (house QN6; Scerrato, 1979, p. 727, fig. 16; Genito, 1987, pp. 480-81, pls. 1, 3) and a complete "fire altar" with stepped top and hemispherical depression for fire or fire container (near building QN16; Scerrato, 1979, p. 727, figs. 18-19; Genito 1987, pp. 480-81, pls. 2-3). The excavators feel that both are Achaemenid in date (Houtkamp, p. 36).
The site of Bünyan (in central Turkey) has produced a possible altar (now in the Museum of Anatolian Civilizations, Ankara): a limestone pedestal with a two-stepped plinth (with relief decoration) and top (Houtkamp, pp. 34-35). The exact context and date of the monument are not known.
A four-stepped top of a pink sandstone altar has been found at Khâkrîz (Khakriz; Kermânšâh; Gaube, p. 155, fig. 2, pl. 32.4; Houtkamp, p. 36). Dates for the monument vary from Achaemenid (Kleiss) to Sasanian (Vanden Berghe, p. 515).
The conquest of Persia by Alexander and the subsequent Hellenization of the country evidently brought a great deal of destruction of Persian religious sites (Boyce and Grenet, Zoroastrianism, pp. 3-22). Classical sources describe, however, actual fire temples, thus providing contemporary documentation for the temple cult of ever-burning fire (Yamamoto, 1979, pp. 41-42, 49-50; Boyce and Grenet, Zoroastrianism, pp. 235-38, 269-71). Strabo (15.3.15) describes an ever-burning fire altar and its associated upkeep and ritual in Cappodocia; Pausanias (5.27.5-6) in Hierocaesareia and Hypaepa in Lydia. The late source Isidor of Charax also refers to the survival of a fire in the town of Asask, home of the Arsacid dynasty (Schippmann, 1971, pp. 33-34; Boyce, 1975, pp. 461-62); perhaps this fire was the dynastic fire of the Arsacid family (Yamamoto, 1979, pp. 40-41). Owing to the large number of independent regional powers in the Parthian period, scholars postulate an increase in local "dynastic fires" (e.g., Boyce, Zoroastrianism I, p. 461), a practice which may have its origin in the Achaemenid period. Depictions of stepped fire altars occur on at least one series (the third series: Darius II, Ozathres, and Artaxerxes II) of silver coins of the so-called frataraka rulers of Persis (Houtkamp, pp. 44-45; Yamamoto, 1979, pp. 45-49, pl. 53; Boyce and Grenet, Zoroastrianism, pp. 110-16; Stronach, 1966, pp. 220-23, 227; Sellwood, pp. 300-6, pls. 10/12, 11/2), as well as on the coins of the Arsacid rulers Vologeses I (Yamamoto, 1979, p. 449, pls. 54-55) and Vologeses IV (Duchesne-Guillemin, in Camb. Hist. Iran, p. 867). Fire altars of other shapes occur in the coinage of Hierocaesareia, Hypaepa, Zela and a few other Greek cities (Boyce and Grenet, Zoroastrianism, pp. 230-31, 235, 288, 300), on some tomb reliefs in Media (Boyce and Grenet, Zoroastrianism, pp. 95-106) and perhaps on rock-cut reliefs at Bîsotûn and Tang-e Sarvak (Schlumberger, pp. 1041, 1043-44, pl. 64).
The innermost room of the fire temple at Kûh-e Khúâja (Sîstân, Schippmann, 1971, pp. 57-70, figs. 10 and 83; Yamamoto, 1979, p. 43), a monument with a still uncertain structural history, yielded (in later levels) a stepped fire altar which Ernst Herzfeld dated to the first century B.C.E. (Figure 1; Yamamoto, 1979, p. 43, pl. 34, identified it as the oldest surviving Zoroastrian altar; but now cf. Huff).
Site IV at Šahr-e Qûmes (near Dâmghân, q.v.) shows a plan with some similarities to that of the temple at Tepe Nûš-e Jân (Yamamoto, 1979, p. 44-45, fig. 7). The building was abandoned in the first century B.C.E. In room 5 was discovered a small plinth (plastered mudbrick) which supported a "fire bowl."
The room described as a small sanctuary in the temple at Sorkh Kotal (Afghanistan) preserved a long "bench-altar" with attached bird figures and large amounts of ash (Schippmann, 1971, pp. 492-96, fig. 81; Yamamoto, 1979, 44-45; Boyce, 1975, pp. 461-62). Although this structure has been linked with an ever-burning fire ritual, neither the altar nor the layout of the sanctuary conform to Zoroastrian types. The square central room of the complex had a large stone platform in its center whose function is unclear (fire altar, support for cult image, or both?). The date of the complex is uncertain (Boyce, 1975, pp. 461-62; Stronach, 1985, pp. 622-23).
At Takht-e Sangîn (on the right bank of the Oxus River) each of two subsidiary rooms in the temple contained a square fire altar (Boyce and Grenet, Zoroastrianism, pp. 173-79). In the corners there were also small fire altars, and large amounts of ash covered the floors. All the altars were mud brick and had received numerous coats of plaster. According to the excavators, these chambers were a feature of the temple from its inception. Boyce considered the complex the "earliest clear example of temple fires kept according to Zoroastrian regulations" (Boyce and Grenet, Zoro astrianism, p. 178). She did not consider the structure a fire temple, but an image temple with accessory fire chambers.
In the lower city at Persepolis (Fârs) about 300 m. from the terrace, Herzfeld uncovered a structure which he called the "Fratadara Temple" (PLATE IV; Schippmann, 1971, pp. 177-85, fig. 24). Room 5 of the western part of the complex contained a stepped-stone pedestal, perhaps the base of a stepped fire altar (fig. 4; Kleiss, figs. 1-2). The date (Achaemenid-Seleucid) of the entire complex has often been debated (Stronach, 1985, pp. 612-17, for recent survey of the evidence). Boyce (Zoroastrianism II, pp. 226-27), preferring a Achaemenid date for the whole of the western complex, identified it as an image shrine dedicated to Anahita, and characterized Room 5 as a "side chapel, with a different dedication." In a subsequent publication (Boyce and Grenet, Zoroastrianism, pp. 117-18), she was less certain of both the date and function of the building.
The enigmatic structure unearthed by Marcel-Auguste Dieulafoy (q.v.) at Susa (Khûzestân) has often been identified as a temple (Schippmann, 1971, pp. 266-74, 496, fig. 38, for discussion and references). Dieulafoy's plan may show the possible presence of an altar between the four columns of the main room, but his text makes no reference to such an installation (Stronach, 1985, p. 621; Schippmann, 1971, fig. 38). Dieulafoy (pp. 411-15) dated the building to the Achaemenid period based upon the style of the architectural fragments, but a Parthian date has also been suggested by David Stronach (1985, pp. 619-21, suggesting a date in the 2nd cent. B.C.E; Yamamoto, 1979, pp. 37-38, follows Dieulafoy's dating).
Various literary sources, such as the inscriptions of Shapur I and his high priest Kirdêr (Boyce, Zoroastrianism II, pp. 222-23; Duchesne-Guillemin, in Camb. Hist. Iran, pp. 876-79) and Nâma-ye Tansar (ed. Mînovî, p. 68) show that the establishment of Sasanian rule involved the destruction of many local dynastic fires. This seems to have been part of the early Sasanian kings' desire to establish their dynastic fire as a symbol of the unity of the whole empire. The Sasanian period also saw, however, the foundation of new fires, many established to eradicate the worship of idols and to substitute in their place the worship of fire (see ÂTAŠ for types of fire identified in the Sasanian period).
Pahlavi books repeatedly mention three famous sacred fires for the Sasanian period: Âdur Burzên-Mihr (in Parthia), Âdur Farnbâg (in Fârs), and Âdur Gušnasp (in Media; qq.v.). Most scholars assume that they pre-date the Sasanian period, and all three had traditions connecting them with the original creation in Zoroastrian myth, but evidence is lacking as to the exact times of their actual founding (see Schippmann, 1971, pp. 86-94, 340-54; Boyce, 1975, pp. 459-60; Boyce and Grenet, Zoroastrianism, pp. 74-81; Yamamoto, 1979, p. 42; idem, 1981, pp. 74-75, 84-85). The relocated Âdur Gušnasp has been identified as the site of Takht-e Solaymân in Azerbaijan (see below); the exact locations of the other two fires are unknown.
Depictions of fire altars are common, especially on seals and coinage (Yamamoto, 1979, p. 46; idem, 1981, pls. 18-35). Seals almost always depict the stepped altar type (one or two steps; Yamamoto, 1981, pp. 68-71). Coins show variations of the familiar stepped altar with fire blazing atop it; the king stands to the left of the altar in an attitude of reverence, a variety of personages to the right (ibid., pls. 2-9). Sasanian issues of the 4th and 5th centuries sometimes show a human bust in the flames on the altar or on the pillar (ibid., pp. 69-70, reviews the evidence). The appearance of fire altars on official coinage shows again the interconnectedness of fire cult and political legitimacy for the Sasanian kings. The rock relief of the investiture of Ardašîr I at Fîrûzâbâd includes a fire altar in the form of a large bowl supported by a pillar on a squared stand (ibid., pl. 1).
Yumiko Yamamoto (1981) and Klaus Schippmann (1971, pp. 475-76, 499-515) have collected the material concerning fire altars and fire worship for the Sasanian period. The stepped altar remained the standard type in the period. Schippmann (p. 505, and Table 3) identified forty-nine or fifty fire temples for the Sasanian period, the great majority of them occurring in Fârs and the neighboring provinces of Khûzestân and ´Erâq-e ´Ajam. Both Schippmann (pp. 499-510) and Yamamoto (1981, p. 79) identified the ubiquitous square structure having an arch in each wall (ch "four arches," q.v.) and piers in each corner supporting on squinches a dome (gonbad) as the place where the sacred fire was housed in the fire temple. Stronach (1966, pp. 219-20, 226; idem, 1985, pp. 623-27) has discussed the standard juxtaposition of an open ch and a protected ataškada (repository for the sacred fire). Especially in southwest Persia, fire temples seem to have been a ubiquitous feature of the landscape. Khosrow II Parvêz alone is reported to have founded 353 fire temples (Duschesne Guillemin, 1983, p. 896).
Of the fifty fire temples Schippmann (1971) identified for the Sasanian period, forty-one of them are ch, the great majority of them located in Fârs and Erâq-e ´Ajam (q.v.; note, however, the possible fire temples in central Asia at Kurgan Tepe [an earlier square altar base replaced later by a circular one; see Pugachhenkova] and Panjîkant [fire cult associated with image worship; see Škoda]). Few of the fifty structures catalogued by Schippmann are well preserved, excavated and/or well published. The preservation of actual fire altars from these sites is also rare.
The most famous (and one of the largest and best preserved) fire sanctuary from the Sasanian period is the complex located at Takht-e Solaymân in Azerbaijan, to which place the famous fire Âdur Gušnasp seems to have been moved sometime in antiquity (Schippmann, 1971, pp. 309-57, figs. 41-46; Yamamoto, 1981, p. 75, figs. 1-2; Duchesne-Guillemin, in Camb. Hist. Iran, p. 895). In a small, well-protected room in the western part of the complex was found a large three-stepped base of an important stepped fire altar, perhaps even that of the main fire itself (Boyce, 1975, pp. 464-65; Yamamoto, 1981, p. 75, pl. 48). The site has also yielded several other possible fire altars (Huff).
In the Sasanian period there were also open-air religious sites which have yielded fire altars. The most celebrated are the twin altars at Naqš-e Rostam near Persepolis in Fârs (PLATE III ). Each of the cubic altars carries a relief of an arch on each of its sides ( Yamamoto, 1981, p. 97, pl. 53; Stronach, 1966, pp. 219-20, 227, fig. 6, defending Erdmann's thesis that the design of the altars emulates the ch). A different form of the same type occurs singularly at Kûh-e Šahrak (Fârs; Stronach, 1966, figs. 2-5). Rock and stone monuments of various shapes which support fire bowls on their top surfaces have been documented at Darra-ye Barra (q.v., Fârs; Stronach, 1966, p. 224, figs. 8-11), Tang-e Karam (Fârs; Stronach, 1966, pp. 224, figs. 12-13; cf. contra, Trümpelmann, pp. 21-22), Naqš-e Rostam (Stronach, 1966, p. 224, fig. 19; Vanden Berghe, p. 514), Kûh-e Ayyûb (Fârs; Stronach, 1966, p. 226), and Bîšâpûr (q.v., Fârs; Stronach, 1966, p. 223). Stepped altars supported on pillars occur at Qanât-e Bâgh (Fârs; Stronach, 1966, p. 224; Vanden Berghe, pls. 60-63), Pengân (Vanden Berghe, p. 514), Kûh-e Hosayn (Fârs; Vanden Berghe, p. 514), Kûh-e Rahmat (Fârs; Vanden Berghe, p. 514), Šamîrûn (Vanden Berghe, p. 514), near Sîrâf (Fârs; Gaube, pp. 154-57, pl. 32.1 and 32.5), and Persepolis (Scerrato, p. 726; Schippmann, 1969; Vanden Berghe, p. 514). Fire bowls in the tops of these monuments indicate that they held fire, but the exact manner in which they were used is unknown. Because all these free-standing monuments stand in the open, certainly none of them could have held an ever-burning fire (Vanden Berghe, pp. 516-18, for speculations on their uses). The platforms in the open air sanctuaries at Bard-e Nešânda (q.v., Khûzestân) and Masjed-e Solaymân (Khûzestân) have also been identified as supports for (now missing) fire altars (Schippmann, 1971, pp. 233-51, figs. 33-36, for Masjed-e Solaymân [dating disputed]; 251-58, fig. 37, for Bard-e Nešânda).
(for cited works not given in detail, see Short References"):
Copyright © 1998-2015 The Circle of Ancient Iranian Studies (CAIS)