MAMMALS iii. The Classification of Mammals and the Other Animal Classes according to Zoroastrian Tradition

The first written information about certain animals in Iran comes from the Zoroastrian literature, according to which the entire animal kingdom is divided into two classes: “beneficent animals” and “evil animals.”



iii. The Classification of Mammals and the Other Animal Classes according to Zoroastrian Tradition

The first written information about certain animals in Iran comes from the Zoroastrian literature, according to which the entire animal kingdom is divided into two classes: “beneficent animals” (Av. gao.spəṇta-; MP. gōspand), such as cattle, dogs, and other domestic animals, which all play important roles in promoting human welfare, and, in the opposing camp, “evil animals” (Av. xrafstra-, MP. xrafstar), creatures potentially harmful to human beings and their agricultural crops or else hideous and obnoxious in aspect. Beneficent animals can be tamed and eaten or sacrificed to the deities. Evil animals are regarded as inedible, unacceptable as sacrificial offerings; they should be killed. The bipartition of animals into beneficent and harmful represents without doubt one of the most original aspects of the Iranian religious worldview as described in the available Zoroastrian literature.


A systematic animal classification in ancient Iranian is found in the Avesta and with more details in Middle Persian texts. In the Avesta there is a distributive description of the whole physical world, with the exception of humans, in Yasna 71.9. It consists of yazamaide (“we worship”) formulae, which only mentions the names of various entities worshipped, without saying much about them:

We worship all the waters (which are) in springs and flowing in rivers,
We worship all the plants in their shoots and roots,
We worship the whole earth, We worship the whole sky, We worship all the stars, the moon, and the sun,
We worship all the endless lights,
We worship all the animals, those living in the water and those living underground, the flying ones, those roaming in freedom (or: wilderness), and those worthy of (or: attached to) the pasture
(Darmesteter,  I, pp. 431-32; Humbach, 1991, I, p. 148)

Yašt 13.74 provides a list of animals divided into domestic (pasuka-) and wild (daitika-); this opposition also occurs in Yasna 39.1-2 (Darmesteter, I, p. 269; Humbach, 1991, I, p. 148). In Yašt 8.48 the creatures are first divided into those living underground and those above the earth, followed by the first four members of the standard list: those living in the water, those living underground, the flying ones, and those roaming in freedom. Here the last member, that is, those worthy of the pasture, is missing. The reason could be either that the quintuple list is an expansion of an older quadruple list preserved only here or that the last member was left out for some unknown reason (Schmidt, 1980, p. 215).

In the Visperad (1.1 and 2.1), the chiefs or prototypes (ratavō) of the animals are invoked and revered. The categories of animals are the same as in Yasna 71.9.  In the Avesta the chiefs are never identified, but the Middle Persian translation of the Visperad gives the following, although the zoological identification of some of these animals is not certain:

ābīg, kar māhīg (living) in water, sturgeon
unīg, kākomag (living) in burrows, stoat/ermine
wāyindag, karšift flying (birds), hawk/falcon
frāx-raftār, xargōš roaming in freedom, hare/rabbit
čarag-arzānīg, xarbūz worthy of the pasture, goat
(Darmesteter, I, p. 444)  

Middle Persian texts other than the translations of the Avesta offer more details on animals (Table 1). A remarkable classification, based on taming degree, foot forms, habitat, color, and morphological differences, is found in the Bundahišn; it demonstrates the method and rigor of the categories and reveals details with which the redactors depict certain animals that were in the process of disappearance.  Several chapters of this book are dedicated to animals: chapter 13 on the nature of the five kinds of animals (TD1, fol. 38r; TD2, fol. 48v; DH, fol. 184v; tr. Anklesaria, pp. 117-26); chapter 22 on the nature of the evil animals (TD1, fol. 59r; TD2, fol. 73r; DH, fol. 200r; tr. Anklesaria, pp. 182-88); chapter 23 on the nature of the wolf species (TD1, fol. 61r; TD2, fol. 75v; DH, fol. 201r; tr. Anklesaria, pp. 188-91); and chapter 24 on fabulous creatures (TD1, fol. 61v; TD2, fol. 76v; DH, fol. 201v; tr. Anklesaria, pp. 190-205).  Chapter 13 of the Bundahišn has a close parallel in another Zoroastrian text, the Selections of Zādspram (chap. 3.50-65). 

According to the lost Avestan Dāmdād Nask “the creating of the creation” (DKM, 3, p. 681, 11-20), of which summaries are given in the Bundahišn (chap. 13.9) and Zādspram (chap. 3.51), the beneficent animals were grouped into five classes, comprising domestic animals and wild animals as well as birds, fishes, and burrowing animals; and they were again subdivided into genera and species. 

A slightly different classification of beneficent animals is given in The Pahlavi Rivāyat Accompanying the Dādestān ī Dēnīg ([PRDD], chap. 46. 22). This classification is described first in the creation myth according to which the beneficent animals are all descendants of the sole-created cow/bull (gāw ī ēwdād), the first animal to live on earth, which was killed by the Evil Spirit; all species of beneficent animals were brought forth from his purified seed. The animal world is first divided into a tripartite classification, then into the traditional quintuple one, and a more modern one, which is also quintuple.  The latter is then further subdivided twice, into a list of eleven or twelve, and finally into individual species.

The Bundahišn, for the threefold division, uses the term kardag, which is absent in Zādspram. For the modern quintuple division, generally the term ēwēnag, genus, is used. For the 11 or 12 subdivisions, the Bundahišn has no term; Zādspram uses bahr, group; for the individual species both have sardag.  Both sources use the same Avestan list. Zādspram follows the order known from the Avesta (see above; in Yasna 71.9, Visperad 2.1). In Bundahišn 13.9, however, the sequence is in part inverted:

čarag-arzānīg   worthy of the pasture
frāx-raftār roaming in freedom
wāyindag flying (birds)
ābīg (living) in water
unīg (living) in burrows

The second quintuple classification is basically the same in both sources, but with different terminology:

Bundahišn (chap. 13) Zādspram (chap. 3.54)
dō-kāft-pāy, double cloven-footed gird-sumb, round-hoofed
xar-pāy, ass-footed dōgānag-sumb, double-hoofed
panj angurāg panjag, five-fingered  panj-čang, five-clawed
wāyendag, flying/bird murw, bird
ābīg, (living) in water māhīg, fish

This is the basis for the further subdivisions into genera and species; the latter are stated to total 282.  Although Zādspram and Bundahišn agree in most respects, there are major differences. One is that the Bundahišn numbers the species from 1 to 12; Zādspram does not give any number, but names select ones: “(Ohrmazd) divided [animals] into genera: as the round-hoofed ones are one single (genus), they are all called horse; the double-hoofed are many, such as the camel, the bovine, the sheep, the goat, and other double-hoofed (animals); the five-clawed are the dog, hare, rat, sable, (and) others; and birds, and then fishes” (chap. 3.54; Gignoux and Tafazzoli, eds., 1993, pp. 50-51). The genera named are eleven; but since it is explicitly stated that there are others in the classes of the double-hoofed and the five-clawed, it is possible that Zādspram did not know the fixed number 12, as the Bundahišn has it.

There is also a discrepancy between the number of species enumerated (178) and the stated total of 282 in the Bundahišn; species are divided within species to make a total of 282. According to Zādspram (chap. 3.55), there are animals, not listed, “of revealed names and unrevealed names” (paydāg-nāmān a-paydāg-nāmān), altogether 282 species, with species within species numbering 10,000 kinds. From this statement it is evident that the names of many species in the older tradition were not known to the later authors, even if “10,000” is regarded as a symbolic number rather than an actual total. The Bundahišn implies that the 95 species missing (that is, the difference between the total of 282 species and the 178 named species) must be made up by a subdivision of species with species, presupposing that the system of 12 or 11 genera is closed.

It should be noted that in the Bundahišn twelve large species are mentioned, while Zādspram lists only eight large species (numbers 2, 3, 4, and 11 of the Bundahišn are absent in Zādspram). However, both texts agree on the number of 282 species, which means that their common source did not name more either and that this number was a well-established tradition.  According to Hanns-Peter Schmidt (1980, p. 225):

the number 282 can be of different provenience than the classificatory list and may be a symbolic number: The purification and dedication recitals for the collection of the Bahrām-Fire number 1128, and this is four times 282.  282 is also very close to the number of days of human pregnancy, that is, 280 or ten lunar months. 282 may be composed of 280 days of gestation and the days of conception and birth. What is still more striking is that Bundahišn XV, 12 states that humans, horse species, and bovines are born in ten months. The numbers 282 and 1128 would then have a specific meaning in the process of creation.

The differences between the two versions suggest that there was a more elaborate system behind it. In the Sasanian period almost all branches of science were influenced by non-Iranian traditions, and the case of zoology is not an exception. The classification found in Middle Persian texts owes much to the ancient Greek tradition. The attempt by redactors of these texts to construct a coherent animal classification may indicate that prior to their work this had not been done. The comparison of the Bible classification of animals with Middle Persian texts yields numerous passages closely parallel, both in form and content. It has been suggested that such correspondence was under the influence of the Academy of Athens, which had moved to Iran in 529 CE after the emperor Justinian closed it. Its philosophers were welcomed in the Persian court by Ḵosrow I (r. 531-579), who, according to the historian Agathias, was a great admirer of the works of Plato and Aristotle (Christensen, 1944, p. 422). 


The evil animals (xrafstars; Table 2) are created by Ahriman, the Evil Spirit. He created his creatures from material darkness, in the form of a frog—black, ashen, worthy of darkness, and evil, like the most sinful-natured xrafstar. (Bund., TD1, fol. 5r.; TD2, fol. 74r; tr. Anklesaria, p. 44).

The word xrafstar (Av. xrafstra-, MP. xrafstar), evil animal, has been variously interpreted, and its exact meaning is much disputed. According to the traditional interpretation Zarathushtra used it in the Gāthās (Y. 28.5; 34.5) pejoratively for “the enemies of the religion” (Bartholomae, AirWb. col. 538). Some scholars see in it a reference to evil animals whom Zarathushtra chose to exclude from the sacred place of worship and hence symbolically from the whole good creation (Humbach, 1991, pt. I, pp. 118, 140; pt. II, pp. 24, 107). H. W. Bailey suggested that the term is a derivation of an Indo-European verbal base *(s)ker, (s)kerp-, (s)krep- “to bite, sting, cut,” which looks promising (Bailey, 1970, pp. 25-28). But, the krt suffix -tra- is never formed on an s-extension. There is a great likelihood that there were two different terms in Iranian daevic vocabulary, as we have *prystr in MMPers., and plstl is attested in the Bundahišn (TD 2, fol. 73v.9). Therefore, the term can be derived from fra-pt-tar, “things that fly-creep” (Gershevitch, 1954, 62, 246; Moazami, 2005, p. 302). Jean Kellens proposed that the term approximately means “affreux, sauvage” (Kellens and Pirart, 1990, II, p. 231).

In the Young Avesta and Middle Persian texts, the term xrafstar is used specifically for reptiles and amphibians such as frogs, scorpions, lizards, and snakes, and insects such as ants, beetles, and locusts. In general, any animal that crept, crawled, pricked, bit, or stung, and seemed hideous and repulsive to human beings, was xrafstar. Predators such as felines and wolves were also creatures of Evil Spirit, but in Middle Persian texts they are referred to as dadān, “wild animals, beasts.”

The Bundahišn states that the physical existence of the evil animals, the shining of their eyes, and their psychic wind are of the Beneficent Spirit, but their spirit of sinfulness and malevolence are of the Evil Spirit, which is of great advantage, since when humans see them they can either kill or avoid them (Bund., TD1, fols. 58v., 59r.; TD2, fols. 73r., 73v., tr. Anklesaria, p. 183).

Evil animals did not receive the same systematic attention as did beneficent animals, but a classification of them is found in the Avesta (Vidēvdād, chap. 14.5–6; Moazami, 2014, pp. 348-49) and later Zoroastrian literatures. This classification starts with snakes, followed by reptiles and amphibians, then insects and worms, and ends with flying insects:

Vidēvdād, Avestan Vidēvdād, Middle Persian
aži udarō. θrąsa az ī uδraišag, snake crawling on its belly
aži spaka kahrpuna az sag karbunag, snake *dog lizard
(dog-shaped lizards?)
kasiiapa kašawag, tortoise
vazaɣa dāδmainiia wazaɣ ān ī zamīg, frog puffed-up/living on earth
vazaɣa upāpa wazaɣ ī abīg, water-frog
maoiri dānō karša mōr dānkeš, corn-carrying ant
maoiri araēka kutaka mōr ī halag, small ant, swift-moving ants
pazdav guθō/gāzō.vərəta pazūg guhward, dung worm
maxši ərəɣaiiti maxš ērang, raging/repelling fly

The Bundahišn assigns separate chapters for xrafstrān “evil animals” and gurgān, “wolves.” It divides evil animals according to whether they live in water, on the earth, or in the air. The worst of each type are the frog (in water), the many-headed dragon (on the earth), and the winged snake (in the air) (Bund., TD1, fol. 59r.; TD2, 73r., tr. Anklesaria, pp. 182-83).

The Pahlavi Rivāyat Accompanying the Dādestān ī Dēnīg ([PRDD], ed. Williams, 1990, chap. 21a.3-46) lists the animals in a different sequence from that given by the Vidēvdād and Bundahišn (TD1, fols. 59r., 59v.; TD2, fol.73 r.; DH, 200r.; tr. Anklesaria, p. 185) and includes species that may be later additions: 

According to the Zoroastrian worldview, evil animals are not always harmful, and their existence is not completely useless. The Beneficent Spirit in his omniscience diverts some of these evil animals to the benefit of his creatures (Bund., TD1, fol. 60v; TD2, fol. 75v; tr. Anklesaria, p. 189). For instance, their bodies are used in the composition of remedies with a mixture of drugs, because they are from the four beneficial elements: water, earth, wind, and fire. (Bund., TD1, fol. 59r; TD2, fol. 73v; Anklesaria, p. 183; Gignoux, 2001, p. 53).

Among wild animals created by the Evil Spirit, the wolf appears to have a close relation with the Evil Spirit and demons (see DĒW). Its creation is directly attributed to the Evil Spirit, which desired to create the species in secret, in the form of fever, disease, and other evils, so that humans would not see them when they come upon them. (DkM, p. 92; de Menasce, chap. 95, p. 101.) The Evil Spirit produced the wolf in fifteen species (Bund., TD1, fols. 60v, 61r; TD2, 75v, 76r; DH, fol. 201r, tr. Anklesaria, p.189.): gurg, wolf; gurg ī syā, black wolf ; babr, tiger; šagr, lion; palang, panther; yōz, cheetah; xaftar, hyena; tōrag, jackal; gār-kan, cave digger; karzang, crab; gurbag, cat; būg, owl; gurg ī ābīg, aquatic wolf; kōsag, shark; and gurg ī čarnak/čahārwāg, wolf (?).

The Bundahišn states that some of the evil wolf species are avoided out of fear, but others can be tamed, such as the elephant and the lion; the lion was created by the Evil Spirit, but according to a model established by the Beneficent Spirit (TD1, fol. 61v.; TD2, fol. 76v.; tr. Anklesaria, p. 191; The Persian Rivayats of Hormazyar Framarz and Others, p. 270).

The lion had a venerable place in Iran as a symbol of kingship; undoubtedly in part for the simple reason that the lion was observable as a powerful and beautiful wild animal (see Root, 2002, pp. 169-209).  The elephant, which was imported to Iran from India, was probably counted among the evil animals because of its unfamiliar appearance, and classifying it with the lion must have been influenced by the association of both animals with royalty, for only kings kept and used them for hunting and warfare.

The creation of the bear, monkey, and some other evil creatures are attributed to the union of Jam (JAMŠID), the mythical king of Iran and his twin sister Jamīy with demons (Bund. TD1, fol. 44r, 44v.; tr. Anklesaria, chap.14B, pp. 136-37; [PRDD] 8e9, p. 13).


B. T. Anklesaria, ed. and tr., Zand-Āāsih: Iranian or Greater Bundahišn, Bombay, 1956.

T. D. Anklesaria, The Bûndahishn TD2, Bombay, 1908.

H. W. Bailey, “A Range of Iranica,” in W. B. Henning Memorial Volume, ed. M. Boyce and I. Gershevitch, London, 1970, pp. 25–28.

A. Cameron, “Agathias on the Sassanians,” Dumbarton Oaks Papers, vol. 23, 1969-1970, pp. 67-183.

A. Christensen, L’Iran sous les Sassanides, Copenhagen, 1944.

J. Darmesteter, Le Zend-Avesta, 3 vols., Paris, 1960.

E. B. N. Dhabhar, The Persian Rivayats of Hormazyar Framarz and Others, Bombay, 1932, repr. 1999.

[DkM] D. M. Madan, ed., The Complete Text of the Pahlavi Dinkard, 2 vols., Bombay, 1911.

I. Gershevitch, A Grammar of Manichean Sogdian, 1954.

Ph. Gignoux, Man and Cosmos in Ancient Iran, Serie Orientale Roma no. 91, Rome, 2001.

Ph. Gignoux and A. Tafazzoli, Anthologie de Zādspram, Paris, 1993.

H. Humbach, The Gathas of Zarathustra and the Other Old Avestan Text (in collaboration with J. Elfenbein and P. O. Skjærvø), 2 vols. Heidelberg, 1991.

J. Kellens and E. Pirart, Les Textes Vieil-Avestiques, 3 vols., Wiesbaden, 1988-91.

J. de Menasce, Le Troisième livre du Dēnkart, Paris, 1973.

M. Moazami, Wrestling with the Demons of the Pahlavi Widēwdād, Transcription, Translation, and Commentary, Leiden and Boston, 2014.

Idem “Evil Animals in the Zoroastrian Religion,” History of Religions 44/4, 2005, pp. 300-317.

[PRDD] A. V. Williams, The Pahlavi Rivāyat Accompanying the Dādestān ī Dēnīg, 2 vols., Copenhagen, 1990.

M. C. Root, “Animals in the Art of Ancient Iran,” in A History of the Animal World in the Ancient Near East, ed. B. J. Collins, Leiden, 2002.

H. P. Schmidt, “Ancient Iranian Animal Classification, ” Studien zur Indologie und Iranistik 5/6, 1980, pp. 209-44.

(Mahnaz Moazami)

Originally Published: May 5, 2015

Last Updated: May 5, 2015

Cite this entry:

Mahnaz Moazami, "MAMMALS iii. The Classification of Mammals and the Other Animal Classes according to Zoroastrian Tradition," Encyclopædia Iranica, online edition, 2015, available at (accessed on 05 May 2015).