Written by The scale of the loss
The Avesta, the sacred corpus of Zoroastrianism, is the oldest body of Indo-Iranian religious literature still in continuous use. In its Sasanian form — the form it had reached by the 6th century CE, after roughly a thousand years of accumulation, redaction, and commentary — it consisted of twenty-one nasks (books), organized by subject, covering ritual, law, cosmology, ethics, eschatology, history, medicine, astronomy, and the lives of the prophets and kings.
What survives today is approximately one quarter of that corpus.
Three quarters of the Avesta was destroyed.
The destruction happened in two phases, separated by nearly a thousand years. The first was the burning of Persepolis by Alexander of Macedon in 330 BCE, which destroyed the royal archives where one major recension of the religious literature was kept. The second, far more comprehensive, was the Arab Muslim conquest of Persia in the 7th century CE and the centuries of attrition that followed, during which the priesthood was killed or scattered, fire temples were destroyed, libraries were burned, and the chains of memorized transmission that preserved the texts were broken.
We know what was lost in part because the Denkard — a 9th-century encyclopedic work compiled by Adurfarnbag-i Farroxzadan and later expanded by Adurbad-i Emedan — preserves a chapter-by-chapter summary of the entire twenty-one nask Sasanian Avesta. The Denkard‘s book 8 walks through each nask, describes its contents, and in many cases lists the topics each chapter addressed. Reading that summary alongside the surviving Avestan texts is reading a catalog of erasure. We have a table of contents for the destroyed library.
This post walks through what was in that library, what survives, and what was lost.
The structure of the Sasanian Avesta
The twenty-one nasks were divided into three groups of seven, corresponding to the three lines of the Ahuna Vairya (the most sacred Zoroastrian prayer):
The seven gāhānīg nasks — concerned with the spiritual and theological dimensions of the religion, drawing most directly on the Gathas of Zarathustra.
The seven hadag-mānsrīg nasks — concerned with the intersection of the spiritual and the material, ritual practice, and the application of theology to lived life.
The seven dādīg nasks — concerned with law, civil order, and the practical regulation of society according to Zoroastrian principles.
What survives today, fully or partially, are: the Yasna (the central ritual liturgy, including the Gathas), the Visperad (a liturgical supplement), the Vendidad (the Videvdad, “Law against the Demons” — laws of purity and ritual), the Yashts (hymns to specific Yazatas), the Khorda Avesta (the “Little Avesta,” prayers for daily use), and fragments preserved in later Pahlavi commentaries.
What was lost, by nask:
The First Group: The Spiritual Nasks
1. Sūdgar Nask. Commentary on the Gathas, organized stanza by stanza, providing the religious-historical context for each verse. Lost. The Denkard preserves a summary of its contents, including legendary material on Zarathustra’s life that we have only in fragmentary form elsewhere. What we lost: the canonical commentary tradition on the oldest Zoroastrian hymns.
2. Warštmānsr Nask. A second commentary on the Gathas, with different emphases — apparently more concerned with the practical ethical applications of Gathic teaching. Lost. The Denkard preserves the outline.
3. Bag Nask. A third Gathic commentary, organized around the divine names and titles. Lost. We have the Bag Yasht — a related liturgical text — but the full commentary is gone.
4. Damdat Nask. Cosmology. The creation of the world, the structure of the spiritual and material realms, the relationships between the seven creations (sky, water, earth, plants, animals, humans, fire), the role of the Amesha Spentas in cosmic order. Lost. The Bundahishn preserves Sasanian-period cosmological material that probably overlaps with the Damdat Nask‘s content, but the full text is gone. What we lost: the canonical Zoroastrian cosmological treatise.
5. Nadar Nask. Astronomy and astrology. The movements of the heavens, the calendar, the religious significance of celestial phenomena. The Greek classical tradition recognized Persian astronomical knowledge as advanced — Eudoxus and others studied with Persian magi. The Nadar Nask was apparently the foundational text of Zoroastrian astronomy. Lost. What we lost: the canonical text of one of the ancient world’s most respected astronomical traditions.
6. Pājag Nask. Religious calendar and the ritual cycle. The structure of the Zoroastrian liturgical year, the festivals, their theological meanings, their proper observances. Lost. We have fragments of calendar material in the surviving Yashts and in later Pahlavi sources, but the systematic treatment is gone.
7. Ratuštāitīh Nask. The structure of priestly authority and the religious hierarchy. The roles of different orders of priests, the conditions of valid priesthood, the lines of religious succession. Lost. What we lost: the canonical text on the institutional structure of the Zoroastrian priesthood — particularly painful given that it was the priesthood itself that was decimated in the conquest.
The Second Group: The Intermediate Nasks
8. Spand Nask. The life of Zarathustra. Birth, childhood, calling, prophetic mission, teaching career, death. The canonical biography of the prophet. Lost. Fragments survive in Denkard book 7 and in the Zaratusht-nameh (a much later New Persian poetic version). Most of what we now claim to know about Zarathustra’s life as preserved in the religious tradition came from this nask. What we lost: the prophet’s own scripture-preserved biography.
9. Bagān Yašt Nask. Hymns to the Yazatas (divine beings under Ahura Mazda). The surviving Yashts in the Khorda Avesta are partial — twenty-one of an unknown larger original set. The Bagān Yašt Nask contained the full canonical collection. Lost (in part). What we lost: the majority of the hymnic literature to the Yazatas, the lesser divinities of Zoroastrian theology.
10. Nikādum Nask. Civil and criminal law. The Zoroastrian legal code on theft, murder, assault, contract, marriage, inheritance. Lost. What we lost: the canonical Zoroastrian civil law of the late antique world’s most populous and longest-lived empire.
11. Ganabā-sar-nigad Nask. Specific applications of law to particular cases. A case-law supplement to the Nikādum. Lost.
12. Hūspāram Nask. Priestly law — the legal regulations specific to the priesthood, including ritual purity, fees, succession, discipline. Partially preserved through quotation in later Pahlavi legal texts (Madigan-i Hazar Dadestan) and in the Vendidad. What we lost: most of the canonical priestly law.
13. Sakādum Nask. The legal code’s relationship to religious doctrine — the theological foundations of Zoroastrian law. Lost. What we lost: the canonical Zoroastrian theology of jurisprudence.
14. Vidēvdād (Vendidad). The “law against the demons” — laws of ritual purity, regulations on corpse-handling, treatment of the body, definitions of pollution and purification. This one survives. It is one of the few complete nasks preserved. The fact that the most ritualistic and (to modern readers) most strange of the nasks is the one that survived in full, while the cosmological, astronomical, and legal nasks were lost, is its own commentary on what kinds of texts the late surviving priesthood prioritized for preservation under conditions of catastrophe.
The Third Group: The Legal Nasks
15. Cidrast Nask. Religious law in its broadest theological frame. Lost.
16. Spand Nask (a different Spand from #8). Confused in the manuscript tradition; some material may overlap with the biographical Spand. Lost.
17. Bagān-yašt Nask (sometimes confused with #9). Lost or partially overlapping with surviving Yasht material.
18. Nigādum Nask. The right ordering of religious life. Lost.
19. Duzd-sar-nigad Nask. Penalties for theft and related offenses, with religious-legal commentary. Lost.
20. Husparum Nask (variant of #12). Lost.
21. Sakātūm Nask. The final nask. The summation of the legal corpus. Lost.
The numbering and exact content of the third group is itself uncertain because the Denkard‘s summary is less detailed for these nasks, and modern scholarship has had to reconstruct what they contained from cross-references. The fact that we are uncertain even about the table of contents of the third group is itself evidence of how thoroughly the corpus was destroyed — not only the texts but the bibliographic memory of them is degraded.
The first burning: Persepolis, 330 BCE
The first major destruction came at the hands of Alexander of Macedon. In 330 BCE, after defeating Darius III and capturing the Persian heartland, Alexander entered Persepolis — the ceremonial capital of the Achaemenid empire and the repository of one major recension of the Zoroastrian religious literature, kept in the royal archives.
What happened next is recorded with remarkable consistency across the Greek and Roman historians, though with varying interpretations. According to Plutarch (Life of Alexander 38), Diodorus Siculus (17.72), Quintus Curtius (5.7), and Arrian (Anabasis 3.18) — four independent traditions, drawing on Alexander’s own court historians — Alexander’s army, after a feast, set fire to the palace complex. The Athenian courtesan Thais is named in several sources as urging the burning, framed as revenge for Xerxes’ burning of the Athenian Acropolis in 480 BCE. Alexander participated.
The fire destroyed not only the palace but the archives. Pliny the Elder (Natural History 13.21) and other later sources record that vast quantities of Persian written material — religious, administrative, scientific — were lost in the destruction.
Plutarch records that Alexander later regretted the burning. Whether the regret tradition is historical or a literary moralization by later historians is debated, but it is preserved consistently enough that something like it likely circulated in Alexander’s own circle. Arrian, writing more critically, treats the burning as straightforwardly a moral failure on Alexander’s part — an act of cultural devastation that even the conqueror’s most loyal historians could not fully justify.
The Zoroastrian tradition’s own memory of this event is preserved in several Pahlavi sources, most notably the Denkard and the Arda Viraf Nameh. These texts identify Alexander — Alexandar i guzastag, “Alexander the accursed” — as one of the great catastrophes of Iranian religious history, alongside the Arab conquest. The Denkard records that Alexander destroyed the Avesta that had been written down in the time of Vishtaspa (Zarathustra’s royal patron) and dispersed the priests who knew the texts. After the destruction, the religion had to be reconstructed from oral memory and from such partial copies as had been preserved in outlying temples.
The Parthian and Sasanian dynasties, over the following nine centuries, undertook this reconstruction. The Sasanian Avesta — the twenty-one nasks described above — was the result of this thousand-year project of recovery. Texts were collected from surviving priestly memory, from outlying communities, from temples that had escaped Alexander’s destruction. The corpus was redacted, organized, written down in the specifically devised Avestan script (developed in the Sasanian period precisely to preserve the texts accurately), and copied into the great fire-temple libraries.
The second burning: the Arab conquest, 633–651 CE and after
When the Arab Muslim armies entered Sasanian Persia in 633, the Avesta they encountered was the rebuilt corpus — the work of nearly a thousand years of restoration after Alexander. What followed across the next several centuries was the destruction of that work.
The pattern was systematic, though not always centrally directed:
The royal libraries were lost. The Sasanian capital Ctesiphon fell in 637. The royal libraries, which held the most authoritative recensions of religious and scientific literature, were dispersed or destroyed. Some material was looted, some translated into Arabic in subsequent centuries (particularly under the Abbasids’ translation movement), but the bulk was lost.
The fire-temple libraries were destroyed or converted. Each major fire temple held its own copies of religious literature. As fire temples were destroyed or converted to mosques across the 7th–10th centuries, their libraries went with them. The Persian historian al-Biruni (973–1048), writing several centuries later in al-Athar al-Baqiya, refers to wholesale destruction of Zoroastrian books in the early conquest period.
The priesthood was decimated. The Zoroastrian religious literature was preserved through a combination of written manuscripts and oral memorization. Priests memorized vast quantities of Avestan text — the entire Yasna, the Vendidad, the Yashts — and transmitted them through hereditary chains of teacher-to-student. When the priesthood was killed, exiled, or pressured to convert, those memorization chains broke. Texts that depended on living priestly memory for their preservation were lost when the memorizers died.
Conversion pressure stripped the lay community of literacy in the sacred languages. As populations converted to Islam, knowledge of Avestan and Pahlavi atrophied. Even where physical manuscripts survived, the readership shrank to a small priestly remnant. By the medieval period, much of the original meaning of even the surviving texts had become obscure to the communities that still nominally possessed them.
The Parsi flight to Gujarat saved a remnant. The Qissa-i Sanjan records the migration of a Zoroastrian community from Iran to Gujarat in the 8th–10th centuries. They carried their sacred fire and what manuscripts they could. The Parsi community in India became, over subsequent centuries, the primary preserver of the surviving Avestan corpus. Most of what we have today — including the manuscripts that allowed European scholars from Anquetil-Duperron onward to access the Avesta at all — was preserved by this community in conditions of exile.
What it took to build, and what it took to destroy
The Sasanian Avesta represented:
A thousand years of accumulation, redaction, and commentary across the Achaemenid, Parthian, and Sasanian dynasties.
A specially devised script, created in the Sasanian period, which recorded Avestan with an unprecedented level of phonetic precision — preserving even fine acoustic distinctions in the recitation that had been transmitted orally for centuries.
The institutional infrastructure of an empire’s worth of fire temples, each with its library and its priestly establishment.
A hereditary priesthood numbering in the tens of thousands, organized in chains of memorization stretching back across generations, holding in living memory texts that had been transmitted continuously for a millennium.
A class of religious scholarship comparable to what the Library of Alexandria represented for the Greek world or what the Talmudic academies represented for the Jewish world — a sustained, institutionalized intellectual culture working on the foundational texts of a civilization.
Three quarters of this was destroyed in the Arab conquest and its aftermath.
The destruction was not a single event. It was the cumulative effect of conquest, occupation, conversion pressure, legal disability under the dhimmi system, the long economic strangulation of Zoroastrian communities through jizya and inheritance laws that pressured conversion at every generational transition, and the eventual reduction of the Zoroastrian population from the religious majority of a vast empire to a tiny remnant in scattered enclaves.
What was destroyed was not only books. It was the institutional capacity to produce more books. It was the chain of transmission. It was the readership. It was the knowledge of what had once been known. By the time the Denkard was compiled in the 9th century, its compilers were already reconstructing from partial sources what had been complete a few centuries before. The Denkard‘s book 8 — the summary of the lost nasks — is itself a salvage operation, an attempt to record at least the table of contents of a library that had ceased to exist.
What was lost specifically
It is worth dwelling on what specific kinds of knowledge were destroyed, because the abstract phrase “the Avesta was lost” can obscure what that meant in particular:
The canonical biography of Zarathustra (the Spand Nask). The prophet’s life as preserved in his own religion’s scripture is gone. What we now claim to know about Zarathustra comes from fragments in the Denkard, much later New Persian poetry, and reconstruction from the Gathas themselves.
The canonical cosmology (the Damdat Nask). The systematic account of how the world was made, why it exists, what its structure is, and how the spiritual and material realms relate. The Bundahishn preserves Sasanian-period cosmological material that probably overlaps, but the full canonical treatment is gone.
The canonical astronomy and calendar science (the Nadar Nask and Pājag Nask). Persian astronomical knowledge was foundational to the Greek and Islamic astronomical traditions. The Avesta’s own astronomical texts are gone.
The canonical legal corpus (Nikādum, Ganabā-sar-nigad, Hūspāram, Sakādum, and others). The civil, criminal, and religious law of the empire that governed western Asia for over a thousand years. Mostly lost.
The canonical priestly literature (Ratuštāitīh and others). The institutional theology of the Zoroastrian priesthood — gone precisely when it would have been most needed for the survival of the religion under conditions of persecution.
The canonical commentary on the Gathas (the Sūdgar, Warštmānsr, and Bag Nasks). The interpretive tradition on the oldest hymns of the religion. Lost, along with the lines of priestly commentary that would have transmitted oral readings.
The full hymnic literature to the Yazatas (Bagān Yašt Nask). What survives in the Khorda Avesta is a partial collection. The full canonical hymnic corpus is gone.
The result is that contemporary Zoroastrians — and contemporary scholars studying Zoroastrianism — are working with the surviving quarter of a corpus, mostly the most ritualistic and liturgical texts, while the doctrinal, narrative, legal, scientific, and historical nasks are reduced to summaries and fragments. The religion’s own self-understanding has been forcibly narrowed to what could be carried out by a fleeing community in the holds of ships.
What this means for what we know
It is worth saying clearly: most of what can be said about Zoroastrianism today is being said on the basis of about a quarter of the canonical literature. The questions that scholars cannot answer about pre-Sasanian Zoroastrianism — the dating of Zarathustra, the original theology of the Gathas, the relationship between Achaemenid royal religion and prophetic Zoroastrianism, the Parthian-period synthesis that produced what became the Sasanian canonical form — many of these questions are unanswerable not because the answers are inherently unknowable but because the texts that contained them were destroyed.
When ancient sources outside the tradition — Greek classical authors, the Hebrew Bible’s references to Cyrus and Persian patronage, Roman writers, early Islamic sources — describe Zoroastrian theology and practice, they are sometimes describing things that the Zoroastrian texts themselves can no longer corroborate or expand on, because the relevant texts no longer exist. The external sources sometimes preserve information about Zoroastrianism that the Zoroastrian sources themselves have lost.
This is the deepest meaning of the destruction. It is not only that texts were burned. It is that the religion’s capacity to remember itself accurately was permanently impaired. What we know of Zoroastrianism today is less than the religion knew of itself at the height of the Sasanian period. The conquerors did not only destroy a civilization. They destroyed that civilization’s ability to be fully known, even by the descendants who carry its remnant.
What survived, and why it matters
The surviving Avestan corpus, despite representing only about a quarter of the Sasanian total, includes texts of extraordinary value:
The Gathas — seventeen hymns attributed to Zarathustra himself, in an archaic Avestan dialect dated by linguistic analysis to the second millennium BCE. These are among the oldest religious poems in any Indo-European language and contain the foundational theology of the religion.
The Yasna — the central liturgical text, recited during the daily ritual.
The Visperad — a liturgical supplement.
The Vendidad — the law of ritual purity.
Twenty-one Yashts — hymns to specific Yazatas, including extraordinary mythological and cosmological material (especially Yasht 19, the Zamyad Yasht, on the Saoshyants and the eschaton).
The Khorda Avesta — the daily prayers.
Pahlavi commentary literature — the Bundahishn, the Denkard, the Selections of Zadspram, the Wizidagiha-i Zadspram, the Shkand-gumanig Wizar, the Arda Viraf Nameh, the Zand-i Wahman Yasn, and others. These are 9th–10th century compositions, written in part as salvage operations during the long aftermath of the conquest, preserving in Middle Persian what the priestly tradition could still reconstruct of the lost Avestan material.
It is from this surviving corpus, supplemented by external sources, that all modern knowledge of Zoroastrianism descends. The corpus is enough to demonstrate the sophistication of the tradition, to articulate its foundational theology, and to support a continuing religious life for the Parsi and Iranian Zoroastrian communities. It is not enough to fully recover what was destroyed.
What this asks of contemporary readers
The standard popular treatment of Zoroastrianism in Western religious education is to mention it briefly, note that it had some “possible influence” on Judaism and Christianity, and move on. This treatment is possible only because the destruction of the canonical literature has already done its work — there is so little surviving Zoroastrian text in popular circulation that the religion can be treated as a footnote rather than what it was, which was the foundational religious tradition of one of the longest-lived empires in human history and the source of much of the theological architecture of the religious world that came after it.
A more honest treatment requires acknowledging the scale of what was destroyed. Twenty-one nasks, of which sixteen are essentially gone. A canonical biography of the prophet, gone. A canonical cosmology, gone. A canonical legal corpus, gone. A canonical astronomy, gone. The institutional infrastructure of priests, libraries, and fire temples that had transmitted this material for a thousand years, dismantled. The community reduced from the majority of an empire to a remnant carrying what manuscripts it could on the road to Gujarat.
The destruction was carried out by Alexander partially, and completed by the Caliphate decisively. The Greek tradition preserved the memory of Alexander’s regret. The Islamic tradition has preserved no comparable record of regret for the destruction of Zoroastrian Persia, in part because the conquest was understood theologically as the spread of true religion rather than as cultural devastation. Acknowledging the loss requires looking at the conquest with eyes other than those of the conquerors — using the Zoroastrian tradition’s own preserved records (the Denkard‘s table of contents, the Zand-i Wahman Yasn‘s lament, the Qissa-i Sanjan‘s flight narrative) as the testimony of what was destroyed.
The library that took a thousand years to build was reduced, in a few centuries of conquest and pressure, to a quarter of itself. What we have is precious. What we have lost is more than what we have. And the shape of the loss — the missing cosmology, the missing biography, the missing law, the missing science — is itself a measure of the scale of the catastrophe that Zoroastrianism survived but did not escape.
The fire was carried to Gujarat. It is still burning. But the library is gone.
Sources
Primary Zoroastrian sources
- Denkard, especially book 8 (summary of the twenty-one nasks) and book 9 (commentary on the Sūdgar, Warštmānsr, and Bag Nasks); standard text in B.N. Dhabhar’s edition; partial English translations in Sacred Books of the East vol. 37 (West)
- Bundahishn (Greater Bundahishn), trans. B.T. Anklesaria (1956)
- Zand-i Wahman Yasn, ed. and trans. Carlo G. Cereti (1995)
- Arda Viraf Nameh, trans. M. Haug and E.W. West (1872)
- Qissa-i Sanjan, trans. Hodivala (1920)
- Shkand-gumanig Wizar (Mardan-Farrukh’s polemical defense), trans. West in Sacred Books of the East vol. 24
Surviving Avesta
- The Avesta in standard editions (Geldner’s Avesta, 1886–1896; Westergaard’s earlier edition); standard English-accessible translations in the Sacred Books of the East series (Darmesteter, Mills) — though these are now superseded by more recent specialist work
Greek and Roman sources on Alexander’s destruction of Persepolis
- Plutarch, Life of Alexander 38
- Arrian, Anabasis of Alexander 3.18
- Diodorus Siculus, Historical Library 17.72
- Quintus Curtius Rufus, History of Alexander 5.7
- Pliny the Elder, Natural History 13.21 (on Persian writings)
Modern scholarship on the Avesta and its history
- Mary Boyce, Textual Sources for the Study of Zoroastrianism (1984) — the standard sourcebook
- Mary Boyce, A History of Zoroastrianism, vols. 1–3 (1975–1991)
- Jean Kellens, Essays on Zarathustra and Zoroastrianism (2000)
- Philip G. Kreyenbroek, “The Zoroastrian Tradition: An Oralist’s Approach,” in K.R. Cama Oriental Institute International Congress Proceedings (1991)
- Almut Hintze, “Avestan Literature,” in A History of Persian Literature, vol. 17 (2009)
- Albert de Jong, Traditions of the Magi: Zoroastrianism in Greek and Latin Literature (1997)
- Michael Stausberg, ed., Zoroastrian Rituals in Context (2004)
On the Arab conquest and the destruction of Sasanian institutions
- Touraj Daryaee, Sasanian Persia: The Rise and Fall of an Empire (2009)
- Touraj Daryaee, “The Fall of the Sasanian Empire to the Arab Muslims,” Journal of Islamic Studies (2003)
- Michael G. Morony, Iraq After the Muslim Conquest (1984)
- Jamsheed K. Choksy, Conflict and Cooperation: Zoroastrian Subalterns and Muslim Elites in Medieval Iranian Society (1997)
- Sarah Stewart, ed., The Everlasting Flame: Zoroastrianism in History and Imagination (2013)
- Hugh Kennedy, The Great Arab Conquests (2007)
On the Parsi preservation
- John R. Hinnells, The Zoroastrian Diaspora (2005)
- Jesse S. Palsetia, The Parsis of India: Preservation of Identity in Bombay City (2001)
On al-Biruni and early Islamic-period Persian sources
- Al-Biruni, al-Athar al-Baqiya (The Chronology of Ancient Nations), trans. C.E. Sachau (1879)