In the late fifth century, a famine swept the Sasanian Empire. The Hephthalites had just destroyed an entire Persian army and its king. Tribute payments drained the treasury. The nobility, tax-exempt, kept their grain stores full. The Zoroastrian clergy, tax-exempt, kept their fire temples endowed. The peasantry, who bore the entire fiscal burden of the empire, starved.
Into this crisis stepped a movement that said: the cosmic order requires sharing. Light and Darkness are mixed in the world, and five demons, Envy, Wrath, Vengeance, Need, and Greed, keep them mixed by concentrating wealth and power in the hands of the few. The cure is redistribution. Open the granaries. Break the monopolies. The cosmic struggle between Light and Darkness is not abstract theology. It is happening in your village, right now, and the nobleman hoarding grain while your children die is on the side of Darkness.
The movement was called Mazdakism. It briefly won a king’s support. Then it was crushed so thoroughly that we know it almost entirely through the writings of those who hated it.
The Man, or the Movement
The first confusion about Mazdakism is in the name. Mazdak son of Bamdad was not the founder. He was the most politically consequential leader of a movement that already existed.
The tenth-century bibliographer Ibn al-Nadim, in his Kitab al-Fihrist (987-988 CE), makes a crucial distinction. He identifies “Bundos” as “Old Mazdak” and Mazdak son of Bamdad as “New Mazdak.” The Old Mazdak is likely identical with or closely related to Zaradusht-e Khuragan, a Zoroastrian priest (mobed) from the area of Fars who began a reform movement within Zoroastrianism in the fourth or fifth century CE. Some sources say Bundos resided in Rome for a time and encountered Gnostic movements there, which would explain the syncretic character of Mazdakite theology.
The movement, then, predated its famous leader by perhaps a century. Mazdak inherited something. What he did with it, in the specific political context of Kavad I’s reign, is what made the name stick.
About Mazdak himself, we know almost nothing with certainty. Al-Tabari places his origin at Mazariya (modern Kut) on the Tigris. Al-Biruni says Nisa in Khorasan. Al-Dinawari says Istakhr in Fars. The disagreement tells you something: by the time anyone cared to record where Mazdak came from, the man had been dead for centuries and the record had been deliberately destroyed. The high priest Kartir had established the precedent of erasing heretics from memory two centuries earlier. Khosrow I applied the same technique to Mazdak.
There is even a radical skeptical position. In 1982, the scholar Heinz Gaube argued in Studia Iranica that Mazdak may not have existed at all, that Khosrow I invented a convenient scapegoat to deflect blame from his own unpopular tax reforms. The scholarly community rejected this, but the argument illustrates the depth of the problem: when all your sources are enemies, even the existence of the person they are writing against becomes uncertain.
The Enemies Who Preserved It
No Mazdakite scripture survives. No hymn, no prayer text, no first-person account. Everything we know about what Mazdakites believed comes from people who wanted the movement destroyed.
The chain of transmission runs like this. The Khwaday-namag (Book of Lords), a Sasanian-era compendium of Persian history, was compiled during or shortly after Khosrow I’s reign. It reflected court ideology and was inherently hostile to Mazdakism. Ibn al-Muqaffa (d. c. 757) translated it into Arabic. That Arabic translation is also lost. But it is the common ancestor of nearly everything written about Mazdakism afterward: al-Tabari’s ninth-century history, Ferdowsi’s eleventh-century Shahnameh, and the accounts of al-Thaalibi, Nizam al-Mulk, and others.
The fullest theological account comes from al-Shahrastani (1086-1153), writing in his comparative study of religions Kitab al-Milal wa al-Nihal. His source was Abu Isa al-Warrag (d. 861), a Manichaean or Zoroastrian convert to Islam who appears to have had access to something closer to genuine Mazdakite teaching. This makes al-Shahrastani’s cosmological description, filtered through multiple transmissions and six centuries, the nearest thing we have to what Mazdakites actually believed about the universe.
The contemporary witnesses are the most revealing for what they do not say. Procopius (mid-sixth century), Malalas, and the Syriac Pseudo-Joshua the Stylite all describe Kavad’s communal property reforms. None of them mentions anyone named Mazdak. They associate the ideas with the king, or with the earlier reformer Zaradusht-e Khuragan. The figure of “Mazdak” as the central protagonist may be, at least in part, a later construction of the hostile Khwaday-namag tradition, which needed a single named heretic on whom to blame a complex and varied social upheaval.
The Architecture of Light
What did Mazdakites actually believe? Al-Shahrastani’s account, the most detailed we have, describes a cosmological system built on numbers, correspondences, and a structural parallel between the divine court and the Sasanian royal court.
At the foundation: two primordial principles. Light is endowed with knowledge and acts by will and design. Darkness is ignorant, blind, and acts randomly. This sounds like Manichaeism. But there is a crucial difference that changes everything about the movement’s character. In Mani’s theology, Darkness actively attacks Light, trapping divine sparks in matter, and salvation requires ascetic withdrawal from the world. In Mazdak’s theology, Darkness is not aggressive. It is stupid. The mixture of Light and Darkness in the world is more accident than assault. And the response is not withdrawal but engagement: humans can advance Light through right conduct in the world itself.
This makes Mazdakism optimistic in a way that Manichaeism is not. The world is not a prison to escape. It is a field to work.
Above the two principles, the Supreme Being sits enthroned. He governs through four Powers: Discernment (al-tamyiz), Understanding (al-fahm), Memory (al-hefz), and Joy (al-sorur). The scholar Mansour Shaki (1985) noticed something remarkable: these four powers correspond structurally to the four chief dignitaries of the Sasanian court. The mowbedan mowbed (chief priest), the herbed (chief legal scholar), the esbahbed (military commander), and the rameshgar (court musician/master of ceremony). The divine court mirrors the earthly court. Or the earthly court mirrors the divine court. Either way, if the cosmic order runs on discernment, understanding, memory, and joy, then an earthly order that runs on hoarding, ignorance, violence, and misery is cosmically wrong.
The four Powers govern through seven viziers, corresponding to the seven planets. The seven viziers act within twelve spiritual forces, corresponding to the twelve signs of the zodiac. When all of these, 4+7+12, come together in a single human being, that person becomes rabbani, godlike, and is no longer bound by religious observances. This is the antinomian endpoint of Mazdakite gnosis: complete spiritual knowledge liberates you from ritual and law.
And then there are the five demons: Envy, Wrath, Vengeance, Need, and Greed. These are not abstract spiritual forces. They are social relations. They are what happens when wealth concentrates, when the powerful hoard, when the poor are left to starve. The five demons are the structural link between Mazdakite cosmology and Mazdakite social action. To fight the demons is to redistribute. To redistribute is to advance Light. Theology and politics are not separate departments.
The Empire That Made It Possible
Mazdakism did not appear in a vacuum. The Sasanian Empire of the late fifth century was a pressure cooker.
Sasanian society was divided into four rigid, hereditary estates. The asronan (priests) sat at the top, controlling fire temples, courts, and customs enforcement. The arteshtaran (warriors and nobility) held vast landed estates and maintained private armies. The dabiran (scribes and administrators) ran the bureaucracy. And at the bottom, bearing the entire fiscal weight of the empire, the vastryoshan (and the related hutuxshan, artisans), the commoners: peasant farmers, cattle breeders, traders. The system allowed no mobility. You were born into your estate and died in it.
The nobility and clergy were effectively tax-exempt. When the state needed revenue, the collection fell on the people who could least afford it.
Then the crisis hit. In 484 CE, King Peroz I led three successive campaigns against the Hephthalites, the dominant Central Asian steppe power. In the third, he and his entire army were annihilated. The Hephthalites seized the major eastern cities and extracted ruinous tribute from the Sasanian state. A multi-year famine compounded the disaster. The peasantry, already overtaxed, now faced starvation while noble granaries remained full and temple endowments remained intact.
This is the world Kavad I inherited when he took the throne in 488 CE at roughly fifteen years old. A ruined treasury. An arrogant nobility that had deposed his predecessor. A priesthood that answered to its own interests. A starving population. And a movement that said: the cosmic order demands that the rich share with the poor, or they serve Darkness.
The political logic was irresistible.
Kavad’s Gamble
Kavad I is one of the most interesting figures in Sasanian history, and the question of what he actually believed about Mazdakism remains unresolved.
During his first reign (488-496), Kavad openly supported the movement. He ordered granaries opened. He allowed or encouraged redistribution. He executed Sukhra, a powerful nobleman, which inflamed the aristocracy. The nobility and the Zoroastrian clergy, recognizing the threat, formed a coalition. In 496, they deposed Kavad and imprisoned him in the Castle of Oblivion (Anushbard) in Khuzestan. His more pliable brother Jamasp was placed on the throne.
What happened next is the stuff of adventure narratives. With the help of his sister and an officer named Siyawush, Kavad escaped. He fled east to the Hephthalite kingdom in Bactria. The Hephthalite king agreed to provide him with an army of roughly thirty thousand troops in exchange for territorial concessions. In 499, Kavad returned and reclaimed the throne. The nobility, unwilling to risk civil war, submitted.
His second reign (499-531) was more cautious. Kavad made peace with the religious establishment. He prosecuted wars against Byzantium. But in the 520s, he initiated a cadastral survey, a comprehensive land register designed to reform the tax system from harvest-proportional payments (flexible, payable in kind) to fixed cash assessments. For subsistence peasants embedded in a non-cash economy, this was devastating.
Patricia Crone (1991) argued that this tax reform was the real trigger for Mazdakite agitation in the later phase: Kavad used the movement to push through changes that benefited the state at the expense of the nobility, then abandoned the movement when it was no longer useful. The evidence supports a cynical reading. Contemporary Byzantine sources credit Kavad himself, not any charismatic preacher, with the communal property ideas. He withdrew support from Mazdak in the 520s. And he allowed, or arranged, for his son Khosrow to lead the suppression.
If Kavad was a believer, he was an inconsistent one. If he was a politician, he was a ruthless one who used a genuine popular movement as a weapon against his domestic enemies and then let it be destroyed when the weapon was no longer needed.
The Garden
The suppression of Mazdakism is one of the most graphically described events in Persian literary history.
Kavad’s son Khosrow I Anushirvan (r. 531-579) organized what multiple sources describe as a formal trial or debate. Mazdak was summoned to defend his doctrines before the Zoroastrian mobads (priests) and, in some versions, Christian and Jewish representatives. The multi-confessional condemnation was rhetorically powerful: even the empire’s minority religions agreed that Mazdak was dangerous.
What followed the verdict has become one of Persian literature’s most disturbing set pieces. In Ferdowsi’s Shahnameh (c. 1010 CE), Khosrow has three thousand Mazdakites buried head-first in his garden, their feet pointing upward like trees. He then invites Mazdak to see “a garden of a kind no one has seen before.” When Mazdak arrives and sees the bodies, he cries out and faints. Khosrow has him strung up on a gallows, head downward, mirroring his buried followers, and killed with a shower of arrows.
Nizam al-Mulk’s Siyasat-nama (1091) offers a different version: Khosrow invites the Mazdakites to a banquet under the pretense of honoring them with robes, then has them slaughtered.
The literary details are almost certainly invented. No contemporary sixth-century source confirms the garden execution method. The scene comes from the Khwaday-namag tradition, compiled by the court that ordered the killing. The inversion imagery, men planted as trees, the leader hung upside-down, is symbolically dense: it mirrors and reverses what the Sasanian elite believed Mazdak had done to the social order. He turned the world upside down. So they turned his followers upside down.
What is not literary is the massacre itself. Mass suppression is confirmed across multiple independent traditions. The casualty figures reported (80,000 to 150,000) are medieval hyperbole without corroboration. But that a large-scale killing occurred, that organized Mazdakite communities were destroyed, and that Mazdak himself was executed, is as well-attested as anything from this period.
The question of timing remains genuinely unclear. Some sources place the persecution in the 520s, during Kavad’s lifetime, with the king’s tacit approval. Others place Mazdak’s execution at the start of Khosrow’s reign in 531. The most likely synthesis: there were two phases. A persecution under Kavad in the 520s killed many followers. The final elimination of the leadership occurred at or just after the transition of power to Khosrow. The sources blur the distinction deliberately, allowing both father and son to emerge favorably.
The Accusation
The most famous claim about Mazdakism, the one that every hostile source fixates on, is that Mazdakites advocated sharing women and property in common.
The fullest statement comes from al-Thaalibi: “God placed the means of subsistence on earth so that people divide them among themselves equally, in a manner that no one of them could have more than his share; but people wronged one another and sought domination over one another. It is absolutely necessary that one take from the rich for giving to the poor. Whoever possesses an excess of property, women or goods, he has no more right to it than another.”
This text treats “women” as a category of property alongside goods and wealth. The hostile framing is obvious. But what was the reality behind the polemic?
The scholarly consensus has shifted substantially. Mansour Shaki (1978) and early Patricia Crone took the claims largely at face value. More recent work is more skeptical, for several reasons.
First, “sharing women” is a standard heresiological accusation. The same charge was leveled at early Christians by Roman writers, at various Gnostic sects by the Church Fathers, and at Shia Muslims by Sunni heresiographers. It is the default polemical weapon against any group that challenges conventional social boundaries. Its appearance tells you the accusers felt threatened. It does not tell you what the accused actually did.
Second, no source describes rules, regulations, or institutional mechanisms for this alleged sharing. If Mazdakites had actually organized communal marriage or sexual sharing, some trace of the system’s regulation would appear somewhere. None does.
Third, later Mazdakite-descended communities, the Khorramites, show no evidence of promiscuity. What they do show is opposition to elite polygamy and restrictions on women’s autonomy.
The revisionist interpretation, supported by Crone’s later work: Mazdak likely attacked the harem system of the wealthy. In Sasanian society, rich men maintained enormous harems, effectively monopolizing available women and using bride prices as a form of wealth storage. Poor men could not afford to marry. Mazdak’s “reform” probably meant prohibiting elite polygamy, reducing dowry costs, and enabling marriage across class lines. Hostile sources collapsed this into the most scandalous possible interpretation.
The Khorramite practice of “fraternal polyandry,” brothers sharing a wife to prevent property division, is a distinct, regionally specific custom from northwestern Iran. It is not the same thing as “sharing women” and should not be confused with it.
The Joyous Religion
Mazdakism did not die in the garden. It went underground.
The Khorramites (Khurram-Dinan, “Those of the Joyous Religion”) were the direct descendants of post-massacre Mazdakism. They survived in mountain communities of Azerbaijan, the Jibal region of western Iran, and the Caspian coastal areas, populations that had practiced local forms of Zoroastrian tradition predating the Sasanian orthodoxy. Mazdakism gave these traditions a political articulation. After the massacre, the movement persisted in places where Sasanian, and later Arab, administrative control was weakest.
When Abu Muslim, the Persian revolutionary commander, was assassinated by the Abbasid Caliph al-Mansur in 755 CE, the reaction was immediate. Sunpadh, a Zoroastrian nobleman and former ally of Abu Muslim, launched a revolt in Khorasan. He claimed Abu Muslim had not died but had been transformed and now dwelt “in a citadel with the Mahdi and Mazdak.” Two centuries after the garden, Mazdak had become a messianic figure. His return would inaugurate an age of justice.
Al-Muqanna, the “Veiled Prophet” (c. 768-780), operated from Transoxiana, claiming that the divine spirit had passed successively through Muhammad, Ali, Abu Muslim, and now himself. His movement blended Zoroastrian neo-Mazdakite communalism with Islamic apocalypticism. He died around 780, allegedly immolating himself to prevent capture.
The longest and most militarily consequential of the Khorramite revolts was led by Babak Khorramdin (816-837). From the fortress of Badhh in the mountains of Azerbaijan, Babak commanded a guerrilla force that held out against the Abbasid Caliphate for twenty years. His program included redistribution of noble estates and an end to Arab colonial administration. The Abbasid caliph al-Mu’tasim eventually assigned his best general, al-Afshin, to crush the revolt. The campaign was methodical: al-Afshin advanced only four miles per day, establishing garrisons and gathering intelligence. In August 837, Badhh fell. Babak was captured, paraded through Baghdad on a decorated elephant, then taken to the Abbasid capital at Samarra and executed publicly: his limbs were cut off, then he was disemboweled.
The chain did not end there. Some scholars trace the transmission further: Mazdakism to Khorramism to heterodox Shi’ism to the Qizilbash (the Turkic-speaking Alid warrior tribes) to the Safavid Empire that defined modern Iran. Whether this represents continuous institutional transmission or a series of independent revivals drawing on the same cultural reservoir remains debated. But Mazdak’s name persisted. As late as the seventeenth century, the author of the Dabestan-e Mazaheb (1645-1658) claimed to have personally met secret Mazdakites who still preserved a text called the Desnad in Middle Persian, containing Mazdak’s teachings.
What We Actually Know
Mazdakism studies face the same fundamental problem as Zurvanite studies: the people who believed it did not leave their own account, and the people who preserved the record wanted the movement destroyed.
The modern scholarly debate about what Mazdakism “really was” has produced four schools.
The economic-determinist school (Nina Pigulevskaya, Otakar Klima): Mazdakism was a peasant revolt dressed in religious clothing. The real engine was resistance to Kavad’s land survey and the tax reforms it enabled. The cosmic vocabulary gave the revolt shape, but the motor was economic.
The political-instrument school (Patricia Crone, 1991): Kavad engineered the movement, or at least directed it, as a weapon against the nobility. The revolt was about state-building: breaking noble power to create a stronger royal-bureaucratic state. The “heresy” framing was imposed afterward.
The religious-movement school (Ehsan Yarshater, Mansour Shaki): Mazdakism was a genuine religious movement with its own internal logic, cosmology, and ethical system, which happened to have political consequences. The theology is not reducible to an economic program or a political instrument.
The post-orthodoxy school (Khodadad Rezakhani, 2015): The entire framework of “heresy” is anachronistic. Sasanian Zoroastrianism was not a monolithic orthodoxy. Multiple competing religious traditions coexisted. Mazdakism’s designation as heresy is a retroactive judgment made by post-Sasanian scholars constructing an orthodoxy under pressure from Islam. In the Sasanian period itself, the boundaries were much more fluid.
Each school captures something real. Mazdakism was a genuine religious tradition. It was politically instrumentalized. It had economic drivers. And calling it “heresy” distorts a more complicated reality. The problem, as always, is that the evidence is filtered through centuries of hostile transmission. We are reading the Mazdakite position through the words of people who killed them for holding it.
One observation from the Bogomil parallel is relevant. Dualist movements across very different cultures, from Sasanian Iran to medieval Bulgaria to Cathar Languedoc, share a recurring structural feature: they emerge when rigid social hierarchies claim divine sanction, and they challenge those hierarchies by reinterpreting the cosmic order. The content varies. The pattern persists. Whether this means these movements influenced each other across centuries and continents, or whether it means that radical dualism is simply what happens when people with religious imaginations confront entrenched inequality, is a question worth sitting with.
The Oldest Argument
Mazdakism is fifteen centuries old. The argument it represents is older still, and it has not been settled.
Can religion demand redistribution? Should cosmic order translate into social equality? Is it heresy to say that the rich serve Darkness, or is it heresy to deny it?
The Zoroastrian priesthood answered one way. Kavad answered another, then changed his mind. Khosrow answered with a garden full of inverted corpses. The Khorramites answered with twenty years of guerrilla war. Al-Shahrastani, six centuries later, called the myth “infantile” in literal form but acknowledged that behind it stood “a mystery of what is figured in the mind.”
Mazdak’s theology, as far as we can reconstruct it, said something specific. It said that the five demons (Envy, Wrath, Vengeance, Need, Greed) are not spiritual abstractions. They are social structures. They are what happens when wealth concentrates and power calcifies. And fighting them is not a private spiritual exercise. It is a collective, material, political act. You fight Greed by opening the granary. You fight Envy by ending the monopoly. You fight Need by sharing what you have.
The Sasanian state found this intolerable. Every state that has encountered a similar argument since has found it intolerable. The argument keeps being made anyway.
Whether that persistence reflects a deep truth about cosmic order, or just about human nature, is, like so much in this territory, a question we present without resolving. The evidence is here. The pattern is clear. What it means is up to you.
By the Author
Selections from the Memoirs of Satan by Wilhelm Hauff, trans. Rade Kolbas
Sources
Bibliography. The same list is held in the article’s frontmatter for the citation tools that read it programmatically.
- Ibn al-Nadim, Kitab al-Fihrist (987-988 CE)
- Al-Tabari, Tarikh al-Rusul wa al-Muluk (9th century)
- Al-Biruni, Chronology of Ancient Nations (c. 1000)
- Al-Dinawari, Kitab al-Akhbar al-Tiwal (9th century)
- Al-Shahrastani, Kitab al-Milal wa al-Nihal (12th century)
- Abu Isa al-Warraq (d. 861), source preserved via al-Shahrastani
- Al-Thaalibi, Ghurar Akhbar Muluk al-Furs (11th century)
- Ferdowsi, Shahnameh (c. 1010 CE)
- Nizam al-Mulk, Siyasat-nama (1091)
- Procopius, History of the Wars (mid-6th century)
- Malalas, Chronographia (6th century)
- Pseudo-Joshua the Stylite, Syriac Chronicle (early 6th century)
- Ibn al-Muqaffa (d. c. 757), Arabic translation of the Sasanian Khwaday-namag
- Dabestan-e Mazaheb (1645-1658)
- Heinz Gaube, ‘Mazdak: Historical Reality or Invention?’ Studia Iranica 11 (1982)
- Otakar Klima, Beiträge zur Geschichte des Mazdakismus (Prague, 1977)
- Nina Pigulevskaya, Les villes de l’état iranien aux époques parthe et sassanide (Paris, 1963)
- Mansour Shaki, ‘The Social Doctrine of Mazdak in the Light of Middle Persian Evidence,’ Archiv Orientální 46 (1978); and ‘The Cosmogonical and Cosmological Teachings of Mazdak,’ Acta Iranica (1985)
- Ehsan Yarshater, ‘Mazdakism,’ in The Cambridge History of Iran, vol. 3(2) (Cambridge University Press, 1983)
- Patricia Crone, ‘Kavad’s Heresy and Mazdak’s Revolt,’ Iran 29 (1991); and The Nativist Prophets of Early Islamic Iran (Cambridge University Press, 2012)
- Khodadad Rezakhani, ‘Mazdakism, Manichaeism and Zoroastrianism: In Search of Orthodoxy and Heresy in Late Antique Iran,’ Iranian Studies 48 (2015)
- Mary Boyce, Zoroastrians: Their Religious Beliefs and Practices (Routledge, 1979)
