Entangled Religions 11.2 (2020) er.ceres.rub.de

The Co-formation of the Manichaean and Zoroastrian Religions in Third-Century Iran

Jason BeDuhn Northern Arizona University, United States

The assumption that an already established Zoroastrian religion served as the source for terms, concepts, and themes which Mani and Manichaeans appropriated and altered is due for reassessment. Building on the work of P. O. Skjaervø, this study argues that (1) Zoroastrianism and Manichaeism arose together, side by side, in the third century (2) against the background of older Iranian religious cultural traditions, (3) each fitting those antecedent cultural artifacts into different systems of interpretation and application.

Zoroastrianism, Manichaeism, Avesta, pantheon, ritual, legend, ethos, eschatology



A little over twenty years ago, Prods Oktor Skjærvø published a set of four articles in which he surveyed the “Iranian Elements in Manicheism” as well as the possible impact Mani and Manichaeism may have had on institutional Zoroastrianism in the early Sasanid period (Skjærvø 1995a, 1995b, 1996, 1997). In these articles, Skjærvø performed the service of pushing the study of Iranian religion out of a comfort zone that viewed Manichaeism as dependent on an already well-established “Zoroastrianism.” In the latter view, an organized and orthodox Zoroastrianism served as a source for terms, concepts, and themes that Mani and Manichaeans appropriated and altered to fit into Manichaean “syncretism” (see e.g. Scott 1989). In contrast, Skjærvø envisions the origin of Manichaeism occurring at a “time when the Mazdayasnian religion was being redefined and consolidated” (Skjærvø 1995a, 267). Confronting the late date of nearly all primary sources on this religion, he cautions that there is “no reason a priori to think […] that all of the cosmology of the Pahlavi books was part of the official doctrine in the third century or, indeed, that they contain everything that belonged to the official third-century doctrine” (Skjærvø 1995a, 268), and recognizes that “it is quite possible that elements of the Manichaean myth permanently entered the Zoroastrian myth at this time” (1995a, 267). While not completely unprecedented, Skjærvø’s recognition of the possibility of an exchange of ideas in both directions between these two communities (1995a, 281), in a period that was formative for both, marks a major advance away from an understanding of Manichaean origins that consists simply of normative heresiological tropes put in the fancy dress of modern academics. The present contribution is intended to build on Skjærvø’s observations in light of new sources, but especially within a fresh conception of the religious landscape of the era, by which the sources that Skjærvø himself marshaled take on a different significance.


The traditional, medieval Zoroastrian account of the community’s history, as read through modern (nineteenth- to twentieth-century) Religious Studies discourse, continues to have a powerful hold over how historical materials are interpreted. Hence, scholars across the field, including Skjærvø, continue to speak in terms of a Zoroastrian “religion” in the centuries BCE, at a time when nowhere else on earth was there such a thing as a “religion”; and continue to speak in terms of “doctrine,” even “official doctrine,” at a time when priestly institutions authorized myths, ritual scripts, purity codes, and other elements of regional religious culture, held together at most by a loosely-defined theory of efficacy in relation to divine beings. The traditional Zoroastrian narrative of a religion founded, lost, and reconstructed (in the Sasanid period) only to be shattered again, and again reconstructed (in the medieval period), is deeply enmeshed in anachronism, compounded by modern scholars of religion anxious to find parallels to other, later religions. Indeed, the Manichaean narrative of religions founded, lost, and reconstructed is much the same, and told in no more fanciful detail—and yet most scholars would regard it as a tendentious, ideological construct, while according the Zoroastrian narrative it has at least a basis in history, however embellished by legendary elements. Yet, when set within the larger context of historical developments, the evidence suggests that Zoroastrianism was in its formative (not reformative) period at the same time as Manichaeism was; that both religions emerged only in the third century against the background of Iranian religious cultural traditions; that both religions laid claim to those older traditions, and appropriated them selectively, tendentiously, within the bounds of their distinct hermeneutics; that this process of cultural interpretation and consolidation in third-century Iran resulted in precisely the emergence of “religions” as distinct institutional entities for the first time in this part of the world; but that both religions wanted to read their present into the past, and set about constructing a narrative of an earlier time much like theirs, of an earlier founding of Zarathustra’s “religion.”

The Religious Landscape of Late Antiquity


Recent discussions in the field of Religious Studies have made us more aware of the anachronism involved when we speak of communities in the ancient world as “religions” (e.g., Nongbri 2013; Barton and Boyarin 2016). We should not impose this category on cultures that had no such concept. As much as we may be accustomed to speaking of “Zoroastrianism” as something dating back to the Achaemenid period, this needs to be carefully qualified. It has been commonplace to speak even of an ancient “Zoroastrian Church,” implying an institutional organization with an orthodox “dogma” or “official/standard doctrine” (e.g., Skjærvø 1995a, 269). This is surely wrong for the dawn of the Sasanid period and the centuries preceding it. The elements we see in Achaemenid remains that we might identify as “Zoroastrian” are cultural artifacts later taken up into the emerging Zoroastrian religion as it took shape in the Sasanid period. Avestan literature itself is taken up into later Zoroastrian usage in much the same way that the legal codes and temple psalms of the kingdom of Judah, for example, were taken up and employed in later Judaism and Christianity.


“Religions” are entities that incorporate and repurpose older cultural material in a new, more systematic setting, and it is historically inaccurate to take the previous existence of that cultural material as an indication of the presence of the religion that will later make use of it. While obviously its later use suggests some continuity of ideas and themes, we must pay close attention to the great discontinuities and reinterpretations that are often involved in religious re-use of older cultural material. The emergence of something we can fairly call “religions” occured in West Asia only two millennia ago, for very specific historical reasons. The category “religion” emerges only when it has to, when it is summoned forth by conditions on the ground that no longer abide by previous assumptions. Those previous assumptions are that every ethnic group has its culturally distinctive ways of relating to the forces that govern the universe, the gods; and that one is born into this identity and dies in this identity, and one cannot opt out of these traditional ways without literally leaving home and going to some foreign land and marrying into a different ethnic group and acculturating to its ways. The category “religion,” then, only emerges where there are options, alternative ways within a single ethnicity or state for relating to the gods. To be more precise, it is not simply a matter of variety, since variety of belief and practice is found everywhere and always. Rather, the key development is what might be called religious pluralism, where distinct and mutually exclusive identities exist that are not interchangeable or coterminous with ethnic identity. Native religious practices, deeply embedded within particular socio-cultural identities, began at particular points of history, under specific conditions, to be distilled into or displaced by religions: organized systems of belief and practice disembedded from particular societies and cultures. It is only when such organized communities began to exist or to be thought of as distinct entities, side-by-side and in competition with each other, that we can speak of religions in the proper sense for the first time (BeDuhn 2015b).


As far as evidence suggests, this awareness of religious pluralism (which no doubt had been noted informally in the cities and marketplaces of Iran for some time) had its first formal statement in the words of Mani addressed in the mid-third century to Shapur I in the Šābuhragān:


Wisdom and knowledge have been brought repeatedly by the messengers of God in one period after another. Thus they appeared at one time through the messenger called Buddha in the land of India, and in another through Zaradusht in the land of Persia, and in another through Jesus in the land of the West. Then in the present time there came this revelation […] through me, Mani, the messenger of the God of truth in the land of Babylonia. (Bīrūnī, Āthār ul-bāqiya, ed. Sachau, 207.14-18; cf. Reeves 2011, 102–3 and n114-115)


Notice the transitional nature of this declaration. In some ways, Mani is still working with the old assumptions: each land has its own distinctive tradition, its own cultural hero who brought the wisdom of knowing one’s place in the universe and how to relate to its governing forces. Mani even identifies himself here with a particular region: Babylonia, as if he were the prophet to the Babylonians—much as Muhammad would at first declare himself to be the latter-day prophet to the Arabs. But both Mani and Muhammad developed in their thinking, and came to see their missions more universally. Mani makes that first step here by conceptualizing each of these historical figures as offering a permutation of the same truth, rendered into distinct systems of expression and practice. This pluralism on the ground in Mani’s world offered food for thought, on the basis of which Mani comes to think of these traditions as something like what we refer to when we speak of “religions.” Mani was able to imagine his own permutation of the truth behind these traditions as a world religion, one that will transcend geographic, linguistic, and ethnic boundaries (Lieu 2006). He was able to take that step because he had conceptualized “religion” as something disembedded from those other identities.


This incremental advance in the conceptualization of religion can be seen in passages from the Chester Beatty Kephalaia:


Behold, I will [tell] you about each one of the apostles by name, they who came (and) appeared in this world. Zarades was sent to Persia, to Hystaspes the king. He revealed the truly-founded law in all of Persia. Again, Bouddas the blessed, he came to the land of India and Kushan. He also revealed the truly-founded law in all of India and Kushan. After him again, Aurentes came with Kebellos to the east. They also revealed the truly-founded law in the east. Elchasai (?) came to Parthia. He revealed the law of truth in all of Parthia. Jesus the Christ came to the west. He (also?) revealed the truth in all of the west. […] For they were seized from this place; they were taken up; they went, they saw, they came (back), they bore witness; they have told [that the] land of light exists and that we have come from it. Also, hell exists, and we have seen the place where it is […]. Their testimony exists till now in their writings. […] I, myself, whom you are looking at: I went to the land of light. Indeed, I have seen the land of light with my eyes, the way that it exists. Again, I have [seen] hell with my eyes, the way that it exists. I have received […] from God. I came; I have revealed this place (i.e. the land of light) in this world; I preached the word of God. (2Ke 422.28 – 424.12; Gardner, BeDuhn, and Dilley 2018, 164–69)


In another extended passage (not yet published), Mani surveys these other traditions, enumerating key parallel aspects of their formation and institutionalization. Mani crafts a working definition of religion as the product of revelation,1 authorized by a founding authority, organized as a community, guided by textual resources (BeDuhn 2015b).


Towards the end of the third century, another Iranian religious leader, Kerdīr, shows that he has recognized the same pluralistic conditions as Mani, and he addresses them accordingly.


And from province to province, place to place, throughout the empire the rites of Ohrmezd and the gods became more important and the Mazdayasnian religion (dyn) and magians were greatly honoured in the empire and great satisfaction befell the gods and water and fire and beneficent creatures, and great blows and torment befell while Ahreman and the demons and the heresy (qyš) of Ahreman and the demons departed and was routed from the empire. And Jews (yhwdy) and Buddhists (šmny) and Hindus (blmny) and Nazarenes (n’čl’y) and Christians (klystyd’n) and Baptists (mktky) and Manichaeans (zndyky) were smitten in the empire, and idols were destroyed and the abodes of the demons disrupted and made into thrones and seats of the gods. (Translation from MacKenzie 1989, 58)2


Many of the same groups are referred to in Mani’s and Kerdīr’s declarations: Kerdīr’s Mazda-worshiping dēn as Mani’s followers of Zoroaster, Kerdīr’s Šamans as Mani’s followers of Buddha, Kerdīr’s Nazareans and Christians as Mani’s followers of Jesus, and Kerdīr’s Zandiqs as Mani’s own Manichaeans. The other groups Kerdīr mentions—Jews, Brahmans, and the enigmatic Maktaks—overlap with groups that Mani speaks of elsewhere.3 Moreover, Kerdīr’s references to both personal revelation and textual sources of his tradition echoes factors Mani considered essential components of a “religion.” Indeed, Kerdīr’s primary authorizing narrative involves a visionary journey very similar to the one claimed by Mani. Skjærvø and others have suggested Kerdīr was consciously responding to, and competing with, Mani in his inscriptions (see Skjærvø 1997; Russell 1990).


Both Mani and Kerdīr, then, report a rich and vibrant religious pluralism within third-century Sasanid Iran. Kerdīr, for his part, was forced to acknowledge a changed condition where religious traditions had begun to consolidate as at least in part distinct from ethnic identity. Yet he also holds a rather transitional view, in that he still sees only one tradition as natural and right for the Iranian people; the alternatives represent to him just a many-headed form of a single demonic lie, as they already did for Xerxes. Kerdīr echoes much of the latter’s phrasing about the daivas or temples where they are served in a land that should belong only to the supreme god. But Kerdīr now sees these opponents of truth manifested in autonomous, named religious traditions and communities. Mani had recognized the same solidification of religious identities, but by emphasizing their parallelism of structure and purpose, succeeded in achieving a more abstract concept of “religion,” even if his own was still “better.” In any case, it is clear that Kerdīr represents a response to the catalyst of this religious pluralism, in which a distinct Zoroastrian religion is gaining self-consciousness out of prior Iranian religious discourse and practice.

Competing Religious Appropriation of Iranian Cultural Traditions


The conceptual and historical framework established above sets the stage for a reconsideration of the origins of religions in Late Antiquity, some of which we have been in the habit of thinking about as much older. In this light, Zoroastrianism and Manichaeism, among other religions, can be understood to have arisen together, side by side, in the third century. Both of these religions drew upon earlier Iranian traditions and materials, but they each fitted those antecedent artifacts into different systems of interpretation and application. That specific process of cultural appropriation and repurposing by these two emerging religions requires closer scrutiny.


Within the broader understanding of how “religions” first emerge into conceptualization in the third century, we must move away from treating Manichaeism as a heresy that breaks with an existing religion, or even as a religion that is unusually syncretistic compared to others, one that, so to speak, has disassembled an existing Zoroastrianism and scavenged it for parts. Rather, Manichaeism comes into existence on a field of traditional Iranian discourses and practices that can be arranged and assembled in any number of ways to make a religious system; there is no necessary arrangement or normative template in place that Manichaeans must defy to make something different. Nor should the Zoroastrian use of those materials be taken as their normative or default meaning and purpose, one that existed prior to Mani’s creative misprision of them. Instead, Manichaeans and their Zoroastrian peers were undertaking the same work of systemizing at the same time, working with the same raw materials, and creating from them alternate religious systems for the first time.


In what follows, therefore, sets of such raw materials will be briefly examined for their antecedent character as part of Iranian culture and for their respective redeployment in the two new and competing religions in third-century Iran.

Dualistic Universe


The overall dualistic worldview of Iranian religion is well attested in the centuries before Mani and Kerdīr, and is a dominant characteristic of the two interpretations of Iranian religion they represent. Even though Manichaean mythology retains traces of non-dualistic Jewish and Mesopotamian sources, these are overwritten in every case by a dualistic structure. Those who wish to argue that Manichaeism is essentially Judeo-Christian with only a veneer of Iranian elements cannot overcome this contrary fact. Likewise, even though Zoroastrian mythology retains traces of polytheistic complexity, this has been subordinated to a dualistic supremacy that had already asserted itself well before Sasanid times. Among contemporaneous sources for this are Diogenes Laertius, Proöm. 6.8 and Plutarch, de Iside 47 (König 2020). Yet, there is evidence for forms of Iranian religion that at least qualify and subordinate this dualism to a higher order: eternal time personified as Zurwān and as parent of both Ohrmazd and Ahriman (Rezania 2010; Zaehner 1971).4 While probably not an organized alternative “orthodoxy,” these non-dualist traditions were part of the larger Iranian religious culture that were explicitly rejected by both Zoroastrian and Manichaean authorities, who agreed in polemicizing against them (Skjærvø 1995a, 271–72).5

Myth of Primordial Combat


The dualistic view of the universe shared by the Zoroastrian-Manichaean wing of Iranian religious culture includes close parallels between the two religions’ myths of primordial combat between the forces of good and evil. Evil is in both cases the aggressor against a perfect world, and must be repelled, resulting in a “mixture” of both good and evil in our current world that needs to be distinguished and sorted out. This mixture permeates nature itself, and is not just a human moral feature. As Skjærvø points out, allusions to this myth are few and far between in Avestan materials; he cites Videvdad 22.1-2 as “the only Avestan reference to the myth” of Angra Mainyu catching sight of Ahura Mazda’s realm and craving to possess it, which finds its full expression only in medieval Pahlavi texts such as Bundahišn 1.7, but is found in its essentials in third-century Manichaean sources (Skjærvø 1996, 604 and n26). Skjærvø notes additionally a reference to primordial combat in Yašt 13.76-78 where, “when the Evil Spirit passed through (into) the domain of good Order, Vohu Manah and the fire came down between. Those two overcame his aggressions” (Skjærvø 1996, IV:604–605). Similarly, in the Manichaean myth, the king of darkness faces off with the divine Primal Man and his five elements, of which he fashions fire into a weapon (2Ps 10.10-10; 1Ke 126.31-127.11, 129.6-12). The two myths differ, however, in the immediate outcome of this combat, with the forces of good victorious in the Zoroastrian version but (temporarily) defeated in the Manichaean one. Consequently, the Manichaean version of the myth jumps directly to the disintegration of a primordial animate being that provides the building blocks of the natural world: Primal Man himself. The Zoroastrian version of the myth has an interlude following the primordial combat, which resumes when Angra Mainyu recovers his aggression and attacks Gayomard, the disintegration of whose body produces the raw materials for the first humans and much else in the world (Bundahišn 14.1-6). A close reading of Avestan sources raises the possibility that the earlier form of the myth had a single assault of Angra Mainyu on the world created by Ahura Mazda. The fully developed Zoroastrian myth of Pahlavi literature has grafted in an earlier assault on the divine world, prior to the creation of this world, perhaps by appropriating elements of Manichaean myth.


Of course, many cultures of West Asia had a myth of primordial combat, whether set in a monotheistic, polytheistic, or dualistic conception of cosmic powers. Many specific details of the Manichaean myth show continuity with Mesopotamian myths, and others with Jewish para-biblical narratives, such as the Enoch corpus (which itself has points of continuity with Mesopotamian myths; see Reeves 1992, 1993). Ancient Iranian myths belonged to this larger West Asian mythological culture, with many common themes traceable from the Mediterranean to India. Avestan materials contain allusions to all sorts of mythic narratives, including, as one might expect, a number with themes of divine combat and conflict. It would be wildly anachronistic to imagine that they all fit seamlessly into a single standard, “official” mythology of the time. It has long been recognized in Zoroastrian Studies that this material represents traces of a rich and diverse polytheistic mythology that is not always identical to what came to be Zoroastrian “orthodoxy,” and much of which shares mythic plots and themes with neighboring cultures.


Zoroastrians in the Sasanid period codified this antiquated, mostly hymnic material, including many mythic allusions that were no longer fully understood by the living tradition. By selecting portions of this cultural heritage to emphasize, elaborate, and read in light of their Late Antique “orthodoxy,” they superimposed an “official” mythology upon it, just as the Manichaeans did. By the time this occurred, neither community was in a position to sort out an “original” Iranian myth from elements adopted from neighboring cultures; they worked with a mythological heritage that had a long history of cultural exchanges already, and made it into “Zoroastrian” or “Manichaean” myths. Mani and the Manichaeans actively sought connections with the myths of other cultures; the Zoroastrians might be credited with being more culturally purist, except for the testimony of the Dēnkard, which suggests to the contrary that, in the early Sasanid period, a similar multicultural appropriation was at work.6 They were evidently less successful than the Manichaeans in crafting a monolithic official mythology, as indicated by the inconsistencies and disarray in medieval Zoroastrian sources. Perhaps it is anachronistic to assume they even tried to with any persistence; they may have been traditionalists in the sense of being more concerned with conformity of practice than doctrine. Much depends on the use to which the full Avesta as described in the Dēnkard was put, since what was carefully preserved through the traumas of the Islamic conquest represented only the core ritual nasks (individual books of the hypothetical complete Avesta discussed in later sources).



There has been a considerable amount of attention in previous studies to the Manichaean use of the Iranian pantheon and to the independent witness these sources may provide to a “popular” pantheon outside of Zoroastrian control.7 Both Manichaeism and Zoroastrianism superimposed a hierarchy of divine order on an underlying Iranian polytheism. Here is not the place to belabor the relative uselessness of “monotheism” as a term that captures how anyone in the ancient world understood the universe to be governed. Most religious cultures of Late Antique West Asia held the idea of a divine sovereign overseeing the work of many divine subordinates, whether those were termed “gods” or “angels” or something else. Zoroastrianism and Manichaeism were not different in this regard. Both made selective use of an older Iranian pantheon as suited their respective theologies, adjusting the definitions and roles of specific deities within a coordinated divine order. That older pantheon is attested in inscriptions and art from the Achaemenid, Seleucid, and Parthian periods, on Kushan coins and inscriptions, and in evidence from the first century BCE for deities associated with the months and days of the “Zoroastrian” calendar.8


The most significant difference between Zoroastrian and Manichaean interpretations of the Iranian pantheon is the latter’s identification of Ohrmazd not as the supreme deity, but as the agent of the divine that enters into dualistic combat with evil. A higher god stands behind Ohrmazd, which in some contexts is identified by the name Zurwān (or its equivalent, e.g., Sogdian Azrua). This dyad of the father Zurwān and the son Ohrmazd is found already in Mani’s Šābuhragān.9 The discovery of the Turfan texts containing this theology fed a hypothesis about the existence of “Zurvanism” in Sasanid-era Zoroastrianism, supported by a small number of other testimonies. There are a number of anomalies in Zoroastrian texts that might reflect a background in older Iranian religion for Mani’s theology.10 But Skjærvø is correct to point out that, in any case, other than using the name “Zurwān,” Manichaeism has very little to do with theology reconstructed for any supposed “Zurvanism” (Skjærvø 1995a, 269–72).


The proposition that “Zurvanism” represented a significant form of Iranian religion in late Antiquity was bolstered by the discovery of the same theological hierarchy in Sogdian translations of Buddhist texts, in which Azrua (=Zurwān) and Ohrmazd correspond with Brahma and Indra, respectively, in the original Sanskrit source texts. This evidence has been interpreted to mean that Sogdiana belonged to that non-dualist sphere of Iranian religious culture from which “Zurvanite” traditions emanate (e.g., Zaehner 1971, 22). On the other hand, given the late date of Sogdian Buddhist texts relative to the early and lasting presence of Manichaeism among Sogdian populations, it could be that Buddhists relied on the Manichaean pantheon in rendering their Indian gods into something familiar to Sogdians. It is important to note that most Sogdian Buddhist texts appear to have been translated outside of Sogdiana, from Chinese versions in the Tang period, rather than in Sogdiana directly from Indian originals (Dresden 1983, 1221–4). It is not a matter, then, of an ancient Sogdian pantheon featuring Zurwān and Ohrmazd with a recognized correlation in the Indian pantheon to Brahma and Indra from centuries of cultural contact. Rather, it appears to have been a situation where Buddhist missionaries, translating from Chinese Buddhist texts, sought appropriate identifications for the Indian gods in the contemporary medieval religious culture of the Sogdians—and that culture had a significant Manichaean component. If this is the case, then a major pillar of the “Zurvanism” hypothesis must be set aside; and it may have been due to Manichaeans that a Zurwān-Ohrmazd divine dyad was available to translators of Buddhist texts. Other Iranian religious leaders set about reducing Zurwān traditions to abstractions that would not threaten their primary dualism, with only traces of more personified versions of the entity preserved largely in polemics.


A wide array of other Iranian deities were incorporated into Iranian Manichaeism at one stage or another. There is every reason to believe that Manichaeans were adopting these divine figures from broader Iranian regional pantheons, and not directly from the place they later would hold in a more systematized Zoroastrianism.11 There remains a need to examine closely the different underlying pantheons reflected in the different roles assigned to Iranian gods in Middle Persian versus Parthian versus Sogdian Manichaean texts. Mani’s own Middle Persian composition, the Šābuhragān, provides not only the earliest record of the Iranian identifications of the Manichaean pantheon, but even the founder’s own understanding of how the deities he recognized related to the gods worshipped by the Iranians. These gods of the Šābuhragān include Zurwān (for the supreme god Father of Greatness), Ohrmizd (for the Primal Man), the Amahraspandān (for the five divine elements), including Frawahr for one of those elements, Mihr (i.e. Mithra, for the demiurge Living Spirit),12 Nerisah (i.e. Nēryōsang, for the Third Messenger),13 Gēhmurd (i.e. Gayōmard, for Adam), Murdyānag (i.e. Mašyānag, for Eve), and Ahrimēn (for the King of Darkness).


Other Iranian deities found in Manichaean texts include Ardawahišt (for one of the divine elements in Sogdian texts), Wēšparkar14 (for the Living Spirit in Sogdian texts), Wahrām (for the divine warrior Adamas in Sogdian texts), Spendārmad (for the ruler of the foundations of the earth, the King of Glory, in Sogdian texts), Srōš (for the Column of Glory in Middle Persian and Sogdian texts), Sadwēs (for the Maiden of Light in Parthian texts), and Wahman (for the Light Nous in Middle Persian and Sogdian texts). From its relative absence from this second list, it becomes clear that Parthian Manichaean literature displays an intriguing independence; for the most part, it only carries over the identifications made by Mani in the Šābuhragān (Zurwān, Ohrmized, Frawahr, Nerisaf, Mihr, Gēhmurd, and Ahrimēn), and does not expand upon them by drawing in more Iranian gods the way Middle Persian and Sogdian Manichaean literature does. Parthian Manichaeans preferred to use descriptor titles for the gods, often direct translations of Mani’s Syriac titles, rather than names drawn from the Iranian pantheon. We cannot say whether this tells us something significant about pre-Manichaean Parthian religious culture or reflects a peculiar preference of the leadership of the Manichaean mission to Parthia (initially, Mani’s disciple Ammo). Parthian Manichaean literature also diverges from its Middle Persian cousin in one key character of the Iranian pantheon, even defying Mani’s own identification of Mithra with the demiurgical Living Spirit, and instead recognizing in Mithra the sun god Third Messenger. This divergence has been plausibly connected to cultural differences between Parthians and Persians on the role ascribed to Mithra (Boyce 1962).15


The theory that all these Iranian deities in Manichaean texts amount to mere window-dressing and missionary cultural accommodation fails to take into account Mani’s own declarations about his relationship to prior religious traditions. In crediting “Zarades” with establishing the “truly-founded law” in Persia, the same law “truly-founded” by other messengers of God, including himself, Mani credits the Iranian pantheon with revealing at least in part the administrators of the universe. He assumes that some distortions and misunderstandings of their identities and roles have arisen in the process of transmission over time; but his own identifications and clarifications address and resolve that problem. As Mary Boyce explains,


Mani, believing as he did that the prophets who were his forerunners had taught the truth, necessarily also thought that the gods whom they had preached were true gods, made known to diverse people under different names. The “translation” of the names of Manichaean gods by those of other deities must therefore be held an attempt by him to discover his own gods under their older, local guises. (Boyce 1962, 44)


In short, Mani’s own teachings affirm the reality and worthy-of-worship status of these traditional Iranian deities, just as Zoroastrianism does. He highlights certain features of their identity in the existing lore, and selects mythic material regarding them, redacting it according to his religious views, just as Zoroastrian leaders did in their own work systematizing Iranian theology within a system where they were subordinate to Ahura Mazda and were made to embody the values of “Zoroastrianism.” Iranian people, won to Mani’s faith, brought with them additional popular understandings and mythologies of particular deities, which at times overpowered the deity’s officially-sanctioned place and role, just as occurred in the Zoroastrian tradition. Both religions are heirs to a prior Iranian assortment of gods, which they seek to interpret in line with their respective theologies.

Veneration and Ritual Support of Natural Elements


Both Manichaeans and Zoroastrians have strict rules about contact with natural elements, believe that one contracts damning sin by polluting the elements, and do ritual work to support those natural elements. Fire and water in particular are ritually fed in the Zoroastrian ātaš-zōhr and āb-zōhr rituals (Boyce 1966). These rituals were also known and explicitly referred to by the Manichaeans, e.g. in the ritual script, Gwyšn ‘yg gryw zyndg. The latter text explicates older Iranian ritual traditions in terms of the divinity of the elements and the identification of oneself with their plight.


I am the fire which Zardrušt built, and which he bade the righteous build. From the seven consecrated, sweet-smelling fires bring to me, the fire, purified fuel. Bring clean firewood, and delicate and fragrant incense. Kindle me with knowledge, and give me clean zōhr. I am the water which is fit that you should give me the āb-zōhr, that I may become strong. (M 95.v.1-12; Andreas and Henning 1933, 317–18)


Previously, one might have contended that such references arose in Manichaeism as it interacted and competed with “Zoroastrianism” in Iran in the centuries following Mani. New evidence, however, suggests otherwise.


The Chester Beatty Kephalaia contains reports of Mani’s teaching activity that similarly references these rituals. A judge named Adurbat, who holds court “outside the gate of the fire temple,” speaks of the ātaš-zōhr and āb-zōhr rituals, saying, “We ourselves also gather the sticks and the [flowers (?)] […] and we speak over them and we give power […] the fires and the waters” (2Ke 358.12-359.12; Gardner, BeDuhn, and Dilley 2018, 36–39). The passage goes on to speak about ritual acts connected to maintaining a consecrated fire, of which Mani gives an esoteric interpretation (2Ke 360.25-364.3; Gardner, BeDuhn, and Dilley 2018, 40–49). The contents of this late fourth-century Coptic codex, while certainly hagiographical and idealized, go back to stories about Mani already set down in writing and put into circulation in the late third century. The transmission of such material even in settings where it had no missionary value, and would not even be fully understood outside of the Iranian context, demonstrates that it does not represent a local cultural adaptation but concepts considered core to Manichaean teaching.


Rather than see these rituals as integral parts of an organized “Zoroastrian church,” appropriated and reinterpreted by Manichaeans as parasitical heretics, we need to consider them as traditional Iranian religious practices that both religions took up and fit into their developing ritual and ideological systems. After all, Yašt 5.8 associates the āb-zōhr with the worship of Anāhitā, and there is every reason to consider it a traditional rite of that deity appropriated and incorporated into Zoroastrianism. This rite is attested in an Avestan passage quoted in Nērangestān 2.30, concerned with technical details of its proper performance. The Middle Persian Nērangestān illustrates how a Sasanid-period (or later) commentator quoted ritual rules and procedures from an Avestan source (or sources), and commented and expanded upon them to fit them into a system of orthopraxis, quite evidently altering the sense of the original in many cases to produce a different understanding of proper procedure (see Kotwal and Kreyenbroek 1995–2009).16 The Avestan material embedded in the text corresponds with the type of priestly ritual instructions found in ancient Mesopotamian texts, as well as in Indian ritual manuals and commentary. The Nērangestān itself is more akin to the Mishnah, by which older Jewish ritual practices were recalled, codified, and commented upon as a means of standardizing the practice of “Judaism.”


Mani, too, undertook a process of codifying and commenting upon older ritual practices from Iranian culture. He preserved the meaning and significance of such rituals as constituting aid to the divine elements, but he subsumed the actual ritual acts of sacrificial libation within a ritual meal; the divine elements take the form of food, and are strengthened by passing through the Manichaean priests, the Elect, who consume the food. Since the ātaš-zōhr and āb-zōhr as preserved in Zoroastrian practice include the consumption of part of the offering by the priests performing the ritual, Mani’s ritual reform can be seen as merely emphasizing that part of the ritual (due to comparison to other regional ritual meals) at the expense of the accompanying libations directly to fire and water (BeDuhn 2000). Mani understood such ritual reform to be necessary, convinced that in their traditional form the rituals had become corrupted and useless (1Ke 217.2-20).


The retention of the traditional rites more or less in their ancient Iranian form can be seen as one of the principal points of separation between an emergent Manichaeism and an emergent Zoroastrianism. Kerdīr, in his inscription, appears to report his efforts to confirm the validity of the traditional rites as efficacious, implicitly in the face of those, such as the Manichaeans, who had declared them worthless or misunderstood. The inscription therefore attests the formation of a defined “traditionalist” reaction to ritual innovation, which was beginning to consolidate itself as the “Zoroastrian” religion, by reacting against the new prophet Mani and his claims to better understand the intentions of the cultural hero Zarathustra. To resist Mani’s interpretations, some subset of the traditional priests had to organize, systematize, and justify those practices they wished to maintain.

Use and Interpretation of Iranian Religious Literature


Kerdīr, in his inscription, appears to refer to a “nask,” a sacred text (Skjærvø 1983, 276, 290–91), and a similar reference to a “nask” is made in the early Manichaean texts, e.g. in an account of Mani’s exchange with King Bahram and in the Sermon on the Soul (Sundermann 1981, 72, 1997a, 76–81). Both traditions draw on this earlier Iranian religious literature, interpreting and applying it. For Kerdīr, it provides an antecedent to his own investigations of the other world, apparently in its descriptions of that other world. For the Manichaean Sermon on the Soul, the “nask” contains the five Gāthās, each correlated to one of the five “children of Ohrmazd,” who constitute the natural elements: the air called Frawardin is “called in the Nask the Ahunavaiti-Gāthā (’whnwyt g’ẖ),” the wind is “called in the Nask the Uštavaiti-Gāthā (’wyštwyt g’ẖ),” the light is “called in the Nask the [Spǝntamaniiuš-Gāthā?],” the water is “called in the Nask the Vohuxšathra-Gāthā (whwxštr g’ẖ),” the fire is “called in the Nask the [Vahištoišti-Gāthā?]” (Sundermann 1997a, 76–81). This Parthian Manichaean composition, which there is good reason to date to the third century, thus attests by that date the well-established reverence for the five Gāthās as a canonical set contained within a “nask” or “nasks” that might be referenced elsewhere in Manichaean literature as “the books of Zaradēs.”


The exact meaning of such Avestan ritual texts was already obscure in Sasanid times, as the Middle Persian glosses make evident (Skjærvø 1995a, 265–6). As Pallan Ichaporia has demonstrated, the creators of the Pahlavi “translations” misunderstood much of the language of the gāthās, or perhaps creatively read into them their own later religious culture (Ichaporia 2006). In these conditions, Manichaeans were as free as Zoroastrians to make what sense of these obscure chants they could, and to employ them in whatever ritual context they thought appropriate. Before we dismiss the associations made by the Sermon on the Soul between specific gāthās and elements as arbitrary, we should investigate a possible ritual basis for them known to third-century Iranian religious practice but lost to later Zoroastrian tradition. We owe to a later Arab Muslim writer the suggestion that Manichaeans were called Zandiks because they supplied an interpretation (zand) to the gathic texts that their religious rivals—that “traditionalist” reaction referred to above—chose to recite verbatim without interpretation.17 For some parts of their tradition, at least, Zoroastrians were learning, memorizing, and reciting Avestan texts without even an oral instruction to elucidate their meaning (Skjærvø 1997, 319–21).


In other Manichaean texts, we see reflections of particular passages of the Gāthās, Yašts, and other Avestan literature. They were apparently quite capable of simply transmitting and using traditional Avestan texts, such as the Ashem vohu prayer, preserved in a Sogdian Manichaean text, as recognized by Ilya Gershevitch (Gershevitch 1976). The “Complaint of the Cow” of Yasna 29 appears to lie behind an allusion in the Coptic Psalms of Thomas 20 (Allberry 1938, 226–27). Werner Sundermann ingeniously demonstrated that Mani knew Yašt 10 in a Middle Persian version, based on the Šābuhragān’s identification of four of the five sons of the deity Living Spirit (Mihr) by epithets associated with Mihr in Yašt 10.115, not in their Avestan form, but in their Middle Persian equivalents: mānbed, wisbed, zandbed, dahybed (Sundermann 1979b). This discovery supports Skjærvø’s view that Middle Persian versions of the Avestan texts were set down in writing by the early Sasanid period even before the original Avestan ones were, later in the fifth or sixth century CE (Skjærvø 1997, 319–21). He argues that, when Kerdīr cites “the nask” in Middle Persian in his inscriptions, he is quoting directly from such vernacular versions of the Avestan texts, which were the ones that could actually be read for their sense, not just recited by rote, as the original Avestan ones were. This view of the transmission history of Avestan texts is further supported by references in both Manichaean Kephalaia volumes to books holding traditions of the teachings of Zarathustra. In the Chester Beatty Kephalaia, interlocutors of Mani quote “written” words of Zarathustra and ask Mani to interpret them (2Ke 415.25-417.14; Gardner, BeDuhn, and Dilley 2018, 150–55).


But were these texts identical to those that became canonical in Zoroastrianism? The words of Zarathustra quoted in the Chester Beatty Kephalaia do not seem to correspond to anything preserved in Zoroastrian literature. Moreover, as mentioned, Mani’s identification of the five sons of the Living Spirit correspond only in four out of five cases with the epithets of Mihr in Yašt 10.115; it diverges on the fifth. Mani’s fifth term is pāhragbed, “head of the frontier post,” whereas in the Zoroastrian version of the Yašt the fifth epithet is zarathuštrōtǝma in Avestan, and in Middle Persian would have been Zarduštrōtum (the Middle Persian of Yašt 10 is not preserved), the one who carries the aegis of Zarathustra.18 Sundermann assumes that Mani changed a source corresponding to a Middle Persian rendering of the current Zoroastrian Avestan text (Zarduštrōtum > pāhragbed; Sundermann 1979b, 785, cf. 1997b, 345). But it is also possible that Mani preserved the wording of the version of the Yašt known to him. He follows the exact order of terms, even though it forces an unusual sequence of the five sons of the Living Spirit. In the context of praise of Mihr, Mani’s wording has a more consistent sense (i.e., Mihr as “lord of wide pastures” is “head of the house, head of the clan, head of the tribe, head of the country, head of the frontier post” vs. “head of the house, head of the clan, head of the tribe, head of the country, Zarathustra-bearer”). In light of parallel references to a corresponding five ranks of authority elsewhere in Avestan literature that differ precisely in the fifth term,19 an alteration of “head of the frontier post” to “Zarathustra-bearer” would then be a Zoroastrian redaction of this passage of the Yašt, perhaps under the influence of a parallel passage from Yasna 19.18 (cf. 71.22), which in its Middle Persian version uses the five epithets with Zardušt as the fifth term in a context not related to Mihr.20 In other words, we cannot assume that where Manichaean and Zoroastrian versions of Avestan passages differ, the (medieval) Zoroastrian texts always preserve the original reading.



Manichaeans knew traditions of Zarathustra as the ritual hero of Iranian culture. In Manichaean hands, these stories are not only recounted through the lens of Manichaean views, but begin to fashion a hagiography of Zarathustra as a prophet to the Iranian people earlier than similar developments in the Zoroastrian religion.21 As in many younger Avestan texts and the content summarized from the Sasanid Avesta in medieval Zoroastrian literature, Zarathustra has question-and-answer dialogues with God in Manichaean texts. But, in the fashion of ancient Near Eastern wisdom literature, he also has a dialogue with his own soul (which reveals both its divine identity and its identification with all souls; see Asmussen 1975, 47–49). Manichaean texts, like Zoroastrian ones, report his confrontation with evil opponents in Babylon (Hom 11).22 Skjærvø scours both Greek and Pahlavi sources for comparable stories, but connections are tangential at best. Only Dēnkard 7.4.72 alludes to Zarathustra destroying the sorcery and idolatry created by the dragon Dahāg in Babylon (Skjærvø 1995b, 611). He handles fire without himself being burned by it (2Ke 363.2-3; Gardner, BeDuhn, and Dilley 2018, 46–47). His death and funeral are alluded to (Hom 68-70). Zarathustra’s associates also find mention, including Vīštāspa/Hystaspes,23 his queen Hudōs (T II D 58; Henning 1943, 73–74), and Jāmāsp (TM 393; Henning 1944, 141). A fascinating aspect of this construction of Zarathustra as a full-bodied prophet is the explicit comparison Manichaeans made between his teachings and those of other prophet figures, such as Jesus (2Ke 415.25-420.28; Gardner, BeDuhn, and Dilley 2018, 150–61). The Manichaeans crafted an anachronistic depiction of him founding a “church,” choosing “disciples,” and introducing rites (2Ke 362.1-363.25; Gardner, BeDuhn, and Dilley 2018, 44–47)—a “Mani before Mani” or “Jesus before Jesus” (Sundermann 2005, 66). Zoroastrianism followed suit, similarly turning their legendary priest-preceptor into a “prophet,” while carefully highlighting or inventing aspects of his image in reaction to Manichaeism to make clear his advocacy of anti-Manichaean values connected to hearth and home (Hutter 2009).

Heroic Legend


It has recently become increasingly evident that early Manichaeans also took up the broader set of Iranian heroic legends, perhaps already with Mani’s own Book of Giants. It must be noted that Skjærvø was not in a position to confirm that these Iranian heroes had any place in Mani’s own writings; all that was certain twenty years ago is that Iranian translations of Mani’s Book of Giants used the names of Iranian heroes in place of originally Semitic names, and this could have been simply a matter of cultural translation for Iranian audiences of a narrative that owed nothing to prior Iranian traditions. Only in recent years has Enrico Morano published additional fragments of Mani’s Book of Giants that confirm that the Iranian heroic narratives themselves played a role in that book (Morano 2009). Confirmation of this engagement with the Iranian heroic tradition from the inception of Manichaeism, and not merely as a regional adaptation, comes, quite surprisingly, from the Coptic Chester Beatty Kephalaia, which contains a detailed recounting of the legend of King Khusrau’s abdication of his throne in quest of Paradise, best known from the Shah-nama, but now in a text some five centuries earlier than the medieval Iranian epic. It bears noting that orthodox Zoroastrianism never had much use for many of these figures; they remained marginal intruders from popular Iranian culture.


Skjærvø has demonstrated that Manichaean Sogdian fragments preserve a version of a legend about Vīštāspa, his brother Zarēr, and the counselor Žāmāsp found in the Zoroastrian text Ayādgār-ī Zarērān (Skjærvø 1995b, 614–18). Although Skjærvø attributes the wide divergences between the Manichaean and Zoroastrian versions to “Mani, who distorted it to suit his own version of mythical history,” the differences look more like the sort of changes typical of orally transmitted narratives. Key dramatic acts are retained, but the role of characters in connection with those acts shift and even invert. Whereas in the Zoroastrian tale the three aforementioned characters are all arrayed on one side against the evil hordes of the Xyōns, in the Manichaean version Zarēr and Žāmāsp oppose Vīštāspa and Zarathustra. Here again, we cannot assume either that the Zoroastrian version is older than the Manichaean one or that there is always an ideological motive for variations in traditional material.



Even though both Manichaeism and Zoroastrianism share a theoretical dualism, how they apply that dualism to rules of conduct in the world differs considerably. Both religions employ the structural moral triad of “thought, speech, action,” which of course goes back to the Gāthās that they shared. The difference lies in the specific content of moral injunctions included in each category. Zoroastrian polemical literature strongly vilifies Manichaean asceticism. Besides the sharp division over sexuality, Manichaeans differed from Zoroastrians in how to apply antecedent Iranian traditions about the sanctity of life. Those traditions included reverence and protection of “good animals,” such as cattle and otters, and Manichaeans extended this attitude to all animals, while Zoroastrians identified a category of “bad animals,” the khrafstra, whom they would systematically slaughter as a manifestation of Ahriman. The confrontation of these two views is expressed in two stories in the Chester Beatty Kephalaia. In one, Mani encounters and confronts the slaughter of a wolf as part of a festival khrafstra hunt (2Ke 345.10-350.13; Gardner, BeDuhn, and Dilley 2018, 10–21; cf. Henning 1945, 476–77). In another, the aforementioned tale of King Khusrau, the latter’s companion Iuzanes (= Vēžan) asks about this Paradise to which Khusrau wishes to go: “This land to which you will go, is there food and drink in it? Does one marry women there, and do they become pregnant and give birth? Is there gold and silver, war and hunting?” To which the king replies, “There is not a single one of these things in that place” (BeDuhn 2015a). This passage encapsulates the Manichaean critique of popular Iranian values.


Of course, there were ascetic strands in broader Iranian culture that even worked their way into some branches within medieval Zoroastrianism, where they were sharply contested. Buried deep within Zoroastrian literature are hints of attitudes closer to those of the Manichaeans. At the coming of the future restorer Hōšēdarmāh, Zoroastrians expect humankind to give up killing and eating animals (Dēnkard 7.10; Bundahišn 34.2), exactly as Mani espoused. Here, too, rather than assume Mani imposed a reversal of values on Iranian traditions, we need to consider how both Manichaeans and Zoroastrians chose to emphasize certain elements from a complex and ambivalent antecedent religious culture. One might even suggest that their competitive origins contributed to their emphasis of different elements as a means of demarcation and differentiation.



Mani correlated Judeo-Christian and Iranian eschatological traditions in his Šābuhragān. Of course, it has been long accepted that Judeo-Christian eschatology already reflects the influence of Iranian eschatological traditions. On the basis of the latter, both Manichaeism and Zoroastrianism posit an immediate post-mortem judgment unknown to orthodox Judaism and Christianity at the time, as well as the idea held in common by all four religions of an eschatological final reformation of the world. We are still in need of a careful comparison of these different strands of eschatology in Late Antique West Asia, including a convincing account of how an immediate post-mortem judgment like that found in Zoroastrianism and Manichaeism came to be added to Christian eschatology that did not originally have it. We have the advantage that Kerdīr includes an account of a visionary journey that emulates the path of the dead; it differs in a number of details from later orthodox Zoroastrian descriptions, and in some details it is closer to Manichaean accounts (see Skjærvø 1983). Especially notable is a judgment before an enthroned “prince” (šahryār) who measures the deceased’s deeds with scales. This figure corresponds with Rašn in later Zoroastrian literature (e.g., Mēnōy-ī Xrad 1.118-120; Ardā Wirāz-nāmag 5.5), and the Great Judge in Manichaean teachings; but no such figure or weighing in scales appears in Avestan texts, and it appears to be borrowed from western religious culture—via the Manichaeans? Once again, we see a formative stage where older Iranian traditions are being parsed in different ways by different emerging communities of interpretation. In all such cases, we need to refrain from treating either Mani or Kerdīr as the norm from which the respective other deviates, but do what we can to determine possible earlier Iranian traditions from which both lines of interpretation and application may derive.


The idea of an immediate post-mortem judgment appears already in Avestan literature, even though many of its details appear in medieval Zoroastrian texts for the first time. The recently discovered tomb of the Sogdian couple Wirkak and Wiyusi from Xi’an provides significant confirmation of the existence of some of these elements by the sixth century (Gulácsi and BeDuhn 2012). Similar ideas of post-mortem judgment were embedded deeply in Hellenic and Hellenistic culture, however. In the century or two before Plato, “Orphic” groups had transformed older notions of judgments, rewards, and punishments awaiting the special dead into afterlife experiences everyone should expect at death, offering guides and remedies for avoiding possible bad outcomes.24 Plato’s inclusion of these ideas in his works spread the belief in intellectual circles. His Gorgias speaks of “suffering throughout eternity the greatest and most excruciating and terrifying tortures because of their misdeeds,” following judgment after death (525c-526d), while his Myth of Er (Republic 614b-621d) combined judgment, punishment, and transmigration. Jewish and Christian accounts of judgment of the dead made use of this imagery for eschatological judgment, even before embracing the idea of an immediate post-mortem judgment, however much in tension with their earlier views. The Apocalypse of Paul, whose original date is debated,25 has the rewards and punishments doled out by a “just judge” immediately after death, as in Manichaean belief. Michel Tardieu has argued that both Manichaean and Zoroastrian accounts of judgment in Late Antiquity derive some of their crucial details from this long-standing Hellenistic tradition, via Christian apocalyptic literature (Tardieu 1985). If that argument proves persuasive, the next question to investigate is whether both Manichaeans and Zoroastrians adopted these ideas directly and independently from western sources, or if one of them was the mediator of the ideas to the other. If Tardieu is correct that Christian apocalyptic literature serves as the primary source, then it would be much more likely that Manichaeans mediated the material to Zoroastrians, rather than the reverse.



This survey, therefore, both invites further investigation into Manichaean use of Iranian traditions and cautions against approaching that subject through the assumption that our medieval Zoroastrian sources give us a reliable picture of what those traditions looked like already before Mani. This reorientation in our approach parallels recent developments in the study of Christian origins. In the nineteenth and twentieth centuries, historians relied heavily on Rabbinic literature to reconstruct the Jewish religion from which Christianity developed and diverged. Only in the last generation have more and more researchers come to realize that Rabbinic texts are well removed from earlier conditions, and that Judaism and Christianity co-developed in the first half-millennium CE, both making use of prior Israelite religious traditions.


In light of this dawning awareness, we can begin to think of Zoroastrianism as something that came into existence as a nativist and traditionalist reaction to conditions of religious options and innovations that existed in the third century. Like Judaism developing against the challenge of Christianity, or Hinduism developing against the challenge of Buddhism, such a nativist and traditionalist reaction has the quality of reinforcing the traditional interchangeability of religious and ethnic identity. Any “converts” must be incorporated into the ethnicity, not just into a theoretically separable “religion.” The Zoroastrian “religion” emerged through a process in the third to seventh centuries CE out of the traditional religious practices of Iranian culture, in self-conscious response to the presence and challenge of alternative religious identities. Third-century Iranian religious leaders found themselves caught in the middle of the cultural tensions that exist within multi-ethnic and multi-cultural empires, such as the new Sasanid Empire was. The traditional local priesthoods could parlay their existing status into access to imperial power. This can be seen clearly in the rise to power described in Kerdīr’s inscriptions. Although he repeatedly claimed being given absolute authority over religious rites, he acknowledged a progression of becoming ‘more absolute’ from reign to reign, acquiring higher titles, and only eventually being given authority over some of the key traditional pilgrimage and ritual sites of the empire. Aligned with state power, Kerdīr could aspire to suppress alternative religious practices and even to impose the native Iranian religion on non-Iranians. This aspiration came with various normativizing and institutionalizing moves that more sharply defined approved practices over against deviant ones, regularizing and rationalizing something we can begin to think of as an emergent “Zoroastrianism.”


That which later Zoroastrian tradition describes as the gathering of scattered fragments of their sacred literature in the Sasanid period needs to be understood as the formative process of Zoroastrianism against the background of cultural materials they were selecting and codifying. As with any new religion or reform, they imagined a past they were simply rediscovering and reinstituting, projecting their own situation and needs onto their cultural forebears. Read in that light, the Zoroastrian account is quite informative. Still, we must be cautious about picturing a highly centralized and organized process until very late in Sasanid history. Notoriously, the later tradition celebrates a certain Tansar wholly missing from inscriptions, while the loudest voice in the inscriptions, Kerdīr, is completely absent from the later tradition. We now recognize that a precise written form of Avestan texts was undertaken only later in the Sasanid era, and the supposedly ancient content of those texts shows signs of redaction and interpolation, even fresh if stilted composition, right up to the end of that era.


Manichaean sources, including the newly edited material highlighted here, put us in a new position with regard to reconstructing and understanding the state of Iranian religion in the third century, and extrapolating from that which aspects go back to earlier centuries and which arose in the process of forming two rival religions on the basis of those earlier traditions. That material should be taken full advantage of by those working on the history of Zoroastrianism, and of broader Iranian religious culture, even if they hesitate to embrace the particular historical argument of this article.


Allberry, C. R. C. 1938. A Manichaean Psalm-Book: Part II. Stuttgart: Harrassowitz.

Andreas, Friedrich Carl, and Walter Bruno Henning. 1933. “Mitteliranische Manichaica aus Chinesisch-Turkestan: II.” In Sitzungsberichte der Preussischen Akademie der Wissenschaften 1933, 292–363. Berlin: Akademie der Wissenschaften.

Asmussen, Jes Peter. 1975. Manichaean Literature: Representative Texts Chiefly from Middle Persian and Parthian Writings. UNESCO Collection of Representative Works: Persian Heritage Series 22. Delmar, NY: Scholars’ Facsimiles & Reprints.

Barton, Carlin A., and Daniel Boyarin. 2016. Imagine No Religion: How Modern Abstractions Hide Ancient Realities. New York, NY: Fordham University Press.

BeDuhn, Jason D. 2000. “Eucharist or Yasna? Antecedents of Manichaean Food Ritual.” In Studia Manichaica: IV. Internationaler Kongress zum Manichäismus; Berlin; 14.-18. Juli 1997, edited by R. E. Emmerick, Werner Sundermann, and Peter Zieme, 14–36. 4. Berlin: Akademie Verlag.

———. 2015a. “Iranian Epic in the Chester Beatty Kephalaia.” In Mani at the Court of the Persian Kings: Studies on the Chester Beatty Kephalaia Codex, edited by Iain Gardner, Jason D. BeDuhn, and Paul Dilley, 136–58. Nag Hammadi and Manichaean Studies 87. Leiden: Brill.

———. 2015b. “Mani and the Crystallization of the Concept of ‘Religion’ in Third Century Iran.” In Mani at the Court of the Persian Kings: Studies on the Chester Beatty Kephalaia Codex, edited by Iain Gardner, Jason D. BeDuhn, and Paul Dilley, 247–75. Nag Hammadi and Manichaean Studies 87. Leiden: Brill.

Bivar, A. D. H. 1988. “An Iranian Sarapis.” Bulletin of the Asia Institute 2: 11–17.

Boyce, Mary. 1962. “On Mithra in the Manichaean Pantheon.” In A Locust’s Leg: Studies in Honour of S. H. Taqizadeh, edited by W. B. Henning and E. Yarshater, 44–54. London, UK: Percy Lund, Humphries.

———. 1966. “Ātaš-Zōhr and Āb-Zōhr.” Journal of the Royal Asiatic Society 98 (2): 100–118.

Carter, Martha L. 2006. “Kaniska’s Bactrian Pantheon in the Rabatak Inscription: The Numismatic Evidence.” In Proceedings of the 5th Conference of the Societas Iranologica Europæa Held in Ravenna: 6–11 October 2003, edited by Antonio Panaino and Andrea Piras, 351–57. Milan: Mimesis.

Dresden, Mark. 1983. “Sogdian Language and Literature.” In The Seleucid, Parthian and Sasanian Periods, edited by E. Yarshater, 1216–29. The Cambridge History of Iran 3. Cambridge: Cambridge University Press.

Funk, Wolf-Peter. 2009. “Mani’s Account of Other Religions According to the Coptic Synaxeis Codex.” In New Light on Manichaeism: Papers from the Sixth International Congress on Manichaeism, edited by Jason D. BeDuhn, 115–27. Leiden: Brill.

Gardner, Iain, Jason D. BeDuhn, and Paul Dilley. 2018. The Chapters of the Wisdom of My Lord Mani: Part III; Pages 343-442 (Chapters 321-347). Nag Hammadi and Manichaean Studies 92. Leiden: Brill.

Gershevitch, Ilya. 1976. “The Sogdian Fragments: Appendix.” Indo-Iranian Journal 18 (1-2): 75–82. https://doi.org/10.1163/000000076790079492.

Grenet, Franz. 2006. “Mithra ii: Iconography in Iran and Central Asia.” In Encyclopaedia Iranica. http://www.iranicaonline.org/articles/mithra-2-iconography-in-iran-and-central-asia.

Gulácsi, Zsuzsanna, and Jason D. BeDuhn. 2012. “The Religion of Wirkak and Wiyusi: The Zoroastrian Iconographic Program on a Sogdian Sarcophagus from Sixth-Century Xi’an.” Bulletin of the Asia Institute 26: 1–32.

Hennecke, Edgar, and Wilhelm Schneemelcher. 1965. New Testament Apocrypha. Translated by R. McL. Wilson. Philadelphia, PA: Westminster.

Henning, Walter Bruno. 1943. “The Book of the Giants.” Bulletin of the School of Oriental and African Studies 11 (1): 52–74.

———. 1944. “The Murder of the Magi.” Journal of the Royal Asiatic Society, nos. 3-4: 133–44.

———. 1945. “Sogdian Tales.” Bulletin of the School of Oriental and African Studies 11 (3): 465–87.

Himmelfarb, Martha. 1983. Tours of Hell: An Apocalytpic Form in Jewish and Christian Literature. Philadelphia, PA: University of Pennsylvania Press.

Humbach, Helmut. 1975. “Vayu, Śiva und der Spiritus Vivens im ostiranischen Synkretismus.” In Monumentum H. S. Nyberg, edited by J. Duchesne-Guillemin, 1:397–408. Leiden: Brill.

Humbach, Helmut, and Prods Oktor Skjærvø. 1983. The Sassanian Inscription of Paikuli: Part 3.2. Commentary. Wiesbaden: Reichert.

Hutter, Manfred. 1992. Manis kosmogonische Šābuhragān-Texte. Studies in Oriental Religions 21. Wiesbaden: Harrassowitz.

———. 2009. “Das so genannte Pandnāmag ī Zardušt: Eine zoroastrische Auseinandersetzung mit gnostisch-manichäischen Traditionsgut?” In Literarische Stoffe und ihre Gestaltung in mitteliranischer Zeit: Kolloquium anlässlich des 70. Geburtstages von Werner Sundermann, edited by Desmond Durkin, Christiane Reck, and Dieter Weber, 123–33. Beiträge zur Iranistik 31. Wiesbaden: Reichert.

Ichaporia, Pallan. 2006. “The Gāthās in the Pahlavi Tradition of Late Sasanian and Early Islamic Periods: Exemplified by Yasna 30.3 and the Pahlavi Text of the Ahunavaitī Gāthā as Given in the Dēnkard Book IX.” In Proceedings of the 5th Conference of the Societas Iranologica Europæa Held in Ravenna: 6–11 October 2003, edited by Antonio Panaino and Andrea Piras, 519–41. Milan: Mimesis.

Kotwal, Firoze M., and Philip G. Kreyenbroek. 1995–2009. The Hērbedestān and Nērangestān. Nērangestān, Fragards 1-3. Paris: Association pour l’avancement des études iraniennes.

König, Götz. 2020. “The Teaching of the ‘Two Principles’ (dō bun) and the Motif of Light and Darkness in Late Antique Iran.” Entangled Religions 11 (2).

Lieu, Samuel N. C. 2006. “‘My Church Is Superior . . .’ Mani’s Missionary Statement in Coptic and Middle Persian.” In Coptica, Gnostica, Manichaica: Mélanges offerts à Wolf-Peter Funk, edited by Louis Painchaud and Paul-Hubert Poirier. Bibliothèque copte de Nag Hammadi 7. Leuven: Peeters.

MacKenzie, David Neil. 1979. “Mani’s Šābuhragān.” Bulletin of the School of Oriental and African Studies 42 (3): 500–534.

———. 1989. “Kerdir’s Inscription.” In The Sasanian Rock Reliefs at Naqsh-I Rustam, edited by Georgina Herrmann, 35–61. Berlin: Reimer.

Morano, Enrico. 2009. “‘If They Had Lived . . .’ A Sogdian-Parthian Fragment of Mani’s Book of Giants.” In Exegisti Monumenta: Festschrift in Honour of Nicholas Sims-Williams, edited by Werner Sundermann, Almut Hintze, and François de Blois, 325–30. Iranica 17. Wiesbaden: Harrassowitz.

Nongbri, Brent. 2013. Before Religion: A History of a Modern Concept. New Haven, CT: Yale University Press.

Panaino, Antonio. 1990. “Calendars: i. Pre-Islamic Calendars.” In Encyclopaedia Iranica. http://www.iranicaonline.org/articles/calendars#pt1.

Reck, Christiane, Dieter Weber, Claudia Leurini, and Antonio Panaino. 2001. Manichaica Iranica: Ausgewählte Schriften von Werner Sundermann. Rome: Istituto Italiano per l’Africa e l’Oriente.

Reeves, John C. 1992. Jewish Lore in Manichaean Cosmogony: Studies in the Book of Giants Traditions. Cincinnati, OH: Hebrew Union College Press.

———. 1993. “Utnapishtim in the Book of Giants?” Journal of Biblical Literature 112: 110–15.

———. 2011. Prolegomena to a History of Islamicate Manichaeism. Sheffield: Equinox.

Rezania, Kianoosh. 2010. Die zoroastrische Zeitvorstellung: Eine Untersuchung über Zeit- und Ewigkeitskonzepte und die Frage des Zurvanismus. Wiesbaden: Harrassowitz.

Rosenfield, John M. 1967. The Dynastic Arts of the Kushans. Berkeley, CA: University of California Press.

Russell, James. 1990. “Kartir and Mani: A Shamanistic Model of Their Conflict.” In Iranica Varia: Papers in Honor of Professor Ehsan Yarshater, edited by D. Amin and M. Kasheff, 180–93. Acta Iranica 30. Leiden: Brill.

Scott, D. A. 1989. “Manichaean Responses to Zoroastrianism.” Religious Studies 25 (4): 435–57. https://doi.org/10.1017/S0034412500020035.

Shaki, Mansour. 1981. “The Dēnkart Account of the History of the Zoroastrian Scriptures.” Archív Orientální 49: 114–25.

Skjærvø, Prods Oktor. 1983. “‘Kirdir’s Vision’: Translation and Analysis.” Archaeologische Mitteilungen aus Iran 16: 269–306.

———. 1995a. “Iranian Elements in Manicheism: A Comparative Constrastive Approach; Irano-Manichaica I.” In Au carrefour des religions: Mélanges offerts à Philippe Gignoux, edited by Rika Gyselen, 263–84. Res Orientales 7. Bures-sur-Yvette: Groupe pour l’étude de la civilisation du Moyen-Orient.

———. 1995b. “Iranian Epic and the Manichean Book of Giants: Irano-Manichaica III.” Acta Orientalia Academiae Scientiarum Hungaricae 48: 187–223.

———. 1996. “Zarathustra in the Avesta and in Manichaeism: Irano-Manichaica IV.” In Convegno internazionale sul tema: La Persia e l’Asia Centrale da alessandro al X secolo, edited by S. Moscati, 597–628. Rome: Accademia Nazionale dei Lincei.

———. 1997. “Counter-Manichaean Elements in Kerdīr’s Inscriptions: Irano-Manichaica II.” In Atti del Terzo Congresso Internazionale di Studi “Manicheismo e Oriente Cristiano Antico”, edited by Luigi Cirillo and A. von Tongerloo, 313–42. Manichaean Studies 3. Turnhout: Brepols.

Sundermann, Werner. 1979a. “Namen von Göttern, Dämonen und Menschen in iranischen Versionen des manichäischen Mythos.” Altorientalische Forschungen 6: 95–133.

———. 1979b. “The Five Sons of the Manichaean God Mithra.” In Mysteria Mithrae, edited by Ugo Bianchi, 777–87. Leiden: Brill.

———. 1981. Mitteliranische manichäische Texte kirchengeschichtlichen Inhalts. Berliner Turfantext 11. Berlin: Akademie Verlag.

———. 1997a. Der Sermon von der Seele: Eine Lehrschrift des östlichen Manichäismus. Berliner Turfantexte 19. Turnhout: Brepols.

———. 1997b. “How Zoroastrian Is Mani’s Dualism.” In Atti del Terzo Congresso Internazionale di Studi “Manicheismo e Oriente Cristiano Antico”, edited by Luigi Cirillo and A. von Tongerloo, 343–60. Manichaean Studies 3. Turnhout: Brepols.

———. 2005. “Zoroastre, prêtre et prophète dans la doctrine des manichéens.” In La figure du prêtre dans les grandes traditions religieuses, edited by André Motte and Patrick Marchetti, 59–72. Collection d’études classiques 20. Leuven: Peeters.

Tanabe, Katsumi. 1995. “Earliest Aspect of Kanīṣka I’s Religious Ideology: A Numismatic Approach.” In In the Land of the Gryphons: Papers on Central Asian Archaeology in Antiquity, edited by Antonio Invernizzi, 203–15. Monografie di Mesopotamia 5. Florence: Casa editrice le lettere.

Tardieu, Michel. 1985. “L’Ardā Vīrāz Nāmag et l’eschatologie grecque.” Studia Iranica 14: 17–24.

Zaehner, R. C. 1971. Zurvan: A Zoroastrian Dilemma. New York, NY: Biblo and Tannen.

  1. This claim to a visionary confirmation of religious truth is closely paralleled in the inscriptions of Kerdīr, where he recounts his own vision of heaven and hell; see Skjærvø (1983). A very similar personal visionary journey is described in a classic Zoroastrian text set in Sasanid times, the Arda Viraf Namag, and both of these relate to even older regional tropes, as attested in, e.g., Plato’s Myth of Er and the Books of Enoch.↩︎

  2. I have added to the translation some of the original Middle Persian terms in parentheses.↩︎

  3. For passages regarding such other groups from Mani’s Gospel, see Funk (2009, 115–27).↩︎

  4. The passage Zaehner provides from Mēnōk i Xrat (text Z 8, 367-368) provides perhaps the clearest and most succinct representation of this theological tendency, with a transcendent Zurvan overseeing and mediating the time-bound dualist conflict of Ohrmazd and Ahriman.↩︎

  5. Cf. Zaehner’s texts (1971) F3(a), 429-431, from the Dēnkard, and F3(b), 431-432, and F7(b), 439, from Manichaean texts.↩︎

  6. On Shapur’s collection of Avestan material, see Shaki (1981).↩︎

  7. Still the most important of these studies, due to its systematic analysis, is Sundermann (1979a).↩︎

  8. On the latter, see Panaino (1990). While clearly reflecting a “Zoroastrian” hierarchy, and thus a theological distillation pre-dating Mani, the adaptation of the pantheon to the number of months and days imposed an arbitrary character on the selection; it includes a number of deities who have relatively minor places in Zoroastrianism, while omitting a dozen significant figures of Zoroastrian theology, as well as deities (e.g., Anahita) highly prominent in the period.↩︎

  9. MacKenzie (1979, 506, line 76), but the context is very damaged. In the cosmogonic section, “Ruler of Heaven” (whyštw šhry’r) is used instead (Hutter 1992).↩︎

  10. Note, e.g., that Ohrmazd creates himself as the seventh Amahraspandān, suggestive of a duplication to replace another deity in an older form of the narrative (Bundahišn 1.53).↩︎

  11. The evidence of the Kushan pantheon is crucial here, with Rosenfield (1967, 82) arguing that their identification as “Zoroastrian deities” is misleading and should be discarded. Rather, they represent a broader Iranian religious culture. Kanishka I’s coinage, offering by far the largest catalog, includes Miiro (Mihr, as sun god), Mao (Mah), Oado (Vāta), Athsho (Atash), Nana, Oesho (Vāyu), Ardoxsho, Ashaeixsho, Lrooaspo, Manaobago, Oanindo, Oaxsho, Oshragno, Pharro, Rishto, Shaoreoro, Iamsho, and on a single surviving example, Mozdooano [Mazda Vano, “Mazda the Triumphant,” see Tanabe (1995), p. 206). See Carter (2006).↩︎

  12. Mithra does not have a demiurgical role in Avestan materials, but Porphyry, de antro nymph. 6, reports such a role for him, and Hellenistic-Roman Mithras imagery suggests mythic traditions of such a demiurgical character rather close to Living Spirit material in Manichaean myths.↩︎

  13. By which choice Mani appears to show familiarity with traditions of the desirable beauty of this Iranian messenger god and his association with reproductive drives and substances. See Bundahišn 35.59-60; 14.5.↩︎

  14. Humbach (1975) argues for the derivation of this prominent Sogdian deity from Avestan Vaiiuš uparō.kairiiō, a form of the god Vayu (“Vayu who acts in the heights”).↩︎

  15. Parthian identification of Mithra with the sun may be reflected in the report of Strabo, Geographica 15.13.732. He appears as a solar deity in Kushan coinage and, indeed, already on Greco-Bactrian coins. Moreover, he appears with solar attributes in official Sasanid reliefs and coins, and on Sasanian period seals (see Grenet 2006).↩︎

  16. On the question of the date of the composition, see Kotwal and Kreyenbroek (1995–2009, III:17–18).↩︎

  17. This understanding of the term is found in Ma‘sūdī (Reeves 2011, 167), Ibn al-Āthīr (Reeves 2011, 186), Nuwayrī (Reeves 2011, 188); cf. remarks of Sam‘ānī (Reeves 2011, 104).↩︎

  18. Sundermann bases this hypothetical reconstruction on a structurally parallel passage in the Middle Persian version of Yasna 19.18 (1979b, 784–85).↩︎

  19. It is typical for the fifth term to vary: see Yasna 9.27; 13.1. Most significantly, the fifth term varies elsewhere in the Mihr Yašt (10.18; 10.87) where it is a secular governing figure, as the other four are.↩︎

  20. The underlying Avestan passage must be corrupt, with zarathuštrō instead of the expected zarathuštrōtǝma. Cf. Bivar (1988), who argues that Zarduštrōtum in the Middle Persian five ranks passages is a Zoroastrian interpretation/substitution for an original fifth secular administrative title, xšathra-pati. This very term (in the form ’xšyšpt) appears to have been used in Sogdian Manichaean versions of these titles for the five sons of the Living Spirit (Humbach and Skjærvø 1983, 100). I owe both of these references to the Addenda et Corrigenda appended to Sundermann (1979b), in its reprint in Reck (2001, 811).↩︎

  21. E.g. the Čīdag andarz ī pōryōtkēsān Pandnāmag ī Zardušt and Dēnkard book 7. See Sundermann 1997a, 24, regarding his view that such literature follows and borrows from Manichaean precedents in some of its elements.↩︎

  22. See also the Turfan fragment T II D 175/U4 discussed in Skjærvø 1995b, 618-620; cf. Dēnkard 4.72.↩︎

  23. E.g. 1Ke 7.27-33 and 12.16-19; Hom 70.2-15; as well as an unpublished passage from the Chester Beatty Kephalaia.↩︎

  24. These ideas of post-mortem judgment, followed by reward and punishment, are found, for example, in the fifth century BCE in Pindar’s Second Olympian Ode, 58-68, and in Aristophanes’s Frogs, 452-477.↩︎

  25. For an introduction and translation, see Hennecke and Schneemelcher (1965, 755–98). Most versions contain an account of the miraculous discovery of the text in 388 CE; but this passage is lacking in the Coptic version and is moveable in other versions (at the end of the Syriac and the beginning of Greek and Latin versions), and may be a secondary addition to a text possibly already known to Origen in the first half of the third century (see Himmelfarb 1983, 16–19).↩︎