Introduction

European colonial empires created both religion and secularism (that is, the absence of religion). This is the conclusion of much of the recent work in religious studies. Here, creating religion and secularism refers to several different processes: grouping practices found among a single people as a ‘religion’; separating out a certain set from the range of practices and their practitioners that will be classified as a species in the genus of religion; separating those religious practices from other, that is, ‘secular’, aspects of adherents’ lives; and defining certain religions as ‘world religions’, that is, they seem to exist in a transcultural, transnational space. Empires engage in ‘religion-making’ (Dressler and Mandair Reference Dressler and Mandair2011) through a process that is both directed from above and dictated by administrative needs, and also generated from below, as different communities and their leaders engage the imperial apparatus and seek protection, immunities, and advantages through being classified as ‘religions’.

In the colonial empires formed in the post-Enlightenment second wave of European colonialism, the concept of the secular played a range of crucial roles in the process of religion-making.Footnote ¹ By separating the ‘religious’ and the ‘secular’, the European colonial state created a secular space in the lives of the colonized, where the colonial state's jurisdiction could not be legitimately challenged by local ‘religious’ authorities. By deploying the concept of secularity, the colonial apparatus also distinguished those aspects of the metropole's culture which could be appropriately imposed on the colonial subjects as universally applicable and those which were to be treated as ‘religious’ and hence irreducibly culture-bound. Finally, the colonial regime's own ability to distinguish secular and religious served as a key marker of the colonial state's greater rationality and universality, marked by the colonial administrator's ability to be self-conscious about his own culture, value commitments, and own relation to the transcendent. In a Hegelian sense, this critical distance marked the colonial administrator as a mature person from a mature culture. By contrast, the colonial subject, bound with religion, was still ‘asleep’ and immature.

Similarly, colonial subjects found that cooperating in the distinction between religion and secular could enable certain spheres and practices to be walled off from the interference of the colonial state.Footnote ² Embracing this distinction could also mediate crucial alliances between communities that were held to be separate by the colonial authorities as different ‘religions’. Those colonial subjects shaped and educated within the colonial regime could deploy the secular–religious divide both to enhance their own claim to leadership as ‘secular’ and hence able to operate in the modern, international sphere, yet also maintain their ability to claim an organic and natural (because ‘religious’) connection to their colonial constituency.

Contemporary research in the history of religions has heavily stressed the contingent European roots of this secular and religious distinction. In particular, the Protestant distinction of inward faith and outward works, in its Weberian reading, has been taken as the necessary grounds for the imperial distinction of religion (internal, holy, spiritual) and secularism (external, profane, corporal).Footnote ³ Yet the question remains: given the demonstrable administrative utility of the classification of populations into particular communities, and the inevitable negotiation between ruling and ruled populations of certain predictable boundaries of control and autonomy, was there anything analogous to the distinction of religion and secularism drawn in previous empires and, if so, how? While the particular cultural background of second-wave European colonial powers such as Britain, France, and Russia undoubtedly shaped the distinction of religious and secular, which all of them drew in their colonial territories (British India, Algeria, Indochina, Central Asia), do we find an analogy of the religion-making process in other, non-European empires? Have non-modern, non-European, and non-Christian empires found it useful to designate both a genus of ‘religion’ (presumed to be universal such that all of the empire's subjects should belong to one) and a set of species belonging to that generic category? And does this process of non-European imperial religion-making similarly create a concept analogous to ‘secularity’ and ‘secularism’?

The Mongol empire is a particularly advantageous site from which to investigate this question. The extent of its empire made the Mongols rulers over many communities that European colonial powers would later parcel out among ‘world religions’: Buddhism, Christianity, Islam, Daoism, Judaism, Confucianism, and Bon. The empire was also famous for its policy of religious ‘tolerance’ which involved some form of separation of spheres whereby adherence to a religion on the part of rulers was legitimate, but only if it did not influence governance. It is this imperial policy of ‘tolerance’ that seems to have been revived in a modified form in the policy of ‘peace with all’ (sulh-i kull), as Moin's framework article in this special issue argues. Indeed, this might well be the case because in Islamic areas under Mongol control, the distinction of the ĵasaq or töre, on the one hand, and the sharī‘ah, on the other, created a seeming analogue of secular and religious law.Footnote ⁴ In all of these ways, the Mongol empire seems to present an example of another wave of ‘secularism and religion-making’ that offers a fruitful field for comparison with the more familiar wave that began in the eighteenth and nineteenth centuries.

The clerical service estates

Despite the vast gap in the quality of documentation—no archives of the Mongol empire have survived and most Mongol-language private writings have been lost as well—there are many prima facie parallels between the religion-making of the Mongol empire and that of the European colonial empires. To take the largest of these empires—the British—as an example, both the Mongol and the British empires conquered vast, culturally diverse territories in which they engaged in extensive mapping and census activities. In both empires, the census activities recognized, in some sense, specific religious communities, which are named in list form, evidencing the imperial understanding of them as comparable categories. Both empires saw a number of conflicts over the specific boundaries of these communities, in which particular clerical figures sought to use the census positions to advantage their own community and disadvantage others. This classification of religions also came with an implicit theory of what ‘religious’ behaviour was—in the Mongol case, it involved prayer to the supreme god, with the aim of securing benefits, particularly long life (Atwood Reference Atwood2004).

The Mongol rulers, even if members of one or another of the religions so classified, held themselves to a rule of impartiality, refusing to show bias to co-religionists of that same religion. To this extent there appears to have been something analogous to secularity or a space where membership of a particular religion would not be allowed to influence public behaviour.Footnote ⁵ This space included a number of technical fields that were favoured by the Mongol rulers: ‘economics, accountancy, for “balancing the budget” and the estimates of revenues and expenditures, medicine to keep bodies and constitutions (in health), and astrology’ (Kolbas Reference Kolbas2006, 313).

Yet examined more closely, there are as many differences as there are similarities. First of all, the Mongol imperial records hardly record lay communities at all. The lists of religions are recorded not by the abstract name of particular religions, but by the collective names of their clergy: toyit (Buddhist monks), erke’üt (Christian priests), sinšing-üt (Daoist priests), and dašmat (Muslim clergy). Terms for lay believers are much rarer: for example, only occasionally mentioned are Musulman for Muslims or Kiristan irgen, literally ‘Christian commoners’, for Christendom.Footnote ⁶

This rarity of terms for lay members of the community is explained by the context in which these lists of different clergy appear and by the Mongol registration practices. Ordinary subjects among the conquered peoples were, as far as is known, never listed by religion. Instead, they were listed by occupational categories, which hereditarily determined the specific duties they owed to the imperial government. These include civilian households, military households, post-road households, as well clerical and professional households. Thus legally, at least, those outside the ranks of the clerical registration had no recognized religious status. However, many lay folk were affiliated with religious establishments as menial labourers on estates belonging to monasteries, churches, temples, lodges, and shrines.Footnote ⁷

It is an open question as to what degree the Mongol understanding of different religious categories was assimilated by the lay adherents of the clergy. In Western Eurasia, the categories of clergy the Mongols employed, particularly Christian and Muslim, corresponded to pre-existing religious communities with relatively strong boundaries. There the main effect of Mongol rule appears to have been to throw Christians of various communions into a sometimes welcome, and sometimes not, new familiarity.Footnote ⁸ In China, however, where distinctions between religions among the laity had previously been quite ambiguous, Mongol rule, with its broad exemptions given to trusted clergy, heightened clerical involvement in secular affairs enormously. One can also find evidence in the social history of the time that lay populations in North China increasingly defined themselves and their communities in terms of Daoist or Buddhist affiliations.Footnote ⁹ Mongol rule also separately recognized the clergy of three new Buddhist-based religions which under the previous regimes had been subject to periodic persecution as heretical sects. The Dhūtaist community in North China and the ‘White Lotus’ and ‘White Cloud’ communities flourished under Mongol rule and reached an apex of self-awareness as separate religions only to disappear in the bloody fall of the Mongol empire and the renewal of persecution under the Ming that followed.Footnote ¹⁰

Here the closest parallel is with the estate (soslóvie) paradigm in Russian history (Freeze Reference Freeze1986). This paradigm combined broad occupational divisions (nobles, clergy, peasants, and merchants) with groups that shaped modern-day ethnicities such as among the Muslim peoples of the Russian empire (Frank Reference Frank, Zimonyi and Karatay2016). An estate was a

The negative definition of an estate category exactly fits the Mongol categories for clergy, physicians, and ortaq (‘partner’) merchants (that is, merchants trading with capital supplied from the treasury of the state or the Mongol ruling class), all of which are defined in surviving documents precisely by being not subject to some or all of the duties expected of common residents.Footnote ¹¹ Yet the basis for the estate classification in the expanding Russian empire was quite different from that of the Mongol empire. In its pre-Petrine form, among the Volga peoples and Siberia, as Allen Frank (Reference Frank, Zimonyi and Karatay2016, 147, cf. 142 n. 8) writes, ‘The most enduring and common types of the estate organization among Muslims were community-based and community-derived. These often coincided with the boundaries of religious communities and were based on ancestral affiliation.’ Later on in Central Asia, however, they used ‘broad socially undifferentiated colonial legal categories’ (Frank Reference Frank, Zimonyi and Karatay2016, 142).

Either way, however, Russian estate practice differed from Mongol estate practice, which used finely graded occupational categories precisely among the imperial subjects, while the ruling population of the Mongols appears to have been given much less specific differentiation in census documents.Footnote ¹² The only known categories applied to the Mongols themselves seem to be those of vertical status, separating out from the common mass the higher commanders (captains of a thousand and above), the imperial family and in-laws, and those with rights of darqan (tax exemption) for services to the khan.Footnote ¹³ One may say that the service given by the Mongol population was always the same—taxation, corvee, and military service; the only question for the head of a Mongol household was to establish whether or not his family was liable to provide those services or if they were among the fortunate few of the rank not to provide them, but to receive them.

This absence of occupationally divided service estate categories within the Mongolian population explains why people who could be seen as the Mongolian clergy—the bö’e (shamans) and the iduġan (shamanesses)—are never listed alongside the other clergy in tax exemption documents. As far as we know, the indigenous clergy were not distinguished from other Mongolian members of the population. Thus, legally speaking, the concept of clergy, which was viewed as a subset of the category of professionals, did not include those who made a living from practising the Mongol religion. This is the exact opposite of the case in Imperial Russia, where it was the Russians’ own Orthodox clergy which formed the paradigmatic clerical estate (dukhóvnoe soslóvie), which was extended only partially and with reservations to the non-Russian clergy (Werth Reference Werth2010).

While clergy are most commonly listed separately, they are sometimes also listed next to physicians, yin-yang specialists, and diviners.Footnote ¹⁴ These other categories of what we can call the ‘technical professions’ also received certain tax exemptions like the clergy as well as legal autonomy; lawsuits involving physicians, like those involving Buddhist, Christian, Daoist, or Muslim clergy or other separately numbered household categories, would be judged by Joint Courts (yuehui 約會) staffed by representatives of the community in question (Cho Reference Cho2014; Shinno Reference Shinno2016, 30–32, 51–52). Yet, while the tax exemptions and legal autonomy for physicians have significant parallels to those granted to the clergy, they were less extensive and not debated in the same contexts in which those of clergy were discussed. In other words, the Mongol empire clearly did have an implicit category that united the four main groups of clergy and others that would later receive exemption, such as Jewish rabbis, Tibetan Bon priests, and certain Buddhist and Daoist splinter groups. Yet this category of clergy was seen as a sub-set of those who delivered elite, professional services to the empire, not as a category of belief or mass membership—although untrained persons could be included within clerical establishments as subjects under the protection and administration of a temple, shrine, or monastery.

Precisely what united the service estate of Christian, Buddhist, Muslim, or Daoist clergy into a single genus appears from the exemption decrees they received, where their freedom from service requirements is explicitly said to have been given in return for prayers to Heaven for the long life of the khan (Atwood Reference Atwood2004). This type of prayer was not something that ordinary Mongols were considered to be able to do; as John of Plano Carpini said, ‘They believe in one God . . . they do not, however, worship Him with prayers or praises or any kind of ceremony.’Footnote ¹⁵ To this one may add that these four religions had received written scriptures from Heaven. As Möngke Qa'an said to William of Rubruck, ‘We Mo'als […] believe that there is only one God, through whom we have life and through whom we die and towards him we direct our hearts . . . To you God has given the Scriptures and you Christians do not observe them . . .’Footnote ¹⁶

In this ‘confession of faith’ Möngke Qa'an undoubtedly used möngke tenggeri for God and nom for scriptures. Taking this and the epigraphical sources together, one may define the common features of the clerical service estate as those who have received scriptures from Heaven, who have the specialized skill of praying effectually to Heaven, and who can perform this action on behalf of others. This Mongol analogue to religion might better be termed ‘clerical service estates’. In this sense, alongside clergy are those who have certain technical skills that do not involve praying to Heaven, but which can improve and lengthen life: these include physicians, diviners, and Confucians.Footnote ¹⁷ These we could call ‘professional service estates’.

Just as the definition of religion in European colonial empires carried with it an understanding of the secular, so too did the definition of a clerical service estate include an explanation of another category, one analogous to ‘secular’, although certainly not the same. Obviously, anyone who did not belong to the clerical service estate might be seen in some sense as ‘secular’. But as I will show, the clerical service estate did not exhaust the concepts of religion in the Mongol empire, and it would thus be a mistake to see its negation as the only sense in which ‘secularity’ existed in the Mongol empire.

Nom: Customs of communities

As I mentioned above, while terms for religious affiliation outside of the clerical service estate are rare, they are not unknown. Terms such as Musulman ‘Muslim’, Juqud ‘Jewish’, and Kiristan ‘Christian’, while not common, are attested to in ways that make it clear they are referring not to clergy but to broad masses of the subject people. But at the same time, these terms for what we might think of as more purely religious communities (that is, ones that are not connected with service estates), on the other hand, show a pattern of usage that indicates a meaning that is not distinguished from what we might call ethnicity.

This ambiguity is centred on the use of ‘nom’, which is the principal Mongolian term used in the Mongol empire for what we would call ‘religion’ in the abstract. Another term, šaśin, is the modern Mongolian term that translates as ‘religion’. This word, however, is derived from the Sanskrit śāsana and before the eighteenth century it was only used for Buddhism and in Buddhist contexts (Kollmar-Paulenz Reference Kollmar-Paulenz2012, 95–100). Before the modern era, one could speak of the Buddha's šaśin, but not a šaśin of shamans or Christians or others.

Where multiple norms or religions are in question, the term nom was used, borrowed from Uyghur and derived eventually from Greek nomos (‘law’). In Uyghur, nom had come to be used to translate Sanskrit dharma and Tibetan chos, and had also acquired the sense of ‘book’. Nom is thus used for religion; in the 1290 letter of the Il-Qan Arġun to Pope Nicholas IV, Christianity as a teaching is referred to as Misiqa-yin nom or the ‘nom of the Messiah’.Footnote ¹⁸ Nom is also used to distinguish, for example, the monastic way of life or the nom of the Buddha (burqan-u nom), as opposed to the householder's way of life or the nom of people, or the nom of samsara (kümün-ü nom or sansar-un nom).Footnote ¹⁹

Nom thus appears to refer not to distinct religions, but to particular ways of life, such that householders and the monks they support can be understood to be different nom. Not having adopted the European distinction of an interiorized religious realm and an externalized secular realm, sources in Mongolian, and reflecting the Mongolian point of view, do not draw a sharp distinction between those customs that are associated with religion and those that relate to customs and habits often seen as ethnic. This lack of a distinction was very frustrating to missionaries like William of Rubruck. Thus, while the prince Sartaq was baptized, William of Rubruck notes that he did not wish to be called a ‘Christian’, quoting a Mongol saying to him: ‘Do not say our master is a Christian. He is not a Christian; he is a Mo'al.’ William comments unhappily that ‘they regard the term Christendom as the name of a people’.Footnote ²⁰

To a certain degree, the objection may have turned on the fact that such terms for religious communities shared a certain subject-class connotation with those for the clerical service estate. To call Sartaq a Kiristan irgen or ‘Christian commoner’ or ‘Christian subject’ would have been obviously inappropriate, since he was a member of the Golden Lineage (Altan uruq) and hence far from common. Even less appropriate would have been calling him Erke’ün, which would have meant calling him a member of the service category of Christian clergy—as a Mongol, he could not be a member of that service category, even if he was a baptized believer. Similarly, when Möngke Qa'an got the (false) impression that William of Rubruck had called him a toyin (Buddhist), he was quite upset (Jackson and Morgan Reference Jackson and Morgan2009, 236)—not just because he was not, in fact, a Buddhist, but because what toyin meant was not someone who had taken refuge in the Three Jewels, but one who belonged to the service estate of Buddhist clergy.

But the linkage of ethnicity and religion was deeper than that. First of all, the Mongol habit of moving ethnic groups around and employing them as an intermediate classFootnote ²¹ meant that those from one region created immigrant minorities in another that were often denominated by terms that flowed seamlessly between ethnic and religious meanings. Thus, in the eastern part of the empire the term Uyghur ended up acquiring the sense of ‘Buddhist’, such that a person of known Mongolian background could be called a Uyghur just for adhering to Buddhism. Similarly, in the eastern part of the empire, the term Sarta'ul, translated into Chinese as Huihui, which in legal usage always meant ‘Westerner’ without regard to religion, would end up meaning Muslim. And in sociological reality, both the Muslim and Christian Sarta'ul in the Yuan dynasty did their best to secure clerical privileges for all of their adult male members.Footnote ²² As a result, ethnically based legal categories like semuren or ‘various peoples’ used in China for those moved by the Mongol empire (whether by force or incentives) and occupationally based categories like the clerical service estate tended to merge and be seen as a single body.

The results of this merger can be seen in Marco Polo's description of Mongolia's ‘religion’. In the chapter titled ‘Here Is Described the God of the Tartars and Their Law’, Marco Polo describes not just the Mongols’ practice of worshipping the household deity Načiġai,Footnote ²³ but also their clothing, obedience to their lords, and their military organization. He then writes:

After this caveat, he adds a description of the Mongol judicial system. In other words, the chapter title ‘the God of the Tartars and Their Law’; the phrase quoted above about the ‘habits and customs of the true Tartars’, including their military system, clothing, and national character; and their cult of Načiġai all seem to be referring to roughly the same thing.

At this level, the distinction of religious and secular appears particularly inappropriate for the Mongol empire. At the heart of the European definition of religion is the idea that it is not the same thing as race, ethnicity, or nationality, and that while these things pertain to the body and outward action, religion pertains to the soul and inward belief. (As noted in the introduction, we see here we see the Protestant origins of the post-Enlightenment European colonial religion-making.) But, Mongol nom, in this sense, appears to be squarely on the ‘ethnic’ side of this line, even if it also has reference to gods like Načiġai and founders like the Buddha and the Messiah.

Ritual activity versus moral behaviour

So up to this point one would have to say that the analogy between religion-making in the Mongol empire and the European colonial powers has not been very illuminating, or if so, only in a negative way. Legally, the most salient classification of religion did not include lay adherents at all, but only clergy, and embedded clergy within a larger set of technical professionals. It thus has only very limited comparisons with such concepts found in the European colonial empires, where the distinction of, for example, Hindu and Muslim was vastly more salient than that of ordinary Hindu versus sādhu, or ordinary Muslim versus imams and Sufi pīrs. And the concept that includes lay folk—that of nom—shows such slippage between religious and ethnic categories, and between customs that would in post-colonial contexts be called religious and those that would be called secular, that it seems rather to illustrate the absence of ‘religious’ and ‘secular’ in the Mongol empire.

But an analysis of two letters of Mongol khans to the popes shows that the clerical service estates and the idea of nom as the ensemble of customs and habits is not the whole story. One of these letters is the famous one of 1246 from Güyük Qa'an to Pope Innocent IV, while the other is the less famous, but equally (if not more) revealing, letter from the Il-Qan Arġun written in 1290 to Pope Nicholas IV. I will discuss this later letter first.

It was written in response to Pope Nicholas's charge to the Il-Qan that he ought to be baptized and become a Christian. In response to this challenge, Arġun replies first that ‘You are right to say that I should receive silam [baptism]. We descendants of Chinggis Qan say that, if our common Mongols are willing, whether they receiving silam [baptism] or not, Eternal Heaven alone decides.’Footnote ²⁵

Here, entering or not entering into a ritual community is stated to be one of those things that Eternal Heaven ‘decides’ (mede). It is thus implicitly placed alongside victory in war and the fate of parolees placed in the front lines, matters which the khans formally and institutionally refused to take as their own responsibility. Within the Mongols’ peculiar vocabulary of statecraft, this is the same as expressing the idea that entry into a nom, whether by silam (baptism) or suna (circumcision) or some other practice, is a matter in which the Chinggisid state abjures competence.Footnote ²⁶ Then he politely assures the Pope that he was one of those few baptized people who had sincere and pure thoughts, and that neither he nor his Christian religion (Misiqa-yin nom) would violate the legitimate imperial command (ĵrliq). But he observed that those who violate the commands and commit crimes are more common than such pure persons. He then concludes with a criticism of the Pope's attitude: ‘Now since I have not receiving silam [baptism], you are also unhappy in your mind about me. If [one was/you were] to only pray to Eternal Heaven and think rightly, would not that be like receiving silam?’Footnote ²⁷ It should be noted that the last sentence has no subject, so that it is a little ambiguous as to whether Arġun is saying that if he himself were to simply pray to Heaven and think rightly, then it will be just as if he were baptized, or that if the Pope would stop worrying about the Mongol khans’ baptism (which after all Heaven alone would decide) and think rightly, then he would be doing a work as good as his baptism.

Regardless of the interpretation, here the khan Arġun is clearly opposing two methods of achieving merit or a good conscience—what Möngke Qa'an called in the version translated for William of Rubruck ‘an upright heart’ towards Heaven. The first is through ritual action and membership of a specific community headed by one of the clerical service estates, for example, the Christian Church. The second way is through thinking right thoughts and obeying the ĵrliq, by which must be understood the ensemble of commands and powers ordained through Chinggis Qan and his legitimate successors. It is tempting to read Arġun's comments as entirely ironic, as if he was saying that empty rituals and religious group membership are obviously bogus and false methods of being a good person. Yet it is important to remember that the receiving of baptism is something that Heaven decides, even among the Mongols, and that those who dispense, for example, baptism—the Christian clergy—are among the most honoured members of the Mongol imperial elite, alongside other members of the clerical service estate, each disposing of their own rituals of entry into their particular nom. In other words, only an interpretation of Arġun's letter as accepting the validity of both ritual activity and moral behaviour would make sense of known Mongol statecraft. Nicholas's fault is not that he is peddling baptism, rather than simply acknowledging obedience to the ĵrliq, but that he is doing the former in specific contexts in which the latter is more appropriate.

This same analysis makes sense of Güyük's letter to Pope Innocent almost 50 years earlier. Here too, the Mongol khan opposes the demand for moral behaviour, understood as obedience to Eternal Heaven, to the Pope's call to baptism. The various Magyars and others in Christendom (probably Kiristan irgen here as well) had killed envoys and rejected the call to surrender and so were being justly chastised by the khan. Güyük paints a contrasting picture of being a ‘trembling Christian’ who ‘worships God’ and is ‘an ascetic’, and implicitly contrasts that with his own situation as one who ruled ‘from the rising of the sun to its setting’ precisely because he has followed the commands of Eternal Heaven, that is, God.Footnote ²⁸ Once again, one might be inclined to think that Güyük has no use or role for ‘trembling’ clergymen and ‘ascetics’—except that we know very well that those of that description, especially Christians and Buddhists, received both full tax exemptions as well a great measure of the khan's favour. Once again, Pope Innocent's fault is not that he is a clergyman and trembling ascetic who worships God, but that he seeks to extend to those following the other ethic—that of moral behaviour and upright heart towards Heaven—the demands of ritual and communal justification that is proper only in its particular limited sphere.

In-between these two letters comes Möngke's famous confession of faith. In this statement as well, we see the assertion that members of the clerical service estates mostly do not keep the rules of their scriptures, but that the Mongols do indeed obey Heaven. Clearly, the idea is that the rules preached by clerical service estates were incapable of broad application: that it is impossible for the masses to actually follow the various nom (scriptures) peddled by the clergy is a fixed and acknowledged part of the Mongol ideology of rule. What is new, however, is the designation of the Mongolian shamans as the medium through which Heaven speaks to the obedient Mongols. Möngke Qa'an said to William of Rubruck, ‘God has given the Scriptures and you [Christians] do not observe them; whereas to us he has given soothsayers and we do as they tell us and live in peace’ (Jackson and Morgan Reference Jackson and Morgan2009, 237). While this statement clearly envisions some kind of comparability between Mongol shamans and the clerical service estates, it is notable that Möngke does not place Mongolian clergy in an analogous position with the Christian and other clergy, but rather with their scriptures. The focus on the scriptures is significant as all of the exempt religions were well-known to have written scriptures which played a prominent role in their cult.

There is one final example of a contrast between what may be characterized as a confident, ethical type of justification before Heaven and an insecure sectarian type of justification. After describing Qubilai Qa'an's acts of charity to the poor, Marco Polo adds (section 104) that ‘the Tartars according to their first custom, before they knew the idol law, did no alms’. Rather, they would drive away the poor, saying ‘Go with the bad year which God give thee; for if he had loved thee as he loves me he would have done thee some good.’ But because the ‘wise men of the idolators’, and in particular the baqśi or Buddhist monks, suggested to the great khan that ‘the provision for the poor was a good work for him, and that their idols would greatly rejoice at it, he thereupon provides like this for the poor’.Footnote ²⁹ This passage, found only as an additional passage in Ramusio's text, was appended to Polo's description of Mongol ‘law’ (in Mongolian, nom) and its changes due to ‘idolators’ and ‘Saracens’ (that is, Buddhists and Muslims). The original nom of prosperous Mongols was based on confidence in God/Heaven that their prosperity was due to right living; under the influence of the ‘idol law’ (or Buddhist nom), they sought to leverage a now ethically ambiguous prosperity into divine approval through alms-giving.

Yet the confident assertion that success is adequate proof of moral behaviour in the eyes of Eternal Heaven, to which Polo's God is certainly referring, links this statement closely to the arguments Güyük made against Pope Innocent. Once again pleasing Heaven by moral behaviour, which is then directly rewarded by power and prosperity, is contrasted with sectarian ways of justification. To the Pope, this meant silam or baptism, but to the baqśis or Buddhist monks, it is alms to the poor which please their ‘idols’—not, be it noted, Eternal Heaven. In other words, what we see here is another example of a broader opposition between a universal ethical route to ‘an upright heart’ and a sectarian ritual route to that same destination. Both were important parts of the Mongol system, but the two were not equal; rather, the universal and ethical pleasing of Eternal Heaven stood above, in a position to control and limit—but not abolish—the sectarian and ritual practices of the clerical service estate.

This opposition of sectarian ritual versus universal ethics went far beyond the limits of the clerical service estate. Not just the conquered people, but Mongols themselves were to be allowed to freely enter sectarian rituals as Heaven determined. Yet if they did, they were still subject to the demand that they obey the decree of Chinggis Qan and follow only the universal heavenly ordained ĵrliq in exercising public authority. The Persian historian Juvainī, in Mongol service, summarized his rulers’ practice in this way:

The Mamluk historian Yunini (d. circa 1325) described the same policy as followed by Ket-Buqa, the Mongol general defeated by the Mamluks at ‘Ayn Jalut: ‘[Ket-Buqa] tended towards Christianity, but did not show an inclination towards the Christians, due to his belief in the laws of the Yāsā of Chinggis Khān. All of the believers in religions were equal in his eyes and this is from the laws of the Yāsā.’Footnote ³¹

This distinction between the impartial ĵasaq (Persian yāsā) that the Mongols followed and the narrower sects was inherently hierarchical; it was only to the ruling Mongols, not the conquered population at large, to whom the universalist ethic of obedience to the ĵrliq of Chinggis Qan was open. Precisely because other peoples were incapable of following universal ethics, the Mongols became their rulers. We have seen how both Möngke Qa'an and his grand-nephew Arġun charged the vast majority of Christians with not following their own law. Similarly Chinggis Qan famously claimed in the pulpit of Bukhārā that he was the punishment of God sent to discipline the Muslims of Central Asia for the treachery of their sultan.Footnote ³² Only a few years later, the Mongol commander Ögelen Čerbi explained to his Muslim interlocutor how ‘You TajiksFootnote ³³ do such things, and tell lies. A [Mongol], were a thousand lives at stake, would choose being killed, but would not speak false; but false speaking is your occupation; and, on account of such things, it is that Almighty God hath sent a calamity like us upon you.’Footnote ³⁴ Although it is not explicitly drawn out in the sources, one may hypothesize that it was precisely the inferior moral quality of Christians, Muslims, and the people of China that made ritual methods of appeasing God necessary, since they were incapable of living without lies and other moral faults. The Mongols, however, could do so and thus in the imperial ideology deserved to rule.