by Julian Gee, Harvard

I have recently been reading manuscripts of instruction manuals (risala) for rainmakers (yadači) written in Turki and Persian in East Turkestan/Xinjiang during the late nineteenth and early twentieth centuries.[i] The manuscripts offer a fascinating snapshot of magico-religious practices, containing as they do much that is common to broader Islamo-Turkic traditions of rainmaking, such as the well-known use of yada taš or “rain stones” (often bezoars/gastroliths) during the ceremonies, but also several features seemingly unique to East Turkestan.

The tradition among Turks (and other Inner Asian peoples such as Mongols) of rainmakers calling down rain by immersing “rain stones” of different colors in water is of course both well attested from the Middle Ages onwards and also well known to modern scholars. The earliest definition of the term yad is probably to be found in the eleventh-century dictionary of Mahmud Kashgari, while the practice itself is also mentioned in the Zayn al-akhbir by Mahmud’s contemporary Gardizi (the latter’s description is possibly based upon earlier sources).[i] While most likely a cultural institution far preceding the propagation of Islam among the Turks, and indeed attested among peoples and tribes outside the realm of Islam such as Buryats, the practice appears to have been quickly assimilated within Islamic historiographical tradition. Thus, Gardizi attributes the origin of the tradition to a stone used by Japheth, the son of Nuh (Noah), to summon rain and heal the sick. Claims of such an elevated prophetic origin for the practice may not always have been accepted in East Turkestan (to say nothing of the wider Islamic world). For instance, sixteenth-century Kashgari historian Muhammad Haidar Dughlat, eponymizes the practice’s origin to a magician named “Jadah,” rather than Japheth.[ii] However, the Islamic nature of the practice was apparently accepted by many in East Turkestan at the time of writing of the risala manuscripts, being found for instance in the histories by Mulla Musa Sayrami (1836-1917), who describes the original rain stone as having been given by Jibra’il/Gabriel to Japheth, who used it alongside an incantation in Arabic, a claim consistent with the risala manuscripts’ directions to the practitioner to pronounce certain Arabic prayers (both Quranic and extra-Quranic).[iii]

Apart from the use of the rain stone, the rainmaking risala also contain numerous invocations, such entreaties in fact constituting the majority of the lines in the risala. While most of the figures whom the risala direct the rainmaker to invoke consist of individuals well-known from Islamic traditions, such as the Prophets Muhammad, Nuh, Ilyās (Elijah) and al-Khiḍr, not everyone called upon is so easily identified. In particular, the manuscripts refer to an entity named “Qara Khawaja” قراخواجه, literally the “Black Lord” (sometimes also granted the honorific ḥazrat حضرت).

The nature and identity of this Qara Khawaja remains a vexing problem for me, and I have yet to find a satisfactory answer. While one text’s attribution to Qara Khawaja of “spirit” indicates that the former is most likely an individual, little else can be directly inferred from the texts. Indeed, Russian ethnographer Sergei Malov (1880-1957), present in East Turkestan from 1913 to 1915, and the only scholar I have found to have authored a work on rain prayers from East Turkestan, although mentioning Qara Khawajain his summary of a rainmaking risala fails to provide any explanation of the figure in an article otherwise filled with copious footnotes, perhaps implying even the Turkestani rainmaker he interviewed before writing his article could not identify the figure.[iv] Similarly, having examined secondary literature on Iranian, Turkish, and Uzbek rain prayers, I have failed to locate mention of Qara Khawaja or an equivalent figure in any texts from outside East Turkestan, implying that Qara Khawaja, whoever he may be, may well be unique to the Tarim Basin.

Lacking success in obtaining clear evidence about the identity and origins of Qara Khawaja, several possibilities have suggested themselves to me, which I outline below. I am by no means committed to any of these speculations and offer them as potentially interesting conjectures.

Perhaps the most straightforward supposition for the identity of Qara Khawaja is that he is an unnamed member of either the Qarakhanid Dynasty (840-1212), whose monarch Satuq Bughra Khan (d. 955) has always been one of the most popular Islamic saints in East Turkestan, or alternatively one of the Makhdumzada Khawajas. The latter were descendants of the Naqshbandi Sufi Shaykh Jalāl ad-Dīn al-Kāsānī (d. 1542), who from the late sixteenth century increasingly dominated the politics of the Tarim Basin. One of their lineages formed the so-called “Black Mountain” faction (opposed to their relatives in the “White Mountain” faction). However, I have yet to encounter an East Turkestani text in which an individual Makhdumzada is given the title Qara Khawaja or indeed is not personally identified in such a way that his identity is clear, while the “Black Mountain” terminology itself, largely limited to the eighteenth century, seems too ephemeral to have left a lasting mark on separate rain prayers. Similarly, one cannot entirely dismiss the possibility that Qara Khawaja is simply an otherwise unknown local Islamic saint. While none of the works on Uyghur Islamic saints or shrines (mazar) that I have consulted mention such a figure, such would not necessarily be decisive when dealing with a region which (as Muhammad Haidar Dughlat remarked) contained many otherwise unknown and spurious alleged saints and tombs.[i]

Turning to hypotheses of pre-Islamic survival or borrowings from non-Islamic cultures, one alternative possibility that cannot perhaps be totally discounted, albeit in my view remaining highly unlikely, is that the rain prayers’ invocation of Qara Khawaja reflects the survival of a borrowing by pre-Islamic Uyghurs of a Chinese “black” deity (the colour black being associated with water in traditional Chinese cosmology). Such deities could include either Heidi 黑帝 (the Black Thearch) or Heilong Dawang 黑龍大王 (the Great Black Dragon King), the latter of whom is often invoked in rain rituals in Shanxi and other places in north-west China. Substantial contact between the inhabitants of the Tarim Basin and north-west China means that transmission of such a cult is not impossible, although the absence of any other features characteristic of Chinese rain prayers in the text, to me argues against such a possibility. Certainly, nineteenth-century Chinese observers of Turkestani rain prayers, while noting similarities between Turki Muslim and Mongol methods of rain invocation, seem to have seen no similarities with their own rainmaking practices.

If not Chinese, then could Qara Khawaja reflect Mongol or Tibetan influence? Certainly, Mongol rainmaking practices, in particular their use of rain stones, are described by numerous pre-twentieth-century observers as resembling those of the Muslims of East Turkestan. Among the Mongols in recent centuries, “shamanic” religious practices have been divided into “white” and “black” just as their “ninety-nine gods” are also classified by Mongols as either “white” or “black gods.”[ii] Similarly, one of the two risala attributes its rainmaking traditions in part to the “Tanghut” تانغوت, possibly meaning either the ruling ethnicity of the state of Xi Xia (ca. 1038-1227), which is known to have been under strong Tibetan cultural influence, or possibly Tibet itself. It is thus not impossible to imagine that at some point, Mongol or Tibetan invocation of a “black” deity (such as Mahakala) may have found itself into rain rituals in the Tarim Basin, especially if adopted as part of a larger rainmaking ritual complex.

Casting the net further back in time, another (albeit more remote) possibility is that Qara Khawaja could be a remnant of a pre-Turkic, Iranian cult. In particular, based on onomastic and other evidence, historian of medieval China, Chen Sanping, argues for the existence during the fourth and fifth centuries CE of a cult transmitted to China from Central Asia, worshipping a “Black God” (likely the Avestan Sāma), linked by some scholars to both the black-faced Iranian folk-character Haji Firouz and the Persian epic-hero Siyavash (himself long connected in Persian epic and local traditions with the city of Khotan in East Turkestan), and also associated with the harvest.[i] Could the rainmaking texts’ Qara Khawajabe a survival of this Iranian cult? Such an hypothesis could perhaps be strengthened by the views of some scholars that Turkic rainmaking traditions are ultimately of Iranian origin, although conversely, while it is not impossible to imagine invocations to such an obscure figure surviving Turkification and Islamization of the Tarim Basin, such a survival would be quite remarkable.

Ultimately, in the absence of further evidence, I suspect that the issue of Qara Khawaja’s identity will remain unresolved. However, I find the question interesting for raising broader methodological issues about how far back into the pre-Islamic past (“Shamanic” or otherwise) one is justified in delving when seeking to explain obscure Central Asian religious phenomena, which at least on the surface belong entirely to the mental world of Islam, and indeed when such phenomena, unable to be explained by reference to Quran or Hadith, should still be called “Islamic.”


[i] Sanping Chen and Victor H. Mair, “A “Black Cult” in Early Medieval China: Iranian-Zoroastrian Influence in the Northern Dynasties,” Journal of the Royal Asiatic Society 27, no. 2 (2017): 201-24.

[i] For instance, Rahileh Davut راھىلە داۋۇت, UyghurMazarliri ئۇيغۇر مازارلىرى (Urumqi: Xinjiang renmin chubanshe, 2001).

[ii] Yönsiyebü Rinchen, “White, Black and Yellow Shamans Among the Mongols,” URAM, 4, no. 2 (1981): 94-102.

[i] John Andrew Boyle, “Turkish and Mongol Shamanism in the Middle Ages,” Folklore 83, no. 3 (1972): 177-93.

[ii] Muhammad Haidar Dughlat, The Tarikh-i-rashidi, trans. N. Elias (London: S. Low, Marston and Co., 1895), ch. 14, p. 33.

[iii]Tarikh-i amniyya, Jarring Prov. 478, 10A.

[iv] С.Е. Малов, Шаманский камень «яда» у Тюрков Западного Китая, Советская Этнография 1 (1947): 151-60, esp. 159.

[i] For an example of such rainmaking risala, see the Jarring Collection’s Jarring Prov. 53.