Turks and Iranians: An historical sketch Peter B. Golden The eleventh century Turkic lexicographer Mahmūd al-K šγarī, summing up the long-standing tradition of Turko-Iranian interaction, cites the Turkic saying: bašsız börk bolmas, tatsїz türk bolmas ‘without a head there can be no hat, without a Tat (Iranian) there can be no Turk’. While this saying implies some sense of Turkic dependency on the ‘Tat’, there was also a long tradition of wariness and even disdain. Punning on the Turkic word tat (‘rust that appears on a sword or other’), K šγarī quotes the expression: qїlїč tatїqsa iš yunčїr, er tatїqsa et tinčir ‘when rust overtakes a sword the condition (of the warrior) suffers, (just as) when a Turk assumes the morals of a Persian his flesh begins to stink’. K šγarī further comments that “this is coined to advise a person to be steadfast and to live among his own kind.” (Dankoff & Kelly 1982-1985, 1: 273 and repeated 2: 103).1 The original sense of the term tat seems to have been an “alien”, most probably one who occupied an inferior political status, which was then applied to the Iranian peoples that the Turks encountered in Central Asia and later in the Middle East (Clauson 1972: 449). In the fourteenth and fifteenth century Mamluk-Kipchak glossaries, now reflecting a different ethno-cultural setting, it continued to denote non-Turkic, in this case Arab, sedentary populations.2 According to K šγarī “a Persian who does not know any Turkic” was termed a somlin Tat (Dankoff & Kelly 1982-1985 1: 361).3 This implies that there were Iranians, perhaps quite sizable in numbers, who did know Turkic and bespeaks a certain familiarity of the two groups with each other’s languages and cultures. K šγarī clearly underscores this, remarking that those Turks who did not mix with the Persians of the cities had the “most elegant” speech, whereas those who were bilingual and mixed with the Iranian-speaking urbanfolk “have a certain 1 My translation differs slightly from that of Dankoff who renders it: “A Turk is never without a Persian (just as) a cap is never without a head.” 2 Cf. Caferoğlu (1931: 62, Arabic text) tat “al-fall h” (“peasant”) with the interlinear addition “al-ªarab wa«l-f risī” (the Arab and Persian). The entry for tatıqtı is defined as “s ra lisanuhu wa luġatuhu luġat al-fall h” [his language and idiom became that of a peasant (i.e. an Arab)]. The at-tuhfat az-zakiyya fı«l-luġat at-Turkiyya defines “hadarī” [villager] as sart wa tat (see Atalay 1945, fol 12b [=12a]). 3 Somlin was a general term for “anyone who does not know Turkic” < somli- “to talk unintelligibly, er somlidi “the man spoke in a non-Turkish language, one which only the speaker understood” (Dankoff & Kelly 1982-1985 2: 302). 2 Peter B. Golden slurring in their utterances – for example Soγdaq, Kenček and Arγu.”4 The people of the Karakhanid city, Balasaγun, “speak both Soghdian and Turkic. The same is true of the people of Tir z (Talas) and the people of Madīnat al-Bayd « (Isbīj b).” The whole of the Arγu country, which extended from Isbīj b to Balasaγun, was afflicted with this slurring of speech. In the villages around Kashgar, Kenčekī was spoken, “but in the main city [they speak] Kh q nī Turkic”. K šγarī did not consider the people of Khotan and Kenček “among the Turks, since they insert into the speech of the Turks what does not belong to it” (Dankoff & Kelly 1982-1985, 1: 84-85). The pronunciation of the Turkic acquired by these Iranians betrayed substratal elements as well as non-Turkic vocabulary. Turks, as we have seen, were not unaffected by this interaction. Thus, the Oghuz are cited as having adopted Persian words (e.g. ören ‘bad’ < Pers. vīr n ‘ruined’) and having “mixed with the Persians, they forgot many Turkic words and used Persian instead” (Dankoff & Kelly 1982-1985, 1: 115). The uneasy symbiosis of Turk and Iranian reaches deep into the past. In the period up to the mid-fourth century A.D., Iranian nomads had largely dominated the Central Eurasian / Inner Asian steppelands and Iranian agrarian and urban societies developed in regions contiguous to the steppe which subsequently were closely associated with the Turkic peoples. The areas that are today termed Turkistan were largely Iranian or Tokharian (in the eastern zone, present-day Eastern Turkistan / Sinkiang [Xinjiang]) in speech (see Mallory 1989: 48-63 and Litvinskij 1992). The question of early Indo-European contacts with speakers of Altaic languages, while undoubted, remains problematic.5 However, there is little doubt that Iranian nomads (the Scythians of the Greeks, the Saka of the Persians) penetrated Mongolia and South Siberia,6 where they had contact with Turkic peoples whose ancient habitat most probably lay in the adjacent territories extending to the Baikal region.7 Turkic tribes may have been among the subject groupings of the Scythians (Aalto 1971: 29). It seems likely that pastoral nomadism, whose probable origins are to be traced to the Western Eurasian steppes, was brought to the Turkic peoples by these eastern Iranians (Kiselev 1949: 315, 357; Grjaznov 1969: 219, 233; Xazanov 1984: 89ff). It is quite possible that religious and attendant concepts of royal ideology came to the Turkic peoples in this same fashion from Iranian or even earlier Indo-Iranian contacts (cf. Turkic yat / yada / yadu ‘rainstone’ < Avestan y tu, Sanskrit y tu ‘witchcraft’) 4 So daq is clearly some now Turkicized Soghdian grouping. Kenček is associated with the region around K šġar (see Bailey 1985: 54). K šγarī (Dankoff & Kelly 1982-1985, 3: 284) has a list of “Kenčekī” words used in the Kenček Turkic dialect which reflect the original Iranian tongue of the region, e.g. kenbe “a plant” (Clauson 1972: 727), rabčat “unpaid, forced labor” (Clauson 1972: 780) etc. 5 See discussion in Gamkrelidze & Ivanov (1984, 2: 500-501, 554, 579, 585-589, 626-627, 637, 639, 657-658, 660, 940). 6 See Rolle (1989) and Rudenko (1970). Claims have been put forward for the Turkic affiliations of the Scythians (see Miziev 1996, esp. pp. 212ff). 7 For various reconstructions, see Golden (1972: 124-126). Turks and Iranians: An historical sketch 3 (Aalto 1971: 31-32).8 Also of considerable antiquity may be terms such as Turkic sїra ‘beer’ < *Sanskrit sur , Avestan hur ‘kumys’?, Turkic t na ‘heifer’ < Avestan daēnav ‘female of large quadruped’, Sanskrit dhena ‘cow’, etc. (see Aalto 1971: 3033).9 The rise and fall of the Hsiung-nu state10 in Mongolia (ca. 176 B.C.-mid-second century A.D.) produced the westward movement of Iranian and perhaps Tokharian peoples at its inception and of Turkic peoples that had been subjects of the Hsiung-nu with its collapse. It also brought to an end the domination of Mongolia by Iranian tribes and marked the emergence of this region as a center of Altaic speech. By the mid-fourth century, jolted again by the formation of the probably proto-Mongolian Jou-jan Empire in Mongolia, we see these “Hunnic” peoples (the Hyōn / Xiyōn – Xwn) beginning to trouble the borders of Iran and subsequently Northwestern India. The ethno-linguistic identity of the Xiyōn is uncertain. They may well have been a confederation of the earlier Iranian nomads and Altaic-speaking elements (very possibly Proto-Mongolic and Turkic).11 Similarly unclear are the affiliations of the Hephthal dynasty that came to rule them by the mid-fifth century and whose name came to be the political designation of the peoples under their rule. In any event, the polities that took shape here used the Iranian Bactrian tongue as their official state language. By the 370s, these “Huns” had crossed the Volga and under their later leader, Attila (d. 453) entered European history and legend.12 In the aftermath of the European Hun collapse (454) and spurred on by disturbances in the Jou-jan realm, the movement of Turkic peoples into the hitherto cis- and trans-Volgan Iranian Alano-As territories increased. Ca. 463, several Oghuric tribal groupings, speaking 8 Not fully discussed in Clauson (1972: 883): yat. 9 Although neither of these terms is attested in early texts, both are found as loanwords in Hungarian sör ‘beer’ and tinó. The latter points to an Oghuric borrowing (see Benkő et al. 1967-1976, 3: 920-921), while the former is somewhat more problematic (see Benkő et al. 1967-1976: 580), but the initial s (š) would seem to point to Oghuric as well. Oghuric terms were borrowed by the Hungarians in the period prior to their occupation of presentday Hungary (ca. 895) as well as afterwards from local Bulgharic populations resident in Pannonia. Since Oghuric diverged from Common Turkic probably in the aftermath of the breakup of the Hsiung-nu state which sent Turkic elements westward, the OghuroBulgharic elements in Hungarian must be of considerable antiquity within Turkic. 10 Hsiung-nu ethno-linguistic affiliations remain the focus of much speculation. Bailey (1985, 7: 25-41) views them as Iranians. Kettic and Altaic affiliations have also been proferred, see discussion in Golden (1992: 57-59). 11 Frye (1966: 169-171) recognizes the name “Chionites”, etc. as a form of the ethnonym “Hun”, but was inclined to view those “Huns” on the borders of Iran as “one of the last of Iranian-speaking nomads mixed with Altaic speakers who were called Huns”. On the War (Avar, presumably Proto-Mongolic) and Hun elements of the Xiyōn and Hephthalite groupings and their migrations, see Czeglédy (1983). 12 On the Huns of Europe and the complexities of their Inner Asian connections, see Maenchen-Helfen (1973); Dąbrowski (1975: 27ff). 4 Peter B. Golden forms of Turkic distinct from Common Turkic (and surviving today only in Chuvash), entered the region and made contact with the Eastern Roman / Byzantine state (see the account in Priskos (Blockley 1981-1983: 344-345)). Here, they entered into close relations with the remaining Alano-As nomads, both groups contributing to the Saltovo-Majackaja culture which came to be associated with the later Khazar Kaganate (see Pletneva 1999). The tribes that came westward supplanted or assimilated over time the Iranian nomads of Central Asia. The land of “Tur n”, later associated with the Turkic steppe nomads, had earlier been the domain of Iranian pastoralists. There were several clusters of Iranian city-states that had formed in this region. Situated in the western zone of present-day Uzbekistan was Khw razm with a population speaking a distinct Northeast Iranian language. Related to it was the much more widespread Soghdian tongue associated with the urban centers of Bukh r , Samarqand and Č č (today Taškent) and diffused in colonies along the Silk Route extending into Northern China. To the southeast of them was the region of Tux rist n, with Bactrian and other Iranian tongues that had been associated with the Kušan Empire.13 These city states were each ruled by their own kings (e.g. the Khw razmš h). The political and commercial relations of the Turkic tribes that were in contact with these trading states would deepen in time, as did also their interaction with the As and other Iranian steppe peoples previously associated with the region and with the PontoCaspian steppe zone to its West. The origins of the Kök Türk state, founded in 552 with the overthrow of the Joujan and centered in the latter’s territories in Mongolia, was also associated with Iranian and Tokharian peoples. The ruling clan of the Türks, recorded as A-shih-na in the Chinese sources, before coming to the Altay, had lived in the Iranian Saka and Tokharian regions of East Turkistan. In the Soghdian Bugut inscription (580s), the earliest official inscription from the Türk state, gives their name as tr-«wkt «(«)šyn«s for Türküt Ašinas (Moriyasu & Ochir 1999: 123; Kljaštornyj 1973: 257; Kljaštornyj & Livšic 1972: 85). The name A-shih-na itself has been explained as Khotanese Saka s na ‘worthy noble’ or sseina – ššena ‘blue’ (cf. Turkic kök ‘blue’ probably used here as a geographical color referent designating the ‘East’ cf. the Kök Türk of the Köl Tegin Inscription, E3) (Tekin 1988: 8).14 Whether they were Iranian, Tokharian or Turkic is unclear, but clearly there were important non-Turkic ethno-cultural elements present. The names of the early kagans (Bumїn, Ištemi, Muqan, Taspar, Nivar / Ñevar) were not Turkic. The earliest surviving official monument from the 13 For an overview, see Frye (1975: 27ff). 14 See the most recent discussion in Kljaštorny (1994: 445-447), Kljaštornyj & Savinov (1994: 13-14). A derivation from the Tokharian *aršila (cf. the Tokharian title ršil ñči) as suggested by Beckwith (1987: 206-208), while a possible explanation for the Ά σίλα “the most ancient monarch of the Turks” noted in Menander, in light of the Bugut form, Ašinas, seems unlikely as the form behind the Chinese A-shih-na. Turks and Iranians: An historical sketch 5 Türk state, the Bugut inscription, as we have noted, is written in Soghdian. The titulature of the kaganate, which followed many of the patterns of its predecessors, the Hsiung-nu and Jou-jan, borrowed from a variety of sources: qatun ‘wife of the kagan’ < Soghdian xwt«yn [*xw ten] ‘wife of the ruler’, šad, a title below that of qa an (cf. Soghdian «ġšyδ, Saka šao, Middle Iranian š δ, cf. Pers. š h < xš yaθīya ‘king’ or Avestan xšaēta ‘chief’), beg ‘head of a clan, or tribe, a subordinate chief’ < Iranian *bag, Soghdian bġy. These may have come through Jou-jan mediation or directly. Some titles, such as Išbara < Sanskrit iśvara ‘prince, lord’ probably came via Tokharian,15 although Indian colonies did exist in East Turkistan and could have been a direct source (Litvinskij 1992: 77-114). Elements of imperial ideology also had Iranian connections. Fundamental to the power of the kagan, who was “heavenlike, heaven-conceived” (tengri-teg tengride bolmїš), was the notion of divine ordination for the throne and the granting of qut “heavenly good fortune” (tengri yarlїqadїn üčün özüm qutum bar üčün qa an olurtum “because heaven ordained, because I myself possessed heavenly good fortune, I became kagan”) (Tekin 1988: 2, 4, Köl Tegin inscription, S1,9). Qut, in its conceptualization was very close to the Iranian notion of xw rəna (Bombaci 1965: 284-291, 1966: 13-43). A particularly important role was played in the Türk state by the far-flung Soghdian trading colonies16 that stretched across the Silk Route from Inner Mongolia to the Crimea (cf. Sogdaia / Sugdaia, later Qıpčaq Sudaq, Rus’ Surož’).17 Centered in Soghdia proper in the Zeravšan valley and Ustruš na, with Bukh r and Samarqand as their major cities, the Soghdians and the Khw razmians to their west, formed a loose union of mercantile trading states under their own rulers and possessed considerable military power. The Chinese sources, in particular, make note of the che-chieh (Soghdian č«ġr, Pers. č kar ‘a servant’), special military forces attached to the elite houses. These, it has been suggested, served as the prototype for the later ġul m institution of military slavery in the ªAbb sid Caliphate that was largely staffed by Central Asian Turks. They were experienced long-distance merchants of whom the Chinese sources report that “they excel at commerce and love gain” (Chavannes 1900: 134, 136-137; Abaev 1958, 1: 286).18 As the Türks advanced westward, they crushed the Hepthalite state in 557 or 565 after having first established an alliance with S s nid Iran, the famed ¾usraw Anōšīrv n marrying a Türk princess (Pellat 1966-1979, 1: 307). Turkic conquests extending to the Crimea marked a second wave of Turkic-speaking peoples entering 15 See Clauson (1972: 257, 322-323, where he derives beg from Chinese po; 602-603, 866) and discussion in Golden (1992: 148); Aalto (1971: 34-35). 16 The history of the Soghdians has yet to be written. A brief overview of their society and culture can be found in Zeimal (1983, Chapter 6) and Frye (1966, Chaper 13). 17 See Haussig (1983). The Rus’ form shows Bulgharic intermediation, Su daq > *Suwdag > *Surog with Slav. g > ž’. 18 On the č kar-ġul m connection, see Beckwith (1984: 29-43); Frye (1984: 352-353). 6 Peter B. Golden the region and represented an important stage in the Turkicization of much of Central Asia, a process that would take centuries. The Türks established their control over the lands north of the Oxus (Amu Dary ), the region that the Arabs would later term M war « an-nahr (‘That which is beyond the river’ i.e. Transoxiana), while the Sasanids held the region to the South. This was an uneasy and short-lived arrangement doomed from its very inception because of commercial and political rivalries. After several unsuccessful embassies to Iran (the last of which ended with the poisoning of most of the Türk-Soghdian party), in 568 the Türks directed their first embassy to Constantinople, Iran’s principal rival in the Near East. Türk relations with all the major neighboring sedentary empires, Iran, Byzantium and China at the other end of the steppe, were largely conducted through their Soghdian subjects. The first Türk diplomatic mission to Constantinople was headed by the Soghdian Maniax. The latter brought a letter in “Scythian writing”, very probably Soghdian, which was translated for the Emperor (Blockley 1985: 110-115). The Byzantines who needed the Türks as a counterweight to Iran proved to be largely ineffective partners. The Türks, who were soon engaged in internecine strife with different groupings within the ruling clan battling for power, were badly defeated in 589 in the Herat region by Iran (Gumilev 1967: 126-131; Sinor 1990: 305-306). The Byzantine accounts portray Maniax as the kagan’s constant adviser. The Chinese sources, with their well-known disdain for the northern nomads, say that the Türks are “simple-hearted and not possessed of foresight” and the familiar game of “divide and conquer” could be used to control them if not for the “many Soghdians” living among them who are “wily and experienced and they teach and lead them” (Liu Mau-tsai 1958, 1: 87). Already having much imperial experience (they had previously been part of the Achaemenid and Alexandrine empires), the Soghdian ruling houses intermarried with the royal Türks, and their men of commerce and diplomacy, as we have seen, served as intermediaries with the surrounding sedentary empires. It was in the aftermath of Alexander’s conquest that the Soghdian diasporan trading colonies had begun. Thus, by the time of the formation of the Türk state, the Soghdians were international merchants of long standing. Their contacts with the Türks may well have considerably antedated the Türk rise to imperial status. In any event, the Soghdians now needed the military power of the Türk, which could be projected to the borders and beyond of the great empires lying to the south of the steppes. From the Soghdian perspective, the Türk could be used to prod China, often a reluctant trading partner, to open its markets (Barfield 1989: 158). The Soghdians, like other mercantile peoples, also engaged in extensive agriculture, animal husbandry (the Chinese sources credit them with producing excellent horses (Chavannes 1900: 134)) and manufactures. In addition to their mercantile pursuits―and very likely because of them, the Soghdians developed what Richard Frye has termed a “mercantile secularism with more tolerance for religions and for foreigners in general” than was typical of their Sasanid imperial neighbors. Religions here also often took on a syncretistic coloration, reflecting the diverse Turks and Iranians: An historical sketch 7 influences in the region. The same was true of the Khw razmians, who developed close commercial ties with the Volga region (Frye 1984: 351-355). As a consequence, these city-state Iranians were also major Kulturträger across Eurasia, and Soghdian, written in a variety of Semitic (Syriac) scripts, became a lingua franca of the Central Asian Silk Route (Schafer 1963: 12, 281, n. 43).19 It was through them that various religions from the Eastern Mediterranean (Manichaeanism, Christianity) and India (Buddhism) made their way into Central Asia. Texts in Soghdian dealing with those religions have been found, many stemming from Turfan (Rastorgueva et al. 1981: 350ff). Evidence for the penetration of Christianity (most probably Nestorian) among the Türks may, perhaps, be seen in a notice in Theophylactus Simocatta that the Persians sent some Turkic captives (presumably taken at Herat) to the Emperor Maurice (582-602) because they had crosses tattooed on their foreheads (de Boor 1972: 208).20 Not surprisingly, important religious concepts in Old Turkic associated with these universal religions are derived from or entered Turkic through Soghdian: Turkic učmaq ‘heaven, paradise’ < Soghdian «štm«x [uštmax], tamu, tamuq ‘hell’ < Soghdian tamw, Turkic yek ‘the Devil’ < Indic yakka (Clauson 1972: 257, 503, 910). Zoroastrianism, coming from the Iranian world, undoubtedly, had an impact as well. A reflection of Zurv nist notions may be seen in the “God of Time” (öd tengri) noted in the Köl Tegin inscription (KT, N10) (Tekin 1988: 22; Aalto 1971: 36; von Gabain 1983: 621). It is clear that the Soghdians had a far greater cultural impact on the Türk than did China. The Türk writing system, for example, the runiform scripts found in a variety of forms across Eurasia, may have derived from Soghdian or some other Semitic alphabet that entered the region through them or another Iranian people.21 The entry of Soghdian elements into Turkic is even more closely associated with the Uyghur Kaganate (744-840), the successor state of the Türk Empire in Mongolia. This occurred, however, already after a major shift in the Central Asian Iranian world with the conquest of that region by the Umayyad Caliphate. Before turning to the latter events which had a far-reaching, indeed, determinative impact on Turco-Iranian relations, we must say something further about the Uyghur-Soghdian symbiosis. This is characterized by an expansion and enrichment of the early Türk-Soghdian relationship. Soghdian missionaries played a direct role in the conversion of Bögü, the Uyghur kagan, to Manichaeanism in 762. In Uyghur documents, the Soghdian followers of this religion are referred to as niġošaklar ‘listeners’ (< Soghdian nġwš«k ‘listener’), the lowest level in Manichaean society (Nadeljaev et al. 1969: 358-359), 19 Schafer contends that it continued in this capacity until the thirteenth century, when it was replaced by Persian. This seems unlikely as Soghdian was giving way to Persian in many places long before then, see below. 20 Barthold (1963-1977, 2/2: 271-272), however, did not find any conclusive evidence for it during the S s nid era. 21 See discussion in Róna-Tas (1991: 55ff) and Kyzlasov (1994). 8 Peter B. Golden i.e. the catechumens or auditores, who supported the Elect, and as sartlar. The latter term derives ultimately from Sanskrit sartha ‘merchant’, the social grouping among the Soghdians with whom the Uyghurs most closely associated the religion. This term most probably entered Turkic through Soghdian, as Soghdian merchants were some of the most “ardent followers” of this faith. The Manichaean dindarlar ‘Manichaean elect’(< Soghdian δynδ«r ‘electus’, cf. Pers. dīnd r ‘pious, religious’) were settled along the Silk Route and played an important role in the transmission of texts into a variety of languages (Clauson 1972: 846; Lieu 1985: 184ff). More directly, it was from the Soghdian variants of the Syriac script that the Uyghurs received their alphabet which, in turn, was passed on to the Mongols and from them to the Manchus. Directly borrowed into Uyghur (although some of these terms may have been taken earlier into Türk) are terms such as ažun ‘life, living, being state of existence’ (< Soghdian ««zwn, later in an Islamic context it came to mean ‘this world, ad-duny ’) (Clauson 1972: 28), midik ‘layman’ (< Soghdian myd«k) (Clauson 1972: 765-766). Many of the terms stem from Sanskrit but entered Turkic through Soghdian: darm ‘Buddhist teaching, law’ (< Soghdian δ«rm < Sanskrit dharma), nirwan ‘nirvana’ (< Soghdian nyr «n < Sanskrit nirv na), estup ‘reliquary tower, shrine’ (< Soghdian «astwph < Sanskrit stūpa).22 In addition, a goodly number of technical terms associated with Manichaeanism entered Turkic from other Iranian languages but via Soghdian mediation: awardišn ‘assembly, collection’ (< Parthian «mwrdyšn), periken ‘female sorcerers’ (< Middle Pers. pryg«n). Even terms for important cultural implements, such as kegde / kagda ‘paper’, which was invented by the Chinese, were taken via Soghdian (Soghdian k« δyh) (von Gabain 1983: 620).23 The impact of urban life can be seen in terms such as kend ‘town’ (< Soghdian knδ) (Clauson 1972: 728; von Gabain 1983: 623). All of the subsequent Turkic nomadic states of Eurasia made use of specialists stemming from the surrounding (or distant) sedentary realms. These were individuals or groups who possessed special skills in areas of culture or commerce that were important to the nomads or who could serve as effective intermediaries (by virtue of their linguistic and literary skills) in dealing with sedentary society. The Soghdians, and later other Iranian peoples, would serve these bureaucratic and diplomatic functions in Turkic states in both Central Asia and the Near East after the establishment of Turkic states there. In the latter region, there was already a longstanding tradition of Iranian bureaucrats in state service that had come to the fore during early ªAbb sid times. This pattern would also be brought to Turkic states in the Indian subcontinent. The destruction of the Uyghur state by the Kirghiz drove many Uyghurs into hitherto predominantly Iranian and Tokharian Eastern Turkistan. 22 For these and some of the foregoing, see Aalto (1971: 36) von Gabain (1983: 617-718). 23 Clauson (1972: 710) has it as possibly Soghdian. Turks and Iranians: An historical sketch 9 This region now became increasingly Turkicized (Samolin 1964: 75-76; Bregel 1991: 56).24 While the Soghdians were playing this important and direct role in shaping Uyghur culture, their home territories in present-day Uzbekistan were overrun and conquered by the Arabs, who brought with them a new religion: Islam. This new faith would reshape Central Asia as well. After the collapse of the Sasanids (651), Arab probes of the lands beyond Khur s n began. In the 670s-680s Arab incursions increased, causing the local rulers to negotiate treaties with the invaders. The full Arab conquest of the region was delayed by a power struggle within the Umayyad ruling house. It was only in 705 that the Arab pressure on Iranian Central Asia resumed in earnest. Although the Soghdian princes attempted to fend off the Arabs with the aid of the Western Türks, their efforts were unsuccessful. In 737, the Arabs defeated the paramount Western Türk ruler (the same year in which they defeated and captured the Khazar Kagan on the Volga). The Muslim conquest of Iranian Central Asia was confirmed in 751 with the Arab victory over the Chinese near the Talas River in Kazakhstan, a victory to which the defection of the Karluks contributed. The Karluks went on to take over the Western Türk Kaganate in 766.25 The Türk, however, remained largely pastoral nomads and did not settle in the now Islamizing cities in appreciable numbers. To the West of Transoxiana, extending over the Volga-Caspian-North CaucasianPontic steppes, lay the Khazar Kaganate (ca. 650-965), an offspring of the West Türk Empire. Khazaria’s main points of contact with the Islamic world were Khw razm and Iran.26 The Khazars ruled over a polyglot population of Turkic peoples (including relatively large numbers of Oghuric Turks), Iranians (especially Alano-As tribes), Finno-Ugric, Slavic and Palaeo-Caucasian populations as well as colonies of Jews, Iranians and others, practicing a variety of economic pursuits ranging from pastoral nomadism, hunting and gathering to agriculture, crafts and commerce. The Khazar Kaganate became one of the greatest powers of the region. In the late eighthearly ninth century, elements of the Khazar elite converted to Judaism. In time, this religion became widespread among the Khazars proper. Islam and Christianity, however, were well-represented in the Khazar cities, where they may well have been the dominant faiths. Khazaria was ruled by a dual kingship which consisted of a now sacralized kagan (of probable A-shih-na origin) and his Qa an Beg (also called Yilig / Yele , cf. Turkic ellig / illig ‘king’ (Clauson 1972: 141-142) and *їš d < šad). The presence of sizable Alano-As groupings as well as close commercial ties with neighboring Khw razm and the Islamic world via the Volga-Caspian route, led to important Iranian influences in Khazaria. This was reflected in anthroponyms. Thus, 24 The process was largely completed by the end of the tenth century, see Aalto (1971: 37). 25 On these events, see Chavannes (1990: 142-143, 297-298); Gibb (1923); Frye (1975: 74ff. and 1966: 199); Beckwith (1987: 136-139). 26 For the Khazars the most recent study is that of Novosel’cev (1990). 10 Peter B. Golden one of the Khazar generals noted in 737 was “Haz r Tarx n” (< Persian haz r ‘one thousand’). A Khazar general prominent in Abbasid service was Ishaq b. KundaÅїq, whose father’s name may derive from Pers. kund (k) ‘wise, bold’.27 More importantly, the Khazar rulers, at the zenith of their power, had a large, salaried, standing army, the Ors (Arab. ‫ اﻻرﺳ ﻴﺔ‬deriving ultimately from auruša, the ’Άο σοι of the Classical sources) which was recruited from Muslim Khw razmian immigrants. Moreover, the wazīr of the Khazar ruler was chosen from this community (Pellat 1966-1979, 1: 213).28 Turko-Iranian interaction in Khazaria (one of the primary sources for slaves coming to the Islamic world, see below) has yet to be fully explored. By the later part of the eighth century and into the ninth century, Iranian Transoxiana became further integrated into the Muslim world. In the process, a number of transformations took place. Persian (F rsī), which had been making inroads among Soghdians and speakers of other Iranian languages since S s nid times, now, in the course of the ninth and tenth century, blossomed into New Persian, giving rise to a rich literature in that tongue (now written in the Arabic script) and serving as the language of government in the Irano-Muslim East. A new ethnonym developed as well: T Åīk, deriving from the Arab tribal name Tayyi« + Iranian suffix of people -čik. It appears in Turkic, in the form tezik (cf. Pahlavi t čīk, New Pers. t jīk, t zīk, Soghdian t«zyk, Khw razmian t«ćyk, Saka ttaśika) as early as the 720s (in the Küli Čor inscription) and entered Chinese as Ta-shih and Tibetan as ta-źig) (Orkun 1936-1941, 1: 138, Ixe Xušotu; Bailey 1982: 87-88). Islam, when it came to the Turkic peoples, came in an Iranized garb; the formerly Soghdian city states, now Islamized, playing their customary role of cultural intermediaries. This frontier Irano-Arabian Islam that had developed in Khur s n with the fusion of Arab settlers and Iranians not only played a significant role in the revolution that brought the ªAbb sids to power in 750, but also helped to shape the new, Islamic ecumenism which brought more non-Arab Muslims (especially from the Iranian lands) into the center of government (Frye 1975: 101-102, 126-127). As Iranized Arabs and Iranians began to play a more prominent role in caliphal politics, transforming the Caliphate into an Islamic version of the Sasanid Empire (Yarshater 1998: 11-13, 54-74), the Eastern Iranian territories, with their Central Asian and Turkic connections, loomed ever larger in Near Eastern affairs. The T hirids (821873), who had helped to secure the caliphal throne for al-Ma«mūn b. H rūn ar-Rašīd (813-833), also campaigned against the various Turkic tribal confederations that had established themselves in Central Asia in the aftermath of the fall of the Türk and Uyghur empires. Al-Ma«mūn and his brother and successor al-Muªtasim (833-842) began to build up a special slave guard corps consisting largely of Turkic ġilm n 27 For these names, see Golden (1980, 1: 202-204, 210-213). 28 On Ors-Urus, see Golden (1990: 33-46). Turks and Iranians: An historical sketch 11 (sing. ġul m, lit. ‘boy, servant’ and in this context denoting ‘military slave’).29 The prototype for this type of military organization, it has been suggested, may be seen in the č kar organization of the Soghdian kings (see above). Analogous institutions, it should be noted, can be found in Pre-Islamic Iran (see Zakeri 1995) and may be seen in the Imperial Bodyguard (Hetaireia) of the Byzantine emperors which, in the ninth century, recruited from some of the same sources (Khazaria and Central Asia) (Treadgold 1995: 110, 115; Whittow 1996: 169-170). In the ªAbb sid realm, the Turkic ġilm n, who were used against domestic and foreign foes alike, were kept largely segregated from the rest of the population. AlMuªtasim went so far as to build a new city, S marr (97 km north of Baghdad), in which to house his occasionally overbearing and riotous Turks from the increasingly hostile population of Baghdad. Turkic slave women were bought for them to marry (de Goeje 1892: 258-259, 262). Central Asian Iranians, such as Afšīn (a title used as his given name) Haidar, a descendant of the ruling house of Ustruš na, became some of the leading political figures of ninth century caliphal politics in association with these units (Frye 1975: 48, 115, 117, 189-190). The Turkic ġilm n were acquired through raids and trade with the Central Asian Turkic polities and the Khazar Kaganate (see below). The Samanids (819-1005), originally from the Balkh region, who were vassals of the T hirids and later replaced them, came to control the Muslim lands north of the Oxus and, as a consequence of their location, were one of the principal sources of ġilm n. Patrons of New Persian and the Persian renaissance (reflected in the work of the poet Firdawsī, whose epic Š hn ma had the conflict of Tur n, now representing the Turkic nomads, and Iran as its central theme), the Samanids also had large numbers of ġilm n who were given training in special schools. Indeed, it may be argued that it was the wealth that the Samanids acquired from the lucrative slave trade that helped to finance this rebirth of Persian culture.30 It was in this Iranian milieu that Turks also became acquainted with the Persian language in a significant way. Within the Caliphate, it seems very likely that Persian was one of the principal languages of communication between the Turkic ġilm n and Muslim society. This is illustrated by the account of how Ašin s, one of the leading Turkic ġul m commanders received his name. According to at-Tabarī, s.a. 202 / 818, in a battle against the sectarian ¾aw rīj, one of al-Muªtasim’s ġilm n placed himself in front of the future Caliph to protect him, shouting ašin s m -r (‘know me!’ < Pers. šin s < šin ¿tan ‘to know, to acknowledge’). Whereupon al-Muªtasim gave 29 On the Turkic ġilm n, see Pipes (1978), Töllner (1971), Yıldız (1976). 30 See the comments of Wink (1990, 1997, 1: 21). On the training of the ġilm n, see Niz m al-Mulk in his Siy sat-n ma (Qazwīnī & Čah rdihī 1334 / 1956: 109-110, translated by Darke 1960: 106-107, viewed with some doubts by Bosworth 1963: 280). On the numbers of Turkic ġilm n in Samanid service, see Bregel (1991: 56-58). 12 Peter B. Golden him this name (Abū«l-Fadl Ibr hīm 1967-1969, 8: 558). Whether true or not the story is significant in that it shows that Turks in caliphal service were using Persian. Samanid military activity into Turkic regions had produced many captives, some of whom embraced Islam. But, this was not the principal means by which Islam made its way into the Turkic steppe. Rather, it was the movement and activities of Muslim merchants and mystics (sūfīs) (Barthold 1968: 254- 255; Köprülü 1966: 814) that brought Islam to the Turkic tribal confederations (many of which had already been exposed to other universal religions: Christianity, especially in its Nestorian form, Buddhism, Manichaeanism and Judaism). This Islam came in a largely Persianized form (Köprülü 1966: 16) and expressed many of its fundamental concepts in a Soghdian and Persian vocabulary that had already been assimilated by Turkic. In the 920s the ruler of the Volga Bulghars converted. In 960, a mass conversion involving some “200, 000 tents of the Turks” in Central Asia took place (Amedroz 1920-1921, 2: 181; Tornberg 1851-1876, 8: 532). This was most probably associated with the Karakhanid state which was coming into being in Eastern and Western Turkistan under a dynasty of probable A-shih-na origin. Shortly thereafter, in 961, some Turkic ġilm n broke away from the now fading Samanids to form their own state in Afghanistan which expanded into Khur s n. These Ghaznavids, under the forceful leadership of Mahmūd (998-1030), became one of the great powers of the Eastern Muslim world, unleashing devastating raids on India while creating a powerful state in the Iranian lands. With its Turkic army and Iranian bureaucrats, the Ghaznavid polity was an interesting blend of Islamic, Turkic and Iranian traditions. Samanid and Ghaznavid models would play an important role in shaping Seljuk governance and culture. The Ghaznavids presented themselves as patrons of learning and the arts. Firdawsī, earlier at the Samanid court, finished his Š hn ma in Ghazna (ca. 1010) and presented it to Sultan Mahmūd (Arberry 1958: 43, 53-62). Al-Bīrūnī was one of the stars of Ghaznavid-supported culture. His brilliant study of India (written in Arabic) was made possible (but not, it appears, given actual support) by the Ghaznavid government. The Karakhanids, although ruling from important centers of Iranian culture (Bukh r , K šġar) and certainly influenced by Islamo-Iranian culture, produced a distinctly Turko-Islamic culture. It was from this milieu that the Qutad u Bilig originated, a splendid example of the mirror for princes genre, so prominent in Iranian political writing, but one which has meshed the various Turkic, Iranian and Islamic traditions. It was this same milieu that produced Mahmūd al-K šγarī, the great lexicographer of the Medieval Turkic world, who penned his famous Dīw n Luġ t at-Turk in Arabic, in Baghdad, but with the aim of making the Turks, now masters of the Islamic heartlands, better known to their subjects.31 Nonetheless, Arabic and increasingly Persian were the languages of state (Barthold 1963-1977: 111-112). 31 The principal work on the Ghaznavids is that of Bosworth (1963) and his The Later Ghaznavids (Bosworth 1977). On the Karakhanids, see Pritsak (1951) and (1953-1954), Turks and Iranians: An historical sketch 13 The Samanid decline had brought many confessionally diverse elements into the cities, each represented by their own politico-religious ethnarchs, functioning as the equivalents of the Muslim ªulam «. The “composite” society thus produced was ruled by a Turkic (but occasionally Persianized) royal and military elite with a bureaucracy drawn from a professional scribal class and ªulam « who were at home in both Arabic and Persian. Turkic, Iranian and Arabic traditions, all within the bounds of an Islamic world view, shaped this culture. In many respects, this was a continuation of the Turco-Soghdian symbiosis, but now in Islamic form. This culture would be extended to West and South Asia (Canfield 1991: 12-13).32 On the frontiers of the Karakhanid and Ghaznavid realms lay the unruly Oghuz tribal confederation extending from the Syr Darya to the Volga river. In 985, Seljük, a high-ranking warlord within the confederation, broke with his overlord, the Yab u, and repaired to Jand, where he converted to Islam. His sons and followers soon became embroiled in the ongoing Karakhanid and Ghaznavid strife. By the 1030s, they had been forced out of the region and had gathered their forces in Khur s n, a rich Ghaznavid province, parts of which in the time-honored tradition of desperate nomads they began to plunder. Their depredations brought forth a major Ghaznavid effort led by the Sultan Masªūd (1031-1040) himself. It ended with the complete rout of the latter at Dand nq n in 1040 and the shrinking of this once powerful state to its Afghan-Indian borderlands. The Seljuks, joined by Oghuz tribesmen now streaming out of Central Asia soon made themselves masters of the considerably debilitated ªAbb sid Caliphate. In 1055, the Seljuks, led by Toγrul and Čaγrї (grandsons of Seljük), whose victories had ended Shīªite Buyid control over Baghdad, were declared the champions and defenders of the Sunnī Muslim world.33 In 1071, Alp Arslan b. Čaγrї (1063-1072), defeated the Byzantine emperor in Eastern Anatolia, opening thereby much of Asia Minor to Oghuz penetration.34 The Seljuk ruling house easily adapted itself to the splendiferous IslamoPersianate imperial culture of the Near East. It continued to adhere, however, to the Old Turkic notion of the collective sovereignty of the ruling house over the state. Given the centrifugal inclinations of the tribes that had entered the region under their aegis, the constant domestic strife ultimately weakened the state. At its zenith under Malikš h b. Alp Arslan (1073-1092), the Seljuk state extended from Central Anatolia to the Karakhanid realms in Western and Eastern Turkistan, the former recognizing Seljuk supremacy and the latter paying homage. Golden (1990: 352ff) and (1992: 211ff). For a more detailed discussion of events, see Barthold (1968: 180ff). See also the comments of Köprülü (1966: 16-18). On Persian linguistic influences on the Qutad u Bilig, see Dankoff (1983: 10-12). On Mahmūd alK šγarī and Yūsuf Kh ss H jib, see Bombaci (1969: 89-101). 32 The Persian that came to India was that of Central Asia, not Iran proper (Ekram 1964: 36). 33 There are numerous accounts of Seljuk origins, see Sümer (1980: 61-91); Agadžanov (1969). 34 On the Seljuks in Asia Minor, the “Rūm Sultanate,” see Cahen (1968). 14 Peter B. Golden In many respects, the Seljuk state followed the patterns already established by the Samanids and their offshoots, the Ghaznavids. In addition to the familiar paradigm of a culturally Persianized ruling elite, a Turkic soldiery (both tribal and ġilm n) and a Persianate bureaucracy at home in both Arabic and Persian, we find the “composite” society that had already been established in Central Asia. The madrasa (Muslim theological college), a powerful tool in the strengthening of Islamic institutions and norms which may well have had its origins in Buddhist influenced institutions of learning in Central Asia (Barthold 1963-1977, 5: 60) was brought westward by the Seljuks. It was also under the Seljuks that we see the full flowering of the ªulam «, in particular of the Hanafī school, as a powerful urban force closely associated with the madrasa. This was a process that had begun under the Karakhanids and Ghaznavids and had given the New Persian that had developed in the Samanid realm and its successors in Central Asia a strong Arab lexical element. Typical of this TurcoPersianate style of governance was the greater organization of Sunnī Muslim religious institutions. This was also a style of government that extended from Anatolia to Muslim India.35 On the eve of the Mongol conquests, the most powerful state in Central Asia was that of the Khw razmš hs on the lower Oxus. It extended its authority into Iran and had ambitions on Baghdad. Ruled by a Turkic dynasty (the Anušteginids of ġul m origin) installed by the Seljuks and ultimately coming to rest on Qıpčaq soldiery from the surrounding steppes, Khw razm was a familiar blend of Turk and Iranian, with a largely Iranian urban and agricultural population that was already at this time undergoing considerable Turkicization. Persian, however, not the native tongue of the Iranian population, was the language of government. The ruling house, although intermarried with the Qıpčaqs, never succeeded in fully managing them. The state rested on a very insecure base and was swept away with remarkable ease by the Mongols in 1220 (Barthold 1963-1977, 5: 116ff; Kafesoğlu 1956; Bunijatov 1986). The Mongol conquests of the early and mid-thirteenth century shattered the Turkic world, breaking down and then reconfiguring its earlier tribal structures. It brought in new waves of Oghuz and other Turkic groups into the Near East, greatly advancing the Turkicization of much of Anatolia and Azarbaijan. In the latter region, older Iranian and Palaeo-Caucasian populations were largely assimilated (see Golden 1992: 283-308). In Central Asia, some of the great cities (such as Bukh r ) remained largely Iranian in speech, but the rest of the region was overwhelmingly Turkicized―with many becoming bilingual. Today, nearly 89% of the indigenous population is Turkic-speaking (Bregel 1991: 54, 60). In the Činggissid states that developed to the West of Mongolia proper, Turkic became the lingua franca and 35 Frye (1975: 222-230) suggests that this was because a power vacuum developed in the cities as the Turkic rulers preferred their nomadic encampments. Moreover, there was tension between the often heterodox forms of Islam brought to the nomads by the sūfīs and the more orthodox Islam of the ªulam « (Canfield 1991: 13-18). Turks and Iranians: An historical sketch 15 increasingly the language of state (Barthold 1963-1977, 5: 211, 213). The Činggissid Ilkh nid state that took over the old Seljuk lands in Iran, Iraq and Asia Minor, with its Islamization at the end of the thirteenth century, continued in that same TurkoPersianate tradition. In addition, the Mongol world empire with the pax that it created across Eurasia and with their interest in trade and their policy of moving skilled personnel from among the subject populations (this included literati) around the empire, gave far greater opportunities for cultural exchange between East and West. Thus, a Mongol official, Bolad Aġa, with a history of service in Yüan China, came to Iran and was a major source of Rašīd ad-Dīn’s J miª at-Taw rī¿, an extraordinary “world history” as seen from the vantage point of this Ilkh nid official (Allsen 1994). The conquests of Aqsaq Temür (Tamerlane, 1336-1405) were every bit as disruptive of life and property as that of Činggis Qan, whom he believed himself to be emulating. His empire extended from tributary Turkic Anatolia (he had defeated and captured the Ottoman sultan B yezid in 1402) to Turkistan. Temür, born into a Turkicized and Islamized tribe of Mongol origins (the Barlas), was quite familiar with the Irano-Islamic culture of the Central Asian cities (which many of his nomadic, tribal followers shunned and which he plundered more than once). Indeed, he, like others in the late Činggissid world who lived near the cities, was very much a part, albeit occasionally disruptive, of the system (Manz 1989: 17-18, 38, 90). While Temür may have prowled around the edges and spent most of his time in the saddle, his successors, the Timurids, were models of this Turko-Persianate culture; urbane, urban gentlemen and patrons of the arts and sciences. Poetry, the fine arts and astronomy (Temür’s grandson Uluγ Beg built an observatory) all flourished.36 During this period of relative tranquility, the Chaghatay tribes were drawn further into the Irano-Islamic culture of the region and some of them sedentarized (Bregel 1991: 61). Timurid political fractiousness opened their core lands in Transoxiana to invasion, in 1500, by the Kipchak Uzbek confederation led by the Činggissid Šayb nī Khan. One of the refugees from this conquest was Babur, who, following several failures to regain his patrimony, invaded India. Here, in 1526, he dislodged the Lodi sultans of Delhi, the political successors of a series of Turkic and Afghan dynasties ruling the northern regions of the Indian subcontinent that had begun with the Kipchak slave sultans of the early thirteenth century. These states had also fashioned a Turko-Persianate culture derived from the same Central Asian sources and similar to their Turkic contemporaries in Central and West Asia.37 This tradition was grandly continued and expanded by Babur’s successors, the Mughal dynasty (1526-1858), which presided over a unique fusion of Central Asian Turkic and Iranian elements with the indigenous cultures of South Asia and benefited from a continual influx of Islamo-Persian culture specialists (Persian was the language of 36 See the appropriate chapters in The Cambridge history of Iran (Jackson & Lockhart 1986, 6: The Timurid and Safavid Periods) and Golembek & Subtelny (1992); Bŭriev (1997). 37 On the early Mamlūk or “Slave Kings”, see Wink (1997). 16 Peter B. Golden state) hoping to make their fortune in this Islamic frontier (see Richards 1993, Chapter 1.5: The Mughal Empire and Foltz 1998). The Qıpčaq Özbeks (Uzbeks), coming from the Ulus of Âoči (the Golden Horde) created a khanate (soon to fracture into several states) in what is now called Uzbekistan (see Burton 1997). Here, they imposed their military power on several layers of Turkic peoples that had preceded them into the region as well as the old Iranian population. Sedentarizing Turkic nomads, however, in some instances became Tajik-speaking (e.g. many of the Chaghatay tribal grouping who are today largely Tajik in speech) and were themselves sometime termed sart (a term often used pejoratively by the Turkic nomads to denote the sedentary Iranians, but also in some regions applied to sedentarized Turks) (see Subtelny 1994: 50-51; Bregel 1978; Golden 1992: 333-338, 406-408). Some Tajik dialects, in turn, have become so deeply affected by Turkic that they are becoming Turkic (Doerfer 1998: 239). In keeping with now well-established tradition, the Özbek states had a bilingual Turkic elite, a Turkic soldiery and an Iranian (Tajik) bureaucracy. Bilingualism was relatively widespread. Persian had a profound impact on the Turkic literary tongue (Chaghatay), lexical borrowings in some texts can range from 60-90% (Blagova 1997: 159; Bregel 1991: 62-63). A similar structure could be found in neighboring Iran under the Safavid dynasty (1501-1722). The dynasty’s origins are obscure (of probable Persian or Kurdish origins), but by the time they emerge onto the stage of history, they had become largely Turkic in speech and immediate ancestry. Shah Ismail (1501-1524), who brought the dynasty to power in Iran and established its present-day borders (more or less) and confessional allegiances (he declared Shīªism the state religion), wrote poetry in Azeri Turkic. Despite the weight of Iranian imperial tradition, Turkic was the preferred language at court. The Safavids had come to power at the head of a militant sūfī order largely supported by Turkish tribesmen in Anatolia and Azerbaijan termed the qїzїl baš (‘red head’) because of the distinctive headgear they wore to honor the Shiite imams. The Safavids had adopted Shiite views (or something close to it) by the latter part the fifteenth century, but these were solidified only with their conquest of Iran. In some measure, Shiism was useful to distinguish themselves from their Sunnī neighbors, the Ottomans and the Özbeks, with whom they were periodically locked in struggle. The Ottomans and Safavids initially rested on similar bases, the Turkic ġ zīs (‘fighters for the faith’) and tribesmen of Anatolia and its borderlands. Similarities in state structure can be seen in the subsequent extensive employment of slave troops: the Ottoman Janissaries, mostly recruited from the Orthodox Christian populations of the Balkans, and the Safavid quls, largely introduced by Shah Abbas I (1587-1629) and drawn from the Christian Georgians and Armenians and pagan populations of the North Caucasus. In Safavid Iran, tensions between Iranian bureaucrats and Turkic soldiery were well-known. Iran proper had been under Turkic rule, in one form or another, since Ghaznavid and Turks and Iranians: An historical sketch 17 Seljuk times. This tradition largely continued under the Afshar (1736-1796) and Qajar (1779-1925) dynasties.38 The greatest of all the Turkic states of West Asia was the Ottoman Empire (ca. 1300-1922), which had extensive holdings in Europe and North Africa as well. Emerging from a small but acquisitive frontier principality (beylik) in Anatolia, by 1453 it had conquered Constantinople and in 1516-1517 (after having defeated the Safavids in 1514) overran the Mamlūk state in Syria-Egypt. In control of the Islamic heartlands (including the holy cities of Mecca and Medina), Ottoman sultans could with some justice claim to be the heirs of the Roman-Byzantine, Islamic and Turkosteppe empires (İnalcık 1973: 41, 56-57). Structurally, the Ottoman state, in many respects, runs in a straight line from that of the S s nids, with various accretions over time from the Arabs, other Turkic peoples and the Mongols. Although Turkish became the language of state, men of culture and learning were expected to know Arabic and Persian as well. Ottoman Turkish, like its close kinsman, Azerī Turkic and the somewhat more distantly related Chaghatay Turkic used in the Özbek khanates, was deeply impacted by lexical and syntactic borrowings from Persian.39 All of the major Islamic states of early Modern Eurasia, the Ottomans, Safavids and their successors, the Mughals and the Özbeks shared this close symbiosis with the Iranian world. In this we see a pattern of interaction that dates back to the earliest Turkic states. The role of the empire-building nomads40 as Kulturträger, who, having established a pax, allowed for the free flow of goods and information was put forward by scholars such as Barthold and most recently by Adshead (both with respect to the Mongol Empire) (see Bregel 1985: 388, Adshead 1993: 4-6, 53ff). This thesis has been challenged by Bregel who, in general, attributes to the TurcoMongol nomads a largely negative role in the development of culture in neighboring sedentary societies (Bregel 1991: 69). The most recent studies by Allsen of the role of the Mongols in fostering trade and cultural interaction, however, show that not only did the Mongol Empire provide the setting for this interaction, but the Mongol rulers were not passive bystanders in the process. They were actively engaged and moved personnel, material and intellectual property to suit their interests (see Allsen 1994). Although we are much less well-informed (and they were operating on a smaller scale), earlier Turkic empires (e.g. the Türk, the Uyghurs, the Khazars) were, in all likelihood, similarly active agents in promoting trade and with it cultural exchange. That the Turks played an important role in Činggissid mercantile culture can be seen in terminology. The Turkic term ortoq / ortuq / ortaq ‘partner’ (Clauson 1972: 205) (> Mongolian orto ), came to designate in Mongolian ‘a merchant 38 On the Safavids, see Sümer (1976), Savory (1980), Roemer (1986). 39 See brief discussion in Doerfer (1998: 241-243). 40 Some, e.g. Pritsak (1981, 1: 10-20), argue that it was the long distance merchants from sedentary society that were the real force behind the creation of the nomadic empires. 18 Peter B. Golden operating with capital supplied by a Činggissid prince or court official’. In Persian sources of the era, Pers. sūd ‘profit, gain, interest’ is replaced with Turkic asї ‘profit, advantage’ (> Mongolian asi ) (Barthold 1963-1977, 5: 110; Allsen 1989: 85, 118119; Clauson 1972: 244-245). Similarly, the Turkic term sart (see above) denoting a ‘merchant’ (usually of Iranian origin) was transformed by the Mongols (sartaq, sartaġul / sarta«ul, sartaġtai) to designate ‘Muslim’ and was applied to Turkic as well as Iranian Muslims (Barthold 1963-1977, 5: 109). A similar transformation had, as we have seen, occurred earlier with the Arab tribal name that became T Åīk (see above). The Turks had long interacted with Iranian Kulturträger who, it should be noted, had received significant elements of the culture that they brought to the Turks from the Semitic Middle East. When the Soghdians, now Islamized, became increasingly confined to the Muslim zone of Central Asia, the Uyghurs, who had lost their nomadic empire, became the major bearers of Near Eastern cultures (religions, writing systems) in Mongol Inner Asia (among the Qitan and later in the Činggissid realm). While the Mongols absorbed only small groupings of their subject culturebearers, the Turks assimilated very considerable numbers of the Central Asian Iranians (Janhunen 1996: 137-138). In the millennia-long interaction of Turks and Iranians, the former have borrowed and transformed many elements of culture while steadily absorbing the latter. References Aalto, P. 1971. Iranian contacts of the Turks in pre-Islamic times. In: Ligeti, L. (ed.) Studia Turcica. Budapest: Akadémiai Kiadó.29-37. Abaev, V. I. 1958. Istoriko-étimologičeskij slovar’ osetinskogo jazyka. Moskva, Leningrad: Akademija Nauk. Abū«l-Fadl Ibr hīm 1967-1969. At-Tabarı, Ta«rī¿ at-Tabarī 8. Cairo. Adshead, S. A. M. 1993. Central Asia in world history. London: Macmillan. Agadžanov, A. G. 1969. Očerki istorii oguzov i turkmen Srednej Azii IX-XII vv. Ašxabad. Allsen, T. 1994. Two cultural brokers of medieval Eurasia: Bolad Aqa and Marco Polo. In: Gervers, M & Schlepp, W. (eds.) Nomadic diplomacy, destruction and religion from the Pacific to the Adriatic. (Toronto Studies in Central and Inner Asia 1.) Toronto: Joint centre for Asia Pacific Studies. 63-78. Allsen, T. T. 1989. Mongolian princes and their merchant partners. Asia Major, 3rd series, 2 / 2 [page numbers] Allsen, T. T. 1997. Commodity and exchange in the Mongol Empire. Cambridge: Cambridge University Press. Amedroz, H. F. (ed.) 1920-1921. Ibn Miskawaih, Taj rub al-Umam 2. Oxford. [Translated by D. S. Margoliouth.] Arberry, J. 1958. Classical Persian literature. London: Allen & Unwin. [Reprint: Richmond, 1994.] Atalay, B. (ed.) 1945. Ettuhfet-üz-zekiyye fil-lûgat-it-türkiyye. İstanbul: Bürhaneddin basımevleri. Bailey, H. W. 1982. The culture of the Sakas in Ancient Iranian Khotan. Delmar, New York: Caravan Books. Turks and Iranians: An historical sketch 19 Bailey, H. W. 1985. Khotanese texts 7. Cambridge: The University Press. Barfield, T. 1989. The perilous frontier. Oxford: Blackwell. Barthold = Bartol’d, V. V. 1963-1977. Dvenadcat’ leksij po istorii tureckix narodov Srednej Azii. In: Sočinenija 5. Barthold = Bartol’d, V. V. 1963-1977. Istorija turecko-mongol’skix narodov In: Sočinenija 5. Barthold = Bartol’d, V. V. 1963-1977. O xristianstve v Turkestane v domongol’skij period. In: Bartol’d, V. V. In: Sočinenija II / 2. Barthold, W. 19683. Turkestan down to the Mongol invasion London: Luzac. [Translated by T. Minorsky.] Beckwith, C. 1984. Aspects of the early history of the Central Asian guard corps in Islam. Archivum Eurasiae Medii Aevi 4, 29-43. Beckwith, C. 1987. The Tibetan Empire in Central Asia. Princeton: Princeton University Press. Benkő, L. et al. 1967-1976. A magyar nyelv történeti-etimológiai szótára 3. Budapest: Akadémiai Kiadó. Blagova, F. G. 1997. Čagatajskij jazyk. In: Tenišev, Ė. R. (ed.) Tjurkskie jazyki. Biškek: Izd. Dom “Kyrgystan”. Blockley, R. C. (ed.) 1985. The history of Menander the Guardsman. Liverpool: F. Cairns. Bombaci, A. 1965. Qutluγ Bolzun! Ural-Altaische Jahrbücher 36, 284-291. Bombaci, A. 1966. Qutluγ bolsun (II) Ural-Altaische Jahrbücher 38, 13-43. Bombaci, A. 1969. La letteratura turca. Firenze, Milano: Sansoni. de Boor, C. (ed.) 1887. Theophylactus Simocatta, historiae. [Reprint Stuttgart, 1972.] Bosworth, C. E. 1963. The Ghaznavids. Edinburgh. Edinburgh University Press. Bosworth, C. E. 1977. The later Ghaznavids. New York: Columbia University Press. Bregel, J. 1978. The Sarts in the Khanate of Khiva. Journal of Asian History 12, 120-151. Bregel, J. 1980. Barthold and Modern Oriental Studies. International Journal of Middle Eastern Studies 12, 385-403. Bregel, Y. 1991. Turko-Mongol influences in Central Asia. In: Canfield, R. L. (ed.) TurkoPersia in historical perspective. Cambridge: Cambridge University Press. Bŭriev, . 1997. Temuriylar dawri y zma manbalarida markaziy siy . T škent. Bunijatov, Z. M. 1986. Gosudarstvo xorezmšaxov-anušteginidov 1097-1231. Moskva. Burton, A. 1997. The Bukharans. A dynastic, diplomatic and commercial history 1550-1702. New York: St. Martin’s Press. Caferoğlu, Ahmet (ed.) 1931. Abū Hayy n. Kitab al-Idrak li lisan al-atrak. [Arabic text.] İstanbul. Cahen, C. 1968. Pre-Ottoman Turkey. New York: Sidgwick & Jackson. [Translated by J. Jones-Williams.] Canfield, R. L. 1991. Introduction: The Turko-Persian tradition. In: Canfield, R. L. (ed.) Turko-Persia in historical perspective. Cambridge: Cambridge University Press. [page numbers] Chavannes, E. 1900. Documents sur les Tou-kiue (Turcs) occidentaux. St. Petersburg: J. Glasounof. [Reprinted 1941, Paris.] Clauson, Gerard Sir 1972. An etymological dictionary of pre-thirteenth century Turkish. Oxford: Clarendon Press. Czeglédy, K. 1983. From east to west: The age of nomadic migrations in Eurasia. Archivum Eurasiae Medii Aevi 3, 25-125. 20 Peter B. Golden Dąbrowski K. 1975. Hunowie europejscy. In: Dąbrowski, Krzystof & Nagrodzka-Majchrzy, Teresa & Tryjarski, Edward (eds.) Hunowie europejscy, Protobułgarzy, Chazarowie, Pieczyngowie. Wrocław, Warszawa, Kraków: Zaklad Narodowy im. Ossolinskich. Dankoff, R. & Kelly, J. 1982-1985. Mahmūd al-K š arī. Compendium of the Turkic dialects. (Dīw n Lu t at-Turk.) Harvard: Harvard University Press. Dankoff, R. (ed.) 1983. Yūsuf Kh ss H jib. Wisdom of Royal Glory (Kutadgu Bilig). A TurkoIslamic mirror for princes. Chicago: University of Chicago Press. Doerfer, G. 1998. The influence of Persian language and literature among the Turks. In: Hovannisian, R. G. & Sabagh, G. (eds.) The Persian presence in the Islamic world. Cambridge: Cambridge University Press. [page numbers] Ekram, S. M. 1964. Muslim civilization in India. New York: Columbia University Press. Foltz, R. C. 1998. Mughal India and Central Asia. Karachi et al.: Oxford University Press. Frye, R. 1984. History of Ancient Iran. München: C. H. Beck. Frye, R. N. 1996. The heritage of Central Asia. Princeton: Wiener. Frye, R. N. 1975. The Golden Age of Persia. London: Weidenfeld & Nicolson. von Gabain, A. 1983. Irano-Turkish relations in the late Sasanian period. In: Yarshater, E. (ed.) The Cambridge history of Iran. The Seleucid, Parthian and Sasanian periods 3 / 1. Cambridge: Cambridge University Press. 613-624. Gamkrelidze T. V. & Ivanov, V. V. 1984. Indoevropejskij jazyk i indoevropejcy. Tblisi: Izd. Tbilisskogo univ. Gibb, H. A. R. 1923. The Arab conquests in Central Asia. New York. de Goeje, M. J. 1892. Al-Yaªqūbī. Kit b al-Buld n. Leiden: Brill. Golden, P. B. 1980. Khazar studies 1. Budapest: Akadémiai Kiadó. Golden, P. B. 1990. Cumanica 3: Urusoba. In: Sinor, D (ed.) Aspects of Altaic civilization 3. Bloomington: Indiana University. [page numbers] Golden, P. B. 1990. The Karakhanids and early Islam. In: Sinor, D. (ed.) The Cambridge history of early Inner Asia. Cambridge. [page numbers]. Golden, P. B. 1992. An introduction to the history of the Turkic peoples. (Turcologica 9.) Wiesbaden: Harrassowitz. Golembek, L. & Subtelny, M. (eds.) 1992. Timurid art and culture. Iran and Central Asia in the fifteenth century. Leiden: Brill. Grjaznov, M. P. 1969. Southern Siberia. Geneva, Paris, Munich: Nagel. [Translated by J. Hogarth.] Gumilev, L. N. 1967. Drevnie tjurki. Moskva: Nauka. Haussig, H. W. 1983. Die Geschichte Zentralasiens und der Seidenstrasse in vorislamischer Zeit. Darmstadt: Wissenschaftliche Buchgesellschaft. İnalcık, H. 1973. The Ottoman Empire. The classical age, 1300-1600. London: Weidenfeld & Nicolson. Jackson, P. & Lockhart, L. (eds.) 1986. The Cambridge history of Iran 6. Cambridge: Cambridge University Press. Janhunen, J. 1996. Manchuria. An ethnic history. Helsinki: Finno-Ugrian Society. Kafesoğlu, İ. 1956. Harezmşahlar devleti tarihi. Ankara: Türk Tarih Kurumu. Xazanov, A. M. 1984. Nomads and the outside world. Cambridge, New York: Cambridge Unversity Press. [Translated by J. Crookenden.] Kiselev, S. V. 1949. Drevnjaja istorija južnoj Sibiri. Moskva: AN. Kljaštornyj, S. G. & Savinov, D. G. 1994. Stepnye imperii Evrazii. St. Petersburg: Farn. Turks and Iranians: An historical sketch 21 Kljaštornyj, S. G. & Livšic, V. A. 1972. The Sogdian inscription of Bugut revised. Acta Orientalia Hungarica, 26 / 1, [page numbers] Kljaštornyj, S. G. 1973. Drevnetjurkskaja pis’mennost’ i kul’tura narodov Central’noj Azii. Tjurkologičeskij Sbornik 1972, [page numbers.] Klyaštornyj, S. G. 1994. The royal clan of the Turks and the problem of early Turkic-Iranian contacts. Acta Orientalia Hungarica 47 / 3, [page numbers] Köprülü, F. 19662. Türk edebiyatında ilk mutasavvıflar. Ankara. Kyzlasov, I. L. 1994. Runičeskie pis’mennosti evrazijskix stepej. Moskva. Lieu, S. N. C. 1985. Manichaeism in the later Roman Empire and Medieval China. Manchester et al.: Manchester University Press. Litvinskij, B. A. (ed.) 1992. Vostočnyj Turkestan v drevnosti i rannem srednevekov’e. Ėtnos, jazyki, religii. Moskva: Nauka. Liu Mau-tsai 1958. Die chinesischen Nachrichten zur Geschichte der Ost-Türken (T’u-küe) 1. Wiesbaden: Harrassowitz. Maenchen-Helfen, O. 1973. The world of the Huns. Berkeley et al.: University of California Press. Mallory, J. P. 1989. In search of the Indo-Europeans. London: Thames & Hudson. Manz, B. F. 1989. The rise and rule of Tamerlane. Cambridge et al.: Cambridge University Press. Miziev, I. M. 1996. Istorija Balkarii i Karačaja s drevnejšix vremen do poxodov Timura. Nal’čik: El-Fa. Moriyasu, T. & Ochir, A. (eds.) 1999. Provisional report of researches on historical sites and inscriptions in Mongolia from 1996 to 1998. Osaka: Society of Central Eurasian Studies. Nadeljaev, V. M. et al. 1969. Drevnetjurkskij slovar’. Leningrad: Nauka. Novosel’cev, A. P. 1990. Xazarskoe gosudarstvo i ego rol’ v istorii Vostočnoj Evropy i Kavkaza. Moskva. Orkun, H. N. 1936-1941. Eski Türk yazıtları 1. Ankara: Türk Dil Kurumu. Pellat, C. & de Meynard, B. & Pavet de Courteille, A. (eds.) 1966-1979. Al-Masªūdī, Murūj ad-Dahab wa«l-Maª din al-Jawhar. Beirut. Pipes, D. 1978. Turks in early Muslim service. Journal of Turkic Studies 2, 66-78. Pletneva, S. A. 1999. Očerki xazarskoj arxeologii. Moskva, Ierusalim. Blockley, R. C. (ed.) 1981-1983. The fragmentary classicizing historians of the later Roman Empire, Eunapius, Olympiodorus, Priscus and Malchus. Liverpool: Cairns. Pritsak, O. 1951. Von den Karluk zu den Karachaniden. Zeitschrift der Deutschen Morgenländischen Gesellschaft 101, 270-300. Pritsak, O. 1953-1954. Die Karachaniden. Der Islam 31, 17-68. Pritsak, O. 1981. The origin of rus’ 1. Cambridge, Mass.: Distributed by Harvard University Press for the Harvard Ukrainian Research Institute. Qazwīnī, M. & Čah rdihī, M (eds.) 1334 / 1956. Siy sat-n ma. Tihran: Kitab Furusi-i Tuhuri. [Translation: Darke, Hubert 1960. The Book of Government or Rules for Kings. London: Routledge & Kegan Paul.] Rastorgueva, V. S. et al. 1981. Osnovy iranskogo jazykoznanija. Sredneiranskie jazyki. Moskva: Nauka. Richards, J. F. 1993. The New Cambridge history of India. Cambridge: Cambridge University Press. Roemer, H. R. 1986. The Safavid period. In: Jackson, P. & Lockhart, L. (eds.) The Cambridge history of Iran 6. Cambridge: Cambridge University Press. 189-350. 22 Peter B. Golden Rolle, R. 1989. The world of the Scythians. London: Batsford. [Translated by F. Gayna Walls.] Róna-Tas, A. 1991. An introduction to Turkology. Szeged: Univ. Szegediensis de Attila József Nominata Rudenko, S. I. 1970. Frozen tombs of Siberia. The Pazyryk burials of iron-age horsemen. London: Dent & sons. [Translated by M. W. Thompson.] Samolin, W. 1964. East Turkistan to the twelfth century. The Hague: Mouton. Savory, R. 1980. Iran under the Safavids. Cambridge: Cambridge University Press. Schafer, E. H. 1963. The golden peaches of Samarkand. A study of T’ang exotics. Berkeley: University of California Press. [Reprinted 1985.] Sinor, D. 1990. The establishment and dissolution of the Türk Empire. In: Sinor, D. (ed.) The Cambridge history of early Inner Asia. Cambridge: Cambridge University Press. [page numbers] Subtelny, M. 1994. The symbiosis of Turk and Tajik. In: Manz, B. F. Central Asia in historical perspective. Boulder: Westview Press. [page numbers] Sümer, F. 1976. Safevi devletinin kuruluşu ve gelişmesinde Anadolu Türklerinin rolü. Ankara: Güven matbaası. Sümer, F. 19803. Oğuzlar. Istanbul: Ana. Tekin, T. 1988. Orhon yazıtları. Ankara: Türk Tarih Kurumu basımevi. Töllner, H. 1971. Die türkischen Garden am Kalifenhof vom Samarra. Walldorf, Hessen: Verlag für Orientkunde Vorndran. Tornberg, C. J. (ed.) 1851-1876. Ibn al-Atīr. Al-K mil fı«t-Ta«rı¿. [Reprint: Beirut, 1965-1966 with different pagination.] Treadgold, W. 1995. Byzantium and its army 284-1081. Stanford: Stanford University Press. Whittow, M. 1996. The making of Byzantium 600-1025. Berkeley: University of California Press. Wink, A. 1990. Al-Hind. The making of the Indo-Islamic world 1. Leiden, New York, Köln: Brill. Wink, A. 1997. Al-Hind. The making of the Indo-Islamic world 2. Leiden, New York, Köln: Brill. Yarshater, E. 1998. The Persian presence in the Islamic world. [Introduction.] In: Hovannisian, R. G. & Sabagh, G. (eds.) The Persian presence in the Islamic world. Cambridge: Cambridge University Press. [page numbers] Yıldız, D. 1976. İslâmiyet ve Türkler. İstanbul: Edebiyat Fakültesi Matbaası. Zakeri, M. 1995. S s nid soldiers in Early Muslim society. The origins of ªAyy r n and Futuwwa. Wiesbaden: Harrassowitz. Zeimal, E. V. 1983. The political history of Transoxiana. In: Yarshater, E. (ed.) The Cambridge history of Iran, The Seleucid, Parthian and Sasanian periods 3 / 1. Cambridge: Cambridge University Press. [page numbers]