1 The origins and early history
of the Seljūqs
Western Eurasia and Transoxiana,
c. 900ā1025
The history of the Middle East is intertwined with that of the Eurasian steppe. Long before the emergence of the SeljÅ«qs, the āAbbÄsids had sought to exploit Turkish manpower for their armies. Byzantium, too, relied on its Turkish allies, the Khazar empire, to guard its frontiers to the north of the Black Sea. Geographers, travellers and historians from Byzantium and the Islamic Middle East preserve valuable fragments of information regarding the steppe. At the other extremity of Eurasia, Chinese sources recount the Middle Kingdomās efforts to deal with its difficult ā and often powerful ā nomadic neighbours. These sources allow us to discern the broad outlines of the great migratory movements of the late first millenium from which the SeljÅ«qs emerged.
The SeljÅ«qs were by no means the first group of Turks to come to Islamic Central Asia and KhurÄsÄn;1 nor were they the first nomadic invaders to reach the borders of Byzantium, which had been penetrated by Huns as early as the fourth century, by Sabirs, Avars, Khazars, Bulgars, Pechenegs and other Altaic or Turkish groups.2 Indeed, the SeljÅ«qs were not even the first to reach Anatolia, where Huns and Sabirs had preceded them,3 although they had not stayed. Compared to other Altaic peoples, the SeljÅ«qsā migration was on a fairly small scale, at least initially, and even their success at seizing political power was not unique, for the Ghaznavids and QarakhÄnids were also Muslim, Turkish dynasties who took over the Central Asian lands of the SÄmÄnid state, heir to āAbbÄsid and Iranian traditions of rulership, which collapsed at the end of the tenth century. The historical significance of the SeljÅ«qs lies in their success at dominating the central lands of the Islamic world ā the majority of the Nile to Oxus region described by Marshall Hodgson as the Islamic oikumene4 ā in contrast to the Ghaznavids and QarakhÄnids who held sway over remote, recently converted frontier regions. However, the heritage of the steppe was to play a decisive role in early SeljÅ«q politics and society. In this chapter, we shall review the evidence for the origins of the SeljÅ«qs and their early migrations. The term SeljÅ«qs refers here to the immediate family of the dynastyās ancestor DuqÄq, his son SeljÅ«q, and the latterās children and grandchildren, of whom the most prominent were į¹¬ughril and ChaghrÄ«.
Seljūq origins: the Oghuz connection
The meaning and pronunciation of the word SeljÅ«q is obscure,5 and it first occurs as the name of an individual, not a people or tribe. SeljÅ«q is said to have been the son of a chief named DuqÄq or TuqÄq,6 who was probably based in the steppes lying between the Aral Sea and the Volga in the tenth century. SeljÅ«q and his family are said to have belonged to the Qiniq tribe of the Oghuz, hence the common description of the Seljuqs and their supporters as al-ghuzz or al-atrÄk al-ghuzzÄ«ya (ghuzz being the Arabic version of Oghuz).7 The Oghuz are encountered in the eighth-century Orkhon River inscriptions in early Turkic from Mongolia.8 However, the term Oghuz did not originally denote a people, but was rather a political term meaning a ātribal unionā.9 Groupings called Oghuz played an important role in an earlier Central Asian Turkish empire, the Second Gƶk TĆ¼rk Khanate, but we cannot be sure that these necessarily had any ethnic connection with the SeljÅ«qs or even with the Oghuz we encounter in the tenth century around the Aral Sea.10 A report by an unnamed KhurÄsÄnÄ« historian cited by Ibn al-AthÄ«r discussing the Oghuz of twelfth-century Transoxiana who kidnapped the SeljÅ«q sultan Sanjar provides one of the few pieces of evidence for their origins: ā These Ghuzz are a people who migrated from the border region of the furthest lands of the Turks to Transoxiana in the days of al-MahdÄ« and converted to Islam. ā 11
The original lands of this Oghuz grouping were thus probably in Mongolia, home to several Turkish empires. A migration into Transoxiana (or more probably the Syr Darya steppe to its north) around 158/775ā169/785, the dates of the āAbbÄsid Caliph al-MahdÄ«ās reign, is supported by some circumstantial evidence.12 At any rate, the eighth and early ninth centuries witnessed a chain of migrations across Eurasia; with the collapse of the Second Gƶk TĆ¼rk Khanate in 744, a power struggle among the Turks ensued, and the movement of the Oghuz and other confederations like the Qarluqs into western Central Asia forced out the existing Turkish population, the Pechenegs, into the Ural-Itil region.13
According to Sharaf al-Dīn Marvazī, a physician who served at the Seljūq court, a second major wave of migrations was precipitated by events in southern Manchuria:
To [the Turks] belong the QÅ«n; they came from the land of QitÄy, fearing the QitÄkhÄn. They (were) Nestorian Christians, and they had migrated from their habitat, being pressed for pastures ā¦ The QÅ«n were pursued by a people called the QÄy, who being more numerous and stronger than they drove them out of these [new] pasture lands. They then moved on to the territory of the ShÄrÄ«, and the ShÄrÄ« migrated to the land of the TĆ¼rkmƤns who in their turn shifted to the eastern parts of the Ghuzz country. The Ghuzz Turks then moved to the territory of the BajanÄk [Pechenegs], near the shores of the [Black]14 Sea.15
This valuable but problematic report has provoked much scholarly debate about the identity of the tribes and the date of the migrations.16 The QitÄkhÄn is clearly the Khitan ruler and the QÄy are generally accepted to be the people called Hsi in Chinese. The date for the Khitan attack on the QÄy has been put as early as 870,17 but the Khitans ā also known by their Chinese dynastic name of Liao ā are generally agreed to have emerged as a major power in the first decade of tenth century.18 On the other hand, the movement of the Pechenegs into the region of the Black Sea is attested by Constantine Porphyrogenitus, a well-informed tenth-century Byzantine source;19 his account cannot refer to events any later than c. 900, and it has been argued that they should be placed around 825ā865.20 It is difficult to imagine that the rise of the Khitans could have more or less simultaneously precipitated a chain reaction thousands of miles away at the other extremity of Eurasia, let alone one which may have preceded their appearance by more than 50 years.
Pritsak attempted to solve the chronological difficulties by suggesting that MarvazÄ«ās report conflates two separate but connected migratory movements westwards, the first occuring in the late ninth century, the second precipitated by the rise of the Tanguts on the eastern borders of Tibet, around 1000ā1050.21 The evidence for this is complex and not wholly satisfactory, but the core of MarvazÄ«ās information, that the rise of the Khitan dynasty set off a chain of westward migrations, is likely to be true,22 even if they cannot be held responsible for the Oghuz move into the Black Sea steppe. However, the extent and nature of these migrations is debatable, and other factors may have been at work. Evidence discussed below indicates increased aridity in Central Asia from c. 900 onwards. This may also have contributed to these migrations as sources of pasture, essential for feeding the animals on which a nomadic lifestyle relies, dried up. Chinese sources record that some tribes in the tenth century starved due to drought.23 However, it is not possible to reconcile the precise details given by MarvazÄ« with other historical facts.
Thus over the eighth to tenth centuries there were probably several migrations of groups known as Oghuz who were gradually pushed westwards from Mongolia into Central Asia, and then towards the Volga. This series of migrations lies behind the confusion in the sources about the original homeland of the Oghuz groups, which is placed by some not in Mongolia but in the area approximating the south of modern Kazakhstan and northern Kyrghyzstan, stretching from Lake Issyk Kul to Yangikant on the western extremity of the Syr Darya River, near the Aral Sea,24 and by oth...