The ostensibly Turkish and Russian sources that document the early history of the Crimean Khanate state that it was ruled by a khan in consultation with a council of notables, known as the qarachu begs (also qaraçi bey; commoner commander). These qarachu begs were said to have represented the interests of the ‘land’, or perhaps more accurately the propertied Tatar elite, against the ‘state’ represented by the khan and his household. The most significant powers of the qarachu begs included appointing and deposing khans, co-signing all decrees and letters initiated by the khan and advising him on all significant policy decisions. These powers regularly placed the qarachu begs at odds with the ruling Girayid dynasty, many of whose members campaigned to reduce their power with mixed success.Footnote 1 The origin of the qarachu begs remains obscure due to the lack of detailed information on the Golden Horde, the Mongol state from which the Crimean Khanate emerged. This shortfall has led several historians to search for the origins of the qarachu begs in the early Mongol Empire and its successor states in Iran, China and Central Asia. These studies have met with limited success since no institution or office has been found which corresponds to the duties and powers ascribed to the qarachu begs of the Crimean Khanate.
The present study shall address the lacuna in our knowledge of the qarachu begs by placing the office in the context of time and place. The qarachu begs are generally thought to have emerged in the Golden Horde during the first half of the fourteenth century, at which time the neighbouring Īl-Khānate was fracturing into regional amirates ruled by non-royal commanders (qarachus). In most instances these commanders based their authority upon the laws and customs (jasaq and yosun) attributed to Chinggis Khan. Many of them could claim descent from Chinggis Khan's closest companions and regarded the Mongol Empire as the shared property of those that had helped to build it. Nevertheless, their authority remained controversial and was openly challenged by several of the more powerful Chinggisid rulers of Iran, whose court historians used the derogatory term of qarachu (commoner) to undermine their claim to legitimate authority. The fact that the same title was applied to the non-royal aristocracy of the Golden Horde can be attributed to a similar struggle for power between the Chinggisid rulership and their military aristocracy in the Crimean Khanate.
The first serious attempt to trace the origins of the qarachu begs in the early Mongol Empire was made by Uli Schamiloglu in a PhD dissertation and then an article, which appeared in 1984.Footnote 2 At that time, Schamiloglu observed that historians had employed two key methodologies in studying the genesis of the qarachu begs, which I shall term the ‘etymological’ and ‘functional’.Footnote 3 The etymological approach consists of tracing the titles ascribed to the qarachu in the sources that document both the Mongol Empire and its successor states across Eurasia. This methodology has proved less popular since, as Schamiloglu observed, the many languages used to document the history of the Mongol Empire make it difficult to trace the use of any given term. For example, the Turkish, Arabic and Russian sources that reported on the Golden Horde refer to senior commanders by the alternate titles of amir, beg and knyaz, making it very difficult to determine whether they are referring to the same office. Historians such as Christopher Atwood and Schamiloglu himself have sought to bypass the linguistic barrier to comparative analysis by studying the function and duties of the qarachu begs to find possible antecedents. Atwood's research proved particularly fruitful with regards to the Yuan Empire, where he noted that several of the functions later attributed to the qarachu begs were performed by the captains of the kešik (plr. Kešikten; imperial guard), known as ötögü.Footnote 4 Yet Atwood admits that the two offices still differed significantly and whereas the qarachu represented the most influential families of the Crimea and regularly circumvented the khan's will, the ötögü were often barely visible members of the khan's entourage, who retained unflinching loyalty to the khan.Footnote 5 What follows is an attempt to fill the historical gap between the early ötögü and their transformation into the qarachu begs of the Crimean Khanate.
The Īl-Khān Kešik and the Ulus Amirs
Previous studies of the qarachu begs in the Īl-Khānate have uncovered several possible antecedents for the office. Most notably, Schamiloglu identified the qarachu begs with the ‘ulus amirs’ described by the Syrian encyclopaedist, al-‘Umarī (d. 1349).Footnote 6 According to ‘Umarī there were four ulus amīrs at the Īl-Khān court, one of whom served as the supreme commander, or ‘beklarbeki.’ He noted that these ulus amirs, also known as ‘great commanders’ (noyanān-i buzurg) held the highest military offices in the Īl-Khānate and that they were required to ratify all decrees by affixing their seal below that of the sultan, but above that of the vizier (chief-minister).Footnote 7 More recently, Atwood and Melville have suggested that the ulus amirs were, in fact, the ötögüs of the reigning Īl-Khān based on the similarity of their functions within the court.Footnote 8 Already it seems that the ulus amirs are a perfect match with the qarachu begs of the Crimean Khanate. But this highly promising beginning to research on the qarachu begs in the Īl-Khānate has been encumbered by the rigid parameters set for analysis. The determination of Schamiloglu and others to argue that there were ideally only four qarachu begs at any one time has stunted a more nuanced study of the office, since there were rarely ever four claimants to the position. Moreover, the title of ulus amir is employed both infrequently and inconsistently by the Persian sources, which use a variety of other terms to refer to the most senior military elite, among them noyan-i buzurg, amir-i muʽazzam, and arkān-i dawlat (great commander/pillars of state). Yet perhaps the biggest hurdle to a more detailed study of the qarachu begs in the Īl-Khānate has been the failure of previous studies to acknowledge change within the institutions of the Mongol court as it adapted to life in the Middle East. Until now, most studies of the qarachu begs in the Īl-Khānate have sought to explain away inconsistencies in their number, function and titles instead of embracing them as evidence of change within the Mongol state structure. A more flexible approach is required if these challenges are to be overcome.
A brief survey of the ulus amirs reveals that their power was not derived from service within the ruling khan's kešik. The title of ulus amir is most often encountered in the first half of the fourteenth century, particularly during the reign of Abū Saʽīd Ba'atur Khan (1319-1335). Those said to have held the title during this period included the young prince's atabeg (protector), Sevinj, his paternal cousin, Hasan Buzurg, Sharaf al-Dīn Mahmūdshāh Injü, who managed the khan's demesne lands (injü), and Amir Chupan, a senior commander, famed for his military prowess and bravery.Footnote 9 Of this group only Hasan Buzurg could tentatively be placed amongst Abū Saʽīd's kešik, since Sevinj served in the household of his predecessor, Öljeitü (r. 1304–1317) and Sharaf al-Dīn Mahmūdshāh was a protégé of Chupan.Footnote 10 For his part the Jalayirid historian, al-Ahrī, states that Chupan was granted the title in 1303 when he displayed great heroism against the Mamluk army in al-Shām. Chupan was rewarded for his courage with the title of ‘amir-i ulus’ and had his name inscribed upon the yarliqs (decrees) of the Īl-Khān.Footnote 11 He seems to have retained the title throughout the rule of Öljeitü (r. 1304–1317), and then under Abū Saʽīd, during whose reign he virtually usurped the throne.Footnote 12 The fact that Chupan, Sevinj and Mahmūdshāh Injü all served different masters prior to entering Abū Saʽīd's court means that they were unlikely to have served in his kešik for any significant amount of time.
In most instances the ulus amirs are specifically mentioned as forming a separate group to the Īl-Khāns’ kešik. This division was observed by the scholar-bureaucrat, Ḥamd Allāh Mustawfī, who described the court of the second Īl-Khān, Abaqa (r. 1265–82), as consisting of the ‘pillars of state,’ made up by his father's senior commanders, and his inaqs (companions), who had served in Abaqa's household prior to his enthronement.Footnote 13 Similarly, the provincial tax-official and historian, Waṣṣāf-i Ḥaḍrāt, provides a list of the noyanan-i buzurg (i.e. ulus amirs) at Öljeitü’s court, of which only two, Sevinj and Essen Qutlugh, are mentioned as part of his suite prior to his coronation. The remainder, including the supreme commander, Qutlughshāh, Chupan Noyan, Bolad-Chingsang, and Husayn Küregen, all served as ‘pillars of state’ during the reign of the previous ruler, Ghazan (r. 1295–1304).Footnote 14 Even Geikhatu Khan (r. 1291–5), who presided over one of the most turbulent periods in Īl-Khān history, could blame much of his troubles upon the fact that his ulus amirs belonged to previous rulers and rival factions. Both his naʽib (deputy), Shiktur, and one of his great commanders, Taghachar, had both held offices in the kešik of the second Īl-Khān, Abaqa (r. 1265–82), whilst his supreme commander (amir al-‘umara), Aq-Buqa, was also named as one of the muqarbān (intimates) of Abaqa.Footnote 15 These commanders attained their powers and offices well before the enthronement of the above-mentioned Īl-Khāns and could not, therefore, have been the ötögüs of their personal guard. Rather, they owed their offices and wealth to the appointments and political hierarchies of past rulers.
The entrenchment of the senior kešikten in the offices they held under previous rulers was, however, only one aspect of their political authority. Offices and property were generally transmitted on a hereditary basis within the Mongol Empire, thereby allowing guardsmen to form their own aristocratic dynasties alongside those of the Chinggisid rulers.Footnote 16 The power that these families commanded caused even the Īl-Khāns to seek their favour through marriage alliances and prestigious appointments, which in turn only increased their authority further. These families became so notable that even the fervent royalist, Rashīd al-Dīn Faḍl Allāh, felt obliged to record their lineages alongside those of the Chinggisids in his Jāmiʽ al-Tawārīkh (Collected Histories).Footnote 17 The independence which their hereditary wealth and power afforded these aristocratic families can be seen in the example of Qurumshi Küregen, a senior commander who attempted to seize control of the Īl-Khān court in 1319. Qurumshi's ancestor, Quyidu, was a commander in the Kereyid army of To'oril Ong Khan, but defected to join Chinggis Khan around 1204. Quyidu was rewarded for his support with the command of 1,000 soldiers, which passed to one of his sons, Qurtqah, after his death.Footnote 18 Quyidu's eldest son, Tügür, served as a secretary (bitikchi) and commanded 100 soldiers in the kešik of Chinggis’ grandson, Hülegü when the latter came to Iran in 1256. After Tügür's death, Hülegü granted his son, Alinaq, the command of 1,000 soldiers.Footnote 19 Alinaq's power increased due to his support of the third Īl-Khān Aḥmad Tegüder's bid for the throne in 1282, for which he was awarded pastures in Georgia and the hand of the Īl-Khān's daughter in marriage.Footnote 20 Alinaq was murdered along with Aḥmad Tegüder during a coup in 1284, but his son Qurumshi was not prevented from inheriting his property or soldiers.Footnote 21 His proud ancestry, enormous wealth, and powerful army rendered Qurumshi an important figure in the latter Īl-Khānate, so much so that when he was ordered to join the forces of Chupan Noyan to repel an invading army from the Golden Horde, he refused, saying, “our fathers never served under the banner of his [i.e. Chupan's] father, indeed they were one-hundred times greater than him. Today we would choose death rather than endure his rule.”Footnote 22 We can, therefore, say that, although Qurumshi's ancestor Tügür Bitikchi served in the kešik of Hülegü Khan, by the early decades of the fourteenth century his family had become a rival source of power to the Īl-Khān throne. Qurumshi's authority, like that of the other Īl-Khān magnates, was based upon the hereditary transmission of offices, property, soldiers and prestige, which reduced his dependence upon the Chinggisid rulers.
The fact that effective power no longer resided in the ruling khan's kešik, but an autonomous military aristocracy, suggests a transformation of the ‘patrimonial’-style of government practiced by the early rulers of the Mongol Empire into a quasi-feudal state during the early decades of Īl-Khān rule in the Middle East.Footnote 23 Further analysis of this transition will no doubt help us to better understand the emergence of the qarachu begs in the first half of the fourteenth century.
The Emergence of the Qarachu Commanders
The centrality of the ruling khan's household to the government of the Īl-Khānate endured for only a short time before a group of senior commanders managed to entrench themselves at the top of the political hierarchy. This transfer of authority was, to some extent, only natural given the prominent role of non-Chinggisid commanders in the creation and administration of the Īl-Khānate. The territories that would later constitute the Īl-Khānate were initially ruled by a series of military governors, drawn from amongst Chinggis Khan's early followers, and was only later settled by a small number of Chinggisid princes midway through the thirteenth century.Footnote 24 Yet the civil-war and political executions that followed the death of the Great Khan (Mong. qa'an) Möngke in 1259 further depleted the contingent of royal princes in the Middle East.Footnote 25 The relatively small number of Chinggisids residing in the Īl-Khānate meant that from the very beginning the non-royal noyat played a much more prominent role in the administration of the Īl-Khānate than in any of the other Mongol khanates.
The most prominent members of the Īl-Khān military elite seem to have initially been drawn from the kešik of Tolui, Chinggis Khan's fourth son. Tolui's third son, Hülegü the founder of the Īl-Khānate, brought a contingent of his father's kešik with him when he set out to establish his rule over the Middle East in 1253. Their number included Suqunjaq Aqa, Chupan's kinsman and a descendant of Chila'un Ba'atur, one of Chinggis Khan's earliest companions (nököt). Chila'un's son, Sodon, served as a commander in the right wing of the Toluid army, which passed to Tolui's widow, Sorqaqtani Beki, upon his death.Footnote 26 Suqunjaq was a prominent leader in Hülegü’s early campaign against the Ismāʽīlīs of Maymun Diz towards the end of 1256 and then during the siege of Baghdad in 1258.Footnote 27 When Hülegü died in 1265, Suqunjaq was awarded the government of Fars and Baghdad by his successor, Abaqa. He was described as the amir al-‘umara (supreme commander) and the manager of all government business in Abaqa's state.Footnote 28 Suqunjaq's position was, perhaps, only eclipsed by that of another guardsman, Ilugei Noyan, the son of Jochi Darmala Jalayir, one of the hereditary servants of the Borǰigin family of Chinggis Khan. Ilugei seems to have served in Tolui's guard and later held the title of amir-i ordo (commander of the court/camp) in Hülegü’s suite; a position which Kirakos translates as ‘overseer of the house.’Footnote 29 He, like Suqunjaq, was a senior commander in the right-wing of Hülegü’s army as it marched into the Middle East and he served in virtually every one of the khan's campaigns. His position seems to have been transmitted to his son, Shiktur Aqa, since the latter is mentioned as one of the amir-i ulus during Abaqa's reign and as the naʽib (deputy/agent) of Geikhatu in 1291.Footnote 30 Finally, Hülegü’s vanguard was led by Kitbuqa Noyan, who served as a bavurchi (cup-bearer) in Tolui's guard.Footnote 31 He fought bravely against the Ismāʽīlīs and then in the siege of Baghdad before being entrusted with the protection of al-Shām in 1260, where he was killed only a few months later by the Mamluk army at the battle of ‘Ayn Jalūt.
There were, of course, many commanders drawn from outside the Toluid guard that served under Hülegü in the Middle East, but the most senior positions were given to his trusted family servants. Those who had held positions in Iran prior to Hülegü’s arrival were soon pushed to the periphery in the new order, or otherwise demoted. Baiju Noyan, who commanded a contingent of soldiers guarding the southern Caucasus during the reign of Ögödei, was ordered to uproot and move to Rum so that Hülegü’s arriving army could make use of the pastures there.Footnote 32 Similarly, Arghun Aqa, the former administrator of Iran was sent to Mongolia to answer charges of embezzlement and corruption before being appointed to govern the far-eastern province of Khurāsān.Footnote 33 The families of Baiju and Arghun Aqa continued to wield significant power throughout the history of the Īl-Khānate, but their authority was generally confined to the provincial capitals in which they held offices and property. These older aristocratic families were a constant source of rebellion and instability within the new regime created by Hülegü and his heirs.
Hülegü’s kešik commanders served him loyally throughout his lifetime, but their hereditary privileges and entrenched powers soon became an obstacle for his successors. In this regard the Mongol kešik seems to share much in common with the mamluk slave-soldiers employed by most Middle Eastern rulers during the middle ages.Footnote 34 The Ayyūbid rulers of al-Shām and Egypt, in particular, bought and trained large contingents of Turkish slaves to serve in elite units, which constituted the core of their armies. These ‘mamluk’ contingents were raised in barracks close to the king, often within the walls of his palace, where they developed close bonds of affection and loyalty for their master. Like the Mongol noyat, the mamluks were more than just soldiers. They also served in a variety of roles within their master's court, such as standard bearers, treasurers, and masters of the hunt.Footnote 35 Yet the bonds of fealty and friendship between a master and his mamluk, or even amongst the mamluks of a certain troop, made them a constant source of suspicion for incoming rulers. Competitions erupted between old and new contingents of mamluks within the court of a king, who could never fully trust his predecessor's troops as much as his own. Yet displacing incumbent commanders and office-holders was a dangerous business and could often lead to revolt.Footnote 36
As with the Ayyūbid mamluks, the kešikten of previous Mongol rulers were regularly recycled and incorporated within the administration of incoming rulers. When Ögödei assumed the throne from Chinggis Khan in 1229 he merely changed the captains of Chinggis’ night-guards, whilst retaining the quiver bearers and at least one of the day-shift captains of his father. Similarly, when Güyük came to the throne in 1247 he retained at least half of his father's old guardsmen and even kept one of Ögödei's day-shift captains.Footnote 37 The practice of retaining veteran guardsmen in their former positions meant that there was less competition between the guardsmen of different regimes. But it also resulted in several senior guardsmen entrenching their powers at the summit of the Mongol state.
The emerging power of Hülegü’s kešik commanders was soon demonstrated after his death in 1265. At that time the Ordo had been left almost entirely in the hands of Hülegü’s guardsmen. His eldest son, Abaqa, was absent in Khwārazm, campaigning against the hostile forces of his relative, Berke, and his senior wife, Doquz Khatun, who initially took a proactive role in arranging for the succession, died within months.Footnote 38 Most of Hülegü’s other wives and relatives had remained behind in Mongolia and would not arrive in Iran until 1269.Footnote 39 In the absence of any alternative authority, Hülegü’s household staff assumed one of the functions that would ultimately be attributed to the qarachu begs of the Crimean Khanate, namely appointing and crowning a khan. As the head of Hülegü’s court, Ilugei Noyan assumed the initiative, sending word to Abaqa of his father's death and inviting him back to the Ordo in Marāghah (Azerbaijan) to assume the throne. He received the young prince and informed him of the situation before summoning the other notables of the realm to a quriltai (council) to affirm the succession.Footnote 40
Abaqa had not been summoned to a coronation, but rather to attend a council (quriltai) in which his accession would be acclaimed by the leading members of the Īl-Khān court. I have noted elsewhere that the quriltai was less an elective body than a piece of political pageantry designed to affirm the constitutional makeup and character of the Mongol polity.Footnote 41 The quriltai would affirm the primacy of Chinggis Khan's dynasty over the Empire by appointing one of his family to rule, and the new khan would reciprocate by confirming the military aristocracy in their offices, titles and properties by reaffirming the yasa (laws) and yosun (customs) of the previous ruler. In the case of Abaqa's quriltai, Hülegü’s senior kešikten claimed to be the custodians of the deceased khan's will, which named Abaqa as his heir.Footnote 42 With the nomination complete, Abaqa rewarded the kešikten by confirming them in their former positions. Rashīd al-Dīn informs us that by affirming the yasa and yosun of his father Hülegü, Abaqa ensured that Ilugei Noyan would remain the amir-i ordo, Suqunjaq would retain the governorship of Baghdad and Fars, and Arghun Aqa would continue as governor of Khurāsān.Footnote 43 The Persian sources refer to this ceremonial confirmation of the leading magnates as the ẓabt wa tartīb-i muhimāt-i mamālik (recording and arranging the ordinances of the realm), and it featured prominently in virtually all of the future coronation ceremonies to take place in the Īl-Khānate.
Hülegü’s former kešikten would also play a prominent role in the nomination of the third Īl-Khān, Aḥmad Tegüder, after Abaqa's death in 1282. In this instance, Hülegü’s former guardsmen bypassed Abaqa's eldest son, Arghun, in favour of his brother, Aḥmad, on the basis that he would ‘renew the rulings and orders and agree upon the administration of the yasa’ of his predecessors.Footnote 44 In other words, he would maintain the status quo, whereas Arghun, who had served as the viceroy of Khurāsān, could be expected to bring his own contingent of provincial officials and retainers to occupy positions at court. Yet their hopes were soon dashed as Aḥmad proved himself to be an increasingly despotic ruler. Rashīd al-Dīn reports that Hülegü’s guardsmen, particularly Shiktur and Suqunjaq, felt that they had not been adequately rewarded for placing Aḥmad on the throne.Footnote 45 Aḥmad had also ignored the advice of his senior commanders to continue the conflict with the Mamluks, instead proposing peace with the Mongols’ arch-enemies.Footnote 46 Yet perhaps more important, Aḥmad had begun to deprive Hülegü and Abaqa's guardsmen of their former powers, whilst concentrating authority in the hands of his own household staff and the ostensibly Persian bureaucracy.Footnote 47
Aḥmad's flagrant disregard for the traditional powers of the old kešikten prompted them to depose him in 1284. The coup came when Aḥmad travelled east with his army to suppress an uprising led by his nephew Arghun in Khurāsān. Until that point, Arghun's rebellion had failed to win popular support amongst the leading magnates of the Īl-Khānate. Indeed, Arghun's army had deserted him in the castle of Qilāt in Khurāsān, where he was soon captured by Aḥmad's forces and imprisoned at the royal camp.Footnote 48 According to Quṭb al-Dīn Shirazī, the Īl-Khān's victory made him even prouder in his dealings with the old kešikten. They had initially advised Aḥmad not to move against Arghun and now the Īl-Khān sought to rub their faces in his triumph. He forced one of Abaqa's former guardsmen, Buqa, to publicly admit his poor judgment on the issue and even announced that an inquiry would be held into the conduct of his senior commanders after their return to the Ordo.Footnote 49 But Aḥmad had fatally underestimated the old kešikten. These were no longer the humble quiver-bearers or stewards that had served his ancestor Tolui. They were powerful aristocrats with their own ancient lineages connecting them to Chinggis Khan's companions. They were the commanders of vast armies, who held their titles on the basis of hereditary transmission, not the temporary fancies of an Īl-Khān, and when they saw their traditional powers threatened they determined to depose him. As Aḥmad set out ahead of the main army a group of commanders, led by Buqa, massacred the Īl-Khān's loyalists in their camp, freed Arghun, and sent a contingent to capture the now fleeing monarch.Footnote 50 Aḥmad reached the Ordo ahead of his pursuers, but was arrested by Shiktur and Qarabuqa, both sons of Hülegü’s former commanders who had succeeded to their fathers’ positions. The main rebel force soon arrived at the Ordo, where Aḥmad was tried and convicted of betraying the yasa of his ancestors; a crime for which he was executed.Footnote 51
The deposition and murder of Aḥmad Tegüder was the first regicide to be committed anywhere in the Mongol Empire and decisively demonstrated that the families of Hülegü’s former guardsmen had established themselves as a rival source of power to his descendants. This fact was painfully clear to Prince Arghun, who was nominated to succeed Aḥmad Tegüder on the throne. For the first five years of his reign Arghun was a peripheral figure within his own realm, which remained under the control of the coup leaders. The chief agitator behind the revolt, Buqa, was said to have been ‘awarded every title except that of khan’ and was given authority over both the ‘amirate and the vizierate’ (i.e. both the military and civil administration of the state).Footnote 52 These unprecedented powers afforded Buqa a new title, ‘chingsang,’ which had previously only ever been awarded to the most senior members of the qa'an's Secretariat in China.Footnote 53 He was also entrusted with a tamgha (seal), which he used to make, rather than ratify, official decrees (yarliqs), thereby usurping one of the functions later attributed to the qarachu begs. Indeed, Rashīd al-Dīn observed that it was not long before officials serving in the provinces refused to obey Arghun's orders unless they had been co-signed by the chingsang.Footnote 54 A similar situation existed in the other two major administrative divisions of the Īl-Khānate, namely Arab Iraq and Khurāsān, where Buqa's brother Aruq and his ally Nawrūz b. Arghun Aqa, governed in the ‘fashion of a pādshāh [emperor], rather than an amir.’Footnote 55
Arghun successfully overthrew Buqa in 1289 by manipulating rivalries amongst the senior magnates at his court.Footnote 56 Thereafter, the Īl-Khān assumed a far more aggressive posture towards his leading commanders, placing them under the jurisdiction of his chief-minister, Saʽd al-Dawla and curbing their independence significantly.Footnote 57 Yet his reforms proved deeply unpopular with the military aristocracy. Buqa's former friend and ally, Nawrūz Aqa, led a rebellion against Arghun in 1289, which plunged the east of the realm into chaos until he finally made peace with Arghun's son Ghazan in 1294.Footnote 58 The reaction was equally strong at the royal Ordo in Azerbaijan and when Arghun fell ill early in 1291 his senior commanders leapt upon the opportunity to remove any trace of his reforms. Arghun's household staff were invited to a wine-party by Taghachar, one of Abaqa's former guardsmen, and massacred. With them dead, Saʽd al-Dawla was hunted down in the streets of Tabriz and butchered.Footnote 59 Rumours even abounded that the Īl-Khān had been poisoned by one of his concubines, some said at the behest of the senior commanders, although few details are known about the accusation.Footnote 60 The decisive role of the military aristocracy in Arghun's initial success over Buqa and then his ignominious decline and death speak volumes about the new political environment of the Īl-Khānate. Khans could no longer demand the unswerving loyalty of their senior commanders as they had done from their kešikten. The support of the new military aristocracy had to be cultivated and maintained by the reigning monarch, as the careers of Arghun's successors amply demonstrate.
With the destruction of Arghun's centrally administered government the realm was in danger of devolving into a series of regional powerbases, headed by provincial commanders. Aq-Buqa, another son of Ilugei Noyan, and the commanders based in Rum [Anatolia] favoured the appointment of Arghun's brother, Geikhatu, to the throne.Footnote 61 But his appointment was resisted by Taghachar and the commanders based in Iraq-i ‘Ajam, who feared that Geikhatu would displace them with his Rumi favourites. They preferred the appointment of Arghun's cousin, Baidu, who was based in Baghdad.Footnote 62 Meanwhile, Nawrūz had successfully pushed Arghun's son, Ghazan, out of Khurāsān into Māzandarān and had assumed full control of the eastern Īl-Khānate with the backing of Qaidu Khan, the Ögödeid ruler of Turkistan.Footnote 63 The succession was eventually decided by Baidu, who, fearing the intentions of Taghachar's faction, refused to take the throne. His decision eroded support for the Iraqi commanders and allowed Geikhatu to enter the Ordo at Alataq (in the east of modern-day Turkey) and assume the throne.Footnote 64
Geikhatu had won the crown, but Taghachar and his Iraqi backers remained a constant source of opposition to him. Taghachar was too powerful to simply remove and he remained one of the ‘great amirs’ (‘umara-yi buzurg) of Geikhatu's court. Indeed, the power of the Iraqi commanders may have been one factor in Geikhatu's perplexing decision to leave the Ordo and return to Rum until 1292. His departure only encouraged Taghachar, who offered to seize the realm in the name of another prince, Anbarji; a proposal which the latter refused.Footnote 65 Nevertheless, on his return to Azerbaijan Geikhatu again refused, or was unable, to punish Taghachar, who retained the command of a large army.Footnote 66 Both the power and independence of magnates, such as Taghachar and Shiktur, at the court of Geikhatu mirrors that of the qarachu begs of the Later Golden Horde. These were established aristocrats of property and military influence from outside the household of the Īl-Khān. Taghachar, Shiktur, and even Aq-Buqa, had all served in the kešik of previous rulers and had assumed their offices well before Geikhatu's appointment to govern Rum. Even Geikhatu's overbearing vizier, Ṣadr al-Dīn Aḥmad Khāladī, was a protégé of Taghachar, whom he served during the reign of Abaqa.Footnote 67
The entrenched power of the Iraqi commanders was bound to cause friction as Geikhatu and his Rumi cohort sought to assume control of the state apparatus. The Īl-Khān had already generated anxiety amongst the military aristocracy by granting his Persian vizier, Ṣadr al-Dīn Khāladī, new far-reaching powers and by imprisoning several leading members of the Iraqi contingent.Footnote 68 The Armenian bishop, Stéphannos Orbelian, makes it clear that most of those arrested were guardsmen in the kešik of former rulers, listing their hereditary posts: ‘Eldchita [Elčidai], the keeper of birds and beasts and the chase, Dolata [Dolādai], inspector of food and beverages, Khontchibal [Qonchaqbal], the head of a large army, and many others.’Footnote 69 Matters then came to a head when Geikhatu threatened the life of his former rival, Baidu, in a drunken argument at the Ordo. Baidu returned to Iraq and threw his support behind Taghachar's faction who mobilised their armies against the Īl-Khān. Geikhatu was captured and murdered as he tried to flee back to his powerbase in Rum, whilst Aq-Buqa was arrested by Taghachar shortly after the outbreak of violence.Footnote 70
The Iraqi commanders considered the deposition and murder of yet another Īl-Khān to be a justifiable response to the latter's attacks on their hereditary privileges and property. Shortly after Geikhatu's murder the coup leaders dispatched a decree to the provinces informing the relevant officials that because ‘Geikhatu Khan deviated from the manner of rule and changed the yasa of Chinggis Khan, we removed him with the help of the younger and older brothers and khatuns (queens) and amirs.’Footnote 71 When Prince Ghazan heard the news of Geikhatu's death he sent envoys to the Ordo and they were similarly informed of the military aristocracy's right to appoint and depose khans by Ildar Yarghuchi, one of the coup leaders: ‘we, the aqa (lit. senior) and ini (junior) raised Baidu up to the throne together, and if any are discontent it will lead to conflict.’Footnote 72 These commanders considered Baidu's appointment to coincide with a restoration of their hereditary rights to offices and property, which had been compromised during the reign of Geikhatu. Mīrkhwānd explained that ‘the conflicting ideas and changing of the noyans’ [opinion] towards Geikhatu was caused by the fact that they had been prevented from appropriating and inspecting the wealth and property [of the realm], and since during the time of Abaqa Khan each territory had been placed under the protection of one of his majesty's companions, revenue collection and the rule of the regions was ordered and was protected from the stuff of calamity and the armies were obedient and loyal.’Footnote 73 These sentiments were echoed by Waṣṣāf, who claimed that ‘in the fashion of Abaqa Khan’ the commanders ‘divided [the realm] into several provinces and each one of the commanders was sent to one of these provinces [. . .] and they were independent in every respect.’Footnote 74 The military aristocracy viewed the deposition of Geikhatu as a necessary measure to protect the hereditary privileges their families had enjoyed since the time of Hülegü and Abaqa. These rebels were not simple kešik captains, they were propertied aristocrats whose loyalty the khan could no longer command, but had to earn.
It can, therefore, be seen that by the final decade of the thirteenth century the families of Hülegü’s original kešik had usurped the powers and status later attributed to the qarachu begs. Their authority was first entrenched by Abaqa, who confirmed Hülegü’s kešik in the titles and offices they had held during the reign of his father. In doing so, Abaqa ensured that the original kešikten no longer owed their power to the favour of the ruling Īl-Khān, but rather the political precedent established by Hülegü. By the time of Abaqa's death in 1282 the kešikten’s power had risen to such an extent that they could no longer be classed as humble guardsmen, but rather a military aristocracy, whose authority was built upon the hereditary transmission of the property, offices and armies under their control.
The ‘Qarachu’ in the Sources of the Later Īl-Khānate
One might be forgiven for wondering why, if Hülegü’s kešikten had appropriated the powers later attributed to the qarachu begs in the Crimean Khanate, they were not referred to as such by the Persian sources that documented their careers. The answer is, of course, that the non-royal commanders were known by a variety of grander titles, which suited their status at the pinnacle of the Mongol polity. The term ‘qarachu beg,’ on the other hand, was a derogatory term, coined during the reign of Ghazan Khan (r. 1295–1304), who presided over a short-lived revival of the Chinggisid dynasty's prestige and power. Ghazan's court historians sought to denigrate the ‘common commanders,’ who they believed had risen above their proper stations. Yet Ghazan's reign was only brief and could not halt the rise of the former kešik families. The term ‘qarachu commander,’ therefore, only survived for a short period during the latter Īl-Khānate and died with the last effective Īl-Khān ruler, Abū Saʽīd Ba'atur Khan, in 1335. The title is virtually expunged from the histories of the subsequent Jalayirid and Temürid dynasties, whose non-Chinggisid rulers would have cringed at being reminded of their humble origins. It was only in the Crimean Khanate, that the term was revived as the Girayid monarchs and their non-royal commanders fought their own battle for ascendancy.
A study of the word ‘qarachu’ in the sources penned both during and immediately after the reign of Ghazan suggests that it was most often employed as a derogatory term. A colourful example appears in the Collected Histories of Ghazan's chief-minister, Rashīd al-Dīn, when he describes the invasion of Iran by Baraq Khan of the Chaghadai Ulus. Rashīd relates that a scuffle broke out in Baraq's ranks when a non-royal commander, named Jalayirtai, accused a Chinggisid prince, Qipchaq Oghul, of treating Baraq in a haughty manner. Qipchaq was said to have been incensed by the lecturing of someone he viewed to be a mere commoner and retorted, ‘who are you to come amongst the elders and juniors [of the dynasty]? Which qarachu dares to talk back to the offspring of Chinggis Khan? Who are you, dog, to speak impertinently to me?’Footnote 75 Rashīd al-Dīn claimed that Qipchaq was further outraged by Baraq's failure to adequately punish Jalayirtai and that the dispute eventually led Qipchaq to abandon Baraq shortly before the arrival of the Īl-Khān army. He also states that when another Chinggisid prince, Chapar b. Hurdu, deserted Baraq's army and fled into Transoxiana the latter's governor at Bukhārā refused to apprehend Chapar, insisting that he was only a qarachu and didn't have the right to apprehend a royal prince.Footnote 76 This account differs markedly from the Tārīkhnāma-yi Harāt, of Sayf al-Dīn Harāvī, who drew his evidence from the histories and anecdotes collected in his native Harāt, where the battle between Baraq and the Īl-Khān army was fought. Harāvī fails to record anything of the dispute between Qipchaq and Jalayirtai and even claims that Qipchaq fought bravely during the final engagement with the Īl-Khān army.Footnote 77 Moreover, he omits any mention of the desertion of Chapar, who is completely absent from his version. How are the differences in these two versions to be accounted for? It seems unlikely that the inconsistencies can be attributed to shortcomings on Harāvī’s side, since the latter's one-time master, Malik Shams al-Dīn II of Kart, had personally visited Baraq's camp and initially considered joining the Chaghadaid invasion. His visit, and the state of the Chaghadaid army, would have therefore been well documented in the sources of Harāt, as indeed they were by Harāvī himself.Footnote 78 Rather, the two stories about troublesome qarachu in Rashīd al-Dīn's account were part of a broader attack on the non-Chinggisid aristocracy that permeates his historical writing. The stories served as a subtle warning to his readers, to trust the qarachu commanders at their peril.
The narrative of Rashīd al-Dīn's history is one in which disloyal commanders are constantly seeking to mislead or even undermine their rightful sovereigns. Even in instances when Chinggisid princes seem to have been conspicuously guilty of crimes against the state, Rashīd al-Dīn blames their misdeeds upon devious commanders. The most notable example is Rashīd al-Dīn's account of the conspiracy of the Ögödeid and Chaghadaid princes to topple Möngke Qa'an at his coronation in 1251. The earlier account of Juwaynī merely states that the princes had ‘strayed far from the path of discretion and the acceptance of advice because of the promptings of wicked teachers and the suggestion of impious emirs.’Footnote 79 But Rashīd al-Dīn, who generally copied Juwaynī’s account, seeks to absolve the princes by introducing a new character into the story. He states that when Möngke intended to question the captured conspirators he began by torturing Bābā Kardīdī, the atabeg (protector/guardian) of Shiremün, one of the leading Ögödeid princes. Under duress, Bābā Kardīdī confessed that the princes had been unaware of the plot and that it was the amirs (commanders), who had entered into an alliance to depose Möngke. Rashīd then has Möngke question the leading rebel commanders, who initially deny, but later confirm Bābā Kardīdī’s statement.Footnote 80 His version of the Ögödeid-Chaghadaid conspiracy, therefore, absolves the Chinggisid princes of any guilt and places the blame for their actions squarely upon their commanders. Similar bias is evident in Rashīd al-Dīn's account of Negüder, a Chaghadaid prince who held pastures and revenues in the Caucasus and Georgia and conspired with Baraq Khan during his invasion of the Īl-Khānate in 1269. The Armenian chronicler, Gregory of Akanç, wrote that when Abaqa managed to capture Negüder he publicly humiliated the prince before imprisoning him for the remainder of his life.Footnote 81 This account is echoed by the Mamluk historian, Baybars al-Manṣūrī, who reported that Negüder was held under house arrest, forbidden from coming to court and banned from mounting fully-grown horses or stringing a bow for the rest of his life.Footnote 82 This evidence contrasts sharply with that of Rashīd al-Dīn, who states that Negüder had been misled by his leading commanders, whom Abaqa imprisoned before dividing their armies amongst his own supporters as punishment. Negüder, on the other hand, was simply kept under house arrest until the defeat of Baraq, at which point he was again released.Footnote 83 This hardly credible account, once again illustrates Rashīd al-Dīn's overwhelmingly negative attitude towards the participation of the non-royal aristocracy in the government of the Īl-Khānate.
Rashīd al-Dīn's hostility to the non-royal aristocracy was no doubt born out of Ghazan's own experiences with over-mighty military leaders, which are discussed at length in his history. Ghazan had assumed the throne to restore the prestige and power of the Chinggisid dynasty, which had been usurped by the qarachu commanders. It is, therefore, not surprising that the uprisings against the governments of Arghun and Geikhatu were treated with particular severity by the Jāmiʽ al-Tawārīkh.Footnote 84 Rashīd al-Dīn, probably deriving his information from Ghazan himself, claimed that the prince was outraged when he heard the news that his uncle, Geikhatu, had been deposed by his leading commanders. He wrote an angry letter to Baidu demanding justice and denouncing the actions of the qarachu commanders, saying: ‘it was never in the yasa of Chinggis Khan that the qarachu ‘umara (common commanders) could harm his uruq (family). But now a group of qarachu have killed Geikhatu!’Footnote 85 Here Ghazan uses of the word qarachu to assert the primacy of the Chinggisid line, which had been enshrined in the jasaq of Chinggis Khan. Violation of this law represented a departure from Chinggisid tradition and had to be punished severely. Indeed, his first act after assuming the throne was to hold a trial for the commanders accused of taking part in the uprisings against Arghun and Geikhatu.Footnote 86 The most prominent victim of these trials was Qonchaqbal Noyan, the commander of the central army (qol) and a driving force behind the purge of Arghun's officials in 1291. He, like the other victims of the trial, was subjected to the taunts of Ghazan's supporters as he was questioned; “did you imagine that no one remained from the line of Arghun Khan to defend against your evil and misdeeds?”Footnote 87
Rashīd al-Dīn's deprecatory use of the term qarachu was adopted by other sources after the death of Ghazan in 1304. Again, the term only ever appears when the authors seek to draw a comparison between the illegitimate authority of the non-royal commanders and the legitimate authority of the Chinggisid dynasty. Perhaps the most notable example of such use comes from the Tārīkh-i Waṣṣāf, where the term appears on at least two occasions. The first comes from Waṣṣāf's account of the early years of Arghun's rule, during which time the commander Buqa exercised supreme power over the state.Footnote 88 In addition to listing the many humiliations to which Arghun was subjected by this former guardsman, Waṣṣāf summed up the degradation of the Chinggisids in a story about the arrival of a royal princess, Abish Khatun, at the Ordo to answer accusations that she had murdered one of Buqa's officials. When she reached the court Buqa had her attendants confined to the city square and refused her permission to see the Īl-Khān. Abish was then brought before Buqa, who sarcastically chastised his chamberlain for having forced her, ‘a khatun of the khan,’ to be brought before him, a mere ‘qarachu amir.’ Buqa subsequently had Abish's attendants flogged and tortured until they confessed to murdering his agents in Shiraz; a conviction which ended the political career of Abish Khatun.Footnote 89
Chupan Noyan, who dominated the court of Ghazan's nephew, Abū Saʽid, seems to have found the term less humorous than his predecessor Buqa, as neither he, nor his son's panegyrist, Karīm al-Dīn Āqsarāyī make use of it.Footnote 90 The term was, however, used against him by his enemies on at least one occasion, when Özbeg Khan of the Golden Horde launched an abortive invasion of the Īl-Khānate in 1319. Waṣṣāf noted that during the first confrontation between the two armies Özbeg lamented the absence of Chupan, who was at Pilqan at the time, saying that he had heard that ‘Chupan Qarachu holds the padshah’s throne.’Footnote 91 Indeed, Chupan's seizure of power seems to have been one of the justifications used for Özbeg's attempted conquest of Iran, since he withdrew without fighting the main Īl-Khān army, led by Abū Saʽīd, saying that “the one we are after [i.e. Chupan]” had moved to the north.Footnote 92 In any case, there is little doubt that Özbeg's use of the word qarachu in Waṣṣāf's account was again intended to highlight the fact that Chupan was not of the Chinggisid royal family and, therefore, held a subordinate position to Özbeg.
The term qarachu even seems to have found its way into discussions about legitimate spiritual authority at the Īl-Khān court. Ghazan's claim to be a Muslim ruler was an important component in both his rule of the former heartland of the ‘Abbāsid Caliphate and his ambition to conquer the Mamluk-held territory of al-Shām.Footnote 93 He was, therefore, highly sensitive to the often egalitarian nature of Islamic spiritual authority. It was widely believed that Ghazan had been converted by his one-time enemy and non-royal commander, Nawrūz, who had initially rebelled against his father Arghun, proclaiming a jihad against the heathen Īl-Khāns.Footnote 94 The idea that a common commander could convert, let alone rebel against, his Chinggisid master was repugnant to Rashīd al-Dīn, who informed his readership that ‘after investigating the matter it became apparent that this belief was incorrect.’Footnote 95 Rather, Rashīd claimed that God had moulded the Īl-Khān's essence and ‘placed the ability [to find] salvation at the centre of his being and instinct.’Footnote 96 Rashīd even had Nawrūz admit that ‘it is not customary for a qarachu to give biliks [advice] to an uruq [member of the Chinggisid clan]’ before speaking to him about the Faith.Footnote 97
Both Ghazan and Rashīd al-Dīn wanted to ensure that Islamic doctrine reinforced their political regime as much as possible. To this end, they were eager to stress that spiritual authority, like political authority, was transmitted through blood. Hence, Rashīd al-Dīn informed his readers that it was folly to believe that leadership of either the Faith or the state ‘could be achieved through effort and struggle alone without worthiness, righteousness, guidance or seniority.’Footnote 98 The extent to which Ghazan's conception of spiritual and political authority merged can be seen in a story told by Muʽīn al-Dīn Naṭanzī. He claimed that when Ghazan enquired about the tenets of Sunni Islam and was informed that the Prophet Muḥammad had been succeeded by his companions, Abū Bakr, ‘Umar and ‘Uthmān, the Īl-Khān was enraged and bellowed that ‘the qarachu have imposed this view to serve their own purposes. The true madhhab is the one belonging to the descendants of Muḥammad.’Footnote 99 In this instance Ghazan uses the word qarachu to distinguish between the legitimate authority of the Prophet's family and the supposedly usurped power of the latter's Companions. Ghazan's brother, Öljeitü, took the analogy one step further. When he was informed of the imamate of Abū Bakr he retorted that ‘in the time of my forefathers there were amirs who were older than us and who were companions of my father and whom he had praised, who perhaps should have taken the throne.’ All those present agreed that this would have been a terrible crime and that ‘Alī b. Abū Ṭālib (the Prophet's cousin and son-in-law) had been Muḥammad's true heir and successor.Footnote 100 Ghazan and Öljeitü regarded both political and spiritual authority to be transmitted by blood. Those who lacked the necessary pedigree were to be excluded from achieving supreme power.
Use of the term ‘qarachu commander’ severely declined after the death of Öljeitü in 1317. His son and heir, Abū Saʽīd, was only twelve years old at the time of his coronation and his leading magnates assumed power over the Īl-Khānate for most of his reign. These magnates, mostly descended from the former guardsmen of Hülegü, preferred to be known by less derisive titles. Karīm al-Dīn Āqsarāyī, a provincial bureaucrat in Īl-Khān-ruled Rum, composed a history of the region, which he dedicated to Chupan's son, Temürtash sometime before 1328. In his work, Āqsarāyī uses a variety of titles to refer to Chupan, who at the time was serving as the commander in chief; amongst them noyan-i a'zam (great commander) and amir-i mu'azzam (great commander), in addition to the more grandiose ‘khusrāw jahāngusha’ (conquering king).Footnote 101 Similarly lofty titles can be found in the Jalayirid histories of al-Ahrī and Nakhjivānī, but these were too verbose to include in general discussion and usually gave way to the more simple form of ‘amir.’Footnote 102 Indeed, the Temürid court historian, Naṭanzī, characterised the deposition of the last effective Chinggisid ruler of the Ulus Chaghatay, Ghazan Sultan, and his replacement by a series of common commanders with the brief statement that ‘the sultanate became an amirate.’Footnote 103 In short, the great magnates who came to dominate the Īl-Khānate both during and after the reign of Abū Saʽīd ensured that the demeaning and servile terminology of Ghazan's reign was replaced by a new more grandiose lexicon, which more adequately reflected their true position.
The title ‘qarachu commander,’ with all its implied servility and debasement, was, however, revived in the Golden Horde. Beatrice Manz has suggested that the title probably appeared on the Pontic Steppe during the reign of Özbeg Khan, who tried to concentrate power in the hands of his household staff.Footnote 104 It is entirely likely that Özbeg, seeking to marginalise the commanders who supported his rival, Elbasmish, may have introduced the distinction between legitimate Chinggisid authority and the illegitimate power of the commanders found in the neighbouring Īl-Khānate.Footnote 105 The period of unrest that followed the death of his successor, Jani Beg Khan, and the dominion of the non-Chinggisid commander, known as Mamay, may have provided equally suitable circumstances for the reintroduction of the former terminology.Footnote 106 Such ambiguities about the precise date of the revival will remain, but it is certain that the terminology did return after the establishment of the Girayid dynasty in the Crimea during the 1440s. The Girayids struggled constantly with the leading members of their aristocracy for ascendancy over the khanate. Indeed, one of the most authoritative sources on the qarachu begs, Remmel Khodja, was a close companion of the then Crimean ruler, Sahib Giray Khan (1532-51), and his history provides a detailed account of the suppression of the qarachu begs’ powers.Footnote 107 The purge of the qarachu carried out by Sahib Giray and his brother and predecessor, Saʽadat, would have provided fertile ground for the reintroduction of the demeaning terminology employed by Ghazan Khan's court historians against the military aristocracy.
This paper has suggested that the transition of the kešik guard-captains (ötögüs) into the qarachu commanders of the later Īl-Khānate was symptomatic of a broader transformation of the Mongol Empire in the Middle East from a patrimonial state, headed by the khan's household, into a quasi-feudal state, dominated by an independent military aristocracy. This process began shortly after Hülegü’s death when his heir, Abaqa, confirmed his guardsmen in the offices and powers they had held during the reign of his father. By entrenching Hülegü’s senior commanders at the top of the Īl-Khān administration, Abaqa transformed them into a propertied aristocracy, who owed their titles, wealth and privileges to the precedent set by the Īl-Khān founder and not the favour of the ruling khan. The power of this new military elite grew so great that they were able to depose Abaqa's successor, Aḥmad Tegüder, when the latter threatened their positions in 1284. Indeed, the period between 1284 and 1295 saw no fewer than five Īl-Khāns appointed and deposed by regional and political factions within the military aristocracy. A brief revival of the Chinggisid dynasty during the reign of Ghazan saw his court historians brand these old kešik families as qarachu commanders, who had simply exceeded their authority. But both the terminology and the ideology surrounding it declined with Ghazan's death, only to be restored in the Crimean Khanate when the Girayid dynasty faced a similar challenge from its own qarachu commanders.
