The Formation of Turkey

The Seljukid Sultanate of Rum: Eleventh to Fourteenth Century

Claude Cahen

• Publisher: Routledge
• Year: 2014
• ISBN:
  9781317876250
• Edition: 1
• Topic: History
• Subtopic: Eastern European History
• Index: History

From Byzantium to the Mongols to the Sultans of Rum, this acclaimed book offers an important insight into the evocative history of Turkey before the coming of Ottoman power.Turkey forms a historical bridge between Europe and Asia and as such has played a pivotal role throughout history. The rise of Constantinople and the later Ottoman Empire are well known: less well understood are developments in the three centuries in-between. What led to the decline of the Byzantine Empire and what happened in the intervening years before the rise of the Ottomans? Translated from the original French, this classic work examines the history of the Turkey that eventually gave rise to an imperial power whose influence spanned East and West.

Part One
General History of Turkey before the Mongols

Chapter One
Settlement of the Turks in Asia Minor to 1107

The battle of Manzikert may be regarded as the act which gave birth to Turkey. Not that Alp Arslan wished to conquer Byzantine Asia Minor: he set his august captive free for a ransom, a promise of alliance and some rectification of the frontiers. He did not think it possible to bring the Empire of Rum to an end, and no doubt it was not an unfavourable prospect that the Greeks, left free to fight the Turcomans on their territory, would thereby limit the danger that they presented to himself. Nevertheless the conquest of Anatolia ensued. No administrative or military structure was in a condition to oppose it, even if the Turcomans themselves had no precise political aim at that time. Nothing obliged them to return their accumulated booty to the East. The Byzantines themselves caused them to penetrate further than they would have wished, and opened places to them that they might not have taken on account of their partisan struggles. Moreover the greater part of the population in the eastern half of Anatolia was not Greek but Armenian or Monophysite, little attached to the Byzantine Empire, and sometimes, in view of the weakness of the Christian armies, preferring to submit to the invader for better or worse.

The disorganization of Byzantine Anatolia, and the yet primitive organization of the Turcomans, explain the incapacity of the chroniclers of this period to report even the major facts. Those of the year 1073, of which something is known, give a good idea of it. In this year Roussel of Bailleul, the leader of the Norman mercenaries, sent to Caesarea (Kayseri) with Isaac Comnenus, abandoned his chief, who was then captured by the Turks, while his brother, Alexius, had great difficulty in bringing the army back west of Ancyra (Ankara). The Emperor Michael VII Ducas sent his uncle John Ducas against Roussel, who took him and proclaimed him emperor as a rival to Michael. John then called upon the Turkish band of Artuk. Near to Nicomedia (Ịzmit) Roussel and John Ducas were captured, but Roussel was liberated by Artuk (against the wishes of Byzantium), and withdrew to the east. Alexius Comnenus paid another Turkish chief to deliver Roussel to him at Amasea (Amasya). This did not signify that the way to the west was open, for in order to escape Alexius had with great difficulty to take ship at Heraclea (Ereǧli). At the same time the presence of the Turks was noted near Miletus (Balat) and Trebizond (Trabzon).

Little is known of what happened in these regions in the five or six following years. It can be certain only that the Turkish bands remained there, sometimes mixed with Norman or Armenian contingents left from the Byzantine army but on terms with the Turks. On the other hand, it is in this period that the sons of Kutlumuş appeared on the southern flank of Anatolia. One of them was to be the ancestor of the entire Seljukid lineage of Rūm. Their father, Kutlumuş, had been the chief of a branch of the Seljukids, and often in revolt against his cousins, the Great Seljukids, by whom he had been put to death. What had happened to his four sons during Alp Arslan's sultanate is unknown, but a semi-legendary tradition, which seems to be confirmed by subsequent events, shows them in a condition approaching captivity on the middle Euphrates. When Alp Arslan died, they succeeded in escaping, unless indeed the young new ruler, MalikShāh, facilitated their escape in order to make peace with them. However this may be, a number of Turcomans now gathered around them so that they became a significant factor in regional politics. In 1075 the Turcoman Atsiz, the master of Palestine in conflict with Egypt, had to deal with an alliance of a rival at Acre and the troops of two of the sons of Kutlumuş, on whom he had called for aid. Atsiz defeated them and, proclaiming his loyalty to Malik-Shāh, he sent them to him as prisoners. The other two sons, who had stayed behind, withdrew to southern Anatolia, where their standing as Seljukids no doubt helped them to reconstitute an important, quasi-official force; but their attitude in 1075 and subsequent years proves that they were still adversaries of the Great Seljuks.

Once again Byzantine quarrels contributed to their fortune. In 1078 the general Nicephorus Botaneiates, in revolt against the Emperor Michael VII, denuded Asia Minor of its remaining Byzantine troops. Through the Comneni family he won for his cause Erisǧen (alias Chrysosculos), a relative of Malik-Shāh, who had fled to Constantinople and become a Byzantine. Michael appealed to the sons of Kutlumuş against Nicephorus, who escaped them, and won them over through Erisǧen's mediation. They declared fealty to him as their sovereign, so becoming full-blooded Byzantines. In fact they remained independent, with the tiresome circumstance for the Byzantines that they held all the territory around Nicaea (Ịznik) and the Asian shore of the Bosphorus. They were to stay, for their Emperor Nicephorus III had to struggle against a pretender in Europe, Bryennius. Nicephorus obtained help to fight him from the sons of Kutlumuş, whom of course he had to pay. Shortly afterwards Nicephorus Melissenus, supported almost solely by the Turks, revolted in Asia and won over the sons of Kutlumuş in his turn. Their party was of little importance to them, since profit was to be made from either side. Melissenus opened to them the towns which they had been unable to occupy, and these became their bases of operations henceforth. The arrival of another pretender, Alexius Comnenus, changed nothing for the moment, for he had to deal with a Norman attack in Europe, against which he called the same Turks to aid.

Naturally nothing of all this was by Malik-Shāh's instructions. He had no interest in seeing his dangerous cousins obtaining power of any kind. On the contrary, he sent to Constantinople to demand the capture and return of the two surviving sons of Kutlumuş, Manṣūr and Süleyman - a demand backed by an army under the command of his general, Bursuq (Porsuk), who killed Manṣūr but failed against Süleyman and his men. About this time the usage appeared, entirely unofficially, of styling Süleyman 'sultan'. In brief, the Greeks had created a Turkish sultan within the Byzantine Empire.

Süleyman must nevertheless have realized that to keep his new-found power he was too much at risk in the west, and must maintain his relations with the Turcomans of eastern Anatolia. So he turned back to the east, there also aided by his more or less clear new position as deputy for the Byzantines. Alexius Comnenus, since 1081 the Emperor Alexius I, was not unhappy to encourage him in order to keep him at a distance, and to dispose of the Armenian principality in the Taurus region and northern Syria held by Philaretus, who had aided his rivals. Süleyman passed into Cilicia, then took Antioch (Antakya) in December 1094. Here two significant facts occur: for the appointment of a qāḍī in Cilicia, he turned to the ruler of Tripoli, a Shī'ī hostile to the Great Seljuks; and in Antioch he favoured the native Christians against the unpopular Byzantine Church. Unfortunately for him, he could not avoid involvement in the quarrels of Syria, where he met his death in 1086. Malik-Shāh had no difficulty in occupying almost the whole of Syria-Palestine.

So after a long period of apparent lack of interest, Malik-Shāh was able to concern himself with Asia Minor, where Süleyman's ever-hostile power could only have given him offence. When Süleyman had moved away from western Anatolia, he left lieutenants there such as Abu'l-Qāsim on the Straits and Hasan Buldaci in the central Taurus and Cappadocia. Playing off one of these Turks against the other, Alexius Comnenus tried to control their activities on the Straits and the Aegean coast.¹ Without hazarding himself in Asia Minor Malik-Shāh applied himself to the establishment of a solid base extending from Ādharbāyjān to the middle Euphrates for the incursions of his lieutenants, Sarhang-Savtekin, then Buzan and others. About 1090 he sent an army, again commanded by Bursuq, preceded by an ambassador who was commissioned to offer to Alexius the withdrawal of the Turks from all the coastal regions in return for an alliance. Alexius was mistrustful, and perhaps he did not clearly distinguish the Turks dependent on Malik-Shāh from the others. He won over the ambassador, whom he created a Byzantine dux, an equivocal position which enabled him to regain Sinope (Sinop) in the sultan's name. But when Bursuq arrived, Alexius considered him a more dangerous invader than Abu'l-Qāsim, whose demand for reconciliation he accepted. He received him at Constantinople, and sent him to Nicaea against Bursuq, who had to withdraw.

Malik-Shāh did not however give up hope of bringing about the submission of the Turks of Asia Minor with the help of Alexius, the sole means of obtaining it at that distance. He did not covet any Byzantine territory, but he considered himself the chief of all the Turks wherever they were. In 1092 he sent Buzan to Anatolia with the twofold mission of reducing Abu'l-Qāsim, and of offering Alexius an alliance sealed by the marriage of a son of the sultan to a daughter of the emperor: Byzantium would recover all the territories left to Süleyman, and the military help he might need. Abu'l-Qāsim decided to go and make his peace with Malik-Shāh, but was taken on the way and strangled. Alexius could not as a Christian accept the marriage proposal, but he responded politely to the sultan by sending an embassy in return. Unfortunately Malik-Shāh had died before it reached him, and the sultan's army left Anatolia. In fact the situation in the country hardly changed. Hasan Buldaci inherited Ịznik, Süleyman's young son, Kilic Arslan, lately captured at Antioch, escaped, and gathered around himself the Turks who had once been his father's subjects. While he himself seemed to be occupied above all with securing his links with the east, his lieutenants such as particularly Çaka, with the more or less freely rendered help of the native sailors, occupied the great ports and the Aegean isles Clazomenae, Phocaea (Eskifoca), Smyrna (izmir), Chios, Mitylene and Samos. Alexius succeeded in retaking these places from him, but not in curtailing his piratical activities in the Dardanelles, which were all the more redoubtable in that they enabled Çaka to enter into relations with the Petcheneg Turks of Europe, who encircled Constantinople. Alexius then resorted to trickery: he made peace with Kilic Arslan, and led him to fear the ambitions of Buldaci, who had become his father-in-law. Kilic Arslan invited Buldaci to a banquet where he was killed, without on the whole any great change ensuing. If it had not been for the imminent arrival of the crusaders, the Turks might perhaps have established principalities along the coasts similar to those which came into existence in the fourteenth century; the attachment which was shown to the retention of Nicaea also shows the role which the Turks intended to play in Byzantine affairs. At the same time, the quarrels which faced Malik-Shāh's successors made risky their intervention in Asia Minor.

It was probably in the same period that on the fringes of the possessions of the Seljukids of Iran, zones commanded briefly by the Seljukid Yāqūtī, then by his son Ismā'īl, other principalities took shape; perhaps not just at this point those of the Saltukids of Erzurum or the Menguçekids of Erzincan (although family legend made them present at the battle of Manzikert), but certainly the principality of Danişmend.

It is not known when or how the principality of Danişmend began to be formed. There is an explicit reference to the family only c. 1095 on the eve of the First Crusade, the narrators of which tell of a certain Danişmend. The name, Persian by etymology (dānishmand, wise, learned), conjures up a kind of social and religious chief, and in spite of later legends it is difficult to believe that a chief of Turcomans could be other then a Turcoman himself. But there is no clear knowledge of the family's first activities. It is likely that Danişmend, properly named Tailu, is the one who must have died in 1104. His successor is sometimes called Gümüştekin, a Turkish name, or Muḥammad, an Arab name, or Amīr Gazi, which evokes his martial qualities. They were established in central Asia Minor from Sivas (Gk. Sebastea) to Cappadocia, perhaps at Ankara; for even if the Turcomans were but little sedentarized as yet, they nevertheless tended to group themselves regionally under one or two chiefs. Danişmendid opposition to Kilic Arslan had less to do with the nature of their troops, at this time Turcomans on both sides, than with the fact that Kilic Arslan based his pretensions, which were moreover often recognized, upon his being a Seljukid. In spite of his hostility to the other Seljukids, this gave him the idea of creating for himself a political structure of the Irano-Turkish type like that of the sultanate of his kinsmen. Their relative opposition did not prevent marriage-links between the Danişmendids and the Seljukids of Rūm.² Apart from the Seljukid ruler, Danişmend was the only great Turkish chief known to the chroniclers of the Crusades, who call Kilic Arslan 'Soliman' like his father, although this double name is unknown in the Arabic sources.

The central Taurus region was where the men of Kilic Arslan and Danişmend lived side by side, freely interpenetrating without the possibility of the tracing of a frontier between them. Both rulers aspired to the possession of Malatya (Gk. Melitene), the principal route-centre and strategic keypoint of eastern Anatolia on the way to Upper Mesopotamia. Kilic Arslan directed his activity in this direction, for now he could attach only a reduced importance to the Straits and the Aegean region, and he did not seek to become involved with Alexius Comnenus; hence the way he held to south-west Anatolia. Certainly he left wife, children and treasure at Nicaea, but as a secure base as much as a capital. In these conditions he heard of the arrival of the crusaders, to whom however he had at first no reason to attach more importance than to the Norman contingents with whom the Turks had long been acquainted in the country, sometimes as enemies, sometimes as accomplices. It is well known how the impatience of Peter the Hermit's men resulted in their massacre by the Turks of Nicaea in 1096, an event which contributed to their idea that the crusade was a trivial affair. The crusade proper was thus able to cross the Bosphorus and reach Nicaea without Kilic Arslan receiving notice in time to save the town or his family, which was however well treated by the Byzantines, who knew that they might need them in the future as in the past. Kilic Arslan then tried to intercept the crusaders on their entry upon the Anatolian plateau near Dorylaeum (Eskişehir), having obtained help from Danişmend. It was a very hard-fought battle, but in the end Kilic Arslan was beaten, and gave up the defence of western Anatolia. The Turks were after all still seminomads, and to be driven back or dispersed did not signify their destruction, any more than the Franks' march across the plateau signified their wish to occupy the country. Even the Greeks in the rear could only regain possession of the western valleys. The skirmishes which took place when the Franks reached the Taurus, where the Turks were trying to destroy some Armenian lordships, did not modify the course of events. The crusaders made for Edessa and Antioch, which they were to conquer. These cities did indeed belong to Turks, but to dependants of Malik-Shāh's heirs; the Turks of Anatolia made no attempt to help them.

Danişmend had been much less affected than Kilic Arslan by the passage of the crusaders, and he remained in a position to renew the attack on Malatya, where perhaps for a moment under the very vague suzerainty of Süleyman or Kilic Arslan, the Armenian Byzantine governor, Gabriel, had maintained himself in power, or been reinstalled. He appealed for help to one of the chief crusaders, Bohemond, who had become prince of Antioch, and had ventured in the direction of Maraş (Ar. Mar'ash) to reduce some local Armenian lords, as Danişmend was doing a little further north. Bohemond imprudently took the road to Malatya, and was surprised by a contingent of Danişmend's Turcomans, who sent him off to captivity in Niksar (Neocaesarea), while two Armenian bishops were killed. Malatya was however saved by reinforcements from Baldwin of Edessa, but was finally to yield to Danişmend in 1102.

In the meantime Danişmend had been recalled to the west by the approach of a new crusade, which had chosen this time to penetrate into the heart of Anatolia by the northern route. There the crusaders were annihilated. Other crusaders had however taken the route of the crusaders of 1097, but Kilic Arslan, taught by experience, laid the country waste before them without really confronting them. They also in the end were decimated near Ereǧli.

Danişmend however was not unaware that Bohemond had long been the enemy of Alexius Comnenus, and that he had secured possession of Antioch in violation of the promise he had made to the emperor. Henceforward Alexius regarded the struggle against Bohemond as more important than the struggle against the Turks, who held only the interior of the country without having yet organized a real power there. He made an offer to pay Bohemond's ransom if he would hand him over. Kilic Arslan, who no doubt regarded Danişmend as his subordinate, claimed a part of the ransom. Bohemond was then able to persuade Danişmend to prefer the ransom offered by Baldwin of Edessa, and an alliance with the Franks against Kilic Arslan. Kilic Arslan had drawn closer to Alexius Comnenus, whom he had allowed to destroy the last Turkish nest on the Aegean, held by Tangripermes (Tanrivermiş) at Smyrna and Ayasoluk (now Selçuk) away from the crusaders. Moreover he promised him help against any enemy who might arise. Bohemond returned to Europe to organize an expedition against the Byzantines, and in fact landed in Epirus in 1107; Kilic Arslan sent Alexius the promised Turkish contingent.

In 1104 Danişmend died. There were difficulties over the succession, and in 1106 Kilic Arslan occupied Malatya. It was then that he was tempted by appeals from Upper Mesopotamia, where there had been a revolt of Sultan Muḥammad's amirs. The proximity of Kilic Arslan, whom they knew to be the hereditary enemy of the Great Seljuks, seemed to them an opportunity to be seized. Kilic Arslan was evidently aware of the old rivalry between the two branches of the family. Unlike the Turcomans established in Asia Minor, he had no doubt not given up the idea of maintaining contact with the bases of his people in Iran, and of intervening in the quarrels of his kinsmen. He advanced to Mosul. Muḥammad had however rallied his partisans, while Kilic Arslan was weakened by the absence of the contingent sent to Alexius Comnenus. A battle took place on the Khābūr, a tributary on the left bank of the Euphrates. Kilic Arslan was killed, and his son, Shāhānshāh, made prisoner (June 1107).

This date marks a turning-point in the history of Asia Minor, and we must stop for a moment to see the balance-sheet of this first phase.

One point is that we have reached the end of the first wave of the peopling of Asia Minor by Turks. Apart from small exceptions, there will be no more until the Mongol invasion in the thirteenth century. The Turcomans installed in Anatolia were henceforward to some extent divided among three or four powers, above all the Seljukids and the Danişmendids, but without there being very precisely organized links between them and their rulers. Although their pasture-grounds were limited, the great majority of them were semi-nomadic pastoralists and, except for some garrisons in towns of importance to their rulers, they were still predominantly people of the plains. Agriculture was in the hands of the indigenous inhabitants, but had suffered greatly as a result of the depredations of the invaders and the flight of a great number of the peasants, especially from western Anatolia, whence at least those in the easiest circumstances could reach Byzantine territory. Once the time of disorder was past, attempts were made to hold or bring back the peasants, but this process was only just beginning. The Turcomans were attracted to the frontier-zones, which they could pillage. Scholars gave this activity a colour of Holy War by calling them ghāzīs (T. sing, gazi), but they designated themselves in these zones as men of the frontier (T. uc), corresponding to the Greek akritai. It is obviously impossible to suggest a figure for the number of the immigrants, and consequently the proportion of the new Turkish population; it is equally impossible to determine the extent of the damage or the activity continuing in the country at this time. It must only be said that the trend in the Byzantine Empire towards great estates, exploited for the breeding of cattle, must somewhat have facilitated the establishment of the newcomers. Also one must consider eastern Asia as having on the whole been less disturbed. The Armenians continued to be numerous there, and after the first phase seem to have come tolerably to terms with the new masters. At a later stage the documentation will allow a somewhat more precise treatment.

A fundamental distinction must be made between these Turks of Asia Minor and those further to the south, who occupied the traditional Arab territories. Vague as the frontier was, it corresponded approximately to that of the Byzantine Empire previously and that of modern Turkey. Whatever the indigenous elements in Asia Minor to the north, there was a complete peopling by Turks; in the south, a military occupation without a true peopling.³ Upper Mesopotamia was an intermediate zone, occupied by Turcomans subject for the most part to the Artukid dynasty; but these had been faithful servants of the Great Seljuks, and essentially left in being the structures of a province which was among the richest of t...

Table of contents

• Cover
• Half Title
• Title
• Copyright
• CONTENTS
• Editor’s Note
• Names and Dates
• Author’s Preface
• Publisher’s Acknowledgements
• Map
• Introduction: Summary history of the Turks before Turkey
• PART ONE: GENERAL HISTORY OF TURKEY BEFORE THE MONGOLS
• PART TWO: SOCIETY AND INSTITUTIONS BEFORE THE MONGOLS
• PART THREE: GENERAL HISTORY OF THE MONGOL PERIOD
• PART FOUR: SOCIETY AND INSTITUTIONS IN THE MONGOL PERIOD
• Conclusion
• Notes
• Appendices
• Bibliographical Guide
• Glossary
• Index of Persons
• Index of Places