web counter








The appearance of the Turks, starting with the eleventh century, in most of the Moslem world and then of the Byzantine empire, inaugurated a profoundly new phase in the history not only of eastern Christianity, but also of Islam. It is true that the transformations which it brought about were in some respects the culmination of a previous internal evolution, but it precipitated and, in certain respects, disrupted this. There is thus the problem, to which insufficient attention has been directed, of identifying with precision the circumstances, the characteristics, and the scope of this intervention. But to attempt to offer here an inclusive analysis of its history would force us both to remain on too elementary a level and to depart from the general plan of the present work. We shall, therefore, lay particular stress on those of its aspects which affected the international relations of the occident and the orient.

There had long been Turks within the Moslem world. Some tribal groups had established themselves, well before the eleventh century, on the eastern confines of the Islamic domain, cut off from the main body of their relatives. From the ninth century on, especially heavy recruiting of Turkish slaves had been undertaken in order to enlarge or replace the former unreliable indigenous armies, and from their ranks had emerged numerous governors of provinces, some of whom had become autonomous, as had the Tulunids of Egypt and the Ghaznavids of eastern Iran. It is unlikely that these men had retained no Turkish elements in their memories or, especially, in their characters. Since, however, they had been removed at an early age from their original environments and integrated into the structure of Moslem society, they cannot be considered as representing a real penetration by the Turkish world into that of Islam. When the true Turkish conquest occurred, these elements were no less opposed to it than were the natives, just as "barbarian-born chieftains" had defended the Roman empire against the "barbarians". And even though they may unconsciously have facilitated certain transitions, nothing would have been more foreign to them than any concept of Turkish solidarity. It was the same with the many Turkish mercenaries introduced into the Byzantine army during the eleventh century. During the First Crusade, for example, the troops of the basileus were led by a commander of Turkish origin in their effort to reconquer Anatolia from the Turks.

In order to avoid misunderstanding, however, it should be stated immediately that, in our judgment, the Turkish conquest was achieved as much from within as from without. This was done, as we shall try to make clear, in another fashion.

On their side the Turks were not, in the eleventh century, novices in politics. Almost certainly Turkish in all save name were the Huns who, having been forced towards Europe at the time of Attila, were the indirect forerunners of the Bulgar states on the Volga and the Danube and of the Khazars between the Black Sea and the Aral Sea. In the sixth century, Turks — by this time even in name — founded around the Altai range an empire which formed a link between Byzantium and China and left splendid memories in Central Asia, of which we have an eighth-century record in the first of the famous Orkhon inscriptions. Likewise Turkish, in the same region, were the eighth-century Uyghur realm and the ninth-century Kirghiz kingdom. From the time of the first Turkish empire the eastern Turk in contact with Chinese civilization, are to be distinguished from those of the west, leading nomadic lives to the north of Transoxiana. The pressure of new peoples, largely Mongol, caused a progressive withdrawal of the Turks from the east towards the west and the consequent transformation of the western steppes, until then half-Iranian, into that "Turkestan" which has retained their name ever since. Some groups, such as the Pechenegs, even reached Europe. The majority stayed in Asia, among them most of the Oghuz group who, having already been among the chief actors in the events just related, were to dominate later history.

The Turks, generally shamanistic and hence originally alien to any exclusive or circumscribed religion, had been exposed to Nestorian, Manichaean, and Buddhist influences brought in by pilgrims and by merchants from Soghdia and elsewhere as they crossed Central Asia. The Khazars had similarly been open to Jewish influences. The Arab conquests of the seventh century placed them in contact with Islam, and, once the newly-conquered territory was Islamized, Moslem traders in their turn brought into the Turkish zone the influence of their new faith. In the tenth century large groups of Turks were won to Islam, from the Bulgars of the middle Volga to those whom the Kara-Khanids were about to unite on both sides of the mountains separating Russian from Chinese Turkestan. As had formerly been the case among the Arabs of Arabia, Islam was able to constitute for the Turks a common political bond, so that under this dynasty the first great Turkish Moslem realm came into being.

Most of the Moslems who had ventured among the Turks had come from Transoxiana, from Khurasan, and from Khorezm. Thus it was in the special forms which had been developed in the north­eastern Iranian region that the Turks came to know both Islam as a religion and the general civilization from which they were unable to distinguish it. It should he stressed that its spread had been accomplished not by orthodox theologians but by merchants and plebeian mystics. Although the princes, on adopting Islam, associated themselves with orthodox groups, the mass of Turks remained no less Moslem, but professed a folk-Islam very different from orthodoxy. And naturally the Turks, on adopting the new faith, did not entirely forget all the customs, beliefs, and practices of their non-Moslem ancestors.

Even though the Turks lived, like all nomads, in symbiosis with the sedentary oasis-dwellers, and though some of them had themselves become sedentary, the overwhelming majority remained pastoral migrants from steppe to steppe. It has often been emphasized that nomad societies usually ignore or challenge the property limits which administrative states establish, as well as the frontiers which these states erect in an attempt to reserve for themselves the right to use certain territories. The Oghuz were not different. Like their "Scythian" precursors, they constantly launched against their neighbors and the sedentary inhabitants rapid raids which were hardly more than adventurous episodes in their perpetual wandering, although in times of drought the booty they secured was almost essential to life. The sedentary population referred to the Oghuz nomads and analogous neighboring groups as Turkomans.

Along the northern border of Transoxiana, therefore, the Moslems continued against the nomads the old Iranian tradition of frontier defense. A special military organization provided this, and since their original opponents were unbelievers, it attracted all those whose enthusiasm was aroused by the Moslem ideal of holy war, namely the ghazis. Their tactics, matching those of their adversaries, stressed flexibility and speed, and were adapted to a strategy of incursions. Organized into martial brotherhoods in which the spiritual and military leaders simultaneously encouraged religious fanaticism and developed combat skills, the ghazis often represented, for the rulers of eastern Iran, a source of internal unrest and at the same time a bulwark against external enemies; the Saffarid dynasty in Sistan originated among them.

The conversion to Islam of a growing proportion of the Turkomans adjacent to Transoxiana upset this whole system. Against the others, still non-Moslem, the Moslem Turks became ghazis in their turn. Obviously this entailed an extension of the Islamic domain, but it also meant the disappearance of the former fortified frontier. From place to place along that line the former ghazis and the new Turkish ghazis mingled, all the more readily because in many respects their ways of life and of war were alike. Against such an infiltration, if it should appear menacing, it would be impossible to mobilize the ghazis of the interior, as they would not fight against Moslems. The idleness to which they found themselves reduced aggravated social discontent. The Samanid sovereigns of Transoxiana and Khurasan found themselves compelled, in imitation of the rest of the Moslem world, to increase the slave element in their armies. It was their misfortune that at the same time the Turkish invasions of Russia had ruined commerce on the Volga, from which they and their subjects had derived great profit. Forced on this account to increase tax burdens, the Samanids alienated the mass of the people, and by making an effort to reduce this unpopularity by concessions to heretics, they also alienated the leaders of orthodox Islam. No one but the slaves had any apparent interest in defending the Samanid realm against the Moslem Turkish chieftains. By this combination of reasons is to be explained the conquest of Transoxiana by the Turkish Kara-Khanid princes, while the balance of the Samanid domains fell into the hands of the Ghaznavids, the offspring of Turkish slaves, who kept their warlike elements occupied by inaugurating at the end of the tenth century a new aggressive policy against the Hindu plain. Moreover, the advance of the new Turkish population modified the ethnic character of these hitherto Iranian regions, such as Khorezm, which within two centuries was to become wholly Turkish.

The reciprocal interpenetration of the ghazis and the Turkomans meant for the latter the assimilation of Moslem civilization in the special ghazi form, which was so well suited to their habits. The frontier zones, where they set up a quasi-autonomous government, they called marches. Their moral cohesion, in default of any administration, was assured by the preachers and the learned, heirs of the shamans, who continued to live among them, teaching and judging, and who sometimes succeeded in acquiring the prestige of chieftains.

One of the principal Turkish groups on the Moslem borders who were converted to Islam in the second half of the tenth century had as chief one Selchuk, of the Kinik tribe of the Oghuz. He had established himself on the lower Jaxartes (Syr Darya). At the end of the tenth century this group was hired by the Samanids to resist the Kara-Khanids, and at the start of the eleventh century by the prince of this latter family who held Bukhara and Samarkand to support his revolt against the others. The Sasanids, with their men, therefore settled in Transoxiana, in the old Iranian Moslem land, where they received grazing grounds for their flocks. Closely associated with the princes in their activities, as leaders of one of the principal elements of their military forces, they could begin to familiarize themselves with the traditional Moslem ways of life and administration and to form ties with the orthodox Moslem leaders.

In 1025 a portion of the Selchukid Oghuz were settled in Khurasan itself by Malmud the Ghaznavid who, victorious over their Kara-Khanid protector, was probably desirous of depriving him of their strengths. But very soon these nomads, by the necessary conditions of their life, set themselves up as a troublesome element, destroying harvests around the towns and thus causing misery and unrest, as well as a decrease in tax revenues. Military operations against them, conducted by troops less mobile than they, succeeded in driving them back but not in destroying them; the result was the diversion of their disorderly activity towards central and western Iran. The revolt of Masud, son of Mahmud, against the immediate successor of his father stripped Khurasan of its army; while the tendency of Masud to minimize the danger, which seemed to him merely to call for police action, and to use his army for profitable raids on India left the Turkomans practically uncontrolled.

In 1035 the rest of the Selchukid Oghuz, who had embroiled themselves with the new princes of Samarkand and Bukhara, moved to Khorezm with a rebellious vassal of Masud; then, threatened by a neighboring prince, they crossed the Oxus (Amu Darya) without difficulty, since the principality of Khorezm straddled the river, and in their turn made an unauthorized entry into Khurasan, in the territories left vacant by the departure of their predecessors, where they naturally behaved as had the others.

The population of the commercial cities of Khurasan had no reason to be faithful to the Ghaznavids — whose government, entirely devoted to the military, was fiscally oppressive — except when this dynasty guaranteed their security. When it appeared unable or unwilling to do this, the leaders decided that the most practical way of avoiding disaster would be to recognize Selchukid suzerainty, which could be done without religious qualms since they affected a severe orthodoxy. At least, concerned for the prosperity of these cities, they would deflect elsewhere the disorders of their people. This was done by Merv and then, in 1037, by Nishapur, the capital of Khurasan,

It is scarcely to be doubted that the chiefs of this second Selchukid group, two brothers, grandsons of Selchuk, Tughrul-Beg and Chagri-Beg, were willing to form a state, making use of their Turkomans, but in accordance with concepts strange to them. From the start they had their authority recognized by the caliph under the title "clients of the commander of the faithful", which legitimized in Islamic eyes their actual power over their men, and established a claim to enlarge it. For although the groups which, lured by booty, followed Tughrul and Chagri recognized them as warrior chiefs, they did not consider that this recognition conferred on them any rights in regard to the internal affairs of the tribes, nor that it prevented any Turkoman tribe from leaving the confederation whenever it wished. Tughrul and Chagri were merely first among equals. But, charged by the commander of the faithful with the responsibility for imposing on their men the word of Allah, Tughrul and Chagri found their justification for claiming an authority which they could not otherwise have exercised. After their subsequent accession to the rank of territorial princes, they found themselves automatically integrated into the old Moslem organization. This brought the brothers a new power foreign to their functions as chieftains of nomads, but it led them to desire in their turn to preserve their territories from the depredations of the same men to whom they owed their acquisition.

The capitulation of the great cities opened the eyes of Masud to the political danger threatening him and he led his forces back into Khurasan. This was followed by several years of exhausting struggle in which the enemy always fled into the desert, to reappear unexpectedly and attack in a different quarter. In a country which the nomads had impoverished it was difficult to maintain a large army, poorly prepared for this style of warfare. The soldiers complained and the hard-pressed inhabitants did not assist Masud. At last the Selchukids dared to attack. In 1040 at the battle of Dandaqan in the province of Merv the Ghaznavid army was annihilated. Masud fled to India. Khurasan was lost, and the Iranian plateau was wide open. The evolution of the Iranian and Turkish worlds had led the former to admit the Turks into its own bosom. Like that of the Germans in the Roman empire, the conquest by the Turks, from then on, was accomplished from inside.

Among the simple yet powerful ideas which the Selchukid chieftains found in Iran was that of the scandal involved in the oppression of the caliph by the heretical Buwaihids. Already Mahmud and Masud had spoken of going to his relief, had begun the subjection of the Buwaihids of Iran, and had persecuted heretics. A "crusade" was in the air, and it can scarcely be doubted, from the course of ensuing events, that Tughrul-Beg promptly decided to profit from it. He immediately received the support of the orthodox notables of Khurasan, both for ideological reasons and for the sake of the profits they expected from exercising administrative control over the new conquests. For naturally it was through them that the Selchukids, whose Turkomans had had no administrative experience, would have to govern their territories. In certain respects the entry of the Turks into Baghdad would reproduce the earlier Khurasanian conquests of the Abbasids over the Umaiyads and of al-Mamun over al-Amin.

At the same time, the occupation of Khurasan allowed the Selchukids to add to their Turkoman bands an army of the traditional Moslem type, supplied with weapons suitable for taking cities, which the men of the desert had lacked. Moreover, this army diminished their dependence on their Turkomans. The latter remained, nevertheless, their basic force, which required almost no pay and alone assured their superiority over their adversaries. The main problem of the Turkomans was the locating of new pastures. In religion their attitude was that of the ghazi, which was not that of the orthodox against the heretic but rather that of the Moslem of every description against the unbeliever; and they remained opposed to any Selchukid domination over them except the purely military.

In some respects the two attitudes might in practice coincide. This may be seen in the division of tasks which Tughrul and Chagri agreed upon following DandanqanChagri retained, in addition to most of Khurasan, the Selchukid homelands, to be defended against the Kara-Khanids and the Ghaznavids. He accomplished this by annexing Khorezm and the upper Oxus and, at first through the intermediary of some cousins, the provinces of Herat and Sistan. But in this direction expansion halted there, not only because it was made difficult by the presence of other Turks in the northeast and by mountains to the southeast, but also because in fact the majority of the Turkomans were not oriented thither. Even though, towards the south, a son of Chagri named Kavurd occupied Kerman and went beyond to seize the entrance to the Persian Gulf and impose his power on Oman, these excessively hot deserts could not greatly attract the Turkomans.

On the other hand Tughrul, to whom had been allotted whatever he could conquer towards the west, was able to take advantage of the more normal area of expansion which the steppes of the northern and western portions of the Iranian plateau presented to the Turkomans, as they had to many others before them. The Buwaihids and other Iranian princes, torn apart by dissensions, poorly supported by troops who, more than elsewhere, were attached to the soil by land-grants, were no longer in a position to organize any real resistance. Tughrul had no trouble in taking Rayy or in leaping forthwith to the opposite edge of the plateau to capture Hamadan, at the same time that, on his flanks, he had his "suzerainty" recognized in Tabaristan and, in 1043, Isfahan. This advance was considered menacing by the first wave of Turkomans to have entered western Iran. Fleeing the Selchukids, they spread over upper Mesopotamia where, cut off from their bases, they were annihilated by the Arabs and Kurds, who had formed a coalition against their ravaging pastoral competitors.

The situation of Tughrul with his own Turkomans was complex. It was chiefly in the direction of Azerbaijan that the convergence of the Iranian routes caused them to reassemble, and in addition they were drawn by the proximity of frontiers — Georgian, Armeno-Byzantine, and Caucasian — which suggested the possibility of resuming the ghazi activity which they had had to abandon in the east. In itself this did no harm to Tughrul, who thus assured at slight expense the covering of his northern flank and might look forward to further conquests. In any event, it was preferable for their docks to browse on pastures other than his. But there were disadvantages; Tughrul needed the Turkomans at hand for his own operations, which had become much less attractive to them now that, as we shall see, he often forbade pillage and did not let them take their families for permanent settlement. On the other hand, the Turkomans could give asylum and assistance beyond Tughrul's reach to ambitious rebels or, even without any preconceived plan, might end by founding a separate state. For all these reasons it was essential that Tughrul participate in the activities of the Turkomans in order to direct and channel them, And since, as Saif-ad-Daulah had just shown, the ideal of holy war still inspired the Moslems, he could derive increased prestige, redounding to the benefit of his other undertakings, from engaging in it beside his Turkomans.

Thus can be discerned, amidst the jumble of episodes monotonously narrated in the chronicles, the permanent balancing of two great trends: the expansion northwestward and the consolidation of power within Iran. The former began with intervention in north­western Iran to enforce recognition of Selchukid authority, and in addition, following the traditional invasion routes, isolated Turkoman raids against the Georgians, the Armenians, and the Byzantines. In 1048 came the campaign of Ibrahim Inal (or Yinal), half-brother of Tughrul, into Armenia and the sack of Erzurum, and in 1054 that of Tughrul himself further south, capturing Arjish and besieging Manzikert. There he was also motivated by the desire to reconquer and strengthen the old Moslem frontier against the expansionism of Byzantium, whose response to the first Turkoman pressure from Azerbaijan had been the direct annexation of the hitherto autonomous Armenian kingdoms. The Byzantine government, renouncing a militarist policy, negotiated and purchased a peace which it expected the Selchukid to guarantee, and, by initiating closer ties with the Abbasid caliphate for this purpose, succeeded only in cooling the friendship of the Fatimids, whose support would prove to be lacking at the decisive moment.

Meanwhile a ten-year period was devoted to establishing solid Selchukid dominance over the entire region between Khurasan and Baghdad through the direct annexation of vassal principalities, through the penetration of the Kurdish hill province, where Ibrahim Inal combined military operations with diplomatic play on the rivalries of the indigenous tribal chieftains, and through utilization of similar rivalries and fear of the Turkomans to set up in Mesopotamia itself a faithful circle of petty princelings. In the province of Baghdad itself all pillaging was forbidden; Tughrul knew what he wanted.

At Baghdad, with the decline in the authority of the last Buwaihid of Iraq, the rule had fallen to his Turkish general and fellow­Shiite, al-Basasiri, the oppressor of both people and caliph. The latter, however, took advantage of the Buwaihid collapse to reconstitute a sketchy caliphal government for which he needed orthodox support against al-Basasiri. He had long enjoyed pleasant relations with Tughrul. In 1051 the famous jurist, al-Mawardi, at the same time that he had urged him to restrain the pillagers, had conferred on the Selchukid prince titles superior to those borne by anyone else. Tughrul had spoken of his desire to liberate the caliph, to assure the security of the pilgrimage, to subdue domestic heretics, and to deprive those abroad of Syria and Egypt, while disowning any intention of effecting direct seizure of Iraq. Pushed by al-Basasiri to extreme measures, the caliph thought of summoning the conqueror of Iran as a protector. Even the Buwaihid thought he might deal better with him than with al-Basasiri. The latter, uneasy and too weak, left Baghdad. In 1055, after everything had been solemnly prepared, Tughrul-Beg made his entry into Baghdad at the head of his troops without striking a blow. There he was welcomed by the vizir of the caliph.

This moral triumph, it is true, was soon followed by a very grave crisis. Most of the Arabs, who were worried about their pastures and who were Shiite, gathered around al-Basasiri in his refuge on the Syrian border. From there he appealed to the Fatimids, who sent ambassadors and money, and led him to hope for reinforcements. Difficult operations ensued in upper Mesopotamia. The Turkomans grew discontented. For them, long accustomed to contact with Iranians and to a similar climate, northwestern Iran was not a strange land. But because of its heat and because of the language and customs of its inhabitants, Mesopotamia was. Further, they were prevented from establishing themselves there comfortably by the presence of nomadic Arabs and Kurds and by the policy of Tughrul; they had to leave their women in Iran; they suffered from a lengthy separation uncompensated by adequate booty. Moreover, Tughrul, to gain acceptance from his new subjects, surrounded himself with Arabs and overwhelmed them with favors. He adopted the manner of a sovereign. All these things offended the Turkomans and the Selchukid princes. In the midst of the Mesopotamian war Ibrahim Inal deserted to instigate a revolt among the Turkomans in Iran. Tughrul had to leave Mesopotamia; al Basasiri returned to Baghdad, proclaimed Fatimid sovereignty, and expelled the caliph, who was sheltered by an Arab chief.

The assistance which Tughrul as a last resort obtained from the sons of Chagri-Beg saved him. The Turkoman revolt was stifled, Ibrahim Inal strangled, Iraq retaken, al-Basasiri hunted down and killed, and the caliph restored. All the Mesopotamian chieftains, especially the Uqailid of Mosul, now hurried to make their peace with the omnipotent victor. By 1059, and this time definitively, Tughrul-Beg was master of Mesopotamia as far as the Byzantine and Syrian frontiers.

Obviously thereafter, in Iraq as elsewhere, it was Tughrul who exercised the real power, but not in exactly the way the Buwaihid had; and the caliph was the beneficiary of the change. He was indeed sometimes made to feel that his domains had been left to him as a favor and that his government was subject to the agreement of Tughrul, as when in 1060 he tried to refuse his daughter's hand to the sultan. It was nevertheless noteworthy that he did have a civil government which, with the Turkish garrison, ruled Baghdad, and that he did hold domains commensurate with his rank. Above all, Tughrul, whether sincere or merely aware of the moral authority he derived from him, showed a real respect for the caliph. It was he who, as master, tried to avoid offense by not leaving too many Turks in Baghdad; he who, ill at ease amid the welter of Arab intrigues, preferred not to visit Baghdad often; and he who, above all, fought for the faith and for orthodoxy, and to whom for that reason the caliph gave his sincere support.

The title of sultan which the caliph conferred on him — long since a part of the current vocabulary, though Tughrul seems to have been the first to bear it officially — meant that he exercised all material power, on behalf of Islam in the service of the caliph, who was the supreme religious leader. It was a somewhat novel situation. The ninth-century caliphs had actually ruled; those of the tenth century were not even recognized as their religious superiors by the Buwaihids; and the principalities where they were so recognized, like the Samanids, were so distant that they were forgotten there. Now there was a true symbiosis which might suggest that which had existed in western Christendom between Charlemagne and the papacy.

The two long reigns which followed that of Tughrul-Beg, those of Alp Arslan (1063-1072) and Malik-Shah (1072-1o92), witnessed the development of both the Selchukid empire and the Turkoman power. It is impossible to describe here in detail events the characteristics of which were not new.

The deaths of Chagri-Beg and the childless Tughrul-Beg led to the unification of all the Selchukid domains except Kerman under the rule of a son of Chagri named Alp Arslan. It could have been a source of weakness for the sultan to have to keep watch simultaneously over the whole of so extensive a frontier. In fact, even though Alp Arslan happened to die in Transoxiana, neither the Kara-Khanids, who were disunited, nor the Ghaznavids, whose ambitions were deflected toward India, were to cause him or his successor serious trouble. The bulk of their external affairs concerned the west. Tughrul had received from the caliph the title "king of the east and the west", investing him in advance with all he might conquer from the heretical Fatimid. Alp Arslan, as will be evident, remained aware of this mission, It was not, how­ever, from this quarter that he was to acquire his glory in the eyes of posterity, but from that where he became involved in the expansion of the Turkomans themselves.

Since the later years of Tughrul-Belt's reign, these nomads had been making deep raids into Byzantine Armenia. The weakening of the Byzantine army, the internal revolts, the indiscipline and rivalries of the Armenian frontier chieftains, and especially perhaps the unsuitability of a system of large garrisons in widely-spaced fortresses for intercepting light troops crossing the country­side — for, once across the frontier, these no longer feared any army — these are the explanations of how such raids could have been accomplished with so little risk. Each year they had penetrated a little further. After 1057, when they sacked Melitene (Malatya), those who were perhaps most closely in touch with Selchukid policy had ranged southward along the Byzantine-Moslem border, descending the Euphrates as far as Syria; but the boldest were those who, for whatever reason, had fled Selchukicl authority and who wanted to carve out by main strength a refuge inside Byzantine territory. In 1067-1068 they were to be found in Anatolia proper, at Amorium, at Iconium (Konya), and in Cilicia, and, in 1070 at Chonae. Sometimes they were hired by Byzantines, as was a brother-in-law of Tughrul-Beg in 1070. Another leader, who had served the Marwanid Kurds on the upper Tigris and then the Byzantines, ended by serving the Mirdasid Arabs of Aleppo against the Byzantines. A third, Atsiz, having escaped from Anatolia, landed in Palestine in 1071 and was engaged by the Fatimids to pacify insurgent bedouins. It had long been the practice of "civilized" governments to hire for use against each other whatever "barbarian" bands offered their services.

It can be seen how indispensable it was for Alp Arslan, for the same reasons as for Tughrul-Beg but even more urgently, to intervene on the Byzantine frontiers. In 1065 he took Ani and about1068 annexed some Georgian territory, thus making sure not only of the fidelity of his native vassals in Azerbaijan, but also of firm bases for activities in connection with the Turkomans. At length Byzantium reacted. The soldierly emperor Romanus Diogenes in 1068-1069 conducted a campaign into Syria and then along the upper Euphrates, by which he acquired or strengthened the frontier fortifications. The appearance of bands of Turks far to his rear demonstrated the futility of this method, and his army suffered from the devastation inflicted by the Turkomans on the regions through which it passed. In 1o7o Alp Arslan could consider his realm safe.

It was then that he revived the old project of war with Egypt, to which he was the more receptive because of the welcome found by the Turkoman Atsiz in the Fatimid possessions. Though on his way he occupied several Christian places in consolidating his Euphrates frontier, his real goal was Aleppo. This strategically placed junction, autonomous but under Egyptian influence, he subdued and officially restored to Abbasid control. From there he was prepared to continue southward, but he received word that Romanus Diogenes, profiting from his extended advance, was preparing an offensive in his rear. He reversed his movements in the Turkoman way, leading unprepared observers to assume a rout, but he reunited his troops at the assembly point.

A battle which has been embellished by legend, but which has always been fascinating because it was the first meeting in centuries between a Byzantine emperor in person and a comparable Moslem sovereign, took place near Manzikert in Armenia in the summer of 1071. The Byzantine army, heterogeneous, suffering from the mute hostility of the native population and of the mercenaries composing it, frightened by a poorly known adversary, and fearing treason because of the presence in its ranks of a Turkish contingent, fell victim to the classical nomad maneuver, a simulated flight permitting a return-offensive envelopment. The Byzantine army was annihilated and, for the first time in history, the Byzantine emperor himself was brought captive to the feet of his vanquisher.

The battle of Manzikert marked the beginning of a new period. Not that Alp Arslan had any idea of dismembering the Byzantine empire; he was satisfied to demand a tribute and the cession of the formerly Moslem border towns, provisions which the overthrow of Romanus Diogenes at Constantinople rendered meaningless. What the sultan wanted was a guarantee of neutrality or alliance in his enterprise of unifying the Moslem world, and the eventual aid of the basileus against rebels who fled into Byzantine territory. But Manzikert completed the ruin of the Byzantine military strength; the Turkomans, instead of retiring after each raid, no longer had any reason not to stay in the territory of the empire. The populations of Armenia and Cappadocia, hostile to Byzantium for fiscal and religious reasons, no longer able to rely on the Byzantines for defense, treated with the invader just as had the inhabitants of Khurasan. Certain of their component-elements — military colonists planted on the frontier and others — had less in common with the Byzantines than with the border Moslems with whom for centuries they had alternately had minor battles and courteous exchanges, and who sometimes mingled with the Turkomans. At times these groups joined the newcomers. The Byzantine system had, moreover, become disorganized by the action of the Constantinopolitan government itself in annexing Armenia and Edessa (Urfa) and thereby advancing its frontier beyond the prepared zone. Distrusting its new subjects, it had replaced them as soldiers with mercenaries hateful to the inhabitants, who under the pretext of protecting them from the Turks were deported to Cappadocia and Cilicia. Thus the area where anti-Byzantine quarrels and bitterness prevailed was permanently enlarged. 

A few years sufficed to eliminate the last traces of Byzantine administration from the main routes of Armenia and Cappadocia. It was not that they had been formally expelled, but in a flat land held by nomads and deserted by whatever peasants survived, how could taxes be collected? The cities remained as foreign bodies which surrendered in order to escape famine. And even though. the Turkomans necessarily allowed them to govern themselves, they lost all contact with the Byzantine government. No deliberate seizure of Byzantine territory by the Turkomans had occurred; they were in a land which they knew belonged to "Rome", but that sovereignty had been emptied of any reality.

By themselves the Turkomans could perhaps not have progressed as rapidly as they did. It was the Byzantines who had brought them into the heart of the empire. Since Byzantium had commenced the habitual enrollment of "barbarian" as mercenaries, the Turks who had for a century or two regularly offered their services were strangers no longer. Even if they had been, what difference would it have made to all the generals competing for the throne ? Had Romanus Diogenes himself not called upon his late enemy, the sultan, for aid in regaining power from Manzikert on, and especially from 1078 to 1081, others successively brought them in opening to them the Greek villages of Asia Minor, even establishing them on the shore of the Sea of Marmara and near the Bosporus at Nicaea or along the coasts of the Aegean Sea.

Assuredly these Turkomans, though theoretically responsible to the Byzantines through the leaders who imported them, were none the less autonomous Turks whose perpetual pillaging by land and soon by sea was an obvious danger to Byzantium, and not only to Byzantium but also to the sultan, from whose control they had completely escaped. In the last years of Tughrul-Beg's life one of his cousins, Kutulmish (or Kutlumush), whose father had formerly been the eldest and foremost member of the family, had withdrawn with some Turkomans into the mountains south of the Caspian Sea. Proceeding into open revolt against Alp Arslan, the sons of Kutulmish sought safety in Anatolia amidst some free Turkomans. It was with them in particular that the Byzantines had dealings and doubtless it was they or their Turkomans who wished to set themselves up as a state in Anatolia, or at least as a force capable of resuming the contest with their Iranian cousins. From 1075 on they were involved in Syria as allies of the Fatimids against a Selchukid adherent. In the Taurus mountains a former general of Romanus Diogenes, an armenian named Philaretus (Filardos), had gathered under his authority the people of Cilicia and of the region from Antioch to Edessa and Melitene. The Byzantine emperor Alexius Comnenus allowed — if nothing more — Sulaiman, the last survivor of the sons of Kutulmish, who was installed at Nicaea, to take from Philaretus, in the capacity of a Byzantine lieutenant, Cilicia, Antioch, and Melitene. At Iconium, he was in complete possession of one of the two great east-west Anatolian routes and hence at the border of Selchukid Mesopotamia and Syria, a grave danger to Malik-Shah in 1084–1085.

On a smaller scale, the same problem was posed by Atsiz further south. Though summoned by Egypt, he had promptly embroiled himself with her and, together with Jerusalem and Damascus, had formed an autonomous principality which he now sought to consolidate by recognizing Malik-Shah, thus provoking the Egyptian appeal to the sons of Kutuhnish. Atsiz defeated the coalition and, in his new power, could hardly have inspired much confidence in Malik-Shah.

Indeed, this sultan's policies seemed much more concerned with the avoidance of such dangers than with the further extension of his empire. It is true that he had cleared the Persian Gulf region of the Qarmatians who had infested it, and had disputed with the Fatimids the allegiance of Mecca, but those were minor undertakings. Young, born to the "purple" and not to the steppe like his father, he was less a soldier than a proponent of the diplomacy counseled by his vizir Nizam-al-Mulk. The latter knew that the unity of the empire needed careful safeguarding, and that every prolonged absence of the sovereign in one quarter could be utilized by fomenters of trouble elsewhere.

He also knew that within the Selchukid family itself, where the tradition still lingered of rule by the family rather than by a single sovereign, there could arise new discontents like those of Ibrahim Inal and Kutulmish, recently quelled. Though Malik-Shah had removed, by executing him, the embarrassment of his uncle, Kavurd of Kerman, who had claimed as eldest of the family to supplant him, it appeared that it might be useful to create appanages for the young princes. Such a course would please them, propitiate local sentiment, and avoid unnecessary travel by the sultan. This was the solution Malik-Shah adopted for Syria, among other places, in response to an appeal by Atsiz concerning an Egyptian attack. Tutush, brother of Malik-Shah, received central Syria and Palestine, and in 1079 rid himself of Atsiz. As for Aleppo, distracted by the ravages of the Turkomans, which Tutush was unable to prevent, and deceived by princes incapable of giving protection, it yielded to the Usailid of Mosul, an Arab allied by marriage to the Selchukids and vassal to them. Malik-Shah left Aleppo alone for the time being, but he sent an army to Anatolia to combat the sons of Kutulmish.

Meanwhile the Selchukid government also gradually limited the autonomy of the indigenous population, and that policy of perpetual small-scale local encroachment would continue long after the Selchukicl empire had been broken into fragments. It was thus that the Shabankarah Kurds of Fars were subdued, and that the vassal states of extreme northwestern Iran were, except for Shirvan, annexed little by little. Those situated on frontiers or on main strategic routes were in particular danger. Some remained more or less openly Shiite, like the Uqailid, who clashed with Tutush and intrigued with Philaretus and even with Egypt. On the other hand there was the more vulgar greed of the lieutenants of sultan and caliph, when they knew a treasury was rich. This was one of the reasons for the suppression of the Kurdish principality of the Marwanids, innocuous though it was otherwise. Against the Uqailid an assault was prepared, for he had feared the consequences of the disappearance of the Marwanids, and had come to their aid. But it happened that Sulaiman, the son of Kutulmish, having just taken Antioch, got into a dispute with him and killed him. Sulaiman thought it wise to be reconciled with Malik-Shah, but was attacked and killed by Tutush.

Chance favored Malik-Shah. The bedouin victims of the Turkomans, the victims of Sulaiman, of the Uqailid, and of Tutush, all those who were exasperated by the continued devastations appealed to him, asking only to submit to him. He arrived without striking a blow, annexing Mosul, Aleppo, Antioch, and the rest of Philaretus's holdings. He had been seen at Samarkand; he now appeared on the shore of the Mediterranean. This time Mesopotamia and Syria were wholly incorporated into the Selchukid empire. Tutush remained, but in 1o86-1087 the other captured cities received as governors freedmen from the Selchukid army — Buzan at Edessa, Yaghi-Siyan at Antioch, and Aksungur at Aleppo.

There remained only one dark spot, Anatolia. For the death of Sulaiman, though it had enabled Malik-Shah to occupy Antioch, had not contributed to the subjugation of the Turkomans of Anatolia. Against them Malik-Shah, at the same time that he sent troops, tried to obtain as an ally Alexius Comnenus, whom he recognized as legitimate possessor of all the former Byzantine territories. This was a necessary procedure, for how could one organize a Moslem administration in territories where there were no Moslems except the Turkomans? But Alexius hesitated, not knowing whether to prefer the troops of the powerful Selchukid or the bands which he hoped in the long run to neutralize by playing them against one another. Malik-Shah was to die without having concluded this agreement or having accomplished anything important against the Turkomans. Subsequent events, it is true, were to demonstrate that once they had left the empire, the Turkomans could not make much headway against it.


It is impossible here to consider exhaustively the internal administration and the civilization of the Selchukid empire, subjects about which very little is known. It will suffice to describe certain general characteristics necessary to the understanding of events which will be mentioned in this work.

The Selchukid regime can be defined as an orthodox dictatorship accepted by the majority of the population, administered by Khurasanians, and relying upon a dual military basis of Turkish slaves and Turkomans. In the domain of culture it was the period of Omar Khayyam (Umar al-Khaiyam), when the revival of the Persian language, which began at the end of the tenth century, culminated in the progressive elimination of Arabic from the land of Iran, even as the language of learning. In Selchukid art Khurasanian influences are evident. The administrative personnel, even in the Arab areas on at least the upper levels, was basically Iranian.

The great organizers of the regime, the vizir of Tughrul-Beg, Amid-al-hulk al-Kunduri, and the vizir of Alp Arslan and Malik-Shah, the illustrious Nizram-al-hulk, who left us a Treatise on Government, started as functionaries of the Ghaznavids, and belonged to the petty aristocracy of Khurasan. They were in complete charge of internal administration, for the Turks had had no experience along that line, and the sultan left it in their care. Especially under Malik-Shah, who had become sultan while young and who owed to Nizam-al-Mulles ability his ascendancy over the other princes of his family, the vizir was the actual master. He had an enormous following, mostly Khurasanian, an army of slaves, and numerous sons on whom the most lucrative posts were bestowed, to such an extent that for nearly twenty years after his death it would be almost impossible for the Selchukids to secure vizirs not of his family. This power, it is true, aroused envy among those who, perhaps even with the complicity of Malik-Shah, procured his assassination early in 1092.

The power which the regime derived from its conquests, from the elimination of its foes, and from the unification of a territory almost as vast as that of the Abbasid caliphate at its start —except for the far west nothing important was lacking but Egypt— equipped it for action on a huge scale. Its military strength was its foundation, permitting it, paradoxical though it seems, by holding the military power in leash to restore the primacy of the civil administration, which had fallen into neglect under the Buwaihids. In this administration, as in the whole social structure, it was necessary to construct a solid orthodox framework. While up to this time Jews, Christians, and niches could be found on all levels of the bureaucracy, now the Jews were eliminated as much as possible except in wholly subordinate positions, and the Shiites were rooted out. The training of officials was no longer left to chance.

Education had long been left mainly to private initiative, and had been directed toward the development of learning rather than the inculcation of orthodoxy, This had been altered somewhat, to the benefit of Ismailism, under the fatimid caliphate. Among the Sunnites, perhaps as a reaction, an analogous movement had been spreading through eastern Iran since the time of the later Samanids, and was doubtless further encouraged by the Ghaznavids, resulting in the establishment of schools distinct from the public mosques where until then instruction had usually been dispensed. The Selchukids extended this movement throughout their realm, especially in the former Buwaihid domain, where it was a complete innovation. Even if the idea was not wholly novel, in practice they created a new situation by the vigorous interest they took in the widespread diffusion of the madrasahs and the material help they afforded to the schools, their students, their teachers, and their libraries. The most distinguished of these madrasahs was the Nizamiyah, founded at Baghdad by Nizam-al-Mulk for the great philosopher aba-Isbaq ash-Shiraz. Soon, with the notables competing out of ardor, conviction, or a desire to flatter their master, the Moslem world was covered with madrasahs, Iran from the late eleventh century, the Arab world during the twelfth. Of the four rites two in particular were encouraged, the Shafiite, which was that of most Arabic-speaking easterners and of Nizam-al-Malulk, and the Hanafite, which predominated in Ithurasan and had thus become that of the Turks and of their sultans.

Among the mass of the people the dominant influence was that of the sufis who, because of their indifference to rites and laws, had often been unfavorably regarded by those in power, and who were riddled with heretical tendencies. But a new form of sufism was beginning to appear in the east, organized into congregations. Their rule was indeed outside the classical practices of Islam, but their influence might, according to circumstances, be exercised either in the direction of official orthodoxy or against it. The westward thrust of the Turks and Khurasanians promoted and accelerated the diffusion of these congregations. The Selchukids, their Persian vizirs, and their Turkish officers, sincerely devoted to saintly individuals and aware of their usefulness in the spiritual control of the urban masses, favored certain of these orders. At the same time as the madrasahs appeared the orders dotted the empire with their headquarters.

Finally, it is from this functional point of view, among others, that it is fitting to note the construction of numerous splendid new mosques and richly endowed hospitals, which served indeed to proclaim the glory of the dynasty, but a glory which it attached to all pious institutions susceptible of strengthening the Islamic social structure and binding it to the regime.

Paradoxical though it may sound, however, the Selchukid regime might in certain respects be considered rather non-clerical in comparison with other Moslem states. Power, although exercised in behalf of the Islamic faith, was in the hands of the sultan, whose role, in contrast to that of the caliph, was not primarily religious. It had been the same under the Buwaihids, but the very real priority accorded by the Selchukids to military and political matters, coupled with their intervention in spiritual affairs, meant for the "clerics", as well as material wealth and an enhancement of their social function, a decrease in their independence in that role.

Even in the structure of the Selchukid administration itself this secular characteristic was emphasized by an organic development. In the Abbasid and Buwaihid state, in addition to the daily justice of the magistrates the sovereign exercised a sort of supreme jurisdiction on appeal, the mazalim sessions. In spite of edifying anecdotes told about the great caliphs this justice does not seem to have been very effective. In the Samanid and Ghaznavid states, one has the impression that it acquired more actual importance, being directed by a special functionary named on the same level as the other great heads of state departments, the amir-dad. The Turks conceived of it as continuing their tribal tribunal, the yavlak. The Selchukids adopted and extended to the rest of their empire this institution which seemed so novel and so admirable to officials trained in the Buwaihid state, like the historian ar-Ruzravari. And even though in theory this justice was of course based on the principles of the religious law alone, it was in practice far more flexible than that of the qadis and more responsive to considerations of common sense and political utility.

It is difficult to say, in the present state of our knowledge, whether differences more fundamental than mere nomenclature and the exact division of responsibilities existed between the other great state services of the Samanids and the Ghaznavids – which the Selehukids adopted for their empire with their heads – and their counterparts in the Abbsid and Buwaihid domains. These were the vizir, the director of finances, the controller general, the steward of the palace and the royal domain , the supervisor of the army, and the director of the postal system. The provinces were similarly organized, and their civil governors were recruited, like the heads of the central departments, from an upper category of civil servants entitled amids. The garrison commanders did not, under the great Selchukids, encroach on the civil authority of the amids. In certain instances the administration of a district was farmed out. It could happen that the compensation of the tax-farmers, in place of or in addition to payment of cash, might include a land-grant, but these were never confused, nor was an administrative district ever treated by the official as a land-grant; the state was strong enough to assure respect for its rights. The information and espionage services, which were said to be repugnant to Tughrul-Beg, were nevertheless set up without delay.

It is not easy to disentangle and identify whatever portion of all this might stem from Turkish traditions, that vague heritage of administrative experience derived either from certain Turkish groups in the past or from transfer of nomadic Oghuz usages. The sultans remained Turkish in their personal and family lives, the emancipation of their women for example, and in their language. The seal with which they affirmed their sovereign authority on their decrees was in the form of that bow which had signified such authority among the Oghuz. Their safe-conducts were in the form of arrows for the same reason. More fundamentally, we have seen and shall again see persist among them the tribal idea of the preeminence of a family ruled by the eldest member, beside the contrasting Moslem idea of a monarch who would be succeeded by his sons. Finally, whoever glances over the whole of Turkish history, however cursorily, can hardly avoid receiving the impression that the temperament or the experience of the Turks, as compared to other peoples, had induced a sense of political and military command like that which the first Buwaihids had displayed in a lesser degree.

In the immigrant Turkish population there naturally persisted the traditions and some of the literature, mainly oral, of the Oghuz of Central Asia. It is difficult to determine whether some of this passed into certain aspects of the life and culture of the Iranians and Arabs in the Selchukid period. On the whole, while the Turkish aristocracy tended to adopt Iranian culture and to become diluted in the issue of mixed matings, the bulk of the Turkomans were, because of their nomadic way of life, a body foreign to the society in the midst of which they had come to live, though it seems likely that in certain regions they mingled with the native peasantry. The narrative of events has shown how these nomads were both an indispensable source of strength and a permanent danger to the regime.

It is difficult to compute the number (twenty or thirty thousand?) of these Turkomans or to be sure which among the twenty-four Oghuz tribes of Central Asia, most of whose names reappear among the population of the modern Near East, had already played, before the new ethnic dislocations of the twelfth and thirteenth centuries, a major role in the migration associated with the Selchukid conquest. The Avshar, the Doger, the Salgur and the Iva seem to have predominated. The Turkomans were generally able to stay in tribal groups, but fractions swarmed or were transferred to all corners of the vast new domains, and as a result of these movements, of the chances of war, and of discord, new groupings arose under chieftains who were not always members of the former ruling tribal families. They were numerous in eastern Iran, where many had stayed, in Fars and Khuzistan, on the great arteries of central al-Jibal province, and in Diyar-Bakr province. Above all they were massed in Azerbaijan, which has remained Turkish until today. There were also those who had ranged as far as Palestine or, in evergrowing numbers, had crossed Byzantine Armenia and reached the shores of the Aegean Sea.

In every case, as pastoral nomads, the newcomers had to try to procure grazing lands with a minimum of damage to the rest of the inhabitants. They were aided by their dispersion, by the loose pattern of agricultural utilization of the western Asian countryside, and by their concentration in frontier regions accustomed to receiving military settlers and to relying for food on their enemies. It was necessary to concede to them, or to their chiefs, vast fiefs suitable for grazing, inside which they would live in semi-autonomy.

An attempt to insure their fidelity was made by attracting to the court, through the promise of an education qualifying them for great futures, the sons of their notables, and by using them on occasion on productive military undertakings. Such was the case, for example, with Artuk, chief of a group of the Doger, who as a feudatory of Hulwan on the Mesopotamian edge of al-Jibal was employed by Malik-Shah in Anatolia, in Bahrain, (the Hasa coast of Arabia), and in upper Mesopotamia. There he was circumvented by the Uqailid and thence, at the death of the latter, he fled in fear of Malik-Shah to the service of Tutush, who bestowed Jerusalem on him.

What permitted the Turkornan force to be held in check was the regular army recruited from slaves. It was of the classical type of the armies of almost every nation of Islam at this period, and composed in large part of Turks, but, thanks to the conquests whose further extension it made possible, much larger, with 46,0oo or even 70,000 horsemen, according to unreliable medieval estimates. The economy of the Selchukid domain, which was for many reasons less mercantile than at the start of the tenth century, thereby rendered correspondingly even less practicable the creation of such an army by the sheer expenditure of money, or of property. The Buwaihids had installed and developed a system of supporting troops by the practice of distributing grants of land and its revenues. It is probable that Nizam-al-Mulk, in particular, perfected this system, applying it in a way which ended by interesting the concessionaires in improving their lands and by regularizing the responsibility of certain chieftains holding huge concessions for the maintenance of specified contingents. Thus there was what might be termed a feudal system functioning in the service of the state, which was able to maintain control by reason of the superiority of the resources which remained to it throughout its immense empire.

Although of course the regime functioned on behalf of the military and religious aristocracy, the reappearance of a regular administration and political unity after periods of fragmentation, and in places anarchy, seems to have given a feeling of relief to the people in general. After Alp Arslan, with his aura of military glory, Malik-Shah and Nizam-al-Mulk appeared in the eyes of posterity as the ideal sovereign and vizir.

The remarkable fact is that this was true not only of Moslems but of Christians of all sects. Of course the ecclesiastics deplored the territorial losses sustained by Byzantium, and they all lamented the ravages of the Turkomans, but they generally held the Byzantines responsible for the former and were the more grateful to the great peacemaker Malik-Shah for his praiseworthy suppression of the latter. Whether one listens to the testimony of the Armenians, Matthew of Edessa and Sarkavag, or the opinions of the Syrian Monophysites transmitted later by Michael the Syrian, or those of the Nestorians recorded in the chronicle of Amr, or even those of Copts outside the Selchukid domain as given in the History of the Patriarchs of Alexandria, Christian sentiment is unanimous, even in writings subsequent to the death of Malik-Shah and therefore free of any imputation of venality.

This favorable opinion was even extended, as soon as the Turkoman ravages ceased, to chiefs like Sulaiman at Antioch or Artuk at Jerusalem. This was so because even though the ghazi spirit entailed the subjugation of unbelievers, it did not allow their persecution after submission, thus resembling the spirit of classical Islam. Moreover, in the frontier regions, where Selchukid control was less effective, the struggle between the Turkomans and the old Arab or Byzantine aristocracy worked, insofar as any administration continued, to the advantage of the natives, including numerous Christians of churches happy to be free of the trickery of the Greek clergy. And even the Greek patriarchs of Antioch and Jerusalem could stay in their metropolitan sees; the former was to be imprisoned and the latter expelled by the Egyptians only in reprisal, upon the approach of the army of the crusaders and the Greeks. The sole persecution of Christians inside Islam was that of al-Hakim, the half-insane Fatimid caliph, fifty years before the appearance of the Turks in the region of Iran.

Also it would be absolutely incorrect to imagine, merely because the crusades did occur, that the native Christians had hoped for liberation from outside. Of course some Hellenized elements might have hoped for a Byzantine resurgence in Anatolia or Armenia, but naturally, although they had heard of Frankish mercenaries, they could have conceived of no reconquest other than Byzantine. But the immense majority either were reasonably satisfied or else, if they had anything to complain of, placed their hopes on Malik-Shah. The most that can be said is that in the disorders which were to follow his death, those hopes would no longer have a focus. It has been established that no oriental appeal, except Byzantine, was ever sent westward either to the pope or to anyone else. It may be added that such an appeal would in no respect have corresponded to the mentality of the Christians of the orient. When these latter, after the event, wished to explain the crusade, they borrowed from the occident their explanation, the mistreatment of the pilgrims.

Certainly pilgrims, who often took the land route by Constantinople, suffered from the loss of Anatolia and the anarchy prevalent there. Some of the pilgrims might even have suffered at Jerusalem itself, because of the disorders at the time of Atsiz. But it should not be forgotten that we know of robberies of pilgrims by bedouins before the Turkomans arrived, and we know of none committed by the Turks. In any event, these grievances applied only to certain places at certain times of disorder. By sea Mediterranean commerce and pilgrimages continued. Of course the Turkoman holy war had been a catastrophe for Byzantium, but for it alone. Perhaps it was the very vigor of commerce and pilgrimage which made what had previously been endured without difficulty suddenly seem intolerable, especially since Byzantium was no longer able to extend to Christians in its jurisdiction the protection which it had provided for three generations. On the contrary, the Latin influence among them was increasing. The schism between Constantinople and Rome dating from the middle of the century caused only slight echoes in Antioch and Jerusalem, even among the Melkites, natives Greek in faith and Arab in speech. The idea of taking over in the orient from a weakened Byzantium might have arisen in Rome. It is not extraordinary that in poorly informed western Europe the remote and the recent past should be confused, and that such a confusion, perhaps skillfully induced, should envisage a Byzantine disaster as a great hardship for the eastern Christians.

Not all, however, was strength with the Selchukid empire. The moral cohesion was not complete. It was not that there had been grave moral friction between Turks and natives. But there remained heretical Moslems. The Ismailite propaganda, directed from Cairo, had not disappeared. Hunted down, it had become more secret. The dissensions which, in Egypt in the final quarter of the eleventh century, had ranged those who remained faithful to the ruling Fatimids against the partisans of the ousted prince Nizar had weakened the control of the Fatimids over the propagation of Ismailism.

The dissident faction was reorganized into new autonomous forms and, as was normal in view of the terror hanging over them, its adepts themselves became terrorists. Their history is to be found treated in detail in another chapter. Here it will suffice merely to recall that the new sect, founded by al-asan ibn-as­Sabbah (Persian, Hasan-i-Sabbal) — whose followers, lured by the joys of hashish were termed ashishiyah — succeeded in creating, late in the reign of Malik-Shah, a formidably defended camp around the fortress of Alamut, in the Elburz mountains south of the Caspian Sea. From there action was initiated in the form of those political murders which gave its present meaning to the word "assassin", derived from hashishi. These exploits spread far and wide the dread of the Assassins, whose first victim of note — if indeed they were the perpetrators of the crime, which was incited by his other foes —was the vizir, Nizam-al-Mulk.

A second danger lay in the nature of the dynasty itself. As has been seen, the Selchukids never entirely abandoned the tribal concept of power. Among the Oghuz, as among the Buwaihids, there prevailed the idea of tribal government less by a prince who was to be succeeded by his sons than by a family whose eldest members were chiefs in turn. Nizarn-al-Malk had been able to make the monarchist principle triumph on behalf of Alp Arslan's son, Malik-Shah, by ousting Kavurd, the eldest of the family, but the familial idea was nevertheless to persist visibly until the end of the dynasty. Even during the lifetime of Malik-Shah it was strong enough to force the minister and his sovereign to consent to share the power by distributing appanages to the princes "of the blood" such as Tutush. Even among the sons of the ruler no Moslem dynasty was ever able to fix the order of succession by primogeniture or otherwise, and polygamy aggravated this difficulty by adding the rivalries of the women to those of their sons, Finally, the semi­feudal system gave power to a small number of great chiefs, the danger of which is illustrated by too many examples to need particularization.

These perils were not so great when there was in power a capable prince, wise enough to keep in his possession all the necessary resources. Already the minority of Malik-Shah could have given rise to serious dangers if it had not been for the strong personality of Nizarn-al-Mulk and the resources in his control dating from the reign of Alp Arslan. When Malik-Shah died young in 1092, closely following his great minister, he left only small children with ambitious mothers and no vizir in control of the situation. The caliph, in spite of his desire to do so, could not impose his moral authority to arbitrate. Hence there developed quarrels among the sons of Malik-Shah and between them and his brothers, their uncles, each supported by his adherents and the adherents of vizirial rivals, these uncles being enemies of the family of Nizam-al-Mulk.

This situation resulted in a partition of the empire, devastation, administrative disorder, and universal usurpation. For what had begun in 1092 got worse with every later change of ruler. Each prince in an effort to secure allies disposed of resources and territories and thus weakened himself by that much. They died young and left their infants in the care of military chiefs (Turkish singular, atabeg) whom they judged, or rendered, strong enough to be able to defend their rights; inevitably these atabegs worked above all to secure for themselves the real power and expected some day to liquidate a nominal dynasty which had become useless.

To these struggles the Turkomans, especially in Fars and Azerbaijan, were always ready to lend their weight, for they no longer had other outlets. The road to Asia Minor was blocked by their kinsmen; a stable Christian kingdom had been established in the mountains of Georgia to resist the invader; and a certain attachment to the sail kept them from planning great new migrations. It was doubtless in order to keep these Turkomans under tighter control that the sultans constantly bestowed Azerbaijan as an appanage or an autonomous march, but the scheme invariably boomeranged because the grantee found there an army ready for any revolt. The Kurds, including the Shanbankarah of Fars and others, the Lurs, the bedouins, the Khafajids of Khuzistan, all profited from the disorder, as did especially the Mazyadids of Hila, who ranged from the outskirts of Baghdad itself as far as Basra and who, under Sadaqah and his son Dubais, made life miserable for the caliphs and sultans for the first quarter of the twelfth century.

Asia Minor permanently escaped any effort to incorporate it into the Selchukid empire. The Byzantine administration had disappeared there, but no Moslem administration had yet established itself, for lack of native Moslems. In places the inhabitants had fled. The Turkomans were the rulers and sometimes in the rural districts were the only residents. There, truly, one was outside the classical world to such an extent that for generations the Moslem chroniclers ignored almost everything that happened in that area. But it was this void itself which was to make Asia Minor more important in Turkish history than the Selchukid empire; the Turks flowed thither, and it was there, and not in the empire they had won for their first chiefs, that they created a new "Turkey", which alone bears that name today. From the start, on the Arab side, the limits of Turkish habitation were almost where they are today. Perhaps, if there had been no crusade, the most important of these Turks would have been then, as they were to be later in the time of the Ottomans, those on the shores of the Straits and those who, farther south along the Aegean, joined the traditionally maritime natives to become corsairs. The crusade and the accompanying Byzantine reconquest pushed them back onto the plateau, and Iconium succeeded Nicaea as the residence of their sovereign; the disaster inflicted on the Crusade of 1101 proved that their control of the plateau was effective.

After the death of Malik-Shah the theoretical sovereign of Asia Minor was a son of Sulaiman named Kilij (or Kilich) Arslan, who, being called Ibn-Sulaiman, was known as Solomon to the crusaders. He had escaped from his Selchukid relatives in Iran. But though he directly dominated the road from Nicaea to Iconium and the passes of the northern Taurus range farther east, he was not master of all Asia Minor. In Armenia, facing the Greeks of Trebizond and the Georgians, Turkotnan chiefs who were to attract attention were established at Erzurum — the Saltukids —and at Erzinjan —the Menguchekids. Farther west, on the northern roads, Sebastia (Sivas), Amasya, Caesarea (Kayseri), and Ankara belonged to a man whose descendants would be very important, but whose connections with the Selchukids are obscure. This was a Turkornan chief whose Persian title of danishmend suggests that his power had the spiritual origin which was mentioned earlier in this chapter as attaching to that title. Thus there arose in Anatolia an opposition — which the captivity of Bohemond would illustrate — between the Turkomans, interested primarily in raiding the Greeks, and the Selchukid princes, whose strength rested on the Turkomans, but who sought to organize, with the help of some Iranians in their entourage and of an alliance with the Byzantines, the rudiments of a government, and to return, if opportunity offered, to play their part in the quarrels of their cousins to the east. To do this they had to make sure of their liaison with the Turkish hinterland, but this was also the concern of Malik-Ghazi ibn-Danishmend who was eager to keep open the path of Turkoman reinforcement; hence their rivalry for the possession of Melitene, which after Bohemond's capture in 1100 Malik-Ghazi took in 1103 from its Armenian chief, Gabriel, and which Kilij Arslan occupied in1io6 after the death of his rival.

But the appeal which, as will be seen, the upper Mesopotamian chieftains in revolt against their sultan sent him on that occasion was to culminate for Kilij Arslan in his defeat and death during 1107. Thereafter, the Turks of Anatolia, cut off from their kinsmen to the east, would have to govern themselves in isolation. When non-Turkish Moslems gradually resumed relations with them, these Moslems would be Iranians and not Arabs, because the establishment of crusaders from Cilicia to Edessa impeded communications between Anatolia and Arab Islam, at least in Syria, which was nearest.

Within the Selchukid empire proper, Syria and upper Mesopotamia, regions which the crusaders were to reach, were the first to break up. At the death of Malik-Shah his brother, Tutush, had desired to claim his heritage. He was recognized in Syria and upper Mesopotamia, but, after he had conquered and killed Buzan and Aksungur al-Hajib, who had deserted him, he died in battle in Iran in 1095. His sons, Ridvart and Dukak, fell out, with each taking part of his realm, the former at Aleppo and the latter at Damascus and in the province of Diyar-Baler. New tensions embroiled the former with his atabeg, Janah-ad-Daulah, who entrenched himself at Homs, and with Yaghi-Siyan, still, master of Antioch. None of these chiefs, in these circumstances, had any real power at his disposal. Moreover, the Turkomans had abandoned Syria and Palestine, bringing ruin to the Turkish populations of these lands. Led by Tutush to the conquest of upper Mesopotamia, they had stayed there, mingling with their kinsmen who had never left.

Thus the princelings of Syria, when the crusaders arrived, had for making war only the handful of slaves which the revenues from their meager provinces enabled them to buy. The local pride of the Damascenes, their Sunaism, the protection afforded by their geographical situation, and the skill of Dukak's atabeg, Tughtigin, unified them around these two leaders. But Ridvan, surrounded by Arabs who were largely Shiite, held in check by the armed townsmen, and knowing no other source of help, relied, after a Fatimid interval, on the Assassins, who thus acquired a foothold in Syria. Obviously the crusade, by stripping these princelings of their richest districts, along the coast, and by posing a constant threat to their security, further intensified their impotence. In the cities the real leaders were the notables, Sunnite or Shiite, qadis or headmen, together with their adherents and militia Shiite qadis of the Banu-l-Khashshab and Sunnite headmen of the Banu-Badi of Iranian origin at Aleppo, to a lesser degree headmen of the Banu-s-Sufi at Damascus, and qadis at several coastal ports, of whom the most illustrious were the Banu-Ammar, whom we shall meet again.

On the other hand, Syria and upper Mesopotamia have always been lands of intense geographic, social, religious, and ethnic fragmentation; there had been no opportunity there for the religious unification which elsewhere mitigated the political disunity, but on the contrary the opposition between the new orthodox princes and the frequently Shiite people introduced an extra element of moral division. Dynastic fragmentation often found support in local particularism, and the resulting weakness left a free field for others. Arab lords sprang up, like the Banu-Munqidh at Shaizar on the Orontes, whose life mingled literary diversions with hunting and the petty wars which the Franks were to find so familiar. The Nusairis were fairly independent in their mountains; the Ismailite pro-Fatimid Khalaf ibn-Mulaib set himself up at Apamea; at Tripoli the family of the Banu-Ammar, sheltered between mountain and sea, for a third of a century constituted an autonomous non-Ismailite Shiite principality, spiritually and materially prosperous and untroubled by the Turkomans. The Arab tribes, such as the Numairids around Harran, freed themselves, while the Armenians further north found themselves free by default; at Edessa, at Melitene, at Marash, and elsewhere the crusaders found them under the command of their own leaders, Toros, Gabriel, and Kogh Vasil. And the decline of the Turkish power in the south allowed Egypt, which had been reorganized by the vizirs Badr al-Jamli. and al-Afdal, to regain the ports, though the intervention of the crusaders was required to induce them to retake Jerusalem itself from the sons of Artuk, who had died in 1o91.

The Syria to which the crusaders were to come was thus, of all Islamic regions, the least capable of resistance. The loss of the coastal strip would add to its impotence. It was in upper Mesopotamia, to which it was bound geographically and which had already so often absorbed it politically, that it must find help. As in proportion to the increase of Frankish power such help became more urgent, and as Diyar-Bakr and Mosul had meanwhile become stable local states, it became more and more inevitable that Aleppo at least would rely on their aid and hence come under their sway. The history of the first three decades after the First Crusade was to confirm this conclusion. But it was an irregular process, for these helpers themselves were sometimes paralyzed by the internecine wars of Iraq and Iran, or when this was not the case were arrayed against each other. In any event the Syrians could not view without distrust these "orientals" whom they suspected of aspiring to replace them. This fear was so strong that, as will be seen, it was to lead the Moslems of Syria to ally themselves on occasion with the new Syrians, which in a sense the Franks were to become, against those very foes whom they had on previous occasions summoned for help against them.

Unexpected as it may appear to the westerner, it must be clearly realized that the crusades did not produce much of an impression on the Islamic world in general. In the traditions of the Turkomans of Anatolia almost no trace was left by the crossing of the Frankish army. Of what importance was it, in fact, to the nomads that they had been roughly handled in regions of which they had promptly regained control, or that they had lost some towns outside their grazing area Moreover, at first the crusade was considered as related to those earlier Byzantine expeditions, ephemeral and limited to territories traditionally accustomed to frequent changes of masters, incompletely converted to Islam, distant from Baghdad and Cairo, and negligible since commerce never suffered from the changes. They had supplied the opportunity for worthy exploits and for romantic encounters sung on both sides of the frontiers by the poets in the circle of Saif-ad­Daulah or in the Byzantine Digenis. At most it was deemed necessary to try to reduce the ravages of the unbeliever; his expulsion was hardly imagined. Among the eastern Christians is to be found, in their description of the crusades, a certain amount of oratorical exaggeration, but even there difficult to appraise, as stylistic emphasis was usual with them. Among the Moslems, even in those of their narratives which have survived — all were compiled considerably later than the crusade and had already undergone fundamental revision — the wars with the Franks were invariably treated like any other wars. In the literatures of Iraq and Egypt these wars were scarcely mentioned, in that of Iran not at all. It was to be the length and nature of the Frankish occupation which would gradually provoke a reaction. At the start the crusaders were merely one more pawn on an already overcrowded political chessboard, a pawn indistinguishable from its fellows. The trend of history in the surrounding region was not at all affected by it.


Of the history of the later Selchukids in Iran and Mesopotamia only the broad outlines are appropriate to this work. Before the crusade, Berkyaruk, the eldest son of Malik-Shah, had triumphed successively over an infant brother, who soon died, and, in 1095, over his uncle Tutush. Between 1097 and 1099, while the crusaders were conquering Syria, he was subduing another uncle and various relatives in Khurasan, and taking the grave step of constituting it the appanage of his brother, Sanjar. Scarcely had this situation been thus regulated when he was faced with the revolt of another brother, Muhammad, with whom, in 1103 after four years of war, he decided to share the sultanate. His death in January of 1205 permitted the energetic Muhammad to reunite the remains of the power of western Iran and of most of Mesopotamia in the capacity partly of a sovereign and partly of a leader of a confederation. At least he could now divert the ambitions of certain great chieftains toward the pursuit of a policy of counter-offensive, in the line of Selchukid tradition, against the enemies of Islam whether external — the Franks — or internal — the Assassins, the former perhaps as a pretext and certainly as an occasion to attempt to restore his preponderance in Syria.

Khurasan, however, owed to the longevity of Sanjar, who lived until 1156, a calmer internal history. The reign of this prince, whose last years were so difficult, and whose death was so tragic, had opened with three decades of effective rule; he made laws at Samarkand for the Kara-Khanids and, what Malik-Shah had never done, at Ghaznah for the Ghaznavids. Muhammad's death in 1118 made him the eldest of the Selchukid family. Without aspiring to reunite the whole empire under his sway, he insisted that his nephews accord him a certain primacy. His intervention at the succession of Mahmud safeguarded the unity of the whole; Mahmud could neutralize his brothers Tughrul and Masud and the Mazyadid chief Dubais sufficiently to assist in the war against the Franks with whom Dubais was now allied, and to participate personally in organizing a campaign against the Georgians.

Under Mahmud's successor, his brother Masud (1131-1150), the disintegration was accelerated. Six years of fairly savage warfare against Sanjar, Tughrul, his nephew Daud, the caliphs al­Mustarshid and ar-Rashid, and Dubais ended, it is true, by assuring him of victory and a monopoly of the sultan's title. But of what did this sultanate consist? Fars, Azerbaijan, and soon Iraq, not to mention more distant or smaller territories, constituted autonomous principalities. Even the sultan, at the end of his reign, was the prisoner of chieftains who shared the spoils of the empire and from whom he could only rarely gain an illusory liberty by intriguing to shatter their fragile coalitions. His successors would be mere powerless wards of the atabeg of Azerbaijan whom we should hardly mention except that the last of them, Tughrul, at the end of the century won a final pale reflection of the glory of his ancestors by dying in battle against the troops of Khorezm.

The emancipation of Iraq deserves special mention, because it also involved the emancipation of the caliph. The diminution of the revenues of the sultans had led than to consider Iraq as their last financial reserve, and thus rendered their authority harsher to the caliphs at the same time that it became less justified by services rendered to the Moslem community. But elsewhere, in the rivalries of pretenders, the caliph was sought as arbitrator, and he sold his awards high. Gradually he recovered a real measure of autonomy, at the head of a principality in Iraq analogous to the others. Even the Turkish soldiers, fearing the vengeance of a conqueror, entered his service. But the winning sultan was not always the one he favored, and even when he was, this independence of the caliph at the time that the sultan had greatest need of the resources of Iraq necessarily led to conflict. The gravest of these occurred (1134-1138) during the reign of Masud. It ended with a fierce siege of Baghdad, the successive execution of the two caliphs al-Mustarshid and ar-Rashid, and the forced installation of the candidate of Masul, al-Muqtafi. But the decline of the sultanate nonetheless produced under this same al-Muqtafi the result which Masud had sought to avoid. By the middle of the century the caliph was an autonomous territorial sovereign, perhaps more than he was a real caliph, to judge from his remarkable indifference to the holy war.

Up-river from Iraq, the province of Mosul was, in the first quarter of the twelfth century, a kind of autonomous march whose governor was usually designated by the sultan and charged both with the holy war against the Franks and the reduction to obedience of the Turkomans of the upper Tigris and the Syrian princes. After the occupation of Mosul by Tutush and his subsequent death, the city fell into the hands of a former freedman of Aksungur al­HajibKerbogha, who had had himself recognized by Berkyaruk and was to gain fame among the crusaders through the disastrous campaign he undertook against them in 1098. At his death in1102 he was replaced, thanks to the wars between Berkyaruk and Muhammad, by the governor of Jazirat-Ibn-Umar, Chokumish, whom Berkyaruk approved. But the reconciliation of the two princes, with Mosul falling to the lot of Muhammad, and the subsequent death of Berkyaruk complicated the situation of Chokurmish, who was attacked in 1106 by the successor whom the sultan had designated, Chavli SaqavehChokfumish died in the fighting. It was then that his son appealed to Kilij Arslan, whom almost all the local chiefs at first supported against the return of Muhammad to power but then deserted when they realized that they had merely exchanged one master for another, causing the disaster of Kilij Arslan at the Khabur in 1107. Chavli Saqaveh, however, in his turn quickly became suspect to the sultan by too independent behavior. We shall see how, when the sultan sent Maudud against him, he went so far as to ally himself in 1108 with the Franks of Edessa, but then received his pardon and the governorship of Fars. Maudud conducted four campaigns against the Franks, with uneven results; during the final one he was "assassinated" in 1113 at Damascus.

Aksungur al-Bursuki, who replaced him, remained for only one year, because of the failure of the campaign he undertook in against the Franks. He stayed at Rahba, however, and later, after having been governor of Iraq, regained the governorship of Mosul in 1124, while in 1126 his son was to be the last lieutenant of the sultan there before Zengi. Meanwhile, in 1114, Muhammad named to Mosul Juyush-Beg, as atabeg of his second son Masud, but this time the command of the holy war was entrusted not to the atabeg but to a great emir from Hamadan, Bursuk ibn­Bursuk, who, with the Kurd Ahmad-il of Maragha and Sokman of Akhlat (or Khilat) on Lake Van, had already participated in the preceding campaigns. The campaign of 1115 culminated in disaster, as will be seen, and thereafter for ten years no expedition into Syria would be organized at Mosul or elsewhere. Juyush-Beg spent ten years at Mosul until 1124, when he was compromised by the attempts at insubordination of his pupil Masud against sultan Mahmud, who replaced him with his predecessor Aksungur al-Bursuki. The latter and his son were the last governors of Mosul who can be termed dependents of the sultan. Under Zengi, still in theory atabeg of a Selchukid, the civil war between Masud and the caliph, in which Zengi was to participate, would end in fact in the complete independence of the latter. His successors would retain power without being even in theory atabegs of any Selchukid.

None of the governors of Mosul had succeeded, on his own behalf or on the sultan's, in subjugating the Turkomans of Diyar­Bakr province. On the contrary, the earlier fragmentation of the region had gradually given way to a territorial concentration for the benefit of a Turkoman family, that of the Artukids, whom Zengi would later partially reduce but not evict, and who would survive until the fifteenth century as vassals of all the subsequent empires. The principal city north of the Tigris, Maiyafariqin, successively center of the governments of Dukak, who had inherited it from Tutush, of Kilij Arslan, and finally of Sokman of Akhlat, the vassal of Muhammad, was not destined to fall into the hands of the Artukids until 1118. On the Tigris, Amida (Diyar­bakir) would be until the time of Saladin the capital of a small autonomous principality. Various Turkoman chiefs, between the Tigris and Lake Van, subsisted as vassals either of the Artukids or of the "Shah-i-Armin", like Kizil Arslan, probably the "Red Lion" of the crusade poems. But the most important and most renowned family was always that of the descendants of that Artuk whom we have met in the service of Malik-Shah and Tutush.

Ousted from Jerusalem by the Egyptian and Frankish conquests, and from Saruj, between Syria and upper Mesopotamia, by the Frankish conquest, the Artukids thereafter made a career both in the service of the sultans — as did al-Ghazi in the time of Berkyaruk — and as chiefs of the Turkomans in the land of the upper Tigris where their father had once brought them. To the flat country, which they doubtless dominated very quickly, were added Mardin in 1097, Hisn Kaifa in 1102, Kharput in 1115, and Maiyafariqin in 1118, not to mention Aleppo, which they held six years but did not keep. At the start the best known of them, as much in Diyar-Bakr as in Syria and the Frankish county of Edessa, was Sokman ibn-Artuk. After his death in 1104 his brother il-Ghazi, whose sojourn in Iraq was terminated by the accession of Muhammad, came to Diyar-Bakr to assume the leadership of the family. By the end of his life the family had become a real power, was allied with Dubais in Iraq, and was importuned for aid against the Georgians, and against the Franks to such a degree that, under conditions which we shall detail, the Aleppans were to offer to surrender to them. Il-Ghazi's nephew Belek, who had carved out his own domain around Kharput in spite of a coalition of Danishmendids and Greeks from Trebizond, succeeded him briefly (1122-1124) at Aleppo and in the holy war.

In the struggles of the sultans against each other or of agents against foes of the sultans, as well as in the holy war, the Artukid policy was a perpetual double game with a single goal, the acquisition and retention of autonomous territories. Sokman had participated in the expeditions of Kerbogha in 1098 and of Chokurmish in 1104 against the Franks. Il-Ghazi, embroiled simultaneously with Chokurmish and with sultan Muhammad, but avoiding direct conflict with the latter, fought both Kilij Arslan, an ally of the son of Chokurmish, and partisans of Muhammad like Sokman of Akhlat, who, opportunely for him, died in 1110, and Aksungur al­Bursuki, whom he opposed in 1114. When, however, a large army of the sultan needed his support for the holy war, he joined in 1110, 1113, and 1114 only to desert, contributing by his equivocal attitude to its divisions and setbacks, when Sokman of Akhlat was with it in 1110, for example, or Aksungur al-Bursuki in 1114. Threatened with vengeance by the sultan in 1115, as we shall see, he joined the other adversaries of the sultan in Syria, including the Franks, and, after a grave accident, was one of the architects of their triumph over the sultan's army in that year. Once the danger from the sultan had been cancelled, he deserted the friends of the Franks - now uneasy at the power of the latter — to resume on his own account, at the call of the Aleppans, the holy war which was to redound to his benefit,

After the deaths of Il-Ghazi and Belek, the family remained permanently divided into two branches, one descended from Il­Ghazi at Maiyafarigin and Mardin whose representative at the time of Zengi was to be Timurtash, the other descended from Sokman at Hish Kaifä and Kharput whose representative was to be Daud. The coins of the first Artukids are famous for bearing the Christian creed of the native artisans on whom they had to rely to coin them. Later Artukids were to become ordinary territorial princes and participate in the general movement of orthodox reorganization which the Zengids were to initiate.

In Syria the eviction of Yaghi-Siyan from Antioch by the crusaders and the assassination of Janad-ad-Daulah of Homs in 1103 left Ridvan at Aleppo and Dukak at Damascus practically alone in the field. The death of the latter in 1104 marked the end of the dynasty in that city, for his son and relatives were set aside by his erstwhile close collaborator and atabeg, Tughtigin. Actually this collaboration meant that no change of policy resulted from the change of family. Though Damascus, better governed perhaps, better protected from the Franks, less directly in the path of oriental ambitions, had on the whole a firmer and better balanced policy than Aleppo, and though Dukak and Tughtigin stood for orthodoxy while Ridvan was the accomplice of the heterodox, their policies were similar in the distrust they felt for each other, for the Franks, and for the easterners. Thus there arose all the combinations and shifting alliances, to try to save their little holdings by offsetting and neutralizing one another. This mediocre policy exasperated their subjects, particularly at Aleppo, conscious of the over-riding necessity of forming a common Moslem front to meet the Frankish menace. It is essential, however, for the understanding of the vicissitudes of the Frankish conquest, to summarize also the zigzag politics of Aleppo and Damascus.

Ridvan tried in general to banish the Frankish danger with cash, and not to wage war except with minimal risk. He had scarcely defended Antioch and had not participated in the expedition of Kerbogha, in which, on the other hand, Dukak and Janah-ad­Daulah had figured. Although he had in 1104 risked profiting from the Frankish defeat on the Balikh, he was to lose in the following year the districts he had acquired. Although an "assassination" had rid him of the pro-Fatimid Khalaf of Apamea, he allowed the place to fall into Frankish hands in 1106. In 1106 and 1107 he helped his former subordinate Il-Ghazi against Chokurmish of Mosul, and then against Kilij Arslan of Anatolia, in the hope of getting equivalent reinforcements against the Franks, but in 1108 he was allied with Tancred against Chavil Saqaveh, the new lord of Mosul and ally of Baldwin of Edessa, an alliance which on both sides crossed religious lines to satisfy personal quarrels. Among the Shiites as among the Sunnites, Ridvan had the reputation of being a rapacious miser, but he bought a precarious peace from the Franks at a very high price.

Meanwhile Dukak and Tughtigin were consolidating their power over central and southern Moslem Syria, even installing vassals at Homs —Karaja —and at Hamah—Ali Kurd. They directed their policy of defense against the Franks toward an alliance with Egypt, disregarding former sectarian differences, and helped with the land defense of the Syrian ports which it defended by sea. They did not neglect opportunities for territorial aggrandizement which appeals for help from local rulers offered them; and, having no desire for the expulsion of the Franks, which would leave them in dangerously direct contact with Egypt, readily agreed to arrange truces with the Franks or to abstain from serious hostilities. They had embroiled themselves with ibn-Ammar of Tripoli by supporting Jabala's revolt against him in clot. Ibn-Ammar could no longer count on Egypt, which aspired to reconquer his city. At the start he had, like the others, willingly treated with the Franks. When he had nevertheless to defend himself against them, he was one of the first to send an appeal to Baghdad, where he went in 1108 and would later end his career after the fall of Tripoli. Three years later the Aleppans made a noisy demonstration in the open mosque at Baghdad, to shame the Moslem world for its disunity in the face of the Frankish peril. Like Ibn-Ammar most of these men, who pinned their hopes on the capital of Sunnite Islam, were Shiites, proof that for the people and some at least of their chiefs, sectarian differences were disregarded in times of danger, and that Moslem solidarity was beginning to develop in reaction to past divisions.

After 1110, as we have seen, the sultan, whose policy this newly born movement complemented, was organizing expeditions against the Franks, the first directed only at Edessa, the others into Syria. Ridvan tried to profit from them by participating as little as possible, and by hastily quitting them to buy his pardon from the Franks, dreading a coalition of his subjects and the eastern emirs against him. In 111 he asked these latter to raise the siege of Tell Bashir and hurry to succor Aleppo, and, when they arrived, refused to let them enter the city or to join them in the countryside. When Tughtigin came to meet them in order to try to lead them, by­passing Damascus, to retake Sidon or Tripoli from the Franks, Ridvan tried in vain to have him "assassinated", but then made friends by sending tardy help to save Tyre in return for formal recognition of his sovereignty over Damascus.

In 1113 a double "accident" occurred. At Damascus the commander of the eastern army, Maudud, was "assassinated". Although it was almost certainly an act of vengeance by the Assassins against Maudud, who had been their fierce enemy in the east, public rumor aimed at Tughtigin an accusation symptomatic of the atmosphere of universal distrust. Tughtigin, until then much more favorable to the sultan's expeditions than was Ridvan, since they menaced him less, at once became suspect in connection with the holy war, and was rebuffed into alliance with the Franks. Then Ridvan died. The population, weary of reprisals against the Assassins, forced the young Alp Arslan, his son, to have them massacred; but by doing so, he deprived himself of his only possible support. He tried to put himself under the protection of Tughtigin, but thus aroused the distrust of the Shiite majority; finally he was in his turn slain.

That was practically the end of the Syrian Selchukid dynasty. The slaves of Ridvan and the civic notables who one after the other, in the midst of universal anarchy, tried to take the reins of government had insufficient strength either to impose their authority on all the residents of Aleppo or to raise effective armies for the defense of their territory. Fearing the army of the easterners, in which they well knew the people had put their hopes, they too were thrown into alliance with the Franks. Il-Ghazi, as we have seen, had also broken with the sultan's party. In 1115is Tughtigin, the Aleppans, and Il-Ghazi made common cause with the Franks against the army of the sultan under Bursuk ibn-Bursuk, who had come intending to fight them as much as he had to fight the Franks. It is true that the sultan had found a new partisan in Syria, which he had promised to concede to him, in the person of Kir-Khan, son of karaja, who, at Homs, hoped to liberate himself from the control of Tughtigin, and who once captured Il-Ghazi, though Tughtigin forced his release. It will be seen how this situation culminated —partly because of the jealousy of the easterners for Kir-Khan — in a Frankish victory more complete than Tughtigin had wished. The latter then felt it necessary to visit Baghdad to make his peace with the sultan, bringing back in 1116 an official investiture.

The disaster of 1115 meant for the sultan a permanent check to all his Syrian dreams, and the recognition of Tughtigin was the only way for him to save even appearances. For Tughtigin, now that all the Selchukids of Syria had vanished, it conferred the legitimate succession on him as opposed to his adversaries. Two years later Muhammad died, and under Mahmud there was even less likely to be any resumption of activity in Syria by the easterners, at least before an internal reorganization which the sultan could not accomplish. This does not mean that there was to be no more collaboration between Syria and Iraq; on the contrary; but henceforth it would be with the autonomous princelings of Iraq, whom the policy of the sultan no longer restrained, and who concluded treaties of mutual advantage with the Syrian states, or who at least, being less foreign, were more readily accepted by the Syrians. Before the time of ZengiAksungur al-Bursuki, recently repulsed by some as oriental, but having become semi-Syrian at Rahba in the interval, would be summoned by the same ones in 1124.

The pattern of the Asiatic Moslem world was thus about to be reconstructed on a new framework, Iran, and to a lesser degree Mesopotamia, were to survive almost entirely apart from the western provinces, so much so that in connection with the crusades their further history would be irrelevant were it not for the grave events then being prepared in Central Asia which would, in the following century, brutally reintroduce them into Syrian history. A new alignment of regions, from Mosul to Aleppo, then to Damascus and on to Cairo, would arise and take over the lead, not only in the battle against the Franks, which is self-evident, but also, and perhaps partially because of that, in the whole of Moslem, especially Arab, life.

This is not to say that there did not remain from the Selchukids, in default of political unity, an important heritage, even in the old Moslem countries. In some places a Turkish population, and almost everywhere an aristocracy under Turkish command, had superimposed themselves on the former inhabitants. A policy of orthodoxy had been initiated, and all the subsequent regimes until the Mongol invasion would follow it. The reaction which the Frankish invasion provoked little by little among its Moslem neighbors did not result from it, but the organizational forms it would adopt followed exactly the lines of Khurasanian initiative which the Selchukid conquest had strongly developed in extent and in depth. Nar-ad-Din and Saladin are inconceivable without Tughrul-Beg and Nizarn-al-Mulk.