The historical evolution of medieval Anatolia in the centuries between the decay of Byzantine rule and the Ottoman conquest is closely linked with intricate processes of migration, cross-cultural encounter, and ethnic change. The area in question includes what the Byzantines with a very generic terms used to label ἡ ἑῴα or ἡ ἀνατολή, i.e., “the East”. ¹ After various expansionist stages that culminated in the reign of Basil ii (976–1025) the empire’s eastern provinces stretched from the western coastland of Asia Minor as far as northern Syria, the Upper Euphrates region, and the Armenian highlands. At first, the political, cultural, and ethnic transformation of this area began as a fortuitous side effect of the rise of the Great Seljuk Empire in the central lands of Islam. A ruling clan claiming descent from a common ancestor called Seljuk and superficially Islamized nomadic warriors, who drew their origin from the Turkic Oghuz tribes dwelling in the steppe lands of Transoxania, formed the driving force of this new empire. In the 1040s, Turkmen hosts made their first raids into the region south of the Anti-Taurus range and invaded the Armenian highlands between the Araxes (Aras) and the Arsanias (Murat) Rivers. Soon it turned out that the Taurus Mountains, which for centuries had formed a natural barrier between Christian-Roman and Muslim territories, had become permeable. ²

In what follows I shall present a survey of salient patterns of expansion, migration, and settlement, which Turkish warriors and migrants evinced from the time of their first appearance in the eastern borderland until the emergence of Turkish-Muslim domains in Anatolia. In this context, it is important to examine the correlations and reciprocities between the political and social characteristics of these incoming groups and the internal situation of Byzantine Asia Minor before and during the Turkish expansion. The latter aspect is closely related to the notion of the 11th-century crisis, which modern historians have frequently used as an explicatory model for the rapid collapse of the Byzantine central government and military power in the decades after Emperor Basil ii’s death in 1025. ³

The Turks penetrating Asia Minor were far from being a clearly defined and homogeneous ethnic group. In their efforts to construe an unbroken continuity between the Oghuz tribes of Central Asia and the conquerors of Anatolia, modern Turkish historians highlight the persistence of tribal structures, ethnic characteristics, and behavioral patterns originating from Turkic nomadic and pastoralist traditions. Reports of Muslim authors referring to belligerent groups wandering about the Iranian provinces between Khurāsān and Azerbaijan along with their womenfolk, baggage trains, and livestock can be used in support of these views. ⁴ Likewise, the Seljuk dynastic tradition draws the image of a noble family descending from the Oghuz Kınık tribe and moving with its herds and retinues between summer and winter quarters in central Transoxania. ⁵ Additional evidence for the Turks’ overwhelmingly nomadic character is provided by the reports of Christian authors. ⁶ Their statements, however, represent only segments of the whole picture, and there are many descriptions pointing to more intricate realities. Between the 1040s and 1070s, the sources mention numerous names of Turkish chieftains conducting raids and attacks in various provinces of Byzantine Asia Minor and adjacent Muslim regions. These groups are described either as independently operating units, such as the followers of Arslān b. Saljūq in the 1030s and 1040s and the warriors of Hārūn b. Khān, Atsız b. Uwaq, Qaralū/Qurlū, Shuklī, and the sons of Qutlumush in the 1060s/1070s, ⁷ or as subunits subject to the supreme command of the Seljuk clan, such as the hosts of Samūkh in the late 1050s or of Afshīn in the 1070s. ⁸ Warlords like a brother of the sultan called Aspan Salarios (= isfahsālār) or the sālār of Khurāsān seem to have been granted titles related to military posts in Persian cities and provinces. ⁹ This shows that already in the first decades of Seljuk rule the assimilation of Turkish chiefs to the Iranian military class was well underway.

Claude Cahen proposed a classification based on the commanders’ proximity to the Seljuk sultan, ¹⁰ but this would presuppose the existence of a generally accepted central power, something that was hardly the case before the reign of Sultan Malikshāh (1072–1092). ¹¹ The Seljuk Empire throughout its existence was characterized by incessant rebellions and intra-dynastic conflicts, in which Turkmen chiefs and other military men turned from loyal followers into dangerous rebels and vice versa. ¹² Many scholars ascribe these phenomena to the rise of the emirs, a group of military commanders of disparate origin, who gained power and influence as a result of the consolidation of administrative structures and hierarchical concepts at the Seljuk court. Another factor was the transformation of the Seljuk military forces, which increasingly drew on slave soldiers (mamlūks) in lieu of Turkmen nomadic warriors. ¹³ Modern attempts to sharply distinguish between a traditional Turkmen aristocracy and a new military elite, however, obfuscates the fact that the boundaries between the various socio-ethnic groups included in the Seljuk army were always extremely blurred. The behavioral patterns of early Turkmen chieftains and later Seljuk emirs, who were appointed as governors and iqṭā‘ (“land grant”) holders in the provinces, evince numerous commonalities and continuities. A strong tendency towards a particularization and regionalization of power structures seems to lie at the very heart of the Seljuk expansionist movement and could only partly be curtailed by centralizing attempts on the part of the sultanate.

A characteristic feature of these powerful warlords, be they independent Turkmen chiefs or Seljuk emirs, was their endeavor to effectively interact with the indigenous population and the local elites. The invaders did not confine themselves to raiding and pillaging but aimed at a much broader range of activities by interfering with power struggles, forging coalitions and bonds of marriage with local rulers, and acquiring new sources of income through the extortion of tributes, the exploitation of landed estates, and the release of high-ranking captives. We find numerous examples for these behaviors among the Turks in the Anti-Taurus region, the Armenian highlands, as well as central and western Asia Minor. ¹⁴ It made no difference whether the indigenous aristocracy was Christian or Muslim. Alliances of this kind were usually short-lived and served specific goals so as to support competing groups against their adversaries. Not surprisingly, these coalitions also affected the composition of certain military groups. The available evidence is scarce, but it seems that successful campaigns frequently caused local elements to join powerful Turkish warrior groups permanently. In this way, Turkish nomads merged with Persian, Arab, and Kurdish groups in western Iran, Iraq, Syria, and Azerbaijan. ¹⁵ As regards the Byzantine territories, from early on we have information referring to coalitions with Franks, Armenians, and Greeks. ¹⁶ We may assume that as time went by this process altered the ethnic composition of these groups. The core of Turkish soldiers was gradually supplemented by newly arriving people of different origin. Another factor fostering this development was the presence of captives, who partly assimilated to the Turks. This phenomenon can also be observed reversely with respect to Turkish prisoners who were integrated in the Byzantine cultural environment and the imperial court. We know of Byzantine commanders, who as a result of their captivity were well acquainted with the customs and language of the Turks. ¹⁷ Conversely, there were some high-ranking officers of Turkish descent who reached supreme positions in the Byzantine army. A case in point was the grand primikerios Tatikios, who played an important role in numerous campaigns in the Balkans and Asia Minor throughout the reign of Emperor Alexios i (1081–1118). ¹⁸ Even more renowned was John Axouch, who had been taken prisoner after the capitulation of Nicaea in 1097 and held the rank of megas domestikos, i.e., commander of the eastern and western armies, during the entire the reign of John ii (1118–1143) and in the early years of Manuel i (1143–1180). ¹⁹

Modern nationalistic concepts obfuscate the intricacies of these relations by focusing on binary oppositions, such as thriving Greek-Orthodox communities vs. unruly nomads, who formed a deadly menace to townspeople and peasants, or powerful conquerors in search of a new homeland vs. decadent local groups. ²⁰ Explanations linking the motives of Turkish migration and expansion in Anatolia with the customs and needs of nomadic modes of living, such as climatic fluctuations, the suitability of landscapes, ample opportunities for winter and summer pastures, etc., ²¹ are certainly illuminating with respect to the initial stage in which groups of non-sedentary pastoralists arrived and adapted to the geographical conditions of the Armenian highland and the Anatolian plateau. Yet it is noteworthy that our primary sources rarely refer to these aspects. This lack of information should not be ascribed to the limited scope of outside observers or the distorting effect of literary conventions. The available narratives simply concentrate on those aspects, which determined the ways in which sedentary groups perceived and interacted with nomads and which preconditioned the transition to more permanent forms of settlement and rule. Frequently mentioned phenomena are raiding activities aiming at the accumulation of wealth, military services offered to local potentates, the infiltration of ruling elites, and the forging of coalitions, which opened the way to the acquisition of land and resources. In this way, Turkmen chieftains increased their manpower, developed links with the sedentary communities, and established rudimentary forms of political authority, which under favorable circumstances could result in the creation of lordships based on agreements with the indigenous population.

What were the political concepts, ideological incentives, and conquest strategies that created cohesion, a sense of solidarity, and a common identity among these Turkish warriors? Apparently, there was a strong influence of the Seljuk elite in Iran and Iraq, which under the leadership of Ṭughril Beg and his successors gradually developed a dynastic and imperial ideology drawing on Iranian models of kingship, the title of sultan inherited from the Ghaznavid tradition, and claims to a leading position in the Muslim world as protectors of Sunni Islam legitimated by the authority of the Abbasid caliphate. ²² These ideas undoubtedly underscored conquests and other political ambitions in the central lands of Islam. Turkish chiefs who maintained ties of allegiance with the Seljuk clan imported some of these concepts into the Byzantine-Armenian regions of Asia Minor while carving out their bases of power. The historical memory reflected in later chronicles evokes the notion of cities and territories assigned by the sultan and the caliph to outstanding Turkish emirs. ²³ There are also references to concepts of Muslim jihad, which first appear in reports on campaigns led by Seljuk sultans against the Byzantines and later on with respect to Turkish emirs fighting the crusaders. ²⁴ These features are employed as legitimizing strategies, which retrospectively link the nascent Turkish-Muslim lordships of Anatolia with the traditions and institutions of the Muslim central lands and Seljuk dynastic concepts. They hardly reflect the historical realities of the conquest period. What seems to have been a decisive impetus from the outset, however, was the successful leadership of powerful chiefs, who often gave their names as identifiers to the groups under their command. The frequent mentions of their names in the available narratives indicate the importance of these warlords as leading figures who created cohesion, attracted newcomers, and determined the course of action of their soldiers and followers. ²⁵ At a later stage, they were linked with ideological elements related to Seljuk dynastic traditions and Sunni Islam. In this way, the principalities created by them were embedded in the traditional concepts of Muslim sovereignty.

Reliable Information regarding the size of these groups is extremely scarce. A group of the Iraqi Oghuz is said to have amounted to 2,000 tents, which would mean a total number of 8–10,000 people including women and children. ²⁶ Other reports speak about 3,000 and 5,000 warriors in Armenia and western Iran respectively, ²⁷ while the figures regarding the soldiers under the command of Hārūn b. Khān in northern Syria range between 500 and 1,000. ²⁸ Highly exaggerated are the numbers mentioned with respect to the troops involved in large-scale campaigns of the Seljuk sultans in Armenia, Caucasia, and the Euphrates region. ²⁹ Since most reports refer to military operations, we hardly hear anything about the families accompanying the warriors. We may assume that non-combatant family members stayed with the warriors as soon as the latter acquired fortified camps and permanent strongholds. This happened first in the Diyār Bakr province, in the late 1050s in the Armenian highlands, and from the 1060s onwards in regions of Syria and Palestine. ³⁰ In the rural areas of western and central Anatolia, more permanent forms of Turkish presence are attested to from the mid-1070s onwards. ³¹ The gradual disintegration of Byzantine administrative and military structures, which was caused in various parts of Asia Minor by a series of power struggles among competing factions of the Byzantine aristocracy from 1057 onwards, allowed local lords, army units, ³² and foreign mercenary groups to gain a high degree of independence from the central government in Constantinople and brought about a breakdown of alliances with Muslim emirs in the borderlands. In this situation, Turkish hosts were able to maintain lines of communication with their compatriots in the frontier zones and the Muslim regions and new groups of Turkmen migrants along with their livestock and families invaded Byzantine territories almost unhindered. Again, it remains unclear to what degree these people were still primarily preoccupied with stockbreeding and transhumance or were already acquainted with farming and sedentary modes of living. We may assume that the divides between the two types increasingly blurred and that the Turks, while subjugating the local population and occupying urban centers, quickly adapted to their new environment. This is well attested, for instance, in the case of the Turkish bands of Atsız and Shuqlī, who around 1070 began to carve out their lordships in Syria and Palestine. ³³ In Asia Minor, the Turks needed more time to take hold of urban centers, but the overall instability enabled them to accumulate wealth from rural areas and to infiltrate the existing power structures. Undoubtedly, the turmoil in Asia Minor brought about a great deal of ransacking and pillaging, but there were no clear-cut frontiers or warring parties. Turkish warriors stood in the employ of both Byzantine rebels and the imperial government, while the Greek, Armenian, or Syrian population in towns and the countryside endured raids, sought the protection of Turkish chieftains, or forged alliances with them. Sections of the indigenous population certainly fell victim to these hostilities or fled to safer regions, but the majority kept on living in their hometowns and villages. ³⁴

How did the Turks hold sway over the rural areas in Asia Minor and how did they manage to seize towns? The scholarly literature frequently refers to large devastated zones that had been abandoned by their former Byzantine lords and thus could be easily seized and populated by Turkish newcomers. ³⁵ Certain narratives evoke images of a massive influx of settlers, who within a short period brought about radical changes to the ethnic and demographic composition. ³⁶ This is partly supported by descriptions of Byzantine authors, who speak about a total collapse of the imperial administration, the decay and withdrawal of military units, fatal mistakes committed by the central government, and incidents of inexcusable negligence. ³⁷ Armenian authors, too, refer to pitiless massacres and destructions of apocalyptic dimensions caused by the Turkish raids, ³⁸ while Muslim authors describe stunning amounts of booty and captives seized by the invaders. ³⁹ Taken at face value, all these accounts give the impression of disruptive events with fatal consequences for all preexisting structures, but we should bear in mind that they focus on specific conflict situations and are informed by the intentions and vantage points of their authors. ⁴⁰ Accordingly, they highlight military successes, barbarian menaces, or the idea of divine wrath, or they utter acerbic critique of opposing factions or undesired incumbents of the imperial throne.

A number of studies are devoted to the emergence of Turkish toponyms in Asia Minor, but except for a handful of instances in which geographic names can be related to personalities and sacred sites of the conquest period, the bulk of the known material concerning Turkmen tribes, topographic particularities, or nomadic customs is derived from later waqf (pious foundation) documents or Ottoman tax registers and reflects data of the 15th and 16th centuries. ⁴¹ All surviving monuments, artifacts, and archaeological evidence indicating Turkish-Muslim presence in Anatolia postdate the mid-12th century, when the firm establishment of Turkish domains in the urban centers of the central and eastern highlands was well underway. ⁴² Hence, the only way to reconstruct the dynamics of expansion is to carefully examine the available narratives by taking into account prevailing perceptions and intentions and by comparing data from Anatolia with those garnered from the Muslim central lands. Arabic sources refer to activities of Turkish warriors, the reactions of the local lords, and the interactions between the two sides in much more detail than Byzantine and Eastern Christian sources. Hence, certain gaps of information can, with the necessary caution, filled in with material from other regions.

From a methodological point of view, it is important to sharply distinguish between military activities, on the one hand, and aspects of permanent settlement and territorial rule, on the other. Raids and attacks certainly had a negative impact on the regions affected by them, ranging from limited devastation to a total destruction of social and economic structures. ⁴³ Yet they do not necessarily imply that the invaders from the outset aimed at the acquisition of territories. In fact, Turkish groups who were roaming about Byzantine territories initially evinced no ambitions whatsoever to permanently occupy towns or provinces. They were content with increasing their income from booty, tributes, taxes, and military coalitions. The Turks’ transition from rural to urban presence was not a sudden and violent act but a gradual process resulting from successful interaction and shared objectives between indigenous groups and immigrants.

As regards the various stages of Turkish expansion in Asia Minor, the scholarly bibliography places much emphasis on the early Seljuk campaigns and, above all, the battle of Manzikert in 1071 as decisive events sparking off the Turkification of Asia Minor. ⁴⁴ There is some truth in that, but a closer look at the sources reveals realities that are more intricate. The Seljuk campaigns in the Armenian highlands between the Araxes (Aras Nehri), the Arsanias (Murat Nehri), and the Lykos (Çoruh Nehri) Valleys, in Transcaucasia, and, later on, along the southern flank of the borderland as far as Aleppo did not result in substantial territorial gains or in a massive influx of Turkish warriors into Asia Minor. What these operations actually brought about was a breakdown of the imperial government’s coalitions with Muslim emirs in the frontier region, a dismantlement of the Byzantine defensive structures along the main invasion routes over the Anti-Taurus, the Amanus Mountains, and the Arsanias Valley between the Upper Euphrates and Lake Van. ⁴⁵ This, in turn, enabled independent bands of warriors to extend their raiding activities during the 1050s and 1060s from the Lykos Valley in the Pontus region to Cappadocia and facilitated the further intrusion of Turks into the Diyār Bakr province and northern Syria. ⁴⁶ The Byzantine themata and katepanata/doukata in the east, which had come into being as a result of the Byzantine conquests in the time between the mid-tenth and the early 11th century, turned into a permeable transit region marked by extremely volatile political conditions and the inability of superregional powers to impose centralizing forms of control. ⁴⁷

First invasions into central Anatolia as far as Konya in Lycaonia and the upper Maeander Valley occurred in the years 1068/69, but the advance towards the northwestern fringes of the Anatolian plateau in Phrygia and Galatia and thence along the Sangarios (Sakarya Nehri) Valley as far as Bithynia did not take place before the years 1073–1075. ⁴⁸ In these years, towns were devastated from time to time but not yet permanently occupied. The Turkish presence remained restricted to rural areas close to river valleys and sections of the Anatolian road system. Turks entered towns only on certain occasions with the consent of the local rulers for trade, negotiations, or the bestowal of gifts. ⁴⁹ Yet their intrusion into the regional structures had become more pressing. Contemporary accounts indicate a high degree of control exerted by Turkish warrior groups over certain areas close to urban centers or sensitive points of the local road system. ⁵⁰ This, in turn, suggests the existence of camps in suitable pasture regions as well as rudimentary military structures that afforded protection to their tribesmen and supported the surveillance and exploitation of larger territorial units. Nevertheless, contrarily to what later sources retrospectively claim in their attempt to establish links of legitimacy with the Great Seljuk sultanate, ⁵¹ there is still no evidence for the emergence of proper lordships founded by Turkish emirs in those years.

The period between 1080/81–1097/98 witnessed the establishment of Turkish chieftains in towns of Phrygia and Bithynia as well as in places situated in the coastal areas of western Anatolia, such as Kyzikos (near Ercek), Smyrna, and Ephesos (Selçuk). Unfortunately, the details of this process are only insufficiently known due to a gap in Byzantine historiography between about 1080, where Michael Attaleiates stops his account, and the Komnenian eulogists of Alexios i’s reign, Nikephoros Bryennios and his wife Anna Komnene, who wrote in the 1130/40s. ⁵² Bryennios differs in many respects from Attaleiates in his account of Anatolian affairs during the 1070s and adds numerous accurate observations, but his narrative stops before the outbreak of the Komnenian-Doukas coup in early 1081. ⁵³ Consequently, the decisive transition period between the fall of Nikephoros iii and the consolidation of Alexios i’s regime is only scarcely documented. In particular, it is hardly possible to elucidate the strong impact this event must have had on the power relations between Byzantine aristocrats and Turkish warrior groups in Anatolia. Projecting experiences of her own time back to the expansion period, Anna Komnene depicts the Turks as independent local lords, who were already firmly established in many urban centers of western Anatolia. The incessant revolts of the late 1070s and the fierce infighting among the leading aristocratic clans up to 1081 seem to have created a power vacuum, in which the Turks swiftly intruded. Yet the details of this process can hardly be reconstructed and many regions remain in the dark. We hear of Byzantine rebels handing over their towns to Turkish warriors in exchange for military support, as was the case with Nikephoros Melissenos in late 1080, ⁵⁴ or of proper conquests, as is mentioned with respect to Kyzikos, Smyrna, Konya, and Taxara (Aksaray). ⁵⁵ It is highly improbable though that the Turks at that time would have possessed the necessary equipment and fighting technique to conduct sieges. We may assume that in most cases takeovers resulted from agreements between Turkish chiefs and the local aristocrats. The detailed account of Anna Komnene yields plenty of useful information about the Bithynian city of Nicaea. Sulaymān b. Qutlumush and his warriors seem to have taken hold of the city in early 1081, whereas the settlement of women and children, the appointment of a local governor, and the establishment of a proper residence occurred after Kılıç Arslān i’s takeover in 1093. ⁵⁶ We may assume that the Turks followed similar patterns in other cities and areas. Initially they seem to have been content with establishing small garrisons, which gradually attracted families and other groups of settlers. In the course of time, this brought about an increase and diversification of Turkish elements permeating the social fabric of Anatolian urban settlements. This also implies the creation of a Turkish residence area or district within the walled town, including the erection of mosques or the adoption of pre-existing sacred spaces. The changes of urban spaces resulting from the establishment of Turkish elites and Muslim institutions can be studied more systematically from the second half of the 12th century onwards, i.e., the time from which the oldest monuments of Seljuk art in Anatolia survive. Nevertheless, the garrison-like character of the earliest Turkish settlements is still recognizable in the spatial setting of the earliest architectural monuments, which are mainly concentrated on citadel hills and frequently built in or near sacred areas of the pre-existing Christian substrate, thus visualizing military predominance in conjunction with a local memorial culture of sacredness. The best-known example certainly is the Alaeddin Mosque of Konya (1155), which along with the tomb of Kılıç Arslan ii was erected on the city’s acropolis near the now destroyed Byzantine church of Hagios Amphilochios. ⁵⁷ Likewise, the oldest surviving mosque of Ankara dating from 1178 was constructed just below the peak of the citadel hill, overlooking the whole town and the surrounding steppe land of the Anatolian plateau. ⁵⁸

In the 1080s and 1090s, in western, central, and parts of eastern Anatolia between the Halys (Kızılırmak) basin and the Armenian highlands, Turkish emirs began to develop rudimentary structures of independent lordships. Pieces of evidence are scarce and unevenly distributed, but taken together they reveal a number of recurring patterns that contributed to the transformation of Turkmen raiders into potentates and state builders. Generally speaking, there is a clearly recognizable divide between the Turks in western Anatolia, who were exposed to strong political and ideological influences of their Byzantine cultural environment and the imperial government, and the Turks farther east, who were attached to the Great Seljuk sultanate and other political powers in Azerbaijan and western Iran. Among the most crucial factors fostering the crystallization of state-like entities in western Anatolia we may mark out: (1) alliances with the Byzantine government and local aristocrats; (2) treaties between Turkish chiefs and the imperial government providing for a recognition of certain privileges, territorial rights, or spheres of influence; ⁵⁹ (3) diplomatic contacts between the two sides accompanied by tributes, gifts, and the bestowal of court titles strengthening the ideological ties with the Byzantine ruling elite; ⁶⁰ (4) political coalitions between Turkish lords, which were partly consolidated by intermarriages; ⁶¹ (5) successful forms of accommodation and collaboration with the indigenous population; ⁶² (6) firm control over land, agricultural produce, taxes, and other resources in towns and rural areas. ⁶³ At first, nascent lordships consisted of a number of strongholds along with a highly fluctuating radius of influence stretching over the surrounding areas. Yet there were still no clearly defined territorial or administrative units. The creation of extended realms being under the sway of Turkish emirs resulted from an intricate process, which lasted several decades and was contingent upon changing political constellations in Anatolia in the time before and after the First Crusade as well as upon the balance of power among competing Byzantine, Frankish, Armenian, and Turkish lords. At the same time, the consolidation and expansion of these new domains was closely related to a strengthening of the internal cohesion between the new Turkish ruling elite and the indigenous Christian population. To this end, the Turks offered incentives for local groups to stay and keep on pursuing their professions and economic activities. These included temporary tax exemptions, the protection of their subjects’ lives, property rights, and religious freedoms, effective forms of communication and interaction in administrative and judicial matters, and mechanisms providing a sense of lawfulness and legitimacy by combining newly imported Muslim and pre-existing Byzantine practices and concepts of political authority. ⁶⁴

The First Crusade (1096–1099) brought about a sudden and violent rupture in this development. The hosts of armed pilgrims arriving from France and Italy constituted not only a strong military force that enabled the Byzantines to regain large parts of western Asia Minor and wiped the Turks out of Cilicia and the regions of Antioch and Edessa. They also initiated a new and unexpected migration movement that channeled large numbers of Europeans through Asia Minor to the crusader states in Palestine, northern Syria, and Upper Mesopotamia. The crossing of Anatolia by crusading armies caused an unprecedented scale of devastation that resulted from acts of warfare and the scorched earth strategy applied by the Turks, on the one hand, as well as from the crusaders’ enormous demand for supplies and foodstuff, on the other. ⁶⁵ It can be safely assumed that in the years 1096–1101 parts of Bithynia and Paphlagonia as far as Amaseia (Amasya) and Merzifon, western Phrygia and Pisidia between Dorylaion (Eskişehir) and Lake Eğridir, sections of the Anatolian highlands in Galatia and the southern parts of Lycaonia and Cappadocia, as well as parts of Cilicia and the region around Antioch suffered serious damages in their agricultural zones and economic structures. At the same time, the Byzantine central government extended its sway over substantial parts of western Asia Minor and the southern coastland as far as Seleukeia (Silifke). Those Turks who refused to submit to Byzantine rule managed to gain new footholds in Lycaonia, Cappadocia, and the regions east of the Halys River. ⁶⁶ The towns of Polybotos (Bolvadin) and Philomelion (Akşehir) near the Sultandağları Mountains formed the easternmost points of advance for the Byzantine army and henceforth became nodes in a newly emerging frontier and contact zone where Byzantine and Turkish spheres of influence overlapped. ⁶⁷

The events of 1101 made plain that the Turks on the Anatolian plateau were able to recover from the setbacks of 1097–1098 and to put up effective resistance against invading crusading hosts. Since their first arrival in the 1070s, the Turks had become increasingly familiar with the geography and population of central Anatolia. After their retreat in 1098, they were forced to develop a defensive strategy in order to survive. This prompted them to forge alliances and to increase their control over towns, strongholds, and road networks. One may say that the challenges posed by the First Crusade accelerated the sedentarization and consolidation of the Turks in central Anatolia. Indicative of this profound change is that during the Second Crusade in 1147 both the French and the German contingents failed to fight their way through Turkish-held territories. ⁶⁸ It was not before the Third Crusade in 1190 that another crusading host, the German army of Frederick I Barbarossa, managed to cross the territories subject to the sultanate of Konya. ⁶⁹ They even seized and ransacked the Seljuk capital during their advance, but this was in a time of fierce infighting among Kılıç Arslan ii’s sons just a few years before the old sultan’s death. It is also noteworthy that after 1101 Kılıç Arslan i was on a friendly footing with the Byzantine emperor and even lent him military support against the Norman invasion of Bohemond in 1107. ⁷⁰ At the same time, Kılıç Arslan i turned his entire attention to his ambitions to extend his sway from the recently acquired city of Melitene and the Euphrates region to the Diyār Bakr province and northern Iraq as far as Mosul. ⁷¹ In May 1107, he paid with his life for his far-reaching plans and the surviving members of his family were thrown into a new crisis. ⁷² As a result, the western frontier zone in Anatolia enjoyed a significant respite and could further solidify.

In the years 1109–1116, new waves of Turkish invasions occurred in Bithynia, Mysia, and the western river valleys in the provinces of Lydia and Caria, but Emperor Alexios i successfully warded off these threats by establishing a chain of fortified strongholds and by developing an effective defensive strategy. ⁷³ In this way, he managed to restore centralized control over parts of western and southern Asia Minor. The imperial government blocked new Turkish advances and migration movements from the fringes of the Anatolian plateau towards the western coastland. A peace treaty with the Seljuk lord of Konya, Shāhinshāh, in 1116 put a halt to these large-scale invasions. ⁷⁴ Intra-dynastic rivalries among the sons of Kılıç Arslan i distracted the Turks of Konya from further pursuing their expansionist efforts. This brought about a further stabilization of the Turkish-Byzantine frontier zone between the Sangarios Valley and the Sultandağları Mountains. It was only in the early years of Emperor Manuel i that we hear of new Turkish raids in Bithynia and the Kaystros Valley. ⁷⁵ Again, nothing indicates the existence of a clear-cut boundary. The staging area of Byzantine troops conducting campaigns in the east was situated at a far distance from the conflict area near Lopadion (Ulubat) in the Rhyndakos Valley, and the emperor concentrated on the fortification of strongholds near the Sangarios River, such as Malagina and Pithekas. ⁷⁶ Places situated within the conflict zone obviously changed hands from time to time, but nobody was able to exert permanent control over them. When Manuel i in 1146 successfully proceeded from Akroinon (Afyonkarahisar) to Philomelion (Akşehir), he seized the town but was not able to keep it. Hence, he burned it down and transferred the remnants of the Greek population to Bithynia. ⁷⁷ In this campaign, the Byzantine troops also fought battles further east and advanced as far as Konya, but a siege was considered impossible. The imperial troops retreated via Lake Pousgouse (Beyşehir Gölü) to the eastern extremities of the Maeander Valley. ⁷⁸ First signs of an actual process of Turkification in this area are the Turkish toponyms Andrachman and Tzibrelitzemani as mentioned in the account of John Kinnamos with respect to places east of Philomelion. ⁷⁹ It must remain uncertain whether these names were already in use during the 1140s, but they certainly predate the 1180s.

The central section of the Maeander Valley and regions of Caria and Pisidia further south as far as the hinterland of Attaleia were also exposed to a more or less constant influx of Turkish groups. Fortified places, such as Laodikeia (near Denizli) and Sozopolis (Uluborlu), and smaller fortresses in Pisidia seem to have been seized by Turkish emirs already in the last years of Alexios i’s reign after 1116 and thus constituted targets of John ii’s campaigns in 1119. ⁸⁰ The Turkish presence in these outposts was apparently confined to movements of pastoralist nomads and small bands of warriors under the command of autonomous chiefs. There is no evidence suggesting any effective control exerted by the sultanate of Konya over them. When Emperor Manuel i in 1146 camped with his troops near the springs of the Maeander River, the Byzantine soldiers are said to have considered this region in the vicinity of Apameia (Dinar) as being far from hostile territory. Yet they were suddenly ambushed by a considerable number of Turkish troops. ⁸¹ This indicates that the Byzantines hardly controlled the region in question and failed to maintain any reliable defensive structures or a network of scouts and informants there. The same applies to the region of Lake Pousgouse (Beyşehir Gölü) farther east, which on the occasion of John ii’s campaign in 1142 is described as being of great importance for the emperor because of its proximity to Konya. ⁸² The close relations between the local Christians and the Turkish subjects of the sultanate, however, formed a serious obstacle for the emperor’s attempts to gain access to the lake. This incident illustrates another noteworthy aspect of the social changes resulting from expansionist movements and shifting borderlands. The indigenous local population, which came to be attached to the newly emerging ethnic and political entities, swiftly switched allegiance. The nearby Muslim-Turkish authorities replaced the remote imperial center of Constantinople as primary point of reference for the people living in the borderlands.

In the southern coastland a network of ports between Attaleia (Antalya) and Seleukeia (Silifke) as well as the island of Cyprus as an advanced outpost in short distance of the northern Syrian shores secured a strong Byzantine naval presence in the area between Lycia and the Gulf of Alexandretta. ⁸³ Archaeological evidence bears witness to Komnenian building activities in the towns of the Lycian coastland, whereas in adjacent Pamphylia the port of Side seems to have been deserted because of Turkish pressure and to be replaced by Kalon Oros (Alanya) as new regional center. ⁸⁴ There are traces of nomadic activity in the mountainous regions north of the coastland, as ceramic finds in Sagalassos (Ağlasun) seem to indicate. ⁸⁵

It is difficult to assess the extent of Byzantine control over the Pontus region, but the overall impression on the basis of the available information is that after 1086 the coastland east of Sinope remained in the hands of independent Byzantine and Turkish local lords. ⁸⁶ The region of Trebizond at times was threatened by Seljuk governors based in the Armenian provinces, while in the 1130s it was mainly the Dānishmand emirate in Cappadocia, which exerted control over Neokaisareia (Niksar) in the Lykos Valley and beyond the Halys River as far as Gangra (Çankırı) and Kastamonu. ⁸⁷ Attempts of the Byzantine army to take possession of these places either failed completely or did not bring lasting results. The tomb of Karatekin, a legendary Turkish hero, who in the epical accounts of the Danişmend-name appears as the first Muslim conqueror of Çankırı and other towns of Paphlagonia, is a figurehead for the establishment of a Muslim tradition in the region. ⁸⁸ This monument situated in the castle of Çankırı dates back to the second half of the 12th century and thus belongs to the earliest surviving architectural remains of Turkish provenance in central Asia Minor. The fact that the aforementioned mosques of Konya (1155) and Ankara (1178) date from about the same period clearly indicates that the central Anatolian towns in Paphlagonia, Galatia, and Lycaonia east of the Byzantine-Turkish borderland began to acquire a new character as permanent dwellings of Muslim-Turkish elites from the mid-12th century onwards. At that time both the sultanate of Konya and the Dānishmand emirate had already developed into considerable supra-regional powers covering vast areas as far as the Upper Euphrates River. The transformation of urban centers and the consolidation of political powers were closely intertwined.

The development in western and central Asia Minor during the second half of the 12th century is marked by a further expansion and solidification of political and administrative structures established by the Turkish domains, on the one hand, and by a gradual weakening of the Byzantine defense system especially after the disaster of Myriokephalon in 1176, on the other. ⁸⁹ In addition, the troublesome years after 1180 brought a breakdown of pre-existing bonds of cohesion among Byzantine aristocratic clans and an overall dismemberment of the provincial administration. ⁹⁰ These phenomena did not cause any fundamental shifts in the frontier zone but certainly increased the instability and insecurity of living conditions for both sedentary people and Turkish nomads and thus favored new displacements. The Turks resumed their raids and incursions into Bithynia, Mysia, western Phrygia, as well as the Kaystros and Maeander Valleys. In the 1180s and 1190s, the situation was exacerbated by a series of local uprisings aiming at the establishment of autonomous lordships of Byzantine aristocratic clans, as happened in the Bithynian towns of Nicaea and Prusa (Bursa) under the leadership of the Angeloi family (1184) and in Philadelpheia (Alaşehir) under Theodore Mangaphas (1188). ⁹¹ In these cases, Turkish warrior groups became involved in intra-Byzantine conflicts as auxiliary forces supporting local rebels. Mangaphas, in particular, developed an extensive raiding activity in the Maeander Valley. ⁹² Other rebels, such as a number of Pseudo-Alexioi, i.e., persons pretending to be the murdered son of Emperor Manuel, Alexios ii, and Michael, the doux of Mylassa, strengthened their fighting force by mustering Turkish troops with the consent of the lords of Ankara and Konya. ⁹³ In so doing, they formed a serious threat for the peasantry in the borderland of Ankara and the Maeander region. These new forms of cross-border alliances exhibit an unprecedented level of collaboration between Byzantine and Turkish local lords for the purpose of establishing some extent of regional authority and gaining wealth at the expense of the local population. Unruly Turkish nomads roaming about the countryside were involved in these activities as the rebels’ allies but they cannot be regarded as the driving force underlying the increase of violence in the borderlands after 1180. ⁹⁴ Apparently, the raids were made possible by the progressive inability of the central government and its representatives to exert effective control over the peripheries. Seditious movements related to disturbances in the center-periphery relations fostered the mingling of warrior groups in the borderlands and had disastrous results for the population living within the radius of action of these rebels, irrespective of their ethnic or religious identity.

The Byzantine campaigns of the 1130s and 1140s failed to bring lasting results in terms of territorial gains or to change the balance of power in Anatolia. In the decades after the Second Crusade, the imperial government concentrated on forging more peaceful relations with the Turkish neighbors and on fortifying their strongholds in western Asia Minor. In the region of Chliara (Kırkağaç), Pergamon (Bergama), and Adramyttion (Edremit) in western Mysia, Emperor Manuel i took effective protective measures by transferring the local population from widely dispersed unprotected villages to newly fortified places, the so-called Neokastra. ⁹⁵ In addition, Manuel restored the ruined fortresses of Dorylaion (Eskişehir) and Siblia/Subleon (near modern Evciler). ⁹⁶ Both strongholds were situated in highly sensitive areas of the Byzantine defensive system, the former on the main road leading from the fringes of the Turkish-held regions of Phrygia to Bithynia and the latter in the Upper Maeander Valley at a short distance from Apameia. The Byzantine defeat of 1176 did not cause a total collapse of the Byzantine military system in Asia Minor, but it certainly signaled a turn for the worse in that the emperor was compelled to abandon these strongholds and parts of the Maeander region as far as Tralleis (Aydın) were once again devastated by invasions from the sultanate of Konya. ⁹⁷ The territorial gains, which the Turks achieved in the years after 1176, were limited but included places of high strategic significance. Due to a garbled passage in the manuscripts of Niketas Choniates’ chronicle, it remains unclear whether Dorylaion actually fell into Turkish hands. ⁹⁸ In 1182 Kılıç Arslan ii was able to drastically increase his pressure on Byzantine territories in Phrygia and Pisidia by seizing Sozopolis (Uluborlu) and Kotyaion (Kütahya). ⁹⁹ The Turks even advanced to the southern coastland of Pamphylia and, in the course of a long siege, caused heavy damages to Attaleia and its surroundings. In 1190, the crusader army of Frederick Barbarossa was still able to seize and burn down Philomelion (Akşehir), but a few years later in about 1196 the town seems to have been under firm Turkish control. ¹⁰⁰ Overall, in the time span between 1176 and 1196 the Turks of Konya gained a number of advanced strongholds granting them access to the road system in the central and the southern section of the western frontier zone. Moreover, for the first time they were able to attack the empire’s main port in the coastland of Lycia and Pamphylia, which in 1216 would eventually surrender to the Seljuk sultan. ¹⁰¹

The gradual disintegration of the Byzantine central government in the years after Manuel i’s death in 1180 in conjunction with the concomitant disintegration of administrative structures and regional insurrections in the provinces of Asia Minor created a vacuum of power that could be easily filled by potentates of the adjacent Turkish-held regions. The sultan of Konya and other local emirs quickly came to form the only powers to guarantee some extent of stability and security in the region. Unsurprisingly, this situation prompted indigenous groups to relocate to areas, where they enjoyed more protection and tax privileges. Niketas Choniates mentions a large group of 5,000 captives, who at the sultan’s behest were transferred from towns of the Maeander Valley to Philomelion, where they were granted land, fields, grain, and a five-year tax exemption. ¹⁰² This population transfer apparently aimed at an improvement of the agricultural productivity of the region, something that was highly attractive for other Byzantine subjects as well, who voluntarily set forth to join their compatriots in Philomelion.

Another form of mobility, which in the late 12th century becomes increasingly palpable, results from overland trade between Constantinople and Konya. Niketas Choniates relates an episode referring to precious horses that had been sent by the Ayyubid sultan of Egypt along with an embassy to Constantinople but were intercepted by the authorities in Konya. ¹⁰³ In response to Sultan Kaykhusrau’s offense, Emperor Alexios iii reportedly imprisoned all merchants from Konya and confiscated their belongings. This account affords us only an isolated glimpse, yet it points to the existence of a well-established trading network between Byzantine and Turkish merchants, which must have been based on safe travel conditions and facilities supporting the transport of people and goods along the routes connecting Byzantine and Turkish-held regions between Constantinople and the western Anatolian plateau. We may assume that the people involved in these activities deployed various forms of collaboration and abided by mutual undertakings. The oldest surviving example of such agreements is the Cypriot-Seljuk treaty of 1216, which regulated the seaborne trade between the island and the southern coastland. ¹⁰⁴ Earlier evidence for Byzantine-Seljuk trade is hardly available, but from the 11th century onwards, the sources mention large amounts of gifts granted to high-ranking Turkish dignitaries and fugitive potentates who sought sanctuary in the imperial city. ¹⁰⁵ This shows that Turkish elites from early on became familiar with and held in esteem all sorts of luxury goods of Byzantine origin. When Konya in the early 12th century gradually became the Seljuk court’s main residence, the city must have attracted new economic activities and population groups and thus turned into a hub for trade between Byzantine territories, the Anatolian plateau, and the Muslim lands beyond the Euphrates. ¹⁰⁶ Moreover, the nearby Byzantine-Turkish contact zone certainly encouraged commercial exchange, as is attested by the reports about the tight links between the Greeks at Lake Pousgouse and Konya in 1142. ¹⁰⁷

The influx of Turkish nomads and other ethnic groups from the Muslim central lands as well as the gradual emergence of Muslim-Turkish domains on Byzantine soil between the 1040s and the end of the 12th century set into motion a profound process of social and political transformation. With the incoming migrants, nomadism and pastoralism became predominant modes of living in parts of the Armenian highlands and the Anatolian plateau. Yet it would be inaccurate to assume that it was these nomads who were the bearers of change in Asia Minor by supplanting Byzantine civilization with a Muslim-Turkish culture, as is frequently implied by the secondary bibliography. A comparison of the available primary reports demonstrates that practices of expansion and intrusion observable in Syria, Upper Mesopotamia, and western Iran were also transplanted to Asia Minor. This is to say that sections of the Turkish warrior elite, who took hold of larger territorial units and urban centers, swiftly switched from nomadism to sedentary forms of rule and adapted to the pre-existing social environment. Newcomers mingled with indigenous groups at various levels and, in many instances, a mutual process of integration can be observed. Greek, Armenian, and Syrian Christians forged contacts with Muslim-Turkish officials and merchants and, conversely, many Turks adopted Byzantine cultural habits and ideological expressions or became members of the Byzantine elites. Changing center-periphery relations and the disintegration of Byzantine central rule in the years 1057–1081 fostered the regionalization of political structures so that powerful local factors ousted the influence of the imperial center. A re-stabilization of central control in western Asia Minor and the coastlands under the Komnenian emperors between the First Crusade and 1180 was followed by a new dismemberment of imperial administration in the last quarter of the 12th century. This led to increasing activities of local warrior groups, which included Turkish nomads, troops subject to the sultanate of Konya, and Byzantine rebels. Again, it is hardly possible to recognize clear-cut boundaries between sedentary and nomadic or Christian and Muslim groups. It seems more appropriate to talk about local coalitions of marginalized groups, which frequently assumed the character of Byzantine-Turkish alliances. Byzantine administrative and political structures were not destroyed but lost their links with the elite of Constantinople so as to be integrated into the regional structures of central Anatolia. The events of 1204 and the rise of the sultanate of Rum in the first half of 13th century created a new equilibrium in Asia Minor, in which the new center of Konya replaced Constantinople as focal point of administrative, social, and political structures in Anatolia.