PREFACE
The Turks consist of a group of peoples who, in the course of roughly the last two thousand years, have swarmed across vast territories from their homeland in central eastern Asia, to reach the Indian Ocean, the Mediterranean and eastern and central Europe. There they created or dominated many states and, at the end of what we know as the Middle Ages, they established a multinational State, the Ottoman Empire, which in the event proved to be one of the most enduring known to history. Today they still occupy large expanses of territory in eastern Europe and Asia, and, with the Turkish Republic established in Asia Minor and to some degree several of the Republics of the Soviet Union, they thus have their place in the general evolution of the modern world towards the organization of what are now national States. In this series of facts there is just as much to claim the attention of historians as in the history of other peoples.
However, this expectation has not altogether been fulfilled. It is true that, in the nineteenth century and sometimes even earlier, the role played by the Ottoman Empire, in European history particularly, inspired works which were of importance in their own time and which, in part, can be, and indeed are, still used today (Hammer-Purgstall, Mouradja d’Ohsson, Zinkeisen, Jorga 1909). Since their panegyrist Léon Cahun devoted his Introduction à Vhistoire de l’Asie (1896) to them, the other Turks too have formed the subject of writings of some value. And here, of course, I am referring only to general surveys, without mentioning the very numerous monographs and detailed studies which little by little have built up an independent field of study, Turcology. The fact remains that these studies and general surveys on the whole fall far short of the corresponding works relating to the history of Europe and even to certain branches of orientalism. It is worth noting briefly the reasons for this, in order to help as far as possible to remedy this backward state of affairs.
A primaiy reason is common to all non-European peoples. Through varying circumstances it so happens that no people outside Europe established or preserved before the modern period a general body of documentation such as is available for European history, either because they had not attained the requisite level of civilization, or because, though civilized, an interest in their own past was foreign to them, or, finally, because the preservation of their documents had suffered either as a result of their social organization or from historical catastrophes. Whatever progress may be achieved in the future in the study of these peoples, clearly it will always be impossible to fill in all the gaps in our knowledge.
Moreover, recent history has given Europe a considerable lead over some peoples who, at times in the past, had been further advanced on the path of civilization. As a result, it was generally by Europeans alone that the study of these peoples’ past was initiated, following the methods of modern scholarship. This means, in the first place, that the work could only be performed by scholars who had previously undergone a sometimes long and arduous linguistic apprenticeship, with the consequence that the linguistic point of view gradually took precedence over the historical; furthermore, that in research the emphasis was often placed on what, for a variety of cultural or political reasons, was of interest to Europeans, rather than on what constituted the main issues in the life of these peoples, as they would have appeared if considered on their own merits and from the inside. I do not wish to exaggerate this last defect, since one cannot in fairness deny that European culture in particular has made an effort towards understanding others which still remains unparalleled elsewhere. But it is fortunate that an increasingly large number of peoples are now acquiring the ability to study their own past scientifically, not only to supplement the inadequate number of linguistically qualified Europeans, but also to assess that past in the new light of different points of view, which can only prove to be mutually rewarding. This is as true of the Turks as of others.
In respect of the Turks, the documentary position before the modern period is difficult. In Mongolia and Chinese Turkestan writings that derive from the Turks have been found; to these we shall return later. But in this same region and a fortiori for the whole area of the Eurasiatic steppes over which they spread, apparently without concerning themselves with writing, the essential part of the documentation is furnished by the more advanced peoples with whom they came in contact. This means that there are numerous gaps, for those zones and periods where such contacts were incomplete; and even when some information exists, the points of view are external and disconnected, and there are differences of language which at times make it difficult to reconcile exactly the information provided by Chinese, Arab, Byzantine and, later, Russian sources. It is not even always possible to determine if a specific people is or is not Turkish. For those of the Turks who later swarmed into western Asia, the situation is at the outset no better. At first, it is Arabs, Persians, Greeks, Armenians and others who tell us more about them than they do themselves, and, even when they in their turn begin to write, they often do so in languages other than their own, in Arabic and above all in Persian, thus presenting Turkish things in a somewhat alien garb. For the Ottoman Empire, particularly from the sixteenth century, the situation is potentially much better. Apart from its literature, the Ottoman Empire possessed archives fully comparable with those of the great European States; but too short a time has elapsed since their interest was first appreciated, too few specialists have concerned themselves with them, there are too few inventories and catalogues, there has been too little classification, to have allowed them to provide anything like what could legitimately be expected from them.
To these difficulties others can be added. At almost every stage of their history and in all the lands where their ventures took them, the Turks have been intermixed with so many established peoples, whose history has not ended with their arrival, that the distinction of what is and what is not Turkish is often difficult to make, and indeed devoid of any real significance. While it is true that the history of the Seljukid Empire cannot be omitted from that of the Turkish expansion, nonetheless it forms a phase of Iranian history. The conditions under which modern orientalist historiography has been established have often caused the Turkish factor to be underestimated; and the Turks themselves, as a result of being integrated in more comprehensive communities, did not in fact become fully aware of this factor until the advent of the modern and truly Turkish Republic. But obviously it is not enough to go to the opposite extreme in regard to this inadequacy and to insist in and out of season on the importance of the Turkish factor in order to give a truer picture of it.
It has to be admitted that the scientific difficulty in this respect is increased by wholly extra-scientific considerations. Because the Ottoman Empire was at war with Christian Europe for longer periods than any other ‘oriental’ state, because, during the period of its decline, as contemporary national states were being established, it appeared in the eyes of Europe as the oppressor of brotherly Christian peoples, everything Turkish was often regarded as a priori retrograde, tyrannical and contemptible. The Turks naturally have reacted with vigour, but here too the opposite assertion, also a priori, could not scientifically establish the truth of the matter, even though it inspires the search for it.
The historian must try objectively to establish facts, without considering whom they will please or displease. But the accumulation of accepted ideas makes the task difficult, and one cannot claim that even the most unprejudiced persons can today be sure of ridding themselves of them entirely.
The present book does not aim at giving the complete history of the Turks. Basically it will be concerned with the mediaeval Turkey of Asia Minor, the foundation of modern Turkey (that is to say of the only essentially Turkish state, if one leaves aside the Republics incorporated in the U.S.S.R. which, in themselves, are not of comparable importance). It is true that it will not be possible to embark immediately upon their history, and in an introductory section we shall have to trace, though more briefly, the history of the Turks who lived earlier or were neighbours of that country during the Middle Ages. But, when this is done, we shall in the main devote ourselves to Turkey in Asia Minor. It must be said that this will be virtually for the first time. Not that the historians of the Ottoman Empire have completely ignored this earlier Turkey - but they have treated it as a prelude to that Empire, in relation to which, however, as we shall see, it is difficult to portray it. A considerable number of monographs (some of them of importance) have been devoted to it. But hitherto there has been only one analytical and detailed account, which was premature and unsuccessful for reasons which we shall have to study. It is hoped that the present work will make it possible both to convey the character and intrinsic interest of early Turkey, as it was before the Ottomans, and also to guide future research by a clearer warning against what the author regards as certain misleading points of view.
It must, however, be stated frankly at the start that this work is a provisional synthesis, still incomplete in detail, of the research upon which I have for a long time been engaged. I hope soon to bring out a first volume, in French, giving the more detailed and exhaustive results of this research {Histoire de la Premiere Turquie, I, Des origines à 1243), but it will certainly be some time before the two succeeding volumes are ready. In these circumstances, perhaps no justification need be offered for the decision to provide the educated public - and what is more, in another language -with a work that is simpler to read, and at the same time to present to specialists some provisional facts and ideas: one never knows what may happen, particularly when one is no longer so young . . .
Nevertheless, the fact remains that any constructive criticisms which may be expressed to me will be particularly welcome.
Much of the text will consist of what is now somewhat disparagingly called ‘narrative’ history. This is certainly not to say that the author’s point of view is that of the narrator of anecdotes and episodes. But there are two reasons for this course, one connected with the state of the documentation, the other more general and a matter of principle. As far as the documentation is concerned we can only accept the situation, for it so happens that the sources at our disposal are almost exclusively of a narrative character. It is thus solely by means of narrative accounts that we can hope to penetrate more deeply into the structural facts which are now the historian’s fundamental concern, and I feel that it would be unwise, and even slightly disingenuous, not to provide the reader with the essential part of the narrative material that we are using, but instead merely to present him with ready-made deductions. But there is a further reason which I would like to put forward, even though it lies outside the scope of one particular history. In a reaction against the period when our predecessors confined themselves to a narrative history ‘of wars and kings’, many modern historians are no longer willing to consider anything except the structural facts, and scorn simple events ; and the young writers who are approaching history under their tutelage, in both the East and the West, anxious to be modern, are carrying this contempt to dangerous extremes. Masters are masters, and moreover they have had the support of the solid framework provided by their predecessors who established ‘narrative’ history; pupils are not necessarily all masters, and they no longer always have the framework behind them. History is total, that is to say it combines together inextricably both ‘events’ and ‘structures’, and we have no theoretical right to separate them. History is evolutionary, and it is events which are the landmarks in this evolution. In a mere outline of structures this would run the risk of disappearing, or, with the latter omitted, the account given might appear to lack foundation. Events and structures react upon each other, in both directions, and I do not think it necessary to explain this characteristic to those who witnessed the two World Wars, or even one of them. On a lower scale, a document, even though not by its nature narrative, always has a place, a date, an occasion, in its way it is a small event among important ones: to study it without reference to that place, date or occasion is to run the risk of not understanding it. (Some illustrations of this point will be found later.) It is thus quite deliberately that, without attributing a fundamental importance to events, we have given them a large place.
In these circumstances, the ideal would obviously be to have constantly interconnected the two categories of facts in the text, as they are in reality. In practice this is impossible, and where important connections have to be strongly emphasized, the endless reference from one series to the other would make the book both immense and unreadable. We must therefore be satisfied with divisions which, though they may seem regrettably traditional, are nonetheless inevitable. The only point in question is with which series one should start. A Marxist would answer that economics conditions everything else, the infrastructure determining the superstructure, but only to add immediately ‘in the last analysis’, and that, at all stages, there are reciprocal effects, a fact which no historian would dispute. The reason for the choice made is thus not so much theoretical as practical. In practice, in all societies whose documentation is essentially narrative, it is more convenient to start with ‘political’ events, since they are the raw material of our work and since they furnish the places, dates, the identities of persons and various circumstances useful for the understanding of other documents. This is what we shall do here, even to the point of ultimately proving somewhat tedious, and we hope that there will be no misunderstanding of the reasoning which dictated this choice. It is quite obvious that when the detailed narrative fabric of mediaeval Turkish history has been completed, other writers will be able to refer to it without being obliged to reproduce it, except on special points. But unfortunately we are not yet at that stage, and it would be doing no service to evade the obligations of our generation.
The composite character of this work and the differing nature of its component parts explain the method that has been adopted with regard to references to sources and bibliography. For the introductory section, only a general bibliography is given, with no reference to sources. For the following parts, that is to say the main body of the work and everything relating to the Turks of Asia Minor, I shall give firstly an almost complete list of sources, arranged as methodically as possible according to periods and categories, and then a bibliography for each chapter or group of chapters, by means of which the specialists will have no difficulty in finding the information they need. In these circumstances it seems unnecessary to make use of foot-notes, which would be of value in exceptional cases only (where that is so, the necessary comments have been incorporated in the bibliography for the chapter in question). The sources of the narrative chapters are throughout and almost exclusively the chronicles indicated in the sources. For the other chapters, they are obviously more diverse, and for the most part are mentioned in the text. In any case, it is always easy to trace them from the bibliography and the comments included in it.
The transcription of ancient Turkish proper names cannot always be certain on account of dialectal variations, the inadaptability and diversity of the ancient authors’ notations in the Arabic and other alphabets, and also the uncertain state of our historico-linguistic knowledge. In the text, the accepted forms have been given when they exist, and in other cases those which seemed most probable, expressed in the form most normally pronounceable for the English-speaking reader. This may lead to discrepancies with the transcriptions the reader will find in other publications, particularly the Encyclopaedia of Islam (and to a lesser extent the Turkish Islam Ansiklopedisi), in which the need to maintain a certain unity between the systems of transliteration used for the different Muslim languages has led the editors to admit forms which are at variance with the phonetic transcription. In some cases it may perhaps be thought that I have failed to keep in touch with some hypotheses expressed with regard to the pronunciation of certain names, for example Kutlumush, which I continue traditionally to give in this form, and not Kutalmïsh or the other forms suggested by some scholars. Without claiming any competence in the purely linguistic field on which suppositions of this sort are based, I think however that only too often purely etymological reconstitutions are doubtful and that they cannot in all cases be upheld when transcriptions are at variance with them. The Arabo-Persian alphabet denotes the vowels only poorly, but the Greek and Armenian denote them better, even if on the other hand certain consonantal sounds in these are imperfect for the notation of Turkish. One cannot discard their transcriptions solely because they fail to agree with our ideas about the etymology of the ancient Turkish dialects.
The eight Turkish vowels have been denoted thus: a, e, i, ï, o, o, u, ü. Of the consonants, the only ones requiring explanation are kh (pronounced like the German hard ch or the ch in Scottish loch) and gh (pronounced almost like a simple lengthening of the preceding vowel).
Pre-Ottoman Turkey
Introduction
The Turks and Their Islamization before the Seljukids
The Turks belong perhaps to a branch, known as Ural-Altaic (whose exact limits are in any event highly uncertain), of what are called the ‘yellow’ peoples (who probably also include all or part of the American ‘Redskins’). Relatively more closely related to them than, for example, the Chinese are the Finno-Ugrians (Finns and Hungarians), the Samoyeds, the Tungus and, in particular, the Mongols. It is practically certain however that the earliest Turks known to history - although not called by that name -were the Huns. Known to us from the Chinese Annals as early as the third century b.c., the Huns were finally, after successive migrations, to establish Attila’s Empire (fifth century a.d.) in the heart of Europe. Under the specific name of Turks (the meaning of which is uncertain), the Turks made their appearance, both in the Chinese sources in the East and in the Byzantine sources in the West, in the sixth century a.d. in the territory that is now Mongolia, but very soon also over a wide area, expanding towards the south and west. It is difficult to reconstruct their history in detail in the two or three centuries that follow, since the foreign sources that happen to mention them occasionally and even the few Turkish inscriptions, practically confined to Mongolia, which also appear from the eighth century, provide a succession of different names, in what consequently seems at first sight to be a list of distinct peoples, but which in reality is probably no more than an enumeration of the names of political groups and tribal federations within changing limits but including a part of the same peoples in different guises. Specialists are so hesitant over matters of detail that we can only refer to one or two landmarks. In the sixth century, a Turkish ‘Empire’ existed in the northern part of what in modern times has been called Russian ‘Turkestan’, from the Syr Darya/Jaxartes to the borders of Siberia, from the Altai to the Volga. The memory of this was long to survive among …
