446

 THE FORMATION AND EVOLUTION OF THE IMAGE OF THE TURKS IN
BYZANTINE HISTORIOGRAPHY

The Formation and Evolution of the Image of the Turks
in Byzantine Historiography
This study explores the representation of the Turks in middle and late Byzantine
historiography. It features a corpus covering the 11th to the 15th centuries,
comprising the works of Attaleiates, Skylitzes, Anna Komnene, Kinnamos, Niketas
Khoniates, Akropolites, Pakhymeres, Gregoras, Kantakouzenos, Palamas, Doukas,
Khalkokondyles, Kritoboulos, and Sphrantzes. The Turkic peoples were known to
the Byzantines from Late Antiquity. However, in the 11th century, both at their
eastern and northern frontiers, the Byzantines encountered a rapid expansion of the
Turkic populations. This new encounter made the Turkic peoples an indispensable
object within Byzantine historiography.
The representation of the Turkic peoples in Byzantine historiography bears
both the remnants of the tradition of mimesis, in which the reuse of ancient
ethnonyms, models, and, topoi was an indispensable condition for a respectable
literary work, as well as vivid reflections based on the recent encounter of the
Byzantines with the Turks. The image of the “barbarian” was already present in
Antiquity, in various forms, employed for several foreign populations by the authors
of the Greco-Roman world. In the representation of the “barbarians”, different
ethnonyms such as “Persian”, “Skythian” or “Hun” referred to the varying cultural
memory of the Byzantine authors. These authors left a rich corpus on the different
Turkic states, their rulers, customs, warfare, and the different aspects of the lifestyle
of the Turkic peoples. This dissertation investigates the formation and evolution of
the representation of the Turkic peoples using this material.
v
ÖZET
Bizans Tarihyazımında Türk İmgesinin
Oluşum ve Evrimi
Bu çalışma orta ve geç dönem Bizans tarihyazımında Türklerin temsilini
incelemektedir. 11. ve 15. yüzyıllar arasında eser vermiş Attaleiates, Skylitzes, Anna
Komnene, Kinnamos, Niketas Khoniates, Akropolites, Pakhymeres, Gregoras,
Kantakouzenos, Palamas, Doukas, Khalkokondyles, Kritoboulos, and Sphrantzes bu
çerçevede Türki halkların tarihsel temsili açısından çalışılmıştır. Bizanslılar Türki
halklara Geç Antikite’den beri aşinaydılar. Ancak, 11. yüzyılda, hem kuzey hem
doğudan Türki halkların hızlı bir genişleme hareketiyle karşılaştılar. Bu yeni
karşılaşma, Bizans tarihyazımında Türki halkları olmazsa olmaz bir konu haline
getirdi.
Bizans tarihyazımında Türki halkların temsili, hem eski etnonim, model ve
toposların saygıdeğer bir edebi eser vermek için olmazsa olmaz olduğu mimesis
geleneğini, hem de Bizanslıların Türki halklarla güncel karşılaşmalarına dair ilginç
yansımaları içerir. “Barbar” imajı antik çağdan beri mevcuttur, çeşitli şekillerde
Yunan-Roma dünyasının yazarları tarafından yabancı topluluklar için kullanılmıştır.
“Barbarlar”ın temsilinde, “İskit”, “Hun” ve “Pers” gibi değişik etnonimlerin
kullanımı Bizanslı yazarların çeşitli kültürel hafıza katmanlarına işaret eder. Bizanslı
yazarlar Türki devletler, bu devletlerin hükümdarları, adetleri, savaş yöntemleri ve
hayat tarzlarının değişik yönleri üzerine zengin bir külliyat bırakmıştır. Bu tez
sözkonusu materyali kullanarak Türki halklarının Bizans kaynaklarındaki temsilinin
oluşum ve evrimini incelemektedir.
vi
CURRICULUM VITAE

Tirali, A. (2009) Le roi, la princesse et le compagnon d’armes - Les représentations
des personnage chrétiens dans la littérature épique turque du Moyen Age, EHESS,
Centre d'Etudes Byzantines Néo-Helléniques et Sud-Est Européennes, Paris.
Book Chapters
Tirali, A. (2021) “Fransa’da Sosyalist Parti: Köklerden Bugüne”, in O. A. Yılmaz Ed.
Sosyal Demokrat Geleneklerin Temel Kaynakları: Türkiye, Almanya ve Fransa
Örnekleri, 105-135, İstanbul: Sodev Yayınları.
vii
Conference Proceedings
Tirali A. (2018) “Les Scythes qui s’appellent populairement les Petchénègues, Les
Perses qui s’appellent maintenant les Turcs: Attaleiatès sur les peuples Turciques”,
Circulation des hommes, des biens et des idées à Byzance: Journée des doctorants du
monde byzantin, Université d'Aix-Marseille, 12. 04. 2018.
Tirali A. (2011) “Mise-en-scènes de l’altérité : les non-musulmans dans le cinéma
arabe contemporain”, Atelier international des jeunes chercheurs- La mémoire dans
la pratique sociologique, CADIS, EHESS, 28. 10. 2011.
viii
ACKNOWLEDGMENTS
First of all, I want to express my deepest gratitude to my co-directors: Nevra Necipoğlu
and Paolo Odorico. I would like to extend my sincere thanks to Boško Bojović, Koray
Durak, Ahmet Ersoy, Selçuk Esenbel, Cemal Kafadar, Charis Messis, Dan Mureşan,
Arzu Öztürkmen, Gabriel Poloniecki, Derin Terzioğlu, and Meltem Toksöz.
My old friends Deniz Günce Demirhisar, Alexandra Vukovich, and Esin İleri
were always very helpful and nice; they also merit special thanks.
I am extremely grateful to my parents Esra and Bülent Tirali, to whom I
dedicate this work. And, most importantly, I could not complete this dissertation
without the loving support of Sedef, my darling. Words cannot explain my gratitude
to her.
ix
To my parents, Esra and Bülent Tirali, who noticed my endless interest in history
when I was very young, raised me as an unbiased and free-minded man,
and thus made me turn to Byzantine studies...
x
TABLE OF CONTENTS
CHAPTER 1: INTRODUCTION ................................................................................ 1
1.1 The presentation of the subject .......................................................................... 1
1.2 A note on the terminology: Which Turks, which Turkic peoples? ................... 5
1.3 The presentation of the sources ......................................................................... 6
1.4 The current state of the scholarly literature regarding the subject .................. 26
1.5 Methodology and key concepts ....................................................................... 32
1.6 Ideological geography and frontiers ................................................................ 36
CHAPTER 2: THE HISTORICAL BACKGROUND UP TO THE 11TH CENTURY
.................................................................................................................................... 40
2.1 Early Arab expansion into Byzantine territory ................................................ 40
2.2 The formation of a permanent frontier between the Byzantines and Arabs .... 44
2.3 The Byzantine offensive .................................................................................. 52
2.4 The Byzantine confrontation with Turkic peoples .......................................... 56
2.5 The appearance of Seljuk Turks in the Eastern frontier .................................. 61
CHAPTER 3: OLD MEMORIES, NEW BARBARIANS: TURKIC
“BARBARIANS” IN BYZANTINE HISTORIOGRAPHICAL IMAGINATION .. 69
3.1 The many faces of the eternal barbarian .......................................................... 69
3.2 Classifying the barbarians ............................................................................... 75
3.3 The Turks and Türkmen .................................................................................. 79
3.4 The Huns: The uses of a late antique ethnonym .............................................. 84
3.5 Persians between the Skythians and Saracens ................................................. 91
3.6 Pagans or Muslims: Turks, religion, and religiosity ........................................ 98
3.7 Mongols and Timurids in Byzantine sources ................................................ 107
xi
CHAPTER 4: BYZANTINE NARRATIVES ABOUT THE ORIGINS OF THE
TURKS ..................................................................................................................... 117
4.1 The genre of origo gentis: An overview ....................................................... 117
4.2 Origo gentis narratives about the Seljuks ...................................................... 122
4.3 Origo gentis narratives about the Ottomans .................................................. 130
4.4 Slaves as rulers: Aristotelian thought and the representation of Turkic
rulers ..................................................................................................................... 139
4.5 Conclusion ..................................................................................................... 147
CHAPTER 5: “BARBARIAN” POLITIES THROUGH BYZANTINE EYES...... 149
5.1 The place of barbarian polity in the Byzantine worldview ........................... 149
5.2 Archons and sultans: How the Byzantines saw the Turkic rulers? ............... 152
5.3 The formation and early conquests of Anatolian beyliks in Byzantine
sources .................................................................................................................. 159
5.4 Family intrigues and usurpers: Byzantine commentators on succession
and dynastic struggles in the Turkic states ........................................................... 164
CHAPTER 6: THE MECHANISMS OF CO-EXISTENCE: ANTAGONISM,
ACCULTURATION, ASSIMILATION ................................................................. 173
6.1 A conceptual introduction ............................................................................. 173
6.2 The entry of the Turks into Byzantine service: Pechenegs and Seljuks ........ 174
6.3 Tzachas and Syrgiannes Palaiologos: Two case studies ............................... 192
6.4 Evangelization as a way of integration .......................................................... 199
6.5 The place of Turks in Byzantine society ....................................................... 212
6.6 The Turcopoles and Mixobarbarians: A look into the grey areas ................. 214
6.7 An unsuccessful integration? ......................................................................... 221
6.8 Conclusion ..................................................................................................... 223
xii
CHAPTER 7: TRICKSTERS, MONSTERS, AND INVISIBLE WOMEN ........... 226
7.1 The tricksters at war: The Byzantine narratives about Turkic warfare........... 226
7.2 The representation of Turkic women in Byzantine literature .......................... 238
7.3 Intemperate and lustful: Turks and sexuality ...................................................... 248
7.4 Wild customs of the freaks: On the way of being dehumanized ..................... 261
7.5 Conclusion .............................................................................................................. 264
CHAPTER 8: CONCLUSION ........................................................................................ 267
APPENDIX: LE RÉSUMÉ DE THÈSE ........................................................................ 273
REFERENCES .................................................................................................................. 325
xiii
NOTE ON TRANSLITERATION
The transliteration of Byzantine names is a challenging task and there is no universal
standard. I used the standard transliteration of the Byzantine proper names, but the
names of people in common use in English (such as John, Michael etc.) are left as they
are. For Greek toponyms, I prefer the spellings with “k” instead “c”, for example, I
opted to use Kilikia instead of Cilicia. The spelling from the Encyclopedia of Islam is
used for Islamic names.
1
CHAPTER 1
INTRODUCTION
1.1 The presentation of the subject
This dissertation examines the representation of the Turkic peoples in late Byzantine
historiographic literature, roughly between the 11th and 15th centuries. It aims to
explore various aspects of the formation and evolution of the image of the Turkic
peoples in the Byzantine texts of the abovementioned period. It offers an analysis of
the representation of Anatolian or Seljuk Turks that the Byzantines encountered from
the mid-11th century; and also of the Turkic peoples dwelling in the Pontic steppes
and Northern Balkans of the same period, namely the Pechenegs, Cumans, and
Oghuz. This study is not limited only to Anatolian Turks because Byzantine authors
were well aware of the ethnic ties between the Turkic populations in their eastern and
northern borderlands. More importantly, the formation of the image of the Turks in
Byzantine literature is closely related to the ethnographic digressions about the
Turkic populations of the north that were written long before the arrival of the Seljuk
Turks in the Armenian highlands. Thus, the comparison between the “Eastern” (i.e.,
Seljuk and Ottoman) and “Northern” Turks constitutes one of the important issues of
this dissertation. How were these two groups first perceived and how did their image
evolve in the eyes of the Byzantine authors? How were these populations defined and
located in the Byzantine mind map of “us and barbarians”? What was the
relationship between earlier ethnographic digressions about the Turkic or non-Turkic
steppe peoples and post-11th century descriptions of the Turks? What were the
differences between the representations of “Anatolian” Turks and “Northern” Turks
and what role did religion play in this differentiation? How did Byzantine authors
2
perceive the origins, the state foundation and succession in the Turkic states, and
how did they define their rulers? What were the mechanisms of assimilation and
acculturation for the individual Turks in Byzantine society? Finally, how did
Byzantine authors comment on various aspects of Turkic people, including their
warfare, women, and sexuality? This dissertation seeks to answer these questions.
In the last two centuries of the time frame of this dissertation, the Mongols
take the place of the Pechenegs and Cumans in the comparative analysis with the
Anatolian Turks. Although the Mongols were neither a Turkic nor a Turkophone
people, they shared many sociocultural traits with the Turkic populations of Central
Asia. They preserved the ancient traits of the nomadic-pagan lifestyle of the steppes,
they were not fully Islamized, and their position in the geopolitics of the Near East
did not put them in an antagonistic position to the Byzantines. Their particular place
in Byzantine historiography does merit attention and is a useful subject to compare
with the Seljuk and Ottoman Turks.
In the present chapter, I offer an overview of my corpus and my
methodology. I have limited my corpus to works of historiographical character, in
other words, the histories and chronicles written by Byzantine authors from the
works of Michael Attaleiates and John Skylitzes in the last quarter of the 11th century
to the Histories of Laonikos Khalkokondyles that ends around the year 1467. The
only work of non-historiographic character that has been included in my corpus is the
Diegesis of Gregory Palamas, which narrates his days of captivity at the hands of the
early Ottomans. This work is unique and illuminates so well both the early Ottomans
and the Byzantine intellectual stance toward them that omitting it would have made
this dissertation incomplete.
3
In Chapter 2, I give an account of the historical background, from the early
Islamic conquests until the fall of Constantinople. This chapter also pays special
attention to the frontier as the fluid space where the encounter with the “other”
happens. It shows that the shock of the military encounter with the “barbarian other”
was simultaneous with real-life coexistence.
In the third chapter, I analyze the notion of the barbarian as the “other” of the
civilized self since the ancient Greeks and Romans. The ethnographic digressions
about the Persians, Egyptians, and Skythians in the Histories of Herodotus created
many topoi about the populations considered barbarian. These topoi, passing through
the prism of the Christian worldview, formed the basic lines of the representation of
the barbarian in Byzantine literature. This representation was not limited to the
dichotomy of barbarian and civilized. Rather, it was composed of a complex system
of ethnonyms which reflects a very large classification of barbarian populations
according to their geographical location, lifestyle, and ethnic origin. Moreover, the
elaborate Byzantine system of using archaic ethnonyms for contemporary
populations makes it necessary to find the place of Turkic peoples on the sociopolitical
map in the mind of Byzantine authors.
The next chapter focuses on the narratives about the origin of the Turks in the
Byzantine texts. These narratives reflect an effort to locate these peoples in the
familiar environment of Byzantine historiography, which itself was based on the
material of ancient Greek historiography and ethnography. The association of Turkic
peoples with the notion of slavery had a particular influence on the formation of the
image of the Turks in Byzantine literature. The same chapter also includes a brief
discussion of Aristotelian political thought and its place in the Byzantine worldview,
which contributed to this association.
4
In Chapter 5, I examine three aspects of the Byzantine-Turkic interaction in
the light of the literature: assimilation, acculturation, and antagonism. Many
members of Turkic societies, both from the north and the east, entered Byzantine
service in the 11th century. These individuals were generally fully assimilated into
Byzantine society; however, this did not change their general perception as outsiders.
On the other hand, their conversion to Christianity was the key factor in the process
of their integration into Byzantine society. Many families of Turkic origin existed in
Byzantine society; some were members of the aristocracy, who maintained their
patronymics demonstrating their ethnic roots.
Chapter 6 deals with the place of the barbarian entities in the Byzantine
worldview and then analyzes the Greek titulature used for Turkic rulers and the idea
of hierarchy behind it. This is followed by an examination of the Turkic state
foundation and its relationship with conquest, according to the Byzantine authors.
In Chapter 7, I focus on four particular aspects of the Turkic populations as
presented in our corpus: Turkic warfare, Turkic women, the sexuality of Turkic
peoples, and the cruelty attributed to them in the Byzantine texts. Finally, with a
concluding chapter, I end my study.
In a chronological study of historiography, there exists the risk of
representing the historiographic tradition as a linear and non-personal accumulation
of the works. However, the authors who contributed to our corpus were members of
the bureaucratic elite (such as Michael Attaleiates and Niketas Khoniates), members
of the ruling dynasty (such as Anna Komnene) or members of religious-political
movements (such as Gregory Palamas or Nikephoros Gregoras). Thus, their aim in
their writing career was not merely writing for the sake of literature; their authorial
5
ambitions had obvious political goals. Accordingly, their works will be analyzed by
underlining their personal aims and stances, political struggles and prejudices.
1.2 A note on the terminology: Which Turks, which Turkic peoples?
In this dissertation, the word Turkic is used as an umbrella term for all the
contemporary and historical populations that speak Turkic languages. As it is well
known, the Turkic languages are classified under two essential categories: Common
Turkic and Oghur Turkic. The only extant representative of the Oghur Turkic
languages is Chuvash, spoken in the Chuvash Republic in the Russian Federation, in
the Volga region. All other living Turkic languages (Turkish, Azerbaijani, Kazakh,
Turkmen, Kyrgyz, Uyghur, Uzbek, Yakut, etc.) are part of the Common Turkic
group. Historically, the Old Bulghar language is considered an Oghur Turkic
language. Pecheneg and Cuman, and most probably Khazar are considered Common
Turkic languages. However, several historical Turkic populations of Western Eurasia
could have been speakers of Oghuric Turkic, so the Oghur sprachbund may have
been larger in the early Middle Ages.1
The word Türk appears as a term that designates the founding population of
the Türk Khaganate, under the Ashina dynasty, centered in contemporary Mongolia,
in the 6th century. This Khaganate was an important element of Eurasian politics
until the 8th century, despite everlasting civil wars and Tang China’s attempts of
subjugation. The Türk Khaganate was not an ethnically homogeneous state and, just
like similar political formations in the steppe region; included many Mongolian,
Indo-European, and Uralic populations. The Soghdian language, which was once a
very common language across the urban settlements in the Central Asian steppe
1 Golden, An Introduction to the History of the Turkic Peoples, 17 and passim.
6
region and which belongs to the eastern branch of the Iranian language family, was
used in the chancellery of the Türk Empire, along with the Old Turkic language.
The ethnonym Turk became widespread with the rise of the Türk Empire in
the steppe region as a common term for all the nomadic populations. Particularly,
Muslim geographers used this term persistently for such populations, sometimes
even for peoples such as the Varangians and the Rus’ who were non-Turkic.
According to Peter B. Golden, this usage is similar to the usage of the ethnonym
Skythian by the Byzantine writers as a generic category.2
I use the word Turkish only to refer to the Turkish-speaking population of the
Sultanate of Rum and the later Ottoman Sultanate, as well as to the individual Turks
of Anatolia and the Balkans. The Turkish-speaking dwellers of these regions had a
certain process of acculturation and ethnogenesis that made them distinct from the
other Turkic populations of the region. These Turks could be considered “Turks of
Rum,”3 using the geographical term which they themselves applied to the region of
Anatolia.
1.3 The presentation of the sources
In this dissertation, I examine the Byzantine historiographical texts written between
the late 11th and 15th centuries in order to understand the formation and evolution of
the image of Turkic peoples in Byzantine literature. In this section, I shall present the
details of the life and times of the authors because in my approach to my corpus, I
follow Paolo Odorico’s method and try to understand each author’s aim and audience
2 Golden, “Some Thoughts on the Origins of the Turks and the Shaping of the Turkic Peoples,” 152.
3 On this paradigm, see Kafadar, “A Rome of One's Own: Reflections on Cultural Geography and
Identity in the Lands of Rum,” 7-25; Necipoğlu & Bozdoğan, “Entangled Discourses: Scrutinizing
Orientalist and Nationalist Legacies in the Architectural Historiography of the ‘Lands of Rum’,” 1-7.
7
to identify the goal of the author. As it has been argued by Odorico, Byzantine
literature, including historiography, has an absolutely utilitarian character, for
supporting an ideological position or for defending self-interest. Accordingly, the
motivation of the author, the audience he is addressing, the purpose of the patron
supporting the literary project are elements that are crucial for understanding and
interpreting the text. Thus, Byzantine literature should be read as a manifestation of
the personal and political conflicts of the empire. Furthermore, throughout this
dissertation, special attention is given to the links between these texts because they
cannot be studied as independent literary products; they are all part of the same
cultural universe.4
Below, I offer a chronologically ordered list of the sources I used for my
research. Each entry will discuss the context and personal aim of the authors of these
sources and comment on their development of a particular aspect or “theme” in the
Byzantine representation of the Turkic peoples. In my research, I only used
historiographic works addressed to Constantinopolitan audiences. Hence, I omitted
the works of Michael Panaretos, Leontios Makharias, and the anonymous chronicles
of the Morea and Ioannina. The only exception is Gregory Palamas’ narrative of
captivity at the hands of Ottoman Turks. I used this non-historiographical text
because of its particular character as the only autobiographical account of a
Byzantine who encountered the Turks (under the most inappropriate circumstances).
Other genres of Byzantine literature such as hagiographies, epistolography, poetry, or
rhetorical works also present material useful for the study of history; however, works
belonging to these genres are not part of my corpus. Also, the Byzantine
4 Odorico, “Displaying la littérature Byzantine,” 213-234, particularly 214-215. The articles in the
book of the colloquium held in Paris in 2008 reflect the same approach: Odorico (ed.), La face cachée
de la littérature byzantine, le texte en tant que message immédiat.
8
encyclopedic works are not included in my corpus, but I will refer to them in this
thesis when they are relevant for explaining the “Byzantine notions” and the mind
map of homo byzantinus.5 In short, my corpus is limited to historiographic works of
the middle and late Byzantine periods because history-writing, as a genre, presents a
direct continuity with Classical Antiquity during which we see the beginnings of the
representations of the images of various ethnic groups.
Before introducing the individual works in my corpus, it will be appropriate
to briefly comment on the genres of historiography in Byzantium. Byzantine
historiographic works are generally divided into two groups as histories and
chronicles. This division is a product of 19th-century historiography. Chronicles are
mostly works that have a “year by year” or “emperor by emperor” approach and they
employ a simpler language and are often based on earlier works. They also often tend
to begin with the Creation of the world. Histories are much more complex works that
are characterized by a cause-and-effect approach rather than simple chronology, and
they often use embellished language. They also deal with specific time frames.
However, since the Byzantine audience did not have a conception of such categories,
I chose not to apply such a division of genres to my corpus.6
1.3.1 Historia by Michael Attaleiates
This work covers the period between the years 1034 and 1080. The author is a
bureaucrat from the city of Attaleia (Antalya) who seems to be a rather erudite
5 This conceptualization has been proposed by Alexander Kazhdan and Giles Constable, in their 1982
book People and Power in Byzantium: An Introduction to Modern Byzantine Studies. In this
groundbreaking study, which deals with various aspects of the Byzantine individuals’ lives and the
Byzantine mentality, the authors conceptualize homo byzantinus as a concrete personality inside his
material and spiritual environment, busy with his daily occupations. See, particularly, the introduction
that deals with earlier historiography and chapters 1 and 5. Kazhdan & Constable, People and Power
in Byzantium: An Introduction to Modern Byzantine Studies, 1-36 and 96-116.
6 Odorico, “What’s in a Name? The Byzantine Chroniclers,” 85-86 .
9
person who studied law in the mid-11th century in Constantinople. He then became a
judge and became a part of the political elite of the empire. He was a member of the
faction that supported the young Anatolian aristocrat Romanos IV Diogenes to rise to
the throne. When the latter became emperor, Michael had the position of judge of the
army (κρıτὴς τοῦ στρατοπέδου). He was very active during the reign of Romanos IV.
He participated in his campaigns, including the Battle of Manzikert (1071). His
chapters dealing with the campaigns and other military matters are lively and worthy
of attention. After the fall of Romanos IV and the eruption of civil war, Attaleiates
aligned himself with Nikephoros III Botaneiates, to whom he dedicated his history.
The work ends with a laudatory description of the domestic policies of
Nikephoros III, who lost the throne the following year. The author died around the
year 1085.
Unlike his near-contemporary Skylitzes, Attaleiates is not an invisible author.
He has a powerful personal voice and integrates some autobiographical material into
his History. This fact is particularly visible in his chapters dealing with the
campaigns of Romanos IV. As already mentioned, he also does not hide his political
sympathies for certain figures, most importantly for Nikephoros III. Furthermore,
Attaleiates’ earlier chapters coincide with the narrative of Skylitzes on the same
period, which permits a comparative reading of the two texts.
Apart from his History, Attaleiates authored two other works: the Ponema
Nomikon, a treatise on law, and the Diataxis, the foundation document of his
monastery in Constantinople and poorhouse in Raidestos (Tekirdağ). The latter text
also contains some autobiographical details of the author.
Attaleiates’ History has survived only in two manuscripts (Coislianus gr. 136
and Scorialensis T.III. 9.) The text was put together in both manuscripts with the
10
Synopsis Historion of Skylitzes. The text commonly called “the Continuation of
Skylitzes” reproduces several passages of Attaleiates.7
1.3.2 Synopsis Historion by John Skylitzes
John Skylitzes’ Synopsis Historion was written at the end of the 11th century and it
covers the events from 811 to 1057. This work is essentially a compilation and rewriting
of earlier historical sources. The author seems to be in the shadows and does
not manifest a personal tone in his text. He uses and sometimes mentions some of the
sources that he used, such as the work of Joseph Genesios, Theophanes Continuatus,
George Synkellos and the Vita Basilii. His narrative of the events is simple, not
always chronologically accurate, and divided into chapters according to the reigns of
the emperors. It still reflects some personal views, especially in the case of his moral
comments about the deeds of the emperors or other statesmen; such are the generally
positive narratives of the emperors of the Macedonian dynasty. He is very favorable
of the reign of Basil II and sees the beginning of the Byzantine decline in the reign of
Constantine IX Monomachos.8
His digressions about the Seljuks and Pechenegs are worthy of attention.
They seem to be based on now lost material. The digression dealing with Seljuks is
reproduced almost word by word by Nikephoros Bryennios.9 However, Skylitzes
7 Editions: Attaleiates’ text was first published as part of the Corpus Scriptorum Historiae Byzantinae
(CSHB), the so-called “Bonn Corpus.” Recently Inmaculada Perez Martin prepared a new critical
edition with Spanish translation: Miguel Ataliates: Historia, Madrid, 2002. The most recent edition of
the text is the bilingual edition of Dumbarton Oaks Medieval Library: Attaleiates-Kaldellis-Krallis,
The History. I will use this most recent bilingual edition. Secondary literature: Krallis, Serving
Byzantium's Emperors: The Courtly Life and Career of Michael Attaleiates.
8 Skylitzes-Wortley, A Synopsis of Byzantine History, xxviii.
9 A notable exception is that Bryennios does not identify the Seljuk Turks with the Huns. For a
commentary on the representation of Byzantium’s enemies in the work of Bryennios, see Neville,
Heroes and Romans in Twelfth-Century Byzantium: The Material for History of Nikephoros
Bryennios, 82.
11
does not have a particular interest in ethnography and these digressions reflect
political features or scenes of war, yet not the ethnographical details. Skylitzes’
Synopsis Historion was reproduced frequently in later centuries. Hans Thurn, who
prepared the first modern edition of the work, used nine manuscripts of the text
written between the 12th and 14th centuries. Moreover, George Kedrenos’
Chronographia includes the entire text of Skylitzes. The manuscript commonly
known as “Madrid Skylitzes” (Codex Matrit. Bibl. Nat. Vitr. 26. 2.) has a particular
importance as it contains very important illustrations. Finally, the Synopsis Historion
can be considered as a work representing the historical knowledge of the average
Byzantine literati. Skylitzes’ work is used as a standard reference for mid-Byzantine
history until the 15th century.10
1.3.3 Alexiad by Anna Komnene
Anna Komnene’s Alexiad can be considered one of the most personal historiographic
works of entire Byzantine literature. It is also the first history written by a woman in
European literature, so this unique feature makes it very important for the point of
view of gender relationships in the literature. Anna was the daughter of emperor
Alexios I Komnenos (r. 1081-1118) and her history primarily narrates the life and
deeds of her late father, and it has a distinct laudatory tone. She depicts her father as
a heroic and religious emperor. It covers the period 1081-1118, coinciding exactly
with the tumultuous reign of Alexios I. However, the chapters dealing with Alexios’
early reign are much more detailed than the later ones.
10 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum is the critical edition of the work by Hans
Thurn. English translation: Skylitzes-Wortley, A Synopsis of Byzantine History. Secondary literature:
Neville, Guide to Byzantine History Writing, 155-160.
12
Anna’s husband, Nikephoros Bryennios (known as Nikephoros Bryennios the
younger), also wrote a historiographical work called Hyle Historias, which is an
incomplete and arguably less important work than the Alexiad. Anna also used this
book and integrated many of its materials into her own history. Being a Byzantine
princess, Anna had access to imperial archives, and she used them for her history.
The reign of Alexios I does not have multiple sources like other periods, and
the Alexiad, despite its panegyrical tone, gives us many insights into this very
tumultuous period. Anna’s access to the imperial archives makes possible the
preservation of many minor persons or events of the reign of Alexios I. For my
study, Anna Komnene’s work is of utmost importance because it gives an account of
the early history of the presence of Seljuk Turks in Asia Minor and the activities of
Turkic warlords in the chaotic situation of the peninsula. Many Turkic warlords of
this period could only be known today because they were referred to in the Alexiad.
Anna was an intellectual woman of her period and her writing style reflects
both an elegant style and her erudition. However, her chronology is sometimes
ambiguous, and she constantly omits events which were not important for the career
of Alexios I. Her text is an almost unique work for a crucial period in the history of
the Byzantine Empire.11
1.3.4 Epitome by John Kinnamos
John Kinnamos’ work is of the utmost importance in understanding the Komnenian
period. Little is known about his historical personality. Charles M. Brand defines him
11 Editions: Alexiad was first published in the CSHB series. Its most recent critical edition has been
prepared by Dieter Reinsch and Athanasios Kambylis, as part of the CFHB collection: Komnene-
Reinsch-Kambylis. Annae Comnenae Alexias. Alexiad’s standard English translation is the revised
edition of the translation of E. R. A. Sewter. Komnene-Sewter-Frankopan, The Alexiad. Secondary
literature: Gouma-Peterson (ed.), Anna Komnene and Her Times; Neville, Anna Komnene: The Life
and Work of a Medieval Historian.
13
as an “ordinary Byzantine bureaucrat,” and although his knowledge in the area of the
classics and religion was unquestionable, he has the simple and straightforward
approach of a bureaucrat.12 He was a member of the entourage of the emperor
Manuel I Komnenos (r. 1143-1180). He was present at some of the emperor’s
campaigns including the Battle of Myriokephalon. His work covers the period from
the beginning of the reign of John II Komnenos (r. 1118-1143) and ends
unexpectedly in 1176, before the battle of Myriokephalon. Although his account of
the reign of Manuel I was mostly based on his eyewitness observations, his account
of the rule of John I is taken from oral sources. It is basically a political and military
history of the late Komnenian Empire. The author’s interest in the diplomacy and
politics of the period deserves attention. Being a supporter of Manuel Komnenos, the
author gives very rich details about the ethnic composition of the Byzantine army
and ruling classes. The latter parts of the Epitome cover the periods narrated in the
work of Khoniates. This fact gives us a chance to take a comparative look at what
these two authors wrote about the same events. A remarkable aspect of Kinnamos’
history is his anti-Latin approach.13
1.3.5 Chronike Diegesis by Niketas Khoniates
Niketas Khoniates’ Chronike Diegesis covers the period between the years 1118 and
1207. However, his later chapters are much more detailed than the earlier chapters.
Khoniates was a man of the province from the town of Khonai (Honaz) in Western
Asia Minor. Harry J. Magoulias, who translated the Annals of Khoniates, speculates
12 Kinnamos-Brand, Deeds of John and Manuel Komnenos, 2.
13 Editions: It was first published in the CSHB series in a single volume together with the history of
Nikephoros Bryennios: Ioannis Cinnami Epitome rerum ab Ioanne et Alexio (sic) Comnenis gestarum.
This volume was edited by August Meineke (Bonn, 1836). Charles M. Brand translated the work into
English: Kinnamos-Brand, Deeds of John and Manuel Komnenos. Secondary literature: Neville,
Guide to Byzantine Historical Writing, 186-190.
14
that he most likely came from the provincial lesser nobility. Michael Khoniates, the
brother of Niketas, served as the archbishop of Athens during 1182-1204. Niketas
had a bureaucratic career, even reaching the post of logothetēs tōn sekretōn. As
demonstrated in his work, he was a member of the imperial court in the late 12th and
early 13th century. He lost his office and his ostentatious palace during the Crusader
occupation of Constantinople in 1204 and fled first to Selymbria, then to Nicaea.
However, in his later years, he could not obtain a bureaucratic position in the empire
in exile.
As a writer, he is considered a master of Byzantine prose. His work is rich in
details and descriptions. His comments about the events of his time are also worthy
of attention. His history is particularly anti-Latin or anti-Western, which is not
surprising for a Byzantine intellectual who experienced the trauma of the Latin
occupation of Constantinople of 1204. Apart from his historical writings, he wrote
about religion: he was the author of Panoplia Dogmatike, a polemical text against
heresies.14
1.3.6 Chronike Syngraphe by George Akropolites
George Akropolites (1217-1282) was the author of a historical work that deals with
the history of the Empire of Nicaea. The author belongs to a Constantinopolitan
family that served in the Byzantine upper bureaucracy since the late 12th century.
After the fall of Constantinople into the hands of Crusaders, the author’s family took
refuge in the domain of the Laskaris-Vatatzes family in western Anatolia. His father
was logothete Constantine Akropolites. The author had a good education; he studied
14 Edition: Khoniates-Van Dieten, Nicetae Choniatae Historia; English translation: Khoniates-
Magoulias, O City of Byzantium: The Annals of Niketas Khoniates. Secondary literature: Neville,
Guide to Byzantine Historical Writing, 219-225.
15
philosophy with Nikephoros Blemmydes, the illustrious philosopher of the 13th
century. Similar to his father, the author was appointed megas logothetes of the
Empire of Nicea. After the conquest of Constantinople by Michael VIII Palaiologos
(r. 1259-1282) in 1261, the author was appointed as teacher at the University of
Constantinople. He also served in Byzantine diplomatic missions during which his
most important function was to be part of the delegation sent to the Council of Lyon
in 1274. He died in 1282.
He wrote a brief historical work called Chronike Syngraphe. This work gives
us rich details about Byzantine-Seljuk relations in the first half of the 13th century. It
could be considered a relatively balanced work, which is usually based on personal
observations. His main goal is presenting the Empire of Nicaea as the legitimate
successor of the Byzantine Empire. His work is very illuminating for the Seljuk-
Nicene relationships in the mid-13th century, which is a period where there was no
other historiographical work written in Greek. Moreover, Chronike Syngraphe is a
unique work that represents the perspective of the Empire of Nicaea.15
1.3.7 Syngraphikai Historiai by George Pakhymeres
The great intellectual and cleric George Pakhymeres (1242-1310) was the author of a
history of the early Palaiologan period. His history, covering the period between the
years 1261 and 1307, is called Syngraphikai Historiai. As a man of rich cultural and
intellectual erudition, George Pakhymeres was not only a historian, but he also wrote
treatises on philosophy and theology. He was born in Nicaea, but his family origins
were Constantinopolitan. He was a member of the clergy and attained high positions
15 Akropolites-Bekker, Georgi Acropolitae Annales; English translation: Akropolites-Macrides,
History. Secondary literature: Neville, Guide to Byzantine Historical Writing, 226-231
16
in the Patriarchate. He held many important clerical functions, such as protekdikos
and dikaiophylax. The author, who wrote his work after the death of Michael VIII
Palaiologos, has a somewhat critical approach to the emperor.
As an author, Pakhymeres uses a rich and archaizing prose. His voluminous
historiographical work is especially important for early Ottoman history because it is
the first source in which ʿOt̲h̲ mān I, the founder of the dynasty, is attested by name.
He also gives other interesting details of the early 14th-century Turkish expansion in
Western Asia Minor. He is also an important source for Byzantine-Mongol relations
and the Seljuk sultan Kaykāʾūs II’s (r. 1246-1262) life in Constantinopolitan exile.16
1.3.8 Rhomaike Historia by Nikephoros Gregoras
Nikephoros Gregoras was a famous theologian and writer of the 14th century. He was
born in Herakleia Pontika, where he was educated by his uncle who was the
metropolitan of that district. Then he went to Constantinople and entered the ranks of
high clerics. He had a turbulent political life; first, he was a member of the entourage
of patriarch John XIII Glykis; then he was known as a partisan of Andronikos II and
then John VI Kantakouzenos. Apart from his religious duties, he was also engaged in
diplomatic activities. After 1347, he fell out of favour because of his anti-Palamist
positions, and from 1351 until his death in 1360 he lived under house arrest. Apart
from his historical work, he was the author of many other important works, including
treatises on astronomy and calendar reform proposals, various religious and
philosophical works, and hagiographies.
16 Pachymeres-Failler-Laurent, Relations historiques, 5 vols. Secondary literature: Neville, Guide to
Byzantine Historical Writing, 237-242.
17
His account of contemporary history, known as Rhomaike Historia (Roman
History), is a vast historical study comparable to the work of Pakhymeres. In this
monumental work, Gregoras covers the period 1204-1359 and gives special attention
to religious matters. His interest in the early Ottoman expansion is also noteworthy.17
1.3.9 The works of Gregory Palamas
Gregory Palamas (1296-1357) was not a historiographer but a churchman who
generally wrote about religious issues. He was associated with hesychasm
(ἡσυχασμός), a mystical tradition in Orthodox Christianity that gives special
importance to constant contemplation and inner prayers. Although the practices
associated with hesychasm already existed before his time, Gregory Palamas’
teachings made it into a doctrinal synthesis as Palamism. Gregory Palamas had
family origins in Asia Minor; his father was a senator and died in about 1303,
leaving his son as an orphan boy in the imperial court. Palamas received great
support from the emperor Andronikos II Palaiologos (r. 1282-1328) and had a very
good education. He turned to monastic life at a relatively young age and lived on
Mount Athos. Starting in 1341, Palamism gained the support of John VI
Kantakouzenos, who was the leader of a faction in the Byzantine civil war of
1341-1347.
He wrote intensely on religious and philosophical themes. In 1354, he was
captured by Turkish pirates and remained for nearly one year in captivity, in
Lampsakos (Lapseki), Pegae (Biga), Brusa (Bursa) and Nicaea (İznik), among
which, the last two cities are important urban centers of the early Ottoman state.
17 Gregoras-Schopen-Bekker, Nicephori Gregorae Byzantini Historia; Gregoras-Van Dieten,
Rhomäische Geschichte. Secondary literature: Neville, Guide to Byzantine Historical Writing, 243-
248.
18
There he engaged in theological discussions with the Muslim clergy and an
enigmatic group of men called the Xionai (Χιόναι), who appear to have been
members of a Judaizing religious movement, recently converted to Islam. In his
captivity, he also met Prince Ismail, the grandson of Ork̲ h̲ an (r. 1324-1362), the
second ruler of the Ottomans. His account gives some insight into the mentality of
both the Byzantines and Turks of the 14th century.
Anna Philippides-Braat prepared an edition of the correspondence and
narrative of the theological discussions concerning the captivity of Palamas, with a
textual and historical commentary. Palamas’ narrative of captivity gives us a lively
account of the early Ottoman socio-cultural milieu from the point of view of a
Byzantine intellectual. Because of this unique feature, I have added this text to my
corpus even though it is not historiographic in nature.18
1.3.10 Historiai by John Kantakouzenos
As an emperor and scholar, John Kantakouzenos (c. 1295-1383) was one of the most
predominant personalities of 14th-century Byzantium. He came from an aristocratic
family whose members attained administrative positions in the 12th century. He
appears to be a close collaborator of Andronikos III Palaiologos, who rebelled
against his grandfather Andronikos II Palaiologos, and was appointed megas
domestikos by the emperor. He continued to hold this office until Andronikos II’s
death in 1341. After the death of the emperor, he became the regent to the emperor’s
son John V (r.1341-1376 and 1379-1390), and this fact triggered a civil war (1341-
18 Philippides-Braat, “La captivité de Palamas chez les Turcs: dossier et commentaire,”109-221. There
is also a recent English translation of these texts: Russell, Gregory Palamas: The Hesychast
Controversy and the Debate with Islam. For a biography of Palamas, see Meyendorff, A Study of
Gregory Palamas.
19
1347) between the faction of John Kantakouzenos and the faction of empress Anna
of Savoy and megas doux Alexios Apokaukos. In this war, the faction of John
Kantakouzenos was allied with the adherents of the hesychasm movement. At the
end of the war, John Kantakouzenos was proclaimed co-emperor as John VI. He
reigned until 1354 and then renounced the throne and became a monk. In his
monastical life, he wrote his History, which could be considered his Memoirs.
His work is made up of four books and demonstrates the influence of
Thucydides. This work is particularly important for the representation of the Turks
because of his personal initiative in the relationship with the Turks, including the
first marriage alliance between the Byzantines and Ottomans: his daughter Theodora
Kantakouzene married the Ottoman ruler Ork̲ h̲ an in 1346.19
1.3.11 Historia Turco-Byzantina by Doukas
Doukas is a historian of the transition period. This transition does not only mean the
transition from the Byzantine Empire to the Ottomans but also the formation of the
Ottoman domination in Muslim Anatolia. His text covers the period from the mid-
14th century up to 1466. Doukas, whose first name is unknown, was probably related
to the house of Doukai, who played an important role in Byzantine history during the
11th century. However, when the author was born, his family had already left the
Byzantine capital and migrated to the city of Ephesus, which was under the
domination of Aydinids in that period. The author’s grandfather, Michael Doukas,
was a member of the entourage of İsa, the Aydinid ruler of Ionia. The author spent
19 Kantakouzenos-Schopen, Ioannis Cantacuzeni eximperatoris Historiarum; Kantakouzenos-
Fatouros-Krischer, Johannes Kantakuzenos: Geschichte. Secondary literature: Neville, Guide to
Byzantine Historical Writing, 266-272; Kaldellis, A New Herodotos; Nicol, The Reluctant Emperor: A
Biography of John Kantakouzenos, Byzantine Emperor and Monk c. 1295-1383.
20
the majority of his life outside of the Byzantine capital. In his youth, he appears to
have lived in New Phokaia as the secretary of the Genoese podestà Giovanni Adorno
of this tiny city. Then, he went into the service of the Gattilusio dynasty, the rulers of
Lesbos. He undertook some diplomatic missions on behalf of the Genoese rulers of
Lesbos and personally visited various former Byzantine cities which were now part
of the Ottoman Empire.
As demonstrated in his career, he was a dedicated supporter of the Genoese.
Hence, he could even be considered the most “pro-western”20 author in our corpus.
His takes reflect a very strong agony for the loss of the former heartland of the
Byzantine world to the Ottomans.
His history roughly covers the same period as the Historiai of Laonikos
Khalkokondyles; however, their focus is very different, and Doukas’ work lacks the
digressions about early Ottoman history in the text of Khalkokondyles. However,
Doukas’ text particularly merits attention for his non-Constantinople-centric view of
events and his rich details about western Asia Minor, such as his narratives about the
rebellion of Börklüce Muṣṭafā and the downfall of Junaid Beg, the last independent
ruler of the Aydin Emirate.21
1.3.12 Apodeixis Historion by Laonikos Khalkokondyles
Laonikos Khalkokondyles (c. 1430 – c. 1490), who presents himself as “Laonikos
the Athenian” in his Apodeixis Historion (Demonstration of Histories), was the
author of one of the most important works on the history of the 15th-century Ottoman
20 For a study dealing with late Byzantine political attitudes toward the Ottomans and Westerners, see
Necipoğlu, Byzantium Between the Ottomans and the Latins: Politics and Society in the Late Empire.
21 Doukas-Grecu, Historia Turcobyzantina; English translation: Doukas-Magoulias, Decline and Fall
of Byzantium to the Ottoman Turks. Secondary literature: Neville, Guide to Byzantine Historical
Writing, 298-301; Grecu, “Pour une meilleure connaissance de l’historien Doukas,” 128-141.
21
state. His literary first name Laonikos was probably an amalgam of his personal
name Nikolaos. Like Doukas, his near-contemporary, he was born in an ex-
Byzantine territory ruled by the Latins, in the Duchy of Athens under the Acciaiuoli
family. His family was part of the political élite of this tiny state; however, his father
was forced to leave Athens to the Despotate of Morea because of his unsuccessful
political intrigues. The author grew up in the fragmented and vanishing world of
these last Christian strongholds in the southern Balkans.
Khalkokondyles, reflecting the zeitgeist of the 15th century, was more
interested in ancient Greek civilization rather than the contemporary Byzantine
Orthodox tradition. He was a member of the entourage of Gemistos Plethon, the neo-
Platonist Byzantine philosopher who sought a revival of Hellenic culture.
He clearly follows the narrative model of Herodotus. His history deals with
the rise of the Ottomans as Herodotus’ narrative of the rise of the Persians. Although
it was centered on the narrative of the rise of the Ottomans, Khalkokondyles seems to
be interested also in the rest of the world, and his narrative includes geographical
digressions about Western and Eastern Europe and the Arab world. His account gives
the impression that he had some relationship with the Ottoman ruling classes of the
1460s, very likely for the most part with those of a Greek devşirme origin. According
to Anthony Kaldellis, who translated and edited his work and wrote a monograph
about the Histories, Khalkokondyles’ intended audience was probably
Constantinopolitan Greeks after the city’s fall. This view was recently challenged by
Aslıhan Akışık-Karakullukçu. According to her, the author’s intended audience was
made up of western literati who were educated in classical Greek.22
22 For the discussion about Khalkokondyles’ audience see Kaldellis, A New Herodotos, 199; Akışık-
Karakullukçu, “A Question of Audience: Laonikos Chalkokondyles’ Hellenism,” 1-30.
22
His later life is a matter of discussion: he may have migrated to Italy like his
relative, the Renaissance humanist Demetrios Khalkokondyles, but it seems that
there is no solid evidence about such a fact.
In brief, Khalkokondyles gives us a balanced narrative of the rise of the
Ottomans without a particularly biased regard.23
1.3.13 Historia by Kritoboulos of Imbros
Michael Kritoboulos was a scholar and statesman from the island of Imbros. Similar
to Laonikos Khalkokondyles, he adopted the pen name Kritoboulos by changing his
original surname Kritopoulos. He seems to be erudite in classical culture and served
as the governor of Imbros under the reign of Mehmed II. These qualities make the
author one of the exceptional figures of the Byzantine-Ottoman transition period.
His work survived in only one manuscript, the Topkapı Manuscript, which
seems to be his autograph text. Another copy of the text, published in Germany in the
19th century, is now lost.
The particular feature of the work of Kritoboulos is its political position. His
history starts with a dedication addressed to the Ottoman ruler Mehmed II. The text
could be considered the most pro-Turkish of all the works in the corpus. This fact is
closely related to the political alignment of the author. This pro-Turkish attitude does
not reflect only Kritoboulos’ approach, but also a good portion of the Byzantine elite
in the 15th century, such as Gennadios Scholarios, George Amiroutzes, and, many
others. The collaboration with the Ottomans was a survival strategy for an important
faction of Byzantine statesmen and clergy. Historia covers a relatively short period
23 Khalkokondyles-Kaldellis, Histories. Secondary literature: Neville, Guide to Byzantine Historical
Writing, 312-318; Kaldellis, A New Herodotos.
23
of time, it narrates the events of the years 1451-1467, hence its narrative centers on
the reign of Mehmed II. The author also mentions his will to write another work on
the Ottoman past, a dynastic history that covers the totality of Ottoman history in
Historia, but it seems that this project was not undertaken.
Kritoboulos was very faithful to ancient literary models and wrote his work in
an Atticizing language. His numerous comparisons between ancient history and the
rise of Ottomans, as well as associations between Alexander the Great and
Mehmed II are worthy of attention. In brief, Kritoboulos offers his solution to the
ideological crisis in Greek society after the fall of Byzantium by presenting the
Ottomans as the legitimate successors of the Roman Empire.24
1.3.14 Chronicon Minus by George Sphrantzes
George Sphrantzes was a Byzantine statesman and author who wrote a work that
could be considered both the autobiographical text of a Byzantine diplomat and an
annalistic chronicle about the Byzantine-Ottoman transition in the 15th century.
Sphrantzes was closely related to the Palaiologan court, he served as a diplomat for
the Palaiologan rulers on various occasions. He was a member of the entourage of
Constantine XI, the last emperor of Byzantium. He was also an eyewitness of the
conquest of Constantinople by the Turks, where he was taken prisoner along with his
family. After his captivity, he mostly lived in Greece, in Mistra and Corfu. He spent
his last years as a monk and died toward the end of the 1470s.
24 Kritovoulos-Reinsch, Critobuli Imbriotae Historiae. English translation: Kritovoulos-Riggs,
History of Mehmed the Conqueror. There is a recent Turkish edition that includes the facsimile of the
manuscript with a Turkish translation: Kritovoulos-Çokona, Kritovulos Tarihi. Secondary literature:
Neville, Guide to Byzantine Historical Writing, 308-311.
24
Despite his work lacking ethnographical material, Sphrantzes presents some
valuable material about the 15th -century Turco-Byzantine coexistence, mostly based
on personal anecdotes. Despite the suffering of his family in the hands of Ottomans,
his account has generally a milder style. His narrative is important for the history of
Byzantine diplomacy, particularly for the relations with the Turks. His Chronicon is
also noteworthy because of the use of the Greek language; the author uses a language
closer to colloquial Greek. This fact represents also a departure from the older
literary models.
The work of Sphrantzes, known as Chronicon Minus, survived in three
manuscripts. All three manuscripts are held in Italian libraries. A more
comprehensive text that repeats the content of Chronicon Minus is the Chronicon
Maius, which once was attributed to Sphrantzes. However, now it is well understood
that Chronicon Maius was written by Makarios Melissenos, the metropolitan of
Monemvasia.25
In concluding this sub-chapter on my sources, I present Table 1 below, in
which the authors whose works I examined are categorized according to their social
backgrounds, geographic origins, and political positions. Although such a
categorization is far from expressing the complexity of the characters of these
authors, methodologically it helps us understand and interpret them better.
25 Sphrantzes-Maisano, Georgii Sphrantzae Chronicon. English translation: Sphrantzes-Philippides,
The Fall of the Byzantine Empire. A Chronicle by George Sphrantzes, 1401-1477. Secondary
literature: Neville, Guide to Byzantine Historical Writing, 302-307.
25
Table 1. The Authors
Author Profession Region of Origin Political stance
Attaleiates High-level
bureaucrat
Attaleia Pro Nikephoros III
Skylitzes High-level
bureaucrat
Thracesion
Theme
Pro-Macedonian dynasty
Komnene Imperial
princess
Constantinople
(Paphlagonia?)
Pro Alexios I
Kinnamos Mid-level
bureaucrat
? Pro-Komnenian (generally), antiwestern
Khoniates High-level
bureaucrat
Khonai (Phyrgia) Critical of late Komnenian emperors,
anti-western
Akropolites High-level
bureaucrat
Constantinople Pro-unionist
Pakhymeres Cleric,
theologian
Constantinople Anti- Michael VIII
Gregoras Cleric,
theologian
Herakleia Pontika Pro-unionist
Kantakouzenos Soldier,
emperor
Constantinople Anti-unionist
Palamas Cleric,
theologian
Constantinople Anti-unionist
Doukas Mid-level
bureaucrat
Ephesus Pro-Unionist, pro-western
Khalkokondyles Mid-level
bureaucrat
Athens Slightly pro-Ottoman
Kritoboulos Mid-level
bureaucrat
Imbros Pro-Ottoman
Sphrantzes Diplomat Constantinople Anti-unionist, anti-Ottoman
26
As seen in Table 1, the majority of the authors in my corpus belong to the upper
segments of Byzantine society; they were either high-ranking officials or men of the
church. There is only one woman (Anna Komnene) among them. Two of them (Anna
Komnene, Kantakouzenos) are directly linked with the imperial throne, and a third
(Doukas) possibly has descent from an imperial dynasty. None of the authors had a
non-Greek ethnic origin. Three of them were born and raised in the frontier region
(Khoniates, Gregoras, Doukas) and this experience was slightly reflected in their
works.
1.4 The current state of the scholarly literature regarding the subject
There exists a rich and growing literature about the history of the Late Middle Ages
in Anatolia. From a broader perspective, in recent years there have been flourishing
discussions about individual and group identities, as well as the contemporary waves
of migrations to developed countries from the less developed parts of the world, and
there are questions concerning the integration of the foreign and immigrant
populations to where they moved. Studies focusing on identity have recently become
widespread in academia, and this concept bears a strong influence on the Anglo-
Saxon approach to the social sciences. These discussions naturally influence and
shape the social sciences; mostly political science and sociology, yet also history. In
short, the debates about identities make both the communities formed around these
identities and the "others" excluded by them very current topics in social sciences.
A review of the scholarly literature on the subject of this dissertation must
begin with Gyula Moravcsik’s monumental two-volume work about the Byzantine
sources dealing with the history of Turkic peoples (and the Hungarians and
Mongols), which is still of utmost importance for any researcher interested in the
27
topic. Despite advances in the fields of Byzantine Studies and Turcology, no book
has yet been written to replace this precious work, which was first published in
1942.26 The author’s meticulous attention to marking every reference to a Turkic
group or individual of Turkic origin is very helpful for any study concerning the
people of Turkic origin in Byzantium. However, because of its date of publication, it
does not contain any edition or research published in the last seventy years. Its
bibliography is, therefore, very old and must be supplemented with contemporary
works.
Among major works that deal with the Byzantine-Turkish encounter in
Anatolia, there are two comprehensive studies that stand out, namely The Decline of
the Medieval Hellenism in Asia Minor by Speros Vryonis, published in 1971, and La
Turquie pré-ottomane by Claude Cahen, published in 1988, but based on an earlier
English edition without footnotes dated 1968.27 As their titles indicate, the former
book is centered on the declining Byzantine world and the latter on the formation of
Turkish polities in Anatolia. These two works must, therefore, be used together to
have a comparative picture of the Byzantine-Turkish political and social dynamics in
late medieval Anatolia. Michel Balivet is another scholar who has made significant
contributions to this field. In contrast to the generalist approach of Cahen and
Vryonis, Balivet treated many specific issues regarding the interactions between the
Byzantine and Turkish domains in his numerous articles.28 In a short article
published in 1993, Nicolas Oikonomides examined the Turkish image in Byzantine
26 Moravcsik, Byzantinoturcica I: Die byzantinischen Quellen der Geschichte der Türkvölker;
Moravcsik, Byzantinoturcica II: Sprachreste der Türkvölker in den byzantinischen Quellen.
27 Vryonis Jr, The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from
the Eleventh Through the Fifteenth Century; Cahen, La Turquie pré-ottomane; Cahen, Pre-Ottoman
Turkey.
28 For two volumes containing his collected articles, see Balivet, Byzantins et Ottomans: Relations,
interaction, succession and Balivet, Mélanges byzantins, seldjoukides et ottomans.
28
orations recited in the presence of the emperor by 12th-century writers. More
recently, in an article that appeared in 2011, Angeliki Papageorgiou also discussed
several aspects of the representation of the Turks in the mid-12th century. These two
short articles complete each other.29
A particular work, which I envision myself to be in dialogue with, is the
seminal study of Rustam Shukurov: The Byzantine Turks, 1204-1461.30 In this work
of monumental character, which was published in 2016, Shukurov touches on several
questions that are explored in this dissertation. These include, first, the issue of the
categorization and classification of barbarians – the genoi and ethnoi – and,
secondly, the question of the presence of the Turks in the Byzantine Empire, or with
more accurate terminology, in the Byzantine spaces. The first chapter of Shukurov’s
book is dedicated to the Byzantine classification of the Turks. In the second chapter,
the author presents his database of oriental names in the Byzantine cultural space that
belonged to people of presumably Turkic origin. Then, in the following chapters,
Shukurov meticulously investigates the demographic, social, and cultural
implications of the Turkic presence within the Byzantine borders. The sixth chapter
of his book is particularly important for this dissertation as it discusses the tools of
assimilation for integrating the people of Turkic origin into Byzantine society.
Shukurov introduced with this work several novelties to the field of Byzantine
studies. Firstly, he is the first scholar who dedicated a monograph to the question of
the presence of Turks in Byzantium. Apart from a few articles, these individuals
were mostly overlooked by modern historiography. A second contribution is
29 Oikonomides, “The Turks in the Byzantine Rhetoric of the Twelfth Century,” 149-155 and
Georgiou, “οί δέ λύκοι ώς Πέρσάι: The Image of the “Turks” in the Reign of John II Komnenos
(1118-1143),” 149-161.
30 Shukurov, The Byzantine Turks, 1204-1461.
29
Shukurov’s use of antroponymics as a tool to research his subject. The use of
praktika, chrysobulls, and various religious or private documents has made possible a
new perspective on the prosopography of the Turkic people in Byzantium. The
Byzantine Turks, 1204-1461 and the present dissertation touch sometimes on the
same questions regarding both the Byzantine classification of the Turkic people and
their assimilation into Byzantine society, but the methodologies and aims of these
two works are totally different. Whereas Shukurov enquires about the situation of the
Turkic people in the Byzantine space as a subject of social history, the approach of
this dissertation is the discussion of the representation of the Turkic people in
Byzantine historiography.
The works of Anthony Kaldellis, on the other hand, should be cited as
examples of a new approach to Byzantine ethnographic literature.31 His methodology
is based on discourse analysis and the re-contextualization of texts according to the
circumstances in which they were written and their intended audience(s). This
approach has been a major source of inspiration for the development of my
dissertation’s methodology. Also, the unpublished Ph.D. thesis of Roman Shliakhtin,
From Huns into Persians: The Projected Identity of the Turks in the Byzantine
Rhetoric of Eleventh and Twelfth Centuries (2016),32 has recently put a new light on
the question of the identity of the Seljuk “other” in the early part of the period with
which my dissertation deals. Shliakhtin’s work, like this dissertation, utilizes the
theoretical framework offered by François Hartog. Shliakhtin includes, besides
historiographical narratives, poetry as well as rhetorical works in his corpus and
31 Kaldellis, Le discours ethnographique à Byzance: Continuité et rupture; Kaldellis, Romanland:
Ethnicity and Empire in Byzantium.
32 Shliakhtin, From Huns into Persians: The Projected Identity of the Turks in the Byzantine Rhetoric
of Eleventh and Twelfth Centuries.
30
presents a detailed and persuading analysis. Nevertheless, two shortcomings of his
work are its focus on a relatively short time frame and its lack of a comparative
perspective with the representations of other Turkic populations in the Byzantine
sources. This dissertation has some common themes with Shliakhtin’s work, such as
the lack of Turkish women’s visibility in Byzantine literature. On this point,
however, the conclusions of the two dissertations are different.
Andrew Peacock, Alexander Beihammer, and Dimitri Korobreinikov
represent a new generation of researchers focusing on late medieval Anatolia.
Peacock’s two books on the Great Seljuks are essential for understanding the early
Seljuk state and its policy toward the West. His examination of the autonomous role
of Türkmen groups in the Turkish conquest of Anatolia also illuminates a period for
which the sources are very scarce.33 Beihammer’s recent book on the formation of
Turco-Muslim Anatolia between the second half of the 11th and first quarter of the
12th century completes the studies of Peacock and gives us a new, more accurate
chronology of the early period of the Seljuk Sultanate of Rum and other Turkish
polities of Anatolia.34 Two earlier articles by Beihammer on the role of Komnenian
propaganda on ideological changes in the representation of the Turks in Byzantine
literature and on defection and apostasy between the Byzantines and Seljuks have
considerably improved our understanding of the nature of the relationship between
these two societies.35 Dimitri Korobreinikov’s book on 13th-century Byzantine-
33 Peacock, Early Seljuk History: A New Interpretation; Peacock, The Great Seljuk Empire; Peacock,
“From the Balkhān-Kūhīyān to the Nāwakīya: Nomadic Politics and the Foundations of Seljūq Rule in
Anatolia,” 55-80.
34 Beihammer, Byzantium and the Emergence of Muslim-Turkish Anatolia, ca. 1040-1130.
35 Beihammer, "Defection across the Border of Islam and Christianity: Apostasy and Cross-Cultural
Interaction in Byzantine-Seljuk Relations," 597-651 and Beihammer, “Orthodoxy and Religious
Antagonism in Byzantine Perceptions of the Seljuk Turks (Eleventh and Twelfth Centuries),” 15-36.
31
Seljuk relations also enabled us to understand better the actual situation of the
frontier region between these two powers.36
Gill Page’s study about the evolution of the self-image of the Byzantine
people is a work of utmost importance for the socio-cultural history of the period.37
Buket Kitapçı Bayrı’s work on the representation of warriors, martyrs, and dervishes
in the narratives of the Byzantine-Turkish frontier merits particular attention for its
original approach and meticulous treatment of the subject.38 The papers from the
Fourth International Sevgi Gönül Byzantine Studies Symposium, edited by Koray
Durak and Ivana Jevtić, focus on the notions of identity and the “other” in
Byzantium.39 The volume also includes a very useful introduction by the editors,
with references to earlier works on the subject.
Finally, there are several individual studies to cite on particular topics: two
articles dealing with the nomadic peoples by Elisabeth Malamut,40 three articles
about the Byzantine ethnic terms employed for the Turks – two of them by Alexios
Savvides41 and one by Koray Durak42 – and an article dealing with the 12th-century
Byzantine representation of Seljuk Turks by Aleksandar Jovanović.43 Furthermore,
some recent studies have appeared in Turkish that treat subjects covered by this
36 Korobeinikov, Byzantium and the Turks in the Thirteenth Century.
37 Page, Being Byzantine: Greek Identity Before the Ottomans.
38 Kitapçı Bayrı, Warriors, Martyrs, and Dervishes. Moving Frontiers, Shifting Identities in the Land
of Rome (13th-15th Centuries).
39 Durak & Jevtić (eds.), Identity and the Other in Byzantium: Papers from the Fourth International
Sevgi Gönül Byzantine Studies Symposium.
40 Malamut, “L'image byzantine des Petchénègues,” 105-114; Malamut, “Les peuples étrangers dans
l'idéologie impériale. Scythes et Occidentaux,” 119-132.
41 Savvides, “Some Notes on the Terms Agarenoi, İsmailitai and Sarakenoi in Byzantine Sources,” 89-
96 and Savvides, “Byzantines and the Oghuz (Ghuzz). Some Observations on the Nomenclature,”
147-155.
42 Durak, “Defining the Turk: Mechanisms of Establishing Contemporary Meaning in the Archaizing
Language of the Byzantines,” 65-78.
43 Jovanović, “Imagining the Communities of Others: The Case of the Seljuk Turks,” 239-273.
32
dissertation, including a book by Adem Tülüce44 and an article by Hasan Çolak.45
However, it is difficult to say that these studies in Turkish have made a new
contribution to the subject.
In concluding this review of the secondary literature, I would like to add that
my dissertation is a study of literary representation and follows the path of the abovementioned
works in the field of Byzantine studies. But there are also aspects of my
research that differ greatly from the studies I have mentioned. Firstly, my study aims
to interpret the representation of the Turkic peoples in a broader time frame: from the
11th to the 15th century. Thus, it explores the evolution of the Byzantine
representation of Turkic peoples in the longue durée. Apart from the work of
Shukurov, all the abovementioned studies focus primarily on the Anatolian Turks,
yet I also examine the Byzantine representation of the non-Anatolian Turkic peoples
who have so far occupied a marginal place in Byzantine studies. Finally, my aim has
been to explore the representation of the different Turkic populations in a
comparative manner in order to understand the evolution of the image of these
populations, in other words, their differentiation from each other in the course of the
centuries in the context of a greater Turkic world.
1.5 Methodology and key concepts
The modern historiography of the Byzantine Empire took shape under two strong
influences: the classical philological tradition dealing with Greek literature and the
positivist historiography of the 19th century. This background made the field called
“Byzantine studies” or “Byzantinistik” somewhat conservative and very faithful to
44 Tülüce, Bizans Tarih Yazımında Öteki - Selçuklu Kimliği.
45 Çolak, “Bizans Tarih Yazıcılığında Dönüşüm: Laonikos Chalkokondyles'te Bizanslı ve Osmanlı
İmajı (1299-1402),” 333-352.
33
empiricist and positivist models.46 However, my goal in this study is to put forward a
more interdisciplinary approach. As already mentioned, this dissertation focuses on
the formation and evolution of the image of Turkic peoples in Byzantine
historiography, so I basically offer an anthropological reading of the abovementioned
sources. Despite the fact that these sources were written as the narration of historical
events, and therefore have a historiographical character, they must also be considered
literary texts. Moreover, in Byzantine literature, the differences between the genres
are fluid. Thus, as I already noted, there are even important differences among texts
of the same genre: Anna Komnene’s Alexiad and John VI Kantakouzenos’ History
are essentially memoirs. John Skylitzes’ Synopsis of History is a compilation of
various historical works, while Gregoras’ History has an annalistic character.
In my opinion, the essential idea of structuralism can help us in our context.
There is a binary antagonism between Byzantine and Persian, sedentary and nomad,
and Christian and non-Christian in our texts. In the case of Byzantine literature, these
categories are mutually exclusive. François Hartog, who studied the representation of
the other in the Histories of Herodotus, particularly focusing on the representation of
the Skythians, created a new approach to the study of ancient literature. He
questioned the objectivity and truthfulness in Herodotus’ representation of the
Skythians and defined his study as a study of “Herodotus’ Skythians.” Thus,
according to him, “we may read the text (Herodotus’ Histories) with the assumption
that this or that Skythian practice may be interpreted concerning its correspondent in
the Greek world. When Herodotus speaks of sacrifice among the Skythians, he sets
up an implicit opposition with Greek sacrifice[...]”47 However, my inspiration from
46 For a general discussion of methodology in the field of Byzantine studies, see Haldon, “‘Jargon’ vs.
‘the Facts’? Byzantine History-Writing and Contemporary Debates,” 95-132, particularly 109-122.
47 Hartog, Le miroire d’Hérodote, 28.
34
Hartog’s work is only partial in the context of this dissertation. Hartog was dealing
with only one author, Herodotus, but I am dealing with a range of authors whose
works spanned four centuries. Moreover, the relationship between the practices
attributed to barbarians and Byzantine realities is only one dimension in this
dissertation. However, I think that this binary opposition is a general leitmotiv in the
works of these Byzantine authors. Thus, Hartog’s approach, which has its roots in
structuralism, especially in the anthropology of Claude Lévi-Strauss, is a useful
instrument for my study. Furthermore, the linguistic approach of Ferdinand de
Saussure has also inspired me.48
Encounter and alterity are two of the key notions of this dissertation:
Encounter means the first act of confrontation between two populations, in this case,
the Byzantines and Turkic peoples. However, this encounter is not the first encounter
and the image of the Turkic peoples was not drawn on a tabula rasa. The Byzantines
had recourse to the centuries-old logos of the Skythians and Persians. Thus, the
Byzantine description of the Turkic other does not always represent the actual
encounter, but it is mostly a consequence of the constant evolution of an image.
Nevertheless, there are still very lively depictions of the actual Turks, located in the
literary tradition in the longue durée. To understand it, one must take a look into the
author’s life and question whether he experienced a real encounter with the
individuals belonging to Turkic populations or whether his text was based on another
written or oral historiographical material. Secondly, even an actual encounter with
the Turkic other does not necessarily suggest that the descriptions are accurate. It is
48 Structuralist methods have been used in the field of Byzantine studies since the 1970s. An important
scholar who introduced this methodology to the field was Évelyne Patlagean. See her works,
particularly Patlagean, Pauvreté économique et pauvreté sociale à Byzance, and Patlagean, Un Moyen
Âge grec: Byzance, IXe - XVe siècle, particularly 83-162.
35
well known that in Byzantine literature even the events eyewitnessed by the authors
were narrated by utilizing ancient texts as models, such as in the case of the narration
of the black death by John Kantakouzenos.
On the other hand, the notion of alterity (otherness) represents the essential
differentiation between the self and the other. Saussure explained the notions of
difference and opposition in language as a situation which has a negative character in
itself. According to him, if a is different from b, this essentially means that an a is
not a b, regardless of the degree of non-coincidence; but as soon as a relationship
exists elsewhere between a and b, they are now part of the same system, and their
difference becomes the opposition.49 In our case the foreign peoples, regardless of
their ethnicity, were alien to Byzantine society. Hence, they are others (ἄλλοι).50 The
Turkic peoples, who were perceived as the steppe nomads at first, were of course
foreigners, but only a subgroup of foreigners. This subgroup had the features that not
only distinguished them from the Byzantines but also from other non-Byzantines.
Moreover, the Byzantines already had an idea of a steppe nomad because of their
familiarity with ancient and late antique literature, and they could easily substitute
the old logos of the Skythians with the recently appeared Turkic populations. So the
formation of the Turkish image in Byzantine literature was shaped by the
accumulation of various strata of cultural memories.
Finally, what I have to explain is the progressive formation of the image of
Turkic peoples during the four centuries from the 11th to the 15th century. There is a
continuing process of formation resulting from the representation of such an image. I
underline the notion of representation which I consider also very crucial because, as I
49 His essential work is Saussure, Cours de linguistique générale.
50 Malamut, “Les peuple étrangers dans l’idéologie impériale,” 120.
36
already explained, the difference between “what actually happened” and fiction was
particularly fluid and ambiguous in Byzantine literature. Thus, in every text dealing
with the Turkic peoples, one must understand the context and the writer’s aim.
1.6 Ideological geography and frontiers
The Byzantines considered their empire and the world around it through an
ideological lens. Their empire was the center of the oikoumene, the civilized world, if
not constituting the civilized world itself. The frontiers were the key space to
understand this mentality because they constituted the lines that separated the world
of Romans and Barbarians. After the invasions of Muslim Arabs in the 7th century,
the Byzantine boundaries were limited to the Northern Balkans, Crimea, Eastern
Anatolia, and Taurus mountains. These territories were essentially frontier regions.
In this dissertation, the aim is to deal with two frontiers where the Byzantines
encountered the Turkic peoples: the eastern and northern frontiers. The eastern
border is where the Byzantines confronted the eastern enemies; the Persians, Arabs,
and ultimately Turks. However, they confronted mostly the “Skythian” peoples in the
northern frontier. In other words, Turkic peoples have a special position in Byzantine
history; they are the only ethnolinguistic group to have been neighbors in two
different regions on both ends of the empire at the same time as the Byzantines.
However, there was an essential difference between these frontiers: in the east, the
border has an ideological meaning, firstly, between the Romans and their archenemy,
the Persians, and then between the Christian Romans/Byzantines and Muslim Arabs.
Both of the borders underwent changes through the centuries. In the north,
the river Danube (Istros) constitutes the essential frontier between the Byzantine and
steppe worlds. This river was surpassed frequently by the nomadic populations that
37
ravaged the Balkan provinces of Byzantium, such as Avars, Bulgars, and Pechenegs.
However, the river maintained its symbolic and ideological function as a frontier.
Until the Late Middle Ages, it was also a frontier between Christianity and paganism.
However, just like any other border during the Middle Ages, these borders were
often oversimplifying. The entire frontier areas should be seen as a continuous space
between these two parts of the world.
Just like the ideological meaning of the eastern frontier, the Danubian frontier
had a “lesser” ideological meaning, particularly under the Macedonian dynasty. In
parallel with the military expansion in that period, the Byzantine rhetoricians, such as
John Geometres, pointed out the ideological significance of the empire’s natural
borders and underlined the importance of the “mighty Ister.”51 Furthermore, it seems
that the southern shores of the Crimean Peninsula, which constituted another border
between the Byzantine Empire and peoples of the steppe, had still less ideological
importance.
On the other hand, the eastern border, according to Hélène Ahrweiler, “has
always separated constituted worlds, carrying different messages, each aspiring to
impose its will on the other one, while the other frontiers of the empire have been
erected facing the barbarians, facing the gentes that have access to political forms
and become aware of their personality by and against Byzantine action.”52 Thus, the
empire was confronted with an entity of equal ideological weight on the eastern
front. The meaning and evolution of this ideological weight will be discussed in the
later chapters of this dissertation. However, the northern frontier was not totally
devoid of an ideological meaning, but this meaning was far simpler: it was a frontier
51 Stephenson, Byzantium's Balkan Frontier: A Political Study of the Northern Balkans, 82.
52 Ahrweiler, “La frontière et les frontières de Byzance en Orient,” 225.
38
between the civilized Roman oikoumene and the barbarian world of the Eurasian
steppes. Attaleiates’ complaint about the Danubian region’s becoming a hub of
mixobarbaroi was not groundless; this space became an area of interaction between
the Byzantines and the so-called “barbaric” populations who were mostly of Turkic
origin.
This difference is coherent with the main arguments of my dissertation. The
“Skythian” and “Persian” alterities were two different types of alterity, and the
Byzantine authors gave them different levels of ideological meaning. Thus, my
argument is that the antagonism between the Byzantines and Skythians, which forms
a secondary antagonism, is built mostly on a difference between the notions of
civilization and barbarity, sedentary life and nomadism. On the other hand, the
Byzantine-Persian antagonism is the reflection of a much more fundamental issue: It
was the continuity of an experience that made up a very strong influence on the
Greek intellectuals and then all the Western world, namely the Persian Wars
(499 BC- 449 BC). As it was explained by François Hartog, “Le Barbare, c’est avant
tout, plus que tous et pour longtemps le Perse.”53 The image of the Turk in Byzantine
literature is constructed antagonistically from nearly the beginning, as the Turks are
ultimately treated as enemies on the battlefield. However, the nature and degree of
this antagonism will enable us to understand the position of the Turks in Byzantine
historiography. Moreover, it should never be forgotten that this differentiation in the
antagonism could also be connected with the political-cultural evolution of the
Turkic polities. If the Islamic element overtook the aspects attributed to “Skythians”
in the ideology of these polities in the historical process, and, for example, if an
approach of religiously motivated war based on an anti-Christian antagonism gained
53 Hartog, Altérité, diversité, différence, 2.
39
prominence, this would lead to a change in the Byzantine perception of the Turks.
This dissertation leaves out of the focus the century-long debates about the role of
Islam on the Turks' expansion against the Byzantines, namely Paul Wittek’s “ghaza
thesis” and its opponents, but the question of the role of Islam in changing the
Turkish image in Byzantine historiography will be discussed in different contexts.54
54 For the “ghaza thesis” see Wittek, The Rise of The Ottoman Empire. For a discussion of the
historiographical trends about the early Ottomans see Kafadar, Between Two Worlds: The
Construction of the Ottoman State, 29-59.
40
CHAPTER 2
THE HISTORICAL BACKGROUND UP TO THE 11TH CENTURY
2.1 Early Arab expansion into Byzantine territory
In this chapter, I will give a historical background to better understand the context of
this study, in which the representation of the alterity and the formation of the Turkish
image will be discussed. This chapter is not written as a linear history of the
Byzantine Empire but rather as a history of the Byzantine frontiers from the 7th to the
11th century. The reason I put the concept of the border in the center is because I see
it as a place of encounter with the other. Thus, I wanted to include the history of the
border regions, the place where empirical knowledge of Turkic peoples is produced,
which I see as an element that feeds the literary topoi, into the context of the thesis.
The eastern frontier of the Byzantine Empire was the space of the military
confrontation between Orthodox Byzantines and Muslim Arabs from the mid-7th
century to the mid-11th century, a roughly 400 years period before the arrival of
Seljuks to this region. The Byzantine-Arab confrontation succeeded a long history of
opposition and antagonism between the Greeks and Romans and the Persians under
the Achaemenid, Parthian and Sassanid dynasties. Since the times of Herodotus, the
archetype of the Persian was the embodiment of the eastern barbarian for the Greeks.
Sassanid Iran was dramatically weakened after the long war against the
Byzantines (602-628) and it was destroyed by the unexpected Arab invasion after the
birth of Islam. In the middle of the 7th century, nearly all Persia was annexed by the
Muslim invaders, excluding several areas in the northern and eastern edges of the
Sassanid Empire that continued to resist them still for a while.
41
The penetration of Arab armies into Byzantine soil started in 634 with the
capture of the city of Bostra in the southern margins of the Syrian Desert. In half a
century, all Syria, Palestine, Egypt and North Africa fell into the hands of the
Arabs.55 These territories were Arabized and Islamized in the course of centuries and
lost the cultural basis of their former allegiance to the Roman/Byzantine Empire.
As early as the last quarter of the 7th century, the Muslim raids started to
affect Anatolia. Arabic traditions regarding the earliest memories concerning the
Arabo-Byzantine wars and the takeover of the cities of Rum are collected in al-
Balād̲ h̲ urī’s compilation “Futuh al-Buldan”(The conquests of the realms).56
In the early 8th century, there appeared the first signs of the formation of a
permanent frontier between the Arabs and Byzantines. This frontier passed roughly
through the Taurus Mountains. This range of mountains was the natural border
between Anatolia and Northern Syria since Antiquity, and it could be protected
easily by the fortifications and border guards.57
The Umayyads, the first dynasty of the Muslim Caliphate, were centered in
Syria and according to H. A. R. Gibb, the great historian of medieval Muslim
civilisation, represented “in several respects […] a succession-state to the East
Roman Empire, notwithstanding the ideological oppositions involved in the sphere of
religion.”58 He further argues,
The Umayyad Caliphate however in its attitude to the Empire, was much
more than a provincial succession-state. The two facets of its policy, the
military assault and the administrative adaptation, point clearly to the real
ambition of the first-century Caliphs, which was nothing less than to establish
55 Standard modern study about the early Muslim conquests and its impact on the Byzantine Empire is
Kaegi, Byzantium and the Early Islamic Conquests.
56 Al-Baladhuri, The origins of the Islamic state: being a translation from the Arabic, accompanied
with annotations, geographic and historic notes of the Kitâb futûḥ al-buldân of al-Imâm Abu-l ʻAbbâs,
Aḥmad ibn-Jâbir al-Balâdhur.
57 Lilie, “The Byzantine-Arab Borderland from the 7th to the 9th Century,” 14.
58 Gibb, “Arab-Byzantine Relations under the Umayyad Caliphate,” 232. For a comprehensive work
about various features of the Umayyad Caliphate see Marsham (ed.), The Umayyad World.
42
their own imperial dynasty in Constantinople. Seen in this light, their
administrative imitations and adaptations take on a different character; they
are not merely the tribute paid by raw and parvenu princes to the
achievements of their predecessors, but an almost deliberate effort to learn
the ropes and fit themselves to assume the imperial destiny.59
Gibb also argues that the unsuccessful siege of Constantinople in 718 led the Arabs
to a new cultural and ideological stand to adopt more and more the Persian traditions.
Starting with the caliph His̲ h̲ ām (r. 724-743), the Caliphate designed itself as a
successor to the Sassanid Empire.
In that very period, the Byzantine Empire was faced with the Iconoclast
crisis; the first wave of Iconoclasm lasted from 726 to 787. Despite the climate of
unrest in the Empire, the emperors of the Isaurian dynasty managed to organize a
remarkable resistance on Anatolian soil against the invaders. To discuss the
theological dimensions of Iconoclasm is far beyond the limits of this dissertation, but
it is useful to remember Ahrweiler’s remarks on the political evolution of the
Byzantine Empire under the iconoclast emperors. According to her, this period is
important for the phenomenon of the militarization of Byzantine society, which
could be traced back to the earlier origins of important Byzantine aristocratic
families as the military leaders of this period. Secondly, it was an epoch of the
formation of some kind of Byzantine nationalism, different from the imperial
universalism and expansionist ambitions of earlier emperors; it was embodied in
forming the Byzantine army as a people’s army. The soldiers from modest origins
were fighting not for the ancient ideals now, but for their country and their faith.
Moreover, this new national sentiment was clearly identified with Christianity and
the will to defend the homeland against the unbeliever enemies.60
59 Gibb, “Arab-Byzantine Relations under the Umayyad Caliphate,” 232.
60 Ahrweiler, L’idéologie politique de l’Empire byzantine, 30-35.
43
In 750, the Umayyad dynasty was destroyed by a large coalition of dissidents
including the Shi’ite factions led by the charismatic Khorasanian warrior Abū
Muslim and, as the result, a new caliphal dynasty emerged: the Abbasids. The
Abbasid Caliphs continued the tendency of later Umayyad Caliphs to bring the
center of the Caliphate from the Levant to Mesopotamia and they adopted Persian
imperial customs. The early Abbasid period is marked also with a renewal of jihad
against the Byzantine territories. Almost every year there were Abbasid raids through
the Byzantine territory; these attacks devastated the urban network and agrarian
production. At the end of the 8th century, Empress Irene was obliged to sign a peace
treaty with the Abbasid Caliph Hārūn al-Ras̲ h̲ īd that forced the Byzantine Empire to
pay a considerable amount of money two times a year.61
The caliphs Hārūn al-Ras̲ h̲ īd (r. 786-809) and his sons and successors, al-
Maʾmūn (r.813-833) and al-Muʿtaṣim (r.833-842), personally led their armies into
the Byzantine territories of Anatolia. Particularly the 830s was a decade of
continuous war between the Arabs and Byzantines, in eastern and central Anatolia.
In this decade Anatolia saw many dramatic events which were further remembered in
the religious traditions, such as the “Forty-two Martyrs of Amorion,” which was the
execution of 42 people belonging to the religious/military élite of the empire who
were executed after the orders of Caliph al-Muʿtaṣim.62
The mid-9th century could be considered the decisive end of Arab
predominance in the Byzantine-Muslim conflict.63 In the two subsequent centuries,
the Byzantines were on the attack and the Arabs on the defence. In the most difficult
61 Vasiliev, “Byzantium and Islam,” 311.
62 For this event see Kolia-Dermitzaki, “The Execution of the Forty-Two Martyrs of Amorion:
Proposing an Interpretation,” 141-162.
63 Traditionally, the destruction of the invading army of the amir ʿAmr of Melitene in 864 is
considered the turning point of Byzantine- Arab wars. See Shepard, “Constantine VII, Caucasian
Openings and the Road to Aleppo,” 19.
44
periods of the Arab invasion, the Byzantines managed to save their core lands in Asia
Minor, but they lost their other Mediterranean provinces to different Arab emirates:
Cyprus was already fallen in 650, but later became an Arabic-Byzantine
condominium, Crete fell in 824 (to be reconquered by Nikephoros Phokas in 961),
and the cities of Sicily started to fall in 831.
These campaigns left a vivid ideological impact both on Byzantine and
Muslim sides. The persons, places, and events identified with the early Arabo-
Byzantine wars continued to live both in the literature and peoples’ collective
memory. Therefore, this early period (c. mid-7th- c. mid-9th century) could be
considered as the formative period of both the ethos and antagonism of the
Byzantine-Muslim border.
2.2 The formation of a permanent frontier between the Byzantines and Arabs
The Umayyads organized their northern border regions with the Byzantines under
two distinct administrative units: awasim and thugur. The term awasim is the plural
of the word al-asima, which means the protectress. This region of “the protectresses”
covered Northern Syria, plus Antioch and Kilikia. Its military center was Kinnasrin.
Awasim continued to exist as a region of internal frontier alongside the thugur that
constitutes the outer or real frontier where the Byzantines and Muslims confronted
each other periodically.64
Thughur’s meaning in Arabic is “gap, breach, opening”. Although awasim
was a merely administrative term, thugur was a more ideological term and it was
used for all the frontier zones between dar al-Islam and dar al-harb; several thagrs65
64 Canard, “al-ʿAwāṣim,” 761.
65 It is the singular form of thugur.
45
are mentioned against the Oghuz, Georgians, Alans, Nubians, and Franks (in Spain)
around the Muslim world. While the awasim covers the cities of Northern Syria,
thugur lies from the Taurus Mountains (on the westernmost edge, river Lamos
constituted a natural border between the caliphal and imperial territories) up to the
Northern Mesopotamia. The Arab geographers like al-Istakhri and Ibn Hawkal count
Tarsus, Sis, Adana, Germanikeia (Maraş), Samosata, Melitene (Malatya), and
Kamacha (Kemah) among the important cities of the thugur region.66
The Byzantine key concept of the border was the notion of kleisourarchia.
The word κλεισούρα means “defile,” and these kleisouras were small, fortified
boundary districts that existed since at least the 6th century. As the name suggests,
these fortifications are founded mostly on the mountain passes on the invasion
routes. Perhaps strategically most important of them was the kleisoura at the Kilikian
Gates.67 However, there were other kleisouras, such as Seleukeia, Sebasteia, and
Koloneia in the eastern borders. Many of these districts evolved into little themes
(μικρὰ θέματα) in the mid-10th century. This was the umbrella term for small,
Armenian populated themes in the east.68
There is a paradoxical position of religion and religiosity in the border
regions where both of the imperial states designed themselves as the defenders of
their faiths, Orthodox Christianity and Islam, respectively. Thus, the border was also
a grosso modo religious border, and religion had central importance. Sometimes
66 Bosworth and Latham, “al-T̲h̲ug̲ h̲ ūr,” 446-447.
67 Honigmann, Bizans İmparatorluğunun Doğu Sınırı, 40-41; Haldon and Kennedy, “The Arab-
Byzantine Frontier in the Eighth and Ninth Centuries: Military Organisation and Society in the
Borderlands,” 83-87. Furthermore, the Kilikian frontier was not only a military region but also an
important route for commercial and diplomatic purposes, as demonstrated by Durak, “Traffic across
the Cilician Frontier: Movement of People between Byzantium and the Islamic Near East in the Early
Middle Ages,” 141-154.
68 Ahrweiler, “La frontière et les frontières de Byzance en Orient,” 217-218.
46
there were religious persecutions, but there also existed a mixed and culturally
heterogeneous society that is echoed in the literary sources.69
The borderlands were ethnically and religiously heterogeneous regions.
Although the Muslim cities in the borderlands appear as Arab garrison cities, the
social and demographic realities of these cities are further complicated. An Arab
garrison city was not an Arab colony but rather a surprising anagram of different
elements, such as a high population of slave-soldiers and the volunteers, known as
muttatawwis,70 who immigrated to thugur from the far-away provinces of the
Caliphate to involve in jihad against the Byzantines. A striking element of the thugur
cities were the ribats. The ribats were the guesthouses where the volunteers from
different countries could live during brief periods, and there could be involved in the
ascetic-mystical ways of the religious warriors. Thus, the cities of the thugur were
not only military but also religious centers. There is also evidence about the
circulation and settlement of religious scholars and preachers in the region, thus it
could be speculated that these men were coming to thugur to preach a more militant
form of Islam that stressed the importance of jihad against the unbelievers. However,
at the same time, the existence and the activity of unorthodox sects of Islam echoed
the complaints of the religious scholars.71 Thus, the frontier cities were not free from
socio-religious tensions of the core parts of the respective countries.
The formation of the permanent frontier region also created the main pillars
of an ideological geography, many cities that bear a particular symbolism in the eyes
69 Lilie, “The Byzantine-Arab Borderland from the 7th to the 9th Century,” 20.
70 According to Deborah Tor, in the mid-8th century, large-scale Muslim expeditions into Byzantine
territory were ceased. The jihad was privatized, so smaller-scale raids by independent volunteer
groups became frequent events in the border regions. Tor, “Privatized Jihad and Public Order in the
Pre-Seljuq Period: The Role of Mutatawwi’a,” 558.
71 Brown, “Christians, Muslims and Heretics: Religion and the Arab Byzantine Frontier c. 750-934,”
95.
47
of Arabs or Byzantines. These cities identified with the saints, religious figures,
legendary warriors or rulers created the spatial representation of the frontier ethos.
Tarsus was perhaps the most famous of the frontier cities. As the city of St.
Paul, it had a very particular religious significance in the eyes of the Christians.
Tarsus fell into the Muslim sphere of influence in the 7th century and then
experienced a tumultuous history of raids, takeovers, and lootings. In the early
Abbasid period, it was an important military center, in which a significant number of
Khurasani troops were stationed. With its ribats and volunteers, it represented the
microcosmos of the thugur. When he died during a campaign in Kappadokia, Caliph
al-Maʾmūn’s body was also buried in this city. The city decisively passed to the
Byzantines during the Kilikian campaign of Nikephoros Phokas in 965.72
Melitene was another city identified with the memories of the Byzantine-
Muslim wars. As already stated, it was the center of a tiny Muslim emirate involved
in extensive raid campaigns against Byzantine Kappadokia. It was conquered by
John Kourkuas in 934. In the subsequent centuries, this city became the center of a
hero cult around the legendary figure of Seyyid Battal Gazi. The Arabs and Turks
wrote chivalric romances about this figure: Sirat Delhemma and Battalname,
respectively. The fictional personality of Seyyid Battal Gazi is based on the memory
of ʿAbdullāh al-Baṭṭāl, a general (bearing the title “al-Antaki,” “the Antiochian,” so a
man from the awasim region) serving the dynasty of Umayyads who was killed in
battle in 740 at Akroinon.73 This obscure historical figure’s deeds are enriched with
many details of 9th-century Byzantine-Arab wars involving the Emirate of Melitene.
72 Eger, The Spaces Between The Teeth: A Gazetteer of Towns on the Islamic-Byzantine Frontier, 183-
190; Bosworth, “The City of Tarsus and the Arab-Byzantine Frontiers in Early and Middle Abbasid
Times,”268-286.
73 Grégoire, “Comment Sayyid Battal, martyr musulman du VIIIe siècle, est-il devenu, dans la
légende, le contemporain d’Amer (†863)?” 571-575.
48
Henri Grégoire does not hesitate to call this epic cycle the “Geste de Meliténé”
because of its particular affiliation with this frontier city.
A town near Melitene, called Sozopetra by the Greeks and Zibatra by the
Arabs, was another symbolic space between the Arabs and Byzantines.74 Skylitzes
states that Sozopetra “was the homeland of the Caliph” (πατρίδα τυγχάνουσαν τοῦ
ἁμερμουμνῆ), in a point of his narrative where the Caliph is al-Mu’tasim. However,
there is no evidence of such a relationship between that city and the Caliph, and this
detail seems like an ideological invention to balance the ideological impact of the
sack of Amorion; indeed, emperor Theophilos (r. 829-842) conquered this city in 837
in a campaign presented as the revenge of Amorion. Zibatra’s fortifications were
destroyed and reconstructed at least four times between the years 742 and 872 by the
Arabs and Byzantines who took and retook the city without a permanent control.75 In
872, the city was permanently annexed by the Byzantine Empire under Basil I.76
The religious groups that were seen as heretical were the ever-present
elements of borderlands: Already, Syriac, Coptic or Armenian speaking Christians
who remained in the territories that the Empire lost to the Caliphate in the 7th century
were members of the non-Chalcedonian Churches that were officially seen as
heretical by the Patriarchate. Moreover, there existed other religious movements in
the border regions, such as the Paulicians that appeared in the 8th century, a Christian
sect that was considered Manichean and, therefore, heretical by the Byzantines; it
was centered in the city of Tephrike in the frontier region. Paul Lemerle defines
Paulicianism as the “frontier religion” that belongs to the Syro-Armenian
74 Today the subdistrict Doğanşehir in the province of Malatya.
75 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 74; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057,75.
76 For a general overview of the history of Melitene and Sozopetra, see Eger, The Spaces Between The
Teeth: A Gazetteer of Towns on the Islamic-Byzantine Frontier, 118-123, 197-198.
49
borderlands and has some non-negligible ethnic aspects.77 Many leaders of the
Paulicians were Armenians. The Paulicians were permanently allied with the Arabs
and played the role of the fifth column in the Arab invasions of Anatolia. They were
annihilated by the armies of Basil I that retook Tephrike in 878 and this victory gave
the emperor a long-needed image of the defender of Orthodoxy. The memories of the
war against the Paulicians are perhaps present in the epic of Digenis Akritas.
The Khurramites, on the other side, could be considered a religious
movement that played a similar role in the Caliphate. How the Paulicians were a sect
based on Armenians, the Khurramites were basically a Persian religious movement.
This movement was centered on western Iran and Azerbaijan; their founder Babek
(d. 838) launched a rebellion against the Abbasids. However, after their defeat at the
hands of the caliphal army in 833 and Babek’s execution, one of their leaders, Nasr
(d. 842), went to Constantinople with his Persian followers. There he converted to
Christianity and married the sister-in-law of Emperor Theophilos. His new name
Theophobos demonstrates his alliance with the Byzantine emperor. The followers of
Theophobos formed a Persian division serving in the Byzantine army. They served
the Byzantines in the eastern campaigns of Theophilos until the rebellion they
committed in Sinope in 838; then, they fell from imperial grace, and their leader was
imprisoned and killed. Skylitzes refers to Theophobos as a leader who comes from
the royal bloodline of Persia.78 Just like the representation of the Paulicians in the
Epic of Digenis Akritas, Babek appears as one of the villains in the Turkish chivalric
novel Battalname. In that romance, Babek was a heresiarch and became the Emperor
77 Lemerle,“L’histoire des Pauliciens d’Asie Mineure d’après les sources grecques,” 134.
78 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 66; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 68.
50
of Byzantium!79 The reminiscence of these groups in the epic literature shows that
these popular-religious movements left a deep impact on the collective memory of
both Christian and Muslim populations. The Khurramite and Paulician
movements/rebellions shared a common feature that can help us to understand the
realities of the Arabo-Byzantine border. These populations from different ethnicities
who followed a different interpretation of their state’s main religions were living in
the margins of their states and were the most appropriate candidates for collaboration
with the enemies.
Besides these populations, there were also several less populous communities
that played a certain role in the demographic formation of the region. The first of
them is the Mardaites or Djaradjima, as the Arabs called them. They were Christian,
either Monophysite or Monothelite, and had special relations with the Patriarchate of
Antioch. The Mardaites seem to be talented in irregular warfare; they collaborated
with both sides and enjoyed a high degree of autonomy. Their clothing was similar to
the Muslims, and they were exempted from the jizya and despite their Christian creed
they had the right to get a share of the booty in the wars they were involved with the
caliphal army. However, their situation worsened during the reign of the zealot
Caliph al-Mutawakkil (r. 847-861). Apart from the ones living in Syria, there was a
Mardaite diaspora in the Byzantine Empire, particularly in Pamphylia.80
Another enigmatic population of the borderlands was an ethnic group named
al-Zutt by the Arabs. They appear as a community from Northern India, perhaps
related to Jats in contemporary India. They seem to be converted to Islam and
79 Battalname, ed. Yorgos Dedes, 592.
80 Canard, “Ḏj̲
arād̲j̲
ima,” 456-458. Honigmann identifies them with the Maronites (of Lebanon).
51
transferred from their ancestral lands to several places, including Kilikia and the
Amik plain.81
The Armenians themselves were a population that preserved a degree of
autonomy by following a balance policy between the Byzantines and Arabs. They
were both populous in thugur and Armenia proper, which was organized as a
province of the Caliphate. This province was administered by Arab governors called
ostikans by the Armenians. In 885, the prince of princes, Ashot Bagrationi, was
crowned as the King of Armenia with the permit of Caliph al-Muʿtamid.82 For the
Abbasids, the creation of an independent Armenia would serve as a buffer state
between the Arabs and the Byzantine Empire. However, now an independent
kingdom that was situated in their ancestral lands, the Armenians continued to be an
important ethnic element in Byzantium. Bagratid Armenia, on the other hand, over
the years, disintegrated into small principalities controlled by several local dynasties.
When the Byzantine Empire had completely established its dominance in the areas
surrounding the Armenian highlands in the 11th century, they started to annex the
tiny Armenian principalities of Kars, Ani, and Vaspurakan. On the other part, in
upper Mesopotamia, there were many Muslim emirates governed by Arab or Kurdish
emirs, namely Bitlis, Bergiri, Manzikert, and the Marwanids of Amida and
Mayyafarikin. This annexation policy weakened the Byzantine defenses in the
eastern frontiers of the Empire and left Asia Minor defenseless against the Seljuk
invasion.83
81 Bosworth, “al-Zuṭṭ,” 574-575.
82 Grousset, Histoire de l’Armenie, 394-395.
83 For a discussion of the impact of the annexation of the Armenian principalities on the defense of the
eastern and southern borders of the Byzantine Empire, see Peacock, Early Seljuk History: A New
Interpretation, 129-139.
52
The demographic evolution of the frontier region is highly marked by many
state-sponsored transfers and deportations of population. Various populations
according to their possible allegiances with the imperial regimes were transferred to
the inner parts of the states. For example, there was a continuous westward
immigration movement of Christians of the Orient, regardless of their ethnic identity.
The Byzantine Empire tried to repopulate the recently conquered cities with
Christians from the other provinces of the Empire. Needless to say, just like the
dhimmis in Muslim countries, there were small Muslim communities that continued
to live under the imperial administration. For example, after the conquest of Tarsus
by Nikephoros II Phokas, the emperor allowed all Muslims to leave the city, but for
the ones who did not want to leave, apart from conversion to Christianity, there was
also the option of paying the poll-tax (similar to jizya) in order to remain within
Byzantine territories.84
2.3 The Byzantine offensive
By the late 9th and early 10th century, it could be clearly seen that the Byzantine
Empire regained its strong position in military matters. This revival is generally
identified with the Macedonian dynasty that ruled the empire from 867 to 1056.
Mid-9th-century is the beginning of the disintegration of the Abbasid Caliphate and
the rise of various local dynasties, particularly in Africa, Persia, and Transoxiana:
Tulunids, a ghulam dynasty of Turkic origin, now became de-facto independent
rulers of Egypt, Yaʿḳūb b. al-Layt̲h̲ founded the independent Saffarid Emirate in
Persia, Aghlabids established themselves as an independent emirate in Tunisia. In the
10th century, the Sunni Muslim world entered into a deeper ideological crisis, which
84 Bosworth, “The City of Tarsus,” 279.
53
is marked by the rise of Isma’ili Sh’ia and the formation of the Fatimid Caliphate, the
peasant and slave revolts, and the gradual disintegration of the Abbasid Caliphate
and several Muslim territories’ reconquest by the Byzantine Empire. Patricia Crone
summarizes the symptoms of this crisis as “Contemporaries lamented the
enfeeblement and disappearance of Islam…The Triumph of the Byzantines over the
Muslims, the disruption of pilgrimage, the absence of holy war, the unsafety and
disruption of the roads, and the establishment of independent power by every leader.
Prognostications, such as that a man will come and restore the domination of
Zoroastrianism … and put an end to the power of the Arabs’ were rife.”85
This crisis coincided with a period of Byzantine revival under the
Macedonian dynasty. The reign of Basil I (r. 867-886), who was nicknamed “the
Macedonian” as the eponymous founder of the dynasty, was the humble beginning of
this regeneration of the imperial power. In his campaigns against the Paulicians, he
managed to re-establish imperial rule in the eastern portion of the theme of
Charsianon that, for three decades, had become the center of a de-facto Paulician
state. It seems that their leader Chrysoheir is the model of the personality of
Chrysoberges in the epic of Digenis Akritas. During Basil I’s reign also Armenia
became independent, so Arab military presence diminished in the frontier regions of
the Empire.
However, the two emperors identified most with this aggressive military
policy and expansionism in the east were Nikephoros II Phokas (r. 963- 969) and
John I Tzimiskes (r. 969-976). These warrior emperors drastically changed the
geopolitical situation in Armenia and Northern Syria. Nikephoros, who was already a
famous commander before his coronation as the emperor, having the public image of
85 Crone, “The Rise of Islam in the World,” 29.
54
the “warrior monk,” conquered Crete in 960-961 and launched a series of attacks
against the Hamdanids in Kilikia and Northern Syria. After becoming the emperor in
963, he restarted his Kilikian expedition, which resulted in the conquest of Tarsus
and Mopsuetia. In 969, when he was in Constantinople, his armies led by Michael
Bourtzes took Antioch and turned the Hamdanid Emirate of Aleppo into a vassal of
the Empire. He was murdered after a palace conspiracy. Although during his reign
Nikephoros II was “hated and abominated by everybody”86 because of his
maltreatment of his subjects for extra taxes and his permanent pressure on religious
institutions for the same cause, he would be remembered as one of the most
successful military commanders of the Empire.
John I Tzimiskes, who was the relative of the great commander John
Kourkouas, after organizing the murder of Nikephoros II, personally led two
campaigns against the Arabs. The second campaign’s ultimate goal was to retake
Jerusalem from the Arabs and re-establish Byzantine rule in the Holy Land, but he
could not reach his goal. However, he could impose tribute on the city of Damascus
and obtained a real hegemony over the Muslim tribes of Syria. After his unexpected
death, his energetic policy was continued by one of his generals, Michael Bourtzes.
Yet the rebellions in the late 10th century, committed by Bardas Skleros and Bardas
Phokas, gradually stopped the Byzantine military activities in Syria.87
Basil II (r. 976-1025), who was known as the Bulgarslayer (Βουλγαροκτόνος)
for his vicious campaigns in the Balkans, was the last of these emperors that pushed
the Byzantine-Muslim frontier southwards and eastwards for a century. Although he
86 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum 278-279, Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 262-263.
87 Garrood, “The Illusion of Continuity: Nikephoros Phokas, John Tzimiskes and the Eastern Border,”
27-28.
55
turned his attention to the Balkans, unlike his predecessors, he had occasional
political moves in the east. Basil II’s eastern policy was based on the diplomatic
annexations of the small independent principalities extending on the eastern frontier
of Byzantium.
As it was remarked, Armenia was no more a united realm but an amalgam of
tiny principalities. Among the important ones there were the Kingdom of
Vaspurakan that was the hereditary realm of the House of Ardzruni, and the
Kingdoms of Kars and Ani which belonged to the different branches of the Bagratid
royal family.88 The Byzantines annexed these tiny principalities without involving
any military conflict but using diplomatic measures. They first annexed Taik (Tao),
whose ruler, David the Kuropalates, declared Emperor Basil II as his heir in 1000.
Then they annexed the Kingdom of Vaspurakan in 1022 and the Kingdom of Ani in
1045. Finally, after the annexation of the Principalities of Bgni and Kars, the
Byzantine Empire became the possessor of the near totality of all Armenian lands.
All the Armenian rulers, namely Senacherim of Vaspurakan, Gagik of Kars, Gregory
Pahlavuni of Bgni and their heirs, had new estates in inner parts of the Empire, in
Kappadokia, Charsianon, Kilikia and Mesopotamia. These rulers did not leave their
ancestral fiefs unaccompanied, yet they were followed by thousands of people that
were their subjects on their way to new lands. This immigration caused new
secessionist movements in these provinces, especially in Kilikia, in the years that
followed the Byzantine defeat at Manzikert (1071).89
88 For medieval Armenia, the work of René Grousset is still of utmost importance. His survey,
Histoire de l'Arménie des origines à 1071), encompasses the period from Antiquity until the Seljuk
invasion. See ibid, 483-484 for the internal divisions of the Armenian territories.
89 Charanis, The Armenians in the Byzantine Empire, 48-53.
56
2.4 The Byzantine confrontation with Turkic peoples
The first Turkic people that encountered the Romans were probably the Huns. It is
likely that the ruling élite of the Huns was Turkic speaking, while the rest of this
tribal confederation was made up of Slavic, Germanic, and Iranian peoples of
Western Eurasia. After the dissolution of the Hunnic Federation, there remained
certain tribal groups in the Pontic steppes: Saragurs, Akatzirs, Sabirs, Onogurs,
Utigurs, Kutrigurs, and finally the Bulgars. These populations probably spoke a
Turkic language, which was not Common Old Turkic but may have been a language
belonging to the subgroup known as Oghuric, whose only known successor today is
Chuvash.90
The very limited knowledge about these peoples usually comes from
Byzantine sources. In Agathias’ Histories and Jordanes’ Getica, there are brief
passages that list and locate these peoples. However, a part of the Bulgars, these
populations left very little trace in Byzantine historiography.91 As stated
melancholically by Agathias, “[they] were well-known right upon the time of
Emperor Leo and were considered a force to be reckoned with, but whom we in our
day and age neither know nor, I imagine, are likely to, since they have either
perished or migrated to the ends of the earth.”92 Both of the authors underline that
these populations are originally Huns or somewhat connected with the Hunnic
Confederation.
In the mid-6th century, in today’s Mongolia and Altai region, there formed a
new Turkic state called Göktürk, or Celestial Turks. I shall follow the naming
convention in the seminal work of Peter B. Golden and call them the Türk
90 Golden, An Introduction to the History of the Turkic Peoples, 97.
91 Golden, An Introduction to the History of the Turkic Peoples, 98.
92 Agathias, The Histories, 146.
57
Khaganate.93 This entity substituted the nomadic confederation of Rouran-Rourans,
and expanded itself through the west and had a military presence both in Central
Asia and western Eurasian steppes, including Crimea. The Türk Khaganate also had
diplomatic relations with the Byzantine Empire, and the earliest Byzantine digression
about Central Asia, the itinerary of Zemarchus, was the narrative of a Byzantine
embassy to the court of the Türk Khaganate. Zemarchus, who was magister militum
per Orientem under the reign of Justin II (r. 565-578), accompanied Maniakh, the
Türk ambassador of Sogdian origin, who voyaged to the heartland of the Türk
Khaganate. There he met Silziboulos (Istämi), the ruler of the Türk Khaganate.94 The
Byzantine Empire and the Türk Khaganate had common interests against the
Sassanids of Iran. Hence, they became allies in the late 6th century. In this way, the
Turks were able to organize raids in regions, such as the Crimea and the Caucasus,
which were far west of their own lands. The Türk Khaganate continued to exist until
the mid-8th century as a vassal of Tang China.
Avars also appeared in the mid-6th century in the Northern Balkans. They
were also a heterogeneous group of steppe nomads; their federation probably
included several Turkic, Mongol and Iranian tribes. They raided the Byzantine
territories in Europe in the 580s, and for the next half century, they constituted a
threat to Byzantine domination in the Balkans. In 626, they besieged Constantinople
with the help of their Slavic and Persian allies. After this unsuccessful siege, they left
the Balkans gradually. However, until the end of the 8th century, the Avars still
dominated the Pannonian plain.95
93 Golden, An Introduction to the History of the Turkic Peoples, 127.
94 Dobrovits, “The Altaic World through Byzantine Eyes: Some Remarks on the Historical
Circumstances of Zemarchus’ Journey to the Turks (AD 569–570),” 388-389.
95 Golden, An Introduction to the History of the Turkic Peoples, 112.
58
Bulgars were another Turkic population that appeared in the late 7th century
in the Danubian frontier of the Byzantine Empire. The Bulgars -with the help of
several Slavic tribes- defeated the Byzantine Empire in the Battle of Ongal (680) and
occupied today’s Romania and Bulgaria. Under the khans Asparukh (r. second half
of the 7th century) and Tervel (r. first quarter of the 8th century), they founded an
independent monarchy in the eastern Balkans. Bulgars experienced a rather complex
process of ethnogenesis and cultural shift, and they became Slavic-speaking
Bulgarians.96
Khazars founded a successor state of the Türk Khaganate in the Pontic
steppe. They continued the pro-Byzantine policies of the Türk Khaganate and had
even closer relationships with the Byzantines. The Byzantines and Khazars had two
matrimonial alliances, which is truly exceptional in the context of Byzantine
diplomacy with a non-Christian entity. First, Justinian II, after his dethronement and
exile to Crimea, went to the Khazar court as an asylum, where he married Theodora
(she took this name as part of her baptism, her original name is unknown), the sister
of Busir, khagan of Khazaria. Then, Constantine V (r. 741-775) took the Khazar
princess Tzitzak as a spouse.97 Their son Leo IV is known as Leo the Khazar (Λέων
ὁ Χάζαρος) because of his mother’s ethnicity. A key event in the history of Khazars
is their conversion to Judaism. They must have converted to Judaism in the 830s at
the latest.98 Moreover, the Khazars, because of their unique position as a result of the
religion they chose, managed to remain outside the ideological impact of both the
Byzantines and Muslim Arabs. The khaganate vanished in the 10th century because
of the expansion of Kievan Rus’, Pechenegs, and Oghuz.
96 Golden, An Introduction to the History of the Turkic Peoples, 246-253.
97 Howard-Johnston, “Byzantine Sources for Khazar History,” 168.
98 Golden, “The Conversion of Khazars to Judaism,” 156.
59
Pechenegs were another Turkic nomadic group that appeared in the Pontic
steppes in the late 9th century. The Byzantines were aware of them since their earliest
westward movements; in Constantine Porphyrogennetos’ De Administrando Imperio,
there is an ethnological digression about them. These sections that deal with the
Pechenegs were possibly based on authentic Pecheneg material and give interesting
information about the internal structure of this Turkic population. This text locates
the Pechenegs close to the Byzantine city of Kherson. This nation expanded their
authority in the Balkans in the mid-11th century. Particularly in the last quarter of the
11th century, the Pechenegs posed a real threat to the empire. However, during the
reign of Alexios Komnenos, the Byzantine-Cuman alliance defeated the Pechenegs at
the battle of Levounion (1091); afterwards these people ceased to exist as an
independent entity.99
In the mid-11th century, an important agitation occurred between various
nomadic groups in the steppes. This agitation triggered a westward migration of
these nomads. The Armenian historian Matthew of Edessa explains this fact as
follows:
And there took place days that breathed enormous carnage and bitterness
because of the carrion-eating, godless, unclean people of the Pechenegs, the
mad, blood-drinking beasts. Then the "people of the snakes" (possibly Qay, a
Mongolic population of Eastern Eurasia) drew near and attacked the "pale
ones" (i.e., Cumans) and the "pale ones" were driven out and attacked the Uz
(i.e., Oghuz) and Pechenegs and in concert they were fired up against the
Romans.100
Thus, the Pecheneg, Oghuz, and Cuman invasions of the Balkans were the results of
a common migration movement, similar to what happened in the Migration period.
99 For an extensive study of the history of Byzantine-Pecheneg relations, see Malamut, "L’image
byzantine des Petchénègues," 105-147.
100 Golden, An Introduction to the History of the Turkic Peoples, 274.
60
There was also a brief Oghuz invasion in the northern Balkans in 1064. These
Oghuz were a less populous branch of the Oghuz of Central Asia. These tribes
appear in the Byzantine texts as Οὖζοι. After a series of raids against the princedom
of Pereyaslavl, they were defeated by the prince Vsevolod Yaroslavich (r. 1054-
1073) and were repelled to the Balkans. Then, they briefly raided the Byzantine
frontier on the Danube. Afterwards, they were defeated both by the Byzantines and
Hungarians and then were entered into the service of Kievan Rus. In the Russian
chronicles, this population was known as Torks.101 These Oghuz’ activities in the
Byzantine territories were narrated in the accounts of Attaleiates and Anna
Komnene.
Finally, the Cumans were the final Turkic population of the Balkans that
confronted the Byzantines. They were originally a union of three different groups:
Cumans proper, Kypchaks, and Kangli. However, their westernmost ethnic element
was Cumans. Thus, in Greek and Latin, they were known as Κουμάνοι and Cumani.
They were called “Polovtsy” in the Slavic languages, which means “the pale ones.”
In medieval Muslim geographical literature, they were generally referred to as
Kypchaks. In any case, our knowledge of their ethnogenesis and internal evolution is
very scarce. This population was at the height of its political and military activity at
the end of the 11th and beginning of the 12th century. The Cumans played an
important role in the formation of the Bulgarian and Wallachian states in the 14th
century, and they also served the Kingdom of Hungary. After the Mongolian
invasions in the 13th century, they lost their political influence in the steppe region;
however, because of the presence of a huge number of Cuman-Kypchak slaves in the
Crimea, Khwarizm and, later, Egypt, they preserved an important political network,
101 Agacanov, Oğuzlar, 234-238.
61
especially in the Muslim world. Many early Mamluk sultans of Egypt, such as
Baibars (r. 1260-1277) and Qalawun (r. 1279-1290) were of Cuman-Kypchak origin.
In concluding this subchapter, I should underline some important points: The
Turkic populations who dwelled in the Balkans, Pontic steppe, and Crimea remained
outside the ideological impact of Islam. They were generally pagans, and when they
converted to a monotheistic religion, they were evangelized. (The Khazars’
conversion to Judaism is a unique event.) Secondly, apart from the Khazars and
Bulgars, they could not form an organized state structure. Thirdly, in Byzantine
historiography, these populations (with the exception of the Türk Empire) were more
or less associated with the Skythians because of the geographical location of their
habitat and their lifestyle. This association with the Skythians reflects a long tradition
of historiography, which began with the Histories of Herodotus.
2.5 The appearance of Seljuk Turks in the Eastern frontier
While the permanent enemies in the eastern borders of the Byzantines were the
Arabs, it is possible to see occasional references to the Turks in Byzantine historical
accounts of 9th-century events.102 For example, in a passage concerning the Battle of
Dazimon (838) in the Synopsis Historion, John Skylitzes mentions the presence of
Turks among the forces in the army of the “amermoumnes” (i.e., the caliph al-
Muʿtaṣim) that affronted the imperial army led by Emperor Theophilos:
His thinking was that if the son got the better of the emperor, victory would
surely follow for the father. If the son failed, it were better to stay where he
was. Having considered that advice and come to this decision, he despatched
102 The ethnonym Turk (Τοῦρκος) was not unfamiliar to the Byzantines. It was used appropriately for
the Eastern Turks (Göktürks) and its earliest attestation is in the History of Agathias in the 6th century.
During the 8th and 9th centuries, this ethnonym was utilised mainly for the Magyars who were
following a nomadic lifestyle that was similar to the Altaic populations of the Eurasian steppes. For a
chronological list of the use of this ethnonym, see Moravcsik, Byzantinaturcica, vol. 2, 320-327.
62
his son, who took with him Amr, the then emir of Melitene, ten thousand
Turks, the entire army of the Armenians and their commander-in-chief.103
Skylitzes further writes that “by incessant use of their bows, the Turks deterred the
Romans from pursuing them, which caused the battle to take on a different character.
Unable to withstand the continuous hail of the Turkish arrows, the Romans did an
about-turn and abandoned the emperor.”104
There is no reason to believe that these passages demonstrate an improper use
of the ethnonym Turks, projecting the 11th-century realities that John Skylitzes
himself encountered to the narration of the events of the 9th century. In the Abbasid
armies of the 9th century, there was an important presence of slaves (ghulams) of
Turkic origin. These people were enslaved and brought to Iraq, where they were
recruited as slave-soldiers. They gradually became the king-makers of the Abbasid
capital. Skylitzes’ description of their warfare seems also quite accurate for the
medieval Turks.
In the “Lay of the Emir,” the first part of the epic of Digenis Akritas, the
author lists the populations recruited as soldiers by the Emir, who subsequently will
be converted to Orthodox Christianity and become the father of Digenis, as “Turks
and Daylamites (Διλεβίτας), Arabs and Troglodytes (Τρωγλοδύτας).”105 Leaving
aside the last ethnonym, which was a mere repetition of the name of a legendary
103 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 75; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 77.
104 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 76; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 78.
105 The Daylamites that were a warlike population of Northern Persia were the founders of the ruling
dynasty of Iran prior to Seljuk invasion, the Buyids. There are occasional references to Daylamites in
Byzantine historiography. A passage in Skylitzes concerning the ethnic groups in Ṭog̲ h̲ rı̊l Beg’s army
is interesting: “Καὶ δὴ λαόν ἐπίλεκτον συστησάμενος ἔκ τε Τούρκων καὶ Καβείρων καὶ Διλιμνίτων
περὶ τὰς ἕκατον χιλιάδας…” In this passage the Διλιμνίτοι are also identified as the Daylamites by
John Wortley. The ethnonym Καβείροι seems mysterious. Skyliztes-Thurn, Ioannis Scylitzae Synopsis
Historiarum, 514. Skylitzes-Wortley,A Synopsis of Byzantine History, 811-1057, 422.
63
people attested in Herodotus, these ethnonyms refer to the main ethnicities of the
soldiers combatting in the pre-11th century Muslim armies against the Byzantines. In
the Grottaferrata version of the text, the author adds the Γουλαβίους, an enigmatic
name that could be explained only with the word Μαγούλιοι (that is attested in the
reconstructed version Z) which designates the enemies of the emperor. According to
Henri Grégoire and George Huxley, this word Μαγούλιοι is a metathetic form of the
word Γουλάμιοι; ghulams, and Γουλάβιοι is the incorrect form of the same word.106
Thus, in the epic universe of Digenis Akritas, among the oldest strata of the
reminiscences of the wars against the Arabs, which were mixed with the experiences
of the confrontation vis-à-vis Seljuks in the later periods, there was that name
designating the slave-soldiers, mostly of Turkic stock, placed in the frontier regions.
Actually there were such commanders in the thugur, just like Yazaman al-Khadim,
the eunuch of Turkic origin who was the emir of Tarsus in the mid-9th century.
In the 11th century a new power appeared in the steppes of Central Asia: the
Seljuks. The founder of the dynasty, Seljuk, was an Oghuz in the service of either the
ruler of the Khazars or the Oghuz,107 who left his country because of his conflict with
this ruler, and migrated southward with his tribesmen and followers. There, his
tribesmen and followers (who will be called the Seljuks) established themselves and
gained power, engaging in conflicts between the Samanids and Kara-Khanids.
Seljuk’s grandsons, Ṭog̲ h̲ rı̊l Beg and Čag̲ h̲ rı̊ Beg, after decades of struggles,
managed to dominate Khorasan and adjacent areas. In the mid-11th century, after
106 Digenes Akrites, ed. and tr. Mavrogordato, 4; Grégoire, “L'épopée byzantine et ses rapports avec
l'épopée turque et l'épopée romane,” 481-482; Huxley, “Antecedents and Context of Digenes Akrites,”
329.
107 I am convinced by Peacock’s view that Seljuk was in the service of the Khazar Khagan, not the
Oghuz Yabghu. See Peacock, Early Seljuk History: A New Interpretation, 27-31.
64
their victory over the Ghaznavids at Dandanaqan (1040), they became the dominant
power in the territories extending from Transoxiana to Mesopotamia.108
Earliest Türkmen raids on the eastern edges of Byzantium were dated to
1029,109 although the details of this campaign are obscure. The scarce information
about these campaigns –the only goal of which seems to be pillage- comes basically
from Armenian and Syriac authors. The Seljuks seriously came to the attention of
Byzantine authors in the accounts of the reign of Constantine IX Monomakhos (r.
1041-1055) after their victory over the Byzantines and their Georgian allies at the
Battle of Kapetron in 1048. In this small-scale war, the Seljuk prince and commander
Ibrāhīm Yinal (appears in Skylitzes as Ἀβράμιος Ἀλείμ) captured the Georgian
nobleman Liparites who commanded the Christian armies. Liparites was brought to
the city Rayy in Persia and delivered to Ṭog̲ h̲ rı̊l Beg. He could be ransomed by the
Byzantines in exchange for extravagant gifts and a peace treaty between the two
powers. This episode is mentioned in the chronicles of both John Skylitzes and
Michael Attaleiates.110
As it was mentioned above, the early Seljuk state had a strong tribal and
nomadic element. The Türkmen groups such as Yinaliyan, Nawakiya and Iraqiya had
a certain degree of autonomy from the Seljuk government. They had also close
relations with several members of Seljuk ruling family, such as Ibrāhīm Yinal and
Sulaymāns̲ h̲ āh I of Rum. These groups also played a crucial role in the Seljuk
activity in the west.111
108 Peacock, Early Seljuq History: A New Interpretation, 16-46.
109 Cahen, “La première pénétration turque en Asie Mineure (seconde moitié du Xle s.)," 5-67. For the
earliest campaigns, see 7-10.
110 Attaleiates, The History,78-81 and Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 447-
455; Skylitzes-Wortley, A Synopsis of Byzantine History, 811-1057, 421-426.
111 For these tribal groups see Peacock, Early Seljuq History: A New Interpretation, 47-71 and
particularly, Peacock, “From the Balkhān-Kūhīyān to the Nāwakīya:Nomadic Politics and the
Foundations of Seljūq Rule in Anatolia,” 55-80.
65
The encounter of the Byzantines with the Seljuks was a new encounter. The
Seljuks were a new population in the Byzantine oikoumene, so the authors wrote a
narrative of origo gentis to propose or invent a historical origin for the new
populations and so fulfil this need.112 Of course the need for the production of new
narratives on the newly appeared populations was not limited to the ones regarding
their origin and the re-use of older literary models inherited from Antiquity was also
frequent.113 The creation of such narratives could be interpreted by an explanation by
Christopher Mallan. When he discusses the signification of the narrative about
Ṭog̲ h̲ rı̊l Beg and Liparites, he demonstrates that these narratives follow an older
model closely, the episode of the aftermath of the Battle of Hydaspes (326 BC)
between Alexander the Great and King Porus of India, in which Alexander asks how
should he be treated by the Indian ruler in captivity and he responds “Like a king.”
Attaleiates repeats the narrative putting “Sultan” and Liparites in the places of
Alexander and Porus, respectively. Mallan interprets this narrative as a literary
invention that aims to assimilate the alien Turkish leader into the familiar Roman
worldview.114
The Byzantine defeat at Manzikert (1071) concludes this period and starts a
new political conjuncture that will be radically different from the geopolitics of the
10th and 11th centuries. The invasion of Asia Minor by the Seljuks was quite
unexpected, and the image of these Turks from Persia was not quite different than the
steppe nomads whom the Byzantines were habituated to dealing with since the
invasion of the Avars in the 7th century. However, these new Turks were gradually
112 I shall treat such narratives in detail in Chapter 4. For the evolution of this literary genre, see
Wolfram, “Le genre de l’Origo gentis,” 789-801.
113 Mimesis was a predominant literary technique in Byzantium. The classical introductory text on the
subject is Hunger, “On the Imitation (Mimesis) of Antiquity in Byzantine Literature,” 15-38.
114 Mallan, “A Turkish Alexander? Michael Attaleiates, Porus and Alexander the Great,” 101-107.
66
influenced by the Muslim culture they adopted in the Islamic countries they
conquered, and they started to play the old role of the antagonist against the
Byzantines. Unlike the religious zeal that played a negligible role in the Seljuk
invasion of Asia Minor, the deeds of the Türkmen warriors are considered to fulfill
the ideals of the Arab warriors of long times ago. The adoption of such an ethos
probably helped the cultural integration of the recently Islamized Turks into the
greater Muslim world. There could be further speculation that the ghulams or
mercenaries of Turkic origin played a certain role in this process, acting as cultural
mediators. The works Battalname and Danişmendname reflect a cultural milieu and a
military subculture of the persons involved in the war against the Byzantines. The
Seljuks who entered Baghdad in 1055, there encountered a military élite of Turkic
roots. For example, Arslan Besasiri, the last Buyid military governor of Baghdad,
was originally a Turkic slave. This milieu could have perfectly played the role of a
cultural mediator between the old Muslim warrior communities who embraced the
old frontier ethos of thugur and the newly Islamized Turkic warriors that were aliens
in this new environment and searching new values different than their tribal ones.
In the aftermath of Manzikert, the invasion of Anatolia by the Turks was not
the result of Seljuk state policy. Alp Arslan was not interested in a westward
expansion into the Anatolian plateau. However, the swift disorganization of the
Byzantine state authority in the eastern borders of the empire and sporadic rebellions
and civil wars between 1071-1081 left defenseless the eastern border. Taking
advantage of the chaotic situation, various Turkish warlords and Türkmen groups
invaded these territories. The Armenian nobility who were reluctant to accept the
Byzantine suzerainty in the 11th century had also separatist tendencies and created
67
several breakaway statelets in the last quarter of the century. The combination of all
these element made the invasion of Anatolia possible.
In conclusion, the main aspects of this historical introduction can be
summarized as follows:
i. The early Byzantine-Arab wars resulted in the permanent Arab
domination and subsequent Arabization and Islamization of the oriental
provinces of the Byzantine Empire. However, Asia Minor resisted such an
invasion; the formation of a frontier space between two powers created both an
ethos on two sides and a strong antagonism. The Arab (Saracen/Hagarene/
Ismaelite) figure substituted the archetype of the Persian as the antagonist of the
Byzantine Empire, and even of Greco-Roman identity. Thus, this period until the
9th century was the formative period of the frontier ethos and antagonism.
ii. The invasions by nomadic peoples in the Balkan borders of the
empire, at first glance, had a less significant ideological impact compared to the
ideological effect of the invasions in the east. However, the invasions by
Bulgarians and Slavs had a strong demographic impact in the Balkans and this
fact triggered several socio-political changes, but discussing them falls far
beyond the remit of this thesis. However, the different waves of the “barbarian”
invasions both contributed to the evolution of the image of the “Skythian”
peoples and created social osmosis between the people of the steppes and
Byzantines. This process was decisive for the ethnogenesis of several Balkan
nations, such as modern Bulgarians.
68
iii. The Byzantine eastward and southward expansion in the 9th and 10th
centuries gave a new dimension to the already existing conflict. It inspired new
hopes and new fears that created a new relationship with the (non-Chalcedonian)
Christians of the East and the Byzantines and a new geopolitical system that
included the Byzantine cooperation and alliance with the smaller Muslim entities
of the region. Politically, after the conquests of Nikephoros II and John I, there
was no more a thugur region. This zone was incorporated into Byzantine realms.
iv. Both the ideological geography and the cultural ethos of the
Byzantine-Muslim frontier region predate the arrival of the Seljuk Turks to the
Near East. Thus, the Turks did not create but inherited this ethos from the Arabs.
This fact demonstrates the cultural continuity between the Arab warriors in the
thugur and their Turkic successors. Finally, the Seljuk invasion of Anatolia
ended the 10th-11th centuries’ Byzantine dominance in the Near East and
transferred the Byzantine-Muslim frontier to the heart of Asia Minor.
69
CHAPTER 3
OLD MEMORIES, NEW BARBARIANS: TURKIC “BARBARIANS”
IN BYZANTINE HISTORIOGRAPHICAL IMAGINATION
3.1 The many faces of the eternal barbarian
As it was mentioned before, the Byzantine literary conventions including the
historiographical literature were based on the mimesis of the classical Greek texts.
Cyril Mango called it “a ritualized ballet” in which the contemporary nations and
tribes appear under the names of ancient peoples mentioned in classical texts and
contemporary personalities paraphrase the rhetorical speeches of Classical
Antiquity.115 However, in the Byzantine thinking, there are notions which were not
merely a repetition of ancient clichés, but were inherited from the classical past and
survived in both the collective memory and historiography. Thus, the barbarian
(βάρβαρος) is an archetype which was inherited from classical literature. The
etymology and semantics of this word have been discussed since classical antiquity,
and because of the influence of this subject for our study, hereby I summarize this
semantic evolution as an introduction to this chapter.
Although there is still some discussion on the etymology of the word, it
seems that it is an onomatopoeic word that reflects a notion which exists in other
Indo-European languages. The word essentially indicates an outsider who had a
linguistic barrier with a society where he/she does not belong. A second word which
appears as early as in the Iliad of Homer and seems closely related to barbarian is
barbarophonos (βαρβαρόφωνος). Homer uses this word to describe the Carians in
115 Mango, “Discontinuity with the Classical Past in Byzantium,” 50.
70
the Trojan War, a people of South-western Asia Minor who spoke Greek poorly.116
Strabo also discusses the meaning of the word barbarophonos and defines it in
onomatopoeic terms as any stranger who speaks or pronounces Greek in an improper
or inappropriate way. However, the notion also had a moral dimension. This moral
element which indicates the essential difference between a Greek and a non-Greek
was explained in the Politics of Aristotle. According to Aristotle, the barbarians had
no natural ruling class, hence they were societies of slaves and a barbarian was the
same as slave.117 Since Aristotle was widely read in the intellectual circles of
Byzantium and had a certain influence on Byzantine political thinking, it may not be
incorrect to assume that his thinking was one of the sources of the Byzantine
approach to barbarians. In this case, the approach of Paolo Odorico is also very
helpful to understand these dichotomies: Odorico; cites the following statement from
Aristotle’s Metaphysics; “τὸ δὲ διάφορον τινὸς τινὶ διάφορον, ὥστε ἀνάγκη ταὐτό τι
εἶναι ᾧ διαφέρουσιν,” i.e. “that is different from anything is different in some
respect, so that there must be something identical whereby they differ”118. So
“difference” and “diversity” are distinct notions. The difference of two objects
implies the existence of other aspects which are identical, but the diversity between
them means an ontological diversity. Diversity can lead to situations such as the
early encounters of Spaniards with the American Indians, which created discussions
questioning the humanity of American Indians.119
The category of the “barbarian” in the eyes of ancient Greeks covers all the
non-Greeks, thereupon it is not a part of the antagonism of “civilized versus
116 À la rencontre de l'étranger: L'image de l'Autre chez les Anciens, 8-10. This word appears in the
Iliad only once, in the verse: Νάστης αὖ Καρῶν ἡγήσατο βαρβαροφώνων (Iliad, II, 867).
117 À la rencontre de l'étranger: L'image de l'Autre chez les Anciens, 16.
118 Aristotle, The Complete Works of Aristotle - The Revised Oxford Translation, vol. 2, 141.
119 Odorico, “Différence non diversité: Les Grecs du Moyen Âge faux aux autres Européens,” 1.2.-1.4.
71
uncivilized” or “sedentary versus nomad.” The antagonism was simply between the
“Greeks” and “Non-Greeks.” Thus, Persians, Phoenicians, and even the Egyptians,
who were exceedingly respected by the Ancient Greeks, were still considered
barbarians. To call someone a barbarian is never a neutral approach. It is a pejorative
naming that reflects the construction of a civilized self that obviously claims a certain
feeling of superiority over the one who is called a barbarian. The meaning of this
superiority was somewhat shifted between Antiquity and the Middle Ages, but its
social function remained unchanged.
Following the Christianization of the Roman Empire, the old Greco-Roman
model was somewhat changed, but the antagonism between the “Romans” and
“barbarians” continued to exist. After the fall of the Western Roman Empire, the
Eastern Roman Empire continued to survive as the stronghold of Christianity and
Greco-Roman culture. Justinian I’s imperial project represented its ambition to
reclaim the Empire’s lost lands in the West, particularly the Italian peninsula.
However, the imperial restoration was no more than a short-lived success. A century
later the Byzantines lost even their oriental provinces, namely Syria, Palestine, and
Egypt, to the Arabs.
The evangelization of Roman society naturally affected people’s opinions
about the ones who were considered “others.” Basil of Caesarea (330-379), one of
the Kappadokian fathers, wrote a homily called “On Baptism” which stresses the
importance of baptism as a rite that gives the people a new dignity as “to be dressed
of Christ.” He cites the New Testament passage “Where there is neither Greek nor
Jew, circumcision nor uncircumcision, Barbarian, Skythian, bond nor free: but Christ
is all, and in all.” (Colossians 3:11) and states that whoever baptized, despite he/she
was Jew, Greek, male, female, slave or free, Skythian, Barbarian or from another
72
race, he stripped up old man from old practices and becomes of the blood of
Christ.120
Ahrweiler states that after the 9th century the Byzantine Empire adopted the
idea of the “chosen people” that has a historical role of being the pioneer of the
project of an Orthodox Christian empire. According to her, it was an ideology that
originated in the collective superiority feeling that often takes the form of a “racisme
sui generis.” The evangelization of the barbarians (such as the Bulgarians and the
Serbs) and their further situation in the eyes of Byzantine authors is also a
problematic issue. These peoples converted to Orthodox Christianity thanks to
Byzantine missionaries, such as Saint-Cyril and Saint-Methodius. However, they
could not gain the status of civilized peoples in the eyes of the Byzantines. While the
rulers of these nations were styled as “very Christian” rulers in the documents
prepared in the imperial chancellery, the same nations could appear as a “barbarian
race” or a “corrupted race” in private correspondence.121 This claim of superiority
was not directed only against the recently Christianized Balkan peoples, but also the
populations of Western Europe that were commonly named “Frangoi” by the
Byzantines. A rhetorical work cited by Anthony Kaldellis supports her position: after
the evangelization of the Bulgarians, they say, “we do not want to be called Skythian
or barbarian or by another name anymore; we are now all Christians and children of
God.”122 However, even in the 11th and 12th centuries, there were Byzantine sources
with a hostile tone and barbarizing discourse against the Bulgarians.123 The
Byzantines did not see any people equal to them unless they embraced “Romanitas”
120 À la rencontre de l'étranger: L'image de l'Autre chez les Anciens, 306.
121 Ahrweiler, L'idéologie politique de l'Empire byzantin, 51.
122 Kaldellis, Le discours ethnographique à Byzance: Continuité et rupture, 155. He attributes this text
to Theodore Daphnopates.
123 Kaldellis, Le discours ethnographique à Byzance: Continuité et rupture, 162-163.
73
and adopted Byzantine culture itself. In the Byzantine worldview, the center was
always Constantinople and the capital always represented the real culture of
“Romanitas”. Even the provincials were considered inferior to them in this respect
and non-Greek peoples who converted to Christianity had a rather negative status.124
In that case the situation of the ethnic groups, such as Bulgarians, Armenians
and Vlachs, merits special attention. These three ethnic groups were Christianized in
different periods. The Armenians were one of the peoples who embraced the
Christian religion earlier in the 4th century along with the Georgians, yet they were
generally non-Chalcedonian Christians who had their own Armenian Apostolic
Church. However, there was a large number of Armenians in the Byzantine political
élite who were embraced the Chalcedonian creed.
The Vlach people were the socially marginalized post-Roman inhabitants of
the Balkan peninsula who spoke a language of Latin origin and lived a semi-nomadic
life in various parts of the Balkans. They were Christians and, despite their
socioeconomic marginalization, they were remnants of the Latin-speaking
inhabitants of the Roman Empire in the Balkan peninsula. So, theoretically, they
could claim the legacy of “Romanitas” as much as the Byzantine Greeks. However,
probably because of their social marginalization, they were still considered inferior
and quasi-barbarian.125
The abovementioned Bulgarians were for a long time considered Skythian by
origin126 and created troubles for the Empire, particularly until their conversion to
124Tremblay, “L’identité romaine est-elle exclusive à Constantinople? Dichotomie entre Byzance et
les Balkans à l’époque médiobyzantine (vie–xiie siècles),” 25-40.
125 For their ethnogenesis and early literary representation, see Charanis, “John Lydus and the
Question of the Origin of the Vlachs in the Greek Lands,” 103-107. For the sources concerning the
Dacians and Vlachs see Fontes Historiae Daco-Romanae, 4 vols.
126 Even in the 14th century; when the ethnonym Skythian was employed only for the Mongols,
Nikephoros Gregoras states that the Bulgarian people are indeed of the Skythian stock and they took
their name from a river basin north of Danube (sic) called Bulga. This name could be a Greek
74
Christianity. However, their ruler Boris I (r. 852-889) received the holy baptism in
864 and this event created new expectations for the Byzantine side. For the
Byzantine ruling élite, the signification of this baptism was the subordination of
Bulgars to the empire because the Bulgarian ruler had the Byzantine emperor
Michael III (r. 842-867) as his godfather. However, the event’s outcome did not
happen as the Byzantines had wished and Boris I created his own Bulgarian
Orthodox Church, while his son, Simeon of Bulgaria (r. 893-927) even had the
ambition of becoming emperor. After that point, the Byzantine authors adopted a
hostile anti-Bulgarian tone. There was a hierarchy between the two entities and the
Bulgarian ruler could be only a son of the emperor. Although there was a spiritual
link between them, a Bulgarian ruler could not wear the imperial crown. Even if
there was a marital link between the Macedonian dynasty of the Byzantine Empire
and the Dulo dynasty of Bulgaria, Bulgarian rulers were considered somewhat
inferior to Byzantine emperors.127
A striking expression of this difference between the Byzantines and other
“Orthodox barbarians” can be found in John Kanaboutzes’ Commentary to Dionysius
of Halicarnassus. The 15th-century author states that the Greeks considered the
Trojans barbarians even though they believed in the same gods, and the
contemporary Byzantines (the author addresses them proudly as “we”) consider the
Bulgarians, Vlachs, Albanians, and Russians barbarians, because to be a barbarian is
not a matter of religion, but of race, language, life-style, and culture.128
rendering of the Slavic hydronym Volga. Gregoras-Schopen-Bekker, Nicephori Gregorae Byzantina
historia, 26 and Gregoras-Van Dieten, Rhomäische Geschichte, 75.
127 Malamut, “Les peuples étrangers dans l'idéologie impériale,” 128-130. See also Page, Being
Byzantine: Greek Identity before the Ottomans, 53-55.
128 Odorico, “Identité et craintes. Théodore Pédiasimos à Serrès au XIVe siècle,” 171.
75
Apart from the socio-political change of the perception of the notion of the
barbarian, the historical lexicology of the word “barbarian” in Byzantine cultural
history is still worthy of attention. In the Lexicon of Patriarch Photios I (810-893) the
word βάρβαρος was not present, yet there is the verb βαρβαρίζω which is defined as
“to be minded like barbarians.”129
On the other hand, it would be wrong to say that the Byzantine identity
remained static in history.130 As Gill Page puts it, the Fourth Crusade is an important
turning point for Romanitas. As early as the 12th century, a new insulting language
began to be used to describe western Catholics. Latins residing in the Byzantine
Empire were massacred from time to time, especially during the reign of
Andronikos I Komnenos. The trauma of 1204 triggered changes in Roman
identity.131 In the 13th century, the differences between the Byzantines and the
westerners were emphasized, while in the next century, Gregoras and Kantakouzenos
began to emphasize their Christian identity. The issue of which identity will be
established against which community also determines which elements of identity will
be underlined.
3.2 Classifying the barbarians
As mentioned previously, Byzantine authors hardly used the contemporary names of
foreign nations. Their commonly accepted literary convention was using ancient
ethnonyms for contemporary populations. This usage of ancient ethnonyms was
never arbitrary and these terms were organized within a complex system of
129 Photios, Photii Patriarchae Lexicon: A-D, 326.
130 See Stouraitis, “Roman Identity in Byzantium: A Critical Approach,” 175-220; Stouraitis,
“Reinventing Roman Ethnicity in High and Late Medieval Byzantium,” 70-94.
131 Page, Being Byzantine: Greek Identity before the Ottomans, 69-70.
76
interrelations based on the prior representations of the peoples, which corresponded
to “predecessors” of the contemporary populations. Moreover, the current political
conjunctures also had a certain role in the attribution of ancient ethnonyms to
contemporary nations. Kaldellis, who criticizes the over-use of mimesis to explain
the usage of ancient ethnonyms only as an archaizing literary effort, further claims
that thinking these nations had one “true” or “authentic” and one “false” or “exotic”
name is misleading. According to him, all these names could be true and any
objectivity about this question is impossible because in the Middle Ages there were
no universally accepted official names.132
These ethnonyms could be perceived as signifiers (a term coined by
Ferdinand de Saussure) used by the Byzantines to define any foreign ethnos. The
cause of their divergence from the actual names of the populations to whom they
refer is the difference between the literary language, examples of which could be
found in written texts, and non-written language. In this context, the works written in
Greek that are closer to the spoken language can give some idea about the Greek
names used in the spoken language or even these populations’ endonyms in their
proper language. A good example of such a work is De Administrando Imperio.
The first step to understanding the formation of the image of the Turks in
Byzantine literature is to classify and analyze these ethnonyms. The ethnonyms can
be categorized under three main groups as chronological/historical, geographical, and
composite ones (see Table 2).
132 Kaldellis, Le discours ethnographique à Byzance: Continuité et rupture, 126-127.
77
Table 2. Ethnonyms
The two chronological categories of ethnonyms were the ones inherited from
classical antiquity and those that appeared later. The most important source of
classical ethnonyms is the History of Herodotus. The Persian (Πέρσης), Skythian
(Σκύθης), Egyptian (Αἰγύπτιος), and Indian (Ἰνδικός) are the most frequent examples
of classical ethnonyms.
The group of post-classical ethnonyms is made by the terms which were first
attested in Late Antiquity. Turk (Τοῦρκος), Slav (Σκλάβος), Goth (Γότθος), and Hun
(Οὖννος) could be considered examples of this second category. Finally, there were
contemporary ethnonyms, the words used during the lifetime of Byzantine authors,
terms which were possibly the endonyms of such peoples. The terms, such as
Pecheneg/Patzinak (Πατζινάκος), Oguz (Ὀγούζιος), and Turcoman (Τουρκομάνος)
were the contemporary names of these populations.
Ethnonyms
Geographical
Mysians
Karians
Triballi
Historical
Classical
Eastern
Persians
Indians
Northern
Scythians
Sauromatae
Massagetae
Postclassical
Huns
Turks
Pechenegs
Cumans
Composite
Persoarmenians
Tauroscythians
Turcoarmenians
78
There were also ethnonyms that did not reflect a clear ethnicity, yet only a
geographical indication. The words, such as Mysoi (used for the Bulgarians) and
Triballoi (used for the Serbs), are examples of such terms. These terms are basically
derivatives of geographical regions. However, it would be wrong to say that these
ethnonyms did not have a particular semantic weight because they were the names of
the ancient imperial provinces, and such an appellation could legitimize the
reincorporation of these lands into the Empire.133
Still, a final –and a hybrid– category is constituted by composite ethnonyms,
which were constituted by two elements like the Perso-Armenian, Tauro-Skythian
and Perso-Turks. The first ethnonym, used by Khoniates for the Danishmendids,
could imply the possible Armenian origin of the dynasty.134 In the second case, the
term is constituted by one geographical indication and one ethnonym; it was used for
the Crimean Tatars. The third term is constituted with the same formula and it was
used to describe the Akkoyunlu Türkmen and emirate of Erzincan in Eastern
Anatolia.135 An even rarer composite ethnonym, Perso-Skythian, is found in a text
outside the corpus of this thesis, an oration by George Tornikes dated 1192-1193.136
This word was used for Türkmen. Furthermore, a striking feature of these ethnonyms
is that they never reflect a linguistic family. Even the generic ethnonym Skythian,
which was used for the “northern barbarians,” does not reflect the direct
representation of the Altaic peoples; it was also used for the Russians and Goths.
Kaldellis explains this usage with the formula “the geography rules the
133 Kaldellis, Le discours ethnographique à Byzance: Continuité et rupture, 133.
134 Khoniates-Bekker, Nicetae Choniatae Historia, 27; Khoniates - Magoulias, O City of Byzantium,
12.
135 Tauro-Skythian and Turko-Persian (Perso-Turk) were both used by Doukas. See Doukas-Grecu,
Historia Turcobyzantina, 91 and 89, 163, 281; Doukas-Magoulias, Decline and Fall of Byzantium to
the Ottoman Turks, 90 and 127.
136 Oikonomides, “The Turks in the Byzantine Rhetoric of the Twelfth Century,” 150-151.
79
ethnography.”137 In the lack of ethnographic information about these peoples, their
ethnonyms were determined by their homeland. According to him, also the ethnonym
Hun is used –more or less- interchangeably with Skythian; however, I shall deal with
the unique semantic baggage of this word in the next section.
3.3 The Turks and Türkmen
The ethnonym Turk (Τοῦρκος) also merits some attention for its rather unique
position. It could be considered a late antique ethnonym. It was first used in the 6th
century for the Türk Khaganate, which was allied with the Byzantine Empire against
Sassanid Iran. The itinerary of Zemarchus into the court of the Türk khagan is
preserved in Theophylactus Simocatta.138 In later centuries, the ethnonym Turk is
used for the Khazars and Hungarians. It is easy to understand its use for the Khazars:
The Khazar were a Turkic population. However, the use of the same ethnonym for
the Hungarians poses some problems; the Hungarians are a Finno-Ugric, not a Turkic
ethnic group. There are various theories that try to explain this situation. According
to one theory, the Byzantines called the Hungarians as Turks because they were
originally nomads in the region between Don and Caucasus. Thus, they had a
“Turkish” lifestyle. Other theories explain the fact by arguing that the use of the
ethnonym Turk for the Hungarians is an archaizing literary convention or by
asserting that this ethnonym was attached to Hungarians because of their tribal union
with some Turkic groups, such as Kabars. Gyula Moravcsik explains it by arguing
137 Kaldellis, Le discours ethnographique à Byzance: Continuité et rupture, 135.
138 These fragments are translated and analyzed in Dobrovits, “The Altaic World through Byzantine
Eyes: Some Remarks on the Historical Circumstances of Zemarchus’ Journey to the Turks (AD 569-
570),” 373-409.
80
that in the late 9th century, the Hungarians were calling themselves as Turks, so the
word Turk was their endonym.139
After the Seljuk invasions of the 11th century, this ethnonym is also used for
the Seljuk Turks. The chronicle of Skylitzes is a striking example of the different
uses of the same ethnonym in the same work. The author uses the term for the Turkic
slaves/mercenaries in the Abbasid armies, for the Hungarians, and, in the end, for the
Seljuks.140 Attaleiates uses the terms Huns and Turks interchangeably for the
Seljuks.141 Anna Komnene generally refers to the Seljuks as Turks; however, she
sometimes employs the term Persian, but generally in the context of the “sultanate of
Persia,” i.e., Great Seljuks.142 In her work, the term Persia reflects a political entity,
not the homeland of an ethnic group. Yet she seems to be aware of the difference
between the Turkish and Persian languages: when she speaks of Abu’l Qasem, the
Seljuk governor (and regent) of Nicaea, she points out that he was “commonly called
a satrap by the Persians and emir by the Turks who are now masters of the Persian
lands.”143 However, in this context, beyond the difference between the two
languages, the term satrap is used anachronistically and was part of the lexicon about
Antiquity. Finally, she also uses the ethnonym Turk for the Vardariote Turks, a
people that could either the Hungarians or the remnants of the Western Oghuz settled
in Byzantine Macedonia.144
139 Constantine Porphyrogenitus, De Administrando Imperio, commentary, 13-14.
140 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 75-76 (Turkish slave soldiers in the
Abbasid army), 176-177 (Hungarians), 484-485 (Seljuks); Skylitzes-Wortley, A Synopsis of Byzantine
History 811-1057, 77-78, 170-171 and 451-453.
141 Attaleiates, The History, 142-143, 252-255 and 436-439.
142 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias,11 and 186; Komnene-Sewter-Frankopan,
The Alexiad, 9 and 169.
143 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 222; Komnene-Sewter-Frankopan, The
Alexiad, 202.
144 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 126-127 ; Komnene-Sewter-Frankopan,
The Alexiad, 115.
81
The ethnonym Turk was closely related to Skythian. Khalkokondyles brings
some explanation to the semantic relation between the terms Turk and Skythian:
I do not know by what ancient name to call the Turks that would not fall short
of the truth about the matter. Some believe that the Turks are descendants of
the Skythians, which is quite a reasonable conjecture about them, given that
their customs are not all that different and that their languages are even now
closely related... Even today, so they say, it is possible to see numerous
offshoots of this people roaming about in many parts of Asia, who tend to
follow the ways and customs of the nomadic Skythians and have clearly not
settled down in any particular part of Asia. And they also add that the
barbarian nations of the Turks who live in Asia Minor, I mean in Lydia,
Karia, Phrygia and Kappadokia, speak the same language and have the same
dress as the Skythians who roam the lands from the Don into Russia.145
Khalkokondyles then gives two more theories about the origin of the Turks; one
relates them with the Parthians, the other with the Arabs, which I shall deal with in
the next chapter on the origo gentis narratives about the Seljuk and Ottoman Turks.
However, in this passage, there is one of the rare cases that gives some details about
the relationship between the actuality and historical tradition. The author clearly
states that Turk is the contemporary name given to this population, so any other
appellation reflects a narrative strategy or origin theory about Turks.
An ethnonym closely related to Turk, the Turkomanoi (Τουρκομάνοι), is also
attested in Byzantine texts. This ethnonym was used only for the Türkmen, the tribal
populations which acted somewhat independently from the court of Seljuks of Persia
or Rum. The origin of this term is still a matter of dispute: According to Maḥmūd al-
Kashgari, the author of the 11th-century compendium of Turkic dialects, the word
Türkmen is a deformation of the Persian expression Türk manand that means “(those
who are) similar to Turks.”146 However, this etymology is far from being
145 Khalkokondyles, The Histories, 10-13.
146 Kaşgarlı Maḥmūd, Divanü Lügati’t Türk, ed. Erdi - Yurteser, 608.
82
satisfactory. Modern scholarship gives preference to the possible etymology Türk +
“man”, where the suffix “man” strengthens the meaning.147
In the 11th century, this word was mostly used for the Oghuz who converted
to Islam. Afterwards, the words’ meaning changed and was used to refer to the
Oghuz nomads who acted outside the authority of the Seljuks or any other
centralized state in the region. It seems that when the Turks settled in Asia Minor, the
meaning of the ethnonym was as such.
The different uses of the words Turks and Türkmen in the early Ottoman
chronicles also merit some attention. In Aşıkpaşazade’s Tevarih-i Ali Osman, the
Christians always refer to the Ottomans as the Turks, thus suggesting that in the 15th
century, the standard Greek exonym for the Ottoman (and other Anatolian) Turks
must have been Τοῦρκοι. Moreover, although Aşıkpaşazade asserts that the
Ottomans are of Türkmen origin, he uses this word only to define Turkic nomads
outside of the Ottoman state authority, and, generally, with negative connotations.148
This term is first attested in the Epitome of John Kinnamos, when he
describes them as “those who lay beneath his [Qiliç Arslān II] authority, but who are
clever at living by thefts and customarily are called Turkomans.”149 George
Akropolites, carefully distinguishes the Turcomans from the Persians and describes
them as “this is a people who occupy furthest boundaries of the Persians and feel
implacable hatred for the Romans, delight in plundering them, and rejoice in booty
from wars; this especially at the time when Persian affairs were agitated and thrown
into confusion by the Tatar attacks.”150 In this account, the Türkmen are depicted as a
147 Golden, An Introduction to the History of the Turkic Peoples, 212-213. For their social
organization see Peacock, The Great Seljuk Empire, 28-29.
148 Mengüç, “The Türk in Aşıkpaşazâde: A Private Individual’s Ottoman History,” 59.
149 Kinnamos, The Deeds of John and Manuel Comnenus, 158.
150 Akropolites-Heisenberg, Georgi Acropolitae Opera, 136; Akropolites-Macrides, The History, 315.
83
destabilizing force on the edges of the Seljuk sultanate of Rum. Doukas uses the term
twice for the rulers of the Anatolian emirates: İsfendiyar the Turcoman (Σφεντιὰρ
τοῦ Τουρκομάνου) and [Meḥmed] of Dulkadir, the man who rules the Türkmen of
the upper side of Kappadokia (Τουργατὴρ ἀνδρὸς ἀρχιγοῦ τῶν ἐκεῖσε
παρακοιμομένον Τουρκομάνον ὑπεράνω Καππαδόκων).151 Thus, the Byzantine
system of ethnonyms was based mostly on the geographical position of the real or
imaginary homelands of the related populations. However, the lifestyles of the
peoples also had a certain role in determining their location in this system. Thus, an
ethnonym as Skythian, which is identified with “north”, could be used for the Seljuk
Turks that came from the “east”.
The association of the barbarian populations with geographical regions, or
more generally with the cardinal directions, is not limited to either the ancient Greek
or the Byzantine historiographical tradition. The same pattern is present in Chinese
historiography and traditional worldview: in ancient China, all the barbarians
considered non-Chinese were referred to as Yi. The eastern barbarians were called
Dongyi, the western barbarians Xirong, the northern barbarians Beidi, and the
southern barbarians were called Nanman. Despite China’s distance from the Greco-
Roman world, a similar phenomenon of the formation of a civilized central area that
distinguished itself from the various groups they considered not civilized has created
similar patterns of the classification of the “barbarians”.152
151 Doukas-Grecu, Historia Turcobyzantina, 123, 279; Doukas-Magoulias, Decline and Fall of
Byzantium to the Ottoman Turks, 105, 186. In the translation of Magoulias, the word Turcoman was
also used for Uzun Hasan, the ruler of Akkoyunlu. However, this seems to be an addition by the
translator, because in the original Greek text published in the CSHB series, this ethnonym does not
appear in that passage: Doukas-Bekker, Historia Byzantina, 339; cf. Doukas-Magoulias, Decline and
Fall of Byzantium to the Ottoman Turks, 257.
152 Drocourt, “Des simples sauvages aux redoutables étrangers: la notion de «barbares » en Chine
ancienne, à travers leurs dénominations,” 18-19.
84
3.4 The Huns: The uses of a late antique ethnonym
The Huns were a nomadic population whom the Romans saw in Late Antiquity. Like
many nomadic tribal confederations, they had no ethnic homogeneity; they were
formed by several Turkic, Slavic, Iranian, and probably Germanic clans. The Hunnic
invasion of Eastern Europe in the 4th and 5th centuries was especially disastrous for
the later Roman Empire, both its eastern and western branches, and left strong traces
for the European imagination of the Barbarians. Particularly their ruler Attila’s
(r. 434-453) campaigns triggered the ultimate downfall of the Western Roma Empire.
Although the Huns proper no longer played an important role in European politics
after the Battle of Nedao (454), their invasion left a long-lasting effect on the
historical memory of both the Roman and Germanic peoples. According to Denis
Sinor, the name Hun “has become synonymous with that cruel, destructive invaders,”
and it “has been used pejoratively to stigmatize any ferocious, savage enemy.”153
The ethnonym Οὖννοι appears in the Greek texts since -at least- the late
4h/early 5th century. According to Gyula Moravcsik, the earliest text in which the
term was attested was the fragmentary remnants of the lost history of Eunapius.154
Because of the Huns’ crucial historical role in late antique European history, the
ethnonym Hun continued to survive in the later centuries. Moreover, many possibly
Turkic nomadic tribes of western Eurasia were defined as Huns by the Byzantine
historians of Late Antiquity. According to Agathias, “All these peoples were referred
to by the general name of Skythians or Huns, whereas individual tribes had their own
particular names, rooted in the ancestral traditions, such as Kutrigurs, Utigurs,
153 Sinor, The Cambridge History of Early Inner Asia, 177.
154 Moravcsik, Byzantinoturcica, vol. 2, 231-237.
85
Ultizurs, and Burugundi.”155 Jordanes, defines the Huns as “like a fruitful root of
bravest races, sprouted into two hordes of people.” He called these two hordes as
Altziagiri and Sabiri.156 The ethnonym Hun was synonymous with the nomadic
barbarians, although it was not an ethnonym inherited from Classical Antiquity and it
was not part of Herodotus’ lexicon; it was used frequently for many nomadic nations,
such as Bulgars, Avars, Turks, and Cumans, until the 13th century.157
Another population that was associated with the Huns were the Hephthalites
(Ephtalitae) who founded a kingdom in the historical Bactria region. This population
seems to be a multi-ethnic tribal federation that contains both Turkic and Indo-
Iranian elements. Hephthalites were also referred to as the White Huns. Furthermore,
there were also Red Huns: Kermichiones (Κερμιχιῶνες) were another obscure
population attested in Theophanes. Their name seems to be a Greek rendering of the
Armenian expression “Karmir Hiyon,” which means Red Huns.158 While recording
these names, it should also be remembered that the ancient Turkic peoples had a
geographical nomenclature system that identified cardinal directions with colors.
These “Hunnic” populations furthermore adopted Buddhism and ruled parts
of contemporary Afghanistan until the Arab invasions. One of the last local dynasties
of the region, distinguished by their resolute resistance against the Muslim invaders,
was known as Turk-Shahis. Their dynastic name demonstrates the possible
interchangeable use of the ethnonyms Hun and Turks in a socio-cultural circle very
155 Agathias, The Histories, 146. The Burugundi referred to in the text could be either Bulgars or
Germanic Burgundians (Burgundi). If they were Burgundians, this fact suggests how these nomadic
confederations were heterogeneous.
156 Jordanes, Gothic History, 60.
157 Moravcsik, Byzantinoturcica, vol. 2, 234-235.
158 Theophanes, Chronicle, 351. The author presents the ruler of Kermichiones as “Askel, king of the
Hermichiones, who dwell inland of the barbarian nation near the Ocean.” For further information
about the Turkic peoples in the 6th-century Byzantine sources, see Macartney, “On the Greek Sources
for the History of the Turks in the Sixth Century,” 266–75.
86
far from the Byzantine Empire, in the frontiers of the Eastern Iranian – Indian
worlds.
In a passage from the History of the Wars (Ὑπὲρ τῶν Πολέμων Λόγοι),
Procopius of Caesarea describes the Ephtalitae Huns as follows:
The nation of the Ephthalitae Huns, who are called White Huns (Οὔννων τῶν
Ἑφθαλιτῶν ἔθνος, οὕσπερ λευκοὺς ὀνομάζουσι), gathered an imposing army
and marched against them. Ephthalitae is of the stock of the Huns in fact as
well as in name; however, they do not mingle with any of the Huns known to
us […]. Because they are not nomads like the other Hunnic peoples, but for a
long period have been established in a goodly land (ἀγαθῆς χώρας).159
In this passage, nomadic lifestyle and Hunnic identity were bounded inseparably.
Thus, nomadism is the sine qua non condition for being a Hun. Consequently, the
unique situation of the Ephtalitae Huns, who were isolated from the other branches
of the Huns and settled down on a “goodly land,” made them different from the other
Huns.
This differentiation from the main body of the Hunnic genos brought other
positive aspects to the Ephtalitae:
As a result of this, they never made any incursion into the Roman territory
except in company with the Median army. They are the only ones among the
Huns who have white bodies and countenances which are not ugly. It is also
true that their manner of living is unlike that of their kinsmen, nor do they
live a savage life as they do; but they are ruled by one king, and since they
possess a lawful constitution, they observe right and justice in their dealings
both with one another and with their neighbours, in no degree less than the
Romans and the Persians.160
Procopius, thus, provides significant information on the perception of the Huns in the
Byzantine imagination of the 6th century, which can be summarized as follows:
159 Procopius of Caesarea, History of the Wars, vol. 1, books 1-2, (Persian War), 13-15.
160 Procopius of Caesarea, History of the Wars, vol. 1, 15.
87
i. The Ephthalitae are ethnically Huns, but culturally not, because they
are sedentary people. Their differentiation in lifestyle and isolation (endogamy?)
from the other Huns make them less ugly and dark-skinned than their other
kinsmen. Even the word “white” could be understood as a marker of
differentiation which could possibly have a positive meaning for the Byzantines.
ii. The nomadism (νομάδες) and savage (θήριον) are the social aspects
which make a population free from civilized life. To have a lawful constitution
and a monarchy (contrary to nomadic confederations ruled by multiple archons)
is another aspect of the social life that differentiates Ephthalitae from the other
Huns.
iii. A population which is not nomadic meant that it was less dangerous
and hostile to the Romans; the sedentary peoples do not engage in pillage
campaigns toward imperial territories. They only appear as invaders or looters as
allies or mercenaries (“in the company of the Median army”).
Another passage from the History of the Wars can complete the image of the
Huns. In this passage, the Huns are referred to the Massagetae (Μασσαγέται), which
is the name of a Skythian nation attested in Herodotus. This passage from Procopius
deals with a trial for a murder committed by two Hunnic mercenaries in the
Byzantine army, which is concluded with their execution by order of the commander
Belisarius. The scene is very well-constructed with a literary taste; the narrative
touches on the marginal position of the Hunnic mercenaries in the Roman army, the
anxiety of soldiers, and the wisdom of Belisarius as a judge-soldier.
Procopius writes that “two Massagetae killed one of their comrades who was
ridiculing them, in the midst of their intemperate drinking; for they were intoxicated.
For of all men, the Massagetae are the most intemperate drinkers.” Belisarius
88
punishes the mercenaries in a way which evokes the barbarians: they are impaled on
a hill near Abydos. The Hunnic elements in the army dissent by saying that “it was
neither to be punished nor to be subject to the laws of the Romans that they had
entered into an alliance (for their own laws did not make the punishment for murder,
such as this, they said).” Belisarius makes a speech to all soldiers of the army in
which he says that “if any barbarian who has slain his kinsman expects to find
indulgence in his trial on the ground that he was drunk, in all fairness he makes the
charge so much the worse by reason of the very circumstance by which, as he
alleges, his guilt is removed.”161
The passage above shows that, according to Procopius, the Huns have also a
law, but it is such a barbarian law that could permit that if a drunken man kills him,
comrade-in-arms could have impunity. The drunkenness of the barbarians stressed by
Procopius demonstrates another aspect of the perception of the Huns: that they were
regarded as intemperate persons, unlike Greco-Roman people who are moderate.
Procopius’ text was re-elaborated by Michael Attaleiates six centuries later in
his chapter, which deals with the rise of the Seljuks. I shall deal in detail with that
account in the next chapter, as an example of origo gentis narrative; however, a brief
summary of its first passage will be given here. The ethnonym Ephtalitae Huns in
Procopius’ History of the Wars now became the rather mysterious Nepthalite Huns
(Οὖννοι Νεφθαλῖται).162 According to Attaleiates, although the Romans started
formal diplomatic relations and exchanged embassies and gifts with them, they did
not stop raiding the Byzantine territories because the Huns had a “rapacious nature”.
For this fact, even the Sultan (Ṭog̲ h̲ rı̊l Beg) “excused himself by saying that not even
161 Procopius of Caesarea, History of the Wars, vol. 2, Books 3-4. (Vandalic War), 113-115.
162 Attaleiates, The History, 80-81.
89
he knew the identity of these plunderers who, like wild wolves, were making the
raids.”163 Later, Attaleiates also described them as “Nephtalite Huns, that is to say
the Turks” (Νεφθαλιτῶν Οὔννων, ἤτοι τῶν Τούρκων),164 which clearly demonstrates
that Huns and Turks were basically the same people in the eyes of an 11th-century
Byzantine intellectual. Nevertheless, he states that the ethnarch of the Huns is called
“sultan” in the Persian language and refers to the Seljuk forces that sacked
Neokaisereia in 1068 as “the Persians, who are now called Turks.”165 He calls the
forces of the rebellious Seljuk prince, “the Turks of Koutloumous (οἱ
Κουτλούμουσιοι Τοῦρκοι) who were encamped at Chrysopolis with the Hunnish
host,” and writes that their leaders are called “emirs” and “selarioi” (σελάριοι) in the
Turkish language.
A similar narrative also appears in the Synopsis Historion of John Skylitzes,
who defines the Seljuk Turks as “τὸ τῶν Τούρκων ἔθνος γένος μέν ἐστιν Οὐννικόν,”
i.e. “the Turkish nation which is Hunnic by race.”166 Skylitzes uses the same
ethnonym a second and last time, when he narrates the deeds of the Pecheneg
warlord Kegenes, by writing that he “on many occasions routed and repelled the
Oghuz (a Hunnic people)” [γένος δὲ Οὐννικὸν οἱ Οὖζοι].167 These Oghuz were not
the tribes who were engaged in Seljuk state formation in Transoxiana. Instead, they
were another branch of the Oghuz who left their homeland earlier in the 9th century
163 This passage refers to the Turcoman tribes and other centrifugal forces inside the Seljuk realm that
were hard to control and, contrary to the will of their sultan, insisted on plundering Byzantine
territories. The metaphor of wild wolves is also noteworthy because of the role of the wolf in Turkic
mythology as a totem. On the usage of animal metaphors for the representation of Turks, see
Shliakhtin, From Huns into Persians: The Projected Identity of the Turks in Byzantine Rhetoric of
Eleventh and Twelfth Centuries, 192.
164 Attaleiates, The History, 142-143.
165 Attaleiates, The History, 78-79, 192-193.
166 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 442; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 416.
167 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 455; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 427.
90
and migrated westward into the Pontic steppes and stayed untouched by the influence
of Islam. Then, they migrated further to the Balkans in the 11th century and interacted
with the Byzantines. Thus, as these passages demonstrate, Hunnic identity was not
considered to be limited within the boundaries of nationhood; rather it was regarded
as a genos (γένος), a race which is based on a nomadic lifestyle, savagery, and a
roughly Asiatic provenance. As was mentioned, before the peoples or nations (ἔθνοι)
are constructed as the branches of a genos, as demonstrated in the relationship
between the identities, such as Ephtalitae, Massagetae, Oghuz and Seljuks and the
Huns. This model could explain the Byzantine worldview regarding the classification
of the foreign peoples according to their lifestyles. Neither ethnic identity in a
modern sense nor linguistic vicinities did play any significant role in that
classification. Another point to be noted is that the Huns had no a permanent
fatherland and in every narrative their region of origin is different. Although there is
a Persia of Persians and even a Skythia of Skythians, there is no Hunnia. Thus, in
every narrative, the place of origin of the Huns varies: from the misty Lake Maeotis
to the banks of the Ganges River.168 The Huns possessed a special situation as being
a non-territorial and non-geographical nation.
In later sources like the histories of Niketas Khoniates and John Kinnamos,
the abovementioned ethnonym Οὖννοι is used persistently for the Hungarians. The
new identification of the term firstly with the Hungarians (Magyars), who were a
Finno-Ugric tribal confederation that was settled in the Pannonian Basin (Alföld) in
the 10th century where the Huns had settled there roughly four centuries earlier, and
168 Lake Maeotis is proposed by the late antique authors Priscus and Ammianus Marcellinus as the
homeland of the Huns, and Attaleiates put the Hunnic homeland in a country “separated from the land
of Persia by the Ganges River.”
91
then with the Turkic Cumans, who were closely related to the Hungarian crown,
demonstrates a new mechanism based on the standard principle of geography.
3.5 Persians between the Skythians and Saracens
As I already cited Kaldellis, the ethnonym Skythian was semantically close to Hun,
and these two words could be used interchangeably. Both of the ethnonyms reflected
the pastoral and nomad outsiders of the Greco-Roman oikoumene. Since the 7th
century, the ethnonym Skythian has been employed to refer to various Turkic
populations engaged in a nomadic life and looting activities, mostly in the Balkan
frontier of the Byzantine Empire.
As mentioned before, Skythian is one of the oldest ethnonyms; thus its
semantic connotations can be traced back deep in history. This ethnonym is
employed originally to define an Iranian nomadic population of the Pontic Steppe,
the population known as Saka by Persians. Skythians occupy an important part of the
Histories of Herodotus as the barbarian nomads whose lifestyles were in contrast
both with Greeks and Persians.
The moral connotations of the Skythians were bad as the Huns. Michael
Attaleiates describes the Skythians (in this case, the Pechenegs) as follows:
But the Skythians, who are popularly called Pechenegs, crossed the Danube
with all their people and soon established themselves on Roman territory.
This race practices armed raids more than any other skill or art and makes its
living by continuous use of the sword, bow, and arrow. They are loathsome in
their diet and the other aspects of their life, and do not abstain from eating
foul foods. By some evil chance, they poured over the Roman borders and
later on caused many hardships that it would not be possible to enumerate in
detail here.169
169 Attaleiates, The History, 52-53.
92
The ethnonym Skythian will be further used for the Cumans and Mongols in
Byzantine historiography in later centuries. Still, it should be noted that Skythian is
the ethnonym which has created the biggest problem for researchers because of the
excessive use of the word for all the steppe peoples. It is not easy to distinguish
between a Pecheneg or Cuman in the 12th century or a Mongol or Cuman in the 14th
century Balkans.170
Skythia is the country of Skythians. Since Herodotus, it was represented as a
realm of the wilderness without well-defined borders. It corresponds very roughly to
the Eurasian steppes. As it was stressed by François Hartog, it is “the land of ἐρημία
and ἐσχατιά,” in other words, it is the desert (and wilderness) and the edge. Σκυθῶν
ἐρημία (Skythian Desert) was an expression used for rough persons without social
relations.171 This usage is attested in Aristophanes. However, in the late 13th-early
14th century George Pakhymeres employed this term to indicate the territories ruled
by the Golden Horde. Thus, Skythia and Skythian Desert are terms indicating
unknown, quasi-mythical lands with mostly negative connotations.
A frequently attested subgroup of the Skythian is the Sauromatae. The
Sauromatae or Sarmatians were originally a population of Skythian origin dwelling
in today’s Ukraine in Late Antiquity. Their homeland, Sarmatia, was also called
Skythia Minor (Little Skythia) by the Romans. This ethnonym was later employed
for the Hungarians, Pechenegs, and Oguz.172 It has a mixed geographical and social
character. Thus, being a branch of the Skythians, it indicates a nomadic way of life.
However, it only contains the nomadic populations that appeared in the northern
170 This confusion is particularly remarkable in the work of Kinnamos, which covers the period 1118-
1176, in which Pechenegs and Cumans were present together in the Balkans. The author still insists –
nearly always- on classical ethnonyms.
171 Hartog, Le miroir d'Hérodote: Essai sur la représentation de l'autre, 31
172 Beyond the limits of the timeframe and corpus of this thesis, but it should be noted that some minor
sources of later periods employ this term also for Seljuks and Ottomans.
93
frontiers of the Byzantine Empire, i.e., the regions close to the ancient Sauromatae
homeland.
The abovementioned ethnonym Persian used for the Seljuk Turks reflects the
idea that the Seljuks took over a well-defined realm that existed since Herodotus’
times and became the new rulers of this country. Thus, this identification puts the
Seljuks on a geographical field and makes them related to a country/state that had
existed since Classical Antiquity. This could be further formulated as a translatio
imperii in which the imperium of Persia, once governed by the Medes, Achaemenids,
Parthians, and Sassanids, and even by the Arabs after the Muslim invasion of Persia,
was now taken over by the Turks. However, the ethnonym Persian is employed to
design the Great Seljuks that ruled over Persia and for the Seljuk Sultanate of Rum
based on Ikonion. Rustam Shukurov explains this situation as an exception that the
term is used for a population not based in Persia. The term Persian acquires an
ambivalent status after the Byzantine loss of Asia Minor, and it becomes a generic
term for all the Turks in the Muslim Near East.173 However, slightly before using the
term Persians for the Seljuk Turks in Asia Minor, which became a convention; an
author like Anna Komnene could carefully separate the Grand Seljuks and Seljuks of
Rum by employing the term Persian for the first and Turk for the second.174
It should be still noted that the Seljuks, even its branch in Rum, was a
Persianate state that embraced the Persian culture and used Persian as the language of
the court and bureaucracy; then self-identified with the ancient Persian heroes whose
173 Shukurov, The Byzantine Turks, 1204–1461, 43.
174 Durak, “Defining the 'Turk': Mechanisms of Establishing Contemporary Meaning in the
Archaizing Language of the Byzantines,” 59, 76. Shliakhtin calls the Byzantine use of the term
Persian for the Seljuks in the 12th century as the “persification” of the Turks in the Byzantine
discourse. Shliakhtin, From Huns into Persians: The Projected identity of the Turks in the Byzantine
Rhetoric of Eleventh and Twelfth Centuries, 57-69
94
deeds Firdausi narrated in his S̲ h̲ āhnāmah. Moreover, they used the titles and
symbols of the ancient Persian monarchy as an instrument of dynastic legitimacy.
The figure of the arch-rival Persian could also regenerate the memories of a
distant past. These memories could be expressed even as a prophecy: when
Constantine Monomachos deployed the Macedonian forces on the eastern front,
according to Skylitzes, the Turks were rumoring that they would “be overturned by a
force similar to that with which Alexander the Macedonian overturned the
Persians.”175 Gregory Palamas refers to the earthquake of Gallipoli in 1354 which
caused the Ottoman takeover of the city by saying: “This earthquake brought this city
into the hands of the Achaimenids (Ἀχαιμενίδαı) that we now call Turks.”176 This
phrase clearly draws a historical parallelism between the Persian crossing of
Dardanelles and the invasion of Thrace and the early Ottoman expansion into the
Thracian territories of Byzantium.
However, this ethnonym implies not only a geographical belonging but also a
moral connotation; as it was told in the Strategikon attributed to the emperor Maurice
(r. 582-602): “Persians are perverted, dissembling and slavish, but they love their
country and are also obedient.”177
An ethnonym closely related to the Persians, the Parthians (Πάρθοι) was also
sometimes employed to refer to the Turks. They were an Iranian nomadic ethnic
group that successfully invaded Iran in the 2nd century BC and founded there a
dynasty that ruled this country until the rise of the Sassanids.178 The dynasty that
ruled the Parthian Kingdom was the Arsacid dynasty. Niketas Khoniates presents
175 Skyliztes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 479, Skylitzes-Wortley; A Synopsis of
Byzantine History 811-1057, 447.
176 Philippides-Braat, “La captivité de Palamas chez les Turcs - Dossier et commentaire,” 138.
177 Odorico, “L'étranger et son imaginaire dans la littérature byzantine,” 66.
178 Moravcsik, Byzantinoturcica, vol. 2, 247. In Chapter 4, I shall deal with Laonikos
Khalkokondyles’ origo gentis narrative that relates the Ottoman Turks with Parthians.
95
Danishmend Gazi, the eponymous founder of the Danishmendid dynasty, as
Περσαρμένιος Ταϊσμάνιος and affiliates the Danishmendid rulers with this
dynasty.179 The Armenian branch of the dynasty, called Arshakuni, ruled this country
until the 5th century AD. Leaving apart the theory of the Armenian origin of
Danishmendids, which was claimed by the Armenian historian Matthew of
Edessa,180 who was contemporary of Danishmend Gazi, this ethnonym probably
indicated a symbolism similar to Persian, with further emphasis on nomadism.
However, the Islamic religious affiliation of the Seljuks was putting them on new
ground: on a religious identity. In fact, the Byzantines never used the word Persian
(or Median) for the Arabs. They were always Saracens (Σαρακηνοί), Agarenes
(Ἀγαρηνοί) or Ismaelites (Ἰσμαηλῖται).181 The first ethnonym’s etymology is not
clear, however it is a term used for different groups of Arabs since Late Antiquity.
Agarenes and Ismaelites are terms related to the biblical genealogies, in which the
Arabs descend from Abraham’s son Ismael, who was not born from the legitimate
wife of the biblical patriarch but her slave Agar.182Saracen is an ethnic term for the
Arabs, which is never used for the Turks, as Koray Durak brilliantly demonstrated it
in his article that challenges Moravcsik’s claim that it was a term that is used to
indicate the Turks. The sporadic use of the other two ethnonyms for the Seljuk and
Ottoman Turks is attested after the 12th century; this use seems related to an
ideological change in the motivations of the Seljuk-Byzantine conflict in Asia Minor.
179 Khoniates-Bekker, Nicetae Choniatae Historia, 27; Khoniates-Magoulias, O City of Byzantium, 12.
180 Mathieu d'Édesse, Chronique de Mathieu d'Édesse, 256: “Danishmend, grand émir du pays des
Romains, et Arménien d’origine, cessa de vivre.”
181 Durak, “Defining the 'Türk': Mechanisms of Establishing Contemporary Meaning in the
Archaizing Language of the Byzantines,” 72. See also Savvides, “Some Notes on the Terms Agarenoi,
Ismailitai and Sarakenoi in Byzantine Sources,” 89-96.
182 Savvides, “Some Notes on the Terms Agarenoi, Ismailitai and Sarakenoi in Byzantine Sources,”
90-95.
96
Still, one can speculate that there could be imagined a possible antagonism
between Persians and Skythians as the one between the Romans and Skythians
according to their ways of life. While the Skythian identity reflected
pastoral/nomadic society, Persia implies an established “imperial” order, not similar
but somewhat parallel to the Roman/Byzantine Empire. Thus, the naming change
from Skythian to Persian also indicated a passage from the nomadic federation to the
sultanate (see Table 3).
Table 3. Persians and Skythians
Persian Skythian
Sedentary Nomad
Empire Tribal Federation
Eastern Northern
Byzantine sources hardly mention any king, ruler or monarch of the peoples called
Skythian. They have archons (ἄρχων) or hegemons (ἡγεμών). It is not the result of
the mimesis of the ancient sources. Indeed, the nomadic steppe societies generally do
not have a centralized authority, yet every tribe has their begs who govern over them.
This social structure –according to Skylitzes: “they are divided into thirteen tribes all
of which have the same name in common, but each tribe has its own proper name
inherited from its own ancestor and chieftain”183- creates the appearance of a deeply
fragmented and even chaotic society in the eyes of Byzantine authors and it fits
perfectly the image of a barbarian society.
183 Skylitzes-Thurn, Ioannis Scylitzae Synopsis historiarum, 455; Skyliztes-Wortley, A synopsis of
Byzantine history, 811-1057, 426
97
Another social aspect of Skythians is whether they have a law or not. In the
eyes of Byzantines, the presence of a law (νόμος) is a sine qua non-feature of people
that makes them barbarian or non-barbarian. As it was seen in the passage from
Procopius, in which the difference between Ephalitae and other Hunnic peoples was
discussed, the author stressed the presence of law and a king (“but they are ruled by
one king, and since they possess a lawful constitution”) as the aspects that
differentiated them from the rest of the Huns. The notion of law is also related to the
affiliation to a monotheistic, i.e., non-pagan religion. In the texts studied for this
dissertation, there is a correlation between the law and monarch as the institutions of
the civilized nation.
In concluding this section, a final remark should be made on the relationship
between toponyms and ethnonyms. By definition, Skythia is the land of the
Skythians, and Tourkia is the land of the Turks. However, especially after the 12th
century, the use of the Persian ethnonym for Seljuk Turks in Byzantine texts
becomes widespread. But the Sultanate of Rum corresponds to a different place than
Persia. Central Anatolia is not Persia geographically. However, the cultural
geography perception of Byzantine intellectuals was not shaped by strict boundaries.
The Seljuks came from Iran and the geographical location of the Sultanate of Rum is
in continuity with Iran. In a sense, this overlaps with cultural memories from
Antiquity, the Achaemenid and Sassanid era, like a palimpsest. Moreover, as
Shukurov indicates, the language widely spoken in the city centers of Anatolia
during the Seljuk period was Persian.184 According to an anecdote quoted by
Shukurov from Eustathios of Thessaloniki, so many Anatolian Turkish prisoners
were settled in Thessaloniki in 1178 that this region began to be called "New
184 Shukurov, The Byzantine Turks, 40.
98
Persia".185 In short, calling the Turks in Anatolia Persians does not completely repeat
the pattern mentioned, but is understandable in terms of wider cultural geography.
3.6 Pagans or Muslims: Turks, religion, and religiosity
The Islamization of the Oghuz Turks is still a subject of scholarly debate. However,
the sources agree that the 11th century was the period when the Oghuz Turks’
conversion to Islam became a massive phenomenon.186
The formation of the Seljuk proto-state and their conversion to Islam were
probably related. Seljuk, the eponymous founder of the dynasty, was a subashi in the
court of an Oghuz or Khazar ruler that could be Jewish or pagan. The early dynastic
narrative about the formation of the Seljuk Sultanate stresses a religious conflict
between Seljuk and his ruler. This tradition echoes a possible political conflict in the
élite of the abovementioned state which results in the self-exile and later conversion
of the dissidents and their later memory of this moment as a retrospective ethos.187
The Seljuk Turks, independently from the level of their religious zeal, entered
into the classical Islamic oikoumene as the saviour and allies of the Abbasid Caliphs
who were disturbed by the military control of Iraq by Buyids. Starting with Ṭog̲ h̲ rı̊l
Beg, who was their first sultan to rule Persia and Iraq, they presented themselves as
the devout Sunnis. According to Arabic traditions in the earliest phase of the
Byzantine-Seljuk relations, in negotiation between Constantine IX Doukas and
Ṭog̲ h̲ rı̊l Beg, the latter stipulated the renovation of the mosque of Constantinople and
the nomination of the name of the Abbasid Caliph (instead of Fatimid Caliph) in the
185 Shukurov, The Byzantine Turks, 245.
186 Here, the Oghuz Turks in question were the ones who were involved in the Seljuk state apparatus.
Before the 11th century, there were earlier conversions of the Volga Bulgars and Karluks.
187 Peacock, Early Seljuq History: A New Interpretation, 16-47 for their early history, .99 ff. about the
Seljuks and religion. He argues basically that the zealous Sunni image of the Seljuks was a
mechanism of political legitimization, not the reality itself.
99
Muslim prayers in that mosque. Whether they were pious Sunnites or not, the Seljuks
paid special attention to giving the impression of their affiliation to Sunnism, in order
to enforce their authority and legitimacy in their recently-conquered territories.188
In the early Byzantine depictions of the Seljuk Turks, however, there is no
clear reference to their adherence to the Muslim faith. Furthermore, there is no
reference to a Muslim holy war that the Byzantines knew of it since the early Arab
invasions. Seljuk raids are depicted as standard looting campaigns of the nomads;
there is hardly any mention of either religious motivation or religious zeal.
In Attaleiates’ history, the author stresses the similarity and near-equality
between the lifestyle of Seljuks and Pechenegs although the former are Muslims and
the latter are still faithful to their ancient pagan beliefs despite the evangelization of
some of their leaders. According to Attaleiates, “The Skythian mercenaries,
moreover, resembled the Turks in all respects.”189
Furthermore, this resemblance could nourish Byzantine doubts on possible
defection to Seljuk’s side on the battlefield: “That same day, a band of Skythians
commanded by a certain Tamis went over to the enemy, which threw the Romans
into some real consternation because they suspected that the rest of those people,
whose way of life was so similar to that of the Turks, might join them and fight on
their side.”190 Furthermore, when he narrates the battle of Hierapolis191 (1068)
between the Byzantines and Syrian Bedouin dynasty Mirdasids that were allied to
Seljuks, he points out the Arabs’ motivation to combat by saying, “They (Saracens)
fought in their traditional way for the defense of their religion and city but were
188 Beihammer, “Orthodoxy and Religious Antagonism in Byzantine Perceptions of the Seljuk Turks
(Eleventh and Twelfth Centuries),” 15-36, here 9.
189 Attaleiates, The History, 284-285.
190 Attaleiates, The History, 286-287.
191 Today, it is the city of Manbij in Syria.
100
unable to hold out forever.”192 This particular emphasis on religiously motivated war
was never employed for Seljuks in his work.
Thus, as it was already mentioned before, the lifestyle is the strongest way
that binds the different nations (or tribes) that were members of the same genos to
each other and this affiliation to the same genos can provoke defections between
these populations during the times of war.
This passage demonstrates that although they were converted to Islam, the
general view of the Seljuk army resembles more a Pecheneg horde that practiced
Turkic paganism than the armies of the Caliphates which were present in Byzantine
historical memory, or Syrian Arab Emirates, such as the Hamdanids of Aleppo,
conflicts with whom constituted a recent layer of history.
Alexander Beihammer also stresses the similar perception of Seljuk Turks
shared by Michael Attaleiates and Armenian and Syriac writers who wrote on the
early Seljuk raids on the eastern border of the empire and explains their common
tendency to represent the Seljuk Turks as “first and foremost as fierce barbarians, not
as representatives of a new Muslim threat. We may safely assume that the Islamic
faith of the Turkish warriors that the Byzantines were confronting in the first phase
of the conquest period, at least from the perspective of outside observers, was not
very visible, nor a determining factor of their behavior.”193
This representation began to change in the 11th century. Anna Komnene
sometimes used the terms Agarenoi or Ismaelitai for the Turks. However, in general
terms, it does not differ from earlier texts, those of Skylitzes and Attaleiates, which
pre-date this period by half a century. Like many Byzantine authors, Anna Komnene
192 Attaleiates, The History, 200-201.
193 Beihammer, “Orthodoxy and Religious Antagonism,” 20-21.
101
did not have a clear idea of Islam. The Byzantine intellectuals, generally perceived
Islam as a kind of heresy. However, she also attributes to them the aspects associated
with Greco-Roman paganism. According to her, “Ismaelitai” are “slaves to
drunkenness, wine and Dionysos,” “dominated by Dionysos and Eros” and “nothing
more than slaves of the vices of Aphrodite.”194
John Kinnamos, whose history starts off where Anna Komnene’s Alexiad
ends, tends to see the Pechenegs’ lifestyle (he calls them Skythians) as the way of
life that Persians (Seljuk Turks) used to live: “Since they (Persians) were still
untrained in agricultural labours, but gulped milk and devoured meat, like Skythians
and were always uncamped in scatterings on the plain, they were ready prey to
whoever wished to attack them. Thus, the Persians had previously lived.”195 He states
the actual lifestyle of Pechenegs was a level of civilization that Seljuks recently got
through. However, this depiction says nothing about the religious beliefs or practices
of Seljuks. Kinnamos still explains the religion of Seljuks by noting its similarities to
the beliefs of other steppe people. The author mentions “Halisians” (Χαλισίοι), an
ethnoreligious group among the Christian Huns (i.e., Hungarians) who practice
“Mosaic Law,”196 and says that they agree on the same doctrine as the Persians.
According to the author Halisians’ Judaism and Seljuks,’ Islam is basically the same
194 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 298; Komnene-Sewter-Frankopan, The
Alexiad, 276. For an overview of Byzantine ideas about Islam, see Meyendorff, “Byzantine Views of
Islam,” 115-132. Meyendorff stresses that there are two Byzantine approaches on Islam, the first
views it as paganism, the second views it as a heresy that shares the basic Monotheist aspects of
Christianity. It seems that the latter view is much more common.
195Kinnamos-Meineke, Rerum ab Ioanne et Alexio Comnenis Gestarum, 9, Kinnamos-Brand, Deeds of
John and Manuel Komnenos, 17. I slightly changed Brand’s translation because the translator
employs contemporary ethnic and geographical terms and sometimes loses the nuances between the
notions.
196 Kinnamos-Meineke, Rerum ab Ioanne et Alexio (sic) Comnenis Gestarum, 10; Kinnamos-Brand,
Deeds of John and Manuel Komnenos, 86,
102
belief, and it is a heterodoxy.197 However, Kinnamos was also aware that the
Caliphate was a source of spiritual authority for the Muslim world, which also covers
the Sultanate of Rum. He notes that in the correspondence between Manuel I and
Qiliç Arslān II, when the former asks the cause of the latter’s infidelity to their treaty
and hostile military actions against Byzantium, the sultan responds by saying that the
Caliph (Kinnamos defines him as μέγας ἀρχιερεύς, the high priest) was angered to
him because of his amicable relations with Byzantium.198 This reference
demonstrates that the religious antagonism between the Seljuks and Byzantines
resulted from an exterior factor, rather their own zeal.
To see a more obvious and religiously antagonistic depiction of Islam as the
religion of the Anatolian Turks, one must wait until the 14th century; first, the letters
of the captivity of Gregory Palamas, then the Historia Turco-Byzantina of Doukas
provide us with some more detail about the evolution of this image. Doukas’ history
is the text that emphasizes most on the religion’s role in the Turkish military
expansion. Particularly before the 14th century, there was no significant reference to
the visible symbols of Islam. No Byzantine authors mention Islamic rituals; neither is
there any reference to the segregation of sexes which was frequent in classical
Muslim societies.
Gregory Palamas, who was a prominent theologian himself, after falling
captive to Turkish pirates in the Aegean Sea, as was narrated in chapter 1,
experienced a short-time stay at the Ottoman court. There he first meets Ismail, the
grandson of Ork̲ h̲ an, a young Ottoman prince who seems quite interested in
197 Kinnamos-Meineke, Rerum ab Ioanne et Alexio (sic) Comnenis Gestarum, 247; Kinnamos-Brand,
Deeds of John and Manuel Comnenus, 186 “και εισιν ἑτερὁδοξοı, καθἁπερ ᾒδή ἔφην, πἑρσαıς
ταυτοφρονοῦντνεσ”.
198 Kinnamos-Meineke, Rerum ab Ioanne et Alexio (sic) Comnenis Gestarum, 289, Kinnamos-Brand,
Deeds of John and Manuel Comnenus 216.
103
theological discussion. Then, he was brought to Nicea and engaged in theological
debates with the Danişmends and a group of enigmatic religious men called
Χιόναι199 whom the author presented as “the men who studied and learned nothing
else than blasphemy and impudence against Jesus-Christ by (their teacher) Satan.”
He carefully differentiates the Χιόναι and the Danişmends who seem to be Muslim
scholars. He knows that the Islamic faith is a branch of Abrahamic religions and –
although he never uses this word- he sees it as a heresy created by a false prophet
whom Christians cannot accept. Palamas never uses the word “Agarene” or
“Musulman” but constructs the religious antagonism between the Christians and
Turks.
Furthermore the author’s opinion about the Xionai is more negative than his
opinion about the Turks: “After what I heard about them (Xionai) and what they said,
they are obviously Hebrews and me, under these circumstances I do not speak to
Hebrews.” Thus, Palamas’ anti-Semitism is stronger than his defiance against the
Turks.200
Doukas, who wrote roughly in the 1460s, stresses the Muslim identity of the
Ottomans. When he describes the Ottoman sultan Bāyazīd I, one of the villains of his
History, he presents him as “a feared man, precipitate in deeds of war, a persecutor
of Christians as no other around him, and in the religion of the Arabs a most ardent
disciple of Muhammed, whose unlawful commandments were observed to the
199 The identity of Χιόναι is still a matter of discussion. G. G. Arnakis identified this word with the
Turkish word “ahi” and Paul Wittek with “hoca;” however, both J. Meyendorff and G. M. Prohorov
stressed that this group is not considered really Muslim by Palamas and must be a heretical Jewish
sect. Anna Phillipides-Braat interprets other sparse references to this group and defines them as a
Jewish group who converted to Islam. According to Michel Balivet, this word is the Greek rendering
of kühhan (the plural of kâhin=oracle): Balivet, “Byzantins judaïsants et Juifs islamisés. Des ‘Kühhân’
(Kâhin) aux ‘Xiónai’ (Xiónos),” 24-59. For an overview of the discussion, see Phillippides-Braat, “La
captivité de Palamas chez les Turcs: Dossier et commentaire,” 214-218.
200 Phillipides-Braat, “La captivité de Palamas chez les Turcs - Dossier et commentaire,” 170.
104
utmost, never sleeping, spending his nights contriving intrigues and machinations
against the rational flock of Christ.”201
This description, which re-elaborates Procopius’ portrait of Justinian II in his
Historia Arcana, has remarkable details to interpret. Firstly, Doukas does not build
up the antagonism on the duality between the barbarians and Greeks. However, he
stresses the persecution of Christians (similar to the persecutions of Christians in the
hands of the pagan Roman emperors). Secondly, although the sultan is “the ruler of
the Turks” (ἀρχηγὸς τῶν Τούρκων), he practices the “cult of the Arabs” (τῶν
Ἀράβων θρησκεία). Thus, Islam is the religion of the Arabs, which –in its deepest
roots– is identified with another population that lives in the countries very far from
the vanishing Byzantine world. However, the Ottoman Turks appear as the
spearheads of this faith, which is still the “cult of the Arabs.” Thus, Doukas clearly
distinguishes religion and ethnicity.
Khalkokondyles, who wrote roughly in the same decade as Doukas, gives
even a brief digression of early Islamic history. It seems that, despite sometimes
making factual errors on the subject, he has satisfactory knowledge of the tenets of
Islam and the geography of the Arabian Peninsula. 202 However, Khalkokondyles
employs a much milder language both on the religion, and the empire of the
Ottomans. He was a student of the Athenian philosopher Plethon and interested much
more in the antiquity of Greeks than Byzantine and Christian culture. His history was
based on the Herodotean model, and the history of the Ottomans covers the central
part of the text, just like the Persians in the History of Herodotus. One can
furthermore dare to say that as an early Greek example of renaissance,
201 Doukas-Grecu, Historia Turcobyzantina, 39; Doukas-Magoulias, Decline and Fall of Byzantium to
the Ottoman Turks, 62.
202 Khalkokondyles The Histories, 193-205.
105
Khalkokondyles’ approach to the Ottoman world was beyond the antagonisms of the
Byzantine worldview.
The following passage in the work of Sphrantzes also merits attention:
Lord Manuel was greatly admired by the Anatolian Turks in the retinue of
Prince Mustafa, who thought that in appearance he resembled Mohammed,
the founder of their faith. Bayazid, Manuel’s enemy, had once remarked that
even if one did not know the emperor, Manuel’s appearance would make:
“This man must be emperor.”203
This passage sheds light on the 15th-century Turks’ imagination of the prophet
Muhammad. Sphrantzes clearly mentions the attitudes of “Anatolian Turks in the
retinue of prince Mustafa,” so these soldiers are not janissaries or soldiers of Greek
or other origins that recently converted to Islam who still had respect for the
Byzantine emperor. However, the association of a Christian ruler with the prophet,
even for his physical characteristics, is unusual. The passage probably demonstrates
that these “Anatolian Turks” lacked religious fanaticism and had a fluid religious
approach.
In brief, the Turks’ affiliation with the Islamic faith becomes more visible,
especially after the 14th century and this fact could be followed closely in Byzantine
historiography dealing with the Turks. It is obviously related to a deeper
appropriation of the institutions and rituals identified with Islam by the Turks.
However, the evolution of the Byzantine-Seljuk conflict must not be understood
independently from the internal evolution of the Seljuk polity, which once had a
more “Hunnic” or “Skythian” structure, at least in the eyes of the Byzantine authors,
but later had a more “Persian” administration; I use the term to describe a wellestablished,
Islamic, Persianate court organization. Thus, as I already mentioned, the
203 Sphrantzes-Maisano, Georgii Sphrantzae Chronicon, 22; Sphrantzes-Philippides, The Fall of the
Byzantine Empire. A Chronicle by George Sphrantzes 1401-1477, 28.
106
ethnonym Persian hardly indicates any ethnicity but a geographical and sociocultural
association with the country, culture, and political heritage of Persia. The
12th-century poet John Tzetzes (d. 1180) wrote a poem in which he employed phrases
from seven different languages, which also included verses in Persian (Old Anatolian
Turkish) and Skythian (Cuman or Pecheneg), employing expressions in the Turkish
vernacular, not in Persian itself.204 However, it should be noted that the Byzantines
were never too interested in others’ religions. They were not interested in an
ethnographic depiction of the others’ religion. They hardly give any hint on a
particular religious practice or a belief of a foreign nation. This fact does not permit
us to understand or to reconstruct the real nature of the “idolatry” and “polytheism”
of the barbarians. However, they were much more aware of Islam. Although they
generally regarded it as a kind of paganism, it could be used as an ex nihilo argument
to the religious character of the 11th century Turks. However, as it was pointed out by
Shukurov, in the Byzantine canonical literature, the epithets “heathen” and
“Hagarene” were very often used together. The fact that the Turks were Muslim or
pagan did not mean much to the Byzantines. They were non-Christian barbarians,
they adhered to a “barbarian doctorine.”205
Now, a question that should be asked is, under these circumstances, how did
religion differentiate the Seljuk Turks in Anatolia from Turkic pagan nomads in the
Balkans in the eyes of Byzantine authors?
204 The abovementioned verses of Tzetzes’ poem are cited in Moravcsik, Byzantinoturcica, vol. 2, 19.
A new interpretation is offered in Shukurov, The Byzantine Turks, 1204-1461, 49-51.
205 Philippides-Braat, “La captivité de Palamas chez les Turcs: Dossier et commentaire,” 140.
107
3.7 Mongols and Timurids in Byzantine sources
The Mongols were a nomadic ethnic group that, until their unexpected expansion in
the 13th century, lived outside the borders of Byzantine geographical knowledge. So
until the 13th century, there was no reference to Mongols in Byzantine sources. The
Mongol invasion of Western Asia was one of the key events that determined the
human landscape of the whole region, particularly Asia Minor. It triggered new
waves of Turkic migration to Anatolia. The decline of the Seljuk Sultanate of Rum
after the defeat at the battle of Kösedağ (1243) at the hands of the Mongols also
loosened the sultanate’s authority on the various Türkmen groups in the southern and
western edges of Turco-Muslim Anatolia and permitted the corrosive looting of
Byzantine territories by the Türkmen.
The Mongol monarchy of Persia, called the Ilkhanate, was the suzerain of the
Seljuks of Rum. Their early rulers Hulagu (r. 1256-1265), Abaqa (r. 1265-1282),
Arghun (r. 1284-1291), and Gaykhatu (r. 1291-1295) were not Muslim and they had
somewhat revoked the traditional privileges of Sunni Muslims in Iraq and Persia.
Furthermore, there were also Mongols in the Balkans, Eastern Europe, and Crimea.
This subdivision of the Mongol Empire was known as Ḏj̲
oči Ulus by their
contemporaries and today it is commonly called the Golden Horde. The authority of
Mongols in the northern Balkans was also very strong. The principalities of
Wallachia and Moldavia and the Kingdom of Bulgaria were the tributary states of the
Golden Horde. Even a Mongol prince of the imperial dynasty of Borjigin, Chaka
(r. 1299-1300), once became the king of Bulgaria. In the following chapters, I shall
also address in detail the close relationship of Kaykāʾūs II, Sultan of Rum in exile,
with the khans of the Golden Horde. The Golden Horde was fragmented as a result
of its internal crises in the mid-15th century, and there formed several entities that
108
claim the legacy of this state. The Khanates of Crimea and Kazan were the most
important successors of the Golden Horde. Apart from Ḏj̲
oči Ulus and Ilkhanate, the
other successor states of the Mongol Empire were the Khanate of Chagatai in
Transoxiana and the Yuan Empire in China. In the early 14th century, all of these
monarchies were still ruled by the khans of Genghisid descent.
Secondly, similar to the western expectations about the future role of the
Mongols as a force that balanced the Muslims in the east, also the Byzantines
expected that the Mongols could be appropriate allies against the Muslims. The early
Mongol rulers’ rather sympathetic approach to Christianity probably contributed to
these hopes. Some Mongol (or Turco-Mongol) nomadic groups in Central Asia
adopted Nestorian Christianity in the 12th and 13th centuries. Sartaq Khan, the second
ruler of the Golden Horde (r. 1256-1257), was a staunch follower of Orthodox
Christianity. However, the Ilkhanate rulers subsequently converted to Islam and the
country's native Persian culture took over Mongolian institutions.
The first news about the Mongols appears in the work of Akropolites: he uses
the ethnonym Ταχάρıοı, a word that echoes both Τόχαροι and Τάταροι, Tocharians,
and Tatars (see Table 4). The Tocharians are an ancient Indo-European people whose
name is attested in ancient geographical sources. They were living in the Tarim basin
and the cities along the Silk Road, until their assimilation by the Turkic nomads,
particularly by Uyghurs. Tatar is the name of a Mongolian tribe of the Pre-Genghisid
era. The use of the ethnonym Tocharian demonstrates the classicizing approach, yet
the word Tatar is probably a Turkish loanword in Byzantine Greek.
The second Byzantine text dealing with the Mongols is the work of
Pakhymeres. Pakhymeres wrote his text in a period where the Byzantine-Mongol
alliance seems to be a possible survival strategy. Pakhymeres explains the emperor
109
Michael VIII’s strategy regarding the Mongols with two principles: use the
“Persians” (i.e. Seljuk Turks) as a barrier against the Mongols and conclude a
marriage alliance with the latter. The emperor succeeds in these two policies: the
Turks in Western Anatolia constitutes a buffer zone between the Byzantines and
Mongols in Asia Minor and he sends his illegitimate daughter Maria Palaiologina to
the court of Ilkhanate as a bride. The princess marries with Abaqa Khan.206 Then the
Byzantine Empire makes a second marriage alliances with the Mongols, this time
with the Golden Horde by marrying another illegitimate daughter, Euphrosyne with
the de-facto ruler of this polity, Nogai. However, this policy of rapprochement did
not lessen the Byzantine people’s fear of the Mongols. The ordinary Byzantines do
not see the Mongols as an ally. The rumors of a Mongol raid to Nicaea creates a
terrible wave of panic in the Bithynian city.
However, when Pakhymeres evaluates the history of the Mongols in a
retrospective way, he could interpret their past using historical concepts and figures
familiar to Byzantine intellectuals: Their first lawgiver (νομοθέτης) was neither
Solon, Lycurgus or Draco, because the Mongols are not Athenians or
Lacedaemonians. Despite they are brave at war, they are still the savages and they
live in a barbarian way. However, their lawgiver Genghis Khan -the author stresses
that he remembers his name- commanded them justice and truth, and they are faithful
to his legacy and they live without intrigues and deception.207
The Mongols were referred to as Μουγούλιοι, Τόχαροι and Ἄταροι (probably
a misspelling of Τάταροι, Tatars) in Pakhymeres (see Table 4). The use of this
ethnonym demonstrates the ideological function associated with the Northern Turkic
206 Pachymeres, Relations Historiques, 234-235.
207 Pachymeres, Relations Historiques, 444-447.
110
nomads passed to Mongols. Finally, Pakhymeres knows well the internal divisions of
the Mongol Empire, and refers the Ilkhanate as the Eastern Tocharians (Ἀνατολικοί
Τόχαροι ) and the Golden Horde as Northern Tocharians (βόρειοι Τόχαροι).
Pakhymeres stresses on the unclean nutrition habits of the Tatars, similar to the
description of Pechenegs by Attaleiates.
Gregoras gives an ethnographic digression about the Mongols whom he calls
Skythians. This digression is based completely on ancient literature. He states that
the Skythians are originally from the furthest northern parts of the world and they do
not eat anything but meat, blood, and milk. He identifies them with the Cimmerians,
Cimbri, and Teutons which are mentioned by the ancient authors. The Cimmerians
are an ancient Iranian people, and the Cimbri and Teutons are Germanic tribes that
fought the Romans in the first century B.C. However, the author speaks without
precision, he complains that the ancient authors gave them Greek names and
arbitrarily used them.208 He then gives an account of the 13th-century Mongol
conquests, mentions Σιτζıσχᾶν (Genghis Khan) and his sons Χαλαοῦ (Hulagu) and
Τελεπουγᾶς (Talabuga) and their victories against other Skythians, Huns, and
Cumans in Central Asia.209
His comment about the acculturation of the Mongols in the territories they
conquered also merits attention: “Later they abandoned the irreligion they inherited
from their ancestors and converted to the religion of Assyrians, Persians, and
Chaldeans. And they adopted also the luxuries of clothing, food and drink and other
aspects of these people’s lifestyles and customs.”210 This approach is not only an
208 Gregoras-Schopen-Bekker, Nicephori Gregorae Byzantina historia, 32 and Gregoras-Van Dieten,
Rhomäische Geschichte,78-79.
209 Gregoras-Schopen-Bekker, Nicephori Gregorae Byzantina historia, 35-36 and Gregoras-Van
Dieten, Rhomäische Geschichte,80-81.
210 Gregoras-Schopen-Bekker, Nicephori Gregorae Byzantina historia, 40 and Gregoras Van Dieten,
Rhomäische Geschichte, 83.
111
interpretation of the cultural and administrative “Persianization” of the Ilkhanids but
also the reflection of the Byzantine mentality about the acculturation of “barbarian”
populations: their conversion to monotheism and adoption of the various aspects of
the culture of the population of the country where they settled.
Kantakouzenos’ observations about the Mongols were shorter and because of
the work’s date of composition, they were generally limited with the Golden Horde
Tatars whom the author defines as the Skythians from Hyperborea. By the middle
14th century, when Kantakouzenos wrote his work, the might of the Mongols in the
Balkans was somewhat waned. He portrays the them as mostly the raiders that
devastate the imperial territories in Thrace.
In Doukas and Khalkokondyles’ works, there is information about Tamerlane
(r. 1370-1405) and the state he founded. This polity, centered in Transoxiana, could
be considered a successor of the Mongol Empire in Persia and Central Asia.
Tamerlane, being a Turco-Mongol warlord from the Barlas tribe, used the nominal
suzerainty of Khanate of Chaghatai in Central Asia for his legitimacy. As an
aggressive conqueror, he engaged in war with the Ottomans in the early 15th century.
He defeated the Ottomans in 1402 at the battle of Ankara and triggered a period of
dynastic struggles of the Ottoman State, commonly known as the Ottoman
Interregnum (1402-1413). Doukas refers to him as the “sultan of Persia and
Babylon” and his “nation” as the Skythians. But he occasionally uses adjective
Persian to describe Timur’s army.211 He mentions the Tatars of Crimea once, as the
Tauro-Skythians. It must be said that, despite the semantic function of the ethnonym
Skythian, the identification of Timur as the “sultan of Persia and Babylon” directly
211 Doukas-Grecu, Historia Turcobyzantina, 87, 99,101; Doukas-Magoulias, Decline and Fall of
Byzantium to the Ottoman Turks, 88, 94, 95
112
evokes the Persian alterity and a cultural memory going back to the time of the
Abbasids. However, in Doukas’ text, Timurids do not represent the primary
antagonism. The primary antagonism is reserved for the Ottomans.
In Khalkokondyles’ Histories, Tamerlane was occasionally presented as the
“king who subjugated Asia” or “king of Samarkand” and his ethnic origin is defined
as “from the race of Massagetae”. However, the author uses neither the ethnonym
Skythian, nor Persian to refer to the polity of Tamerlane. Khalkokondyles is wellinformed
about the geopolitics of post-Genghisid Eurasia. He has a noteworthy
knowledge about the Golden Horde which he calls simply Horde (Οὐρδάς) by using
the original Turco-Mongol term. He also knows that it is a political, not ethnic term
and he gives his Greek translation as “ἀγορά”, i.e. an assembly.212 Furthermore, he
defines the population of the Golden Horde as the “Skythians of the assembly”. He
describes them as follows:
The rest of the Skythians are united and are ruled by one king; they have their
court at the so-called assembly of the Horde; and they appoint as their king a
member of the most ancient royal family. There is a branch of them elsewhere
in Europe, toward the [Crimean] Bosporos; it is quite large and they
are dispersed throughout the land, subject to a king from the royal family,
whose name is Ἀτζικερίης.213
The “branch” referred to here is obviously the Khanate of Crimea. However,
although he does not give the information at a similar level, he also mentions another
branch of this population, namely Chaghatai.214
212 Khalkokondyles, The Histories, 206-207.
213 Khalkokondyles, The Histories, 210-211. Ἀτζικερίης is Ḥād̲j̲
d̲j̲
ī I Girāy (r. 1441-1466) of Crimea.
214 In the Histories of Khalkokondyles, the Greek rendering of the name Chaghatai is Τζαχατάἴδες (as
nominative plural). However, he uses the word once as Σαχαταῖοι; in the phrase Σαχαταῖοι
ἐκλήθησαν, ὑπὲρ τὴν τῶν Περσῶν χώραν ἐς τοὺς Σάκας τε καί Καδουσίους. So it seems that
Khalkokondyles links the ancient name Σάκαι that refers to a group of Skythians with the
contemporary term Chaghadai that has nothing to do with this ancient ethnonym. The word Chaghadai
comes from a personal name, one of the sons of Genghis Khan. Khalkokondyles, The Histories, 208-
211, 236-237.
113
Khalkokondyles’ ethnographic observations, clearly based on ancient literary
material, repeat the description of the Skythians in older texts such as Attaleiates. His
Skythians of Bosporos, indeed, are not very different from Attaleiates’ Pechenegs:
they dwell in wagons and they do not eat wheat or barley, but millet and rye.
However, Khalkokondyles adds three other elements to his description: they wear
linen clothes and they are rich in gold. So, Khalkokondyles stresses the impact of
luxury, just like Gregoras. Finally, another phrase by the Athenian author merits
attention. He says: “They use bows, barbarian swords, and shields like those of the
Dacians. They usually wear felt hats, but not like those who live around Sarmatia,
nor garments made of wool because they do not use linen.”215
In conclusion, the Byzantine representation of the Mongols is far from
negative. In Akropolites’ narrative the Mongols only had a marginal role.
Pakhymeres, despite his evoking the legends about cannibalism among the Mongols,
uses a balanced language regarding the Mongols, also because of the political context
of the early 14th century: He was writing in a period in which the Byzantine Empire
sought an alliance with the Mongols, against the Turks in Asia Minor and the Serbian
and Bulgarian Kingdoms in the Balkans.216 Gregoras and Kantakouzenos did not
have the same approach, because in the mid-14th century, the Mongols constituted
neither a threat, nor the potential of an important ally any more. Doukas was not
interested in the Mongols, except the Anatolian campaign of Tamerlane which
triggered the Ottoman interregnum. Finally, Khalkokondyles’ interest in the
contemporary Eurasian world can be interpreted regarding two factors: Firstly, the
author takes the historiographical model of Herodotus as an example to his Histories.
215 Khalkokondyles,The Histories, 222-223. I slightly changed Kaldellis’ translation by replacing the
words Wallachians and Russia with Dacians and Sarmatia.
216 Pachymeres, Relations historiques, vol. 1, 186-187.
114
Secondly, the Crimean Khanate formed as a successor state of the Golden Horde in
1441 as its territory consisted of the Crimean peninsula. This new polity lacked the
greater human resources of the Golden Horde and it had to rely on stronger alliances.
After the death of Ḥād̲j̲
d̲j̲
ī I Girāy (1466), the Crimean princes engaged in interal
struggles to seize the throne. Subsequently, the Ottomans intervened there to put
Mengli Girāy I on the throne. This intervention made the Crimean Khanate a vassal
or a protectorate of the Ottoman Empire. Hence Khalkokondyles’ interest in Crimea,
probably demonstrates the Ottoman political projects toward this country in the mid-
15th century. It may well be that the author was not far from the circles where these
issues were discussed.
Table 4. Ethnonyms Used for Mongols
Author End date of the text Ethnonym used for
Mongols
Akropolites 1261 Ταχάρıοı
Pakhymeres 1308 Tocharians, Mongolians
Gregoras 1359 Skythians
Kantakouzenos 1465 Skythians
Doukas c. 1463 Skythians, Tatars (for
Timurids)
Tauroskythians (for
Golden Horde)
Khalkokondyles c. 1469 Skythians (for Golden
Horde/Khanate of
Crimea) Massagetae (for
Timurids)
115
In conclusion, this chapter presented the formation of the grand categories to
understand the configuration of the Turks in the Byzantine mentality, and my
findings can be summarized as follows:
i. When the Seljuk Turks first appeared on the eastern borders of the empire in
the mid-11th century, their image was indistinguishable from any other
nomadic invader who created troubles in the Balkan frontier. They were
pastoral nomads that belonged to the genos Hun/Skythian. This genos is
considered the most representative example of the barbarian in classical and
Byzantine literary traditions.
ii. The main branch of the Seljuk dynasty conquered Persia in the mid-11th
century and established a Persianate sultanate there. This context triggered
the use of the ethnonym Persian for the Great Seljuks and then for the Seljuk
Sultanate of Rum. Persian was an ethnonym that had a completely territorial
character. The Seljuks acquired this ethnonym due to a translatio imperii
after their conquest of Persia. Byzantine authors were aware of the dynastic
ties between the Seljuks of Rum and Persia.
iii. In the eyes of the Byzantine authors, visible aspects of the religious affiliation
of the early Seljuks were indistinguishable from the other Turkic populations
that remained faithful to their ancestral beliefs. However, in the course of
time, the Seljuk-Byzantine conflict acquired a more religious outlook. This
fact could be explained both by the ideological evolution of the Byzantine
Empire under the Komnenoi and a possible stronger self-identification of the
116
Seljuks with the Muslim faith. Starting in the 14th century, with the rise of the
Ottomans, the religious element became a sine qua non element of Turkish
identity in the eyes of the Byzantine authors.
iv. Finally, the Seljuks had a particular ideological function for the Byzantines.
Since they were settled in Central Anatolia, they started to play a double
ideological role. They substituted both the ancient Muslim enemy in the
borderlands and the ancient Persian rival of the Byzantine Empire in the
centuries before the arrival of Islam. The Ottomans represent somewhat the
continuation of this image. They are still occasionally called Persians.
However, the relationship with the Byzantine literati, as it can be understood
from what the authors wrote, suggests a completely new situation. There is no
more the antagonism of two ancient rivals of somewhat equal strength;
instead, a bygone world has been swallowed by a new Empire that has taken
the place of the old one.
117
CHAPTER 4
BYZANTINE NARRATIVES ABOUT THE ORIGINS OF THE TURKS
4.1 The genre of origo gentis: An overview
In this chapter, I will analyze several aspects of the representation of the Turks in
Byzantine sources. Firstly I shall focus on origo gentis narratives about the Seljuk
and Ottoman Turks. Then I shall discuss the narratives about two aspects of Turkic
polities in Byzantine texts: the social base of the ruling classes and the notion of
dynasty, dynastic patrimony, and ruler succession among the states ruled by Turkic
populations.
To begin, what is an origo gentis narrative? Origo gentis means “origin of
people,” so it is a historical/meta-historical text that deals with the alleged origin of a
people. This origin narrative could be based on legends or could be entirely fictional.
It serves basically to introduce a recently appeared population into an established
historiographical tradition.
This genre has long existed as a sub-element of ethnographical texts since
Antiquity. Starting with the works of Julius Caesar dealing with the populations and
tribes of Gallia, a visible interest in the traditions of autochthonous peoples who were
subordinated under the interpretatio romana,217 and simultaneous to their
subordination to the victorious Roman legions, was occasioned. Tacitus’ monumental
work De origine et situ Germanorum (commonly known as Germania), which deals
with the ethnography of the Germanic tribes dwelling beyond the limes of the
empire, is another key text that has many themes in common with origo gentis texts.
217 Interpretatio romana means “Roman translation” and indicates the identification of foreign gods
with Roman deities.
118
By the fourth century, new populations appeared at the eastern frontier of the
Roman world. These new invading hordes were Germanic (as Vandals or Goths),
Iranian (as Sarmatians and Alans) or, at least, predominantly Altaic (like the Huns).
These populations contributed to an important change in the ethnic and linguistic
landscape of Europe in the subsequent centuries. This fact caused a new necessity to
produce origo gentis texts in order to demonstrate and locate the primordial history
of such peoples in a long-established Greco-Roman world. Consequently, Late
Antiquity was a period of revival for this genre. Beginning with Cassiodorus (in the
6th century), who wrote an Origo Gothica on the origins of Goths, in Western
Europe several texts were produced to fulfill this aim. Here I mention the Getica of
Iordanes (which deals not with the Getae or the Gets, who were a Daco-Thracian
folk of Antiquity, but with the Goths), Gesta Saxonica of Saxo Grammaticus, and
Res Gestae Saxonicae of Widukind.218
Cassiodorus is considered the foremost writer who redefined the rules of the
genre. Being a Roman by birth, his Origo Gothica in Latin provided legitimization to
the Gothic kings of Italy in the 6th century. In his text, he even attributes a phrase to
the child king Athalaric (r. 526-534) that “he (Cassiodorus) raised (the narration of)
the Gothic origins to the rank of Roman History.”219
Prior to the 6th century, all the narratives of origines gentium were exclusively
written from a civilized point of view, so in these accounts there is always an obvious
opposition between the Roman and the barbarian. The Roman world occupies the
civilized part of the earth, whereas the world of barbarians is a realm of chaos
without a history. However, the revival with Cassiodorus was the result of a new
218 Wolfram, “Le genre de l’Origo gentis,” 789-791.
219 Wolfram, “Le genre de l’Origo gentis,” 791.
119
conjuncture in which the Roman Empire fell and the barbarians established their new
kingdoms on the ruins of the Empire. As it was mentioned before, Cassiodorus
himself was the subject of the Gothic Kingdom of Italy. Although the victorious and
conquering barbarian societies were reluctant to a métissage with the remnant
population of the Empire, still there was the need for a compromise between the two
worlds.
These texts, being written generally in Latin, contain also fragments of
barbarians’ own traditions regarding their legendary history, but these traditions are
generally reworked. Herwig Wolfram, in his article “Le genre de l’Origo gentis,”
identifies three faits primordials as the events worthy of remembering which are
regular components of an origo gentis narrative. These are:
i) Military victory against strong enemies
ii) The crossing of a river
iii) Conversion to a new religion220
These events (together or separately) fulfill the necessary condition for the
legitimization of a core of traditions and create a new gens by providing the
attraction of new elements to them. Military victory could be interpreted as the
founding event which gives enormous prestige to the groups who were on the
winning side. The defeated enemies, just like the Huns in the aftermath of the Battle
of Nedao (454) or the Alemanni in the Battle of Tolbiac (496), could easily come to a
point of dissolution. In such a case, many tribes that had once belonged to the
defeated party could easily pass to the winning side.
River crossing could be a symbolic narrative of massive immigration in a
world where rivers constitute the visible geographic borders between the worlds. For
220 Wolfram, “Le genre de l’Origo gentis,” 800.
120
example, the rivers cited in the origines gentium narratives about the Goths,
Lombards, and Saxons –such as the Danube, Rhine, and Elbe– marked different
levels of frontiers between the Roman and barbarian worlds. This recurring theme
has obviously biblical origins: the crossing of the Red Sea by Moses and the Jordan
River by Joshua. It can be furthermore assumed that this act reflects both a political
and a spiritual meaning: the exodus from Egypt and the arrival to the land of Canaan
are the stages of the formation of a common identity among the Israelites. So, the
themes of crossing the sea/river, the military successes against the Canaanites, and
the Mosaic covenant constitute the origo gentis story of the Israelites. It can be
assumed that in the origo gentis stories of Christian writers, these biblical narratives
were always present as a model. The biblical inspiration in the Frankish and
Lombard origo gentis accounts, written by Gregory of Tours and Paul the Deacon
respectively, was underlined by Walter Goffart.221
Lastly, converting to a new religion, always Christianity in this case,
represents a new step into the civilization or even the entrance into a new symbolic
cosmos where new values prevail. This fact also indicates a break with legendary
time or the heroic age. Thus, conversion to Christianity constitutes a chronological
milestone between real history and legendary history.
There are other typical elements in the origo gentis narratives such as the
deeds of brave men, the existence of social mobility in the population with which the
text deals, and the lack of obstacles in the warrior career for those that have different
ethnic or social origins. These elements could be interpreted as the motifs that
demonstrate the basic difference between the well-established hierarchies of the
Greco-Roman world and the bellicose nature of the barbarian tribal populations, in
221 Goffart, The Narrators of Barbarian History (A.D. 550-800), 220, 380.
121
which social mobility was relatively easy and closely related to success on the
battlefield. Therefore, such societies permited men from humble social origins to
obtain high positions because of their distinctions in war.
Byzantine history followed a different path. In the eastern part of the empire,
both the institutions and military power were more resistant to the waves of
invasions. Throughout the centuries, Byzantium suffered similar invasions by the
Slavs, Arabs, and Bulgarians, but the Empire managed to preserve the Byzantine core
lands from the invaders. In Byzantine literature, there are partial and generally short
narratives dealing with these populations that could be considered their origo gentis
accounts. Digressions from some of these narratives survived in De Administrando
Imperio, such as the chapters dealing with the history of Dalmatia, the early history
of Serbs and Croats, and their conversion to Christianity, all of which demonstrate
the characteristics of an origo gentis narrative.222 It is true also for the account of the
origins of the Turks (Hungarians) in the same work.223 There is also a similar
narrative about the origins of the Bulgarians, which survived in the chronicle of
Theophanes the Confessor and in the breviarium of patriarch Nikephoros I.224
However, the loss of Asia Minor to the Seljuks was a trauma that cannot be
compared with these earlier experiences.
Secondly, as it was pointed out by Christopher Mallan, there is a strong need
to assimilate the new populations into the familiar Greco-Roman worldview. So
reading the Byzantine historiographical texts, it should be remarked carefully on
each passage that relates the deeds of a Turkic ruler with a person of Antiquity. In
that case, in the episode between an unnamed Seljuk sultan and the Georgian
222 Constantine Porphyrogenitus, De Administrando Imperio, 138-165.
223 Constantine Porphyrogenitus, De Administrando Imperio, 170-174.
224 Todorov “Byzantine Myths of Origins and Their Function,” 66.
122
commander Liparites, in which the clemency of the sultan is underlined, Attaleiates’
parallelism between the sultan and Alexander the Great is crucial. On the other hand,
this fact could be thought of as a reflection of the author’s criticism of the Byzantine
emperor, notably against Constantine IX Monomachos. However, Attaleiates deals
with many other aspects of the “barbarian” invasions, particularly those by the
Seljuks. So this passage has an introductory function to upcoming events.225
4.2 Origo gentis narratives about the Seljuks
In this section, I will analyze two Byzantine origo gentis narratives about the Seljuks,
both written in the second half of the 11th century. The first narrative constitutes a
part of the account of John Skylitzes about the reign of Constantine IX Monomachos
(r. 1042-1055). This period coincided with the earliest Seljuk raids that targeted the
oriental frontiers of the Empire, resulting in the devastation of the easternmost theme
of the Empire, the katepanate of Vaspurakan. Skylitzes gives a rather detailed
narrative of Seljuk origins as a historical introduction to the appearance of the
Seljuks on the borders of Iberia:
I will now explain who the Turks are and how they came to fight against the
Romans. The Turkish people are Hunnic by race, living to the north of the
Caucasus mountains, populous and autonomous, never enslaved by any
nation.226
This introduction is the typical beginning of an origo gentis narrative. The author
first gives the context of his narrative, then defines the race of the Turks (Hunnic),
their ancestral homeland (a region north of the Caucasus mountains), and their
national characteristics (populous, autonomous, never enslaved by any nation).
225 Mallan, “A Turkish Alexander? Michael Attaleiates, Porus, and Alexander the Great,”.106–107.
226 Skylitzes-Thurn, Corpus Fontium Historiae Byzantinae, 442; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 416.
123
As already seen in the previous chapter, there is a common consensus on the
idea that the Turks were included in the Hunnic genos. Skylitzes’ reference to an
ancestral land “north of the Caucasus mountains” is worthy of attention. In the 6th
century, there are numerous references both in Byzantine (Procopius, Theophanes,
Malalas) and Armenian texts to a Hunnic population living in the same area. Peter
Golden cites an Armenian text (Pseudo Moses-Xorenac’i) which states that “North of
Darbant is the Kingdom of the Huns.” Other oriental sources of the 6th and 7th
centuries give us some more information about this population. A Syriac source gives
notice of a successful Armenian missionary effort among the North Caucasian Huns
in 535 or 537. Furthermore, the narrative of a bishop from Caucasian Armenia,
preserved in the chronicle of Moses Dashuranc’i, describes the Huns of North
Caucasia as a barbarian population that practiced paganism. These Huns sacrificed
horses to a “gigantic savage monster,” which they called “Tangri Xan”. So the author
probably deals with a Turkic population who were remnants of the once glorious
nomadic federation of Attila.227 The choice of “north of Caucasia” as the homeland
probably reflects the reminiscences of these old narratives or is taken from Skylitzes’
oriental sources.
Moreover, a region that could be defined as “north of Caucasia” was also the
heartland of the territories of Khazars, who are called “Turks” in several Byzantine
texts. The relationship between the early Seljuks and the ruling dynasty of the Khazar
Khaganate was a theme that existed in some earlier Muslim sources but has since
disappeared. In the lost dynastic history of the Seljuks, the Malik-name, Dukak, the
father of Selçuk, was a counselor at the court of the Khazar ruler. His son Selçuk was
subsequently appointed a commander by the ruler. The Khazar connection disappears
227 Golden, An Introduction to the History of the Turkic Peoples, 107-108.
124
in the official Seljuk historiography after the 12th century but has survived in several
works written outside of the Seljuk court circles. According to Andrew Peacock, the
nomenclature of the second generation of the members of the Seljuk dynasty could
possibly reflect a Khazar influence. Names from the Old Testament, such as Mikail,
İsrail, and Musa, could be a reflection of the impact of Khazar Judaism. However,
these names were also used among the Muslims. Peacock concludes as follows:
Given the dubious nature of the sources associating the Seljuks with the
Oghuz Yabgu, the lack of any obvious reason to invent the Khazar story, and
the names of Selçuk’s sons, we must conclude that the extant evidence
suggests that the origin of the Seljuks did indeed lie in the Khazar Empire.
(...) Most probably Selçuk or Dukak was a local chief who perhaps split away
from the empire around the time of its collapse, in the late tenth century,
which shortly preceded the migration of the Seljuks to Transoxiana.228
The fact that Skylitzes put the original homeland of the Seljuks in the Caucasus may
be a reflection of this Khazar connection. Skylitzes’ source about the early Seljuks
may have been an oriental Christian text that utilized the Malik-name as a source.
The idea of the independence of the Turks is still a manifestation of the ethos
attributed by Byzantines to Hunnic/Skythian nomads. However, Skylitzes’ statement
that the Turks were “never enslaved by any nation” contradicts the general historical
narratives which stress the ghulams and Turkish military slavery in the medieval
Muslim world. This expression is surprisingly compatible with Ibn Hassul’s
(d. 1058) comparison of the Seljuks with the Ghaznavids: “As for the genealogy of
this sultan [Ṭog̲ h̲ rı̊l]; its honor does not go back to a low slave and someone
completely obscure, as other’s [Ghaznavids] do.”229
Skylitzes’ account continues as follows:
Once domination of the Persians had passed to the Saracens, the Saracens
went on to rule over not only Persia and Medea and Babylon and Assyria but
228 Peacock, Early Seljuq History: A New Interpretation, 34-35.
229 Peacock, Early Seljuq History: A New Interpretation, 29.
125
also Egypt and Libya and a considerable part of Europe. Then it came about
in various circumstances that they rose up against each other and that one
great empire was torn into many segments. Spain had one ruler, Libya
another, likewise Egypt, Babylon, and Persia. And these neighbors did not
share a common mind but rather waged war on each other. He who was the
ruler of Persia, the Khwarizmians, the Oretanes,230 and the Medes (ἀρχηγὸς
Περσίδος καὶ Χωρασμίων καὶ Ὠρητανῶν καὶ Μηδίας) in the time of the
emperor Basil was Mouchoumet (Μουχούμετ), son of Imbrael. Waging war
against the Indians and Babylonians and getting the worst of it in battle, he
decided that he should treat the ruler of Tourkia, requesting some allied forces
from that source.231
Here the author gives a rather ambiguous narrative of the early territorial expansion
of the Caliphate and its later disintegration. The ruler of Persia, whom Skylitzes
presents as the contemporary of Emperor Basil II, must be Maḥmūd of Ghazni
(r. 999-1030). Furthermore, in this passage, there is an interesting reference to the
“ruler of Tourkia”. The information –and the usages of the names– demonstrates that
this account is based on some oriental sources. The use of the topoynm Tourkia
(Τουρκία) to refer to a place in Asia was rare in Skylitzes’ time. In middle Byzantine
historiography the word Tourkia always refers to Hungary, but in the same period
Arab and Syriac authors used the ethnonym Turk frequently for the Turkic
populations of Asia. It is thus hard to identify the person referred to as the “ruler of
Tourkia”. As already mentioned, Tourkia is a rarely attested geographical indication
in the Byzantine sources of the period. Here the “ruler of Tourkia” may refer to either
the Karakhanids or the petty rulers of Transoxiana. However, the most likely
candidate is Shah-Melik of Jand, the Ghaznavid vassal in Transoxiana who had a
rather complex relationship of suzerainty with the early Seljuk rulers.232
230 Oretanes is probably a misspelling of Oreitans, who were an ancient Indian population that lived in
the region of Oreitis, in the contemporary region of Makran in Pakistan. Their name is attested in
Arrian’s Anabasis of Alexander. See Arrian, Alexander the Great - Anabasis and Indica, tr. Martin
Hammond, 185, 187.
231 Skylitzes-Thurn, Ioannis Scylitzae Synopsis historiarum, 442-444; Skylitzes-Wortley, A Synopsis
of Byzantine History, 811-1057, 416-417.
232 Barthold, Moğol İstilasına Kadar Türkistan, 313-314.
126
After Maḥmūd’s request for auxiliary forces from the ruler of Tourkia, the
latter sent him, according to Skylitzes, “three thousand men under the command of
Tangrolipex Moukalet, son of Mikael, to Mouchoumet”. This Tangrolipex Moukalet
(Ταγγρολίπηκα Μουκάλετ) must be Ṭog̲ h̲ rı̊l Beg, whose Muslim name was
Muhammad. Skylitzes then continues with an explanation of why the ruler of
Tourkia agreed to help Maḥmūd:
He did this in the hope that, if they succeeded in repelling the enemies of the
Saracens, they would quite easily render passable the bridge on the river
Araxes (which was preventing the Turks from entering Persia since it had
guard-towers at either end and it was always watched by guards). After doing
away with its garrison, they could subject the land of the Persians to his
rule.233
In this passage, one sees first the image of the river as the boundary between the
barbarian and civilized worlds. In this case, the river which divides the worlds is the
river Araxes in Southern Caucasia. However, the country of Tourkia must be located
somewhere in Central Asia in this context, so there is the possibility that there is
confusion between the two hydronyms and Araxes is used for the Oxus or most
probably for the Jaxartes, which lay on the historical route that the Seljuks used for
their early invasion of Khorasan in the mid-11th century.
The insistence of Maḥmūd to his soldiers to cross the river Araxes triggers an
unexpected mutiny among the Seljuk mercenaries in the “Saracen” army and these
bands pillage and loot the “Saracen lands.” After Maḥmūd sent an army of about
twenty thousand men “under the command of ten of the most noble and wise
Saracens,” they were defeated in the hands of the mercenaries. Following this event,
[Tangrolipex] no longer conducted his raids surreptitiously like a refugee and
a thief, but openly disputed possession of fortified positions. Some of those
criminals who feared for their lives, some slaves and some of those who took
233 Skylitzes-Thurn,Corpus Fontium Historiae Byzantinae, 448; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 417.
127
pleasure in robbery with violence joined his camp; in very short time a large
force of about fifty thousand congregated around him.234
As seen here, despite the Turks being a “never enslaved” nation, Ṭog̲ h̲ rı̊l’s army was
made of thieves, criminals, slaves, and riff-raff. After this battle Ṭog̲ h̲ rı̊l defeats again
the Ghaznavid army, “an army of about fifty thousand by arming Saracens, Persians,
Kabirs and Arabs.” In Skylitzes’ account of this battle at a place called Aspachan,235
Maḥmūd himself dies falling from his horse. In the aftermath of this battle,
Tangrolipex was universally declared to be king of Persia. Once he was
proclaimed, he sent and eliminated the guard on the crossing of the Araxes,
giving free access into Persia to any Turk who wanted it. Freed of this
impediment, the entire host of them rushed in (except for those who preferred
their own homeland) killing Persians and Saracens. Thus [the Turks] became
masters of Persia, naming Tangrolipex Sultan; that is, absolute ruler and king
of kings.236 He relieved all the indigenous governors of their commands and
transferred them to Turks, among whom he divided out the whole of Persia,
entirely crushing and humiliating the people of the land.237
This narrative may be interpreted as a distorted version of the Seljuk-Ghaznavid wars
of 1038-1040.238
The second Byzantine text that could be called an Origo gentis narrative is
Michael Attaleiates’ text about the origins of the Seljuks:
During those same years, the Nephtalite Huns, neighbors of the Persians, who
are separated from the land of Persia by the Ganges River, which is four and
half miles wide, crossed the river at its narrowest crossing point, when their
leader showed them the way. This man, though he had previously been a
234Skylizes-Thurn, Corpus Fontium Historiae Byzantinae, 444; Skylitzes-Wortley A Synopsis of
Byzantine History, 811-1057, 418.
235 As noted by Wortley (419, n. 113), this toponym must refer to the city of Isfahan. However the
battle of Dandanaqan (1044), in which the Seljuks defeated the Ghaznavids, did not take place near
Isfahan as Wortley claims, and not even in Iran but in actual Turkmenistan. Skylitzes has probably
confused this event with the Seljuk occupation of Isfahan, which happened seven years later in 1051.
Neither was Maḥmūd present in the battle of Dandanaqan; he died in 1030. The Ghaznavid ruler who
commanded against the Seljuks was Masʿūd I, the son of Maḥmūd. Despite his defeat, he was not
killed on the battlefield but survived.
236 This recalls the historical Iranian title, shahanshah, which was used by Achaemenid and Sassanid
rulers.
237 Skylizes-Thurn, Corpus Fontium Historiae Byzantinae, 445; Skylitzes-Wortley A Synopsis of
Byzantine History, 811-1057, 419.
238 Peacock discusses in Early Seljuq History: A New Interpretation, 79-81.
128
captive and came from a humble and servile origin, became the lord of Persia
after the death of its ruling despot. Making a display of their unconquerable
strength tho all the people in that part of the east, they approached the borders
of Iberia.239
In this passage, Attaleiates locates the ancestral homeland of the Seljuks on the
eastern coast of the Ganges River, i.e. in Northern India. However, the Ganges River
does not constitute a frontier of Persia; even in its largest borders. The Ganges River
originates from the Himalayas and flows into the Indian Ocean in the Bay of Bengal.
So it is placed in the eastern part of India. It could be speculated that Attaleiates used
the hydronym Ganges for the river Indus. According to the classical texts, the river
Indus constitutes the border between the lands of Persia and India. Alexander the
Great crossed the river Indus on his way to conquer the land of India. Thinking about
the parallelisms of the anecdote of Alexander and Porus with the Seljuk ruler and
Liparites, it may be safe to assume that Attaleiates confused the names of the rivers
known from the classical texts because of his lack of real geographical information
about India.
If we leave aside the question of the river, there is another reference worthy
of analysis: the bizarre ethnonym “Ounnoi Nephtalitai” (Οὖννοι Νεφθαλῖται) which
is possibly a misspelling of Οὖννοι Ἑφθαλῖται (Ounnoi Hephtalitai), i.e. Hephtalite
Huns. In Chapter 3, information was given about this ancient Hunnic population of
Asia, which is described by Procopius240 as the only civilized branch of this genos.241
This population vanishes from Byzantine literature after the mid-7th century, with the
239 Attaleiates, The History, 76-77.
240 Shliakhtin claims that Attaleiates’ source of inspiration is the Strategikon rather than the works of
Procopius. Shliakhtin, From Huns into Persians: The Projected Identity of the Turks in the Byzantine
Rhetoric of Eleventh and Twelfth Centuries, 50.
241 As it was expressed earlier, it is wrong to identify the ancient populations, particularly the nomadic
federations that were made up of different tribes, with contemporary linguistic groups. The Ephtalites,
although considered to be a part of the Hunnic genos, were probably a mixture of Iranian, Tocharian,
and Altaic tribes. The names of their rulers were almost totally of Iranian origin.
129
downfall of their kingdom located in contemporary Afghanistan and Northern India.
Moreover, the misspelling “Nephtalitai” could not be a simple coincidence, because
it seems not to be a ghost word. It is a biblical demonym that indicates one of the
twelve tribes of Ancient Israel: the tribe of Naphtali, who descended from Jacob’s
son Naphtali. This tribe is frequently referred to in the Old Testament. Therefore, this
misspelling could be related to a linguistic confusion hard to illuminate.
The second part of Attaleiates’ origo gentis narrative is based on the account
of a raid by the Seljuks into the west and their victory over a Byzantine-Georgian
army under the leadership of the Georgian noble Liparites:
That nation then made continual raids on an annual basis, doing no small
damage to Roman territory. The Romans in charge of the borders tried to
resist them but were defeated because the enemy knew well how to use the
bow and hit targets accurately, which made their opponents fear the wounds
inflicted by bows. (…) At one point, a large army was assembled by imperial
order on the border of Iberia, having as its joint commander a famous man
named Liparites. A fierce battle was joined between it and the Huns and for a
while the outcome hung in the balance, but in the end the opposing side
prevailed and defeated the Romans, capturing Liparites alive and taking him,
like some kind of splendid prey, to their ethnarch. He is called sultan in the
Persian language. But when he saw him and learned of his family –for the
fame of the man’s bravery had preceded him –he asked him how he thought
he should be treated. And he said, “Royally.”242
This passage sets out another common theme of origo gentis narratives. Here –as
was demonstrated by Christoph Mallan243– the narrative perfectly fulfills the aim of
such a narrative: by giving to the sultan the role of Alexander the Great, it locates the
Seljuks in a known historiographical tradition. As in Cyril Mango’s “ritualized
ballet,” new barbarians wear the costumes of ancient heroes (or villains), henceforth
becoming familiar.
242 Attaleiates, The History, 78-81.
243 Mallan, “A Turkish Alexander,” 107.
130
Finally, there is also the third criterion of origo gentis: the conversion to a
new religion. This motif is not present in these narratives about the Seljuks, because
in old origo gentis stories, the barbarian populations converted to Christianity, so
they share the same religion with Greco-Roman authors who wrote their origin
stories. The Seljuks changed their religions two generations before the foundation of
the sultanate, but they converted to Islam. As I have already mentioned, Byzantine
authors were not interested in the religion of foreign peoples, so the conversion to
Islam of nomad Turkic pagans was not a noticeable fact for Byzantine authors.
In conclusion, the digressions on the origins of the Seljuk Turks in the
accounts of Skylitzes and Attaleiates were certainly based on material from different
origo gentis stories. It may be assumed that the two authors used different oriental
sources. Thus, the differences in their narratives come not from their political
function, but from their sources.
4.3 Origo gentis narratives about the Ottomans
Byzantine origo gentis narratives about the Turks are not limited to those that deal
with the Seljuks. Byzantine historiographers also produced the same type of
narratives about the Ottomans. The early formation of the Ottoman beylik happened
in the border regions of Bithynia, so several events of the early development of the
Ottomans were realized under Byzantine eyes. George Pakhymeres (1242-1310)
gives us the only contemporary depiction of the deeds of ʿOt̲h̲ mān I, the first beg of
the Ottomans. However, his account regarding ʿOt̲h̲ mān can hardly be described as
an origo gentis narrative. Pakhymeres introduces ʿOt̲h̲ mān (as Ἀτμᾶν) as one of the
Persian chiefs who attacked and devastated Byzantine territory. The other two chiefs
named by him are Lamises (Λαμίσης) and Amourios (Ἀμούρıος). This is the earliest
131
reference to the founder of the Ottoman dynasty in any known historiographic text.
Then he narrates the battle of Bapheus (1302) and in the next chapter counts
ʿOt̲h̲ mān among the chiefs who invaded and pillaged the upper parts of Bithynia,
Mysia, Phrygia, and Lydia. Nevertheless, there is no narrative of the origin of the
Ottomans and the author focuses only on the military deeds of ʿOt̲h̲ mān.244
The Seljuks formed their sultanate first in Transoxiana and Persia, in the
territories far from the borders of the Byzantine Empire. Their history before their
first raids into the Byzantine oikoumene reached the Byzantine authors as rumors
from the Orient, probably via Armenian or Syriac interlocutors. However, the
Ottoman beylik developed under the eyes of the Byzantines. This difference explains
the absence of the Byzantine origo gentis narratives about the Ottomans in the 14th
century.
Doukas is totally silent about the origins of the Ottoman dynasty and its early
history, yet he narrates a rather strange prophecy about the historical parallelism
between two dynasties, Ottomans and Palaiologoi:
While still a youth I learned from old and venerable men that the end of the
Ottoman tyranny would take place with the extinction of the Palaiologan
dynasty. These two began together, ʿOt̲h̲ mān in tyranny and Michael
Palaiologos in sovereignty. Michael’s reign ended shortly after ʿOt̲h̲ mān’s
began. ʿOt̲h̲ mān’s tyranny coincided with the reign of Michael’s son,
Andronicus Palaiologos. ʿOt̲h̲ mān ruled as a tyrant in the latter days of
Michael but he was also a brigand. According to this prophecy, the end of the
emperors and of the City (Constantinople) was to occur first, followed by the
cessation of the Ottoman reign.245
Although this narrative does not present us with any legendary or ethnographic
explanation of the early Ottomans, in a symbolic space it equalizes them with the
Palaiologan dynasty. Because Michael VIII, the first emperor of the Palaiologan
244 Pachymeres, Relations historiques, vol. 4, 346 and 358-369.
245 Doukas-Grecu, Historia Turcobyzantina, 399-401; Doukas-Magoulias, Decline and Fall of
Byzantium to the Ottoman Turks, 244.
132
dynasty was a usurper who blinded the legitimate heir John Laskaris, he and
ʿOt̲h̲ mān represent the same vileness in the eyes of Doukas. The former, despite the
fact that he is Greek and Christian, has usurped the imperial throne, and the latter is a
tyrant who originally was a brigand (λῃστικός).
However, the most extant account of origo gentis about the Ottomans was
written by Laonikos Khalkokondyles in the 1460s. Khalkokondyles’ interest in the
early history of Ottomans and their historiographical traditions was not a
coincidence. He lived in a vanishing once Byzantine world now dominated by the
Ottomans. As pointed out in Chapter 1, in his Histories, Ottoman history is the center
of the narrative and the work itself constitutes a transition between Byzantine and
post-Byzantine historiography.
Khalkokondyles offers four origo gentis explanations about the Turks. The
first one ascribes a Skythian origin to them:
Some believe that the Turks are descendants of the Skythians, which is quite
a reasonable conjecture about them, given that their customs are not all that
different and that their languages are even now closely related.246
Here he uses an argument based on the similarities of the customs and the languages
of the Turks of Asia Minor and the Turkic populations of the Eurasian steppes. He
continues with a historical explanation and remarks that the Skythians conquered
Asia, including territories ruled by the Parthians and even Asia Minor, where they
subjugated “specifically Phrygia, Lydia, and Kappadokia.”247 He then adds,
Even today, so they say, it is possible to see numerous offshoots of this people
roaming about in many parts of Asia, who tend to follow the ways and
customs of the nomadic Skythians and have clearly not settled down in any
particular part of Asia. And they also add that the barbarian nations of the
Turks who live in Asia Minor, I mean in Lydia, Karia, Phrygia, and
246 Khalkokondyles, The Histories, 10-13.
247 Khalkokondyles, The Histories, 12-13.
133
Kappadokia, speak the same language and have the same dress as the
Skythians who roam the lands from the Don into Russia.248
In this narrative, there are several remarkable points. Firstly, Khalkokondyles
constructs a parallelism between language and way of life. This seems to be a
completely new idea. As it was seen in the account of Attaleiates, the Seljuks were
Huns but spoke the Persian language. Also, Anna Komnene clearly distinguishes
genos from language. These are intersectional but not forcefully corresponding
categories. Khalkokondyles makes a break with the medieval historiographical
tradition to which he belongs and interprets nationhood and language with an
approach that echoes early modern European ideas. Secondly, when he lists the
regions in which the barbarians settled, he does not count Bithynia, Mysia, or even
Paphlagonia, even though these regions sheltered an important Türkmen
population.249 He distinguishes the Anatolian space belonging to the beyliks and the
territories that witnessed the early Ottoman expansion. However, it should be
remarked that Khalkokondyles uses careful words to indicate the source of this
knowledge. He implies that these are not his own opinions, yet these are the ideas
that were circulating.
Khalkokondyles’ second origin story makes the Turks the “descendants of the
Parthians” who were “pursued by the nomadic Skythians and moved down into Asia
Minor.” Furthermore, he adds, “turning to a more nomadic way of life, they became
dispersed among the cities there, and since then these people have been known as the
248 Khalkokondyles, The Histories, 12-13.
249 The region around Kastamonu, which correponds to Byzantine Paphlagonia, harbored 100,000
Turcomans according to the Arab author Ibn Said. See Cahen, “Questions d’histoire de la province de
Kastamonu au XIIIeme siècle,” 146.
134
nomadic Turks.”250 The relationship with the Parthians, the Iranian nomadic group
that ruled post-Seleucid Persia, merits a little more attention.
Khalkokondyles’ third origin story is as follows: “Others again say that this
people had its origin in Tourke, a large and prosperous city of the Persians. They
affirm that they left it for Asia Minor and became scattered there, maintaining control
over Asia.”251 Here Tourke seems to be an imaginary toponym, which is probably a
derivation of the ethnonym Turk. However, this origo gentis story presents a distant
relationship with the Ottoman narrative of origins which claims that the ancestors of
the Ottomans were the padishahs of the city of Mahan in Persia.252
Khalkokondyles’ last narrative of Ottoman origin is the most striking one:
There are some, however, who would have it that the Turks came to this land
from Koile Syria and Arabia, rather than from the Skythians, and that they did
so in the company of ‘Umar, who succeeded as lawgiver, and so established
their realm in Asia; when they had been left behind there by him, however,
they turned to a more nomadic way of life.253
The origo gentis tradition which relates the Ottomans with Omar and the Arabs was
strange, yet common for a brief period in the mid-15th century.254 This story’s origin
could be traced back to the historical/biographical work Anba al gomr fi Abna al omr
of the Egyptian religious scholar and chronicler Ibn Hajar al-Askalani (1372-1449).
In this work, Ibn Hajar al-Askalani states that the lineage of the rulers of Rum, i.e
House of Ottomans (Ibn ʿOt̲h̲ mān), comes from the Arabs of Hejaz.255 The 15thcentury
chronicler and poet Enveri, who compiled the Düsturname (c. 1466), a world
chronicle in verse, elaborates this story and states that the ancestor of the Ottomans
250 Khalkokondyles, The Histories, 12-13.
251 Khalkokondyles, The Histories, 12-13.
252 This tradition is preserved in the history of Oruç Bey: “Oruç Beğ Tarihi” (ed. N. Atsız), in Üç
Osmanlı Tarihi, 21. Mahan is located in the Kerman province, in Central Iran.
253 Khalkokondyles, The Histories, 12-15.
254 Imber, “The Legend of Osman Gazi,” 73-75.
255 İnalcık, “İbn Hâcer'de Osmanlı'lara Dair Haberler,” 189-191.
135
was a certain Ayaz (or Iyaz) from the tribe of Quraish. This Ayaz was a warrior of
Caliph Umar who was serving in the army of commander Sa’d ibn Vaqqas. After the
Arab victory over the Persians at Ctesiphon, Sa’d and Ayaz encounter a group of
Oghuz nomads. The beautiful daughter of the ruler of pagan Oghuz, Turunç Khatun,
falls in love with Ayaz and they have a son called Oghuz or Suleiman. Oghuz
(Suleiman) has six sons and twenty-four grandsons, so he realizes the deeds
attributed to Oghuz Khan in the Oghuznama tradition. (Here Enveri tries to make the
Oghuz and Arab stories compatible.) His father Ayaz dies around the cities of Homs
and Hama of Syria, and the descendants of his son Oghuz become the ancestors of
Ertog̲ h̲ rul Ghazi after years of rivalry with the Seljuks.256
Khalkokondyles’ last digression clearly comes from the same source with the
Düsturname, as indicated by the common references to Omar, Koele Syria (Hama
and Homs), and Turkification/nomadization of the descendants of Arab ancestors.
Having an ancestry from the House of the Prophet (Ahl-al bayt) or from the first
generation of Muslims (sahaba) was a typical legitimization instrument of the
Islamic Middle Ages. Of course, this approach reflects a much simpler legitimization
strategy in comparison to the elaborate references to the reminiscences of Oghuz
Khan legends of the steppe world. This way of legitimacy could exist in the mid-15th
century as an alternative to the regular usage of the symbols identified with the Qayi
tribe as a legitimacy tool during the reign of Murād II. Moreover, these legitimacy
tools should be interpreted within the heated ideological debates of the reign of
Meḥmed II. But it seems that this story circulated for a short time and disappeared
quickly. Kaldellis points out this situation, by remarking that Enveri’s patron was the
256 Fatih Devri Kaynaklarından Düsturname-i Enveri: Osmanlı Tarihi Kısmı, 1299-1465, ed. Öztürk ,
5-13.
136
Ottoman grand-vizier Mahmud Paşa Angelović, whose deeds are recounted in the
work of Khalkokondyles.257 So there could be an intersecting network here: since
Mahmud Paşa was a Byzantine aristocrat from the Balkans, just like Laonikos
Khalkokondyles, these ideas could be associated with a certain circle to which the
grand-vizier also belonged.
After giving these four origo gentis stories, Khalkokondyles concludes his
discussion as follows:
I am not able to say with certainty how much truth each of these views
contains or to what degree one should trust in each. But this much, at least,
can be said, that it would be better to side with those who ascribe a Skythian
origin to these people because the Skythians who even now remain in the
eastern parts of Europe in the so-called Horde have no difficulty in
understanding the Turks of Asia. Both nations have the one and the same way
of life and use the same dress even now, because the Skythians prevailed
throughout Asia. Anyway, the name Skythian itself obviously designates
anyone who follows a nomadic way of life and spends most of his time doing
this.258
Khalkokondyles’ conclusion demonstrates that he also chose the linguistic
explanation of the existence of a greater Skythian nationhood. The entity which he
designates as “Horde” (ἀγορά) is obviously the Khanate of the Golden Horde, the
most powerful successor of the Genghisid Empire in the Pontic steppes.
In order to interpret these explanations of origin, there is a need to
contextualize the milieu in which Khalkokondyles lived. He was a noble Athenian by
origin, who lived at least in his youth in the Principality of Athens; therefore he did
not share the lived experiences of the Greeks of Asia Minor during the early Ottoman
expansion. He had a certain access to the inner circles of the Ottoman powerhouse;
for example, he cites some information about the Ottoman budget that came from the
257 Kaldellis, A New Herodotos, 134
258 Khalkokondyles, The Histories, 14-15.
137
accountants of Meḥmed II.259 Although there is not much information about his adult
life apart from what can be traced in his Histories, he seems to have accepted the
Ottoman domination of the post-Byzantine world.
At this point, it is possible to do a more advanced analysis of the provenance
of Khalkokondyles’ origo gentis stories. It seems that the Skythian theory, which was
supported also by the author himself, is a “Byzantine” theory, because, in spite of the
mid-15th century revival in the interest of Oghuz genealogies in the court of
Murād II, the Ottomans did not have friendly sentiments toward the Mongols, who
are called “Tatars” by them. Early Ottoman historiographical material demonstrates a
clear hostility against Tatars. As has been demonstrated by Rudi Paul Lindner, in
early Ottoman chronicles the Tatar tribes were the arch-rivals of the Ottomans.260
Apart from this, the invasion of Anatolia by Tamerlane and the defeat of Ankara
(1402) were the biggest traumas of the Ottoman ruling class. During
Khalkokondyles’ lifetime, it is very possible that the anti-Tatar feelings were still
strong, despite the vassalization of the Khanate of Crimea in 1475. Consequently, it
seems that such a “pro-Mongol” origin story could have hardly come from Ottoman
sources.
Finally, Kritoboulos gives some hints of a narrative about the origin of the
Ottomans. These passages, however, cannot be considered proper origo gentis
accounts. In his first narrative, Kritoboulos states that Murad II was “the sixth of the
brilliant line of the Ottomans, a nobleman of noblemen.”261 He explains the origin of
259 Vryonis, “Laonicus Chalcocondyles and the Ottoman Budget,” 423-432.
260 Lindner, Explorations in Ottoman Prehistory, 90-101.
261 Kritovoulos-Reinsch, Critobuli Imbriotae Historia, 15; Kritovoulos-Riggs, History of Mehmed the
Conqueror, 13.
138
this noble line as the descendants of Achaemenes and Perses, and hence from the
royal line of all Persian kings. He gives a further account:
So too the Greeks are descended from Danaus and Linges, who were in origin
Egyptians, from the town of Chemis, situated in the marsh land. They
migrated into Greece. Ages afterwards, the descendants of these people, who
were called Achaemenidae and Persidae, crossed over into Asia and settled at
first in Persia. And when they died, they left their race and name to that
place.262
So in the world of Kritoboulos there is no longer the antagonism between the
Persians and the Greeks. These two elements are reconciled with each other in a
common historical narrative. According to him,
[The] rule has gone from nation to nation and from place to place in
succession, always changing and passing, now to the Assyrians, the Medes,
the Persians, and then to the Greeks and Romans, according to the times and
epochs establishing itself in a place and never returning to the same.263
The second passage concerning the origin of the Ottomans is the scene where
Mehmed II visits the ruins of Troy. According to Kritoboulos, the sultan said:
God has reserved for me, through so long a period of years, the right to
avenge the city and its inhabitants. For I have subdued their enemies and have
plundered their cities and made them the spoils of the Mysians. It was the
Greeks and Macedonians and Thessalians and Peloponnesians who ravaged
this place in the past, and whose descendants have now through my efforts
paid the just penalty, after a long period of years, for their injustices to us
Asiatics at that time and so often in subsequent times.264
The attribution of Trojan origins to a certain people or dynasty is a common
genealogical motif since Antiquity. Vergilius’ great epic Aeneid connects Rome and
Troy through the figure of Aeneas. In the chronicle of Fredegarius, a Trojan origin of
262 Kritovoulos-Reinsch, Critobuli Imbriotae Historia, 15-16; Kritovoulos-Riggs, History of Mehmed
the Conqueror, 13.
263 Kritovoulos-Reinsch, Critobuli Imbriotae Historia, 16; Kritovoulos-Riggs, History of Mehmed the
Conqueror, 13.
264 Kritovoulos-Reinsch, Critobuli Imbriotae Historia, 170; Kritovoulos-Riggs, History of Mehmed
the Conqueror, 181-182.
139
the Franks is also mentioned. Moreover, Geoffrey of Monmouth attributes a Trojan
origin to the Britons, through Brutus, the grandson of Aeneas.265 The narrative of
Kritoboulos seems similar, except for a little modification of the motif. According to
the author, the Ottomans are not the offspring of Trojans, but they are descendants of
the royal line of Persians. The feature which connects the Ottomans and Trojans is
their common Asiatic belonging. Kritoboulos’ stress on Asian vs Greek is a
Herodotean motif. This dichotomy substitutes the antagonism between Christianity
and Islam and offers a new antagonism. However, this new antagonism, as the
sultan’s alleged statement demonstrates, is already over: the Asian power has won
and their superiority will be continued until an undefined future.
Thus, Kritoboulos’ approach reflects his aim in writing his History. His goal
was to legitimize the takeover of the Byzantine space by the Ottomans. This task
could be achieved by using the models of Antiquity. The instrumentalization of
Antiquity not only fit his political goals but also fit the intellectual pursuits of his
time.
4.4 Slaves as rulers: Aristotelian thought and the representation of Turkic rulers
In this sub-chapter, the formation of the image of the Turkic peoples in the origo
gentis narratives in the Byzantine texts will be discussed. This time another founding
element of the image of the Turks will be analyzed: their social origin and its
relationship with a well-known notion of the medieval world; slavery.
Here the starting point is still in Antiquity. To understand the genealogy of the
thought which relates the barbarians and slavery one must go back to the ancient
265 MacMaster, “The Origin of Origins: Trojans, Turks and the Birth of the Myth of Trojan Origins in
the Medieval World,” 1-12.
140
Greek philosopher Aristotle (384-322 BC), who stated that barbarian societies had no
real ruling class and were societies of slaves. Aristotle was widely read during the
Byzantine era. Along with Plato, he was one of the rare ancient philosophers whose
works continued to be appreciated in Christian intellectual circles. Both Patriarch
Photios and Michael Psellos extensively studied the works of Aristotle. When Psellos
was teaching at the University of Constantinople, he gave lectures on his philosophy.
One can even talk of a Byzantine school of Aristotelianism to which several
intellectuals belonged, such as Nikephoros Blemmydes, Theodoros Metokhites, and
George Pakhymeres. The latter wrote an “Epitome of the Philosophy of Aristotle,”
consisting of twelve books on Aristotelian subjects, in which he directly copied or
summarized several passages from the works of Aristotle.266 Although they probably
did not go as deep as Pakhymeres, other authors belonging to the corpus of this
dissertation were also influenced by Aristotelian thought. A near contemporary of
Psellos, Michael Attaleiates referred to Aristotle twice in his history.267
Furthermore, Anna Komnene was an avid reader of Aristotle. She was part of
a circle of scholars interested in Aristotelian philosophy in the early 12th century and
even commissioned commentaries on several works of Aristotle.268 Even Kinnamos,
who generally deals with more concrete aspects of military and political issues, states
that he frequently discussed the works of Aristotle with emperor Manuel I
Komnenos. So it is possible to state that Aristotelian thought had definitive
importance and formative role in the worldview of the Byzantine intellectual class.
Moreover, in the Komnenian period, which roughly coincides with the first century
266 Oehler, “Aristotle in Byzantium,”138-139. Pachymeres, Relations historiques, vol. 1, xxi. Only the
first book of the Epitome survived as an extant text. For slavery in Aristotle’s thought, see also Heath,
“Aristotle on Natural Slavery,” 243-270.
267 Attaleiates, The History, 394-395, 566-567. He cites Meteorology and commentary of Rhetorics.
268 Komnene-Sewter-Frankopan, The Alexiad, xiii.
141
of the time frame of this dissertation, there is a strong interest and appreciation for
the works of Aristotle.
As a social phenomenon, slavery was not unknown in the Byzantine Empire,
though in a lesser frequency than in Antiquity. The Byzantine Empire, particularly its
capital and most important port Constantinople, was an important center of the slave
trade in the Middle Ages. Although generally the captives of war constituted the
biggest source of slaves, this fact was not true in the cases of the Byzantine Empire
and the Abbasid Caliphate that constantly waged war against each other, because
between the two entities there were often treaties for the exchange of prisoners of
war. So, in the Byzantine Empire, it seems that an important part of the slaves came
from the North of the Black Sea, i.e. the Skythian slaves who could be of Slavic or
Turkic stock. These slaves were also employed in agriculture in the rural parts of the
Empire.269
In Byzantine Greek, the generic term that means the slave is δοῦλος. But after
the 11th century, a new term appears: σκλαυος. This term existed in Byzantine Greek
as an ethnonym since the 6th century, it means the Slav. This word passed to the
Arabic language as saqaliba, as a term for people of Slavic origin or generally white
slaves. The word gained the meaning of slave, probably with the impact of the Arabic
language. This word σκλαυος is the origin of the words for slave in certain European
languages: such as slave in English, esclave in French, and schiavo in Italian.
Thus, in Byzantine society there was a visible presence of slaves, who
generally came from the world commonly known as Skythia. This Skythia could
sometimes coincide with the Slavic homeland. On the other hand, slavery in the
Islamic world is also very important. There is a voluminous scholarly corpus about
269 Rotman, “The Medieval Mediterranean Slave Trade,” 129-142.
142
the different aspects of slavery in medieval Islamic societies. Slavery was a
socioeconomic status that was determined by religion, and in medieval Islamic
societies there are three types of slavery: domestic (which includes concubines), field
labor, and military slavery. All three categories were frequent, but the latter is a
phenomenon particularly identified with medieval Muslim societies.270
The Turkic populations of Central Asia were the abundant sources of slaves
for the Caliphates for centuries. Thus, long before the entry of the Seljuks into the
core lands of the Abbasid Caliphate, slave troops of Turkic background were present
in the Muslim armies that the Byzantines encountered in the thugur region. I already
cited the passages of John Skylitzes concerning the Turks in the Abbasid army in the
Battle of Dazimon (838). The existence of Turkic slaves in the Abbasid armies is also
echoed in the epic of Digenis Akritas.
So the very first Muslim Turks that the Byzantines knew were the slave
warriors in the service of the Caliph. In the eyes of Byzantine historiographers, the
Turks were identified with slavery. However, it is doubtful whether this identification
comes from real and direct contact with these people or from the oriental sources
used as materials for Byzantine historical texts. The Turks’ identification with slavery
constitutes not only a social, but also a moral category.
If one focuses on the Turkic state formations after the 9th century, two
dominant models can be seen. The first one is the tribal-state formation, which
repeats old patterns under a new narrative of legitimization. The Karakhanids,
Seljuks, and Anatolian beyliks could be considered examples of the first model. In
this case, there is a clear source of legitimacy based on dynastic prestige and tribal
270 For the evolution of this institution, see Crone, Slaves on Horses: The Evolution of the Islamic
Polity, 74-91.
143
origins. This legitimacy is empowered by new means of legitimization taken from
Islamic culture. The second model is the mamluk or “slave sultans” model. This
model has a long past in the Islamic oikoumene; since the early Abbasid Caliphate,
slave warriors from different origins played an important role in the Muslim armies.
The warriors of slave origins such as Aḥmad b. Ṭūlūn (r.868-884) and Muḥammad b.
Ṭug̲ h̲ d̲j̲
(r.935-946) became the provincial governors of the caliphal state that
managed to establish their secessionist sultanates in Egypt and Syria. Even though
their short-lived states soon vanished, after the 11th century several dynasties of
Turkic mamluk origin –such as Ghaznavids, Khwarazmians, the Mamluks of Egypt,
and the Mamluks of India– represented a similar state formation model. In such a
model, the warrior of slave origin, a self-made man who takes the power, could
establish his dynasty (such as Alptigin r. 962-963, the founder of the Ghaznavid
dynasty) or could form a kind of military oligarchy (such as the Mamluks of Egypt).
If there was no established dynasty, a new amir of slave origin, generally from the
close entourage of the late sultan, took power.
Here I can offer a generalization about these processes tof state formation.
The tribal model was prevalent in the territories where the nomadic lifestyle and
Turkic populations were dominant, and the “mamluk” model was dominant where
the sedentary lifestyle and non-Turkic populations were dominant. Still, it could be
said that the Seljuk state represented a hybrid of these two models. Although the
founders of the state were not slaves, they were part of the nomad élite of the society
to which they belonged, and they founded the nucleus of their state in Khorasan, a
country where the majority of the inhabitants were Persian. So they adopted the
customs of Persianate state administration; they had a Persian chancellery, adopted
the bureaucratic traditions of the earlier Muslim states of Iran, and organized ghulam
144
troops. However, these facts triggered tension between the ruling élite and nomad
Türkmen, as well as between the members of the ruling dynasty where the non-ruling
princes allied to the Türkmen or other non-sedentary warbands against their central
authority. This fact is a founding event of the immigration policy on the margins of
the Great Seljuk Sultanate which also triggered the formation of the Sultanate of
Rum.
So slavery is a well-established institution in the Islamic world and the
phenomenon of slave soldiers was quite an authentic feature of Muslim societies.
However, these facts should not be considered unique features of the Muslim world.
The easy social ascension in the barbarian societies, as mentioned above, was a
cliché. Naturally, it was a Byzantine (or more correctly a Greco-Roman) idea and it
does not demonstrate that social mobility is unknown in the Byzantine world. There
is no need to mention that emperors like Basil I and Michael IV (r. 1034-1041) were
men of very humble origins. The former was the son of a poor Armenian family
settled in Macedonia and the latter came from the peasantry of Paphlagonia. Basil I,
a poor young man from Macedonia, could enter the entourage of Michael III because
of his talent in wrestling, and then could become the emperor’s parakoimomenos. By
assassinating the emperor, he could take the throne. Michael IV could marry Empress
Zoe and become the emperor because of his brother John Orphanotrophos, who was
an influential court eunuch. In brief, seeing men of humble origins in higher
positions was not very strange to Byzantines.
However, there is still a striking difference between the upward social
mobility of a peasant and a slave. A journey from slavery to the throne was a strange
destiny in the eyes of the Byzantines. It seems that, at least in the mid-to-late 14th
century, they had information about this institution of Mamluks because George
145
Pakhymeres states that: “the sultan of Ethiopians is from Cuman origin and he was
one of those who were sold as slaves (εἷς τῶν εἰς δουλείαν ἀποδεδομένων).”271 He
furthermore adds that because of the effects of climates on the human characters, the
Ethiopians revere so much the Skythians, so they buy them as slaves and employ
them in military matters. So when a Skythian takes the power, he searches for
Skythians to compose his army.272 In this explanation, there is an echo of the
Aristotelian theory of the climates.
As it was seen in the narrative of Attaleiates, if there is a Turkic ruler of
Persia, he must be an ex-slave of the sultan of Persia. However, Skylitzes stresses
that “(they were) never enslaved by any nation.”
These two different approaches regarding early Seljuk history demonstrate
the co-existence of two contradictory narratives about their social origin. Attaleiates’
version could be understood as a continuation of tradition and Skylitzes identified
them with free men. Still according to Attaleiates, when a family of servile origin
becomes a ruling dynasty, it becomes noble. When he presents the nephews of sultan
Alp Arslan, he calls them two nobles from Persia (εὐπατριδῶν τῆς Περσίδος) and
indicates that they are from the royal lineage (γένους ὄντες βασιλικοῦ).273
Among the later narratives dealing with the same issue, Anna Komnene’s
representation of the Seljuks merits attention because, as it was already said, she
clearly differentiates two Seljuk entities, the sultanates of Rum and Persia (Great
Seljuk Sultanate), and she indicates the tumultuous relations of suzerainty and
vassalage between them.274 However, she seems to be quite unaware of the dynasty’s
271 Pachymeres, Relations historiques, vol. 3, 236.
272 Pachymeres, Relations historiques, vol. 3, 236.
273 Attaleiates, The History, 484-485.
274 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 186; Komnene-Sewter-Frankopan,
Alexiad, 169. “I must now describe how Emir Solymas (Sulaymāns̲ h̲ āh), having left Nicaea, appointed
this Apelkhasem governor of the city; how Pouzanos was sent by the Persian sultan to Asia…”
146
early history. John Kinnamos seems also unaware of such an issue, and he never
mentions the early roots of the dynasty. Niketas Khoniates is also silent about the
early Seljuks.
A dialogue between the emperor Manuel I Komnenos and a Turkish warrior
gives a detail about the Byzantine imaginary regarding the Seljuks. In this passage
(which is probably fictional) the emperor sends an ultimatum to sultan Masʿūd I
(r. 1116-1156) via a Turkish soldier on the battlefield and he complains that the sultan
is being withdrawn into the inner parts of Asia Minor by saying he “fled continually,
like runaway slaves.”275
The Byzantine exonym employed for the Muslims, Agarene, has an implied
reference to slavery. The Arabs (and generally Muslims) were not considered as the
legitimate descendants of the patriarch Abraham by his wife Sarah, yet they were
considered his descendants of the inferior status of her Egyptian slave Agar. Ismael,
the progenitor of the Arabs, is Abraham’s less respected son.276 This word originally
signifies an ethnic Arab, but later became a pejorative term used for all the Muslim
populations. In that context, it could be assumed that the term Ismaelitai has also a
very indirect reference to slavery because Ismael’s mother was a slave.
Niketas Khoniates, who was among the first Byzantine authors that employed
the word Agarenoi for the Seljuk Turks, reflects the agony of being prevailed over by
a community that they see as inferior to themselves:
O Lord of vengeance, thy taking revenge? How long shall these calamities
follow one another and the descendants of the bondwoman Agar continue to
subjugate those of us who are free and destroy and kill thy holy nation which
above every name has called upon thine? How long shall we endure this longcontinued
servitude and suffer the reproaches and buffetings of these
275 Kinnamos-Meineke, Rerum ab Ioanne et Alexio Comnenis Gestarum, 58; Kinnamos-Brand, Deeds
of John and Manuel Comnenus, 52.
276 Savvides, “Some Notes on the Terms Agarenoi, Ismailitai and Sarakenoi in Byzantine Sources,”
94-95.
147
accursed foreigners? Let the affliction of those in fetters, O Master, lover of
goodness, come before thee at last. Let the blood shed by your servants cry
out to you, o merciful God, as did Abel's blood in the beginning.277
In short, the reference to Agar is not a random reference. The use of this term
implicitly indicates a nobility-slavery tension. This expression takes on an even more
bitter meaning as the nobles weaken against those who are attributed to them as
slaves. The discourse of alterity is based on this tension.
In concluding this sub-chapter, I want to stress that the identification with
slavery constitutes a partially independent part of the pejorative rhetoric used for
Turkic peoples. It is closely related to the general discourse against the northern
barbarians but has a distinct moral meaning. Secondly, the issue of Agarenoi adds a
new dimension to this discussion, because Agar’s servile origin is a theological issue.
4.5 Conclusion
In this chapter, I dealt with some questions regarding the representation of the
mythic, historical, and social origins of the Ottoman and Seljuk ruling élites and
populations. Above all, I discussed the origo gentis narratives about the Seljuks and
Ottomans, because these narratives had two important functions. Firstly they located
new populations in the human geography of an already known world. So they made
them familiar to Byzantines. Secondly, the explanation of the origin introduced these
peoples into a general historiographical narrative that included later events related to
them.
Aristotelian thought is still very important to understand the Byzantine
intellectual world. As was stressed above, his texts were read in the Byzantine
277 Khoniates-Van Dieten, Nicetae Choniatae Historia, 117; Khoniates-Magoulias, O City of
Byzantium, 66.
148
intellectual circles and had a strong influence on the Byzantine worldview. The
slavery issue is also related to the abovementioned explanation of origins. The
existence of a servile origin or any reference to it demonstrates a lack of valor
attached to an individual in the Byzantine worldview. The image of the Skythian is
closely related to this approach. There was an abundance of “Skythian” slaves in the
Byzantine Empire and these rulers could be potentially considered contradictory to
the Byzantine concepts of nobility.
The last subchapter tried to complete the question of slavery with an analysis
of ruler succession and dynastic conflicts in Turkic entities, as they were reflected in
the Byzantine texts. These narratives are important for a better understanding of how
the Byzantines perceived these entities, their internal structure, and the sources of
their legitimacy.
149
CHAPTER 5
“BARBARIAN” POLITIES THROUGH BYZANTINE EYES
5.1 The place of barbarian polity in the Byzantine worldview
In this chapter, I will focus on several aspects related to the literary representation of
the political hierarchy between the Byzantines and Turkic peoples whom Byzantines
considered barbarians. Byzantine civilization, very much like the other great
civilizations of the medieval world, was a self-centered civilization. As was already
discussed in Chapter 2, the dichotomy between the Byzantines and “barbarians”
indicated a cultural hierarchy between the two groups. Similar to this cultural
hierarchy, there also existed a political hierarchy between the Byzantine Empire and
the rest of the world. The Byzantines called their state Βασιλεία τῶν Ῥωμαίων - The
Empire of Romans. The other entities surrounding the Byzantine Empire were
considered hierarchically inferior by the Byzantines. The title “emperor” (βασιλεύς)
was generally employed to define the Byzantine emperor because the Byzantine
Empire was considered not a successor of the Roman Empire, but the empire itself.
Unlike modern historiography, Byzantine historians believed in an uninterrupted
continuity between the ancient Romans and medieval Byzantines. The Byzantine
Empire considered itself the only legitimate empire in the world and did not want to
share this title with the western monarchies that saw themselves as the inheritors of
the Roman heritage, such as the Carolingian dynasty in France. In 800, when
Charlemagne was crowned by Pope Leo III, the Frankish king’s coronation as the
emperor of Romans was perceived as an offense by the contemporary Byzantine
empress Irene (r. 797-803). Additionally, the Byzantine emperor had a spiritual
150
dimension in his duty, and his authority over the people was believed to be divinely
ordained.278
George Ostrogorsky explains this notion in the Byzantine model of
international relations as the “Byzantine hierarchy of states.” He further defines the
imperial office as such:
The emperor and omnipotent ruler of the Romans will be the leader of all the
world and the guardian and protector of the Christian faith, because he is the
only legitimate emperor on earth, being the Chosen of God and the successor
of Roman emperors. The idea that there may be only one single legitimate
empire is the basic principle, the alpha and omega of all Byzantine political
doctrines.279
However, this traditional approach was criticized recently by a new generation of
researchers such as Anthony Kaldellis. Kaldellis argued that the Byzantine Empire
was essentially a republic. Yet, it was a republic not in a contemporary sense, but in a
strictly Roman sense. According to him, the Byzantine Empire was a politeia, not a
basileia, and the legitimacy of Byzantine emperors was deeply connected with
popular support, i.e., the support of the Roman masses.280
In any case, it should be admitted that the emperor (βασιλεύς) occupied the
central role in the Byzantine worldview. The Greek title βασιλεύς was used mostly
for the Byzantine rulers until the late Byzantine era, its most important exception
being the Persian shahs. But the same title was also claimed by Bulgarian and
Serbian kings who sought the expansion of their authority in the Balkans, and this
challenge was perceived as a threat by the Byzantine ruling class. Yet these rulers
used titles such as “basileus of Romans and Bulgarians” or “basileus of Romans and
Serbs” in their official titulature. In the 12th century, the Byzantine exclusivity about
278 Ostrogorsky, “The Byzantine Emperor and the Hierarchical World Order,” 1–14.
279 Ostrogorsky, “The Byzantine Emperor and the Hierarchical World Order,” 5.
280 Kaldellis, The Byzantine Republic: People and Power in New Rome, 6-8.
151
the imperial title was relaxed, and authors like Niketas Khoniates and George
Akropolites employed this title for foreign Christian rulers. This change was closely
related to the evolution of the imperial idea throughout the centuries. The Byzantine
emperor no longer represented the same authority he used to represent through the
11th and 14th centuries. The political ethos of the Komnenian dynasty was far
different from that of the Palaiologans.
Another ideological feature that accompanied this Byzantine superiority was
the diplomatic language which put the Byzantine emperor into a spiritual kinship
with the rulers of neighboring countries. Christianity further strengthened these ties.
For example, when the Bulgarian prince Boris converted to Christianity, he
recognized the Byzantine emperor as his spiritual father.281 According to Khoniates,
Manuel I honored Qiliç Arslān II by adopting him as a son. In their official
correspondence, the sultan addressed the emperor as his father, and the emperor
addressed the sultan as his son.282 There is a similar reference in Akropolites’ work,
dealing with the baptism of Izz al-Dīn Kaykāʾūs I: “The said Iathatines had escaped
from the hands of his brother Azatines, then ruler of the Muslims, and had fled to the
city of Constantine; he was received by the emperor Alexios and was baptized by
him and adopted.” 283 So, the dichotomy between the father and the son was the
Byzantine expression of vassalage.
281 Malamut, “Les peuples étrangers dans l’idéologie impériale,” 128.
282 Khoniates-Van Dieten, Nicetae Choniatae Historia, 121; Khoniates-Magoulias, O City of
Byzantium, 68.
283 Akropolites-Heisenberg, Georgii Acropolitae Opera, 14; Akropolites-Macrides, History, 124.
152
5.2 Archons and sultans: How the Byzantines saw the Turkic rulers
In this sub-chapter, I shall deal with the Byzantine perception of the Turkic rulers. As
I already mentioned, the barbarian rulers were perceived as inferiors in the Byzantine
worldview. So, a Turkic ruler, just like any ruler from a non-Byzantine background
could never be equal to an emperor and always has a lower reputation than the
Byzantine emperor. A most useful way to understand this relationship is to focus on
the titulature for the foreigners in the Byzantine texts. Here I shall offer a systematic
study of these titles through the centuries.
In the early centuries; the title used for the rulers of Turkic people is
“khagan” (χαγάνος).284 This ancient term is used for the supreme rulers of steppe
nomads since Late Antiquity and it could be traced back to the fragment about the
voyage of Zemarchus to the court of the Türk Khaganate which has survived in the
chronicle of Theophylactos Simocatta. In this text, the Türk khagan Silziboulos was
referred to also as “ the ruler of so many peoples” (ὦτοσούτων ἐθνῶν ἡγεμών).285
The title khagan was used to designate the rulers of Avars, Turks, Khazars, and
Bulgars in the later centuries. The Mongol expansion in the 13th century revived the
use of this title. Interestingly, the Byzantines never used this word to refer to the
rulers of Mongolian states. Furthermore, this title was never used by Seljuks and the
Byzantines never used this title to designate their rulers. Khan (χάν, χάνης, κανάς)286
is a derivative form of khagan which was part of the titulature of the Ottomans and
Genghisid states of Eurasia, such as Ilkhanate and Khanate of Crimea. Khagan, in its
284 For an etymological discussion of the title, see Clauson, An Etymological Dictionary of Pre-
Thirteenth-Century Turkish, 611.
285 Dobrovits,“The Altaic World Through Byzantine Eyes: Some Remarks on the Historical
Circumstances of Zemarchus’ Journey to the Turks (AD 569-570),” 388.
286 Moravcsik, Byzantinoturcica, vol. 2, 339.
153
original context, was an equivalent of the title emperor in the world of steppe, hence
it represents a claim of universality.
However, starting with Seljuks, the most common titles to define the leaders
of the Turks were “sultan” and “amir”.287 These Arabic titles were adopted by the
Turks themselves with Islamization. In Byzantine use, the term “sultan” is generally
employed to define the ruler of Seljuks and Ottomans, while “amir” designates any
petty ruler. Interestingly, the traditional Turkish title “beg”, which designates
whoever rules a beylik, appears very scarcely in the Byzantine texts. The terms
archon and hegemon are frequently used instead of the title “beg”. When this word is
used, it appears nearly always as a part of a personal name, such as Χασάνμπεγις. A
composite title that is derivative of the title “beg”, “atabeg” (ἀταπάκας, ἀτάπακος),
seems to be more frequent in the Byzantine texts. This title is a composition of two
words, “ata” (which means father in Turkish) + “beg” and it represents a rather
original institution of Seljuks. The “atabeg”s were tutors of the Seljuk princes who
were employed with the provincial governorships in their youth. Many times, the
atabegs got rid of their princes and became independent rulers, such as the atabegs of
Aleppo and Mosul, respectively ʿImād al-Dīn Zangī (r.1127-1146), the founder of
the Zangid dynasty of Syria, and Badr al-Dīn Luʾluʾ(r.1234-1249). A third title
frequent among Muslim Turks was “melik” which means literally “king” in Arabic.
This title is also found in the Byzantine sources under the form of μελίκης. The last
title that should be mentioned here is the title “shah”. It is a Persian title that was
familiar to Byzantine authors also before the arrival of the Turks. In the Byzantine
287 As it was seen earlier, it is a generic term employed to designate the Muslim petty rulers by the
Byzantines. The form ἀμερμουμνῆς (amīr al-muʾminīn) was used to refer to the Caliphs.
154
texts related to the Turks, this title is attested rarely, in the form of σάχ or σιάχ, and
always as part of a personal name.
There also should be briefly mentioned the feminine forms of these titles. In
the traditional Turkic titulature, the feminine form of Khagan is Khatun. However,
because of the scarcity of mentions of Turkic women in the Byzantine texts, there is
not much reference to this title by the Byzantine authors. There is one exception:
Doukas speaks about Fatma Khatun (Φατμάκατουν), who is the daughter of Emir
Süleymān and the sister of Demetrios-Yusuf. In concluding this introduction, I must
remark that the Byzantine writers were never eager to use authentic titles of foreign
populations and apart from the sultan and beg, the use of these titles was particularly
rare.
Attaleiates calls the ruler of the Seljuks as the ethnarch who became the ruler
of Persia (δεσπότου τῆς Περσικῆς) and he is called sultan in the Persian language.288
He always refers to the Seljuk ruler (both Ṭog̲ h̲ rı̊l and Alp Arslan) as the sultan and
never explicitly mentions their names. When he mentions the army of the sons of
Kutlumus (Sulaymāns̲ h̲ āh I and Alp-İlek?) he states that the leaders of the army were
called “emirs” and “selarioi” in the Turkish language.289 The title salar, as it was
already seen in the text of Skylitzes, was a Persian military title which designates
generally a military governorship, similar to strategos in the Byzantine Empire. The
salars were generally referred to with the province which they were charged, such as
Chorosalaris (salar-e Khorasan).290
In Anna Komnene’s Alexiad there is a more detailed panorama of the Seljuk
world: in Asia Minor; the rulers from the cadet line of the Seljuks, the descendants of
288 Attaleiates, The History, 76-79.
289 Attaleiates, The History, 504-505.
290 Attaleiates, The History, 142-143.
155
Kutlumus; were called sultans. (Sulaymāns̲ h̲ āh, Qiliç Arslān I and S̲ h̲ āhins̲ h̲ āh)
However, the author also differentiates the Seljuk Sultanate of Rum and the Great
Seljuk Sultanate; by designing the latter as the Sultanate of Khorasan.291 Their capital
Nicaea is soultanikion.292 The Great Seljuk sultanate soon declined and became
unable to intervene in Anatolian politics and remained out of sight of Byzantine
historians. In the history of Kinnamos, there is no reference to this entity.
The title of sultan was widely known and used by Byzantine authors. This
term was used often for the Seljuk rulers. In the medieval Muslim world, the rulers
could not use the title of sultan without the confirmation of the caliph. However,
after the 14th century this situation changes a bit: In the work of Pakhymeres, the
title of sultan was used both for the sultans of Rum and the sultans of the Ethiopians,
which means the Mamluk sultans of Egypt. He also designates the Mamluk sultans
as the “sultans of Babylon”. He never uses the title of sultan for the descendants of
Izz al-Dīn Kaykāʾūs II, but rather designates them as satraps. The title satrap is
indeed a common title to designate the Turkish petty rulers of Asia Minor. Shliakhtin
traces the first use of the term back to the poetry of Theodore Prodromos in the
1130s and 1140s. According to him, the use of this word refers not to Classical
Antiquity, yet to the Old Testament and the Greek version of the Romance of
Alexander.293 The term satrap remains in use until the 15th century. Kantakouzenos
and Doukas use these terms to refer to different people. Kantakouzenos refers to the
Anatolian begs (such as Saruhan, satrap of Lydia and Umur, satrap of Ionia) by using
291 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 196; Komnene-Sewter-Frankopan, The
Alexiad, 179-180.
292 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 196-197; Komnene-Sewter-Frankopan,
The Alexiad, 180.
293 Shliakhtin, From Huns into Persians: The Projected Identity of the Turks in the Byzantine Rhetoric
of Eleventh and Twelfth Centuries, 144.
156
this term, but Doukas uses it exclusively for the governors (Musa’s satraps,
Meḥmed’s satraps) of the Ottoman state.
I think that the title of satrap has an interesting meaning in terms of otherness.
It reflects a perfectly Persian image of the Turks, but this Persian image is not
antagonistic, it is so out of time that it could be considered exotic. Moreover, it has
always a certain ambiguity: it could refer to both a subordinate or an autonomous
ruler.
The Türkmen dynasties of Anatolia, such as the Danishmendids minted their
coins in the Greek language.294 Following closely the Byzantine model; in these
coins, there were the Byzantine style visual representations of the rulers and the
writings in Greek. This fact is particularly worthy of attention because of the
representation of the Melik Danishmend Ghazi, the eponymous founder of the
Central Anatolian dynasty as the most excellent ghazi in the epic romance dedicated
to him, the Danişmendname. However, despite the existence of the historical
material in the Danişmendname, this work is neither a contemporary work (it was
written in the 14th century as a new version of a now lost romance), nor was it
commissioned by the Danishmendid dynasty; aiming to present their self-image. In
these coins, the Danishmendid rulers define themselves always as megas melikis or
megas amiras, they use neither the title of sultan, nor basileus. The only exception is
the Alexiad of Anna Komnene in which, Danishmend Ghazi was mentioned as Sultan
Tanisman.295
294 See the article by Oikonomides “Les Danishmendides, entre Byzance, Bagdad et le sultanat
d'Iconium.”
295 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 331; Komnene-Sewter-Frankopan, The
Alexiad, 305.
157
The toparch is another similar notion: Emir Saltuq of the Saltuqids in Chaldia
was designated as toparch (τοπάρχης) of that region. The term toparch is a composite
word made by topos and archôn; it means the lord of a region, a locality. It was used
not only for the Turkic or other foreign rulers but also for the Byzantine provincial
magnates. The word topos indicates the locality; thus this title was employed for the
local and non-independent rulers. Jean-Claude Cheynet compares this title with
ethnarch and phylarch and points out that the title toparch is mostly employed by the
rulers of the entities at the imperial periphery or former Byzantine territories. The
territories ruled by a toparch are somewhat modest in terms of size. So in our
context, this title was mostly used for the Turkish begs of Asia Minor.296
Tyrant (τύραννος) is another title employed to designate the rulers of Turkic
states; differently than the toparch; it does not imply a lower degree in a so-called
hierarchy of rulers, but moral inferiority of the barbarians to the Romans. As it was
already pointed out, in the Aristotelian thinking there is an obvious parallelism
between barbarism and tyranny. The tyrannical governments are appropriate for the
barbarians, by their very nature. However, the issue of tyranny and the rule by tyrants
could not be reduced only to barbarians. There exist not only the Byzantine usurpers
but also the emperors who are described as tyrants. For example, according to
Khoniates, Andronikos I Komnenos is a tyrant. As I mentioned above, tyranny has
also a moral dimension independent from politics. A tyrant typically commits
sadistic acts against innocents and has very bad sexual morals.
Interpreting the titulature employed by the Byzantine authors for Turkic
rulers; one must first distinguish the Greek titles and the titles of
Turkish/Arabic/Persian origin.
296 Cheynet, “Toparque et topotèrètès à la fin du 11e siècle,” 215-216.
158
Another interesting point is the Byzantine use of honorifics in place of the
names of Seljuk sultans. For example, when the Byzantines refer to various Seljuks
sultans named G̲ h̲ iyāth al-Dīn Kayk̲ h̲ usraw, they always refer to them as Iathatines
(Γιαθατίνης or Ἰαθατίνης); however, the sultan’s real name was Kayk̲ h̲ usraw, and
G̲ h̲ iyāth al-Dīn is only a Muslim honorific title, that means the protector of the
religion. The same is true for the uses of the honorifics such as Azatines and
Alatines, to refer to the sultans Kaykāʾūs and Kaiḳobād.
How much did the Byzantines know what these titles mean to the Turks, and
the hierarchical relationship between them? They were obviously aware of the
general hierarchy of the titulature, if not of the finer nuances. They utilized this
titulature as accurately as possible for a culture to which they were unfamiliar.
The use of the title basileus for a foreign ruler demonstrates the degree of
equivalence between this ruler and his state and the Byzantine Empire. In our
context, the most significant example of such usage is the persistent use of the title
basileus for the Ottoman rulers by Laonikos Khalkokondyles. Khalkokondyles’ work
could be considered a post-byzantine work because of the date of its composition.
This usage demonstrates clearly its context. The Byzantine Empire was over, and the
title of basileus now passed to the Ottomans, the new rulers of post-Byzantine space.
However, this may be a too hasty judgment. Khalkokondyles’ utilized the same title
also for Tamerlane who was perceived as the dynasty’s arch-nemesis in the early
Ottoman chronicles. The use of the same title for both the Ottoman rulers and
Tamerlane could not be explained with a projected pro-Ottoman approach to him.
Despite the use of the title autocrator (αὐτοκράτωρ) for the Holy Roman Emperors
whom normally a Byzantine author would not use. He uses once the title for the King
Sigismund (r. 1387-1437) of Hungary, as “the emperor and autocrator” of the
159
Romans. So, it may be more reasonable to interpret this as the absence of the old
imperial hierarchy in the post-Byzantine world. So there is no more only one
basileus. Furthermore, Khalkokondyles was personally part of the circle of Plethon,
and as he demonstrated in his work, he defined the Byzantines as Hellenes, not as
Romans. So the word basileus did not give him the sense of uniqueness it
represented to other Byzantine writers.
5.3 The formation and early conquests of Anatolian beyliks in Byzantine sources
The role of the banditry and free bands of warriors in the Turkish conquest of
Anatolia was already mentioned. Now it is possible to focus more on the historical
outcome of such activities in the ex-Byzantine space. In this case, I shall deal with
the narratives about the formation of the Turkish beyliks, in Byzantine
historiography.
Until the end of the 12th century, it is possible to speak of a certain anti-Seljuk
resistance by the centrifugal policies in Turco-Muslim Anatolia. Naturally, the
Seljuks of Rum were always the most dominant force in this region and, because of
being a branch of the prestigious dynasty in Iran and had a somewhat primus-interpares
position in the political reality of the late medieval Anatolia. However,
Danishmendids established themselves in a relationship with the ghaza tradition of
Anatolia and the followers of the tradition of Sayyid Battal Ghazi. They were not less
prestigious than the Seljuks, particularly in the early 12th century.
One can assume that several Turkish warlords present in Western Anatolia
such as Karatikes or Elchanes in the Alexiad, could probably form a statelet in the
territories under their domination (such as Kyzikos and Sinope), but the vicinity of
the Rum Seljuk powerbase and Alexios I’s reconquest of these cities has prevented
160
such a formation. Yet, there are also Mengujekids and Saltuqids of the Eastern
Anatolia which were out of the focus of the Byzantine authors.297
After the destruction of these principalities by Seljuks and Ayyubids; until the
mid-13th century, there was a consolidation of Asia Minor under the Seljuks of Rum.
However with the gradual disintegration of the Seljuk sultanate, there was a new
wave of the centrifugal forces in Turkish Anatolia: in the last quarter of the 14th
century, several Anatolian beyliks appeared in the margins of the territories ruled by
Seljuks.
The formation of the beyliks in Asia Minor is a theme found in four
Byzantine historiographical texts (Pakhymeres, Gregoras, Doukas, Khalkokondyles).
The end of the 13th century and the beginning of the 14th century is the period of
both the total disintegration of Seljuk authority in Central Anatolia and the Byzantine
loss of Western Anatolia. The Seljuk Sultanate was already a puppet state, the vassal
of the Mongols of Persia, and was unable to control the Türkmen of frontier regions.
In the last quarter of the 14th century, there appeared some statelets that were called
beyliks, in these regions.
Germiyan and Karaman were the first beyliks formed, during the power
vacuum because of the decline of the Seljuks. Differently from the later beyliks, both
seem to have a tribal origin. Karaman, the eponymous founder of the beylik, was
depicted as a brigand in several Seljuk sources.298 Germiyan whose capital is
Kütahya (Kotayeion) in the Phrygia, appears to be suzerain of the other beyliks of the
western Asia Minor in the beginning.
297 For their history, see Sümer, Doğu Anadolu’da Türk Beylikleri.
298 Hopwood, “Peoples, Territories, and States: The Formation of the Beğliks of Pre-Ottoman
Turkey,” 132.
161
The Ottomans annexed these emirates, starting with the annexation of Karasi
in Mysia in 1345-1346. The Ottoman expansion toward Muslim Anatolia took a
century and a half. In the early 16th century, the Ottomans integrated the emirates of
Dulkadir and Ramazan, the last independent Türkmen beyliks in southern Anatolia
and they consolidated their rule in the whole region.
Pakhymeres’ account is chronologically the first narrative of these events. His
narrative is rather detailed and contains information absent from the other sources. In
his account, it is possible to see the Turkish warlords’ entry to the scene, not as the
founders of the dynasties, but as simple warlords:
The higher regions of Bithynia and Mysia, of Phrygia and Lydia, and of
famous Asia, with the only exception of the strongholds, had completely
ruined them. The perpetrators of these acts were the Amourios and the
Osmans, the Atines and the Alisher, the Menteshe, the Salampaxis, the Alaïs,
the Ameramanes, the Lamises, the Sphondyles, and the Pagdines, and any
other with a fatal and cursed name. Excited in their audacity by an
extraordinary arrogance and terrible, like a wildfire they occupied everything
and devoured it, prevented by this sea alone from advancing even further.299
Nikephoros Gregoras depicts the events as follow:
The Turks agreed and divided afterwards, the country in Asia under Roman
rule by lot. Karmanos Alisurios (Germiyan) received most of the Phrygian
inland and also the territory of Antiochia on the Meander to Philadelphia with
the entire environs. All country from this region to Smyrna, and the Ionian
coast in between, was given to another who was named Sarchanes; the area
around Magnesia, Priene and Ephesus, had previously been taken away by
another satrap, Sasan. The territory from Lydia and Aeolia to Mysia on the
Hellespont was given to Kalames and his son Karases, that around Olympus
and all of Bithynia again another one, Atman (ʿOt̲h̲ mān), and the territories
between the river Sangarios and Paphlagonia was distributed by the sons of
Amourios among themselves.300
299 Pachymeres, Relations historiques, 424-425.
300 Gregoras-Schopen-Bekker, Nicephori Gregorae Byzantina historia, 214-215; Gregoras-Van
Dieten, Rhomäische Geschichte, vol. 1, 174.
162
Doukas’ account is rather short:
During his (Andronikos II’s) reign Ephesus, the metropolis of Asia, and the
province of Caria fell to Menteshe. Lydia as far as Smyrna was taken by
Aydın. Magnesia as far as Pergamon and the entire province of Magedon fell
to Saruchan. All Phyrgia fell to Germiyan. Phrygia Magna, extending from
the city of Assos to the Hellespont, fell to Karasi. All Bithynia and part of the
land of the Paphlagonians fell to Osman. All were Turkish leaders.301
Finally, Khalkokondyles narrates these events in the context of the re-foundation of
these statelets by Tamerlane:
When Basileus ʿAlāʾ al-Dīn died and his leading men started disputing among
themselves, ʿOt̲h̲ mān is said to have entered into negotiations with them, and
they among themselves. He managed to forge a mutual military alliance with
them and took an oath that he would wage war in common with them all. They
would subjugate as much territory as they possibly could, and however much
land they conquered they would divide among themselves in accordance with
their common agreements. And so he marched out with them and subjugated a
large area, performing great deeds and amassing much money, so that in a
short time he acquired a considerable realm. There were seven leaders and
after this, they divided among themselves whatever territory had come into
their power. Karaman was allotted interior of Phrygia all the way to Kilikia to
Philadepheia, and Saruhan the coast of Ionian region as far as Smyrna.
Kalamshah and his son Karasi were allotted Lydia as far as Mysia, while
Mount Olympos and Bithynia were given to ʿOt̲h̲ mān and Teke. The sons of
Umur were allotted the lands toward the Black Sea and Paphlagonia. They say
that Germiyan was not among the original seven but had already become the
King of Iconion, a city in Karia, where they used to have their court for a long
time. But when he was driven out from there, he went to Ionia, where he lived
a peaceful private life. So these seven were the ones who together subjected
this whole land to themselves. However, there is no point in concerning
oneself with whether each acted on his own or in agreement with someone else
in some other way, when each obtained his realm.302
He later turns again on this subject, now adding two more principalities: Turgut303
and Metin.304
301 Doukas-Grecu, Historia Turcobyzantina, 33; Doukas-Magoulias, Decline and Fall of Byzantium to
the Ottoman Turks, 59.
302 Khalkokondyles, The Histories, 18-21.
303 Turgut was a small beylik in the Taurus mountains; it was a vassal polity under the suzerainty of
Karaman.
304 This name is impossible to identify. According to Kaldellis, it is a deformed version of the name
Hamid. So this is the principality of this name in Pamphylia.
163
After that, he moved against the remaining rulers (hegemons) in Asia, namely
Aydın, Saruhan, Menteşe, Teke and Metin. He stripped them of their realms,
driving them out, and appropriating their territory. Driven from their own
lands, they went off to King Timur, in Skythia. (…) I should add that
Saruhan, who governed the coast of Ionia, Menteşe who was the descendant
of Kalemşah; and Teke, who held Mysia, were descended from the seven
rulers who jointly assisted Osman in conquering the realm of Asia, and they
are said to have been servants of Basileus ʿAlāʾ al-Dīn. I have no specific
information as to how Metin and Aydin obtained their realms. It is said only
that Aydin ruled the land from Kolophon to Karia. I know clearly, however,
that the subjects of Turgut, Karaman, Metin and Aydın are Turks and are
called that.305
Khalkokondyles saw the Turkish conquest of Western Asia Minor as a result of a
coalition of several Turkish hegemons who jointly invaded the region. So he projects
anachronistically such an alliance of seven rulers against the Byzantine Empire in
early 13th century. The last phrase is also interesting: “the subjects of Turgut,
Karaman, Metin and Aydın are Turks and are called that”. What could it possibly
mean? It probably refers the ethnic composition of these beyliks: in the first three of
these emirates; there were a high number of nomadic populations in their territories.
Returning to our subject, the conquest of western Anatolia by Turkish beyliks,
it could be said that the idea found in the texts of Gregoras and Khalkokondyles, a
coalition of the Turkish warlords that allocated western Asia Minor to each other, is a
Byzantine historiographical myth. The work of Pakhymeres, the earliest of all four
accounts, does not provide such a piece of information.
Also in Doukas’ narrative the message is clear: there is no common strategy
or alliance among these Turkish leaders; every leader acted independently and
occupied former Byzantine regions. Among these four writers, Doukas was the one
who knew the recent history of western Anatolia the best. Thus, it seems that starting
305 Chalkokondyles, The Histories, 104-105.
164
with the work of Gregoras, Byzantine authors created such a teleological account to
explain the early expansion of the Anatolian beyliks.
5.4 Family intrigues and usurpers: Byzantine commentators on succession and
dynastic struggles in the Turkic states
In pre-modern monarchies, the event of ruler succession is a critical process that is
closely related to dynastic legitimacy. In medieval western societies, there were
several succession rules, such as primogeniture, tanistry or agnatic seniority, which
were used to make the order of succession predictable. On the other hand, in eastern
societies, the order of succession was not based on a generally accepted rule or an
official text like the Lex Salica. This often provoked succession crises.
In both Turkic and Muslim states, the same problem occurred as there was
not a universally accepted rule of succession. In Turkic states, there was another
important aspect, which was that all the state’s territory was considered the domain
of the royal family. So as the territory belonged to the dynasty, any member of the
dynasty could claim the throne. This kind of succession without strict rules can be
called “open succession”. In this system, the prince who manages to gain the support
of the different factions of the court and eliminate his brothers can inherit the throne.
It may be speculated that in such situations the older brothers were slightly more
advantageous because of their experience and possibly larger networks, but the
outcome was not always in favor of them. Such are two cases in Ottoman history;
when Bāyazīd I eliminated Prince Yakub and Selim I eliminated Prince Ahmed. In
the case of the Ottomans, the historian Doukas explained this principle as follows:
The new ruler need only be a descendant of ʿOt̲h̲ mān. The Janissaries looked
upon the Ottoman rulers as their patrons and the latter treated them as their
own freedmen. When it concerned the succession of one Ottoman ruler to
165
another or of father to son or of brother to brother, the regiment of slaves
faithfully served him whom Fortune favored.306
Below, I offer an overview of the evolution of succession, taking into account the
usurpers and rebellions in Turkic states, from the eyes of Byzantine authors who
were familiar with similar crises within their society in which succession crises were
not rare.
Since the earliest apparition of Seljuks in the Synopsis Historion of Skylitzes,
the Byzantine authors were aware of that notion of patriarchal state of the Turks, and
they remarked the lack of common consensus in their royal families and their endless
intrigues to get a higher part of royal authority.
In Skylitzes’ narrative, when the Seljuk Prince Ibrāhīm Yinal took Liparites
captive and brought him to the city of Re (Reyy) where his brother Tughril was
ruling as sultan, the latter gives “the appearance of rejoicing and gladness” but he
was jealous of his brother’s achievement and looks for some pretext to get rid of his
brother.307
A while after, Ibrāhīm Yinal notices the Sultan’s plots against him and rebel
against his authority with his nephew Koutloumous. However, Ṭog̲ h̲ rı̊l defeats the
insurgents and Ibrāhīm Yinal is executed. Afterwards, the other dissident escapes
with 600 men and with Melech, the son of Ibrāhīm Yinal.308
Skylitzes’ Synopsis Historion concludes with the end of the reign of Michael
VI Stratiotikos in 1057, but it is possible to follow the aftermath of the rebellious
306 Doukas-Grecu, Historia Turcobyzantina, 81; Doukas-Magoulias, Decline and Fall of Byzantium to
the Ottoman Turks, 136.
307 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 454; Sklitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 426.
308 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 474; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 442.
166
Seljuk Prince Koutloumous in other sources. According to the Seljuk sources, he was
killed in 1063 by his cousin Alp Arslan after another rebellion.
In Attaleiates, it is possible to see other dissident Seljuk nobles.
Chyrsoskoulos, Sultan Alp Arslan’s brother-in-law who arrives at the Byzantine
court, is one of them. The two “nobles of Persia” who “had inherited the name of
Koutloumous from their father” meet Nikephoros Botaneiates in Nicaea. The two
brothers who set a nucleus of state in the city - what will later become the Sultanate
of Rum - bend their knee before Nikephoros Botaneiates.309 As it was already
mentioned, the author emphasizes their Seljuk royal lineage. The author calls this
cadet branch of the dynasty “Κουτουλμούσιοι” and notes that their commanders call
these princes “ἀμηράδες” and “σελάριοι” in the Turkish language. These titles are
obviously “amir” and “salar”, military titles of Arabic and Persian origin.310
These two brothers were Sulaymāns̲ h̲ āh and Mansur who immigrated
westward after the execution of their father with a group of Türkmen nomads, called
Nawakiya collectively. As mentioned in Chapter 2, Sulaymāns̲ h̲ āh operated in
Anatolia in the 1080s and became the ruler of the Anatolian branch of Seljuks.
It is possible to follow later exploits of Sulaymāns̲ h̲ āh in the Alexiad. Anna
Komnene, though she presents Sulaymāns̲ h̲ āh’s operations in the Asia Minor as the
career of a warlord, carefully notes that he had the title “Amir” in the beginning, and
later promoted to sultanate. She seems aware of the internal divisions within the
Great Seljuk Empire and their dynastical struggles. Anna draws the portrait of
Τουτουσης (Tutush), the Seljuk sultan of Syria, as an ambitious and arrogant man
who has killed Sulaymāns̲ h̲ āh in battle and sought to overtake the Seljuk throne. For
309 Attaleiates, The History, 484-485.
310 Attaleiates, The History, 504-505.
167
this purpose, he summons “twelve bloodthirsty individuals called in the Persian
dialect Χάσιοı (the assassins) and send them at once as envoys to him (Malik-S̲ h̲ āh).
He gave them instructions as to the killing: Go, he said, and first of all make an
announcement that you have certain secret information for the sultan, and when you
are granted the right of entry, approach as if you desire to speak with him privately
and massacre my brother then and there.”311
The assassins go to the court of sultan and during a festive occasion, they get
over his guards and cut sultan to pieces. The author concludes the scene by writing
“The Χασιοı delight in that sort of bloodshed, their idea of pleasure is merely the
plunging of a sword into human entrails. Furthermore, should anyone happen to
attacks them at the very same moment and cut them up into mincemeat, they regard
such a death as an honour, passing on these bloody deeds from one generation to
another like some family heritage.”312
Sulaymāns̲ h̲ āh’s sons were hostages at the court of Malik-S̲ h̲ āh, and after the
murder they left the court and fled to Nicaea, their father’s capital. According to
Anna, “Poulkhases (Πουλχάσες) handed over the town to them as if it were a family
inheritance (πατρῷον κλῆρον). The elder son, Qiliç Arslān by name, received the
title of sultan (προχειρίζεται δε σουλτᾶν).” The plots in the House of Seljuk continue
with the next generation. After the death of Qiliç Arslān, his son Saisan (S̲ h̲ āhins̲ h̲ āh)
takes his place. However, his “bastard brother Masout,”313 who is “jealous of his
brother and plots to murder him,” gets support from a group of his satraps. Saisan
who is described as a “fool” by Anna, is then dethroned and blinded.
311 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 196; Komnene-Sewter-Frankopan,
Alexiad, 179.
312 However Malik-S̲ h̲ āh was not murdered as depicted by Komnene; he was poisoned.
313 This reference to bastardism could imply that Masʿūd was a son born by a concubine and Saisan
was born by the principal wife of Qiliç Arslān.
168
Thus, Anna suggests that the Seljuks perceived their state as a dynastic
patrimony which belonged to all the members of the branch of the family. However,
the ruler title generally passed to the elder son, hinting at a light system of
primogeniture. She also implies that the throne naturally belonged to the legitimate
son of the sultan, and not to the bastard. Her account is at times less accurate, for she
confuses the names of the rulers, but it is very illuminating for a reading which
focuses on the representation of the alterity. She presents the Seljuks as barbarians
who tended to be cruel and violent, yet who had a system of dynastic legitimacy and
succession.
John Kinnamos describes Qiliç Arslān II’s voyage to Constantinople where
he signs a peace treaty with Emperor Manuel I. According to this treaty, the Sultan
becomes an ally of the emperor, therefore, he is obliged to punish Türkmen tribes
that pillage Byzantine territories. However, the rumors of the treaty spreads “from
Europe to Asia”, and the tribal leaders (φύλαρχοι) become discontented of the deal.
Therefore, the leaders of the Türkmen tribes challenge the authority of the Sultan.314
Kinnamos’ narrative about the preparations of Manuel I of an unrealized
Anatolian campaign against Seljuks demonstrates that the Byzantine ruling élite
knew the inter-family conflicts in the ruling dynasty and how to use it for their
strategy to reconquer their lost land:
He (Manuel) wrote to his brother S̲ h̲ āhins̲ h̲ āh who governed Gangra and
Galatian Ancyra and to his son-in-law Yaghi-Basan who ruled both Kaisereia
and Amasia and other outstanding cities which are situated in Cappadocians’
land. After he rendered them suspect to the sultan, he was in a short time
ready for war.315
314 Kinnamos-Meineke, Ioannis Cinnami Epitome rerum ab Ioanne et Alexio Comnenis gestarum,
208; Kinnamos-Brand, The Deeds of John and Manuel Comnenus, 158.
315 Kinnamos-Meineke, Ioannis Cinnami Epitome rerum ab Ioanne et Alexio Comnenis gestarum,
200; Kinnamos-Brand, The Deeds of John and Manuel Comnenus, 151.
169
Yaghi-Basan mentioned by Kinnamos in this passage was the Danishmendid emir of
Sivas. Danishmendids, particularly the later ones, had a complex relationship with
the Seljuk rulers of Konya. The two states were rivals who both wanted to dominate
Turco-Muslim Asia Minor, yet at times were allied via dynastic marriages. The
Byzantines knew not only the internal tensions within the Seljuks but also the
Danishmendids. According to Niketas Khoniates,
Yaghi-Basan assembled his troops and drew up his forces in battle order, but
he was checked in his eagerness by death. Since Yaghi-Basan’s throne was
vacant, Dhul’Nun (Δανούνης) secretly entered the satrapy of Amaseia. There
he was repulsed and there he was the cause of the death of Yaghi-Basan’s
wife, who had secretly made Dhul’Nun ruler by marrying him; after he had
sent for her, the Amaseians rebelled and killed her. Dhul’Nun, whom they
held in contempt as a ruler, they expelled.316
As the passage demonstrates, the legitimacy of the succession was so important for
the Turkic states that even satrapies, as named by the author, and a usurpation
attempt could provoke urban rebellions.
Khoniates also discusses the Seljuk Sultan Qiliç Arslān II’s dividing of his
realms into small portions for his sons. After he counts which territory passed to
which prince, he narrates the civil war between the princes:
Ruknaddin (Sulaymāns̲ h̲ āh II) who was more clever by nature and exulted
exceedingly in warfare, outdistanced his brother (Masʿūd, the ruler of
Amasya and Ankara) and rival and carried off the victory. Since Masʿūd
submitted and agreed to a covenant of friendship, the more powerful
Ruknaddin took possession of only a portion of Masʿūd’s toparchy and
allowed him to govern there as before. He was especially maddened, however
by Kayk̲ h̲ usraw and suffered a burning passion for Ikonion, the paternal seat
of government; he also loathed him for having a Christian mother. Through
envoys, he advised Kayk̲ h̲ usraw to withdraw form Ikonion and remove
himself from all power if he wished to perform a good service and spare the
cities and the individuals and nations therein from the horrors of war. Thus
did the barbarian boast, unsurpassed in his arrogance, his eyebrows raised
316 Khoniates-Van Dieten, Nicetae Choniatae Historia, 122; Khoniates-Magoulias, O City of
Byzantium, 69.
170
above the clouds in scorn, as he poured out and scattered his deadly venom in
many directions.317
Later, Kayk̲ h̲ usraw accepted the defeat and fled to Constantinople where he would
remain until the death of his brother Sulaymāns̲ h̲ āh II in 1204. He would soon return
to Konya and take power.
The activities of the members of the Seljuk dynasty who were in exile can be
further traced in the Syngraphikai Historiai of George Pakhymeres. After the death
of Kayk̲ h̲ usraw II (1246), his three sons were proclaimed co-rulers by the Ilkhanate
that was the sultanate’s suzerain. His youngest son Kayqubad III died in childhood,
and then the Mongols divided the sultanate in two; they made Kaykāʾūs II the sultan
of western provinces and the frontier regions and Qiliç Arslān IV the sultan of
Eastern Anatolia. However, the conflict occurred between Kaykāʾūs II and the
Mongols, and after a military defeat, he fled to Constantinople with his family and
entourage. Pakhymeres provides a detailed narrative of Kaykāʾūs II’s life in exile and
the aftermath of his sons. After he gives a brief and ambiguous narrative of the
Mongol invasion of Iraq and Anatolia, he states that the Sultan, whom Pakhymeres
calls Αζατινεσ using only his first name İzzeddin, came to Constantinople with his
wives, children, sister, and his mother who was an “excellent Christian.” Emperor
Michael VIII Palaiologos receives Kaykāʾūs II in such a friendly way that he even
permits him to comport as a ruler in the Byzantine capital. The sultan was seated
next to the emperor in imperial stands with guards around him and used royal
insignia peculiar to Seljuks, such as wearing red shoes.318 Kaykāʾūs II spent the rest
317 Khoniates-Van Dieten, Nicetae Choniatae Historia, 521; Khoniates-Magoulias, O City of
Byzantium, 286.
318 Pachymeres, Relations historiques, 184. The use of red shoes as regalia in the Sultanate of Rum is
also noted in Oriental sources.
171
of his days in Constantinople “feasting and drinking,” resembling the Festivals of
Dionysus.319
As was seen in the abovementioned passages, the image of the members of
the Seljuk dynasty evolved in the eyes of Byzantine authors. The two authors of the
11th century –Skylitzes, who possibly based his account on oriental sources, and
Attaleiates, who presented the Kutlumusioi as leaders of a band of mercenaries that
managed to invade Nicaea, a city very close to the imperial capital– had very
superficial information about the inner mechanisms of the Seljuk Sultanate. Anna
Komnene’s representation had some similarities to theirs, but she accepted the
existence of a sultanate in Asia Minor.
Kinnamos remains silent about the inter-family relations within the Seljuk
dynasty, but as was demonstrated in the case of the Türkmen chieftains, he viewed
the sultan of Rum as a ruler of the federation of tribes who shared his authority with
the Türkmen chieftains.
Khoniates’ approach to the dynastic relations of the Sultanate of Rum was
also similar to that of Kinnamos, but he was clearly more informed on the political
realities within the Sultanate of Rum. As was stated in the previous chapter, the
Islamic allegiance of the Seljuks was slightly more emphasized in the History of
Khoniates. Being the brother of Michael Khoniates, the archbishop of Athens,
Niketas Khoniates seems to be more interested in religious issues.
Pakhymeres’ text demonstrates the ultimate image of a Seljuk sultan in the
eyes of a Byzantine intellectual: the ruler of a well-defined domain (Persia) that was
vanished and became a vassal of the Mongols, a tired and heavy drinking man who
unsuccessfully intrigues to recapture his throne, and who was unlike other Persian
319 Pachymeres, Relations historiques, 284.
172
brigands in Asia Minor that would later evolve into the founders of Anatolian
beyliks. Despite the pitiful tone in the representation of the sultan in exile, his vitality
and ressentiment, which ultimately brought him to treason against the emperor
Michael VIII, was demonstrated in a lively way in the Syngraphikai Historiai of
Pakhymeres. After his plot in 1264, in which he provoked the Bulgarians and Tatars
to attack Byzantine soil, the sultan was given to Tatars who transferred him to
Crimea.320
In Gregoras’ text, there are some observations on succession in the Ottoman
state: while the author mentions Prince Khalīl, the son of Ork̲ h̲ an, he makes some
statements about the mechanism. According to Gregoras, Khalīl inherited the area
around the bay of Nikomedeia from his father and ruled there in an autonomous way.
This description is consistent with the Ottoman practice of appanage. The author then
mentions the death of Prince Süleymān (d. 1357) and states that he was the eldest son
and successor. However, Prince Khalīl was the son of Ork̲ h̲ an and Theodora
Kantakouzene and betrothed to one of the daughters of John V Palaiologos. So his
Byzantine imperial descent and the marriage alliance with the Palaiologoi made him
a more adequate successor to the Ottoman throne for the Byzantines. The author
mentions that the emperor wanted Ork̲ h̲ an to proclaim Khalīl as the official heir;
nevertheless, Gregoras’ History ends before the death of Ork̲ h̲ an and there could not
be read the conclusion of this unsuccessful project.321
320 Pachymeres, Relations historiques, 300-312. The details of Kaykāʾūs II’s plot are discussed in
Vásáry, Cumans and Tatars: Oriental Military in the Pre-Ottoman Balkans, 1185-1365, 72-79.
321 Gregoras-Schopen-Bekker, Nicephori Gregorae Byzantina historia, vol. 3, 558-566, 501-510;
Gregoras-Van Dieten, Rhomäische Geschichte, vol. 6, 171-187.
173
CHAPTER 6
THE MECHANISMS OF CO-EXISTENCE:
ANTAGONISM, ACCULTURATION, ASSIMILATION
6.1 A conceptual introduction
This chapter will examine the entry of people of Turkic origin into Byzantine society
and their further interactions with the Byzantines. As an introduction to the three
major concepts I will be dealing with –namely, integration, assimilation, and
acculturation– I would like to begin with a brief semantic discussion of them.
Integration is a modern sociological term that generally implies the successful entry
of migrants or outsiders into contemporary societies. Integration comes from the
Latin root integrare, which means to fuse or to merge. It implies a group of people’s
fusion with another group of people, of whom the latter seem to be more populous or
dominant. The terms integration and assimilation must be distinguished. The word
assimilation etymologically comes from the Latin root similis, and according to The
Cambridge Dictionary of Sociology it means “the process in which outsiders
(especially migrants) give up their distinctive culture and adopt the cultural norms of
the host society.”322 However, the term integration indicates a relationship where the
identity and culture of the “integrated” are still the original. Today generally
integration has a positive, but assimilation a negative meaning. Moreover, the
processes of assimilation and integration cannot be considered mutually exclusive,
and assimilation generally follows a successful process of integration. Acculturation
is semantically closer to assimilation, but it refers to a dynamic process rather than an
event or a situation. It is a term strongly associated with American sociology, and it
322 Ray, “Assimilation,” 24.
174
was used primarily for the social evolution of the immigrant communities in the
United States.
6.2 The entry of the Turks into Byzantine service: Pechenegs and Seljuks
Turkic peoples offered occasional mercenary service to the Byzantine Empire at least
since the 6th century. As it was already seen in Chapter 3, there were Hunnic
mercenaries in the army of Justinian II. In later centuries, the Byzantines employed
Avars, Bulgars, and Khazars for various military services. In the 11th century, the
Pecheneg warriors are attested among the Byzantine ranks; they served faithfully
even at the Battle of Manzikert. It is possible to speculate that this “barbarian
manpower” was dependent on the Turkic nomads who occupied various parts of the
Balkan peninsula throughout the early medieval centuries. This usage of foreign
manpower in wars secured the service of the “barbarian populations” to the
Byzantine Empire. As it was seen before, De Administrando Imperio demonstrates
that the Byzantine Empire was the most dominant power in its region because of its
rather complex diplomatic policy toward its neighbors, particularly with the steppe
peoples dwelling beyond its northern borders. This diplomacy included both an
elaborate system of alliances with these populations (which involved the use of a
population or tribal group against the other) and the use of their military forces as a
source of military manpower.
However, before dealing with this question, the terminology regarding these
acts should be explained. To enter into service and to defect have slightly different
meanings: The latter term implies an antagonism between two parties. This
antagonism could be the reflection of an ideological antagonism (as it was seen in
earlier chapters; such as the antagonism between the Christians and Muslims) or the
175
harshness of a momentary confrontation. Furthermore, it indicates a member of the
ruling family or elite’s act of passing to the other side. However the former term has
a much lighter signification, and I use the verb for the individuals who were not part
of the ruling élites of their societies. In our case, many active Turkish warlords in
Asia Minor in the 11th century did not have any important ties, either with the Great
Seljuks or the emerging Sultanate of Rum. In my conceptualization of the primary
antagonism and secondary antagonism, they represented the secondary antagonism,
in their confrontation with the Byzantines. As I have already mentioned, this attribute
is also suitable for the pagan Turkic populations in the steppe region, such as the
Pechenegs.323
Both John Skylitzes and Michael Attaleiates demonstrate to us the essential
patterns of this imperial effort to integrate Pechenegs into the Empire. These patterns
were not fundamentally different from Byzantium’s earlier experiences with other
nomadic groups.
Skylitzes gives a rather detailed narrative of the internal conflicts of Pecheneg
tribes who were dwelling in the north of the Danubian frontier of the Byzantines. He
presents two Pecheneg leaders as the protagonists; Tyrach (Τυράχ) and Kegenes
(Κεγένης). Kegenes is described as a self-made man, a man who does not belong to a
noble house but is known for his warlike qualities. On the other hand, Tyrach is
“highly distinguished by birth, but otherwise unremarkable.”324 So Tyrach was a
member of the Pecheneg tribal nobility. Among the Pechenegs, although Tyrach
323 There are several studies focusing on coexistence and fluidity of identities between the two sides.
The most important ones are: Brand, "The Turkish Element in Byzantium, Eleventh-Twelfth
Centuries;” Necipoğlu, “Turks and Byzantines (Eleventh and Twelfth Centuries);” Beihammer,
"Defection across the Border of Islam and Christianity: Apostasy and Cross-Cultural Interaction in
Byzantine-Seljuk Relations."
324 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 455; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 427.
176
attained respect for his family origins, Kegenes was much more popular. While
Tyrach was trying to get rid of his rival, Kegenes started a rebellion with the support
of a small portion of Pecheneg society. However, Kegenes lost the first round of war
and fled to Constantinople. Skylitzes gives an account of this event:
Kegenes came to the capital where he was generously and graciously
received in audience by the emperor. On promising to accept baptism himself
and to persuade his followers to do likewise, [Kegenes] was raised to the
dignity of patrician; he received three of the fortresses standing on the banks
of the Danube and many hectares of land. Finally, he was inscribed among
the friends and allies of the Romans, all this because he and his followers
accepted baptism (as he promised). Euthymios, a devout monk, was sent to
administer the sacred bath by the Danube river, giving them all holy
baptism.325
So basically Kegenes, who came from modest roots, goes back to the frontier region
where his people were dwelling, after having obtained a Byzantine noble title and
administrative authority. In return, he leaves his ancestral paganism and converts to
Christianity. This return could have threatened the traditional Pecheneg society on
three different grounds. Firstly, obtaining the patrikios title was a challenge to the
traditional nomadic aristocracy of the Pechenegs. Secondly, the conversion to
Christianity represented a challenge to the Pecheneg religious landscape, which
appears to be similar to other versions of Tengrism or paganism practiced by Turkic
peoples. His settlement of unnamed castles on the banks of the Danube, however,
was the greatest threat. It marks the passage to sedentary life and it is a challenge to
the traditional nomadic social order of Pechenegs.
Some twenty years later, Attaleiates described this region as “the region of
the mixobarbaroi (μιξοβάρβαροι) who dwell by the Danube,” adding that “[t]here are
numerous and large cities by its shores whose inhabitants constitute a multilingual
325 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 457; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 428.
177
crowd and support a large number of soldiers. To those cities the Skythians who had
previously crossed the river have introduced their ways of life.”326 So the Danubian
frontier region is a zone of passage between the Byzantine oikoumene and the steppe
world. I dealt with this frontier in Chapter 2 and presented a comparison with the
empire’s eastern/south-eastern frontier where the confrontation with the Muslims
happened.
According to Skylitzes, Kegenes, who once settled in the frontier zone, now
started to engage in raids with his 2,000 men against the Pechenegs who remained
with Tyrach, and “they would slaughter the men they encountered, but enslave the
women and children and then sell them to the Romans.” So Tyrach sent a delegation
to emperor Constantine IX Monomachos and requested the ending of these
incursions. The Byzantine refusal of this request triggered a Byzantine-Pecheneg
war. Although in the early stages of the war Tyrach’s horde made some gains, they
surrendered in the end.
When the Pechenegs surrendered, Kegenes advises to kill all men who could
bear arms, but this advice was seen as barbaric and impious. In conclusion, Tyrach’s
Pechenegs were told to settle in Bulgaria, in the plains of Sardike, Naissos, and
Eutzapolis. As described by Skylitzes: “They were all well spread out and completely
stripped of weapons to guard against uprisings.” The author concludes his narrative
by saying: “as for Tyrach and his hundred and forty followers, these were brought to
the emperor who received them benevolently, had them baptized and awarded them
highest honors, entertaining them in luxury.”327 This policy also divided Pecheneg
society in two: while the followers of Kegenes were quartering in the Danubian
326 Attaleiates, The History, 372-373.
327 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 459; Skylitzes-Wortley, A Snopsis of
Byzantine History, 811-1057, 430.
178
frontier, Tyrach’s Pechenegs settled in Central Bulgaria.328 Moreover, the emperor
formed a Pecheneg division of 15,000 men, aiming to use them mostly on the eastern
borders. He appointed four Pecheneg warriors to command them: Soultzous (his
name appears as an accusative: Σουλτζοῦν), Selte (Σελτέ), Karamas or Karaman (his
name appears as an accusative: Καραμᾶν) and Kataleim (Καταλείμ). Although the
author only states that the emperor “then showered them with gifts, providing them
with first-rate weapons and excellent horses,” it can be guessed that they also
converted to Christianity.
However, this treaty did not secure the peace in the Balkans, inasmuch as the
Pechenegs did not cease to make incursions on the Roman territories. Moreover, the
Pechenegs are now dwelling in a region much closer to the Byzantine capital. Under
these conditions the emperor Constantine IX called Kegenes to Constantinople,
however before this meeting, he became the target of an assassination attempt by
other Pechenegs. Strangely, Valtzar (Βαλτζάρ), the son of Kegenes takes the
assassins as the prisoners, but he does not punish them right away. This fact garners
the attraction of the emperor; Valtzar explains the impunity by saying “because they
were invoking your name.” After this answer, Constantine IX questions the would-be
assassins by asking them why they wanted to kill a patrician and they state: “Because
he was evilly disposed towards your reign and to the city; he was intending to enter
the city at dawn, to slaughter everybody in it, pillage the city and return to the
Pechenegs.” At this point, Skylitzes makes a moral comment that he expresses rarely
in his work and criticizes the emperor for his lack of good willingness: “He ought to
have examined these statements to find out the truth, but that is not what he did; he
328 See Florin Curta’s comments on the Byzantine attempts to settle Pechenegs. Curta, Eastern Europe
in the Middle Ages (500-1300), 166-167.
179
put his faith in some irresponsible and inconsistent accusations.”329 Furthermore, he
arrests Kegenes and his sons and releases the would-be assassins. The author
describes this attitude as “a clear sign of malevolence.” The same night, the would-be
assassins leave Constantinople in secret and return to the Balkans, establish
themselves at Aule, near Adrianople and begin to raid the region. Slightly later, these
Pechenegs defeat the armies of Constantine Arianites near Dampolis and their
insurgency becomes a great threat. Under these circumstances, the emperor decides
to use Tyrach and his entourage to pacify the Pechenegs. However, during the new
campaign against Pechenegs, Tyrach and his comrades leave the Byzantine army and
join their compatriots.330 Two more campaigns aimed to pacify the Pechenegs,
ending with Roman defeat.
At this point, the emperor Constantine IX decides to set Kegenes free and
utilize him against the Pechenegs. Kegenes helps the Byzantine to pacify the
Pechenegs violently. He makes a treaty with the Pechenegs who “promise him with
oaths to do whatever he wanted;” however, once the treaty is made, his compatriots
do not honor it, and “he was promptly murdered and cut up into small pieces.”331
This murder triggers another Byzantine campaign to punish the Pechenegs and the
commanders Nikephoros Bryennios and Michael the Akolouthos massacre large
groups of Pechenegs. Skylitzes concludes the passage by stating: “This reverse put
fear and caution into the Pechenegs; in the fourth and fifth years of indiction they no
longer raided with impunity as before, but sporadically.”
329Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 466; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 435.
330Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 468; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 437.
331Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 472; Skylitzes-Wortley, A Synopsis of
Byzantine History, 811-1057, 440.
180
This account gives us some insight into how a nomadic population could be
integrated into Byzantium: Generally there is a difference between the imperial
strategy in dealing with the tribal elite and commoners. The imperial strategy
prioritizes the agreements with (at least a clique of) the notables. The Byzantine
policymakers were never secure of the fidelity of the tribal chiefs and because of it,
they never cease to use a faction against another. (As it was seen, firstly Kegenes
against Tyrach, then an agreement with Tyrach, then another alliance with Kegenes,
then using the faction of murderers against Kegenes, then Tyrach against the faction
of murderers, then Kegenes against Tyrach and the faction of murderers.) The
imperial authorities frequently call the leaders of factions to their capital and try to
make them content with titles and gifts. The presence of the members of the notables
of a nomadic society in Constantinople is useful for two reasons:
i) Their presence as refugees at the imperial capital makes them good candidates
for a pro-Byzantine tribal leadership.
ii) Their presence as vassals (similar to the foederati of Late Antiquity) makes
them easy to control and prevents them from revolting.
In this strategy, there was nothing related to the Pecheneg lower classes. They are
only passive elements of society.
Despite Skylitzes only mentioning the baptism of the Pecheneg élite, this fact
must not be limited to the nobility. Also, the lower classes of the Pecheneg nation
must be evangelized to some point as a way to integrate into the Byzantine peasantry.
Michael Attaleiates gives an alternative narrative about these events.
However, he mentions neither Tyrach nor Kegenes. Indeed, he does not give any
narrative about the internal structure or struggles of Pechenegs. According to him,
after the oppression of the first Pecheneg invasion of the Balkans (c. 1047), these
181
nomads started again to raid the Byzantine territories as “the snakes warmed up by
the heat.” Emperor Constantine XI Monomachos’ resolution to this agitation was a
rapprochement with the Pecheneg leaders: The emperor’s plan was to send their
leaders in the hope that they might bring their people to their senses. He had honored
them with the rebirth of holy baptism and the greatest gifts, and hoping to use them
to avert war, or so he thought, he spared their lives and restored them to their own
clans.332
However, the Pecheneg warriors –evangelized or not- were not regarded
without suspicion. Attaleiates says that “wanting to lift the suspicion that hovered
over the Skythians, I myself advised the emperor to bind them an oath. He accepted
my advice and right away appointed me to execute and oversee the matter.”333 So to
secure the fidelity of the Pechenegs, there are always necessary extra measures such
as the oaths.
In the Alexiad, there are some more references about the Byzantine policy
toward the Turkic nomads. In this work, the Cumans, a Turkic population recently
arrived at the Balkan frontier of the Empire, enter the narrative. It seems that there is
an obvious hostility between Cumans and Pechenegs that creates sporadic wars.
Anna Komnene narrates early wars between Alexios I Komnenos and the Pecheneg
chieftain Tzelgou (Τζελγού). Then later these Pechenegs were ultimately defeated by
Cumans who appeared beyond the northern frontiers and fled toward the Lake of
Ouzolimne (Οὐζολίμνη).334 According to Anna Komnene, the imperial strategy was
332 Attaleiates, The Histories, 54-55.
333 Attaleiates, The Histories, 288-289.
334 Anna Komnene explains its etymology with the ethnonym Oghuz: “It has been called Ozolimne,
not because it emits an evil or unpleasant odour, but because an army of Huns once visited the lake
(those who used to be called Huns are now commonly known as Ouzes) and camped by its banks. The
name Ouzolimne was given to it, with the addition of the vowel ‘u’. No congregation of Huns in that
area has ever been mentioned by ancient historians, but in the reign of the Emperor Alexios there was
a general migration there from all directions – hence the name. Komnene-Reinsch-Kambylis, Annae
182
“to make use of Scythians against the Cumans, if the latter again approached the Ister
and tried to seize territory beyond it.”335
Another Turkic population that appeared on the northern edges of the Balkan
frontier of the Byzantine Empire were the Oghuz (Οὖζοι). I already mentioned these
Oghuz who were the westernmost fraction of Oghuz nomads that were neither joined
to the founders of the Seljuk state, nor remained in the ancestral pastures in
Transoxiana, but followed the westward movement of Pechenegs and Cumans in the
Pontic Steppe. They had also a limited presence in the region.336 During the ill-fated
Balkan expedition of the Norman prince Bohemond Hauteville (he had not become
prince of Antioch yet), a certain man called Ouzas (Οὐζᾶς) is described by Komnene:
(A man) who owed his name to his race, a man famed for courage and one
who knew how to wiled the dried bull’s hide to right and left, as Homer says,
when he emerged from the pass, swerved slightly to the right, swiftly turned
and struck at the Latin behind him. The man at once fell head first to the
ground. Nevertheless, Bohemond chased them to the River Salabrias. In the
flight, however, Ouzas, whom I mentioned already, speared Bohemond’s
standard-bearer, snatched the insignia from his hands, waved it around a little,
and then pointed it towards the ground.337
The same Ouzas appears again during a battle against the Pechenegs; he is always
described as a “Sarmatian” and his comrade Karatzas, another commander of the
Byzantine Empire, a “Skythian”. Ouzas could secure the place of his descendants in
the Byzantine army; at the end of the Alexiad, the son of Ouzas appears in the
description of a battle against the Seljuks of Rum. As it was stressed by the author,
the name Ouzas seems to be identical to the demonym Oghuz.338
Comnenae Alexias, 217; Komnene-Sewter-Frankopan, The Alexiad, 198.
335Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 216; Komnene-Sewter-Frankopan, The
Alexiad, 197.
336The majority of these Oghuz entered the service of princes of Kievan Rus and are collectively
known as Torks or Cornije Klobuki in medieval Russian sources.
337Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 160 ;Komnene-Sewter-Frankopan, The
Alexiad, 145.
338Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 476 ;Komnene-Sewter-Frankopan, The
183
Naturally, the background or exact way to enter into Byzantine service of
every soldier of Turkic origin is not known in detail. However, the essential patterns
of entry can be seen as
i) the members of the ruling élite of their nomadic populations/tribes
ii) the leaders of independent war bands/mercenary groups.
The entry of the persons who were once members of Seljuk society into
Byzantine service follows basically the same patterns. In the moment of a political
conflict, the ones who want to seek an alliance with the empire or provide Byzantine
assistance to his cause could easily approach the empire. Nonetheless, an essential
difference between Pechenegs and Seljuks should be mentioned: The Pechenegs
were a tribal federation. There was no real “Pecheneg State”. This Pecheneg polity
which appears as a federation of multiple tribes could be defined only as a “protostate”
or “state nucleus”. However, the Seljuks –which were a coalition of Türkmen
warrior bands assembled around the charismatic warlords of the Seljuk family– could
become a state in a few decades. Despite their tribal origins and early relations with
the Turkic homeland, their state was developed rapidly into a sultanate, similar to
Buyids or Samanids, and adopted traditional Islamic institutions.
Could religion also diminish the enthusiasm of Muslim Seljuks to enter
Byzantine service? One can reject easily this argument. Regarding the cases of the
11th century, it is hard to associate any Seljuks with a deeper understanding of Islam,
as it was already discussed in Chapter 3. The lifestyle of the early Seljuks was not
very different from that of Pechenegs, as it was seen in the earlier chapters. In the
subsequent centuries, although one can presume that Islam became much more
established in Anatolia, these entries continued. Many members of the Seljuk ruling
Alexiad, 446.
184
family went to exile in Constantinople or defected to Byzantium. Rustam Shukurov
explains this fact with his “dual identity” hypothesis; he argues that an important part
of the members of the ruling elite of the Seljuks of Rum, including the sultans, had a
dual identity (Muslim-Christian) and because of this they could live comfortably in
the Byzantine world. He explains it as follows: “Dual identity supposes that one of
the two identities is in active mode while the other is in deferred mode when in a
Christian environment, such persons would identify themselves as Christian,
deferring their Muslim identity. They would, however, embrace their Muslim identity
when in a Muslim space, in turn deferring their Christian self for the time being.” So
when they went to Constantinople, they behaved just like other Greek Orthodox
Christians.339
The first Seljuk commander who entered Byzantine service was
Chyrsoskoulos or Arisghi, who was the brother-in-law of Seljuk sultan Alp Arslan.
His story is somewhat different from the later defections. Chyrsoskoulos defeats
Manuel Komnenos in battle and then “had of its own accord decided to join the
emperor and was bringing the general (Manuel Komnenos) along with him. He
would rather be known as a servant of the emperor than the grand commander of the
Huns.” But then the author explains the real reason for Chrysoskoulos’ defection:
“He came to the Imperial City having left behind his own forces and made his
decision to change sides, though it was not fully voluntary. The reason was that the
sultan governing Persia (Alp Arslan) was ill-disposed toward him as though he were
a traitor, and had sent out one of his captains with an army against him. He was
seized with fear and could think no other way of escaping the danger than to seek
339 Shukurov, "Harem Christianity: The Byzantine Identity of Seljuk Princes," 134.
185
refuge with the emperor of the Romans.”340 So his defection follows the same pattern
as the early similar events. The difference between the way of the defection of
Kegenes and Chrysoskoulos was related to the administrative system of their
respective societies. Kegenes was part of a society in which a tribal aristocracy
controlled the structures of power and he could gain support from a smaller portion
of it. On the other hand, Chrysoskoulos came from a society where only one ruling
family dominated central power and he was related to this family by the bond of
marriage.
This defection of Chrysoskoulos was the beginning of a chain of defections
from Seljuks to the Byzantine Empire. This defection reflects a pattern that will
repeat in later generations: the individuals who do not possess the power to challenge
successfully the central authority of the sultanate or those who feel themselves under
a threat. In these times there was also always - at least one – pretender who was the
ally of the Byzantines.
But even Chrysoskoulos had a precedent: Amertikes (Ἀμερτικῆς) a Turkish
warrior who came as a refugee to the Byzantine court. Although Attaleiates presents
him as “an energetic man...claimed to be of the imperial family of Persia,” it seems
that he is not a member of the House of Seljuk. Claude Cahen explains his name as
Harun Ibn Khan. He was the leader of a Turkish war band that looted Kilikia. After
talking about the atrocities of Amertikes’ band in Kilikia, the author returns to the
past and narrates his early moves: “He was very hostile to the Romans because he
had been deceived in his dealings with them. He had formerly come to the emperor
of the Romans, who was then the Old Man [i.e. Michael VI Bringas] and was
340 Attaleiates, The Histories, 258-259.
186
splendidly received during his stay in the Reigning City. But he was accused before
the emperor Konstantinos [X] Doukas of plotting to stab him and was condemned to
exile.” Even so, after his return from exile, he was employed to fight against the
Seljuks, but he defected to the Mirdasid ruler of Aleppo because he could not get
money for his soldiers’ rations.341 Amertikes appears a decade later in Northern
Syria, as the commander of a Turkish contingent that was allied with the
Mirdasids.342 Amertikes was an one of the leaders of free war bands who had no
indispensable tie with any political entity. Such warlords could easily move between
different states or centers of power.
Furthermore, Anna Komnene’s Alexiad gives us also several details of
mechanics of entry of the Turks into Byzantine service. Among the three examples
she has given, Elchanes (Ἐλχάνης), the satrap of Kyzikos, Skaliarios (Σκαλιάριος),
and an unnamed comrade of arms of him were already mentioned.
She also mentions a certain Siaous (Σιαούς), who was the ambassador of the
Seljuk sultan to the emperor Alexios I. His name could be both a Greek rendering of
the Persian given name Siyavuş or the Turkish military rank çavuş. He was a Turk on
his father’s side and a Georgian on his mother’s side. Being the ambassador of the
Great Seljuks, he brought the letter of sultan Malik-S̲ h̲ āh to establish peace between
two forces. After his meeting with the emperor at Constantinople, he went to Sinope
and some other cities under the control of Turkish warlords (such as Karatekin,
Χαρατικής, the ruler of Sinope) and “showing the sultan’s order, removing the
satraps and reinstating the emperor’s satraps in their place.” After his mission in Asia
341 Attaleiates makes a bitter comment about Amertikes’ defection: “Now, whether it was prudent to
send this barbarian against his own people, especially when he had been ill-treated and would not
have even received the promised rations money, I leave to my readers to investigate.” Attaleiates, The
History, 172-173.
342 Attaleiates, The History, 198-199.
187
Minor, he returned again to Constantinople, entered Byzantine service, converted to
Christianity, and became the doux of Ankhialos.
In Kinnamos’ work, there is rich material about the Turks who entered
Byzantine service: The first of them is Prosuch (Porsuk), who could be the same
person as the one referred to in the Alexiad.343 The author describes him “as a Persian
by birth, but who had enjoyed a Roman upbringing and education” and he appears as
a commander who served in campaigns with John Kontostephanos. Another Turkish
soldier who entered Byzantine service was Poupakes (Abu-Bakr?), who is also
described as “a Persian by birth.” He was employed for spying on the Turks near the
Byzantine camp. The commanders who ordered Poupakes to fulfill this duty were the
brothers John and Isaac Axouchos. There should be no doubt on the dangerous nature
of this task. A person assigned to such a job must be free from any suspicion of
treason. Axouchoi’s choice demonstrates that they see him as totally trustworthy. It is
possible to imagine that the Turkic origin of the Axouchoi may have created an
affinity between them and Poupakes and there may have been an ethnically based
solidarity of Turks in the Byzantine court. Kinnamos’ description of Poupakes is
totally positive; he states that he is a man who “possessed great courage and
activity.”344 The same person also appears in the work of Khoniates, in the battle of
Kerkyra between the Byzantines and Normans. In a difficult moment of the war, after
the proclamation of “He who loves the emperor and is eager to distinguish himself in
the face of danger, let him ascend,” Poupakes impresses the emperor by ascending
the ladder on the besieged citadel. He is described as a devoted and brave warrior,
343 Kinnamos-Meineke, Ioannis Cinnami Epitome rerum ab Ioanne et Alexio Comnenis gestarum, 33-
34; Kinnamos-Brand, Deeds of John and Manuel Comnenos, 35.
344 Kinnamos-Meineke, Ioannis Cinnami Epitome rerum ab Ioanne et Alexio Comnenis gestarum, 47-
48 ;Kinnamos-Brand, Deeds of John and Manuel Comnenos, 45.
188
without any reference to his barbarian origins. Afterward, he unintentionally helps
the escape of Andronikos Komnenos and is publicly scourged for that crime. In front
of the crowd, he defends himself by saying: “Let my shame be before every man
who so wishes for not having betrayed my benefactor who came to me, for not
having dismissed him harshly, but instead attending rightly to his needs and sending
him rejoicing on his way.”345
In the later part of the narrative, a Poupakes346 appears again, possibly the
same person as the former one who was presented as the nephew of a certain
Süleymān who was the governor or chief of the westernmost territories of the Seljuk
sultanate. This Poupakes serves as an envoy between the emperor Manuel I and his
uncle.
Among the captives taken on the battlefield in 1146, he mentions a certain
Pharkousas (Φαρκουσᾶς), who was the “cupbearer”347 of the Seljuk sultan.
Kinnamos describes him as “an outstanding man among the Turks,” and even though
he never mentions him again, the knowledge and positive statement about him could
also imply that this Pharkousas also passed to the Byzantine side. Another Byzantine
officer who appears to be of Turkic origin is Ishaq (Ἰσάχ), who was also in charge of
a very intriguing task, to inform the emperor Manuel I Komnenos about the
conspiracies of his cousin (later emperor) Andronikos Komnenos. However, before
Ishaq arrived at the emperor, he became aware of the conspiracy. The author
describes Ishaq as “a man of barbarian descent who was a particular favorite of the
345 Khoniates-Van Dieten, Nicetae Choniatae Historia, 172; Khoniates-Magoulias, O City of
Byzantium, 75.
346 According to Brand, they are not same persons. See Brand, “The Turkish Element,” 8.
347 This must be a translation of the Seljuk/Ottoman court title şarabdar. Doukas uses the original
Turkish title in his History, nearly three centuries later. Doukas-Grecu, Historia Turcobyzantina, 235;
Doukas-Magoulias, Decline and Fall of Byzantium to the Ottoman Turks, 164.
189
emperor.”348 We also learn that at a certain point this Ishaq converted to Christianity
and took the baptismal name Michael.349 Furthermore, Kinnamos cites Bayram
(Παϊράμης) as “a man tried in battle” and “a Persian by race” who fought against the
Normans of Sicily. Along with his comrade-in-arms John Kritoples and Georgian and
Alan units, he managed to take a Norman standard and brought it to the Byzantine
headquarters in Brindisi.350
Despite their conversion to Christianity, a typical accusation used against the
Turks is sorcery. Michael Italikos, in his letters that addressed Tziknoglou, warned
his friend of Turkish origin not to employ a sorcerer to practice Chaldean magic to
heal his sick sister. The same accusation is echoed also in Kinnamos when he
narrates the trial of Alexios Axouchos, who served as protostrator under Manuel I
Komnenos. Among the crimes that Axouchos was accused to have committed,
Kinnamos counts “he frequently invited into his presence a man, a Latin by birth but
a magician and outstanding in wizardry, unfeignedly conversed with him, and
communicated monstrous plots. These were as to how the emperor might always be
unfortunate in lack of an heir; he used to receive many drugs from the wizard for the
said purposes and the wretch did not leave off doing such things.” Khoniates, who
describes the emperor Manuel I’s attitude against Alexios Axouchos as “wrongdoing
and disgraceful”, finds these accusations baseless.351 According to him, these claims
were fabricated by a group of people that includes Aaron Isaakios of Corinth and
348 Kinnamos-Meineke, Ioannis Cinnami epitome rerum ab Ioanne et Alexio Comnenis gestarum, 129
Kinnamos-Brand, The Deeds of John and Manuel Comnenus, 102
349 Kinnamos always uses the term barbaros for Ishaq, he never states that he is Persian (or Turkish).
His name is obviously Islamic, perhaps there is a possibility that he is Arabic or from another
predominantly Muslim ethnicity.
350 Kinnamos-Meineke, Ioannis Cinnami epitome rerum ab Ioanne et Alexio Comnenis gestarum, 167;
Kinnamos-Brand, The Deeds of John and Manuel Comnenus, 128.
351 Khoniates-Van Dieten, Nicetae Choniatae Historia, 187-188 ; Khoniates-Magoulias, O City of
Byzantium, 82.
190
they “secretly induced to accuse Alexios of using his powers of witchcraft against the
emperor, powers which were so illusory and efficacious that the sorcerer could fly in
the air and remain invisible to those upon whom he wished to swoop down with
sword in hand; their other buffooneries and vulgarities to which sound ears ought not
to listen were such as those of which the Hellenes, fabricating fables, accused
Perseus.”352
This accusation of sorcery suggests that the Byzantines had some doubts
about the Turks’ dedication to their new religion, Orthodox Christianity. Although in
the second case the wizard is a Latin (i.e. Italian), the Turks’ connection with the fact
is obvious.
Of course, though it was seen as a sin in Byzantine society, sorcery was everpresent
in social life. In Byzantine society, as in other societies, generally, sorcery is
identified with the alien populations, particularly with those who did not share the
Orthodox religious practices of the majority. Also in the ancient Turkic religion,
which is sometimes defined as shamanism, there were religious practices that could
be defined as sorcery.353
However, Alexios Axouchos had committed two bigger crimes: in his
mansion, there were paintings that depicted the martial deeds of Qiliç Arslān II and,
more importantly, he bribed the Skythian mercenaries to assault the tent of the
emperor during his campaign against the Hungarians.354 These details are only found
in Kinnamos’ work. Apart from the accusations of sorcery, it is clear that both
352 Khoniates-Van Dieten, Nicetae Choniatae Historia, 187-188; Khoniates-Magoulias, O City of
Byzantium, 82.
353 Brand, “The Turkish Element,” 7 (for the case of Tziknoglos) and 9 (for the case of Alexios
Axouchos). Kinnamos-Meineke, Ioannis Cinnami epitome rerum ab Ioanne et Alexio Comnenis
gestarum, 267-268; Kinnamos-Brand, Deeds of John and Manuel Comnenus, 200-201. For Turkic
shamanism see İnan, Tarihte ve Bugün Şamanizm.
354 Kinnamos-Meineke, Ioannis Cinnami epitome rerum ab Ioanne et Alexio Comnenis gestarum, 267-
268; Kinnamos-Brand, Deeds of John and Manuel Comnenus, 200-201.
191
Alexios’ alleged sympathies for the Seljuk ruler and his intrigues with the Skythian
(i.e. Cuman) mercenaries were related to his Turkic origins. Alexios Axouchos, even
though he was a second-generation member of an aristocratic family that had close
ties with the imperial court, could not escape from such allegations. Independent
from the degree of truth of such accusations, his Turkic background associated him
with sorcery and collaboration with the Seljuks and Cumans.
There is a striking difference between the representation of the Turks in the
histories of Kinnamos and Khoniates. Kinnamos always stresses the ethnic origin of
the Byzantine officers he mentions, but Khoniates generally ignores it.355 Brand
interprets the different approaches of these authors as a sign of Kinnamos’ anti-
Turkish attitude. However, I can hardly see any particularly anti-Turkish comment of
Kinnamos distinguishable from other Byzantine authors. So the passages concerning
the trial of Alexios Axouchos must be read according to the methodology proposed in
this dissertation. These authors had political aims for writing their histories:
Kinnamos had sympathy for emperor Manuel I, so he chooses to narrate these
accusations in detail, in order to convince the reader that Alexios Axouchos is the
villain. However, Khoniates who was critical of Manuel I constructs his narrative on
the arbitrary and unjust attitude of the emperor. So in Khoniates’ narrative, the details
of the trial of Alexios Axouchos were omitted, because they do not serve the aim of
his work.
In summary, from the discussion above, the following patterns of the entry of
the Seljuks (not only the members of the House of Seljuk, but all persons related to
the Great Seljuks or the Sultanate of Rum) and Pechenegs into Byzantine service
355 However, he mentions that John Axouchos is of Persian origin: Khoniates-Van Dieten, Nicetae
Choniatae Historia, 14; Khoniates-Magoulias, O City of Byzantium, 7.
192
emerged.
i. The visibility of the Seljuks in the Byzantine ranks was much greater
than the Pechenegs. Apart from the tribal leaders who had obtained aristocratic
titles, there were no important officers of Pecheneg origin, nor were there any
families such as the Axouchoi or Chalouphes who had important positions in
Byzantine society for more than one generation.
ii. One can speculate that in the lesser military positions and among the
peasantry there were more “Skythians”. However, there is little concrete
evidence about it. Because of the baptismal names used in the register, it is not
easy to understand how many of the peasants were ethnic Greeks and how many
belonged to recently converted foreign populations.
6.3 Tzachas and Syrgiannes Palaiologos: Two case studies
Çaka (Τζαχᾶς) is an interesting example of the Turks who entered Byzantine service.
The only account of his life is found in the Alexiad. In addition to Alexiad, in
Danişmendname a certain warrior called Çavuldur Çaka356 is mentioned among the
comrades of Danişmend Gazi. Çavuldur is the name of an Oghuz clan, one of the
twenty-four clans that were considered descendants of the legendary Oghuz Khan in
Turkic genealogical legends. However, Danişmendname gives no details of the deeds
of Çavuldur Çaka.357 Moreover, the personage of Çaka is totally absent in oriental
sources dealing with Seljuk history.
356 Melikoff, La Geste de Melik Dānişmend: Étude critique du Dānişmendnāme, vol. 1, 72-73, 85-88.
357 Faruk Sümer put forth that Çavuldur Çaka is the same person as a certain Emir Çavuldur, attested
in the Persian history of Zahir-al-Din Nishapuri (12th century), who captured the cities of Maraş and
Sarız during the Seljuk invasion of Anatolia. So he does not identify him with Çaka of Smyrna.
Sümer, Oğuzlar (Türkmenler): Tarihleri-Boy Teşkilatı, Destanları, 324.
193
Çaka appears in the Alexiad as a warlord/pirate who established himself in
the city of Smyrna. He built a pirate navy there and then captured the towns of
Phokaia, Klyzomenai, and Mitylene. Even though his family origins are obscure, his
brother Galabatzes (Γαλαβάτζης)358 is with him at Smyrna. In an interesting passage,
Anna Komnene tries to give Çaka’s own voice about his early life: “You should
know that I am the young man who used to make incursions into Asia. I fought with
great spirit, but because of my inexperience, I was deceived and captured by the
famous Alexander Kabalikas. He offered me as a prisoner of war to the emperor
Nikephoros Botaneiates. I was at once honored with the title of protonobelissimos
and after being rewarded with liberal gifts I promised obedience to him. But ever
since Alexios Komnenos seized power, everything has gone wrong.”359
After he defeated Niketas Kastamonites, sent by the emperor to subjugate
Tzachas, he became the controller of the region. Tzachas had thousands of Turks
who were following him. He managed to control Smyrna, despite the imperial efforts
to recapture the region. Emperor Alexios I sent two of his most distinguished
generals, Constantine Dalassenos and John Doukas, but despite their initial
successes, they could not subjugate Tzachas.
The author stresses the ambitious character of Tzachas. It seems that these
ambitions were not only remarked by the emperor, but also by the sultan. Anna
Komnene cites a letter written by Alexios I to sultan Qiliç Arslān I of Rum. The style
gives the impression that it is a product of the Byzantine imperial chancellery. This
letter distinguishes Qiliç Arslān as the “most illustrious sultan” and states that the
sultanate is his right of inheritance. However, it designates Tzachas as a usurper who
358 It must be a Greek rendering of the Turkish word for envoy and prophet, yalavaç.
359 Komnene-Reinsch-Kambylis,Annae Comnenae Alexias, 225; Komnene-Sewter-Frankopan, The
Alexiad, 205.
194
claims the purple robe and warns the sultan by saying “the whole mischievous plan is
directed against you.” This letter convinces Qiliç Arslān to kill Tzachas. They meet
in the city of Abydos in Thrace and after hours of heavy drinking at the banquet
table, the sultan murders his father-in-law with a sword blow.360 Tzachas’ time on the
stage is relatively shorter, but it leaves bitter traces in western Anatolia. Anna
Komnene describes the dramatic effects of these years as follows: “When Tzachas
had ravaged the Smyrna area, he had reduced it to rubble and wiped it out
entirely.”361
Why was Çaka’s memory forgotten in medieval Seljuk historiography? As it
was seen before, he was a major protagonist in the later 11th century Asia Minor.
Furthermore, he was the father-in-law of Qiliç Arslān I, so he was a relative by
marriage of the Seljuk dynasty. Three explanations can be offered for the silence of
Seljuk sources about Çaka. Firstly, he was very probably converted to Christianity.
This idea was first proposed by Charles M. Brand in his article, “The Turkish
Element in Byzantium, Eleventh-Twelfth Centuries,” in which he made the following
argument:
(Tzachas’) pretension to the Byzantine throne would be unthinkable if he
could not at least claim to be Christian. Without this primary qualification, he
would have been unable to attract support; if he had taken Constantinople, he
would have been utterly unacceptable to the Byzantines save as a Christian.
Tzachas’ Christianity was barely skin-deep, and the same is probably true for
a good many others who entered Byzantine service as adults.362
So Tzachas could be qualified not even as a Turkish beg, but a Byzantine local
magnate. However, it seems that he had a network in which both the Pechenegs and
360 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 263; Komnene-Sewter-Frankopan, The
Alexiad, 243.
361 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 326; Komnene-Sewter-Frankopan, The
Alexiad, 398. The aftermath of the descendants of Tzachas is largely unknown. It was already
mentioned that his daughter married Qiliç Arslān I. It seems that he had also a son.
362 Brand, “The Turkish Element,” 17.
195
Seljuks were included. Even his daughter was the wife of Qiliç Arslān I of Rum. He
could be both converted from Turkic paganism or Islam. If he converted from Islam
to Christianity, this could be the cause of a damnatio memoriae.
A second explanation could be his hostility with Qiliç Arslān I of Rum
Seljuks and his assassination by his son-in-law. So he was perceived as an enemy of
the Seljuk dynasty and the later historiographers related to Seljuks of Rum did not
consider his memory worth remembering.
Thirdly it can be said that his area of activity was geographically far from the
center of weight of Rum Seljuks. Nonetheless, this argument must be carefully
examined. Sulaymāns̲ h̲ āh I, whose deeds were recorded in Seljuk historiographical
tradition, was based in Nicaea. This city was the capital of Rum Seljuks until the
First Crusade. An important part of the raiding activity of the early Seljuks of Rum
was realized in Northwestern Anatolia.
Ibn Bibi began his history by saying “how Sulaymāns̲ h̲ āh I (Suleiman b.
Kutlumus b. Israil) invaded the realm of Rum and the deeds of grand amirs such as
Mengujek, Artuk, and Danismend were not very obvious,”363 so he does not mention
Çaka among these begs. Indeed he begins his account with Qiliç Arslān II’s choice of
Kayk̲ h̲ usraw I as his heir. The Anonymous Selçukname, which deals more with the
early history of Seljuks of Rum, also never mentions Çaka.364
As mentioned, in the Danişmendname a Türkmen commander called
Çavuldur Çaka is referred to among the comrades of Melik Danişmend. However, he
is mentioned only three times in the whole work, and there is no particular similarity
to historical Tzachas. But the affiliation with the Çavuldur clan could be a reflection
363 Ibn Bibi, el-Evâmirü'l-Alâiyye fi'l-umûri'l-Alâiyye, vol. 1, 43.
364 Anonim Selçukname, ed. Coşguner-Gök.
196
of reality. The use of tribal names before the personal names was a custom among the
medieval Turks, as the examples such as Salur Kazan, one of the main protagonists
of the Kitab-ı Dede Korkut demonstrates. However, it should be remembered that he
was not a “tribesman”, Çaka lived all his adult life on Byzantine soil and his
followers do not appear as the members of a clan or tribe, but as Turkish (and Greek)
riff-raff who mingle in the chaotic Western Anatolia of late 11th century.
In conclusion, the brief and adventurous life of Çaka demonstrates the
potential and the limits of the integration of a Turkish warrior into Byzantine society.
Theoretically, after his conversion to Christianity and having a military force, he
could dream of everything, even the purple robe. But he was still perceived as a
barbarian. Even Çaka himself does not hesitate to acknowledge this fact while he
announces his desire to form a marriage alliance between his family and Dalassenoi:
“...let the marriage contract be committed to writing, agreeable to both parties, as is
the custom of the Romans and of us barbarians.”365
Syrgiannes Palaiologos Philantropenos366 (d.1334) was a Byzantine
statesman of Cuman origin born roughly two centuries and a half later than Tzachas.
He was the son of a Cuman tribal chief called Sytzigan (Συτζιγάν), who entered
Byzantine service, received holy baptism, and became part of the entourage of the
Palaiologan emperors. Sytzigan then married a Byzantine noble lady, Eugenia
Palaiologina.367 His father was presented as “a remarkable man from to Cumans who
came from the region of Hyperborean Skythians came to the emperor” by
365 Komnene-Reinsch-Kambylis,Annae Comnenae Alexias, 225; Komnene-Sewter-Frankopan, The
Alexiad, 205.
366 For his career see Kyriakidis, “The Portrayal of Syrgiannes Palaiologos Philanthropenos in the
Historical Works of Nikephoros Gregoras and John Kantakouzenos,” 221-238; Vasary, Cumans and
Tatars : Oriental Military in the Pre-Ottoman Balkans, 1185-1365, 120-121.
367 Kyriakidis, “The Portrayal of Syrgiannes Palaiologos Philanthropenos in the Historical Works of
Nikephoros Gregoras and John Kantakouzenos,” 221-238.
197
Gregoras368 and “one of the noblest of the Cumans” by Kantakouzenos. So the
nobility of the Cuman paternal descent of this Byzantine aristocrat was stressed by
both authors.
Syrgiannes Palaiologos was the governor of Macedonia around the year 1315.
He was considered a brilliant commander and became part of the faction formed
around future emperor Andronikos III. His military role in the rebellion against
Andronikos II, during the Byzantine Civil War 1321-1328 is emphasized by
Gregoras, yet relatively underestimated by Kantakouzenos whose aim is to stress his
role in these events.369 During the civil war, Syrgiannes Palaiologos changes the
sides and defects to the faction of Andronikos II. The old emperor promotes him to
the office of megas doux. Furthermore, he was also appointed as strategos to the
western provinces of the empire. He was known as a plotter, he was accused of
treason several times, and there were further doubts that he wanted to establish an
independent state in the Balkans. After many accusations and trials, he finally flees
from Constantinople to the court of King Stefan IV Dušan of Serbia. Having the
support of the Serbian king, Syrgiannes starts a rebellion in the western frontiers of
the Byzantine Empire, however before his rebellion becomes a threat to the empire;
he was murdered by Sphrantzes Palaiologos.
It seems that Syrgiannes Palaiologos had a long-lasting friendship with the
Serbian kings of the Nemanjić dynasty, namely Stefan II Milutin and Stefan IV
Dušan. Just before the rebellion of Andronikos III, Stefan II Milutin sends an envoy
368 Gregoras-Bekker, Nicephori Gregorae Byzantina historia, 296; Gregoras-Van Dieten, Rhomäische
Geschichte, 28.
369 Kyriakidis, “The Portrayal of Syrgiannes Palaiologos Philanthropenos in the Historical Works of
Nikephoros Gregoras and John Kantakouzenos,” 234.
198
to the emperor Andronikos II to call back the Cuman mercenaries whom he borrowed
him. The ambassador of Stefan II Milutin also meets the faction of Andronikos III
and they negotiate an alliance. Kantakouzenos attributes the successful negotiation
with the Serbians to the friendship of Syrgiannes with the king.370
Here I offer a hypothesis about the political project of Syrgiannes
Palaiologos. Both Vasary and Kyriakidis underline the extraordinary militarypolitical
activity of Syrgiannes in the Byzantine-Serbians frontier regions, but they
restrain themselves from commenting further. I think that Syrgiannes’ real aim was
forming an independent statelet in Macedonia. In Kantakouzenos’ narrative, the
references to the Cuman mercenaries in Serbia and Stefan II Milutin’s alliance with
the Byzantine rebels follow each other and the role of Syrgiannes in the latter was
emphasized, yet there was no causality between the two episodes. In my opinion, it
can be speculated that the network of Cuman mercenary bands played a certain role
in the friendship of Syrgiannes with the rulers of Serbia and the same populations in
the Macedonia region were the dynamic forces on which the state he envisioned
would be based.
So it could be assumed that Syrgiannes Palaiologos Philantropenos aimed to
form a secessionist statelet in Macedonia, similar to Tzachas’ breakaway statelet on
the western coast of Asia Minor. Syrgiannes, unlike Tzachas, had ties with the
Byzantine aristocratic cycles, because his mother was a member of the Palaiologan
dynasty. This fact gives him a bigger playground in Byzantine politics. Furthermore,
Syrgiannes was a member of the Cuman aristocracy on his paternal side, yet Tzachas
was an ordinary war prisoner of Turkish origin. As was demonstrated above, for all
370 Kantakouzenos-Schopen, Ioannis Cantacuzeni eximperatoris Historiarum Libri IV, 35-38;
Kantakouzenos-Fatouros, Geschichte, 33-34.
199
three authors (Komnene, Gregoras, and Kantakouzenos) the noble or non-noble
origin of a barbarian is something to be remarked on. Finally, it should be noted that
both of the secessionist projects ended unsuccessfully. This fact demonstrates that in
the Byzantine regions with a presumably Greek majority, such as Macedonia and
Ionia, the formation of an independent entity under a ruler of foreign origin was not
easy.
6.4 Evangelization as a way of integration
The adoption of Orthodox Christianity was a sine-qua-non condition for integration
into Byzantine society, as it was seen in the case of Tzachas. The imperial dignities
and titles were exclusively available to Orthodox Christians. The unique and
dominant position of Orthodox Christianity as a central feature of Byzantine identity
is above any discussion. Byzantium was New Jerusalem, as much as it was New
Rome. As already discussed above, one should distinguish conversion to Christianity
from paganism or Islam. Because though paganism is the natural enemy of
Christianity and Islam was seen basically as a version of paganism by some
Byzantine theologians, this duality is slightly different for the Muslims. As it was
already mentioned, the majority of Byzantine theologians considered Islam as an
heresy. Islam, like other Abrahamic religions, gives a strong self-image to believers
and makes it harder to convert the other religions. Furthermore, the conversion from
Islam to Christianity –called tanassur in traditional Muslim religious literature- is an
issue that merits careful examination. In Islamic jurisprudence (fiqh) the conversion
of a Muslim to any other religion (i.e. the apostasy, irtidad)371 is forbidden and it is a
371 It is the Greek rendering of Arabic mürted (apostate). There are a number of references to murtatoi
as military contingents, similar to Turcopoles.
200
crime punished by death. It is known that in the Muslim countries in our timeframe,
such as Egypt under the Mamluks, such punishments were taken place. However, in
the frontier regions where the identities were fluids, such as the borders, such
punishments were probably less frequent. Furthermore, the punishment of such an
act could be possible only by a decision of a Muslim court. So if the apostasy takes
place after the person became a subject of a Christian state where in the moment of
apostasy, there is no authority to punish it, it remains unpunished and possible.372
In the 10th century, during the heyday of the Byzantine reconquest toward
Syria and Mesopotamia, the Empire managed to annex the cities such as Edessa and
Antioch which were hosting an important Muslim population. In the sources of this
period, there are several references to the fate of the Arab Muslim inhabitants of the
region. Some of these inhabitants left their cities and immigrated to the core Muslim
lands. On the other hand, other inhabitants remained in this region. In 941, some
10.000 members of the Arab tribe Banu Habib converted to Christianity with their
families and slaves. In subsequent years they were followed by other tribesmen.
These tribesmen have also obtained lands and gifts from the Empire.373 It seems that
similar conversions en masse were frequent in the cities conquered by the
Byzantines; such as Melitene after the reconquest of John Kourkouas.
The medieval epic romances of the borderlands can throw light on the place
of conversion in the social history of Anatolia. Both the story of the father of Digenis
Akrites and a story of a Christian convert in Danişmendname could demonstrate that
Muslims’ conversion to Christianity –at least in early centuries or in conjunctures
372 For the general question of apostasy see Beihammer, "Defection across the Border of Islam and
Christianity: Apostasy and Cross-Cultural Interaction in Byzantine-Seljuk Relations," 597-651. For an
introduction to the notion of apostasy in Islamic jurisprudence, see the article murtadd (W. Heffening)
in Encyclopaedia of Islam (2nd ed.), vol. 7, 635-636.
373 Kaldellis, Romanland, 129.
201
where the Byzantines were dominating were frequent.374 These sources of course
were not historical, yet they still echo the vivid memories of such facts in the
collective memory of the peoples of Asia Minor. The father of Digenis Akritas is an
Arab amir from Syria (“Amir, servant of God and prince of Syria”) who spoused a
Greek noblewoman and defected to the Byzantine side (“I will become a Christian in
Romania/and listen to the truth, by the great prophet”). 375 The name Digenis (twogenes)
came from this very situation that the Byzantine epic hero belongs to two
different and antagonistic origins. However, his mixed origins do not keep him from
becoming a hero of Christianity.376
However, the most interesting personality of this kind was Artuhi, the close
friend and comrade of Melik Danişmend in Danişmendname. Artuhi is depicted as
both a warrior of faith and an interlocutor between two sides in the epic narrative of
Danişmendname. His ethnic origin was Türkmen and he belongs to a tribe of 12.000
tents. He is portrayed as a Christian who was convinced to convert to Islam by Melik
Danişmend, though there is always a doubt about the veracity of his dedication to
Islam. As was both stressed by Iréne Melikoff377 and Nicolas Oikonomides,378 his
name has a strong similarity with the name of Artuk, a contemporary Türkmen
warlord from the tribe Döğer who is the eponymous founder of the Artukid dynasty
(1102-1409) in Upper Mesopotamia. As it was pointed out the continuity of the
frontier ethos, these conversions and reconversions could be reflected in the
collective memory of the frontier regions as a historical reminiscence.
374 Also in the Battalname, though presented as a Muslim, there is a Turcoman nomad called Yuhanna
b. Afshin. Yuhanna is a very frequent name among the Nestorian and Syriac Christians of
Mesopotamia. Battalname, ed. Dedes, 632.
375 Digenis Akrites, Mavrogordato, 20-21.
376 Digenis Akrites, Mavrogordato, 9 and 21.
377 Melikoff, Le Geste de Melik Dānişmend: Étude critique du Dānişmendnāme, vol.1, 122-125.
378 Oikonomidès,“Les Danishmendides, entre Byzance, Bagdad et le sultanat d'Iconium,” 195-196.
202
A striking scene in the Alexiad is the prayers of the Turkish soldiers besieged
by the armies of Constantine Dalassenos in Chios. These Turks who were the
followers of Tzachas, after they realized that the resistance to Byzantine forces is
impossible, began to utter prayers “in the Roman tongue”.379 This scene merits
attention: First of all, it is understood from here that the Turks did not go through an
Islamic acculturation enough to pray in Arabic. Whether these people were Muslims
throughout their lives, even if superficially, is extremely uncertain. Furthermore, this
scene demonstrates that even the Turks who were razing the Anatolian and Aegean
countryside had a certain familiarity with the Christian religion or they even attended
church services. It can be said that the Turks around Tzachas were outside the
shaping influence of the two great monotheistic religions in terms of upbringing, and
perhaps the influence of Christianity was a little more on them due to the time they
spent in Asia Minor.
Returning to our subject, it can be said that not only the Turks but all the
foreigners who were successfully integrated into Byzantine society were baptized. It
seems that there is no exception to this fact. This was the traditional pattern of
Byzantine society. However late 10th-early 11th century the religious missionary
activity in the Empire obtain a new zeal.
The reign of Alexios I Komnenos marks a turning point in the religious
history of Byzantium. Alexios I who announces a reform edict in 1107, forms several
new clerical offices regarding religious education. Among the positions he creates,
there is an office called didaskalon ton ethnon (the teacher of the gentile). These
didaskaloi were employed to convert the barbarians.380 Of course, these efforts did
379 Komnene-Reinsch-Kambylis,Annae Comnenae Alexias, 223; Komnene-Sewter-Frankopan, The
Alexiad, 203.
380 Stone, “The missionaries of Manuel I,” 253-254.
203
not begin with this edict, since centuries Byzantine Empire has a determined
missionary agenda targeting the pagan populations, which managed to convert the
Bulgarians and Russians among the other populations.
A traditional practice among the Turks of Asia Minor is also worthy of
attention: The Turks, despite their adherence to Islam, were baptizing their children.
This fact seems related to a popular belief that the non-baptized children could be
possessed by demons or stank like dogs. The mothers of some of these children were
Orthodox Christians. Despite they were baptized, the religious identity of these
children is highly doubtful. The baptism became a part of the practices of their folk
religion and do not reflect their devotion to Christianity.381
However, in the Komnenian period, there sparked new zeal for missionary
activity. Needless to say, the Seljuk invasion of Asia Minor and the Crusades were
two important events that triggered this renowned zeal. Anna Komnene states in the
Alexiad –concluding his father’s generosity towards the Seljuk officers Elkhanes and
Skaliaros who contributed to their conversion to Christianity– “He (Alexios I) was an
excellent teacher of our doctrine, with an apostle’s faith and message, eager to
convert to Christ not only the nomad Scythians but also the whole of Persia and all
the barbarians who dwell in Egypt or Libya and worship Muhammad in their
extraordinary ways.”382 The target of this activity was, without any doubt, the Turkic
populations of Asia Minor and Persia.
This activity arrives at its zenith in the reign of Manuel I Komnenos who was
personally very interested in religious matters. In the orations delivered by the
381 The source is Theodore Balsamon. The fact is recorded during the patriarchate of Loukas
Chrysoberges (1157-1169/1170), see Brand, “The Turkish Element,” 16.
382 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 199; Komnene-Sewter-Frankopan, The
Alexiad, 182.
204
churchmen during his reign and even in his epitaphio delivered by Eustathios of
Thessaloniki there were obvious references to this missionary effort and the
evangelization of the Persians. According to Eustathios, Manuel I “brings together
unto God what is alien to the faith and leads the rebellious into a familiarity with
God” and “fills the court of God with sheep.”383 Euthymios Malakes states, “O
Persian….that according to the habit of the Persians of old you renewed the practice
of sending your children to shared schools to learn justice….It is well, that you look
back to the true emperor of the earth and the Persians three days before…to enquire
personally into the manger now placed by the Jordan…”384
Andrew Stone states that these missionaries could not be only related to the
Turks in Byzantine territory, and they must have gone into Seljuk territories. There is
no independent attestations of such a missionary activity there, but the population of
the frontiers, particularly nomadic Türkmen who had a very superficial adoption of
the Muslim religion, they could be an appropriate target for such an activity of
proselytization. So the Komnenoi planned to integrate the Seljuk realms in Asia
Minor and beyond, via a project of evangelization to Byzantine oikoumene, as they
did with Russia and Bulgaria. This project could have tied the Seljuks to the
Byzantine world pacifically.
This project triggered also a theological discussion in Constantinople.
Already during Qiliç Arslān II’s visit to Constantinople (1162), his entry to Saint-
Sophia as an “infidel” disturbed the high ranks of clergy, including the patriarch
Luke Chrysoberges. In the course of the reign of Manuel I Komnenos, to make the
evangelization of Seljuks easier, he proposed a small change in the Orthodox
383 Stone, “The Missionaries of Manuel I,” 255-256.
384 Stone, “The Missionaries of Manuel I,” 256.
205
catechism, the removal of the anathema against the God of Muhammad. This
anathema says that the God of Muhammad is “neither begat nor was begotten” and
was a solid (holosphyros).385 Manuel I must have thought that without such a formula
he could imply that the Muslims and Christians believe in the same God and the
Turks’ conversion to Christianity could be easier. With the aim of removing the
anathema, he summoned the holy synod with the members of the clergy. As it could
be expected, this proposal made the patriarch Thedosios Boradiotes and other
hierarchs discontent. As it was described by Khoniates, “they all shook their heads in
refusal, unwillingly even to listen to his proposals, which they considered slanderous
and detracting from the most true glory of God.” According to these hierarchs, the
God of Muhammad was not a God, but “a solid God fabricated by the deluded and
demoniacal Muhammad.” One of the champions of opposition against Manuel I’s
formula was Eusthatios of Thessalonike. With such a rigid opposition from the
clergy, the tension between the emperor and the clergy has risen. However, in the
end, the synod found a resolution, they did an agreement to remove the anathema of
Muhammad’s God from the catechism and put there the anathema of Muhammad
and his teachings.386
Nonetheless, during the controversy and crisis in this holy synod, Manuel was
already very ill. He died in 1180 and during the brief and tumultuous reigns of the
latter Komnenian and Angelos emperors, the missionary zeal died out. The possible
halt in the evangelization of the Turks, however, must be related to the political
conjuncture, not to the theological formulas. The Byzantine Empire, under the
385 Khoniates-Van Dieten, Nicetae Choniatae Historia,213; Khoniates-Magoulias, O City of
Byzantium, 121.
386 Angold, Church and Society under the Comneni,108-113; Khoniates-Van Dieten, Nicetae
Choniatae Historia, 213-214; Khoniates-Magoulias, O City of Byzantium, 121.
206
difficult conditions it encountered after 1185, could not support any more such an
ambitious religious/political agenda.
In the later centuries, particularly the period 1261-1453 which was poetically
called “a protracted death agony of the remnants of the Empire” by Gibb,387 the
dimension of the conversions was dramatically changed. However, in a very late
period such as the 1410s it is possible to see an Ottoman prince in Constantinople,
Yusuf, the son of Sultan Bāyazīd I, accepting the holy baptism. This conversion is
ignored in Ottoman sources and it is attested only in the histories of Doukas and
Khalkokondyles.388 So though the dimension and the frequency of conversions were
changed over time, there was no impassable barrier between the Muslim and
Orthodox Christian identities. The absolute Ottoman domination of the region marks
the definitive end of these conversions. As it was already said, according to fiqh the
irtidad which in this case a tanassur is punished by death. The history of Ahmet the
Calligrapher or St. Ahmet, the Orthodox neo-martyr who chose to convert to
Christianity and was killed for this act in Constantinople in 1682 in the zenith of
Islamic zealotry in the Ottoman Empire, was an example of which fate awaits such
an act.
The individual evangelizations among the members of the Seljuk ruling
family are also worthy of attention. The most famous member of the Seljuk dynasty
who converted to Christianity was the founder of the monastery of Koutloumoussiou
(Κουτλουμουσίου) at Athos. Both Byzantine and Seljuk sources are absolutely silent
about the name and identity of this prince. The earliest mention of the monastery is a
document from 1169. Here the name Kutlumus must be a patronymic rather than a
387 Gibb, “Byzantium and Islam,” 323.
388 Doukas-Grecu, Historia Turcobyzantina, 135-137; Doukas- Magoulias, Decline, 112;
Khalkokondyles, The Histories, 292-293.
207
personal name. As it was already seen, since the 11th century the Byzantines were
employing the term “Kutlumusians” (Κουτουλμουσίοι) for the branch of Seljuks
who founded the Sultanate of Rum. According to Michel Balivet, the founder of the
monastery could have defected to Byzantium in the early 12th century. If the
conversion was realized in such an early period, before the frequent marriages
between the members of the Seljuk dynasty and Greek women, it could hardly be the
case of dual identity, but a simple conversion such as Chrysoskoulos. As was said
above, it is very difficult to trace the apostates in Muslim sources.389
Another member of the Seljuk dynasty, who was called G̲ h̲ iyāth al-Dīn,
married Rusudan, queen of Georgia (r.1223-1245), and converted to Christianity.
This prince does not belong to the main line of the dynasty which was ruling in
Konya, yet a cadet branch of the Seljuks that was installed at Erzurum
(Thedosioupolis) after the Seljuk conquest of this ancient capital of the local
Türkmen dynasty Saltuqids. His father, Mughis al-Dīn Ṭog̲ h̲ rı̊l-S̲ h̲ āh (d.1225),
governed Erzurum as the vassal of his nephew ʿAlāʾ al-Dīn Kaiḳobād I of Rum. The
motivations of this arranged marriage were purely political; hence this branch of the
sultanate must have the support of the Kingdom of Georgia which was at the zenith
of its power in this period and expanded its territories toward the realms of Muslim
emirates in the Armenian Highlands. There was no mention of this marriage in Seljuk
sources; it is recorded only in the Georgian sources and Arabic sources of Ibn al-
At̲h̲ īr and Baybars al-Manṣūrī who were written outside of the Seljuk world and in
the chronicle of S̲ h̲ ihāb al-Dīn Muḥammad al-Nasawī who was the official chronicler
of D̲ j̲
alāl al-Dīn K̲ h̲ wārazm-S̲ h̲ āh who was a rival of Seljuks.390
389 Balivet, “Deux monastères byzantins fondés par des Turcs: Koutloumoussiou/Kutulmuş et
Dourachani/Turahan,” 82-83.
390 Turan, Selçuklular Zamanında Türkiye, 434.
208
Moreover, there is the curious case of the family and entourage of sultan Izz
al-Dīn Kaykāʾūs II who after a number of dynastic struggles in Asia Minor left their
country and went to Constantinople as refugees. Their religious identity is hard to
define and perhaps this ambiguity could be considered an example par excellence of
Shukurov’s dual identity thesis. As it was already seen in the last chapter, their life in
exile was documented well in the History of Pakhymeres. Surprisingly there is also a
Turkish text about them: Yazıcıoğlu Ali’s Selçukname. Despite its name, this work
was compiled a century and a half after the vanishing of the Sultanate of Rum. This
Turkish text acknowledges the sultan’s entourage’s evangelization. However, the
situation of the sultan himself could be hardly considered a definitive conversion to
Christianity, because when he escaped to Crimea, in the court of Berke Khan (r.
1257-1266), the first Muslim ruler of the Golden Horde, he returned to Islam.
Kaykāʾūs II’s alleged conversion to Christianity provoked a conflict among the
Byzantine clergy because the sultan was attending religious ceremonies with the
patriarch Arsenios. It is not considered appropriate for someone whose adherence to
Orthodox Christianity is not certain to participate in such ceremonies. However, in
the narrative of Gregoras, the sultan is presented as the son of Christian parents
(χριστιανῶν τε ὑπῆρχε γονέων υἱός) and as a Christian who during his reign was
observing his religious duties in secret. So the situation could be unusual. It could be
further speculated that the very existence of the political ambitions of the ex-sultan
made him even more untrustworthy to the Byzantines.391
The aftermath of the descendants of Kaykāʾūs II was carefully examined by
391 Gregoras-Bekker, Nicephori Gregorae Byzantini Historia, v.1, 94 ; Gregoras-Van Dieten,
Rhomäische Geschichte, v.1, 111. On the other hand, according to Shukurov, this expression means
that the sultan is “a son of Christian ancestors.” See the discussion in Shukurov, The Byzantine Turks,
1204-1461, 107.
209
Paul Wittek.392 In the 14th century, these Christianized Seljuks seem to live in the city
of Karaferye (Veria) in Greek Macedonia. According to Selçukname, these
descendants of the Seljuk dynasty encountered also the Ottoman sultan Bayazid I,
after the conquest of Karaferye. Despite Yazıcıoğlu Ali’s slightly discontent tone
about the Christianity of Rum’s ancient rulers, he depicts them as noble men with
dignity. Since the work of Paul Wittek, this Christian descendency (and possible
Türkmen grouping around them) are considered the forefather of the Gagauz people,
an Orthodox Christian Turkic ethnic group, now living in the Bessarabia.393
To return to our starting point, how Byzantine were considered these newly
converted Turkish men? According to Kinnamos, despite the barbarity in their roots,
they were still Byzantines, and according to Pakhymeres they were dubious
Christians.
Now let us look closely at the conversion of the Ottoman prince, a son of
Bayazid I, Yusuf Çelebi (called “Isa the younger” by Khalkokondyles) who
converted to Christianity in the 1410s and received the baptismal name Demetrios.
According to Doukas,
(He) acquired a passion for Greek learning. He accompanied John, the
emperor’s son, to school, and there as a student he was introduced to
intellectual matters. So absorbed was he by the love of learning when he
attended school with John that he came to Emperor Manuel (Manuel II
Palaiologos) and requested to be baptized according to Christian law. Daily
he professed to the emperor that he was a Christian and not a believer in
Muhammad’s doctrines. The emperor did not wish to listen because it might
cause scandal. Then when the dreaded disease continued to consume and
destroy bodies, neither respecting nor sparing any age, it attacked to
Bāyazīd’s adolescent son. The stricken youth sent the following message to
Emperor John, “O Emperor of the Romans, you who are both master and
father to me, my end is near. Against my wishes, I must leave everything
behind and depart for the Heavenly Tribunal. O confess that I am a Christian
and I accuse you of not granting me the warmest of faith and the seal of the
Spirit. Know, therefore, that as I must die unbaptized I shall bring accusations
392 Wittek, “Yazijioghlu ‘Ali on the Christian Turks of Dobruja”, BSOAS, XIV/3 ,639-668.
393 Wittek, “Yazijioghlu ‘Ali on the Christian Turks of Dobruja”, 667-668.
210
against you before the Judgement Seat of the impartial God.” Yielding finally
to his plea, the emperor sent for him and as his godfather sponsored his
baptism. He died the next day. The emperor buried him with great honor in a
marble sarcophagus near the church and within the gate of the Stoudite
Monastery of Prodromos.394
This narrative is quite striking in comparison with Khalkokondyles’ brief and dry
statement about the same Ottoman prince: “Isa, the younger of the Bayazid’s sons,
also came to the Greeks and even converted to the religion of Jesus, and died shortly
afterward.”395 The same is true also for Sphrantzes: “Bayazid’s five sons -Sülayman,
Musa, Isa, Mehmed, and Yusuf- arrived in Europe; Yusuf converted and took the
Christian name Demetrios.”396 However, both texts are dealing with essentially the
same situation. During the bloodshed of the Ottoman Interregnum (1402-1413) the
youngest of the Ottoman princes came voluntarily or as an asylum to Constantinople,
where he was strongly influenced by Byzantine culture and Orthodox Christianity
and in the middle of a plague, he decided to convert to Christianity. Doukas’ stress on
Manuel II’s reluctance to such a conversion by an Ottoman prince is also
comprehensible under the conditions of the delicate balance between the Ottomans
and Byzantium. This example is remarkable and it serves as a reminder that one
should not look at historical facts with a socio-political determinism. Such personal
decisions could also be the result of an individual’s faith crisis.
In conclusion, conversion to Orthodox Christianity appears to be the
condition of utmost importance to be integrated into Byzantine society. Our sources
generally accept the newly converted Christians as good Romans, though they never
394 Doukas-Grecu, Historia Turcobyzantina, 135-137; Doukas- Magoulias, Decline and Fall of the
Byzantium to the Ottoman Turks, 112.
395 Khalkokondyles, The Histories, 292-293.
396 Sphrantzes-Maisano, Georgii Sphrantzae Chronicon,7; Sphrantzes-Philippides, The Fall of the
Byzantine Empire. A Chronicle by George Sphrantzes 1401-1477, 22.
211
cease to stress their “barbarian” provenance, they do not have an exclusionary
approach toward them. Only their approach turns negative in the cases of
confessional ambiguity, such as in the case of Kaykāʾūs II and his sons. This fact
must be considered perhaps with the phenomenon of Crypto-Muslims, in the
periphery of Byzantine society.
This phenomenon was studied by Rustam Shukurov recently. His research
demonstrates that in marginal areas of ex-Byzantine oikoumene (like Chaldia) there
are a certain number of crypto-Muslims who were Christian in appearance, but
continued secretly to adhere to Islam.397 Shukurov tries to explain this phenomenon,
by using an unpublished Persian geographical text of a well-known polymath and
author Ḥāfiẓ-i Abrū (d. 1430). He was a Persian of Khorasan and spent his life in the
Timurid court, so not a native of Asia Minor. Hafiz-i Abru, in a brief passage of his
work, mentions a Frankish kingdom north of Western Armenia (Shukurov interprets
it as the Empire of Trebizond), where there was a population that was Christian in
appearance but practiced Islam in secret.
As it was mentioned before, both the ambiguities and spontaneous
conversions en masse are typical features of frontier regions. The frequency of
conversions in a region where different cultures and religions co-exist could be an
explanation for the presence of the syncretic faiths and “heterodoxies” altogether.
The Paulician creed was also the fruit of the same frontier atmosphere of the
encounter and co-existence of Christianity and Islam. However, when this creed was
transferred to the Balkans as a result of the imperial system of population transfers, it
can find a powerful interaction with the Pechenegs who were recently Christianized.
These new Christians supported Lekas’ Paulician rebellion in Bulgaria with great
397 Shukurov, “The Crypto-Muslims of Anatolia,” 135–158.
212
enthusiasm. In the religious and hagiographical literature, there can be found further
examples of co-existence and interaction between the religions in the frontier zones.
But the details of such processes remain outside of my focus.
6.5 The place of Turks in Byzantine society
Between the 11th and 14th centuries, a certain number of Turks in Byzantine high
offices are attested. Alexander Kazhdan has stated that the individuals of Turkic
origin constitute 1% of the Byzantine aristocracy.398 The Axouchos family, without
any doubt, was the most successful family of Turkish origin in the Byzantine upper
class; their presence in the Byzantine upper classes continued for three generations.
As was stated by Brand, the Turks were more frequent in lesser ranks of the
Byzantine army, so the Turks in Byzantine high society were only a minority of their
community. Other Turks generally obtained more modest positions in Byzantine
society. Many Turks were originally not from a particularly noble lineage and they
were preferred by the emperors for their lack of social connections. This fact recalls
the rise to prominence of the devşirmes in 15th-century Ottoman society, where they
were preferred because they were representing an alternative to the old Turkish
families who were leading figures in early Ottoman civil administration, such as
Çandarlis.399
Finally, an unpleasant way to enter Byzantine society (or any other society in
the Middle Ages) is slavery, such as in the cases of war captives or the other victims
of similar incidents. Our historiographical texts are generally silent about this reality,
398 He counts 23 individuals of Turkish (Seljuk) origin in a total of 2,300 aristocrats. So they are 1% of
all the Byzantine aristocracy. Brand, “The Turkish Element,” 19. He refers to Kazhdan’s Армяне в
составе господствующего класса Византийской империи в XI-XII вв. 5 as his main source.
399 Brand, “The Turkish Element in Byzantium,” 25.
213
however, in the cases of Tatikios and Tzachas, their captivity was the beginning of
their career in the Byzantine Empire. Probably there were still more numerous people
of Turkic origin, particularly women, employed as house servants and slaves.400
The Byzantine Empire was not a nation-state, yet it has the abovementioned
strong national dimension. These Turks were integrated into society and they hold
offices. During these centuries the Sultanate of Rum and the beyliks formed in its
borderlands continued to have a sometimes antagonistic, sometimes friendly
relationship with the Empire. There is no surviving first-person account of a Turk
who was “integrated” into Byzantine society. However, the evidence in the Byzantine
historiographical works demonstrates that in the second or at least third generation
they became Byzantines proper. In other words, they were assimilated. However,
they were still blamed for their Turkishness, in times of crisis, like John Komnenos
the Fat. The communal character of the settlements could help to the preservation of
the distinguished ethnic character of the populations in the countryside. The
existence of Turkic personal names in Ottoman tahrir defteris for Orthodox
Christians in regions such as Bithynia could be reconsidered with this fact.401
However, our material is insufficient to write an ethnic history of these communities.
6.6 The Turcopoles and Mixobarbarians: A look into the gray areas
400 This subject is treated excellently in Shukurov, The Byzantine Turks, 1204-1461, 244-249.
401 Beldiceanu, “La population non-musulmane de Bithynie,” 18. After citing a long list of Christian
tax-payers who bear Turkic names in 15th century Bithynian countryside, she concludes as “Ce
phénomène est amplement connu et a fait l’objet de plusieurs études. On sait également que les
Byzantins ont christianisé les communautés turques venues soit par le nord de la Mer Noire, soit des
parties de l’Asie Mineure dépendant de l’Etat Seldjoukide. Il ne faut cependant jamais perdre de vue
que le passage au Christianisme était a Byzance soumis à des règles, très strictes et exigeait, entre
autres l’abandon du nom “barbare”. A cette catégorie appartiennent probablement les habitants du
village Tchepni dont les noms sont tires du calendrier grec. Quant aux autres, il ne peut s’agir ni de
populations autochtones grecques, ni de musulmans convertis. A en juger d’anthroponymes, nous
avons affaire à une population turco-tatare (même si d’autres éléments ne sont pas à exclure),
chrétienne de père en fils, mais dont le christianisme, voire parfois la religion, reste à définir.”
214
Between the categories of the Greeks and Turks, there are hybrid identities of
Mixobarbaroi (μιξοβάρβαροι) and Turcopoles (Τουρκόπουλοι). The terms Turcopole
and Mixobarbaros are quasi-ethnic terms used in Byzantine historiography to denote
the half-barbarians.
Turcopole means “son of Turk”.This term was generally explained as the
offspring of the unions of Turkish men and Greek women, moreover, it was used
mostly to describe the Turkic origin military contingents in the Byzantine army. It
can be speculated that the Turkish warbands who were dwelling in Asia Minor
around the 11th century could be seen as the prototype of Turcopole contingents.
Their leaders, such as Elkhanes in the Alexiad, could defect to Byzantium and these
contingents (who once were gangs of freemen living by the sword) could become a
part of the Byzantine army.
In the 12th century, the Turcopole units also appeared in the Crusader Armies.
The Crusaders first encountered these warriors among the ranks of Byzantine armies.
In 1101 there were 500 Turcopoles in the army of Raymond St. Gilles. These
warriors were presented as a gift by Alexios Komnenos to the Frankish Nobleman.
Anna Komnene talks about the destiny of this Turkish contingent, without using the
term Turcopole: a big part of these soldiers were slaughtered by passing the province
of Armeniakon. The commander of Turcopoles was a man called Tzitas (Τζιτας) who
was probably a Turk.
These forces were organized autonomously, under a Turcopolier who
commanded the Turks.402 Nonetheless, in the next decades, the Crusader States
started to recruit their own Turcopole forces in the Levant. They generally served as
402 For their role in the Crusader Armies, see Harari, “The Military Role of the Frankish Turcopoles: A
Reassessment,” 75-116, particularly 76-79 and 102-114.
215
light cavalry units. However, the majority of the Turcopoles in Crusaders’ forces
were not of Turkic but of Arab (Muslim or Christian) origin. Though the Muslimorigin
warriors generally converted to Christianity, there are rarer examples that
demonstrate the presence of Muslims in the Crusader armies. Finally, the Turcopoles
of the Levant do not seem the have a distinct ethnic identity; they were essentially
native warriors combating in Crusader armies. On the other hand, it should be noted
that there were also “non-Turcopole” native troops fighting for the Crusaders.403
Though its vernacular usage is certain, as it was demonstrated by its passage
to Latin as a loanword from Greek; the term Turcopole is not attested in the
Byzantine sources until the 14th century: it was first used by George Pakhymeres.
However, its usage as a patronymic is attested as early as the 11th century: there is a
certain Sergios the Turcopole attested in 1082.404
Mixobarbaros is generally translated as semi-barbarian, but this translation
does not reveal the nuanced meaning of the word. Unlike the term Turcopole, which
seems to appear in the last quarter of the 11th century, mixobarbaros is an older term.
It was attested in Greek texts since Antiquity. Alexander Kazhdan, in his article in the
Oxford Dictionary of Byzantium, cites Hesychios of Alexandria (c. 6th century),
according to whom the term refers to “men who were neither Hellenes nor barbarians
but had qualities of both.”405 This citation can be a starting point. The terms
“mixobarbaros” and “Tourkopoulos” both had the meaning of ethnic admixture. As it
was seen before, Michael Attaleiates employed this term for the populations of the
Danubian frontier. In the later Middle Ages, this term is used with special reference
403 However, there are also Turks serving in the Crusader Armies of the Levant: For example, a
“Bohemond the Turk” appears during the First Crusade; he was a Turkish warrior who took the
baptismal name Bohemond after Bohemond I Hauteville, the Norman ruler of Antioch.
404 Harari, “The Military Role of the Frankish Turcopoles: A Reassessment,” 76.
405 ODB, vol. 2, 1386.
216
to this region. Hélène Ahrweiler explains this fact as follows:
The term Μιξοβάρβαροι refers to cultural issues, and is used for those who
filtered across the Danube and whose nomadic way of life interacted with
sedentary traditions. However, the terms Μιξέλληνες and Μιξοβάρβαροι used
by Byzantine authors of the eleventh and twelfth centuries should be studied
in connection with the practice of mixed marriages in that area, which was
inhabited by Christianized nomadic groups.406
Interfaith marriages seem to be common in late medieval Asia Minor. However, the
vast majority of these unions must be made between Muslim men and Christian
women because Muslim law strictly forbids marriages between Muslim women and
Christian men. However there are examples in Asia Minor where these restrictions
were broken; Abu’l Fida, the Arabic geographer of the 14th century states that in the
city of Melitene, Muslim women marry Christian men. Furthermore, it seems that in
–at least some– interfaith marriages, the daughters could be considered of their
mothers’ religion or at least free to choose their religious identity.407
There are other similar terms used to denote specific situations. For example,
the term Gasmuloi (Γασμοῦλοı) is employed for Latinized Byzantines or the
offspring of Greek-Latin (mostly Italian) marriages.408 Similar to Turcopole,
Gasmuloi was not only an ethnic denomination, but implies military contingents who
were employed both in the land army and navy, mostly in the 14th century.
Among these terms, Turcopole is the term that reflects a clearer ethnic
character. It was used always for the Turkic mercenaries from the east, the ones that
seem related to the Seljuks.
George Pakhyhmeres deals a lot with the Turcopoles in his work. He uses the
406Ahrweiler, “Byzantine Concepts of the Foreigner: The Case of the Nomads,” 13.
407 Vryonis, The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from
the Eleventh Through the Fifteenth Century, 228.
408 According to ODB, its etymology is unknown and it generally implies the offspring of Latin men
and Greek women. ODB, vol. 2, 823.
217
term first to denote the band of mercenaries used by the Byzantines c. 1305,
remarking that the “Roman” army fears the defection of the Alan (mercenaries) and
old Persian contingent which is also called Turcopoles (τὸ ἐκ παλαιοῦ Περσικόν, οὓς
καὶ Τουρκοπούλους ὠνόμαζον).409 The Alan and Turcopole troops were commanded
by a Bulgarian prince called Vojsil, who was the brother of the late Bulgarian tsar
Smilets (r. 1292-1298).410 However, the Turcopoles have also their chiefs.
Pakhymeres then mentions the Turcopoles in the case of the Battle of Hemeros where
the Byzantines and their Alan and Turcopole allies fought against the Catalan
Company. The first commander who is described as the commander of Turcopoles is
Melik Isaak (Ἰσαὰκ Μελήκ); however, this war band, just like the Catalan and Alan
bands, is very undisciplined and hard to control. Melik Isaak is presented as a Persian
satrap.411 The usage of the ethnonym Persian demonstrates that he is an Anatolian
Turk and not a “Skythian”. The author further mentions two other Turcopole
commanders: Tzarapes (Τζαράπης) and Taghatziaris (Ταγχατζιάρις).412
Koutzimpaxis (Κουτζίμπαξις) is a rather enigmatic figure. He is presented as
a Tocharian (Τόχαρος), i.e. a Mongol-Tatar, and he was among the most powerful
magicians (περὶ ἐκεῖνον μάγων τὰ κράτιστα) of the Nogai Khan. But he was also “a
believer in the cult of the Persians,” so a Muslim. After the death of Nogai Khan, he
tried to return to Asia Minor, but his ship drifted to Herakleia Pontika which was a
Byzantine territory and after he landed, he converted to Christianity with all his
family.413 Then he became part of the emperor’s entourage and was appointed
409 Pachymeres, Relations historiques, vol. 4, 572-573.
410 Pachymeres, Relations historiques, vol. 4, 572.
411 Pachymeres, Relations historiques, vol. 4, 650-651.
412 Pachymeres, Relations historiques, vol. 4, 696-697.
413 Pachymeres, Relations historiques, vol. 4, 378-379. Elizabeth Zachariadou interprets the name
Koutzimpaxis as the Greek rendering of the Turkish name “Koca-Bahşi”. Hence this is not a personal
name, but a title. “Bahşi” is a frequent term in 13th-14th century Genghisid Eurasia to denote the
shamans. Zachariadou, “Observations on Some Turcica of Pachymeres,” 261-267.
218
governor of Nikomedeia. He was both occupied with the Turcopoles and the
diplomacy with the Turkish beyliks of Asia Minor.
According to Pakhymeres, the Turcopoles were recently Christianized (τοῖς
ἐξ ὑπογύου χριστιανοῖς Τουρκοπούλοις).414 This remark contradicts the idea that the
Turcopoles were sons of Turkish fathers and Greek mothers and they somewhat grew
up as Christians. Interestingly these Turcopoles were not moving in the way a band
of mercenaries must move, i.e. as a group of men, but brought with them their wives
and children. So the identity of Pakhymeres’ Turcopoles could be slightly different
from the earliest examples and they could be for example a Christianized Türkmen
(or Tatar) population.
These Turcopoles must be related to the military activity of Turks, directed to
Byzantine Thrace. Their relationship with the beylik of Karasi is a matter of
discussion. The legendary accounts of Sarı Saltuk, who was an Alevi-Bektashi
saintly figure and posthumously considered a forbearer of Islam in pre-Ottoman
Balkans, bore the memories both of the Seljuk immigration to Dobruja and Crimea in
the 1260s and the later activities of the bands of Turcopoles in Thrace in the early
14th century.
The account in Nikephoros Gregoras’ Roman History is slightly different. He
identifies Turcopoles directly with the Seljuk sultan Kaykāʾūs II. According to him,
the main body of Turcopoles emigrated to Byzantine territories with him. After he
escaped to Crimea, the remaining Turcopoles were organized under the command of
two Turks: Melik (Μελήκ) and Khalīl (Χαλήλ). These Turkish warriors – Gregoras
mentions their conversion to Christianity twice – were active in Byzantine
Macedonia, Thessaly, and Thrace in the first decade of the 14th century. Although
414 Pachymeres, Relations historiques, vol. 4, 626-627.
219
they were first employed by the Byzantine emperor Michael VIII, they looted
sporadically the region and had a sometimes friendly, sometimes hostile relationship
with Catalans and Alans, other important mercenary groups in the region. Finally, the
band of Melik allied with King Milutin of Serbia and the band of Khalīl, after many
exploits, was massacred and enslaved by the Byzantine troops, before they attempted
to cross the Dardanelles. It should also be remarked that Gregoras uses the terms
Turks and Turcopoles interchangeably.415
The memory of the Turcopoles is also present in the History of Laonikos
Khalkokondyles. However, he never uses the term Turcopole and confuses the events
narrated in the works of Pakhymeres and Gregoras, with the Anatolian expedition of
Catalan mercenaries and later exploits of Umur of Aydin. He describes the abovementioned
Turcopoles as follows:
During his reign (of ʿOt̲h̲ mān I), eight thousand Turks crossed over into
Europe at the Hellespont and seized a Greek fort in the Chersonese. They
made it their base and advanced through Thrace all the way to the Danube,
devastating the land as they overran it. They looted most of it and, taking as
many prisoners as they could enslave, transported them over to Asia; and so
they despoiled the Greeks and Serbs. At this point, however, a large
contingent of Skythians advanced from Sarmatia to the Danube. They crossed
the Danube and met the Turks in Thrace where they routed them in battle.
Except for a few, they mercilessly slaughtered them all. Those who were not
killed sought refuge in the Chersonese, and then they crossed over into Asia
and never returned. (…) It was at this time that Prousa was besieged, starved
out, and taken by ʿOt̲h̲ mān, and other cities of Asia were captured. Thus the
Turks acquired great power in Asia and crossed over into Europe, where they
caused trouble in Thrace. There were many of them, including Khalīl who
was blockaded by the Greeks in a fort of the Chersonese and summoned
Turks over from Asia. He defended himself against the attacks of the
Emperor and then marched out and heavily plundered Thrace. (…) As these
kings (of Greece and Serbia) had bad relations with each other over other
matters, they did not make good use of the Turkish leaders who had defected
to their side, such as Izz al-Dīn (Ἀζατίνης) and the others.416
415 Gregoras-Bekker-Schopen, Nicephori Gregorae Byzantini Historia, 229-232; Gregoras-Van Dieten,
Rhomäische Geschichte. vol. 1, 182-183.
416 Khalkokondyles, The Histories, 22-27.
220
He concludes his long narrative over Turcopoles by stating: “But then the Turks with
Izz al-Dīn, however many there were, went over to the king of Serbs, while those
from Asia turned around and went back on foot to the Chersonese with the intention
of crossing straight over to Asia, in whatever way they could.”417
In Khalkokondyles’ account of events, which was written roughly in the
1460s under the rule of Meḥmed II (r. 1444-1446, 1451-1481), the tone and ethos of
the deeds were changed. It differs clearly from Pakhymeres and Gregoras’ tone, the
memories of the massacres and sporadic violence of the mercenary bands are further
away, and the appearance of these Turks in the Balkans is presented as an
introduction to the later expansion of the Ottomans. In the narrative of
Khalkokondyles, the confusing reminiscences of the 14th century were integrated into
a teleological narrative of Ottoman ascension.
It should be furthermore remarked that the interaction between the various
strata of the presence of Turkic elements in the Balkans; the Seljuks-in-exile under
Kaykāʾūs II, the Mongols on the northern edge of the Balkans, the Second Bulgarian
Empire under the domination of the Turco-Mongol elements, various Turkic
mercenary bands and finally the military intervention of Turkish beyliks (Ottomans,
Aydin, Saruhan) into the Byzantine Civil War of 1341-1347. In this period, these
“barbarian” military elements were present in the margins of the declining Byzantine
Empire, but the Empire was still a political and civilizational axis that bound all these
elements.
A term that could be interpreted as the counterpart of “mixobarbaros” seems
to be “iğdiş” or “ikdiş”. This Persian origin term in contemporary Turkish means
eunuch, but it seems that in medieval Muslim Anatolia, it had a different meaning.
417 Khalkokondyles, The Histories, 26-27.
221
Claude Cahen, discussing this term, remarks that it can mean any animal from
interbreeding, for example, a mule. Therefore, in that case, it can imply men of
mixed origins. For example, the children born of a union of a Muslim and non-
Muslim (just like the Turcopoles) can be called iğdiş. But it can also basically mean
the recently converted people, i.e. mühtedis. Furthermore, there is an administrative
position called “iğdiş başı” or “emir ül-eğadişe” in 13th-century Konya. The iğdiş
had responsibilities in the urban administration. However, the semantic evolution of
the word is not clear.418
6.7 An unsuccessful integration?
It is very hard to judge how successfully integrated the Turks were into Byzantine
society because the criterion of such an integration is ambiguous. In this subchapter,
I shall dare to speculate about the Byzantine grand strategy toward the Turks of Asia
Minor. In ideal terms, the successful integration of Turkic populations into Byzantine
society could be similar to the integration of the Slavs in the Balkans in the 8th and
9th centuries. These Turks must have been evangelized and must not have preserved
an autonomous political structure. So they must not have archons or ethnarchs. The
aftermath of Pechenegs, Cumans, and Oghuz in the Balkans could be considered a
successful integration. After their subjugation to the Byzantine authorities, they
continued to exist as rural communities, mostly with a distinct ethnic identity, yet
they lost their political vitality. They would be assimilated in the later centuries,
mostly by the Bulgarians and Greeks in the region.419 It should also be noted that the
418 Turan, Türkiye Selçukluları Hakkında Resmi Vesikalar, 178; Cahen, La Turquie Pre-Ottomane,
151-152. Speros Vryonis speculates that they could not be the children of mixed marriages, but
children of ghulams.
419However, the Cumans played a certain role in the Second Bulgarian Uprising (rebellion of Ivan and
Asen), which demonstrates that they did not easily lose their political collective identity.
222
rural settlements, such as the ones where Pechenegs and Cumans settled in the
Balkans, constituted small social units probably resistant to the cultural influences of
the Byzantine center. In the mid-14th century, John Vatatzes gave arable land to the
Cuman soldiers in Asia Minor and these soldiers formed military colonies there.
Such military colonization has been an ancient custom of Roman administration
since Antiquity. The formation of such military colonies was advantageous both for
the state and “foreigners” for several reasons. According to Mark Bartusis, such
colonies could accommodate the social needs of the Cumans (and other peoples such
as Tzakones) and make them easier to administrate. Moreover, they were also
preserving their social organization there.420
Acculturation is inseparable from social change. Hélène Ahrweiler interprets
the acculturation of Turkic nomads in the Danubian region as a mechanism by which
these nomads force the Byzantine Empire and society to try to control them. This fact
creates the acculturation process and the osmosis between the nomads and the
Byzantine settled population, and this interaction makes the nomads evolve into
semi-nomads. Meanwhile, they became Christians.421
Another path could be the formation of a Persia or Turcomania, a Greek
Orthodox Turkish state in Central Anatolia. Such an entity could be perceived as an
autonomous unit of the Byzantine oikoumene. It is possible to speculate that the
policy of establishing a position of didaskalon ton ethnon is aimed to convert the
subjects of the Sultanate of Rum and the Türkmen population living in the frontier
zone. I am making this comment based on the experience of Bulgarian and Serbian
state formations. The nature of the political relations between the Byzantine Empire
420 Bartusis, The Late Byzantine Army: Arms and Society, 1204-1453, 158-159.
421 Ahrweiler, “Byzantine Concepts of the Foreigner: The Case of the Nomads,” 15.
223
and Seljuks, particularly until the 13th century demonstrates a hierarchical structure
where the emperor is a father and the sultan is his son. The strong Christian element
around the Seljuk court and the dual identity of the sultans of Rum probably made
this option more feasible for the Byzantine ruling elites. However, when the
Sultanate of Rum started to dissolve, the Byzantine Empire had no more the power to
ideologically dominate Asia Minor. Moreover, Turkish beyliks substituted the Seljuks
in Turco-Muslim Anatolia with a more aggressive military agenda than Seljuks.
In the mid-14th century, the Byzantine Empire was already a rump state,
surrounded by the Ottomans in the east and by the hostile Bulgarian and Serbian
kingdoms in the north and west. The empire experienced a sharp decline in military
power. The decline of military power and the lack of stability affected negatively the
empire’s ideological power. After the mid-14th century, the empire basically relied on
western military aid against the Turkish advance. The Turkish troops continued to
play a role in Byzantine politics, even in a later period such as the return of John V
Palaiologos to Constantinople with the help of the troops of Murād I; however,
Byzantium was no more a center of attraction for Turks or any other foreigners.422
6.8 Conclusion
In this chapter, the essential patterns of the Turks’ entry into Byzantine service,
Byzantium’s potential to acculturate and assimilate, and the possibility of integration
were discussed. My findings are summarized below:
i. T
he entry of different groups or individuals belonging to the Turkic peoples into
422 Bartusis, The Late Byzantine Army: Arms and Society, 1204-1453, 102-107.
224
Byzantine service is a frequent phenomenon in the 11th and 12th centuries. It
must be considered a continuum of traditional relations between the Byzantine
Empire and Turkic nomadic populations in the past. Individuals from different
social origins, both from the Seljuk and Pecheneg societies, entered into imperial
service.
ii. F
rom the beginning, the Byzantine world represented “a better life” for the Turks.
The early Turkish expansion into Western Asia was driven not by religious zeal
or a well-defined political project, but by the warlords’ desire to pillage or take
over new territories in a relatively civilized and prosperous, yet chaotic land.
Even though the Seljuk army defeated the Byzantines at Manzikert, Alp Arslan
did not demand territories in Asia Minor apart from three castles in the frontier
region. Furthermore, the Seljuks’ grand strategy for westward expansion was the
reunification of the Muslim oikoumene which was divided into different
caliphates and sultanates since the early Abbasid period. They did not seek the
annexation of Byzantine Asia Minor, yet they aimed to defeat the Fatimids of
Egypt and to invade Fatimid Palestine and perhaps Egypt. The campaign of
Atsiz into Syria and Palestine and his capture of the city of Jerusalem (1073)
from the Fatimids demonstrate this project. So the invasion of Asia Minor by
Turks was largely a centrifugal and spontaneous event realized by independent
warbands.
iii. I
n the Komnenian period, a new imperial project was launched: conversion of the
Turks (and other Muslim peoples in the east) and the formation of a new model
of political superiority not necessarily with military domination. Manuel I
225
Komnenos particularly dedicated himself to this project. However, the project
ended with a failure, not only because the time of troubles started in the 1180s
and later the Latin takeover of Constantinople in 1204, but also because of the
reluctance of Byzantium’s religious authorities to loosen the theological dogmas
of Orthodox Christianity. This reluctance could not be explained only by a zealot
approach about an abstract adherence to the religious fundaments, but also by
the lack of will for the inclusion of these newcomers to their society.
iv. T
he Turks’ entry into Byzantine service dropped off largely after the mid-14th
century. This fact could be explained by the decline of the Byzantine Empire
both as a political and spiritual center and its lack of power to attire new foreign
populations. Using the words of Charles M. Brand, I can say that Byzantium lost
its “power to attract and absorb.”423 Meanwhile, the Ottomans began to expand
into the core territories of the Byzantine Empire. The Hesychast controversy and
the Byzantine Civil War (1341-1347) further contributed to the socio-political
dissolution of Byzantium. That was the beginning of the end.
423 Brand, “The Turkish Element,” 1.
226
CHAPTER 7
TRICKSTERS, MONSTERS, AND INVISIBLE WOMEN
7.1 The tricksters at war: The Byzantine narratives about Turkic warfare
In this chapter, I explore four elements that are considered to be the key elements
forming the image of the Turks and Turkic peoples in Byzantine literature. These
four elements are warfare, women, sexual behavior, and the violent practices
attributed to the Turkic peoples.
A remarkable genre of Byzantine literature is military writing. Between the
6th and 11th centuries, Byzantine authors produced many treatises that explore
military science. These texts reflect a serious understanding of military theory, and
they follow the ancient Greek model of military writing. The deep-rooted tradition of
military writing in which these texts were formed goes back to Aelian (2nd century
BCE) and Onasander (1st century BCE). In addition to their value for the military
history, Byzantine military manuals also contain important materials regarding the
Byzantine representation of the “other” peoples. To summarize, it can be said that
these texts are not only narratives of military art, but also ethnographic texts.424
However, just like in the other Byzantine literature genres, the use of mimesis
is an important feature of military writing. These texts contain ancient material
produced long before the period in which they were written. Many clichés were
employed to describe the military tactics and methods of various nations throughout
the centuries. For example, the Byzantine representation of Turkic warfare was full
of previously written strata of a centuries old topos about the Skythian warfare. The
nomadic populations of the Eurasian steppe were known for their dedication to
424 McGeer, “Military Texts,” 907-914.
227
mounted archery. This feature was noted in the military manuals since Late
Antiquity. Taking a closer look at these works, it is possible to see that the military
tactics and methods that were identified with the “Skythian” peoples were repeated
constantly throughout the centuries. However, this repetition demonstrates how
foreigners were unimportant in the Byzantine worldview.
Gilbert Dagron who studied the representation of the foreign ethnoi in
military manuals, lists several different populations which had different martial
features. He emphasizes the clarity and the descriptive value of the brief ethnological
passages in these texts.425 His approach inspired me to write this sub-chapter.
Therefore, I will first discuss the representation of Turkic peoples in the tradition of
military writing, and then examine the continuation of this representation in the texts
in the corpus of this thesis.
The Strategikon, written around the year 600 and attributed to Emperor
Maurice (r. 582-602), can be considered the first Byzantine military manual. Some
later manuals repeated the contents of this treatise. It contains a chapter about the war
tactics and attitudes of foreign nations and narrates the features of Persians,
Skythians, Huns, “Fair-Haired Peoples” (i.e., Franks and Lombards), Slavs, and
Antes. In the sub-chapter which bears the title “Dealing with the Skythians, that is,
Avars, Turks and others whose way of life resembles that of the Hunnish peoples,”
the author describes and analyzes various war tactics of steppe nomads. As was
already seen in Chapter 3 exploring the late antique ethnonym “Hun,” this name was
repeatedly used as an umbrella term for the steppe nomads. In Maurice’s text, the
ethnonym Turks (Τοῦρκοι) designates the population of the First Türk Khaganate (in
the 6th and 7th centuries).
425 Dagron, “Ceux d’en face: Les peuples étrangers dans les traités militaires byzantins," 207-232.
228
The author describes these nations as:
The Skythian nations are one, so to speak, in their mode of life and in their
organization, which is primitive and includes many peoples. Of these peoples,
only the Turks and the Avars concern themselves with military organization,
and this makes them stronger than the other Skythian nations when it comes
to pitched battles. The nation of the Turks is very numerous and independent.
They are not versatile or skilled in most human endeavors, nor have they
trained themselves for anything else except to conduct themselves bravely
against the enemies. The Avars, for their part, are scoundrels, devious and
very experienced in military matters.426
He then narrates the military equipment used by the Hunnic nations as he counts,
“mail, swords, bows and lances”427 and their way of fighting: “They prefer battles at
long range, ambushes, encircling their adversaries, simulated retreats and sudden
returns, and wedge-shaped formations.”428 The author then makes an interesting
distinction between Turks and Avars, stating that the former is an independent and
populous nation that knows no other art than fighting bravely against its enemies,
and the latter is very experienced in military matters, yet is a group of devious
scoundrels.429
To gain an advantage over Turks on the battle pitch was also closely related
to exploiting their moral weaknesses. The Turks are not homogeneous and loyal to
their agreements, so they can easily defect and desert. The author of the Strategikon
describes this fact as such: “They are seriously hit by defections and desertions. They
are very fickle, avaricious and, composed of so many tribes as they are, they have no
sense of kinship or unity with one another. If a few desert and are well received,
many more will follow.”430
426 Maurice’s Strategikon, 116.
427 Maurice’s Strategikon, 116.
428 Maurice’s Strategikon, 117.
429 Maurice’s Strategikon, 116.
430 Maurice’s Strategikon, 118.
229
The author also emphasizes the relationship between climate, people’s
character, and warfare. This approach is basically a repetition and vulgarization of
Aristotle’s ideas. According to him, climate determines a people’s character and their
war methods. Turks “endure heat and cold, since they are nomadic peoples.”431 He
describes the social order of Turks by asserting that they have a monarchical model
of government and they are not governed by love, but by fear. He also underlines that
“their rulers subject them cruel punishments for their mistakes.”432 This
representation is antagonistic to the image of the Persians who are represented as
wicked, obedient, and servile. However, he asserts that the Persians were also ruled
by fear, just like the Skythians.433 On the other hand, the “light-haired peoples”
respect liberty, but they disobey their kings and lack the discipline in the
battlefield.434 Finally, the Slavic peoples lack a government and since they live in
absolute freedom, they refuse to be enslaved or governed. However, they also have
“many kings among them always at odds with one another.”435 The Slavs are also
undisciplined in the battle, just like the Germanic and Turkic peoples.
These moral weaknesses make it difficult to reach agreements with
Skythians, because “they scorn their oath, do not observe agreements, and are not
satisfied by gifts.”436 Moreover, if they demonstrate themselves to get an agreement
431 Maurice’s Strategikon, 116.
432 Maurice’s Strategikon, 116.
433 Maurice’s Strategikon, 113.
434 The same theme appears, some five centuries later, in Anna Komnene’s description of the Celtic
“race” (i.e. the Crusaders from France and Belgium): “The truth is that the Celtic Race, among other
characteristics, combines an independent spirit and imprudence, not to mention an absolute refusal to
cultivate a disciplined art of war; when fighting and warfare are imminent, inspired by passion they
are irresistible, evident not only in the rank and file, but in their leaders too, charging into the midst of
the enemy’s line with overwhelming abandon – provided that the opposition everywhere gives
ground; bu if their foes chance to lay ambushed with soldier-like skill and if they meet them in a
systematic manner, all their boldness vanishes.” Komnene-Reinsch-Kambylis, Annae Comnenae
Alexias, 339; Komnene-Sewter-Frankopan, The Alexiad, 313.
435 For Germanic peoples, see Maurice’s Strategikon, 119; for the Slavs, see 123.
436 Maurice’s Strategikon, 116.
230
and accept gifts, they can easily betray their pact anytime. It can be said that
Strategikon became the model of military writing for later generations. It divides the
foreigners into various groups by their ethnicity and lifestyle and asserts that there
are certain relationships between the warfare and the social order, and the climate
and the physical geography of the countries where these populations lived.
In the subsequent centuries, several military treatises were produced. The
Taktika, attributed to emperor Leo VI, paraphrases the chapters concerning the
foreigners of the Strategikon, yet it also adds a digression about Byzantium’s new
enemies: the Arabs. The Taktika does not say anything new or special about the
Turks. However, the author employs the ethnonym Turk and uses the old material
regarding Göktürks, and explicitly refers to Hungarians. Furthermore, he changes the
ethnonym Avar with Bulgar. When he explains the difference between the Turks and
Bulgars, he states that Bulgars converted to Christianity and adopted Roman customs
while Turks remained pagans. He also adds that Turks fought against the Bulgarians
on the Danubian shores. However, there is no striking difference between the
methods of warfare of the 7th century Göktürks and the 10th century Hungarians.437
Another military manual is attributed to Emperor Nikephoros Phokas. This
text is commonly known as De Velitatione (or De re militari) and is dated to the 10th
century. It is a book dedicated exclusively to the military matters of the empire’s
eastern frontier, so it mentions only the Arabs and lacks a chapter concerning other
ethnoi.438 The last important military manual is the Taktika of Nikephoros Ouranos
dated to the late 10th century. This work also lacks an ethnographic section.439
437 The Taktika of Leo VI, 452-453. For the evangelization and subsequent change of the Bulgars, see
section 42.
438 Le Traité Sur La Guérilla (De Velitatione), ed. Dagron, Mihăescu, and Cheynet.
439 McGeer, “Military Works,” 912.
231
Having thus summarized the representation of Turkic peoples in Byzantine
military manuals, I can return to the historiographical sources. In the Synopsis
Historikon, Skylitzes makes a digression about the early Seljuk expansion, which has
already been discussed. This digression is followed by an episodic narrative of the
early Seljuk-Byzantine encounters. The author describes the early Seljuk army as a
band of robbers, criminals, and riff-raff. He mentions a battle between Katakalon
Kekaumenos and Hasan the Deaf, the nephew of Ṭog̲ h̲ rı̊l, near the river Stragna
(Great Zab). The Byzantine commander, this time, deceives the Seljuks:
At dawn Hasan emerged from his own encampment on the river Stragna and
advanced ready to do battle. When he encountered nobody, he approached the
Roman stockade. No guards could be seen, no voice was heard; it was
completely devoid of forces. Thinking the Romans had taken to flight, he
breached the fortification at several points and ordered the seizure of booty to
begin. Towards evening, the Romans emerged from their hiding places and
hurler themselves on the Turks, who were now scattered and disorganized.
They were immediately routed, for they could not withstand the irresistible
force of the Roman charge. Hasan was the first to fall, fighting in the front
line; every stout-hearted man in the army fell too. The very few who survived
the fray fled unarmed through the mountains and found refuge in the cities of
Persarmenia.440
Michael Attaleiates defines the early Seljuk activities as a series of raids and gives no
important military detail. On the eve of the Battle of Manzikert, he describes the
Seljuk raid near the river Euphrates as “The barbarians, who were prepared to shoot
form a distance, easily wounded them from afar while remaining untouched
themselves, to the point where they forced them to go into the river and fight them
there. At the same time, the enemy who stood on the banks kept shooting at the
Romans, causing many casualties and forcing them to turn and run.”441 He also refers
440 Skylitzes-Thurn, Ioannis Scylitzae Synopsis Historiarum, 449; Skylitzes-Wortley, A Synopsis of
Byzantine History, 422.
441 Attaleiates, The History, 171-172.
232
to Emperor Romanos IV’s opinion that the Seljuks “were not strong enough to meet
the Romans in close combat.”442
In the case of the Battle of Manzikert, most probably the author’s eyewitness
account, the Turks are described as “wicked by nature and masters of deceit” for
“they accomplish everything by trickery and unabashed reversals.”443 However, in
the narrative of Attaleiates, there is a detail which contradicts the features attributed
to Skythian peoples in the Strategikon. Just before the battle of Manzikert, the
Byzantines became suspicious of the possibility that the Pechenegs might defect to
the other side. However, the author himself made the Pechenegs swear oaths in their
traditional manner and managed to lift the suspicion over them. The author was
proud of his success: “I made them into firm guardians of the agreement. Nor did I
fail in my purpose, for not one of them defected to the enemy during the battle.”444 I
assume that the suspicion over the fidelity of the Pechenegs demonstrates that the
stereotype about the attitudes of the Skythians were frequent among the Byzantine
military officers.
The descriptions of the tactics in the military texts also shaped the Byzantine
expectations in the battlefield. Emperor Romanos IV assumed that “the Turks would
make an ambush and attack the unguarded camp” and if he continued his pursuit
much longer “(they) would reverse their flight and shoot (arrows) from a
distance.”445
As it was seen, the author does not see a big difference between the military
tactics of the Seljuks and the Pechenegs. They both followed a common model of
442 Attaleiates, The History, 234-235.
443 Attaleiates, The History, 284-285.
444 Attaleiates, The History, 286-289.
445 Attaleiates, The History, 292-293.
233
steppe warfare in which horse archery, tactical withdrawals, and various tricks played
a key role. In harmony with this early material, Anna Komnene expresses her idea
about the nature of the Turks, as the thoughts of Constantine Dalassenos, who was
sent by Alexios I Komnenos against Tzachas: the Byzantine nobleman “knows” that
the Turks are “of treacherous nature.”446 During a conversation with the crusader
leader Count Baldwin of Bouillon, later King Baldwin I of Jerusalem (r. 1100-1118),
Alexios I states, “I strongly recommend you not to take up position in the rear of the
army, nor in the van; stand in the centre with the junior officers. I know the enemy’s
methods and have had much experience of combat with the Turks.”447
Both John Kinnamos and Niketas Khoniates give us accounts of Byzantine
military clashes against the Seljuk Turks in the 11th century. The depiction of
Kinnamos follows the old model; the Seljuk Turks are tricksters. In Manuel
Komnenos’ campaign against the Seljuks in 1146, the Byzantine army did not find
the chance to have a direct encounter with the Seljuk army on the battlefield. The
emperor warned his soldiers: “Romans, do not let barbarian trickery turn your
shrewdness to fear: while there is a lack of standards in the army visible in front of
us, you should not imagine that they are elsewhere with another force.”448 During
this campaign, the Seljuk ruler Masʿūd I escaped constantly from the Byzantine
forces, so at one point the emperor humiliated him by saying to his ambassador,
“Report this to your sultan (…) You, however, fled continually, like runaway
slaves.”449 Apart from the Seljuks, there is one unique feature of the Pecheneg
446 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 225; Komnene-Sewter-Frankopan, The
Alexiad, 205.
447 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 317; Komnene, Sewter Frankopan, The
Alexiad, 292.
448 Kinnamos-Meineke, Ioannis Cinnami epitome rerum ab Ioanne et Alexio Comnenis gestarum, 43;
Kinnamos-Brand, Deeds of John and Manuel Comnenus, 42.
449Kinnamos-Meineke, Ioannis Cinnami epitome rerum ab Ioanne et Alexio Comnenis gestarum, 58-
59; Kinnamos-Brand, Deeds of John and Manuel Comnenus, 52.
234
warfare which was underlined by Khoniates: the use of wagons. These war wagons,
similar to tábors used by the Hussites of Bohemia in the 15th century, gave an
advantage against the enemy infantry in the battlefield. However, when the ramparts
of the wagons were destroyed, the Pecheneg troops could be scattered: “When the
rampart of wagons had been demolished and the fighting had turned into hand-tohand
combat, the enemy was put to inglorious flight, and the Romans pursued them
boldly.”450 The author even uses the expression “wagon-dwellers” (hamaxobion) to
define the Pechenegs. This expression that reflects the nomadic life, evokes a similar
epithet used for the nomads: σκηνῖται, which means “tent-dwellers.”451
However, the next year, the German king who participated in the Crusade,
King Konrad, risked being captured by the Turks:
The Turks turned tail and pretended flight; but when their [the Germans’]
cavalry was exhausted and they were far from camp, they [the Turks] made
rapid charges and slew horses and men. The same thing which happened
frequently cast them into immeasurable terror. Then it was possible to
observe those who were formerly lash braggarts who attacked in the fashion
of irresistible brutes, cowardly and ignoble and incapable of either doing or
planning anything. Then Konrad452 (for he was courageous in warfare) rushed
against the Turks, lost the particularly swift horses which the emperor had
presented to him, and came close to being captured by those barbarians.453
This passage is reminiscent of the warning of Alexios I to Count Baldwin: the Turks
are particularly dangerous for the armies who do not have the experience of fighting
them.
Khoniates’ depiction of Turkish warfare fits the earlier standards. He
underlines the barbarian customs of the Turks in his narrative about the aftermath of
450 Khoniates-Van Dieten, Nicetae Khoniatae Historia,16; Khoniates-Magoulias, O City of Byzantium,
11.
451 Shukurov, The Byzantine Turks, 1204-1461, 28.
452 Konrad III Hohenstaufen (r. 1138-1152) King of Germany and Italy.
453 Kinnamos-Meineke, Ioannis Cinnami Epitome rerum ab Ioanne et Alexio Comnenis gestarum, 81-
82; Kinnamos-Brand, Deeds of John and Manuel Comnenus, 68.
235
the battle of Myriokephalon: The Seljuk soldiers cut the phalluses of the corpses of
the Byzantine soldiers on the battlefield. The author tries to offer a rational
explanation about this act: “It was said that the Persians took these measures so that
the circumcised could not be distinguished from the uncircumcised and the victory
therefore disputed and contested since many had fallen on both sides.”454
In the description of the first large scale battle between the Byzantines and
Ottomans, the Battle of Bapheus (1302), Pakhymeres repeats basically the same
narrative: ʿOt̲h̲ mān, the eponymous founder of the Ottoman state, ambushes
Mouzalon and his troops and encircles them. ʿOt̲h̲ mān’s troops, enforced with the
Persian auxiliaries from the Meander region, defeat the Byzantines with their
superior archers and make them withdraw.455 The Ottomans who fought against the
Byzantines in this battle employed the infantry forces among their troops. These very
episodes document an important point of evolution of warfare among the Turks of
Asia Minor. Moreover, it can be assumed that, at least in the early stages of the
Seljuk invasion of Asia Minor, the difference in warfare and the technological
superiority of Turkish mounted archery could have played a key role in the Byzantine
loss of Anatolia. Further changes in the Byzantine military demonstrate that the
empire tried to renew its armies to counter the mounted archery tactics of the Turks.
However, interpreting the Seljuk conquest of Anatolia from a detailed military
perspective is beyond the limits of this thesis.456
In Doukas’ Historia Turco-Byzantina, although the author mentions various
episodes of Ottoman wars, there is no particular emphasis on the cavalry and nomad
454 Khoniates-Van Dieten, Nicetae Khoniatae Historia, 190; Khoniates, O City of Byzantium, 107.
455 Pachymeres, Relations historiques, 364-365.
456 Kaegi, “The Contribution of Archery to the Turkish Conquest of Anatolia,” 96-108, particularly
107-108.
236
warfare. The author is well informed about the Ottoman military; he demonstrates his
knowledge about different Ottoman military units (janisseries, azabs) and he just
does not repeat the old narrative of nomad archers. Such details were not present
even in the first chapters of his work, he does not depict even the early Ottomans in
that way.
Finally, in the Histories of Khalkokondyles it can be seen how the Turkish
warfare has changed through the centuries. He never describes the contemporary
Ottoman military by using the topoi in the military manuals and early military texts.
Moreover, in a dialogue he invented, Bāyazīd says:
My men, it seems as though you are afraid of their (Timurids’) numbers, that
is how I interpret it. (…) We too in Europe have often gone into battle and
routed the most courageous races in the world, the French, and the
Hungarians. Therefore, do not belittle our bravery or declare us to be worse
and less significant than the Skythians and Chaghatai, who have never ever
used swords but who only shoot with a bow and arrows, as they positively do
not want to come to blows. 457
In this passage, the typical feature of the nomadic barbarian –warfare based on horse
archery– is identified exclusively with the Timurids, and the Ottomans are placed
within the borders of the “civilized world.”
Before ending this sub-chapter about the representation of Turkic warfare in
Byzantine texts, it will be appropriate to add a brief section about Turkish
seamanship in the Byzantine sources. Being a people from Central Asia, the Turks
had no familiarity with the sea or seamanship when they first arrived in Asia Minor.
However, in subsequent centuries, they progressively gained familiarity with this
craft. As it was discussed in Chapter 5, Anna Komnene gives an account of the deeds
of Tzachas, a Turkish-Byzantine warlord who ruled a maritime principality for a
457 Khalkokondyles, The Histories, 250-253.
237
short period of time. She adds that he collaborated with the Greeks to construct his
fleet.458 As it was seen, Greco-Turkish cooperation was necessary when it came to
seamanship.
The traces of Tzachas’ model of the maritime emirate can be followed in later
Turkish principalities of the 13th century formed along the western coast of Asia
Minor. In that region, four principalities were formed from north to south: Karasi,
Saruhan, Aydın, and Menteshe. All four of these beyliks possessed a fleet and were
engaged with piracy activities, often against the commercial vessels belonging to
Italian city-states and the Aegean islands ruled by Italian dynasties. Among the rulers
of these emirates, Umur of Aydin was known to be the greatest seaman who was
involved actively in “maritime ghaza.” His life is also narrated in the History of
Doukas, whose family took refuge in the emirate of Aydin in the 14th century. By the
15th century, Turkish seamanship was already taken for granted by the Byzantines.
This led Khalkokondyles to attribute victories in unreal and anachronic sea battles to
Ertog̲ h̲ rul, the ancestor of the Ottoman dynasty.459 However, Turkish seamanship was
not distinguishable from piracy in the eyes of the Byzantines.
In concluding this sub-chapter, the findings can be summarized thus:
i. First of all, in the medieval world, the battlefield is an important place
to encounter the other. According to Byzantine authors, the “national traditions”
in the art of war were mere reflections of the national character determined by
climate, lifestyle, and social order which were also interrelated with one another.
458 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 222; Komnene-Sewter-Frankopan,The
Alexiad, 202.
459 “(Ertog̲ h̲ rul) enjoyed success while fighting wars in many places and he even built ships in order to
sail to the islands of Aegean, both those near Asia and those near Europe, and pillage them. He
ravaged Europe: among other exploits, he even entered the river Tearos, the one by Ainos, and sailed
his ships up a long stretch of it. He is also said to have made landings in at many other places in
Europe, reaching the Peloponnese, Euboia and Attica, where he plundered the land and made huge
profits by carrying off as many captives as possible as slaves.” Khalkokondyles, The Histories, 16-17.
238
ii. The Byzantine representation of Turkic warfare reflects the Byzantine
perception of the national character of the Turks. They were represented as
faithless, treacherous, wild, cunning, deceitful, yet independent people. Their
original element in warfare was fake withdrawal. This tactic was also in line
with the “national character” of the Turks because it represented their general
faithlessness and lack of morality according to Byzantine texts.
iii. Turkic armies were depicted as horse archer armies until the 14th
century; however, as Turkish warfare evolved, later authors gave up repeating
this traditional description.
7.2 The representation of Turkic women in Byzantine literature
A striking feature of the Byzantine narratives about the Turks between the 11th and
the 14th centuries is the scarcity of narratives regarding women. As pointed out by
many scholars who studied the history of women in the Muslim world, the
representation of women was rather limited in the medieval historiography of that
part of the world, because of the segregation of the sexes and the isolation of women
from public life.460 However, the full segregation of women and men in the Muslim
world must also be considered a historical issue and cannot be generalized to all the
Muslim societies of all times, ignoring the nuances.
Foreign women occupied a negligible place in Byzantine historiography,
except as brides who came to Constantinople.461 In that sense, the role of Turkic
women did not differ. In this context, Islam could also be a factor in this invisibility.
460 For an introduction, see Keddie, Women in the Middle East: Past and Present and Keddie &
Baron, Women in Middle Eastern History – Shifting Boundaries in Sex and Gender.
461 For a recent study which also deals with late Byzantine empresses of foreign origin, see Melichar,
Empresses of Late Byzantium: Foreign Brides, Mediators and Pious Women.
239
Before moving on to the rare references to Turkish women in Byzantine texts, I will
start by investigating Ibn Battuta's travel narrative, which displays the status of
Turkish women before Islam.
The earliest references to the status of women in Turkic societies were the
testimonies of Ahmad Ibn Faḍlān (877-960) of Baghdad, the Arab traveler of Turkic
and Slavic lands in the 10th century. His travel narrative is of utmost importance both
for the history and ethnography of Turkic peoples in the eve of Islamization and for
the early settlements of Slavs and Vikings during the formation of the State of Rus’.
He visited several Turkic peoples beyond the borders of Muslim oikoumene such as
Khazars, Oghuz, Volga Bulgars, and Bashgirts.
Ibn Fadlan visited the Oghuz settlements along the Oxus River in 921. In that
period, the Oghuz were still pagan, yet the influence of the Muslim traders along the
river could be observed. The author narrates the following episode:
The women hide no part of their body from anyone. One day, we descended
on one of them (an Oghuz) and we sat together. The wife of that man was
also among us. When we were chatting, she opened up her sexual parts and
scratched them while we looked at her. We hid our faces with our hands and
said ‘May God forgive us.’ Her husband laughed and said to our interpreter:
‘Tell them, she opens up her sexual parts in your presence and you see them,
but you look without reaching them. That is better than if she covers them but
someone can reach them.’462
However, despite his lack of sympathy for the Oghuz, Ibn Fadlan does not interpret
their lack of gender segregation as a moral inferiority. He underlines that among the
Oghuz people, adultery is very rarely committed: “They do not know about adultery,
but if they get to know an act of this nature, they split both adulterers in two.”463 Yet
this passage regarding adultery can be read rather as a criticism of his contemporary
462 Ibn Fadlan, Voyage chez les Bulgares de Volga, 38-39.
463 Ibn Fadlan, Voyage chez les Bulgares de Volga, 39.
240
Muslim society than a sincere praise of the Oghuz tribes who were merely
unbelievers in his eyes. Ethnography was frequently used for social criticism of
similar issues that occur within the writers’ society since Tacitus. Comparable to Ibn
Fadlan’s emphasis on the sexual morality of the Turks, Tacitus considered the sexual
morals of Germanic peoples superior to that of the Romans.
Another Turkic people that surprised the Arab traveler with the coexistence of
men and women in daily life were the Volga Bulgars. The author complains that the
men and women of this nation go all naked in a river and bathe together. Yet,
confessing that he could not be successful to remove these customs, he adds that they
never commit the sin of adultery, and if any two of them do so, they kill both the
woman and man brutally.464 Ibn Fadlan’s depiction of pre-Islamic Turkic societies is
unique with its emphasis on the status of women and the gender relations in daily
life. Therefore, this narrative can be compared with the Byzantine narratives
concerning Turkic peoples.
I began this sub-chapter with these citations from an Iraqi traveler to portray
an earlier encounter between the Turks, who converted to Islam later, and the Arabs.
Eventually, it can be assumed that there is a certain change in the status of Turkish
women after converting to Islam. Turks gradually adopted the social norms of
Muslim societies that had institutionalized segregation of the sexes, obedience to
men, and modest dress. These social norms of Islamic society limited women’s
public life. Nevertheless, this change was gradual, and it did not affect certain strata
of the society, particularly the ones who continued the nomadic lifestyle of their
ancestors. The representation of the women warriors in the Turkish epic narratives in
the Middle Ages also demonstrate that the ancient ethos lived among Turks for a very
464 Ibn Fadlan, Voyage chez les Bulgares de Volga, 65.
241
long time. Even in the Great Seljuk Sultanate, where the Persianate Muslim culture
was gradually embraced by the ruling dynasty, the women of court were active in
public life.465
The Byzantine sources remain silent about Turkic women; neither Skylitzes,
nor Attaleiates mentions a single Turkish woman. One can expect Anna Komnene to
include the deeds of women in her writing as a female author; however, she too
remained silent about Turkic women. There are only scarce references to Turkic
women in the Alexiad; she only refers to unnamed wives and daughters of Turkic
rulers.
There are very brief references to Seljuk noblewomen in the context of the
Byzantine campaign against Seljuks in Kinnamos’ and Khoniates’ texts (1146). In
Kinnamos’ work the unnamed wife of the Seljuk Sultan Masʿūd I appears. The
existence of a correspondence between her and the emperor Manuel I Komnenos
gives the impression that she was of Greek origin. In her letter to the emperor, she
states that she has in readiness around two thousand sheep and a vast quantity of
oxen and many other sorts of edibles to welcome the emperor, yet he will not be
welcomed because of the Byzantine army’s burning of the dwellings of the city.466
However, Niketas Khoniates mentions an unnamed daughter of the Seljuk Sultan
Masʿūd I. This princess was also the spouse of John Komnenos who is the cousin of
Manuel I. Masʿūd I, as stated above, avoided direct encounters with the emperor and
withdrew his army to the inner parts of Anatolia. Khoniates narrates the encounter at
the gates of Ikonion by stating “one of his (Masʿūd’s) daughters, reportedly married
465 Peacock, The Great Seljuk Empire, 178. The author states: “Far from being restricted to the domain
of the harem, Seljuk women played a much more public role than many of their counterparts in other
contemporary Muslim dynasties.”
466 Kinnamos-Meineke,Ioannis Cinnami epitome rerum ab Ioanne et Alexio Comnenis gestarum, 46;
Kinnamos-Brand, The Deeds of John and Manuel Comnenus, 44.
242
to the emperor’s cousin, John Komnenos, the son of the sebastokrator Isaac who
become some trifling vexation against his uncle, the emperor John I Komnenos, had
fled and defected to Masʿūd, peered out from above the walls and delivered a
persuasive defense on behalf of her father, the sultan.”467 Both of these narratives,
whether they represent the reality or are merely fiction, reflect the Byzantine authors’
thoughts of the Seljuk noblewomen having a public role.468
Moreover, Niketas Khoniates narrates a bloody episode which takes place in
the Danishmendid city of Amaseia. After the death of Yağıbasan, the ruler of
Danishmendid emirate, his wife marries his brother and rivals Zünnun in secret.
However, the marriage provokes a rebellion in the city and the Amaseians kill the
woman. The author does not mention the ethnic identity of these Amaseians,
however by the brutality of the event, it can be supposed that they were barbarians.469
Finally, in Khoniates’ Annals, in the context of Seljuk dynastic struggles after
the reign of Qiliç Arslān II, it was stated that the title-claimant Rükneddin
(Sulaymāns̲ h̲ āh II) loathed his brother Kayk̲ h̲ usraw (later became the Sultan
Kayk̲ h̲ usraw I), because the latter’s mother was Christian.470
That being the case, there must also be distinguished the Turkish women born
Muslim and who were supposedly ethnic Turks, from the women who have a Greek
origin and a Christian background, who then became a part of the Seljuk society via
467 Khoniates-Van Dieten, Nicetae Khoniatae Historia, 53; Khoniates-Magoulias, O City of
Byzantium, 31.
468 Roman Shliakhtin explains the visibility or more precisely “the right to speech” of the princess by
stating that she is half-Byzantine, because she married a member of the Komnenos dynasty. However,
I do not share his opinion, because she is married to the son of a Byzantine dissident who escaped
from Constantinople and took refuge in Ikonion. So, although her husband is a Byzantine noble, she is
still talking in the name of the legitimacy of where she grow up. Shliakhtin, From Huns into Persians:
The Projected Identity of the Turks in the Byzantine Rhetoric of Eleventh and Twelfth Centuries, 175.
469 Khoniates-Van Dieten, Nicetae Khoniatae Historia, 122; Khoniates-Magoulias, O City of
Byzantium, 69.
470 Khoniates-Van Dieten, Nicetae Khoniaiate Historia, 521; Khoniates-Magoulias, O City of
Byzantium, 286.
243
marriage or slavery. These women, therefore, had more access to Byzantine society,
but as it was seen in the case of Kaikhusraw’s mother, they can attire the hate of their
new compatriots. In Pakhymeres’ Syngraphikai Historiai, the description of the
mother of Kaykāʾūs II is worthy of attention. The author describes this noblewoman
of Greek origin as “an excellent Christian”.471 In Pakhymeres’ work, there are other
sporadic references to Turkic women: such as the kidnapping of the Turcopoles’
women by the Almogavars, the Catalan mercenary band in the service of the Empire.
In Gregoras’ Roman History, there is a narrative about a Skythian woman
who converts to Christianity: A Skythian woman buys a Christian captive from
Thrace and marries him. Then this man sees his ex-wife who was also a slave in the
hands of Skythians and his new Skythian wife purchases her and makes her a maid.
The woman converts to Christianity and, together with her husband and her
husband’s ex-wife, settles in Constantinople. When they settle, the maid complains to
the Patriarchate about the Skythian woman by saying that she was unfair to her. Then
the Skythian woman by her own will decides that she will release the maid and
permit her to find her ransom in her native region of Thrace; however, there she will
fall captive to Skythians again. The Skythian woman’s personality is appreciated by
the Patriarch. She acted in a noble and honest manner; therefore, she deserves to live
happily with her husband.472 The protagonist of this episode is similar to the
extremely Christian mother of Kaykāʾūs II. If someone is a good Christian and an
honest person, could be depicted in a positive light. However, as I have already
mentioned, the scarcity of depictions of Turkic women makes it difficult to propose a
general explanation about this situation.
471 Pachymeres, Relations historiques, 182-183.
472 Gregoras-Bekker-Schopen, Nicephori Gregorae Byzantina historia, Gregoras-Van Dieten,
Rhomaïsche Geschichte, 284-285.
244
Doukas narrates the Turks’ interest to mingle with foreign women as such: “if
they seize a Greek woman or Italian woman or a woman of another nation or a
captive or a deserter, they embrace her as an Aphrodite or Semele, but a woman of
their own nation or of their own tongue they loath as though she were a bear or
hyena.”473
This passage could be read as a reflection of the Turks’ general attitude
towards polygamy and exogamy. In early Ottoman chronicles, there are several
references to ghazis who marry the women of the town they conquered.474 It can be
assumed that polygamy among Muslims constituted the first and foremost difference
between the Byzantines and the Muslim peoples. Furthermore, as it was
demonstrated by the same sources, the 14th and 15th centuries were a period in which
the Ottomans enslaved Christians massively. However, these arguments do not
explain why the Turks loathed their own women. Moreover, what was the difference
between “the woman of their own nation” and “the woman of their own tongue”?
In the Historia Turco-Byzantina, Doukas draws a distinction between the
Turks, which included the Ottoman population and the Anatolian beyliks, and the
speakers of Turkic languages, who were described with different ethnonyms. Doukas
addresses the Akkoyunlu as "Persarmenian", and the Timurids as "Skythian", and
underlines that they speak Turkish, but does not identify them as Turks. The author
thinks that these populations spoke roughly the same language, but differed in terms
of their nations. Doukas’ Turks dislike not only the women of their own country, but
also Turkic women from other lands. This statement could contain some
exaggeration, yet it should be assumed that in a period when there was an influx of
473 Doukas-Grecu, Historia Turcobyzantina, 59; Doukas-Magoulias, Decline and Fall of Byzantium to
the Ottoman Turks, 73.
474 See the case of the conquest of Nicea in Aşıkpaşazade, Tevarih-i Al-i Osman, 313.
245
foreign women who were nothing but slaves, these men did not opt for Muslim
women who even under the patriarchal structure of Islam, were requesting more
responsibility.475
Finally, Kantakouzenos, in his history, when criticizing the empress Anna of
Savoy, refers to a Turkish proverb to reinforce his own patriarchal and misogynistic
approach to women: κἂν μέχρι νεφελῶν ἀφίκηται ἡ κεφαλὴ τῆς γυναικός, οὐδὲν
ἔλαττον ἢ πρότερον ἐφάπτεται τῆς γῆς.476 This proverb could be translated as “if the
head of a woman could arrive as far as to the cloud, she is always attached to the
ground.”477 As in many other cases in Byzantine literature, the context of the
reference of such a saying is a critique directed at the internal politics of Byzantium.
But, in a broader context, as a person who personally met the Turkish rulers and had
a certain knowledge of the Turkish societies of his time, Kantakouzenos’ use of such
a proverb, could also demonstrate that he personally thinks the Turkish society is
more patriarchal than the Byzantine one. At least, according to the context of
Kantakouzenos’ critique, this approach could have sense: during the life of
Kantakouzenos, there was no Turkish female ruler. The woman closest to the position
and prestige of Anna of Savoy in Turkish Asia Minor could be Nīlūfer K̲ h̲ ātūn
(d. first half of 1380s), the first consort of Ork̲ h̲ an, who was also a Greek
noblewoman from Bithynia. More information about her will be given below.
475 In Doukas’ Historia Turco-Byzantina, there are some more references to Turkish women: he
mentions the wife and daughter of Junaid, the ruler of Smyrna, and her marriage with a certain
Abdullah, of Albanian origin, who is a slave of him, the women of Junaid’s family who obey
Meḥmed I, he mentions also that though Murād II first married the daughter of Isfandiyar, “he longed
more for this new wife (Mara Brankovic) who was beautiful in both body and soul.” Murād II seems
to follow the abovementioned pattern toward foreign women. Doukas-Grecu, Historia
Turcobyzantina, 141, 143, 259-261; Doukas-Magoulias, Decline and Fall of Byzantium to the
Ottoman Turks, 116, 117, 176.
476 Kantakouzenos-Schopen, Ioannis Cantacuzeni eximperatoris Historiarum, vol. 2, 48;
Kantakouzenos-Fatouros, Geschichte, vol. 3, 23.
477 Despite the existence of several misogynistic proverbs in Turkish, I could not identify it.
246
In concluding this sub-chapter, it may be suggested that the situation of
women among Turks deteriorated during Islamization and the adoption of the
dominant social models of the Muslim oikoumene. Maybe a comparative approach
between the two authors from different cultural backgrounds can enlighten the
situation of the women among the Turks. In Palamas’ narrative of his captivity, he
does not mention any woman whom he saw or talked to in the Ottoman court.
However, the Moroccan traveler Ibn Battuta (1304-1369), who visited the Ottoman
territories some twenty years before the Byzantine theologian, mentions the Nilüfer
Hatun, whom he saw and referred to as Bayalun, the wife of Ork̲ h̲ an in Bursa. This
noblewoman of Greek origin was left as regent in the Ottoman capital, during a
campaign of his husband. The author describes Nilüfer as a wise woman who helps
them, without mentioning her Greek origin.478 In this respect, it may be more
revealing to pay attention to what the authors intended to see.
In Khalkokondyles’ Histories, there are a few more women who are in our
field of interest: His narrative about the rivalry between Bāyazīd I and Tamerlane
gives not only some remarks on the wife of Tamerlane, but also some details about
the authors’ knowledge about Muslim marriage. Bāyazīd I already had a conflict with
Tamerlane because of the latter’s anger to the Ottoman annexation of Anatolian
beyliks. During the correspondence between the two rulers, Bāyazīd I says: “if he
does not come to fight against me now, let him renounce his wife three times.”
Khalkokondyles obviously refers to the Islamic practice of triple talaq.479 The author
explains this fact as follows: “This is an insult among this race, for Muhammad
ordained that one should renounce his wife three times if she is not obedient. This
478 Ibn Battuta, The Travels of Ibn Battuta, 453-454.
479 Khalkokondyles, The Histories, 170-171.
247
happens because there is a law among them that prohibits a person who has rejected
his wife from taking her back into his household, for this is considered improper.
When a man has said that his marriage is dissolved on three spleens the law prevents
him from entering again the same marriage, unless he does so after she has
committed adultery with another man in the meantime who has also thrown spleen
three times.”
His messenger brings this insult to Tamerlane. However, the wife of
Tamerlane whom the author introduced as such: “they say that (she was) an
especially devout woman” would not allow her husband to attack Bāyazīd I’s
territories, because “he was a man who was worthy of praise in their religion and was
fighting against the faction of Jesus.”480 Then, Tamerlane makes the messenger
repeat this insult before his wife and asks her if it was right to allow the Ottoman
ruler to say such things. Khalkokondyles states: “He made it clear that if, on the one
hand, she still thought Bāyazīd right, he could no longer live with her in the future.
If, on the other hand, she had changed her mind and would now favor war, she would
be considered his wife and would assent to whatever it was that he was forced to do.”
Tamerlane’s wife responds by saying that Bāyazīd is mad, and his husband would
punish him justly. However, she adds that Bāyazīd still fights “on behalf of our hero”
against “the Hellenes and other peoples on the other continent.” Therefore it is not
right to wage a war against him, yet it is enough to occupy Sivas (Sebasteia) as a
revenge of his taking of the city Malatya (Melitene).
This narrative is interesting as it presents Tamerlane as a man who is ready to
wage a war because of an insult targeting his conjugal honor and his wife as a devout
noblewoman who is attached deeply to her religion and who insists her husband to
480 Khalkokondyles, The Histories, 170-171.
248
wage a low-profile war against the Ottomans. In his depiction of events, Tamerlane
represented a gentle husband who listens to his wife. However, the upcoming events
were to be different and the Ottoman-Timur conflict would provoke a great war
between two states.
In conclusion, as it was also stated by Roman Shliakhtin, Turkish women
were almost invisible in Byzantine historiography.481 Considering the abundance of
references to individual medieval Turkish women in Persian and Syriac chronicles,
and live depictions of upper-class Turkish women in hagiographies, and the
ambitious patronage projects of Seljuk noblewomen, this fact is contradictory.
Furthermore, despite the cultural change that came with Islam, which is increasingly
internalized by society, it is difficult to say that women were isolated from daily life
with strict gender segregation in Turkish Anatolia of the late Middle Ages.
Bertrandon de la Broqiuére, the Burgundian pilgrim who traveled Anatolia in a late
date as 1432-1433, remarked on the existence of women troops in the army of the
Beylik of Dulkadir: "Surkadiroly (Dulkadiroglu) who was 30.000 armed Turcoman
men and some 100.000 women brave and valiant as much as men..."482 In the end,
the scarcity of references to Turkic women in Byzantine texts can be only explained
by lack of interest of the Byzantine authors.
7.3 Intemperate and lustful: Turks and sexuality
A significant element of the barbarian image is its lack of moderation in various
aspects of life. The concept of moderation is closely related to civilized populations,
and the barbarians are notoriously intemperate in life. Thus, sexual behavior is an
481 Shliakthin, From Huns into Persians: The Projected Identity of the Turks in the Byzantine Rhetoric
of Eleventh and Twelfth Centuries, 172.
482 Bertrandon de la Broquère, Le Voyage d’Orient, 115.
249
aspect of daily life in which moderation and intemperance could play a
distinguishing role between the civilized and the barbarian. Since Herodotus, the
representation of barbarian sexuality is both the demonstration of weirdness and
inferiority from the civilized one.483
However, there is the need to contextualize this separation in an era in which
sexuality was perceived differently than our contemporary standards, formed mainly
in the 19th century, particularly in the Victorian Age. Michel Foucault’s colossal
L'Histoire de la Sexualité (1976)484 started a wave of discussions about whether the
seemingly extreme elements of sexuality were repressed or not. In this sub-chapter, I
shall deal with the sexual behavior attributed by the Byzantine authors to the Turks,
whom they consider perverse or deviant. Indeed, the Turks were identified with such
behaviors in some later Byzantine texts.
Anna Komnene stresses that the Ishmaelites “are indeed dominated by
Dionysos and Eros; they indulge readily in every kind of sexual license, and if they
are circumcised in the flesh, they are certainly not so in their passions. In fact, the
Ishmaelites are nothing more than slaves -trebly slaves – of the vices of Aphrodite.
Hence, they revere and worship Astarte and Astaroth, and in their land, the figure of
moon and the golden image of Khobar are considered of major importance.”485 This
expression clearly reflects the idea about the origin of the Muslims in De
Administrando Imperio. In the origin of the Muslim religion, there is a cult of
Khobar, or Koubar, a female divinity of uncontrollable lust. Though in De
Administrando Imperio, Islam was described as a heterodoxy, and the cult of Koubar
483 Wenghofer, “Sexual Promiscuity of Non-Greeks in Herodotus' Histories,” 515-534.
484 Foucault, Histoire de sexualité: La volonté de savoir, vol. 1.
485 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 298; Komnene-Sewter-Frankopan, The
Alexiad, 276.
250
– or Aphrodite – was reflected as a secondary trait of this religion; Anna Komnene
directly put this cult in the center of her narrative. Furthermore, she explains the
sexual aggressiveness of the Muslims as a result of their worshipping Astarte. As it
was known, the goddess Astarte is a Syrian divinity of fertility and love who was
identified with Aphrodite of the Greeks and Venus of the Romans.
One of the most striking Byzantine topoi about the Muslims is the attributed
relationship between their religion and the cult of Aphrodite/Venus. As it was known,
Aphrodite was the female divinity of love of Ancient Greeks. They considered the
Ka’aba in Mecca as an ancient temple of Aphrodite in Arabia. This narrative first
appears in the chapter about the origin of the Arabs and Islam in Constantine
Porphyrogennetos’ De Administrando Imperio:
And they (Arabs) pray, moreover, to the star of Aphrodite, which they call
Koubar (Κουβάρ), and in their supplication cry out “Alla wa Koubar”, that is,
“God and Aphrodite” (Θεὸς καὶ Ἀφροδίτη’). For they call God “Alla”, and
“wa” they use for the conjuction “and”, and they call the star “Koubar”, and
so they say “Alla wa Koubar.”486
This is an argument which was not invented by the author(s) of De Administrando
Imperio; it reflects a tradition of anti-Muslim polemics going back to John of
Damascus (d. 749) and Niketas of Byzantium (9th century) which alleges that inside
the Ka’aba there was an idol of Aphrodite that was worshipped by the Arabs.487
A small trace of this idea can be found in Khalkokondyles’ digression about
the history of Islam, in which he describes the Islamic prayer (sal’at) as “their
custom is to pray to God four times a day and they let nothing prevent them from
praying. On the day of Aphrodite [Friday], they all go into the shrines together to
pray.” He adds that “this race is especially devoted to prayer and for no reason at all
486 Constantine Porphyrogenitus, De Administrando Imperio, 78-79.
487 Meyendorff, “Byzantine Views of Islam,” 118.
251
will they agree to neglect it.”488 The author implies yet again an obscure connection
with Aphrodite but emphasizes the devotion and piety of the Muslims. However, he
continues by stating: “In other matters in their way of life and overall conduct
nothing is regarded as so reprehensible that it would prevent them from living
pleasurably; thus, they do not curb nature in any way. For they marry more than one
wife and have concubines from among their captives, however many as each man is
able to support and feed.” He then summarizes the Muslim marriage customs,
erroneously saying that a man can take up to five wives and there is the tradition of
bride price among them. He repeats the narrative of “the three spleens” which was
mentioned above, in the episode with Tamerlane. However, he also describes
Tamerlane as a man of sexual excesses:
He (Tamerlane) appointed the eldest son S̲ h̲ āh Ruk̲ h̲ (Σαχροῦχος) to be king
after him, while he himself indulged in sex and died preoccupied with that. In
fact, it is said that Tamerlane was tormented by his nature more than any
other person, to such a degree that he ordered young men to copulate with
women in front of him in order to become aroused enough himself to act. But
when he set sex aside, he would immediately turn to war against his enemies,
so that he was never at rest. It is said he committed offenses against his nature
with his sexual habits.489
However, this explanation could be an ideological fabrication to explain the sexual
aggression of Muslims. The centrality of polygyny in Islam was already mentioned.
This kind of polygamy clearly brings an image of hypersexuality. However, such
representations are very rare before the 13th century. For example, Niketas Khoniates,
who condemns the sexual vices of Byzantine emperors such as Manuel I, Andronikos
I Komnenos, and even accuses the sister-in-law of the Bulgarian King Asan of
488 Khalkokondyles, The Histories, 196-197.
489 Khalkokondyles, The Histoires, 270-273. This representation of Tamerlane does not match with
Khalkokondyles’ first representation of the Mongol warlord as a “gentle husband.”
252
adultery with a certain man called Ivanko, never constructs a sexual narrative of the
Seljuk Turks.490
In Akropolites’ History, the author condemns the Seljuk sultan G̲ h̲ iyāth al-
Dīn Kayk̲ h̲ usraw II of being licentious and incompetent:
A son [Kayk̲ h̲ usraw II] of the sultan Azatines [Kaiḳobād I] a bad leader who
was born of a good one. For he took pleasure in drinking and licentiousness,
in strange and unnatural sexual intercourse, and was always in the company
of creatures who no longer knew reason or indeed anything of human nature.
His father was not this sort, although he did give way to licentiousness, but
not very much.491
As it was seen, the author accuses also Kaiḳobād I of being involved in licentious
sexual acts, however in a lesser level. Ruth Macrides, who prepared the present
edition of work, states that the Byzantine representation of the Muslim rulers as
persons who engage in the promiscuous sexual behavior is a literary cliché. Keeping
this explanation in mind, I shall explore other similar depictions below.
In the 14th century, the issue of sodomy becomes the subject of even
theological polemics. Particularly the mid-14th century is a period of ideological
turmoil in the declining Byzantine world. The appearance of the movement of
Hesychasm and the reactions against it, the Byzantine Civil Wars and the pro-Turk
and pro-Latin fragmentation of the Byzantine society clearly demonstrate a deep
ideological crisis. Nikephoros Gregoras, who was a staunch supporter of the pro-
Latin, unionist faction, accuses his rival Gregory Palamas for being involved in
sodomy with Turks when he had fallen captive to them. Despite the hesychast
theologian who had fallen captive in 1354, after the zenith of the Hesychasm
490 Khoniates narrates the relationships of Manuel I with many different partners on p. 32, Andronikos
I’s relationship with his niece Eudoxia on pp. 59 and 80, Ivanko’s adultery with the Bulgarian princess
on p. 257. For a general narrative of sex in Byzantium see Laiou, “Le désir, l’amour et la folie: Les
rapports sexuels vus par les Byzantins,” 67–89.
491 Akropolites-Heisenberg, Georgii Acropolitae Opera, 69; Akropolites-Macrides, The History, 220.
253
controversy, the inquietude in the society was still there. As it was expressed by
Charis Messis, accusations of being a sodomite generally targeted the members of
the pro-Turkish faction in late Byzantine society. In that context, the accusations of
alleged sodomy were instrumental to denigrate people who supported pro-Turkish or
anti-Latin groups. Indeed, the people who were denigrated with such arguments –
Nikephoros Blemmydes (1197-1272),492 Gregory Palamas, and Loukas Notaras
(1402-1453) – were all anti-unionists. Hence, being a victim of sodomy appears to be
a punishment for showing a pro-Turkish attitude.
Gregoras, who describes Palamas as an advocate of impiety, narrates what
happened to his opponent in an obscene manner: Palamas is captured by pirates and
taken to the eldest son of the satrap Hyrcanos (Ork̲ h̲ an). There he is mocked, stripped
of his clothes, flogged and raped. It is hard to understand whether Gregoras’ narrative
about Palamas was entirely his invention or whether it was built on a rumor in
Constantinople.493
Palamas does not confirm such an incident, but in his narrative states that:
“These ungodly, infamous people who hate from the God (…) who live by the arrow,
sword and debauchery, find pleasure in making slaves, enjoy murdering, pillaging,
plundering, lust, adultery and love against nature.”494
Another text of the mid 15th century, a satirical narrative written by John
Argyropoulos -who was also a unionist- mocks a certain man of Serres called
Katablattas, who came from a lower social origin and had a pro-Turkish attitude, also
had “the barbarian customs” which included homosexuality. This person was
492 The anti-unionist theologian was accused of homosexuality during his residence at Smyrna. See
Messis, “Lorsque la périphérie assiège et conquiert le centre: certains aspects des relations entre
Byzantins et Turcs”, XIIe et XVe siècles,” 80.
493 Gregoras-Schopen-Bekker, Nicephori Gregorae Byzantina historia, vol. 3, 228-234; Gregoras-Van
Dieten, Rhomaïsche Geschichte, vol. 5, 175-178.
494 Palamas, La captivité de Palamas, 142-143.
254
represented as someone who corrupted the Thessalonican youth and became a judge
towards the end of the text.
However, Doukas is the author who stresses more the “perversion” of the
Turks. He describes at least three Ottoman sultans with homosexual tendencies and
attributes these acts to all Turks. In the passage where he mentions the marriage
between the daughter of John Kantakouzenos and Ork̲ h̲ an, he uses the generalized
discourse on the sexual behavior of Turks to explain Ork̲ h̲ an’s will to marry the
emperor’s daughter: “This nation is intemperate and lustful as no other people,
incontinent beyond all races and insatiate in licentiousness. It is so inflamed by
passion that it never ceases unscrupulously and dissolutely from having intercourse
by both natural and unnatural means with females, males and dumb animals.”495
He describes Bāyazīd I’s palace as a place wherein “boys and girls, selected
for their unblemished bodies and beauty of countenance, were there young and
tender youths, and girls outshone the sun” and the ruler’s daily life as “(he lived) idly
and wantonly, never ceased from lascivious sexual acts, indulging in licentious
behavior with boys and girls.”496
He narrates an even grimmer episode of Meḥmed II:
After the tyrant had traversed most of the City, he celebrated by holding a
banquet on the palace grounds. Full of wine and in a drunken stupor, he
summoned his chief eunuch and commanded him, “Go to the home of the
grand duke and tell him, ‘The ruler orders you to send your younger son to
the banquet.’” The youth was handsome and fourteen years old. When the
boy’s father heard this, his face turned ashen as though he had been struck
dead. He protested to the chief eunuch, “It is not our custom to hand over my
own child to be despoiled by him. It would be far better for me if the
executioner were sent to take my head.” The chief eunuch advised him to
surrender his child or otherwise the tyrant would be wrathful.497
495 Doukas-Grecu, Historia Turcobyzantina, 59; Doukas-Magoulias, Decline and Fall of Byzantium to
the Ottoman Turks, 73.
496 Doukas-Grecu, Historia Turcobyzantina, 87; Doukas-Magoulias, Decline and Fall of Byzantium to
the Ottoman Turks, 87-88.
497 Doukas-Grecu, Historia Turcobyzantina, 385; Doukas-Magoulias, Decline and Fall of Byzantium
to the Ottoman Turks, 234.
255
The episode ends with the execution of both the grand duke Notaras and his son.
Doukas also states that, after the execution of the members of Byzantine aristocracy
and high-ranking officials, “among their wives and children” the ruler selected
“beautiful maidens and handsome boys and entrusted them to the watchful care of
the chief eunuch.”498
The emphasis on the eunuch merits also some attention. Eunuchs were
present for centuries in the Byzantine palace. However, in Byzantine literature these
people also have a very negative image.
In Laonikos Khalkokondyles’ Histories, the same episode is presented
without its sexual implications:
When it was announced to the sultan that Notaras’s son was a child of twelve
years, he sent one of his wine pourers to request the child. When he heard the
wine pourer’s request, Notaras grew angry and considered it an insult, saying
‘Wine pourer, it is utterly outrageous for the sultan to remove my children
when he has nothing at present time for which to reproach us, given that he
has forgiven our offenses by ransoming us himself. If that is what he intends
to do with us, why does he not just order that we be delivered to a horrible
death?’ That is what Notaras said, and he said that he was himself blameless,
he would never willingly surrender his son.499
Although their narratives are quite similar and both authors agree on the dignity and
heroism of the late Loukas Notaras, there is a big difference between the tone of
narration and allusions in the text. In Doukas’ text the boy is fourteen years old and
in Khalkokondyles’ text he is twelve years old. Doukas underlines that he is
handsome; however, the latter gives no detail about his physical appearance. The
former openly alludes that the son of Notaras will go to Meḥmed II’s banquet and
“will be despoiled” by him; the latter leaves the purpose of the sultan’s request
498 Doukas-Grecu, Historia Turcobyzantina, 387; Doukas-Magoulias, Decline and Fall of Byzantium
to the Ottoman Turks, 235.
499 Khalkokondyles, The Histories, 204-205.
256
unclear. One can also think that Meḥmed II wanted to take the boy to recruit him as a
future bureaucrat, just like the officers of devshirme origin of the period. In the
former text, the person who came to take the boy is a eunuch, and was mentioned
before, the eunuchs’ bad reputation in Byzantine literature. On the other hand, in
Khalkokondyles’ text the person is a wine pourer, so despite being a servant to the
sultan, has an occupation without such a bad repetition as a eunuch. Finally, in
Khalkokondyles’ narrative, there is no allusion to the sultan’s tyranny or perversion.
Doukas and Khalkokondyles wrote roughly in the same period (mid-15th
century). The difference between the narratives could be explained by the context in
which they were written. Doukas was a secretary under the Gattilusio family, the
Italian rulers of the island of Lesbos, and was a member of the pro-Latin party in the
post-Byzantine Greek world. He wrote his work in the Genoese dominions of the
Aegean Archipelago. As for Khalkokondyles, he lived in the Duchy of Athens and
later in the Ottoman Empire, and even though he cannot be considered a pro-Turkish
author, he is not anti-Turkish either. But even in his text there was a reference to
Meḥmed II’s homosexual tendencies:
The sultan spent that winter in his palace and summoned Vlad, the son of
Dracul and the ruler of Wallachia, as he already had his younger brother at the
court, keeping him as his lover and maintaining him. It happened that the
sultan was almost killed by the boy when he had wanted to have sex with
him. This was when he had first gained the throne and was preparing to
campaign against Karaman. He was in love with the boy and invited him for
conversation, and then as a sign of his respect he invited him for drinks to his
bedchamber. The boy did not expect to suffer such a thing from the sultan,
and when he saw the sultan approaching him with that intention, he fought
him off and refused to consent to intercourse with him. The sultan kissed the
unwilling boy, who drew a dagger and struck the sultan on his thigh. He then
fled in whatever direction he could find. The doctors were able to thread the
sultan’s wound. The boy had climbed up a tree there and was hiding. When
the sultan packed up and left, the boy came down from the tree, began his
journey, and shortly afterward, and arrived at the Porte and became the
sultan’s lover. The sultan was used to having relations no less with men who
shared his own inclinations. For he was always spending his time in close
company of such people, both day and night, but he did not usually have
257
relations with men who were not his own race, except for brief periods of
time.500
This narrative is worthy of attention, because Khalkokondyles, despite his lack of
anti-Ottoman fervor, still represents Meḥmed II as a homosexual. Yet his narrative is
still different from Doukas’ narrative of Turks because these events are presented not
with a propagandist tone, but rather a calmer style. He also gives details about the
events between Meḥmed II and Vlad III of Wallachia, as an explanation of the later
Ottoman-Wallachian conflict. The final remarks also merit attention because the
explanation of the sexual life of Meḥmed II is not written in a moral tone, nor for
vilifying him. However, it should not be forgotten that these texts cannot be accepted
as historical reality, and that Byzantine historiography is based on representation.
It could be assumed that these narratives build a dichotomy between the
Christian Greeks and Muslims, considering that Christian values such as chastity and
virtue are the norms, and the barbarians have the tendency to be perverts.
Furthermore, it can be concluded that there is a correlation between martial violence
and sexual violence since both seem to be the expression of uncontrollable
masculinity and untamed barbarism according to the Byzantines. Therefore the
Turkish invasion does not only target the territory of the Byzantines but also poses a
threat to their bodies. At this point, I propose a conceptual discussion regarding the
performance of sexuality.
These references, which I call homosexual hypersexuality, could be
interpreted as “Persian” traits, rather than “Skythian”. As was already seen in the
earlier texts of Skylitzes and Attaleiates, extreme cruelty and sadistic acts were
500 Khalkokondyles, The Histories, 366-369.
258
generally presented as traits of the Skythians. To make a comparison, Doukas’
description of Tamerlane should also be taken into account. He records Tamerlane’s
massacres and tortures in a cold-blooded way, but there is nothing sexual about these
descriptions. This fact is understandable according to the cultural model he fulfills.
However, I utilize the formula of “Persian” and “Skythian” traits as a conceptual
model to understand these cultural elements. Here, these terms are not related with
the use of the ethnonyms. Both Doukas and Khalkokondyles refer to the Ottomans as
Turks.
Charis Messis interprets the Byzantine representation of the Turks as a reuse
of the themes used in the past to identify the Arabs.501 The Arabs and Muslims were
already accused of sodomy and pederasty according to a long tradition of narratives
that also reflects the characteristics of anti-Muslim polemics. A source of the 10th
century, John Kaminiates’ narrative of the sacking of Thessaloniki by Arab pirates in
904, clearly reflects how the Byzantines perceived the sexual behavior of the Arabs:
What must they all have felt in such a situation, when they were being led off
to slavery in a foreign land, where the worship of our faith is treated as an
abomination and the most senseless passions are revered, where whoredom is
held in high repute, where madness is honoured and shamelessness prized,
where males are made to play the part of females and creation is violated, and
everything is topsy-turvy, confused, distorted and directed towards evil?502
So this accusation of sodomy has its roots not in the Turkish lifestyle, but in the
Byzantine imagination of the Arabo-Islamic world. This literary tradition goes back
to the 8th century, to the writings of pseudo-Stephan of Alexandria, and identifies the
Arabs with “uncontrollable masculinity.”503
501 Messis cites a letter written in Latin by Alexios I Komnenos to count Robert of Flanders, in which
the Turks are accused of sodomy and sexual aggression against all the elements of Byzantine society,
including the monks. However, the author himself states that this letter is of dubious authenticity. For
the citations from this letter, see Messis, “Lorsque la périphérie assiège,” 78.
502 Kaminiates, The Capture of Thessaloniki, 121.
503 This expression is a formulation of Messis. Similarly Christian spaces such as monasteries and
259
The Byzantine authors construct the identity of the Byzantine man, homo
byzantinus, based on sexual normativity and exclude the Turks. Finally, one must
never overlook the existence of different gender relations between the Greeks and
Turks. The Turks, particularly after their conversion to Islam, adopted a certain
Islamic code of gender relations. In the medieval Islamic world, where gender
segregation was a very common element of social life, homosexual relationships
between men were frequent. There is abundant literary evidence about this fact. It
seems that these relationships were more tolerated in the Persianate cultural sphere
which included the Seljuks. In the 11th century, just before the rise of the Seljuks, the
relationship between Maḥmūd of Ghazna and his slave Ayaz was a well-known
episode of Persian literature. A Persian Mirror of Princes of the 11th century, the
Kabusname, written by Kaykāʾūs B. Iskandar, suggests that having homosexual
relationships with ghulams are legitimate acts for Muslim statesmen.504
Although Islamic law punishes homosexual relationships, called liwāṭ in
Arabic, meaning sodomy,505 it seems that these relationships were frequently seen
and tolerated in the medieval Islamic world. The fact that such relationships were
particularly frequent among the élite circles of the society in Medieval Islam could
be compared with Classical Greece. It can be safely said that homosexual
relationships were probably less condemned in Muslim Asia Minor, than in the
Byzantine territories.
nunneries were imagined as places of pleasure and licentiousness by medieval Arabic poets. See
Wood, “Christians in the Middle East 600-1000: Conquest, Competition and Conversion,” 23-51.
504 The 15th century Anatolian Turkish translation of this book is considered as one of the first
important works of Turkish prose. Keykavus, Kabusname, tr. Mercümek Ahmed, ed. Orhan Şaik
Gökyay, 96.
505 “Liwāṭ”, Encyclopaedia of Islam (2nd ed.), vol. 5, 776-779. The authors conclude the article by
stating: “It is indeed difficult to measure precisely the extent of the phenomenon, but it should be
recognized that the separation of the sexes, which is a particular feature of Islam, has played a
significant role in promoting it.”
260
Nevertheless, one cannot equate the homosexual relationships attested among
the elite circles of medieval Muslim Anatolia with the alleged sexual aggression of
Turks, depicted by the authors mentioned above. Yet the contemporary differences in
gender relations between the Byzantine and Muslim societies of the era could have
made the attribution of the behavior identified with the 14th and 15th century Turks
easier. However, even accepting this difference in gender relations, these accounts of
private life could not be read as objective narratives of historical facts. Because as
was stated before, the Byzantine historiography is totally based on representation and
has a very large fictional component. So these scenes should be perceived mostly as
the reflections of several aspects associated with Muslim Arabs on the Turks.
In conclusion, the formation of the narrative of aggressive sexuality could be
summarized with the following strata:
i. There is an ancient narrative and belief regarding the Arabs and Islam
based on the idea that Islam is interrelated with the cult of Aphrodite/Astarte,
and the veneration of this cult is related to their unsatisfied lust and aggressive
sexual behavior. It can be said that this cliché was already existing before the
Byzantine-Turkish encounter in the 11th century and it had nothing to do with the
Turks.
ii. This theme was rarely used for the Turks in the beginning, because
particularly in the Byzantine literature of the 11th and 12th centuries regarding the
Turks, there is no clear identification between the Turks and the Arabo-Islamic
civilization. However, in the process of identification of Turks with the cultural
attitudes the Byzantines related to Islam, such narratives started to appear.
iii. Finally, in the 14th and 15th centuries, the increase of narratives
regarding sodomy could somewhat reflect a crisis of masculinity triggered by the
261
invasion. However, this scandalizing theme was instrumentalized as a part of the
polemics by the Byzantine authors for the unionist cause. Hence, it may rather
be a literary motive than an everyday phenomenon. Another simple but thoughtprovoking
question is how an author like Doukas, who lived far from the
Ottoman capital, could obtain objective information on such a private matter as
the sultan's sexual life. One can think that this information source could be the
Greek network around the palace. But this time, the problem arises of
distinguishing between true or false rumors and literary models. This is an
interesting but probably unsolvable question.
7.4 Wild customs of the freaks: On the way to being dehumanized
As seen above, some Byzantine narratives about sexual violence almost dehumanize
the Turkic peoples, whom they consider barbarians. Other narrative elements of
dehumanization could be further found in the Byzantine literature of our timeframe.
In ancient Greek ethnography, uncanny realms are always associated with monsters,
and this tradition also influenced Byzantine writers. For example the κυνοκέφαλοι of
Pakhymeres was already mentioned by Herodotus, Ktesias, and Megasthenes.506
The individual persons of Turkic stock were not considered beautiful in the
Byzantine sources: moreover, they were considered particularly ugly, even as freaks:
Attaleiates describes Chrysoskoulos, Alp Arslan’s brother-in-law who escaped to
Constantinople, as “young, but almost a pygmy in height and his face was that of a
Skythian and ugly because this people are of Skythian ancestry and have inherited
their depravity and deformity.”507 Anna Komnene describes, still more strikingly, a
506 For monsters in Ancient Greek literature see Mitchell, Monsters in Greek literature: Aberrant
Bodies in Ancient Greek Cosmogony, Ethnography, and Biology.
507 Attaleiates, The Histories, 258-259.
262
Pecheneg warrior who captured the cousin of Bohemond II of Antiochia: “It was
indeed an extraordinary sight - this huge giant, a really monstrous man, the prisoner
of a pygmy of a Scythian. (...) In came the Scythian leading this tremendous Kelt on
a chain, barely as tall as his waist. Of course, there was an instant outburst of
laughter from all. The rest of the counts were committed to prison.”508 In Khoniates’
account, the ugliness and bad-looking body of sultan Qiliç Arslān II was remarked:
“(he) was not a physically well-proportioned man but maimed in several of the vital
parts of his body. His hands were dislocated at the joints, and he had a slight limp
and traveled mostly in a litter.” Andronikos Komnenos calls the sultan “Koutz-
Arslan” (Güç-arslan?) because of his defects. Khoniates still underlines the sultan’s
energy and cunning, despite his physical problems.509
It seems that the Byzantine misrepresentation of the physical appearances of
Turks could be explained by a mixture of prejudice and their odd feeling about the
Asiatic appearance of the Turkic peoples. This theme is also not new, even in
Jordanes’ description of Attila (which was based on the narrative of Priscus) there are
the same features: “He was short of stature, with a broad chest and a large head; his
eyes were small, his beard thin and sprinkled with gray and he had a flat nose and a
swarthy complexion, showing the evidence of his origin.”510 So the phantasm of the
Asiatic barbarian is an existing topos in the Byzantine literary culture.
Another motif related to the dehumanization of the Turks has also been
pointed out by Oikonomides, Papageorgiou, and Shliakhtin. Turks are often
associated with wild animals, especially wolves. The depiction of a group as being
508 Komnene-Reinsch-Kambylis, Annae Comnenae Alexias, 402; Komnene-Sewter-Frankopan, The
Alexiad, 374.
509 Khoniates-Van Dieten, Nicetae Khoniatae Historia, 122; Khoniates-Magoulias, O City of
Byzantium, 69.
510 Jordanes, Gothic History, 102.
263
identical to an animal, especially a wild animal, is an indication that they are not
considered human. The wolf figure also has a biblical meaning: wild wolves against
the flock of Jesus. However, this parallelism may be a reference to the wolf totem
among the ancient Turks, as emphasized by Papageorgiou, by the Byzantine authors,
and may indicate a specific war cry of the Turks, as stated by Shliakhtin. Moreover,
these two elements are most likely related: that is, the war cry of the Turks is a
remnant of an ancient pagan ritual.511
The depiction of extreme cruelty completes this image. This feature is mostly
the reflection of a Skythian trait. According to Attaleiates, the Pechenegs “are
loathsome in their diet and the other aspects of their life, and do not abstain from
eating foul foods.”512 Furthermore, on the battlefield they commit sadistic acts: when
they captured Michael Dokeianos, the commander of the Byzantine army in the
campaign against them, they cut him in pieces, slit open his belly, pulled out his guts
and replaced them with his hands and feet.513 As it was already seen, after the battle
of Myriokephalon, the Seljuk warriors cut the penis of the fallen Byzantine soldiers.
Pakhymeres reports that, before the start of diplomatic relations with the Mongols,
the Byzantines thought that they were cannibals or dog-headed (κυνοκέφαλοι).514
At this point, I shall revisit an important article by Speros Vryonis Jr. which
deals with the possible evidence for human sacrifice among the early Ottoman Turks.
In his article, he analyzes several passages in Contra Mohametem Apologia of John
Kantakouzenos and in the histories of Doukas and Khalkokondyles, arguing that in
511 See Oikonomides, “The Turks in the Byzantine Rhetoric of the Twelfth Century,” 150;
Papageorgiou, “οί δέ λύκοι ώς Πέρσάι: The Image of the “Turks” in the Reign of John II Komnenos
(1118-1143),” 150-152; Shliakhtin, From Huns into Persians: The Projected Identity of the Turks in
the Byzantine Rhetoric of Eleventh and Twelfth Centuries, 192-196.
512 Attaleiates, The History, 52-53.
513 Attaleiates, The History, 60-61.
514 Pachymeres, Relations historiques, vol. 1, 186-187.
264
the early Ottoman period, there was a tradition of human sacrifice to honor the dead.
Khalkokondyles states that after a battle in Isthmus Murād II buys about six hundred
slaves and sacrifices them to his father performing as an act of piety.515 Kaldellis
states that this passage is problematic but not impossible.516 However, in the account
of Khalkokondyles, there is another reference to human sacrifice that Vryonis
overlooked. When the author deals with the Khataians who seem to be the
northeastern neighbors of Tamerlane, he argues that they sacrifice every year the
children who reached the age of puberty to honor Artemis.517
As I already stated in the introduction, I am not interested in whether the data
in these historiographical works are empirically true or false. Human sacrifice was a
practice once common in different parts of the world. Since the time of Herodotus,
the Skythians and other steppe peoples associated with the human sacrifice.518 The
Turks in Anatolia, perhaps, practiced the custom of human sacrifice in their new
homeland, even though centuries had passed since their conversion to Islam. I do not
consider myself competent to speculate on this. I am interested in the Byzantine
mentality behind it. In my opinion, these passages demonstrate that there is a
dehumanizing discourse against the barbarian populations. The discourse about the
sexual aggressiveness and physical depictions as the freaks could not be imagined
apart of this. However, the arguments like human sacrifice or cannibalism are the
ultimate arguments to antagonize and dehumanize an alien population.
7.5 Conclusion
515 Vryonis Jr, “Evidence of Human Sacrifice Among the Early Ottoman Turks,” 145.
516 Khalkokondyles, The Histories, 496.
517 Khalkokondyles, The Histories, 268-269.
518 Vryonis Jr, “Evidence of Human Sacrifice Among the Early Ottomans Turks,” 141.
265
In this chapter, four features of the representation of the Turks in Byzantine
historiographical narratives were discussed: warfare, status of women, sexuality, and
cruelty. I have investigated these four elements as features that draw a distinction
between the two populations.
The representation of warfare could be both realistic and full of clichés.
Indeed, there is an ancient tradition of military writing, centering on warfare
techniques. However, battlefields were also the places where an encounter between
the Byzantines and foreigners happened. Some of the authors witnessed the battles
against Turkic invaders, yet the ancient material was also widely used in the warfare
narratives about the nomadic peoples.
As we have seen, Turkish women are nearly invisible in Byzantine
historiography. The representation of sexuality in Turks has a similar paradoxical
point as well. Several Byzantine sources represent both the Turkish commoners and
the Ottoman rulers as pederasts. On one hand, this representation is an element of the
polemics between the pro-Turkish and anti-Turkish factions within Byzantine
society. Nevertheless, the Turks who had a much different culture and lifestyle were
probably perceived as perverts by some Byzantines. It is also true that the
abovementioned way of representation has its roots in the traditional representation
of the Arabs noticed in Byzantine texts. So it can be said that the Byzantines
reflected their impressions and ideas about the Arabs on the Turks to some extent.
It can be furthermore said that the literature about Turkish warfare is the most
realistic part of the Byzantine accounts on the Turks, yet the accounts about sexuality
are probably less realistic. This may be because the former can be considered a
reflection of what was actually encountered in the battlefield, and the latter is mostly
the reflection of a literary topos.
266
In conclusion, the Byzantine representation of the Turks is an amalgamation
of testimony and fiction. It is very difficult to distinguish the so-called historical
reality from fiction. The unique way to find a solution is to read these texts according
to the alleged aims of the authors.
267
CHAPTER 8
CONCLUSION
This chapter will conclude our dissertation on the formation of the image of Turkic
peoples in Byzantine historiography. My study had several limits, I omitted all nonhistoriographical
works such as sermons, hagiographies, panegyrics, romances, etc.,
and limited my corpus to historiographical works, with only one exception – the
narrative of the captivity of Gregory Palamas. I omitted the historiographical works
written in the periphery of the Byzantine world, such as the works of Panaretos or the
Chronicle of the Morea. Under ideal conditions, a research that aims to deal
systematically with the representation of the Turkic peoples in Byzantine literature
must contain also these materials.
Also, the time frame of the dissertation spans four centuries. However, the
texts of my corpus did not contain these periods in equal detail. For example, the first
part of the 13th century was only covered by the work of Akropolites, which, despite
its rich material on the Seljuks of Rum, is not a very comprehensive text. The late
14th century poses the same problem as well. So the quality of the sources can
determine the outcome of the sources. Furthermore, the passages dealing with the
“northern” Turks are often dry and without details, in comparison to the materials on
the Seljuks and the Ottomans.
My first important finding is that forming a collective image of a people or an
ethnic group is closely related to cultural memory. The difference between the
images of Northern (i.e., Skythian) Turks and Eastern (i.e., Persian) Turks was the
result of their geographical location, which was the basis of the Byzantine
nomenclature of foreign peoples, and of the Aristotelian viewpoint that correlates the
268
geographical region in which people live with their national character. However,
these two points are meaningful only with the presence of cultural memory.
Byzantine literature had no direct relation to anything that could be called
“objective reality.” The basic concept of Byzantine historiography was
representation. Hence, Byzantine historiography can only be understood through this
notion. Byzantine authors followed Greek and Roman models and nourished an
already created image of the Skythians and Persians. This image was a logos and was
not an image that was fulfilled by an ordinary Byzantine man, but it had continuity
and an ideological function. This ideological function was the logos’ role in alterity.
In the Byzantine worldview, Northern and Eastern Turkic peoples were presented as
the “others” to the Byzantines.
However, real-life often does not follow the literary topoi. Beginning with
Michael Attaleiates, Byzantine authors encountered Turkic individuals with different
backgrounds. The authors from the frontier region, such as Niketas Khoniates or
Doukas, experienced the agony of losing their ancestral lands. A great theologian
such as Gregory Palamas experienced life in the Ottoman court as a captive.
Thus, Byzantine authors wrote their personal experiences with the Turks.
These narratives were neither without an ideological aim nor were they independent
of the earlier literature about Turkic peoples. It can be said that all the Byzantine
literature regarding Turkic peoples was a dialogue with the ancient masters. The
authors who contributed to this literature were not hommes de lettres who made a
living by writing; they were bureaucrats or men of politics in its broader sense. Their
political aims and views inevitably affected their approach to the Turks. Within the
last centuries of Byzantium (c. 1350-1453), the essential axis of Byzantine politics
consisted of taking position between the Turks and the Westerners.
269
A key variable in the representation of the Turks in Byzantine historiography
is the role of Islam. Between the 11th and 15th centuries, the weight of the Islamic
religion visibly increased in the Byzantine texts, and it shaped the image of the
Turks, coinciding with the association of the Turks with the Persian people.
Moreover, Islam assembled the contemporary representation of the Turks and the
cultural memory of the Arab invasions. The “Skythian” nature of the Turks was
forgotten by some, but was eventually brought to light by a skillful ethnographer,
Laonikos Khalkokondyles.
After the fall of Constantinople and the destruction of the empire, the
ordinary Byzantines continued their lives as the Rum milleti, the Christian subjects of
the Ottoman Empire. The empire was at an end, and the secular institutions were
destroyed. Yet, the Ecumenical Patriarchate of Constantinople enabled them to live in
a certain sense of unity as a religious community.
Therefore, I can speak of some essential strata in the formation of the image
of the Turks in Byzantine texts:
i. The first stratum is the cultural memory of Skythians and Persians.
The ancient representation of both nations played a significant role in the
formation of the image of the Turks. The genealogy of this representation goes
back to the times of Herodotus. The images formed around these two signifiers
represented two different levels of antagonism for the Byzantines. The Persians
in the east represented the primary antagonism, and the Skythians in the north
represented the secondary antagonism.
ii. The second stratum is the cultural memory of the Arabs and the early
Muslim invasions. The associations of the Turks with Agarenes and Ismaelites
are
270
iii. typical manifestations of the use of this cultural memory for Turks.
The traces of this motive were seen firstly in the Alexiad of Anna Komnene.
iv. The third stratum consists of the remnants of the cultural memories of
the Byzantine encounters with the “Northern Turks” in the regions of the
Danube and possibly Crimea. This cultural memory should be considered as a
continuation of the reminiscences of the Huns of Late Antiquity. Since the time
of Constantine Porphyrogennetos, some knowledge about these populations
existed, and the 11th-century encounters with the Pechenegs added new
information to this stratum based on the experiences of the field.
v. The fourth stratum is the Byzantine literature regarding the Seljuk
Turks, namely the works of Anna Komnene, John Kinnamos, Niketas Khoniates,
and George Akropolites. The literary works produced during the period provide
us with details not only on the history of Seljuk-Byzantine relations but also on
the internal issues of Seljuks and other Turkish entities of Asia Minor. The
heritage of the works of this period is also formative regarding the image of the
Turks in Byzantine literature.
vi. The fifth stratum is the corpus regarding the Ottomans. This corpus
starts with the work of George Pakhymeres, in which can be seen a nucleus of
the Ottoman state, although it was yet another one of the Turkish warbands in
Bithynia, and the phenomenon of “the rise of Ottomans” had not yet taken place.
Authors such as Pakhymeres and Khalkokondyles represented the last
generations of Byzantine historiography, and their representations of the
Ottomans were based on personal experience and often biased because of the
political aims of their authorship. The Turks were not newcomers anymore; since
271
the 11th century, there had been a Turco-Greek coexistence in Asia Minor. In
addition to the biased attitude of the abovementioned authors, their
representation of the Ottomans had traces of earlier strata.
Focusing on these findings about the process of the formation of the image of
the Turks, some essential patterns can be observed:
The association of the Turks with Persian traits has several causes.
Geographically, the Great Seljuk Sultanate was established in Iran and Khorasan.
The Seljuk Turks, thus, took over the historical memory of the Persian Empire
because of their geographical location, without being ethnically Persian. The rise of
the Seljuks, though, they primarily defeated and annexed the Persian states of the
“Iranian intermezzo”, ultimately the Abbasids and the Syrian and Iraqi Arab states of
the 11th century; so historically the Byzantine authors could have been indicating the
rise of the Seljuks as the revival not of the ethnic Persians, but of the “Empire of
Persia” over the Arabs. Particularly, the narrative of Skylitzes on the rise of the
Seljuks and Komnene’s stress on the “sultanate of Khorasan” demonstrate this idea.
The Persianization of Turks meant becoming antagonists of the Byzantine
Empire which has two dimensions to underline; the first is the Turkish invasions and
crusades after the rise of Islam, particularly after the 11th century. There was a
growing sentiment over a religious antagonism between Christianity and Islam, and
the religious identity of the Turks in Asia Minor became more visible to the
Byzantines. The increased awareness caused the emergence of ideas such as the
Komnenian project concerning the evangelization of the Turks. The association of
the Turks with notions of the Ismaelitai and Agarenoi became frequent, and the
difference between the Skythian and Persian Turks was now a matter of discussion.
When the Turks began to be identified with the Persians, they were burdened with
272
the topoi associated with the Persians, such as sexual perversions.
It should also be added that there are always Skythians somewhere. If a group
of people loses the socio-cultural aspects that make it “Skythian”, a new population
could be the new Skythians. The steppe region which is associated with these peoples
is a very turbulent and culturally heterogeneous zone. Every time a new group of
nomads came from some obscure part of innerAsia, they became the new Skythians,
as it was seen with the Pechenegs, Cumans, and Mongols.
The historiography of the 14th century demonstrates the Byzantines’ bitter
agony of accepting the dominion of the Ottomans and other Turkish emirates. The
texts written in this context may be considered narratives that demonstrate their
authors’ goal more directly, as in the case of Doukas and Khalkokondyles and,
additionally, Kritoboulos and Sphrantzes, who focused on the 15th century. The
Byzantine representation of the Turks as “the others” transformed into a narrative of
self-victimization (in Doukas) and accepting the domination of Turks (in
Khalkokondyles).
Finally, the frontier ethos and the martial culture of the Byzantines vanished,
and the military ethos of the society gradually decreased after the 13th century. A
frontier ethos has a function only in a society where there is a peaceful space in the
center and a militarily active zone in the frontier. By the late 13th century, nearly all
Byzantine territory became the target of the Turks and other nations, and the
Byzantine Empire based itself more and more on mercenaries, until eventually there
was no difference between the frontier and the center.
273
APPENDIX
LE RÉSUMÉ DE THÈSE
INTRODUCTION
Cette thèse a pour sujet la construction et l’évolution de l’image des divers
peuples turciques dans la littérature historiographique byzantine. Les auteurs du XIe
et XIVe siècle sont étudiés dans cette thèse doctorale. Cette recherche ne s’agit pas
d’une étude du corpus de sources chronologiques, cependant elle constitue une
enquête axée plutôt sur l’anthropologie, qui se concentre sur les différents thèmes et
motifs qu’on retrouve dans les oeuvres du corpus.
Cette thèse est composée de huit chapitres. Le premier chapitre est une tentative
d’introduction, comprenant une revue détaillée de littérature et une discussion de la
méthodologie. Le deuxième chapitre traite du contexte historique. Le troisième
chapitre est consacré à la notion de « barbare » et de son usage dans l’historiographie
byzantine. Le quatrième chapitre étudie les narratives sur l’origine des Turcs dans la
littérature Byzantine. Dans le cinquième chapitre, on a étudié les modalités de
coexistence des Turcs dans la société byzantine, et donc les thèmes comme
l’assimilation, l’acculturation et l’antagonisme. Le sixième chapitre discute la place
des états barbares dans le point de vue byzantin tandis que le septième chapitre
discute divers aspects individuels de la représentation des Turcs. Finalement, le
dernier chapitre propose une conclusion à notre recherche.
Quand on analyse les quatre siècles étudiés lors de cette recherche (du XIe au
XIVe) on peut observer une défense contre la pression des différents pouvoirs par ses
frontières. Les peuples turciques font partie éléments des plus importants qui ont
274
contribué à ce fait. Au milieu du XIe siècle, différentes populations turques
apparaissent, presque simultanément, aux frontières impériales. Les tribus nomades
des Petchenègues et Uzes sont apparus dans les confins nord-ouest de l’empire, plus
précisément dans le bassin de Danube et les Balkans du nord. Dans le même temps,
les Seldjoukides ont pris le pouvoir aux Perses. Les Seldjoukides appartiennent à la
ligne des Oghuz : Ils sont conquis l’Asie Mineure à partir du dernier quart du XIe
siècle et les Seldjoukides de Rum, une branche cadette de la dynastie des
Seldjoukides de Perse, ont constitué un sultanat en Anatolie. Donc, à partir de ce
point, les conflits et la coexistence avec les Seldjoukides sont devenus les sujets les
plus courants dans l’historiographie Byzantine. Au XIIe siècle, les Coumans
s’emparent des terres autrefois dominées par les Petchenègues et remplacent ce
peuple. Après l’invasion mongole au XIIIe siècle, le sultanat Seldjoukide de Rum
devient un état marionnette sous la suzeraineté des Mongols de l’Iran ; et les
différents groupes Turcomans forment les émirats, dits beyliks, aux frontières
occidentaux du sultanat. Donc, l’émirat Ottoman était un de ces petits émirats, qui à
partir du milieu de XIVe siècle a réussi à s’étendre vers les territoires Byzantines en
Europe.
Dans cette thèse, on utilise le mot “turcique” comme un terme générique pour
toutes les populations contemporaines et historiques parlant des langues turques.
Dans la littérature, les langues turques sont divisés en deux catégories essentielles : le
turc commun et le turc oghour. Le seul représentant vivant des langues turques
oghours est le tchouvache, parlé dans la République Tchouvache de la Fédération de
Russie, dans la région de la Volga. Toutes les autres langues turciques vivantes font
partie du groupe turc commun. Historiquement, la langue ancienne bulgare est
considérée comme une langue turque de la branche oghour. Petchenègue et Couman,
275
et très probablement Khazar sont considérés comme des langues turques communes.
Cependant, plusieurs populations turques historiques de l’Eurasie occidentale
pourraient être des locuteurs du turc Oghuric, de sorte que le sprachbund oghour
pourrait être gaspillé au début du Moyen Âge.
Le mot « Türk » apparaît comme un terme qui désigne la population fondatrice de
Khaganat Türk, sous la dynastie Ashina, centrée sur la Mongolie contemporaine, au
VIe siècle. Ce khaganat était un élément important de la politique eurasienne
jusqu’au VIIIe siècle, malgré leurs guerres civiles et les tentatives d’assujettissement
de la Chine sous la dynastie Tang. Khaganat Türk n’était pas un État ethniquement
homogène et tout comme les formations politiques similaires de la région des
steppes, c’était une fédération nomade qui contenait des nombreuses populations
mongoles, indo-européennes et ouraliennes. La langue soghdienne, qui était autrefois
une langue très courante dans les agglomérations urbaines de la région des steppes
d’Asie centrale, qui appartient à la branche orientale de la famille des langues
iraniennes, était utilisée dans la chancellerie dans le Khaganat Türk, avec l’ancien
turc.
L’ethnonyme “Turc” s’est répandu avec la montée de l’Empire Turc dans la
région des steppes comme terme commun à toutes les populations nomades. Les
géographes musulmans en particulier ont utilisé ce terme avec persistance pour de
telles populations, parfois même pour des peuples non turcs comme les Varègues et
les Rus. Selon Golden, cet usage est similaire à l’usage de l’ethnonyme scythe par les
écrivains byzantins en tant que catégorie générique.
Dans cette thèse on utilisera le mot turc uniquement pour désigner la population
turcophone du sultanat de Rum et du futur sultanat ottoman, ainsi que les Turcs
individuels d’Asie Mineure et des Balkans. Les habitants turcophones de ces régions
276
avaient subi un certain processus d’acculturation et d’ethnogenèse qui les distinguait
des autres populations turques de la région. Ces Turcs pourraient être considérés
comme des “Turcs de Rum” – en utilisant le terme géographique qui s’applique luimême
à la région de l’Asie Mineure.
Les historiens qui sont étudiés dans ce travail sont: Jean Skylitzès, Michel
Attaleiatès, Anna Comnéne, Jean Cinnamus, Nicetas Choniatés, Georges Acropolite,
Georges Pachymère, Nicéphore Gregoras, Jean Cantacuzène, Grégoire Palamas,
Doukas, Laonicos Chalcondyle, Critobule d’Imbros et Georges Sphrantzès. Ces
auteurs représentant presque tout le corpus historiographique byzantin, du dernier
quartier de XIe siècle jusqu’à milieu du XVe siècle. Ces auteurs sont classifiés dans
la thèse suivante selon leur provenance, métier et tendances politiques.
Selon Paolo Odorico, la littérature Byzantine a un fort caractère utilitaire, donc
chaque texte doit être lu en regardant l’objectif de l’auteur et l’objectif du mécène
qui soutient la composition de l’oeuvre et l’audience. Ainsi, toute oeuvre
historiographique byzantine doit être pris en compte dans son contexte. La
représentation est aussi l’un des concepts les plus centraux de cette thèse. La
littérature Byzantine ne peut être lue comme un narrative objective d’une réalité, la
fiction a toujours une place importante dans ce genre. En historiographie byzantine,
toute la narrative est construite autour des représentations et parfois, cette narrative
ne reflète pas la réalité. L’altérité est une notion philosophique qui décrit la relation
entre soi et l’autre. La différence et la diversité son deux conceptions pour définir le
degré d’altérité. Donc, la dichotomie entre la différence et diversité est aussi
importante pour nos conceptualisations. La différence de deux objets signifie qu’il y
a quelque chose de commun entre les deux, mais diversité indiques qu’ils sont divers,
aussi dans un contexte ontologique.
277
Les Scythes et les Perses sont deux différentes catégories des barbares employées
dans la littérature grecque, depuis l’antiquité. L’image de “Scythe” représente les
populations nomades qui vivaient dans le nord du monde gréco-romain. Les Scythes
historiques étaient une population indo-européenne, mais l’ethnonyme scythe est
employé systématiquement pour les populations turco-mongoles d’Eurasie dans la
littérature Byzantine. D’autre part, l’ethnonyme perse est aussi employé
originairement pour les Perses Achéménides et Sassanides. Mais après la chute de
l’Empire Perse, les écrivains byzantins ont continué à utiliser « perse » pour désigner
les habitants de la région dans le Moyen Âge tardif ; ce mot (perse) est aussi
fréquemment utilisé pour les Turcs d’Orient, c’est-à-dire les Seldjoukides. Il est
employé plus rarement pour les Ottomans.
Les altérités “scythes” et “perses” étaient deux types différents d’altérité, et les
auteurs byzantins leur ont donné différents niveaux de signification idéologique :
ainsi, mon argument est que l’antagonisme entre les Byzantins et les Scythes (qui
forme un antagonisme secondaire) se construit surtout sur une différence entre la
civilisation et la barbarie, la sédentarité et le nomadisme. En revanche, l’antagonisme
byzantin-persan est le reflet d’un enjeu bien plus fondamental : c’est la continuité
d’une expérience qui a exercé une très forte influence sur les intellectuels grecs puis
sur tout le monde occidental, les guerres médiques (499 avant JC - 449 avant JC).
Comme l’a expliqué François Hartog : “Le Barbare, c’est avant tout, plus que tous et
pour longtemps le Perse.” L’image du Turc dans la littérature byzantine est de
manière antagoniste depuis presque le début, car les Turcs sont finalement traités
comme des ennemis sur le champ de bataille. Cependant, la nature et le degré de cet
antagonisme permettront de comprendre la position des Turcs dans l’historiographie
byzantine. De plus, il ne faut jamais oublier que cette différenciation dans
278
l’antagonisme pourrait également être liée à l’évolution politique et culturelle des
politiques turques. Si l’élément islamique dépassait les aspects attribués aux
« Scythes » dans l’idéologie de ces régimes politiques dans le processus historique,
et, par exemple, si une approche de la guerre à motivation religieuse basée sur un
antagonisme antichrétien gagnait en importance, cela conduirait à un changement
dans la perception byzantine des Turcs.
LE CONTEXTE HISTORIQUE
Le rival historique de l’Empire Romain et de son continuateur, l’Empire
Byzantine, était l’Empire Perse. L’Iran Sassanide et les Romains étaient les
adversaires principales jusqu’à la destruction de l’Empire Sassanide par les Arabes
Musulmans. La mémoire collective sur l’antagonisme avec les Perses, est un motif
très fort dans l’historiographie gréco-romaine. Mais la naissance de l’Islam a changé
dramatiquement le vieil ordre du Moyen-Orient. Après la guerre perso-byzantine de
602-628, tous les deux états ont épuisé leurs ressources, donc leurs territoires sont
devenus les cibles de l’expansion arabe. Les Arabes ont rapidement conquis tous
l’Empire Perse et les provinces Byzantines du sud-est (Syrie, Palestine, Égypte).
Donc le pouvoir arabe a remplacé les Sassanides aux yeux des Byzantins. La
formation d’un très grand espace de frontière entre les Byzantines et Arabes a causé
de la formation d’une région de guerre sporadique entre les deux pouvoirs. Cette fois
entre les Byzantines et Arabes un nouvel antagonisme avec une forte dimension
religieuse est formé.
Dans la région de frontière entre les Byzantines et Arabes qui s’appelait
« thugur » en Arabe, s’est formé une culture particulière et un ethos de frontière, qui
279
se sont préservés surtout dans les oeuvres littéraires dit les “narratives de frontière”
comme les épiques de Digenis Akritas et Sayyid Battal Ghazi. Ces narratives sont
remarquables dans leur description de la réalité quotidienne et de l’idéologie de
guerre. Donc la formation d’une telle frontière bâtit aussi une géographie
idéologique, les villes de cette région (Tarse, Melitene, Sozopetra) ont une
importance symbolique pour les deux partis.
Les premières guerres arabo-byzantines ont entraîné la domination arabe
permanente et l’arabisation et l’islamisation subséquentes des provinces orientales de
l’Empire byzantin. Cependant, l’Asie Mineure a résisté à une telle invasion ; la
formation d’un espace frontière entre deux puissances a créé à la fois un ethos des
deux côtés et un fort antagonisme. La figure arabe (sarrasine/agaréne/ismaélite) s’est
substituée à l’archétype du persan comme antagoniste de l’empire byzantin voire de
l’identité gréco-romaine. Ainsi, cette période (jusqu’au IXe siècle) a été la période de
formation de l’ethos frontalier et de l’antagonisme. Mais dans Xe siècle, l’Empire
Byzantine sous la dynastie macédonienne a commencé une offensive vers cette
région de frontière. Les Abbassides qui étaient dans un processus de fragmentation,
étaient très faibles pour pouvoir résister à expansion Byzantine. Les régions
frontalières sont déjà laissées aux dynasties autonomes locales qui sont
progressivement annexées par les Byzantines.
Ce processus est continu jusqu’à moitié de XIe siècle. L’Empire Byzantine a aussi
annexé les principautés arméniennes aux confins orientales Byzantines. Cette
annexion a engendrée une expansion des territoires Byzantines vers l’est, mais avec
cela les frontières orientales sont devenues plus vulnérables. Les Seldjoukides
d’origine Oghuz ont consolidé leur pouvoir dans la même période. Seldjouk,
l’ancêtre de la dynastie Seldjoukide, était un notable dans le Khaganat Khazar. Il a
280
quitté son pays pour immigrer vers la Transoxiane, où ses descendants ont constitué
un rassemblement tribal qui est devenu progressivement le plus grand pouvoir de
monde musulman au XIe siècle. Les guerriers sous la direction de Tughrul et Çağrı,
qu’étaient les petits-enfants de Seldjouk, ont réussi à contrôler la région de Khorasan.
Cet état a facilement occupé le reste des domaines des Ghaznavides, Abbassides et
Bouyides et leurs armées commençaient de faire des razzias aux territoires
byzantines.
Le tournant dans les relations byzantines-seldjoukides fut la bataille de Manzikert
(1071). La défaite Byzantine a entraîné un vide de pouvoir en Asie Mineure.
L’absence d’autorité impériale, les révoltes des magnats locaux, les conflits entre les
prétendants au trône ont rendu possible l’invasion turque en Anatolie. Dans les deux
décennies l’empire a perdu presque tous les territoires anatoliens, à part des côtes et
certaines villes fortifiés. Il faut dire que cette invasion n’était pas centralisée et
dirigée par les Grands Seldjoukides, elle était généralement accomplie par les
activités individuelles de différents individus et groupes qui ne reposaient pas sur une
grande stratégie. Donc, au fil des siècles, les Turcs d’Anatolie ont repris la fonction
idéologique de leurs anciens ennemis arabes.
Les invasions de peuples nomades aux frontières balkaniques de l’empire ont, à
première vue, un impact idéologique moins important, car ces invasions n’ont pas un
effet idéologique équivalent à celles de l’Est. Cependant, les invasions par les
Bulgares et les Slaves ont eu un fort impact démographique dans les Balkans et ce
fait a déclenché plusieurs changements sociopolitiques, mais les discuter dépasse
largement le cadre de cette thèse. Pourtant, les différentes vagues d’invasions
barbares ont contribué à la fois à faire évoluer l’image des peuples “scythes” et à
créer une osmose sociale entre les peuples des steppes et les byzantins. Ce processus
281
a été décisif pour l’ethnogenèse de plusieurs nations balkaniques, comme les
Bulgares modernes.
La géographie idéologique et l’ethos culturel de la région frontalière byzantinemusulmane
sont antérieurs à l’arrivée des Turcs Seldjoukides au Proche-Orient.
Ainsi, les Turcs n’ont pas créé, mais hérité cet ethos des Arabes. Ce fait démontre la
continuité culturelle entre les guerriers arabes du thugur et leurs successeurs turcs.
Enfin, l’invasion seldjoukide de l’Anatolie mit fin à la prédominance byzantine des
Xe-XIe siècles au Proche-Orient et transféra la frontière byzantine-musulmane au
coeur de l’Asie Mineure.
VIEUX MÉMOIRES, NOUVEAUX BARBARES: LES BARBARES
TURCIQUES DANS L’IMAGINATION HISTORIOGRAPHIQUE BYZANTINE
Dans la littérature byzantine, le “barbare” est un archétype qui est hérité de la
littérature classique. C’est un mot onomatopéique qui désigne les étrangers qui sont
distingués par la langue qu’ils parlaient. Dans l’antiquité, la catégorie du “barbare”
désigne les gens qui n’apparentent pas à la société gréco-romaine.
Lorsque les Turcs Seldjoukides sont apparus pour la première fois aux frontières
orientales de l’empire au milieu du XIe siècle, leur image était indiscernable de tout
autre envahisseur nomade qui a créé des troubles à la frontière des Balkans. C’étaient
des nomades pastoraux qui appartenaient au genos Hun/Scythe. Ce genos est
considéré comme l’exemple le plus représentatif du barbare dans les traditions
littéraires classiques et byzantines.
Suite à la christianisation de l’Empire romain, l’ancien modèle gréco-romain est
quelque peu modifié, mais l’antagonisme entre “Romains” et “barbares” persiste.
282
Après la chute de l’Empire romain d’Occident, l’Empire romain d’Orient a continué
à survivre en tant que bastion du christianisme et de la culture gréco-romaine. Le
projet impérial de Justinien II représentait son ambition de récupérer ses terres
perdues à l’ouest, en particulier la péninsule italienne. Cependant, la restauration
impériale ne fut qu’un succès de courte durée. Un siècle plus tard, les Byzantins
perdirent même leurs provinces orientales, à savoir la Syrie, la Palestine et l’Égypte
au profit des Arabes.
L’évangélisation de la société romaine a naturellement affecté les opinions des
gens sur ceux qui étaient considérés comme « les autres ».
L’évangélisation des barbares (tels que les Bulgares et les Serbes) et leur situation
ultérieure aux yeux des auteurs byzantins sont également problématiques. Ces
peuples se sont convertis au christianisme orthodoxe grâce aux missionnaires
byzantins, comme Saint-Cyrille et Saint-Méthode. Cependant, ils n’ont pas pu
obtenir le statut de peuples civilisés aux yeux des Byzantins. Alors que les dirigeants
de ces nations étaient qualifiés de dirigeants “très chrétiens” dans les documents
préparés à la chancellerie impériale, les mêmes nations pouvaient apparaître comme
une “race barbare” ou une “race corrompue” dans les correspondances privées.
Dans ce cas, la situation des groupes ethniques tels que les Bulgares, les
Arméniens et les Valaques mérite une attention particulière. Ces trois groupes
ethniques n’ont pas été christianisés en même temps. Les Arméniens étaient l’un des
peuples qui ont embrassé la religion chrétienne plus tôt, comme au IVe siècle, avec
les Géorgiens, mais ils étaient généralement des chrétiens non chalcédoniens qui
avaient leur propre église apostolique arménienne. Cependant, il y a un grand
nombre d’Arméniens dans l’élite politique byzantine qui ont adopté la croyance
chalcédonienne.
283
Les Valaques étaient les habitants post-romains socialement marginalisés de la
péninsule balkanique qui parlaient une langue d’origine latine et vivaient une vie
semi-nomade dans diverses parties des Balkans. Ils étaient chrétiens et malgré leur
marginalisation socio-économique, ils étaient des vestiges des habitants de langue
latine de l’Empire romain dans la péninsule balkanique. Ainsi, théoriquement, ils
pourraient revendiquer l’héritage des “Romanitas” autant que les Grecs byzantins.
Cependant, probablement par leur marginalisation sociale, ils étaient encore
considérés comme inférieurs et quasi-barbares.
Une expression frappante de cette différence entre les Byzantins et les autres
“barbares orthodoxes” se trouve dans le Commentaire de Jean Kanaboutzes à Denys
d’Halicarnasse. L’auteur du XVe siècle déclare que les Grecs considéraient les
Troyens comme des barbares, bien qu’ils crussent aux mêmes dieux, les Byzantins
contemporains considèrent les Bulgares, les Valaques, les Albanais et les Russes
comme des barbares ; car la barbarie n’est pas une question de religion, mais de race,
de langue, de style de vie et de culture.
Ces ethnonymes pourraient être perçus comme des signifiants (dans le sens
inventé et employé par Ferdinand de Saussure) utilisés par les Byzantins pour définir
toute ethnie étrangère. La cause de leur divergence avec les noms réels des
populations qui les référaient est la différence entre la langue littéraire, dont on
pouvait trouver des exemples dans les textes écrits et la langue non écrite. Dans ce
contexte, les ouvrages écrits en grec proche du grec vernaculaire peuvent donner
quelques idées sur les noms grecs utilisés dans la langue parlée ou encore sur les
endonymes de ces populations dans leur propre langue. Un bon exemple d’un tel
ouvrage est De Administrando Imperio.
284
La première étape pour comprendre la formation de l’image des Turcs dans la
littérature byzantine consiste à classer et à analyser ces ethnonymes :
Les deux catégories chronologiques d’ethnonymes sont celles héritées de
l’Antiquité classique et celles apparues plus tardivement. La source la plus
importante d’ethnonymes classiques est l’Histoire d’Hérodote. Le Persan, le Scythe,
l’Égyptien et l’Indien sont les exemples les plus fréquents des ethnonymes
classiques.
Le groupe des ethnonymes postclassiques est formé par les termes qui ont été
attestés pour la première fois dans l’Antiquité tardive. Turc, Slave, Goth et Hun
pourraient être considérés comme des exemples de cette deuxième catégorie. Enfin,
il y avait les ethnonymes contemporains, les mots utilisés du vivant des auteurs
byzantins, terme qui était peut-être l’endonyme de ces peuples. Les termes, tels que
Petchénègue, Oguz et Turcoman étaient les noms contemporains de ces populations.
Il y avait aussi les ethnonymes qui ne reflétaient pas une appartenance ethnique
claire mais une indication géographique. Les mots, tels que Mysoi (utilisé pour les
Bulgares) et Triballoi (utilisé pour les Serbes), sont des exemples de tels termes. Ces
termes sont essentiellement des dérivés de régions géographiques. Cependant, il
serait erroné de dire que ces ethnonymes n’avaient pas un poids sémantique
particulier, car il s’agissait des noms des anciennes provinces impériales ; et une telle
appellation pourrait légitimer la réincorporation de ces terres dans l’Empire.
Pourtant, une dernière catégorie –et hybride– est celle des ethnonymes
composites, qui étaient constitués des deux éléments comme le tauro-scythe, le turcoarménien
et le perso-turc. Dans le premier cas, le terme est constitué d’une indication
géographique et d’un ethnonyme ; il a été utilisé pour les Tatars de Crimée. L’autre
285
terme est constitué avec la même formule et il a été utilisé pour décrire les
Turcomans d’Akkoyunlu en Anatolie orientale.
Certains de ces ethnonymes sont très révélateurs pour notre objet d’étude.
L’ethnonyme “scythe” par exemple – indépendamment des Scythes du temps
d’Hérodote, employé pour différentes populations nomades et pastorales d’Eurasie.
Ce terme est employé pour de nombreuses populations turciques et slaves, pour les
Mongoles et même pour les Russes. Cet ethnonyme fait référence à un mode de vie
et à une localisation géographique assez ambiguë où ces populations sont censées
vivre.
La mémoire des Huns a également survécu dans la littérature byzantine en tant
que stéréotype historiographique. L’ethnonyme “Hun” est apparu à l’antiquité
tardive, mais a été utilisé au cours des siècles suivants pour, par exemple, les Turcs,
les Hongrois et d’autres peuples des steppes. Comme on peut le voir dans cette thèse,
Attaleiatès a également défini les Seldjoukides comme des Huns. Mais une fois
qu’ils se sont installés en Iran et en Anatolie et ont entamé une vie sédentaire,
l’ethnonyme utilisé pour les Seldjoukides s’est transformé à « persan ». Après la
chute de l’empire perse, cet adjectif a été utilisé sur une base géographique, et les
Seldjoukides étant appelés par ce nom peuvent également être considérés comme une
sorte de translatio imperii. Mais il faut encore noter que les Seldjoukides (même sa
branche à Rum) étaient un État persan qui embrassait la culture persane et utilisait le
persan comme langue de la cour et de la bureaucratie et qui s’identifiait aux anciens
héros perses dans la Shahnama. De plus, ils utilisaient les titres et les symboles de
l’ancienne monarchie perse comme instrument de légitimité dynastique. La figure de
“ennemi par excellence” persan pourrait aussi régénérer les souvenirs d’un passé
lointain.
286
Dans ce chapitre, on traite aussi la représentation des Mongols dans la littérature
byzantine: Les Mongols sont une population nomade qui, jusqu’à leur expansion
inattendue au XIIIe siècle, vivait hors des frontières du savoir géographique
byzantin. Ainsi, jusqu’au XIIIe siècle, il n’y avait aucune référence aux Mongols
dans les sources byzantines. L’invasion mongole de l’Asie occidentale a été l’un des
événements clés qui ont déterminé le paysage humain de toute la région, en
particulier de l’Asie Mineure. Cela a déclenché de nouvelles vagues de migration
turque vers l’Anatolie. Le déclin du sultanat seldjoukide de Rum après la défaite à la
bataille de Kösedağ (1243) aux mains des Mongols a également relâché l’autorité du
sultanat sur les divers groupes turkmènes des bords sud et ouest de l’Anatolie turcomusulmane
et a permis la corrosion corrosive pillages des territoires byzantins par
les Turcomans.
Les Mongols sont attestés dans les oeuvres d’Acropolite, Pachymère, Gregoras,
Cantacuzène, Doukas et Chalcondyle. Dans les deux derniers ouvrages, il y a un
matériel très riche sur les affaires militaires et politiques du célèbre chef de guerre
mongol Tamerlan.
Dans ces textes, la représentation byzantine des Mongols est loin d’être négative.
Dans le récit d’Acropolite, les Mongols n’avaient qu’un rôle marginal. Pachymère,
bien qu’il évoque les légendes sur le cannibalisme chez les Mongols, utilise un
langage équilibré concernant les Mongols, également en raison du contexte politique
du début du XIVe siècle : il écrivait à une époque où l’Empire byzantin cherchait une
alliance avec les Mongols, contre les Turcs en Asie Mineure et les Royaumes Serbes
et Bulgares dans les Balkans. Doukas ne s’intéresse pas aux Mongols, sauf la
campagne anatolienne de Tamerlan qui déclenche l’interrègne ottoman. Enfin,
l’intérêt de Chalcondyle pour le monde eurasien contemporain pourrait être
287
interprété selon deux facteurs. Premièrement, l’auteur prend le modèle
historiographique d’Hérodote comme exemple pour ses Histoires. Deuxièmement, le
Khanat de Crimée s’est formé en tant qu’État successeur de la Horde d’Or en 1441
car son territoire se composait de la péninsule de Crimée.
Aux yeux des auteurs byzantins, les aspects visibles de l’appartenance religieuse
des premiers seldjoukides étaient indiscernables des autres populations turques
restées fidèles à leurs croyances ancestrales. Cependant, au fil du temps, le conflit
seldjoukide-byzantin a pris une tournure plus religieuse. Ce fait pourrait s’expliquer
à la fois par l’évolution idéologique de l’Empire byzantin sous les Comnènes et par
une éventuelle auto-identification plus forte des Seldjoukides à l’identité musulmane.
À partir du XIVe siècle, avec la montée des Ottomans, le religieux devient aux yeux
des auteurs byzantins, l’élément indispensable de l’identité turque.
Enfin, les Seldjoukides avaient une fonction idéologique particulière pour les
Byzantins. Depuis leur installation en Anatolie centrale, ils ont joué un double rôle
idéologique. Ils se substituaient, tous les deux, à l’ancien ennemi musulman dans les
régions frontalières et à l’ancien rival perse de l’empire byzantin dans les siècles
précédant l’arrivée de l’islam. Les Ottomans représentent un peu la continuation de
cette image. Ils sont encore parfois appelés Persans. Cependant, la relation avec les
lettrés byzantins, telle qu’elle peut être comprise à partir de ce que les auteurs ont
écrit, suggère une situation complètement nouvelle. Il n’y a plus l’antagonisme de
deux anciens rivaux de force à peu près égale, mais un monde révolu a été englouti
par un nouvel Empire qui prend la place de l’ancien.
288
LES RÉCITS BYZANTINES SUR L’ORIGINE DES TURCS
Les narrations origo gentis sont les textes qui traitent l’origine d’un peuple. Ces
narratives peuvent avoir des éléments historiques ou ils peuvent être totalement
fictifs ou légendaires. Les textes origo gentis sont produits depuis l’antiquité
classique, mais ils sont progressivement augmentés et diversifié au Haut Moyen-
Age, pars que les invasions barbares ont changé la situation démographique de
l’Europe et il fallait situer les histoires primordiales de ces nouveaux peuples dans le
monde gréco-romain.
Dans cette période, Cassiodore, Jordanes, Widukind et Saxo Grammaticus ont
écrit les oeuvres pour le but susmentionné. Dans ces narratives c’est possible de voir
trois motifs communs:
a) Victoire militaire contre des ennemis puissants
b) La traversée d’une rivière
c) La conversion à une nouvelle religion
Il y a deux récits byzantines d’origo gentis sur les Seldjoukides. La première
narration se trouve dans l’histoire de Jean Skylitzés. Le deuxième récit se trouve
dans le texte de Michel Attaleiatés.
Jean Skylitzès localise la patrie originale des Seldjoukides au nord du Caucase.
Selon Skylitzès les Seldjoukides sont de race hunnique. Après cette introduction, il
donne une brève digression sur la fragmentation du pouvoir abbasside et dit que un
certain Mouchoumet était le souverain de la Perse et de ses environs. Ce souverain
289
demandait des soldats par le souverain de Tourkia (probablement les Karakhanides
de Transoxiane) et il envoya un groupe des trois mille soldats. L’auteur donne les
noms de ces soldats comme Tangrolipex Moukalet. L’auteur mentionne évidemment
Tughrul Beg. Après avoir utilisé ces soldats contre ses ennemis sarrasins, il a voulu
les envoyer dans leur pays d’origine. Mais les Turcs se rebellent et occupent la Perse.
Les troupes Turcs massacrent les perses et les sarrasins, Tughrul devient sultan et
partage le pays entre les commandants turcs.
Attaleiates construit un narratif diffèrent sur l’origine des Turcs: il dit que les
Turcs sont les Huns Nephtalites. Leur pays origine est séparé du Perse par la rivière
Gange. Le chef des Huns Nephtalites était un captif d’origine servile et devient le
souverain de Perse, après la mort du “despote” au pouvoir. Ces Huns alors
commencent des razzias annuelles aux territoires byzantines et battent les défenseurs
romains. Donc, un commandant géorgien nommé Liparites va défendre ces territoires
contre eux, mais il perd la guerre et il tombe captif entre les mains des Turcs.
L’ethnarque des Turcs, dit sultan dans leur langue, voit le courage de ce commandant
et le traite de manière très respectueuse et honorable.
Dans ces narrations il y a des différences intéressantes: Skylitzès localise le pays
d’origine des Turcs en Caucase, mais Attaleiates le met dans l’Inde. Skylitzès est
beaucoup plus informé sur l’histoire des Turcs tandis que la connaissance
d’Attaleiates est beaucoup plus limitée. Skylitzès avait probablement utilisé des
fonds orientaux que Attaleiatès ne pouvait pas accéder.
Dans ce chapitre, le deuxième thème étudié concerne les récits d’origine sur les
Ottomans. Les premiers Seldjoukides se trouvaient dans les steppes asiatiques, loin
des frontières byzantines, mais la formation du beylik ottoman s’est produite dans les
régions frontalières de la Bithynie, de sorte que plusieurs passages du développement
290
des Ottomans ont été réalisés sous les yeux byzantins. Pachymère nous a donné la
seule représentation contemporaine des actes d’Osman I, le premier bey des
Ottomans. Cependant, son récit concernant Osman pourrait difficilement être qualifié
de récit origo gentis. Pachymère le présente comme l’un des chefs persans qui
attaquent et dévastent le territoire byzantin. Il s’agit de la première référence au
fondateur de la dynastie ottomane dans un texte historiographique connu. Puis
l’auteur byzantin raconte la bataille de Bapheus (1302) et dans le chapitre suivant, il
compte Osman parmi les chefs qui ont envahi et pillé les parties supérieures de la
Bithynie, de la Mysie, de la Phrygie et de la Lydie. Néanmoins, il n’existe aucun
récit sur l’origine des Ottomans et l’auteur s’est concentré uniquement sur les faits et
gestes militaires des Osman. Doukas est silencieux sur les origines de la dynastie
ottomane et ses débuts, mais il transmet une prophétie plutôt étrange sur le
parallélisme historique entre deux dynasties, les Ottomans et les Paléologues. Cette
divination indique que la fin des dynasties paléologue et ottomane se succédera. Les
règnes d’Osman Ier et de Michel VIII étaient presque contemporains. Le règne de
Michel VIII a pris fin, lorsque le règne d’Osman I a commencé. Ainsi, les
Paléologues cesseront d’exister d’abord, puis les Ottomans tomberont. On peut dire
que les Chalcondyle, contemporain de Doukas sont plus intéressés par l’histoire des
débuts des Ottomans. Cet intérêt donne également des informations sur les milieux
ottomans dont Chalcondyle était proche. L’auteur, qui inclut des digressions
ethnographiques dans son travail, fait quatre affirmations sur l’origine des Ottomans
(plus précisément, des Turcs). La première d’entre elles est que l’origine des Turcs
est les Parthes. Chalcondyle ne s’attarde pas trop sur cette affirmation. Le deuxième
récit raconte que le pays d’origine des Turcs était une grande ville de Perse appelée
Tourke. Une troisième histoire origo gentis prétend que les Turcs sont venus
291
d’Arabie et étaient des descendants du Calife Omar. Cette affirmation coïncide avec
le chroniqueur égyptien Ibn Hajar et l’écrivain ottoman Enveri, qui ont déclaré que la
patrie d’origine des Ottomans était l’Hedjaz. Le dernier récit est que les Turcs sont
d’origine scythe. L’auteur est également le plus enclin à cette explication:
Chalcondyle dit que, comme preuve de l’origine scythe des Turcs, les Scythes (c’està-
dire les tribus turco-mongoles d’Eurasie) et les Turcs d’Anatolie avaient des
coutumes et des traditions similaires et parlaient des langues qui étaient proches les
uns des autres.
Donc les Seldjoukides ont d’abord formé leur sultanat en Transoxiane et en Perse,
dans les territoires éloignés des frontières de l’Empire byzantin. Leur histoire avant
leurs premières razzias dans l’oikoumene byzantin pourrait être parvenue aux auteurs
byzantins sous forme de rumeurs venues d’Orient, probablement via des
interlocuteurs arméniens ou syriaques. Cependant, le beylik ottoman se développe
sous les yeux des Byzantins. Cette différence explique l’absence de récits byzantins
origo gentis sur les Ottomans au XIVe siècle.
Une autre sous-section de ce chapitre est consacrée aux origines sociales des
Turcs et à leur rapport à l’esclavage. Il sera résumé comment ce sujet a été perçu par
les écrivains byzantins dans le cadre de leurs propres visions du monde. L’esclavage,
sous ses diverses formes, est l’une des institutions les plus répandues du monde
médiéval. Cette institution existe aussi bien à Byzance que dans le monde islamique.
Bien que généralement, les captifs de guerre constituent la plus grande source de ces
esclaves, ce fait n’était pas vrai dans le cas de l’Empire byzantin et du califat
abbasside qui se faisaient constamment la guerre, car entre les deux entités il y avait
souvent des traités d’échanges de prisonniers de guerre. Ainsi, dans l’Empire
byzantin, il semble qu’une partie importante des esclaves sont venus du nord de la
292
mer Noire, c’est-à-dire ils sont des esclaves scythes qui pouvaient être de souche
slave ou turque. Ces esclaves étaient également employés dans l’agriculture dans les
parties rurales de l’Empire.
En grec byzantin, le terme générique désignant l’esclave est δουλος. Mais après le
XIe siècle, un nouveau terme apparaît : σκλαυος. Ce terme existait en grec byzantin
en tant qu’ethnonyme depuis le VI siècle, il signifie le slave. Ce mot est passé à la
langue arabe sous le nom de saqaliba, en tant que terme désignant l’origine slave ou
généralement les esclaves blancs. Le mot a pris le sens d’esclave, probablement avec
l’impact de la langue arabe. Ce mot σκλαυος est à l’origine des mots pour esclave
dans certaines langues européennes : comme slave en anglais, esclave en français et
schiavo en italien.
Ainsi, dans la société byzantine, il y avait une présence visible des esclaves qui
venaient généralement du monde communément appelé Scythie. Cette Scythie
pouvait parfois coïncider avec la patrie slave. D’autre part, l’esclavage dans le
monde islamique est également très important. Il existe un corpus scientifique
volumineux sur les différents aspects de l’esclavage dans les sociétés islamiques
médiévales. L’esclavage était un statut socio-économique déterminé par la religion et
dans les sociétés islamiques médiévales. İl existe trois types d’esclavage : l’esclavage
domestique (qui comprend les concubines), le travail des champs et l’esclavage
militaire. Les trois catégories étaient fréquentes. Mais ce dernier est un phénomène
particulièrement identifié aux sociétés musulmanes médiévales. Ici, je peux proposer
une généralisation de ces processus à la formation de l’État ; le modèle tribal est
prédominant dans les territoires où le mode de vie nomade et les populations turques
étaient dominants et le modèle “mamelouk” était dominant là où le mode de vie
sédentaire et la population non turque étaient dominants. Pourtant, on pourrait dire
293
que l’État seldjoukide représentait un hybride de ces deux modèles. Bien que les
fondateurs de l’État n’étaient pas des esclaves, ils font partie de l’élite nomade de la
société à laquelle ils appartenaient, ils ont fondé le noyau de leur État au Khorasan,
un pays où la majorité des habitants étaient persans. Ils ont donc adopté les coutumes
de l’administration de l’État persan; ils avaient une chancellerie persane, adoptaient
les traditions bureaucratiques des premiers États musulmans d’Iran et organisaient
des troupes de ghulam, c’est-à-dire les guerriers esclaves. Cependant, ces faits ont
déclenché des tensions entre l’élite dirigeante et les Turcomans nomades, ainsi
qu’entre les membres de la dynastie au pouvoir où les princes prétendants se sont
alliés aux Turcomans ou à d’autres bandes de guerre nomades contre leur autorité
centrale. Ce fait est un événement fondateur de la politique d’immigration aux
marges du Grand Sultanat seldjoukide qui a également déclenché la formation du
Sultanat de Rum.
Dans ce chapitre, j’ai traité certaines des questions relatives à la représentation
des origines mythiques, historiques ou sociales des élites dirigeantes et des
populations ottomanes et seldjoukides. J’ai surtout évoqué les récits origo gentis sur
les Seldjoukides et les Ottomans, car ces récits avaient deux fonctions importantes :
Premièrement, ces récits situent de nouvelles populations dans la géographie
humaine d’un monde déjà connu. Ils les ont donc rendus familiers aux Byzantins.
Deuxièmement, l’explication de l’origine introduit ces peuples dans un récit
historiographique général qui inclut des événements ultérieurs liés à ces populations.
La pensée aristotélicienne est très importante pour comprendre le monde
intellectuel byzantin ; comme cela a été souligné ci-dessus, ses textes ont été lus dans
les cercles intellectuels byzantins et ont eu une forte influence sur la vision du monde
byzantine. La pensée aristotélicienne définit les gens dans certaines sociétés du
294
monde comme sujets à l’esclavage. La question de l’esclavage est également liée à
l’explication des origines susmentionnée. L’existence d’une origine servile ou toute
référence à celle-ci démontre un manque de valeur attaché à un individu dans la
vision du monde byzantine. L’image du Scythe est étroitement liée à cette approche.
Il y a une abondance d’esclaves "scythes" dans l’Empire byzantin et ces dirigeants
pourraient être potentiellement considérés comme contradictoires avec les concepts
byzantins de noblesse.
Dans le dernier sous-chapitre, j’ai tenté de compléter la question de l’esclavage
par une analyse de la succession des souverains et des conflits dynastiques dans les
entités turques tels qu’ils étaient reflétés par les textes byzantins. Je pensais que ces
récits étaient importants pour comprendre comment les Byzantins percevaient ces
entités, leur structure interne et les sources de légitimité.
LA PLACE DES ÉTATS BARBARES DANS LA VİSİON DU MONDE
BYZANTINE
Les autres entités entourant l’Empire byzantin étaient considérées comme
hiérarchiquement inférieures par les Byzantins. Le titre “empereur” (βασιλεύς) était
employé exclusivement pour définir l’empereur byzantin, car l’empire byzantin
n’était pas considéré comme un successeur de l’empire romain mais comme
l’empire-même. Contrairement à l’historiographie moderne, les historiens byzantins
croient en une continuité ininterrompue entre les anciens Romains et les Byzantins
médiévaux.
Dans le centre de la vue du monde byzantine, il y avait l’empereur. Ce titre était
utilisé exclusivement pour les dirigeants byzantins jusqu’à la fin de l’ère byzantine,
295
son exception la plus importante étant les shahs persans. Mais le même titre était
également revendiqué par les rois bulgares et serbes qui cherchaient à étendre leur
autorité dans les Balkans et ce défi était perçu comme une menace par la classe
dirigeante byzantine. Pourtant, ces souverains utilisaient des titres tels que “basileus
des Romains ou des Bulgares” ou “basileus des Romains et des Serbes” dans leur
titulature officielle. Au XIIe siècle, l’exclusivité byzantine sur le titre impérial a été
assouplie; les auteurs comme Nicetas Choniatés et George Acropolite ont employé ce
titre pour les dirigeants chrétiens étrangers. Ce changement était étroitement lié à
l’évolution de l’idée impériale à travers les siècles. L’empereur byzantin ne
représentait pas la même autorité qu’il représentait aux XIe et XIVe siècles. L’éthos
politique de la dynastie Comnène était très différent de celui des Paléologues.
Donc les rois barbares étaient perçus comme des inférieurs dans la vision du
monde byzantine. Ainsi, un dirigeant turc, comme tout dirigeant d’origine non
byzantine, ne peut jamais être égal à un empereur et a toujours une réputation
inférieure à celle de l’empereur byzantin.
C’est une tendance courante dans les sources byzantines de voir la formation des
principautés turques comme l’activité d’un groupe de bandits. La formation des
beyliks en Asie Mineure est un thème que l’on retrouve dans quatre byzantins textes
historiographiques. (Pachymère, Gregoras, Doukas, Chalcondyle) Cet événement est
simultané à la dissolution de l’État seldjoukide. Germiyan et Karaman ont été les
premiers beyliks formés, pendant la vacance du pouvoir à cause du déclin des
Seldjoukides. Contrairement aux derniers beyliks, les deux semblent avoir une
origine tribale. Karaman, le fondateur éponyme du beylik, a été décrit comme un
brigand dans plusieurs sources seldjoukides. Germiyan dont la capitale est Kütahya
(Kotayeion) en Phrygie, apparaît comme suzerain des autres beyliks de l’ouest de
296
l’Asie Mineure à l’origine. Pachymère, témoin de l’époque, raconte l’invasion de
l’Anatolie occidentale par les beys turcs, comme une simple activité de bandits. Son
récit est assez détaillé et contient des informations absentes des autres sources. Dans
son récit, il est possible de voir l’entrée en scène des seigneurs de guerre turcs, non
pas en tant que fondateurs de dynasties, mais en tant que simples seigneurs de guerre.
Cependant, dans les sources ultérieures Pachymère et Gregoras, il y a un récit selon
lequel cet événement a eu lieu en tant que coalition, dans le cadre d’un certain plan
stratégique. Sans aucun doute, ce récit est une légende historiographique. Toujours
dans le récit de Doukas, le message est clair : il n’y a pas de stratégie ou d’alliance
commune entre ces dirigeants turcs ; chaque chef a agi de manière indépendante et a
occupé d’anciennes régions byzantines. Parmi ces quatre écrivains, Doukas était
celui qui connaissait le mieux l’histoire récente de l’Anatolie occidentale. Il semble
donc qu’à partir des travaux de Gregoras, les auteurs byzantins aient créé un tel récit
téléologique pour expliquer l’expansion rapide des beyliks anatoliens. XIIIe siècle
est un siècle des beyliks en Asie Mineur. Mais les Ottomans occupent ces émirats, à
commencer par l’annexion de Karasi en Mysie en 1345-1346. L’expansion ottomane
vers l’Anatolie musulmane a duré un siècle et demi. Au début du XVIe siècle, les
Ottomans ont intégré les émirats de Dulkadir et de Ramazan, les derniers beyliks
turkmènes indépendants du sud de l’Asie Mineure et ils ont consolidé leur
domination dans toute la région.
Dans les sociétés occidentales médiévales, il existait plusieurs règles de
succession, comme la primogéniture, qui serve à rendre prévisible l’ordre de
succession. Mais, dans les États turcs et musulmans, il n’y avait pas de règle de
succession universellement acceptée. Dans les États turcs, il y avait un autre aspect
important, à savoir que tout le territoire de l’État était considéré comme le domaine
297
de la famille royale. Ainsi, comme le territoire appartient à la dynastie, tout membre
de la dynastie peut revendiquer le trône. Ce type de succession sans règles strictes
pourrait être appelé “succession ouverte”. Dans ce système, le prince qui parvient à
gagner le soutien des différentes factions de la cour et à éliminer ses frères peut
hériter du trône. On pourrait supposer que dans de telles situations, les frères aînés
sont légèrement plus avantagés en raison de leur expérience et éventuellement de
réseaux plus importants, mais le résultat n’est pas toujours en leur faveur.
Ci-dessous, j’ai donné un aperçu de l’évolution de cette succession, en tenant
compte des usurpateurs et des rébellions dans les États turcs, du point de vue
d’auteurs byzantins qui connaissaient des crises similaires au sein de leur société
dans laquelle les crises de succession n’étaient pas rares.
Depuis la première apparition des Seldjoukides dans l’oeuvre de Skylitzès, les
auteurs byzantins étaient conscients de cette notion d’état patriarcal des Turcs, et ils
ont remarqué l’absence de consensus commun dans leurs familles royales et leurs
intrigues sans fin pour obtenir une plus grande partie de l’autorité royale. Dans le
récit de Skylitzès, lorsque le prince seldjoukide Ibrahim Yinal a pris Liparites en
captivité et l’a amené à la ville de Reyy où son frère Tughrul régnait en tant que
sultan, mais il est devenu jaloux de son l’exploit de son frère et cherche un prétexte
pour se débarrasser de son frère. Peu de temps après, Ibrahim Yinal remarque les
complots du sultan contre lui et se rebelle contre son autorité avec son neveu
Koutloumous. Cependant, Tughrul bat les insurgés et Ibrahim Yinal est exécuté. Par
la suite, l’autre dissident s’échappe avec 600 hommes et avec Melech, le fils
d’Ibrahim Yinal.
Synopsis Historion de Skylitzès se termine avec la fin du règne de Michel VI en
1057, mais il est possible de suivre les séquelles du prince rebelle seldjoukide
298
Koutloumous dans d’autres sources. Selon les sources seldjoukides, il fut tué en 1063
par son cousin Alp Arslan après une autre rébellion.
Dans l’histoire d’ Attaleiatès, il est possible de voir d’autres nobles seldjoukides
dissidents. Chyrsoskoulos, le beau-frère du sultan Alp Arslan qui arrive à la cour
byzantine, est l’un d’entre eux. Les deux “nobles de Perse” qui “avaient hérité le
nom de Koutloumous de leur père” rencontrent Nicéphore Botaniates à Nicée. Les
deux frères qui établirent un noyau d’État dans la ville - ce qui deviendra plus tard le
Sultanat de Roum - fléchirent le genou devant Nicéphore Botaniates. Comme il a
déjà été mentionné, l’auteur insiste sur leur lignée royale seldjoukide. L’auteur
appelle cette branche cadette de la dynastie "Κουτουλμουσίοι" et note que leurs
commandants appellent ces princes "αμηράδασ" et "σελαριοι" en langue turque. Ces
titres sont évidemment “amir” et “salar”, les titres militaires d’origine arabe et
persane.
Les deux frères étaient Süleyman et Mansur qui ont immigré vers l’ouest après
l’exécution de leur père avec un groupe de nomades turkmènes, appelés
collectivement Nawakiya. Süleyman est ensuite devenu le fondateur de la branche
anatolienne des Seldjoukides. C’est possible de suivre les exploits ultérieurs de
Süleyman dans L’Alexiade. Anna Komnene, bien qu’elle présente les opérations de
Süleyman en Asie Mineure comme la carrière d’un chef de guerre, note
soigneusement qu’il portait le titre d’émir au début, puis promu sultanat. Elle semble
consciente des divisions internes au sein du Grand Empire seldjoukide et de leurs
luttes dynastiques. Anna dresse le portrait de Toutuche, le sultan seldjoukide de
Syrie, comme un homme ambitieux et arrogant qui a tué Süleyman au combat et a
cherché à s’emparer du trône seldjoukide.
299
Cinnamus est resté silencieux sur les relations interfamiliales au sein de la
dynastie seldjoukide, mais comme cela a été démontré dans le cas des chefs
turkmènes, il considérait le sultan de Rum comme un dirigeant de la fédération de
tribus qui partageait son autorité avec les chefs turkmènes.
L’approche de Choniatès vis-à-vis des relations dynastiques du sultanat de Rum
était également similaire à celle de Cinnamus, mais il était clairement plus informé
des réalités politiques au sein du sultanat de Rum. Comme indiqué dans le chapitre
précédent, l’allégeance islamique des Seldjoukides a été légèrement plus soulignée
dans l’Histoire des Choniatès. Donc, étant le frère de Michel Choniatès qui était
l’archevêque d’Athènes, Nicetas Choniatès semble plus intéressé par les questions
religieuses.
Le texte de Pachymère montre l’image ultime d’un sultan seldjoukide aux yeux
d’un intellectuel byzantin : le sultan d’un domaine bien défini (la Perse) qui a disparu
et est devenu un vassal des Mongols, un homme fatigué et buveur qui intrigue sans
succès pour reprendre son trône, et qui était à la différence d’autres brigands persans
en Asie Mineure qui évolueront plus tard les fondateurs des beyliks anatoliens.
Malgré le ton pitoyable de la représentation du sultan en exil, sa vitalité et son
ressentiment qui l’amenèrent finalement à trahir l’empereur Michel VIII, furent
démontrés de manière vivante dans l’oeuvre de Pachymère. Après son complot de
1264 où il provoqua l’attaque des Bulgares et des Tatars sur le sol byzantin, le Sultan
fut livré aux Tatars qui le transférèrent en Crimée.
Dans le texte de Gregoras, il y a quelques observations sur la succession dans
l’État ottoman : tandis que l’auteur mentionne le prince Halil, fils d’Orhan, il fait
quelques déclarations sur le mécanisme. Selon Gregoras, Halil a hérité de la région
autour de la baie de Nicomédie de son père et y régnait de manière autonome. Cette
300
description est conforme à la pratique ottomane de l’apanage. L’auteur mentionne
ensuite la mort du prince Süleyman (né en 1357) et déclare qu’il était le fils aîné et le
successeur. Cependant, le prince Halil était le fils d’Orhan et de Theodora
Cantacuzène et fiancé à l’une des filles de Jean V Paléologue. Ainsi, sa descendance
impériale byzantine et l’alliance de mariage avec les Paléologues font de lui un
successeur plus adéquat au trône ottoman pour les Byzantins. l’auteur mentionne que
l’empereur voulait qu’Orhan proclame Halil comme héritier officiel; Néanmoins,
l’Histoire de Gregoras s’achève avant la mort d’Orhan et on ne pouvait y lire la
conclusion de ce projet infructueux.
LES MECHANISMES DE COEXISTENCE: ANTAGONISME,
ACCULTURATION, ASSIMILATION
Ce chapitre traite l’entrée des Turcs dans la société byzantine. Pendant les quatre
siècles de coexistence des états Turcs et l’Empire Byzantine, un nombre remarquable
de Turcs passèrent du côté byzantin. La première manière de l’entrée des Turcs au
Byzance, c’est emploi impérial des bandes de mercenaires turciques. Ce fait n’est pas
commencé au 11ème siècle. Les peuples turciques ont offert un service mercenaire
occasionnel à l’Empire byzantin au moins depuis le 6ème siècle. Comme on l’a déjà
vu dans le troisième chapitre, il y avait des mercenaires hunniques dans l’armée de
Justinien II. Au cours des siècles suivants, les Byzantins employèrent des Avars, des
Bulgares et des Khazars pour divers services militaires. Au 11ème siècle, les
guerriers Petchénègues étaient attestés parmi les rangs byzantins, ils ont servi
fidèlement même à la bataille de Manzikert. Il est possible de supposer que cette
«main-d’oeuvre barbare» dépendait des nomades turcs qui occupaient les différentes
301
parties de la péninsule balkanique tout au long des premiers siècles médiévaux. Cette
utilisation de la main-d’oeuvre étrangère dans les guerres assura le service des
“populations barbares” à l’Empire byzantin.
Une variation de cette manière de l’usage des peuples Turciques c’était
l’utilisation des différents éléments des sociétés nomades comme allies contre les
autres éléments de leur société. Un exemple de cette politique c’était la diplomatie
byzantine en regard la société petchénègue en XIe siècle. Comme on a déjà dit, les
tribus Petchénègues occupaient les territoires Byzantines dans le bassin de Danube.
Deux chefs petchénègues rivalisaient pour diriger leur société: Tyrach et Kegenes
(ou Kegen).
Dans un sous-chapitre on discute deux individus d’origine turcique qui sont
abordés dans la société byzantine: Tzachas et Syrgiannes Paléologue Philantropene.
Tzachas était un captif de guerre turc qui dans la situation chaotique d’Asie Mineur
après la bataille de Mantzikert, il a réussi de former un principauté dans la cote
Égéenne d’Anatolie. Il était honoré avec le titre “protonobellisimos” par l’empereur
Nicephore III, mais après Alexis I Comnene devenu l’empereur, il devient un rebelle
et forme une principauté séparatiste sur les rives de la mer Égée.
Alexis I le décrit comme un usurpateur qui veut gagner le trône byzantin dans une
lettre qu’il a écrit au sultan Kılıçarslan I qui est aussi le gendre de Tzachas.
L’empereur byzantin convainc Kılıçarslan que Çaka désire non seulement le trône
byzantin, mais aussi le trône du sultanat de Rum. En conclusion, le sultan tue son
beau-père. Tzachas, malgré il vient de société Seldjoukide, il cherchait son futur
politique dans l’Empire byzantin.
Un deuxième exemple similaire a Tzachas, c’est Syrgiannes Paléologue
Philantropene qui est le fils d’un noble coumane et une fille de la dynastie
302
paléologue. Il est connu comme un commandant compétent qui est aussi devenu le
gouverneur de Macédoine. Il est devenu un des principaux acteurs de guerre civile
byzantine 1321-1328. D’abord partisan du Andronic III, en pleine guerre civile, il
devient un partisan du Andronic II. Il a été promu mégas doux à cause de sa trahison
envers sa faction. Il a eu une carrière politique mouvementée qui se termine par son
évasion à la cour de Stefan Dusan, le roi de Serbie. Ayant le soutien du roi serbe,
Syrgiannes déclenche une rébellion aux frontières occidentales de l’Empire byzantin,
mais avant que sa rébellion ne devienne une menace pour l’empire; il a été assassiné
par Sphrantzes Paléologue.
À mon avis, il y a des parallèles dans la vie politique et militaire de ces deux
individus.Tous deux ont développé un projet d’État séparatiste basé sur certains
éléments turciques avec lesquels ils ont coopéré. Tzachas et Syrgiannes Paléologue
avaient des liens au sein de l’élite dirigeante byzantine et ils voulaient réaliser ce
projet politique sur les terres byzantines. Finalement, tous deux n’ont pas réussi à
réaliser le projet politique qu’ils visaient.
L’adoption du christianisme orthodoxe était une condition indispensable pour
l’intégration dans la société byzantine, comme on l’a vu dans le cas de Tzachas. Les
dignités et titres impériaux étaient réservés exclusivement aux chrétiens orthodoxes.
La position unique et dominante du christianisme orthodoxe en tant que
caractéristique centrale de l’identité byzantine est au-dessus de toute discussion.
Byzance était la Nouvelle Jérusalem, combien elle ressemble à la Nouvelle Rome.
Comme il a déjà été discuté ci-dessus, il faut distinguer la conversion du
christianisme du paganisme ou de l’islam. Parce que si le paganisme est l’ennemi
naturel du christianisme et que l’islam était essentiellement considéré comme une
version du paganisme par certains théologiens byzantins, cette dualité est légèrement
303
différente pour les musulmans. Comme cela a déjà été mentionné, la majorité des
théologiens byzantins considéraient l’islam comme une hérésie. L’islam, comme les
autres religions abrahamiques, donne une forte image de soi aux croyants et rend
plus difficile la conversion des autres religions. En outre, la conversion de l’islam au
christianisme est une question qui mérite un examen attentif. Dans la jurisprudence
islamique, la conversion d’un musulman à une autre religion est interdite et c’est un
crime puni de mort. On sait que dans les pays musulmans de notre époque, comme
l’Égypte sous les Mamelouks, de telles punitions ont eu lieu. Cependant, dans les
régions frontalières où les identités étaient fluides, comme les frontières, de telles
punitions étaient probablement moins fréquentes. De plus, la punition d’un tel acte
ne pourrait être possible que par une décision d’un tribunal musulman. Donc, si
l’apostasie a lieu après que la personne est devenue un sujet d’un État chrétien où, au
moment de l’apostasie, il n’y a aucune autorité pour la punir, elle reste impunie et
possible.
Au Xe siècle, lors des grandes heures de la reconquête byzantine vers la Syrie et
la Mésopotamie, l’Empire réussit à annexer des villes telles qu’Edesse et Antioche
qui abritaient une importante population musulmane. Dans les sources de cette
période, on trouve plusieurs références au sort des habitants arabo-musulmans de la
région. Certains de ces habitants ont quitté leurs villes et ont immigré vers les
principales terres musulmanes. Par contre, d’autres habitants sont restés dans cette
région. En 941, quelque 10 000 membres de la tribu arabe Banu Habib se sont
convertis au christianisme avec leurs familles et leurs esclaves. Au cours des années
suivantes, ils ont été suivis par d’autres membres de la tribu.
Le règne d’Alexis Ier Comnène marque un tournant dans l’histoire religieuse de
Byzance. Alexis I qui annonce un édit réformateur en 1107, forme plusieurs
304
nouveaux offices cléricaux concernant l’enseignement religieux. Parmi les postes
qu’il crée, il y a un bureau appelé didaskalon ton ethnon (l’enseignant des gentiles).
Ces didaskaloï furent employés pour convertir les barbares. Bien sûr, ces efforts
n’ont pas commencé avec cet édit, depuis des siècles Empire byzantin a un
programme missionnaire déterminé ciblant les populations païennes, qui ont réussi à
convertir les Bulgares et les Russes parmi les autres populations.
Une pratique traditionnelle chez les Turcs d’Asie Mineure mérite également
l’attention : les Turcs, malgré leur adhésion à l’Islam, baptisaient leurs enfants. Ce
fait semble lié à une croyance populaire selon laquelle les enfants non baptisés
pourraient être possédés par des démons ou « puer comme des chiens ». Les mères
de certains de ces enfants étaient chrétiennes orthodoxes. Bien qu’ils aient été
baptisés, l’identité religieuse de ces enfants est très douteuse. Le baptême est devenu
une partie des pratiques de leur religion populaire et ne reflète pas leur dévotion au
christianisme.
Cependant, à l’époque comnénienne, un nouveau zèle pour l’activité
missionnaire a émergé. Inutile de dire que l’invasion seldjoukide de l’Asie Mineure
et les croisades ont été deux événements importants qui ont déclenché ce zèle
renommé. Anna Komnene déclare dans Alexiad - concluant la générosité de son père
envers les officiers seldjoukides Elkhanes et Skaliaros qui ont contribué à leur
conversion au christianisme - "Il (Alexis I) était un excellent enseignant de notre
doctrine, avec la foi et le message d’un apôtre, désireux de se convertir à Christ non
seulement les Scythes nomades mais aussi toute la Perse et tous les barbares qui
habitent en Égypte ou en Libye et adorent Mahomet dans leurs manières
extraordinaires." La cible de cette activité était –sans aucun dout – les populations
turques d’Asie Mineure et de Perse.
305
Cette activité atteint son apogée sous le règne de Manuel I Comnène qui
s’intéressait personnellement beaucoup aux questions religieuses. Dans les oraisons
prononcées par les ecclésiastiques pendant son règne et même dans son épitaphe
prononcée par Eustathios de Thessalonique, il y avait des références évidentes à cet
effort missionnaire et à l’évangélisation des Perses.
Ce projet a également déclenché une discussion théologique à Constantinople.
Déjà lors de la visite de Qiliç Arslān II à Constantinople (1162), son entrée à Sainte-
Sophie comme “infidèle” dérangea les hauts gradés du clergé, dont le patriarche Luc
Chrysoberges. Au cours du règne de Manuel I Comnène, pour faciliter
l’évangélisation des Seldjoukides, il proposa un petit changement dans le catéchisme
orthodoxe, la suppression de l’anathème contre le Dieu de Mahomet. Cet anathème
dit que le Dieu de Mahomet c’est un Dieu qu’il “n’a jamais engendré, n’a pas été
engendré non plus” et était un solide (holosphyros). Manuel I a dû penser que sans
une telle formule il pourrait laisser entendre que les musulmans et les chrétiens
croient au même Dieu et que la conversion des Turcs au christianisme pourrait être
plus facile. Dans le but de lever l’anathème, il convoqua le saint synode avec les
membres du clergé. Comme on pouvait s’y attendre, cette proposition fait le
mécontentement du patriarche Thedosios Boradiotes et d’autres hiérarques. Comme
il a été décrit par Khoniatès, "ils ont tous secoué la tête en signe de refus, même à
contrecoeur pour écouter ses propositions, qu’ils considéraient comme calomnieuses
et portant atteinte à la plus vraie gloire de Dieu". Selon ces hiérarques, le Dieu de
Mahomet n’était pas un Dieu, mais "un Dieu solide fabriqué par le trompeur et
démoniaque Mahomet". L’un des champions de l’opposition à la formule de Manuel
I était Eustathe de Thessalonique. Avec une opposition aussi rigide du clergé, la
306
tension entre l’empereur et le clergé a augmenté. Cependant, à la fin, le synode a
trouvé une résolution : ils ont conclu un accord pour retirer l’anathème du Dieu de
Mahomet du catéchisme et y mettre l’anathème de Mahomet et ses enseignements.
Néanmoins, pendant la controverse et la crise de ce saint synode, Manuel était
déjà très malade. Il mourut en 1180 et pendant les règnes brefs et tumultueux des
derniers empereurs Comnéne et Ange, le zèle missionnaire s’éteignit. L’arrêt
possible de l’évangélisation des Turcs doit cependant être lié à la conjoncture
politique, non aux formules théologiques. L’Empire byzantin, dans les conditions
difficiles qu’il rencontra après 1185, ne pouvait plus soutenir un programme
religieux/politique aussi ambitieux.
Dans les siècles suivants, en particulier la période 1261-1453 qui a été
poétiquement appelée "une agonie prolongée des restes de l’Empire" par Gibb, la
dimension des conversions était radicalement changée. Cependant, dans une période
très tardive comme les années 1410, il est possible de voir un prince ottoman à
Constantinople, le prince Yusuf qui est le fils du sultan Bāyazīd I accepte le saint
baptême. Cette conversion est ignorée dans les sources ottomanes et elle n’est
attestée que dans l’Histoire des Doukas et des Chalcondyle.
En conclusion, la conversion au christianisme orthodoxe apparaît comme la
condition de la plus haute importance pour s’intégrer dans la société byzantine. Nos
sources acceptent généralement les chrétiens nouvellement convertis comme de bons
Romains, bien qu’elles ne cessent de souligner leur provenance “barbare”, elles n’ont
pas une approche d’exclusion à leur égard. Seule leur approche devient négative dans
les cas d’ambiguïté confessionnelle, comme dans le cas de Kaykāʾūs II et de ses fils.
Ce fait doit être considéré peut-être avec le phénomène des Crypto-musulmans, à la
périphérie de la société byzantine.
307
Comme il a été mentionné précédemment, les ambiguïtés et les conversions
spontanées en masse sont des caractéristiques typiques des régions frontalières. La
fréquence des conversions dans une région où coexistent différentes cultures et
religions pourrait expliquer la présence des croyances syncrétiques et des
“hétérodoxies”. Le credo paulicien était aussi le fruit de la même atmosphère
frontalière de la rencontre et de la coexistence du christianisme et de l’islam.
Cependant, lorsque ce credo a été transféré dans les Balkans à la suite du système
impérial des transferts de population, il peut trouver une puissante interaction avec
les Petchénègue récemment christianisés. Ces nouveaux chrétiens ont soutenu la
rébellion paulicienne de Lekas en Bulgarie avec beaucoup d’enthousiasme. Dans la
littérature religieuse et hagiographique, on trouve d’autres exemples de coexistence
et d’interaction entre les religions dans les zones frontalières. Mais les détails de ces
processus restent en dehors de mon objectif.
Entre les catégories des Grecs et des Turcs, il existe des identités hybrides de
Mixobarbaroi (μιξοβάρβαροι) et de Turcopoles (Τουρκόπουλοι). Les termes
Turcopole et Mixobarbaros sont des termes quasi-ethniques utilisés dans
l’historiographie byzantine pour désigner les demi-barbares.
Turcopole signifie "fils de Turc". Ce terme était généralement expliqué comme la
progéniture des unions d’hommes turcs et de femmes grecques, de plus, il était
principalement utilisé pour décrire les contingents militaires d’origine turque dans
l’armée byzantine. On peut supposer que les bandes de guerre turques qui vivaient en
Asie Mineure vers le XIe siècle pourraient être considérées comme le prototype des
contingents turcopoles. Leurs chefs, comme Elkhanes dans Alexiad, pourraient faire
défection à Byzance et ces contingents (qui étaient autrefois des gangs d’hommes
libres vivant par l’épée) pourraient faire partie de l’armée byzantine.
308
Au XIIe siècle, les unités Turcopole apparaissent également dans les armées
croisées. Les croisés ont d’abord rencontré ces guerriers dans les rangs des armées
byzantines. En 1101, il y avait 500 turcopoles dans l’armée de Raymond Saint-
Gilles. Ces guerriers ont été présentés en cadeau par Alexis Komnenos au noble
franc. Anna Komnene parle du destin de ce contingent turc, sans utiliser le terme
Turcopole : une grande partie de ces soldats ont été massacrés en passant la province
d’Armenikon. Le commandant des Turcopoles était un homme appelé Tzitas
(Τζιτασ) qui était probablement un Turc.
Ces forces sont organisées de manière autonome, sous la direction d’un
Turcopolier qui commande les Turcs. Néanmoins, au cours des décennies suivantes,
les États croisés ont commencé à recruter leurs propres forces turcopoles au Levant.
Ils servaient généralement comme unités de cavalerie légère. Cependant, la majorité
des Turcopoles dans les forces des croisés n’étaient pas turques mais d’origine arabe
(musulmane ou chrétienne). Bien que les guerriers d’origine musulmane se soient
généralement convertis au christianisme, il existe des exemples plus rares qui
démontrent la présence de musulmans dans les armées croisées. Enfin, ces
Turcopoles du Levant ne semblent pas avoir une identité ethnique distincte et ce sont
essentiellement des guerriers indigènes combattant dans les armées croisées. D’autre
part, il faut remarquer qu’il y avait aussi des troupes indigènes “non-turcopoles”
combattant pour les croisés.
Bien que son usage vernaculaire soit certain, comme il a été démontré par son
passage au latin comme emprunt au grec ; le terme Turcopole n’est attesté dans les
sources byzantines qu’au XIVe siècle : il fut d’abord utilisé par Georges Pachymère.
Cependant, son usage comme patronyme est attesté dès le XIe siècle : il existe un
certain Sergios le Turcopole attesté en 1082.
309
Mixobarbaros était généralement traduit par semi-barbare, mais cette traduction
ne révèle pas le sens nuancé du mot. Contrairement au terme Turcopole qui semble
apparaître dans le dernier quart du XIe siècle, mixobarbaros est un terme plus ancien.
Elle est attestée dans les textes grecs depuis l’Antiquité. Alexander Kazhdan, dans
son article de l’Oxford Dictionary of Byzantium, cite Hésychios d’Alexandrie (c. VIe
siècle) en déclarant “les hommes qui n’étaient ni Hellènes, ni barbares, mais avaient
les qualités des deux” . Cette citation pourrait être un point de départ : les termes
“mixobarbaros” et “tourkopoulos” avaient le sens de mélange ethnique. Comme on
l’a vu précédemment, Michel Attaleiates employait ce terme pour désigner les
populations de la frontière danubienne. À la fin du Moyen Âge, ce terme est utilisé
avec une référence particulière à cette région. Hélene Ahrweiler explique ce fait ainsi
: “Le terme Μιξοβάρβαροι fait référence à des enjeux culturels, et est utilisé pour
ceux qui ont filtré sur le Danube et dont le mode de vie nomade interagissait avec les
traditions sédentaires. Cependant, les termes Μιξέλληνες et Μιξοβάρβαροι utilisés
par les auteurs byzantins des XIe et XIIe siècles doivent être étudiés en relation avec
la pratique des mariages mixtes dans cette région, qui était habitée par des groupes
nomades christianisés”.
Les mariages interconfessionnels semblent être courants à la fin du Moyen Âge en
Asie Mineure. Cependant, la grande majorité de ces unions doivent se faire entre
hommes musulmans et femmes chrétiennes car la loi musulmane interdit strictement
les mariages entre femmes musulmanes et hommes chrétiens. Cependant, il existe
des exemples en Asie Mineure où ces restrictions ont été enfreintes ; Abu’l Fida, le
géographe arabe du XIVe siècle déclare que dans la ville de Melitene, les femmes
musulmanes épousent des hommes chrétiens De plus, il semble que dans les
mariages interconfessionnels –du moins certains–, les filles pourraient être
310
considérées comme appartenant à la religion de leur mère ou au moins libres de
choisir leur identité religieuse.
Il existe d’autres termes similaires utilisés pour désigner des situations
spécifiques : par exemple, le terme Gasmuloi (Γασμουλοı) est employé pour les
Byzantins latinisés ou la progéniture de mariages gréco-latins (principalement
italiens). Semblable à Turcopole, les Gasmuloi n’étaient pas seulement une
dénomination ethnique, mais cela impliquait des contingents militaires qui étaient
employés à la fois dans l’armée de terre et dans la marine, principalement au XIVe
siècle.
Parmi ces termes, la Turcopole était le terme qui traduisait un caractère ethnique
plus marqué. Il a toujours été utilisé pour les mercenaires turcs de l’Est, ceux qui
semblent apparentés aux Seldjoukides.
Idéalement, l’intégration réussie des populations turques dans la société byzantine
pourrait s’apparenter à l’intégration des Slaves dans les Balkans aux VIIIe et IXe
siècles. Ces Turcs doivent avoir été évangélisés et ne doivent pas avoir conservé une
structure politique autonome. Ils ne doivent donc pas avoir d’archontes ou
d’ethnarques. Les séquelles de Petchénègue, Coumans et Oghuz dans les Balkans
pourraient être considérées comme une intégration réussie. Après leur
assujettissement aux autorités byzantines, ils ont continué à exister en tant que
communautés rurales, la plupart avec une identité ethnique distincte, mais ils ont
perdu leur vitalité politique. Ils seront assimilés dans les siècles suivants,
principalement par les Bulgares et les Grecs de la région. Il convient également de
noter qui les établissements ruraux, tels que ceux où Petchénègue et Coumans se sont
installés dans les Balkans, constituaient de petites unités sociales probablement aux
influences culturelles du centre byzantin. Au milieu du XIVe siècle, John Vatatzes a
311
donné des terres arables aux soldats Couman en Asie Mineure et ces soldats y ont
formé des colonies militaires. Une telle colonisation militaire est une ancienne
coutume de l’administration romaine depuis l’Antiquité. La formation de telles
colonies militaires était avantageuse à la fois pour l’État et les «étrangers» pour
plusieurs raisons. Selon Mark Bartusis, de telles colonies pourraient répondre aux
besoins sociaux des Coumans (et d’autres peuples tels que les Tzakones) et les rendre
plus faciles à administrer. De plus, ils y préservaient aussi leur organisation sociale.
Dans ce chapitre, les schémas essentiels de l’entrée des Turcs dans le service
byzantin, le potentiel d’acculturation et d’assimilation de Byzance et la possibilité
d’intégration ont été discutés. Je vais résumer mes conclusions ainsi :
1. L’entrée de différents groupes ou individus appartenant aux peuples turciques
dans le service byzantin est un phénomène fréquent dans les XIe et XIIe siècles. Il
doit être considéré comme un continuum de relations traditionnelles entre l’Empire
byzantin et les populations nomades turques dans le passé. Les individus d’origines
sociales différentes, à la fois des sociétés seldjoukide et péchenègue, sont entrés au
service impérial.
2. Dès le début, le monde byzantin représentait “une vie meilleure” pour les
Turcs. La première expansion turque en Asie occidentale n’était pas motivée par le
zèle religieux ou un projet politique bien défini, mais le désir des seigneurs de la
guerre de piller ou de s’emparer de nouveaux territoires dans une terre relativement
civilisée et prospère, mais chaotique. Bien que l’armée seldjoukide ait vaincu les
Byzantins à Manzikert, Alp Arslan n’a pas demandé de territoires en Asie Mineure à
l’exception de trois châteaux dans la région frontalière. De plus, la grande stratégie
des Seldjoukides pour l’expansion vers l’ouest était la réunification de l’oikoumene
musulman qui était divisé en différents califats et sultanats depuis le début de la
312
période abbasside. Ils ne cherchent pas l’annexion de l’Asie Mineure byzantine, mais
ils visaient à vaincre les Fatimides d’Égypte et à envahir la Palestine fatimide et
peut-être l’Égypte. La campagne d’Atsiz en Syrie et en Palestine et sa prise de la
ville de Jérusalem (1073) aux Fatimides témoignent de ce projet. Ainsi, l’invasion de
l’Asie Mineure par les Turcs a été en grande partie un événement centrifuge et
spontané réalisé par des bandes de guerre indépendantes.
3. A l’époque Comnène, un nouveau projet impérial est lancé : la conversion des
Turcs (et d’autres peuples musulmans de l’Est) et la formation d’un nouveau modèle
de supériorité politique sans nécessairement la domination militaire. Manuel Ier
Comnène s’est particulièrement consacré à ce projet. Cependant, ce projet se termina
par un échec, non seulement parce que le temps des troubles commença dans les
années 1180 et plus tard la prise de contrôle latine de Constantinople en 1204, mais
aussi à cause de la réticence des autorités religieuses de Byzance à assouplir les
dogmes théologiques du christianisme orthodoxe. Cette réticence ne s’expliquait pas
seulement par une approche fanatique d’une adhésion abstraite aux fondements
religieux, mais aussi par le manque de volonté d’inclusion de ces nouveaux venus
dans leur société.
4. L’entrée des Turcs dans le service byzantin a chuté en grande partie après le
milieu du XIVe siècle. Ce fait pourrait s’expliquer par le déclin de l’Empire byzantin
en tant que centre politique et spirituel et son manque de pouvoir pour attirer de
nouvelles populations étrangères. En utilisant les mots de Charles M. Brand, je peux
dire que Byzance a perdu son "pouvoir d’attirer et d’absorber." Pendant ce temps, les
Ottomans ont commencé à s’étendre dans les territoires centraux Empire Byzantin.
La controverse hésychaste et la guerre civile byzantine (1341-1347) ont encore
contribué à la dissolution socio-politique de Byzance.
313
LES AUTRES ASPECTS DE LA REPRÉSENTATION DES TURCS
Dans ce chapitre, on discute les autres aspects de la représentation des Turcs: la
guerre, les femmes, les comportements sexuels et la violence extrême. Les Byzantins
ont rencontré les Turcs, premièrement dans les champs de bataille. La littérature
militaire byzantine est important, du point de vue ethnographique, pour les peuples
étrangers. Donc, dans les manuels de guerres Byzantines, il y a des riches matériaux
en regard de l’image des Turcs. Dans le manuel de guerre attribué à l’empereur
Maurice, Strategikon, les Turcs (le Khaganat Türk) sont présentés comme une des
nations scythes. Selon l’auteur, les Turcs sont une nation peuplée et indépendante, ils
combattent avec l’armure, épées, arcs et lances. Ces gens réussissent à la guerre mais
sont moralement faibles. Ils peuvent être trompés par des cadeaux et rompre leurs
voeux.
Le même narratif sur les Turcs est employé aussi dans le Taktika, attribué à Leo
IVe. Une caractéristique intéressante de ce texte est le suivant : l’auteur utilise
l’ancien récit et parle des Turcs, mais cependant il s’agit des Hongrois. Il y a aussi
les manuels militaires plus tardifs comme De Velitatione qui est attribué à Nicéphore
Phocas et un autre Taktika attribué à Nicéphore Ouranos, mais dans ces oeuvres les
digressions ethnographiques concernant les Turcs sont plus rares. Donc, la
représentation byzantine de la guerre turque reflète la perception byzantine du
caractère national des Turcs. Ils étaient représentés comme des gens infidèles,
traîtres, sauvages, rusés, trompeurs, mais indépendants. Leur élément original dans la
guerre était le faux retrait. Cette tactique était également conforme au “caractère
national” des Turcs, car elle représentait selon les textes byzantins, l’infidélité
générale et le manque de moralité des Turcs. De plus, les armées turques étaient
314
représentées comme les armées d’archers à cheval jusqu’au XIVe siècle, mais à
mesure que la guerre turque évoluait, les auteurs ultérieurs avaient renoncé à répéter
cette description habituelle.
Dans la littérature byzantine, il y a très peu des références aux femmes turques.
En fait, cela est compréhensible, car dans cette littérature il y a très peu de références
aux femmes étrangères en général. Dans ce cas, la ségrégation spatiale des hommes
et des femmes, qui est courante dans les sociétés musulmanes, peut également jouer
un rôle. Mais les Turcs du Moyen Âge, même après qu’ils soient devenus
musulmans, ne peuvent être considérés comme des sociétés dans lesquelles ces
règles islamiques ont été appliquées strictement. Avant l’Islam, lorsque les Turcs
étaient encore un peuple païen, il n’il n’y avait pas de tels tabous dans le domaine de
la vie des femmes et des hommes dans la société. Le carnet de voyage d’Ibn Fadlan
le montre clairement. L’évolution après l’islamisation est également restée relative.
On voit que les femmes de la classe supérieure jouaient un rôle plus important dans
la vie publique à l’époque seldjoukide que dans les États musulmans contemporains.
Finalement, on peut supposer qu’il y a un certain changement dans le statut des
femmes turques après la conversion à l’İslam. Les Turcs ont progressivement adopté
les normes sociales des sociétés musulmanes qui avaient institutionnalisé la
ségrégation des sexes, l’obéissance aux hommes et la pudeur vestimentaire. Ces
normes sociales de la société islamique limitaient la vie publique des femmes.
Néanmoins, ce changement a été progressif et n’a pas affecté certaines couches de la
société, en particulier celles qui ont continué le mode de vie nomade de leurs
ancêtres. La représentation des femmes guerrières dans les récits épiques turcs au
Moyen Âge démontre également que l’ethos ancien a vécu parmi les Turcs pendant
très longtemps. Même dans le sultanat seldjoukide d’Iran où la culture musulmane
315
persane a été progressivement adoptée par la dynastie au pouvoir, les femmes de
cour étaient actives dans la vie publique. Dans les textes étudiés, on voit un certain
nombre des femmes turques, elles ne sont généralement pas nommées. Si l’on
considère l’abondance de références à des femmes turques médiévales dans les
chroniques persanes et syriaques, les représentations vivantes de femmes turques de
la classe supérieure dans les hagiographies, et les ambitieux projets de mécénat des
femmes nobles seldjoukides, ce fait est contradictoire. Donc ce fait peut-être
expliqué par le manque d’intérêt des auteurs byzantins.
Un élément significatif de l’image barbare est son manque de modération dans
divers aspects de la vie. Le concept de modération est étroitement lié aux populations
civilisées, et les barbares sont notoirement intempérants dans la vie. Ainsi, le
comportement sexuel est un aspect de la vie quotidienne dans lequel la modération et
l’intempérance pourraient jouer un rôle distinctif entre le civilisé et le barbare.
Depuis Hérodote, la représentation de la sexualité barbare est à la fois la
démonstration de l’étrangeté et de l’infériorité par rapport à la civilisée.
Cependant, il est nécessaire de contextualiser cette séparation à une époque où la
sexualité était perçue différemment de nos normes contemporaines, formées
principalement au XIXe siècle, en particulier à l’époque victorienne. L’Histoire de la
Sexualité (1976) de Michel Foucault a lancé une vague de discussions sur la question
de savoir si les éléments apparemment extrêmes de la sexualité étaient réprimés ou
non. Dans ce sous-chapitre, je traiterai de l’attribution aux Turcs du comportement
sexuel des auteurs byzantins, qu’ils jugent pervers ou déviant. En effet, les Turcs ont
été identifiés avec de tels comportements dans certains textes byzantins ultérieurs.
L’homosexualité, ou plutôt la sodomie, est associée aux Turcs dans les sources
byzantines. Ces pratiques attribuées aux Turcs sont en fait une répétition de certaines
316
de leurs accusations contre les Arabes, bien plus anciennes et plus courantes dans la
littérature byzantine. Comme l’identité musulmane des Turcs est devenue évidente
au fil des siècles, ces accusations ont également été utilisées contre eux. Ces
références se multiplient au XIVe siècle et la question de la sodomie devient l’objet
de polémiques même théologiques. Le milieu du XIVe siècle, en particulier, est une
période de troubles idéologiques dans le monde byzantin en déclin. L’apparition du
mouvement de l’hésychasme et les réactions contre lui, les guerres civiles byzantines
et la fragmentation pro-turque et pro-latine de la société byzantine démontrent
clairement une profonde crise idéologique. Nicéphore Gregoras, qui était un fervent
partisan de la faction unioniste pro-latine, accuse son rival Gregory Palamas d’avoir
été impliqué dans la sodomie avec des Turcs alors qu’il était devenu captif d’eux.
Malgré le théologien hésychaste qui était tombé en captivité en 1354, après l’apogée
de la controverse de l’hésychasme, l’inquiétude dans la société était toujours présent.
Comme l’a exprimé Charis Messis, les accusations de sodomie visaient généralement
les membres de la faction pro-turque de la société byzantine tardive. Dans ce
contexte, les accusations de sodomie présumée ont contribué à dénigrer les personnes
qui soutenaient des groupes pro-turcs ou anti-latins.
Gregoras, qui décrit Palamas comme un partisan de l’impiété, raconte ce qui est
arrivé à son adversaire de manière obscène : Palamas est capturé par des pirates et
conduit au fils aîné du satrape Hyrcanos (Orhan). Là, il est moqué, dépouillé de ses
vêtements, fouetté et violé. Il est difficile de comprendre si le récit de Gregoras sur
Palamas était entièrement imaginé ou bien construit sur une rumeur à
Constantinople. On pourrait supposer que ces récits établissent une dichotomie entre
les Grecs chrétiens et les musulmans, considérant que les valeurs chrétiennes telles
que la chasteté et la vertu sont les normes, et que les barbares ont tendance à être des
317
pervers. De plus, on peut conclure qu’il existe une corrélation entre la violence
martiale et la violence sexuelle puisque les deux semblent être l’expression d’une
masculinité incontrôlable et d’une barbarie indomptée selon les Byzantins. Par
conséquent, l’invasion turque ne vise pas seulement l’espace de vie des Byzantins,
mais constitue également une menace pour leur corps.
La déshumanisation est un autre aspect frappant de la représentation byzantine
des Turcs. Il y a des récits dans l’historiographie byzantine qui mettent l’accent sur la
laideur physique, disent que la nourriture que les Turcs mangent n’est pas propre et
comestible, et qui attribuent un comportement sadique à ce peuple, surtout pendant la
guerre. De plus, ils les accusent même de sacrifice humain. Les références
susmentionnées à la sodomie doivent être comprises aussi comme un aspect de cette
représentation qui dépeint les Turcs comme des monstres aux limites de l’humanité.
Comme nous l’avons déjà mentionné dans l’introduction, notre objectif n’st pas
de savoir si les données de ces travaux historiographiques sont empiriquement vraies
ou fausses. Le sacrifice humain était autrefois une pratique courante dans différentes
parties du monde. Depuis l’époque d’Hérodote, les Scythes et d’autres peuples des
steppes étaient associés au sacrifice humain. Les Turcs d’Anatolie pratiquaient peutêtre
la coutume du sacrifice humain, dans leur nouvelle patrie, même si des siècles se
sont écoulés depuis leur conversion à l’islam. On ne spéculera pas ci-dessus. Dans
cette thèse on s’intéresse plutôt à la mentalité byzantine qui se cache derrière ces
recits. Ces passages démontrent qu’il y a un discours déshumanisant contre les
populations barbares. Le discours sur l’agressivité sexuelle et les représentations
physiques en tant que monstres ne pouvait être imaginé en dehors de cela.
Cependant, les arguments comme le sacrifice humain ou le cannibalisme sont les
arguments ultimes pour contrarier et déshumaniser une population étrangère.
318
La représentation de la guerre pouvait être à la fois réaliste et bourrée de clichés.
En effet, il existe une ancienne tradition d’écriture militaire, centrée sur les
techniques de guerre. Cependant, les champs de bataille étaient aussi les lieux où se
produisait une rencontre entre les Byzantins et les étrangers. Certains des auteurs ont
été témoins des batailles contre les envahisseurs turcs, mais le matériel ancien est
également largement utilisé dans les récits de guerre sur les peuples nomades.
Comme il a déjà été mentionné, les femmes turques sont presque invisibles dans
l’historiographie byzantine. La représentation de la sexualité chez les Turcs a
également un point paradoxal similaire. Plusieurs sources byzantines représentent à
la fois les roturiers turcs et les dirigeants ottomans comme des pédérastes. D’une
part, cette représentation est un élément de la polémique entre les factions proturques
et anti-turques dans la société byzantine. Néanmoins, les Turcs qui avaient
une culture et un mode de vie très différents étaient probablement perçus comme des
pervers par certains Byzantins. Il est également vrai que le mode de représentation
mentionné ci-dessus trouve ses racines dans la représentation traditionnelle des
Arabes relevée dans les textes byzantins. On peut donc dire que les Byzantins
reflétaient dans une certaine mesure leurs impressions et leurs idées à propos des
Arabes sur les Turcs. On peut, en outre, dire que la littérature sur la guerre turque est
la partie la plus réaliste des récits byzantins sur les Turcs, mais les récits sur la
sexualité sont probablement moins réalistes. Car le premier peut être considéré
comme le reflet de ce que l’on rencontre réellement sur le champ de bataille et le
second est surtout le reflet d’un topos littéraire.
En conclusion, la représentation byzantine des Turcs est un amalgame de
témoignage et de fiction. Il est très difficile de distinguer la soi-disant réalité
319
historique de la fiction. L’unique façon de trouver une solution est de lire ces textes
en fonction des prétendues visées des auteurs.
CONCLUSION
Dans cette section de conclusion, je résume mes découvertes. Ma première
découverte importante est que la formation d’une image collective d’un peuple ou
d’un groupe ethnique est étroitement liée à la mémoire culturelle. La différence entre
les images des Turcs du Nord (c’est-à-dire les Scythes) et des Turcs orientaux (c’està-
dire les Persans) résultait de la situation géographique, qui était à la base de la
nomenclature byzantine des peuples étrangers, et du point de vue aristotélicien qui
corrèle la région géographique dans laquelle les gens vivent avec leur caractère
national. Cependant, ces deux points n’ont de sens qu’avec la présence de la
mémoire culturelle.
La littérature byzantine n’avait aucun rapport direct avec ce que l’on pourrait
appeler la “réalité objective”. Le concept de base de l’historiographie byzantine était
la représentation. L’historiographie byzantine ne peut donc être comprise qu’à
travers cette notion. Les auteurs byzantins suivaient les modèles grecs et romains et
nourrissaient une image déjà créée des Scythes ou des Perses. Cette image était un
logos et n’était pas une image remplie par un homme byzantin ordinaire, mais elle
avait une continuité et une fonction idéologique. Cette fonction idéologique était le
rôle du logos dans l’altérité. Dans la vision du monde byzantine, les peuples
turciques du nord et de l’est étaient présentés comme les autres aux Byzantins.
Cependant, la vie réelle ne suit souvent pas les topoï littéraires ; à commencer par
Michael Attaleiates, les auteurs byzantins ont rencontré des individus turcs
320
d’horizons différents. Les auteurs de la région frontalière, tels que Nicetas Khoniates
ou Doukas, ont connu l’agonie de perdre leurs terres ancestrales. Un grand
théologien tel que Grégoire Palamas a vécu la vie à la cour ottomane en tant que
captif.
Ainsi, les auteurs byzantins ont écrit leurs expériences personnelles avec les
Turcs. Ces récits n’étaient ni sans but idéologique ni indépendants de la littérature
antérieure sur les peuples turciques. On peut dire que toute la littérature byzantine
concernant les peuples turciques était un dialogue avec les anciens maîtres. Les
auteurs qui ont contribué à cette littérature n’étaient pas des hommes de lettres vivant
de l’écriture ; ils étaient des bureaucrates ou des hommes politiques au sens large.
Leurs objectifs et opinions politiques ont inévitablement affecté leur approche des
Turcs. Au cours des derniers siècles de Byzance (c. 1350-1453), l’axe essentiel de la
politique byzantine consistait à prendre position entre les Turcs et les Occidentaux.
Ainsi, à cette époque, toute la vie politique byzantine prend sens selon cet axe.
Il existe une variable clé dans la représentation des Turcs dans l’historiographie
byzantine - le rôle de l’islam. Entre le XIe et le XVe siècle, le poids de la religion
musulmane s’est visiblement accru dans les textes byzantins, et il a façonné l’image
des Turcs, coïncidant avec l’association des Turcs avec le peuple persan. De plus,
l’Islam a réuni la représentation contemporaine des Turcs et la mémoire culturelle
des invasions arabes. La nature "scythe" des Turcs a été oubliée par certains, mais
elle a été soulignée par un ethnographe-historien comme Chalcondyle.
Je peux expliquer la formation de l’image des Turcs dans les textes byzantins
dans cinq strates:
321
1. La première strate est la mémoire culturelle des Scythes et des Perses.
L’ancienne représentation des deux nations a joué un rôle important dans la
formation de l’image des Turcs. La généalogie de cette représentation remonte à
l’époque d’Hérodote. Les images formées autour de ces deux signifiants
représentaient deux niveaux différents d’antagonisme pour les Byzantins. Les Perses
à l’est représentaient l’antagonisme primaire, et les Scythes au nord représentaient
l’antagonisme secondaire.
2. La deuxième strate est la mémoire culturelle des Arabes et la première invasion
musulmane. Les associations des Turcs avec les Agarénes et les Ismaélites sont des
manifestations typiques de l’utilisation de cette mémoire culturelle pour les Turcs.
Les traces de ce motif ont été vues d’abord dans l’Alexiade.
3. La troisième strate est constituée des vestiges de la mémoire culturelle des
rencontres byzantines avec les “Turcs du Nord” dans les régions du Danube et peutêtre
de la Crimée. Cette mémoire culturelle doit être considérée comme une
continuation des réminiscences des Huns de l’Antiquité tardive. Depuis l’époque de
Constantine Porphyrogenitus, certaines connaissances sur ces populations existaient,
et les rencontres du XIe siècle avec les Petchénègues ont ajouté de nouvelles
informations à cette strate basées sur les expériences de terrain.
4. La quatrième strate est la littérature byzantine concernant les Turcs
seldjoukides, à savoir les oeuvres d’Anna Comnène, Jean Cinnamus, Nicetas
Choniates et George Acropolite. Les oeuvres littéraires produites au cours de la
période nous fournissent des détails non seulement sur l’histoire des relations entre
322
les seldjoukides et les byzantines, mais aussi sur les problèmes internes des
Seldjoukides et d’autres entités turques d’Asie Mineure. L’héritage des oeuvres de
cette période est également formateur quant à l’image des Turcs dans la littérature
byzantine.
5. La cinquième strate est le corpus concernant les Ottomans. Ce corpus
commence par l’oeuvre de Georges Pachymère, dans laquelle on peut voir un noyau
de l’État ottoman, même s’il s’agissait encore d’une autre des bandes de guerre
turques en Bithynie, et le phénomène de "la montée des Ottomans" n’avait pas
encore eu lieu. . Des auteurs tels que Pachymère et Chalcondyle représentaient les
dernières générations de l’historiographie byzantine, et leurs représentations des
Ottomans étaient basées sur une expérience personnelle et souvent biaisées en raison
des objectifs politiques de leur paternité. Les Turcs n’étaient plus des nouveaux
venus ; depuis le XIe siècle, il y avait eu une coexistence turco-grecque en Asie
Mineure. Outre l’attitude biaisée des auteurs susmentionnés, leur représentation des
Ottomans avait des traces de strates antérieures.
En se concentrant sur ces découvertes sur le processus de formation de l’image
des Turcs, on pourrait trouver quelques modèles essentiels :
- L’association des Turcs aux traits persans a plusieurs sources :
géographiquement, le Grand Sultanat seldjoukide s’est formé en Iran et au Khorasan.
Dès lors, les Turcs seldjoukides se sont emparés de la mémoire historique de
l’Empire perse, du fait de leur situation géographique, sans être ethniquement
persans.
La persanisation des Turcs signifiait devenir antagonistes a l’Empire byzantin, ce
qui a deux dimensions à souligner : le premier est les invasions turques et les
323
croisades après la montée de l’islam - en particulier après le XIe siècle. Il y avait un
sentiment croissant sur un antagonisme religieux entre le christianisme et l’islam, et
l’identité religieuse des Turcs en Asie Mineure est devenue plus visible pour les
Byzantins. La prise de conscience accrue a provoqué l’émergence d’idées telles que
le projet Comnène concernant l’évangélisation des Turcs. L’association des Turcs
avec les notions d’Ismaelitai et d’Agarenoi est devenue fréquente, et la différence
entre les Turcs scythes et persans était maintenant un sujet de discussion. Lorsque les
Turcs ont commencé à être identifiés aux Perses, ils ont été accablés par les topoï
associés aux Perses, comme les perversions sexuelles.
Il faut aussi ajouter qu’il y a toujours des Scythes quelque part. Si un groupe de
personnes perd les aspects socioculturels qui le rendent “scythe”, une nouvelle
population pourrait devenir les nouveaux Scythes. La région steppique associée à ces
peuples est une zone très turbulente et culturellement hétérogène. Chaque fois, un
nouveau groupe de nomades venu d’une partie obscure de l’Asie intérieure pouvait
devenir les nouveaux Scythes, comme on l’a vu dans le cas des Pechenegs, Coumans
et Mongols.
L’historiographie du XIVe siècle démontre l’amère agonie des Byzantins
d’accepter la domination des Ottomans et d’autres émirats turcs. Les textes écrits
dans ce contexte pourraient être considérés comme des récits qui démontrent plus
directement le but de leurs auteurs, comme dans le cas de Doukas et Chalcondyle,
car ils se concentraient au XVe siècle. La représentation byzantine des Turcs comme
“les autres” s’est transformée en un récit d’auto-victimisation (dans l’oeuvre de
Doukas) et d’acceptation de la domination des Turcs (dans l’oeuvre de Chalcondyle).
Enfin, l’ethos frontalier et la culture martiale des Byzantins ont disparu, et l’ethos
militaire de la société a progressivement diminué après le XIIIe siècle. Un ethos
324
frontalier n’a de fonction que dans une société où il y a un espace pacifique au centre
et une zone militairement active à la frontière. À la fin du XIIIe siècle, presque tout
le territoire byzantin est devenu la cible des Turcs et d’autres nations, et l’Empire
Byzantin s’est de plus en plus basé sur des mercenaires jusqu’à ce qu’il n’y ait
finalement plus de différence entre la frontière et le centre.
325
REFERENCES
PRIMARY SOURCES
Agathias. (1975). The Histories (J. Frendo, Ed.). Berlin: De Gruyter.
Aḥmad ibn Yaḥyā ibn Jabir al-Balādhurī. (1916). The origins of the Islamic state :
being a translation from the Arabic, accompanied with annotations, geographic
and historic notes of the Kitâb futûḥ al-buldân of al-Imâm Abu-l ʻAbbâs, Aḥmad
ibn-Jâbir al-Balâdhuri (P. K. Hitti, Trans.). New York: Columbia University
Press.
Akropolites, G. (1978). Georgii Acropolitae Opera (A. Heisenberg & P. Wirth, Eds.;
2 vols.). Stuttgart: B.G. Teubner.
Akropolites, G. (2007). The History (R. J. Macrides, Trans.). New York: Oxford
University Press.
Anonim Selçukname. (2014). (F. Coşguner & H. İ. Gök, Eds.). Ankara: Atıf.
Arrian. (2013). Alexander the Great: Anabasis and Indica (M. Hammond, Trans.).
Oxford: Oxford University Press.
Aşıkpaşazade. (2007). Tevarih-i Al-i Osman (K. Yavuz & M. A. Y. Saraç, Eds.).
İstanbul: Gökkubbe.
Attaleiates, M. (2012). The History (A. Kaldellis & D. Krallis, Ed. & Trans.).
Cambridge, MA: Harvard University Press.
Battalname. (1996). (Y. Dedes, Ed.). Cambridge, MA: Harvard University Press.
Bertrandon de la Broquière. (2010). Le voyage d’Orient (H. Basso & J. Paviot, Eds.).
Toulouse: Anacharsis.
Constantine Porphyrogenitus. (1967). De Administrando Imperio (G. Moravcsik,
Ed.; R. H. Jenkins, Trans.). Washington, D.C.: Dumbarton Oaks Center of
Byzantine Studies.
Cusset, C., & Salamon, G. (Eds.). (2008). À la rencontre de l'étranger: L'image de
l'Autre chez les Anciens. Paris: Les Belles Lettres.
Digenes Akrites. (1956). (J. Mavrogordato, Trans.). Oxford: Clarendon Press.
Doukas. (1958). Historia Turcobyzantina (1341-1462) (V. Grecu, Ed.). Bucharest:
Editura Academiei Republicii Populare Romîne.
Doukas. (1975). Decline and fall of Byzantium to the Ottoman Turks (H. J.
Magoulias, Trans.). Detroit: Wayne State University.
326
Enverî. (1954). Le Destan d’Umur Pacha (Düsturname-i Enveri) (I. Mélikoff, Ed. &
Trans.). Paris: PUF.
Enverî. (2003). Fatih devri kaynaklarından Düsturname-i Enverî: Osmanlı tarihi
kısmı, 1299-1465 (N. Öztürk, Ed.). Istanbul: Kitabevi.
Fontes Historiae Daco-Romanae. (1964). 4 vols. Bucharest: Academiei Republicii
Socialiste România.
Gregoras, N. (1829-1855). Nicephori Gregorae Byzantina historia (L. Schopen & I.
Bekker, Eds.; 3 vols.). Bonn: Weber.
Gregoras, N. (1973-2000). Rhomäische Geschichte (J. L. Van Dieten, Trans.;
6 vols.). Stuttgart: Anton Hiersemann.
Ibn Battuta. (1958). The travels of Ibn Battuta (H. A. R. Gibb, Trans.). London:
Hakluyt.
Ibn Bibi. (1996). el-Evâmirü'l-Alâiyye fi'l-umûri'l-Alâiyye (M. Öztürk, Ed. & Trans.;
vol. 1). Ankara: Kültür Bakanlığı.
Ibn Fadlan. (1988). Voyage chez les Bulgares de Volga (M. Canard, Trans.). Paris:
Sindbad.
Jordanes. (1915). Gothic history (C. C. Mierow, Ed. & Trans.). Princeton: Princeton
University Press.
Kaminiates, J. (2000). The capture of Thessaloniki (D. Frendo & A. Fotiou, Eds.; G.
Böhlig, Trans.). Perth: Australian Association for Byzantine Studies.
Kantakouzenos, J. (1828). Ioannis Cantacuzeni eximperatoris Historiarum (L.
Schopen, Ed.; 3 vols.). Bonn: Weber.
Kantakouzenos, J. (1982-2001). Geschichte (G. Fatouros & T. Krischer, Trans.; 3
vols.). Stuttgart: Anton Hiersemann.
Kaşgarlı Mahmud. (2005). Divanü Lügati’t Türk (S. Erdi & S. T. Yurteser, Eds.).
İstanbul: Kabalcı.
Keykavus. (2007). Kabusname (O. Ş. Gökyay, Ed.; M. Ahmed, Trans.). Istanbul:
Kabalcı.
Khoniates, N. (1975). Nicetae Choniatae Historia (J. L. Van Dieten, Ed.). Berlin: De
Gruyter.
Khoniates, N. (1984). O city of Byzantium: Annals of Niketas Khoniates (H. J.
Magoulias, Trans.). Ohio: Wayne State University Press.
Kinnamos, J. (1836). Epitome rerum ab Ioanne et Alexio [sic] Comnenis gestarum
(A. Meineke, Ed.). Bonn: Weber.
327
Kinnamos, J. (1976). Deeds of John and Manuel Komnenos (C. M. Brand, Trans.).
New York: Columbia University Press.
Komnene, A. (2001). Annae Comnenae Alexias (D. Reinsch & A. Kambylis, Eds.).
Berlin: De Gruyter.
Komnene, A. (2009). The Alexiad (P. Frankopan, Ed.; E. R. A. Sewter, Trans.). New
York: Penguin.
Kritovoulos. (1954). History of Mehmed the Conqueror (C. T. Riggs, Trans.).
Princeton: Princeton University Press.
Kritovoulos. (1983). Critobuli Imbriotae Historiae (D. Reinsch, Ed.). Berlin: De
Gruyter.
Kritovoulos. (2012). Kritovulos Tarihi (1451-1467) (A. Çokona, Trans.). Istanbul:
Heyamola.
Maurice’s Strategikon – Handbook of Byzantine military strategy. (1984). (G. T.
Dennis, Ed. & Trans.). Philadelphia: University of Pennsylvania Press.
Matthew of Edessa. (1858). Chronique de Mathieu d'Édesse (E. Dulaurier, Trans.).
Paris: A. Durand.
Mélikoff, I. (Ed.). (1960). La Geste de Melik Dānişmend: Étude critique du
Dānişmendnāme. 2 vols. Paris: Librairie Adrien Maisonneuve.
Oruç Bey. (2011). Oruç Beğ Tarihi. In N. Atsız (Ed.), Üç Osmanlı Tarihi (Oruç Beğ-
Ahmedi-Şükrullah) (pp. 11-127). Istanbul: Ötüken.
Pakhymeres, G. (1984-2000). Relations historiques (A. Failler & V. Laurent, Eds.
&Trans.; 5 vols.). Paris: Les Belles Lettres.
Photios. (1982). Photii Patriarchae Lexicon: A-D (C. Theodoridis, Ed.; Vol. 1).
Berlin: De Gruyter.
Phillippides-Braat, A. (Ed. & Trans.). (1979). La captivité de Palamas chez les
Turcs: Dossier et commentaire. Travaux et Mémoires, 7, 109-221.
Procopius of Caesarea. (1914). History of the wars, Vol. 1: Books 1-2 (Persian War)
(H. B. Dewing, Ed. & Trans.). Cambridge, MA: Harvard University Press.
Procopius of Caesarea. (1916). History of the wars, Vol. 2: Books 3-4 (Vandalic
War) (H. B. Dewing, Ed. & Trans.). Cambridge, MA: Harvard University Press.
Russell, N. (Trans.). (2020). Gregory Palamas: The Hesychast controversy and the
debate with Islam. Documents relating to Gregory Palamas. Liverpool: Liverpool
University Press.
328
Skylitzes, J. (1973). Ioannis Scylitzae Synopsis historiarum (H. Thurn, Ed.). Berlin:
De Gruyter.
Skylitzes, J. (2010). A Synopsis of Byzantine History, 811-1057 (J. Wortley, Trans.).
Cambridge: Cambridge University Press.
Sphrantzes, G. (1980). The fall of the Byzantine Empire. A Chronicle by George
Sphrantzes (M. Philippides, Trans.). Amherst: University of Massachusetts Press.
Sphrantzes, G. (1990). Georgii Sphrantzae Chronicon (R. Maisano, Ed.). Rome:
Academia Nazionale dei Lincei.
The Taktika of Leo VI. (2010). (G. T. Dennis, Ed. & Trans.). Washington, D.C.:
Dumbarton Oaks Resarch Library and Collection.
Theophanes. (1997). The Chronicle of Theophanes Confessor - Byzantine and Near
Eastern history AD 284-813 (C. Mango & R. Scott, Ed. & Trans.). Oxford:
Clarendon Press.
Le Traité Sur La Guérilla de l’empereur Nicéphore Phocas (De Velitatione). (2011).
(G. Dagron, H. Mihăescu & J.-C. Cheynet, Eds.). Paris: CNRS.
SECONDARY SOURCES
Agacanov, S. G. (2013). Oğuzlar (E. Necef & A. Annaberdiyev, Trans.). Istanbul:
Selenge.
Ahrweiler, H. (1974). La frontière et les frontières de Byzance en Orient. In Actes du
XIVe congrès international des études byzantines - Bucarest, 6-12 septembre
1971 (pp. 209-230). Bucharest: Editura Academiei RSR.
Ahrweiler, H. (1975). L'idéologie politique de l'empire byzantine. Paris: PUF.
Ahrweiler, H. (1998). Byzantine concepts of the foreigner: The Case of the nomads.
In H. Ahrweiler & A. E. Laiou (Eds.), Studies on the internal diaspora of the
Byzantine Empire (pp. 1-15). Washington, D.C.: Dumbarton Oaks Research
Library and Collection.
Akışık-Karakullukçu, A. (2019). A Question of audience: Laonikos Chalkokondyles’
Hellenism. Byzantinische Zeitschrift, 112,1-30.
Angold, M. (1995). Church and society under the Comneni 1081-1204. Cambridge:
Cambridge University Press.
Badenas, P. (1998). L’integration des Turcs dans la société byzantine (XIe–XIIe
siècles): Échecs d’un processus de coexistence. In S. Lampakis (Ed.), Byzantine
Asia Minor (6th-12th Cent.) (pp. 179–188). Athens: National Hellenic Research
Foundation.
329
Balivet, M. (1982). Byzantins judaïsants et Juifs islamisés. Des ‘Kühhân’ (Kâhin)
aux ‘Xiónai’ (Xiónos). Byzantion, 52, 24-59.
Balivet, M. (1986). Deux monastères byzantins fondés par des Turcs:
Koutloumoussiou/Kutulmuş et Dourachani/Turahan. Osmanlı Araştırmaları /
The Journal of Ottoman Studies, 5, 51-59.
Balivet, M. (1999). Byzantins et Ottomans: Relations, interaction, succession.
Istanbul: The Isis Press.
Balivet, M. (2005). Mélanges byzantins, seldjoukides et ottomans. Istanbul: The Isis
Press.
Barthold, V. V. (1981). Moğol istilasına kadar Türkistan (H. D. Yıldız, Trans.).
Istanbul: Kervan Yayınları.
Bartusis, M. C. (1992). The Late Byzantine army: Arms and society 1204-1453.
Philadelphia: University of Pensylvania Press.
Beihammer, A. D. (2011). Defection across the border of Islam and Christianity:
Apostasy and cross-cultural interaction in Byzantine-Seljuk Relations.
Speculum, 86(3), 597-651.
Beihammer, A. D. (2011). Orthodoxy and religious antagonism in Byzantine
perceptions of the Seljuk Turks (Eleventh and twelfth centuries). Al-Masāq,
23(1), 15-36.
Beihammer, A. D. (2017). Byzantium and the emergence of Muslim-Turkish
Anatolia, ca. 1040-1130. London: Routledge.
Beldiceanu, I. (1993). La population non-musulmane de Bithynie. In E. A.
Zachariadou (Ed.), The Ottoman Emirate (1300-1389) (pp. 7-22). Rethymnon:
Crete University Press.
Bosworth, C. E. (1992). The City of Tarsus and the Arab-Byzantine frontiers in early
and middle Abbasid times. Oriens, 33, 268-286.
Bosworth, C. E. (2000). al-Zuṭṭ. Encyclopaedia of Islam (2nd ed.), 11, 574-575.
Bosworth, C. E., & Latham, J. D. (2000). al-T̲ h̲ ug̲ h̲ ūr. Encyclopaedia of Islam (2nd
ed.), 10, 446-447.
Brand, C. M. (1989). The Turkish element in Byzantium, eleventh-twelfth centuries.
Dumbarton Oaks Papers, 43, 1-25.
Brown, R. (2013). Christians, Muslims and heretics: Religion and the Arab-
Byzantine frontier, c. 750-934. In A. Kralides & A. Gkoutzioukostas (Eds.),
Byzantium and the Arab world: Encounter of civilizations (pp. 91-104).
Thessaloniki: Aristotle University of Thessaloniki.
330
Cahen, C. (1948). La première pénétration turque en Asie Mineure (seconde moitié
du Xle s.). Byzantion, 18, 5-67.
Cahen, C. (1971). Questions d’histoire de la province de Kastamonu au XIIIe siècle.
Selçuklu Araştırmaları Dergisi, 3, 145-158.
Cahen, C. (1988). La Turquie pré-ottomane. Istanbul: Institut français d'études
anatoliennes d'Istanbul.
Canard M. (1960). al-ʿAwāṣim. Encyclopaedia of Islam (2nd ed.), 1, 761.
Canard, M. (1965). Ḏj̲
arād̲j̲
ima. Encyclopaedia of Islam (2nd ed.), 2, 456-458.
Charanis, P. (1963). The Armenians in the Byzantine Empire. Lisbon: Calouste
Gulbenkian Foundation Armenian Library.
Charanis, P. (1979). John Lydus and the question of the origin of the Vlachs in the
Greek lands. In S. Dufrenne (Ed.), Mélanges Ivan Dujčev (pp. 103-107). Paris:
Association des Amis des Études Archéologiques.
Cheynet, J-C. (1984). Toparque et topotèrètès à la fin du 11e siècle. Revue des
Études Byzantines, 42, 215-224.
Clauson, G. (1972). An Etymological dictionary of pre-thirteenth-century Turkish.
Oxford: Oxford University Press.
Crone, P. (1980). Slaves on horses: The Evolution of the Islamic polity.
Cambridge:Cambridge University Press.
Crone, P. (1996). The Rise of Islam in the World. In F. Robinson (Ed.), The
Cambridge Illustrated History of the Islamic World (pp. 2-31). Cambridge:
Cambridge University Press,.
Curta, F. (2019). Eastern Europe in the Middle Ages (500-1300). 2 vols. Leiden:
Brill.
Çolak, H. (2008). Bizans tarih yazıcılığında dönüşüm: Laonikos Chalkokondyles'te
Bizanslı ve Osmanlı imajı (1299-1402). Uludağ Üniversitesi Fen-Edebiyat
Fakültesi Sosyal Bilimler Dergisi, 9(15), 333-352.
Dagron, G. (1987). Ceux d’en face: Les peuples étrangers dans les traités militaires
byzantins. Travaux et Mémoires, 10, 207-232.
Dobrovits, M. (2011). The Altaic World through Byzantine eyes: Some remarks on
the historical circumstances of Zemarchus’ journey to the Turks (AD 569-570).
Acta Orientalia, 64(4), 373-409.
331
Drocourt, Z. (2010). Des simples sauvages aux redoutables étrangers: la notion de
«barbares » en Chine ancienne, à travers leurs denominations. In I. Rabut (Ed.),
Visions du “barbare” en Chine, en Corée et au Japon (pp. 13-28). Paris:
Publications Langues O’.
Durak, K. (2009). Defining the 'Turk': Mechanisms of establishing contemporary
meaning in the archaizing language of the Byzantines. Jahrbuch der
Österreichischen Byzantinistik, 59, 65-78.
Durak, K. (2013). Traffic across the Cilician frontier: Movement of people between
Byzantium and the Islamic Near East in the Early Middle Ages. In A. Kralides
& A. Gkoutzioukostas (Eds.), Byzantium and the Arab World: Encounter of
Civilizations (pp. 141-154). Thessaloniki: Aristotle University of Thessaloniki.
Eger, A. (2016). The Spaces between the teeth: A Gazetteer of towns on the Islamic-
Byzantine frontier. Istanbul: Ege.
Garrood, W. (2013). The Illusion of continuity: Nikephoros Phokas, John Tzimiskes
and the eastern border. Byzantine and Modern Greek Studies, 37, 20-34.
Gibb, H. A. R. (1958). Arab-Byzantine relations under the Umayyad Caliphate.
Dumbarton Oaks Papers, 12, 219-233.
Goffart, W. (2012). The Narrators of barbarian history (A.D. 550-800). Notre Dame-
Indiana: University of Notre Dame Press.
Golden, P. B. (1992). An Introduction to the history of the Turkic Peoples:
Ethnogenesis and state-formation in medieval and early modern Eurasia and
the Middle East. Wiesbaden: Harrassowitz Verlag.
Golden, P. B. (2006). Some thoughts on the origins of the Turks and the shaping of
the Turkic Peoples. In V. H. Mair (Ed.), Contact and exchange in the Ancient
World (pp. 136-157). Honolulu: University of Hawai’i Press.
Golden, P. B. (2007). The Conversion of the Khazars to Judaism. In P. B. Golden, H.
Ben-Shammai & A. Róna-Tas (Eds.), The World of the Khazars: New
perspectives (pp. 123-162). Leiden: Brill.
Gouma- Peterson, T. (Ed.). (2000). Anna Komnene and her times. New York and
London: Garland.
Grecu, V. (1948). Pour une meilleure connaissance de l’historien Doukas. Mémorial
Louis Petit (pp. 128-141). Paris: Peeters.
Grégoire, H. (1931). L'épopée byzantine et ses rapports avec l'épopée turque et
l'épopée romane. Bulletin de la classe des lettres et des sciences morales et
politiques de l’Académie Royale de Belgique, 17, 463-491.
332
Grégoire, H. (1936). Comment Sayyid Battal, martyr musulman du VIIIe siècle, estil
devenu, dans la légende, le contemporain d’Amer (†863)? Byzantion, 11, 571-
575.
Grousset, R. (1947). Histoire de l'Arménie des origines à 1071. Paris: Payot.
Haldon, J. (1985). ‘Jargon’ vs. ‘the facts’? Byzantine history-writing and
contemporary debates. Byzantine and Modern Greek Studies, 9, 95-132.
Haldon, J., & Kennedy, H. (1980). The Arab-Byzantine frontier in the eighth and
ninth centuries: Military organisation and society in the borderlands. Zbornik
Radova Visantološkog Instituta, 19, 79-116.
Harari, Y. (1997). The military role of the Frankish Turcopoles: A Reassessment.
Mediterranean Historical Review, 12(1), 75-116.
Heath, M. (2008). Aristotle on natural slavery. Phronesis, 53, 243-270.
Hartog, F. (1980). Le miroir d'Hérodote: Essai sur la représentation de l'autre.
Paris: Gallimard.
Hartog, F. (2021). Altérité, diversité, différence: Quelques jalons. DIVE-IN – An
International Journal on Diversity and Inclusion, 1(1), 1–11.
Heffening, W. (1993). Murtadd. Encyclopaedia of Islam (2nd ed.), 7, 635-636.
Honigmann, E. (1970). Bizans İmparatorluğunun doğu sınırı (F. Işıltan, Trans.).
Istanbul: İstanbul Üniversitesi Edebiyat Fakültesi Yayınları.
Hopwood, K. (1993). Peoples, territories, and states: The Formation of the beğliks of
Pre-Ottoman Turkey. In C. E. Farah (Ed.), Decision making and change in the
Ottoman Empire (pp. 129-138). Kirksville: Thomas Jefferson University Press.
Howard-Johnston, J. (2007). Byzantine sources for Khazar History. In P. B. Golden,
H. Ben-Shammai & A. Róna-Tas (Eds.), The World of the Khazars: New
Perspectives (pp. 163-194). Leiden: Brill.
Hunger, H. (1969-1970). On the imitation (Mimesis) of Antiquity in Byzantine
literature. Dumbarton Oaks Papers, 23-24, 15-38.
Huxley, G. (1974). Antecedents and context of Digenes Akrites. Greek, Roman and
Byzantine Studies, 15, 317-338.
Imber, C. (1991). The Legend of Osman Gazi. In E. A. Zachariadou (Ed.), The
Ottoman Emirate (1300-1389) (pp. 67-75). Rethymnon: Crete University Press.
İnalcık, Ş. (1948). İbn Hâcer'de Osmanlı'lara dair haberler. Ankara Üniversitesi Dil
ve Tarih-Coğrafya Fakültesi Dergisi, 6(3), 189-195.
İnan, A. (1954). Tarihte ve bugün Şamanizm. Ankara: Türk Tarih Kurumu.
333
Jenkins, R. J. H. (Ed.). (1962). De Administrando Imperio: A Commentary. London:
Athlone.
Jovanović, A. (2018). Imagining the communities of others: The Case of the Seljuk
Turks. Byzantina Symmeikta, 28, 239-273.
Kaegi, W. E. (1964). The Contribution of archery to the Turkish Conquest of
Anatolia. Speculum, 39(1), 96-108.
Kaegi, W. E. (1992). Byzantium and the early Islamic conquests. Cambridge:
Cambridge University Press.
Kafadar, C. (2007). A Rome of one's own: Reflections on cultural geography and
identity in the lands of Rum. Muqarnas, 24, 7-25.
Kaldellis, A. (2013). Le discours ethnographique à Byzance: Continuité et rupture
(P. Odorico & C. Messis, Trans.). Paris: Les Belles Lettres.
Kaldellis, A. (2014). A New Herodotos: Laonikos Chalkokondyles on the Ottoman
Empire, the fall of Byzantium, and the emergence of the west. Washington,
D.C.: Dumbarton Oaks Research Library and Collection.
Kaldellis, A. (2015). The Byzantine Republic: People and power in New Rome.
Cambridge, MA: Harvard University Press.
Kaldellis, A. (2019). Romanland: Ethnicity and empire in Byzantium. Cambridge,
MA & London: The Belknap Press of Harvard University Press.
Kazhdan, A. P. (Ed.). (1991). The Oxford Dictionary of Byzantium (3 vols.). Oxford:
Oxford University Press.
Keddie, N., & Baron, B. (1992). Women in Middle Eastern history: Shifting
boundaries in sex and gender. New Haven-London: Yale University Press.
Kitapçı Bayrı, B. (2020). Warriors, martyrs, and dervishes: Moving frontiers,
shifting identities in the land of Rome (13th-15th centuries). Leiden: Brill.
Kolia-Dermitzaki, A. (2002). The Execution of the forty-two martyrs of Amorion:
Proposing an interpretation. Al-Masaq, 4(2), 141-162.
Korobreinikov, D. (2014). Byzantium and the Turks in the thirteenth century.
Oxford: Oxford University Press.
Krallis, D. (2019). Serving Byzantium's Emperors: The courtly life and career of
Michael Attaleiates. Cham: Palgrave Macmillan.
Kyriakidis, S. (2021). The portrayal of Syrgiannes Palaiologos Philanthropenos in
the historical works of Nikephoros Gregoras and John Kantakouzenos.
Byzantinische Zeitschrift, 114(1), 221-238.
334
Laiou, A. E. (1992). Le désir, l’amour et la folie: Les rapports sexuels vus par les
Byzantins. In A. E. Laiou, Mariage, amour et parenté à Byzance aux XIe– XIIIe
siècles (pp. 67-89). Paris: De Boccard.
Lemerle, P. (1973). L’histoire des Pauliciens d’Asie Mineure d’après les sources
grecques. Travaux et Mémoires, 5, 1-113.
Lilie, R.-J. (2005). The Byzantine-Arab borderland from the 7th to the 9th Century.
In F. Curta (Ed.), Borders, barriers, and ethnogenesis: Frontiers in Late
Antiquity and the Middle Ages (pp. 13-21). Turnhout: Brepols.
Lindner, R. P. (2007). Explorations in Ottoman prehistory. Ann Arbor: University of
Michigan Press.
Macartney, C. A. (1944). On the Greek sources for the History of the Turks in the
sixth Century. Bulletin of the School of Oriental and African Studies, 11(2),
266–275.
MacMaster, T. J. (2014). The Origin of origins: Trojans, Turks and the birth of the
myth of Trojan origins in the medieval world. Atlantide, 2, 1-12.
Malamut, E. (1995). L’image byzantine des Petchénègues. Byzantinische Zeitschrift,
88(1), 105-147.
Malamut, E. (2000). Les peuples étrangers dans l'idéologie impériale. Scythes et
Occidentaux. In: L'étranger au Moyen Âge. Actes du XXXe congrès de la
Société des historiens médiévistes de l'enseignement supérieur public
(Göttingen, 1999) (pp. 119-132). Paris: Publications de la Sorbonne.
Mallan, C. (2017). A Turkish Alexander? Michael Attaleiates, Porus and Alexander
the Great. Histos, 11, 101-107.
Mango, C. (1981). Discontinuity with the classical past in Byzantium. In M. Mullett
& R. Scott (Eds.), Byzantium and the classical tradition (pp. 48-57).
Birmingham: University of Birmingham.
Marsham, A. (Ed.). (2020). The Umayyad World. London: Routledge.
Mc Geer, E. (2008). Military texts. In E. Jeffreys, J. F. Haldon & R. Cormack (Eds.),
The Oxford Handbook of Byzantine studies (pp. 907-914). Oxford: Oxford
University Press.
Mecit, S. (2014). The Rum Seljuqs: Evolution of a dynasty. London: Routledge.
Melichar, P. (2019). Empresses of Late Byzantium: Foreign brides, mediators and
pious women. Berlin: Peter Lang Verlag.
Mengüç, M. C. (2014). The Türk in Aşıkpaşazâde: A Private individual’s Ottoman
History. Osmanlı Araştırmaları / The Journal of Ottoman Studies, 44, 45-66.
335
Messis, C. (2020) Lorsque la périphérie assiège et conquiert le centre: Certains
aspects des relations entre Byzantins et Turcs, XIIe et XVe siècles. Études
Byzantines et Post-Byzantines, Nouvelle série, 2, 75-88.
Meyendorff, J. (1964a). Byzantine views of Islam. Dumbarton Oaks Papers, 18,
113–132.
Meyendorff, J. (1964b). A Study of Gregory Palamas (G. Lawrence, Trans.).
London: The Faith Press.
Mitchell, F. (2021). Monsters in Greek literature: Aberrant bodies in Ancient Greek
cosmogony, ethnography, and biology. New York: Routledge.
Moravcsik, G. (1958). Byzantinoturcica, vol. 1: Die byzantinischen Quellen der
Geschichte der Türkvölker, vol. 2: Sprachreste der Türkvölker in den
byzantinischen Quellen. Berlin: Akademie Verlag.
Necipoğlu, G., & Bozdoğan S. (2007). Entangled discourses: Scrutinizing orientalist
and nationalist legacies in the architectural historiography of the “lands of
Rum”. Muqarnas, 24, 1-7.
Necipoğlu, N. (2006). Turks and Byzantines (Eleventh and Twelfth Centuries). In E.
Çağatay & D. Kuban (Eds.), The Turkic speaking peoples: 2,000 Years of art
and culture from Inner Asia to the Balkans (pp. 254-265). Munich-Berlin-
London-New York: Prestel Verlag.
Necipoğlu, N. (2009). Byzantium between the Ottomans and the Latins: Politics and
society in the late Empire. Cambridge: Cambridge University Press.
Neville, L. (2018). Guide to Byzantine historical writing. Cambridge: Cambridge
University Press.
Nicol, D. M. (1998). The Reluctant emperor: A biography of John Cantacuzene,
Byzantine emperor and monk, c. 1295-1383. Cambridge: Cambridge University
Press.
Odorico, P. (1993). L'étranger et son imaginaire dans la littérature byzantine. In L.
Mayali & M. M. Mart (Eds.), Of Strangers and Foreigners (Late Antiquity -
Middle Ages) (pp. 65-79). Berkeley: Regents of the University of California.
Odorico, P. (2006). Displaying la littérature Byzantine. In E. Jeffreys, F. K. Haarer &
J. R. Ryder (Eds.). Proceedings of the 21st International Congress of Byzantine
Studies (London 21-26 August 2006) (vol. 1, pp. 213-234). Aldershot: Ashgate.
Odorico, P. (2010). Différence non diversité: Les Grecs du Moyen Âge faux aux
autres Européens. Conférence présentée aux Dialogues d’Athénes en 2010.
Odorico, P. (Ed.). (2012). La face cachée de la littérature byzantine, le texte en tant
que message immédiat. Paris: Centre d’études byzantines, néo-helleniques et
sud-est européennes - École des Hautes Études en Sciences Sociales.
336
Odorico, P. (2017). Identité et craintes. Théodore Pédiasimos à Serrès au XIVe
siècle. In A. Cuomo & E. Trapp (Eds.), Toward a historical sociolinguistic
poetics of medieval Greek (pp. 161-173). Turnhout: Brepols.
Odorico, P. (2022). What’s in a name? The Byzantine Chroniclers. In L. James, O.
Nicholson & R. Scott (Eds.), After the Text: Byzantine Enquiries in Honour of
Margaret Mullett (pp. 85-94). London-New York: Routledge.
Oehler, K. (1964). Aristotle in Byzantium. Greek, Roman and Byzantine Studies, 5,
133-146.
Oikonomides, N. (1983). Les Danishmendides, entre Byzance, Bagdad et le sultanat
d'Iconium. Revue numismatique, 25, 189-207.
Oikonomides, N. (1993). The Turks in the Byzantine rhetoric of the twelfth century.
In C. E. Farah (Ed.), Decision making and change in the Ottoman Empire (pp.
149-155). Kirksville: Thomas Jefferson University Press.
Ostrogorsky, G. (1956). The Byzantine emperor and the hierarchical world order.
The Slavonic and East European Review, 35(84), 1–14.
Page, G. (2008). Being Byzantine: Greek identity before the Ottomans. Cambridge:
Cambridge University Press.
Papageorgiou, A. (2011). οί δέ λύκοι ώς Πέρσάι: The image of the “Turks” in the
reign of John II Komnenos (1118-1143). Byzantinoslavica, 69(1-2), 149-161.
Patlagean, É. (1977). Pauvreté économique et pauvreté sociale à Byzance. Paris-La
Haye: Mouton.
Patlagean, É. (2007). Un Moyen Âge Grec: Byzance, IXe-XVe siècle. Paris: Albin
Michel.
Peacock, A. C. S. (2010). Early Seljuq History: A new interpretation. London:
Routledge.
Peacock, A. C. S. (2013). From the Balkhān-Kūhīyān to the Nāwakīya: Nomadic
politics and the foundations of Seljūq rule in Anatolia. In J. Paul (Ed.), Nomad
aristocrats in a world of empires (pp. 55-80). Wiesbaden: Dr. Ludwig Reichert
Verlag.
Peacock, A. C. S. (2015). The Great Seljuk Empire. Edinburgh: Edinburgh
University Press.
Pellat, C. (1983). Liwāṭ. Encyclopaedia of Islam (2nd ed.), 5, 776-779.
Ray, L. (2006). Assimilation. In B. S. Turner (Ed.), The Cambridge Dictionary of
Sociology (p. 24). Cambridge: Cambridge University Press.
337
Rotman, Y. (2016). The medieval Mediterranean slave trade. In P. Magdalino & N.
Necipoğlu (Eds.), Trade in Byzantium: Papers from the Third International
Sevgi Gönül Byzantine Studies Symposium (pp. 129-142). Istanbul: Koç
University Research Center for Anatolian Civilizations.
Saussure, F. de (1995). Cours de linguistique générale (T. De Mauro, Ed.). Paris:
Payot. (First edition 1916).
Savvides, A. G. C. (1993). Byzantines and the Oghuz (Ghuzz). Some observations
on the nomenclature. Byzantinoslavica, 54, 147-155.
Savvides, A. G. C. (1997). Some notes on the terms Agarenoi, Ismailitai and
Sarakenoi in Byzantine sources. Byzantion, 67, 89-96.
Shepard, J. (2001). Constantine VII, Caucasian openings and the road to Aleppo. In
A. Eastmond (Ed.), Eastern approaches to Byzantium: Papers from the Thirty-
Third spring symposium of Byzantine studies (pp. 19-40). Aldershot: Ashgate.
Shliakhtin, R. (2016). From Huns into Persians: The projected identity of the Turks
in the Byzantine rhetoric of eleventh and twelfth centuries (Unpublished PhD
thesis). Central European University.
Shukurov, R. (2004). The Crypto-Muslims of Anatolia. In D. Shankland (Ed.),
Archaeology, anthropology and heritage in the Balkans and Anatolia: The Life
and Times of F.W. Hasluck, 1878-1920 (vol. 2, pp. 135–158). Istanbul: The Isis
Press.
Shukurov, R. (2013). Harem Christianity: The Byzantine identity of Seljuk princes.
In A. C. S. Peacock & S. N. Yıldız (Eds.), The Seljuks of Anatolia: Court and
society in the medieval Middle East (pp. 115-150). London: I.B. Tauris.
Shukurov, R. (2016). The Byzantine Turks, 1204–1461. Leiden: Brill.
Sinor, D. (1990). The Cambridge History of early Inner Asia. Cambridge: Cambridge
University Press.
Stephenson, P. (2000). Byzantium's Balkan frontier: A political study of the Northern
Balkans. Cambridge: Cambridge University Press.
Stone, A. F. (2008). The Missionaries of Manuel I. Revue des Études Byzantines, 66,
253-257.
Stouraitis, I. (2014). Roman identity in Byzantium: A critical approach.
Byzantinische Zeitschrift, 107(1), 175-220.
Stouraitis, Y. (2017). Reinventing Roman ethnicity in high and late medieval
Byzantium. Medieval Worlds, 5, 70-94.
Sümer, F. (1980). Oğuzlar (Türkmenler): Tarihleri-boy teşkilatı, destanları (3rd
revised ed.). Istanbul: Ana Yayınları.
338
Sümer, F. (1990). Doğu Anadolu’da Türk Beylikleri. Ankara: Türk Tarih Kurumu.
Todorov, B. A. (2008). Byzantine myths of origins and their function. Studia Slavica
et Balcanica Petropolitana, 2(4), 64-72.
Tor, D. (2005). Privatized jihad and public order in the Pre-Seljuq period: The Role
of Mutatawwi’a. Iranian Studies, 38(4), 555-573.
Tremblay, V. (2016). L’identité romaine est-elle exclusive à Constantinople?
Dichotomie entre Byzance et les Balkans à l’époque médiobyzantine (VIe–XIIe
siècles). In N. S. M. Matheou, T. Kampianaki & L. M. Bondioli (Eds.), From
Constantinople to the Frontier: The City and the Cities (pp. 25-40). Leiden-
Boston: Brill.
Turan, O. (1958). Türkiye Selçukluları hakkında resmi vesikalar. Ankara: Türk Tarih
Kurumu.
Turan, O. (1971). Selçuklular zamanında Türkiye. Istanbul: Ötüken.
Tülüce, A. (2011). Bizans tarih yazımında öteki: Selçuklu kimliği. Istanbul: Selenge.
Vásáry, I. (2009). Cumans and Tatars: Oriental military in the Pre-Ottoman
Balkans, 1185-1365. Cambridge: Cambridge University Press.
Vasiliev, A. A. (1948). Byzantium and Islam. In N. H. Baynes & H. St. L. B. Moss
(Eds.), Byzantium (pp. 308-325). Oxford: Clarendon Press.
Vryonis, Jr, S. (1971a). The decline of medieval Hellenism in Asia Minor and the
process of Islamization from the eleventh through the fifteenth century. Los
Angeles: University of California Press.
Vryonis, Jr, S. (1971b). Evidence of human sacrifice among the early Ottoman
Turks. Journal of Asian History, 5, 140-146.
Vryonis, Jr, S. (1976). Laonicus Chalcocondyles and the Ottoman budget.
International Journal of Middle East Studies, 7(3), 423-432.
Wenghofer, R. (2014). Sexual promiscuity of Non-Greeks in Herodotus' Histories.
The Classical World, 107(4), 515-534.
Wittek, P. (1952). Yazijioghlu ‘Ali on the Christian Turks of Dobruja. Bulletin of the
School of Oriental and African Studies, 14(3), 639-668.
Wittek, P. (1958). The rise of the Ottoman Empire. Londan: The Royal Asiatic
Society of Great Britain and Ireland.
Wolfram, H. (1990). Le genre de l’Origo gentis. Revue Belge de Philologie et
d’Histoire, 68(4), 789-801.
339
Wood, P. (2015). Christians in the Middle East, 600-1000: Conquest, competition
and conversion. In A. C. S. Peacock, B. De Nicola & S. N. Yıldız (Eds.), Islam
and Christianity in medieval Anatolia (pp. 23-51). Farnham: Ashgate.
Zachariadou, E. (1978). Observations on some Turcica of Pachymeres. Revue des
Études Byzantines, 36, 261-267.