Basil II and The Government of Empire (976-1025)
Basil II and The Government of Empire (976-1025)
Basil II and The Government of Empire (976-1025)
Catherine J.Holmes
Short Abstract
The reign of Basil II (976-1025) is widely accepted as the high point of medieval
Byzantium. When the emperor died, imperial frontiers were at their most far-flung
since the seventh century. Yet despite the territorial significance of Byzantium in
this period, there is no comprehensive modern history of the reign. This thesis
develops two important foundation stones for a new narrative history of Basil II: a
better understanding of the relevant medieval historiography, and an analysis of the
economic and administrative structures which underpinned contemporary political
society. The first three chapters analyse the main Greek narrative account of the
reign composed by John Skylitzes at the end of the eleventh century. The first
chapter is a detailed textual study. The second chapter explores the literary, social
and political contexts behind Skylitzes' text. The third chapter compares Skylitzes'
coverage of Basil's reign with the rest of the medieval historical record, and
identifies a hitherto unacknowledged source in the Greek tradition. Read together,
these chapters demonstrate how the demands of history writing in the later eleventh
century conditioned Skylitzes' narrative. In order to gain a more contemporary
view of the reign, chapters four to six examine the economy and administration of
the eastern half of the Byzantine empire during the tenth and eleventh centuries.
These chapters argue that from the middle of the tenth century onwards, the
administration of the eastern half of the empire was predicated on an imperial
desire to exploit increasing regional economic prosperity. However, successive
emperors, most notably Basil II himself, recognised the substantial practical
constraints on the penetration of imperial authority in the locality. As a result the
administration of the Byzantine east was characterised by considerable flexibility,
and was able to adapt with surprising ease to local conditions.
Long Abstract
The principal deterrent to a modern analysis of the reign has been the problem of
evidence. There is no contemporary appraisal of the whole reign in Greek.
Coverage by later medieval Greek historians is meagre in quantity, and inconsistent
in quality. Isolated references to the reign in literary materials outside the Greek
historical record are difficult to interpret in the absence of a sustained Greek
11
narrative. Although contemporary historians located on the eastern periphery of the
empire, writing in languages other than Greek, offer more reliable dated
information than their Greek counterparts, large chronological periods of the reign
and substantial geographical regions are almost entirely neglected by the historical
record. To some extent the short comings of the written sources can be offset by
archaeological, sigillographical and numismatic evidence. Yet, the material record,
which is often so difficult to date, should not be used simply to plug geographical
or chronological lacunae left gaping by the medieval historians.
The ultimate ambition of any fresh investigation into Basil's reign must be the
development of a new narrative of the political, military and diplomatic history of
the Byzantine empire in the later tenth and early eleventh centuries. Yet, the
chronological difficulties inherent in the primary sources mean that such a narrative
cannot simply be constructed by synthesising the extant written materials and
adding occasional details from the material record. Such an approach runs the risk
of replicating the chronological and geographical lacunae of the medieval
historians, and is unlikely to improve substantially on Schlumberger's analysis of
the reign. Instead, a convincing narrative needs to be predicated on a better
understanding of the extant literary evidence, and a strong sense of the structures
which underpinned political society in medieval Byzantium. This thesis develops
these two essential foundation stones to the construction of a new narrative. The
first half of the thesis examines the main medieval Greek account of the reign in the
light of the wider literary, social and political contexts in which it was written. The
second half looks more directly at the economic and administrative structural bases
to political power in tenth- and eleventh-century Byzantium by analysing a wide
range of contemporary literary and material evidence from the eastern half of the
empire.
111
The first three chapters of this thesis investigate the 'Synopsis Historion' of John
Skylitzes. This long synoptic history, which was compiled towards the end of the
eleventh century and covers the period 811 to 1057 (or 1079 in those versions of
the text which contain the 'Continuation'), contains the earliest and longest
connected narrative of the reign of Basil II in Greek. At the very beginning of the
first chapter of the dissertation, the importance of Skylitzes' testimony to any
understanding of the reign of Basil II is considered in general terms. The discussion
summarises the contents of Skylitzes' coverage of the reign, stressing the extent to
which the first half of the account is dominated by the civil wars waged by the
generals Bardas Skleros and Bardas Phokas between 976 and 989, and the second
half by Basil's campaigns in Bulgaria. The text's many geographical and
chronological confusions and lacunae are highlighted. The chapter then argues that
it is impossible to approach Skylitzes' coverage of this period simply by trying to
amplify or verify his information and interpretation with material from other
medieval historians. Instead of attempting to assess Skylitzes' coverage of Basil's
reign on a fact by fact basis, the modern historian of this period should try to
understand the principles of selection, presentation and interpretation which
underpin Skylitzes' text. Such an approach requires a more profound understanding
of how Skylitzes' coverage of Basil's reign relates to the text of the 'Synopsis
Historion' as a whole, and of the literary, social and political contexts behind the
author's compilation. It is this relationship between authorship, text and context,
which lies at the heart of the first three chapters of this thesis.
Having outlined the analytical approach which frames discussion of the 'Synopsis
Historion' within the thesis, the first chapter goes on to summarise the small body
of existing Skylitzes' scholarship. It outlines past research into the manuscript
history of the 'Synopsis Historion', Skylitzes' biography, and the author's source
IV
materials and working methods. However, given the extremely limited ambition
and scope of much of the secondary literature, the chapter quickly moves on to a
detailed and original textual analysis of Skylitzes' use and abuse of source
materials. Since none of Skylitzes' underlying sources for Basil's reign survive, this
analysis involves moving backwards through the 'Synopsis Historion' to the reign
of Romanos Lekapenos (920-44), and comparing Skylitzes' testimony for this
earlier tenth-century period with that of his principal source, Theophanes
Continuatus. Through this comparative process, Skylitzes' working methods and
treatment of source materials are elucidated. It becomes clear that while Skylitzes
copies the root narrative closely, he makes several significant alterations to both the
content and the structure of the underlying text. Some alternations are unwitting
copying errors, others are more deliberate, the result of conscious intervention. Two
elements of Skylitzes' reshaping of Theophanes Continuatus' text are particularly
striking. First, his devoted attention to the careers of leading members of the
Byzantine political elite; and second, his tendency to describe administrative and
military matters through the use of a homogenised short hand which renders the
meaning of the underlying text opaque. At the end of the first chapter, it is argued
that many of the distinctive features visible in Skylitzes' coverage of the reign of
Romanos, are also to be found in his treatment of Basil's reign. Particularly
conspicuous is his deployment of a stereotypical military vocabulary to describe the
civil wars waged by Phokas and Skleros and Basil's campaigns in the Balkans.
The second chapter of the dissertation sets the discussion of author and text
developed in the previous chapter in the broader literary, social and political
contexts of the later eleventh century, the period when Skylitzes wrote the
'Synopsis Historion'. By assessing textual evidence in the light of Skylitzes' career
as a senior civil servant within the imperial government of the later eleventh
century, the chapter argues that the 'Synopsis Historion' was written for an elite
Constantinopolitan audience at the court of the emperor Alexios Komnenos (1081-
1118). It also suggests that the conditions of internal and external political
insecurity, which prevailed in the 1080s and 1090s, may explain the text's
characteristic interest in the ancestors of the most important families of the later
eleventh century, and its obsession with Balkan history. At the end of the chapter
the implications of these later eleventh-century contexts for the text's treatment of
Basil's reign are explored. Above all, it is argued that the text may overstate the
significance of the hereditary aristocracy, while underestimating the role of the
emperor, thus fundamentally misrepresenting the balance of power within the
Byzantine body politic in the later tenth and early eleventh centuries.
In chapter three attention focuses more directly on the reign of Basil itself. The
chapter takes as its subject matter the revolts of Bardas Skleros and Bardas Phokas
(976-989). Although most of Basil's reign is poorly covered in the medieval
historical record, these insurrections represent an exceptional case, since they are
described in detail by several historians both in Greek and in other languages. As a
result, these revolts constitute the only period of Basil's reign when it is possible to
compare Skylitzes' coverage with accounts presented by other historians. Such a
comparison is the undertaking of this chapter. However, the purpose of this analysis
is not to produce a new comprehensive narrative of the revolts themselves, but to
learn more about the composition of Skylitzes' text. Thus, while the chapter is
prefaced by a short narrative appraisal of the political history of the tenth century
and a synopsis of the revolts themselves, its principal concern is to illustrate how
literary genre shapes the historical record. Attention focuses most closely on the
Greek accounts of John Skylitzes and Michael Psellos. A hitherto unacknowledged
source, attributable to the general Bardas Skleros, is identified in the testimonies of
VI
both Psellos and Skylitzes. However, it is argued that the Greek medieval
historians' decision to utilise this source was driven more by the demands of history
writing in the second half of the eleventh century, than by the accuracy of the
source's presentation of the history of Basil's reign.
In order to gain a more accurate view of the structures and processes of power
which underpinned Byzantine elite society during Basil's reign, the second half of
this thesis looks behind the eleventh-century Greek historical record at more
contemporary forms of evidence. At the heart of this investigation is the question of
government in the eastern half of the empire. Given the long war of attrition
between Byzantium and Bulgaria during Basil's reign, this decision to look east
rather than west may seem unusual. However, in order to understand Balkan
relations in this period, long-term economic, social, and political developments in
the eastern half of the empire, the region where Byzantine military expansion had
been at its most pronounced in the century before Basil's reign, must first be
established. The fourth chapter sets the scene by outlining the geography and
economy of the eastern half of the empire in the tenth and eleventh centuries. It
begins with a general discussion of the relevant primary sources and an overview of
modern ideas about the medieval Byzantine economy. It then outlines the relief,
climate and communications of the Byzantine east. In more detail, it demonstrates
that arable cultivation was possible throughout the eastern half of the empire,
including much of the Anatolian plateau. The chapter argues that once Arab raiding
activity ceased and Byzantine armies began to cross the Taurus and Anti Taurus
mountains in the second quarter of the tenth century, the whole of the Byzantine
east experienced rapid economic growth. Trade was of fundamental importance to
that expansion. However, the chapter emphasises that the Anatolian plateau lacked
the demographic and capital resources necessary to inspire the economic recovery
VI1
of the eastern plateau and eastern frontier regions of the empire. Instead, these areas
were revivified by a heterodox, non Greek-speaking, non-Chalcedonian, and even
Chapters five and six deal more directly with the articulation of imperial authority
in the eastern half of the empire by focusing on administration. Having stressed the
importance of sigillographical evidence to any understanding of Byzantine
administration, the first half of chapter five offers a background summary of
administrative changes which took place in the decades before Basil came to
power. It suggests that Byzantine conquests of regions east of the Taurus and Anti
Taurus mountains in the second and third quarters of the tenth century had two
important ramifications for local government in the eastern half of the empire as a
whole: the gradual demilitarising of the Anatolian provinces (or themes], and the
development of a larger, centralised army. The second half of chapter five discusses
administrative change in the Anatolian heartland of the Byzantine east during
Basil's reign, arguing that long standing tenth-century trends of demilitarisation and
centralisation continued. The most visible evidence of these developments was the
appearance within the locality of increased numbers of civil officials appointed by
Constantinople. The extent to which the principle underpinning local government
shifted from the provision of defence to the exploitation of resources by a
centralising, imperial authority is emphasised. However, the chapter argues that the
exploitation of the Anatolian themes was probably concerned more with the
indirect collection of taxes than with the direct management of imperially-owned,
immovable assets. Despite Basil IPs legislative onslaught against the most
'powerful' members of the political elite after the civil wars against Skleros and
vin
Phokas, there is little sign that large swathes of land were confiscated from private
ownership and transferred on a permanent basis to the control of the state.
Chapter six moves eastwards to those regions beyond the Taurus and Anti Taurus
mountains which had been conquered in the fifty years before Basil's accession.
Although the chapter is primarily concerned with how imperial authority was
expressed in these newly conquered regions, this analysis is prefaced by a brief
summary of the empire's dealings with its eastern neighbours, both Muslim and
Christian, during the reign of Basil. The chapter then moves on to consider the
military administration of the frontier. It argues against the idea that the later tenth-
century administrative manual, the 'Escorial Taktikon', reflects the quotidian
administration of the eastern frontier. Instead, by analysing each of the key border
commands known as katepanates, the chapter argues that during the first half of
Basil's reign military administration was typified by a series of ad hoc
arrangements, which were constantly shaped and reshaped by political tensions
within and outside the empire. Military administration only became more stable
after a peace accord was reached with the Fatimids of Egypt in 1000. Nonetheless,
flexibility remained a characteristic of frontier command as the administrative
experience of the new katepanates of Iberia and Vaspurakan established in western
Caucasia at the end of the reign proved. Finally, the chapter considers civil
administration in these eastern regions, particularly in the former Muslim emirates.
It concludes that during the first half of the reign, local intermediary figures
supervised a quasi tribute relationship between Constantinople and the eastern
periphery. Although this relationship became subject to greater central control in
the second half of the reign, the fundamental tribute principle underpinning local
administration remained unaltered.
IX
The conclusion to this thesis summarises the major arguments developed during the
Basil's reign should progress. It suggests that the relationships between the emperor
and other members of elite society should be analysed closely, above all those
senior officials who were responsible for executing imperial authority on the
ground. However, while the conclusion to this thesis acknowledges that such
research has yet to be undertaken in detail, it presents a provisional picture of
political relations during Basil's reign, suggesting that the key structural threat to
the stability of the Byzantine state was the army. This problem, however, was
substantially mitigated by the peace of 1000 with the Fatimids. With the cessation
of a hostile threat from Egypt, Basil was able to divert Byzantine military energies
away from the eastern frontier to the Balkans, a region where the field armies and
their generals could be more closely supervised by the emperor himself.
Acknowledgements
I would like to thank the members of the Oxford 'Byzantine Seminar' for making
Byzantine Studies such a rewarding and pleasing pursuit, as well as the students
and fellows at Balliol College for their friendship and intellectual stimulation
between the years 1994 and 1998. The British Academy sponsored me during this
four-year period in Oxford. I was also the beneficiary of a Jowett scholarship from
Balliol in the period 1995-7. My final year of doctoral work has been facilitated by
a research fellowship from Gonville and Caius College, Cambridge. The fellows of
Caius have offered me warm support throughout this academic year (1998-9). I
would particularly like to thank Anita Bunyan and Nicholas Hammond for their
bibliographical advice.
I would also like to thank my parents, Patricia and George, for always encouraging
me to imagine and discover other worlds. My mother, Patricia Willgrass Holmes,
not only had the patience to read the thesis, but even checked my spellings. Finally,
I could thank Doug Rogers for helping me with the maps. However, my gratitude to
him extends far beyond cartography.
XI
Table of Contents
Short Abstract /
Long Abstract //
Acknowledgements xi
Table of Contents xii
Transliteration xv
Abbreviations xvi
Introduction 1
I. Introduction 109
II. Historical synopsis 112
III. Historiographical discussion
xn
/. General historiographical interest in the revolts 116
ii. Pro-Skleros interest in the Greek historiography
a. Bardas Skleros: the focus of attention 122
b. Bardas Phokas: the more dangerous rebel 126
c. A pro-Skleros or pro-Phokas source? 132
d. Dating the pro-Skleros source 136
IV. Conclusion: the pro-Skleros source and the writing of history 141
at the end of the eleventh century
Xlll
//'. Civil administration
a. Judges 248
b. Fiscal lands 256
c. Crown lands 266
//'/. Conclusion 271
Conclusion 342
Maps:
1. Anatolia during the revolts of Bardas Skleros and Bardas Phokas 354
2. Anatolia: relief and communications 355
3. Anatolia: cities and towns 356
4. Western Cappadocia 357
5. Anatolian themes and kleisourai c.917 358
6. Byzantium's eastern frontier lands in the tenth and eleventh
centuries 359
Bibliography 360
xiv
Transliteration
Greek names and place names are transliterated with '' and 'as-', except in cases
where a Latinate or Anglicised version is very familiar. Turkish place names follow
current Turkish usage. Arabic names and place names follow a simplified version
of the system used by the Encyclopaedia of Islam, New Edition (Leiden, I960-).
Armenian names and place names follow the spelling adopted by Robert Thomson
in History of the House of the Artsrunik' (Detroit, 1985); Georgian names and place
names follow the spelling adopted by Thomson in Rewriting Caucasian History:
The Georgian Chronicles (Oxford, 1996)
xv
Abbreviations
AB Analecta Bollandiana
Anatol Stud Anatolian Studies
B Byzantion
BCH Bulletin de Correspondance Hellenique
BHG F.Halkin, Bibliotheca Hagiographica Graeca, 3 vols. (Brussels
1957)
BK Bedi Kartlisa
BMGS Byzantine and Modern Greek Studies
Byz Forsch Byzantinische Forschungen
Byz Slav Byzantinoslavica
BZ Byzantinishe Zeitschrift
Cah Arch Cahiers Archeologiques
CFHB Corpus Fontium Historiae Byzantinae
CSHB Corpus Scriptorum Historiae Byzantinae
CSCO Corpus Scriptorum Christianorum Orientalium
DMA Dictionary of the Middle Ages, (ed.) J.R.Strayer (13 vols., New
York, 1982-)
DOP Dumbarton Oaks Papers
El Encyclopaedia ofIslam, New Edition (Leiden, I960-), Vols. I-
IX, Supplement and Indices
EEBS
JHS Journal ofHellenic Studies
JOB Jahrbuch der Osterreichischen Byzantinistik
MGH SS Monumenta Germaniae Historica. Scriptores
ODB Oxford Dictionary ofByzantium, (ed.) A.Kazhdan et al. (3
t
XVI
PG Patrologiae Cursus Completus, Series Graeco-Latina, (ed.) J.-
P.Migne (Paris, 1857-66, 1880-1903)
PO Patrologia Orientalis, (eds.) R.Graffin and F.Nau (Paris, 1930)
r
REArm Revue des Etudes Armeniennes, New Series
f
REB Revue des Etudes Byzantines
SBS Studies in Byzantine Sigillography, (ed.) N. Oikonomides (4
vols., Washington, 1987-1995)
TIB Tabula Imperil Byzantini: Verlag der Osterreichischen
Akademie der Wissenschaften (Vienna, 1976-1998)
TM Travaux et Memoires
Viz Vrem Vizantijskij Vremennik
ZRV1 Zbornik radova Vizantoloskog instituta
xvn
Introduction
Byzantium. When Basil died in 1025 the boundaries of empire were at their most
far-flung since the seventh century. Bulgaria had been conquered in 1018 after
more than thirty years of attritional warfare. In Caucasia the Georgian princedom of
Tao was annexed in 1000 and Armenian Vaspurakan absorbed in the final decade
of the reign. Just before the emperor's death, attention had shifted to the
Mediterranean and plans were afoot for the conquest of Sicily. Outside the
territorial borders of the empire, Byzantine religious and cultural influence was
988. Nor was expansion purely external. When Basil died his wealth was
legendary: a labyrinthine treasury was rumoured to extend under the Great Palace
where the booty of his numerous military campaigns was stored. So great indeed
was his wealth that he remitted taxation for the final two years of his reign. * This
success seems all the more remarkable given the highly inauspicious start to the
reign. Between 976 and 989 Basil had faced two intense periods of civil war waged
by the generals Bardas Skleros and Bardas Phokas. Basil himself suffered a
humiliating military defeat against the Bulgarians in 986. The scale of Basil's
recovery and his subsequent military conquests persuaded later medieval Byzantine
writers to claim that he was the greatest emperor since Herakleios.2 Yet Basil's
golden legacy was relatively short-lived. Within thirty years of his death the empire
began to fragment amid Turkish attacks in the east, and Norman and nomad raids
in the west. By the early 1040s the empire once again became prone to coups
1 Psellos (Michael): Michel Psellos. Chronographie, ed. E.Renauld (2 vols., Paris, 1926), i, 19-20;
Skylitzes (John): loannis Skylitzae Synopsis Historiarum, ed. I.Thurn (Berlin and New York, 1973),
p.373
d'etat. By the 1070s revolt became all out civil war as leading aristocratic dynasties
The importance of Basil's long reign to the internal and external history of the
Byzantine empire could hardly be clearer. Yet, there has been no general
the reign since Gustave Schlumberger published his long analysis of this period in
the final years of the last century. 3 Furthermore, most regional narratives which
contain detailed coverage of Basil's reign are also often antique. Gay's study of
Byzantine southern Italy was published shortly after Schlumberger's analysis. Most
surveys of Byzantine relations with Bulgaria during Basil's reign were written in
the first half of this century.4 The only area to attract more substantial coverage in
more recent times has been the eastern frontier and in particular relations with the
empire's Muslim neighbours. Unfortunately, many of these analyses are either only
partial or unpublished. Felix's study of the Byzantine relations with the Muslims
begins in 1000 and thus omits the first half of Basil's reign. 5 Forsyth's more general
analysis of Byzantine relations with its eastern neighbours, both Muslim and
2 Choniates (Michael): Mi%a^i 'AxofAivarou TOU Xcavtdrou -m trtt*6fAcvcL, ed. S.P.Lampros (2 vols.,
Athens, 1880), ii, 354.
3 G. Schlumberger, L 'Epopee byzantine a la fin du dixieme siecle (2 vols., Paris, 1896-1900), i, 327-
777; ii, passim. Schlumberger was able to build on the work of V.R.Rozen, whose Imperator Vasilij
Bolgarobojca. Izvlechenija iz letopisi Jach-i Antiochijskago: the Emperor Basil Slayer of the
Bulgarians, Extracts from the Chronicle ofJahja ofAntioch (St Petersburg, 1883; reprint London,
1972) had drawn attention to the fact that the chronicle of Yahya ibn Sa'id ofAntioch contained a
vast array of material relevant to the reign of Basil.
4 J.Gay, L 'Italic meridionale et I'empire byzantin depuis I'avenement de Basil I jusqu 'a la prise de
Bari par les Normands, 867-1071 (Paris, 1904), pp.324-429 for narrative coverage of southern Italy
in Basil's reign. Subsequent analyses of later tenth- and eleventh-century Byzantine southern Italy
have usually been dedicated to the structures of administration and government: see for example,
V.von Falkenhausen, Untersuchungen uber die byzantinische Herrschaft in Suditalien vom 9. bis ins
11. Jahrhundert (Wiesbaden, 1967). S.Runciman, History of the First Bulgarian Empire (London,
1930), pp.219-257 contains some coverage of Basil's relations with Bulgaria; see also N.Adontz,
'Samuel I'armenien roi des Bulgares', in Etudes armeno-byzantines (Lisbon, 1965), pp.347-407. The
only more recent publication to deal with Bulgaria in Basil's reign is W.Seibt, 'Untersuchungen zur
Vor- und Friihgeschichte der "bulgarischen" Kometopulen', Handes Amsorya 89 (1975), pp.65-98.
Christian, is unpublished.6 Indeed, most recent published analyses of the history of
Basil's reign have been narrow in compass. On the one hand detailed attention has
which the reign is divided into two distinct temporal and geographical phases: first,
Nonetheless, despite its relative neglect the reign has traditionally been accepted
Byzantine political society. At one level Basil's reign has been typified as the
the land and manpower resources of the state. Basil's decision in 996 to strengthen
existing legislation prohibiting the greater subjects of the empire, the dunatoi or
"powerful", from seizing the lands of their poorer neighbours, has been interpreted
aristocracy. 9 Yet, while the reign has been seen as the finale to tenth-century
5 W.Felix, Byzanz unddie islamische Welt imfruheren 11. Jahrhundert (Vienna, 1981).
6 J.KForsyth, 'The Chronicle of Yahya ibn Sa'id al-Antaki' (Univ. of Michigan PhD. Thesis, 1977).
Another unpublished doctoral thesis dealing with the eastern frontier during the reign of Basil is that
of W.A.Farag, 'Byzantium and its Muslim Neighbours During the Reign of Basil n (976-1025)'
(Univ. of Birmingham PhD. Thesis, 1979). I would like to thank Professor John Haldon for this
reference.
7 See, for example, A.Poppe, The Political Background to the Baptism of the Rus', DOP 30 (1976),
pp. 196-244; D.Obolensky, 'Cherson and the Conversion of the Rus: An Anti-Revisionist View',
BMGS 13 (1989), pp.244-256 (see below, p.27, n.16)
8 See for example G.Ostrogorsky, History of the Byzantine State (trans. J.Hussey), 3rd edn. (Oxford,
1968), pp.298-315; W.Treadgold, A History of the Byzantine State and Society (Stanford, California,
1997), pp.513-33.
9 Ostrogorsky, History of the Byzantine State, pp.305-7, idem, 'Agrarian Conditions in the Byzantine
Empire in the Middle Ages', in M.Postan (ed.), Cambridge Economic History of Europe Vol. 1, The
Agrarian Life of the Middle Ages, 2nd edn. (Cambridge, 1966), pp. 216-221.
centripetal eleventh-century Byzantium, characterised by a greater political and
administrative focus on Constantinople, and the ascendancy of a new urban and
civilian aristocracy. 10 More recently Jean-Claude Cheynet has suggested an
alternative model of political action in the tenth and eleventh centuries, but one in
which Basil's reign remains the crucial turning point. Cheynet argues that after the
revolts of the early years of Basil's reign, aristocrats ceased to hold public office in
regions where they were also estate owners, and instead exercised official
command in areas where they had no landed interest. Put crudely, those who owned
estates in the eastern half of the empire were appointed to serve the emperor in the
west, and vice versa. 11 The net effect, however, of this geographical partition of
private resources and public authority was that Constantinople, rather than the
regions, became the political fulcrum of empire, with aristocrats increasingly
choosing to be absentee landowners and relocating their households to the capital. 12
10 H.Ahrweiler, 'Recherches sur la societe byzantine au Xle siecle: nouvelles hierarchies et nouvelles
solidarites', 7M 6 (1976), pp.99-124; and in the same volume, N.Oikonomides, 'devolution de
1'organisation administrative de 1'empire byzantin au Xle siecle', pp. 125-52. The implosion of the
Byzantine empire in the second half of the eleventh century has sometimes been attributed to
irreconcilable tensions between the new civilian elite based in Constantinople, and a more established
military aristocracy based in the provinces: see for example S.P Vryonis, 'Byzantium: the Social Basis
of Decline in the Eleventh Century', in Byzantium: its Internal History and Relations with the
Muslim World (Variorum, London 1971), number 2, pp. 160-4.
11 J.C.Cheynet, Pouvoir et contestations a Byzance (Paris, 1990), pp.303-09, 333-36.
Michael Psellos who argued in the 'Chronographia' (his wide-ranging analysis of
fourteen eleventh-century Byzantine emperors) that the revolts of the early years of
Basil's reign caused the emperor to undergo such a profound alteration of character
that he was turned from sybaritic dilettante to austere military man of steel.
According to Psellos after the death of Phokas and the surrender of Skleros, Basil
crushed the greater families of the empire, took public affairs into his own hands,
appointed a series of low-born men as his subordinates, and led his armies in
structural changes, Psellos' bipartite model of the reign, and his allegations that the
emperor was able to re-engineer the government of the empire after the early
There are manifold reasons why modern scholarship has been reluctant to look
more closely at the reign of Basil himself, accepting instead with minimal criticism
the bipartite interpretation offered by Psellos. Paul Magdalino has suggested, for
powerful, warrior emperors such as Basil not only because they are reluctant to
engage with narrative history, but also because they mistrust the odour of autocracy
12 MAngold (ed.), Byzantine Aristocracy (BAR International Series, Oxford, 1984), p.3 stresses the
extent to which Basil II encouraged the great families of the empire to move to Constantinople on a
permanent basis.
13 Psellos, pp.4, 11, 18-22
14 M.Angold, for example, following the model established by Psellos, believes that Basil was a
"complete autocrat" who created an idiosyncratic and personalised style of government which his
eleventh-century imperial successors were unable to follow (The Byzantine Empire 1025-1204
(London and New York, 1984), pp.4-5).
which hovers about their memory. 15 However, in the case of Basil, more prosaic
Despite the scale of the empire's territorial aggrandisement during the later tenth
and early eleventh centuries, the reign of Basil is sparsely covered by medieval
historians writing in Greek. Although Michael Psellos indicates that historians were
at work during Basil's reign itself, the only extant contemporary account is that of
Leo the Deacon. 16 A member of the palace clergy during the first two decades of
Basil's reign, Leo wrote a detailed history of the reigns of Nikephoros Phokas (963-
revolts of Bardas Skleros and Phokas (976-89), as well as an eye witness account of
testimony, written after the mid-990s, terminated with the defeat of Phokas in
Michael Psellos in the third quarter of the eleventh century is the earliest extant
account of the entire period from 976 to 1025. This brief account is to be found at
15 P.Magdalino (ed.), New Constantines: the Rhythm of Imperial Renewal in Byzantium, 4-13
Centuries (Variorum/Aldershot, 1994), pp. 1-2.
16 Leo the Deacon: Leonis Diaconi Caloensis Historiae Libri Decem, ed. C.B.Hase (CSHB, Bonn,
1828), pp. 169-76. For references to other historians active during Basil's reign, see Psellos, p.4. One
of these historians may have been Theodore of Sebasteia (see below, pp.43-45). It is possible that a
chronicle of the reign of Basil II owned by the library of the monastery of St John, Patmos, was
written by another historian working in the later tenth or early eleventh centuries. Mention of this
chronicle is made in an unpublished early fourteenth-century manuscript (K.Snipes, 'The
'Chronographia' of Michael Psellos and the Textual Tradition and Transmission of the Byzantine
Historians of the Eleventh and Twelfth Centuries', ZRVI 27-28 (1989), p. 57).
17 The exact date when Leo wrote his history is not known. However, it is likely that he was writing
after 995. Leo himself mentions that repairs to Hagia Sophia damaged in an earthquake took six years
to complete. According to both Leo and Yahya ibn Sa'id this earthquake happened in 989 (Leo the
Deacon, pp. 175-6; Yahya ibn Sa'id al-Antaki, 'Histoire', ed. and trans. I.Kratchkovsky and
AVasiliev, PO 23 (1932), p.429). The widespread belief among modern historians that Leo wrote
some three years earlier c.992 is based on John Skylitzes' erroneous dating of the 989 earthquake to
986 (Skylitzes (John): loannis Scylitzae Synopsis Historiarum, ed. I.Thurn (Berlin, 1973), pp.331-2.
VII (1071-78). However, while Psellos' assessment of the reign of Basil has
the emperor's personality and the nature of his government, his account is not a
the entire reign. For example, although it contains some brief coverage of the
revolts of Skleros and Phokas, it omits any mention of Basil's wars with Bulgaria. 18
The first surviving narrative account of the reign is that of John Skylitzes compiled
in the later eleventh century as part of a much longer historical synopsis. 19 Yet,
even Skylitzes' treatment is less than comprehensive. The first half of his testimony
is dominated by the revolts of Skleros and Phokas, the second by Basil's campaigns
the eastern frontier, Kiev or Italy. Moreover, his account contains many
historians writing in Greek. From John Zonaras onwards, most accounts represent a
Indeed, in order to amplify the political, diplomatic and military history of the reign
Among those historians who have used Skylitzes' reference to date Leo's historical writings is
H.Hunger, Die hochsprachlicheprofane Literatur der Byzantiner (2 vols., Munich, 1978), i, 368).
18 Psellos (Michael): edition and translation into French Michel Psellos. Chronographie, ed.
E.Renauld (2 vols., Paris, 1926), i, 1-24 (see above, p.l, n.l). There is also a more recent edition
with an Italian translation, Michele Psello Imperalori di Bisanzio (Cronografid), ed. S.Impellizzeri
and trans. S.Ronchey (2 vols., Rome, 1984), i, 8-40. However, since most modern historians
continue to cite the Renauld rather than the Impellizzeri edition, I too have chosen to use Renauld for
all references in this thesis.
19 Skylitzes (John): loannis Scylitzae Synopsis Historiarum, ed. I.Thurn (CFHB, Berlin, 1973),
pp.314-69 (see above, p. 1, n. 1)
record. Yet these references are scarce and rarely yield any firmly datable evidence.
In the eastern half of the empire, Greek written materials pertinent to the long reign
handful of letters from Philetos Synadenos, the judge of Tarsos, and the general
Nikephoros Ouranos, and a few chapters from Ouranos's own military manual. 21
Further west, an anonymous later tenth-century military manual sheds some light
on the strategies and tactics developed by Byzantine armies during warfare against
the Bulgarians. Some miscellaneous reflections about Basil's reign and his
campaigns in the Balkans are also recorded in the later eleventh-century advice
early years of Basil's reign and his dealings with the Balkans can be gleaned from
as well as from saints' lives which reflect on contemporary Greece and Macedonia.
Sparta, Saint Athanasios founder of the Lavra monastery on Mount Athos, and
Thessalonika.25 Apart from the testimony in Skylitzes little is known about the
20 Zonaras (John): loannis Zonarae Epitomae Historiamm Libri XIII-XVIII, ed. T.Biittner-Wobst
(CSHB, Bonn, 1897), iii, 538-569
21 For discussion of the written sources relating to the east of the empire see below, pp. 110, 209-11
22 G.T.Dennis, Three Byzantine Military Treatises (Washington, 1985), pp.246-326. This military
manual is often known as the 'Taktikon Vari' in honour of its first editor, R.Vari. Dennis chooses to
call it 'Campaign Organisation and Tactics'.
23 Kekaumenos: Cecaumeni Consilia et Narrationes, ed. and trans. G.Litavrin (Moscow, 1972)
24 For example a document from the Lavra relates details of a Serbian embassy to Basil in 993: Actes
de Lavra I. Des origines a 1204, Archives de 1'Athos V, eds. P.Lemerle, AGuillou, N.Svoronos,
D.Papchyrssanthou (Paris, 1970), no. 10.
25 Saint Nikon: The Life of St. Nikon, ed. and trans. D.F.Sullivan (Brookline, Mass. 1987); Saint
Athanasios: 'Vitae Duae Antiquae Sancti Athanasii Athonitae', ed. J.Noret, Corpus Christianorum
Series Graeca 9 (Brepols-Turnhout, 1982). Saint Phantinos was born in Calabria in southern Italy,
but spent part of his life in Thessalonika: La vita di San Fantino il Giovane, ed. and trans. E.Follieri
(Brussels, 1993). Saint Photios was the spiritual advisor to Basil H during the emperor's Balkan
campaigns (B.Crostini, 'The Emperor Basil II's Cultural Life', B 64 (1996), p.78). The panegyric
celebrating his life was written by an anonymous author (BHG 1545). The existence of this text,
contained in a manuscript from the Synodal Library in Moscow, was first signalled at the end of the
8
history of Bulgaria after Basil's conquest of 1018, although a series of sigillia
produced before May 1020 give some idea of the ecclesiastical arrangements that
followed annexation. 26
However, if references to the east and to the Balkans are rare, materials written in
Greek concerned with relations with the west, the north and events in
Constantinople are even scarcer. The letters of Leo, Metropolitan of Synada, the
envoy who was sent to Italy to negotiate a marriage alliance between the Byzantine
empire and the Ottoman emperor of Germany in the later 990s, shed light on a
short period of diplomatic dealings with the west.27 Byzantine reaction to closer
affiliations with the Rus of Kiev is limited to references in the poems of John
Geometres.28 Meanwhile, only the life of Saint Symeon the New Theologian, who
spent most of his career in later tenth and early eleventh-century Constantinople,
provides any sense of quotidian life among the civilian elite of the capital during
the reign of Basil.29 Finally, the imperial novels of 988 and 996 are the only non-
narrative sources which reflect on the contemporary governance of the empire. The
30 Zepos and Zepos, lus, i, 259-72; N.Svoronos, 'Remarques sur la tradition du texte de la novelle de
Basile II concernant les puissants', Recueil des Travaux de I'lnstitut d'Etudes Byzantines, Melanges
G.Ostrogorsky II (Belgrade, 1964), pp.427-34
31 A lack of backgrourid narrative context is a problem that has affected E.McGeer's studies of the
career and literary interests of Nikephoros Ouranos, M.Vinson's analysis of the correspondence and
career of Leo of Synada, and A. Cutler's study of Basil's psalter: E.McGeer, 'Tradition and Reality in
the Taktika of Nikephoros Ouranos'. DOP 45 (1991), pp. 129-40; Vinson, The Correspondence of
Leo, passim, A.Cutler, 'The Psalter of Basil IT, in Imagery and Ideology in Byzantine Art
(Variorum/Aldershot, 1992), number III. Very little research has been published on the 'Menologion'
since S.Der Nersessian, 'Remarks on the Date of the Menologium and the Psalter written for Basil
IT, B 15 (1940-1), pp. 104-125 and I.Sevcenko, 'The Illuminators of the Menologium of Basil II',
DOP 16 (1962), pp.243-276. As Barbara Crostini ('The Emperor Basil ITs Cultural Life', pp.53-80)
has so appositely pointed out, the other reason why the arts and literature of the later tenth and early
eleventh-century have been so widely neglected, is the naive belief on the part of many modern
10
In some senses the difficulties of distinguishing the chronological outlines of
Basil's reign can be overcome by consulting the testimonies of historians writing in
languages other than Greek. Greater chronological precision and a more
sophisticated view of the Byzantine east, for example, can be gleaned from a
variety of historians writing in Arabic, Armenian, Syriac and Georgian. Of
particular significance are the contemporary histories of Yahya ibn Sa'id and
Stephen of Taron. Yahya was a Christian Arab doctor who migrated to Antioch
from Cairo during the second half of Basil's reign, a period when members of the
indigenous Christian and Jewish administrative elite were persecuted by the
Fatimid caliph al-Hakim. The extant version of Yahya's chronicle begins in 937/8
and ends with the reign of Romanes HI (1028-34). His historical writings not only
display great chronological, patronymical and toponymical accuracy, but they also
range across most of the contemporary Near East from Egypt to Syria, Iraq,
Byzantium and various Christian Caucasian states. Moreover, as far as Basil's reign
is concerned, Yahya's migration to Antioch allowed him to consult a variety of
histories written in Greek which are no longer extant but which reflect on the
internal history of Byzantium and, to a more limited extent, on Basil's warfare in
Bulgaria. In addition, his use of local chronicle and hagiographical materials
provides a unique view of events in Antioch during the later tenth and early
eleventh centuries. 32
scholars that since Michael Psellos alleged that Basil himself had no interest in the arts, there were no
arts at all.
32 Yahya: Yahya ibn Sa'id al-Antaki, 'Histoire'. ed. and trans. I. Kratchkovsky and A.Vasiliev, PO 23
(1932), pp.372-520, contains coverage from 976 to 1013/4 which is translated into French. From this
date onwards there is no French translation, and one needs to consult in the Arabic original, Yahya
ibn Sa'id al-Antaki, ed. L.Cheikho, CSCO Scriptores Arabia, Series 3, Vol. 7 (Paris/Beirut, 1909),
pp.209-250.1 would like to thank Feras Hamza of Wolfson College, Oxford, for translating excerpts
from the Cheikho edition. All references in this thesis are taken either from the Kratchkovsky or
Cheikho editions. The recent translation of Yahya's text into Italian appeared too late for me to be
able to make all the necessary alterations to the references in this thesis (Yahya al-Antaki Cronache
dell'EgittoFatimide e dell'ImperoBizantino 937-J033, trans. B.Pirone (Ban, 1998)).
11
While the world chronicle of the Armenian Monophysite historian Stephen of
Taron is less finely honed than the testimony of Yahya, it also contains an
the reign of Basil.33 Although its principal concern is with the domestic histories of
Stephen pays close attention to relations between these powers and the Byzantine
from Caucasia who entered Byzantine service during Basil's reign, many of whom
fought in imperial campaigns in the Balkans. Moreover, although the extant version
Thus, the material included in Aristakes' account of Basil's wars against Georgia at
the end of his reign is probably taken from the history of Stephen. 34
Apart from Yahya, Stephen and Aristakes, several other historians writing in
languages other than Greek can also illuminate the history of the Byzantine east in
Basil's reign. Some of these writers were near contemporaries such as Ibn
33 Stephen of Taron: Des Stephanas von Taron armenische Geschichte, trans. H.Gelzer and
A.Burckhardt (Leipzig, 1909), pp. 137-217.
34 Aristakes Lastivert: Recit des malheurs de la nation armenienne, trans. M.Canard and H.Berberian
according to the edition and Russian translation by K.Yuzbashian (Brussels, 1973), pp.2-26.
35 Ibn Miskawayh: Eclipse of the Abbasid Caliphate, ed. and trans. H.Amedroz and D.Margoliouth
(6 vols., Oxford, 1920-1), v, 424-5, 436-9. Ibn Miskawayh was a servant of the Buyid emir Adud ad
Daula who ruled in Baghdad between 978 and 983 (see below, p. 110).
36 Elias of Nisibis: La chronographie de Mar Elie bar Sinaya, Metropolitain de Nisibe, ed. and trans.
L.J. Delaporte (Paris, 1910), pp. 134-142. This text composed in two columns, the first in Syriac and
the second in Arabic, is mainly a short list of entries concerned with Mesopotamia under the rule of
the Bedouin Uqalids. However, it was written by a contemporary and occasionally refers to events in
Byzantium. It confirms, for example, that Basil annexed Bulgaria in 1018.
12
historians and hagiographers writing in Georgian.37 In addition, the accounts of
later chroniclers, such as the Armenian histories of Matthew of Edessa38 and the
continuator of Thomas Artsruni,39 the Syriac accounts of Michael the Syrian and
Bar Hebreus,40 and the Arabic text of the Baghdad historian Abu Shudja al-
Yet, while the eastern sources can add chronological backbone, shed light on the
eastern frontier, and on rare occasions offer insights into the Balkans, it is striking
that when they are aggregated with the Greek sources, large chronological and
regional gaps are still very conspicuous within the overall political history of the
reign. Thus, while many of the sources, both Greek and non-Greek, are liberal in
their coverage of the civil wars of the first thirteen years of the reign led by Bardas
Skleros and Bardas Phokas, their treatment of events after 989 is much thinner.42
During the decade following the civil wars, the Armenian and Arabic records,
above all Stephen of Taron and Yahya, reflect on relations with Fatimid Egypt and
37 The 'Life of John and Euthymios', composed c. 1040, is perhaps the most valuable of the Georgian
materials for the internal history of Byzantium (The Life of John and Euthymios: B.Martin-Hisard,
'La Vie de Jean et Euthyme: le statut du monastere des Iberes sur 1'Athos', REB 49 (1991), pp.67-
142). Material composed by Sumbat Davitidze c. 1030 which is included in the much longer text of
the Georgian Royal Annals, comments on relations between Byzantium, the Black Sea kingdom of
Abasgia and various Caucasian princedoms (Georgian Royal Annals: R.Thomson, Rewriting
Caucasian History: The Georgian Chronicles (Oxford, 1996), pp.274-285; S.H.Rapp, 'Imagining
History at the Crossroads: Persia, Byzantium and the Architects of the Written Georgian Past' (Univ.
of Michigan Ph.D. thesis, 1997), pp.492-3. See below, pp. 119-120, for further discussion of the
Georgian materials.
38 Matthew of Edessa: Armenia and the Crusades in the Tenth to Twelfth centuries: the Chronicle of
Matthew of Edessa, trans. A.E.Dostorian (Lanham/New York/London, 1993), pp.34-50.
39 Thomas Artsruni: History of the House of the Artsrunik', trans. R.W.Thomson (Detroit, 1985),
pp.368-71.
40 Michael the Syrian: Chronique de Michel le Syrien, Patriarche Jacobite d'Antioche (1169-99), ed.
and trans. J.B.Chabot (Paris 1905-1910), pp. 132-146; Bar Hebreus: The Chronography of Gregory
Abu'l Faraj, the Son of Aaron, the Hebrew Physician, Commonly Known as Bar Hebreus, ed and
trans. E. A. Walk's Budge (London, 1932), pp. 175-189.
41 Al-Rudhrawari: Eclipse of the Abbasid Caliphate, ed. and trans. H.Amedroz and D.Margoliouth (6
vols., Oxford, 1920-1), vi, 6-7, 23-35, 115-119. For more on documents from Buyid Iraq relevant to
the domestic and external history of Byzantium in Basil's reign, see below, p. 110, n.4.
13
the northern Syrian client state of Aleppo, and on the absorption of the Iberian
(Georgian) princedom of Tao. Meanwhile, Skylitzes offers sporadic and confused
coverage of warfare with the Bulgarians during the same period. Between the early
1000s and 1014 there is almost complete silence about Asia Minor, the east, the
Balkans and Constantinople itself. Further west, a few dated references can be
extracted from the testimonies of chroniclers writing in Latin, such as Lupus
Protospatharius. Nonetheless, such chronicles tend to be very short and concerned
predominantly with Byzantine Italy and relations with the Ottoman emperors of
Germany rather than with events in the heartlands of the Byzantine empire.43 The
historiographical gloom elsewhere in the empire only begins to lift after 1014,
when Yahya includes some very parochial material on relationships between
Byzantium and Aleppo, and Skylitzes contains some detailed snapshots leading up
to the annexation of Bulgaria. However, it is only at the very end of the reign that
we find more sustained coverage. Many of the narrative sources comment on the
absorption of the southern Armenian kingdom of Vaspurakan (1019-21), Basil's
campaigns against George of Abasgia and Iberia in 1021/2, and the
contemporaneous revolt against imperial authority in central Anatolia led by
The obvious short comings of the primary written sources mean that the historian
of Basil's reign cannot afford to ignore the material record of early medieval
Byzantium. And here, indeed, there are real signs for hope. In the century which
has passed since Schlumberger produced his analysis of the reign, many hundreds
of lead seals and coins from the later tenth and early eleventh centuries have been
42 For coverage of the Skleros and Phokas revolts in the primary written sources see below, pp. 109-
110.
43 Lupus Protospatharius, MGH SS V, pp.55-7.
44 See below, pp.283-5
14
discovered, analysed and published. In more recent years archaeological
excavations and surveys in many of the former provinces of the Byzantine empire
have begun to expand in scale and ambition. Gradually scholars have begun to use
this ever-expanding material archive to investigate the structures underpinning
medieval Byzantine society. A series of recent studies about Byzantine
administration, the army, and the economy, have illuminated the resource-base of
the medieval Byzantine state.45 Yet, in order to maximise the potential of the
material archive the right questions must be asked of it. For example, it is
unrealistic to expect material evidence, which so often cannot be dated accurately
itself, to plug geographical or chronological lacunae in the written sources. Nor
should the material record be asked to provide answers to very specific political,
administrative or chronological problems about which the written texts are silent.
Such inquiry will either fail through the lack of appropriate evidence, or will
simply result in the highly selective use of material to support preconceived
models. Instead, the material archive is likely to provide the greatest insight when it
is used to provide long-term structural backdrops, against and within which, the
existing narrative sources can then be examined.
The ultimate ambition of any investigation into the reign of Basil II must be to
develop a new analytical narrative of the political, military and diplomatic history
of the Byzantine empire in the later tenth and early eleventh centuries to replace
Schlumberger's extremely old synthesis. Yet, in the time that has passed since
Schlumberger wrote his account, no new substantial narrative source from the tenth
or eleventh centuries has been uncovered by historians. In the absence of such a
new contribution from the medieval historical record, it would be unwise to
attempt to write the history of Basil's reign by simply synthesising the extant
45 This secondary literature will be examined in more detail in chapters four to six.
15
written sources and adding ephemeral details from the material record. Such an
undertaking is unlikely to expand significantly on Schlumberger's very competent,
if old-fashioned, general narrative, or Forsyth's more recent and skilful
reconstruction of relations with the eastern frontier. Instead, a new critical narrative
of Basil's reign can only be constructed by turning the very limitations of the
sources to best advantage. The historian of the reign must find new ways of
analysing existing written sources, in particular the medieval historical record,
while at the same time discovering how to integrate such texts with the material
archive. Furthermore, in order to provide some analytical coherence to such a long
reign, the ambition behind the reconstruction of political history must be expanded.
The two-dimensional concatenation of personalities, battles, treaties, and coups
described in the medieval sources, must be underpinned by a consideration of how
political society worked in the context of the structures and resources of the
empire, and in the face of the capacities and motivations of Byzantium's
geographical neighbours. In other words central to any new analysis should be an
understanding of how Byzantine government in all its senses, political,
administrative, military and diplomatic, functioned.
My doctoral research has been shaped by many of the prescriptions outlined above.
However, the thesis in which this research is presented does not represent the final
analytical narrative of the whole of Basil's reign. Instead it amounts to six
preliminary studies upon which such a narrative should be built. Yet, while the
approach adopted and the evidence considered in each of these studies varies
widely, each aims to contribute to a better understanding of the nature, changes and
continuities in Byzantine government during Basil's reign. In all cases the approach
employed has required an expansion of the canvas to a perspective wider than that
of Basil's reign alone. Whether dealing with literary or material sources, the focus
16
of each analysis has extended chronologically beyond the reign of Basil himself.
Moreover, while the principal concern of this thesis is with government within
Yet, while there has been broadening, there has also been narrowing, so that the
project could be compatible with the time constraints of a doctoral thesis. The
subject-matter of all six studies is predominantly secular, with less attention paid to
the church, either in the capital, or in the provinces. Furthermore, the thesis
concentrates mainly on the eastern half of the empire, that is to say Anatolia and
the frontier regions beyond the Taurus and Anti Taurus mountains. Given Basil's
conquest of Bulgaria and his sobriquet the "Bulgarslayer", the decision to look east
may seem odd. However, the nature of Byzantine relations with Bulgaria will never
be understood without first establishing the background of long-term economic,
social, political and military developments in the eastern half of the empire both
before and during Basil's reign. Geographically Anatolia and the eastern frontier
regions formed the largest land mass within the Byzantine empire. The east was the
area where Byzantine armies in the decades preceding Basil's reign had achieved
their most significant territorial conquests. It was also here that the greatest
aristocratic families held estates and exercised official command during the tenth
century. More directly relevant to Basil's reign, it was from the eastern half of the
empire that the revolts, led by Bardas Skleros and Bardas Phokas, which
punctuated the first thirteen years of Basil's reign, were launched. It was also in
the east of the empire, in Cappadocia, that Nikephoros Phokas and Nikephoros
17
The first three chapters of the thesis are concerned with the medieval
historiography of Basil's reign. They are predicated on the understanding that the
medieval historical record must be at the heart of the construction of any modern
narrative of the reign. Although this record contains vast lacunae and serious flaws,
it is the only source of evidence which offers a chronological spine to the reign and
a series of explicit contemporary interpretations of the political history of the later
tenth and early eleventh centuries. As such, it cannot be circumvented. Yet, at the
same time as acknowledging the centrality of the medieval historians to any
understanding of Basil's reign, the first three chapters are predominantly concerned
with the problem of how to utilise the medieval historical record in the construction
of an analytical narrative, which allows us to move beyond Schlumberger's existing
model. In examining this problem, these chapters argue that a new narrative cannot
be constructed simply by comparing and contrasting information and interpretation
from different historical accounts. Instead, they suggest that to get the most out of
the historical record we need to consider the methods and motivations behind each
medieval historian's presentation of the reign. These chapters argue that such a
model of analysis requires investigating the surviving medieval narratives in two
related ways: first, by looking directly at each text and assessing how its appraisal
of Basil's reign relates to the wider narrative as a whole; and second by considering
how each text is conditioned by the contemporary literary, social and political
contexts in which it was written. This second approach is particularly relevant to
the reign of Basil, since so many of the historical accounts which report on this
In the case of Basil's reign there are at least four historians whose narratives should
be subjected to such rigorous analysis: Michael Psellos, John Skylitzes, Yahya ibn
Sa'id and Stephen of Taron. Yet, such analysis is both difficult and time-
is
consuming, since it requires a detailed knowledge, not only of the entire text of
each historian, but also of the very different background contexts against which
they were composed. It is a task which is complicated further by the fact that
relatively little sustained, modern scholarship has been dedicated to the texts of the
historians in question. During the period when this thesis was being researched and
compiled, the only accessible detailed study was Forsyth's unpublished Ph.D.
Thus, while it is an ambition of future research to look at all the major historical
accounts of the reign of Basil II in their wider contexts, the three historiographical
chapters of this thesis begin this in-depth appraisal by concentrating on just one
historian: John Skylitzes, whose account is the first connected narrative of the
Since Skylitzes is not well-known to most historians of Byzantium, the first of these
chapters begins by summarising the small body of scholarship which has been
dedicated to the author and text of the 'Synopsis Historion'. It offers a resume of
existing research into the manuscripts of the 'Synopsis', Skylitzes' biography, and
the author's working methods. The chapter then moves on to a detailed textually-
based analysis of Skylitzes' use of source materials. At the end of the chapter, the
past including the reign of Basil are highlighted. The second chapter considers how
the literary, social and political milieux in which Skylitzes was writing influences
his text as a whole, and his coverage of the reign of Basil in particular. The third
chapter focuses more directly on the reign of Basil itself. It takes as its raw subject
46 Forsyth, 'The Chronicle of Yahya ibn Sa'id', chapters 1-6 contain a detailed discussion of Yahya's
sources and his place within the traditions of Arab and Byzantine historiography
19
matter the revolts of Bardas Skleros and Bardas Phokas (976-89), that period of the
reign which is covered in most detail by all the historians of Basil's reign including
Skylitzes. It then compares Skylitzes' coverage of this period with the accounts
presented by other historians. The purpose of this analysis, however, is not to
produce a new, comprehensive narrative of the revolts themselves, but to learn
more about the composition of the 'Synopsis Historion'. By comparing the various
accounts of these rebellions, this analysis will suggest that a hitherto
unacknowledged source, attributable to the general Bardas Skleros, underpins the
first thirteen years of Skylitzes' account of the reign of Basil.
Taken together the first three chapters of this thesis will illustrate the extent to
which Skylitzes' appraisal of the reign of Basil is shaped by the demands of history
writing at the end of the eleventh century, demands which were not only social and
literary in nature, but also political. These chapters will suggest that contemporary
historiographical requirements compelled Skylitzes to take a retrospective view of
the Byzantine past. In the case of Basil's reign Skylitzes' backward-looking
perspective means that most attention is focused on the aristocratic families of the
empire, and least on the person of the emperor himself. This anachronistic
treatment can obscure a clear understanding of the nature of relations between the
constituent members of the Byzantine political elite.
In order to discuss the nature of Byzantine government during the reign of Basil II,
the historian needs to transcend the anachronisms of the representation of relations
within the political elite offered by the later eleventh-century Greek
leading members of the elite during Basil's reign are to be meaningfully analysed,
20
attention first needs to be paid to the structures and processes which underpinned
contemporary political authority. The second half of this thesis, again arranged in
three chapters, begins this structural analysis by looking at the economy and the
administrative institutions and practices of the eastern half of the Byzantine empire
in the later tenth and early eleventh centuries. Although this study will draw upon
the medieval historical accounts where appropriate, the material record, above all
the lead seals, will be of crucial importance.
The principal question asked in these final three chapters is to what extent did
Byzantine administrative structures and processes change in the course of Basil's
reign? The scene is set in the fourth chapter which investigates the variables of
geography, climate and economy both in Anatolia and in the eastern territories
lying beyond the Taurus and Anti Taurus mountains. This chapter considers how
economic expansion both facilitated and constrained administrative change.
Chapters five and six deal with administration more directly. Chapter five begins
with a synopsis of the background to administrative developments in the eastern
half of the empire in the decades before Basil came to power, concentrating on the
bureaucratic ramifications of military expansion during the second and third
quarters of the tenth century. The second half of chapter five goes on to discuss
administrative changes in the Anatolian heartland of the Byzantine east during
Basil's reign itself, outlining the processes by which the locality was gradually
demilitarised and penetrated by civil officials. Chapter six moves eastwards to
those regions beyond the Taurus and Anti Taurus which had been conquered
shortly before Basil's accession. In this instance the flexibility of administrative
practice will be stressed, and key continuities between imperial bureaucracy and
argued that there were crucial differences between administration in the Anatolian
21
heartland of Byzantine Asia Minor and on the eastern periphery, both chapters five
and six will stress how bureaucratic practice in all regions was predicated on the
imperial desire to extract maximum financial gain from the Byzantine provinces.
However, it will be suggested that the penetration of the state into the locality may
have been less heavy-handed than is sometimes believed.
Clearly the analysis undertaken in chapters four to six of this thesis can only
partially illuminate the government of the Byzantine empire during the reign of
Basil n. It will shed light on the abstract articulation of the authority of the emperor
in the locality, but will not deal closely with the more practical issues of how that
authority was used and abused by its executors. However, during the course of the
second half of the thesis it will be argued that it is only once the structures behind
imperial authority are established, that the complexities of the relationships
between the public power of the state and the private power of the state's
functionaries can be fully understood. While exploration of these more intricate
relationships remain the ambition of future research, some preliminary findings are
22
Chapter One
I. Introduction
empire during the reign of Basil II cannot be overstated. Skylitzes' appraisal written
towards the end of the eleventh century is the first surviving connected narrative of
the reign in Greek. It is the principal source for several of the most politically
significant events of the reign, including the revolts of the generals Bardas Skleros
and Bardas Phokas, and the long war of attrition against the first Bulgarian empire.
It is the primary source around which most later historians, both medieval and
modern, have shaped their chronological presentations of the later tenth and early
eleventh centuries. l
1 Those medieval historians who based their accounts on Skylitzes' narrative include. Kedrenos
(George): Georgius Cedrenus, ed. I.Bekker (CSHB, Bonn, 1938-9), ii, 416-480; Pseudo-Psellos:
Michaeli Pselli Historia Syntomos, ed. and trans. W.Aerts (CFHB XXX, Berlin, 1990), pp. 105-109;
Zonaras (John): loannis Zonarae Epitomae Historiarum Libri XHI-XVIII, ed. T.Biittner-Wobst
(CSHB, Bonn, 1897), iii, 538-569; Manasses (Constantine): Constantini Manassis Breviarum
Chronicum, ed. O.Lampsidis (CFHB XXXVI, Athens, 1996), pp.314-321; Glykas (Michael):
Annales (Biblos Chronike), ed. I. Bekker (CSHB, Bonn, 1836), pp.575-9. Skylitzes' account also
provides the main template for Ostrogorsky, History of the Byzantine State, pp.298-315; Treadgold,
History of the Byzantine State and Society, pp. 513-33. Schlumberger (L 'Epopee byzantine, i, 327-
777, and ii, passim) adopts a more complex chronological structure which integrates the narrative
outlines of both Skylitzes and Yahya ibn Sa'id. This is an approach also followed by M.Whittow in
The Making of Orthodox Byzantium (Oxford, 1996), pp.358-90.
23
account of the later tenth and early eleventh centuries. Instead it is a fusion of the
writings of other earlier historians. Thus, one of the most significant problems
Skylitzes' narrative and the different source materials which underpin it.
Furthermore, the 'Synopsis' represents a highly abbreviated account of two-and-a-
half centuries of the Byzantine past. All versions of Skylitzes' 'Synopsis Historion'
run from 811 to 1057, and if the author of the 'Continuation' of Skylitzes
Another important consequence of the text's brevity is that its geographical and
chronological coverage tends to be extremely uneven. In the case of Basil's reign,
the first half of Skylitzes' account is dominated by the civil wars of the first thirteen
years of the reign, and the second half by warfare against Bulgaria. As far as the
first half of the reign is concerned, chapters one to ten (more than twenty percent of
the forty-seven chapters Skylitzes dedicates to Basil) are concerned with the first
three years of the reign and the revolt of the general Bardas Skleros (976-9).4 The
next two chapters deviate briefly from the theme of internal revolt by dealing with
warfare between Byzantium and Bulgaria in the first decade of the reign,
2 The 'Continuation' of the 'Synopsis' has been published by E.T.Tsolakes (Skylitzes Continuatus:
'H 2uv%eia r!j$ xpovojpaji'ag TOU 'latawou S/cvAfrgy., 'Eraipe/a MaKeftoviKojv STOU&MV (Thessalonika,
1968). For a summary of the arguments about whether Skylitzes Continuatus should be identified
with Skylitzes himself see below, p.36, n.41. Because it is mutilated at the start the only manuscript
of the main text of the 'Synopsis' (as opposed to the 'Continuation') which does not begin in 811 is
U [Vindob Hist. Gr. 74]. This manuscript begins with the reign of Basil n (Thurn, Joannis Scylitzae,
p.xxvi)
Skylitzes (John): loannis Scylitzae Synopsis Historiantm, ed. I.Thurn (Berlin, 1973) contains
Skylitzes1 narrative to 1057. The coverage of the reign of Basil is to be found between pages 314 and
369.
24
concentrating on Basil II's defeat in the Haimos mountains in 986 against the army
of the Tsar Samuel.5 However, this deviation is short. Chapters fourteen to nineteen
return to the theme of internal insurrection, and cover the second revolt of Bardas
Skleros and the contemporaneous rebellion of Bardas Phokas, events which lasted a
little over two years, between 987 and 989.6 With the death of Phokas and the
surrender of Skleros, the second half of Skylitzes' testimony for the reign is almost
exclusively concerned with the Bulgarian wars. Seventeen of the twenty-eight post-
989 chapters are concerned with conflict in the Balkans.7 Meanwhile, most of those
chapters dedicated to other matters are extremely short, often comprising little more
o
than a handful of lines. Moreover, of the chapters dealing with Bulgaria, three-
quarters are concerned with the period 1014 to 1018.9 There is very little material at
all in Skylitzes' account which deals with the period between 1005 and 1014.
The net result of Skylitzes' twin focus on early internal strife and the final stages of
warfare in the Balkans, is that large regions of the internal Byzantine world and the
empire's relations with its neighbours are rarely touched upon. For example,
Skylitzes' coverage of events within Constantinople itself is limited to short notices
concerning natural disasters, the accessions and deaths of patriarchs, and occasional
urban improvements, such as the rebuilding of Hagia Sophia after the earthquake of
989 and the restoration of the Aqueduct of Valens towards the end of the reign. 10
After the emperor's victories over Skleros and Phokas in 989 political relations
within the elite of the empire are treated brusquely. Basil II's novel against the
'powerful' is cited in the same short, undated chapter which deals with the
4 Skylitzes, pp.314-328
5 Ibid., pp.328-331
6 Ibid., pp.332-339
7 The Bulgarian testimony is to be found between pp.341-66 of Skylitzes' account of the reign
(chapters 23 to 44).
8 For example chapter 33 dealing with the destruction of the church of the Holy Sepulchre in
Jerusalem by Fatimid forces in 1009 extends to only seven lines (Skylitzes, p.347).
9 All of chapters 35 to 44, bar chapter 39 (Skylitzes, pp.348-66)
25
emperor's decision to imprison Eustathios Maleinos, one of the rebels who
supported Bardas Phokas in 987. n At the end of his account of the reign, Skylitzes
includes a cursory survey of the 1022 revolt of Nikephoros Phokas and Nikephoros
Xiphias in Cappadocia. 12 Virtually no mention is made of Basil II's fiscal policy,
beyond two short references to the emperor's imposition of a measure known as the
allelenguon. The first reference notes that this decree, issued c. 1002, ordered that
the taxes of those 'poor' landowners who defaulted should be paid by their
'powerful' neighbours. The second reports that the emperor refused a request by the
patriarch Sergios that the tax should be lifted after the end of the Bulgarian wars in
1018. 13
Further afield, references to the empire's dealings with its neighbours, especially
*
those outside the Balkans, are extremely cursory and confused. For example
relations between the empire and its eastern neighbours from 990 to 1022 are
described in a garbled account which is less than a page long. 14 The surrender of the
Armenian princedom of Vaspurakan in the final decade of the reign is to be found
in a compressed chapter, which also contains a notice dealing with a joint
Byzantino-Rus expedition against Cherson. 15 More significantly, the events
surrounding the alliance between Basil and Vladimir prince of Kiev in 988-9 are
summarised in a two-line parenthesis inserted into Skylitzes' coverage of the revolt
of Bardas Phokas. Skylitzes merely mentions that Basil's armies included Russian
troops, which had been dispatched after the emperor had married his sister Anna to
Vladimir archon of the Rus. Thus, no reference is made to the official conversion
26
of the Rus to orthodox Christianity which accompanied Basil and Vladimir's
alliance, nor to the mysterious siege of Cherson conducted by the Rus after the
the second decade of the eleventh century, organised by a local notable from Ban
called Meles. 17 Finally, just before Skylitzes reaches the end of his testimony of the
reign, he refers to the advance expedition to Sicily led by the eunuch Orestes, which
was intended to prepare the way for Basil's own invasion of the island, a campaign
ic
which was brought to a premature end by the emperor's death in 1025. Apart
from these brief references, Skylitzes makes no reference at all to the empire's
dealings with the Ottoman emperors of Germany, the early French Capetians or the
Pope.
16 Ibid., p.336. Skylitzes' lack of interest in relations with the Rus is reflected in the accounts of
other historians writing in Greek about Basil's reign. Neither Leo the Deacon nor Michael Psellos
mention the conversion of the Rus. Leo the Deacon does not even refer to the participation of
Russian troops within the imperial army during the Phokas revolt, while Psellos makes only a
passing allusion to their presence (Leo the Deacon, pp. 173-4; Psellos, p.9). For more details on the
alliance struck between Basil and Vladimir in 988-9 one must turn to Yahya ibn Sa'id, Stephen of
Taron, and the 'Russian Primary Chronicle' (Yahya ibn Sa'id, PO 23, pp.423-4; Stephen of Taron,
p.211; The Russian Primary Chronicle, Laurentian Text, ed. and trans. S.H.Cross and
O.P.Sherbowitz-Wetzor (Cambridge (Mass.), 1953), pp. 110-119). However, the uneven nature of
the source materials makes it difficult to establish the chronology of all the principal events
surrounding the Byzantino-Rus alliance. The accounts presented by Yahya and Stephen are brief,
while the author of the much fuller story in the 'Primary Chronicle' was writing considerably later,
in the early twelfth century. The motivation for the Rus attack on Cherson has proved particularly
problematic to elicit from the sources. Poppe, 'The Political Background to the Baptism of the Rus',
pp. 196-244, believes that the Rus attacked Cherson in fulfilment of the military agreement with
Basil II. He maintains that their attack was aimed against insurrectionists within Cherson who
favoured the rebel Bardas Phokas. A more traditional reading is preferred by Obolensky, 'Cherson
and the Conversion of the Rus', pp.244-256, who argues that Vladimir agreed to send Basil troops in
return for a marriage with Anna and conversion to Christianity. Vladimir dispatched his troops to
Byzantium but Basil failed to send his sister in return. As a result Vladimir attacked Cherson. The
modern historiographical debates surrounding the conversion of the Rus are summarised in
S.Franklin and J.Shepard, The Emergence of Rus 750-1200 (Cambridge, 1996), pp. 159-69
17 Skylitzes, p.348
27
Byzantine political and military history in the later tenth and early eleventh
achieved? The most obvious way of approaching Basil's reign through Skylitzes'
text is to compare the material in his account with information and interpretation
contained in other written sources, independent of the 'Synopsis', which also report
on the same period. As we shall see later in this chapter, this is an approach which
has been fruitful in the investigations conducted by Jonathan Shepard into
this thesis it will be argued that the most fruitful method of approaching Skylitzes'
18 Ibid., p.368
19 See below, pp.45-8
20 See above, pp.7, 11-14
28
treatment of Basil II lies less in trying to improve upon the testimony he transmits
of selection, presentation and interpretation which underpin his text. In this sense
the key questions to be asked are: how and why does Skylitzes offer his reader this
particular text? Behind these questions lies the explicit acknowledgement that all of
of Basil II. In chapter two this textual analysis will provide the basis for a broader
29
discussion of how Skylitzes and his narrative relate to the wider literary, social, and
indeed, political contexts of the period in which the 'Synopsis' was compiled.
Since Skylitzes' text is often cited, but rarely studied in detail by modern scholars,
the first chapter will be prefaced with a brief overview of the current state of
Skylitzes' scholarship, including a summary of recent research into the author's
own biography. Two additional caveats, however, need to be made explicit.
where relevant, the principal engagement will be with the main 811-1057 section of
the 'Synopsis'. Since this thesis is predominantly concerned with the eastern half of
the empire and political elites, the interpolations that pertain to western Bulgaria in
manuscript U [Vindob Hist. Gr. 74] will not be examined in detail.22
It is only in the past thirty years that John Skylitzes has come to be widely
recognised as an independent historian. Before I.Thurn published a critical edition
of the main body of the 'Synopsis Historion' (811 to 1057) in 1973 and
E.T.Tsolakes produced an edition of the 'Continuation' (1057-79) in 1968, most
modern scholars only had access to Skylitzes through the world chronicle of
George Kedrenos. This text compiled in the later eleventh century and published in
the Bonn Corpus edition of 1838/9 includes a verbatim copy of Skylitzes'
testimony from 811 to 1057.23 The obscurity of the 'Synopsis Historion' was
22 See below, p. 31
23 John Skylitzes: loannis Skylitzae Synopsis Historiarum, ed. I.Thurn (Berlin and New York, 1973)
(see above, p. 1, n. 1); Skylitzes Continuatus: 'HSwe% ia rfc %povoTpa4ia$ roS 'luawouSKuJM/q, ed.
E.T.Tsolakes, 'Eraipcia Mcurcdoi/iKoh' "Snoud&v (Thessalonika, 1968); George Kedrenos: Georgius
Cearenus, ed. I.Bekker, (CSHB, Bonn, 1938-9), Vol. 2
30
reflected by a long-standing lack of scholarly interest in Skylitzes and his text. In
audience of credulous monks.24 Further comment usually only came from those
historians interested in the early history of Bulgaria. Attention in this case was
focused on a fourteenth-century manuscript of the text, U [Vindob Hist. Gr. 74],
which contains some information about the western Balkans during the reigns of
Basil II and Michael IV not found in other manuscripts. This extra material was
believed either to have been interpolated into the 'Synopsis' by the Macedonian
bishop Michael of Devol in the early twelfth century, or to represent a fuller form
of the 'Synopsis' closer to Skylitzes' original account.25
It was only with the appearance of the critical editions of the main text and its
'Continuation' between 1968 and 1973 that the 'Synopsis' became the object of
more urgent investigation. Since the early 1970s a small body of research has been
published pertaining to Skylitzes' biography, his working methods, his sources, the
manuscripts of his testimony, and his relationship with the author of the
'Continuation'. However, it should be stressed that few historians have displayed a
sustained interest in Skylitzes despite the existence of the critical editions. For
example, it is striking that Skylitzes and his text have not been as closely examined
as the writers of other historical synopses in Greek, such as John Malalas or
31
Theophanes the Confessor.26 A translation and commentary in French is promised,
but has yet to appear in print.27 Only the Madrid manuscript [M] attracts consistent
attention in the secondary literature. As the only illustrated Byzantine history, the
'Madrid Skylitzes' is frequently discussed by art historians. Moreover, the fact that
the manuscript was copied and illustrated in the mid-twelfth century outside the
political borders of Byzantium in Norman southern Italy means that it constitutes
invaluable evidence for the transmission of manuscripts between the empire and
western Europe. Nonetheless, investigation into the 'Madrid Skylitzes' has tended
to be narrowly focused: much attention has been paid to the visual imagery, little to
the relationship between this manuscript and the other copies of the 'Synopsis'. As
a result, research into the 'Madrid Skylitzes' has not contributed greatly to a more
general understanding of the production and dissemination of Skylitzes' 'Synopsis'
08
in the centuries after the text was written.
When I.Thurn published his critical edition the complicated manuscript history of
the 'Synopsis' was finally elucidated. Thurn listed nine twelfth- to fourteenth-
century manuscripts containing the narrative from 811 to 1057. Although he
('John Scylitzes and Michael of Devol', ZRVI 10 (1967), pp. 163-70). Ferluga supported the idea that
Michael of Devol was responsible for the additional material in U.
26 E.Jeffreys, R.Scott, B.Croke et al. (trans.), The Chronicle of John Malalas (Melbourne, 1986);
E.Jeffreys et al. (eds.), Studies in John Malalas (Sydney, 1990); C.Mango and R.Scott (trans.), The
Chronicle of Theophanes Confessor (Oxford, 1997); E.Jeffreys ('The Attitudes of Byzantine
Chroniclers towards Ancient History', B 49 (1979), pp. 199-238) has discussed Malalas and
Constantine Manasses; R.Macrides and P.Magdalino, "The Fourth Kingdom and the Rhetoric of
Hellenism', in P.Magdalino (ed.), The Perception of the Past in Twelfth-Century Europe (London,
1992), pp. 120-39) discusses eleventh- and twelfth-century synoptic historical writing, but only deals
with Constantine Manasses, Michael Glykas and John Zonaras in detail. The standard summary of
Skylitzes and the 'Synopsis' is found in H.Hunger, Die hochsprachliche profane Literatur der
Byzantiner, (2 vols., Munich, 1978), i, 389-93.
2 In conversation with Bernard Flusin in September 1998 I learned that he is producing a
translation. His colleague Jean-Claude Cheynet is working on a commentary.
28 For example: I. Sevcenko, 'Poems on the Deaths of Leo VI and Constantine VII in the Madrid
Manuscript of Skylitzes', OOP 23-4 (1969-70), pp.185-228; idem., The Madrid Manuscript of the
Chronicle of Scylitzes in the Light of its New Dating', in I.Hiitter (ed.), Byzanz und der Westen
(Vienna, 1984), pp.117-130; N.Wilson, 'The Madrid Scylitzes', Scrittura e Civilta 2 (1978), pp.209-
32
acknowledged that later manuscripts also existed, it was from the medieval
manuscripts that he compiled his edition. 29 Of the manuscripts in question, none is
a contemporary autographed copy, although three [A,M, and O] are datable to the
twelfth century and thus, as we shall see from the biographical discussion below,
were copied within a century of the author's own lifetime.30 Of the three twelfth-
century manuscripts, A is probably the oldest. 31 Thurn also suggested that the
surviving medieval manuscripts fell into three main families: ACE, VBO, and
MNU. It was to this final family that the version of the text contained in the world
chronicle of Kedrenos was connected.32 Thurn also produced a summary of the
interpolations to be found within the medieval manuscripts. Apart from registering
the additional Bulgarian material in manuscript U, he drew attention to the
insertions included in the other manuscripts, such as a eulogy to the emperor John
Tzimiskes in the ACE family, and eleven poems describing the deaths of several
tenth-century emperors in the Madrid Skylitzes. 33 Recently manuscript O, which
was known to Thum but which he had not seen for himself, was rediscovered in a
museum at Ochrid in Macedonia. The fact that this manuscript also contains the
'Continuation' increases to five the number of manuscripts known to continue
Skylitzes' testimony as far as 1079. Thurn himself had suggested that only four
texts included the 'Continuation'. 34
33
/'//". Skylitzes' biography
At the same time as critical editions of the 'Synopsis' and the 'Continuation' were
34 J.M Olivier, 'Le "Scylitzes" d'Ochrid retrouve', BZ 89 (1996), pp. 417-19; Thurn, loannis
Scylitzae, pp. xxii discusses manuscript O. The other four manuscripts containing the 'Continuation'
are A,V,B and U.
35 Thurn, loannis Scylitzae, pp.vii-viii contains a very short biography of Skylitzes; see also
S.Antoljak 'Wer konnte eigentlich Joannes Skylitzes sein?' Acts of the 14th International Congress
1971, (Bucharest, 1974), pp.677-82; W.Seibt, 'loannes Skylitzes - Zur Person des Chronisten', JOB
25(1976), pp. 81-6
36 The title sequence to the 'Synopsis' in manuscript A [Vind. Hist. gr. 35] has a lacuna where the
word megas was once inscribed In the title sequence contained in manuscript C (Coisl. 136]
Skylitzes is also called the megas droungarios (Skylitzes, p.3). The prefix megas was added to the
office of droungarios of the Vigla in the second half of the eleventh-century (R.Guilland,
Recherches sur les institutions byzantines (2 vols., Amsterdam, 1967), ii, 573).
For the memorandum (hypomnesis) sent by John and the reply (lysis) sent by the emperor see
J.Zepos and P.Zepos, lus Graecoromanum, i, 319-325.
34
time he identified John the Thrakesian at a slightly earlier stage in his career.
According to a novel dated to June 1090 one John proedros and droungarios of the
vigla was also eparch of Constantinople. Seibt believed this figure was John the
Nicholas Skleros was droungarios in 1084, Seibt concluded that Skylitzes' must
have been appointed in the second half of the 1080s. Taking all these materials
together, Seibt concluded that the author of the 'Synopsis Historion' was, by the
1090s, a high-ranking government official, who in his position as megas
droungarios of the vigla, occupied the most senior position within the Byzantine
judiciary. 38
For Seibt the fact that other Byzantine historians knew the author of the 'Synopsis
Historion' as John the Thrakesian was additional proof that the John Skylitzes cited
in the manuscript title sequences was the same individual as John the Thrakesian,
the high-ranking Komnenian official. One of these other historians was George
Kedrenos, who referred to the synoptic historian John the Thrakesian in the preface
to his own world chronicle. Since he copied the 'Synopsis' of Skylitzes verbatim
into his own text, it can be safely assumed that Kedrenos equated John the
Thrakesian with Skylitzes.39 In addition, Seibt noticed that the twelfth-century
synoptic historian John Zonaras also knew Skylitzes as the Thrakesian. In his
coverage of the death of Isaac Komnenos, Zonaras mentions that one story of the
38 Seibt, 'Zur Person', pp.81-3 and Die Skleroi (Vienna, 1976), pp.96-7. In the ninth and tenth
centuries the occupant of the senior position at the vigla was the emperor's military chief of palace
security. By the end of the eleventh century this official had become the senior judicial officer
within the Byzantine empire (N.Oikonomides, 'L'evolution de 1'organisation administrative de
1'empire byzantin au Xle siecle', pp. 133-4). Guilland (Recherches, ii, 573) lists the droungarioi of
the Vigla during the latter part of the eleventh century but does not mention Skylitzes.
35
the account conveyed in the 'Continuation' of the 'Synopsis Historion'. 40 With this
observation Seibt also provided the most convincing evidence that the author of the
Seibt's neat fusion of the manuscript title sequences, the evidence from the imperial
and patriarchal documents, and the information conveyed by other Byzantine
historians, was however, subject to one difficulty: George Kedrenos identified John
the Thrakesian not as a kouropalates, but instead as a protovestiarios, a position
which by the later eleventh century was almost always held by a member of the
ruling imperial dynasty. 42 Seibt's solution was to argue that Kedrenos had made a
transcription error, and that rather than protovestiarios he meant to write
protovestes or protovestarches. Armed with this emendation, Seibt suggested that
John Skylitzes, also known as John the Thrakesian, wrote the main 811-1057
section of the 'Synopsis' in the 1070s when he held the relatively lowly title of
protovestarches or protovestes. Shortly after this date his 811-1057 text was
39 Kedrenos, i, 5
40 Seibt, 'Zur Person', p.81.
41 Seibt's conclusion brought to an end a long-standing debate over the relationship between the two
texts. The arguments against seeing a single author were first put by de Boor, 'Weiteres zur
Chronik', pp.460-7. The case was restated by Moravcsik, Byzantinoturcica, pp.340-1. Moravcsik,
argued that since Kedrenos, who copied Skylitzes so slavishly, only included the 811-1057 section
of the 'Synopsis' in his text, the 'Continuation' must have been written by another author at a
different time. Moravcsik saw support for the idea of different and later authorship for the
'Continuation' in the fact that Michael Attaleiates, whose 'Historia' was the principal source of the
'Continuation', was not named as one of Skylitzes'sources in the preface to the original 'Synopsis'.
Tsolakes in Skylitzes Continuatus: 'H 2wc%eia rife %povoTpeufrias TOV 'latawou S/ft/Arrgq, pp.76-99,
countered Moravcsik's case by arguing that the list of historians in the preface to the 'Synopsis' is
not an exhaustive enumeration of Skylitzes' sources. He believed that the similarities in working
methods, vocabulary, and phraseology between the two texts pointed to a common author. Back
references within the 'Continuation' to events included in the main text of the 'Synopsis' and a
concentration on Balkan and Italian matters in both compositions also suggested a single author. See
also M. Hicks, 'The Life and Historical Writings of Michael Attaleiates' (Oxford Univ. M.Litt.
thesis, 1987), pp.60-7. However, as J.Shepard 'A Suspected Source of Scylitzes' Synopsis
Historiarum: the Great Catacalon Cecaumenus' BMGS 16 (1992), pp. 180-1 n.28 points out, it is
Zonaras' reference to the 'Thrakesian' which provides the most convincing evidence that a single
author wrote both the 'Synopsis' and the 'Continuation'.
36
incorporated into the world chronicle of Kedrenos. By 1090 Skylitzes was
proedros, megas droungarios of the vigla and eparch. Two years later he was a
participants at the Synod of Blachernai held in this year, the megas droungarios of
the vigla was Nicholas Mermentolos rather than Skylitzes. 43 Seibt concluded that it
was in his retirement in the later 1090s that Skylitzes added the 1057-79
'Continuation' to the 811-1057 narrative he had compiled some twenty years
earlier.
Since Seibt constructed his biography, some of his conclusions have been
confirmed, others questioned. For instance the identification of John Skylitzes with
John the Thrakesian is now certain. Attention has recently been drawn to the fact
that an anonymous commentary on the twelfth-century canon lawyer Balsamon,
explicitly names John Skylitzes as "the Thrakesian".44 However, in
contradistinction to Seibt's opinion, it now seems likely that the 'Synopsis' and
'Continuation' were not written at separate points in the 1070s and 1090s, but were,
instead, composed at approximately the same time. This likelihood is raised by the
fact that material from the same underlying source appears in both the 'Synopsis'
and the 'Continuation'. As we shall see shortly, Jonathan Shepard has suggested
that one of the most important sources behind the mid-eleventh century narrative of
Skylitzes' 'Synopsis' was an encomiastic biography of the general Katakalon
Kekaumenos 45 It is clear that this panegyrical biography not only informs the main
text of the 'Synopsis' but also the 'Continuation'. For example, in the
43 While he held the position of droungarios of the Vigla Mermentolos was a regular correspondent
of archbishop Theophylact of Ochrid (Theophylacte d'Achrida Lettres, ed. and trans. P.Gautier
(CFHB XVI/2, Thessalonika, 1986), letters 25, 29, 33, 47, 76; Mullett, Theophylact of Ochrid,
.101, 103, 118, 121, 183,271,275.
TOU TfjviKavTa, KoupomtAaTou KOA fAeyaXou ftpouvyafiou rfe B/VAaj xupov Twawou TOU SicuAW/ij TOU KO.I
iov: V.Tiftixoglu, 'Zur Genese der Kommentare des Balsamon', in N.Oikonomides (ed.),
Byzantium in the 12th Century (Athens, 1991), pp. 528-9. I am grateful to Paul Magdalino for this
reference.
37
'Continuation' Skylitzes mentions that Katakalon was raised to the rank of
kouropalates during the reign of Isaac Komnenos (1057-9). This detail must have
its origins in the pro-Kekaumenos source since it is an allusion absent from the
main source for the 'Continuation', the 'Historia' of Michael Attaleiates. 46 As far as
the composition of the 'Synopsis' and its 'Continuation' is concerned, the fact that
Skylitzes used the biography of Kekaumenos for the composition of both texts
raises the possibility that he wrote both narratives within a short time of each other.
Since the 'Continuation', which covers the period 1057 to 1079, must have been
written after 1079, this would mean that the main 811-1057 body of the 'Synopsis'
was also composed after 1079. In these circumstances, it makes most sense to see
Skylitzes working on both sections of his text during the first half of the reign of
Alexios Komnenos (1081-1118), possibly during the 1090s, either when he was
still megas droungarios of the vigla, or shortly after he had retired from public
office. Indeed, it is possible that the 'Synopsis' and the 'Continuation' were not
composed as separate texts at all, but instead form part of one continuous synoptic
history composed by Skylitzes in the last decade of the eleventh century.
person of the author. Apart from the prosopographical nuggets contained in the title
38
sequences, the text itself provides virtually no biographical data. The only hint of
personal detail occurs in the preface where the author indicates that he was a
wrought from other written testimonies. Little overt information about the career of
the compiler can be expected from such a literary production, and modern scholarly
attempts to amplify Skylitzes1 biography by isolating biographical clues from
within the text are not convincing. For example, it has been argued that a reference
in the 'Continuation' to the Serb leader Bodin, who died in 1101, as if he were still
alive, offers a terminus ante quern to Skylitzes1 historical writings. 50 However,
rather than demonstrating that Skylitzes was writing before 1101, this passage
merely indicates that it was the author of Skylitzes' source who was active before
this date. Comparable evidence from his coverage of the reign of Michael VI
indicates that Skylitzes' most likely contribution to such an apparently biographical
allusion was his verbatim copying of an underlying source rather than personal
reflection. At one point in his testimony he copies the biography of Katakalon
Kekaumenos so closely that he implies that Michael VI was still alive. In fact
Michael VI died before August 1059. Quite clearly Skylitzes was not at work
48 Skylitzes, p.4
49 Having entered imperial service during the reign of Michael IV (1034-41), Psellos was still active
in court circles in 1075. Although it is usually assumed that Psellos died during the later years of the
reign of Michael VH (1071-8), there is some evidence to suggest that he was still alive in the final
decade of the eleventh century (P.Gautier, 'Monodie inedite de Michel Psellos sur le basileus
Andronic Doucas', REB 24 (1966), pp. 153-164; AKazdan, 'An attempt at Hagio-Autobiography:
the Pseudo-Life of "Saint" Psellos', B 53 (1983), pp.546-56).
50 Antoljak, 'Johannes Skylitzes', p.679 following the example given by Tsolakes in Skylitzes
Continuatus: 'HSwexcia -rfc %povo7pcufria rou 'ItudwouSicuhMfl, pp.75-6.
39
before 1059; instead this reference indicates that it was his source, Katakalon
Kekaumenos, who was writing before the death of Michael VI. 51
51 Shepard, 'A Suspected Source', p. 178; idem., 'Isaac Comnenus' Coronation Day', Byz Slav 38
(1977),p.24
40
Investigations of Skylitzes' tenth-century testimony have been equally short,
In another short study by Frei, a different and more substantial relationship between
Skylitzes and Theodore Daphnopates has been suggested. On the basis of a
comparison between Skylitzes' text and a sermon by Daphnopates, Frei argued that
Theodore Daphnopates was responsible not for the sixth book of Theophanes
Continuatus, but instead for a lost history of the reigns of Constantine VII (945-59)
and Romanes II (959-63) which was used much later by Skylitzes. His argument is
based on certain similarities in narrative structure, vocabulary and word order
between Skylitzes' account of the arrival of the relic of the hand of John the Baptist
in Constantinople during his testimony of the reign of Constantine VII, and a
homily composed by Theodore Daphnopates to celebrate the first anniversary of
this event. It was Frei's belief that Daphnopates used his own speeches as primary
sources in the compilation of a history, and that it was from this intermediate
D.Polemis, 'Some Cases of Erroneous Identification in the Chronicle of Scylitzes', Byz Slav 26
(1975), pp.74-81
41
Daphnopates history that Skylitzes in turn compiled his own testimony for the
period 948-63. 54 However, the fact that Frei only dealt with one episode, covering
less than six lines of Greek in Skylitzes' 'Synopsis', in which only the most general
parallels of vocabulary and content with the sermon of Theodore Daphnopates are
visible, leaves his conclusions open to doubt.
42
Meanwhile, another possible source has been identified for the reign of John
Tzimiskes. It has been suggested that a war diary may have provided Skylitzes with
material for his extensive descriptions of this emperor's campaign against the Rus
in Bulgaria in 971. Skylitzes' testimony for this passage of warfare will be
considered towards the end of the second chapter of this thesis. 56
56 S.McGrath, 'The Battles of Dorostolon (971): Rhetoric and Reality', in T.S.Miller and J.Nesbitt
(eds.), Peace and War in Byzantium: Essays in Honor of George T.Dennis (Washington, D.C.
1995), pp. 152-64); see below, pp.94, 101, 104
57 Skylitzes, p.4
58 Skylitzes, p.313
59 Traite: 'Le traite des transferts', ed. J.Darrouzes, REB 42 (1984), p. 181. This entry contains
erroneous information. Agapios was in fact bishop of Aleppo when he was translated to Antioch
(Yahya, PO 23, pp.375-6). See below, pp.334-5. The 'Peri Metatheseon' contains another corrupt
entry which refers to Agapios. This entry states that he was translated to Jerusalem rather than to
Antioch (Traite, p. 181).
43
Yet while these two allusions suggest that Theodore of Sebasteia composed a
history about Basil II, they are too insubstantial to establish the nature of this text
and its exact relationship to Skylitzes' own production.
In an article written just before his death in 1997 Panagiotakes argued that two
miracle stories of Saint Eugenics of Trebizond compiled in the fourteenth century
by John Lazaropoulos, but set during the revolt of Bardas Phokas (987-9) and Basil
II's campaigns against the Iberians in 1021-2, contain material extracted from the
lost eleventh-century history of Theodore of Sebasteia.60 He noted that the miracle
stories were prefaced with narrative passages of political history, some of which
resembled passages in Skylitzes and the twelfth-century world chronicle of
Zonaras, but others of which had no parallel with the testimonies of these
historians. Thus, during the revolt of Phokas, the Eugenios Miracles mirror
Skylitzes' and Zonaras' account of the deployment of rebel troops on the Asian side
of the Bosphoros, but go on to include additional material about the emperor's
plans to resupply the capital by sea from Trebizond, and the rebels' decision to raid
the Pontus coast using an Iberian army.61 In the case of Basil's Iberian offensive of
1021-2, the Miracles allude to the emperor's decision to winter in Trebizond during
a break in his campaign, information absent from the accounts of Skylitzes and
Zonaras.62 The principal conclusion that Panagiotakes drew from his comparison of
the various narratives was that Lazaropoulos, Skylitzes and Zonaras all drew on the
62 Pangiotakes, 'Fragments', p.330. Once again, although Panagiotakes is correct to say that no
Greek source mentions Basil's residence in Trebizond, his presence in the city is recorded by a
variety of eastern narratives: Yahya (Cheikho), p.240; Matthew of Edessa, p.46; the Georgian Royal
44
same underlying historical account for at least part of their coverage of the reign of
Basil II. Yet, in identifying Theodore of Sebasteia as the author of this lost history,
Panagiotakes provided no other supporting evidence apart from the fragile allusions
to Theodore in Skylitzes' preface and the 'Peri Metatheseon' discussed above.
So flimsy is the evidence linking the lost history of Theodore of Sebasteia to the
account of Basil's reign contained in the 'Synopsis', that it is unlikely that any
greater understanding of Skylitzes' sources for Basil's reign will come from further
investigation into the putative relationship between these two texts. In the third
chapter of this thesis the question of Skylitzes' source materials for the reign of
Basil II will be explored further, with discussion focusing on the likelihood that an
encomium of the general Bardas Skleros underpins Skylitzes' coverage of the
period 976-989. However, at this stage it is simply important to note the lack of
sustained research into the source materials and working practices behind Skylitzes'
testimony as a whole, and his narrative of Basil's reign in particular.
Indeed, it should be pointed out that the only substantial analysis of Skylitzes'
working methods and treatment of source materials is that conducted by Jonathan
Shepard in the course of a series of investigations into the 1028-1057 section of the
'Synopsis'. The context for Shepard's research was a series of rigorous
examinations of Byzantine diplomatic relations with neighbouring powers during
the mid-eleventh century, during which he compared Skylitzes' testimony, often the
only account of the relevant events in Greek, with narratives composed in other
languages. Partly through these inter-textual comparisons, and partly by looking at
the internal structures of the 'Synopsis' itself, he then identified several key
diagnostic elements integral to understanding how Skylitzes' text was composed.
Annals, p.283. The Armenian historian Aristakes Lastivert does not refer to Trebizond by name, but
45
His research is of particular interest to the historian of Basil's reign since it
illustrates how Skylitzes' text can be approached when none of the underlying
sources survive. Since Shepard's analysis has to be collated from several different
articles, and since so many of the diagnostics he isolated are germane to the first
three chapters of this thesis, the main points of his analysis and their implications
for an understanding of the material in Basil's reign are summarised briefly here. 63
says that Basil wintered in Chaldia, the theme of which Trebizond was the capital (Aristakes, p. 15).
63 The Shepard articles in question are: 'John Mauropous, Leo Tornices and an Alleged Russian
Army: the Chronology of the Pecheneg Crisis of 1048-9', JOB 24 (1975), pp.61-79; 'Byzantium's
Last Sicilian Expedition: Skylitzes1 Testimony', Rivista di Studi Bizantini e Neoellenici 14-16 (1977-
9), pp. 145-159; 'Isaac Comnenus' Coronation Day', Byz Slav 38 (1977), pp.22-30; 'Scylitzes on
Armenia in the 1040s and the Role of Catacalon Cecaumenus', REArm (1975-6), pp.296-311;
'Byzantinorussica', REB 33 (1975), pp.211-225; 'A Suspected Source of Scylitzes' Synopsis
Historiarum: the Great Catacalon Cecaumenus', BMGS 16 (1992), pp.171-81.
64 Shepard, 'Scylitzes on Armenia', pp.292-5
65 Skylitzes, pp.339-40
46
the 1048 campaign against the Turks, the 1048-9 battles against the Pechenegs, and
Isaac Komnenos' coup of 1057.66 Similar examples are also visible in Skylitzes'
treatment of the later tenth century. Nearly one third of Skylitzes' coverage of John
In the process of tracing the internal structures of Skylitzes' text, Shepard also
began to identify different genres of source material that underpinned his historical
testimony. On the one hand he argued that some of the most vivid passages of
narrative action were excerpted from contemporary panegyrics produced by senior
commanders within the Byzantine army. One of these panegyrical accounts was an
apologetic pamphlet produced by associates of George Maniakes at the time of that
general's revolt in 1043. More important, however, to the overall composition of
Skylitzes' eleventh-century narrative was the encomiastic biography of the general
Katakalon Kekaumenos. According to Shepard this forms the core of Skylitzes'
coverage of the period between 1042 and 1057. Several of the distinguishing
features of the encomia used by Skylitzes will be discussed in greater length in the
third chapter, where it is argued that an apologetic text produced by the general
Bardas Skleros underpins the coverage in the 'Synopsis' of the reign of Basil II.
Skylitzes used other sources. The proliferation of annus mundi and indiction dates
66 Shepard, 'Byzantium's Last Sicilian expedition', pp. 155-8; 'Scylitzes on Armenia', pp.270-79, 'A
Suspected Source', pp. 172-6.
67 Skylitzes, pp.298-309
47
in the material between 1029 to 1043 suggested to Shepard that Skylitzes had
access to a set of annals that consisted of short entries about politics, diplomacy and
natural disasters. Shepard argued that is possible that these annals were the source
for some of the brief notices found at the end of Skylitzes' coverage of Basil II's
reign, such as the description of Orestes' expedition to Sicily in 1025.69
senior generals that contained few, was that the overall chronology of the narrative
could easily become disturbed. For example, during his coverage of Sicilian matters
in the 1030s Skylitzes attempted to integrate an undated section of the Maniakes'
encomium into an annalistic entry. As a result he implied that Maniakes was
appointed to lead an expedition to Siciliy in 1034-5. In fact, Maniakes only took up
this position in 1037-8. 70 In view of such confusions in the 'Synopsis', Shepard
concluded that where Skylitzes relies on a single source, his chronological and
factual details are likely to be at their most trustworthy. But if the text has been
synthesised from a mixture of materials inaccuracies may occur.
48
III. Textual analysis: Skylitzes and Theophanes Continuatus
/. Method
'Synopsis' is the only extant narrative source in Greek, such as the reign of Basil,
49
source, the sixth book of the 'Continuation' of Theophanes. In its scope and detail
this comparative study is, as far as I am aware, innovative.71
In the very first sentence of his preface Skylitzes presents his preferred model for
the writing of synoptic history: the "epitome of history" compiled by George the
Synkellos and Theophanes at the beginning of the ninth century. However, having
identified his historiographical ideal, Skylitzes then goes on to explain why more
recent historians have fallen short in their attempts to continue the work of the
Synkellos and Theophanes. Some, such as Michael Psellos (described here by
Skylitzes as the "hypertimos Psellos") and the didaskalos Sikeliotes, have failed to
deal with history in sufficient detail:
But having undertaken the task in a desultory way, they both lack accuracy, for
they disregard very many of the more important events, and they are of no use
to their successors, since they have merely made an enumeration of the
emperors and indicated who took imperial office after whom, and nothing
more.73
71 A very limited comparison between Skylitzes Continuatus and its principal source the 'Historia'
of Michael Attaleiates has been conducted by A.Kazdan ('The Social Views of Michael Attaleiates',
in A.Kazdan and S.Franklin (eds.), Studies on Byzantine Literature of the Eleventh and Twelfth
Centuries (Paris, 1984), pp.23-86). However, this investigation is predominantly concerned with the
text of Attaleiates. There is little direct concentration on either the content, vocabulary, level of
language or style of the 'Continuation' produced by Skylitzes, nor of the wider literary, intellectual
or political contexts in which the author was writing.
72 The preface to the 'Synopsis' is to be found in Skylitzes, pp.3-4. A full translation is included in
the appendix to this thesis.
oAAo, naptpyu*; cu/ift v0f TOU epyou T% TC aiepi^eia^ aTvonemwKauri, ra Tr/Veierra TWV KOipiarrepiov
*w avmnrroi roig /xer'a-uTouf yeybvauow, wnapiQivrpw itovrp mrr^nievoi r&v /SacnAcwv xai
rig /w*ra TIVO, TW tncrjnrpwiv jeyavev erKpa-nfc, KCW nkefov ou&v (Skylitzes, p.3). It is usually
argued that the historical writings of Michael Psellos to which Skylitzes applies this criticism are not
50
In other cases historians are criticised for bias and short sightedness:
....each [historian] had his own agenda, the one proclaiming praise of the
emperor, the other a psogos of the patriarch, another the encomium of a
friend.......... They wrote histories at length of the things which happened
during their times and shortly before: one sympathetically, another with
hostility, another in search of approval, another as he had been ordered. Each
composing their own history, and differing from one another in their narrations,
they have filled the listeners with dizziness and confusion.74
Some of these offenders are listed. They include those whose texts are still extant
today, such as Joseph Genesios, as well as other historians whose compositions are
now lost, such as Theodore of Sebasteia and Theodore Daphnopates.75
Psellos' long appraisal of fourteen emperors known as the 'Chronographia', but instead, the much
shorter 'Historia Syntomos'. This second text is a list of Roman and Byzantine emperors extending
from Romulus to Basil II. Attached to each emperor's entry is a very brief account of the principal
events of his reign. Modern scholarly opinion is, however, divided on the issue of whether Psellos
was responsible for the 'Historia Syntomos'. Snipes and Ljubarskij believe that he was, and it is to
this text that Skylitzes refers in his preface. On the other hand Aerts, the editor of the critical edition,
believes that the text was written by another eleventh-century author, possibly John Italos (K. Snipes,
'A newly discovered history of the Roman Emperors by Michael Psellos', JOB 32.2 (1982), p.55;
J.Ljubarskij, 'Some Notes on the Newly Discovered Historical Work by Psellos', pp.213-228;
W.J.Aerts (Pseudo-Psellos: Michaeli Pselli Historia Syntomos, ed. and trans. W.JAerts (CFHB
XXX, Berlin, 1990), pp.i-xxv). Turning from Psellos to Sikeliotes, it should be noted that there is no
extant Byzantine historical work by a didaskalos of this name. However, it is possible that Skylitzes
is referring here to lost historical writings by John Sikeliotes, a rhetorician active at the end of the
tenth century. Sikeliotes once made a speech (which is no longer extant) in the presence of Basil II
at the Pikridion monastery near Constantinople (Hunger, Hochsprachliche profane Literatur, i, 45-
6; ii, 476-7; ODB, ii, 1068)
... oiKeiav e'jcaoTos urrofletnv Trpotmjotiftcvoi, 6 /xev enatvov fcpe enrefv jSeuriXeiog, 6 tie \l^yov "rraTpiapxpv,
arepo$ #e <^/Aoy eyKWfjuov........... aTmra^rjv yap ra Kara TOUJ aurwv %povou$ aweveyfievra,
ava)6ev, ioTopiKto^ <TU77pat/tyt vo, jcai o /ttev oT;/M,7ra0wj , o ffavrnraOa^, o de Kai Kara %,aptv, aAAog #6
fe> irpoa-ereraKTo, Trp eaurov owfaTg 'urropiav Kai -npog aAAijAouj ev TTJ TW aurotv a&i\frprt\
/Ai^-you KOI rapaxfa roug aKpoara^ epTrerrMiKao-iv (Skylitzes, p.4)
75 See above, pp.41-4
51
histories, he then removed "that which was written in a state of emotion or in the
search of approval", "disregarded differences and inconsistencies", "shaved off
whatever we have found which is too close to legend", and ignored rhetoric. In his
own opinion his final product is "a nourishment which is soft and finely ground in
language". This literary fare he believes will be to the taste of his audience whom
he divides into four categories: "those who love history"; those "who prefer that
which is very easy to that which is more wearisome"; those "who are acquainted
with histories"; and finally, those "who are not yet acquainted with histories".
76 Coverage of the reign of Romanes Lekapenos is to be found in Skylitzes, (Thurn edition), pp.213-
232; and Theophanes Continuatus, ed. LBekker (CSHB, Bonn, 1838), pp.398-435. Several small
textual points confirm that Skylitzes' main source is the 'Continuation' of Theophanes rather than
one of the many versions of the Logothete such as George the Monk Continuatus. First, Skylitzes
includes a compressed version of a eulogy of the Kourkouas family and a notice about marital
relations between the Lekapenoi and the Argyroi which are only found in Theophanes Continuatus
(Skylitzes, pp.229-30; Theophanes Con., pp.426-9; Skylitzes, p.213; Theophanes Con., p.399).
Second, Skylitzes mentions that Peter the emperor of Bulgaria suffered a revolt by his brother
Michael: this too is discussed only by the 'Continuation' of Theophanes (Skylitzes, p.226;
Theophanes Con., p.420). Finally, in the case of second marriage of Constantine, third son of
Romanos Lekapenos, the wedding is recorded in all three texts, but Skylitzes follows Theophanes
Continuatus in identifying the first name of the bride, Theophanu, and her family name, Mamantos,
whereas the Logothete fails to record the bride's identity (Skylitzes, p.229; Theophanes Con., p.423;
George the Monk Con., p.914).
77 Skylitzes, pp.226-7
52
verbatim. A clear example of close verbal parallels occurs in his account of a
Bulgarian victory near the palace of Pegai at the beginning of Romanes1 reign. The
appropriate passage from Skylitzes' account is cited here, with those phrases taken
directly from Theophanes Continuatus underlined:
78 Skylitzes, p.215: Theophanes Con., p.401. My translation of Skylitzes' text: On the one hand the
rector John fled, whereas the patrikios Photeinos the son of Platypodos, who was fighting for him,
was killed as were several others. And so the rector having barely escaped, boarded the dromon
(warship). And although Alexios, the admiral (droungarios) wanted to do the same thing, he was not
able to climb up on the deck of the dromon; he fell into the sea and was drowned together with his
protomandator. Verbatim copying can also be observed in Skylitzes Continuatus1 coverage of the
second half of the eleventh century. As Shepard has pointed out Skylitzes often copies his
underlying text so closely that he retains the first person singular voice of the root source (Shepard,
'Byzantinorussica', p.217)
79 Skylitzes, p.213; Theophanes Con., p.399; see also Polemis, 'Some Cases of Erroneous
Identification', p.77
53
miscopied Tw^AwfleVre?. Another single word error involves an accident in the
forum, which according to Theophanes Continuatus resulted in the deaths of six
o i
indiction; the original source records February the second of the sixth indiction.
54
Kalomaria.85 Equally Skylitzes usually transmits dates accurately. The error
concerning the patriarch Theophylact is the only glaring mistake in the nineteen
dates included in the 'Synopsis'. Furthermore, Skylitzes very rarely omits a date
the date of the second sea battle involving the Rus in 941 (September 10th). More
frequent than his complete failure to register a date are his omissions of the precise
day when an event occurred. For example, although he records the month and
indiction date when the Union of the Church was confirmed in the early part of
Taken as a whole such plentiful evidence for Skylitzes' faithful copying of content
and vocabulary lends weight to the contention put forward by Thum, the editor of
the critical edition of the 'Synopsis', that Skylitzes is little more than a transcriber,
who adheres so faithfully to his source material that it is impossible to attribute any
OQ
55
that Skylitzes was merely a passive copyist and abbreviator whose testimony can be
accepted as an accurate transmission of the materials he collates. Instead, further
inspection of Skylitzes' treatment of Theophanes Continuatus' coverage of the
reign of Romanos Lekapenos, reveals a number of subtle adaptations which when
aggregated demonstrate that the compiler of the 'Synopsis' exercised an active
authorial role. On some occasions Skylitzes' interventions are compatible with the
intentions he outlines in his preface. On others his manipulations appear to deviate
from his own statement of purpose. More important for the historian of medieval
Byzantium Skylitzes' active authorship can impose serious distortions on the
contents and interpretations of the underlying materials he transmits. 00
At the most basic of levels Skylitzes takes measures to ensure that his history is a
synopsis rather than a simple copy. Thus, he sometimes elides two main verbal
clauses from the underlying text into a single clause containing a main verb and a
participle construction; the latter may take the form of a genitive absolute.91 His
enthusiasm for abbreviation is most visible whenever he tries to combine so many
phrases and sentences from the underlying source into a single unit that the
meaning of his narrative becomes elusive. For example, on several occasions he
combines several main verbs from the underlying text into a more elaborate single-
verb sentence, with the result that a large slice of prose is expressed in a case other
89 Zonaras (John): loannis Zonarae Annales, ed. M.Pinder (CSHB, Bonn, 1841), i, 8-9; Hunger,
Hochsprachlicheprofane Literatur, i, 417.
90 Useful comparisons can be drawn with the ninth-century synoptic historian Theophanes the
Confessor. Although Theophanes usually follows the word order and phraseology of the texts which
underpin his narrative very faithfully, he can make interventions of a very subtle order, sometimes
involving no more than the insertion of single word, the omission of a phrase, or the repositioning of
a date. Some of these alterations are accidental, however others are deliberate, undertaken with the
purpose of altering the sense of the text (Mango and Scott, The Chronicle of Theophanes, pp.xcii-v).
For example Skylitzes uses two genitive absolutes to describe the defeat of the Arabs and the
flight of Leo of Tripoli at the hands of the admiral (drotmgarios) John Radenos in the early 920s:...
o 'Po#evo KaT9vo/Aoo/A v0 cufaifaov ercHJxiveig ppd'icos erpei^aro, T&V ' Ayap^ivatv <rxf$bv navrtov
avwmuvtavt rov & TptrroAfTou yubvov <frurn f^v (rarrnpiav nopunwevov (genitive absolutes underlined;
Skylitzes, p.218); whereas Theophanes Continuatus uses two main verbs (underlined) to describe
the actions of the Arabs and Leo: ...... b 'Padivb$ K(novona6i*evo<; aujtvidittx; erreflero aimjr KCU
56
than the nominative. Thus, when Skylitzes decides to make the emperor the subject
of a long sentence about the dismissal in 944 of John Kourkouas, the domestikos of
the scholai, the subsequent description of the career and exploits of the general has
during the reign of Romanes II (959-63). Once again, because emperor Romanos
occupies the nominative position, Nikephoros Phokas and his many military
exploits against the Arabs have to be expressed in a very long accusative phrase. 93
...... 6eov ouvepy'up, 01 I/TT'CH/TOV rpenovrat 'Ayapwoi, f^oAtg #6 /jtovog 6 TptTroA/ViK ___
(Theophanes Con., p.405)
92 Skylitzes, p.230: <i>66vov # Kivydevros Kara rov o^eoriKOv ratv c^oAftiv 'Itoawou rov KoupKova -rrapa
r&v d}Jtfiiv patnXetov (efiouXero 'yap 'P(t)na.vo$ o BatriAeuj Ew^poowijv Tt)v rov dojieariKou Bvyarepa
TQ) oiKeitf) eKyovtf) 'Puifjutv^ (ra) uit^ rov eoyc/arov TTCUOO$ aurov Ktworavrivov)
qj auTov TTapaXv(rai, erri ftv(ri KOU eiKOO~i %p6voi KCLI ynyri'v enra, 001006^.10^ Ttjy TQU
__________}6vva>vra,, KQA TTCUTCLV, fa>g erneiv, nrjv Supmv /cttTttopafAovra KO,I ra/rreivtiXTtLvra,'. the first
two underlinings highlight Kourkouas in the genitive case; the last underlining indicates the long
phrase in the accusative.
Skylitzes, p.249: rovrq) TQJ erei Nuoxfropov uta/Yi<rrpov rov OtuKav,
^
TK ^ ^
avaToAtjs napa^ T7" '
KqjvcrravTtvoi; TOV** peunAeajg,
ft ^ '
<cai* "noM-a
^^^ '
-rponaia ___'
errngavra KOTO,* TMV
^
, KO,I TOV re -rife Tapcnou an'npa.v K.a0a4Ut)vw> KOJ Xa^8^&v rov TOU XaAcrr icai rov
57
("let him seek"). In contrast, Theophanes Continuatus uses a simple third
person plural verb to inform "those readers who wish to learn more" of Kourkouas'
exploits that they will find the information they are seeking in the books of Manuel:
'lC*>i0 oXotrxepax raireiviixravTa, TTepnei Kara raw ev Kpiyrjj Sopcuctyv&v, Tdsr)6o$ emAeicTwv
erridous ai/r<J> KOI o-roAov Ka-rjpriovevov KaXwj: Nikephoros in the accusative case underlined.
94 Skylitzes, p.230, Theophanes Con., pp.427-8.
95 Skylitzes, p.214 and p.216; Theophanes Con., p.400 and p.402.
96 Skylitzes, p.214; Theophanes Con., p.399; see also Skylitzes, p.270 for a similar example from the
reign of Nikephoros Phokas
97 Skylitzes, p.213; Theophanes Con., p.398
98 Skylitzes,, pp.259,
., 286
99 In a review article of Thurn's
critical edition of loannis Skylitzae Synopsis Historiarum, Cyril
Mango reflects on the inconsistencies of Skylitzes1 Attic prose (JHS 95 (1975), pp.304-5)
58
sometimes employs two main verbs where Theophanes uses a genitive absolute. On
at least one occasion he replaces an optative with a simple main verb. 100
i f\n _
During the story of the revolt of Bardas Boilas, thestrategos of Chaldia, Theophanes Continuatus
informs the reader that the emperor forgave the poorer rebels and allowed them to go wherever they
wanted: OOTJ jSouAoivro whereas Skylitzes renders this phrase as OTTO fZoufovrai (Theophanes Con.,
p.404; Skylitzes p.217)
1 Theophanes Con., pp.423-4; Skylitzes, p.229
102 Theophanes Con., p.402; Skylitzes, p.215
59
of Theophanes himself. Much more conspicuous are Skylitzes' attempts to remove
the excesses of encomium from the underlying text. At one level this operation
John Kourkouas. For instance while Theophanes alleges that Kourkouas, "....
became unrivalled in matters of war, and established many great trophies, and
extended the Roman boundaries and sacked very many Agarene cities", and makes
reference to his "outstanding virtue", Skylitzes rather more drily comments that he
"..... overran and humbled, so it is said, the whole of Syria". 105
60
the emperor Peter, or of their grief over the death of the eldest son Christopher in
931. 107
However, it is in the section dealing with Romanos' meeting with the Bulgarian
leader Symeon in 924 that Skylitzes' reconditioning of the encomium of the
emperor is at its most conspicuous. First, Skylitzes omits those elements in
Theophanes Continuatus' account which contribute to an aura of sanctity. Although
he mentions that Romanos entered the church of Blachernae before he met Symeon
in order to pray and to put on the protective omophorion of the Virgin, he excises
all references to Romanos weeping and imploring the Mother of God for her
assistance. 110 Later in the account of the meeting of the two leaders Skylitzes
107 Reference to John of Bulgaria. Theophanes Con., p.419, Skylitzes, p.225; reference to Mary
Lekapene's departure to Bulgaria: Theophanes Con., p.415, Skylitzes, p.224; reference to
Christopher's death: Theophanes Con., p.420, Skylitzes, p.226
ineopnanes Con., pp.415-7; Skylitzes, p.225
108 Theophanes
109 TViQUo.o /-.., ~-> A^C\ -5/v oi~.i:* . o-> i
109 Theophanes Con., pp.429-30; Skylitzes, p.231
110 Skylitzes, p.219; Theophanes Con., pp.406-7
61
excludes all allusions to the bravery of Romanes. 111 It is in his concluding editorial
comments about Romanes' encounter with Symeon, that Skylitzes' deviation from
Theophanes Continuatus is at its most obvious. For where the original text
maintains that Symeon went back to his camp praising, "....the intelligence and
humility and ..... appearance of bodily strength and ... dauntless spirit" [of the
emperor], Skylitzes alleges that the Bulgarian leader returned to his associates and
commented on the, ".... moderation of the emperor and his lavishness and
generosity in matters of money".' 12 One explanation for Skylitzes' reshaping of
Lekapenos' role could be that the compiler had access to anti-Lekapenos material,
which he used to counter the rhetorical hyperbole of Theophanes Continuatus.
However, given the extremely tight congruence between the narrative structures of
the 'Synopsis' and the account of Theophanes Continuatus, it seems superfluous to
suggest that Skylitzes draws on an alternative primary source. Instead his treatment
of the Lekapenoi family is almost certainly the result of his own willingness to
comment upon, reorder and reshape his core source using his own powers of
interpretation.
Nonetheless, it is should be noted that Skylitzes does not completely jettison the
laudatory excesses of his underlying sources. The encomium of the bridegroom of
Agatha Lekapene, a member of the Argyros family, is reduced in length in
Skylitzes' version, but the essence of the praise of the protagonist's physical and
intellectual merits is retained. 113 Equally, although Skylitzes brutally curtails the list
of the military activities of John Kourkouas, his brother Theophilos and his son
Romanes, he retains all the information relating to the intra-familial links between
62
the three Kourkouas commanders. Moreover, he even updates the text so that
whereas Theophanes imparts the information that Theophilos was the grandfather
of John Tzimiskes, "who became domestic of the schools under the emperor
Nikephoros", Skylitzes tells us that he was the grandfather of John Tzimiskes, ".....
who was emperor after these things". 114 Furthermore, even if he dismisses or dilutes
the more obvious passages of panegyric, Skylitzes cannot entirely escape the
viewpoint of the original subject of the encomium. Thus, although he reshapes
Theophanes Continuatus' praise of Romanos Lekapenos and John Kourkouas,
Skylitzes still has to accept the underlying source's identification of these two
actors as the most important protagonists in the history of this period.
63
Other link phrases have the additional purpose of trying to make sense of the
structure of the source Skylitzes is processing. For example, at the end of the
Skylitzes inserts the simple phrase: "and matters in the City came to an end in this
way". 116 This phrase is intended as a pointer to the next episode in the text, namely
Byzantine relations with Melitene, which certainly were not matters pertaining to
the City, but instead took the reader out to the eastern frontier. Another one-line
explanatory interpolation occurs in Skylitzes' account of the conspiracy of Bardas
Boilas, the strategos of Chaldia. When Theophanes Continuatus reports on this
revolt he offers no explanation as to why this episode is included in his text, but
simply begins with the allegation of a plot. Skylitzes, in contrast, wishes to make
explicit the fact that this revolt is mentioned at this point in the narrative because it
represents yet another rebellion of the sort that peppered the early years of
Romanes' reign. Therefore, at the beginning of the episode he adds the explanatory
phrase: "another revolt happened against the emperor in Chaldia". 117
Skylitzes, p.229; Theophanes Con., p.425. For a discussion of link phrases in synoptic history as
signs of active editing see Jeffreys, Studies in John Malalas, p.21. On Skylitzes' fondness for such
devices see Shepard 'Byzantium's Last Sicilian Expedition', pp. 147-8
64
explanation has no apparent background justification in the text of Theophanes. In
contrast, Saktikios' early success against the Bulgar camp is explained by the
absence of most of the Bulgars, who were away raiding the surrounding
countryside for supplies. While Theophanes Continuatus does not actually say that
this was the case, Skylitzes' hypothesis is at least partially supported by the
allegation of the underlying source that once the Bulgars heard of the assault of
Saktikios they all returned to the camp. 118 One might suggest that Skylitzes1
treatment of Symeon's post-conference report in 924 on the qualities of Romanes is
an explanatory addition in the same vein. For having alleged that Symeon was
enormously impressed with the personal qualities of the emperor, Theophanes
Continuatus goes on to highlight the largesse which Romanes displayed on
Symeon's departure: "and so having embraced one another they parted, with the
emperor having bestowed magnificent presents on Symeon". In these circumstances
Skylitzes1 decision to make Symeon stress the largesse rather than the virtue of
Romanes seems perfectly justified. 119
Yet, the reader of Skylitzes' should note that even the simplest tightening of the
structure of the text to comply with thematic rigour may easily eliminate the deeper
nuances of the core source. For example, we have already seen that in his treatment
of the revolts, which plagued the early years of Lekapenos' reign, Skylitzes not only
retains and accurately records the names of most of the conspirators listed by
Theophanes Continuatus, but also alerts the reader's attention to the thematic
integrity of these early passages by interpolating explanatory phrases. Yet, he also
chooses to control the overwhelmingly large dramatis personae of his core source
by selectively omitting certain minor personalities. Unfortunately this has the effect
65
of destroying fragile clues supplied in the underlying text about the workings of
Byzantine high court politics in the early tenth century. For instance, in the case of
the conspiracy of Arsenics and Paul the Manglabites, Skylitzes retains the names of
the plotters but excises a minor character called Leo, the anthropos of Arsenics,
who acted as an informer to the imperial authorities. 120 Not only does this omission
mean that the reader of the 'Synopsis' is furnished with less information about the
chiaroscuro of rumour, coup and counter-coup in the embryonic period of a key
reign, it also deprives the narrative of vital information about the role of the elusive
John the Rector and Mystikos. Theophanes Continuatus tells his reader that it was
John who had originally recommended Leo to the emperor and secured his
appointment in imperial service. In this scant information provided by Theophanes
Continuatus, John the Rector emerges as a key political broker at court.
Unfortunately Skylitzes1 omission of a minor character such as Leo means that
much less can be deduced about major figures such as John. 121
One of the reasons why Skylitzes omits minor characters like Leo is because his
prime concern is to focus the text more narrowly on the more prominent personnel
of the narrative. This desire may also determine Skylitzes1 enthusiasm for
attributing additional personal details such as names, titles and offices, to the most
important figures within the history, even where they are missing in the core
source. Just as Polemis noted that Skylitzes was willing to insert patronymic details
into his narrative with no support from his underlying texts in the course of his
ninth- and early tenth-century coverage, evidence of a similar nature appears in his
66
treatment of the reign of Romanes Lekapenos. 122 Alexander Kazdan noticed that
Skylitzes is the first historian to record the family name Lekapenos in connection
fondness for embellishing the personal details of the main characters in his
narrative include awarding individuals titles, which cannot be corroborated from
the root text of Theophanes Continuatus, nor indeed from other tenth-century
sources such as the Logothete. Thus, the Arsenios mentioned above is for no
apparent reason given the title ofpatrikios by Skylitzes. The same title is awarded
to Bardas Boilas, the rebellious general (strategos} of Chaldia. During Byzantine
military actions against Melitene, Melias, the leader of the Armenian troops, is
I A A
given the additional label magistros. The most likely explanation for Skylitzes'
tendency to award titles out of thin air, is that he may have tried to grant officials
the rank he believed they deserved on the basis of comparative evidence from
elsewhere in the underlying text. Thus, Bardas Boilas is probably given the title
patrikios because other strategoi during the reign of Romanes were described as
having this title by Theophanes Continuatus: for example, Bardas Phokas is
described by Theophanes during the invasion of the Rus in 941 as a former
strategos with the title of patrikios. 125
Although Skylitzes1 decision to award Boilas the title at patrikios appears to reflect
his sensitive understanding of the administrative history of the empire in the early-
to-mid tenth century, elsewhere his presentation of the administrative structure of
empire imposes serious distortions. Thus, where Theophanes Continuatus refers to
67
a certain Michael son of Myroleo as a topoteretes, a senior officer within the
professional and centralised tagmatic army forces of the early tenth century,
Skylitzes uses the rather generalised term tagmatarchon 126 One suspects that
Skylitzes' generalised phrase may either reflect his own ignorance of military and
administrative structures in the early tenth century, or may be an attempt to make
124 Skylitzes, pp.213, 217, 224; Theophanes Con., pp.399, 404, 416
125 Theophanes Con., p.424
126 Theophanes Con., p.400; Skylitzes, p.214
68
Continuatus1 account of Byzantine action against the eastern emirate of Melitene in
the later 920s and early 930s, that he provides no indication of the annual
campaigns that were waged by imperial armies against the Arabs, nor of the events
of the final siege which eventually forced the city to capitulate in 934. Whereas
Theophanes Continuatus describes the Byzantines burning the countryside of the
emirate, their use of siege equipment, and the general John Kourkouas1 impatience
at his initial failure to take the city, Skylitzes summarises the twists and turns of the
drama in a single bland phrase: "having confined those inside by siege he
_ i oft
[Kourkouas] compelled them to look for agreements".
69
suggested that the entertainment of a later-eleventh century audience is more
important to Skylitzes than a sense of strategy. 129
Equally frustrating for the modem historian who wishes to extract reliable military
material from the 'Synopsis' is Skylitzes' tendency to compress the underlying
narrative by applying homogenising cliches. These have the effect of suppressing
the uniqueness of the events in question, erasing specific detail, and transforming
each military engagement into a string of impenetrable stereotypes. Thus, in
Skylitzes' text the joining of two sides in battle is frequently represented by the
phrase m^T^o/^f revoftevy?. 131 One of the protagonists, particularly in a hand to
hand engagement, is always likely to be mortally wounded (TrA^^y) Kaipiav #e
1 T^
A protagonist will conduct a siege with or without care eT
70
eTroAiop/cei. 133 The recipient/s of a siege always resist with spirit u4/v
, until the protagonist presses them too hard
When they surrender it is usually because they are in need 77? evfaiqb of essential
supplies. 136 Camps are always established crrparvnedov Tnjfe. 137 Those encamped
will often scour the surrounding countryside for booty or spoils e-ni ^tapna^v
owAtov. 138 The term evedpa is preferred when denoting an ambush. 139 Those who
triumph in battle always do so easily p^iw?. 140 Equally their triumph is often
achieved with unstoppable strength pupy avimoaTar^41 Those who chose to rebel
often "hole up" at a well fortified castle: cfrpovpiov e
71
triumphed easily". It should be noted that this phrase contains two of the
commonplace generalisations we identified in the paragraph above: a
tf\ \ t> _> \ t %/ > f 143
enenuero ovv puii/n dvimocrrarQ) KCLI pcto/cog erpenero.
One final example from the reign of Romanes Lekapenos will show how Skylitzes'
substitution of a standardised cliche in the service of narrative compression leaves
the reader entirely uninformed about the underlying details represented by that
standardised phrase. In his encomium of the Kourkouas family Theophanes
Continuatus devotes considerable praise to the achievements of John's brother
Theophilos. Although Theophanes Continuatus' text is bestrewn with rhetorical
hyperbole, including a comparison between Theophilos and the Biblical King
Solomon, it also conveys the key information that Theophilos was the strategos of
Chaldia and Mesopotamia, and that during his tenure of the former position he was
involved in the capture of Theodosioupolis (modern day Erzerum) on the Caucasian
frontier. 144 Thus, although Theophanes Continuatus overdoses with rhetoric, his
text identifies Theophilos1 geographical sphere of military operations with some
accuracy. In contrast, although Skylitzes removes the hyperbolic allusion to
Solomon, he also completely excises all the substantive detail of the general's
career by articulating Theophilos' achievements in a standardised and anodyne
cliche. He sums up Theophilos' achievements as strategos in Mesopotamia with the
phrase: raTreivayaug KOA reXeax; a^avlaug TOV<; eic TT^ "Ayap ("having humbled and
finally destroyed the sons of Hagar") 145
Skylitzes, p.218; Theophanes Con., pp.404-5. It should be noted that Skylitzes idiosyncratically
uses the middle voice of the verb trepo in an active sense on many occasions.
144 Theophanes Con., p.428; see below, pp.222, 229, 290, for the fall of Theodosioupolis to
Byzantine armies in 949
145 Skylitzes, p.230
72
IV. Conclusions: Skylitzes and the reign of Basil II
This chapter has taken the preliminary steps towards understanding how those parts
of the 'Synopsis Historion', where Skylitzes' underlying source materials fail to
survive, should be approached. The method developed has been to examine another
section of the 'Synopsis', where the underlying text does survive, in order to
uncover the most characteristic features of Skylitzes' treatment of source materials.
Obviously, the next stage in the process is the application of these general
conclusions to sections of the text, such as the reign of Basil II, for which none of
the underlying sources are extant. This is an enterprise which will be undertaken at
greater length in subsequent chapters of this thesis. However, in concluding the
textual analysis in this chapter, I would like to summarise the most important
problems that Skylitzes' testimony presents to the historian of medieval Byzantium
in general, and of Basil's reign in particular.
73
by Basil II. Yet, as Mark Whittow has recently pointed out, the rest of Skylitzes'
account of this campaign does not suggest that the Bulgarians suffered an
annihilating defeat in 1014. 146 Moreover, the continuation of the war between
Bulgaria and the Byzantine empire for another four years indicates that Basil's
Balkan adversaries retained considerable fighting capacity. One way of explaining
the anomaly between the scale of the Bulgarian defeat at Kleidion reported in the
'Synopsis' and the subsequent recovery of Bulgarian fortunes, might be to suggest
that Skylitzes simply miscopied, or else deliberately exaggerated, the figure of
Bulgarian casualties.
In addition to his frailties as a copyist, the textual analysis developed in this chapter
suggests that Skylitzes' more active authorial interventions, such as his tendency to
compress, omit, expand, explain, and add homogenising glosses to his underlying
source material, can cause substantial obfuscation. Three particular problems
arising from such interventions are of especial relevance to the historian of the
political and military history of medieval Byzantium. The first concerns
prosopography, or the "Who's Who", of the Byzantine empire. As the detailed
textual analysis of Skylitzes' treatment of Theophanes Continuatus has
demonstrated on several occasions, our compiler is frequently willing to draw
genealogical connections between individuals and to bestow offices and titles on
the basis of little supporting evidence from his underlying sources. This cavalier
attitude to evidence must warn the prosopographef against using Skylitzes too
slavishly in the reconstruction of the careers of the most politically significant
individuals and families within the Byzantine empire. The second problem relates
to the administrative history of the empire. Here, Skylitzes' ubiquitous tendency to
146 Whittow, Making of Orthodox Byzantium, pp.387-8. He points out that while Basil was
victorious at Kleidion, another contemporaneous Byzantine attack led by the general Theophylact
Botaneiates, the doux of Thessalonika, was heavily defeated by Samuel's forces. Moreover, after
74
replace precise administrative terms with homogenised generalisations, means that
his text represents exceptionally insecure evidence for any historian attempting to
explore changes and continuities within the history of the medieval Byzantine
Byzantine empire during the later tenth and early eleventh centuries. 147
Kleidion Basil was forced to return to his base rather than press on into Bulgarian-held territory (see
also Skylitzes, pp.348-51)
See below, chapters five and six
75
compress and abbreviate long campaigns. As a result, his treatment of warfare in
the Balkans is customarily denuded of any sense of long-term strategy or
geographical context. Instead, his narrative is frequently expressed in a language so
uniform and uninformative, that the reader is presented with little more than a litany
of ill-defined invasions and sieges against geographically decontextualised targets.
As a result, it is difficult to distinguish one campaign from the next.
76
Chapter Two
The previous chapter began an investigation into how the later eleventh-century
synoptic historian John Skylitzes shapes his presentation of the history of the
Byzantine empire in the ninth to eleventh centuries. At the heart of this
investigation was a close textual comparison between a small section of the
'Synopsis Historion' and one of its underlying texts, the sixth book of the
Continuation of Theophanes. On the basis of evidence offered by the text alone,
Skylitzes' working methods and his use and abuse of source materials were
investigated. Some of the problems which his selection, presentation, and
interpretation of materials pose for modern historians, above all those interested in
the reign of Basil II, were highlighted. This chapter continues the analysis of
Skylitzes' presentation of the Byzantine past by widening the discussion to consider
the broader literary, social and political contexts within which the author was
working. Central to this discussion is the question of how Skylitzes, the author who
emerges from the close textual analysis in chapter one, relates to Skylitzes, the later
eleventh-century, senior Constantinopolitan civil-servant employed by the Emperor
Alexios Komnenos, the Skylitzes whose career has been outlined by the
77
characteristics of Skylitzes' testimony of the reign of Basil II, above all his
conspicuous interest in the Byzantine aristocracy and his fascination with the
Balkans.
monasteries.2
78
However, if we look more closely at Skylitzes' composition in the light of the
textual analysis undertaken in the Chapter One of this thesis and in the light of the
biography outlined by Seibt, then it becomes evident that many aspects of the
Krumbacher model are of little relevance. Indeed, rather than belonging to a literary
and social world that was in every way distinct from that of the production of high-
style history, Skylitzes' historical writings were composed in the same milieux as
high-style productions, and were in many ways their literary counterpart. In the
first place it is important to realise that by the eleventh and twelfth centuries most
chroniclers, or synoptic historians as I would prefer to call them, did not live and
work in monasteries, but in Constantinople, the imperial court, the higher echelons
of state administration, and in the law courts: in other words in the same
professional environment as many high-style historians. 3 Thus, as Seibt's
reconstruction of Skylitzes' biography demonstrates, John Skylitzes was eparch of
Constantinople in the later eleventh century. He also held the senior judicial
position of megas droungarios of the vigla.4 John Zonaras served the emperor
Alexios Komnenos (1081-1118) in the same position of megas droungarios of the
vigla and also as protoasekretis 5 Although Zonaras dedicated himself to writing
history once his career was at an end and he had retreated into a monastery, other
synoptic historians completed their compositions while still active in
Constantinople and in the employ of the emperor. For example, Constantine
Manasses who served Manuel Komnenos (1143-80) as a diplomat, was
commissioned to compile a synoptic history in verse by the emperor's sister-in-law
For Cyril Mango the differences between chronicle and high-style history are part of a more general
distinction between the thought world of the "average" Byzantine and the views of a small clique of
intellectuals who "exerted no appreciable influence on the thinking of the public at large" (C.Mango,
Empire of the New Rome (London, 1980), pp. 8-9, and also chapters 10 and 13).
3 H.G.Beck, 'Zur byzantinischen "Monchskronik"', in Speculum Historiale: Geschichte im Spiegel
von Geschichtsschreibung und Geschichtsdeutung (Festschrift K.Adler) (Freiburg/Munich, 1965),
pp. 188-97.
See above, pp.34-5
79
Irene, before she died in 1153, and long before he left court to take up his position
as metropolitan of Naupaktos.6 In these circumstances there is no reason to suppose
that Skylitzes himself could not have worked on his synopsis, either composing it
However, having noted that synoptic historians like Skylitzes worked in the same
professional milieu as high-style historians, the more important question to be
asked is whether there were any other significant overlaps between the two genres
of history writing apart from personal and professional proximity? At one level the
answer to this question ought to be positive. Alexander Kazdan, a vociferous critic
of over-schematic approaches to different manifestations of Byzantine culture,
frequently argued that Byzantine literature of all forms was typified more by
fluidity and innovation than by conservatism and inertness, and that different genres
did not remain immutable in time and isolated from each other.7 Within
historiography itself he noted parallels in material, presentation and interpretation
between higher and lower style productions. For example, as Kazdan pointed out,
eleventh- and twelfth-century high-style historians such as Michael Attaleiates and
Nikephoros Choniates, are as fond of including notices about omens and natural
o
80
Lekapenos to the Bulgarian Tsar Symeon which is riddled with high style rhetorical
devices.9
Furthermore, it is clear that whenever authors were at work in a shared physical and
professional environment, overlaps between literary genres could become even
more pronounced. For example, in the mid-twelfth century, several authors who can
be identified to varying degrees with service within the imperial court or with
members of the imperial family, participated in a groundswell of contemporary
literary activity which was characterised by considerable innovation and cross-
fertilisation of language levels and genres. 10 Many demonstrated a willingness to
experiment with different registers of language, introducing elements of vernacular
grammar and vocabulary into higher level productions. Many, including several
historians, also moved between genres. For example, Michael Glykas, a secretary at
the court of Manuel Komnenos, was a poet, the writer of theological treatises, and a
synoptic historian. Constantine Manasses composed panegyric, ekphrasis, a verse
romance 'Aristandros and Kallithea', and a verse account of his diplomatic mission
to Palestine in 1160, as well as his verse synoptic history. * ]
Although scholarly interest has only recently begun to focus on literary production
in the later eleventh and early twelfth centuries, it is worth noting that here too
innovation and cross-fertilisation has been identified, particularly in works
et al. (eds.), Studies in John Malalas (Sydney, 1990), pp.67-85; ACameron, Procopius and the Sixth
Century (London/New York, 1996), pp.24-32).
9 George the Monk Con., p.900. This speech is probably taken from the pro-Lekapenos family
history that was one of the Logothete's sources.
10 See for example R.Browning, 'Byzantine Scholarship', Past and Present 28 (1964), pp. 13-17;
A.P.Kazdan and A. Wharton Epstein (eds.), Change in Byzantine Culture in the Eleventh and Twelfth
Centuries (California Berkeley, 1985), pp. 83-6. Apart from historians, other writers who moved
between genres and switched registers included Theodore and Manganeios Prodromes, the
anonymous author of the 'Spaneas', and the authors of the twelfth-century novels.
The literary achievement and career of Glykas are summarised in: Hunger, Hochsprachliche
profane Literatur, i, 422-6; for Manasses see idem., pp.419-21; ODB, ii, 1280; and Jeffreys, 'The
Attitudes of Byzantine Chroniclers', pp.202-15.
81
associated with the imperial court, the higher ranks of the civil service, and
Constantinople. For example, some scholars have noted the emergence in this
period of a new interest in the telling of vivid stories. Although writers in locations
outside Constantinople, such as Kekaumenos, made use of entertaining narratives in
their literary productions, this new interest was most conspicuous among those
authors who moved in the highest social and political circles in the capital. A
fondness for elaborate anecdotes, particularly those describing military endeavour,
has been noted among high-style historians such as Bryennios and synoptic
10
historians such as Skylitzes. It has even been argued that this greater interest in
martial narrative was inspired during the latter years of the eleventh century by the
arrival in Constantinople of provincial aristocrats fleeing the contemporary Turkish
invasions of Asia Minor. Beaton has suggested, for example, that the epic/romance
'Digenes Akrites' was composed in Constantinople in this period, as emigres from
central Anatolia sought to make a permanent written record of much older oral
poems describing the daring-do of life on the ninth- and tenth-century Arabo-
Byzantine frontier. 13 Although other scholars, most notably Elizabeth Jeffreys and
Paul Magdalino, have suggested that the Digenes epic was written down in
Constantinople during the mid rather than the early twelfth century, the importance
of the arrival of aristocratic refugees from central Anatolia for developments within
Byzantine literature throughout the Komnenian period has been widely accepted. 14
12 Mullett, Theophylact of Ochrid, pp.69-78; C.Roueche, 'Byzantine Writers and Readers: Story
Telling in the Eleventh Century', in R.Beaton (ed.), The Greek Novel (London/Sydney, 1988),
pp. 123-132; J.D.Howard-Johnston, 'Anna Komnene and the AlexiacT, in M.Mullett and D.Smythe
(eds.), Alexios I Komnenos (Belfast, 1996), pp.260-302. It should, however, be noted that this
enthusiasm for military anecdotes was not an entirely new phenomenon. There are stories about
heroic exploits in texts which predate the later eleventh century, such as Theophanes Continuatus
and Leo the Deacon.
13 R.Beaton, 'Cappadocians at Court: Digenes and Timarion', in M.Mullett and D.Smythe (eds.),
Alexios I Komnenos (Belfast, 1996), pp.329-338.
14 Digenes Akrites: Digenis Akritis: the Grottaferrata and Escorial Versions, ed. E.Jeffreys
(Cambridge, 1998), pp.xvii, Ivi-lvii; P.Magdalino, 'Digenes Akrites and Byzantine Literature: the
82
Nonetheless, there is a potential objection to the view that synoptic and high-style
histories must necessarily be seen as part of the same cultural phenomenon simply
because they were produced by historians working within the same physical and
temporal environment. This is the very fact that contemporary Byzantine writers
indicate a quite explicit generic difference between these two literary productions.
In the preface to his world chronicle the ninth-century historian George the Monk,
criticised the writers of secular history for their ostentation, loquacity,
incomprehensibility and overweening desire for applause. 15 The prefaces to
eleventh and twelfth-century synoptic histories contain echoes of George's
criticism. Despite his considerable classical erudition and his use of Dio Cassius as
a source for his coverage of the Roman Republic, the mid twelfth-century synoptic
historian John Zonaras chastised those historians who indulged in detailed
descriptions of military matters, lengthy and irrelevant digressions, and improbable
dialogues. Here Zonaras appears not only to have been criticising ancient
historians, but also the classicising historians of more recent generations such as
Anna Komnene and Nikephoros Bryennios. In contrast, Zonaras indicates that his
own work belongs outside this tradition when he says that it is his ambition to
produce a synopsis which will present a short but clear view of important past
events. 16 Even Constantine Manasses mirrors the sentiments of contemporary
synoptics when he promises his patroness Irene "..... a clear and comprehensible
treatise........ giving plain teaching in history", which will remedy the contradictory
accounts of the writers of histories and chronographies. 17 However, it is Skylitzes
twelfth century background to the Grottaferrata Version', in R.Beaton and D.Ricks (eds.), Digenes
Akrites: New Approaches to Byzantine Heroic Poetry (London, 1993), pp. 1 -14
15 George the Monk: Georgius Monachus Chronicon, ed. C.de Boor (and P Wirth) (2 vols, Stuttgart,
1978 edition), i, 1-2.
Zonaras, ed. Finder (Bonn, 1841), pp.4-6. Hunger, Hochsprachliche profane Literatur, i, 417 sees
Zonaras specifically criticising Anna Komnene and Nikephoros Bryennios. For Zonaras' good
knowledge of classical sources see M. di Maio, 'Smoke in the Wind: Zonaras' use of Philostorgius,
Zosimos, John of Antioch and John of Rhodes in his Narrative of the Neo-Flavian Emperors', B 48
(1988), pp.230-255
Manasses, p. 5.
83
himself who seems to provide the clearest distinction between synopsis and high-
style history. As we saw in the previous chapter, Skylitzes uses his preface to draw
an explicit contrast between his short and readable continuation of George the
Synkellos and Theophanes with the long psogos- and encomium-riddled appraisals
of recent generations of historians. 18
Yet, it can be argued that the eleventh and twelfth-century synoptic historians
protest their differences from their high-style counterparts too much. In many cases
they exaggerate the differences in form and content between their writings and the
high-brow histories of their contemporaries, while failing to highlight the
conspicuous similarities between the two genres. And indeed, in many cases it is
the similarities which are the more striking. For example Zonaras couches his
reasons for writing in the same terms as the high-style historian Michael Psellos.
Both historians explain that it is only the encouragement of friends that has
persuaded them to take up their pens. 19 Zonaras' negative portrayal of Alexios
Komnenos is a blatant psogos for which he, rather than his source, appears
responsible.20 Moreover, at a general level, many synoptic historians of the eleventh
and twelfth century deviate greatly from the idealised model they claim to follow.
They often ignore the annalistic chronological structure of their synoptic
predecessors such as Theophanes, employ more elevated language, and take a less
providential view of history: features they share in common with more high-brow
historians.21 Taken as a whole the evidence suggests that by the eleventh century
the composition of synoptic history had become a genre of court literature
expressed in a middling to high-style register of Greek with its own particular
i ft
Skylitzes, pp.3-4; see above, pp.50-2
19 Zonaras, ed. Finder (Bonn, 1841), p.4, Psellos, pp. 127-8
20 Zonaras, ed. Biittner-Wobst (Bonn, 1897), Bk. XVHI, pp.726-768
Even Krumbacher accepted that Zonaras was skilled in his appreciation and use of Attic Greek
(Geschichte der byzantinischen Literatur, i , 370-4). Mango and Scott (Chronicle of Theophanes,
84
rhetoric. If a court official in this period wanted to write a history with a longer
compass than his own lifetime, he adopted the chronicle form at a superficial level.
In order to indicate the synoptic nature of his production to his potential audience, it
was important that he should state in the preface that he intended to produce a short,
unbiased, and easily understandable account. Thereafter, considerable innovation
and overlap with the methods and materials of higher style histories was possible.
I.Sevcenko has indicated the frequency with which high-style texts were reshaped
into handbook paraphrases expressed in a middling level of Greek, precisely so that
p.Hi) reflect on the failure of the synoptic successors of Theophanes to adopt his strictly annalistic
structure.
22
Skylitzes, pp.3-4; see above, pp.50-2
23 Ibid., p.4; for a full translation see the appendix
85
subsequent readers and writers could absorb the contents of verbose texts without
being forced to read the originals.24
Yet, the main question which arises from the recognition that synoptic historians
such as Skylitzes and Zonaras were educated civil servants who used high-style
histories in their production of historical handbooks, is whether their audience
should also be located within the environs of the imperial court? Since recent
research has begun to present a more optimistic picture of levels of literacy and
access to literature outside the higher echelons of the Byzantine civil service, it
would be premature to limit the reception of Skylitzes' 'Synopsis Historion' purely
to those most senior ranks of the imperial administration and court familiar with the
writing of high-style histories.25 Certainly the incidence of nine Skylitzes'
manuscripts dating from the twelfth to fourteenth centuries indicates that his text
had a long-term audience which far exceeded the narrow world of the Komnenian
*\
elite. The fact that the Madrid manuscript was copied in Norman southern Italy
during the mid-twelfth century indicates that its appeal extended beyond the
political borders of the empire itself.27
24 I.Sevcenko, 'Levels of Style in Byzantine Prose', JOB 31.1 (1981), pp.309-10; idem., 'Some
Additional Remarks to the Report on Levels of Style', JOB 32.1-2 (1982), pp.228-9
25 It has usually been assumed that the high cost of book production meant that few individuals
outside elite government circles had regular access to books in medieval Byzantium (C.Mango and
I. Sevcenko (eds.), 'Byzantine Books and Bookmen', DOP 29 (1975). However, more recently
Margaret Mullet has argued for a more optimistic view of literacy levels and the dissemination of
literature outside the court elite (M.Mullett, 'Aristocracy and Patronage in the Literary Circles of
Comnenian Constantinople', in M.Angold (ed.), Byzantine Aristocracy (BAR International Series,
Oxford, 1984), pp. 173-201, and eadem., 'Writing in Early Medieval Byzantium', in R.McKitterick
(ed.), The Uses ofLiteracy in Early Medieval Europe (Cambridge, 1990), pp. 156-184).
In 1201 the library of the monastery of St John on Patmos owned a copy of Skylitzes' history
(syngraphe) (C.Diehl, 'Le tresor et la bibliotheque de Patmos au commencement du 13e siecle', BZ
1 (1892), pp.500, 522; C.Astruc, 'L'inventaire dresse en septembre 1200 du tresor et de la
bibliotheque de Patmos, edition diplomatique', TM 8 (1981), pp.28-29; Snipes, 'The
'Chronographia' of Michael Psellos and the textual transmission of the Byzantine historians', p.57).
86
libraries, readers and interpolators, circumstantial and textual evidence suggests that
the author's principal, original audience was based among the Constantinopolitan
elite. The circumstantial evidence relates to the fact that synoptic histories such as
Skylitzes are often preserved in the same manuscripts as contemporary, high-style
literary works. Thus, one of the earliest manuscripts of the 'Synopsis' of Skylitzes,
manuscript C (Paris BN Coislin 136) which is usually dated to the twelfth century,
contains a variety of high-style materials including the 'Historia' of the later
eleventh-century lawyer and courtier Michael Attaleiates, and the speeches of
Manuel Straboromanos, the megas hetaireiarches (head of the imperial bodyguard),
*
during the final decade of the reign of Alexios I. Manuscript E also conveys
-%o
Attaleiates' history.
87
connections between the language levels of synoptic texts and the professional
character of the audience remains valid.
However, the best evidence that Skylitzes was writing in response to the
competences and concerns of his contemporaries within Constantinopolitan high
society, is his conspicuous interest in the political elite of the empire, which for
convenience sake I shall term 'the aristocracy'. During the analysis of his treatment
of the reign of Romanos Lekapenos contained in the last chapter, Skylitzes'
fondness for swashbuckling military anecdotes and his willingness to embroider
heroic incidents was noted. This enthusiasm seems compatible with the almost
obsessive interest in martial valour which Alexander Kazdan has identified among
the Byzantine ruling families of the later eleventh century. 31 Further convincing
evidence that Skylitzes was writing with an aristocratic audience in mind, is the
emphasis that he places upon carefully recording the names, titles and pedigrees of
the leading families of the empire during the ninth to eleventh centuries. This
emphasis is particularly visible in his retention of the family trees and marriage
alliances of the Lekapenos, Argyros and Kourkouas families during his coverage of
the reign of Romanos Lekapenos. Nor is his interest in such clans solely confined to
the early tenth century. Instead, in the case of the Argyros family his attention
persists into the later tenth and early eleventh centuries. Thus, in his coverage of the
reign of Constantine Porphyrogenitus (945-59), he includes an extensive report on
Marianos Argyros' expedition against Sicily. During the reign of Basil II, Basil
Argyros is cited twice: first as the strategos of Samos who was defeated by the
Italian rebel Meles in 1011; second as the first Byzantine commander of
See above, pp. 56-9, for discussion of Skylitzes' syntax and language register
Jeffreys et al., Studies in John Malalas, pp.6-11
See, for example, Kazdan, 'Der Mensch', p.lSff; idem., 'Aristocracy and the Imperial Ideal', in
M.Angold (ed.), Byzantine Aristocracy (BAR International Series, Oxford, 1984), pp.43-58; Kazdan
and Wharton Epstein, 'Change in Byzantine Culture in the Eleventh and Twelfth Centuries",
pp. 104-117, Kazdan, 'The Social Views of Michael Attaleiates', pp.23-86.
Vaspurakan. Also recorded is the marriage of the sister of Romanos Argyros to the
Doge of Venice. She is explicitly described as "the sister of Argyros who was
emperor after these things."32
Indeed, Skylitzes' fascination with the personnel of the elite families of the empire
is visible elsewhere in his coverage of the reign of Basil II. His narrative of the
revolts of Bardas Skleros and Bardas Phokas is replete with names of leading
than a fleeting presence. For example, at the battle of Kleidion in the Haimos
32 Skylitzes, pp.266-7, 348, 354, 343; cf. Lupus Protospatharius, MGH SS V, p.58; see below, p.320
89
Bulgarians, the success of imperial armies is not accorded to Basil himself, but to
the inspiration of the commander Nikephoros Xiphias. When the Byzantine armies
could not break through the blockade which the Bulgarians had established in the
pass of Kleidion, it was Xiphias who recommended that a party of men should be
led through the mountains to attack the enemy from the rear. Xiphias himself then
led this operation.34 This concentration on the aristocracy at the expense of the
emperor may make sense of the criticism included in Skylitzes' preface that
historians such as Michael Psellos: "have fallen short of accuracy, having passed
over the most opportune things, ....having made an enumeration of the emperors
and having taught who happened to hold imperial rule after whom and nothing
more".35 Here, Skylitzes may be implying that the deeds of the emperor were not
the more "opportune things"of the past, and that he himself in the course of his
compilation was about to depart from this traditional interpretation of history. This
is not to suggest that Skylitzes entirely ignores the person of the emperor as a unit
around whom the rest of his text is organised. His narrative structure is, for
example, based on individual imperial reigns. He usually retains imperial accession
dates and lengths of reigns. Yet, within the narrative itself, the focus falls as much
on individual members of the elite as on emperors.
90
sense, but with the exception of the principal actors such as Bardas Skleros or
Bardas Phokas, individual members of the elite are almost never identified. Thus,
during his coverage of the first revolt of Skleros, Zonaras omits all Skylitzes'
aristocrats from his account of Basil's reign, and refocuses the narrative around the
person of the emperor himself.
One explanation for why Skylitzes' 'Synopsis' emerges as the history of the
aristocrats is that it was composed in the second half of the eleventh century at a
time of immense political tension within the Byzantine political elite. Whether
Skylitzes was writing the 'Synopsis' in either the 1070s as Seibt suggests, or the
1090s as I would maintain,37 this was a period characterised by endemic
competition for imperial power among the more important families of the
Byzantine empire. The past in such circumstances was an important legitimising
tool. Michael Attaleiates' elaborate genealogy of the emperor Nicephoros III
Botaneiates (1078-81), which traces the emperor's descent from the tenth-century
Phokades and the first-century Fabii, demonstrates clearly how the writers of
history looked to both the immediate and the remote past to bolster dynastic
prestige. Within his generally encomiastic treatment of the emperors Constantine X
(1059-67) and Michael VII Doukas (1071-78) in the 'Chronographia', Michael
91
JO
Attaleiates and Psellos, since the 'Synopsis' is a survey which ranges far wider than
the achievements and pedigrees of a single family, it is possible that his text was
composed within the very specific political context of aristocratic rivalry at the end
It is possible to unravel this political context by starting from the observation that
Skylitzes' 'Synopsis' often resembles a 'Who Was Who' of the 'Who Is Who' of
the last quarter of the eleventh century. 39 That is to say Skylitzes' text frequently
displays a conspicuous interest in those figures from the Byzantine past whose
ancestors enjoyed political significance in the later eleventh century. Thus, in
Skylitzes' narrative of the first Skleros revolt during Basil's reign, he pays
particular attention to Anthes Alyates. He describes Alyates as one of Skleros' most
effective servants. He notes that it was Alyates whom Skleros sent to rescue his son
Romanes Skleros from the imperial palace before the revolt began. He also records
Alyates' death during a battle in the Anti Taurus mountains a few months after the
outbreak of revolt.40 Yet, Skylitzes' detailed interest in Alyates' role in Basil's
reign may be dictated more by the political significance of his family in the later
eleventh century than during the later tenth. In this context it is worth noting that
38 Attaleiates (Michael): Historia, ed. I.Bekker, (CSHB, Bonn, 1853), pp.216-27; Psellos, pp.134,
141; Kazdan, 'Social Views of Michael Attaleiates', pp.28-9; idem., 'Aristocracy and the Imperial
Ideal', p.45; P.Magdalino, 'Byzantine Snobbery', in M.Angold (ed.), Byzantine Aristocracy (BAR
International Series, Oxford, 1984), p.64. For late eleventh-century and twelfth-century interest in
genos see also P.Magdalino, 'Honour among the Romaioi: the Framework of Social Values in the
World of Digenes Akrites and Kekaumenos', BMGS 13 (1989), pp. 183-4, 193-6.
Analogous to the case of Skylitzes may be the thirteenth-century synoptic history written by
George Akropolites, the megas logothetes of Michael VIII Palaiologos (1258-82). Ruth Macrides
argues that this history resembles a 'Who's Who' of the Empire of Nikaia. Like Skylitzes
Akropolites demonstrates concern that the full names and titles of all officials and aristocrats should
be accurately recorded: R.Macrides, 'The Historian in the History', in E.Jeffreys et al. (eds.),
v, Studies in Honour of Robert Browning (London, 1996), pp.221-2.
92
one of Alyates' descendants, a certain Theodore, had been a leading supporter of
Romanes IV during the civil wars which overtook the Byzantine empire in 1071.
He was taken prisoner in battle and blinded by the forces of Michael VII Doukas.
His career in the political ferment of the third quarter of the eleventh century is
described by Michael Attaleiates. Significantly, he is a character in whom Skylitzes
himself took an interest. When Skylitzes came to use Attaleiates' 'Historia' to write
93
the Anemai were leading figures in the army.44 The prominence of the Anemai in
early Komnenian political society is confirmed in the letters of Theophylact, the
was Nicholas Anemas who held a senior military position in Macedonia, possibly
as doux of Skopje.45
Byzantine elite of certain "outsiders" during the later tenth and early eleventh
centuries may also have been determined by the political significance of aristocratic
families with the same name at the end of the eleventh century. Thus, the inclusion
of the name Pakourianos among the list of Iberians who entered the service of Basil
II in 1000, may stem from his perceived association with the later eleventh-century
general Gregory Pakourianos. The role of this latter-day Pakourianos as a
powerbroker within the Byzantine polity at the time of the Komnenian coup against
Nikephoros III Botaneiates in 1081 is described in detail in the 'Alexiad'. Created
megas domestikos by Alexios as a reward for his loyalty, he was the founder of the
Georgian monastery at Backovo in Bulgaria. ^ Although he himself appears to
have died without children in 1086, one of his younger relatives, also called
Gregory, was the son-in-law of Nikephoros Komnenos, the brother of the emperor
Alexios. In the mid-1090s Archbishop Theophylact of Ochrid noted that despite his
youth this junior Pakourianos enjoyed free access to and counsel with (parrhesia)
94
the emperor Alexios.47 A comparable example concerns Skylitzes' record of the
Basil II's reign. Once again a member of a family with the same name was closely
associated with the Komnenos regime at the end of the eleventh century. In May
1089 Theodore Senecherim, described as a close associate (oikeios) of the emperor
Alexios, was sent to oversee the restoration of the Xenophon monastery on Mount
Athos to its original founder.48
Of course both these late eleventh-century personalities may have had very little to
do with their late tenth- and early eleventh-century namesakes. Indeed, in the
typikon of his monastery at Backovo, Gregory suggests that his family had still
been located in Iberia until very recent times. He identifies his father as
Pakourianos, the "archon of the archontes of the very noble race of the Iberians."49
After his father's early death he himself wandered through Armenia, Iberia and
Syria before seeking employment within the empire. 50 His Caucasian background
was also noted by Anna Komnene.51 However, the probity of genealogical
connections did not have to be Skylitzes1 principal concern. The more important
46 Skylitzes, pp.339-40; Anna Komnene, i, 73-4; Typikon: P.Gautier, 'Le typikon du sebaste
Gregoire Pakourianos', REB 42 (1984), pp.5-145; P.Lemerle, Cinq etudes sur le Xle siecle byzantin
(Paris, 1970), pp. 114-191; Skoulatos, Lespersonnages byzantins, pp. 112-5.
47 Theophylact of Ochrid, letter 68. Gregory was appointed to command in Macedonia during the
mid-1090s possibly as governor of Ochrid (Mullett, Theophylact of Ochrid, pp.94, 130, 146, 186,
215,276).
48 Skylitzes, pp.354-5; see below, chapter six, p.284; Actes de Xenophon, Archives de 1'Athos XV,
ed. D.Papachryssanthou (Paris, 1986), p.71. This later eleventh-century Senecherim may also be the
senior fiscal official of whom Theophylact of Ochrid complained c. 1094/5 (Theophylact of Ochrid,
letter 77; Mullett, Theophylact of Ochrid, p. 130).
49 Typikon (Pakourianos), p. 21
50 Ibid., p.92
5 'Anna terms Pakourianos an Armenian (Anna Komnene, i, 74). Pakourianos identifies himself
rather more broadly. According to his typikon he founded his monastery explicitly for the Iberians
who had served with him in the Byzantine army and who only spoke Georgian (Typikon, p.21).
However, he added his signature to the typikon in Armenian script (ibid., p. 130), while at the same
time mentioning that he had Armenian Monophysite relatives (ibid., p. 129); cf. N.G.Garsoian,
'Armenian Integration into the Byzantine Empire", in H.Ahrweiler and A.E.Laiou (eds.), Studies on
the Internal Diaspora of the Byzantine Empire (Washington, 1998), pp. 89-91.
95
point is that these names from the tenth and early eleventh centuries fitted the
political landscape of the later eleventh-century world very neatly.
The most conspicuous group of "outsiders" to attract Skylitzes' interest in the reign
of Basil II was the Bulgarian royal family which was removed from imperial office
in 1018 and absorbed within the Byzantine aristocracy. As we have already noted,
the second half of Skylitzes' testimony for the reign of Basil II concentrates almost
exclusively on relations with the first Bulgarian empire. Within this narrative most
attention is paid to the last four years of the war and the final Bulgarian
capitulation.52 Dominating the narrative of the surrender itself is a detailed analysis
of the size and personnel of the Bulgarian royal family. Skylitzes notes that when
Maria, the widow of the last Bulgarian tsar John Vladislav, was brought before
Basil II at Ochrid, she was accompanied by three of her own sons and six
daughters, as well as by two daughters and five sons of the previous tsar, Gabriel
Radomir, and an illegitimate son of the tsar Samuel, Gabriel's father. Skylitzes goes
on to record the surrender shortly afterwards of three more sons of John
Vladislav.53 Moving beyond Basil's reign, Skylitzes continues to chart the progress
of the principal members of the family, stressing above all their participation in
conspiracies and their marital connections with the Byzantine aristocracy as a
whole. Two episodes of aristocratic unrest involving Prousianos, one of John
Vladislav's sons, during the reigns of Constantine VQI and Romanes III are
described by Skylitzes. In the first of these accounts the reader is told that that the
sister of Prousianos was married to Romanos Kourkouas who was blinded by
Constantine VIII on suspicion of conspiracy.54 During the reign of Michael IV,
Skylitzes records the participation of Alousianos, the brother of Prousianos, in the
96
Balkan revolt of Peter Deljan. He begins his description with the reflection that
Alousianos was the second son of Aaron (that is John Vladislav), and that he was
married to a wife with an estate in the theme of the Charsianon.55
It is possible that Skylitzes displays such long-term interest in the family history of
the house of John Vladislav because by the end of the eleventh century all the
principal families of the empire, including the Komnenoi, had genealogical
connections to the former Bulgarian ruling dynasty. During the revolt of Isaac
Komnenos in 1057 Skylitzes cites the magistros Aaron, one of the sons of John
Vladislav, as one of the commanders who remained loyal to the emperor Michael
VI. Nonetheless, he also mentions that Aaron's sister was married to Isaac. 56 This
sister was called Catherine. She became empress when her husband Isaac overthrew
Michael VI and became emperor. Two years later Isaac was forced to abdicate.
According to Michael Psellos, Catherine held Psellos himself responsible for her
husband's demise. However, Psellos seems to have had grudging admiration for
Catherine, whom he noted was descended "from a very noble family". 57 Anna
Komnene confirms the continuing significance of Bulgarian aristocratic ancestry to
status within Byzantine political society at the end of the eleventh and beginning of
the twelfth century. One obscure revolt against Alexios, she notes, was led by, "....
a man who traced his ancestry back to the famous Aaronoi on one side". The fact
that this rebel descended from an illegitimate branch of the family does not appear
53 Skylitzes, pp.359-60. All of the medieval manuscripts record that one of the six sons was
Prousianos. Manuscript U adds the names Alousianos, Aaron, Traianos, Radomir; manuscript E
adds the name Klimen.
54 Skylitzes, pp.372, 376, 384
"Skylitzes, pp.413-415
56 Skylitzes, p.493
"Psellos, pp. 131-3
97
to have diminished the lustre of his pedigree or the extent of his political threat in
Anna's eyes. 58
Skylitzes' treatment of the Bulgarian royal family makes it particularly evident that
he did not simply conceive of the 'Synopsis Historion' as a genealogical address
book of the past. Instead, it is clear that his wider interpretation of the history of the
Byzantine empire, the "more opportune things" of the past, was centred on dynastic
competition and revolt, particularly among those families who were famous at the
end of the eleventh century. The regularity with which aristocratic insurrection had
punctuated the senior levels of elite political society in Byzantium during the
decades following the death of the last Macedonian empress, Theodora, in 1056,
does not make this interpretation surprising. Nonetheless, the idea that Skylitzes
produced a compendium of revolt for an audience composed of predominantly of
aristocrats, sits uneasily with his position as a highly placed official within the
administration of the emperor Alexios I Komnenos.
In order to appreciate fully the ambiguity between Skylitzes' historical interests and
his professional role as a civil servant, it is worth stressing the seniority of the
position which he enjoyed within Komnenian government at the beginning of the
1090s and his close association with the emperor Alexios himself. In the first place,
Skylitzes held the title of kouropalates. Although the author of the late
eleventh/early twelfth century 'Historia Syntomos' of pseudo-Psellos claims that
rank of kouropalates "is common now and borne by many people", it was still an
important secular title during the reign of Alexios Komnenos. 59 At the 1094-5
CO
Anna Komnene, iii, 88-91. Cheynet, Pouvoir, p. 102. In 1107 this Aaron accompanied the emperor
Alexios on his campaign in the western Balkans against Bohemond (Skoulatos, Les personnages
byzantins, pp.3-4).
59
Pseudo-Psellos, p. 100. This rank held a much more exclusive cachet in the tenth century. Leo
Phokas, the brother of the emperor Nikephoros Phokas (963-9), held this title, as did Bardas Skleros
after he surrendered to Basil II in 989. When Skleros was kouropalates he was described as second
98
Synod of Blachernae only seventeen laymen held a title of superior rank, and many
of these were members of leading aristocratic families related by marriage to the
ruling dynasty.60 Moreover, although Skylitzes' position of megas drungarios of
the vigla was usually held by someone of legal training, evidence from the twelfth-
century 'Ecloga Basilicorum' suggests that important officers of the Komnenian
judiciary were in the first instance political place men. They were appointed, ".....
not because of their legal knowledge but because of their rank and their loyalty
...... to the emperor. The many and frequent cares which beset officials and their
attendance on the emperor excuse their ignorance."61 When Constantine
Keroularios was droungarios of the vigla during the reign of Michael VII (1071-8),
he held the senior title of protoproedros and was one of the emperor's senior
advisers. Such was his political influence that he was accused by contemporary
historians of controlling grants of titles and offices.62 Another sign of Skylitzes'
own political importance to the Komnenian regime was his exercise of the office of
eparch of Constantinople, during the early 1090s, the period Alexios was under
constant threat from internal rivals as well as external foes.63 At such a sensitive
time the eparch, responsible for public order and the administration of justice
within the capital city of the empire, had to be a servant of the utmost loyalty and
discretion. In these circumstances, the question of how one of the most senior
Komnenian officials, and a man of tried and tested loyalty to the incumbent
in rank only to the emperor (Skylitzes, p.339; Psellos, p. 16). R.Guilland, 'Curopalate', in litres et
fonctions de I'empire byzantin (London/Variorum, 1976), no HI, pp. 190-2, discusses the decline in
the prestige of this title during the eleventh century; he recognises that Skylitzes and John
Thrakesios held the title kouropalates, but does not seem to realise that they were the same man.
60 Paul Magdalino, The Empire ofManuellKomnenos (1143-1180), (Cambridge, 1993), appendix 2,
pp. 501-3.
" Ibid., pp.262-3
62 Zepos and Zepos, lus, i, 279-83; Bryennios, pp.211-13; Hicks, 'The Life and Historical Writings
of Michael Attaleiates', pp.238-9.
63 Cheynet, Pouvoir, pp.96-9, 361-7 argues that by the 1090s Alexios' regime was widely criticised.
Despite enjoying some respite from Norman attack after the death of Robert Guiscard in 1085, and
defeating the Pechenegs in 1091/2 at Levounion, domestic discontent was rife. Opposition ranged
from criticisms contained in the 1090-1 speeches of John the Oxite, the rebellion of the general
Humbertopoulos in the same year, and the conspiracy of Nikephoros Diogenes in 1094.
99
emperor, could write a history which interpreted the past in terms of endemic
ranking Komnenian official could have been compatible with his interpretation of
remote and recent history as the story of the rival aristocratic families. Magdalino
has suggested that the regime of the emperor Alexios Komnenos was characterised
by its inclusive policies towards other leading aristocratic families. One
manifestation of this greater inclusivity was frequent intermarriage between the
great dynasties of the empire, such as the Komnenos, the Doukas, the
Kontostephanos, and the Bryennios families. In the early years of Alexios1 reign
Magdalino believes that this strategy of intermarriage was not simply a form of
imperial reward for loyal supporters, but was a conciliatory policy explicitly
designed to placate recent opponents.64 While one would not wish to imply that
Skylitzes' history was sponsored or commissioned by the emperor, it is possible that
the articulation of history presented in the 'Synopsis Historion' by one of the
emperor's most loyal servants may be related to this official policy of division
healing within the aristocracy. In the emphasis which it places on the predecessors
of contemporary aristocratic families, their valiant deeds, their long-standing
associations with high politics, and their frequent intermarriages, the 'Synopsis' of
Skylitzes may contain echoes of the Komnenian propaganda which accompanied
production may help to explain the retention and elaboration of other materials
100
within the 'Synopsis'. For example at several key junctures in his later tenth- and
saints, Theodore Stratelates^ George and especially Demetrios. He records that after
the fall of Preslav in 971 to the emperor John Tzimiskes, the Rus were defeated in
open battle on the feast day of Saint George. During that emperor's final victory
over the Rus at Dristra in 971, Saint Theodore was said to have appeared on a white
horse offering assistance to the imperial armies. Tzimiskes' subsequent support for
65 Skylitzes, pp.300, 308-9. The emperor rebuilt the church containing the tomb of Saint Theodore at
Euchaneia and renamed the town "Theodoropolis". N.Oikonomides, 'Le dedoublement de Saint
Theodore et les villes d'Euchaita et d'Euchaina', AB 104 (1986), pp.327-35 argues that
Euchaina/Euchaneia, the cult centre of Saint Theodore the general (Stratelates), was probably a
hilltop site near the city of Euchaita, the location of the tomb of Saint Theodore the Recruit (Tirori).
The tombs of both saint Theodores were popular pilgrimage sites during the eleventh century
(E.Malamut, Sur la route des saints byzantins (Paris, 1993), p.42).
Skylitzes, p.339. It was probably during this pilgrimage to Thessalonika, which preceded his 991
campaign against the Bulgarians, that Basil met Saint Photios, the monk who became his spiritual
guide and accompanied him on campaigns thereafter. During his lifetime Basil made pilgrimages to
the tombs of the warrior martyrs Saint George, Saint Theodore Tiron and Saint Theodore Stratelates
(see above, pp.8-9, n.25; Crostini, 'The Emperor Basil II's Cultural Life', p.78; Schlumberger,
L 'Epopee byzantine, i, 646 and ii, pp.46-7).
Skylitzes, p.413. Kekaumenos, who conveys a shorter narrative of the same siege, does not
mention the intervention of Saint Demetrios (Kekaumenos, pp. 160-2)
101
Adrian feature saints Theodore and George respectively. 68 Many Komnenian
the desirability of cohesion within the early Komnenian body politic, however, is
the fact that most of the miraculous interventions cited by Skylitzes occur in Balkan
contexts. At the end of the eleventh century the Balkans were the principal theatre
of Byzantine warfare. In the first two decades of Alexios' reign Byzantine armies
were engaged with Normans in the west of the region and nomads in the north.
Anna Komnene's accounts of Alexios' Balkan expeditions in this period make it
clear that many leading aristocrats occupied positions of high command within the
Byzantine campaign armies. Yet, recent research by Cheynet has argued the Balkan
policy of the early years of Alexios' reign may have been unpopular with many
members of the Byzantine elite. He notes that most incidents of aristocratic
discontent against the Komnenian regime occurred shortly after imperial victories
in the Balkans; whereas, once Alexios turned his attention to Asia Minor in the
second half of his reign such dissent evaporated.70 If Skylitzes was at work on the
'Synopsis Historion' in the 1090s, then he was writing at a time when imperial
attention was still predominantly focused on the Balkans. In these circumstances, it
seems reasonable to argue that one of Skylitzes' motivations in writing may have
been to allay contemporary aristocratic suspicion of the geographical trajectory of
Alexios' campaigns. Such a motivation would explain two conspicuous
preoccupations within Skylitzes' text. First, the overwhelming interest which his
narrative takes in the Balkans; and second, the author's frequent demonstration that
68 See for example G.Zacos and A.Veglery, Byzantine Lead Seals (3 vols., Basle, 1972), i, nos.2701-
2 (Isaac brother of Alexios); 2704-07 (Alexios before he became emperor); 2708 (Adrian).
102
it was the bravery of earlier aristocratic generations which had regained the Balkans
for the empire in the reigns of John Tzimiskes and Basil II.
Both of these preoccupations are visible throughout the whole of Skylitzes' text. As
we have already seen the second half of Skylitzes' treatment of the reign of Basil is
almost entirely devoted to the Bulgarian wars. During the course of his narrative he
lionises senior army generals such as Nikephoros Xiphias.71 He also dedicates more
than two-thirds of his coverage of the reign of John Tzimiskes to the Balkan sphere,
the context in which the heroics of Anemas are recorded.72 Other aristocratic names
who receive Skylitzes' attention during his Balkan coverage in the later tenth and
early eleventh centuries include Gregory and Ashot Taronites, Theophylact and
Michael Botaneiates, Constantine Diogenes, and David Areianites. 73 These were
exactly the families who enjoyed immense political authority at the end of the
eleventh century and whose loyalty Alexios needed to consolidate. For example,
the Taronites were a high-profile family related to the Komnenoi by marriage.74
The Diogenes family had already provided one emperor of recent times, Romanes
IV (1068-71), and by the 1090s his son Nikephoros was perceived to be the most
dangerous of Alexios' rivals. Despite serving with the imperial campaigns of the
1080s against Normans and Pechenegs, and being appointed governor of Crete,
Nikephoros was eventually blinded in 1094 on charges of conspiracy. 75 Meanwhile,
the scion of another leading family, Nikephoros III Botaneiates had occupied the
imperial throne before Alexios became emperor. Even families featured in
69
Mullett, Theophylact of Ochrid, p.51 and n.213. The Saint Demetrios issue was struck at
Thessalonika.
70 Cheynet, Pouvoir, pp.362, 368.
71 See above, p.90
72 See above, p.93
73 Skylitzes, pp.339-42 (Taronites); pp.350 (Botaneiates); pp.352, 355-6, 365 (Diogenes); pp.345,
354-5, 358 (Areianites).
74 John Taronites was the brother-in-law of Alexios Komnenos (Cheynet, Pouvoir, p.277)
75 Nikephoros participated in military actions against the Normans in 1081 and Pechenegs 1087
(Anna Komnene, i, 155; ii, 90, 96, 100); his conspiracy (ibid, ii, 169-184).
103
Skylitzes' tenth and early eleventh-century Balkan coverage with less obvious later
eleventh-century imperial credentials remained influential in the period when
Alexios came to power. Towards the end of the eleventh century, for example, a
certain Leo Areianites held the senior rank of protoproedros and was katepan of the
theme of the Optimatoi.76
76 J.Nesbitt and N.Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks and in the Fogg
Museum ofArt, (3 vols., Washington D.C., 1996), iii, no.71.27
77 McGrath, 'The Battles of Dorostolon', pp. 152-62; Skylitzes, pp.304, 308; Leo the Deacon, p. 148
(Kourkouas), p. 153 (death of Anemas). The possibility that Leo the Deacon and John Skylitzes used
the same source, probably a war diary (Kriegesbuch), for their accounts of warfare in the Balkans
during the reign of John Tzimiskes, was also raised by G.Moravcsik, Byzantinoturcica, Die
bvzantinischen Quellen der Geschichte der Turkvolker, 2nd edn (Berlin, 1958), pp. 3 98-9.
See above, pp.70-73
104
standardisation. I would suggest that the clue to this emendation is to be found in
the career of one of Kourkouas' late eleventh-century descendants. In 1091-2, the
year when Alexios finally defeated the nomad Pechenegs at Levounion, Gregory
Kourkouas was doux of Philippoupolis, a crucial position in the defence system of
the Balkans.79 With such an important Balkan city in the hands of Kourkouas,
Skylitzes had a powerful motive for obscuring the fact that during that glorious
campaign which had first taken Byzantine frontiers back up to the Danube in 971,
one of Gregory's forebears had been drunk at his post.
This chapter has outlined some of the later eleventh-century literary, social and
political contexts within which John Skylitzes composed the 'Synopsis Historion'.
It has been argued that the text itself was compiled by an author, who worked
within the upper echelons of the imperial administration during the first two
decades of Komnenian rule, for an aristocratic audience which was also based in
the Constantinopolitan court. This hypothesis is supported not only by the fact that
Skylitzes held a senior position within the imperial government of Alexios
Komnenos, but also by textual evidence from within the 'Synopsis Historion' itself.
105
suggested that synopses and high-style histories were complementary texts rather
than polar opposites: they were both read and written by those who lived and
worked in the milieu of the early Komnenian Constantinopolitan court. Moreover,
the likelihood that Skylitzes was writing for an elite audience based in the imperial
capital is supported by several of the key diagnostics which emerge from his
his enthusiasm for the genealogies of key aristocratic families, above all those
dynasties which were of political significance towards the end of the eleventh
century.
In the second half of this chapter it has been suggested that Skylitzes' devotion to
the deeds and pedigrees of the leading aristocratic families of the Byzantine empire
was rooted in a very distinct political, as well as social and literary, context. This
political context was underpinned by the conditions of internal and external
insecurity which prevailed in the Byzantine empire during the first two decades of
the reign of Alexios Komnenos. Externally, the greatest danger came from nomad
and Norman attacks on the Balkans; internally, the greatest perils were posed by
those aristocratic families which rivalled the Komnenoi. Both these tensions shaped
Skylitzes' historical writing. On the one hand hostility between leading families at
the end of the eleventh century encouraged Skylitzes to interpret the Byzantine past
in terms of dynastic connections and competition. On the other hand, his interest in
aristocratic achievement in the Balkans may have been shaped by an imperially-
sanctioned desire that all the major aristocratic families should work together to
106
At various points in the chapter the ways in which these later eleventh-century
literary, social, and political contexts shaped Skylitzes' presentation of the reign of
Basil II have been highlighted, as have the distortions which these contexts can
impose. It has been argued, for example, that the prosopography of Basil's reign is
often conditioned by the 'Who's Who' of the Komnenian period. This later
eleventh-century filter can distort our understanding of the relative importance of
individuals and families within political society during the reign of Basil. For
example, it is possible that Skylitzes' reports on the entrance of several Armenian,
Georgian and Bulgarian "outsiders" to imperial service during the reign of Basil,
are determined more by the political significance of their later-eleventh century
descendants, than by their role within Byzantine political society in the later tenth
and early eleventh centuries. Likewise, some of the rebels involved in the first
Skleros revolt (976-9), may be retained within Skylitzes' narrative on the basis of
the political significance of their eleventh-century descendants, rather than their
own importance to the political history of the reign of Basil. At a more general
level, it has been suggested that Skylitzes' enthusiasm for interpreting the
Byzantine past through the achievements of the great aristocratic families can
obscure the role of the emperor. The marginalisation of the emperor as a principal
character in the narrative is particularly visible during Skylitzes' coverage of the
Balkan wars in the second half of the reign.
As a final point, it is worth considering the most important danger that Skylitzes'
aristocratic articulation of the Byzantine past, and his concomitant neglect of the
person of the emperor, presents to the historian of Basil's reign. This danger is that
Skylitzes' account can convey a false impression of the balance of power between
imperial authority and leading members of the political elite. That is to say, by
looking at the later tenth and early eleventh centuries through Skylitzes' eyes, it is
107
easy to overstate the power and prominence of aristocratic families and
underestimate the authority of the emperor. While such a balance of power in
favour of the aristocracy was the context within which the later eleventh-century
politics of Skylitzes' own lifetime were played out, the political situation in the
reign of Basil II may have been quite different. Indeed, as the rest of this thesis will
demonstrate, the reign of Basil was a period when imperial authority was far from
moribund.
108
Chapter Three
I. Introduction
The first two chapters of this thesis have constituted historiographical explorations
into the main Greek narrative of Basil's reign, the 'Synopsis Historion' of John
Skylitzes. Both discussions have been predicated on the premiss that Skylitzes'
how his text as a whole was put together at the end of the eleventh century. One of
the key reasons for adopting this broader historiographical focus is the lack of other
coverage of Basil's reign can be compared directly. However, there is one period
of Basil's reign where the problem of scarcity of evidence is less acute, and a
relatively substantial section of Skylitzes' account can be set against several other
detailed historical narratives. This period comprises the first thirteen years of the
reign (976-89), when the emperor Basil, and his brother and co-emperor
Constantine VTQ, were challenged for imperial power by the generals Bardas
Skylitzes' coverage of Basil, these revolts represent the dominant narrative in the
analyses of the reign presented by Leo the Deacon and Michael Psellos. 1 They also
languages other than Greek, including Stephen of Taron (Armenian), Yahya ibn
Sa'id (Arabic), and the author of the Georgian 'Life of John and Euthymios'. 2 Even
historians writing in Arabic in locations far away from Byzantium, whose interest
in the domestic history of the empire was rather more occasional, such as Ibn
Byzantine affairs during this period. 3 Moreover, the Arabic evidence pertaining to
the revolts is strengthened by the survival of a series of documents from Iraq, which
relate to the period between 980 and 987, when Bardas Skleros was held captive by
the Buyid authorities in Baghdad. These documents include letters, treaties and
One implication of this plethora of sources is that a more detailed narrative of this
period can be constructed than for any other phase of the reign. Moreover,
against other sources. It is, therefore, no coincidence that this is the only part of the
However, this relative abundance of sources not only allows for the reconstruction
of a coherent narrative, but also offers the unique opportunity to use the history of
2 Stephen of Taron, pp. 140-3, 187-9; Life of John and Euthymios, pp. 67-142; Georgian Royal
Annals, pp.373-4; Yahya, PO 23, pp.372-89, 398-402, 418-427; see above, pp.11-13
3 Ibn Miskawayh, pp.424-5, 436-9; Al-Rudhrawari, pp.6-7, 23-35, 115-119.
List of 'documents' under consideration: Ibn Shahram's embassy to Constantinople in 981/2. Al-
Rudhrawari, pp.23-34; Propaganda letter of Adud issued in the aftermath of the embassy of Ibn
Shahram contained in the letter collection of Al-Shirazi: J.C.Burgel, Die Hqfkorrespondenz Adud al-
Dawlas (Wiesbaden, 1965), pp. 155-6; Eyewitness account of the release of the Skleroi: Al-
Rudhrawari, pp. 116-17; The treaty of the release of the Skleroi 987 and a letter to Skleros from the
Buyid general Chutur in March 990 contained in a diplomatic treatise of the Egyptian Mamluk jurist
al-Kalkashandi: M.Canard, 'Deux documents arabes sur Bardas Skleros', Studi Bizantini e
Neoellenici 5 (1939), pp.55-69.
110
Skylitzes' narrative compares with those of other Greek and non-Greek accounts of
this period, we may learn more about the construction of the 'Synopsis Historion'
noted that the method adopted here does not constitute a chronological step-by-step
comparison of information and interpretation presented by Skylitzes and the other
historical accounts. Such an approach is likely to provide little insight into the
background contexts or historiographical traditions within which the individual
texts themselves were written. Instead, the chapter will compare the various ways
in which the main narratives respond to a series of more general questions. Why
were so many contemporary and later medieval historians so intrigued by these
rebellions? To what extent did they consciously interpret these revolts as pivotal to
the political history of Basil's reign? Which dimensions of the revolts attract their
attention? To what extent is their work conditioned by the accident of survival or
the constraints of literary genre? Where appropriate this discussion will be
illuminated with insights from the historiographical analysis presented in the first
two chapters of this thesis. Finally, although the discussion in this chapter is
characterised by a constant interweaving of history and historiography, it is
necessary that an initial outline of the main political events of this period be
established at the beginning. Thus, the chapter will be prefaced by a summary of
the internal and external history of the Byzantine empire in the century before Basil
came to the throne, and a brief synopsis of the revolts themselves.
The most detailed and well-rounded account of the revolts is to be found in Forsyth, 'The Chronicle
of Yahya ibn Sai'd', pp.370-462. See also N., Adontz, 'Tornik le moine', B 13 (1938), pp. 143-64;
111
II. Historical synopsis
When Basil II and his younger brother Constantine VIII became senior emperors in
976 they inherited an empire which was much changed since the foundation of the
However, with this greater expansion came the threat of greater instability within
the upper echelons of Byzantine political society. This threat was presented by the
very agent of territorial conquest: the army. As we shall see in the introduction to
112
chapter five, a larger, more centralised, professional, and mobile army had probably
generalship of John Kourkouas. It was then enlarged and refinanced during the
the army is to the state as the head is to the body; if it changes the whole must
change with it, and whoever does not carefully surpervise it, endangers his own
safety. 8
The political importance of control of the military to the stability of the Byzantine
state was first made manifest after the death of Constantine's son, the emperor
Romanos II, in 963. Romanos left two sons as heirs to the imperial throne, two sons
who would eventually assume power in 976 as the emperors Basil II and
Constantine VTH. However, in 963 these brothers were still minors, with their
mother Theophano acting as their regent. Their minority rule did not last for long.
Within a few months of the death of Romanos, the general Nikephoros Phokas,
marched at the head of the army to Constantinople. The boy emperors' great-uncle
Basil Lekapenos, the parakoimomenos, the chief eunuch in charge of the Great
Palace, allowed Phokas to enter the city. On the same day Phokas was crowned
emperor. 9 Six years later he was assassinated. His replacement as emperor was his
assassin, another leading army commander, John Tzimiskes.
113
In January 976 Tzimiskes himself died with the result that Basil and Constantine
finally assumed full rule as adult emperors. However, within six months they were
threatened by another revolt led by another senior general. In the spring or early
summer of 976, Bardas Skleros, the doux of Mesopotamia, the general in charge of
the army based east of the Anti Taurus mountains, declared himself emperor. His
initial revolt lasted for three years and was punctuated by several pitched battles
between imperial and rebel forces. Victory for the emperors was only achieved
after Bardas Phokas, the nephew of the emperor Nikephoros, was recalled from
internal exile to lead a Byzantine army reinforced with troops from the Georgian
princedom of Tao. Skleros was finally defeated by Phokas and this Georgian army
in March 979.
But the suppression of the rebels did not preface a period of peace and stability.
Skleros and his immediate retinue of about three-hundred men took refuge with, or
were imprisoned by, the Buyid ruler of Iraq, Adud al-Dawla, and remained a
potential threat off-stage in Baghdad. Skleros' residence in Arab territory was the
subject of intense diplomatic exchange between the Byzantine court and the Buyids
during the early 980s. It is clear from the diplomatic account of Ibn Shahram, a
Buyid envoy sent to Constantinople in 981/2, that Basil himself was willing to
contemplate ceding the empire's client state of Aleppo in northern Syria to the
Buyids of Baghdad if Adud surrendered Skleros. 10 However, this was a policy that
attracted considerable opprobrium both from his military chiefs such as Bardas
Phokas, and his great-uncle and chief advisor Basil the parakoimomenos.
114
These internal political tensions reached a denouement in 985. The first signs of
serious trouble came in the form of a rumour that Basil Lekapenos was about to
incite a palace coup against his great-nephew. In response the armies of the east
removing the emperor, the parakoimomenos found himself dislodged from power
grew even greater when Basil attempted to wrest the "foreign policy" initiative for
himself by attacking Bulgaria the following year. When Basil's foray across the
war ensued. Skleros was released by Baghdad in the winter of 987. He soon
relaunched his rebellion in the area around Melitene. Phokas announced his own
imperial candidature and seduced Skleros into a military alliance. He then reneged
on the terms of the alliance and imprisoned Skleros in the Anti Taurus. For two
years Phokas' forces threatened Constantinople from the Asian side of the
Bosphoros.
At some point in late 988 or early 989 Phokas' forces led by Kalokyros Delphinas
were defeated at Chrysopolis. 12 In April 989 Phokas himself was killed at the battle
This imperial army was heavily reinforced with Russian troops sent to the emperor
as part of a deal in which Basil's sister Anna was despatched as a bride for
115
Yet even after the defeat of Phokas, the spectre of rebellion endured. Bardas'
younger son Leo held out at Antioch until November 989. Meanwhile, Skleros was
released from prison by Phokas' widow, and with his own supporters and the
remnants of the Phokas party, launched his third insurrection. This revolt however
was of short duration and the Skleroi soon capitulated. Although the exact date of
Skleros' final surrender is unknown, it must have happened before 6 March 991,
the date when Bardas died. 13
For the problems associated with dating the battle of Chrysopolis see Forsyth, "The Chronicle of
Yahya ibn Sa'id', pp.439-40.
116
Asia Minor littoral and in the Sea of Marmara. Localised fighting was reported in
Antioch. 14 During the Phokas revolt military action was less widespread, but once
again Abydos came under siege. 15 Civil war also enfolded Arab, Armenian and
Iberian buffer states into alliances with imperial and rebel armies, and offered
encouragement to hostile powers beyond the empire, such as the Fatimids of Egypt
According to other historians, the collapse in the Byzantine position was even more
profound: not only had the offensive against the neighbours ceased, but the
117
neighbours had begun to fight back on all fronts. Both Yahya and Skylitzes report
that the prolonged period of Byzantine civil war allowed the armies of the
Bulgarian state, based in western Macedonia and led by the Kometopoulos family,
to raid deep into mainland Greece and the Peloponnese. 21 Stephen of Taron points
out that Bad ibn Dustuk, a Kurdish emir who controlled Chliat on Lake Van, took
advantage of the general mayhem in Caucasia to sack the town of Mus. <^7 But
perhaps it is Leo the Deacon who conveys most dramatically the incomprehensible
scale and suddenness of the reversal in Byzantine fortunes during the early years of
Basil's reign. As a member of the palace clergy and a writer of speeches at the court
of Basil n, Leo represents a contemporary, Constantinopolitan perception of the
disasters of the post-976 period. His summary of the Skleros and Phokas revolts,
the 986 failure in Bulgaria, and the humiliating intervention of the Rus, stand in
elegiac contrast to his more detailed accounts of the martial achievements of Basil's
imperial predecessors. 0^ His sombre reflections are mirrored by some of the poems
written by John Geometres in the last quarter of the tenth century which lament the
evil consequences of civil war, the shame of the arrival of Rus troops in
Constantinople, and the opportunities which internal weakness afforded to the
Byzantine empire's Bulgarian adversaries. 24
Yet, there are significant reasons to doubt whether these revolts are accorded so
much attention in the primary sources solely because medieval historians
20 Ibn Miskawayh, pp.326-9; Yahya ibn Sa'id, PO 23, pp.353-8; M.Canard, 'La date des expeditions
mesopotamiennes de Jean Tzimisces', Melanges Henri Gregoire, Ann. de I'lnst. de Phil, et d'Hist.
Or. etSlav. 10 (1950), pp.99-108. See below, pp.295-6
21 Yahya, PO 23, p.430; Skylitzes, pp.329-30, 339. For the rise of the Kometopouloi see Seibt,
'Kometopulen', pp.65-98.
22 Stephen of Taron, p. 141; see below p. 279
23 Leo the Deacon, pp. 169-78; Mango, Empire of the New Rome, p.211.
See for example the poems, V ryv aimarcuriv; eig ra$ r&v 'Ifrqputv apnayas; c/j TOUJ BouXyapovg; e/$
rov KoiuTomuhov (Geometres, iv, 271-3, 282-3; Skylitzes, pp.282-3). See also Schlumberger,
L 'Epopee byzantine, i, 641-6, 725; ii, 34, 43-4; Poppe, 'The Political Background to the Baptism of
118
recognised their fundamental importance to the political history of the reign. In the
first place it is always possible that the physical processes involved in the
transmission of texts may, over time, either enhance the supply of information
the Rus', pp.214-7; Sevcenko, I., 'Poems on the Deaths of Leo VI and Constantine VII in the Madrid
Manuscript of Skylitzes', DOP 23-4 (1969-70), p. 189
The Life of John and Euthymios, pp.89-92; Adontz, Tornik le moine', pp. 143-64;
P.M.Tarchnishvili, 'Le soulevement de Bardas Skleros', BK 17-18 (1964), pp.95-7. Imperial
patronage for the Iberians on Athos in the aftermath of the Skleros revolt can be corroborated by
various Georgian manuscript colophons and by documents from the archives of the Iviron monastery
itself: Actes d'lviron I. Des origines au milieu du Xle siecle, Archives de 1'Athos XTV, eds. J.Lefort,
N.Oikonomides, D.Papachryssanthou, H.Metreveli (Paris, 1985), pp.7-31, 117ff.
119
historiographical materials. However, the description provided in this text is
nothing more than a paraphrase of the historical background contained in the Life
of John and Euthymios which was interpolated into the Annals during the
eighteenth century.26
In addition part of the significant Arabic contribution to the history of the revolts,
which at first sight appears to be based on a solid core of documentary evidence,
may also be something of an illusion. For while Skleros1 flight to Iraq and seven-
year detention in Baghdad can be traced in several pieces of first-hand evidence,
now contained in a variety of Arabic literary sources, the survival of this evidence
is probably determined as much by the dictates of genre, as by its significance for
the domestic politics of Byzantium and relations between Baghdad and
Constantinople.27 For example, the treaty containing the terms of the release of the
Skleroi from Baghdad in 987 is to be found out of all historical context in the later
fourteenth- or fifteenth-century diplomatic manual of the Egyptian Mamluk
*")Q
120
surrounding the release of the Skleroi five years later.29 Both these records could be
the foreign, and the exotic. This enthusiasm had its roots in a genre of Arabic
literature known as adab. Taking material from a variety of other literary genres,
including poetry, oratory and grammar, adab was primarily concerned with
rendering its reader more courteous, urbane, and erudite. As Arab contact with non-
Arab peoples grew, so had adab come to be informed by the written traditions of
interest in the cultures which fostered these exotic literary forms. 30 Marius Canard,
the great Byzantino-Islamic scholar of the pre- and post-Second World War
literature were conditioned by the traditions of adab 31 Certainly, in the case of al-
fascination with the exotic as it applied to Byzantium. This evidence comes in the
explicit reference to the fact that the Skleroi spoke Greek during the ceremonial
As Ward (ie Bardas [Skleros]) approached he bowed his head slightly and
kissed the prince's hand. A chair with a cushion was placed for him and he sat
down thereon. Samsam al-daulah proceeded to make civil inquiries, and he
[Ward] invoked a blessing on the prince and thanked him in Greek, the
conversation being conducted through an interpreter." 32
121
//. Pro-Skleros interest in the Greek historiography
demonstrate why this is so, we must start by noting the degree to which these Greek
historians concentrate on the rebellion of Bardas Skleros at the expense of that of
Bardas Phokas. Here three principal diagnostic elements are of great significance:
33 The most detailed account of the key battles fought in the extreme west of Asia Minor between
987 and 989 is to be found outside the Greek tradition in the history of Stephen of Taron (Stephen of
Taron, pp. 188-9; Yahya ibn Sa'id's account contains the crucial information that the final battle of
Abydos was fought on 13 April 989 (Yahya, PO 23, pp.419-31). However, Yahya's more detailed
coverage of the revolt is usually concerned with events on the eastern frontier, especially in Antioch
and Trebizond.
122
information about personnel, place-names, and titles, (highlighted in bold font in
the text produced below), as well as a reasonably coherent picture of a key battle.
On the other hand, Skylitzes1 account of the outbreak of the Phokas revolt as it
involved Phokas amounts to no more than two very short passages (also produced
below). Not only are these passages short, they convey exceptionally little detail
123
went over with the whole fleet to Michael Kourtikios, sent by Skleros to be
strategos of the Kibyrrhaiotai. 34
But the greatest men of the Romans were very angry with the
emperor............they gathered together in Charsianon in the house of
Maleinos, on the 15th August, of the 15th indiction, and proclaimed Bard as
Phokas emperor, having set the diadem around him and the rest of the
recognisable regalia of empire....................
He [Skleros] was uncertain and changeable in his thoughts. For on the one hand
he judged that he was too weak to continue and hold fast to the revolt on his
own. But he thought it ignoble and unmanly to move over to Phokas or to the
emperor. And so having discussed at length with his supporters, in the end he
calculated that for him to be hailed as emperor on his own was reckless and
Skylitzes, pp.318-20. See Map 1 for all geographical references in this passage
35 Skylitzes, pp.332, 336
124
unprofitable, because it was impossible, and yet he hated the idea of going over
to one of the dynasts and rejecting the other, because of the uncertainty of the
future. And so he decided that as far as was possible, he would win over both
powers, in order that in any unfortunate eventuality he should have the help and
support of at least one of them. And so he himself sent letters to Phokas asking
for a joint plan of action and for a partition of the empire, if they were able to
conquer the emperor. But with very clever judgement and calculation, he
secretly sent his son Romanos to the emperor, as though he [the son] were a
deserter, so that if Phokas prevailed, he himself [Skleros] would be the saviour
of his son, and if the emperor proved the stronger, he himself [Skleros] would
be saved from danger, as the beneficiary of the intercession of that man
[Romanos]. And Romanos, having assumed the appearance of flight, went to
the emperor. 36
3) The third diagnostic element in a tradition which seems to emphasise the actions
and views of Skleros rather than Phokas is the frequency with which the Greek
coverage draws attention to the cunning and acumen of Skleros. Two examples
occur in passages already cited above. In his analysis of Skleros1 first march across
Asia Minor, Skylitzes describes how Skleros cleverly used deception to gain an
advantage at the battle of Lapara. In the passage which expatiates on Skleros'
vacillations in 987, the general's decision to send his son Romanos to the emperor
is explicitly interpreted by Skylitzes as a sign of shrewdness. It is noteworthy that
Michael Psellos, in his appraisal of the 976-89 revolt period of Basil's reign, also
picks up on this general theme of Skleros' cunning and strategic awareness. He
favourably compares Skleros' abilities as a commander with those of Phokas:
Psellos expatiates on this resourcefulness in greater detail, describing the skill with
which Skleros deprived both Constantinople and the imperial field armies of
36 Skylitzes, pp.334-5
"Psellos, p. 14
125
essential supplies. For example, he tells his reader that Skleros diverted naval
convoys, blocked roads, and siphoned off merchandise for his own troops. 38
Phokas, for example, was almost certainly able to mobilise much larger private
resources in support of his own insurrection. By the middle of the tenth century, it
is likely that the Phokas family owned considerable estates in Cappadocia in central
Anatolia. 39 While the exact location and scale of these Cappadocian estates remains
unknown, an index of Phokas family wealth in the locality is provided by the
luxurious gold and lapis lazuli decoration of the Cappadocian rock church, Tokah
Kilise, which has been identified as a Phokas foundation. It is usually argued that
126
the nearby church of the Great Pigeon House at Qavuin, which includes a fresco
depicting the emperor Nikephoros Phokas and his brother the kouropalates Leo,
was excavated and decorated by local adherents of the Phokas family.40
Considerable private Phokas resources and local support may also explain why
Nikephoros Phokas, the eldest son of Bardas, was able to lead a rebellion based in
central Anatolia against the emperor as late as 1021/2, despite not having held
office since his father's defeat in 989.41
In contrast, Skleros seems to have relied more upon those resources at his
immediate disposal in his capacity as a military commander. For example, at the
beginning of his first revolt, he established his campaign headquarters at Charpete
(also known as Harput in Armenian, and Hisn Ziyad in Arabic), a strong point in
the Anzitene, the region east of the Anti Taurus which he controlled in his capacity
as doux of Mesopotamia.42 Next, he sequestered the fiscal revenues of nearby
Melitene. Finally, he secured additional troops from outside the empire by
contracting military alliances with the regional powers which neighboured
Byzantine Mesopotomia. The Armenian prince of Mokh, a region south of Lake
127
Van, fought with Skleros forces; the Hamdanid emir of Mosul, Abu Taghlib,
This is not to argue, however, that the Skleros alliance did not possess considerable
resources or present considerable danger to the emperor. For example, in the eyes
of contemporary Arab witnesses, Skleros' principal ally, Abu Taghlib, was
exceptionally wealthy. The traveller and geographer Ibn Hawkal makes several
43 For Skleros' Armenian support see: Stephen of Taron, pp. 140-1; Skylitzes, pp.320-1. Skylitzes
also mentions that the Armenians of the Byzantine army were the first to hail Skleros as emperor
(Skylitzes, pp.315-6). For relations with Abu Taghlib see Skylitzes, pp.315-6; Yahya, PO 23,
p.398ff.; and Ibn Miskawayh, pp.424-6. Seibt, Die Skleroi, p.38, and Forsyth, 'The Chronicle of
Yahya', p.377, both refer to the marriage alliance between Abu Taghlib and Skleros, but do not
discuss it in detail. This lack of interest on the part of modern historians is surprising since such an
alliance must have involved the marriage of a Christian to a Muslim.
44 Ibn Hawkal: La Configuration de la terre, trans. J.H.Kramers and G.Wiet (Beirut/Paris, 1964),
p.205; Al-Tanukhi: Table Talk of a Mesopotamian Judge, trans. D.Margoliouth (London, 1922),
pp. 195-201; Ibn Miskawayh, pp.312, 415, 421-5, 431-4; al-Rudhrawari, pp.26-7. Ibn Miskawayh and
Yahya make indirect references to the legendary wealth of Abu Taghlib in their reports about the
expulsion of the Hamdanids from Mosul by Adud al-Dawla in 978/9. They allege that Adud was
afraid that Abu Taghlib would use rumours about his immense fortune to persuade the Fatimids of
Egypt to help him regain control of northern Syria and the Djazira from the Buyids (Ibn Miskawayh,
p.434; Yahya, PO 23, p.402). See below, p. 131, for expulsion of Abu Taghlib from Mosul by Adud.
128
part of the thematic navy, and besieged the key point of Abydos by land and by sea,
Nevertheless, as the revolt progressed, it became clear that Skleros' only real hope
of success had been a swift knockout punch. The longer the campaigns went on, the
more apparent became the structural weaknesses in Skleros1 challenge to imperial
authority. Fundamental to Skleros1 ultimate lack of success was that he was never
able to present a permanent threat to Constantinople. Whenever he approached the
capital, a new imperial field army always emerged to drive him back. For example,
having defeated imperial forces at the battle of Lapara in the eastern theme of the
Lykandos at the beginning of his revolt, Skleros and his army marched west and
camped at Dipomaton on the western reaches of the Anatolian plateau, near the
Lake of the Forty Martyrs, modern Aksehir Golii. In response to the arrival of
Skleros in the west, an imperial army under the command of Peter the
Stratopedarch and Leo the protovestiarios, set out from Kotyaion, marched past the
rebel camp at night, and so drew Skleros1 forces south and east towards Ikonion and
away from Constantinople. Although Skleros defeated his enemies in the
subsequent pitched battle at Rhageai, his victory was achieved at the price of
turning eastwards.46 Less than a year later, Skleros was back in the west, now
geographically much closer to Constantinople: not only was Abydos under siege,
but Nikaia had also been taken. Yet, he was still forced back onto the plateau, when
imperial armies crossed the Bosphoros under the command of Bardas Phokas and
45 Skylitzes, pp.315-27; Yahya, PO 23, pp.372-89, 398-402; J.T.Teall, 'The Grain Supply of the
Byzantine Empire', DOP 13 (1959), pp.104, 119, on the importance of Abydos in the passage of
basic foodstuffs into Constantinople from the eighth to the tenth century. See below, p. 196
Skylitzes, pp.319-32; Skleros' itineraries across Asia Minor during his first revolt can be traced in
several of the volumes produced by the Tabula Imperil Byzantini project: F.Hild and M.Restle, TIB,
Kappadokien (Verlag der Osterreichischen Akademie der Wissenschaften) (Vienna, 1981), pp.93-4;
K.Belke and M.Restle, TIB, Galatien undLykaonien (Vienna, 1984), p.71; K.Belke and N.Mersich,
TIB Phrygien undPisidien (Vienna, 1990), pp. 96-7. For place names, see Map 1; see below, chapter
four, pp. 160-3, for further discussion of these routes.
129
began marching towards Kaisareia. Once again Skleros was victorious in the field
battles which ensued; but once again these victories were achieved at the price of
moving eastwards. For while Skleros defeated Phokas at the battle of Pankaleia
near Amorion in the theme of the Anatolikon in June 978, he was unable to prevent
Phokas from regrouping imperial forces, and continuing his march away from
Constantinople. When Skleros won a second engagement later in 978, it was far to
the east at Basilika Therma in the theme of the Charsianon, and despite his military
success, Skleros was forced to spend the winter in this area of Anatolia. Thus, even
though Skleros had remained undefeated during 978, his threat to the capital had
been thoroughly dissipated by Phokas1 march east. Moreover, during the winter of
978, Skleros' position worsened considerably. By March 979 the imperial forces
led by Phokas had been reinforced by the Iberian army dispatched by David of Tao.
On 24 March 979 this joint Ibero-Byzantine army defeated Skleros at Sarvenisni
(Aquae Saravenae) in Charsianon. Skleros and his immediate retinue fled across
the Diyar Bakr to Mayafariqin, where they were taken into custody by the Buyid
forces of Adud ad-Dawla. 47
Meanwhile, even though his first revolt had lasted more than two years, Skleros'
power-base had always been prey to the uncertain loyalty of many of his supporters.
His rebellion had, from the very beginning, been afflicted by defections. Before
Skleros had even crossed the Anti-Taurus at the start of his revolt, he was forced to
have his chief of staff (hetaireiarches) murdered to prevent him from swapping
sides. As early as 977 the Hagiozacharites brothers defected to the protovestiarios
Leo; when they were captured by Skleros at the battle of Rhageai they were blinded
as a punishment for their disloyalty. An even more significant defector was the
47 Skylitzes, pp.323-5, Yahya pp.375, 399, Stephen of Taron, pp. 142-3; Hild and Restle,
Kappadokien, pp.94, 143-4, 156-7.
130
AQ
diplomatic alliance in the east. Severe stresses began to appear as soon as Abu
Taghlib was attacked by the Buyid armies of Adud al-Dawla, and was forced to flee
from Mosul in 978. At first Abu Taghlib took refuge at Hisn Ziyad, Skleros1 main
command centre in the east, and requested military aid from his Greek ally. Skleros
responded by telling Abu Taghlib that more Hamdanid troops were needed to
secure a victory against Bardas Phokas before he could come east to fight the
Buyids. Although Abu Taghlib despatched some fresh troops, he himself left Hisn
Ziyad and fled back to the Djazira, and thence to Syria the moment he heard that
Skleros had been defeated by Phokas and the Iberians of Tao in March 979.49
In contrast, fragility of support and limited resources do not seem to have been
typical of the 987-9 Phokas revolt. While information about the events of the
Phokas rebellion is much less readily available, rebel forces do appear to have
presented a much more consistent danger to imperial authority in Constantinople.
The very fact that the key battles between rebel and imperial armies were fought at
Chrysopolis and Abydos, in the extreme west of Asia Minor close to
Constantinople itself, suggests that it had been impossible for Basil and his
ministers to muster the forces from within the empire necessary to drive Phokas
back onto the Anatolian plateau. Secondly, the fate of the Phokas party in the
aftermath of their defeat points to Phokas as the greater threat. After Phokas was
48 Skylitzes, pp.318, 322; Yahya, PO 23, pp.376-7. See below, pp.332-3, for an outline of the career
of the basilikos Obeidallah.
49 Yahya, PO 23, pp.397-402; Ibn Miskawayh, pp.420-443. The chronology of Abu Taghlib's
wanderings on the Djazira frontier is confused since the accounts of Yahya and Ibn Miskawayh which
deal with his expulsion from Mosul and his flight into exile do not coincide exactly (Forsyth, 'The
Chronicle of Yahya', pp.315-6). However, it is clear that Abu Taghlib's fortunes were on the wane
from the moment that Adud al-Dawla took control of Baghdad at the end of May 978. Buyid troops
loyal to Adud were besieging Mosul by the end of June 978 (Ibn Miskawayh, pp.420-1)
131
killed at Abydos in 989 his head was conveyed around the empire as a warning to
other rebels. A few months earlier after the defeat at Chrysopolis, Kalokyros
Delphinas, one of Phokas' generals had been impaled.50 In contrast, when the
Skleroi finally surrendered in 989, they were allowed to retain their offices and
titles.51 Later in the reign they can be observed in top ranking military posts. For
example, Romanos Skleros, the son of Bardas, led a Byzantine army against
identified by Yahya ibn Sa'id as a Byzantine army commander in the Antioch area
The likelihood that the accounts of Skylitzes and Psellos artificially enhance the
role of Bardas Skleros at the expense of Bardas Phokas in the history of the first
thirteen years of Basil's reign, has important implications for our understanding of
the source material used in the medieval Greek historical record. As we saw in the
first chapter of this thesis, many modern scholars believe that a lost Phokas family
chronicle was utilised by several later Greek historians, including John Skylitzes, in
their coverage of the second half of the tenth century. 54 However, the minimalist
treatment of Phokas and the expansive coverage of Skleros in both Skylitzes' and
132
Psellos' appraisals of the revolt period of 976-89, must cast some doubt on whether
a pro-Phokas source was widely utilised to report on the reign of Basil II.
Suspicions that a non-Phokas source is at work in the Greek tradition surface even
on those occasions when Phokas is presented with greatest favour. The most
This man [Phokas] on seeing him and recognising who he was, went forth to
meet him and struck him on the head with his mace, and he [Gauras]
immediately fell down lifeless from his horse with the unstoppable force of
the blow. 55
However, the most striking element of Phokas-related passages such as these is that
their content is usually very brief, vague, and repetitive. In the course of the final,
and decisive, battle at the end of the first Skleros revolt in 979, Skylitzes describes
a duel fought between Phokas and Skleros in very similar terms to that between
Phokas and Gauras:
Then Phokas having seen his own army giving ground little by little and on the
point of flight, and judging that a renowned death was better than an ignoble
and reproached life, broke through the enemy phalanxes and attacked Skleros
with all his might....................Skleros cut off the right ear of the horse of
Phokas and the bridle with a blow from his sword. And Phokas, having beaten
ov OUTOJ idon/ KCU ocrrig en) Karavoycra$, -ryje/AO, rov Irrrnov TrapeveyKcav KOI unavriacras naiei
Kara rijg KopuQog. KOI o ftev heiiroQufMfaras TJJ avu7TO0TaT<t> <bop. rijs TrXtfffa -n'mrei napauriKa rov IVnrou
.....(Skylitzes, p.325)
133
him on the head with his mace, threw him down over the neck of his horse
with the weight of the blow. 56
Likewise, when Skylitzes describes Phokas' final battle against Basil II at Abydos in
989, he explains both the general's motivation for fighting, and his actual military
action, in terms which are identical to his account of Phokas's earlier engagement
with Skleros in 979.
...., having nobly selected an honourable death rather than an ignoble life,
when Phokas saw the emperor from a distance..... he reasoned with himself that
if he were to triumph over this man (ie Basil) he would easily vanquish the rest;
having spurred on his horse, he charged violently against him, and slicing the
enemy phalanxes in two, to everyone he appeared unstoppable. 51
Indeed, it seems likely that rather than emanating from a family chronicle, scenes
such as these were constructed from phrases out of Skylitzes' bland, generalising,
and homogenising military lexicon, which was discussed in the first chapter of this
CQ
thesis. For example, it is not surprising to note that two of these excerpts depict
either Phokas or his military initiatives through the use of one of Skylitzes1
favourite adjectives: avi/mwrraTos (unstoppable).
Yet, if the case for pro-Phokas material in the Greek record of Basil's reign is
proven to be rather weak, the overwhelming interest in the activities and attitudes
of his rival, Bardas Skleros, points to the possibility that a pro-Skleros source may
evravQa rov eaurou Xaov Beatro^evog o 3>u)Kag*\ Kara,* fjuKpov evb&ovra KCU npog (fruyqv fikenovra, fieXrtov
r i
' *\ P *~< r
*?
eivcu Kpivag rov
i
oavarov TTJJ
"\ ^ /i /
evjcAei} ayewovg KOI enoveioiorov
%'
^&/i)> T<* T|WJ/ evavriotv
t<^
owyKowag
<baXayyag npog avrov /Aera O^O^OTTJTOJ lerai rov SKAijpoj/. ........... KCU o tiev 2KAijpoj rov rnnov rov
TO fafyov ovg o~w TQ> xaXivty naicra$ artoKoirrei T^> r<^ i. o 5e $ta>tca rjj Kopvvj) rrara^a^ avrov
Kara rife c<^aA% rovrov /*ev erri rov rpaypfivov rov rrmov pmrcf rq> fiapei -rqj
Karcvexflevra. .... . ..(Skylitzes, p.326)
.... o Q>ioK<L$ rov tyjv ayewtog TO -ycwcwwg airofaveTv evyev&s npoKpivag, rov (ZcunXea
mpptaSev. . . .KOI npog eavrov JuryicraiJbevog, wg, ci rovrov entrvjchs 'yevyrai, pfrov av KOI rovg Aorrrouj
Kara/ycwierrai, rov Irmmv ^vomiort^ paydaia}; iKtro Ka.r'avrov, rag evavriag ^aXayyag diaKomrutv KOU
avvmararog ro?g TTOO~I (fratvopevog (Skylitzes, p.337)
See above, chapter one, pp. 7 1-2
134
underpin the testimonies of both Skylitzes and Psellos. 59 Indeed, the most
provided, paradoxically, by those other texts which focus upon Skleros to the
greatest extent, that is to say various narratives and documentary records from
Buyid Iraq. For example, a comparison between the Buyid and Byzantine accounts
of Skleros' departure from Baghdad in 987 demonstrates the extent to which
An alternative story emerges from the Buyid evidence. Both the eyewitness
accounts of the release of the Skleroi in the history of al-Rudhrawari, and the treaty
in the diplomatic handbook of al-Kalkashandi which details the terms of their
departure from Baghdad, indicate that the former rebels were set free under the
cq
The only historian who has ever raised this possibility is Roueche, 'Byzantine Writers and
Readers', p. 127.
60 Skylitzes, p.333
61 Ibid., pp.332-4; Psellos, pp.7-8
135
strictest of conditions, and provided by the Buyids with the necessary troops and
s~*\
supplies to enable them to reach the empire. Skleros promised, in the event of his
becoming emperor, to set free all Muslim prisoners of war, to protect the property
and families of Muslims, to prohibit any Byzantine army from marching on the
eastern frontier, and to hand over seven border fortresses to the control of the
was prepared to bankroll Skleros' enterprises once the rebel was back in
Byzantium.64
62 Al-Rudhrawari, pp. 116-7; Canard, 'Deux documents', pp.59-62, 65-8. Yahya also reports that the
Skleroi returned to Byzantium under the escort of Buyid troops (Yahyan, PO 23, pp.419-20)
63 Canard, 'Deux documents', pp.59-62, locates these seven fortresses in the extreme north of the
Diyar Bakr in the area of the Upper Tigris. Canard argues that they had fallen into Byzantine hands
during Tzimiskes' campaigns on the Mesopotamian frontier in 972 (see above, p. 117, and below,
pp.295-6). They were at the centre of negotiations between the Buyids and Byzantines at the time of
Ibn Shahram's embassy in 981-2 (see above, p. 114); see also Forsyth, 'The Chronicle of Yahya ibn
Sa'id', pp.402-409, 426.
64 Canard, 'Deux documents', pp.63-4, 68-9
136
There is persuasive evidence for the first interpretation. Like all rebels, Skleros had
always sought to manipulate public relations. During his first revolt he let it be
known that his spiritual guide, a certain monk, had foreseen Bardas' elevation to
the imperial purple in a dream.66 Both Bardas Skleros and Bardas Phokas are said
to have taken great comfort in the popular contemporary prophecy that the name of
the future emperor would begin with the letter beta 61 However, by the final stages
of his revolt in 989, Skleros had rather different propaganda needs. Instead of being
engaged in a quest for the imperial purple, he was involved in a game of diplomatic
bluff to ensure his own survival. Indeed, since his return to the empire in Baghdad
in 987 Skleros had been struggling to keep afloat. Despite being escorted back to
the empire by Bedouin and Kurdish horsemen and seizing the revenues from his
erstwhile base at Melitene, he had been unable to build the "auld alliances" which
had underpinned his first revolt a decade earlier. The Bedouin and Kurdish troops
fro
had soon left his service, and in his vulnerability he had joined Phokas. On his
release from captivity by the Phokas family in March/April 989, Skleros' support
base was fragile and his scope for military action limited. His correspondence with
the Buyid general Chutur indicates that as late as June 989, at least three months
after his release from imprisonment by the Phokades in the Anti Taurus mountains,
Skleros was still to be found in the plain of Larissa in eastern Cappadocia.69 There
is no evidence that he ever advanced any further west. Instead, he seems to have set
up a laager in eastern Anatolia from which he negotiated his surrender with the
emperor.
See above, p. 116 and n. 13 for Yahya's dating of the death of Skleros
66 Skylitzes, pp.316-7, Cheynet, Pouvoir et Contestations, pp. 161-2, points out the need for rebels
against imperial authority in the tenth to twelfth centuries to disseminate effective propaganda.
Skylitzes, p.338, interpolation in manuscripts U and E.
68 Yahya, PO 23, pp.421-3; Stephen of Taron, pp. 187-8; Skylitzes, pp.334-6
Canard, 'Deux documents', pp.63, 68-69.
137
Central to Skleros' strategy was a need to persuade the emperor that despite his
lack of resources he remained a potent foe. Delaying in Larissa, safely distant from
the hinterland of the capital was one element to this strategy. Allowing the emperor
to believe that preparations were underway for a resumption of revolt was another
strand.70 Yet another might have been the production of a narrative which
contained examples of the heroism of the Skleroi, above all extensive coverage of
their fabulously exotic military exploits in the distant regions of Buyid Iraq.
Certainly there is independent evidence to suggest that Basil feared Skleros'
qualities as a general, especially his ability to raise light cavalry troops from
neighbouring powers. Although Skleros' return to Byzantium in 987 proved to be a
false dawn, Yahya ibn Sa'id reports that the emperor was terrified of the arrival of
the erstwhile usurper and his nomad escort. So great indeed was his concern that he
reappointed Bardas Phokas to the position of domestikos of the scholai,
commander-in-chief of the army, only to see Phokas himself organise a much more
potent rebellion.71 Furthermore, Skleros' continuing correspondence with the Buyid
authorities, after his release from captivity by the Phokades in 989, might have
suggested that this long-term dissident still had the contacts necessary to build a
future eastern alliance.
Yet, although it is possible that the pro-Skleros encomium was composed before
the rebel and emperor came to terms, there is more convincing evidence that it was
written after the Skleros surrender. That evidence comes from the fact that that
some of the most overtly pro-Skleros material in the Greek tradition occurs in that
It is interesting that in referring to the renewal of Skleros' revolt in 989, Skylitzes uses a verb
meaning 'to get into training for war': o SicAipo mikiv avehavfiavw eaurov KO.I njv nporepav
airoo-nuriav (Skylitzes, p. 3 3 8).
71 Yahya, PO 23, pp.42 1-3
138
part of Michael Psellos' account which describes the conference between Basil and
Bardas, that took place after the general had capitulated to the imperial authorities.
Early on in his description of the eventual meeting between Basil and Skleros,
Psellos contrasts the experience and dignity of the rebel favourably with the
juvenile irascibility of the emperor. He vividly describes Skleros approaching the
emperor still wearing the red shoes symbolic of imperial rule. According to Psellos
Basil, "..... saw all this from a distance and shut his eyes in annoyance, refusing to
see him at all until he first clothed himself like an ordinary citizen."72 More
significantly, however, this is the occasion at which Skleros' ingenuity is presented
v
by Psellos as the quality which allows him to transcend the ignominity of defeat.
Despite failing to take the throne himself, the battle-hardened general is still able to
call upon his idiosyncratic cunning, and thus give the emperor some Machiavellian
advice on how to prevent future disorder:
But why should such encomiastic material have been necessary after Skleros'
surrender? One of the reasons may have been that Skleros needed to justify his
political career to the Byzantine political elite at large, to members of his own
family, and even to himself. Amid the complexities and confusions presented by
the Skleros narrative in the Greek record, it is easy to forget the enormous cost of
Bardas' revolt to his immediate supporters. By refusing to surrender to Basil at the
72 Psellos, p. 17
139
end of his first revolt in 979 with the rest of the rebel party, Skleros had led his
immediate family and entourage into a ten-year political cul-de-sac. This impasse
confinement by Phokas, and more important still, many years without the benefits
that service within the upper echelons of the Byzantine military could bring.
before the emperor." In such unpropitious circumstances it was imperative for the
future of the Skleros family that the imperial authorities should be reminded of the
military daring and expertise of the Skleros family. More important, it was vital
that these qualities were recognised as indispensable to the future well-being of the
Byzantine state.
Ibid, (translation adapted: E.R.A.Sewter, Michael Psellus: Fourteen Byzantine Rulers (London,
1966), p.43). This story is also briefly alluded to by Kekaumenos, although Skleros is not actually
named in his account (Kekaumenos, p.284)
74 Skylitzes, pp.338-9; Psellos, p. 17
140
However, while the appearance of other members of the Skleros family at the head
of Byzantine armies during the subsequent decades of Basil's reign proved that
military expertise was the route back to political fortune, Bardas Skleros and his
brother Constantine did not live to see this revival. 75 Having been resettled in
Thrace by the emperor, where it is possible that they were kept under house arrest,
both died before the end of March 991, in circumstances that at least one later
medieval Arab historian found suspicious.76 Given the brevity of the time that
intervened between the surrender and death of Skleros, it is likely that the
encomium preserved in the accounts of Skylitzes and Psellos was written during the
latter part of 990.
IV. Conclusion: the pro-Skleros source and the writing of history at the end of
Having identified traces of a pro-Skleros source behind the Greek historical record
of the first thirteen years of Basil's reign, the most pressing question must be, why
did historians such as Skylitzes and Psellos choose to articulate their appraisals
around this source? One reason for their choice could of course have been the
accident of survival: when they came to write their respective accounts in the mid-
to-late eleventh century this was all they had available. But this is not the only
possibility. Skylitzes himself says that an alternative account about the activities of
the Skleroi in Baghdad existed: that they fought for the Buyids, were well-treated,
and that as Adud al-Dawla lay dying he ordered his son to grant the Romans their
75 See above, p. 132 and below, p.310-11 for the future careers of Bardas' son Romanes and the rest
of the Skleros entourage.
Al-Rudhrawari notes that some reports alleged that Bardas Skleros died of poison (al-Rudhrawari,
p. 119)
141
freedom.77 In addition, an alternative Byzantine narrative also seems to have
Inserted into his text just before mention of the Ibn Shahram embassy of 981/2 is a
summary account of domestic Byzantine history from the death of Romanos II in
yo
963 to the arrival of Skleros in Baghdad. This account seems to have been taken
from a Greek original since it displays an accurate rendering of Byzantine names
and titles, which is unusual in most Arab historiography.79 More striking, however,
is the fact that this narrative does not portray Skleros the astute military hero, but
rather Skleros, the leader of a motley band, limping into Baghdad, having been
abandoned by most of his supporters:
Before the arrest of Ward his chief followers assembled in his presence, and
told him that they saw no likelihood of their negotiations with Adud ending in
the latter's furnishing effective aid. They said....... our right course is to return to
Byzantine territory peacefully if we can, or if we must fight then doing our
utmost with the prospect of winning, or else leaving this world with honour
unimpaired. He replied that this proposal was worthless, that he had a high
opinion of Adud...........when he resisted their proposals ..............many of them
O/"\
abandoned him.
However, while the contemporary source that made its way into the Iraqi historical
tradition may have been a more accurate account of events at the end of the tenth
century, it is possible that Psellos and Skylitzes selected their underlying source
materials with criteria other than precision and truthfulness in mind. In the previous
two chapters it has been suggested that Skylitzes' 'Synopsis Historion' was
compiled in accordance with many of the requirements and expectations of history-
77 Skylitzes, p.334
Al-Rudhrawari, pp.4-7
For typical instances of inaccuracy in such matters see M.Canard, 'Les sources arabes de 1'histoire
byzantine aux confins des Xe et Xle siecles', REB 19 (1961), pp.296-7, n.7; idem., 'Quelques noms
de personnages byzantins dans une piece du poete Abu Firas', B 11 (1936), p.452, n.5; cf. Miquel,
La Geographic humaine du monde musulman, ii, 387-91 comments on the habitual inaccuracy
displayed by Arab geographers in describing Byzantium.
142
writing of the latter stages of the eleventh-century. In concluding this chapter I
would like to suggest that the use of the pro-Skleros source in Skylitzes' and
Psellos' accounts, both of which were written in the second half of the eleventh
century, may be explained more by the political and literary demands of the period
in which these historians themselves were writing, rather than by the source's
accurate representation of the political history of the first half of the reign of Basil
II.
At the most basic level of levels we should note the extent to which the pro-Skleros
source meets the general requirements of history-writing in the second half of the
eleventh century. In the second chapter of this thesis, a widespread enthusiasm
among both authors and audiences in the later eleventh century for lively narratives
which contained large doses of vivid military heroism was noted. The long and
detailed military accounts within the Skleros source, in particular the narrative
describing the general's heroic flight from Iraq in 987, certainly meet this
o1
fascination. Moreover, the ubiquity of praise for Skleros' ingenuity and cunning
within this source may also have satisfied a contemporary interest in military
acumen and clever strategies, an enthusiasm which Alexander Kazdan has
identified as characteristic of much eleventh- and twelfth-century historical
literature.82
If we look more specifically at Skylitzes, then it is clear that here too a pro-Skleros
source would fit with his later eleventh-century working methods and use of source
materials. As we saw in the first chapter of this thesis, Jonathan Shepard has argued
XO
Al-Rudhrawari, pp. 7-8
See above, chapter two, pp.82, 88
Kazdan, 'Aristocracy and the Imperial Ideal', passim, idem., Change in Byzantine Culture, pp.99-
118, and especially pp.104, 112, idem., 'Social Views of Michael Attaleiates', pp.63-73
143
that eulogistic narratives of senior generals were among Skylitzes' preferred source
materials in the writing of history. Shepard concluded from his assessment of
Skylitzes' coverage of the mid-eleventh century, that two of his key sources were a
laudatory account of the general Katakalon Kekaumenos, and an apologetic text
0-5
produced to justify the revolt of George Maniakes (1043). If Skylitzes was
prepared to use encomia produced by senior generals as the main vehicle for his
articulation of the history of the empire in the mid eleventh-century, then there is
no reason why he should not have used similar source material in his account of
Basil's reign. In this context it is worth pointing out that the ways in which the
panegyrical accounts of the generals Kekaumenos and Maniakes shape Skylitzes'
eleventh-century testimony closely resemble the impact of the Skleros narrative on
Skylitzes' coverage of the later tenth-century. Thus, narrative passages connected to
the exploits of Kekaumenos and Maniakes tend to be long and detailed. The careers
of these generals are minutely explained. Their tactical cleverness and strategic
awareness are frequently stressed. It is their viewpoint which is most
overwhelmingly audible. All these are diagnostic traits which have been detected in
those parts of Skylitzes' coverage of Basil II which appear to be taken from the
Skleros source.
Komnenian period. In this context, it is worth noting that the Skleroi were not only
an important family during the reign of Basil II, but were also still a very important
0-5
See above, p.47
144
dynasty at the end of the eleventh century during the early decades of the reign of
which was inferior only to the rank of sebastos found among those families which
had marriage links with the Komnenoi. Another member of the family who
1100-01. Moreover, it is clear that the family were too politically important to be
irreparably damaged by their association with this conspiracy. Instead, their star
continued to rise. By 1104, the Andronikos, who was a protonobelissimos in 1094,
held the title of sebastos. Seibt, the principal modern prosopographer of the Skleros
family, has suggested that this elevation in status may even reflect the fact that
Andronikos had recently married into the imperial family.86 While this hypothesis
has yet to be substantiated by other evidence, Andronikos' senior title clearly
indicates his importance within the highest echelons of Byzantine political society
in the first decade of the twelfth century.
The idea that Skylitzes selected pro-Skleros source material in his analysis of
Basil's reign on the basis of the political importance of this family in his own
lifetime, gains support from the appearance of similar material in the account of
Basil's reign in Michael Psellos' 'Chronographia'. Although the exact date of the
composition of Psellos' 'Chronographia' has yet to be established, it is likely that
Psellos was at work on his history several decades before Skylitzes wrote the
145
O7
'Synopsis Historion', probably during the third quarter of the eleventh century.
However, just as the Skleroi enjoyed political prominence at the end of the eleventh
century when Skylitzes was writing, so they occupied the highest echelons of
Byzantine elite society when Psellos was active. Among the leading Skleroi of the
mid eleventh century, were Basil Skleros, strategos of the theme of the Anatolikon
in the late 1020s and early 1030s; Romanos Skleros, doux of Antioch on two
separate occasions during the 1050s, and a leading protagonist in the 1057 coup
__ QQ __
which unseated Michael VI; and two female members of the family. The first of
these women was the wife of Constantine Monomachos, who died before her
husband became emperor in 1042. The second, called Maria, was Constantine's
OQ
long-standing mistress both before and after his accession to the imperial purple.
The prominence of the family at this time was certainly recognised by Psellos
himself. For example, in the early stages of his account of the reign of Constantine
Monomachos (1042-55) he refers at length to the emperor's Skleraine mistress. 90
However, it is his much briefer reference to the Skleraine who had been
Monomachos' wife, which is the more interesting for our purposes. According to
Psellos this woman was, "..... descended from the very famous Skleros family".91
In this brief phrase the reader gains an elusive hint of the extent to which the
political identity of mid eleventh-century members of the Skleros family was
articulated through the fame and renown of their ancestors. This conclusion is
substantiated by the contents of a letter written by Michael Psellos to Romanos
Skleros, former doux of Antioch, in the mid 1050s. In this letter Psellos notes: "....
146
I neither saw your grandfather nor lay eyes on your father. But reputation has it that
they were noble (gennaioi) men and the paragons of good birth (eugeneia}.^2
While the prosopographical research of Seibt suggests that Bardas Skleros, the
rebel of Basil's reign, was too old to have been either the father or grandfather of
the doux Romanes,93 the comments of Michael Psellos offer strong circumstantial
evidence that stories glorifying the earliest members of the Skleros dynasty were
ubiquitous in mid eleventh-century Byzantium. Given the continuing importance of
the Skleroi within Byzantine political society at the time Psellos was writing, it may
not be entirely surprising that he chose to use one such Skleros narrative, that of
Bardas, the founder of the Skleros family, as one of his sources for his appraisal of
the reign of Basil II.
Although any connection between Michael Psellos and the pro-Skleros source must
remain provisional until I have been able to conduct more detailed research into the
historiographical background to the 'Chronographia', it nonetheless offers further
support for the general argument that has underpinned the first three chapters of
this thesis. That argument maintains that the first step towards uncovering the
history of Basil's reign, lies in understanding more of the social, literary and
political contexts behind the much later historiographical traditions which describe
that reign.
92 > ^ ^
aAA'cyct) (rot ovre rov ueumw elftov ovre rov narepa reSeatiat' ij fyfosi\ tie rovrov$ 'yewaiovg av8pa>g KOI
evycveias <bc& aya^iara. (Psellos (Michael), Scriptora Minora: Michaelis Pselli Scripta Minora, ed.
E.Kurtz and F.Drexl (2 vols., Milan, 1937-41), ii ,102). Romanos Skleros was not the only member
of this family with whom Psellos corresponded. At a rather later date, in the early 1070s, Psellos
wrote to Nicholas Skleros, at that time aproedros and krites of the theme of the Aegean Sea (Psellos
(Michael), Scriptora Minora, ii, letters 37, 44, 56 and 63; Seibt, Die Skleroi, pp.94-5). By 1084
Nicholas had been appointed to the senior judicial position of megas droungarios of the vigla. He
may even have been Skylitzes' predecessor in this post (see above, chapter one, p.35)
Romanos' father was probably the brother of the Basil Skleros, the strategos of the Anatolikon
mentioned above. His grandfather was Romanos, the son of Bardas Skleros (Seibt, Die Skleroi
pp.64-5, 76)
147
Chapter Four
I. Introduction
The over-arching ambition of this thesis is to analyse the political history of the reign
of Basil II by investigating how the Byzantine empire was governed in the later tenth
and early eleventh centuries. The first three chapters began that inquiry by considering
the medieval historiographical background to the reign. In the specific case of John
Skylitzes it has been suggested that the later eleventh-century literary, social, and
political contexts within which the author was working, meant that he presented the
reign in terms of the rivalries, deeds, and pedigrees of the aristocratic families, above
all those which were famous at the end of the eleventh century. At a more general
level, the extent to which all historians, whether writing in Greek or in other eastern
languages, were interested in the revolts of the first thirteen years of the reign, has been
noted.
In trying to overcome the difficulties presented by the medieval historical record, there
are a number of ways the historian could approach the question of Byzantine
government during the reign of Basil H The first might be to look at the relationship
between members of the political elite and the emperor on a case-by-case basis, in
order to assess the sustainability of Michael Psellos' contention that the emperor
suppressed the greatest families of the empire after the revolts of Skleros and Phokas.
This is an approach which has been adopted most recently by Sifonas. He has
suggested that with the exception of the Phokas family and their close allies the
Maleinoi, most leading families were not purged from government by Basil, and that
indeed, the emperor fostered the later tenth- and early eleventh-century aristocracy
149
within his administration.2 Thus, at a structural level, Sifonas sees little change in the
balance of power between emperor and aristocracy during the reign as a whole.
Instead, if the veil imposed over Basil's reign by historians at work in the later
eleventh century is ever to be lifted, it is necessary to start with a different conceptual
framework, and with alternative forms of evidence. These are the precepts
underpinning the inquiry into Byzantine government in the reign of Basil presented in
the rest of this thesis. In terms of conceptual framework, the analysis offered in the
next three chapters will focus on imperial authority, since it is only once the structures
and processes fundamental to the power of the emperor have been identified, that
relationships between the emperor and leading members of the political elite can be
meaningfully analysed. In other words, it is only once the outlines of the authority of
the state have been determined, that the use, abuse, and appropriation of that authority
150
by members of the political elite can be investigated, and the relationship between the
exercise of private and public power established. Given the time limits of a doctoral
dissertation, this analysis of imperial authority during Basil's reign will concentrate on
the eastern half of the empire, the region that was at the centre of the revolts of Bardas
Skleros and Bardas Phokas. As far as evidence is concerned, the thin historical record
will be set against other literary and material sources. However, since much of the
evidence from outside the historical record is difficult to date, the chronological
parameters of the investigation will extend beyond the limits of Basil's reign,
backwards into the mid-tenth century, and forwards into the mid-eleventh. Chapters
five and six will deal with imperial administration. This chapter will provide the
context to that discussion by looking at the economy of the eastern half of the empire
during the tenth and eleventh centuries, thus establishing the potential and limitations
of the resource-base upon which the bureaucratic framework of the Byzantine state
was built.
II. Sources
Unfortunately any study of the economy of the eastern half of the Byzantine empire in
the tenth and eleventh centuries is hampered by a variety of source material problems.
A dearth of documentary evidence, such as wills and land grants, means that the
agricultural exploitation of the countryside, the size and economic life of urban
settlements, and patterns of trade remain obscure. Indeed, the only archive materials
comparable to those found in much greater quantities in contemporary early medieval
western Europe, amount to the will of Eustathios Boilas, a member of the landed
gentry of mid eleventh-century Mesopotamia, and a small selection of documents
151
connected to the later eleventh-century landed arrangements of various monasteries
from the islands and littoral of western Asia Minor. 3 In conditions of such
documentary dearth, any interpretation of the economy in the Byzantine east has to be
based on occasional references in a kaleidoscope of literary sources and the material
record. Literary evidence includes Greek and non-Greek historical writings, saints'
lives, letters written by contemporaries, 'Digenes Akrites' (the solitary Byzantine
epic/romance), military handbooks from the second half of the tenth century, and the
accounts of ninth- and tenth-century Arab geographers. The material record is more
extensive and includes coins, seals, art, architecture and archaeology. However, many
of these forms of evidence are often not well-documented or well-known by historians.
Within this miscellany of literary and material sources some regions are better
represented than others. Areas of coastal western Asia Minor and the Pontus, where
Byzantine occupation continued into the thirteenth century, are represented in later
medieval archival documents. With care, conclusions about the patterns of land tenure
and methods of agricultural cultivation visible in these documents can be used to
reflect on conditions in the same localities in earlier periods such as the tenth and
eleventh centuries. These areas of the littoral of western Asia Minor are also relatively
Testament (Eustathios Boilas): P.Lemerle, 'Le testament de Eustathius Boilas (April 1059)', in Cinq
etudes sur le Xle siecle byzantin (Paris, 1970), pp.40-62. None of the eleventh-century monastic
document collections is very big. The largest comes from the monastery of Nea Moni on the island of
Chios off the Aegean coast of Asia Minor. Yet, this collection only amounts to nineteen texts available in
poorly edited versions of early nineteenth-century copies. Only four pre-1080 documents are extant from
the monastery of Stylos which was located on Mount Latros, south-west of Ephesos. These are to be
found in a fifteenth-century copy of a fragmentary cartulary. Only one document survives from the
monastery of Xerochoraphion, which was located to the north of Mount Latros. To this slim repetoire
can be added five pre-1080 documents which pertain to western Asia Minor, and survive in the archives
of the monastery of Saint John on the island of Patmos. These monastic collections are summarised in
M.Whittow, 'Social and Political Structures in the Maeander Region of Western Asia Minor on the Eve
of the Turkish Invasion' (Oxford Univ. D.Phil, thesis, 1987), pp.371-2.
152
rich in saints' lives.4 Moving slightly inland, the economy of those transitional regions
stretching from the uplands of the coastal river valleys onto the western and northern
reaches of the plateau is less well served in the written record. Here, references have to
be culled from a patchy hagiographical record and a handful of letters written by
contemporary churchmen.5 However, throughout the littoral and plateau regions of
western Asia Minor, written materials can also be supplemented with archaeological
findings from a variety of sources: excavations at several urban sites in the river
valleys of western Anatolia, including Ephesos, Sardis, Priene, Pergamon, and
Hierapolis (Pamukkale); the excavation at the late Roman and Byzantine city of
Amorion on the western reaches of the plateau; surveys of Byzantine fortifications and
regional landscape histories in which the medieval period is but one phase in a longue
duree perspective.6 However, it should be stressed that the archaeological record
4 The archive record of the river valleys of thirteenth-century western Asia Minor is analysed in
M. Angold, A Byzantine Government in Exile. Government and Society under the Lascarids of Nicaea
1204-6 (Oxford, 1975) and H.Ahrweiler, 'L'histoire et la geographic de la region de Smyrne entre les
deux occupations turques (1018-1307) particulierement au XIQe siecle', TM \ (1965), pp. 1-204; for the
Pontus see A.Bryer, 'The Estates of the Empire of Trebizond', Archaion Pontus 35 (1979), pp.370-477.
Saints' lives that reflect on everyday life in western Asia Minor include Saint Peter of Atroa (ninth
century: BHG 2364), Saint Paul of Latros, Saint Nikephoros of Miletos (both tenth century; BHG 1474
and 1338), and Saint Lazaros of Mount Galesion (eleventh century; BHG 979); from the Pontus, see the
Miracles of Saint Eugenics of Trebizond (various periods; BHG 610-13), Saint George of Amastris (late
eighth/early ninth century; BHG 688), the earlier sections of both versions of the life of Saint Athanasios,
the founder of the Athonite monastery of the Lavra, who grew up in Trebizond (tenth and early eleventh
century; BHG 187-8; 'Vitae Duae Antiquae Sancti Athanasii', ed. Noret, pp.5, 128, 130). Some of the
letters of Nikephoros Ouranos, one of Basil ITs closest advisers, refer to this region: Ouranos
(Nikephoros): Epistoliers byzantins du Xe siecle, ed. J.Darrouzes (Paris, 1960), pp.44-48.
5 Saints' lives from the northern or western periphery of the plateau: Saint Theodore of Sykeon (sixth to
seventh century; BHG 1748), Saint Philaretos (eighth to ninth century; BHG 1511), Saint Luke the Stylite
(tenth century; BHG 2239), and the Miracles of Saint Theodore (BHG 1764). The letters of the
metropolitans Leo of Synada and John Mauropous reflect on the economy of these transitional regions
(Leo of Synada, letters 43 and 54; Mauropous (John): The Letters of Joannes Mauropous Metropolitan
ofEuchaita, ed. A.Karpozilos (Thessalonika, 1990), pp.25, 148-93; see below, pp. 170-1).
The most detailed published interpretations of the Byzantine phases of several urban sites in western
Asia Minor have been produced by C.Foss: see, for example, Ephesus after Antiquity: A Late Antique,
Byzantine and Turkish City (Cambridge, 1979) and idem., Byzantine and Turkish Sardis (Cambridge
(Mass.), 1976); see also the unpublished thesis by Whittow for analysis of some of the less well-known
urban sites of the Maeander Valley ('Social and Political Structures in the Maeander Region', especially
chapters 2-6, pp.26-265). For Amorion see annual reports by R.M.Harrison and C.S.Lightfoot in Anatol
StodVoh. 38-46 (1988-1996). C.Foss is also the principal authority on medieval Byzantine fortifications:
153
almost always reflects the history of urban settlement rather than the fortunes of the
Byzantine countryside, since with the exception of some recent surveys, most
excavations have focused on the great city sites of Late Antiquity rather than smaller
towns or villages. 7
Further east the archaeological record is patchier, literary sources are rare, and the
approach in the secondary literature to the material culture of these areas has tended to
be encyclopaedic rather than interpretative, with a stress on the recording of
monuments and their decoration rather than on the wider economic, social, and
Q ____
political contexts of the visual record. This is an approach particularly visible in many
'Late Byzantine Fortifications in Lydia', JOB 28 (1979), pp.297-320; idem., 'The Defenses of Asia Minor
against the Turks', Greek Orthodox Theological Review 27 (1982), pp. 145-204; idem., 'Sites and
Strongholds of Northern Lydia', Anatol Stud 37 (1987), pp.81-101; idem., 'Strobiles and Related Sites',
Anatol Stud 1% (1988), pp. 147-74; idem., 'Byzantine Malagina and the Lower Sangarius', Anatol Stud 41
(1991), pp. 161-183; idem., Survey of Medieval Castles of Anatolia (3 vols., Oxford, 1985-1996); and
idem, with D.Winfield, Byzantine Fortifications: an Introduction (Pretoria, 1986). For the fortifications
of the Maeander region in western Turkey see H.Barnes and M.Whittow, 'The Oxford University/British
Institute of Archaeology at Ankara Survey of Medieval Castles of Anatolia', Anatol Stud 43-44 (1993-4)
dealing with Mastaura and Yilanli kalesesi respectively; for the Pontus, J.Crow and S.Hill, 'Byzantine
Fortifications of Amastris in Paphlagonia', Anatol Stud 45 (1995), pp.251-65; and for the northern
reaches of the plateau, J.Crow, 'Alexios Komnenos and Kastamon: Castles and Settlement in Middle
Byzantine Paphlagonia', in M.Mullett and D.Smythe (eds.), Alexios I Komnenos (Belfast, 1996), pp. 12-
36. Regional surveys in which the Byzantine period will be contextualised within a broader chronology,
include the projects underway in Paphlagonia and on the Konya (Ikonion) plain: R.Matthews, 'Project
Paphlagonia', Anatolian Archaeological Reports 3 (1998) pp.20-1; D.Baird, 'Konya Plain Survey',
Anatolian Archaeological Reports 1-3 (1995-7). Many of the volumes produced by the Tabula Imperil
Byzantini project present a site-by-site guide to the eastern half of the empire. Each volume also contains
a summary of the historical geography of the regions it covers. The relevant volumes are: 2
(Kappadokien), 4 (Galatien und Lykaonien), 5 (Kilikien und Isaurian), 7 (Phrygien und Pisidien), 9
(Paphlagonien und Honorias). Volume 10 (Aigaion Pelagos), deals with the islands lying off the coast of
western Asia Minor (TIB: Verlag der Osterreichischen Akademie der Wissenschaften (Vienna, 1981-
1998). A.Bryer and D.Winfield, The Byzantine Monuments and Topography of the Pontus (2 vols.,
Washington D.C., 1985) also present a historical geography of the Pontus.
7 S.Mitchell, Anatolia: Land, Men and Gods in Asia Minor (2 vols., Oxford, 1993), i, 9
8 The Life of Saint Michael Maleinos (first half of the tenth century; BEG 1295) contains occasional
references to Cappadocia and Ankara. Although most of the Life of Saint Nikon Metanoeite of Sparta is
set in Greece and the Peloponnese in the second half of the tenth century, it contains some references to
the saint's childhood in the theme of the Armeniakon (BHG 1366). Other literary sources include the
Digenes epic which comments on early medieval Cappadocia and the Euphrates region (Digenes Akrites:
Digenis Akritis: the Grottaferrata and Escorial Versions, ed. E.Jeffreys (Cambridge, 1998). The standard
guide to the monuments of the Byzantine east is T.Sinclair, Eastern Turkey: an Architectural and
154
modern analyses of Cappadocia in eastern Anatolia. Although hundreds of rock-cut
churches, monasteries, and other troglodyte complexes in this region, many richly
decorated with wall-paintings and sculpture, have been rigorously documented and
discussed by art historians, the significance of this material record for the broader
political history of the medieval Byzantine empire is rarely considered.9 This is a
curious omission since Cappadocia is widely believed by many modern historians to be
the region where the most powerful aristocratic families held their estates, and where,
indeed, they regularly fomented rebellion against imperial authority. 10
The dispersed and fragmentary nature of the evidence has given rise to very different
interpretations among modern scholars about the nature of the economy in Asia Minor
during the period which followed the loss of Egypt, Syria and Palestine to the Arabs in
the mid-seventh century and the Turkish invasions of Asia Minor itself at the end of
the eleventh century. Some historians maintain that consistent seasonal raiding by Arab
armies between the seventh and the tenth centuries caused such widespread
depopulation and deurbanisation that Anatolia was turned into a zone characterised by
Archaeological Survey (4 vols., London, 1987-90). Relatively few studies integrate geography,
monuments, and literary texts in these regions: among the rare exceptions is Howard-Johnston,
'Byzantine Anzitene', pp.239-90.
The standard art history guides to the rock complexes of Cappadocia include G. de Jerphanion, Une
Nouvelle Province de I'art byzantin. Les eglises rupestres de Cappadoce (2 vols., Paris, 1925-42) and
M.Restle, Die byzantinische Wandmalerei in Kleinasien (3 vols., Recklinghausen, 1967). In recent years
the architectural and artistic record of this region has attracted enormous interest from art historians.
Some of the most noteworthy studies include: A.Wharton-Epstein, Tokali Kilise: Tenth-Century
Metropolitan Art in Byzantine Cappadocia (Washington, 1986), and C.Jolivet-Levy, Les Eglises
byzantines de Cappadoce: le programme iconographe de Vabside et de ses abords (Paris, 1991). The
basic historical geography of Cappadocia is provided by Hild and Restle, Kappadokien, and F.Hild, Das
byzantinische Strassensystem in Kappadokien (Verlag der Osterreichischen Akademie der
Wissenschaften) (Vienna, 1977). Various publications among the vast oeuvre of N.Thierry have made
some attempt to place the artistic and architectural record of Byzantine Cappadocia in wider social and
political contexts: see, for example, Nouvelles Eglises rupestres de Cappadoce. region du Hasan Dag
(Paris, 1963); idem., 'Les enseignements historiques de I'archeologie cappadocienne', TM 8 (1981),
pp.501-19; idem., Haut Moyen-dge en Cappadoce: les eglises de avuin (Paris, 1994); see also
L.Rodley, Cave Monasteries of Cappadocia (Cambridge, 1985).
155
scattered village communities. According to this model Late Antique city sites were
reduced to hill-top forts, home only to imperial tax collectors and army garrisons. 11
Others believe that although raids drained the vitality of the economy of Asia Minor,
city sites continued to be occupied and act as centres of economic exchange for their
agricultural hinterlands. 12 Less contentious is the belief that as Arab raiding activity
waned in the later ninth and early tenth centuries the economy of Asia Minor began to
recover. However, it should be stressed that much less attention has been paid to this
period of expansion than to the retraction that followed the defeats of the seventh
century. 13
10 See, for example, Kaplan, 'Les grands proprietaires de Cappadoce', pp. 125-158
11 Proponents of widespread and dramatic deurbanisation in the seventh to ninth centuries include Mango,
Empire of the New Rome, pp.71-3; AP.Kazdan and G.Constable, People and Power in Byzantium: An
Introduction to Modern Byzantine Studies, 3rd edn. (Washington, 1996), p.57; J.F.Haldon and
H.Kennedy, 'The Arabo-Byzantine Frontier in the Eighth and Ninth Centuries: Military and Society in the
Borderlands', ZVRI 19 (1980), pp.85-95; M.Angold, 'The Shaping of the Medieval Byzantine 'City",
ByzForsch 10 (1985), pp. 1-8. Clive Foss's analyses of the archaeological records of a variety of seventh-
to ninth-century urban centres have often been cited by those scholars who wish to support theories that
the Late Antique city completely disappeared and that settlement retreated to hilltop sites. However,
although Foss suggests that the Persian invasions at the beginning of the seventh century and regular
incursions by Arabs over the next two centuries dramatically reduced the size of many urban sites of Asia
Minor, he does not envisage the complete desertion of such locations, and admits that the literary and
material evidence of urban decay is both fragmentary and sometimes contradictory: see for example Foss,
Byzantine and Turkish Sardis, pp.53-62; Ephesus After Antiquity, pp. 103-113.
12 S.P.Vryonis, The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from
the Eleventh through the Fifteenth Century (Berkeley and Los Angeles, 1971), pp.6-24, adhered to this
more optimistic model. His views have been supported to some extent by more recent research which
argues that the lower cities of many Late Antique sites continued to be occupied, with their inhabitants
only fleeing to hill-top citadels during raids: see, for example, Crow and Hill, 'Byzantine Fortifications of
Amastris', pp.251-65; Crow, 'Alexios Komnenos and Kastamon', pp.26-33; C.Zuckerman, 'The Reign of
Constantine V in IheMiracles ofSt Theodore the Recruit (BHG 1764)', REB 46 (1988), pp.191-210.
13 AHarvey, Economic Expansion in the Byzantine Empire 900-1200 (Cambridge, 1989), chapters 2 to
6; Mango, Empire of the New Rome, pp.81-3; Lemerle, Cinq Etudes, pp.272-93, deal with the expansion
of the economy in the ninth to twelfth centuries throughout the Byzantine empire, that is to say, in the
Balkans as well as in Asia Minor. Their conclusions overturned previously held beliefs that internal
political collapse and external military defeat in the last third of the eleventh century were predicated on
longterm economic decline. For this older view, see for example, N.Svoronos, 'Societe et organisation
interieure dans Pempire byzantin au Xle siecle: les principaux problemes', in Etudes sur I'organisation
interieure, la societe et I 'economic de I'empire byzantin (Variorum Reprints, London, 1973), number 6.
156
This chapter will be primarily concerned with the changes to the economy that were
set in train by the waning of Arab raids into the heartland of Anatolia during the tenth
century, and the eastern conquests that were secured by Byzantine armies in and
beyond the Taurus and Anti Taurus mountains during the half-century preceding
Basil's reign. Given the sheer size of the geographical area under consideration and the
effort required to collate and synthesise such a heterogeneous array of source
materials, this investigation does not represent a detailed synthesis of all the available
material and literary evidence, but instead offers a backdrop against which
administrative and political developments during Basil's reign can better be
understood. 14
At the heart of the eastern half of the Byzantine empire lies the central Anatolian
plateau. Ringed on all sides by mountains, and roughly triangular in shape, the plateau
extends five-hundred kilometres from Akroifnos (modern-day Turkish Afyon) in the
west to the Halys river (Kizil Irmak) in the east, and three-hundred kilometres from the
mountains of the Pontus in the north to the Taurus range in the south. It rises from a
height of one thousand metres in the west to one-and-a-half thousand metres in east.
Although some of the terrain, particularly the area around the salt lake, Tuz Golu, is
flat and barren, much of the plateau is characterised by rolling hill country. It is
punctuated by high peaks, notably the 4,000 metre Mount Argaios south of Kaisareia
It should also be noted that while most references in the remainder of this chapter are taken from
original primary material, the considerable weight of evidence considered means that some citations are
extracted from secondary authorities.
See Map 2 for all references in this physical geography section
157
(Kayseri), and isolated fertile plains, including those around Ikonion (Konya) in the
south and Kaisareia in the east. Surrounded by a rim of mountains on all sides, the
plateau experiences a continental climate of hot dry summers and cold winters. Snow
falls between November and February for an average of twenty-five days a year in
western regions, rising to thirty-six days further east. Rainfall is low averaging 230
millimetres per annum in the area around Ankara and 180 millimetres in the Ikonion
plain. Much of the terrain is treeless, but in springtime, the season when most of the
annual rain is precipitated, much of the plateau is covered with grass and wild
flowers. 16
To the north the plateau is bordered by the heavily-wooded Pontus mountains, which
while continuous and steep in the east, gradually give way to the lesser ranges of
Paphlagonia further west. Beyond these mountains lies a narrow coastal strip bordering
the Black Sea where the climate is typified by mild, wet winters and relatively humid
summers. To the north-west and west of the plateau mountains rise to a height of two-
and-a-half thousand metres, before yielding to the plains and river valley systems of
the coasts of the Sea of Marmara and the Aegean. Here the climate is predominantly
Mediterranean: hot dry summers followed by mild, wet winters, with little frost,
although winter conditions can be more severe around the Sea of Marmara. 17 Steep,
rugged mountains shield the southern side of the plateau. South of these mountains lies
For the relief and climate of the plateau see Mitchell, Anatolia, i, 43-7, M.Kaplan, Les Hommes et la
terre a Byzance du Vie au Xle siecle (Paris, 1992), pp.7-8,17-18; MF.Hendy, Studies in the Byzantine
Monetary Economy c.300-1450 (Cambridge, 1985), pp.26-28, 40-46; X. de Planhol, Les Fondements
geographiques de I'histoire de /'Islam (Paris, 1968), pp. 197-201, 207-209, 221-4. J.D.Howard-Johnston,
'Studies in the Organisation of the Byzantine Army in the Tenth and Eleventh centuries', (Oxford Univ.
D.Phil, thesis, 1971), pp.xix-xxi, 194-236.
For the coastal river valleys in general see: Kaplan, Les Hommes et la terre, pp. 15-16, 23; Whittow,
Orthodox Byzantium, p.30; Hendy, Studies, pp.28-32. For western Asia Minor: Angold, Byzantine
158
the Pamphylian plain which surrounds the port town of Attaleia (Antalya). Once again
1 ft
this region is typified by a Mediterranean climate. However, while marked
differences in climate and relief exist between the coasts of western Asia Minor and
the central plateau of the interior, there are large transitional areas between these two
regions where a mixture of littoral and plateau conditions prevail. For example, in the
regions of the headwaters of the Maeander river, fertile plains exist up to a height of a
thousand metres. A little further east around the Pisidian Lakes the month of April can
bring an average of three days of frost, and yet daytime temperatures can reach thirty
degrees centigrade. 19
To the south-east of the plateau lies the steep, narrow rim of the Taurus mountains.
Beyond this range stretches the broad Cilician plain where a Mediterranean climate is
is intensified by very torrid summer heat.20 The Cilician plain continues eastwards to
the Amanos mountains, which run in a north-east to south-west direction and separate
the plain of Antioch from the coast of northern Syria. East of the Taurus lies the rather
broader range of the Anti Taurus extending one-hundred-and-thirty kilometres from
west to east, and two-hundred kilometres from north to south. Between the major
ridges of this range are located alluvial depressions, such as the plains of Lykandos
Government, p.!02ff. For the Pontus: Bryer, Empire of Trebizond, p.372ff.; Bryer and Winfield,
Topography and Monuments of the Pontus, pp.2-4.
C. Foss, 'The Cities of Pamphylia in the Byzantine Age', in Cities, Fortresses and Villages of
Byzantine Asia Minor (Variorum, Aldeshot, 1996), number 4, pp. 1-4; X. de Planhol, De la Plaine
pamphylienne aux lacspisidiens: nomadisme et viepaysanne (Paris, 1959), pp.27-42.
Hendy, Studies, pp. 56-7 following the fourth-century 'Expositio Totius Mundi et Gentium' draws too
sharp a distinction between the climate and economies of interior plateau regions such as Cappadocia,
Galatia, Phrygja, Armenia Minor, Paphlagonia and the Pontus and the coastal areas of Pampytia, Lycia,
Caria, Asia, Hellespont and Bithynia. De Planhol, La Plaine pamphylienne, p.32, Mitchell, Anatolia, i, 5-
7 and Kaplan, Les Hommes et la terre, p. 17 stress the existence of transitional areas between coast and
plateau.
"J{\
Philetos Synadenos, the judge (krites) of Tarsos during Basil's reign, comments on the vicious heat of
the summer in Cilicia: Synadenos (Philetos): Epistoliers byzantins du Xe siecle, ed J.Darrouzes (Paris,
I960), letter 12.
159
71
(Elbistan), Melitene (Malatya) and the Anzitene. Eastwards beyond the Euphrates
and its principal tributary the Arsanias (Murat Su) stretch the mountains and lava flows
160
Kaisareia, or south-eastwards to Podandos. 25 However, once on the plateau the
traveller could branch away from these main roads and take several more minor routes
north-south routes ran along the western edge of the plateau from Dorylaion, via
Furthermore, although the interior of western Anatolia was separated from the western
littoral by mountains, it was relatively easy to pass from the western reaches of the
plateau into the coastal river valleys. For example, some routes led from Amorion on
the western plateau, via Apameia (Dinar), into the Maeander valley. Others led
westwards from Amorion, via modern-day Uak, before descending into the Hermos
valley system.26
Much more difficult than travelling westwards from the plateau was crossing the
mountains that bordered the Anatolian interior to the north and the south. In the
Byzantine period most routes from the southern coast led up from Attaleia, and having
crossed the mountains, headed either in a westerly direction towards coastal valleys of
the Aegean, or northwards along the western edge of the plateau and on to
25 The more southerly route of these routes (that from Constantinople to Podandos via Amorion and
Ikonion) was described in detail by the ninth-century Arab geographer Ibn Khurradadhbih: Bibliotheca
Geographicorum Arabicorum, ed. M.J.de Goeje (Leiden, 1889), vi, 73-5, 82, 85.
Whittow, 'Social and Political Structures in the Maeander Region', p. 194
These routes are described by the twelfth-century geographer al-Idrisi on the basis of earlier Arab
accounts: M. Kordosis, 'The road network in Asia Minor and Armenia according to Edrisi. The stop of
"Meldini"', in EJeffreys et al. (eds.), Q/AeMm Studies in Honour of Robert Browning (London, 1996),
especially p. 182, which describes the route between Attaleia and Amorion and lists as stopping places a
variety of points in Pisidia including Philomelion and Polybotos. At the turn of the tenth and eleventh
centuries Saint Lazaros of Mount Galesion followed a route which led inland from Strobiles in south-
west Asia Minor to Chonai in the interim zone between coast and plateau. He then journeyed southwards
to Attaleia (Malamut, Sur la route des saints byzantins, p.41; see also Hendy, Studies, p.26, and Vryonis,
The Decline ofMedieval Hellenism, p.33)
161
mountains and linking coast and plateau were rare. In the far north-east many roads
*\Q
were open only on a seasonal basis. As recently as the beginning of this century,
caravans could not journey over the Pontic Alps to Trabzon (Trebizond) on the Black
impenetrability of these northern and southern ranges, meant that the coastal strips
they sheltered were more accessible by sea than by land. 30 Harbour towns, many
heavily refortified by imperial authorities during the seventh to the tenth centuries,
punctuated both shorelines: Attaleia, Korykos and Seleukia in the south; Amastris,
Sinope and Trebizond on the Pontic coast. 31 To the south-east of the plateau the Taurus
mountains present an almost impenetrable barrier. The only major passage through this
range is at the Cilician Gates linking Tyana in southern Cappadocia with the Cilician
1*5
plain. The Anti Taurus mountains are more easily traversed. One important
Byzantine route ran from Kaisareia, past the mountain-top fortification of Tzamandos,
eastwards to the plains of Melitene and the Anzitene. 33 A Roman road from Kaisareia
to Melitene took a more circuitous route through the Anti Taurus via the plain of
fJO
The easiest passage ran between Amisos on the shores of the Black Sea and Amaseia on the northern
reaches of the plateau: Vryonis, The Decline ofMedieval Hellenism, pp.32-3.
29 Bryer and Winfield, Topography and Monuments, pp.39-56; T.Mitford, 'Military Road from Satala to
Trabzon' Anatolian Archaeology Reports 2 (1996), p.21-2.
30 Bryer and Winfield, Topography and Monuments, p. 18: a messenger travelling from Constantinople to
Theodosioupolis at top speed by an overland route in the medieval period took fifteen days; the journey
could be completed in a third of the time if the Constantinople-to-Trebizond leg of the journey was by
sea.
For many of these sites see Foss and Winfield, Byzantine Fortifications, pp. 15-16, 21-2, 30; Foss,
'Cities of Pamphylia', pp. 1-62; Crow and Hill, 'Byzantine Amastris'. The principal medieval
refortification of Trebizond occurred in the fourteenth century. However, parts of the Roman defences in
the upper city show signs of having been repaired during the ninth century: Bryer and Winfield,
Topography and Monuments, pp. 186-93.
Hild, Strassensystem, pp.51-63, 118-129; Ibn Hawkal reflected on the impenetrability of the Taurus
(IbnHawkal,p.l81)
Hild, Strassensystem, pp.90-100. Ibn Hawkal describes an Arab raiding expedition from Mayafariqin
which travelled via the Anzitene, Melitene and Tzamandos (Ibn Hawkal, pp. 191-2). A similar route
linking the regions beyond the Anti Taurus with Kaisareia is mentioned by Ibn Hawkal's close
contemporary al-Mukaddasi (E.Honigmann, 'Un itineraire arabe a travers le Pont', Annuaire de I'lnstitut
de Philologie et d'Histoire Orientales et Slaves: Melanges Franz Cumont (Brussels, 1936), pp.270-1).
162
Lykandos (Elbistan). In the first year of his revolt, Bardas Skleros seems to have
crossed the Anti Taurus from Charpete (Harput) in the Anzitene using this Roman
Further east still the Anti Taurus range itself could be approached from western
Caucasia by routes which followed the principal rivers of the region. The more
southerly route ran alongside the Arsanias (Murat Su) from Manzikert north of Lake
Van, via the plain of Mu, towards Melitene. The northerly route kept close to the
Araxes, passed Theodosioupolis, and approached the northern Anti Taurus along the
upper Euphrates.35 The importance of these two routes lay in the ease of passage they
allowed for east-west movement between Caucasia, the Anti Taurus and Anatolia. But
while lateral movement across Caucasia along these east-west arteries was relatively
easy, it was much more difficult to move into this region from those areas which lay to
the south. The only major passes connecting the isolated plains of the Anti Taurus and
the Armenian highlands with the northern Djaziran cities of Amida and Mayafariqin
located on the upper Tigris river, were the Bitlis pass south-west of Lake Van, and the
Ergani pass south of the Anzitene. A subsidiary pass crossed the Bingol Dagi
appoximately halfway between the Ergani and the Bitlis passes. 36
34 Skylitzes, pp.317-9. See chapter three, pp. 123-4, for the route followed by Skleros in 976/7.
35 Bryer and Winfield, Topography and Monuments, p.20. A tenth-century Armenian itinerary refers to a
trading route which ran from Dvin in eastern Armenia to Karin (Theodosioupolis), and thence to
Constantinople via a series of towns along the northern reaches of the Anatolian plateau including
Koloneia, Amaseia, Gangra and Ankara (H.A.Manandian, The Trade and Cities of Armenia in Relation
to Ancient World Trade, (trans. N.G.Garsoian), (Lisbon, 1965), pp. 168-9).
J.D.Howard-Johnston, 'Procopius, Roman Defences North of the Taurus and the New Fortress of
Citharizon', in D.H.French and C.S.Lightfoot (eds.), The Eastern Frontier of the Roman Empire (BAR
International Series, Oxford, 1989), pp.215-8.
163
IV. Land use 37
The most productive areas of the landmass of Anatolia are the coastal plains and river
valleys of the north and west, which enjoy a mild Mediterranean climate. Traditionally
these regions have been characterised by a mixed agriculture of arable, arboriculture
(vines, olives, figs and other fruit trees) and small-scale livestock rearing. In the
thirteenth century the agricultural economy of the empire of the Laskarids based in the
river valleys of western Asia Minor included the rearing of animals and the cultivation
of arable crops, olives, vines and orchards. The main grain-producing area was around
3R
Smyrna. Although the evidence for land use in the tenth and eleventh centuries is less
plentiful than during the Laskarid period, contemporary accounts comment upon the
agricultural productivity of the western littoral of Asia Minor in the second half of the
tenth century. Leo, the metropolitan of Synada, a diocese located on the western
periphery of the plateau, declared that he was accustomed to receiving supplies of
wine, olives, and wheat from the western theme of the Thrakesion and the region of
Attaleia. The Arab geographer Ibn Hawkal noted that the territory which separated the
two cities of Constantinople and Attaleia :
is fertile and well populated and the passage of trade is uninterrupted along the
whole route all the way from the suburbs of Attaleia and its rural hinterland,
which flourishes greatly and is very productive, as far as the Bosphoros.39
37
For all place names henceforth in this chapter, see Map 3, except for those in western Cappadocia,
which are indicated on Map 4.
38
Angold, Byzantine Government, p. 103 ff; see below, pp. 194-5and n. 150
Leo of Synada, letter 43; Ibn Hawkal, p. 196; see also Kaplan, Les Hommes et la terre, p.23; Hendy,
Studies, p.28; Whitlow, 'Social and Political Structures in the Maeander Region', pp.369, 389.
164
A similar pattern of intensive market gardening characterised the landscape of the
medieval Pontus. In 1292 the envoys of Edward I of England, who stopped off in the
empire of Trebizond on their way to Tabriz in northern Iran, were able to purchase
wine, meat, cereal, rice, oil, fruit, wax, herbs, spices, cloth, and hemp.40 Some of these
products were produced in this region in the ninth and tenth centuries: the Book of the
Eparch, compiled during the reign of Leo VI (886-912), mentions that linen and spices
were both exported from the Pontus to Constantinople.41 The capacity of the northern
shore of Asia Minor to support a string of towns from Early Antiquity on into the
medieval period, despite the narrowness of the cultivable area between the Black Sea
A *")
and the Pontic Alps, is a further index of the natural productivity of this region.
Inland from the fertile coasts it is usually argued that the principal economic activity of
the central Anatolian plateau during the medieval Byzantine period was the rearing of
livestock. Such arguments are predicated on the belief that until advent of modern
irrigation and transportation, the altitude, the scrubby terrain, and above all the
intensity of the winter cold and the irregularity of annual rainfall, made the cultivation
of crops unduly precarious.43 The best contemporary evidence in support of this view
comes from a letter written by Leo of Synada. He claimed that the altitude of the theme
of the Anatolikon, the province in the western plateau where his diocese was located,
was such that neither wheat, grapes, nor olives would grow there, and that the only
40 r>
Bryer, 'Empire of Trebizond', pp.375-8; Bryer and Winfield, Topography and Monuments, pp.5-7.
*' Eparch. Das Eparchenbuch Leos des Weisen, ed. J.Koder (CFHB XXXIII, Vienna, 1991), pp. 106.
42
Whittow, Making of Orthodox Byzantium, p. 30
43
Kaplan, LesHommes et la terre, pp.23-4; Hendy, Studies, pp.26-8, 34-5, 53-8.
165
viable arable crop was barley. 44 Furthermore, when Crusading armies crossed the
central Anatolian plateau on their way to Palestine in the eleventh and twelfth
There are also several references within other Byzantine and Arab sources to the
ubiquity of the rearing of livestock all over the tenth- and eleventh-century plateau.
Saint Philaretos, who owned forty-eight estates in the late eighth century in Galatia and
Paphlagonia on the north-west periphery of the plateau, possessed vast flocks and
herds.46 The ninth-century life of Saint Peter of Atroa refers to a wealthy landowner in
upland Lydia whose flocks died prematurely after succumbing to the malevolence of a
demon.47 In the early tenth century Saint Luke the Stylite worked as a swineherd near
When Arab jihad raiders attacked the plateau from east of the
AQ
Kotyaion (Kutahya).
Taurus and Anti Taurus mountains in the ninth and early tenth-centuries they often
rustled livestock on the plateau. In 931 raiding parties from Tarsos headed for
Amorion and Ankara to capture live animals. Eight years previously a huge tally of
eight thousand cattle and twenty thousand sheep had been captured by Tarsiote
forces.49 In the mid-tenth century the Arab geographer Kudama ibn Djafar advised
44 Leo of Synada, letter 43; L.Robert, 'Sur les lettres d'un metropolite de Phrygie au Xe siecle', Journal
des Savants (1961), pp.97-166; (1962), pp. 1-74.
45 For instance, the German contingent of the Third Crusade could find no ready supplies of grain or fruit
on the plateau during their march between Laodikeia (Denizli) in the west and Karaman in the south. Odo
of Deuil, a participant in the Second Crusade of 1147, had observed a similar lack of supplies in the
regions east of Laodikeia (Hendy, Studies, pp.42-3).
46 Philaretos' livestock amounted to 600 cattle, 100 pairs of oxen, 800 mares, 80 mules and 12,000 sheep:
Saint Philaretos: 'La vie de S.Philarete', ed. and trans. M.H. Fourmoy and M.Leroy, B 9 (1934), p. 113;
Kaplan, LesHommes et la terre, p. 3 32.
47 H.Maguire, Ikons of their Bodies: Saints and their Images in Byzantium (Princeton, 1996), p. 1 14.
48 Saint Luke the Stylite: 'Vie de saint Luc le Stylite (879-979),' ed and trans. F.Vanderstuyf, PO XI
166
raiding parties to brave the winter snows of Anatolia and the mountain passes, in order
to round up flocks and herds which had been brought down from their upland summer
pastures and left to graze on the more accessible regions of the plateau. 50 By the ninth
century baggage animals for imperial expeditions were reared in the western-plateau
districts of Phrygia, on private estates as part of a tax in kind called the mitaton. When
imperial expeditions were due to depart from Constantinople, the logothetes of the
herds and flocks would transport pack animals reared on these estates north and
westwards to Malagina in the region of the lower Sangarios, the first stopping place on
the route for armies heading towards the eastern frontier. The animals would then be
left in the care of the homes of the Stables. 51 Further east the plain of Lykandos was
described by Constantine Porphyrogenitus as suitable for the raising of stock, the
fattening of cattle, and grazing of horses. 52 The historian Michael Attaleiates records
in the eleventh century that livestock markets were held at Ikonion on the southern
reaches of the plateau.53
Yet, it would be premature to assume that the whole of the Anatolian interior had by
physical necessity to be given over to the rearing of livestock. In the first place it is
clear that the interior of Anatolia could support other agricultural activities. In all but a
few areas of the plateau the soil cover is suitable for the growing of grain. In some
regions such as the area around the Pisidian lakes, a rich red loam even makes some
varieties of arboriculture possible. Moreover, while drought can affect the plateau on
50 Kudama ibn Dja'far: Bibliotheca Geographicorum Arabicorum, ed. M.J.de Goeje (Leiden, 1889), vi,
pp. 199-200; Howard-Johnston, 'Byzantine Army', p.213.
51Military Expeditions: J.Haldon, Constantine Porphyrogenitus Three Treatises on Imperial Military
Expeditions (Vienna, 1990), text A; see also Foss, 'Byzantine Malagina', p. 163; Howard-Johnston,
'Byzantine Army', p. 168. Ibn Khurradadhbih described the region around Dorylaion in Phrygia as "the
pasture of the asses of the emperor"( Ibn Khurradadhbih, p.73).
52 Hendy,Sfc/dr<?5, p.55.
167
occasions, average annual precipitation levels are adequate for the cultivation of grain.
Furthermore, during the Late Antique period all but the most barren and arid zones of
Asia Minor were typified by a sedentarised arable agriculture. 54 According to the life
of Saint Theodore of Sykeon, cereal production was ubiquitous in the north-west
plateau in the sixth and seventh centuries. 55
168
underground complexes where grain could be stored in order to sustain local
communities during famine or invasions. These depositories were often guarded with
large millstones.59 One of the biggest underground complexes is at Derinkuyu
But if arable farming was possible and indeed practised on the contemporary plateau,
how are Leo of Synada's reflections on the limited agricultural potential of western
Anatolia to be interpreted? Here we should be cautious of accepting Leo's statements
as accurate autopsy. Leo's complaints about the desolate nature of his diocese are
formulated in a letter to Basil n requesting the restoration of an allowance which
enabled him to give generous hospitality to his guests. Despite the guarantee of this
allowance by previous imperial chrysobulls, it had recently been rescinded. If, as its
modern editor suggests, this letter can be dated to the late 990s, then it may be that Leo
was victim of the novel promulgated in 996, in which all chrysobulls issued by Basil's
great uncle and former advisor Basil Lekapenos were declared void unless
Arab geographers noted that rock was quarried and shaped into mill stones in the area of western
Cappadocia characterised by underground complexes (Ibn Khurradadhbih, p.80)
Hild and Restle, Kappadokien, p.227. See Map 4, for the exact location of Derinkuyu
Hendy, Studies, p. 54. The difficulty of knowing whether Crusaders reflected on short term devastation
caused by Turkish invasions or long-term economic malaise has been pointed out by Foss, 'The Defenses
of Asia Minor against the Turks', p. 151.
169
countersigned by Basil himself.62 In these circumstances it is distinctly possible that
Leo chose to exaggerate the barbarity of the wilderness in which he was forced to live,
in order to get his erstwhile privileges restored. That Leo's comments should be
interpreted within the wider context of a petition to the emperor, was first suggested
several decades ago by Louis Robert. He noted that the archbishop's testimony was
steeped in a rhetoric of humiliation, "profitable pour une reclamation fiscale".63
It is clear that hyperbole is relevant in other contexts where letter writers comment on
the poverty of their provincial locales. In one letter, the mid-eleventh century
metropolitan of Euchaita John Mauropous describes his see on the northern plateau as
an uninhabited, treeless desert devoid of shade, wood, and vegetables. Yet in the same
letter John grudgingly concedes that the region "abounds..... in the production of
grain."64 And in his sermons he praises the well-watered hinterland of Euchaita, the
crops produced by local villages, the fat herds of local cattle, and the bustling fair
(panegyris} held on the feast day of the Saint Theodore.65 Such a dichotomy only
makes sense if it is understood within the literary topoi practised in epistolography in
medieval Byzantium. Tenth- to twelfth-century letter writers such as Nikephoros
Ouranos, Philetos Synadenos, Theophylact of Ochrid and Michael Choniates all
reflected on the poverty of their provincial postings as a matter of good literary taste.66
62 Leo of Synada, letter 43; Zepos and Zepos, lus, i, 270-1; see below, p.258 and n.165.
63 Robert, 'Sur les lettres d'un metropolite de Phrygie', pp. 150-6.
64 Mauropous (John), letter 64, pp. 172-3
65 Hendy, Studies, p. 141; S.Vryonis, 'The Panegyris of the Byzantine Saint', in S.Hackel (ed.), The
Byzantine Saint (Birmingham, 1981), p.202. In a summary of the medieval evidence relating to Euchaita,
J.Crow suggests that interpretations which see Euchaita reduced to little more than a livestock-rearing
ranch have been unduly pessimistic ('Alexios Komnenos and Kastamon', pp.25-33).
During military expeditions Nikephoros Ouranos (Constantinopolitan, keeper of the imperial inkstand,
diplomat and leading general during the reign of Basil n) displayed conspicuous nostalgia for
Constantinople and distaste for the hardships of life on campaign (Ouranos, letters 19, 41 and 47).
Philetos Synadenos, krites of Tarsos during Basil's reign, complained that residence among the Cilicians
170
Yet, as the correspondence of Theophylact and Michael Choniates demonstrates, the
same letter writers could also write on other occasions in praise of the locales under
Beyond the Taurus and Anti Taurus, in the plains which had been conquered by
Byzantine armies in the tenth century, there is less ambiguity about the use to which
the land was put. Here, as in the fertile river valleys of coastal Anatolia, mixed
agriculture was practised, and the landscape was characterised by cultivated fields,
pastures, trees, and mills. Most of these plains were quite large but isolated; their
agricultural produce supported a single, but sizeable, urban settlement. Only the very
extensive Cilician plain sustained more than one city: Tarsos, Mopsuestia, Adana and
Anazarbos. Unfortunately most descriptions of the agricultural production of these
regions in the early Middle Ages were composed by Arab geographers and historians
who wrote in the period before the Byzantine conquests of the mid- to later tenth
was turning him into a barbarian (Synadenos, letter 7). See below, pp.313-5, 336-9, for further discussion
of the careers of Nikephoros and Philetos.
67 M.Mullet, 'Originality in the Byzantine Letter: the Case of Exile', in A.R.Littlewood (ed.), Originality
in Byzantine Literature, Music and Art (Oxford, 1995), pp.39-58.
171
century.69 However, a rare snapshot of life in Byzantine rural northern Syria is
provided by Ibn Butlan, a Christian Arab doctor who travelled from Baghdad in
The evidence pertaining to land use suggests that all regions of the eastern half of the
Byzantine empire, including the Anatolian plateau, were able to support a mixed
agriculture of arable production and livestock rearing. Furthermore, it is clear that
mixed husbandry was maintained, albeit at a subsistence level, even in regions which
suffered severe Arab raiding. Once peace came to the eastern half of the Byzantine
empire, agriculture of many varieties had the potential to prosper. However, the degree
to which prosperity could be realised would be dependent on other economic variables,
not least the size of the labour force available to manage the landed resources of
Anatolia and the eastern frontier.
V.von Falkenhausen, 'Bishops', in G.Cavallo (ed.), The Byzantines (Chicago, 1997), p. 179.
69 For example Ibn Hawkal, writing in the later tenth century, describes Antioch and the cultivated fields,
pastures, trees and mills of its immediate hinterland. However, he explains that the city is now in enemy
(ie. Byzantine) hands. It is clear that his description draws on the observations of earlier Arab
geographers rather than contemporary eye witness testimony (Ibn Hawkal, pp. 176-7)
Ibn Butlan: The Medico-Philosophical Controversy between Ibn Butlan of Baghdad and Ibn Ridwan
of Cairo, trans. and ed. J.Schlacht and M.Meyerhof (Cairo, 1937), pp.54-5.
172
V. Economic growth
Turning from land use to economic expansion, there is clear evidence from across the
eastern half of the Byzantine empire that the tenth and eleventh centuries represented a
period of significant growth. In the regions of the coastal plains and river valleys the
best evidence of expansion comes from the archaeology of a variety of "urban" sites in
the river valleys of the Aegean and Mediterranean coasts. At Sardis during the ninth
century the main basilical church was refurbished. By the tenth century domestic
dwellings and small industrial units began to appear with greater frequency in many
different locations within the city: on the acropolis, at the temple of Artemis, in the
gymnasium, and at the sites of several Roman suburban villas. In the same period the
city of Ephesos also began to expand. The church of the Virgin was remodelled, and
the area around it became a new residential precinct. At several "urban" sites in
western Asia Minor finds of stray copper coins, the small change of everyday
economic exchange, are more plentiful in the tenth and eleventh centuries than in the
preceding Dark Age period. Although the numbers involved are patchy because of the
at Aya Suluk, (two kilometers inland from the ancient urban settlement at Ephesos), as
71 G.Bates, Byzantine Coins (Cambridge (Mass.), 1971) deals with the coin finds at Sardis; C.Foss,
Ephesus after Antiquity, p. 112, 130ff. deals with those at Ephesos; C.Morrisson, 'Monnaie et finances
dans 1'empire byzantin Xe-XTVe siecle', in Hommes et richesses dans I 'empire byzantin (2 vols., Paris,
173
More controversial than whether these "urban" sites were expanding is how this
expansion should be interpreted. In his assessment of Sardis in the early Middle Ages
Clive Foss emphasises the modesty of the expansion of the "city" in the ninth and tenth
centuries. Rather than a viable urban settlement, Sardis was nothing more than a loose
network of villages scattered in the ruins of the Late Antique city. The purpose of the
agricultural activity of these villages and their small industrial units, such as potteries,
was merely to provide food and goods for local consumption. Yet, this highly localised
model may misrepresent the economy of tenth- and eleventh-century Byzantine
western Asia Minor. In the first place, it is possible that some towns were considerably
larger than the current state of their excavations indicate. For example, much of
medieval Sardis may still be underground: neither the area around the Byzantine
cathedral nor the western reaches of the site, where two Byzantine fortifications are
located, have yet been excavated.72 More important, however, it would be premature to
assume that dispersed settlement was necessarily incompatible with urban vitality.
Anthony Bryer has suggested that although medieval Byzantine towns often took the
form of an agglomeration of villages separated by fields and orchards, these discrete
settlements were not entirely independent or self-sufficient, but instead performed
quite distinct functions that contributed to the overall prosperity of the town as a
whole. Within this loose urban matrix, certain sectors could be devoted to specialist
industries or commercial enterprises; others were dominated by residents with
monastic or military functions; others by particular religious or ethnic groupings. This
flexible, disaggregated structure enabled the inhabitants of these different component
1992), ii, 298-308 figure 2 summarises the copper coin finds in graphical form from a variety of ninth- to
eleventh-century archaeological sites including those in western Asia Minor.
Foss, Byzantine and Turkish Sardis, p.60
174
villages to identify themselves with the city as a whole, and to contribute to a level of
Indeed, elsewhere in coastal western Asia Minor, literary evidence from the tenth and
eleventh centuries suggests that urban sites could act as genuine centres of commercial
exchange for extensive agricultural hinterlands. For example, Ephesos was populated
in the tenth and eleventh centuries by a cosmopolitan population of Jews, Bulgars,
Arabs and Armenians. Traders included booksellers, painters, perfumers and
translators. Monks from Mount Latros journeyed to the city to collect essential supplies
and luxury goods. The remains of early medieval rural settlements in the countryside
around Ephesos, including a farmhouse and an eleventh-century unwalled site at
Arvalia, suggest that the city on the coast was supplied from its agricultural
hinterland.74 A Jewish marriage contract from the Cairo Geniza archive, dated to
1022, indicates that Jews lived at Mastaura, close to modern-day Nazilli in the
Maeander Valley, during the reign of Basil II. 75 A Jewish community is also attested in
Attaleia on the south coast during the same decade.76 Further north at Nikaia the life of
the ninth-century monk Constantine, a convert from Judaism, confirms the existence of
a Jewish community in this town too. 77 Seals of customs officials (kommerkiarioi)
from Nikaia and xenodochoi (officials in charge of hostels where merchants were
A.Bryer, 'The Structure of the Late Byzantine Town: Dioikismos and the Mesoi\ in A.Bryer and
H.Lowry (eds.), Continuity and Change in Late Byzantine and Early Ottoman Society,
(BirminghamAVashington, 1986), pp.263-279.
Foss, Ephesus after Antiquity, pp. 120-34; Vryonis, Decline ofMedieval Hellenism, pp. 10-11.
T.Reinach, 'Un contrat de mariage du temps de Basil le Bulgaroctone', in Melanges offerts a
M.Gustave Schlumberger, (Paris, 1924), ii, 123.
76
Foss, 'Cities of Pamphylia', pp.9-10; see below, p. 194, for Attaleia's importance as a trading centre
77
Vryonis, Decline of Medieval Hellenism, p. 12; BHG 370.
175
"752 __
accommodated), also indicate the commercial importance of the town. During the
reign of Basil II it was endowed with large grain silos, big enough to feed the garrison
of the city for two years. Awed by the size of these silos Bardas Skleros was persuaded
not to undertake a full scale siege against the city during his first revolt. Although he
was tricked into believing that the city was well provisioned by the fact that the silos
had been filled with sand and covered with grain, the sheer size of the storage facilities
in question indicate that the city must have been supplied with basic foodstuffs from its
hinterland, rather than by crops grown on a small scale within the urban site itself.79
Yet, while the economy of the littoral of Asia Minor expanded during the tenth and
eleventh centuries, it would be unwise to overstate the extent of this growth. This is
because retraction during the seventh to ninth centuries may not have been so
pronounced here as in other regions of Anatolia. A series of literary references, for
example, indicate the local economy in these regions enjoyed a certain vitality even
during the eighth and ninth centuries. When the Anglo-Saxon bishop Willibald visited
Ephesos in 721, the town was still a thriving port.80 In 795 the fair held at the city
yielded at least one hundred pounds for the imperial authorities.81 Meanwhile, on the
Black Sea coast, the early ninth-century life of the eighth-century Saint George,
metropolitan of Amastris, includes allusions to Greek merchants from the Euxine
176
travelling beween different local port towns. In the later ninth century Niketas the
Paphlagonian described Scythian, probably Rus, merchants from the Crimea travelling
to the Pontus to trade.82 Archaeological evidence also suggests that the lower cities of
many Late Antique urban sites in these coastal and river valleys regions continued to
be occupied during this period, and that flight to hilltop sites was a phenomenon
connected more to Turkish invasions in the later eleventh and twelfth centuries than to
Arab incursions in the eighth and ninth centuries.83 The reason for a degree of
economic prosperity was almost certainly the fact that these regions were more
sheltered from Arab raids than other districts of Anatolia. The towns and hinterlands of
the north coast were always protected from raiding parties by the Pontic mountains. 84
places such as Dorylaion and Kotyaion during the seventh and eighth centuries, which
were located along the edge of the western plateau. Although not immune from
invasion, these regions were less liable to be ravaged every year.85 Raids conducted by
dualist Paulicians from the Anti Taurus regions which reached as far as Ephesos in 867
seem to have been unusual. Certainly, they provoked a swift imperial response. In 871
R9
Vryonis, Decline ofMedieval Hellenism, p. 15.
83 Whittow, 'Social and Political Structures in the Maeander Region', pp.26-244 passim, but especially
p.245; Barnes and Whittow, 'The Survey of Medieval Castles of Anatolia: Mastaura', p. 131.
84 See above, p. 158
or *
RJ.Lilie, Die byzantinische Reaktion aufdie Ambreitung der Araber: Studien zur Strukturwandlung
des byzantinischen Staates im 7. und 8. Jahrhundert (Munich, 1976), pp.325-6, n.93; Whittow, Making
of Orthodox Byzantium, p. 178.
177
ii. Anatolian plateau
On the western reaches of the plateau, enemy incursions had been more frequent and
more intense than in the river valleys, continuing indeed until the first decade of the
reign of Romanos Lekapenos (920-944).** However, once Arab raids ceased, a very
visible economic recovery ensued. Good evidence for this expansion comes from the
material remains of tenth- and eleventh-century churches found at sites and museums
across the transitional zone between plateau and river valleys, and on the western
plateau itself. For example, at Sebasteia, a Byzantine urban site located in Phrygia fifty
kilometres south-west of Afyon (Byzantine Akroinos), two sixth-century basilicas were
rebuilt in the tenth century.88 At a neighbouring site, where the stone for the Sebasteia
churches was quarried, there is a rock-cut church with tenth-century frescoes.89 At
Beycesultan in the same region of Phrygia, church fittings ressembling those at
Sebasteia were found with pottery sherds which can be dated to the same period.90
Spolia of church furnishings from the eleventh century have been discovered at the
hilltop site of Tabala on the road between the head of the Hermos valley and the
Phrygian plateau.91 The fragment of an epistyle dated to the reign of Romanos
Lekapenos has been found reused in a Muslim sufi tekke at Afyon; an architrave dated
o>*
178
to 1063-4 was discovered in the region of Synada.92 Furthermore, similar undated
pieces abound at museums in the cities on the western reaches of the plateau such as
Afyon (Akromos), Usak, Kutahya (Kotyaion), Eskiehir (Dorylaion), and Yalva?
(Pisidian Antioch).93
However, the best evidence for increased prosperity in western Anatolia comes from
the excavations at Amorion, the capital of the Anatolikon theme, which despite its sack
by Arab armies in 838, experienced a significant revival during the tenth and eleventh
centuries. The late antique three-aisled basilical church in the lower city was
completely rebuilt as a cross-domed basilica, and decorated with terracotta tiled and
opus sectile floors, plastered and frescoed walls, tenth-century impost capitals, stained
glass windows, and extensive ceiling mosaics. In the style and quality of the marble
furnishings and its glass windows the Amorion basilica has been compared to several
churches from tenth- and eleventh-century Constantinople built by figures of great
i -i
prestige at the imperial court. 94
The prosperity of Amorion is evident in other forms of material evidence. Both the
lower and the upper city were occupied after 838: finds of stray copper coins from the
ninth to eleventh centuries have been made at locations in both sites; glass bowls and
92 C.Barsanti, 'Scultura anatolica di epoca mediobizantina', Milion: Studi e richerche d'arte bizantina,
(Rome, 1988), p.281. Two other dated epistyle fragments, from 960 and 967 respectively, have been
found further west, at Ionia (960) and in a mosque at Manisa (967).
93 Ibid., pp.275-307. Firath, 'Une eglise byzantine a Sebaste', p. 165. When Cyril and Marlia Mango
visited the museum at Afyon in the 1980s they discovered more than thirty pieces of carved marble from
the tenth and eleventh centuries which may have come from as many as twenty separate churches
(J.D.Howard-Johnston and N.Ryan, The Scholar and the Gypsy (London, 1992), p.223). These
observations were confirmed by Professor Mango in conversation.
The marble carving ressembles that in the tenth-century Constantinopolitan church of Constantine Lips;
the stained glass, the early twelfth-century imperial Komnenian church, Christ Pantokrator (Lightfoot,
179
bracelets of the same period, some of extremely good quality, have been found
everywhere; a pottery workshop was located in the upper city.95 Discoveries of seeds
of wheat and grape pips during archaeo-botanical investigations at the site indicate that
the inhabitants of the city had access to products which Leo of Synada claimed could
not grow in these interior regions. Signs of prosperity in the hinterland around
Amorion indicate that most of the food products consumed in the city were cultivated
locally rather than being imported. Not far from the main urban site excavators have
surveyed three rock-cut church complexes and two churches constructed from stone. 96
Further east the remote reaches of the plateau and the isolated plains (ovalar) of the
Anti Taurus had been the areas most severely affected by annual Arab raids from the
later seventh to the early tenth centuries. As Byzantine, Arab, and Persian sources
indicate, this region had become sparsely populated, characterised by isolated rural
communities which grew crops and raised flocks at a purely subsistence level. The
tenth-century geographer Ibn Hawkal reflected that, ".... most of their [Byzantine]
country comprises mountains, citadels and fortresses, troglodyte villages and hamlets
with houses dug in the rock or buried underground." From as far afield as tenth-century
Afghanistan the author of the 'Hudud al Alam' noted that, "...... in the days of old
cities were numerous in Rum [the Byzantine empire], but now they are few. Most of
the districts have ........ an extremely strong fortress on account of the raids. To each
village apertains a castle, where in time of flight they take shelter". 97 While Arab and
Persian accounts may exaggerate the demise of Byzantine Anatolia for their own
'Amorium', Anatol Stud 44 (1994), pp. 120-2; 45 (1995), pp. 128-131; Anatolian Archaeological Report
1 (1995), p.6).
Lightfoot, 'Amorium', Anatol Stud 43-6 (1993-6).
96 Idem., AnatolStud43 (1993), pp.151-3; 45 (1995), pp.126-7, 134-6.
180
propagandist purposes, some support for their analysis comes from contemporary
sources written in Greek within the empire itself. For example, the tenth-century
military manual 4On Skirmishing' instructed Byzantine military commanders serving
in the eastern regions of the plateau to ensure that local villagers and their flocks were
QO
Yet, here too, both literary and material evidence suggests that as Arab raiding activity
began to wane in the tenth century, and the Byzantine eastern frontier moved beyond
the Taurus and Anti Taurus ranges, the economy began to expand. The most visible
evidence of that expansion comes from western Cappadocia. Here, in a triangle of land
between modern-day Nigde, Nevehir and Aksaray, lie a series of valleys gouged out
of the lava produced by the isolated volcanoes of this region such as Mount Argaios
QQ ___
(Erciyes Dagi). The cliffs and outcrops of these valleys are riddled with hundreds of
churches, chapels, monasteries, and hermits' cells, cut into the soft volcanic tufa.
Although most of these churches and complexes have to be dated according to the style
of their architecture and their wall paintings, criteria which frequently divide art
historians, it is generally agreed that between the mid-seventh and mid-ninth century
very few churches or monasteries were either excavated or decorated. However, from
the second half of the ninth century onwards the excavation and decoration of churches
and associated complexes once again began to pick up. The few surviving dated
inscriptions support this chronology. Of the fourteen churches which bear dated
97 Ibn Hawkal, pp. 194-5; HududalAlam (The Regions of the World: a Persian Geography 372 A.H./982
A.D.), trans. V.Minorsky, 2nd edn. (London, 1970), pp. 156-7.
98 Skirmishing: G.Dagron and H.Mihaescu, Le Traite sur la guerilla (De velitatione) de I'empereur
Nicephore Phocas (Paris, 1986), pp.51-3; Hendy, Studies, pp.28, 54 ff, 557; Harvey, Economic
Expansion, pp.30-2, 149-50; Kaplan, Les Hommes et la terre, pp.23-4; Howard-Johnston, 'Byzantine
Army', chapter 3.
99 See Map 4, for sites located in western Cappadocia
181
inscriptions, eight were painted in the tenth and eleventh centuries, the rest in the
thirteenth century. None was decorated before the minority of Constantine VII (913-
19) ioo
Furthermore, it is clear that the most elaborately carved and expensively decorated
churches, such as New Tokah Kilise and the so-called Column churches located in the
vicinity of the Goreme Valley ten kilometres east of Nevsehir, date from the mid-tenth
to the mid-eleventh century. 101 A similar date has been applied to a series of little
understood complexes scattered across the rocky valleys, which some scholars have
seen as monasteries, others as the homes of the local gentry or even Byzantine hans
(hostels for travellers). The characteristics of these large complexes usually include a
three-sided courtyard, an inverted "T" shaped vestibule and reception hall, and a cross-
in square side chapel. 102 While it is difficult to know whether all the Cappadocian rock
100 J.Lafontaine-Dosogne, 'Nouvelles notes cappadociennes', B 33 (1965), pp. 182-3, lists the following
five dated churches from the tenth and eleventh centuries: Tavanli kilise near Urgiip (Constantine VII:
probably 913-19); the Great Pigeon House at Cavusm (reign of Nikephoros Phokas: 963-9); Direkli kilise
in the Ihlara valley (Basil II/Constantine VIII: 976-1025); Saint Barbara in the Soganli valley (either 1006
or 1021); Karabas. kilise in the Soganli valley (1060-1). To these may be added: Ayvali Kilise (Saint John
at Giillu Dere) near Cavusin (913-9), Saint Michael in the Ihlara valley (either 1025-8 or 1055-6)
(N.Thierry, "Notes critiques a propos des peintures rupestres de Cappadoce", REB 26 (1968) pp.357)
and Egri Ta kilise in the Ilhara valley (921-7) (N.Oikonomides, 'The Dedicatory Inscription of Egri Ta
Kilisesi', Harvard Ukranian Studies 1 (1983), pp.501-6). Furthermore, none of the three churches with
datable graffiti inscription predates the eleventh century: Ballik kilise in the Soganh valley (1031); Kizlar
kilise near Goreme (1055); Saint Eustathios near Goreme (1148-9) (R.Cormack, 'Byzantine Cappadocia:
the Archaic Group of Wall Paintings', Journal of the British Archaeological Association Third Series 30
(1967), p.22)
101 Tokah kilise is a church of three phases. The first phase was excavated in the later ninth or early tenth
century. The second phase, Old Tokali kilise, was a single-aisled basilica decorated in the early tenth
century. Haifa century later, a transverse nave was then excavated behind the single-aisle church to form
a new construction, New Tokah kilise. This final phase was decorated using luxury materials such as gold
and lapis lazuli. The high quality of the frescoes of New Tokah suggests that an artist from either the
capital or from the provincial centre of Kaisareia was employed (Wharton-Epstein, Tokah Kilise, passim);
see above, chapter three, pp. 126-7 for the connection of this church with the Phokas family.
102 Rodley, Cave Monasteries of Cappadocia, pp. 11-120, sees most of the courtyard complexes as
monasteries, although she believes that several courtyard complexes at Acik Saray may be hans (ibid
pp. 121-50). These seven complexes are located on a key north-south route running through Nigde and
Nevsehir, and are thus ideally placed to act as hostels for merchants. Furthermore, since they only have
182
units of this period are monasteries, particularly these so-called courtyard complexes,
it is the rate at which rock installations of all varieties were produced that should be
stressed. For example, more than thirty churches, chapels and associated domestic
complexes were excavated in and around the small valley of Goreme in the tenth and
eleventh centuries. 103
Since relatively few of the churches and other buildings bear inscriptions identifying
their donors, it is difficult to know whether the proliferation of churches, monasteries,
and other complexes in western Cappadocia reflects a boom in the local economy
during the tenth and eleventh centuries, or investment from outside sponsors. Lyn
Rodley has suggested that many of the monastic complexes in this region were
sponsored by officers in the Byzantine field army who served on the eastern frontier.
Their reasons, according to Rodley, for undertaking this action were manifold: to
express personal piety; to provide themselves with a place of shelter on their way to
the frontier, where they might also take advice from monks who were their spiritual
advisors; to build themselves mausolea where they could be buried when they died. 104
Yet, the foundation of new bishoprics in western Cappadocia during the tenth century
suggests that the proliferation of rock-cut architecture also reflects significant growth
in the local economy. In the early tenth century, three new bishoprics, suffragans of
Kaisareia, were established in the rocky valleys in the region of modern-day Urgiip: at
Dasmendron (Takinpaakoy), Sobessos (ahinefendi/Suve) and Hagios Prokopios
two churches between them, it seems unlikely that they are monasteries. Robert Ousterhuit has recently
proposed that many of the courtyard units of western Cappadocia may be houses belonging to the local
gentry, that is to say the owners of small estates in the immediate locality ('Survey of the Byzantine
Settlement at anh Kilise in Cappadocia: Results of the 1995 and 1996 Seasons', DOP 51 (1997),
pp.301-6). However, his findings are as yet highly provisional.
103 Rodley, Cave Monasteries of Cappadocia, pp. 160-183
104 Ibid., pp.250-54.
183
(Urgiip itself). 105 It is noteworthy that many of the churches excavated and decorated in
the second half of the ninth century and the first half of the tenth, are to be found in the
main drainage valleys contained within the geographic area covered by these three
bishoprics. 106 The establishment of churches, many without associated monasteries, in
reasonably wide fertile valleys close to signficant water supplies, clearly suggests that
they were excavated to serve local villagers rather than monks. A mid tenth-century
inscription on a tomb at the modern village of Zelve near (^avusin which refers to a
chorepiskopos (rural bishop) is further evidence of the development in Cappadocia of a
church hierarchy serving a lay population. 107
In eastern Cappadocia economic expansion was driven by the arrival of settlers from
Armenia, Mesopotamia, and northern Syria. Among these migrants the Armenians
appear to have been the most numerous, part of a much wider later ninth- and tenth-
century phenomenon in which Armenians of all backgrounds migrated to live and
1AR
work within the Byzantine empire. The first signs of migration to the eastern reaches
of the plateau from western Caucasia came in the first decade of the tenth century. At
this time the emperor Leo VI (886-912) was approached by five Armenians with the
request that they should be appointed to military commands over a variety of passes
and small plains in the Anti Taurus mountains west of the emirate of Melitene. The
initial success of the scheme was limited as one by one these commanders were killed
by local Arabs, or exiled for rebelling against the emperor. However, one of their
105 Notitiae. Notitiae Episcopatuwn Ecclesiae Constantinopolitanae, ed. J.Darrouzes (Paris, 1981),
pp.66-7, 274; Hild and Restle, Kappadokien, pp. 116-7. Although these three bishoprics are not
mentioned in later tenth-century notitiae, by the 1030s yet another new bishopric had appeared in the
rocky valleys at Matiana (modern day Goreme): Notitiae pp.109, 324; Hild and Restle, Kappadokien,
p. 116.
106 Giovannini, L., Arts of Cappadocia (Geneva, 1971), plans 1-4
107 Thierry, Haul moyen-dge en Cappadoce, pp.329-33; idem, 'Les enseignements historiques', p.518.
184
number, Melias, was more fortunate. Initially named as tourmaches of Euphrateia,
Trypia and the desert, by 913 he had occupied and fortified the ancient kastron of
mountain top at Tzamandos, which also became a kleisoura, and developed the
previously deserted area of Symposium into a tourma. By 916 Lykandos had been
upgraded to a theme, and Melias was its first full military governor, or strategos. A
patrikios at the time of his appointment, he was soon raised to the rank ofmagistros "
for his many and infinite feats of daring against the Saracens", which included
campaigning against the emirate of Melitene in the late 920s and early 930s. 109
prosperity to those "desert" regions under his command. When Bardas Skleros
marched westwards in 976 or 977 from Melitene, he fought imperial forces on the
plain of Lykandos (Elbistan) which lay below the main castle built by Melias. Skylitzes
notes that the alternative name for Lykandos was "Lapara" because of the richness (TO
Amapov) and the abundance (TO Tro^opov) of the region. Next, having won an
"city (polis) lying on a precipitous rock, well populated and dripping with wealth." 110
At first sight this ekphrasis seems unjustified. If the Byzantine fortress of Tzamandos
185
is to be identified with the site of the extant Seljuk castle of Kuskalesi, it is difficult to
see how it could be described as a town. 111 The site is far too elevated and barren to be
anything more than a mountain-top fortress. However, the "city" to which Skylitzes
refers may have been located below the fortress. Tzamandos was raided by Sayf al-
Dawla in 957 and was described in Arab sources as a "city". 112 By the early tenth
century it was the site of a Chalcedonian bishopric; by 954 a Syrian metropolitan had
taken up residence there. 113
186
Cappadocia. 116 But the arrival of the Artsruni clan and their fourteen thousand
followers was clearly prefaced by a more low-level, but long-standing, migration of the
Armenians to eastern central Anatolia. As early as the second half of the tenth century
Leo the Deacon alleged that the land of the Armenians began at Lykandos. 117 Syriac
sources note that Armenians were migrating to Cappadocia in large numbers by the
beginning of Basil's reign. 118 Indeed, this long-term migration may have dictated the
locations where the Artsruni were granted estates by Basil. For example, the Armenian
princes were given new territories near or in Sebasteia, Larissa and Tzamandos, sites
where Armenian settlement was already firmly established. Since the Artsruni were
also granted lands at Kaisareia, Gabadonia, arid Abara, it is possible that significant
numbers of Armenians were already in residence in these areas by the turn of the tenth
century, before the great migrations from Vaspurakan. 119 Epigraphical evidence from a
rock-cut church in western Cappadocia confirms the presence of a large Armenian
population by the early tenth century at one of the future Artsruni sites. According to a
Greek inscription in the church of Ayvali Kilise (Saint John at Giillii Dere) (913-20),
the district of Gabadonia located to the east of Kaisareia was known as "Great
Armenia". 120
187
Another group of incomers to the eastern plateau were the Banu Habib, a tribe living
near Nisibis in northern Mesopotamia, who became so disaffected with the level of
taxation demanded by the Hamdanids of Mosul, that they defected to the Byzantines,
converted to Christianity, and were settled at an unknown location on the western side
of the Anti Taurus. During the next twenty years they participated in Byzantine raids
aristocratic patrons. Meanwhile, although it is argued that Armenians often appropriated sites built by
Greek Chalcedonians in the ninth and tenth centuries, the evidence for these claims tends to come from
much later medieval historians. It is, therefore, difficult to know the scale of such appropriation in the
tenth and eleventh centuries (Thierry, infra, passim).
121 Ibn Hawkal, pp.205, 214; M.Canard, Histoire de la Dynastie des H'amdanides de Jdzira et de Syne
(Paris,1953), pp.737-9. Haldon and Kennedy, 'Arabo-Byzantine Frontier', p. 101, characterise the Banu
Habib as "semi pastoralists". The later tenth- and eleventh-century fate of the Banu Habib is not clear. It
has been suggested that the Amiropoulos family were direct descendants of the Banu Habib (J.C.Cheynet,
"L'apport arabe a 1'aristocratie byzantine des Xe-XIe siecles", ByzSlav 61 (1995), p. 141, n.34); or that Ibn
Baghil, a Skleros supporter active at Antioch, was a member of the family (Seibt, Die Skleroi, p.46)
However, neither of these suggestions can be corroborated
188
authorities. They were also given six nomismata to buy oxen, and fifty-six modioi of
grain for sowing. 122
East of the Anatolian plateau the Byzantine military advance in the third quarter of the
tenth century had caused significant damage to the local economy of the former
Muslim emirates in Cilicia, Syria and northern Mesopotamia. During his campaigns
against Melitene in the 920s and 30s the domestikos of the scholai, John Kourkouas,
repeatedly "destroyed the surrounding hamlets and villages by fire." 123 The same
strategy was adopted by Nikephoros Phokas in the 960s. By burning and destroying
crops on a yearly basis and capturing the people and animals of the villages, ".... he
seized by this manner of action all the frontier towns of Syria and Mesopotamia" 124
Tenth-century Arab geographers and historians were insistent that frontier regions
either controlled by the Byzantines, or subject to regular Greek invasions, were by the
third quarter of the tenth century "ruined". The policy of forcing Muslims who would
When Tarsos fell to Nikephoros Phokas in 965 many Muslims left the city, first for
Antioch, and then in 969, when that city succumbed to imperial attack, to Balanias
189
Yet, there are signs that the negative impact of these conquests on the economy was
short-lived. The migration of Armenians to Cilicia and Syria, as well as to Cappadocia,
is documented by Michael the Syrian, suggesting that while there were outgoing
populations, there were also incomers. 126 At some point between 972 and 992, during
Growth in monastic activity on the fertile terraces of the mountains around Antioch,
points to a similar recovery of the countryside of northern Syria. Monks of all
190
denominations and nationalities seem to have been involved. When Romanes III
campaigned against Aleppo in 1030 he recruited "heretic" anchorites (presumably
either monophysite Syrians or Armenians) into his armies from the mountains around
Antioch and in the Orontes valley. 132 During the reign of Basil II Armenian
communities were attested in the Amanos mountains. 133 Eleventh-century manuscript
132 Canard, 'Les sources arabes de 1'histoire byzantine aux confins des Xe et Xle siecles', p.309.
133 Matthew of Edessa: Armenia and the Crusades in the Tenth to Twelfth centuries: the Chronicle of
Matthew of Edessa, trans. A.E.Dostorian (Lanham/New York/London, 1993), 47-8
134 W.Z.Djobadze, Materials for the Study of Georgian Monasteries in the Western Environs ofAntioch-
on-the-Orontes (1976, Louvain); idem., Archaeological Investigations in the Region West of Antioch-on-
the-Orontes (Stuttgart, 1986). At Saint Baarlam's monastery on Mount Kasios. 65 kilometres south-west
of Antioch, three anonymous copper folles of issues Al and A2 have been discovered (Archaeological
Investigations, pp.218-9.) These issues are usually dated to the period 969-76 (Al) and 976-1028 (A2):
V.Ivanisevic, 'Interpretations and Dating of the Folles of Basil II and Constantine VIII - the Class of A2',
ZJW27(1989),pp.37-39).
135 Ibn Butlan, p. 56. Inscriptions demonstrate that Greek and Georgian monks were resident at the
monastery of Symeon the Younger Stylite. Epigraphical evidence also points to an eleventh-century
restoration of the site, possibly sponsored by Bagrat IV, king of Iberia and Abkhazia (Djobadze,
Archaeological Investigations, pp.204-11). However, Greek sponsorship was also important since the
patriarch Theodore III (1034-42) founded a library of 420 books there (Djobadze, Archaeological
Investigations, p.59). By the end of the eleventh century Saint Nikon of the Black Mountain noted the
presence of Chalcedonian Armenian monks at Saint Symeon's monastery (Garsoian, 'Armenian
Integration', pp. 106-08). Several eleventh-century seals issued by the monastery survive: Cheynet,
Morrisson and Seibt, Sceaux byzantins: Seyrig, no.288; V.Laurent, Le corpus des sceaux de I'empire
byzantin, Vol. V, L 'Eglise (Paris, 1963-72), no. 1559.
191
Furthermore, it is clear that while there were incoming populations, indigenous
inhabitants, including Muslims, also often remained in the newly conquered territories
of the east. Hints that conquered regions were not entirely denuded of their local
populations occur even at times of surrender. When John Kourkouas seized Melitene
in 934 he set up two tents. Atop one tent was a cross. Those Muslims who wished to
have their goods and families returned to them were told to gather at this tent; those
who remained loyal to Islam were to go to the other tent. The majority of the
population went to the tent with the cross, thus chosing to stay at Melitene and convert
to Christianity. 136 Moreover, in some reconquered regions conversion to Christianity
may have been unnecessary. Many local inhabitants chose to stay in northern Syria
happy to pay the local head tax demanded of those who were not Christians. 137 Many
former inhabitants of Tarsos chose to return to the city once the price inflation caused
by a lack of supplies in the immediate aftermath of Byzantine conquest had
1 ^ft _
subsided. The early eleventh-century general Nikephoros Ouranos recommended
relatively mild treatment for Muslims who surrendered voluntarily: they should be
138 Yahya, PO 18, p.797; Dagron, 'Minorites ethniques', pp. 182-4, acknowledges that indigenous
Muslims either stayed in their homes or returned after the Byzantine conquests were completed.
However, he believes the phenomenon was relatively rare, and that most Muslims were either taken
captive and sold into slavery, or forced to flee. Yet, most of the figures Dagron uses in referring to the
mass deportation or enslavement of Muslims, are derived from the fourteenth-century Syrian chronicler
Bar Hebreus, who customarily inflates exponentially statistics of all varieties. Thus, in another context,
Bar Hebreus reports that in 994 ten thousand Byzantine prisoners were taken to Egypt after a heavy
defeat by Fatimid armies (Bar Hebreus p. 181). The Egyptian historian Maqrizi, using eye-witness
evidence, which is now lost, from the contemporary Egyptian annalist al-Musabbihi, presents a figure of
two-hundred-and-fifty (Forsyth, "The Chronicle of Yahya ibn Sa'id", p.491).
192
allowed to keep their possessions and their leaders among the urban elite should
130
receive presents.
Byzantine Syria, Ibn Butlan noted that the village of Imm had a mosque as well as four
churches. At the coastal town of Laodikeia (Lattakiah) an old pagan temple previously
reused as a mosque had been reconverted into a church. However, there was an
alternative mosque for the local Muslim population who also retained their own qadi
(judge). Although Ibn Butlan's indicates that life for Muslims was not easy - at Imm
the call to prayer was conducted secretly, and in Laodikeia it was smothered by the
ringing of church bells - there is little evidence of sustained resistance to Byzantine
rule. 140 The only revolt by local populations occurred at Laodikeia in 992, at the same
time as armies from Fatimid Egypt attacked northern Syria. 141 Indeed, Basil II himself
seems to have attracted only favourable reports from Muslim writers. An encomium
contained in the eleventh-century Iraqi chronicle of al-Rudhrawari explicitly praises
Basil n for his justice and affection for Muslims, his willingness to keep out of Muslim
territory and his kindness to Muslims who entered his. 142
VL Trade
139 Ouranos Taktika (a). J.A.de Foucault, 'Douze chapitres inedits de la 'Tactique' de Nicephore
Ouranos', TM 5 (1973), pp.298-9; E.McGeer, Sowing the Dragon's Teeth: Byzantine Warfare in the
10th Century (Washington D.C., 1995), p. 158. Indeed, Ouranos reserved the harshest treatment for local
Christians. If Armenians, Syrians, or Arab Christians within a besieged city did not defect to the
Byzantines, they could expect no mercy from the imperial armies, and would be beheaded.
140 Ibn Butian, pp.54, 57
141 Yahya,/)O23,p.439
193
i. Coastal plains and river valleys
The greater prosperity of the eastern half of the Byzantine empire during the tenth and
eleventh centuries cannot be divorced from trade, whether local, regional, or
international. A variety of sources reflect on the importance of regional trade on the
littoral of western and northern Asia Minor. According to the De Administrando
Imperio Cherson in the Crimea was supplied in the tenth century with
"grain.....wine......[and] other needful commodities and merchandise" by ships sailing
from a variety of sites along the southern Black Sea coast: Paphlagonia, the theme of
the Boukellarion, and the Pontus. 143 Turning to the west coast of Asia Minor, the life of
Athanasios relates how monks from Mount Athos sailed from the Chalkis peninsula to
Smyrna for supplies. 144 The wealth of Attaleia located on the southern coast of Asia
Minor was drawn from its position on regional and international trade routes. Seals of
customs officials (kommerkiarioi) with joint responsibility for Attaleia and Cyprus
indicate the importance of trade between these two locations. 145 When Saint Lazaros of
Mount Galesion visited Attaleia in the eleventh century he was almost sold as a slave
into the hands of an Armenian sea captain. 146 The fact that Leo of Synada's diocese on
the western plateau was supplied with foodstuffs from Attaleia indicates that trade also
142 Al-Ruhdrawari,p.ll9
143 ZW, p.286.
144 Vryonis, Decline ofMedieval Hellenism, p.l 1
145 Tenth- to eleventh-century seals of officials connected to the joint kommerkia of Attaleia and Cyprus
include John the semiophoros of the kommerkia (J.Nesbitt and N.Oikonomides, Catalogue of Byzantine
Seals at Dumbarton Oaks and in the Fogg Museum of Art, Vol.2 (South of the Balkans, the Islands,
South of Asia Minor) (Washington D.C., 1994), no.64.2) and Leo a kommerkiarios (Schlumberger,
Sigillographie, p.305). Some kommerkiarioi were responsible for Attaleia alone (A.Szemioth and
T.Wasilewski, Sceaux byzantins de Musee National de Varsovie (Warsaw, 1966), no.52; Nesbitt and
Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, ii, no.64.1). There were also
kommerkiarioi further east along the south coast at Seleukeia (Schlumberger, Sigillographie, pp.271-3).
A small number of the documents from the voluminous Cairo Geniza archive, (from which the trading
activities of a network of eleventh-century north African Jewish traders can be reconstructed) refer to
ports along the southern shores of Asia Minor including Attaleia, Seleukeia, and Tarsos (S.J.Goitein, A
Mediterranean Society (5 vols., California, 1967-93), i, 214).
194
travelled from this city by land. Leo's supplies almost certainly came via the route
which cut through the southern mountains of Asia Minor, continued past the Pisidian
lakes, and then stretched along the western edge of the plateau. 147 Agricultural products
from coastal Asia Minor may also have helped provision Byzantine enterprise further
afield during the tenth and eleventh centuries. The seal of an eleventh-century
horreiarios (an imperial official probably connected to grain supplies) from Kios in
Bithynia, has been discovered in the region of modern-day Serbia between the Danube
and the Sava rivers. It is is tempting to see in this seal the physical proof of the
injunctions of a contemporary military manual dealing with warfare in the Balkans that
Byzantine armies operating in Bulgaria should be provisioned by food supplies
produced inside the empire. 14*
However, the biggest market for the coastal regions of northern and western Asia
Minor was Constantinople. The grain trade was particularly important, with cereals
transported in the main by sea. 149 There are tenth- and eleventh-century seals of
horreiarioi attached to sea ports along the western and northern shores of Asia
Minor. 150 Although the precise responsiblities of these imperial officials are unknown,
146 Malamut, Sur la route des saints byzantins, p.41; Foss, 'Cities of Pamphylia', pp.8-9.
147 See above, p. 161, for this route
148 L.Maksimovic and M.Popovic 'Les sceaux byzantins de la region danubienne en Serbie', in
N.Okonomides (ed.), SBS 3 (Washington D.C., 1993), pp. 116-7, no.l. The manual in question is
'Campaign Organisation and Tactics': Dennis, Three Byzantine Military Treatises, p.304; see above, p.8,
n.22. Grain was sent from Anatolia to a variety of military frontiers including the Danube in Late
Antiquity (Mitchell, Anatolia, i, 250).
149 It has been widely argued that there was a rapid increase in the demand for grain in Constantinople
between the tenth and eleventh centuries (Hendy, Studies, pp.47-9; Angold, 'Medieval Byzantine 'City",
pp.6-14; P.Magdalino, "The Grain Supply of Constantinople in the Ninth to Twelfth Centuries', in
C.Mango and G.Dagron (eds.), Constantinople and its Hinterland (Cambridge, 1995), pp.35-48).
150 Amisos (Zacos, Byzantine Lead Seals II, no.611; Cheynet, Morrisson and Seibt Sceaux byzantins:
Henri Seyrig, no. 153; Dumbarton Oaks Unpublished F1856 and 55.1.2549); Amastris (Dumbarton Oaks
Unpubttshed 55.1.2408); Nikomedia: (Nesbitt and Oikonomides, Catalogue of Byzantine Seals at
Dumbarton Oaks, iii, no.83.2); Opsikion: (Dumbarton Oaks Unpublished F1390); Smyrna
195
it is likely that they were connected with supplying grain to the capital. 151 The Miracles
of Saint Eugenios indicate the importance of the traffic of grain from the Pontus region
to Constantinople during Basil's reign. While the forces of Bardas Phokas were
attacking Constantinople in 988, Basil n wrote to "all the villages and towns lying
around the Black Sea as far as Trebizond to send a cargo of all sorts of grain to the
capital of cities via ships quickly." 152 Preventing grain reaching Constantinople by sea
from Pamphylia may have been one of the reasons why Bardas Skleros seized Attaleia
during his first revolt. 153 Certainly both Skleros and Phokas went to great lengths
during their respective rebellions to ensure that Constantinople was deprived of grain
supplies from either the Mediterranean or the Black Sea regions. For example, both
rebels tried to seize the town of Abydos, which from its location on the Hellespont
guarded the seaways between the Aegean and the imperial capital. 154
(Konstantopoulos, Molybdoboulla, no. 138; Zacos Byzantine Lead Seals II, no.550; Nesbitt and
Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, iii, nos.35.1-2); Kios: (Nesbitt and
Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no. 50.1; Dumbarton Oaks
Unpublished 55.1.4601-2)
151 S.G.Mercati, 'Oriarios-Horrearius', Aegptus 30 (1950), pp.8-13, discusses the etymology and history
of the term horreion. Horreiarioi were not only found in imperial service, but also within the
management of monastic houses. Although associated with the management of grain supplies, horreiarios
often seems to have been a generic term connected to the supply and storing of all foodstuffs, including
wine and oil.
152 Panagiotakes, 'Fragments of a Lost Eleventh-Century Byzantine Historical Work', pp.348. Phokas'
lieutenant at Chrysopolis, Kalokyros Delphinas, controlled both sides of the entrance to the Bosphoros
from the Black Sea in order to prevent the relief of Constantinople by a grain fleet from Trebizond; see
above, p.44). Bryer and Winfield, Topography and Monuments, p. 18, note the plethora of ports in the
many small deltas along the southern shores of the Black Sea. The tenth/eleventh century seal of a
kommerkiarios of Hieron (located at the mouth of the Bosphoros) and Pontus illustrates the passage of
trade from the Black Sea coast through the Bosphoros to Constantinople (Nesbitt and Oikonomides,
Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no. 72.1).
153 See below, chapter five, pp.234-5; during Ibn HawkaTs description of the prosperity of the region
around Attaleia (see above, p. 164) he reports that the sea journey from Attaleia to Constantinople took
fifteen days with a favourable wind.
154 Skylitzes, pp.322, 324 (Skleros); p.336 (Phokas); Leo the Deacon explicitly states that by his control
of the sea and his seizure of Abydos Bardas Skleros tried to prevent grain supplies getting to
Constantinople (Leo the Deacon, p. 170). For Phokas' unsuccessful attempt to take Abydos see Leo the
Deacon, pp. 173-4. For the importance of Abydos to the passage of basic foodstuff into Constantinople
see Teall, 'The Grain Supply of the Byzantine Empire', pp.104, 119. The growth in trade passing from
the Aegean to Constantinople during the tenth and eleventh centuries seems to be reflected in the material
embellishment of the town of Abydos: in the eleventh century a roof was added to the agora (Angold,
196
/'/. Anatolian plateau
of the western reaches of the plateau during the tenth and eleventh centuries. Leo of
Synada, for example, describes the driving of pigs, asses, sheep, horses, and cattle to
the market at Pylai on the southern shores of the Sea of Marmara, from whence they
were transported to Constantinople. It is usually assumed that the animals seen by Leo
had been brought to this livestock emporium from the western reaches of the
plateau. 155 A commercial relationship between Constantinople and the owners of
livestock on the western plateau is also indicated by the exhortation of the early tenth-
century 'Book of the Eparch' that Constantinopolitan butchers should go beyond the
Sangarios river in north-west Asia Minor to buy sheep at a reasonable price. 156
In Late Antiquity many other products such as grain, wine, wool and salt were
transported to the coasts from the interior of Asia Minor, despite the transportation
costs involved in moving bulk goods of these varieties. 157 In the medieval Byzantine
period the evidence for such traffic is less substantial. However, on the western edge of
the plateau and in the transitional zones it is clear that arable production could
'Medieval Byzantine 'City", pp. 14-5). Customs officials, kommerkiarioi, are recorded in the
sigillographical record at Abydos (Nesbitt and Oikonomides, Catalogue ofByzantine Seals at Dumbarton
Oaks, iii, nos.40.11,12,13,15,17,19; Zacos, Byzantine Lead Seals H, nos.627, 728, 1000) and at places
on the opposite side of the Hellespont such as Hexamilion (Nesbitt and Oikonomides, Catalogue of
Byzantine Seals at Dumbarton Oaks, i, no.54.2)
155 Leo of Synada, letter 54.
Eparch, p. 124.
197
transcend levels of mere subsistence production. Official government memoranda
dealing with the Cretan expedition of 911 record that the western plateau theme of the
Anatolikon supplied this military enterprise with wheat grain and flour as well as
1 ^8
barley biscuit. Unfortunately it is impossible to know whether this grain was a one-
off requisition by the imperial authorities, or acquired from private traders who
regularly tranported grain from the interior to the littoral.
The possibility that the plateau produced other goods commercially is more
speculative. However, it has been argued that brocades, texiles, and woollen products
may have been made in the Anatolian interior. The best evidence for this assertion
comes from the much later testimony of the fourteenth-century traveller Ibn Battuta,
who saw Greek workmen manufacturing carpets at the town of Denizli (Laodikeia)
located in the transitional zone between the Upper Maeander and the western
plateau. 159 Some indirect evidence that this interim zone between river valleys and
plateau produced such goods in the tenth and eleventh century comes from the church
furnishings discovered in this area which were discussed above. Cyril Mango has
suggested to me that some of the fantastical depictions of plants and birds carved on
these furnishings may reflect carpet designs manufactured in the immediate locality.
157 Mitchell, Anatolia, pp. 147, pp.245-7: however, Mitchell thinks that these bulk products were part of
a tax in kind, and that the transportation costs involved in moving them may have been included within
this fiscal due. Hendy, Studies, pp. 5 54-61, argues against the viability of transporting grain in Anatolia.
158 De Ceremoniis, p.659; Foss, 'Cities of Pamphylia', p.7, n.28.
198
early eleventh century the most famous monastic patrons in Melitene were the Banu
Abu Imran who came from Takrit, a town located far away from Byzantium on the
west bank of the Tigris river in Iraq. Their wealth was such that when Basil II arrived
in Melitene in the winter of 1022 with the imperial field army, the Banu Imran were
able to lend the emperor enough money to support the entire Byzantine expeditionary
force while it stayed in the city. 160 If the Banu Imran belonged to the same socio-
199
Mosul was Takrit. 164 If Takritan merchants in Melitene, such as the Banu Imran,
retained links with the Tigris cities, then it could be this trade route which made them
so wealthy. Such wealth and such family connections may also help to explain why the
later tenth- and eleventh-century Marwanid rulers of the Djazira south of the Anti
Taurus mountains chose to expel the Arab populations of Amida and Mayafariqin in
Trade with Muslims was also important for the economy of the eastern frontier. The
first official document recording long-distance, overland trade between the empire and
its Arab neighbours is the treaty of Safar, drawn up between Byzantium and the
northern Syrian emirate Aleppo in 969/70. According to the commercial clauses of this
silk, precious stones, silks, linen, Greek brocades and animals. 166 Such caravans
during the civil wars of the early part of Basil's reign. Indeed, imperial and forces loyal
to Skleros competed for control of the annual caravan as it crossed the Taurus
However, as well as the more exotic long-distance trade between the empire and
neighbouring Arab states, commerce occurred at local and regional levels too. During
the ninth and tenth centuries Byzantine spies in Arab lands had often been
164 Al-Muqaddasi, p. 133 (Amida); Ibn Hawkal, p.219 (Djazirat ibn Umar), p.209 (Mosul). Takrit was
famous for its wool and sesame (al-Muqaddasi, p.111). It was a city populated by many dififerent
Christian sects (Ibn Hawqal, p.223).
165 Stephen of Taron, p.203
166 W.Farag, 'The Truce of Safar A.H. 359', Paper from the Eleventh Spring Symposium held at
Birmingham University (Centre for Byzantine Studies, University of Birmingham, 1977): Professor John
Haldon kindly gave me a copy of this off-print; Canard, 'Les relations politiques et sociales', p. 52.
200
I O
merchants. Ships from Attaleia went to Tarsos, ".... to commit themselves there to
commerce, while their agents scoured the country collecting ......precise
intelligence." 169 At the end of the tenth century the general Nikephoros Ouranos
acknowledged that local trade in essential items was difficult to prevent even during
wartime. Cheese, grain and meat were among the products that Muslim cities under
siege would try to buy from local Byzantine merchants. 170 In addition, the arrival of
large expeditionary forces in a locality created ideal conditions for trade since the army
needed to be fed. When Romanes IQ campaigned in Syria his camp was equipped with
a market. 171 Localised trade between Byzantium and the Arab states was important
enough to be used as a political tool. After the explusion of Mansur ibn Loulou, a
Byzantine ally, from Aleppo in 1016 by the Fatimid governor of Apameia, Basil n shut
the border between Aleppo and Antioch to local trade. The only regional power to gain
an exemption from this prohibition was a local bedouin tribe called the Mirdasids.
However, longer distance trade with Egypt and Syria, much of which must have
travelled by sea, was not affected. 172
201
One of the most important cities within the trading networks which linked the
Byzantine Near East to its Muslim neighbours was Laodikeia (Lattakia). When Ibn
Butlan visited the town he observed the many foreigners who stayed at "hans". The
number of such "foreigners" must have been significant since an organised prostitute
business was run for their benefit. 173 The majority of traders in Syrian Laodikeia must
have been involved in sea trade along the Levant and in the Aegean. Other evidence for
maritime trade between Byzantium and the Islamic east comes from references to
mosques in key ports along the coasts of Asia Minor and Byzantine Greece as well as
in Constantinople itself. Ibn Khurradadhbih indicates that a mosque was to be found in
ninth-century Ephesos. 174 Epigraphical evidence suggests that there was also a mosque
in Athens in the second half of the tenth century. 175 In 1027, only two years after Basil
IPs death, a new mosque was established in Constantinople itself. 176 However, the best
evidence for sea trade between the empire and neighbouring Muslim states during
Basil's reign comes from a wreck which sank off Serce Liman on the southern coast of
modern-day Turkey. On board were copper coins of Basil n and gold quarter dinars of
the contemporary Fatimid caliph al-Hakim (996-1021). 177 The ship itself was carrying
glass to be reprocessed. Although it is unclear where this reprocessing was due to take
place, it is possible that the ship was in transit between the Peloponnese and Egypt.
202
Another important Byzantine city on international, early medieval trading routes was
Trebizond. An important staging-post for travellers journeying to Constantinople from
Caucasia, Mesopotamia and Iraq, it was also a key entrepot where Muslim merchants
collected Byzantine linens, silks, woollens and cloaks. 179 Its prosperity was promoted
by a series of fairs, including that established by Basil I in the ninth century to
celebrate the festival of the local saint Eugenics. 180 Arab geographers were favourably
impressed by the enormous customs revenue of Trebizond. 181 Indeed, so great was this
income that at the beginning of the tenth century it paid half of the twenty-pound
annual salary of the strategos (governor) of Chaldia, the theme (province) of which
1 o^ __
Trebizond was the capital. The impression of a busy trading centre is confirmed by
the large number of seals of customs' officials (kommerkiarioi) that survive from
Chaldia. 183
Given the considerable problems associated with the source materials, any conclusions
about the nature of the Byzantine economy during the early Middle Ages must of
necessity be impressionistic. In particular it is difficult to measure rates of growth in
given regions, or to compare the economic fortunes of one region with another.
203
However, it is clear that the end to incessant Arab attacks during the tenth century
allowed the whole of the eastern half of the Byzantine empire to undergo economic
expansion. This upturn in prosperity was probably at its most marked on the plateau,
the region that had consistently borne the brunt of Arab raids. Furthermore, despite
early depopulation and devastation caused by annexation, the eastern territories beyond
the Taurus and Anti Taurus mountains enjoyed an economic recovery during the later
tenth and early eleventh centuries. Trade was of fundamental importance to increased
prosperity across the eastern half of the empire. On the coasts and the western reaches
of the plateau this trade was closely connected to provisioning Constantinople with
basic foodstuffs. On the eastern plateau and on the eastern frontier, trade with
neighbouring Muslim and Caucasian powers was more important.
However, it is important to draw attention to one key difference between the economic
fortunes of the eastern and western regions of the area of the Byzantine empire under
discussion in this chapter. In western Asia Minor, whether in the coastal plains, the
river valleys, or the western reaches of the plateau, economic revival appears to have
been driven by an indigenous Greek-speaking population. The situation in the eastern
regions of the plateau and on the eastern frontier itself was quite different. Here,
prosperity was driven by non Greek-speaking populations, many of them migrants
from outside the empire. A contemporary explanation for this phenomenon is offered
by the twelfth-century chronicler Michael the Syrian. Commenting on the repopulation
of Melitene, he claims that Nikephoros Phokas had been forced to turn to Syrian
migrants because Greeks were too afraid to settle in these eastern regions lest they
204
should be attacked by Muslim forces. 184 However, a more plausible explanation is that
the emperor was compelled to take this course of action because the core Anatolian
territories of the Byzantine empire simply did not have the necessary resources of
manpower and capital to repopulate the eastern reaches of the plateau, and to revivify
the sophisticated market and trading economy of those regions which lay beyond the
Taurus and Anti Taurus mountains.
205
their lands were liable to be forfeit after three years' absence, if they returned to fight
with imperial armies, they were to be provided with new lands. 187
This, of course, is not to argue that there was no Greek migration eastwards. For
example, by the early eleventh century, Melitene was not only populated by Jacobite
Syrians. It also had a Greek speaking population and a Chalcedonian metropolitan. 188
However, as the history of the city of Edessa indicates, it is unlikely that the numbers
of Greek speakers in the east during the tenth and eleventh centuries ever matched the
annexed by the Byzantines in 1032 seven years after Basil's death. In the next decades
several Greek Chalcedonians migrated to the city and its hinterland. Among them was
Eustathios Boilas, who moved from Cappadocia in 1051 to the hinterland of Edessa,
east of the Euphrates. Yet, when Eustathios compiled his will in 1059, he commented
on the unfamiliar world of foreign languages and alternative religious rites that he had
1 SO
encountered on his arrival in the east. In 1071/2 a contemporary estimate of the
see in the next chapter was a key Byzantine administrative centre on the frontier,
retained a large population of Arabs, Syrians and Armenians, many of whom adhered
to their languages and their faiths throughout the period the city was under Byzantine
186 De Ceremoniis, pp.694-5; Vryonis, Decline of Medieval Hellenism, pp.66-7; see above, p. 188, for
earlier discussion of the Banu Habib.
187 Zepos and Zepos, lus, i, 247-8; E.McGeer, 'The Legal Decree of Nikephoros Phokas Concerning
Armenian Stratiotai\ in T.S.Miller and J.Nesbitt (eds.), Peace and War in Byzantium: Essays in Honor
of G.T.Dennis (Washington, D.C. 1995), p. 126.
188 Michael the Syrian, pp.136, 140-5; Dagron, 'Minorites ethniques', pp.200-205.
189 Testament (Eustathios Boilas): pp.20, 22, 27. Among his neighbours were several Armenians.
190 According to Sawiras ibn al Mukaffa the population comprised 20,000 Syrians, 8,000 Armenians,
6,000 Greeks and 1,000 Latins (Dagron, 'Minorites ethniques', p. 195, n.80).
206
control. For example, although Michael the Syrian lamented that many local Syrian
within the city walls as late 1053/4. Meanwhile, wealthy Syrian inhabitants exercised
influence over local government inciting the envy of the local Greeks. 191 Moreover,
despite living in Antioch for at least three decades of the first half of the eleventh
century, the well-educated doctor and historian Yahya ibn Sa'id appears never to have
learnt Greek. 192
The challenge facing successive tenth and eleventh-century emperors in the eastern
half of the empire, was how to manage the relatively sophisticated economy and
heterogeneous populations of the frontier, in the context of the more limited resources
of the heartland of Asia Minor. This challenge was of central importance to the
administration of Anatolia and the eastern frontier in the tenth and eleventh centuries,
the subject with which the next two chapters are concerned.
191 Michael the Syrian, p. 131; Matthew of Edessa, pp.84-5; see below, chapter six, pp.334-5, for the
activities of Agapios.
192 Yahya (Cheikho), pp.263-4. There are many signs throughout Yahya's history that he was familiar
with Byzantine officials who spoke Greek, but that he did not know the language well himself. One such
example comes from his coverage of the arrival of a famous Edessan relic in Constantinople during the
reign of Romanes HI (1028-34). This relic was the correspondence between Christ and King Abgar of the
Armenians. According to Yahya this correspondence was first translated from Syriac into Greek for the
emperor Romanes himself. However, Yahya himself gained access to the letters through an Arabic
translation, which he then reproduced in his own chronicle: "they [the letters] were translated into Arabic
for us by the scribe whose task was to copy them into Greek." [translation Feras Hamza].
207
Chapter Five
I. Introduction
half of the empire before Basil's accession in 976. This overview will make
208
extensive use of recently published research, which has been conducted into the
administrative history of the Byzantine empire in the later ninth and tenth centuries.
Following this preliminary overview, the main body of this chapter will go on to
n. Sources
military expeditions, which are preserved in the 'De Ceremoniis', one of the great
encyclopaedic collections collated by the imperial scriptoria in the mid-tenth
1 Both these taktika are published in: Listes. N.Oikonomides, Les Listes de preseance byzantines des
IXe etXe siecles (Paris, 1972). The 'Kleterologion' is to be found between pp.80-235; the 'Escorial
Taktikon' between pp.262-277. Oikonomides' publication also contains two more minor taktika. the
mid ninth-century 'Taktikon Uspenkij' and the mid tenth-century 'Taktikon Benesevic' See below,
chapter six, pp.285-9, for further discussion about dating the 'Escorial Taktikon'
209
century. These documents are memoranda concerned with the logistics of the
imperial expeditions to Crete in 911 and 949, and to Lombard Italy in 935. 2 Further
information on the organisation of the army in the first three-quarters of the tenth
century is discernible within a variety of military manuals which also bear the name
taktika. Foremost among these manuals are 'On Skirmishing' and the 'Praecepta
Militum', both of which appear to have been written by close associates of the
emperor Nikephoros Phokas (963-9).3
Unfortunately the student of the administrative history of the reign of Basil itself is
hampered by the fact that the relatively rich seam of written administrative
evidence available for the first three-quarters of the tenth century peters out fast
after the 970s, and fails to recover in the eleventh century. Moreover, this paucity
of written evidence is particularly acute in the eastern half of the empire. Thus,
there are no extant lists of precedence with which the 'Escorial Taktikon' can be
compared. The only substantial military manual which has been dated to Basil's
reign, 'On Campaign Organisation and Tactics', refers predominantly to warfare in
the Balkans.4 Although ten chapters of a military taktikon compiled by Basil IPs
close political associate, the general Nikephoros Ouranos, reflect on his experience
as the supreme commander of Byzantine forces on the eastern frontier in the first
decade of the eleventh century, most of his vast compendium is a paraphrase of
2
De Ceremoniis, pp.650-9 (Crete 911); pp.660-2 (Italy 935); pp.662-9 (Crete 949). Another of the
mid tenth-century encyclopaedia, the 'De Administrando', may also contain archive material relating
to the expedition of 935 against the Lombards in Italy. Chapters 50 and 51 refer to the commutation
of military service owed by soldiers from the Peloponnese, who were due to fight with a campaign in
Lombardy during the reign of Romanes Lekapenos (920-44); (DAI, p.257; see below, p.232).
3 Skirmishing: G.Dagron and H.Mihaescu, Le Traite sur la guerilla (De velitatione) de I 'empereur
Nicephore Phocas (Paris, 1986), pp.32-135; Praecepta Militum: E.McGeer, Sowing the Dragon's
Teeth: Byzantine Warfare in the 10th Century (Washington, 1995), pp. 12-58; see below, pp.218-9
224.
4 Campaign Organisation and Tactics. Dennis, Three Byzantine Military Treatises, pp.246-326; see
below, p.8.
210
much earlier military manuals. 5 Further insight into administration of the eastern
frontier comes only from sporadic allusions in the correspondence of Ouranos and
his friend, Philetos Synadenos, the judge (krites) of Tarsos, or from brief references
in historical accounts written in languages other than Greek.6 In central and western
Anatolia, the written record is even more meagre. Here, the principal written source
is Skylitzes. As the discussion in the first chapter of this thesis demonstrated,
Skylitzes is so inconsistent in his transmission of administrative terminology that
his text cannot be accepted as a reliable reflection of administrative realia at the
end of the tenth century.7
5 Very little of Ouranos' vast taktikon is published. A.Dain, Aa 'tactique' de Nicephore Ouranos
(Paris, 1937) provides only a background analysis of the contents, sources and manuscript tradition
of the manual. In 1973 de Foucault published chapters 63 to 74 in the belief that these represented a
hitherto unidentified source for the taktikon (Ouranos Taktika (a): de Foucault, 'Douze chapitres
inedits de la 'Tactique' de Nicephore Ouranos', pp.281-310). More recently McGeer has argued that
chapters 66 to 74 are derived from the Classical tactician Onasander, while chapters 56-62 are a
paraphrase of the 'Praecepta Milirum' of Nikephoros Phokas ('Tradition and Reality in the Taktika of
Nikephoros Ouranos', DOP 45 (1991) p. 132). However, he has also argued that chapters 63 to 65,
represent an original attempt to update the 'Praecepta' in the light of Ouranos' own experience on
the eastern frontier. Thus, taken as a whole, chapters 56 to 65 constitute a treatise which blends
Ouranos' reading of earlier tactical works and his own eastern military experience (McGeer,
'Tradition and Reality', pp. 132-4). These chapters are published in Ouranos Taktika (b): McGeer,
Sowing the Dragon's Teeth, pp.88-162.
6 J.Darrouzes, Epistoliers byzantins du Xe siecle (Paris, 1960), pp.217-48 (Nikephoros Ouranos);
pp.249-59 (Philetos Synadenos).
211
of Byzantine administrative history, the most fundamental of which is the difficulty
of dating such material. Unless a seal is attached to a dated document, or unless its
owner can be identified with a known individual within the written historical
record, establishing a precise date for any given seal is almost impossible. In such
cases, dating relies on comparing the iconography and epigraphy of the seal in
question with the very small number of specimens whose dates are known because
they are attached to dated documents.8 A further problem concerns the question of
provenance: where seals were struck and where they were found. If these details are
unknown, the circulation pattern to which a seal belongs, and its position within the
particularly at Preslav and various Danubian sites, makes this difficulty less
pronounced in the western half of the empire10, seals from further east rarely come
Yet, with the publication and intensive study of many more seals in the last twenty
212
connections between particular ranks and offices, can all now be used as sensitive
dating tools. * * Models have also been suggested for the circulation of seals within
the locality itself, and between the centre and the provinces. 12 Nonetheless, despite
In the survey of Basil's reign which follows in the next two chapters, I shall use a
wide range of seals to reflect on the picture produced by Ahrweiler and
Oikonomides, and to reconstruct a picture of the administration of the Byzantine
east which takes better account of the ever-expanding sigillographical record.
Seals from as many major published collections as I have been able to consult
213
within the time limits of doctoral research will be considered. 14 In addition, a large
number of as-yet unpublished seals from northern and central Anatolia and the
eastern frontier held within the collection at Dumbarton Oaks, which were studied
by my supervisor Dr James Howard-Johnston in the summer of 1981, will, with his
permission, be cited. 15 Discussion of the eastern frontier will be facilitated by the
published seals from another collection, that belonging to the late Henri Seyrig.
This collection is of particular relevance to the eastern frontier because
approximately fifteen percent of its seals were bought in Syrian and Lebanese
markets, and thus have a certain provenance from the south-east of the empire. 16
Recent publication of seals at the museums of Tarsos and Antakya (Antioch)
amplifies the body of sigillographical material with a known eastern origin. 17
empire
In the second quarter of the seventh century Arab invasions had shorn the
Byzantine empire of its provinces of Egypt, Syria, Palestine, and Mesopotamia. As
14 The principal collections I have consulted in detail during my research are, in order of date of
publication: G.Schlumberger, Sigillographie de I'empire byzantin (Paris, 1884);
K.M.Konstantopoulos, (Molybdoboulld) BU&VTIOKO, /AoAwj8#o|8ouAAa TOW ev 'A07jvew 'E0viKou
NopioyMTiKou Mowrc/ow (Athens, 1917); J.Ebersolt, Musees imperiaux ottomans. Catalogues des
sceaux byzantins (Paris, 1922); V. Laurent, Les Sceaux byzantins du Medailler Vatican (Vatican,
1962); G.Zacos and A.Veglery, Byzantine Lead Seals (3 vols., Basle, 1972); W.Seibt, Die
byzantinischen Bleisiegel in Osterreich I: Kaiserhof (Vienna, 1978); V.Laurent, Le Corpus des
sceaux de I'empire byzantin, n, L 'Administration centrale (Paris, 1981); G.Zacos, Byzantine Lead
Seals II, compiled by John Nesbitt (Berne, 1985); J.Nesbitt and N.Oikonomides, Catalogue of
Byzantine Seals at Dumbarton Oaks and in the Fogg Museum of Art, (3 vols., Washington D.C.,
1991-96) (I: Itafy, North of the Balkans, North of the Black Sea; II. South of the Balkans, the
Islands, South of Asia Minor; IE: West, North-West and Central Asia Minor and the Orient), The
George Zacos Collection of Byzantine Lead Seals, 1 (Auction 127, Spinks catalogue, 7 October
1998); Vol.2 (Auction 132, Spinks catalogue, 25 May 1999)
15 1 have accepted the provisional dates given to the unpublished seals by the editors and curators of
the Dumbarton Oaks Collection in 1981
16 J.C.Cheynet, C.Morrisson and W.Seibt, Sceaux byzantins de la collection Henri Seyrig (Paris,
1991)
17 J.C. Cheynet, 'Sceaux byzantins des musees d'Antioche et de Tarse', TM 12 (1994), pp.391-479;
idem., 'Sceaux de plomb du musee d"Hatay', REB 54 (1996), pp.249-70
214
a result the eastern half of the empire had been reduced to little more than the rump
of western and central Asia Minor. In the next three centuries provincial
administration in the east was predicated on the need to defend this Anatolian
landmass against invasion and permanent occupation by Arab armies. At the heart
of this system of defence was the so-called 'theme', an administrative unit of
exceptionally obscure origins and controversial development. For example, when
Leo VI discussed the term at the turn of the ninth and tenth centuries, he clearly
18 Leo VI, "Tactica", PG CVH (Paris, 1863), col. 680. Map 5 illustrates the location of the Byzantine
empire's Anatolian themes c.917, shortly after Leo's reign. This map is based on Whittow, 'Making
of Orthodox Byzantium', p. 166, map DC.
19 Theophanes, pp.300, 303-4; J.F.Haldon, Recruitment and Conscription in the Byzantine Army
550-950 (Vienna, 1979), pp.28-29.
20 The classic articulation of this model of the development of the themes and the close relationship
between military service and peasant land tenure is to be found in Ostrogorsky, History of the
Byzantine State, pp.92-109; idem, 'Agrarian Conditions', pp.206-09. See also Hendy, Studies,
pp.619-26, 634-45.
215
More recently some historians, although not all, have been eager to refine certain
elements in this model of army, administration and local society. According to this
revisionist position, the 'big bang' revolution of Herakleios has been discounted.
Rather than being peasant militias, the original theme armies were the former Late
Roman field armies withdrawn from the eastern and western frontiers in order to
defend the landmass of Anatolia against Arab attack. As these armies settled in
Asia Minor they lent their names to the regions they occupied. Thus, when the army
of Thrace withdrew to the western reaches of Anatolia, it gave its name to the
Thrakesion; the army of Armenia settled in the north-east of the plateau and lent its
name to the theme of the Armeniakon.21 Moreover, evidence from early medieval
hagiography suggests that although the soldiers of theme armies often became
farmers and owned landed estates, they did not receive territory in return for
military service. Instead, service was hereditable and attached to the person of the
soldier; troops were paid in coin; recruits came from gentry rather than peasant
backgrounds.22 In the late eighth century cavalry men (stratiotai) were expected to
turn up to muster with their own horse and equipment, conditions which excluded
many peasants from serving in the theme army.23 One of the soldiers from the army
of the Anatolikon which campaigned against the forces of Symeon of Bulgaria in
the early tenth century was Saint Luke the Stylite. As we saw in the previous
chapter, Saint Luke's father was a landowner in the interim region between the
216
river valleys of western Asia Minor and the Anatolian plateau. When famine came
to this area, the grain reserves held by the saint's family were sufficient to feed an
entire local community. The saint's father was, indeed, wealthy enough to be able
to purchase the local bishopric of Sebasteia for his son. 24 The family names of the
landowning classes of Anatolia in later centuries often recalled their ancestry of
service among the thematic cavalry. For example, the Kaballourioi owned property
in the town of Strobiles on the south-west coast of Asia Minor, and on the nearby
ic
islands of Kos and Leros in the mid-eleventh century.
217
emirates such as Kalikala (Theodosioupolis/Erzerum) in western Caucasia,
Melitene in the western reaches of the Anti Taurus, and Tarsos in Cilicia.27 Arab
For a detailed appraisal of how this apparatus of defence worked in practice, there
is no better source than the military treatise 'On Skirmishing'.29 Although this text
was written in the context of incursions of the Hamdanids of Aleppo during the
mid-tenth century, the defensive strategy it expounds was rooted in the cumulative
experience of many centuries of warfare between Byzantium and its Arab
27 See Map 5 for the proximity of these attack bases to the eastern themes of Anatolia
28 The fact that the Muslim forces enjoyed superior reserves of manpower, material wealth, and
morale was widely acknowledged among Byzantine writers. The emperor Leo VI, for example, drew
attention to the Byzantine inferiority in all these respects in his taktika written at the turn of the ninth
and tenth centuries (G.Dagron, 'Byzance et le modele islamique au Xe siecle a propos des
Constitutions tactiques de 1'empereur Leon VI', Comptes rendus des seances de I'Academie des
Inscriptions et Belles-Lettres (Paris, 1983), pp.219-43) The Arab geographer Ibn Hawkal describes
the ribat sponsored by waqfs which existed in the frontier emirate of Tarsos before the fall of the city
to the Byzantines in 965 (Ibn Hawqal, pp. 181-2).
29 Skirmishing: Dagron and Mihaescu, Le Traite sur la guerilla, pp.32-135. All references will come
from this edition. However, it should be noted that another edition with an English translation is to be
found in G.T.Dennis, Three Byzantine Military Treatises (Washington, 1985), pp. 144-250.
218
neighbours.30 It stresses the importance of forward intelligence provided by spies,
including merchants and border guards, as well as the wisdom of moving villagers
and animals to fortified refuges or remote hill-top spots whenever an invasion was
expected.31 It warns against open battle, recommending instead that Arab forces
should be shadowed and only attacked if they broke up into smaller raiding
parties. 32 Highly praised is the tactic of ambushing the enemy in the passes of the
Taurus and Anti Taurus, either as they started out on their raids, or as they returned
eastwards from the plateau laden with booty. 33 Incursions by the Byzantines into
territory beyond the frontier are only countenanced in order to force the enemy
back to defend his home country.34
Yet, by the time that "On Skirmishing" was compiled, the need for traditional
defensive theme armies was fading. With the collapse of the authority of the
Abbasid caliphate of Baghdad in the first quarter of the tenth century, Muslim
incursions into Anatolia from the east had become much rarer. The last raid from
Iraq was dispatched from Baghdad in 923/4 and from Tarsos in 931/2.35 Even when
Arab raiding resumed in the later 940s and 950s under the leadership of Sayf al-
Dawla, Hamdanid emir of Aleppo, few jihad incursions penetrated very deep into
the Anatolian plateau: no raid reached further than Koloneia on the north-eastern
reaches of the plateau.36 And although the Hamdanid court poets rejoiced in the
30 Whittow, Making of Orthodox Byzantium, p. 176; Ahrweiler, 'La frontiere at les frontieres',
pp.216-9.
31 Skirmishing: Dagron and Mihaescu, pp.39-41, 51-2
32 Ibid., pp.51-63
33 Ibid., pp.41-5, 63, 73-5
34 Ibid., pp. 111-117.
35 Vasiliev, Byzance et les Arabes, ii(l), 250, 265-6; ii(2), 148, 152 (Ibn al-Athir). Ibn al-Athir
mentions a Muslim incursion in 924/5, but does not indicate whether this invasion came from
Baghdad, Tarsos or another of the frontier emirates. See above, p. 172, n.87
36 Sayf reached Koloneia in 940 according to the historians Ibn Zafir and Ibn al-Azraq (Vasiliev,
Byzance et les Arabes, ii(l), 284-90; ii(2), 122-3, 289-90); see Map 3. The expedition is also
described by the contemporary court poet Abu Firas, Sayf s own cousin (Vasiliev, ii(2), 357-8). Yet,
while the Arab sources praise Sayf for his fortitude in reaching so far west, alleging that news of his
219
heroism of Sayf s attacks, the emir's military record was patchy: Sayf himself was
lucky to escape from defeats by Byzantine armies in 950, 958 and 960. 37 Moreover,
once the Byzantines went onto the offensive against Sayf in the later 950s, his
threat immediately diminished. By 962 the suburbs of Aleppo had been raided.
Sayf was forced to flee northern Syrian and take refuge at Mayafariqin in the
Djazira south of the Anti Taurus mountains. He died shortly after returning to
OQ
Aleppo in 967. After the fall of Antioch to the Byzantines early in 969, Aleppo
client.39
Indeed, as the military threat of attack from the east waned and the possibilities for
Byzantine military expansion waxed, the military enterprises of the empire were
increasingly undertaken by forces other than the theme armies. Although it would
be premature to suggest that the theme armies had ceased to exist in the mid-tenth
standing army had, of course, long complemented the provincial theme armies.
siege of Koloneia caused terror in Constantinople, the fact that they stress that no Muslim raid had
reached this far west in living memory indicates that the Anatolian plateau had become much less
vulnerable to marauders from the Islamic east by the mid-tenth century. Sayf himself was never to
raid so far west again.
37 The most prominent of the many Hamdanid court poets who lauded Sayf s military achievements
against the Byzantines was Mutannabi (Vasiliev, Byzance et les Arabes, ii(2), 304-348; M.Canard,
'Mutannabi et la guerre byzantino-arabe', M&noires de I'lnstitut frangais de Damas, (Beirut, 1936),
pp.99-114). Dennis, Three Byzantine Military Treatises, p. 157, summarises these defeats in the
defiles of the Taurus and Anti Tarus mountains; see also Vasiliev, infra., ii(2), 70-1, 96, 111, 181 and
E.Honigmann, Die Ostgrenze, pp.84-5; Canard, M., Histoire de la dynastic des H'amdanides de
Jdzira et de Syrie (Paris, 1953) , pp.763-8, 795, 801-3; RJ.Bikhazi, "The Hamdanid Dynasty of
Mesopotamia and North Syria 254-404/868-1014' (Univ. of Michigan PhD. Thesis, 1981), pp.716-
23, 785-8, 845-9.
38 Bikhazi, 'The Hamdanid Dynasty of Mesopotamia and North Syria', pp.856-68, 934-5.
39 See above, p.200, for discussion of the commercial clauses of the treaty that guaranteed this
relationship.
40 Ahrweiler, 'Recherches', p.2; Whittow, Making of Orthodox Byzantium, pp.323-7.
220
duties had traditionally been twofold: to protect the emperor from armed
insurrections led by provincial generals, and to act as a rapid reaction force which
could be sent to the frontiers more quickly than thematic armies.41 By the end of
the ninth century the core of this centralised army comprised four battalions
(tagmata) of cavalry troops: the Scholai, the Exkoubitai, the Hikanatoi, and the
Arithmos. There were also two infantry battalions: the Walls (Teichos) and the
Noumeroi. The hetaireia, the palace body guard, was a vital augment to this
centralised army. Particularly conspicuous among the troops of the hetaireia were
overseas mercenaries including Bulgarians, Russians and Turks from central Asia.43
tenth century, new units were added to this centralised army. Before the end of the
reign of the emperor John Tzimiskes (969-76) several new tagmata had been
established such as the Athanatoi (the Immortals) and the Stratelatai** The
221
domestikos of the scholai usually assumed overall control of the tagmata and
hetaireia, and indeed of the whole Byzantine army including thematic forces,
during a major expedition. From the middle of the tenth century onwards the
Byzantine field army was divided into western and eastern contingents led
respectively by the domestikoi of west and east.45 Whenever there was no
domestikos, the army could be led by a stratopedarches or a stratelates. 46
The date at which the size and the ambition of this more centralised and
professional army began to expand is unclear. During the reign of Romanos
Lekapenos (920-44), the domestikos of the scholai John Kourkouas led the
Byzantine field army on long distance raids against the Arab emirates of Dvin and
Chliat in Armenia (927/8) and Edessa in northern Mesopotamia (943/4).47 He also
orchestrated more sustained campaigns against Melitene, which fell to Byzantine
forces in 934, and Theodosioupolis (Kalikala), which finally surrendered to his
AV
brother Theophilos in 949. Although these are the only two large-scale permanent
conquests noted in the historical record during the first half of the tenth century, it
is clear that during the 930s and 940s, armies led by the Kourkouas brothers,
gradually brought many of the small, isolated, upland depressions within the Anti
Taurus mountains under Byzantine control. These areas were created into small
of the new tenth-century tagmata see Kuhn, Die byzantinische Armee, pp.68-9, 243-49; Ahrweiler,
'Recherches', pp.27-32.
45 Kuhn, Die byzantinische Armee, pp. 135-57
^uhn, Die byzantinische Armee, pp.262-5; Oikonomides, 'L'evolution de 1'organisation
administrative', p. 142; Ahrweiler, 'Recherches', p.26. For example, Peter the Stratopedarches, a
servant (doulos) of the emperor Nikephoros Phokas, led the main Byzantine army of the east at the
fall of Antioch in 968/9 and at the siege of Dristra in Bulgaria in 971, and was in charge offerees
loyal to the emperors Basil and Constantine during the first year of the Skleros revolt (Skylitzes,
pp.272, 300, 308, 315-22; also see above, p. 123, 129, and below, p.300)
47 Vasiliev, Byzance et les Arabes, ii(l) 262-3, 296-303. Those Armenian historians who refer to a
Kourkouas raid against Dvin in 922, are almost certainly recounting the raid of 927 but applying a
date that is five years too early (Vasiliev, ii(l) 250). See Map 6 for all place names along, and
beyond, the Byzantine eastern frontier in the tenth and eleventh centuries.
222
border themes, such as Chozanon and Asmosaton, many of which are listed in the
Yet, the scale of Byzantine military activity in the first half of the tenth century
should not be overstated. The more spectacular of the Kourkouas-led enterprises, at
Dvin and Edessa, were not expeditions of permanent conquest. Instead, their
objectives were more closely connected to fuelling triumphalist imperial
propaganda. On the one hand, they were designed to terrify the empire's eastern
neighbours; on the other, to bolster imperial prestige within Byzantium. For
example, the principal aim of the Edessa expedition was not the conquest of
northern Mesopotamia, but instead the acquisition of the Mandylion of Christ, a
precious relic (or possibly an ikon) on which was imprinted the face of Jesus. This
was brought back to Constantinople amid great ceremony, greeted by the emperor
and the patriarch, paraded from the Golden Gate to Hagia Sophia, and then taken
and stored in the Great Palace. Having handed over the Mandylion to the forces of
Kourkouas, the inhabitants of Edessa were themselves given a guarantee against
future attack and left in peace. 50
Indeed, it seems more likely that pace of military innovation and territorial
expansion only really increased when Nikephoros Phokas, the future emperor
Nikephoros n (963-9), became domestikos of the scholai in the latter years of the
reign of Constantine Porphyrogenitus (945-959). At this point in the tenth century,
48 Melitene. Vasiliev, Byzance et les Arabes, ii(l) 266-70; Theophanes Con., pp.426-9; and see
above, pp.185, 189, 192; Theodosioupolis. Vasiliev, ii(l) 318-19; DAI, pp.212-15; see below,
chapter six, p.290
49 Honigmann, Die Ostgrenze, pp.74-80; Howard-Johnston, 'Crown Lands', pp.86-91; see above,
p.209 and Map 6.
50 George the Monk Con., pp.918-9; Vasiliev, Byzance et les Arabes, ii(2) 156-7 (Ibn al-Athir);
Yahya, PO 18, pp. 730-33; A.Cameron, 'The History of the Image of Edessa: the Telling of a Story',
Harvard Ukrainian Studies 1 (1983), pp.80-94.
223
according to Greek sources, Nikephoros abandoned the defensive tactics of the
previous domestikos, his own father Bardas Phokas. Now, instead of".... hiding
....... and avoiding action", as had been the Byzantine wont, the imperial armies
"lived in enemy territory as if it were their own." The Arab historian Yahya ibn
Said was of a similar opinion:"...... .les incursions de Nicephore devinrent comme
un plaisir pour ses soldats parce que personne ne les attaquait......; il marchait ou il
voulait."51
A significant growth in the size and development of the martial capacities of both
cavalry and infantry forces during the middle of the tenth century underpinned tthe
greater confidence of Byzantine field armies operating in enemy territory. A key
formation was the hollow infantry square, composed of several thousand spearmen,
archers and slingers, which protected the cavalry and baggage train during marches
through enemy terrain, while at the same time acting as a base from which
Byzantine horsemen could attack the enemy. The fact that the infantry square was
first discussed within military manuals dated to the period a 940-50 suggests that
the outlines of this manoeuvre may have been developed by the generals John and
Theophilos Kourkouas during the eastern campaigns of the reign of Romanos
Lekapenos.52 Yet comparison of the tactic described in these early manuals with the
strategy outlined in the Traecepta Militaria', a military taktikon compiled during
the reign of Nikephoros Phokas (963-969), possibly from the field notes of the
emperor himself, indicates that this formation only began to be employed on a large
scale after Nikephoros took control of the Byzantine army. For example, while one
224
of the earlier manuals, the 'Sylloge Tacticorum', recommends an infantry force of
three thousand men, the 'Praecepta' presents a figure which is four times this
heavy coverings, these were shock troops whose principal task in open battle was to
charge in a triangular wedge formation, smashing their way through the ranks of
their opponents, and scattering the enemy before them. 54 The success of these
tactical developments was soon apparent. Between the period when Nikephoros
Phokas was appointed domestikos (c.955) and his death as emperor in 969,
Byzantine field armies crossed the Taurus mountains conquering and annexing
Cilicia in 964-5, and Antioch in northern Syria in 969. By 971 Nikephoros' imperial
successor, John Tzimiskes, had translated this success westwards. By defeating
Russian and Bulgarian forces at Preslav and Dristra in the Balkans, he was able to
add eastern Bulgaria to the Byzantine empire.55
Simultaneous with the growth of a more centralised and heavily-armed land force
was the development of a more integrated fleet. In the tenth century the main
components of the Byzantine naval force were the imperial fleet based in
Constantinople, and the three thematic fleets of Samos, the Kibyrrhaiotai and the
Aegean located in the islands and litoral of western Asia Minor. Increased attention
to maritime warfare had been demanded in the ninth and first half of the tenth
centuries by the Arab occupation of the Mediterranean islands of Crete and Sicily,
225
and by the sporadic appearance of Rus fleets in the Black Sea.56 The seriousness of
these naval threats is apparent in the literary and archaeological records of the
northern and western shores of Asia Minor. 57 For example, posthumous miracles
related by the Life of George of Amastris, describe local inhabitants taking refuge
from Rus raids during the first half of the ninth century in the fortified citadel of
Amastris on the Black Sea coast. Archaeological survey has suggested that this site
was an important harbour for the imperial navy.58 Epigraphical evidence indicates
that the fortifications at Attaleia on the southern coast were strengthened to
withstand Arab attack between 911 and 916.59
At various points during the tenth century Byzantium tried to take the naval
offensive, in particular against the Arabs of the Mediterranean. Expeditions to
conquer Crete were launched in 911 and 949, both of which required the
participation of large fleets. Neither, however, was successful. It was only in 961,
under the command of Nikephoros Phokas, that Byzantine forces were able to
regain Crete. By 965 Cyprus had also fallen.60 Yet, the success of these offensive
operations at sea was limited. By the mid-tenth century the principal naval rivals to
the Byzantines in the Mediterranean were the Fatimids of North Africa and Sicily.
In the final years of the reign of Constantine Porphyrogenitus a series of
inconclusive naval engagements between Byzantine and Fatimid navies were joined
226
off the coast of Sicily.61 In 965 Nikephoros Phokas' cousin Manuel and the eunuch
Niketas, the admiral of the imperial fleet, led an unsuccessful invasion of Sicily.
After this Byzantine defeat, it is possible that the Fatimids dispatched a fleet to the
eastern Mediterranean in an attempt to regain the island of Cyprus, or at the very
least disrupt Byzantine control of the seas in this area.63 Although this obscure
expedition does not appear to have been successful, the fall of Egypt, Palestine and
southern Syria to the Fatimids in 969/70 brought the naval threat of the North
African Muslims into the very heart of the Levant.
The more ambitious military enterprises of centralised Byzantine forces in the tenth
century, whether at sea or on land, were also supported by an expansion in the
administrative organisation of the periphery of the empire. As far as the eastern half
of the empire is concerned, this expansion took the form of a proliferation of new
administrative and territorial units. These units were created both from the core
Byzantine territories in the eastern reaches of the plateau, and from the lands
newly-conquered by imperial armies. As early as the mid-ninth century the eastern-
most regions of the plateau, including Cappadocia, Charsianon, Chaldia and
Koloneia, had been detached from the administrative structures of the sprawling
themes of the Anatolikon and the Armeniakon, becoming in the process themes in
their own right.64 East of these themes lay the kleisourai, small administrative blocs
which commanded vital border passes. By the early tenth century many of these
61 See Vasiliev, Byzance et les Arabes, ii(l), 371-6, for a summary of this warfare and the difficulty of
establishing the chronology of Fatimid-Byzantine relations between 956-59; Theophanes Con.,
pp.454-5; von Falkenhausen, Untersuchungen uber die byzantinische Herrschaft, p. 165
62 Leo the Deacon, p.65ff; Skylitzes, p.261; Kleinkroniken, i, 338-340, no.45; Canard, 'Quelques
noms de personnages byzantins', p.457; von Falkenhausen, Untersuchungen uber die byzantinische
Herrschaft, pp.99, 126-7
"Yahy^POlS, pp.794-5
64 AToynbee, Constantine Porphyrogenitus and His World (London, 1973), pp.252-74; see Map 5
for the Anatolian themes and kleisourai early in the tenth century.
227
kleisourai, including Sebasteia and Lykandos, were upgraded to themes.65 As we
have already seen, new small border themes were added to this administrative
tessellation when the Kourkouas brothers led Byzantine field armies into the Anti
Taurus during the reign of Romanes Lekapenos.66 The armies of these border
enclaves were often staffed with Armenian infantry and cavalry troops recruited
from the hill country of western Caucasia. As a result these small themes were
respects they were like the kleisourai of earlier periods, since one of their principal
tasks was to watch over mountain passes and send intelligence to commanders in
the interior of the empire.68 On the other hand they also supported Byzantine field
The harrying role of first the garrisons of the kleisourai, and later the Armeniaka
during the tenth century. In the earlier part of the century the armies of kleisourai
within the Anti Taurus mountains allowed the Byzantines to apply constant military
pressure against the emirate of Melitene. Annual raids conducted by the Byzantine
such as Melias, the kleisourarches of Lykandos and Tzamandos.69 The most daring
workmen into the emirate during the winter of 928, so that when the main
Byzantine army arrived the following year, these fifth columnists would be able to
Skirmishing: Dagron and Mihaescu, pp.39-41; Ahrweiler, 'Recherches', pp.49, 82; idem., 'La
frontiere at les frontieres', p.217; Kiihn, Die byzantinische Armee, pp.61-4.
69 See above, pp. 184-5 for further discussion of the career of Melias.
228
deliver the city without a protracted siege. Although the plot failed when the true
identity and purpose of the Armenian workmen was uncovered, the unremitting
pressure of annual attack by field armies and consistent harrying by border troops
had forced the emir of Melitene, Abu Hafs, to enter a tribute arrangement with the
empire in 931. When this arrangement lapsed on the death of Abu Hafs, the city
once again came under sustained military pressure from Byzantine armies. It was
finally conquered by John Kourkouas in 934.70
Further north and east, the emirate of Kalikala (Theodosioupolis) came under
similar pressure from raiding parties operating out of the fortress of Hafdjidj (in
Greek known as Chauzizion) on the Araxes river during the decade before its fall in
949.71 Even Nikephoros Phokas' more rapid, ambitious, and far reaching conquests
in Cilicia and northern Syria in the 960s relied on small scale operations from hill-
top castles to provide back-up for the main expeditionary forces. At Antioch the
hinterland of the city was softened up by a large raid in the autumn of 968. As the
main army withdrew to Cappadocia for the winter, Michael Bourtzes was left at the
castle of Baghras in the Amanos mountains as strategos of the newly created theme
of Mauron Oros. He was instructed to "prevent the inhabitants of Antioch from
coming out and collecting the supplies necessary for living" during the winter. In
the spring the main army under Peter the Stratopedarches returned; by the autumn
70 Theophanes Con., pp.415-6; Vasiliev, Byzance et lesArabes, ii(l), 266-7; see above, pp.68-9, 185,
189, 192
71 DAI, pp.206-214. The historian Ibn Zafir notes that the Byzantines had already built a fortress at
Hafdjidj by 939 (Vasiliev, Byzance et les Arabes, ii(l) 284; ii(2) 122). This fortress has yet to be
identified although it has been tentatively located north of the Bingol Dagi, near the source of the
Araxes river (Howard-Johnston, 'Procopius, Roman Defences North of the Taurus', p.203;
Honigman, Die Ostgrenze, pp.79-80, 195). See Map 6.
72 Yahya, PO 18, p.816; Leo the Deacon, pp.73-4. See Map 6 Skylitzes gives the name of the theme,
Mauron Oros, although he wrongly locates it in the Taurus rather than the Amanos range, a mistake
which some modern historians of Byzantium have copied (Skylitzes, pp.271-2; Ahrweiler,
'Recherches', p.46). Skylitzes suggests that the castle controlled by Bourtzes was built from scratch
229
However, the growth in the number of small garrisons on the eastern frontier and
the centrally-organised Byzantine field army which they supported, came with a
high financial price. Finding an adequate financial footing for the military had long
been a difficult matter. Even when Byzantine armies had consisted predominantly
of locally-recuited thematic troops mustered on only an occasional basis, many
stratiotai had been unable to supply their own horses, weapons and equipment. In
the early ninth century Nikephoros I had tried to alleviate this difficulty by ruling
that impoverished soldiers were to be supported by their neighbours within their
villages (choria) 73 Leo VI recommended that new stratiotai should only be
recruited from among the well-off (euporoi)14 Nonetheless, by the middle of the
tenth century the position of many theme soldiers remained parlous. In his novel
dealing with military finance the emperor Constantine Porphyrogenitus claimed:
"the affairs of the stratiotai, from which their existence and livelihood are drawn,
have, in time, sickened and declined."75 Yet, at the same time as many soldiers
were oppressed by the costs of warfare, the basic equipment required to fight in the
increasingly aggressive Byzantine army, particularly within the cavalry, was
in 968. However, Ibn Hawkal, indicates that the site was already fortified before the period of
Byzantine rule (Ibn Hawqal, p. 182). Although this theme is not listed in the later tenth-century
'Escorial Taktikon', a later eleventh-century seal of a strategos of Mauron Oros, and mention of
Mount Mauros "and all the castles dependent on it" in the 1108 Treaty of Devol, suggest that it
continued to exist throughout the eleventh and into the early twelfth centuries (Cheynet, Morrisson
and Seibt, Sceaux byzantins: Henri Seyrig, no. 183; Anna Komnene, iii, 133-6; Honigman, Ostgrenze,
p. 127). Mauron Oros was almost certainly only one of a string of small forts ringing Antioch
garrisoned by Byzantine raiding parties during the two years which preceded the fall of the city.
Among other sites fortified and used as forward attack bases was Qalat Siman, the monastery of the
fifth-century stylite Symeon, located between Aleppo and Antioch (W.B.R.Saunders, 'Qalat Siman: a
Frontier Fort of the Tenth and Eleventh centuries', in S.Freeman and H.Kennedy (eds.), Defence of
the Roman and Byzantine Frontiers (BAR International Series, Oxford, 1986), pp.291-305)
73 Theophanes, p.486. It has been widely argued that between the eigth and ninth centuries thematic
stratiotai fell into one of two groups, distinguished one from another other by levels of personal
wealth. Richer soldiers, such as Saint Luke the Stylite (see above, p.217), supplied their own
equipment; poorer soldiers were supported either by state provisions or by contributions from
wealthier members of their own local community (Haldon, 'Military Service, Military Lands', pp.21-
6; Ahrweiler, 'Recherches', pp.5-6, 14-15, 19-20).
74 Dagron, 'Byzance et le modele islamique au Xe siecle a propos des constitutions tactiques de
1'empereur Leon VI', pp.234-5
230
becoming much more expensive. Nor was the cost purely financial. The burden of
taxation required to finance the more expansionist Byzantine military apparatus of
the third quarter of the tenth century also carried considerable political risk.
Despite his conquests in the east, the emperor Nikephoros Phokas was pelted with
stones during a procession in Constantinople by citizens enraged by his military
expediture. To ensure his own security he had to fortify the imperial palace. 76
231
unless they owned estates in excess of twelve pounds, did entail that a much greater
pool of immovable property was connected on a permanent basis with military
service. 79
Yet, these legislative initiatives were not latter-day attempts to create or expand
land-based peasant militias. Instead, the new arrangements represented a novel
property tax designed to finance the armed forces more efficiently, known as the
Qf\
the aims underpinning the mid tenth-century decision to link fiscalised military
service to land rather than to the person, was the hope that this problem of
individual impoverishment could be circumvented. Under the new land-based
79 For this more nuanced approach to the novel of Nikephoros Phokas see Kaplan, Les Hommes et la
terre, pp. 251-3.
80 Haldon, Recruitment and Conscription, pp.59, 65; idem., 'Military Service, Military Lands', pp.37-
9; Ahrweiler/Recherches', pp. 12-14.
232
system the strateia would always be paid in full. For example, in cases where the
lands were subdivided because they had been inherited on a partible basis, the fact
that all the heirs were obliged to contribute on a. pro rata basis meant that the full
amount due would still enter the coffers of the state. 82
travelled from the And Taurus to fight in the expedition. This arrangement
contrasts with the earlier Cretan expedition of 911 when the entire complement of
one thousand stratiotai from the Thrakesion fulfilled active service.84
233
HI. The administration of Anatolia in the reign of Basil II
In the century before Basil n came to the throne the administration of the Anatolian
heartland of the Byzantine empire had changed radically. With the waning of Arab
raids a system of local government based on the exigencies of defence had slowly
begun to be eroded. By the mid-tenth century there was little need for part-time
theme armies recruited from the ranks of local landowners. Instead, military service
was fiscalised with greater frequency, and the receipts of commutation channelled
into the more adventurous military expeditions of the centralised armed forces. The
remaining sections of this chapter examine how these deeply-rooted trends of
demilitarisation, fiscalisation and centralisation developed in the Anatolian themes
during the reign of Basil.
/. Military administration
Any investigation into the administrative history of the Anatolian themes during
Basil's reign, must begin with some consideration of the officials who had
exercised both civil and military gubernatorial powers over the locality in the ninth
and tenth centuries, namely the thematic strategoi*5 Unfortunately, tracing the
development of this position during the reign of Basil II is rendered very difficult
by a lack of solid information in the historical record. This is a problem which is
particularly pronounced in the first half of the reign, when even identifying
contemporary strategoi is almost impossible. Despite the seriousness of the armed
struggle which engulfed Asia Minor during civil wars waged by Bardas Skleros and
Bardas Phokas, only one strategos can be named and indisputably linked to a theme
in the whole 976-89 period. That strategos was Michael Kourtikios, who was
84 De Ceremomis, p.652
85 Continue to consult Map 5 for all references to themes in western and central Anatolia
234
nominated to the naval theme of the Kibyrrhaiotai in south-west Asia Minor by
Bardas Skleros shortly after the outbreak of his first revolt in 976.86 For further
information about strategoi during the first half of the reign, one has to turn to
more indirect literary clues, most of which are offered in John Skylitzes' coverage
of the first Skleros revolt. Thus, from Skylitzes' narrative we may deduce that
Karantenos fought for the imperial side during Skleros' first revolt. He defeated
Kourtikios in a naval battle in 977 near the port of Phokaia which lay within the
unsuccessfully defended Nikaia from a siege by Skleros forces in 977/8, was the
komes of the Opsikion, the theme of which Nikaia was the capital.88 When Erotikos
QS'
Skylitzes, pp.319, 322. It is possible that an unpublished seal belonging to Michael Kourtikios,
magistros in the Dumbarton Oaks Collection may refer to this strategos [58.106.1360]. At an earlier
stage in his career Kourtikios appears to have been a strategos and protospatharios (Schlumberger,
Sigillographie, p.380). Cheynet, Pouvoir, p.28 n.7., believes on the basis of the Schlumberger seal,
that Kourtikios was a protospatharios when he was strategos of the Kibyrrhaiotai. However, since
most of the senior Skleros commanders held the title of either magistros orpatrikios, it is more likely
that he was a magistros.
87 Skylitzes reports on the Karantenos:Kourtikios engagement off Phokaia (Skylitzes, p.322).
However, Skylitzes mistakenly implies that Karantenos was the admiral of the imperial fleet based at
Constantinople. In fact, the contemporary Leo the Deacon relates that Bardas Parsakoutenos was the
head of the central fleet. This fleet used Greek fire to destroy a navy loyal to Skleros off Abydos (Leo
the Deacon, p. 170). It is likely that this confusion between Karantenos and Parsakoutenos in
Skylitzes' narrative results from the author's characteristic tendency to conflate events (see above,
chapter one, pp.46-7). Here, he has confused two battles, the first conducted off the western coast of
Asia Minor by Kourtikios and Karantenos, the commanders of two thematic navies; the second
fought much closer to Constantinople between the imperial navy of the capital led by Parsakoutenos
and another Skleros maritime force whose commander is unknown. The likelihood that Skylitzes
conflates the two battles has also been suggested by Seibt (Die Skleroi, p.42).
88 This is the argument presented by Seibt on the basis of evidence provided by Skylitzes (Seibt, Die
Skleroi, p.42; Skylitzes, p.323). Anna Komnene, however, maintains that the office fulfilled by
Erotikos was that of"autokrator, strategos of the whole east" (Anna Komnene, Hi, 9-10). According
to Anna, Erotikos was charged by Basil II with the responsibility of either putting an end to the civil
war with Skleros, or to resisting the rebel by force. A similar story is transmitted by Nikephoros
Bryennios, Anna's husband (Bryennios, p.75). However, it should be remembered that Erotikos was
the tenth-century founder of the Komnenian dynasty. It is possible that later members of the
Komnenian clan, such as Bryennios and Anna, may have deliberately exaggerated the political
importance of Erotikos' role in the Skleros revolt in order to enhance their own family history. It is
striking that none of the more contemporary sources, such as Yahya, Stephen of Taron, or Leo the
Deacon, indicate that Erotikos held an unusually elevated position during the civil wars of the late
tenth century. The governor of the theme of Opsikion was by tradition called komes rather than
strategos.
235
was forced to evacuate the city with his garrison, Skleros appointed an Arab called
Pegasios as commander of the city. It seems probable that he also held the position
of homes Finally, from Skylitzes' account it may be possible to identify one more
strategos towards the end of Skleros' first revolt. This was Nikephoros
Parsakoutenos. He negotiated the surrender and amnesty of several rebels from the
Skleros party, who had taken refuge in several castles in the theme of the
Thrakesion after the revolt collapsed in 979/80. It is likely that Parsakoutenos
performed this task as strategos of the Thrakesion.90
In contrast to the dearth of information about the officials who held the position of
strategos in the first half of the reign, references in the historical record are a little
more plentiful in the second, although most have to be extracted from Skylitzes'
testimony. Thus, from Skylitzes we learn that Basil Argyros campaigned in
southern Italy in 1011 against the rebel Meles while exercising command over the
maritime theme of Samos.91 Further east Skylitzes relates that when Senecherim
king of Vaspurakan handed over his hereditary lands in southern Armenia to Basil
n, he became strategos of the central Anatolian theme of Cappadocia. Although the
period when Senecherim held this office is uncertain, because of the difficulties in
applying a precise date to the cession of Vaspurakan, he must have excercised
responsibilities in central Anatolia before Basil's death in 1025.92 Another
236
Skylitzes reference to an Anatolian strategos in the second half of Basil's reign
comes just before the emperor died. At this point the historian notes that a strategos
of Samos defeated a Rus fleet off the Aegean island of Lemnos. This strategos was
93 Skylitzes, p.368. The strategos of Samos was called David of Ohrid. He may be the same person as
David Nestoritzes, a Bulgarian commander, who raided Thessalonika in 1014. Called simply
Nestoritzes at another point in Skylitzes' text, he appears to have been among those senior Bulgarian
generals who surrendered to Basil H at Ohrid in 1018 (Skylitzes, pp.350, 359)
94 Skylitzes, p.373. The strategos of Samos was George Theodorokanos. He was almost certainly a
relative of an Armenian called Theodorokan, who served Basil n during the first decade of the
eleventh century as strategos of Philippoupolis and doux of Adrianople (Skylitzes, pp.343-5;
Cheynet, Morrisson and Seibt, Sceaux byzantins: Henri Seyrig, p. 150; H.C.Evans and W.D.Wixom,
The Glory of Byzantium: Art and Culture of the Middle Byzantine Era 843-1261 (New York
(Metropolitan Museum) 1997), pp.357-8)
95 Skylitzes, p.372
96 Skylitzes, pp.366-7; Yahya (Cheikho), p.241. The 'Georgian Royal Annals' mention the
involvement of Xiphias in the revolt but not the nature of his command (Georgian Royal Annals,
p. 283). Previously Xiphias had exercised command in the Balkans In 1000-1 he led a raid against the
principal eastern towns of eastern Bulgaria in the company of Theodorokan, strategos of
Philippoupolis. Shortly afterwards he replaced Theodorokan at Philoppoupolis. He still held this
office at the time of the battle of Kleidion in 1014 (Skylitzes, pp.343, 348; see above, chapter two,
237
Theophylact Dalassenos.97 One of the co-conspirators in the Xiphias revolt was a
Georgian called Pherses, who had entered Byzantine imperial service when the
princedom of Tao was annexed by Basil in 1000. He was executed for his role in
the Phokas-Xiphias revolt. Sigillographical evidence suggests that at some point
between 1000 and his death in 1022, Pherses had held the office of strategos of
QO
Cappadocia.
pp.89-90, and below p.315). After the revolt was over, Xiphias was confined to the Studios
monastery (Skylitzes, p.376). Yahya claims that his punishment was relatively mild, precisely because
of his previously loyal service to the emperor in Bulgaria (Yahya (Cheikho), p.243).
97 Yahya (Cheikho), p.241
98 Skytitzes, p.367; Aristakes, p.21; Georgian Royal Annals, p.283; Dumbarton Oaks Unpublished
55.1.3367. Pherses entered Byzantine service in 1000 with his brother Phebdatos/Theudatos. They
were both honoured as patrikioi (see above, p.89) The family had an egregious military pedigree.
Pherses himself had led the Iberian contingent of a joint Ibero-Armenian army which defeated the
emir of Azerbaijan in 998 (Stephen of Taron, pp.205-6). He himself held the position prince of
princes (eristav eristavi), the office inferior only to that of the ruler in the polities of western
Caucasia. The office was hereditary. In 979 Pherses' father Tzotzikios had exercised the same
position in the Iberian army which joined forces with the imperial armies led by Bardas Phokas to
defeat Bardas Skleros (Stephen of Taron, p. 142; see above, p. 114, 119-20). Once the family took up
residence in Byzantium, it served in many locations across the empire, including the Balkans. In
1016/7 Tzotzikios, the son of Theudatos and nephew of Pherses, was strategos of Dristra on the
Lower Danube (Skylitzes, p. 3 56; see above, p. 89). Pherses himself was executed in 1022 because he
238
QQ
Boukellarion and the Opsikion in the middle of the eleventh century. Several
further examples can be added to Cheynet's list from the seals I have studied. A
exercised this responsibility at the same time as he was the anthropos of the kaisar
_ I t\f\ _
John Doukas. The title of kaisar was held by John Doukas between 1081 and
1088. 101 Furthermore, seals of another eleventh-century strategos of the Thrakesion
have been found far from Asia Minor at Sougdaia in the Crimea. 102 To the list of
Cappadocian strategoi can be added a certain Michael, whose seal is decorated
with a patriarchal cross standing on an orb, surrounded by curling, ornate leaves, a
design often found on seals at the end of the tenth century. 103 Sigillography also
allows three eleventh-century strategoi of Seleukeia to be identified: Leo Blangas;
an Arab called Soulaios; and an Armenian called Apellarib Arsakides. 104
However, although sporadic literary references and lead seals indicate that the
position of strategos continued to be occupied throughout Basil's reign and deep
into the eleventh century, there is little evidence to suggest that this survival
signified the revival of local, indigenous armies. Instead, although vestiges of the
theme armies lingered into the early eleventh centuries, literary and sigillographical
evidence indicates that many of the troops stationed in the core Anatolian themes
under the command of local strategoi increasingly tended to be professionals
killed four kouratores and an imperial eunuch during the revolt of Xiphias and Phokas (Skylitzes,
p.367).
99 Cheynet, 'Du stratege de theme au due', pp.187-191
100 See above, p. 66, n.120 for the meaning of anthropos
101 Nesbitt and Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no.2.39.
102 V.Sandrovskaya, 'Die Funde der byzantinischen Bleisiegeln in Sudak', in N.Oikonomides (ed.),
SBS 3 (Washington, 1993), pp.94-5
103 Zacos, Byzantine Lead Seals //, no.234
104 Leo Blangas. Cheynet, 'Sceaux byzantins des musees d'Antioche et de Tarse', no.57 and
Dumbarton Oaks Unpublished F1316; Soulaios: Zacos, Byzantine Lead Seals II, no.705; Apellarib
Arsakides. The George Zacos Collection of Byzantine Lead Seals (Auction 132, Spinks catalogue,
25 May 1999), no. 112. The latter was the uncle of Gregory magistros, who was katepan of
239
recruited from outside the locality. 105 The deployment of professional forces
recruited from outside Anatolia had already begun before Basil's reign. The
memoranda concerned with the Cretan expedition of 949 indicate that by the mid-
tenth century some detachments of the tagmata of the central field army, once
resident in Constantinople, were now based in the themes. In the case of the
tagmata of the Scholai they were to be found in the European provinces of Thrace
and Macedonia. In the case of the Hikanatoi and Exkoubitai, their base was in
north-west Asia Minor, and they were accordingly known as the peratika
10ft
tagmata. Furthermore, elsewhere in western Asia Minor professional troops
employed from outside the empire were increasingly ubiquitous. As early as 911,
five hundred Armenian troops were stationed in the Thrakesion and another five
hundred in the Anatolikon. 107 By 949 the number of Armenian troops in the
Thrakesion had expanded to six hundred, while two hundred and thirty Slavs were
located in the Opsikion. 108 It has been convincingly argued by James Howard-
Johnston that the number of Armenian troops resident in the Thrakesion during the
first half of the tenth century was deliberately increased, in order to provide some
limited compensation for a short-fall in recruits of indigenous thematic
cavalrymen. 109 The seal of Peter, the domestikos of the western Anatolian theme of
the Optimatoi, and katepan of the Iberians, datable to the later tenth century
indicates that troops from western Caucasia were still being deployed in the
Vaspurakan in the mid-eleventh century (see below, pp.298-9). Like his nephew he was a member of
the Armenian Pahlawuni family (Cheynet, Morrisson and Seibt, Sceaux byzantins: Henri Seyri, p.43).
105 The case for the gradual withering of the theme armies and their replacement by professional
troops in the later tenth and early eleventh centuries has received wide support in the secondary
literature. See among others Kiihn, Die byzantinische Armee, passim; Ahnveiler, 'Recherches', pp.2,
23; M.Grigoriou-Ioannides, 'Themata et tagmata. Un probleme de 1'institution de themes pendant les
Xe et Xle siecles', ByzForsch 19 (1993), pp.35-41.
106 De Ceremoniis, p.666; Howard-Johnston, "The Byzantine Army', p. 116
107 De Ceremoniis, p.652; Howard-Johnston, 'The Byzantine Army', pp. 105-112
1 /Ifi
De Ceremoniis, p.666
109 Howard-Johnston, 'The Byzantine Army', p. 112
240
western themes of Asia Minor, during the reign of Basil n himself. 1 10 By the mid-
eleventh century large numbers of mercenary troops from outside Byzantium were
organised into ethnic tagmata and stationed in the provinces. 111 It is reasonable to
assume that these troops were usually paid from the receipts of the fiscalised
strateia first introduced, as we have seen, during the middle of the tenth century. 112
Nevertheless, the increased use of permanent mercenaries within the themes during
the tenth and eleventh centuries does not mean that Anatolian soldiers had entirely
ceased to join the Byzantine armed forces. Some continued their traditional service
within local theme armies, while others either participated in campaigns with the
imperial field armies in enemy territory, or were stationed on the frontiers. At least
two examples of the involvement of troops from the core Anatolian themes in the
military endeavours of the centralised armed forces can be identified during the
reign of Basil. According to a confirmation of a land grant issued to the Athonite
monastery of the Iviron in the mid-990s, a certain John Chaldos was doux of the
Armeniakon, Boukellarion and Thessalonika. 113 Given the proximity of the Athos
peninsular to Thessalonika, it seems safe to assume that Chaldos was stationed at
Thessalonika in the Balkans in command of a garrison staffed by troops from the
110 Nesbitt and Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no. 71.11.
111 It is possible to identify the ethnic roots of these troops, and to locate where they were stationed
from eleventh-century lists of exemptions from the mitaton, the obligation (or commutable tax) to
quarter these soldiers in the winter. In 1044 the exemption lists refer to Rus troops; in 1060 to Rus,
Varangians, Franks and Saracens. By 1088 thirteen ethnic groups were represented (Oikonomides,
'devolution de 1'organisation administrative', p. 144 Ahrweiler, 'Recherches', pp.21-3, 32-4).
112 Ahrweiler, 'Recherches', p.23. More recently Magdalino has pointed out that certain members of
the Norman mercenary tagma settled in the theme of the Armeniakon had, by the mid-1050s,
acquired estates (The Byzantine Background to the First Crusade (Toronto, 1996), pp. 10-11).
However, there is no evidence to demonstrate that they held these estates in return for their military
service. Instead, the oikoi which they bought in the locality were almost certainly acquired from the
salaries which they were paid in their capacity as mercenaries.
113 Actes d'lviron, p. 153. Although the year in which this document was issued is not certain, 995
seems the most likely. Certainly it had to be issued before Chaldos, doux of Thessalonika, was taken
prisoner by the Bulgarians in 996 (Skylitzes, pp.347, 357). Although the Iviron document contains no
annual date, it does indicate that it was issued in the ninth indiction. This indiction tallies with the year
995.
241
Boukellarion theme in north-western Anatolia, and the Armeniakon in the north-
east. Troops from the core of Asia Minor were also to be found on the eastern
frontier. For example, shortly before the death of Basil II in 1025, the
Although none of the sources cited above apply a generic name to the units of
troops they describe, it is likely that they are reporting on early examples of the
thematic (or provincial) tagmata, a phenomenon more usually attributed by
historians to the middle of the eleventh century. These thematic tagmata become
particularly visible in the historical record during the campaigns waged by
Byzantine armies in southern Italy during the early 1040s. For example, in 1042 the
army of Michael Dokeianos, the katepan of Italy, contained a tagma from the
Opsikion, as well as the so-called tagma of the Phoideratoi. The Phoideratoi were
Pisidians and Lykaonians, troops from the theme of the Anatolikon. In the same
period Katakalon Kekaumenos commanded a garrison at Messina in Sicily
described by Skylitzes as the tagma of the Armeniakoi. By 1056 troops from Pisidia
and Lykaonia were called the tagmata of the Anatolikon, and a year later a tagma
242
from the Armeniakon joined the rebel army of Isaac Komnenos. Moving on twelve
Comparison between this list of examples and the cases cited during the reign of
Basil n, suggests that thematic or provincial tagmata recruited from within the core
Anatolian themes were usually deployed on frontiers of the empire, or within the
Byzantine field army. As such the troops who served within these units were, like
other tagmatic troops, full-time professionals. Moreover, the location of their
deployment makes it clear that they were not, as HJ.Kuhn has argued, stationed in
the Anatolian heartland as the direct replacements for the former thematic
1 -I o
armies. Indeed, if there was a time when such full-time professional units were
located in the Anatolian provinces where they were recruited, it was much later,
during the third quarter of the eleventh century, when this region came under
greater threat from Turkish invasion. In contrast, during the first half of the
eleventh century, the lack of an external threat to the prosperous core of Anatolia,
meant that elite troops from these regions could be more profitably employed on
the frontiers.
With the main body of soldiers recruited in Asia Minor serving on the frontiers, and
in the absence of any significant external martial threat, the size of the thematic
armies located in Anatolia probably continued to diminish throughout the reign of
116 Cheynet and Morrisson, 'Lieux de trouvaille et circulation des sceaux', p. 130.
117 Kiihn, Die byzantinische Armee, pp.251-7; see also the brief discussion in Ahrweiler,
'Recherches', pp.29-35. To the examples listed here, can be added the instance of Romanes Skleros
serving simultaneously as doux of the Anatolikoi and domestikos of the scholai of the west during the
mid-eleventh century (Seibt, Die Skleroi, p.82). The most reasonable explanation for Skleros'
combination of responsibilities is that he was head of a field army deployed in the western half of the
empire which included tagmata from the the theme of the Anatolikon.
118 Kiihn, Die byzantinische Armee, p.252
243
Basil, although establishing the exact magnitude of that decline is impossible.
While the documentation of the Cretan expeditions of 911 and 949 allows us to
track the contraction of numbers of active thematic stratiotai in the first half of the
tenth century, no such written evidence exists which facilitates a similar calculation
for the reign of Basil II. Furthermore, it is difficult to know the extent to which
those professional troops from the centralised tagmata and overseas mercenaries
stationed within the themes made up any shortfall in locally-recruited troops.
Nonetheless, while the written record leaves such questions open, evidence from
lead seals suggests that a marked demilitarising of the core Anatolian themes
occurred during the later tenth and eleventh centuries. Most conspicuous within the
sigillographical record is the extent to which those officials within the secretariat of
the strategos who had been most intimately connected with the logistics of active
military service during the ninth and early tenth centuries, either developed new,
fiscally-related responsibilities, or simply disappeared.
Among the officials who seem to have disappeared was the thematic count (komes}
of the tent, whose duties had once included helping to put up the imperial tent
during campaigns. In the course of my research I have been able to find no seals
dated to later than the tenth century for this office. 120 Meanwhile, the seal of an
eleventh- century thematic domestikos, the official in charge of the general staff of
the thematic headquarters, provides further evidence for the contraction of military
infrastructure. By the time he fulfilled his position he was responsible for two
119 A significant decline is postulated by J.V.A.Fine, 'Basil II and the Decline of the Theme System,
Studio Slavico-Byzantina et Medievalia Europensia I (Ivan Duijcev Center for Slavo-Byzantine
Studies, Sofia, 1989), pp.44-7. However, his model of a complete collapse of an active military
presence in the themes may be a little too severe, since it does not account for the incidence of
professional garrisons within the provinces.
120 This official appears within the office of the strategos in the late ninth-century administrative
taktikon of Philotheos (Oikonomides, Listes, pp.109, 153, 341; J.B.Bury, The Imperial
Administrative System in the Ninth Century (London, 1911), p.43). The latest specimen I found dates
244
separate themes, the Opsikion and Anatolikon, rather than simply one as in
previous centuries. 121
and sigillographical evidence. For instance, the Cretan inventory of 949 enumerates
three such officials from the Thrakesion. 122 In addition, we have the seals of a
tenth-century merarches from the Charsianon, and eleventh century tourmarchai
from Seleukia and Paphlagonia. 123 However, other seals demonstrate the extent to
which the responsibilities of even the most senior of thematic military officials had
widened to include non-military duties. For example, one eleventh-century
tourmarches of Abydos supplemented his military duties with the tax-raising
powers of a kommerkiarios, a collector of customs revenues.
1 ^A
An eleventh-
century tourmarches of the Anatolikon was also a dioiketes, a local tax-collector. 125
Although the brevity of the information conveyed on these seals leaves the exact
practice of such fiscal responsibilities ambiguous, they verify the observations
made by the social commentator Kekaumenos that during the eleventh century
many local military commanders frequently undertook tax-raising duties. 126
to the mid-tenth century, and belonged to an official from the theme of the Kibyrrhaiotai (Nesbitt and
Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, ii, no. 59.17).
121 Nesbitt and Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no.86.1. Once
again this official was listed in the late ninth-century taktikon of Philotheos (Oikonomides, Listes,
pp. 109, 157; Bury, The Imperial Administrative System, p.43; see note above).
122 De Ceremoniis, p.666.
123 Charsianon: Dumbarton Oaks Unpublished F3397; Paphlagonia: Cheynet, 'Les sceaux du musee
d'Iznik", REB 49 (1991), no. 13; Seleukeia: Dumbarton Oaks Unpublished 55.1.3409.
124 Nesbitt and Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no.40.13
125 Szemioth and Wasilewski, Sceaux byzantins: Varsovie, no.45
126 Kekaumenos warns strategoi of the risks of taking on fiscal duties (demosiake douleia) of all sorts
(Kekaumenos, p. 154). He cites as an example a salutary case concerning John Maios, a strategos
who decided to manage an estate (episkepsis) belonging to the Orphanotropheion in the expectation
that any surplus profits he accrued would enable him to acquire a bigger house. However, he soon
found himself ruined and imprisoned. The fact that Kekaumenos writing in the 1070s, relates that the
245
A diversification of responsibilties also characterises two other offices associated
with the secretariat of the strategos which persisted into the eleventh century. One
was the chartoularios of the theme, the official who had once been the local
Maios affair happened in the time of his father, suggests that the diversification of the interests of
military officials into civilian offices was certainly current during the first half of the eleventh century
(Kekaumenos pp. 194-6; P.Lemerle, 'Prolegomenes a une edition critique et commentee des 'Conseils
et Recits' de Kekaumenos', Academic royale de Belgique, Classe des Lettres Memoires 54 (1960),
pp. 82-5)
127 Chartoularioi are not only listed in the ninth-century taktika (Oikonomides, Listes,
pp. 109,115,153, 341) but also in the later tenth-century Escorial Taktikon (ibid., p.273). For the
ninth-century duties of the thematic chartoularioi see Bury, The Imperial Administrative System,
p.44; Ahrweiler, 'Recherches', p.43. Seals of eleventh-century chartoularioi are common. A small
sample would include: from the Anatolikon (Nesbitt and Oikonomides, Catalogue of Byzantine Seals
at Dumbarton Oaks, iii, nos.86.11, 86.7; Dumbarton Oaks Unpublished 55.1.2716); from the
Armeniakon (Dumbarton Oaks Unpublished 55.1.2904; F1152); from the Boukellarion (Dumbarton
Oaks Unpublished 55.1.270; 55.1.1568). See above, p.232-3, for discussion of the fiscalised strateia.
246
discussed at greater length below. 129 One was a kourator (manager) of imperial
estates and a horrearios, a government agent concerned with the supply of
performed by many officials attached to the secretariat of the strategos in the core
themes of Anatolia, must be that during the course of the later tenth and eleventh
centuries the underlying principle of provincial government shifted away from
military defence, and turned instead towards the fiscal exploitation of the
localities. There is evidence in the later eleventh-century historical record that this
growth in the judicial and fiscal competences of military officals was closely
connected to a desire on the part of central government in Constantinople to extract
maximum benefit from the greater prosperity of the localities within the Byzantine
empire. According to Michael Attaleiates so overwhelming was Constantine X
Doukas' (1059-67) wish to increase public finances and hear private lawsuits that,
"the Roman state was shaken by specious vexations and sophistic methods and a
1 ^*\
swarm of litigious questions and the complexities of fiscal investigations". As a
result, ".... even soldiers were transformed and instructed themselves in the basics
of such knowledge". 133 The fact that Attaleiates was a propagandist for the emperor
Nikephoros Botaneiates (1078-81), a rival to the Doukas family, means that his
247
comments may be deliberately exaggerated. However, it is worth noting that
a. Judges
A shift away from the provision of local defence towards the more intensive
exploitation of resources by the Byzantine state is also the most reasonable
explanation for the greater ubiquity within the locality, during the tenth and
eleventh centuries, of officials who were purely concerned with judicial and fiscal
duties. Most conspicuous within this new proliferation of civil officials was the
thematic judge or krites. Although judges had always exercised a role within
thematic government, their role throughout the ninth and early tenth centuries had
always been subordinate to that of the strategos. 135 However, several sources of
evidence suggest that by the later tenth and eleventh centuries the position of the
judge within local government began to increase rapidly in signficance, and that as
a result his authority not only began to detach from that of the strategos, but even
began to present a rival threat.
tog KOU auToug (Trparnitrag (AeraftaXeiv KO.I fj&TaiutveTv TO, irpog TI\V rouratv yvf&riv napayovra
I }}
Continuatus, p. 112). For Attaleiates' 'Historia' as a source of Skylitzes Continuatus, see above,
p.38.
135 The mid-ninth century Uspenskij taktikon and the later ninth/early tenth-century taktika of Leo VI
refer to the thematic judge as a praitor, and indicate his subordination to the strategos (Ahrweiler,
'Recherches', p.43). It is likely that the thematic strategos had first taken charge of civil matters in
the second half of the eighth century (Haldon, Byzantium in the Seventh Century, pp. 194-207).
248
At the very simplest of levels there appear to have been more judges in the later
tenth and eleventh centuries than had been the case in earlier centuries. Even the
swiftest perusal of a major seal catalogue indicates a much greater incidence of the
seals of judges, (who may also be called kritai, praitores or dikastai), in the
eleventh century than in earlier periods. 136 Furthermore, the sigillographical
evidence also indicates that the responsibilities of these judges were not limited to
merely judicial affairs, but that they also undertook fiscally-related responsibilities.
Many seals belonging to later tenth- and eleventh-century Anatolian judges indicate
that their owners were also employed by the Genikon, the main bureau in
Constantinople concerned with tax collection. The most frequent office within the
Genikon held by thematic kritai was that of megas chartoularios, the official who
was probably the principal representative of the fisc in the locality. 137 A smaller
number of judges held lesser offices within the Genikon hierarchy such as
i -50
chartoularios, exaktor, epoptes or oikistos.
The correlation of judicial and fiscal responsibilities within the person of one
official, visible within the sigillographical data from later tenth- and eleventh-
136 For example, of the twenty seals of judges of the Thrakesion listed in the third volume of
published seals from the Dumbarton Oaks Collection, five are dated to either the tenth or the eleventh
centuries, while fifteen are dated to the eleventh century alone. In the Opsikion the figure is one from
the ninth/tenth centuries, one from the tenth/eleventh centuries, and ten from the eleventh century; in
the Optimatoi, one from the tenth/eleventh, and three from the eleventh; in the Anatolikon one from
the tenth/eleventh and ten from the eleventh (Nesbitt and Oikonomides, Catalogue of Byzantine Seals
at Dumbarton Oaks, iii, nos.2.11-30; 39.7-20; 71.23-6; 86.23-33).
137 Examples of thematic judges who were also megaloi chartoularioi: from the Anatolikon: Nesbitt
and Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no.86.27; Schlumberger,
Sigillographie, p.266, no.2; from the Armeniakon. Zacos, Byzantine Lead Seals II, nos.432, 797;
from the Boukellarion. Zacos, Byzantine Lead Seals II, nos.376, 388-9, Dumbarton Oaks
Unpublished 58.106.3049, 55.3404-06, 58.106.5703; from Paphlagonia: Dumbarton Oaks
Unpublished 58.106.5642; from the Opsikion. Nesbitt and Oikonomides, Catalogue of Byzantine
Seals at Dumbarton Oaks, iii, no.39.10; from the Charsianon. Szemioth and Wasilewski, Sceaux
byzantins: Varsovie, no.50.
138 Examples of thematic judges who were also chartoularioi of the Genikon: from Paphlagonia:
Schlumberger, Sigillographie, p.299, no.3; from Cappadocia: Dumbarton Oaks Unpublished
55.1.2861; those who held the position of exaktor: from the Opsikion: Nesbitt and Oikonomides,
Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no.39.20; of epoptes: from the Thrakesion:
249
century Anatolia, corroborates the findings of much recent research, which has
suggested that fiscal imperatives were closely associated with the provision of
The most obvious area where fiscally and judicially related competences clearly
overlapped in the single person of the judge was in the enforcement of the
provisions of the 4poor' versus the 'powerful' legislation of the tenth century. This
legislation, which is most easily traced through a corpus of imperial legislation
contained in the appendices of the Synopsis Basilicorum Major, had as its
prevailing theme the defence of the properties of poor landowners of the free
peasant village (the choriori) against intrusions by their more powerful
neighbours. 141 Government concern about this problem first became manifest
Nesbitt and Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no.2.32; of
oikistikos: from the Boukellarion: Dumbarton Oaks Unpublished 58.106.2219.
139 Magdalino, "Justice and Finance", passim. The combination of judicial and fiscal responsibilities
undertaken by eleventh-century judges has been noted elsewhere: Ahrweiler, 'Recherches', pp.70-5.
140 Michel Psellos, (ed.) K.N.Sathas, MB, Vol 5, (Venice-Paris), pp.399-400; Magdalino, 'Justice and
Finance', p.94
141Zepos and Zepos, lus, i, 198-273. Svoronos, La Synopsis Major des Basiliques et ses appendices,
pp. 17-45 (appendix A), pp.91-100 (appendix B)
250
during the reign of Romanes Lekapenos (920-44), when the famine of 927-8 and
the harsh winter of 933-4 had been the catalysts for the 'powerful' to acquire
properties from the 'poor' at prices significantly below market value. In a novel
issued in 934 Lekapenos tried to force the 'powerful' to return their recent
acquisitions to the former owners. 142 This legislation was periodically reissued and
strengthened by his tenth-century imperial successors. Constantine Porphyrogenitus
not only lengthened the time period allowed for the 'poor' to repurchase their
lands from the 'powerful', he also ordered that no compensation was to be paid in
instances where properties had been acquired by force. 143 Nikephoros Phokas
extended the restrictions on acquisitions by the 'powerful' to choria where they
might already hold lands. 144 In 996 Basil n himself introduced a blanket measure,
which stated quite baldly that all property acquired by the 'powerful' from within
free choria since 927 should be restored to its former owners, without
compensation for either the original purchase price or for subsequent
improvements. He also abolished the principle that property ownership was
immune from judicial inquiry after forty years. 145
Although historians have disagreed violently over the motivations for the
introduction and elaboration of this legislation during the tenth century and its
intended targets, it is generally agreed that this conflict should not simply be
interpreted as the high-minded, disinterested protection by the emperor of the
142 Zepo and Zepos, lus, i, 205-214; Theophanes Continuatus, pp.443, 448-9; fLMorris, 'The Poor
and the Powerful in Tenth-Century Byzantium: Law and Reality', Past and Present 73 (1976), pp.8-
10; P.Lemerle, The Agrarian History of Byzantium from the Origins to the Twelfth Century: the
Sources and Problems (Galway, 1979), pp.94-7; Kaplan, Les Hommes et la terre, pp.421-26.
143 Zepos and Zepos, lus, i, 215, 242; Lemerle, Agrarian History, pp.97-8; Kaplan, Les Hommes et
la terre, pp.426-7, 431-2.
144 Zepos and Zepos, lus, i, 254; Lemerle, Agrarian History, pp. 100-2; Kaplan, Les Hommes et la
terre, pp.434-5.
145 Zepos and Zepos, lus, i, 263; Lemerle, Agrarian History, pp. 104-5; Kaplan, Les Hommes et la
terre, pp.437-8.
251
interests of the 'poor' against the territorial rapacity of the 'powerful'. 146 Instead,
prevent the erosion of the fiscal integrity of the chorion, the main tax-paying unit in
the localities, and to ensure that the landed and manpower resources of the locality
remained in the service of the state, rather than being transferred to the private
For the authority of such persons [the 'powerful'] has exulted over the great
misery of the 'poor', by the number of their servants, their hirelings and those
otherwise attending and accompanying their prominent positions, which brings
in prosecutions, forced services, .....oppressions and distresses, and has
introduced no little destruction of the common good ...... For the habitation of
the multitude shows the great profit of its employment, the collective
contribution of the taxes, the joint rendering of military services, which will be
entirely lacking when the common people have perished. 147
One of the most signficant implications of this legislation for the development of
local administration during the tenth and eleventh centuries was the role which
that at the very end of his novel, Romanos Lekapenos explicitly identified judges as
146 According to one view, most forcefully advocated by Ostrogorsky, the anti-'powerful' legislation
represented an attempt by the Macedonian imperial dynasty to prevent the sequestration of the fiscal
and military base of the Byzantine state by an aggressive land-based aristocracy (Ostrogorsky,
'Agrarian Conditions in the Byzantine Empire', pp.215-221). In contrast to Ostrogorsky's views,
another argument maintains that this legislation had little to do with the curtailing of an overweening
feudal aristocracy. Instead, its main purpose was to preserve the fiscal integrity of the Byzantine
state. For Lemerle, Agrarian Historyr, pp.85-115, the chief proponent of this second view, the main
characteristic of the Byzantine fiscal system was the joint contribution of taxes paid by 'poor'
landowners living within free choria. In these circumstances, invasions by the 'powerful', whom
Lemerle defined as public officials rather than feudal aristocrats, meant that the lands from the
chorion might easily be consolidated into private estates, idiostata, and thus become detached from
the basic tax unit. Lemerle's views have been rearticulated at greater length by Kaplan, Les Hommes
et la terre, pp.414-43. A middle position has been taken by Morris, "The Poor and the Powerful",
pp.3-27, who argues that the novels of the tenth century represented a flexible tool which emperors
could wield against a variety of targets whenever their authority was threatened. She suggests that
the legislation was customarily issued at times of crisis for imperial authority: in 934 after a series of
natural disasters; in the reign of Constantine Porphyrogenitus, when the emperor was trying to free
himself of those families who had supported his palace coup in 944/5; and in 996 as part of a package
of retribution against the families involved in the Phokas and Skleros revolts of 976-89.
147 Zepos and Zepos, Ivs, i, 209; translation from C.M.Brand, Icon and Minaret: Sources of
Byzantine and Islamic Civilization (Englewood Cliffs, New Jersey, 1969), pp.83-4.
252
those who were to enforce the protection of the poor: "let each of those who
happens to be a judge constantly maintain these things." 148 Nor was his mandate
mere rhetoric. Various written sources indicate that by the mid-tenth century
localities, dealing with the problems identified in the legislation issued by Romanes
and his successors. Theophanes Continuatus noted that, "having ringing in his ears
the injustices and losses which the pitiable and wretched poor had suffered",
Constantine Porphyrogenitus, "sent pious and virtuous men to lighten the great
burdens of the suffering poor." Those "pious and virtuous men" dispatched to the
principal provinces of central and western Anatolia are named: "And to the
Anatolikon he sent the magistros Romanos Saronites, and to the Opsikion the
magistros Romanos Mousele, and to the Thrakesion thepatrikios Photios and to the
Armeniakon Leo Agelastos". 149
253
in Bithynia, the metropolitan of Klaudioupolis, the bishop of Krateia and the 'poor'
Recently, Michel Kaplan has argued that the issuing of the anti-powerful legislation
of tenth century and its continuing enforcement in the eleventh, indicates that the
Constantinopolitan authorities were engaged in a losing battle to protect the
integrity of the fisc. 155 Yet, it can be argued that the extension of imperial authority
into the localities in the shape of fiscal and judicial officials, was not simply a
defensive and rearguard action on the part of the Byzantine authorities. Although
this is the sentiment encouraged by the defensive tone of the legislation, and by the
tendency of modern scholars to study the novels as an hermetically sealed corpus of
evidence, it should be remembered that the laws themselves reflect an entirely
imperial and Constantinopolitan view-point. When this evidence is set against
broader social and economic contexts, it becomes clear that the promulgation of the
anti-"powerful" legislation, and the civilianising of the localities which that
legislative impulse precipitated, were not merely a defensive and negative response
254
from the imperial authorities. Instead, they also reflected a desire that the authority
of the emperor should be imposed more directly on the localities, not least so that
the state could be a direct beneficiary of the increasing prosperity of urban and
rural western and central Asia Minor.
That the penetration of civilian officials into the provinces was seen in this light by
contemporaries is made clear by the author of the military manual 'On
Skirmishing', produced either during or shortly after the reign of Nikephoros
Phokas (963-9). He alludes disparagingly to the appearance within the locality of
"tribute levying mannikins who contribute absolutely nothing to the common good
[but] ................. store up many talents of gold", and to thematic judges who
dishonour the indigenous stratiotai. More crucially, in his adamant insistence on
the antiquity of the claims ofstrategoi to supreme governance over their themes, he
suggests that these new civilian officials were beginning to present a threat to the
traditional military hierarchy of thematic government:
The law itself stipulates that each officer has authority over his own men and
can judge them. Does anyone else have authority over the men who live in the
theme beside the strategos alone whom the holy emperor has appointed? For
this reason, from the most ancient Romans and from the law, the strategos
possessed authority over his own theme. He judges cases in matters that affect
the soldiers, and he manages affairs that come up in them. He has a judge
(krites) to co-operate with him and with whom he too co-operates. He also co-
operates with the protonotarios and the others appointed in the service of the
fisc (ton demosiou). 156
255
Yet, as we saw in the previous section of this chapter, by the eleventh century those
officials within the secretariat of the strategos, who had traditionally performed
b. Fiscal lands
former rate. Yet, by the end of the tenth century there is evidence from
Constantinople that imperial authorities were becoming more interested in the
direct exploitation of these assets rather than in their immediate sale. By 972/3 a
department can be detected operating within the Genikon, the main fiscal revenue
department, which dealt specifically with the fiscal receipts of klasmata once they
had been sold. 159 By 1030 this office, the sekreton ton oikeiakon, had become
independent. This development has been attributed to Basil n. It is widely believed
that he chose to end the sale of klasmata, and deciding instead to use tenant
256
farmers, known as state paroikoi, to cultivate fiscal lands directly. 160 The sudden
appearance of seals belonging to the staff of the epi ton oikeiakon, which can be
dated to the turn of the tenth-to-eleventh century, certainly corroborates the view
It has been argued by Oikonomides that the major impulse behind this greater
imperial enthusiasm for the direct management of fiscal lands was a sudden and
exponential growth in the amount of immovable property controlled by the state
during Basil's reign. This rapid expansion of fiscal lands was caused, so
Oikonomides maintains, by the emperor's ready willingness to confiscate the
estates of those landowners who contravened the anti-'powerful' legislation, and
those who had fought for the rebels during the civil wars of Skleros and Phokas at
the beginning of the reign. 162 Here, at first sight, the evidence appears to be strong.
A growth in the lands held by the fisc as the result of confiscation could,
theoretically, have begun in 979. When the Skleros parry fled to Buyid Iraq in
March 979, their estates may have been sequestered by the state. The fisc could
also have been further enriched in 985 with the fall of Basil's great uncle, the
parakoimomenos Basil Lekapenos. Both Leo the Deacon and John Skylitzes
indicate that during the reigns of Basil's imperial predecessors, the
parakoimomenos had secured estates for himself in southern Cappadocia and
Cilicia. As a result, "nothing worth speaking of .......... was left to the fisc
160 Oikonomides, 'devolution de 1'organisation administrative', pp. 136-7; idem., 'Terres du fisc et
revenu de la terre aux Xe et Xle siecles', in V.Kravari, J.Lefort, C.Morrisson (eds.), Hommes et
richesses (1991, Paris), pp.321-2; Kaplan, Les Hommes et la terre, p.321
161 Laurent, Vatican, pp. 15-16, reflects on the contrast between the silence of literary sources on the
eixstence of this office before the 1030s, and the frequency with which seals belonging to officials
from this bureau begin to appear after 1000.
162 Oikonomides, 'L'evohrtion de 1'organisation administrative', pp. 136-7; idem., 'Terres du fisc et
revenu de la terre aux Xe et Xle siecles', in V.Kravari, J.Lefort, C.Morrisson (eds.), Hommes et
Richesses (1991, Paris), pp.321-2
257
[demosion]" , 163 Leo identifies Longias and Drizes as the location of these
parakoimomenos was dismissed from office in 985, it is likely that these lands were
returned to the management of the fisc. Furthermore, it is possible that at the time
of Lekapenos' dismissal, his supporters were also deprived of any land grants they
had received from him. According to Basil's novel of 996, all chrysobulls issued
while the parakomoimenos had been in power were to be considered invalid, unless
they had been checked and confirmed by the emperor himself. The fact that this
novel also indicates that this order had first been issued at the time of the
parakoimomenos' dismissal in 985, suggests that any attack on the lands of the
Lekapenos nexus could have been underway for a full decade before Basil issued
Furthermore, the novel of 996 demonstrates that Basil's attack on the 'powerful'
landowners was not simple rhetorical bluster but was implemented at a practical
level. That is to say, it indicates that some territory certainly was confiscated by the
emperor. The main text of the novel itself contains an exemplary story about one
Philokales, a leading court official, who had used the influence and wealth he had
accrued while he was an imperial protovestiarios, to buy all the lands within his
native village (choriori). As a result the village itself was turned into his private
258
estate (proasteiori). Since this consolidation of immovable property flew in the face
of all but his original holding. 166 The location of Philokales' estate is not clear from
Basil's legislation. However, the novel of 996 provides firm proof that landowners
in Asia Minor could easily fall victim to confiscation. In a marginal note to the
novel, the emperor asserts that the Mousele family were among those 'powerful'
malefactors found guilty of using their tenure of public office for the acquisition of
landed estates. As a result of their abuses, the emperor deprived the family of their
properties. These estates lay near Philomelion, close to the Pisidian lakes in the
western plateau region of Anatolia. 167
The details furnished by this scholium do not make it clear whether the Mousele
family suffered the fate of sequestration simply because they disobeyed the anti-
' powerful' legislation, or because they had participated in the revolts of Skleros and
Phokas. However, if evidence from elsewhere in the same novel is set in the
context of the historical record, it becomes clear that it was long-standing
opponents of the emperor who were most likely to be accused of contravening the
anti-"powerful" legislation and forced to suffer the appropriation of their property.
In another scholium to the 996 novel the emperor cites the Phokas and Maleinos
families as members of the 'powerful' who had built estates in defiance of the
1 fSi
legislation of his imperial predecessors. As we saw in chapter three, by rebelling
against the emperor in 987 the Phokas family had presented the most serious threat
to imperial authority during the first thirteen years of Basil's reign. The Maleinos
family were their close allies and fellow rebels. Indeed, it was at Eustathios
one possible casualty of Basil's review of the chrysobulls issued by Basil Lekapenos. See above,
pp. 169-70
166 Zepos and Zepos, lus, i, 265
167 Ibid., pp.266-7
259
Maleinos' house in the central Anatolian theme of the Charsianon that Bardas
Phokas had been declared emperor in August 987. 169 Although the scholium to the
novel merely identifies the families as 'powerful' transgressors, it does not indicate
Yet, while evidence from the novel of 996 provides clear evidence that political
opponents of the emperor and those who contravened the anti-'powerful'
legislation, (frequently one and the same individual or family), were liable to have
260
their estates confiscated, I believe it would be premature to suggest that such
the imperial legislation and were dismembered, their component parts were
returned to their original 'poor' landowners rather retained by the fisc. The novel of
996 indicates that the individual properties comprising the estate of Philokales were
given back to their original village owners. 172 Moreover, it is possible that many
territorial confiscations, even those properties seized from the emperor's leading
political opponents, were much less significant in size than some historians have
assumed. This is because relatively few of those families who rebelled against Basil
in the first thirteen years of his reign were substantial land-owners. As we saw in
chapter three of this thesis, it was the resources that a rebel like Bardas Skleros
could command in his capacity as a regional army commander, namely taxes, the
army and diplomatic alliances, which formed the basis of his authority, and
provided him with the practical means to foment revolt. At no point in his first
rebellion, nor when he returned to the empire after ten years in exile in Baghdad,
does Skleros' political threat appear to have resided in his tenure of immense
private resources. 173
Yet, even if the Skleros party did own substantial immovable property in the
eastern half of the Byzantine empire, it is clear from the historical record that the
emperor did not sequester their lands on a permanent basis, either during or after
their revolt. Instead, those Skleros adherents who surrendered to the emperor in 979
at the end of the first revolt were merely fined. 174 Because of his importance to
relations between the Hamdanids of Aleppo and the empire, one of Skleros' key
261
supporters, Koulei'b, the basilikos of Melitene, did not even suffer this fate. Rather,
he was pardoned for his role in the revolt, and allowed to keep the estates he had
been awarded when he had first entered Byzantine service in 975. 175 Moreover,
when the inner core of the Skleros party finally surrendered to the emperor in
989/90, they too may have been allowed to keep their estates. Psellos tells us that
they were allowed to retain their offices, titles and kteseis 116 Thus, the emperor
does not seem to have been a direct material beneficiary of the end of the Skleros
insurrection.
Furthermore, despite the draconian tone of the emperor's legislation, the extent to
which Basil's other political opponents suffered a substantial loss of landed estates
is not at all certain. Even the Phokas family do not seem to have been entirely
dispossessed in the aftermath of the 987-9 revolt. The Armenian historian Aristakes
Lastivert noted that Nikephoros, the son of Bardas, was deprived of his official
commands in 989, but says nothing of his lands. Yahya ibn Sa'id claimed that
Nikephoros actually received new estates from the emperor. 177 Since the Phokas
family were so closely identified in the scholium to the 996 novel with the
Maleinoi, whose estates certainly were confiscated, it seems likely that the emperor
may have appropriated at least some of the Phokas lands as well. However, this
punishment may not have involved the complete sequestration of the Phokas
The fate of Lekapenos property in the period after the fall of the parakoimomenos
certainly supports the idea that for all his bluster, the emperor's practical measures
Al-Rudhrawari, p.23
Ibid., pp.23-4; see below, pp.331-2, for Kouleib and his role within Byzantine government on the
eastern frontier.
176 Psellos, p. 16
262
were often more moderate. For example, Michael Psellos gives a dramatic account
of Basil's decision to demolish a monastery patronised by Basil Lekapenos at the
time of the parakoimomenos* fall. 178 Yet, the destruction of this monastery was
never fully completed. In 985, the year of the dismissal of Lekapenos, the
monastery continued to produce manuscripts. 179 Seals indicate that the foundation
survived until the twelfth or thirteenth centuries. 180 Even Psellos himself hints that
the destruction was partial:
The emperor now wished to raze this edifice to the ground. However, since he
was careful to avoid the charge of impiety, only certain parts of the monastery
were removed, and not all those at once. 181
In fact Basil only removed the mosaics and the interior furnishings.
The idea that the reign of Basil n may not have seen an exponential increase in
fiscal estates in many parts of the empire, receives some support from the
sigillographical evidence. There is little sign from the lead seal record, for example,
that the fisc controlled vast swathes of the heartland of Asia Minor during the later
tenth and eleventh centuries. Although it is true that large numbers of seals were
issued by the central, Constantinopolitan bureau of the epi ton oikeiakon in this
period, it is noteworthy that very few of these items belonged to officials based in
Anatolia. 182 While it is difficult to rationalise this mismatch between the small
263
number of lead seals in the Anatolian provinces, and the large number produced by
the bureau in the capital, there is one plausible explanation for the anomaly. This
would argue that the growth of the agency based in Constantinople was connected
more to the longstanding problem of what to do with deserted lands in the Balkans,
In this context it is worth noting that the only substantial local evidence for the
development of fiscal lands and state paroikoi in the tenth and eleventh centuries
comes from the archives of the monasteries of Mount Athos, located on the Chalkis
peninsular east of Thessalonika. 183 Yet, the amount of land in the hands of the fisc
may have been much higher here than elsewhere in the Byzantine empire, not least
because this region in southern Macedonia had been vulnerable to hostile attack by
external enemies for much of the tenth century, and as a result, to persistent
desertion by cultivators. These problems continued during the reign of Basil. Until
the annexation of Bulgaria in 1018, raids into Macedonia and western Thrace
conducted by the forces of Tsar Samuel Kometopoulos were frequent, particularly
during periods of domestic unrest in Byzantium. 184 In an order issued by the
patriarch Nicholas Chrysoberges, which united the defunct monastery of Gomatou
near the Macedonian town of Hierissos with the Athonite monastery of the Lavra,
reference is made to the damage caused to the theme of Thessalonika by Bulgar
raids. This order was dated to 988, the period when the rebel armies of Bardas
Phokas were encamped on the Asian side of the Bosphoros. 185 The testimony of
Skylitzes indicates that such Bulgarian incursions continued until at least 1014, and
probably only came to an end when Basil II annexed the empire of John Vladislav
in 1018. Nor were regions along the Byzantino-Bulgarian frontier merely raided.
264
They were also the victims of depopulation. Skylitzes' account contains several
during the reign of Basil. 186 Nor were the Bulgars the only threat. The islands and
shores of the Chalkis peninsular were plundered by Arab pirates during Basil's
1 ft*7 .
reign. One attack in 993 left the islands of Gymnopelagisia deserted. Disorder in
the locality was even fomented by those forces fighting for the Byzantine army. At
some point between 986 and 1005/6, while Basil n was campaigning in the
Balkans, some hegoumenoi from monasteries on Mount Athos were attacked by a
group of Hungarians, mercenary soldiers [hypospondoi] in the pay of the Byzantine
army. 188
Yet, while it is easy to see how this unstable political situtation in Macedonia
resulted in large numbers of deserted lands, or klasmata, which had to be sold or
managed by the fisc, the situation in the themes of Anatolia could hardly be more
different. Secure from external attack and increasingly prosperous, these were
regions where the number of klasmata must have been signficantly lower than in
the regions around Thessalonika. If there were few klasmata to manage, and if, as
we have argued above, relatively few large estates were confiscated from the
emperor's political opponents, then it is hardly surprising that only a small number
of officials connected to the epi ton oikeiakon seem to have been active in the core
themes of Anatolia.
265
c. Crown lands
Another branch of the Constantinopolitan civil administrative machinery which
appears to have been overhauled in the later tenth or early eleventh century was the
management of crown lands. In contradistinction to fiscal lands which provided
revenue to the state, crown estates were part of the private domain of the emperor.
Their income was used to support his personal expenditure. 189 In addition, many
crown lands also bore the running costs of those imperial foundations (oikoi)
established in Constantinople, which fulfilled monastic, charitable, and memorial
i on
functions. During the ninth and tenth centuries management of all these estates
had been relatively diffuse. Contemporary administrative taktika suggest that in this
period there were two officials at the helm of the management of most of the
properties owned and directly cultivated by the crown: the megas kourator and the
kourator of the Mangana. Each had agents in the localities, including managers of
individual estates, episkeptitai (inspectors), and regional co-ordinators, kouratores
or basilikoi. 191 Yet, as Michel Kaplan has pointed out, these two officials did not
have a monopoly on the administration of crown lands. Any bureau within
Constantinopolitan central government during the ninth and early tenth centuries
could be in administrative control of imperial lands anywhere in Byzantium. 192
266
kourator to the megas oikonomos of the euageis oikoi m Circumstantial evidence
(Oikonomides, Listes, p.270). It is mentioned for the last time in an Athonite document of 1012. The
first mention of the oikonomos of the euagoi oikoi occurs in a document issued by the patriarch
267
There is some evidence from Anatolia that centralisation in Constantinople directly
affected the organisation of crown properties in the provinces. This evidence comes
from the western littoral of Asia Minor. This was an area where crown lands are
attested as early as the tenth century. The kourator of an imperial estate is
mentioned in the life of Saint Paul, a mid tenth-century abbot of the monastery of
Xerochoraphion, located near Priene just north of Mount Latros. 199 The life of Saint
Nikephoros, metropolitan of Miletos in the 960s and 70s, indicates that estates
belonging to the Myrelaion, the foundation sponsored by the emperor Romanos
Lekapenos (920-44), were to be found in the region around Miletos.200 However, it
is clear from documentary evidence contained in the archive of the monastery of St
John on Patmos, that by the eleventh century estates located in the Miletos region
were routinely managed by the new bureau of the euageis oikoi. In 1072
Andronikos Doukas, the brother of the emperor Michael VII Doukas, was granted
the episkepsis of Miletos, part of the episkepsis of Alokepai, and the estate
(proasteiori) of Barys. The former managers of these properties had been local
representatives of the euageis oikoi 201 At about the same date, another imperial
estate under the control of the euageis oikoi, was given to the monastery of
Xerochoraphion near Mount Latros. In this instance the local official charged with
268
the administration of the properties was Niketas Chalkoutzes the episkeptites of the
Maeander.202
A paucity of evidence makes it difficult to know to what extent other crown lands
in Anatolia came under the jurisdiction of the new centralised bureaus in
Constantinople concerned with estate management. However, the sigillographical
archive indicates that crown estates were to be found all over Anatolia in the tenth
and eleventh centuries. Seals belonging to kouratores and episkeptitai survive from
*\f\'5
almost every theme in western and central Asia Minor. Moreover, the
geographical designations on these seals indicate that imperial estates were located
in regions ideally situated to reap the benefit of the growing commercial prosperity
of Anatolia. The eleventh-century seal of an episkeptites and xenodochos from
Nikaia suggests that xenodochia (hostels for merchants) in north-west Asia Minor,
which had been managed by the bureau of the megas kourator in the late ninth
century, remained imperial possessions.204 As a result imperial authorities were in a
perfect position to benefit from any increase in commercial trade which passed
*
269
Furthermore, imperial camps (aplekta) located on the western edge of the plateau,
where imperial armies had been accustomed to muster in the ninth and tenth
centuries on their way east to campaign against the Arabs, seem to have remained
imperial episkepseis during the later tenth and eleventh centuries. Thus, Malagina
located on the Lower Sangarios, identified in the ninth century as the first staging
post for the armies marching eastwards, was an episkepsis during the reign of
Romanes IV (1068-71).206 In the early stages of the first Skleros revolt, the battle of
Rhageai was fought near the imperial estate of Dipotamon/Mesanukta in the region
of the Pisidian Lakes.207 An eleventh-century seal points to an episkepsis at
Dorylaion, another ninth-century aplekton and key fortress town on the western
edge of the plateau. The editors of this seal suspect that the owner may have
controlled those estates [or mitata} in Phrygia which once bred livestock for
">/*\Q
imperial expeditions. Given their position on key route systems and their stocks
of good grazing ground, it is possible that the imperial aplekta may have been
turned into farms which raised livestock for the commercial market.
Yet, while there is strong evidence to suggest that imperial estates were to be found
everywhere in Asia Minor during the tenth and eleventh centuries, it is impossible
to know whether the stock of crown land in Anatolia increased substantially during
204 The xenodochia controlled by the state, according to the late ninth-century taktikon of Philotheos,
were located in Nikomedia, Pylai and on the Sangarios river (Oikonomides, Listes, pp. 122-3). For the
Nikaia seal see Schlumberger, Sigillographie, p.250, no. 1
205 See above, 173-80, 193-8.
206Military expeditions: Haldoa p. 75. Attaleiates mentions that Malagina was the site of an
episkepsis in the reign of Romanes IV under the control of a chartularios (Attaleiates, p. 124). For an
eleventh-century seal of a chartoularios of Malagina see Ebersolt, Musees imperiaux ottomans,
no.325.
207 Skylitzes notes that the imperial estate was called Dipotamon, but that the local inhabitants called
it Mesanukta (Skylitzes, p.320). The sillographical record contains the contemporary seals of a
protokourator and an episkeptites of Mesonukta (Nesbitt and Oikonomides, Catalogue of Byzantine
Seals at Dumbarton Oaks, Hi, nos.58.1-2)
208 Nesbitt and Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, Hi, no. 71.22
270
this period, or whether it was more intensively exploited. In other words, it is
impossible to tell whether the greater centralisation of administrative practice was
out that there is some evidence which countermands the idea that the greater
centralisation of the management of crown estates necessarily precipitated a large
increase in revenue. When Andronikos Doukas received the imperial estate of
Barys near Miletos in the early 1070s, the property was said to be very run down,
with vines, olives and orchards overgrown.210 These observations are most
surprising given the general prosperity of this coastal region in the eleventh
century. In these circumstances it is tempting to argue that imperial authorities may
not have possessed the resources of capital or manpower to exploit their immovable
assets to the full.
/'/'/. Conclusion
Any attempt to reconstruct the administrative history of Anatolia during the reign
of Basil n is hampered by a lack of written source materials. The thin historical
record provides little more than the names of a handful of local strategoi. Other
literary sources, such as saints lives, imperial novels and military manuals, provide
only occasional references which are of direct relevance to bureaucratic
developments in Anatolia. However, once these literary clues are examined in the
light of the more substantial sigillographical record, some preliminary conclusions
about the administration of this region during the later tenth and early eleventh
*y/\Q __
For the argument that there was a substantial growth in the stock of crown land elsewhere in the
Byzantine empire during the second half of the tenth century, above all in the former emirates of the
eastern frontier, see the next chapter, p.328.
271
centuries can be drawn. Above all, it becomes clear that Basil's reign represented
the continuation of many deep-seated administrative trends which were already
underway long before the emperor himself came to the throne in 976. The most
fundamental of these trends was the shift away from an administration based on the
needs of defence, towards a form of government devoted to the exploitation of
The greater frequency with which military service was commuted during the later
tenth and early eleventh centuries is one sign of how the imperative behind local
administration began to move away from the provision of defence, towards the
greater financial exploitation of the resources of Anatolia by the emperor in
Constantinople. Another sign of the desire of central government to tap into the
210 Whittow, 'Social and Political Structures in the Maeander Region of Western Asia Minor',
pp.388-9
272
growing prosperity of Asia Minor is the greater prominence within thematic
government of a variety of officials with revenue-raising responsibilities. Many
including judges and the managers of fiscal and crown lands, became more
prominent in local administration. However, the surviving sigillographical record
suggests that an important distinction needs to be drawn. Although civil officials of
all guises emerge with greater frequency in the Anatolian leads seals of the later
tenth and eleventh centuries, it is those functionaries responsible for the indirect
exploitation of the locality rather than those charged with the direct management of
immovable resources, who seem to have been more ubiquitous. Thus, the
sigillographical record suggests that the civil administration of Anatolia seems to
have been characterised more by judges than by officials attached to the office of
the epi ton oikeiakon or the euageis oikoi.
Until further research into the sigillographical record of tenth- and eleventh-century
Anatolia can be conducted, this perceived imbalance within the incidence of civil
officials can be no more than impressionistic. However, if substantiated by
subsequent investigation, this finding has important repercussions. It suggests, in
the first place, that the imperial authorities in Constantinople preferred to extract
material gain from the growing prosperity of Anatolia by the indirect means of
taxation rather than by the direct management of resources. This, in turn, implies
that during the tenth and eleventh centuries most assets of land and manpower in
Anatolia remained in private hands, and were not subsumed by the Byzantine state.
It is also a conclusion which clearly countermands the notion that Basil n himself
the fisc, by confiscating the lands of his principal political opponents and those
273
by the 'powerful'. As this chapter has demonstrated, there is little strong evidence
for the widespread confiscation of landed property during the reign of Basil.
Instead, sequestration was a judiciously employed political tool rather than part of a
sustained policy to enhance the property-base of the state. In the day-to-day
governance of the Anatolian themes, tenth- and eleventh-century emperors,
including Basil n, preferred to exploit the localities through the use of taxation
rather than through the direct management of imperially-owned assets.
274
Chapter Six
I. Introduction
In the second and third quarters of the tenth century Byzantine armies had taken
advantage of the demise of the Abbasid caliphate, and extended the empire's
territorial boundaries over the Taurus and Anti Taurus mountains into Cilicia,
northern Syria, and northern Mesopotamia, thus territorially redefining a Byzantine
east that for the previous three centuries had been limited to the Anatolian plateau.
Yet, when Basil himself came to the throne in 976, few of these annexations had
been consolidated. Melitene, conquered in 934, the oldest of the great territorial
gains of the tenth century, had been under Byzantine suzerainty for less than half a
century. Most permanent conquests were even more recent. Cilicia, absorbed in
964-5, had been under Byzantine rule for little more than a decade. Antioch in
northern Syria had only surrendered to the Byzantines in 969. Other sites in
northern Syria, such as Barzouyah and Saoune, ceded to the Byzantines during the
emperor John Tzimiskes' last eastern campaign in 975, had been under imperial
rule for less than a year when Basil assumed the imperial purple. 1
This chapter considers how imperial authority was consolidated in these eastern
territories during Basil's reign. The first part of the chapter will deal with military
arrangements on the frontier. The second with those matters traditionally perceived
275
to be civilian. The principal question asked will be whether Constantinople sought
to control the periphery according to rigid preconceived administative formulae, or
whether, in the context of the heterogeneous nature of the local populations of these
regions, it admitted a more flexible and devolved relationship. Although discussion
will be primarily concerned with institutions and processes within the empire itself,
the chapter will begin with a brief overview of Byzantine relations with its eastern
neighbours, the crucial military and diplomatic backdrop against which
adminstration on the frontier developed.
/. Muslim neighbours
When Basil came to the throne in 976, the most threatening of the empire's eastern
neighbours were the Fatimids, a militant Shia dynasty from north Africa, who not
only possessed a powerful land army composed of Berber forces from the Maghrib,
but also a large fleet.2 In 969/70 this expansionist Muslim power had invaded and
occupied Egypt. By the autumn of 970, Fatimid armies had seized Palestine and
Syria, and their sights were set on Byzantine northern Syria. During the winter of
970-1 they besieged Antioch for five months. Although this siege was swiftly
disbanded once a Byzantine relief force arrived, Fatimid armies constituted a
perennial threat to Byzantine security in northern Syria during the first half of
Basil's reign, with competition between the two powers focused on two strategic
targets: first, the coastal ports of northern Syria and Lebanon, such as Laodikeia,
See Map 6 for all place names and geographical references in this chapter.
2 For a detailed discussion of imperial relations with Fatimid Egypt and Aleppo, Byzantium' client
state in northern Syria, in Basil's reign, see: Forsyth, 'The Chronicle of Yahya ibn Sa'id', pp.416-
423, 478-557; W.Farag, 'The Aleppo Question: a Byzantine-Fatimid Conflict of Interest in Northern
Syria in the Later Tenth Century', BMGS 14 (1990), pp.44-60; Bikhazi, 'Hamdanid Dynasty of
Mesopotamia and North Syria', pp.973-80; Schlumberger, L'Epopee byzantine, i, chapter 9; ii,
chapters 2,3,7; Whittow, Making of Orthodox Byzantium, pp.367-9, 379-82; H.Kennedy, The
276
Balanias, Tripoli, Tyre, and Beirut; and second, Aleppo, the Hamdanid emirate in
northern Syria which had become a Byzantine, tribute-paying, client state in the
year 969/70.
During Basil's reign the most sustained Fatimid offensive in northern Syria
occurred during the first half of the 990s. In this period Aleppo was repeatedly
besieged, and Byzantine armies based at Antioch under the leadership of Michael
Bourtzes were defeated in open battle in 992 and 994. The second of these reverses
was so serious that Basil II marched from Bulgaria with a detachment of the
Byzantine field army, crossed Anatolia in a little over two weeks, and arrived
unexpectedly in northern Syria in the early spring of 995. The Fatimid army fled in
his path. However, this was not the only occasion during the Byzantino-Fatimid
conflict of the 990s that Basil was forced to intervene personally. In 998 Byzantine
forces based at Antioch under the command of Damian Dalassenos suffered another
defeat in open battle. A year later Basil once again marched east. This time he
ravaged Fatimid-held territory in the southern reaches of the Orontes valley before
cutting westwards to the coast in order to invest Tripoli. Although the siege was
unsuccessful, Basil's swift military response to the defeat of Dalassenos persuaded
the advisors surrounding the young Fatimid caliph, al-Hakim, to come to terms.
The resultant peace lasted without serious rupture from 1000 to 1016. It was
unshaken even by al-Hakim's destruction of the Church of the Holy Sepulchre in
Jerusalem in 1009. Although competition was renewed after Fatimid forces
occupied Aleppo in 1016, this rivalry appears to have remained highly localised,
Prophet and the Age of the Caliphates (London, 1986), pp.282-4, 318-338. See above, pp.226-7, for
the dangers presented by the Fatimid fleet.
277
and did not disintegrate during the remainder of Basil's reign into the full-scale,
armed hostilities which characterised the pre-1000 period.3
Further east, in the Djazira, on the frontier which stretched along the Upper
Euphrates and Upper Tigris rivers, stiff military resistance could still be offered to
Byzantium at the beginning of Basil's reign by local Muslim powers in
Mesopotamia and northern Iraq. 4 For example in 973, less than three years before
Basil came to power, the Hamdanid emirate of Mosul defeated a Byzantine field
army at a battle near Amida, in the Diyar Bakr, south of the Anti Taurus mountains.
In the course of the engagement, the general at the head of the Byzantine army,
Melias, the domestikos of the scholai, was taken captive, and subsequently died in
prison. Although the Hamdanids were expelled from Mosul and the Diyar Bakr
cities of Mayafariqin and Amida in 978-9, shortly after Basil's reign began, they
were replaced by a more potent military threat: forces loyal to Adud al-Dawla.
Adud was a member of the Shia Buyid clan which controlled most of the Iranian
plateau. He himself was the emir of Fars in southern Iran. After ousting his cousin
Izz al-Dawla from power in Baghdad in the spring of 978, he also controlled Iraq.
For the next five years, until he died in 983, Adud threatened to unify a vast
territorial land-mass, which stretched from Persia in the east to the Diyar Bakr in
the west. More important, although Adud's hegemony was brief, it occurred at the
very beginning of Basil's reign, when the emperor himself was threatened by the
first revolt of Bardas Skleros. Indeed, Adud was enthusiastic to exploit his imperial
neighbour's insecurity for his own territorial gain. He gave Skleros sanctuary in
Baghdad when the Byzantine rebel general had been defeated by imperial armies in
3 See above, p.201, and n.172 for discussion of the trade embargo which accompanied the local
deterioration of relations after 1016.
4 Relations between Byzantium and Buyid Iraq are outlined in: Forsyth, 'The Chronicle of Yahya
ibn Sa'id', pp.393-416, 478-557; Canard, 'La date des expeditions mesopotamiennes de Jean
Tzimisces', pp.99-108; idem., "Deux documents arabes sur Bardas Skleros", pp.55-69; Whitlow,
278
979. Adud's hope was that he could use Skleros as a pawn either to recover a series
of mountain fortresses in the Diyar Bakr from the Byzantines, or to realise an even
greater territorial ambition, control over Aleppo in northern Syria. Even after Adud
died, Buyid Baghdad still sought to use Skleros as a political pawn. When the
general returned to Byzantium in 987 he was sponsored by Buyid gold. 5
Nonetheless, by the time that Skleros left Baghdad and re-entered the empire,
Buyid power on the Byzantine frontier was disintegrating in the face of pressure
from Arab and Kurdish nomad tribes. The last Buyid governor left Mosul in 996,
and henceforth this city was under the control of the bedouin Uqalids. Meanwhile,
the Diyar Bakr was absorbed within a Kurdish emirate founded by Bad ibn Dustuk.
This emirate not only encompassed regions south of the Anti Taurus, but also a
variety of urban sites on the northern shores of Lake Van in Armenia. The military
potency of this new Kurdish power was visible during the revolt of Bardas Phokas,
when Bad took advantage of the mayhem inside Byzantium to raid the plain of Mu
in Taron, an Armenian princedom annexed by the Byzantine empire as recently as
966/7. 6 It was only in 992/3, after Bad's death and a series of Byzantine punitive
raids around Lake Van, that Basil was able to negotiate a lasting peace with the
Kurdish emirate.7 In 1000 this deal was further solidified when Basil n offered
Bad's nephew, ibn Marwan, the title of magistros, the office of doux of the east,
and the promise that imperial troops from Taron and western Armenia would assist
o
Making of Orthodox Byzantium, pp.365-7; Kennedy, The Prophet and the Age of the Caliphates,
pp.225-236.
See above, chapter three, passim.
6 Stephen of Taron p. 141; see below, p.291.
Al-Rudhrawari p.262; Stephen of Taron pp.200-3
279
//. Christian Caucasia
8 Yahya, p.260; Stephen of Taron, p.210; Elias of Nisibis, p. 138; Whitlow, Making of Orthodox
Byzantium, pp.381-2
9 For relations between Byzantium and the Caucasian princedoms in Basil's reign see: Forsyth, 'The
Chronicle of Yahya ibn Sa'id', pp.464-478, 557-581; C.Tourmanoff, 'Armenia and Georgia', in
J.Hussey (ed.), The Cambridge Medieval History, Vol. IV, Pt. 1, The Byzantine Empire: Byzantium
and its Neighbours (Cambridge, 1967), pp.615-620; idem., 'The Bagratids of Iberia from the Eighth
to the Eleventh Centuries', Le Musoon 74 (1961), pp.37-42; V.Minorsky, 'New light on the
Shadaddids of Ganja (951-1075)', Studies in Caucasian History (London, 1953), pp. 14-20; A.Ter
Ghevondyan, The Arab Emirates in Bagratid Armenia (Lisbon, 1976), pp. 101-121.
10 Manandian, Trade and Cities ofArmenia, pp. 136-172; see below, p.282, n.19
11 C.Tourmanoff, 'Armenia and Georgia', pp.615-6. In the first half of the tenth century the Artsruni
princes of Vaspurakan constructed several palace complexes of which the most famous was that at
Aght'amar, situated on an island in Lake Van. The palace church is still extant (Thomas Artsruni,
History of the House of the Artsrunik', trans. R.W.Thomson (Detroit, 1985), pp.355-61;
A.Manoukian (ed.), Documents of Armenian Architecture/Document! di Architettura Armena
(Milan, 1974), Vol.8). Later in the tenth century the Bagratids of Sirak sponsored the construction of
a new, walled city at Ani in northern Armenia (Stephen of Taron pp.138, 192, 213; Manoukian,
Documents of Armenian Architecture, Vol. 12). Within Georgian-speaking Caucasia enthusiastic
sponsors of churches and monasteries included David of Tao (d.1000) and Bagrat HI of Abasgia
(d.1014) (W.Z.Djobadze, Early Medieval Georgian Monasteries in Historic Tao, Klarjet'i and
Savseti (Stuttgart, 1992); Evans and Wixom, The Glory of Byzantium, pp.336-349); see below,
pp.280-3 for the careers of David and Bagrat.
Stephen of Taron, pp.138, 192; Georgian Royal Annals, pp.272-7. He was the only Georgian-
speaking member of the Ibero-Armenian Bagratid family to mint his own coins (C.Tourmanoff,
'The Bagratids of Iberia from the Eighth to the Eleventh Centuries', Le Museon 74 (1961), pp.40)
13 See above, p. 114,119-20.
280
Byzantium was concerned, during the revolts of the first thirteen years of Basil's
reign. In 978/9 he supplied Basil II with a large cavalry force which enabled
imperial armies to crush the first Skleros revolt. 13 Among David's rewards was the
stewardship for his lifetime of key imperial territories on the north-eastern reaches
of the frontier, including the city of Theodosioupolis and the plain of Basean. 14 In
the 980s, he used his position of local authority on the frontier to participate in the
Phokas revolt on the side of the rebels. 15 In the aftermath of the failure of this revolt
he was forced to make the emperor Basil n the legatee of his princedom of Tao.
This agreement destroyed a previous arrangement according to which David's
adopted son Bagrat III of Abasgia had been nominated as his heir. 16 Yet, David
remained a powerful force within Caucasian politics until his death in 1000. In the
final decade of the tenth century he not only seized Manzikert near Lake Van from
its Muslim overlords, but also constructed a Christian Caucasian armed alliance
which defeated Mamlan the Rawaddid emir of Azerbaijan in 998/9. 17
When David died in 1000, Basil II marched to the north-east frontier to collect the
inheritance of Tao in person. Having dispersed token resistance to the Byzantine
take-over from the local Iberian nobility, the emperor then garrisoned the key
fortresses of his new territory. During the same expedition he also accepted
obeissance from a variety of neighbouring Caucasian princes, Muslim as well as
Christian, who were rewarded with imperial titles. The following year, one of these
14 Stephen of Taron reports on the territorial rewards offered to David (Stephen of Taron, pp. 141-2).
Some of the regions granted to David, such as Theodosioupolis (referred to as Karin, its Armenian
name, by Stephen) and the plain of Basean were within the north-eastern frontier regions of the
empire (see above, pp.160, 222; see also, Map 6). Others, such as Harkh and Apahunikh, north of
Lake Van, were not under the direct political control of the empire in 979. In the case of these
territories Byzantium seems merely to have acknowledged David's right to launch military attacks
against the Muslim emirates of the Lake Van region. Other areas named within the grant have yet to
be accurately identified. For further analysis see Forsyth, 'The Chronicle of Yahya ibn Sa'id',
pp.464-78, which summarises not only the primary evidence but also the extensive and
contradictory secondary literature concerned with the 979 grant.
15 Yahya, PO 23, pp.424, 429, 460.
16 Georgian Royal Annals, pp.273-4
281
princes, Gurgen of Iberia (K'art'li), unhappy that he had only received the title of
magistros, invaded Byzantine Tao. His attempts were thwarted by a Byzantine
army led by Nikephoros Ouranos, the doux of Antioch. 18 Yet, despite this
Byzantine victory, consolidation of imperial control in Tao was soon eclipsed.
While Basil was busy fighting the Bulgarians in the west, a powerful Georgian state
began to emerge on the north-eastern stretches of the frontier. In part this greater
Georgian political presence was underpinned by dynastic politics: in 1008 Bagrat
III ruler of Abasgia, and erstwhile adopted son of David of Tao, inherited Iberia
(K'art'li) from his natural father Gurgen, thus uniting under a single crown a region
extending from the eastern shore of the Black Sea, to the foothills of the Caucasus
mountains beyond the river Kur. Yet, this regional Georgian hegemony was also
forged through conquest. Among territory gained by the sword was the remote
Georgian princedom of Kakhetia, which lay to the north east of the emirate of
Tiflis. Bagrat also managed to acquire the city of Ardanoutzin, situated a little to
the north of Tao which, as a result of its immense customs revenue, had been
coveted by successive Byzantine emperors throughout the tenth and early eleventh
centuries, including Basil II himself 19 When he died in 1014 Bagrat left his son
282
George a considerable legacy, including a longstanding claim to those territories of
David of Tao which were in Byzantine hands.
It was in the context of disputes over Tao that relations with Christian Caucasia
rose to the top of the Byzantine foreign policy agenda during the final decade of
Basil's reign. With the accession of George in 1014, a disagreement immediately
broke out between Basil and the young Abasgian ruler about the patrimony of
David of Tao. Having warned George to stay out of David's former princedom,
Basil sent an imperial army to crush Iberian resistance in 1014. These Byzantine
forces were decisively defeated. However, once the annexation of Bulgaria in 1018
released Byzantine military energies from the Balkans, preparations for a larger-
scale campaign were set in train, beginning with the refortification of the frontier
city of Theodosioupolis. Three years later Basil himself arrived on the north-eastern
frontier with the imperial field army. Although his first incursion into Iberia in the
autumn of 1021 proved to be inconclusive, another offensive in the spring of 1022
resulted in a crushing victory. In return for peace, George surrendered his son
Bagrat as a hostage to the emperor and handed over several fortresses.20
Meanwhile, Basil's empire-building in Caucasia at the end of his reign did not stop
with the Georgian frontier. It also extended into Armenia. In the winter of 1021/2
John Smbat, prince of the northern Armenian kingdom of Ani, made Basil his heir.
Although the reasons for this decision are not clear, his actions may have been
precipitated by conflict with members of his own Armenian Bagratid family, above
all his brother Ashot IV who was supported by George of Abasgia.21 John Smbat's
283
inheritance was finally secured by Byzantine armies in 1042 nearly two decades
after Basil's death.22
23 Skylitzes, pp.354-5; Aristakes, p. 19; Matthew of Edessa, p.41, 44-6; Bar Hebreus, p. 179; Yahya
(Cheikho), p.240
24 Skylitzes, for example, favours a date of 1016; Aristakes, 1018-19; and Yahya ibn Sa'id, 1021/2
(see note above, for page references). Reconciling these contradictions may, in fact, be impossible.
The accounts of Aristakes and Yahya both refer to the handover of Vaspurakan in undated
digressions contained within their main narratives of Basil's campaign against George of Abasgia.
Meanwhile, Skylitzes' account of the surrender is contained within one of his characteristic
summary chapters, in which several events which happened at different times are telescoped into the
same passage of prose. Thus, the hand-over of Vaspurakan is included in a section of text which
begins with a joint Russo-Byzantine naval expedition to Cherson. It is this campaign which is dated
to 1016. However, it is uncertain whether the same date should also be attributed to the Vaspurakan
episode; see above, p.26, n.15, and below, p.318, n.134.
Aristakes, pp.23-5.
26 See above, pp.228-9 for discussion about the origins of these small themes.
284
When he left Tao, he marched southwards towards his newly acquired Armenian
territory of Vaspurakan. He then headed east to the plain of Her, which was located
between Lake Van and Lake Urmia. Although the emperor was forced to turn
westwards when his army was caught out by early autumn snows, it seems likely
that the original target of this expedition was the emirate of Azerbaijan located to
the east of Lake Urmia.25
285
east respectively. To these would be added during Basil's reign, the katepanate of
Iberia created after the absorption of Tao, and the katepanate of Vaspurakan, which
came into being after the annexation of the eponymous Armenian princedom south
of Lake Van.27
The basis for the belief that a sophisticated and schematised system of military
administration was already in place on the eastern frontier by the time Basil came to
the throne in 976, is the fact that the doukes of Chaldia, Mesopotamia and Antioch
are recorded in the 'Escorial Taktikon', a list of precedence dated by its editor,
Nicholas Oikonomides, to the years 971-75. Yet, the primacy given to the 'Escorial
Taktikon' in many attempts to reconstruct the history of the eastern frontier in the
later decades of the tenth century may be misplaced. At the very simplest of levels,
it should be remembered that lists of precedence by their very nature are not
administrative handbooks, which explain in detail the responsibilities of
contemporary Byzantine bureaucrats. Instead, they are seating plans drawn up by
imperial officials who organised banquets within the Great Palace. As such they
constitute occasional and approximate outline sketches of the hierarchy of the
principal office holders within the Byzantine empire, as that hierarchy was
perceived by palace functionaries in Constantinople. However, they cannot reflect
accurately those ad hoc administrative arrangements which might prevail on the
ground in moments of rapid political change, either at the centre of the empire, or in
AQ
286
frequent political and military upheaval, such as the eastern frontier of the
Byzantine empire in the second half of the tenth century.
More important, however, is the fact that the methods by which such manuals were
composed, make both the manuals themselves, and the information they contain,
extremely difficult to date. For example, such taktika often continue to list positions
which have long fallen into desuetude, while failing to record innovations. As a
result it is impossible to use either the inclusion or omission of administrative
offices as dating criteria. This difficulty is clearly illustrated by the case of the
theme of Mauron Oros. This theme was established, as we saw in the last chapter,
in the Amanos mountains in 968. However, it is not listed in the 'Escorial
Taktikon'.29 If omission was a reliable dating clue, the 'Escorial Taktikon' must
perforce ante-date 968. Yet, since the 'Escorial Taktikon' includes both Antioch,
which fell in 969 to the Byzantines, and an array of small Bulgarian themes such as
Preslav and Dristra conquered by the emperor John Tzimiskes in 971, it was
clearly composed after 968 and thus after the creation of the theme of Mauron
Oros. In this case it is clear that the official responsible for drawing up the list
simply forgot to include the new theme of Mauron Oros. Unfortunately, it is
according to the principle of inclusion and/or omission of administrative offices,
that the dates of the composition of the 'Escorial Taktikon' have been established.
For instance, the terminus post quern of 971 has been justified by the fact that the
taktikon includes those small Balkan themes which were conquered by Tzimiskes
in 971; the terminus ante quern of 975 by the omission of those minor themes in
northern Syria, such as Laodikeia and Balanias, which were absorbed into the
empire during Tzimskes1 final military campaign in the east in the summer of 975.30
Yet, as the case of Mauron Oros indicates, the omission of the Syrian themes may
287
simply reflect a failure on the part of a Constantinopolitan official to include these
offices in the list, rather than the fact that the 'Escorial Taktikon' was produced
before 975.
Once the context in which the 'Escorial Taktikon' was produced is understood, it
becomes apparent that this document cannot categorically be said to pre-date the
period 975, nor, by extension does the information it contains necessarily reflect the
administrative superstructure of the Byzantine state before the death of John
Tzimiskes in 976. As far as the eastern frontier is concerned, the 'Escorial
Taktikon' cannot be used as incontrovertible evidence that a rigid system of frontier
governance based around the three doukes of Chaldia, Mesopotamia and Antioch,
was complete and universally applied on the Byzantine eastern periphery before
Basil came to the throne. Instead, the most that can be said is that this taktikon
offers an approximate, and idealised outline of Byzantine governance in the
decades which followed the fall of Antioch in 969, and the conquest of eastern
Bulgaria by John Tzimiskes in 971.
For a more detailed and nuanced understanding of the development of the military
organisation of the eastern frontier in the later tenth- and early eleventh- centuries,
it is necessary to set the stark blueprint offered by the 'Escorial Taktikon' against
contemporary literary and sigillographical evidence. In the discussion which
follows the frontier katepanates will be examined one by one in the light of this
broader spectrum of evidence. Although interesting parallels could be drawn
between the development of the katepanates in the east with those in the western
half of the empire, such as Thessalonika, Adrianople, Italy, and Bulgaria, such a
288
comparison lies outside the scope of this study and will not be attempted here. 31
Neither will attention be paid to the frontier arrangements which prevailed once
Edessa was annexed by the empire shortly after Basil's death during the reign of
Romanos III (1028-34). 32 Nonetheless, from the evidence under analysis in this
discussion it will become clear that far from resembling the rigid picture presented
by the 'Escorial Tatikon', the military organisation of the frontier was a piecemeal,
ad hoc process, constantly reshaped by changing political, diplomatic, and military
circumstances.
Although problems with dating the 'Escorial Taktikon' mean that it is difficult to
use this document to describe the exact geographical trajectory of the eastern
frontier when Basil II came to the throne in 976, nonetheless by identifying some of
those small themes named in the taktikon, it is possible to gain some idea of the
contours of the Byzantine borders with the east in the final quarter of the tenth
century. Thus, before the annexation of the Ibero-Armenian princedom of Tao in
1000, the most northern section of the eastern frontier began at Soteropolis, an
isolated outpost on the Black Sea coast, located on the left bank of the lower Coruh
(Akampsis) river. Having crossed the Pontus mountains east of Trebizond, the
border then traversed the Basean plain, centre of the former emirate of
Theodosioupolis. It passed the fortification of Hafdjidj (Chauzion) on the north-
west flank of the Bingol Dagi, and headed south towards the erstwhile Armenian
princedom of Taron.33 It is this sector of the frontier that is usually seen as the
responsibility of the doux ofChaldia.34
31 For the western duchies see Kuhn, Die byzantinische Armee, pp.207-240; von Falkenhausen,
Untersuchungen uber die byzantinische Herrschaft in Suditalien, pp.45-50
32 Felix, Byzanz und die islamische Welt, pp.90-1, 100
33 See Map 6, and above, pp. 160-2; the outlines of the frontier at the beginning of Basil's reign are
also sketched out by Forsyth, "The Chronicle of Yahya ibn Sa'id', p.371.
34 Oikonomides, Listes, p.263, 354
289
A lack of primary written materials makes it difficult to know when a commander
officially designated as the doux of Chaldia first appeared. Sigillographical and
literary evidence suggests that such an office certainly did not exist in the first half
of the tenth century, when the region known as Chaldia was still a theme governed
like the rest of Anatolia by a strategos 35 Instead, it is only with the more ambitious
eastern military enterprises which characterised the middle of the tenth century, that
commanders began to appear in Chaldia with broader authority than that of a mere
strategos. These signs emerge for the first time during the conquest of the emirate
of Theodosioupolis in 949, when the victorious Byzantine military commander,
Theophilos Kourkouas, the brother of the general John Kourkouas, and the
grandfather of the emperor John Tzimiskes, is described as monostrategos of
Chaldia, a position which appears to imply greater seniority than that of a mere
thematic strategos 36
35 Many of the seals of these tenth-century strategoi are listed in Bryer and Winfield, Monuments of
the Pontus, p.316; see also Schlumberger, Sigillographie, p.290, nos.2-3; Dumbarton Oaks
Unpublished 55.1.1412-3 and 58.106.2115. One strategos of Chaldia who appears in the literary
sources was Bardas Boilas. He rebelled against Romanes Lekapenos in the third decade of the tenth
century (Theophanes Con., p.404).
36 Theophilos is described as monostrategos of Chaldia by the tenth-century historian Theophanes
Continuatus. The De Administrando Imperio, compiled c.952, names him as strategos of
Theodosioupolis. However, this appellation may indicate that he was commander in Chaldia with
his main base at Theodosioupolis (Theophanes Con., p.428; DAJ, p.212).
37 Skylitzes, p.284; Leo the Deacon, p.96; Oikonomides, Listes, p.354; Kiihn, Die byzantinische
Armee, pp. 184-5. Skylitzes also mentions that Bardas Phokas was a doux on two other occasions
The first reference is to be found in his coverage of Bardas Phokas' revolt against the emperor John
290
in a frontier context in the historical record. Yet, it would be premature to assume
that this appointment necessarily represents a radical and permanent reorganisation
of the administration of the whole of the eastern frontier. Instead, it can be
persuasively argued that the appointment of Bardas Phokas was driven by short-
term military pressures both within and outside the empire. Thus, on the one hand
the greater focus of regional military authority in the hands of Phokas may reflect
the demands of internal security. Bardas' uncle Nikephoros needed a strong and
loyal hand in the furthest reaches of north-eastern Anatolia, a region where the
authority of rival families such as the Lekapenoi and Kourkouai had traditionally
been strong.38 Equally Bardas' appointment may also reflect a growth in the
intensity of Byzantine diplomatic and military relations with neighbouring regions
in western Caucasia during the reign of Nikephoros. In 966-7 Bagrat and Gregory,
the princes of the neighbouring Armenian princedom of Taron, handed over their
territories to Nikephoros Phokas in return for the title ofpatrikios and estates within
the Byzantine empire.39 In 968/9 Byzantine raiding armies passed through Taron on
their way to raid the Arab emirates of Lake Van.40
When Nikephoros Phokas was murdered by John Tzimiskes in 969, his nephew
Bardas was removed from his position as doux. A lack of written evidence means
that it is impossible to know exactly what happened next in Chaldia. However,
using the evidence provided by tenth- and eleventh-century lead seals, it is clear
that administrative power had yet to become settled when Bardas was dismissed.
Instead, frontier authority in Chaldia was consistently characterised throughout the
Tzimiskes in 971. In a passage of direct speech Bardas recalls his former position as a doux,
although without specifiying the geographical location of his responsibilities (Skylitzes, pp.293-4).
The second reference occurs duing Skylitzes' account of the first Skleros revolt, when he notes that
David archon of Tao and Bardas Phokas had become friends while the latter was doux in Chaldia
(Skylitzes, p.326).
38 Cheynet, Pouvoir et Contestations, pp.216, 321-4
39 Skylitzes, p.279
291
tenth and eleventh centuries by a high degree of flexibility. One sign of the ad hoc
nature of this authority is the frequency with which commanders could exercise
responsibility for more than one region. This principle of overlap had, of course,
been visible in the career of Bardas Phokas himself, in his control of both Chaldia
and Koloneia. A similar phenomenon can be detected on the eleventh-century seal
of a certain Nicholas patrikios, who was katepan of both Chaldia and Mesopotamia.
This seal indicates that on occasion the supreme regional military commander of
the northern-most section of the frontier, could also exercise authority over
Mesopotamia, the border region lying further south among the plains and
mountains of the Anti Taurus, and the Upper Euphrates and Upper Tigris rivers. 41
Moreover, the seals of several non-military officials in the later tenth and eleventh
centuries indicate the extent to which administrative authority based principally in
Chaldia could be linked with neighbouring districts. For example, several
protonotarioi and kritai exercised authority over Chaldia and Derxene, a region
lying to the west of Theodosiopolis.42 Other kritai exercised jurisdiction over the
theme of Koloneia as well as Chaldia, thus mirroring the geographical scope of
Bardas Phokas' authority as doux. 43 Meanwhile, in a case not directly involving
Chaldia, but still pertinent to the administrative history of the northern-most sector
of the frontier, a certain Leo, protospatharios, was both strategos of Derxene and
Taron in the last years of the reign of John Tzimiskes.44
Evidence for the development of the office of doux in Chaldia during the reign of
Basil II is extremely meagre. It is impossible to know which of those officials
40 Yahya, PO 18, p.825; Canard, 'La date des expeditions mesopotamiennes de Jean Tzimisces',
p. 100.
Dumbarton Oaks Unpublished 58.106.400; see below, pp.296-7, for the possibility that Skleros
may have held joint command over Chaldia and Mesopotamia.
42 Protonotarios: Dumbarton Oaks Unpublished F3159; kritai. Dumbarton Oaks Unpublished
58.106.5114; 55.1.2933; 55.1.252; F400
43 Krites of Chaldia and Koloneia: Dumbarton Oaks Unpublished 58.106.806; krites of Chaldia,
Derxene, and Koloneia: Dumbarton Oaks Unpublished 58.106.806.
292
whose seals are discussed above held office during Basil's reign itself. Meanwhile,
the only literary evidence for the history of the Chaldia region during the reign
comes from the Miracles of Saint Eugenios. This hagiographical source notes that
during the revolt of Nikephoros Xiphias and Nikephoros Phokas in 1022, a certain
Basil was doux of Chaldia and Trebizond.45 While this example demonstates that
Chaldia remained a duchy/katepanate during the first quarter of the eleventh
century, it is possible that shortly after this date Chaldia reverted to the status of a
theme. There are several seals of the strategoi of Chaldia which appear datable to
the eleventh century.46 One reason why Chaldia might have reverted to thematic
status in the eleventh century would have been the gradual expansion of the new
neighbouring katepanate of Iberia.47 An alternative reason for the incidence of
eleventh-century strategoi could be that the original theme of Chaldia continued to
exist alongside the regional duchy of the same name throughout the later tenth and
eleventh centuries. With command over a much smaller area than the entire duchy,
the strategos of Chaldia would have been subordinate to the regional doux or
katepan**
293
and the Anzitene at its heart, the central sector of the frontier extended as far south
as Germanikeia (Maras) on the Ceyhan (Pyramos) river, and Samosata on the
Euphrates.49 Yet, despite its size, there is little evidence that this frontier region was
governed by a supra-thematic military commander such as a doux or a katepan
before the reign of Basil II began. Instead, during the reign of Basil's predecessor
John Tzimiskes, the only reference to a military commander of Mesopotamia
concerns the strategos of Mesopotamia. Thus, when the Fatimids attacked northern
Syria during the winter of 970-1, an anonymous strategos of Mesopotamia was sent
to deal with the incursion.50 This strategos was almost certainly the military
governor of the theme of Mesopotamia which was established in the Anti Taurus
region north of Melitene during the first decade of the tenth century. The theme
itself covered a considerably smaller geographical area than the central sector of the
frontier as a whole. Moreover, sigillographical evidence also suggests that
strategoi of the theme of Mesopotamia continued to be appointed throughout the
second half of the tenth century. 52
The persistence of the office of strategos of Mesopotamia does not mean in itself,
of course, that a doux/katepan could not have existed at the same time. As we saw
above in the case of Chaldia, it is possible that a doux and strategos could exercise
office simultaneously, with the latter being subordinate to the former.53 Yet in the
case of Mesopotamia there is no evidence that a supra-thematic senior regional
commander, such as a doux or katepan, was appointed to the central sector of the
frontier during the reigns of either Phokas or Tzimiskes. Instead, it is only at the
beginning of the reign of Basil II that the first signs of an overarching military
48
This is a hypothesis first volunteered by Ahrweiler, 'Recherches', p.48
49
See Map 6, and above, pp.222-3 for the campaigns of Kourkouas
50
Skylitzes, p.287; Farag, 'The Aleppo question', p.47; P.E.Walker, 'A Byzantine Victory over the
Fatimids at Alexandretta (971)', B 42 (1972), pp.431-440
51 DAl p.239
52 Schlumberger, Sigillographie, p.288; Dumbarton Oaks Unpublished: F666, F1218; 55.1.2824
294
authority in this region start to emerge. On this occasion the evidence comes from
the 'Synopsis Historion' of John Skylitzes, and his report that Bardas Skleros was
appointed doux of Mesopotamia shortly after Basil came to the throne. Yet, any
understanding of the nature of Skleros' appointment is, unfortunately, prejudiced by
*
Skylitzes1 working methods, and his use and abuse of underlying texts. In this case
Skylitzes' use of a pro-Skleros source in his account of the reign of Basil II means
that it is Skleros' justification of his revolt, rather than the wider strategic context
on the eastern frontier, which underpins the analysis of this appointment in the
'Synopsis Historion'.54 Thus, according to Skylitzes/Skleros' account, Basil
Lekapenos, the parakoimomenos, was so distrustful of Skleros' political ambitions,
that he released the general from his position of stratelates (supreme commander of
the field army), and installed him instead as "doux of the tagmata in Mesopotamia".
In Skleros' place he appointed Peter the eunuch as stratopedarches of all the
tagmata of the east. Meanwhile, Skleros was so vexed by this series of events,
that he was not able to conceal to himself his grief magnanimously, but protested
out loud." 55
295
in 976, or that it had come to represent a regional backwater.56 Instead, this was a
frontier which had seen recent and bloody warfare. In 972 the emperor John
Tzimiskes had led an expedition from the Anzitene in Byzantine Mesopotamia into
the Djazira to raid the cities of Edessa and Nisibis. So convinced were the
inhabitants of Baghdad that Tzimiskes' real target was Iraq that they rioted in the
streets. In response to this Byzantine atack the Hamdanids of Mosul launched a
counter strike through the Bitlis pass, laying waste to the Byzantine territory of
Taron in western Armenia, which had only recently been annexed in 966/7. 58
Another Byzantine offensive followed swiftly, as the domestikos of the scholai
Melias set out from the Anzitene in the summer of 973 and arrived at the gates of
Amida in the Djazira. It was at this battle that Melias was taken captive by the
Hamdanids. 59
With such a recent legacy of bloody military encounters, it is clear that this frontier
in 976 was a highly sensitive region to which only the most competent and trusted
commander, with highly effective and professional troops, would have been
dispatched. Indeed, Skleros' appointment makes more sense when it is interpreted
in this light rather than in terms of a demotion. It is noteworthy that the only other
historian to report on Skleros' arrival on this stretch of the eastern frontier, Yahya
ibn Sa'id, does not consider his appointment a demotion. Instead, he merely notes
that Bardas Skleros was appointed the 'governor' of Bathn Hanzit and Khalidiyat.60
56 Most modem historians have followed Skylitzes in interpreting Skleros' appointment as doux in
976 as a demotion to the margins of the military and political hierarchy: Seibt, Die Skleroi, p.36;
Forsyth, 'The Chronicle of Yahya ibn Sa'id', p.375.
57 See above, p. 117
58 See Map 2
59 See above, p.278; Ibn Miskawayh, pp.326-8; Yahya, PO 23, pp.353-8; Canard, 'La date des
expeditions mesopotamiennes de Jean Tzimisces', pp.99-108.
60 Yahya, PO 23, p.372. The term for 'governor' used by Yahya is the Arabic, 'wall'. Controversy
still persists about the exact geographical regions that Yahya believed to be under Skleros'
command. It is generally agreed that Bathn Hanzit denotes Mesopotamia. This is the Arab name for
the region known in Greek as the Anzitene, the plain in the Anti Taurus at the centre of Byzantine
Mesopotamia. Identification of al Khalidiyat, however, has proved more difficult. Seibt (Die Skleroi,
296
Moreover, Skleros was not simply given command over the light cavalry and
infantry forces of the small, peripheral Armenian themes (Armeniaka themata).
Instead, as even Skylitzes notes, he was placed in charge of tagmata, or field army
troops. The most likely context for his appointment as doux of Mesopotamia, was
that he was expected to continue the struggle against the Hamdanids of Mosul,
which had been set in motion by the campaigns of Melias and John Tzimiskes.
However, instead of fighting the Muslim enemy, Skleros took advantage of his
control over the field army troops and rebelled against the emperor. Forced to
justify his decision to turn the Byzantine field army under his command on
Constantinople rather than on the Djazira, he manufactured the excuse that he was
disappointed with his treatment by the imperial authorities. It is then this excuse
which then enters the historical record via Skylitzes.
p.36) has argued that it should be identified with Kaloudia, one of the small frontier themes on the
west bank of the Euphrates downstream from Melitene. Intuitively, however, it seems most likely
that al-Khalidiyat denotes Chaldia. As noted above, there is sigillographical evidence to support the
idea that command over both the northern and central sections of the frontier could be vested in a
single individual. See above, p.292.
61 This viewpoint contrasts with Kiihn, Die byzantinische Armee, pp. 182-3, who believes that
Nikephoros Phokas was responsible for instituting the katepanate of Mesopotamia. However,
Kuhn's argument is based on supposition. Having attributed the creation of the katepanates of
Chaldia and Antioch to Nikephoros Phokas, he argues that Mesopotamia must have been given this
status at the same time since any other decision would have left a hole in the organisation of the
frontier. However, Kuhn's argument demands that the katepanates of Chaldia and Antioch were
297
centralised army in pursuit of military victory over neighbouring adversaries. At
this point the office of doux was not a sedentary administrative position with
authority over a strictly defined geographical area.
298
as katepan of Mesopotamia, Taron, and Vaspurakan.63 The fundamental principle
that one commander could hold an array of regional commands seems to have
remained constant between the mid-tenth and mid-eleventh centuries.
63 See above, pp.290-2, for Bardas Phokas and Nicholas. For an analysis of the positions held by
Pahlawuni on the eastern frontier see K.N.Yuzbashian, 'L'administration byzantine en Armenie aux
Xe et Xle siecles', REArm 10 (1973-4), p. 147. Yuzbashian expressed doubts about whether a single
individual could have held all three commands at once. However, the evidence produced in this
chapter repeatedly demonstrates that several eastern frontier commands, whether military or non-
military, could be held simultaneously across several regions by the same individual.
299
It is frequently alleged that Antioch was put under the command of a doux or
katepan by the emperor Nikephoros Phokas when it was conquered by Byzantine
armies in October 969.65 Yet, evidence from within the literary and sigillographical
records does not support such an early dating. As long ago as 1962, Laurent pointed
out that neither of the commanders who led the victorious armies in 969, namely
Peter the stratopedarches and Michael Bourtzes, became doux of the city when it
fell.66 Instead, Peter left Antioch at the head of the imperial field army and
preceded to besiege Aleppo.67 Nor did he return to Antioch after the siege of
Aleppo. Instead, he next appears in the historical record as leader of the tagmata of
Thrace and Macedonia during John Tzimiskes' campaign against the Rus at Dristra
fO
64 Yahya, PO 18, p.708; Forsyth, 'The Chronicle of Yahya ibn Sa'id al Antaki', p. 1
65 N.Oikonomides, Listes, p.354; Kiihn, Die byzantinische Armee, pp. 170-1.
66 V.Laurent, 'La chronologic des gouverneurs d'Antioche sous la seconde domination byzantine',
Melangesde I'Universite Saint-Joseph de Beyrouth 38 (Beirut, 1962), p.227
67 Skylitzes, pp.271-2; Yahya, PO 18, pp.816-7, 823-4
68 Skylitzes, p.300; Laurent, 'Gouverneurs d'Antioche', p.227
69 The reason why Bourtzes incurred imperial displeasure is uncertain. According to Skylitzes
Bourtzes exceeded the orders he had been given during the course of the siege. Originally appointed
strategos of Mauron Oros, a small theme based in the Amanos mountains, he was commanded
merely to raid the countryside surrounding Antioch (see above, p.229). However, in search of
personal glory, Bourtzes took the unilateral decision to occupy one of the upper towers of the circuit
walls of Antioch. When his position became desperate, he was forced to send for emergency help
from the imperial field army under the leadership of Peter the stratopedarches. When the field army
arrived, Antioch itself fell. However, as a result of his earlier disobedience Bourtzes was dismissed
from his position as strategos of Mauron Oros (Skylitzes, pp.271-3). Yahya ibn Said conveys a
slightly different story. Although Bourtzes was rewarded for his role in the fall of Antioch, his
recompense was much smaller than expected. This was because Nikephoros Phokas was angry with
him for allowing Antioch to be burned (Yahya, PO 18, p.825). Leo the Deacon confirms that
Bourtzes led an advance party into Antioch and set the city on fire (Leo the Deacon, pp. 81-2).
70 Leo the Deacon, p.85; Yahya, PO 18, p.829; Skylitzes, p.279
300
doux. Instead he was strategos of Antioch and Lykandos. 71 As we saw in chapter
four, Lykandos was a theme established in the early tenth century in the interim
zone between the eastern plateau and the western reaches of the Anti Taurus
mountains.72
301
and well-established eastern Anatolian theme, which determined his appointment at
Antioch.76 As such Maleinos' position at Antioch demonstrates the extent to which
military experience, high-standing within the army, and political loyalty to the
emperor, underpinned the military organisation of the frontier, rather than a
mechanistic adherence to an abstract administrative blueprint such as the 'Escorial
Taktikon'. Furthermore, Maleinos' command over both Lykandos and Antioch
once again illustrates the frequency with which command over several regions in
the Byzantine east could be vested in a single individual.
Powerful analogies exist for Maleinos' position in the immediate aftermath of the
conquest of Antioch. Sigillographical evidence suggests that the first commander of
Anazarbos and Mamistra (Mopsuestia), the Cilician cities conquered by Nikephoros
Phokas in 964, was a certain George Melias protospatharios, the strategos of
Tzamandos.77 Just as at Antioch, therefore, the new conquests of Mamistra and
Anazarbos were put under the command of an officer who already exercised
military command on the eastern-most reaches of the plateau, in this case
Tzamandos. Moreover, George Melias also exercised considerable authority within
the Byzantine army. Like Maleinos, he could call upon a considerable family
pedigree of service in the eastern plateau region. It was his eponymous ancestor,
Melias, who had been responsible for establishing Byzantine control over Lykandos
*7R
and Tzamandos early in the tenth century. Sigillographical evidence indicates that
302
the family continued to exercise military command in the eastern plateau
throughout the tenth and eleventh centuries.79 Moreover, if George, the owner of
the seal, can be identified with the Melias who rose to the position ofdomestikos of
the scholai of the east by 973, then his authority, just like that of Maleinos, was
O/\
built on his own military skills as well as the reputation of his ancestors.
However, while the appointment of officials such as Melias and Maleinos was
driven by short-term domestic and external political needs, such imperatives could
also make their tenure of command on the frontier very brief. Thus, within a year of
the death of his political ally Nikephoros Phokas in December 969, Eustathios
Maleinos found himself transferred from Antioch to the position of strategos at
O -I
Tarsos in Cilicia. His place at Antioch appears to have been taken by Nicholas the
eunuch. He was sent east at the head of an imperial field army by the new emperor
John Tzimiskes during the winter of 970-1 to deal with an attack from Fatimid
79 Apart from George Melias, there also exists a seal of Theodore Melias, kandidatos, who was
taxiarches, infantry commander, of Lykandos (Konstantopoulos, Molybdoboulla, no.224a; Dedeyan,
'Mleh le grand', pp. 101-2) The Zacos collection contains an eleventh-century seal of a strategos
called Melias (Zacos, Byzantine Lead Seals II, no. 572). In 1991 a seal dated to the first half of the
eleventh century appeared at auction which belonged to a Basil Melias, strategos (Oikonomides,
SBS 3, p. 189). In neither of the two latter seals is the geographical region under the command of
these strategoi indicated.
80 See above, p. 278. It is possible that at an earlier point in his career Melias may have been the
strategos of the small theme of Chortzine which lay north-west of the plain of Mus in Taron (Zacos,
Byzantine Lead Seals II, no.227; Oikonomides, Listes, p.359). Melias may have come to Tzimiskes'
attention in the context of the Cilician campaigns of Nikephoros Phokas. As domestikos of the
scholai in the mid-960s, Tzimiskes was a leading protagonist of warfare in Cilicia (Skylitzes,
pp.267-8; Yahya, PO 18, p.793). Certainly fresco decoration of the Great Pigeon House church in
Cappadocia indicates that the two men had already developed close connections during the reign of
Nikephoros Phokas. This church was decorated by local adherents of the Phokas family during the
reign of Nikephoros, and depicts Nikephoros Phokas and the empress Theophano in the north apse.
On the north wall of the church, the forty martyrs of Sebasteia are represented as line of infantry
soldiers. At the head of line ride John Tzimiskes and Melias on horseback (Rodley, 'The Pigeon
House Church at Qavusin', pp. 301-339; Thierry, Haut Moyen-age en Cappadoce, i, 56).
81 Saunders, 'The Reliquary of Aachen', pp.215-6. As governor of Tarsos during the reign of
Tzimiskes Maleinos was responsible for arresting those inhabitants of Antioch who had murdered
Christopher the Melkite Patriarch of Antioch in 966 (Canard, 'Une vie du patriarch melkite
d'Antioche', p.565). Maleinos still occupied this post at the beginning of Basil's reign (Yahya, PO
23, p.373).
303
Egypt.82 However, it is unlikely that Nicholas was in Antioch for long, since as
soon as the Fatimid threat was contained, the imperial field army was recalled to
deal with the Rus in Bulgaria. Thereafter, although the picture of the organisation
of military administration in Antioch is exceptionally unclear, no commander
appears to have exercised control in the city for longer than a few months. During
the reign of Tzimiskes, the only mention of a senior military official at Antioch
after the departure of Nicholas the eunuch, concerns Michael Bourtzes. According
to Yahya, when a severe earthquake caused considerable damage to the circuit
walls of Antioch, Bourtzes was dispatched to the city by Tzimiskes in order to
oo
oversee the necessary repairs. Unfortunately his position of command within the
Byzantine hierarchy at this time is not specified. Yet, he does not seem to have
been given a permanent office or called doux. Instead, when the refortification was
completed, Bourtzes returned to active service within the mobile Byzantine field
army. According to Skylitzes, when John Tzimiskes died Bourtzes commanded the
tagma of the stratelatai within the army led by Bardas Skleros. 84
Indeed, the very first mention of the office of doux in an Antiochene context only
occurs after the death of John Tzimiskes in 976 at the very start of Basil's reign.
Once again this first citation occurs in the testimony of Skylitzes, who reports that
at the same time as Skleros was named doux of Mesopotamia, Bourtzes was
appointed "doux of Antioch-on-the-Orontes".
QC
82 Yahya, PO 23, p.350; Skylitzes, p.287; Saunders, 'The Aachen Reliquary', pp.211-2; Walker, 'A
Byzantine Victory over the Fatimids', B 42 (1972), pp.431-440; Laurent, 'Gouverneurs d'Antioche',
pp.227-8.
W Yahya, PO 23, p.351
84 Skylitzes, p.315; see above, p.221, for the tagma of the stratelatai.
85 Skylitzes, p.315
304
Bourtzes was sent to Antioch so that he would not conspire with Skleros against the
emperors in Constantinople. Reading Skylitzes' interpretation of events, it is
tempting to see Bourtzes' appointment as part of a broader imperial decision in 976
to marginalise the leading figures of the Byzantine field army. Yet, such a view
would be misconceived. Indeed, just as it has been demonstrated that military and
diplomatic relations with neighbouring Muslim powers to the east, rather than
imperial fears of revolt, underpinned Skleros' deployment on the central sector of
eastern frontier, a similar model can be outlined for Bourtzes' appointment. That is
to say, just as Skleros was appointed doux so that he could lead Byzantine field
army detachments on campaigns against the Hamdanids of Mosul, so Bourtzes was
given command at Antioch so that he could keep up the offensive against the
principal Muslim adversary on his sector of the frontier, namely the Fatimids.
into central Syria, and his appearance outside the walls of Damascas. Yet, in
305
many ways the more strategically significant component of the 975 campaign was
its conclusion. Having devastated the interior of Syria, Tzimiskes returned
northwards along the literal of the Levant, raiding or rendering tributary port towns
such as Sidon, Beirut, Djubayl, Tripoli, Djabala and Balanias. The ambition of this
campaign must surely have been to deprive the powerful Fatimid fleet of as many
strategic points on the coast as possible.88 When he was appointed doux of Antioch
in the spring of 976, Bourtzes was expected to continue this strategy.
When the Skleros revolt broke out in the spring of 976, Bourtzes was instructed to
leave Antioch, join forces with Eustathios Maleinos the strategos of Tarsos, and
QQ
prevent rebel armies from progressing westwards across the Anti Taurus. From
this point onwards the military administrative history of Antioch becomes very
obscure. For a short while after Bourtzes left the city, command was held by his
88 The outlines of this coastal campaign are reported in Yahya, PO 23, p.369. The most detailed
account, however, comes from Tzimiskes' own letter to Ashot in Armenian king of Ani, which is
transmitted in the twelfth-century Armenian history of Matthew of Edessa (Matthew of Edessa,
pp.31-2). The campaign itself is analysed in detail by P.E.Walker, "The "Crusade" of John
Tzimisces in the Light of New Arabic Evidence', B 47 (1977), pp.301-327.
89 Yahya,PO23,p.372
306
son. However, as soon as Bourtzes himself decided to defect from the emperor and
join the Skleros party, he instructed his son to leave Antioch and entrust the city
into the hands of the basilikos Koulei'b. w During the remaining years of the Skleros
revolt, the most significant figures in the governance of the city of Antioch were
first Koulei'b, and later Oubeidallah, two Christian Arab administrators, both of
whom held the office of basilikos. The significance of this office and the
responsibilities undertaken by Koulei'b and Oubeidallah will be discussed in the
final section of this chapter dealing with civilian administration on the eastern
frontier.91 However, for now it is worth noting that it is only in September 985, six
years after the end of the first Skleros revolt, that the historical record once again
makes mention of the office ofdoux.
In 985 the holder of the office ofdoux was the general, Leo Melissenos. 92 Yet, there
is little about his appointment which suggests that the office he held had become a
permanent gubernatorial position. Instead, nearly ten years into Basil's reign, the
office was still an ad hoc position shaped by short-term political and military
exigencies. According to Yahya ibn Sa'id, Melissenos's original appointment was
made in the context of a particular, and clearly-defined, military goal: the siege of
the coastal town of Balanias which had recently been occupied by the Fatimids.
However, a conjunction of more pressing political and military conditions both
within and outside the empire ensured that Melissenos' tenure of the position of
doux was brief. For after he had captured and fortified Balanias, he was swiftly
recalled to Constantinople. In part Melissenos' recall was motivated by the distrust
he had inspired in the emperor while he had been on active service in the east. He
had briefly called a halt to the siege of Balanias when he thought that Basil was
90 Yahya,/>O23,p.373
91 See below, pp.330-3
92 Yahya,/>023,pp.416-7
307
about to be deposed by Basil Lekapenos the Parakoimomenos. He had only
resumed military operations after the emperor commanded him to continue with the
siege or face having to pay the soldiers' wages out of his own pocket. However,
another reason for the recall of Melissenos was that Basil II needed this general's
expertise during his forthcoming campaign against Bulgaria, planned for the
summer of 986.93
The flexible and ad hoc nature of command on the eastern frontier is further
demonstrated by the history of the Antiochene region in the year which separated
Leo's departure from Antioch and the outbreak of the revolt led by Bardas Phokas
in August 987. In some senses Leo was replaced at Antioch by Bardas Phokas.
According to Yahya, at the same time as Leo was recalled to Constantinople in 986,
Bardas was relieved of his position of domestikos of the scholai of the east,
supreme commander of the field army. Instead he was appointed "doux of the east
and governor of Antioch and the provinces of the east". 94 Yet Yahya's description
of Phokas' appointment suggests that unlike Leo Melissenos' limited
responsibilities as doux, Bardas' new duties were extremely broad, and not purely
confined to military matters or to command over the immediate environs of
Antioch. So vast, indeed, is the geographical scope implied in this title, that
Phokas' role may have been that of an imperial plenipotentiary on the frontier.
93 Ibid., p.417; Skylitzes, p.330. Laurent, 'Gouverneurs d'Antioche', p.232. It is possible that during
the Bulgarian expedition Melissenos fulfilled the position of domestikos of the scholai of the west.
This argument has been advanced by Jordanov on the basis of two seals belonging to Leo
Melissenos domestikos, which have been found during excavations at Preslav in eastern Bulgaria.
On one seal Leo is attributed with the title magistros, on the other patrikios (Jordanov,
'Molybdobulles de domestiques des scholes du dernier quart du Xe siecle', pp.208-9). However, it
should be noted that Skylitzes maintains that Stephen Kontostephanos was the domestikos of the
west during the 986 Bulgarian campaign. Skylitzes relates that Melissenos was given responsibility
for guarding the city of Philipoupolis during the expedition (Skylitzes, p.331). As yet there is
insufficient evidence to resolve who led the field army of the west and the exact nature of
Melissenos' command.
308
Such an interpretation would be compatible with the broader military position of
the empire at the time. With a summer campaign against Bulgaria planned, a large
part of the professional field army stationed on the eastern frontier was required to
return to Constantinople to fight in the west. Leo Melissenos was among their
commanders as we have seen. In these circumstances there was little sense in
leaving Bardas Phokas in the position ofdomestikos of the scholai when there were
few scholai to lead. Instead, Phokas was given wide-ranging powers to supervise
the eastern frontier while Basil himself was on campaign in Bulgaria. Yet, even
Phokas' appointment proved to be an ad hoc provision which had to be hastily
rearranged when new external and internal political and military pressures arose. 95
These pressures were the ignominious failure of Basil's 986 invasion of Bulgaria,
and the appearance of a new threat in the east, namely the arrival in February 987
of Bardas Skleros, recently released by Buyid authorities in Baghdad. With the
need for a greater armed presence in the east, Phokas was restored to the position of
domestikos at the head of the mobile field army. However, once the imperial field
army returned eastwards to deal with Skleros, Phokas himself rebelled. %
94 Yahya, PO 23, p.417. Laurent sees Bardas' position as simply doux of Antioch (Laurent,
'Gouverneurs d'Antioche', p.233).
95 This is not to say that Phokas was happy with this division of military command. One of the
reasons for Phokas' revolt in 987 was that he had been left behind from the Bulgarian offensive
(Skylitzes, p.332).
96 Skylitzes, pp.330-34; Leo the Deacon, pp. 171-5; Yahya, PO 23, pp.418-421; Stephen of Taron,
pp. 186-90
Yahya, PO 23, p.425; Laurent, 'Gouverneurs d'Antioche', p.233
309
passages concerned with Leo's role in the rebellion. It is, therefore, difficult to
interpret the nature of his command from the historical written record. In the first
Yahya shows Leo tricking Agapios, patriarch of Antioch, whom the Phokades
suspected of treachery, into leaving the city. 98 The second episode occurs in
Yahya's account of the very end of the Phokas rebellion after Bardas' death at
Abydos in April 989. According to Yahya, Leo continued the Phokas insurrection
in Antioch until November 989. Then he surrendered to the citizens of Antioch and
was handed over to Basil's new lieutenant in the city, Michael Bourtzes. At the
time of Leo's surrender he was supported by a small armed force, which included
Armenians and Muslims. Together they had mounted last gasp resistance from a
OQ
tower in the city walls. Nonetheless, such a meagre shard of evidence from the
final, embattled stages of the revolt, does not enable us to reconstruct the nature of
Leo's power earlier in the rebellion.
Equally confusing is the nature of military command in Antioch between the end of
the Phokas revolt in 989 and the early 990s. It is clear that by 992 Michael Bourtzes
had taken control. When Byzantine armies were defeated by Fatimid forces in 992,
I /W\
and later in 994, it was Michael who was at their head. Yet, the situation during
the three years before 992 is less clear. Usually it is argued that Bourtzes was
appointed doux of Antioch when he arrived in the city in 989 to arrest Leo
Phokas. 101 However, a reference in Stephen of Taron's account to an engagement in
991 in the Antioch region between a Fatimid army and a small force led by
Romanos Skleros, the son of the former rebel Bardas, persuaded Werner Seibt to
argue that Skleros was doux in the city between 989 and 991. Seibt believed that
310
Bourtzes only took over as doux in 992. 102 However, modern historians may be
creating a false problem here. In the context of the intensified warfare between
Fatimid and Byzantine armies which characterised northern Syria in the last decade
of the tenth century, it is possible that responsibilities for the defence of the
southern sector of the frontier were shared among a number of senior military
officers, all with experience of warfare in the Byzantine east. Indeed, the narrative
of Yahya illustrates this principle of multiple command in action. When the
Byzantines were defeated in 994 by the Fatimids in the Ruj valley north of
Apameia, their army was led by Bourtzes, but also contained fresh troops recently
sent from Constantinople under the command of Leo Melissenos, himself a former
doux of Antioch. 103 It is possible that an analogous position pertained in the early
990s, with Romanos Skleros being sent to the eastern frontier with responsibility
for a similar mobile relief force. Thus, c. 991-2 Romanos may have held an office
such as domestikos of the scholai at the head of tagmata dispatched from
Constantinople, while Bourtzes was, at the same time, doux of Antioch. 104
Clearer than the nature of the command exercised by Bourtzes in the early 990s is
the fact that his conduct in office did much to dissatisfy the emperor Basil. Not only
did Bourtzes lose two major field army engagements with the Fatimids (in 992 and
994), forcing Basil II himself to campaign in northern Syria in 995, he was also
accused of exacerbating the conflict by imprisoning a Fatimid envoy. 105 By 995
Basil had tired of Bourtzes, and replaced him with Damian Dalassenos. 106
However, Damian's duties are also difficult to interpret. Little can be resolved
from the historical record, since a variety of labels are used to describe his position.
He is called 'Doux of the East' by Yahya, in a phrase which echoes the
311
plenipotentiary position held by Bardas Phokas in 986-7. 107 On the other hand when
two later Arab historians, whose common source was the eleventh-century Iraqi
historian Hilal al Sabi, comment on the death of Dalassenos in battle against the
Fatimids in 998, they merely call him Joza. 108 In contrast, the Armenian historian
Stephen of Taron refers to Damian by his title, magistros, rather than by his
office. 109 Meanwhile, with characteristic vagueness Skylitzes notes that Damian
"ruled" in Antioch. 110 Although no sigillographical evidence has ever been directly
linked to Damian Dalassenos, it is possible that he was the owner of an unpublished
seal in the Dumbarton Oaks collection belonging to a Damian doux ul Yet, even
this seal does not add much to any understanding of the nature of Damian's
command at Antioch. More information is forthcoming from the description of
Damian's actual responsiblities in Yahya ibn Sa'id's narrative. This indicates that
Damian's position was still primarily that of an active military commander. For two
years after his appointment he led raids down the north Syrian coast to Tripoli. In
the third year his attacks were focused further inland, as he tried to capture the town
of Apameia on the east bank of the Orontes, then in Fatimid hands. However, when
the Fatimid governor of Damascas arrived to relieve Apameia in July 998, Damian
was killed in battle. 112
312
Damian's replacement was Nikephoros Ouranos, who took up command late in
999 113 por ^e four years prior to m's arrival jn Antioch he had been domestikos of
the scholai of the west on active service in the Balkans, where he had gained
universal renown as the result of an unexpected, but convincing, victory over a
Bulgarian army at Spercheis near Thermopylae in 997. 114 Moreover, it is clear that
despite the peace that was agreed between Byzantine and Fatimid authorities in
1000-1, military affairs continued to dominate Ouranos' attention once he arrived
on the eastern frontier. Shortly after his appointment to command in Antioch he
accompanied Basil II on his campaign to annex the Ibero-Armenian princedom of
Tao in the spring of 1000. 115 During the following year he returned to Tao to repel
the incursion led by Gurgen of Iberia. 116 Several years later in 1006/7 Ouranos
marched from Antioch to Sarudj in the Diyar Mudar, where he won a victory over
an Arab dervish insurrectionist called al-Acfar and his bedouin allies the Banu
Noumeir and the Banu Kilab. 117 Several letters sent by the contemporary krites of
Tarsos, Philetos Synadenos, to Ouranos while he was stationed in Antioch, praise
110
him for his military valour and may refer to this campaign of 1007.
However, it is clear from both literary and sigillographical materials that Ouranos'
responsibilities were broad in their definition and extended over a vast geographical
area. According to the seals he issued while stationed in Antioch he was:
113 Skylitzes, p.345; Yahya, PO 23, pp.400, 460, 466-7; Laurent, 'Gouvemeurs d'Antioche', pp.235-
6
w
114 Ouranos' campaigns in the Balkans are described by Skylitzes, pp.341-2, 364, and Yahya, PO 23,
pp.446-7. Several seals belonging to a Nikephoros, magistros, domestikos of the west have been
attributed to Ouranos (Jordanov, 'Molybdobulles de domestiques des scholes', pp.210-11, mentions
two such seals: one found at Preslav in eastern Bulgaria, the other at Dristra on the Lower Danube;
see also Zacos, Byzantine Lead Seals //, no.863). Leo, metropolitan of Synada wrote to Nikephoros
congratulating him on his success against the Bulgarians while he himself was serving as the
imperial envoy to the Ottomans in the later 990s (Leo of Synada, letter 13).
115 Yahya, PO 23, p.460
116 Stephen of Taron, p.212
117 Yahya, PO 23, pp.466-7
118 Synadenos, letters 8-13; McGeer, 'Tradition and Reality in the Taktika of Nikephoros Ouranos',
p.131
313
"Nikephoros Ouranos, Magistros and Ruler of the East". 119 The amorphous and
universal nature of Ouranos' power is reflected in the fact that most contemporary
literary sources, including Yahya ibn Sa'id, Stephen of Taron, and Philetos
Synadenos, simply refer to Nikephoros by his title of magistros. m Indeed, when
addressing Ouranos, Synadenos was apt to reflect on the august nature of Ouranos'
position with the superlative invocation, peribleptos magistros. 121 The seniority,
ambiguity, and idiosyncratic nature of the office fulfilled by Ouranos on the eastern
frontier suggests that Nikephoros may have been invested with plenipotentiary
powers, which extended far beyond control over Antioch and field army which was
garrisoned there. It is likely that after he had resolved a peace with the Fatimids in
1000-01, Basil was able to concentrate his full resources on warfare with
Bulgaria. 122 Therefore, just as he had appointed Phokas as his plenipotentiary in the
east during the Bulgarian campaign of 986, he now used Ouranos in this broad-
based eastern office while he himself campaigned in the Balkans. The difference,
however, between Phokas and Ouranos was that the latter was unquestionably loyal
to the emperor, whereas the former had betrayed him. 123
The history of command on the eastern frontier in the region of Antioch is much
less full for the second half of Basil's reign than the first. This is mainly because
119 NnoTT^cpffj juaffcrTpq) TQJ Kpatowri -rfc ' AvaroAifc rcji Oopcwtp (Nesbitt and Oikonomides, Catalogue
of Byzantine Seals at Dumbarton Oaks, iii, no.99.11; also published in McGeer, 'Tradition and
Reality in the Taktika of Nikephoros Ouranos', pp. 139-40)
120 Yahya refers to Ouranos as magistros in the context of his expedition against al-Acfar (Yahya,
PO 23, p.367). However, in his testimony for the period 1000-1 he calls Ouranos daux (ibid, p.460).
Stephen of Taron refers to Ouranos in 1000 by his title, magistros, and by the responsibility he had
fulfilled in imperial service during the early 980s, epi ton kanikleiou (keeper of the imperial
inkstand) (Stephen of Taron, p.212); see below, p.336.
121 Synadenos, letters 8-13, especially letter 11.
122 Although Skylitzes' coverage of Basil's Balkan campaigns is extremely inconsistent, his
testimony does contain several dated entries which suggest that Byzantine armies were engaged in
military action in eastern Bulgaria and on the Middle Danube during the early 1000s. For example,
in 1000 Theodorokan and Nikephoros Xiphias seized Preslav and Pliska. The following year Basil
himself campaigned in Thessaly. A year later he conducted a siege of Vidin on the Danube
(Skylitzes, pp.343-6). Yahya notes that Basil's peace in 1000 with the Fatimids allowed him to
launch a four-year campaign in Bulgaria (Yahya, PO 23, p.461).
314
Yahya ibn Sa'id's coverage of the events in northern Syria is very thin between the
years 1000 and 1016. 124 The most that can be said of this largely unrecorded
sixteen-year period is that Ouranos held power between at least 1000 and 1007, the
year when he campaigned against al-Acfar. 125 By 1011 one Michael koitonites was
doux of Antioch, indicating that Ouranos had lost power in Antioch by the end of
the first decade of the eleventh century. 126 However, the very fact that Ouranos was
in command at Antioch for more than seven years signals an important change in
administrative practice on the frontier. Whereas during the first thirty years of
Byzantine rule in Antioch, external and internal political pressures had entailed a
series of ad hoc military commands and a very rapid turnover in office holders,
now, in the predominantly peaceful conditions which followed the 1000-01 accord
with the Fatimids, short-term expediency was able to give way to a greater degree
of permanent command. Moreover, it is possible that this was not a phenomenon
restricted to the eastern frontier, but was also practised on Byzantium's western
borderlands. Although the evidence which supports this argument comes from
Skylitzes' chronologically-confused account of Basil's Bulgarian wars,
nonetheless, it suggests that many of the emperor's senior commanders in the
Balkans often served in the same post for many years. For example, according to
Skylitzes Nikephoros Xiphias was appointed as strategos of Philippoupolis c. 1000.
He was still strategos of the same place in 1014. 127
123 See below, pp.335-7, for Ouranos' earlier career in imperial service and his responsibilities in
Antioch.
124 In most of his coverage for these years Yahya is concerned with Egyptian history (PO 23,
pp.462-520). In particular he concentrates on the eccentric behaviour of the contemporary Fatimid
caliph al-Hakim, who was responsible for persecuting and expelling some members of the Egyptian
Christian and Jewish communities, including Yahya himself. Most infamous of his attacks on the
Christian communities under his authority was his destruction of the Church of the Holy Sepulchre
in Jerusalem in 1009 (Yahya, PO 23, p.492; Forsyth, 'The Chronicle of Yahya ibn Sa'id al Antaki',
chapter 5; M.Canard, 'La destruction de I1 Eglise de la Resurrection par le Calife Hakim et 1'histoire
de la descent du feu sacre', B 25 (1955), pp. 16-43).
125 See above, p.313
126 Yahya, PO 23, p.501; Laurent, 'Gouverneurs d'Antioche', p.236.
315
To return to the east, by the final decade of Basil's reign, there are further signs of
the consolidation of military administration on the frontier, most notably in the
regularity with which the contemporary literary and sigillographical records refer to
the senior commanders at Antioch as either doux or katepan. For example, in 1016
Yahya ibn Said mentions an anonymous katepan of Antioch who allowed the
former emir of Aleppo, Mansour ibn Loulou, to settle in northern Syria. In 1024 he
cites Constantine Dalassenos as the katepan who offered assistance to the Mirdasid
rulers of Aleppo when the emirate was threatened by Fatimid attack. 128 The seal of
Niketas of Mistheia, who took command in Antioch between 1030 and 1032,
indicates that he was patrikios, rector and katepan 129 Another individual identified
by sigillographical evidence as a doux is Theophylact Dalassenos (1032 to 1034). 130
Sigillographical evidence has added the protospatharios Pankratios (whose term in
office has been dated to the reign of Basil himself), the anthypatos and patrikios
Leo, Constantine Bourtzes magistros, and Michael Kontostephanos magistros to
the list of early to mid eleventh-century doukes/katepanes ni Further
sigillographical evidence that the office only became more stable towards the end
of Basil's reign is provided by the fact that the Seyrig collection, with its impressive
array of tenth and eleventh-century seals from the eastern frontier, contains seven
316
seals of the doux/katepan of Antioch: none predates Theophylact Dalassenos (1032-
4).' 32
Yet, this stability in the structure of frontier command may only have been a
temporary phenomenon limited to the first half of the eleventh century. By the
middle of the century, internal and external political and military pressures once
again rendered the office of doux/katepan more flexible. Evidence of this greater
flexibility is visible in the Antiochene career of Romanos Skleros, the great-
grandson of the rebel Bardas. In 1054 Romanos held the position of doux of
Antioch. However, when he was reappointed to command in Antioch in the latter
part of the decade he held the joint position of stratopedarches of the east and doux.
of Antioch. Two contexts can be suggested for the widening of his responsibilities:
first, that Skleros was appointed to this enhanced military position as reward for
supporting Isaac Komnenos' revolt in 1057; and second, that as Turkish invasions
became more frequent and dangerous, reaching as far as Melitene in 1057, and
Sebasteia in 1059/60, more flexibility was required of the Byzantine military
presence across the whole eastern frontier. 133
132 Cheynet, Monisson and Seibt, Sceaux byzantins: Henri Seyrig, nos. 156-62.
133 Laurent, Vatican, no.94; Cheynet, Morrisson and Seibt, Sceaux byzantins: Henri Seyrig: nos. 158,
159: Seibt, Die Skleroi, pp.79-83; Laurent, 'Gouverneurs d'Antioche', p.242; C.Cahen, Pre-
317
speaking it was from these new territories that the katepanates of Iberia and
Vaspurakan were created.
Despite these problems a certain amount of scholarship has been dedicated to the
history of the Caucasian katepanates, with useful contributions coming from
Yuzbashian, Janssens, Seibt, and Arutjunova-Findanjan. 136 However, many issues
Ottoman Turkey: a General Survey of the Material and Spiritual Culture and History, (trans. J.Jones
Williams) (London, 1968), pp.69-71
134 See above, pp.26, 284 n.24
135 An analysis of political and military relations on the north-east frontier during the 1040s has been
expertly pieced together by Shepard, 'Scylitzes on Armenia in the 1040s', pp.296-311.
K.N.Yuzbashian, 'L'administration byzantine en Armenie aux Xe et Xle siecles', REArm 10
(1973-4), pp. 139-83; F.Janssens, 'Le lac de Van et la strategic byzantine', B 42 (1972), pp.388-404;
W. Seibt, 'Die Eingliederung von Vaspurakan in das byzantinische Reich (etwa Anfang 1019 bzw.
Anfang 1022)', Handes Amsorya 92 (1978), pp.49-66; VA.Arutjunova-Fidanjan, 'Sur le probleme
des provinces byzantines orientales', REArm 14 (1980), pp. 157-69; idem., 'The Social
Administrative Structures in the East of the Byzantine Empire', JOB 32.3 (1982), pp.21-34; idem.,
'Some Aspects of Military Administrative Districts in Armenia During the Eleventh Century',
318
remain unexplored, most notably how these katepanates should be interpreted in the
wider history of the organisation of the Byzantine eastern frontier in the tenth and
eleventh centuries. 137 Although this thesis is not the occasion to write such a
broadly-based history of the Caucasian katepanates, some brief remarks about the
early history of the organisation of the Byzantine eastern frontier in Caucasia are
included here, because they reflect many of the developments in frontier command
already observed in the cases of Chaldia, Mesopotamia, and Antioch. Above all,
they demonstrate the over-riding principle that the military administration of the
Byzantine east was always organised on a highly flexible footing, particularly in the
immediate aftermath of annexation.
Of the two katepanates the origins of Iberia are the most obscure. Not only is it
difficult to ascertain whether Byzantine rule extended into Tao on a permanent
basis in 1000 or only after 1022, it is also impossible to identify any commander in
1 3fi
Iberia before the appointment of Niketas of Pisidia in 1025/6. In contrast, the
early history of imperial rule in Vaspurakan is less opaque. Although the exact date
of the surrender of the southern Lake Van is difficult to establish, it is at least
REArm 20 (1986-7), pp.309-20; idem., 'The New Socio-Administrative Structure in the East of
Byzantium', ByzForsch 19 (1993), pp.79-86
A rare, but brief, attempt to interpret the katepanates of Iberia and Vaspurakan within the wider
military command structure of the frontier is undertaken by Kiihn, Die byzantinische Armee, pp. 186-
193.
138 Skylitzes p.370: Niketas' office is identified as that of doux of Iberia (Kuhn, Die byzantinische
Armee, p. 188). It was probably Niketas who was the anonymous katepan who returned Bagrat, the
son of George of Abasgia, to his homeland in 1025 (Georgian Royal Annals, p.284). Bagrat had
been taken hostage as part of the peace agreement reached between Basil E and George after the
emperor's campaigns on the north-east frontier in 1021-2. It is possible that he was also the owner of
a seal in the Zacos collection belonging to a Niketas, patrikios and katepan of Iberia (Zacos,
Byzantine Lead Seals II, no. 1026). For Yuzbashian, however, Romanos Dalassenos was the first
doux of Iberia, appointed in 1023 in the aftermath of Basil's Georgian campaigns ('L'administration
byzantine en Armenie aux Xe et Xle siecles', p. 156, 183). Yet, this identification comes from an
inscription on the Iberian gate at Theodosioupolis/Erzerum which is now lost. The inscription itself
was dated to 991-2, a date which historians have traditionally rejected because of their belief, largely
founded upon evidence of silence, that the katepanate was founded by Basil in 1023. With the
disappearance of the inscription, it is impossible to know when Romanos exercised authority over
Iberia. On the basis of sigillographical evidence Cheynet and Vannier (Etudes prosopographiques,
pp. 83-4) believe that Romanes' brother Theophylact may have been one of the earlier katepanes of
Iberia, perhaps in 1021. Once again this is a view which cannot be solidly substantiated.
319
possible to identify the first two commanders in this region from the testimony of
John Skylitzes, although it should be noted that with characteristic vagueness
Skylitzes fails to mention either the date of their appointment or their exact office:
Basil Argyros, patrikios, having been sent out to rule this land, and having
failed in all respects was released from office. And Nikephoros Komnenos,
protospatharios, was sent as his replacement, who through using a mixture of
force and persuasion on his arrival there, made the land subject to the
emperor. 139
Despite his lack of specificity about the office exercised by Komnenos and
Argyros, Skylitzes provides some useful clues about the nature of frontier
command in this region of the Byzantine east, particularly when his testimony is
aggregated with evidence from other historians. In the first instance, Skylitzes'
observation about Komnenos' use of force to exert Byzantine rule, as well as more
peaceful methods, demonstrates that military action was at the centre of a
commander's responsibilities. This impression is confirmed by the Armenian
historian Aristakes Lastivert, who comments on the brigades of Cappadocian troops
under Komnenos' command. 140 As we saw in the last chapter these Cappadocians
were almost certainly soldiers recruited in central Asia Minor who were sent to the
frontiers as fulltime, professional troops. 141 As the commander of a full-time
garrison prepared to fight to impose Byzantine authority, Komnenos' role in
Vaspurakan closely resembled that of commanders on the Antiochene frontier
during the first half of Basil's reign, such as Michael Bourtzes and Damian
Dalassenos.
139 -fe apxfiv aTRxrraAeiV Bao-t^eiog rrarpiKio^ 6 'Aprvpo$ KOI rdi$ 0^015 irraiow; TrezpaAuerai -r%
<ccu diado%D auToO -n^tnercu NiK^opog npajrrooTTaBapiog o Kanvyvog, og Kara %ti)pa.v Tcvopteuog KOI ra
neifo?, ra #e fyp xptayuEvog vrrqKOOv TQ> fruriheT TJJV ^wpav eron^ev (Skylitzes, p.355)
140 Aristakes, pp.26-7
141 See above, p.242
320
Furthermore, Aristakes' evidence demonstrates the importance which military
experience and authority within the army itself assumed in the deployment of
commanders to the frontier. Just as these attributes were fundamental to the
command exercised by Eustathios Maleinos and George Melias in Antioch and
Cilicia during the reign of Nikephoros Phokas, so they underpinned the
appointment of the eleventh-century katepanes of Vaspurakan. According to
Aristakes, Komnenos was a brave and bellicose man. He had made himself famous
by his, "..... courageous actions and boldness..... and had become renowned
* A
through all the east". Meanwhile, Basil Argyros was also a military figure with
considerable experience in the field. At the beginning of the second decade of the
eleventh century he had been dispatched to deal with the revolt of Meles in
southern Italy while he was strategos of the maritime theme of Samos. 143 The
frequency with which experienced veterans of the Balkan wars of Basil's reign,
both Byzantine and Bulgarian, later held command in the Caucasian katepanates,
indicates the degree to which military competence remained an essential quality for
commanders on this stretch of the eastern frontier. For example in 1034, Nicholas
Chryshelios served as katepan of Vaspurakan. 144 He was the member of a family of
local notables who had surrendered Dyrrachion to Byzantine control in 1005 and
had been rewarded with titles within the Byzantine hierarchy. 145 The military
pedigree of the Chryshelioi is demonstrated by the fact that an eleventh-century
member of the family became domestikos of the Optimatoi. 146 Although, Nicholas
was removed as katepan of Vaspurakan in 1035, his replacement was another
Balkan war veteran. This was Niketas Pegonites, who had led the Byzantine army
in the battle outside Dyrrachion in 1018 at which John Vladislav, the last Bulgarian
321
tsar, had been killed. 147 Other mid eleventh-century doukes/katepanes of Iberia and
Vaspurakan included at least two of John Vladislav's sons, Aaron and
1 JO
Alousianos.
However, while military pedigree was a prerequisite for command in the Caucasian
katepanates, it is clear that martial experience alone did not guarantee a long career
in one location. Indeed, one of the most striking features of the early history of the
katepanate of Vaspurakan outlined above is the rapid turnover in senior
commanders, a characteristic already observed in other sections of the frontier
during the first half of the reign of Basil. Yet, as the careers of Argyros, Komnenos,
and Chryshelios illustrate, a variety of factors propelled this high turnover. Basil
Argyros was removed from office during the reign of Basil himself on the grounds
of incompetence. Here we can detect clear echoes of the replacement of Michael
Bourtzes at Antioch in 995. 149 Incompetence was also the reason why Chryshelios
was dismissed in 1035. He was blamed for allowing the Lake Van city of Perkri,
which had only recently come under Byzantine control, to fall once again into the
hands of local Muslims. 150 Yet, the reason for the short duration of Komnenos'
command at Vaspurakan was rather different. Nikephoros Komnenos was removed
from office not by Basil, but instead by Constantine VIII, either in 1026 or early in
1027. The reason for Komnenos' dismissal was that he was accused of wishing to
rule fc the east', and allying with George of Abasgia in an attempt to further his
147 Skylitzes, pp.357, 388; Gregoire, "Nicetas Pegonites, vainqueur du roi bulgare, Jean Vladislav",
).289-91. The office held by Niketas in 1018 appears to have been strategos of Dyrrachion.
Aaron: katepan of Vaspurakan (Skylitzes, pp.448-52; Zacos, Byzantine Lead Seals II, no.352);
magistros and doux of Ani and Iberia (Dumbarton Oaks Unpublished 55.1.2179); Kuhn, Die
byzantinische Armee, pp. 189-94; Alousianos: described by Skylitzes as a strategos in
Theodosioupolis in 1040 (Skylitzes, p.413). His location in the town of Theodosioupolis in the far
north-east of Anatolia indicates that he almost certainly held the position of katepan or doux of
Iberia Theodosioupolis appears to have been the centre of the katepanate of Iberia before the
annexation of Ani in 1042. See above, pp.96-7, for further analysis of the role of the sons of John
Vladislav within eleventh-century Byzantine political society and administration.
149 See above, p.311
322
plans. 151 Thus, just as dismissals and appointments at Antioch and Mesopotamia
during the second half of the tenth century had often been shaped by political
tensions between generals on the frontier and the emperor in Constantinople, so in
the short turbulent reign of Constantine VIII, distrust between centre and periphery
also contributed to a high turnover in staff on the borderlands.
vi. Conclusions
150 Skylitzes, p.388. Aristakes claims that Perkri was lost because the troops left to guard the city
became drunk (Aristakes, p.36).
151 Aristakes, pp.26-7; Skylitzes, pp.371-2
323
period between 1000 and 1015/6. However, the finalisation of a peace with the
Fatimids in 1000-01 may have brought more stability to the organisation of the
eastern frontier, particularly in the office of doux/katepan. The origins of the
katepanates of Iberia and Vaspurakan are very unclear. Nonetheless, some of the
features of their early histories display striking parallels with developments
elsewhere on the frontier in earlier periods. Taken as whole, supra-thematic military
command over the frontier during Basil's reign was always more flexible than
administrative blueprints such as the 'Escorial Taktikon' imply.
Administration in all regions of the Byzantine empire was not simply about the
principles and logistics underpinning military defence and attack. Instead, imperial
authority was also articulated in the localities through the provision of justice and
the exploitation of resources. However, the question of the civil administration of
the eastern frontier during the tenth and eleventh centuries has attracted much less
interest among modern scholars than the structures of military organisation on the
periphery. Comment has usually been limited to reflections on the apparent lack of
a civil bureaucracy in the eastern-most regions of the empire, or to generalised
assumptions that functionaries from the former regimes were absorbed within the
superstructure of the Byzantine state. 152 Such a lack of interest is curious. For the
second half of Basil's reign, and the reigns of many of his eleventh-century
successors, peaceful conditions prevailed on the eastern frontier, particularly in
regions bordering Muslim states. In these circumstances it might be expected that
military matters assumed a relatively low administrative significance. In contrast, as
we saw in chapter four, peace facilitated the economic expansion of the east,
324
contributing to a prosperity which the emperors in Constantinople were clearly
eager to foster and exploit. 153 Any understanding of the Byzantine administrative
presence on the eastern frontier during the tenth and eleventh centuries must,
therefore, include an investigation of the bureaucratic structures and processes
which facilitated the exploitation of this local prosperity by the Byzantine state.
One sign that civilian administration in the eastern frontier regions, especially in
those areas which belonged to the former Muslim emirates, differed radically from
152 See, for example, Cheynet, Morrisson, and Seibt, Sceaux byzantins de la collection Henri Seyrig,
p. 120
153 See above, pp. 188-93, 203-07
325
contemporary bureacratic structures in the Anatolian themes, is the paucity of
extant lead seals which belonged to officials with judicial and fiscal responsibilities.
For example, while a plethora of lead seals suggests that judges were a burgeoning
administrative phenomenon in the western and central themes of Asia Minor, seals
of judges in the eastern regions of the empire are much less frequently discovered.
They are also much less conspicuous than those of their military counterparts, the
doukes/katepanes 154 Moreover, even when they do appear within the
sigillographical record, it is clear that the authority of a single judge was very thinly
spread over a vast geographical distance. Most of the surviving seals of judges
belonged to officials whose authority coincided with one of the great katepanates
such as Mesopotamia, Iberia, or Antioch, or with the former emirate of Melitene. 155
On occasion their authority could range even further, extending into neighbouring
districts as well. Single judges could preside over joint themes in the interim area
between the eastern plateau and the Anti Taurus such as Lykandos and Sebasteia, or
Lykandos and Melitene. 156 Equally the small border themes known as Armeniaka
themata do not appear to have had their own judges. 157 Instead, these areas were
1 ^8
grouped together under the jurisdiction of a single judge. Finally, the lack of
seals of judges is mirrored by a more general paucity in eastern regions of seals
154 For instance the Seyrig collection contains the seals of seven doukes/katepanes of Antioch and
only one ofapraitor of the region (Cheynet, Morrisson, and Seibt, Sceaux byzantins de la collection
Henri Seyrig, nos. 156-62 and 163). There are no seals of Antiochene kritai in the collection.
155 Melitene. Zacos, Byzantine Lead Seals II, no.952; Dumbarton Oaks Unpublished 58.106.4365;
Mesopotamia. Dumbarton Oaks Unpublished 55.1.3327; Konstantopoulos, Molybdoboulla, no.!55g;
Antioch: Cheynet, Morrisson and Seibt, Sceaux byzantins: Henri Seyrig, no. 163; Iberia: Zacos,
Byzantine Lead Seals II, no.387; Dumbarton Oaks Unpublished F577 and 58.106.865;
Konstantopoulos, Molybdoboulla, no. 177a
156 Zacos, Byzantine Lead Seals II, no.803; Dumbarton Oaks Unpublished 58.106.5399. The
designation of two frontier regions to the authority of a single judge has been noted by Ahrweiler.
She cites examples of the "twinning" of Melitene and Mesopotamia, Lykandos and Melitene, and
Iberia and Mesopotamia (Ahrweiler, 'Recherches', p.85). She also observes the more general
phenomenon of a lack of judges in the east (ibid., pp. 84-5).
There are very occasional exceptions: for example the Dumbarton Oaks Collection contains the
seal of a krites from Hexakomia, one of the Armeniaka themata located south-west of Melitene:
Dumbarton Oaks Unpublished 58.106.4285; Oikonomides, 'L'organisation de la frontiere orientale',
p.290
Schlumberger, Sigillographie, p.296; Zacos, Byzantine Lead Seals II, no.503; Dumbarton Oaks
Unpublished 55.1.3222; 58.106.5582.
326
belonging to other civilian officials, which are so frequently found in western and
central Anatolian contexts. In the course of my research into the sigillographical
record of the former emirates, I have only found one seal belonging to a customs'
official (kommerkiarios)^ one seal pertaining to an official concerned with fiscal
lands (epi ton oikeiakon), 160 and one seal of a protonotarios }6} The absence of
kommerkiarioi is particularly surprising given the importance of trade to the
Byzantine eastern frontier and the post-conquest economic revival in this region. 162
However, the general absence in the former emirates of lead seals belonging to
officials usually associated with the civilian administration of the locality, should
not be taken as evidence of a lack of civilian administration in these regions.
Instead, when other sigillographical and literary materials are examined, it becomes
clear that the civilian administration of the east was simply configured in different
ways from that in the Anatolian themes. Whereas the west and centre of Asia Minor
in the tenth and eleventh centuries was characterised by the penetration of large
numbers of civilian officials dispatched from Constantinople, the
Constantinopolitan presence in the east was on a much smaller scale. This was
because bureaucracy on the eastern frontier was more indirectly managed. Local
administration, above all the collection of taxes, largely remained in indigenous
hands and was articulated according to indigenous practices. These indigenous
functionaries were then responsible to a thin tier of senior Byzantine officials
appointed by the emperor in Constantinople. As a result, the centrally appointed
official was more like a guarantor of tribute than the collector of fiscal dues or the
manager of imperial assets.
327
The starting point for this hypothesis of a tribute-based form of local governance is
a very particular phenomenon in the sigillographical record which has often been
observed by historians, but only recently investigated in much greater detail by
James Howard-Johnston. This phenomenon concerns the marked incidence of seals
of one particular variety of civil official in many locations along the entire length of
the eastern frontier, namely kouratores or episkeptitai. 163 As we saw in the last
chapter, elsewhere in the empire these officials are usually identified with the direct
management of crown estates. And indeed, historians seeking to explain the
incidence of such seals in this more eastern context, including Howard-Johnston
himself, have hitherto worked within this paradigm of administration. As a result it
has been widely assumed that these seals demonstrate that large areas of the eastern
emirates, in particular those lands deserted by Muslims fleeing Byzantine conquest,
were turned into imperial estates and directly managed for the crown in
Constantinople. 164 Yet, there are reasons to doubt this interpretation. First, given the
marked paucity of civilian officials in all other spheres of local government on the
eastern frontier, it seems odd that the imperial authorities had the manpower to
place such emphasis on a single and relatively specialised area of administration.
Second, when the seals of these officials are set in the context of the historical texts
which record the annexation of former Muslim emirates, it makes much more sense
to see their owners as the guarantors of tribute rather than the managers of estates.
328
The strongest support for the idea that kouratores and analogous officials were the
collectors of tribute comes from the only historical account to mention an eastern
kouratoria. This is the description of the turning of Melitene into a kouratoria when
the city was annexed by the Byzantines in 934. According to Theophanes
Continuatus:
They [the Byzantine army] reduced Melitene to such shortage, that they
suddenly captured it, and razed it to the ground, and not only Melitene but also
its neighbouring cities and districts which were highly productive and very
fertile and <could> yield many other revenues. Having then turned Melitene
into a kouratoria, the emperor had many thousands of [pounds?] of gold and
silver raised annually in revenues from there. 165
Here, the crucial term indicating that many of the kouratoriai of the Byzantine east
were compatible with a tribute paradigm of local government, is the verb used for
the raising revenues: daovofapeTotiai. The principal meaning of daojw<; in Greek is
"tribute". 166
eig rotravrviv orevtocnv Ttjv Mc/jnvyv nepiecrrQcrav wore cwryv crwroiMtrrarov Knop&iyraA KCU et^
Iftfi > \ * r *L \ ' " ' ' ' >/><* \ tl
fia^toug Karcurrp&lxjU, ou pavov dc ravryv oAAa /ecu rag oftopoyg curm TroAeij /cai xtopag TmXu<f>6poug re KCU
mordrttg ovcrag KCU oi'ag nohhag irapexetv irpocroctovg. rauryv ow rr^v McAtrivyv eig Kovparajpiav
anoKarcurrfjcrag o fZacrtXevg TroAAa^ %tXiadag xpwriou KCU apyupiov CKeiSev 8acrjjio<fape?cr6ai errnritog
TrerronjKev Theophanes Con., pp.416-7).
166 H.G.Liddell and R.Scott (H.S.Jones and R.McKenzie), A Greek-English Lexicon (Oxford, 1968,
second edition), p.370
329
attack bases in the mountains and roads that surrounded the city. From these bases
Byzantine commanders were encouraged to raid the countryside around Antioch
itself each day, thus persuading the inhabitants within the city to surrender. 167
However, in a tribute-related context, Leo the Deacon's account of the strategic
reasoning behind this attritional strategy is particularly striking. In a passage of
direct speech which Leo attributes to the emperor himself, Nikephoros Phokas is to
be found arguing that the object of his military policy was to compel Antioch to
__
A tribute relationship between centre and locality in the civil administration of the
eastern frontier also helps to explain the important but rather ambiguous position of
officials described in the historical record as basilikoi. Whenever basilikoi are
discussed by modern historians they are usually attributed with a general role in
fiscal and judicial administration. Sometimes they are seen as analogous to
kouratores. Yet, their role is rarely discussed in detail, and it is widely assumed that
they were lower-ranking functionaries subordinate to more senior officials such as
the provincial krites. 169 However, the position of basilikoi on the eastern frontier in
the aftermath of the tenth-century Byzantine conquests was of much greater
significance than this modest definition implies. Instead, as the careers of two very
famous basilikoi from the later tenth-century frontier demonstrate, these were the
figures on whom the emperor in Constantinople, and even usurpers such as Bardas
Skleros, had to depend in order to mobilise the resources of the great former
emirates.
330
One of these basilikoi was Koulei'b, whose career is predominantly known from the
historical testimony of Yahya ibn Sa'id. 170 He was a Christian Arab and Hamdanid
servant who surrendered the fortresses of Barzouyah and Saoune in northern Syria
to John Tzimiskes in 975, during that emperor's last great eastern campaign. In
return Tzimiskes gave him the title of patrikios and the office of basilikos of
Antioch. 171 During the Skleros revolt Koulei'b surrendered Antioch to the rebels,
and was appointed basilikos in Melitene instead. 172 When Skleros fled to Baghdad
in the aftermath of the failure of his revolt, Koulei'b did not go with him. Instead, he
retained his position at Melitene. When Skleros returned to the empire from
Baghdad in 987, nearly a decade later, Koulei'b still exercised authority in
Melitene. 173
169 Ahrweiler, 'Recherches', pp.73-4. Cheynet, 'L'apport arabe a raristocratie byzantine', pp. 141-2,
discusses Kouleib in this light.
170 According to the spelling of Arabic names in the Encyclopaedia of Islam (the system to which
this thesis has usually adhered), the name 'Kouleib' should be rendered as 'Kulayb'. However, since
the form 'Kouleib' is used by the editors of Yahya ibn Said, I have retained it for ease of future
cross-referencing.
171 Yahya, PO 23, p.369
172 Ibid., p.373
331
It was Kouleib, for example, who was able to ensure the annual delivery of the
tribute of Aleppo. As a result of these intermediary skills, he alone of Skleros
supporters had received a pardon when the first Skleros revolt had collapsed in 979.
Moreover, he had been allowed to keep the estates he had been granted by John
Tzimiskesin975. 174
Further signs that Kouleib was a high profile figure in the frontier world who was
well rewarded by authorities at the centre, can be detected in an early eleventh-
century monastic chronicle from Melitene, which Michael the Syrian inserted into
his history in the twelfth century. According to this contemporary chronicle,
Kouleib, who was also known by his Greek name and title Eutychios the patrikios,
sponsored the monastery of Bar Gagai near Melitene in 987/8. 175 It is even possible
that Kouleib and his family were so important to relations between the centre, the
periphery, and the emirates beyond the empire's eastern border, that they survived
the turmoil of the second phase of Skleros insurrection (987-9) and continued in the
service of the Byzantine state after 989. The Zacos Collection contains a seal
belonging to Bardas the son of Kouleib. 176
Another basilikos of critical political importance at the start of Basil's reign was
Obeidallah, another Arab Christian. 177 In 976 he was basilikos of Melitene. By
173
Ibid., p.420; Laurent, 'Gouverneurs d'Antioche', p.231
174
Al-Rudhrawari pp.23-4; see above, pp.260-1
175
Michael the Syrian, pp. 125-6. The date 987/8 is that provided by Michael the Syrian's account.
However, according to the much later testimony of the thirteennth-century historian Bar Hebreus
(who used Michael the Syrian as one of his sources) Kouleib supported Bar Gagai a decade earlier
in 977/8. Given that Michael is the more contemporary source, his is probably the account to be
accepted. Bar Gagai rapidly became a great centre of Syrian learning. A manuscript from the
monastery dated to 994 is to be found at the monastery of Saint Mark in Jerusalem. Another
manuscript, now found at Mosul, was copied at Bar Gagai in 1013 (Dagron, 'Minorites ethniques',
pp.192, 197)
Zacos, Byzantine Lead Seals II, no.371; Cheynet, 'Du prenom au patronyme', pp.60-2; Guilland,
Recherches, p.288
177 According to the spelling of Arabic names in the Encyclopaedia of Islam, the name 'Obeidallah'
should be rendered as 'Ubayd Allah' However, since the form 'Obeidallah' is used by the editors of
Yahya ibn Said, I have retained it for ease of future cross-referencing.
332
surrendering the city to the rebel forces of Bardas Skleros, he enabled Skleros to
sequester the fiscal revenues of the former emirate, and openly declare revolt
against the emperor. Still in the service of Skleros, Oubeidallah became Kouleib's
1 "7fi
successor as basilikos of Antioch. Basil II was only able to regain Antioch for the
imperial "side" in 977/8 by promising Oubeidallah the position of "governor" for
life. 179 Yahya ibn Sa'id's account of Oubeidallah's actions during the civil war at
Antioch make it clear that the basilikos not only exercised civilian responsibilities,
but even some degree of military power. For example, once he had defected to the
emperor, Oubeidallah defended Antioch against armed attack by two senior Skleros
lieutenants: Sachakios Brachamios and Ibn Baghil. Furthermore, he suppressed a
revolt by local Armenians. Although Yahya ibn Sa'id claims that the citizens of the
city were his chief source of political support, his ability to beat off attack by
leading Skleros commanders such as Brachamios indicates he must also have had
some authority over an armed garrison. 180
Yet, while Kouleib and Oubei'dallahs' authority as basilikoi may have been
heightened by the exigencies of civil war, the very broad nature of their jurisdiction
is echoed in the responsibilities of other basilikoi who exercised power in the
former emirates, but who are only known through the sigillographical record. At
least three such seals are extant: Chosnis, basilikos of Tarsos; John, krites of the
central Constantinopolitan court of the Hippodrome, basilikos of Melitene and the
Armeniaka themata^ and Solomon, basilikos of Melitene, and megas chartoularios
__ 101
of the main Constantinopolitan tax-collecting bureau of the Genikon. The wide
range of responsibilities exercised by the owners of last two examples, John and
Solomon, demonstrate the judicial and fiscal authority of basilikoi, competences
178Yahya,PO23,p.373
179 Ibid., pp.375-7. The term used by Yahya to denote "governor" is the Arabic "wilaya". Laurent,
'Gouverneurs d'Antioche', pp.231-2 .
333
which as we have seen were practised by Oubeidallah and Kouleib. However, it is
worth noting the Constantinopolitan affiliations of the judicial and fiscal offices
held by these eleventh-century basilikoi, a connection which was absent from the
careers of basilikoi active in the tenth century such as Koulei'b or Oubedeillah. This
development suggests that greater political stability within the empire itself, and
peaceful relations with the eastern neighbours during the eleventh century, meant
that the key intermediary functionaries on the eastern frontier could increasingly be
drawn from Constantinople, rather than from former Hamdanid servants or other
local notables.
A brief examination of the ecclesiastical and secular history of northern Syria and
Cilicia indicates that the greater use of Constantinopolitan officials in the exercise
of intermediate power on the eastern frontier almost certainly began in the last
decade of the tenth century, during the reign of Basil II himself. Central to an
understanding of this change is the career of Agapios, the later tenth-century
patriarch of Antioch, another key figure on the frontier whose authority originally
sprang from his ability to mediate between imperial authority in Constantinople and
the local populations of the Byzantine east. Agapios' rise to power began during the
first Skleros revolt with the death of the incumbent patriarch Theodore in May 976.
Motivated by ecclesiastical opportunism Agapios, the local bishop of Aleppo,
travelled to Constantinople to persuade the emperor to appoint him as Theodore's
replacement. In return he promised to persuade Oubei'dallah, the rebel basilikos of
Antioch, to declare for the emperor. 182 Despite Agapios' relatively junior status as
the bishop of Aleppo, Basil and his advisors were so desperate to regain political
334
control of Antioch that they agreed to this plan. 183 Agapios returned to Antioch,
entering the city secretly. He came to terms with Oubei'dallah, and was installed as
patriarch in January 978. 184 During the next decade he used the authority he had
been granted by Constantinople to secure his own position in the locality. At the
heart of Agapios' policy was the promotion of the Antiochene Melkite church at the
expense of the Syrian monophysite faith. According to later Syrian historians
Agapios burnt the books of Syrian churches, forced local notables to have their
children rebaptised as Chalcedonians, and then deployed these converts as local
clergy in rural northern Syria. 185
However, Agapios' power as a mediator between locality and centre, living on the
periphery of the Byzantine empire was short-lived. Twelve years later in the
aftermath of the Phokas revolt, Basil II decided to extricate himself from
dependence on local figures such as Agapios. Accused of colluding with the Phokas
family, Agapios was summoned to Constantinople and secluded in a suburban
t o/r
monastery. As the emperor's authority strengthened during the 990s, he began to
extend his authority even more energetically into the localities. In 996 Agapios was
officially deposed. His replacement was a Constantinopolitan, John the
chartophylax of the Hagia Sophia. 187 Soon John was joined in the east by other
Constantinopolitan officials, including his friends and correspondents, Philetos
Synadenos, krites of Tarsos, and most famous of all, the supreme military
188
commander in the east, Nikephoros Ouranos.
183 Eli, the Chalcedonian patriarch of Alexandria, consistently refused to recognise Agapios as
patriarch of Antioch on the grounds that he was far too junior to have been granted such a lofty
position (Yahya, PO 23, pp.378-89).
184 Yahya, PO 23, p.377
185 Michael the Syrian, pp. 131-2
186 Yahya, PO 23, p.428
187 Ibid, pp.445-6
335
As we saw earlier in this chapter, Ouranos arrived in Antioch in 1000/01 as kraton
of the east with plenipotentiary powers over the whole eastern frontier. 189 At the
most basic of levels this was a military position. Yet, in other respects, Ouranos'
appointment as plenipotentiary represented an imperial desire to use a
Constantinopolitan official to fulfil the intermediary position previous occupied by
local functionaries such as the basilikoi Kouleib and Oubei'dallah. Certainly
Ouranos was ideally suited to such a wide-ranging role. In addition to his
impressive military pedigree, Ouranos was also able to call upon extensive
experience in administrative and diplomatic affairs, competences that had been
fundamental to the authority of Koulei'b and Oubei'dallah. Ouranos' early
professional life had been spent in Constantinople within the imperial palace and
the upper echelons of central administration. By 982 he was keeper of the imperial
inkstand, a position which required him to become competent in the handling of
1QO
sophisticated documents including imperial chrysobulls. His knowledge of the
administrative practices and court politics of Constantinople was so well-regarded
that at some point during the mid- to later 980s, while he was still keeper of the
imperial inkstand, he was appointed epitropos, or lay guardian, of the Athonite
monastery of the Lavra. 191 In the exercise of this responsibility Ouranos must have
gained valuable experience in acting as an intermediary between the interests of a
locality and central government. Moreover, his early career also brought him into
contact with the machinations of high politics and diplomacy with Byzantium's
eastern neighbours. During the early 980s he was involved in the intense diplomatic
188 For letters exchanged between John, Nikephoros and Philetos, see Darrouzes, Epistoliers
byzantins, passim.
189 See above, p.313
190 Several of Nikephoros' own letters seem to date from the period when he was still keeper of the
imperial inkstand (Ouranos, letters 3-6; Laurent, L'Administration centrale, p. 102). In his letter to
Anastasios, the metropolitan of Laodikeia, he displays his familiarity with the handling of imperial
chrysobulls (Ouranos, letter 3). He asks the metropolitan to submit all the chrysobulls of the see for
his perusal. It is possible that a seal of Nikephoros, anthypatos, patrikios and epi ton kanikleiou, may
have belonged to Ouranos (Zacos, Byzantine Lead Seals II, no.861).
336
negotiations with the Buyids which surrounded the exile of Bardas Skleros in
Baghdad. He became a close acquaintance of the Buyid envoy ibn Shahram during
the latter's mission to Constantinople over the winter of 981/2. When ibn Shahram
returned to Baghdad, Ouranos travelled with him as the Byzantine ambassador to
the court of Adud al-Dawla. Shortly after his arrival in the Buyid capital, Adud al-
Dawla died, and Ouranos found himself, like Skleros, confined to prison. 192 He was
eventually released early in 987, the date when Skleros himself returned to the
empire. 193 However, even these relatively barren years may not have been wasted.
As a result of his friendship with ibn Shahram and his captivity in Iraq, it is
possible that he even learnt some rudimentary Arabic. All these skills would have
stood him in good stead for his plenipotentiary role on the frontier in the first
decade of the eleventh century.
191 Ouranos' appointment must post-date 984 and pre-date 999 (Actes de Lavra, pp. 19-20, 45-6, 52,
and doc. 31; McGeer, 'Tradition and Reality in the Taktika of Nikephoros Ouranos'. pp. 130-1).
192 Al-Rudhrawari, pp.25-34; Skylitzes, p.327; Yahya, PO 23, pp.400-02
193 Yahya, P0 23, p.420
194 "Ouranios [the heavenly one] made me come": Synadenos, letter 11. The exact date of Philetos'
arrival in Tarsos is uncertain. But he must have been in the east by 1007 since he wrote to Ouranos
congratulating him on his victory over the Arabs in this year; see above, p.313.
337
1 QX:
Seleukeia and kourator of Tarsos. Moreover, it is possible that he was the owner
of a seal belonging to Philaretos "Krites of the East, Exaktor, and Illoustrios". 191
Much like the office of krites, the office of exaktor was concerned with both the
provision of justice and the exercise of fiscal responsibility. As both a krites and
exaktor, the owner of this seal clearly exercised a host of judicial and financial
competences over a wide geographical area, a formula typical of administration on
the eastern frontier. Furthermore, while it is dangerous to read substantive
meanings into the elusive literary artefacts which passed between senior officials
such as Synadenos and Ouranos, it is possible that an elliptical allusion to the
incompatibility of learning and the bearing of arms contained in one of Synadenos1
letters to Nikephoros, may reflect the wide range of duties, including military
service, that officials on the frontier were expected to undertake in imperial service.
If this is so, Philetos implies Ouranos was better equipped than himself:
On the one hand I have lost the capacity to be wise and to be called wise, and on
the other, I am completely inexperienced in the bearing of arms, the rattling of a
spear, the moving and shooting of an arrow, and the shaking of a spear against
the enemy, and as much as is required to make war against the foe - for I am not
hardhearted or very daring, but someone undaring and feeble -1 have failed at
both: for I am now neither wise, nor daring, in the face of the enemy. And so
tell me who I am, wise Strategos. As for me, what I had I have thrown away,
what I had not, I am unable to take hold of, and that which I am, as you see, I
have lost 198
195 Synadenos' letters appear in manuscript 706 from the monastery of Saint John on Patmos
(Darrouzes, Epistoliers byzantins, pp. 9-12).
196 Eustathios Romaics, krites of Seleukeia and megas kourator of Tarsos (Konstantopoulos,
Molybdoboulla, no.!47a); see above, p.328, n.163 for Euthymios Karabitziotes exaktor, krites of the
Hippodrome and Seleukeia and kourator and anagrapheus of Tarsos and John Hexamilities, krites
of Seleukeia and kourator of Tarsos.
197 Nesbitt and Oikonomides, Catalogue of Byzantine Seals at Dumbarton Oaks, iii, no.86.34. The
editors believe that the owner of the seal was merely the krites of the theme of the Anatolikon on the
grounds that the responsibilities of a single judge could not have extended over an area as great as
the "East". However, as this chapter has demonstrated, individuals in both military and civilian
offices in the eastern-most regions of the Byzantine empire customarily exercised authority over
very large regions indeed.
338
It is clear from Philetos' self-pitying statement that officials appointed to act as
intermediaries on the frontier experienced a profound sense of frustration and
bewilderment at the panoply of commitments which greeted them on their arrival
from Constantinople. Yet, paradoxically, their trauma is also very strong evidence
of the relatively limited nature of the administrative changes set in train by Basil II
during the second half of his reign. For while it is true that officials such as
Ouranos and Synadenos represented a new Constantinopolitan presence in the
locality, nonetheless by acting intermediate plenipotentiaries they continued to
exercise the same role once fulfilled by local notables such as Koulei'b and
Oubeidallah. While the appointment of such Constantinopolitan figures indicates
that the role of intermediary was subject to greater control from the centre after
1000, there is little sign that their presence represented a profound shift in the
governance of the locality at an everyday level. There is no evidence, for example,
to suggest that during the second half of the reign of Basil there was any change in
the basic tribute relationship between locality and centre. Instead, during the
eleventh century, eastern regions continue to be characterised by the lack of seals of
civilian officials found elsewhere in the empire. This trend within the
sigillographical data militates against the possibility that new fiscal and judicial
administrative structures and practices were imposed on the frontier region, or that
large numbers of officials from the capital began to arrive in the east to take up
junior positions within provincial bureaucracy. Instead, it seems more likely that
underneath a thin tier of centrally appointed officials such as Ouranos and
Synadenos, the quotidian management of the frontier remained in the hands of
indigenous officials.
339
Until further research into the little-studied history of the civil administration of the
Byzantine eastern frontier can be undertaken, these conclusions must remain
provisional. However, it is worth noting the evidence from a frontier region north
of the former Muslim emirates, which suggests that Byzantine administration
continued to depend on indigenous officials during the eleventh century. This
evidence comes from the Caucasian katepanate of Iberia. During the reign of
Constantine Doukas (1059-67), the katepan of the east, Bagrat Vxkac'i, introduced
a series of tax concessions for the northern Armenian city of Ani. Notice of these
arrangements was inscribed on the west wall of the city's cathedral. The inscription
itself was written in contemporary vernacular Armenian, and could thus be read by
the local inhabitants. More important, the inscription tells the citizens that the will
of the katepan is to be executed by a series of officials called tanuter-s, all of whom
are identified as local Armenians: Mxit'ar hypatos, Grigor spatharokandidatos, and
Sargis spatharokandidatos. Modern Armenian historians have suggested that these
tanuter-s were the managers of local economic and fiscal affairs, whose duties
resembled those performed by functionaries known as ra 'is within Islamic towns.
That is to say, they acted as the spokesmen for their own communities within
different quarters of the city, and were responsible for managing the fiscal
relationship between those communities and the local representative of centralised
authority. 199 As such they were the key intermediaries around whom a tribute-based
system of taxation could operate smoothly.
340
eastern regions continued to be typified by a tribute relationship between centre and
locality. From the middle of the reign of Basil II onwards, the vital intermediary
representatives of the centre in that locality were usually nominees dispatched by
Constantinople. Yet, the indirect, tribute-paying relationship over which they
presided remained essentially unaltered from arrangements first established in the
tenth century. These arrangements appear to have persisted until the arrival of the
Turks in the second half of the eleventh century. Thus, successive tenth and
eleventh-century emperors, including Basil II himself, acknowledged the logic that
in order to exploit the prosperity of the eastern localities, it was essential to utilise
local officials.
341
Conclusion
The long reign of Basil II (976-1025) and his brother Constantine VIII represents
the apotheosis of Byzantium as a medieval territorial power. When Basil died in
1025 Byzantium was the most powerful state in the eleventh-century Near East,
rivalled only by Fatimid Egypt. Yet, despite its significance to the history of
Byzantium, and indeed to the history of medieval Europe, this long reign has
attracted very little attention from modern historians. No major analysis has
appeared since the publication of Gustave Schlumberger's L 'Epopee byzantine a la
fin du dixieme siecle at the end of the nineteenth century.
This dissertation has developed the preliminary stages in the composition of a new
analytical study of Basil's reign. Although the compilation of a sustained narrative
must be the eventual ambition of this fresh appraisal, this thesis has not been
primarily concerned with piecing together a new chronology. Indeed, it has been
argued that until a new narrative source in the medieval historical record is
discovered, a strictly chronologically-driven analysis is unlikely to expand
significantly on the portrait of the reign offered by Schlumberger a century ago.
Instead, in a series of six studies, this thesis has looked below the chronological
surface of the reign, at the structures and processes of power within later tenth- and
early eleventh-century Byzantium. Fundamental to each study has been a keen
sensitivity to the potential and limitations of the surviving evidence, both literary
and material. In many respects, this thesis has explored how historians should
exploit and deploy evidence in the writing of early medieval history, particularly
for periods when the surviving sources are extremely fragmentary and heterodox in
nature.
342
The first half of the thesis analysed the medieval historians who report on Basil's
reign. This analysis has been predicated on the belief that in order to understand
the political history of the later tenth and early eleventh centuries, it was first
necessary to establish the background to the historical writings which report on this
period. However, rather than devoting only superficial and partial coverage to all
the principal medieval historians in question, this historiographical discussion
focused on the 'Synopsis Historion' of John Skylitzes, the late eleventh-century text
which contains the first Greek narrative appraisal of the entire reign. Three inter-
linked studies looked in detail at the text itself and at the contexts in which it was
written. Skylitzes' working methods, his use of source materials, his purpose in
writing, and his audience were examined. In the course of this investigation it was
suggested that Skylitzes' text was compiled in accordance with the interests of an
elite audience in the Constantinopolitan court of the emperor Alexios Komnenos
(1081-1118). These contexts underpin the text's middling- to elevated language
level, its close affiliation with high-style histories, its stress on military valour, and
its overwhelming interest in the deeds and pedigrees of leading aristocratic families,
especially those which were still politically important at the end of the eleventh
century. The Balkan conflicts of the first two decades of Alexios' reign also explain
Skylitzes' detailed coverage of earlier Byzantine wars in Bulgaria, especially those
of John Tzimiskes and Basil II.
343
policy. Yet, for the historian of Basil's reign, the later eleventh-century concerns
which condition Skylitzes' testimony pose significant problems. The aristocratic
slant to his text means that it is very easy to misjudge the balance of power within
the Byzantine polity during the later tenth and early eleventh centuries. The third
chapter of this thesis illustrated the extent to which the pro-Skleros source, which
underpins both Skylitzes' and Michael Psellos' coverage of the first thirteen years
of Basil's reign, distorts the relative importance of the revolts of Bardas Skleros and
Bardas Phokas, overstating the significance of Skleros, and underplaying Phokas.
More significantly, Skylitzes' emphasis on the deeds of the aristocrats means that
his text seriously underestimates the contemporary authority of the emperor. This
marginalisation of the emperor's role is particularly visible in Skylitzes' Balkan
coverage.
In order to gain a more accurate view of the structures and processes of power
operating within Byzantine elite society during Basil's reign, the second half of this
thesis looked beneath the veil of the later eleventh-century Greek historical record
at other forms of literary and material evidence. In chapters four to six the main
subject matter was the articulation of imperial authority in the eastern half of the
empire during the later tenth and early eleventh centuries. However, before
considering the reign of Basil II itself in detail, two key background contexts were
considered. The first of these discussions, dedicated to the economy, demonstrated
that all regions in the eastern half of the Byzantine empire experienced a significant
increase in prosperity during the tenth and eleventh centuries. The second
discussion, devoted to administration, argued that the expansion of the empire's
territorial borders across the Taurus and Anti Taurus mountains during the tenth
century had two principal ramifications: the demilitarisation of the core Anatolian
344
themes, and the development of a new, mobile, centralised army. However, in both
into the locality were observed. One limitation was the fact that the Anatolian
plateau lacked the demographic and capital resources necessary to repopulate and
revivify the economy of eastern Anatolia and the eastern frontier. Another
constraint was the expense of the more aggressive Byzantine military enterprises of
the second half of the tenth century. Nonetheless, before Basil himself came to the
throne in 976, active imperial responses to both these problems were already
visible. On the one hand, imperial authority sought to work with, rather than
against, the demographic trend, by encouraging non-Greek speaking and non-
Chalcedonian communities to settle in the eastern regions. On the other hand, the
army was refinanced. Military service in the more prosperous western themes of
Anatolia was commuted to a fiscal levy. This tax was then used to subsidise troops
on active service in the east.
A lack of written source materials makes it difficult to examine how these tenth-
century trends in administrative practice developed during Basil's reign. However,
by setting occasional literary references against the much more extensive evidence
provided by lead seals, some outlines of the articulation of imperial authority in the
east during the later tenth and early eleventh centuries have emerged. In the
Anatolian themes the longue duree trends of demilitarisation and centralisation
continued throughout Basil's reign and into the eleventh century. The extent to
which local administration was typified by a civil officials, or military functionaries
345
sigillographical and literary evidence suggest that the exploitation of the localities
was usually indirect rather than direct, concerned with the raising of taxes rather
than with the direct management of assets. There is some evidence to indicate that
the state lacked the manpower and the capital resources to administer large amounts
of immovable property profitably. As far as Basil's reign is concerned, these
observations suggest that despite the rhetoric of his novel of 996, only modest
amounts of private land passed into the hands of the state permanently.
In the east, beyond the Taurus and Anti Taurus mountains, Basil inherited a highly
fluid political, diplomatic, and military situation. This fluidity demanded a more
flexible and ad hoc administrative response than has sometimes been acknowledged
by Byzantine historians. In terms of military administration, political pressures both
within and outside the empire dictated a rapid turnover in senior commanders, and
considerable malleability in the duties they were expected to perform. Civil
administration was left largely in the hands of local intermediary figures, who
supervised a quasi tribute relationship between the eastern periphery and
Constantinople. Greater stability of military command was only achieved in the
second half of Basil's reign, when a series of diplomatic settlements with Muslim
powers brought peace to the eastern frontier. The most important of these
settlements was the peace agreement of 1000-01 with the Fatimids of Egypt. These
more peaceful conditions also enabled civil administration to be supervised more
closely by imperial authorities in Constantinople. However, there is little sign that
the fundamental tribute relationship between centre and locality was altered
significantly either during Basil's reign, or later in the eleventh century.
346
Taken as a whole the administration of the Byzantine east during Basil's reign is
marked by very conspicuous differences between the Anatolian themes and the
eastern periphery. The themes were significantly demilitarised, whereas, even after
1000, the eastern frontier remained a militarised zone. In the sphere of civil
administration the themes were typified by the conspicuous penetration of officials
from Constantinople into the locality; this Constantinopolitan presence on the
eastern frontier was always much less visible. However, underpinning these
differences there is a conspicuous common theme. Fundamental to both systems
was the willingness of the state to work within practical administrative constraints
in order to extract maximum financial gain.
347
those tenth- and eleventh-century texts which reflect on the appropriation and
manipulation of imperial power by the servants of the emperor. Such sources
include the 'poor' versus the 'powerful' legislation, the will of Eustathios Boilas,
the advice manual of Kekaumenos, and the letters of imperial functionaries such as
Nikephoros Ouranos, Leo of Synada, and Philetos Synadenos. The broad economic
and administrative surveys undertaken in the second half of this thesis should
provide the key contexts against which these rather isolated texts can be better
understood.
348
after 989. As chapter five indicated, although Basil confiscated some landed
property from his political foes, sequestration was confined to a small number of
families. It was used for the purposes of example rather than as a systematic policy.
Meanwhile, it is clear that those members of the political elite, whose authority had
once been vested in the army, often returned to public duties. This principle is
visible in the careers of Romanos Skleros (son of Bardas), Michael Bourtzes and
Leo Melissenos, highlighted in the discussion of the katepanate of Antioch in
chapter six.
Yet, while control of land was not the principal cause of political convulsion during
the reign of Basil, there was still a structural problem within the senior echelons of
political society with which the emperor had to deal if domestic insurrection was to
be prevented. That structural problem was control of the army. As the third and
sixth chapters of this thesis demonstrated, the unstable military and diplomatic
situation on the eastern frontier, during the first half of the reign, presented senior
army generals with levers of immense political power. Located on the frontier,
generals such as Skleros, Phokas, Bourtzes, and Melissenos, regardless of their
personal wealth, had access to potent tools, including control of a large mobile
army, and the capacity to construct diplomatic alliances with neighbouring powers.
While usually used to prosecute warfare against Byzantium's external foes, these
tools were all too easy to turn against the emperor in Constantinople. In this
context, the key political change of Basil's reign was not the defeat of Skleros and
Phokas in 989, but the peace of 1000 with the Fatimids. As long as relations
between Byzantium and its Muslim neighbours remained hostile, senior
commanders and large armies would be needed to garrison the eastern borderlands.
From this 'hot' frontier, another powerful general could always emerge, threatening
Basil's authority in Constantinople. However, with peace in the east, this threat was
349
dissipated. Instead of having to employ experienced army generals, many of whom
had much better military records than the emperor himself, Basil could now deploy
Constantmopolitan bureaucrats in eastern commands, such as Nikephoros Ouranos
and Philetos Synadenos, men with a record of proven loyalty. Meanwhile, the
Byzantine field army and its principal commanders could be relocated to the
Balkans to fight the Bulgarians under the close tutelage of the emperor himself.
The centrality of the peace with the Fatimids indicates clearly the extent to which
the domestic political history of the reign of Basil II was conditioned by the
empire's diplomatic and military relations with its geographical neighbours. While
this thesis has concentrated on the east, future research must expand to include the
Balkans, and above all, Basil's dealings with the first Bulgarian empire of the tsar
Samuel Kometopoulos. Only then will it become possible to write the critical
narrative appraisal of the reign of Basil II, which has been lacking for so long in the
secondary literature of the history of medieval Byzantium.
350
Appendix
Skylitzes
First the monk George, who also served the most holy Patriarch Tarasios as
synkellos, and after him the homologetes Theophanes, the hegoumenos of the Agros
[monastery], made an epitome of history excellently in the style of the ancients,
having very persistently pursued historical books, and having summed them up in a
language which was simple and uncontrived, indeed being concerned only with the
essence itself of what happened. One of them, that is George, began from the
Creation and reached the usurpers, I mean Maximianos and his son Maximinos.
The other, that is Theophanes, having made George's end point his beginning,
abridged the rest of the chronography, and having arrived at the death of the
emperor Nikephoros, the former Logothete of the Genikon, stopped his account.
After him no-one else has dedicated himself to such an enterprise. For some have
tried, such as the didaskalos Sikeliotes, and the hypertimos Psellos, hypatos of the
philosophers in our own time, and others in addition to them. But having
undertaken the task in a desultory way, they both lack accuracy; for they disregard
very many of the more important events, and they are of no use to their successors,
since they have made merely an enumeration of the emperors and indicated who
took imperial office after whom, and nothing more. For even if they seem to
351
mention certain actions, even then, since they have narrated them without accuracy,
they hinder those who chance upon them [ie. later readers] and have not helped. For
Theodore Daphnopates, Niketas the Paphlagonian, Joseph Genesios and Manuel,
[that is] Byzantines [ie. Constantinopolitans], and Nikephoros the deacon of
Phrygia, and Leo the Asian, and Theodore of Side who became proedros [ie.
archbishop], and his nephew and namesake [Theodore] archbishop of the church in
Sebasteia, and in addition Demetrios of Kyzikos, and the monk John the Lydian,
each has had his own agenda, the one proclaiming praise of the emperor, the other
a psogos of the patriarch, another an encomium of a friend; but while each one
fulfils his own purpose in the guise of history, each has fallen short of the intention
of those aforesaid men inspired by God [ie. George and Theophanes]. For they
wrote histories at length of the things which happened during their times and
shortly before: one sympathetically, another with hostility, another in search of
approval, another as he had been ordered. Each one composing their own history,
and differing from one another in their narrations, they have filled the listeners with
dizziness and confusion. Having found pleasure in the labour of the aforesaid men
[ie. George and Theophanes], we have hoped that a synopsis would be of not
inconsiderable profit for those who love history and most of all for those who
prefer that which is very easy to that which is more wearisome; [a synopsis which]
gives a very shortened account of the events in different times, which is free from
the weight of documentation. We have read the histories of the writers mentioned
above carefully, and have removed that which was written in a state of emotion or
in search of approval, and have disregarded differences and disagreements, and
have shaved off whatever we have found which is too close to legend, and have
gathered that which is suitable and whatever does not result from rhetoric, and have
also added whatever we have learnt from old men by oral testimony. Having put
them together into one [unit] rapidly we have left behind for posterity nourishment
352
which is soft and finely ground in language, so that on the one hand those who are
acquainted with the books of the aforesaid historians [Theodore Daphnopates etc.]
may have a record by using and approaching this book as a travelling companion -
for reading can foster recollection, and recollection can nourish and increase
memory, just as in contrast neglect and idleness foster forgetfulness, which Lethe
follows, dimming and confusing the memory of deeds completely. [We have also
left this nourishment] on the other hand so that those not yet acquainted with
histories may have this epitome as a guide, and by examining that which is written
with a wide perspective may receive a more complete knowledge of what
happened. But let me now begin.
353
MAP 1: ANATOLIA DURING THE REVOLTS OF BARD AS SKLEROS
AND BARDAS PHOKAS (976-989)
Constantinople
t
*Chrysopolis
(, (m Abydos
Sangarios River
CHARSIANON
Basilika Thcrma
Tzamandos
v t Lykandos
Dipotamon/Mesanukta
* Rhagcai?
* Mayafanqm
KIBYRRHAIOTAI
U) Attaleia
Euphrates
River
MAP 2: RELIEF AND COMMUNICATIONS
ANATOLIA AND THE EASTERN FRONTIER IN THE TENTH AND ELEVENTH CENTURIES
Sinope
NO',0^N
Amastris ,^v A^ A A.
Sea of Marmara Amisos
K, A [\
A Gangra A A t*__
* Trcbizond'
t\
A Amascia > A A j\ A /^ /s '
Sangarios River
Ankara ^ /\ PONTUSMTS AS/N
Dorylaion Koloneia N Araxes River
Kotyaion ^ Dvin
SN ANA TOLIAN PLA TEA U Scbastcia fBasean
p /^ Amorion
AA A
//erwoi- River /* Usak * ticodosioupolis,
AkroTnos Tuz
^^
Golu. Halys River
, Tzamandos
Maeander River ^/ Apamcia Kaisarcia ' Lykandos Arsanias River
o, O% /O * Philomelion
phi Mus*
Chona, ^/\ JJnuto iJMZITENE Lake
^ Ml Argaios , Mclitenc" ^X. *Harput Van^
t Ikonion Bitlis Pass //
v V^ i1"l^
Tyana. N ^ A^ 1 |S It Ergani \ * Mayafariqin
Podandos ( ' Pass I. Amida
PAMPHYLIA ^S V^ Cilician Gates
Stroibilos Attalcia ^
rs f\
CILICIA Tigris
River
Korykos
t^ V-t-
U) ^ 'Sclcukcia r v*
k
<Antioch SYRIA
Euphrates River
Orontes
River
200 kms
J
MAP 3: ECONOMY AND TRADE
ANATOLIA AND THE EASTERN FRONTIER IN THE TENTH AND ELEVENTH CENTURIES
PAPHLAGONIA
Constantinople^
_. .. ..
Claudioupohs
BOUKELLARION ARMENIAKON
Abydos QPSIKION
... . Sykeon
, Malagma %
GALATIA
Pergamon Dorylaion
Kotyaion*
LYDIA
Amorion
__ Akroinos
Smyrna Sardis
Tzamandos Larissa
r- Synada
Ephesos r Stfbasteia
% Mastaura ^X ^Apameia f\ Pisidian Antioch ' ^ , , Abara*
Kaisarcia Gabadonia ,
Mcsanukta CAPPADOCIA Lykandos
Philomelion
Mclitcnc
Mayafariqin
Amida
Anazarbos
Tarsos,, Adana Mopsucstc
**.
Korykos
Sclcukeia
U>
L/t
Kaysen
(KaisareiaV^
4
Dasmendron Mount Argaios
(Erciyas Dag)
Sobessos
Aksaray *X
Derinkuyu
Develi Dag
U)
4.
Hasan Dag
L
Melendiz Daglan
^
Nigde (Magida)
MAP 5: THEMES AND KLEISOURAI OF ANATOLIA c.917
(Whitlow, Making of Orthodox Byzantium, p. 166, adapted)
S \
i
f" 'THRACE
MACEDONIA/ \ PAPHLAGONIA \ CHALDIA \
ARMENIAKON
sOPTIMATOI
KOLONEIA / / Thcodosioupolis
/ (Kalikata)
\ I
I /
\ X
Nikaia /-
SEBASTEIA ^ ' ~ "\
f/ \' BOUKELLARION
OPSIKION
^ ""^^ 11 f ' " \MESOPOTAMIA
f\ CHARSIANON x LEONTO i
/ - " ^ KOME \
Pisidia I x
^ ,~- ' ^ '
' ^ ~ \ '
fcAMOS Tarsos
\ O f KIBYRRHAIOTAI /
V3i \
SELEUKEIA
U)
(j\
oo
Con/ft \ , Ardanoutzin
Trcbizond River
**>-.
CHALDIA
KOLONEIA
ARMENIA
eodosioupolis
Apahunikh'
Chortzinc?
Tcphnke* Manzikcrt
TARON
Charpczikion
4 Mu VASPURAKAN
Halys River Abara Asmosaton
*+ HER
Kaisarcia Mclitenc
Hexakomia? Kaloudia? Mayafariqin
Amida
Gcrmanikciar
Adana
Tarsos
Laodikeiax*
Orontes
Balaniasl ( River
Tripoli/* \
BIBLIOGRAPHY
PRIMARY SOURCES
Greek:
Actes d'lviron I. Des origines an milieu du Xfe siecle, Archives de 1'Athos XIV, eds.
Choniates (Michael): Mi%aifi 'AKo/tivarou TOU XcawaTou TO, (rca&fAcva, ed. S.P.Lampros
360
DAI: De Administrando Imperio, ed. G.Moravcsik and trans. R. J.H. Jenkins (CFHB I
Digenes Akrites: Digenis Akritis: the Grottaferrata and Escorial Versions, ed.
Eparch: Das Eparchenbuch Leos des Weisen, ed. J.Koder (CFHB XXXIII, Vienna,
1991)
George the Monk: Georgius Monachus Chronicon, ed. C.de Boor (and P.Wirth) (2
George the Monk Continuatus: Theophanes Continuatus, ed. I.Bekker (CSHB, Bonn,
1838)
Glykas (Michael): Annales (Biblos Chronike), ed. I. Bekker (CSHB, Bonn, 1836)
1972)
Vienna, 1975-9)
Komnene (Anna): Anne Comnene, Alexiade, ed. B.Leib (3 vols., Paris, 1967)
Leo the Deacon: Leonis Diaconi Caloensis Historiae Libri Decem, ed. C.B.Hase
361
BcuriXeiov B', EEBS 10 (1932), pp.425-34
Listes: N.Oikonomides, Les Listes de preseance byzantines des IXe etXe siecles
(Paris, 1972)
Miracles ofSaint Eugenios: new edition of the material relating to Basil's reign in:
(Paris, 1981)
1960),pp.217-48
Ouranos Taktika (b)\ E.McGeer, Sowing the Dragon's Teeth: Byzantine Warfare in
362
trans. V.Laurent (CFHB XXIV, Paris, 1984)
Peira: J.Zepos and P.Zepos, lus Graecoromanum (8 vols., Athens, 1931), iv, 9-260
Praecepta Militum: E.McGeer, Sowing the Dragon's Teeth: Byzantine Warfare in the
Psellos (Michael), Scriptora Minora: Michaelis Pselli Scripta Minora, ed. E.Kurtz
Pseudo-Psellos: Michaeli Pselli Historia Syntomos, ed. and trans. W. Aerts (CFHB
Regesten: F.Dolger, Regesten der Kaiserurkunden des ostromischen Reiches von 565
1947) Vol. 1
Saint Luke the Stylite: 'Vie de saint Luc le Stylite (879-979)', ed. and trans.
Saint Nikon: The Life ofSt. Nikon, ed. and trans. D.F.Sullivan (Brookline, Mass.
1987)
Saint Phantinos: La vita di San Fantino il Giovane, ed. and trans. E.Follieri (Brussels,
1993)
Saint Philaretos: 'La vie de S.Philarete', ed. and trans. M.H. Fourmoy and M.Leroy, B
363
9(1934),pp.85-167
Saint Symeon the New Theologian: 'Vie de Symeon le Nouveau Theologien (949-
1960), pp.249-59
Straboromanos (Manuel): P.Gautier 'Le dossier d'un haut fonctionnaire d1Alexis ler
1059)', in Cinq etudes sur leXIe siocle byzantin (Paris, 1970), pp.40-62
Zonaras (John): loannis Zonarae Annales, ed. M.Pinder (CSHB, Bonn, 1841), Vol. 1
364
Zonaras (John): loannis Zonarae Epitomae Historiarum Libri XIII-XVUI, ed.
Aristakes Lastivert: Recit des malheurs de la nation armenienne, trans. M.Canard and
Bar Hebreus: The Chronography of Gregory Abu'l Faraj, the Son ofAaron, the
Life of John and Euthymios: B.Martin-Hisard, 'La Vie de Jean et Euthyme: le statut
Matthew of Edessa: Armenia and the Crusades in the Tenth to Twelfth centuries: the
York/London, 1993)
-483
Stephen of Taron: Des Stephanos von Taron armenische Geschichte, trans. H.Gelzer
365
Thomas Artsruni: History of the House oftheArtsnmik', trans. R.W.Thomson
(Detroit, 1985)
1965)
Hudud al Alam (The Regions of the World: a Persian Geography 372 A.H./982 A.D.),
andlbn Ridwan of Cairo, trans. and ed. J.Schlacht and M.Meyerhof (Cairo,
1937)
Beirut/Paris, 1964)
Ibn Miskawayh: Eclipse of the Abbasid Caliphate, ed. and trans. H. Amedroz and
Al-Mukaddasi, The Best Divisions for Knowledge of the Regions, trans. B. A.Collins
(Reading, 1994)
Al-Rudhrawari: Eclipse of the Abbasid Caliphate, ed. and trans. H. Amedroz and
366
Al-Shirazi: J.C.Burgel, Die Hofkorrespondenz Adud adDaulas (Wiesbaden, 1965)
1922)
Yahya ibn Sa'id al-Antaki, 'Histoire', ed. and trans. I.Kratchkovsky and A. Vasiliev,
Yahya ibn Sa'id al-Antaki, ed. L.Cheikho, CSCO Scriptores Arabici, Series 3, Vol. 7
Yahya al-Antaki Cronache dell 'Egitto Fatimide e dell 'Impero Bizantino 937-1033,
Russian Primary Chronicle: The Russian Primary Chronicle, Laurentian Text, ed. and
Adontz, N., 'Samuel rarmenien roi des Bulgares', reprinted in Etudes armeno-
367
occupations turques (1018-1307) particulierement au Xnie siecle', TM 1
(1965),pp.l-204
Ahrweiler, H., 'La frontiere et les frontieres de Byzance en Orient', Acts of the 14th
Angold, M., A Byzantine Government in Exile. Government and Society under the
Angold, M., (ed.), Byzantine Aristocracy (BAR International Series, Oxford, 1984)
Angold, M., The Byzantine Empire 1025-1204 (London and New York, 1984)
Angold, M., 'The Shaping of the Medieval Byzantine "City" ', Byz Forsch 10 (1985),
pp. 1-8.
Antoljak, S., 'Wer konnte eigentlich Joannes Skylitzes sein?', Acts of the 14th
Arbagi, M., 'The Celibacy of Basil IF, Byzantine Studies/Etudes byzantines 2(1)
(1975),pp.41-5
368
Patmos, edition diplomatique', TM8 (1981), pp.15-30
Baird, D., 'Konya Plain Survey', Anatolian Archaeological Reports 1 (1995), pp. 11-
Barzos, K., 'H yeveAoT/a rutv KO/AVTJVWV, Bv&vnva KeifJLeva KO.I ^leXerat 20
(Thessalonika, 1984)
Bass, G.F., 'A Medieval Islamic Merchant Venture,' Archaeological News 1 (1979),
pp. 84-94.
369
Geschichte im Spiegel von Geschichtsschreibung und Geschichtsdeutung
Belke, K. and Mersich, N., TIB Phrygien und Pisidien (Verlag der Osterreichischen
Bivar, A.D.H., 'Bardes Skleros, the Buwayids and the Marwanids at Hisn Ziyad in the
Brand, C.M., Icon and Minaret: Sources ofByzantine and Islamic Civilization
Browning, R., 'Byzantine Literature', DMA (New York, 1983), ii, pp.505-21
Bryer, A., 'The Estates of the Empire of Trebizond', Archaion Pontou 35 (1979),
pp.370-477
370
A.Bryer, 'The Structure of the Late Byzantine Town: Dioikismos and the Mesoi\ in
A.Bryer and H.Lowry (eds.), Continuity and Change in Late Byzantine and
Bryer, A. and Winfield, D., The Byzantine Monuments and Topography of the Pontus
Bury, J.B., The Imperial Administrative System in the Ninth Century (London, 1911)
Material and Spiritual Culture and History (trans. J.Jones Williams) (London,
1968)
Cameron, A., 'The History of the Image of Edessa: the Telling of a Story', Harvard
Cameron, A., (ed.), History as Text: the Writing ofAncient History (London, 1989)
Cameron, A., Procopius and the Sixth Century (London/New York, 1996)
Canard, M., 'Quelques noms de personnages byzantins dans une piece du poete Abu
Canard, M., 'Deux documents arabes sur Bardas Skleros', Studi Bizantini e
(1950),pp.99-108
371
KZayyat, "Vie du patriarche melkite d'Antioche, Christophore (967) par le
(1953),pp.561-69
(Paris, 1953)
Canard, M., 'La destruction de 1' Eglise de la Resurrection par le Calife Hakim et
Canard, M., 'Quelques "a cote" de lliistoire des relations entre Byzance et les
i, 98-119
Canard, M., 'Les sources arabes de ITiistoire byzantine aux confins des Xe et Xle
Canard, M., 'Les relations politiques et sociales entre Byzance et les Arabes', DOP 18
(1964), pp.35-56
224
pp.391-479
Cheynet, J.C., 'L'apport arabe a 1'aristocratie byzantine des Xe-XIe siecles', Byz Slav
372
61(1995),pp.l37-46
Cheynet, J.C. and Morrisson, C., 'Lieux de trouvaille et circulation des sceaux', in
Cheynet, J.C., Morrisson, C. and Seibt, W., Sceaux byzantins de la collection Henri
Cormack, R., 'Byzantine Cappadocia: the Archaic Group of Wall Paintings', Journal
Croke, B., 'City Chronicles of Late Antiquity', in Christian Chronicles and Byzantine
Crostini, B., 'The Emperor Basil H's Cultural Life', B 64 (1996), pp.53-80
Crow, J., 'A Review of the Physical Remains of the Frontier of Cappadocia', in
Crow, J., 'Alexios Komnenos and Kastamon: Castles and Settlement in Middle
Cutler, A., 'The Psalter of Basil IP, in Imagery and Ideology in Byzantine Art
373
tactiques de 1'empereur Leon VI', Comptes rendus des seances de VAcademic
Dagron, G. and Mihaescu, H., Le traite sur la guerilla (De velitatione) de I 'empereur
pp. 109-135
pp.55-60
Zl(1892),pp.488-525
Djobadze, W.Z., Materials for the Study of Georgian Monasteries in the Western
Djobadze, W.Z., Early Medieval Georgian Monasteries in Historic Tao, Klarjet 'i and
374
Ebersolt, J., Musees imperiaux ottomans. Catalogues des sceaux byzantins
(Paris, 1922)
Evans, H.C. and Wixom, W.D., The Glory of Byzantium: Art and Culture of the
von Falkenhausen, V., 'Bishops', in G.Cavallo (ed.), The Byzantines (Chicago, 1997),
pp. 173-194
Farag, W., 'The Truce of Safar A.H. 359', Paper from the Eleventh Spring
Birmingham, 1977)
Felix, W., Byzanz unddie islamische Welt imfruheren 11. Jahrhundert (Vienna,
1981)
Ferluga, J., 'John Scylitzes and Michael of Devol', ZRVI 10 (1967), pp. 163-70
Fine, J.V.A., 'Basil II and the Decline of the Theme System', Studio Slavico-
Firatll, N., 'Decouverte d'une eglise byzantine a Sebaste de Phrygie', Cah Arch 19
Foss, C., Ephesus after Antiquity: A Late Antique, Byzantine and Turkish City
(Cambridge, 1979)
375
Foss, C., 'The Defenses of Asia Minor against the Turks', Greek Orthodox
Foss, C., 'Sites and Strongholds of Northern Lydia', Anatol Stud37 (1987), pp.81-
101
Foss, C, 'Strobiles and Related Sites', Anatol Stud 38 (1988), pp. 147-74
Foss, C., 'Byzantine Malagina and the Lower Sangarius', Anatol Stud4\ (1991),
pp.161-183
Foss, C., 'The Cities of Pamphylia in the Byzantine Age', in Cities, Fortresses and
Franklin S. and Shepard, J., The Emergence ofRus 750-1200 (Cambridge, 1996)
Frei, P., 'Das Geschichtswerk des Theodores Daphnopates als Quelle der Synopsis
French, D., Roman Roads and Milestones ofAsia Minor: Fasc. 1 The Pilgrim's Road
French, D., 'Cappadocia and the Eastern Limes: Aspects of Romanisation at Amaseia
and Byzantine East (BAR International Series, Oxford, 1986), Vol. 1, pp.284-
376
Gautier, P., 'Monodie inedite de Michel Psellos sur le basileus Andronic Doucas',
#24(1966),pp.l53-164
Gautier, P., 'Le dossier d'un haut fonctionnaire d'Alexis ler Comnene, Manuel
Gay, J., L 'Italic meridionale et I 'empire byzantin depuis I 'avenement de Basil Ijusqu'
Gorecki, D., 'The Strateia of Constantine VII: the Legal Status, Administration and
Gregoire, H., Recueil des inscriptions grecques chretiennes d'Asie Mineure (Paris,
1922)
Gregoire, H. and Adontz, N., 'Nicephore au col roide', B 8 (1935), pp. 203-212
Guilland, R.G., Recherches sur les institutions byzantines (2 vols., Amsterdam, 1967)
(London/Variorum, 1976), no in
377
Haldon, J.F., Recruitment and Conscription in the Byzantine Army 550-950 (Vienna,
1979)
Haldon, J.F., 'Military Service, Military Lands and the Status of Soldiers: Current
Haldon, J.F. and Kennedy, H., 'The Arabo-Byzantine Frontier in the Eighth and Ninth
Centuries: Military and Society in the Borderlands', ZVRI 19 (1980), pp. 79-
116
Hamidullah, M., 'Nouveaux documents sur les rapports de PEurope avec 1'Orient
1989)
1985)
Hild, F. and Restle, M., TIB, Kappadokien (Verlag der Osterreichischen Akademie
Honigmann, E., Die Ostgrenze des byzantinischen Reiches von 363 bis 1071,
378
1935)
Howard-Johnston, J.D., 'Procopius, Roman Defences North of the Taurus and the
1989), pp.203-229
Howard-Johnston, J.D., 'Crown Lands and the Defence of Imperial Authority in the
Howard-Johnston, J.D., and Ryan, N., The Scholar and the Gypsy (London, 1992)
Hunger, H., Die hochsprachliche profane Literatitr der Byzantiner (2 vols., Munich,
1978)
Ivanisevic, V., 'Interpretations and Dating of the Folles of Basil II and Constantine
Janin, R., La geographic ecclesiastique de I 'empire byzantin, 1/3: Les eglises et les
Jeffreys, E., 'Malalas' Sources', in E.Jeffreys et al. (eds.), Studies in John Malalas
379
(Sydney, 1990), pp.208-13
Jeffreys, E., Scott, R., Croke, B.et al. (trans.), The Chronicle ofJohn Malalas
(Melbourne, 1986)
de Jerphanion, G., Une Nouvelle Province del'art byzantin. Les eglises rupestres de
Jones, A.H.M., The Late Roman Empire, 284-602 (3 vols., Oxford, 1964)
Jordanov, I., 'Les sceaux de deux chefs militaire byzantins trouves a Preslav: le
Jordanov, I., 'La strategic de Preslav aux Xe-XIe siecles selon les donnees de la
pp. 89-96
Kaplan, M., Les Hommes et la terre a Byzance du Vie auXIe siecle (Paris, 1992)
380
Kazdan, A.P., 'Iz istorii vizantiijskqj chronografii', Viz Vrem 20 (1961), pp. 106-128
Kazdan, A.P., 'Aristocracy and the Imperial Ideal', in M.Angold (ed.), Byzantine
Kazdan, A.P., 'The Formation of Byzantine Family Names in the Ninth and Tenth
Kazdan A.P. and Constable, G., People and Power in Byzantium: An Introduction to
Kazdan A.P. and Wharton Epstein, A., Change in Byzantine Culture in the Eleventh
Kennedy, H., The Prophet and the Age of the Caliphates (London, 1986)
Kordosis, M., 'The Road Network in Asia Minor and Armenia According to Edrisi.
381
Krivocheine, B., Symeon the New Theologian (949-1022) (New York, 1986)
Krumbacher, K., Geschichte der byzantinischen Literatur vonJustinien bis zum Ende
1897 edition)
Kuhn, H.J., Die byzantinische Armee im 10. und 11. Jahrhundert. Studien zur
Laurent, V., 'La chronologic des gouveneurs d'Antioche sous la seconde domination
pp.221-254
Laurent, V., Le Corpus des sceaux de I 'empire byzantin, V, L' Eglise 1-3
(Paris, 1963-72)
(Paris, 1981)
Memoires 54 (1960)
Lemerle, P., Cinq etudes sur leXIe siecle byzantin (Paris, 1970)
Lemerle, P., The Agrarian History ofByzantium from the Origins to the Twelfth
Liddell, H.G. and Scott R., (updated by RS. Jones and R.McKenzie), A Greek-English
382
(1995-7)
Lilie, R.J., Die byzantinishe Reaktion aufdie Ausbreitung der Araber: Studien zur
(Munich, 1976)
Ljubarskij, J., 'Some Notes on the Newly Discovered Historical Work by Psellos', in
McGeer, E., 'Tradition and Reality in the Taktika of Nikephoros Ouranos', DOP 45
(1991),pp.l29-40
McGeer, E., Sowing the Dragon's Teeth: Byzantine Warfare in the 10th Century
Essays in Honor ofG. T.Dennis (Washington D.C., D.C. 1995), pp. 123-37
McGrath, S., 'The Battles of Dorostolon (971): Rhetoric and Reality', in T.S.Miller
McGuckin, J., 'The Symeon the New Theologian and Byzantine Monasticism', in
383
Studies in Honour ofRobert Browning (London, 1996), pp.205-224
Macrides, R. and Magdalino, P., 'The Fourth Kingdom and the Rhetoric of
(1983),pp.326-46
Magdalino, P., 'The Byzantine Aristocratic Oikos', in M.Angold (ed.), The Byzantine
Magdalino, P., 'Honour among the Romaioi: the Framework of Social Values in the
Magdalino, P., 'Justice and Finance in the Byzantine State, Ninth to Twelfth
384
Magdalino, P., The Byzantine Background to the First Crusade (Toronto, 1996)
Maguire, H., Ikons of their Bodies: Saints and their Images in Byzantium (Princeton,
1996)
Mahe, J.P., 'Basile n et Byzance vus par Grigor Narekac'i', TM 11(1991), pp.555-72
di Maio, M., 'Smoke in the Wind: Zonaras' use of Philostorgius, Zosimos, John of
48 (1988), pp.230-255
Manandian, H. A., The Trade and Cities ofArmenia in Relation to Ancient World
JHS95(1975)
Mango, C. A., 'The Palace of the Boukeleon', Cah Arch 45 (1997), pp.41-50
Mango, C.A. and Sevcenko, I., (eds.), 'Byzantine Books and Bookmen', DOP 29
(1975)
Mango, C.A. and Scott, R., (trans.), The Chronicle ofTheophanes Confessor (Oxford,
1997)
385
Manoukian, A., (ed.), Documents ofArmenian ArchitectureDocument! di
pp.20-1
Miles, G.C., 'Byzantium and the Arabs: Relations in Crete and the Aegean Sea', DOP
18(1964),pp.3-32
Mitchell, S., Anatolia: Land, Men and Gods in Asia Minor (2 vols., Oxford, 1993)
Mitford, T., 'Military Road from Satala to Trabzon', Anatolian Archaeology Reports
2 (1996), pp.21-2.
Morris, R., 'The Poor and the Powerful in Tenth-Century Byzantium: Law and
Morris, R., 'Succession and Usurpation: Politics and Rhetoric in the late Tenth
386
Century', in P.Magdalino (ed.), New Constantines: the Rhythms ofImperial
Mullett, M., 'Writing in Early Medieval Byzantium', in R.McKitterick (ed.), The Uses
Mullet, M., 'Originality in the Byzantine Letter: the Case of Exile', in A.R.Littlewood
pp.39-58
Der Nersessian, S., 'Remarks on the Date of the Menologium and the Psalter written
Obolensky, D., 'Cherson and the Conversion of the Rus: an Anti-Revisionist View',
387
siecles et le taktikon de rEscorial', Acts of the 14th International Congress
Oikonomides, N., Les Listes de preseance byzantines des IXe etXe siecles (Paris,
1972)
1986)
1987-95)
Ostrogorsky, G., 'Une ambassade serbe aupres de I'empereur Basile IF, B 19 (1949),
pp. 187-94
Ostrogorsky, G., 'Agrarian Conditions in the Byzantine Empire in the Middle Ages',
388
in M.Postan (ed.), Cambridge Economic History of Europe Vol. I, The
Agrarian Life of the Middle Ages, 2nd edn. (Cambridge, 1966), pp.204-34.
Ostrogorsky, G., History of the Byzantine State (trans. J.Hussey), 3rd edn. (Oxford,
1968)
pp.3-32
Poppe, A., 'The Political Background to the Baptism of the Rus', DOP 30 (1976),
pp. 196-244
389
Restle, M, Die byzantinische Wandmalerei in Kleinasien (3 vols., Recklinghausen,
1967)
Robert, L., 4 Sur les lettres d'un metropolite de Phrygie au Xe siecle', Journal des
Rodley, L., 'The Pigeon House Church at Cavusin', JOB 33 (1983), pp.301-339
Rothman, M.and Kobze, G., 'Mus. in the Early Bronze Age', Anatol Stud47 (1997),
pp. 105-126
Roueche, C., 'Byzantine Writers and Readers: Story Telling in the Eleventh Century',
Antiochijskago: the Emperor Basil Slayer of the Bulgarians, Extracts from the
Salia, K., 'La litterature georgienne du 5e au 13e siecle', Bedi Kartlisa 17 (1964),
pp.28-61
(1982), pp.211-219
Saunders, W.B.R., 'Qalat Siman: a Frontier Fort of the Tenth and Eleventh centuries',
390
Schlumberger, G., L 'Epopee byzantine a la fin du dixieme siecle (2 vols., Paris, 1896-
1900)
pp.61-73
Scott, R., 'Malalas and his Contemporaries', in E.Jeffreys et al. (eds.), Studies in
Seibt, W., 'loannes Skylitzes - Zur Person des Chronisten', JOB 25 (1976), pp.81-6
Seibt, W., 'Die Eingliederung von Vaspurakan in das byzantinische Reich', Handes
Seibt, W., 'Die Darstellung der Theotokos auf byzantinischen Bleisiegel besonders im
pp.243-276
Sevcenko, I., 'Poems on the Deaths of Leo VI and Constantine VQ in the Madrid
Sevcenko, I., 'Levels of Style in Byzantine Prose', JOB 31.1 (1981), pp.289-312
Sevcenko, I., 'Some Additional Remarks to the Report on Levels of Style', JOB 32.1-
2(1982),pp.220-38
Sevcenko, L, 'The Madrid Manuscript of the Chronicle of Scylitzes in the Light of its
391
New Dating', in I.Hutter (ed.), Byzanz und der Westen (Vienna, 1984),
pp. 117-130
Shepard, J., 'John Mauropous, Leo Tornices and an Alleged Russian Army: the
Shepard, J., 'Scylitzes on Armenia in the 1040s and the Role of Catacalon
Shepard, J., 'Isaac Comnenus1 Coronation Day', Byz Slav 38 (1977), pp.22-30
Shepard, J., 'A Suspected Source of Scylitzes' Synopsis Historiarum: the Great
London, 1987-90)
Snipes, K., 'The 'Chronographia' of Michael Psellos and the Textual Tradition and
392
Provincial Aristocracy in the Late Tenth Century', REB 52 (1994), pp. 187-211
Svoronos, N., La Synopsis Major des Basiliques et ses appendices (Paris, 1964)
Svoronos, N., 'Societe et organisation interieure dans I'empire byzantin au Xle siecle:
concernant les puissants', Recueil des Travaux de I'lnst. d'Et. byz,, Melanges
(Warsaw, 1966)
Teall, J.T., 'The Grain Supply of the Byzantine Empire', DOP 13 (1959), pp.89-139
(1985), pp.363-4
Thierry, N., 'Notes critiques a propos des peintures rupestres de Cappadoce', REB 26
(1968),pp.337-71
(1981),pp.501-19
393
commanditaires de la nouvelle eglise de Tokali et d'autres monuments',
Thierry, N., Haut Moyen-age en Cappadoce: les eglises de Cavusin (Paris, 1994)
1981),pp.9-25
I.Thurn (ed.), loannis Skylitzae Synopsis Historiarum (CFHB V, Berlin and New
York, 1973)
Tourmanoff, C., 'Medieval Georgian Historical Literature', Traditio I (1943), pp. 139-
182
Tourmanoff, C., 'Armenia and Georgia', in J.Hussey (ed.), The Cambridge Medieval
History, Vol. IV Pt. 1, The Byzantine Empire: Byzantium and its Neighbours
Tourmanoff, C., 'The Bagratids of Iberia from the Eighth to the Eleventh Centuries',
Treadgold, W., Byzantium and its Army 284-1081 (Stanford, California, 1995)
394
Treadgold, W., A History of the Byzantine State and Society (Stanford, California,
1997)
Vryonis, S., The Decline ofMedieval Hellenism in Asia Minor and the Process of
Islamization from the Eleventh through the Fifteenth Century (Berkeley and
Vryonis, S., 'Byzantium: the Social Basis of Decline in the Eleventh Century', in
Byzantium: its Internal History and Relations with the Muslim World
Vryonis, S., 'The Panegyris of the Byzantine Saint', in S.Hackel (ed.), The Byzantine
Walker, P.E., 'A Byzantine Victory over the Fatimids at Alexandretta (971)', B 42
(1972), pp.431-440
P.E.Walker, 'The "Crusade" of John Tzimisces in the Light of New Arabic Evidence',
47(1977),pp.301-327
26(1968),pp.5-27
(Berlin, 1985)
395
Wright, G.R.H, 'Beycesultan 1954: Some Byzantine Remains', Anatol Stud41
Zacos, G., Byzantine Lead Seals II, compiled by John Nesbitt (Berne, 1985)
(The George) Zacos Collection ofByzantine Lead Seals, I (Auction 127, Spinks
(The George) Zacos Collection ofByzantine Lead Seals, 2 (Auction 132, Spinks
Zacos, G. and Veglery, A., Byzantine Lead Seals (3 vols., Basle, 1972)
Alfeyev, H., 'Saint Symeon the New Theologian and Orthodox Tradition' (Oxford
Bikhazi, R.J., 'The Hamdanid Dynasty of Mesopotamia and North Syria 254-
Farag, W.A., 'Byzantium and its Muslim Neighbours during the Reign of Basil II
Forsyth, J.H., 'The Chronicle of Yahya ibn Sa'id al-Antaki' (Univ. of Michigan Ph.D.
thesis, 1977)
Hicks, M., 'The Life and Historical Writings of Michael Attaleiates' (Oxford Univ.
396
Howard-Johnston, J.D., 'Studies in the Organisation of the Byzantine Army in the
Rapp, S.H., 'Imagining History at the Crossroads: Persia, Byzantium and the
1997)
Stephenson, P., 'The Byzantine Frontier in the Balkans in the Eleventh and Twelfth
Whittow, M., 'Social and Political Structures in the Maeander Region of Western
Asia Minor on the Eve of the Turkish Invasion' (Oxford Univ. D.Phil, thesis,
1987)
397