234567

Greeks and Türkmens: The Pontic Exception
Author(s): Anthony Bryer

Source: Dumbarton Oaks Papers , 1975, Vol. 29 (1975), pp. 113-148
Published by: Dumbarton Oaks, Trustees for Harvard University
Stable URL: https://www.jstor.org/stable/1291371
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Dumbarton Oaks, Trustees for Harvard University is collaborating with JSTOR to digitize,
preserve and extend access to Dumbarton Oaks Papers
This content downloaded from

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
All use subject to https://about.jstor.org/terms
GREEKS AND TfURKMENS:
THE PONTIC EXCEPTION
ANTHONY BRYER

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
This paper is substantially the same as that (entitled then
"The Case of Trebizond: an Exception") read at the Sym-
posium on "The Decline of Byzantine Civilization in Asia
Minor, Eleventh-Fifteenth Century," held at Dumbarton Oaks
in May 1974.
The term 'Tiirkmens' defies definition; by it I intend no

more than Turks who were primarily nomadic or pastoralist;
and by 'Turks' I mean Turks who were primarily settled. Two
appendices, extracts in translation from Panaretos and a
genealogy, are included to justify the argument. Where a
source is quoted, or a statement is made, without reference
in the text, documentation will be found in an appendix.
I am grateful to Mrs. Sue Payne for typing the paper and
to Dr. David Holton for checking the typescript in my absence.
For discussion of it, I am most grateful to Professor H61lne
Ahrweiler, Professor Claude Cahen, and, especially, to Pro-
fessor Speros Vryonis, Jr., Director of the Symposium.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
characteristic of recent studies of Anatolia in the period between
Mantzikert and the rise of the Ottomans is their emphasis on the
Tiirkmens as agents of change. In two formidable, independent, and
distinctly approached analyses, Professors Claude Cahen and Speros Vryonis,
Jr., agree in arguing that the Tiirkmen contribution was decisive in the final
collapse of Byzantine rule, culture, and economy in western and north-
western Anatolia, particularly after a turning point in the late thirteenth
century.' But the more this argument, which is, in general, palpably correct,
passes into wider currency, the more difficult it will be to account for major
exceptions which prove the rule among other local Christian peoples of the
Anatolian coastlands who were also under Tiirkmen pressure. Thus the
Cilician Armenians were able to put an end to both Rustem and the first
Karaman, but could not resist the Mamluks.2 So also, the Georgians were
consistently vigorous in stemming Tiirkmen infiltration on the Caucasian
marches and along the Akampsis (?oruh) river, but succumbed to the Mon-
gols.3 But the most notable exception was on the Pontic shore, where the
Trapezuntines held the Tiirkmens--and outlasted Seljuks, Mongols, and
Byzantines of Constantinople in the process.
An obvious explanation for the relative failure of the Byzantines of Ionian
and Bithynian Anatolia against the Tiirkmens, and the relative success of
the Armenians, Georgians, and Trapezuntines, is the lesson (learned since
the late eleventh century) that where there was the will to resist, backed by
local autonomy, leadership, or identity, and coupled with the restraints of a
centralized state, no Tiirkmens were in a position to surmount it. So the most
substantial Tiirkmen expansion in western and northwestern Anatolia followed
the coincidental transfer elsewhere (in different ways and to differing extents)
1 C. Cahen, Pre-Ottoman Turkey: A general survey of the material and spiritual culture and history,
c. 1071-1330 (London, 1968), 307-14: "... it will come as no surprise to find that the emanicipation
of the Turcomans was accompanied by a new expansion at the expense of the Byzantine Empire"
(after 1243). S. Vryonis, Jr., The Decline of Medieval Hellenism in Asia Minor and the Process of
Islamization from the Eleventh through the Fifteenth Century (Berkeley, Los Angeles, London, 1971),
194: "The invasions, settlements, and raids of the Turkmens played a crucial role in the fate of the
Anatolian peninsula. The impact of this nomadic-pastoral-warrior society, which was at the height
of its heroic age, upon Christians was one of the principal factors in the cultural transformation of
Asia Minor"; idem, "The Byzantine Legacy and Ottoman Forms," DOP, 23-24 (1969-70), 251-308.
P. Wittek, The Rise of the Ottoman Empire (London, 1963), 31; F. Stimer, Oguzlar (Tiirkmenler).
Tarihleri-boy teskildth-destanlarz, 2nd ed. (Ankara, 1972), esp. 327-35; M. Angold, A Byzantine
Government in Exile. Government and Society Under the Laskarids of Nicaea (1204-1261) (Oxford,
1974), 101: "The pressure of the Turkomans upon the Nicaean frontier seems to have become more
noticeable from the 1250s." H616ne Ahrweiler, "L'histoire et la g6ographie de la r6gion de Smyrne
entre les deux occupations turques (1081-1317) particulibrement au XIIIe siecle," TM, 1 (1965), 27.
2Cahen, Pre-Ottoman Turkey, 110-11, 281-82; idem, "Quelques textes n6glig6s concernant les
Turcomans de Rfim au moment de l'invasion mongole," Byzantion, 14 (1939), 133-34; idem, "Notes
sur l'histoire des Turcomans d'Asie Mineure au XIIIe siecle," JA, 239 (1951), 340-49; idem, " Questions
d'histoire de la province de Kastamonu au XIIIe siecle," Selpuklu Arastzrmalarz Dergisi, 3 (1971),
145-58.
3 Vryonis, Decline, 283-84; W. E. D. Allen, A History of the Georgian People from the Begin
down to the Russian Conquest in the Nineteenth Century (London, 1932), 95-108.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
116 ANTHONY BRYER
of two great centralized and local A

to the Mongols after 1243 (and, more
of Nicaea' to Constantinople after 126
of centralized government and the sp
but what is striking is not the evapor
Anatolia and its subsequent supplan
Tiirkmen origin, within half a centu
1261, but the survival (until 1390) o
autonomy and wished to keep it: Phil
By contrast, the Grand Komnenoi o
Cilician Armenian rulers, representat
The Pontos was not some sort of ov
any more than its 'Empire of Trebizo
that, like the Laskarids of Magnesia
tions to Constantinople, but the Sel
a practical end to that by 1214.5 A sm
new Pontic rulers were soon resigned
by equipping it, liturgically, as a lo
formal claim to be emperors of Ro
1282.7 But Pontic separatism had ca
that any western Anatolian identit
Palaiologoi; in modern terms, it migh
to Trebizond's Bangladesh.
Unlike almost any other part of A
for it was hardly touched by the lon
and (later) the Seljuks; whatever thirt
the Pontos had upon the Seljuks an
invasion or occupation, and the Gran
the last decades of their independen
Trapezuntines never experienced w
partly accounted for by, a historic P
Something of the intense localism o
in the very name of the thema of Ch
is almost certainly an Urartu word. T
* Ahrweiler, "Smyrne," 26-28, 137-38; D. M

Neo-Hellenika, 1 (1970), 9-17.
5 Nicetas Choniates, Historia, Bonn ed. (1835),
des Ibn Bzbz (Copenhagen, 1959), 64-68.
I The Panagia Chrysokephalos in Trebizond was
features, which would then have been required
ambo, and a gallery. The argument will be pu
Dumbarton Oaks Study, The Medieval Monumen
D. Winfield.
7 George Pachymeres, Bonn ed. (1835), vol. I, 520-24.
8 Haldi is the Urartu Sun God, who lived over the mountains: C. Burney and D. M. Lang, T
Peoples of the Hills. Ancient Ararat and the Caucasus (London, 1971), 159, 163-64. Constantine VII
naturally sought a (spurious) Chaldaean origin for the unfamiliar word: Costantino Porfirogenito,
De thematibus, ed. A. Pertusi, ST, 160 (Vatican City, 1952), 73.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
GREEKS AND TURKMENS 117
side of the mountains,' a sort of Perateia, a

a number of Trapezuntine villages are still ca
gave administrative expression to local iden
flanking Trebizond became banda, or milit
subculture. From west to east they were (pro
Dere), Trikomia (Akgaabat-Kalenima Dere),
Palaiomatzouka (Marka-Hamsikdy), Gemora
and Rhizaion (Rize). Of these, only the Phil
Tiirkmens entirely; the remaining six valleys r
inherited from the ninth century, in workin
valley became in turn an Ottoman kaza, a d
administrative district. It is not conservatis
that the medieval and modern boundaries must be the same.
The events of the eleventh and twelfth centuries revealed that the polit
boundaries of the Pontos had only been masked by common Roman a
Byzantine rule since the incorporation of the Sinopitan 'empire' into
Roman. The boundaries also serve for the 2,000 m. contour, the high r
mark, and the limits of ancient Greek colonization. Massive thema castles
along the Lykos and at Amaseia, and Kekaumenos' Koloneia went without a
fight, but the inner fortresses and passes and the seven valleys of Chaldia
resisted. They were opposing Constantinople just as much as they were the
Seljuks and Danigmendids. The Gabrades emerged as the first independent
rulers of Chaldia, precursors of the Grand Komnenoi, from the 1070's to the
1140's. It is true that St. Theodore Gabras, duke of Chaldia, died a martyr
to the Seljuks in 1098, but to Constantinople he was an incorrigible rebel,
while in later Turkish heroic poetry he is a clean-limbed hero. In the Melik-
danismendname Gabras is remembered as a combination of heroes (such as the
bishop Metropid), whose Amazonian daughter (Amazons are traditionally
endemic to the Pontos) turns Turk and, attractively disguised as a monk,
insinuates the Danigmendids into the great fortified monastery of St. Gregory
the Wonderworker at Neokaisareia (Niksar). But in all actual probability the
Saint's estates passed through normal family inheritance, within a century of
his martyrdom, to a Mangucak emir of Erzincan (whose dynasty, like the
Danigmendid, issued coins bearing Greek titles and Orthodox saints), and to
Hasan ibn Gavras, vizir to Kili Arslan II, who disputed them in 1192.10
That the Gabrades and Grand Komnenoi were local Anatolian rulers, while
the Palaiologoi were not after 1261, goes some way to explaining why local
resistance to the Tiirkmens was comparatively lively in the Pontos and failed
in the west. But there is more to explain, for if the Tiirkmens were the most
9 Among them X&hAr, one of the last nine Christian villages of Ophis (Of): I. Parcharides, ITrraTlrtKjl
-rilS TrapXias "OpEcoS -To0 vopo TparreloOv-ros, in Tlapvacr6S, 3 (1879), 228.
10 Irene M61likoff, La geste de Melik Danismend, I (Paris, 1960), passim; C. Cahen, "Une famille
byzantine au service des Seljouqides d'Asie Mineure," Polychronion. Festschrift Franz D61lger (Heidel-
berg, 1966), 145-49; A. Bryer, "A Byzantine Family: the Gabrades, c. 979-c. 1653," University of
Birmingham Historical Journal, 12 (1970), 164-87; and A. Bryer, S. Fassoulakis, and D. Nicol, "A
Byzantine Family: the Gabrades. An Additional Note," Byzantinoslavica, 36 (1975), 38-45.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
118 ANTHONY BRYER
insidious threat of all, the classic

local Christians over a very long pe
to the twentieth centuries) is in the
not to the Tiirkmens that the Trape
numerically dominant and retaine
on the face of it, geography and eco
situation in the west to beguiling si
pastoralists faced coastal Greek a
pastures which divided them for
concepts, which form the scenery
Dede Korkut (and, to a lesser extent
almost suspicious neatness to Prof
recent pastoral and agricultural eco
The permanent factors which dist
The extent of ancient, medieval, an
defined by geography, land use, an
ancient Greek trading stations and
west (where olive cultivation resu
hemmed into the Euxine by the Pon
the east. The culture of the coastlan
from Anatolia proper by the mou
Gates (fig. 3), in turn contributed t
The Pontic Gates were open from th
coinciding with the career of the T
a major outlet for Mongol Central
route was then important and the G
Constantinople) kept kommerkion r
overestimated. In fact, the most flo
until the Pontic Gates were reope
Trebizond became the principal por
Canal and rise of Samsun put an
prosperity.12
11 X. de Planhol, "Geographica Pontica," J
graphique du livre de Dede Korkut," JA, 254 (
Plantations c6tibres et vie montagnarde," Bull
(1963), 2-12; idem, "Aspects of Mountain Life i
Methodology by Example, eds. S. R. Eyre and
"Les G6orgiens et les Arm6niens dans la litt
(hereafter referred to as BK), 11-12 (1961), 27
Notes sur le 'Livre de Dede Korkut'," BK, 17
12 W. Heyd, Histoire du commerce du Levant a
Recherches sur le commerce gdnois dans la Mer
argue that the transit trade of Trebizond far o
The Society and Institutions of the Empire of
I, 84-140, argues the reverse. There is, howeve
Issawi, "The Tabriz-Trabzon Trade, 1830-1900:
of Middle East Studies, 1 (1970), 18-27, which d
transit trade," Royal Central Asian Journal, 31
in Pontic history, see A. Bryer, "The Tourkok
Conclusions," Neo-Hellenika, 1 (1970), 33-36.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
To the south of the Pontic Alps, the Armen

dry, and cold: ranch- and thin cornland.
screeching ox carts. In 1292 Edward I of E
found themselves hard put to lay hands on
Alps: no wine, no oil, no fruit, no green stuf
meat. But 55 percent of their budget went on
The Alps themselves, the intermediate z
400-500 km. of the finest grazing in Anatolia
tree line; they are snowbound for four m
shepherds descend to their permanent vill
winter pastures. These summer pastures were
the summer frontier between Tiirkmen and
embassy bought plenty of fresh meat and m
graziers do not have fixed ovens.
Peering north beyond the summer pastur
the rain clouds a 'sea of trees' (agap denizi)
formidable rainfall in Anatolia and, like sim
shielded by mountains in the Lebanon and
Caspian, long immune to settlement from
Korkut, Turali's father warns him against
through the dark valleys of the Pontos, whe
the trees, to Trebizond:
Son, in the place where you would go,

Twisted and tortuous will the roads be;
Swamps there will be, where the horseme
Forests there will be, where the red serpe
Fortresses there will be, that rub shoulde
Your destination is a frightful place. Tur
Between the summer pastures and Trebizond

castles of Mesochaldia (Kovans), Tzanicha (Can
(Torul), and Palaiomatzouka (Hamsik6y), an
only ones to be intimidated by the sudden ch
1405, Clavijo came down from Hempin (wher
"who profess to be Christians," but "all are
maina (Sfirmene) bandon: "The land is of the
the sea coast and is very mountainous, the hi
13 This and items cited later are compiled from account

Public Record Office, London, most accessible in C. Des
Persia nel 1292," Atti della Societd Ligure di Storia Patria,
14 Translation conflated from those in G. Lewis, The Bo
and F. Siimer, A. E. Uysal, and W. S. Walker, The Book
Texas, and London, 1972), 101. See also X. de Planhol, Les f
l'Islam (Paris, 1968), 223-24; E. Efendf, Narrative of Tr
Seventeenth Century, trans. J. von Hammer, II (London,

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
120 ANTHONY BRYER
by forests .... The paths we had t

us the lives of near all the beasts of
But there were compensations. Her
buy on most days of June 1292: br
butter, olive oil, vegetables, fruits,
beef, tongue, and fodder and offal
next year) for a caged leopard, whi
Tabriz to Dover (an embarrassing
a new lock had to be made for it
spent four times as much as they h
bread bill fell to only 5-10 percen
(the famous black wine of Trebizon
record 36 percent one heroic day.
offered exceptionally good living b
charters of the peasantry in the Ac
diet.1e Here they burn wood, not du
narrow paved 'imperial highways"
either fortified towers or wooden c
with the hay piled above. To the
"in trees or pyrgoi";18 Clavijo foun
live in hamlets each of which bears th
the same consisting of well-masone
place and elsewhere others."'9 The
of the Caucasus, but is exception
groups of hamlets and crofts are en
there is a very high ratio betwee
towns (there were no towns inland
more than administrative and com
in size on market days. Unlike the a
never boasted the Hellenistic mar
stadia, and other civic monuments
Sinope and of eastern Trebizond bea
Sinope, Kerasous (Giresun), Tripo
Trebizond, and Rhizaion (Rize) wer
thereby necessarily been places of
instinct was to take to the steep val
Erdebil invested Trebizond; the Gr
15 R. Gonzalez de Clavijo, Embajada a Tam
idem, Embassy to Tamerlane, trans. G. Le Stran
to recognize the importance of the dramatic ch
to Pontic history: Geschichte des Kaiserthums
1e Th. Ouspensky and V. B6n6ch6vitch, Acte
propridtd rurale et monastique d Byzance aux
N. Panayiotakis and A. Bryer are preparing a
Leningrad and Ankara cartularies.
17 Vazelon Acts, no. 100 of 1344. 18 S
19 Clavijo, locc. citt.; A. Bryer, "Some notes
pt. 2, BK, 23-24 (1967), 168.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
reduced to a population of fifty persons, for some 4,000 citiz
melted away-John berated them as "women, cowards and t
they knew what they were doing. When, in 1832, Osman pasa
captured the substantial and then rebel town of Stirmene, its
tion was brought back in triumph to Trebizond: it proved to
old women and one infant.21 The walled towns of the coast never offered
Tiirkmens the rich pickings they found in the cities of Ionia, and they ne
took one for long.
After the Black, or 'Sudden,' Death, I estimate that Tafur's figure o
4,000 for the population of Trebizond is not too low; but it was the cap
of a pocket empire of 200,000-250,000 souls.22 The strength of Pontic socie
therefore lay in the country, in the seven rural banda, their village communitie
(still paying allelengyon levies in the fifteenth century), their 'peacemakin
elders,' their local defense system, and the absence of very large estates
pronoia, or large numbers of paroikoi (a term unknown to the Acts of Vazelo
For once, the legendary Byzantine yeoman, or free peasant, might have be
alive and well.23 The extent of the Trapezuntine rural achievement in resist
Tiirkmens and Turks in the central, but nevertheless 'front-line,' bandon o
Matzouka (Maqka) may be demonstrated by the following population figure
which are approximate, but fairly reliable:24
Date Choria Christians Muslims Christian %

ca. 1520 57 12,080 1,665 88%
ca. 1920 70 16,525 5,335 76%
On the size of Tiirkmen population ther
Mureddin Hamza of Canik a standing arm
before 1348; Panaretos states that Taced
1386; and Clavijo that Altamur (? I) o
1404.25 By contrast, the largest recorde
20 Laonicus Chalkokondyles, Bonn ed. (1843), 462-6
21 Public Record Office, London, FO 524/1; A. B
of the Pontic Derebeys, 1812-1840," BK, 26 (1969), 2
22 Michael Panaretos, fEpi -r Tv PEy6d v Kopvvnvv
74, 80; G. Villani, Historie florentine, Muratori,
Travels and adventures, 1435-1439, trans. M. Let
yiizyil baslannda Trabzon livasi ve dogu Karaneniz
kokratia," 36-41.
23 A. Bryer, "Rural society in the empire of Trebizond," 'ApX.Tf6vr., 28 (1966), 152-60.
24 G6kbilgin, "Trabzon," 314-18; Chrysanthos (Philippides) Metropolitan of Trebizond, H 'EKKXrl)iat
TparrEloovroS, in 'ApX.Tf6v-r., 4-5 (1933) 791-96; G. Zerzelides, TorrowVlKK6 -rT15 "Avw MaTaro'KaGS, in 'ApX.
Tf6vr., 23 (1959), 91-93. Both sets of figures are in hearths or households; a multiplier of five has
been used. The Ottoman figures do not include those for the village of tlaksa, which belonged to the
Padisah. The ca. 1920 figures reveal that the Prytanis valley (off which Vazelon stands) was 88 percent
Greek, and the Panagia valley (in which Soumela stands) was 89 percent Greek. Modern Maqka is
reckoned to embrace 76 settlements.
25 Sihabeddin Abul Abbas Ahmed ben Yahya ibn Fadl Allah el Adawi, al Umari, trans. M. Quatre-
mbre, "Notice de l'ouvrage qui a pour titre: Mesalek Alabsar fi Memalek Alamsar (Masalik-al-absar).
Voyage des yeux dans les diff6rentes contr6es," Notices et extraits des manuscrits de la Bibliothdque
du Roi et autres bibliothdques, 13 (1838), 363-64; Panaretos, ed. Lampsides (supra, note 22), 76, 80;
Clavijo (supra, note 15), trans. Le Strange, 109, ed. Estrada, 73.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
122 ANTHONY BRYER
impress the Akkoyunlu emir in 1

horse. These relative figures may
raised by a Grand Komnenos may
subjects, an entire Tfirkmen peop
from an army. By the 1520's, Muslim
was as slight as that in Matzouka
Trapezuntines were probably overw
I have so far set the reasons for
the simple background of de Plan
zones, for the sake of convenience
the highlands to the Tiirkmens, an
uc. This rule probably holds better
but it is time to admit complications
for example, were all Pontic emi
Tiirkmen group in turn, and to attem
long symbiosis of Trapezuntine and
In the first decades after the eme
fall of the Danismendids to the Selj
26 Two misconceptions, upon which de Planh

immediately. (1) The coastal plantations of haz
but have a continuous medieval history. Alt
greatly developed in the last two decades, and
districts, there is abundant evidence that haz
empire of Trebizond. Nut trees are mentioned f
by the English in June 1292 (although then o
at the end of August and beginning of Septem
Pontos). On 17 September 1405, Clavijo caught
In 1418 Alexios IV arranged to pay a massiv
of that century Barbaro noted that "the wood
as those in Apulia." The first recorded export o
to Alexandria; the 135 nuts found at Novgoro
zouka (Maqka) was Karydia (Cevizhk); and inde
from the name Pontos (/fndzh). (2) De Planho
of abandoned Greek villages which lie on the
the downs north of Ordu (Kotyora) and elsew
final victory of Turkish pasture over former
medieval and older settlements of the valleys)
to the highlands from the late 17th century, w
the growth of the mountain villages of Santa
(Torul), which by credible tradition derived f
the coast. The former Greek villages in the ar
even more recent in origin, perhaps connecte
found no evidence of monuments before the 19
that de Planhol's original description of them
villages" in the English version of his argumen
3; idem, "Aspects" (supra, note 11), 299; Clavij
Notes et extraits pour servir d l'histoire des Croi
Novgorod the Great (London, 1967), 7-8; Iosafa
in India, in Constantinopoli (Venice, 1546), 4
New Jersey, 1950), II, 1073-74; Bryer, "To
D. Winfield, "Nineteenth-century Monuments
and Historical Notes," pt. 1, 'ApX.T6v-r., 28 (1
(1968), 108-9; A. Bryer and D. Winfield, idem,
Isaac, and D. Winfield, idem, pt. 4, 'ApX.Tl6vr.,
(Salonica, 1955), 23-24.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
consolidation and coexistence. When, on 1 November 1214, t
Sinope from the Trapezuntines, they rebuilt its walls, proclaim
in a great bilingual inscription on the acropolis: the 'Seljuk'
called Sebastos; the 'Seljuk' governor was called Hetum. When
tines retook Sinope for the last time in the 1260's, they instal
governor-who else? After all, when Kaykubad wanted an e
Gregory IX and Emperor Frederick II in 1234, he too chose a
Amisos must later have presented a curious spectacle of co
Genoese and the emirs of Aydin had faced each other in their r
at Smyrna,28 but at Amisos there stood up to four apparently
along five kilometers of seashore. Three were walled and on
derelict. The situation was only put to an end by the tidy-min
Reading from east to west, the towns were: the Genoese castle
Simisso; the Turkish settlement of Samsun (later given a c
Aminsos (which soon dropped out of the running-it is not clear
the Grand Logothete George Scholaris of Trebizond fled to
finally, the probably deserted acropolis of classical Amisos. Th
controlled the route south, but territorially it would be fruitl
map around any color at all.29
There are hints of a certain cultural interchange. After his a
on Trebizond in 1222-23, the melik Mugith al-Din Tugrilqa
(son of Kili9 Arslan) may have become some sort of vassal
Komnenos Andronikos I Gidon, and endowed at least one m
Trebizond.30 The 'stalactite' decoration and other reused or
Armenian, Byzantine, classical) elements in the columns, capita
of the Grand Komnenos Manuel I's church of the Hagia Sop
27 Duda, Ibn Bibi (supra, note 5), 64-68; A. Papadop(o)ulos-Kerameus, Fontes

Trapezuntini (hereafter, FHIT), I (all published) (St. Petersburg, 1897; Amsterdam
131; E. Blochet, "Note sur quatre inscriptions arabes de l'Asie Mineure et sur
du sultan mamlouk Kaitbay," Revue sdmitique d'dpigraphie et d'histoire ancienne
but N. Kuruo'lu, "Sinop," in N. Sevgen, Anadolu Kalelert, I (Ankara, 1959), 280-91
tect as Abu Ali ibn Abir-Rakka el Kettani of Aleppo who designed the Alany
N. A. Bees, Die Inschriftenaulzeichnung des Kodex Sinaiticus Graecus 508 (97
Spildotissa-Klosterkirche bei Sille (Lykaonien), mit Exhursen zur Geschichte der Seld
TFByzNgPhil, 1 (Berlin, 1922), 53-54 (for Greek text); Cahen, Pre-Ottoman Turke
Decline, 197, 236; C. Cahen, "Le commerce anatolien au d6but du XIIIe sidcle,
Halphen (Paris, 1951), 91-101; idem, "Quelques textes" (supra, note 2), 138. The
Arabic-Greek inscription known to me also comes from Sinope; dated 1641, it is
later Karamanh inscriptions are notably absent; Maria G. Nystazopoulou, 'H iv T Ta
-r6ots lovySafa &dtw ToO-r I' dXpt T-ro IE' atwCvos (Athens, 1965), 120, entry 17;
181, no. 15; G. Golubovich, Biblioteca bio-bibliografica della Terra Santa e dell' Or
(Florence, 1906-27), II, 298-99; Bryer, "Gabrades," 181 (No. 12).
28 Ahrweiler "Smyrne," 41.
29 Panaretos, ed. Lampsides, 75; the sites will be described in the study by Bry
(supra, note 6). J. Schiltberger describes the situation in 1402, after the Greek tow
"Samson consists of two cities opposite each other, and their walls are distant, on
an arrow's flight. In one of these cities there are Christians, and at that time th
possessed it. In the other are Infidels to whom the country belongs." Between th
berger placed the tale of the battle between the sea-serpents and land-vipers, per
here of the maritime Italians and earthbound Turks: J. B. Telfer, The bondage and
Schiltberger, a native of Bavaria, in Europe, Asia and Africa, 1396-1427 (London,
3o Cahen, Pre-Ottoman Turkey, 125; Papadopoulos-Kerameus, FHIT, I, 131.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
124 ANTHONY BRYER
agreeable cultural confusion.3s It m

that the wife of an emir of Sivas
seek comfort at the tomb of St. At
in 867-86) and in his old monaste
recorded miracle.32
31 Illustrated in The Church of Haghia Sophia at Trebizond, ed. D. Talbot Rice (Edinburgh, 1968),
pls. 9A, 9B, 10A, 10B, 21E, 21F, 22D, 23A-D, figs. 12-53; see also J. M. Rogers, "Recent Work on
Seljuk Anatolia," Kunst des Orients, 6 (1970), 153-61, 165. Sculpturally and architecturally the inter-
change was overwhelmingly between Armenians and Seljuks, not Greeks. Tamara Talbot Rice's
proposition that the Hagia Sophia work is that of Seljuk refugees from the battle of K6se Da' in
1243 is hardly convincing: "Decorations in the Seljukid Style in the Church of St. Sophia at Trebi-
zond," Geddchtnisschrift Diez (Istanbul, 1963), 87-119.
32 The evidence is potentially important, yet of the kind that falls to pieces in the hands, but
there are attestations that it may have some reliability, which will have to be argued out. (1) The
Synaxarion of St. Athanasios is palpably very late, but he is known from earlier Greek sources.
That his tomb was venerated in Trebizond, in 1318 at least, may be confirmed from an unexpected
quarter: Odoric de Pordenone, O.F.M., was shown "the body of Athanasius upon the gate of the
city" then. (Odoric is rightly used with caution, but his opening paragraphs about the Pontic shore-
where he may have picked up his Central Asian Tales--are reliable enough, and even his tale of
4,000 tame 'partridges' at Zigana is credible to anyone who has witnessed the migration of vast flocks
of exhausted quail which blacken the sky there in September.) Later, John Mandeville embroidered
Odoric by reporting that in Trebizond "lieth Saint Athanasius that was bishop of Alexandria, that
made the psalm QVIQVNQVE VVLT." St. Athanasios the Great had (of course) nothing to do with
Trebizond; the only other sainted Athanasios native of the city, the Athonite, was buried in his
Great Lavra. Odoric (and Mandeville) must therefore intend the Exorcist. His Synaxarion states
that his body was translated back to the monastery of St. Phokas, presumably after the emir's wife
and Odoric venerated it, and perhaps after 1461, when the relics of a Greek metropolitan above an
Ottoman city gate would have been inappropriate: there is, in fact, what appears to be a chapel
or shrine above the Gate of St. George of the Limnians to the northeast of the citadel. (2) The
monastery of St. Phokas, perhaps ToO A&itrhov, and not to be confused with that at Kordyle, may
be referred to in 980. It would have had to have been within commuting distance of Trebizond,
where we are told that St. Athanasios took the weekend cathedral services. Ioannides, a 19th-
century Trapezuntine schoolmaster and antiquary, placed its ruins in the Hotz-Kymena district,
south of modern Trabzon airport and the University of the Black Sea. He gives no explanation for his
identification, but his instructions led me in 1973 to a partly rock-cut church (painted, perhaps, in
the 14th century) near Hosmesalos; it will be published in Bryer's and Winfield's study. The Synaxa-
rion gives the monastery an iron-rich estate at TLapwroipou; there is a Cam Burunu east of Sourmaina
(Siirmene), but the Cam Burunu on Cape Jason, in the Chalybian iron lands, seems more likely.
(3) The Synaxarion does not state in which Grand Komnenos Manuel's reign the emir's wife came.
Manuel II reigned very briefly in 1332; in Manuel III's reign (1390-1417), Sivas was wrecked by the
Mongols. I have therefore followed other commentators in taking it to be the reign of the more famous
Manuel I (1238-63). (4) If this is so, and if the other indications cited add sufficient weight for the
Synaxarion to be taken seriously, the tale of the emir's wife fits well enough in a cultural milieu of
late 13th-century Sivas, which was even more hybrid than that of Manuel's Trebizond. Among three
splendid medreses erected there in 1271/72, the G6k was designed by the Greek Kaluyan al-Qunawi,
and the Buriiciye and Gift Minare show Armenianizing influences. A Franciscan house was established
there about eight years later. The emir's wife need not have been Muslim Turk. One candidate would
be the Caucasian Khoshak Mkhargrdzeli, who probably married Sams al-Din Ciivayni (patron of
the Qifte Minare in Sivas) in the last years of Manuel's reign. A. Papadopoulos-Kerameus, Xu pPoXai
EIs Tilv taropiav TpawrrEoiv-ros, Viz Vrem, 12 (1906), 138-41; F. Halkin, Auctarium Bibliothecae Hagio-
graphicae Graecae, SubsHag, 47 (Brussels, 1969), 194, no. 2047t; A. W. Pollard, The Travels of Sir
John Mandeville, with three narratives in illustration of it .... (London, 1900; New York, 1964), 98,
326; Papadopoulos-Kerameus, FHIT, I, 53, 47, 58, 120; K. Lake, The early days of monasticism on
Mount Athos (Oxford, 1909), 103; S. Ioannides, 'I-ropia K Xi T-rTrl-rlK1 TpaSrrEloiv-ros Kil Tf l wEpi T-raj-
Trnv Xcbpas (Constantinople, 1890), 239; Chrysanthos, 'EKKmAlaia (supra, note 24), 217-21; Bryer, Isaac,
and Winfield, "Nineteenth-century Monuments," pt. 3 (supra, note 26), 294; Golubovich, Biblioteca
(supra, note 27), I, 301; A. Bryer, "Trebizond and Rome," 'ApX.nT6v-r., 26 (1964), 294-95, 300 note 2;
Vryonis, Decline, 486; J. M. Rogers, Patronage in Seljuk Anatolia, 1200-1300 (Oxford, 1971) (un-
published Dr. Phil. thesis); 0. Turan, "Les souverains seljoukides et leurs sujets non musulmans,"
Studia Islamica, 1 (1953), 80-83. I am grateful to Miss Barbara Brend and Father Jean Darrouzvs for
discussion and information.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
Things began to change not so much aft

Seljuks and their Anatolian allies went down
(when the Mongols' Anatolian tributaries fai
Komnenos of Trebizond and the pervane (
Samtzkhe as well) both fell to the vengeance
perhaps eaten;33 in the very year of his d
active in his old lands in Samsun and Sinop
Pontic coastal strip was more or less isolat
the doings of greater powers in Anatolia by a
lordships and non-territorial groupings, with
tion. To the east lay principalities on the C
client Gurieli of Guria; then came a sort of L
clerks of Trebizond solemnly entitled a them
and ranches of Greco-Laz dynasties along
Mesochaldia of the Kabazitai,36 or Tzanicha
virtually independent, given great titles at co
in his cheek, duke Leon II Kabazites prote
lived in this barren land ... and had contin
the Turks who were his neighbours on all
and his men had nothing to live on, except i
them by those who passed through their cou
to by plundering the lands of their neighb
caravan cities of Bayburt and Erzincan, whic
Christian majority, a Muslim (not Tiirkmen)
with Trebizond. To the south-southwest l
district, which the Grand Komnenoi were nor
(Siran-Ulugiran). Troubled by Tiirkmens,
organized self-defense under their bishop,
33 Hayton, Flos historiarum terre Orientis, in Recueil des

niens, II (Paris, 1906), 309; A. A. M. Bryer, "The fate o
(1266-1280)," BZ, 66 (1973), 332-50 (where "eaten" shou
346 note 68). On human sacrifice, see Vryonis, Decline, 27
34 Siimer, Oguzlar (supra, note 1), 327.
35 Panaretos, ed. Lampsides, 77; D. Bakradze, Arheologile
Petersburg, 1878), 11-12, 286-335, 337; A. Bryer, "Lud
Anatolian Embassy of 1460-1461," BK, 19-20 (1965), 183
beside a painted tomb figure with scepter, of ca. 1400, in
as: ... 'ApX]yv Fovpia[s: D. Talbot Rice, The Church of H
chrysobulles in6dits des empereurs de Tr6bizonde Alexis I
(1953), 265.
36 John I Kabazites was Grand Logiaristes in 1344-45, Gr
in 1355-56. Leon I Kabazites was Grand Domestic in 134
Kabazites was Duke of Chaldia in 1404. A Kabazites was Pansebastos and Duke of Chaldia in
1461.
37 A Tzanichites was Grand Stratopedarch in 1340. Stephen Tzanichites was Grand Constable
in 1344. John Tzanichites was Epikernes in 1352. Constantine I Tzanichites was Grand Constable in
1386 and Constantine II Tzanichites was Grand Constable in 1415. Perhaps the last office controlled
Tzan (i.e., Laz) irregulars.
38 Clavijo, ed. Estrada, 80; trans. Le Strange, 117-18 (" Quilileo Cauasica," "Qurileo Arbosita" =
"kyr Leon Kabazites," not "Cyril Cabasica").

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
126 ANTHONY BRYER
Fatih in 1461.39 To the southwe

small industrial Greek Christian g
in Trapezuntine hands, but the min
West through Trapezuntine Kerasou
After the bervane's gruesome deat
Tiirkmen groups in west Pontos, ca
are known almost entirely from
appearance on the coast is that the
inconsistent here. The mountains ar
and, in west Pontos, there are two
like passes. They are where the H
can be navigated inland and (as Stra
delta pastures in the winter.41 And
Chalybia and around Omidia, just
fall so abruptly into the sea that
infiltration from the interior. Aft
the Seljuks made great efforts to r
of the twelfth century, matching t
Tiirkmen groups eventually sought
established themselves at the mou
Kerasous. This apparently left the e
enclaves, up to five in number, s
Viewed territorially, this way, it is a
for the old colonial 'empire' of Sinop
disconnected patches, including Tr
as Sinope had a hinterland, it was
mountainous and inhospitable Paph
state. The coastal road, so confiden
reality largely a sea route.44 It is on
has appeared widely in the Pontos
able to use a complete coastal road
ever be possible to travel the Pap
meant little. Of the thirty expediti
Komnenos Alexios III (1349-90), tw
land journeys were performed in th
39 G. Th. Kandilaptes, FEcoypa<PtK6v Kai I(Tr

2 (21-22) (May-June, 1946), 505; Bryer, Isaa
pt. 4, 228.
40 F. Balducci Pegolotti, La pratica della mercatura, ed. A. Evans (Cambridge, Mass., 1936; New
York, 1970), 369; Bryer, Isaac, and Winfield, "Nineteenth-century monuments," pt. 4, 243-52.
41 Strabo, Geographia, I. III. 7; XII. III. 12-15; Muhammad ben Muhammad al-Idrisi, trans. P. A. E.
Jaubert, Gdographie d'Edrisi, Recueil de voyages et de m6moires, publi6 par la Soci6t6 de G6ographie,
6 (Paris, 1840), 393-94; B. Nedkov, La Bulgarie et les terres avoisinantes au XIIe sidcle selon la 'Gio-
graphie d'Al Idrissi' (Sofia, 1960), 96-99.
42 Map in Magie, Roman Rule in Asia Minor (supra, note 26), II, opposite p. 1616.
4* W. Leaf, "The commerce of Sinope," JHS, 26 (1916), 1-10.
44 K. Miller, Itineraria Romana (Stuttgart, 1916), cols. 639-50.
4" A. Bryer, "Shipping in the Empire of Trebizond," The Mariner's Mirror, 52 (1966), 4.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
Thus, the west Pontic-Canik Tiirkmen cent

Halys delta, the Taceddinogullari of the Ir
Hamza, and the Emirogullarl of Chalybia, whic
and Turkish sources, are shown (credibly eno
zuntine field in a recent historical atlas-excep
two of which did not then exist.46
But in west Pontos, at least, matters are more complicated. Here Greek
settlement clung to the coast alone, which itself, exceptionally, became the
uc, for it simultaneously offered winter grazing. In central Pontos, where
Trapezuntine settlement penetrated inland, the situation was reversed, for the
uc lay along the summer pastures. The west Pontic experience is unconven-
tional, for it demonstrated how two states (or rather the empire of Trebizond
and a nest of emirates) could coexist on the same territory, and how, in the
absence of any major Tiirkmen leader or confederacy, the Grand Komnenos
himself could assume a double role. At times, it is no exaggeration to say
that he was simultaneously a Byzantine emperor and a Tiirkmen melik of a
group of small emirates which he had a hand in creating. It is difficult to
suggest how this situation can be depicted in a historical atlas, for Trapezun-
tines and Tiirkmens were doing different things on the same land, which they
could equally claim. An attempt to illustrate this graphically is made in the
map (infra).
West Pontos is called Canik, a puzzling name which is not geographical.
Elsewhere I have attempted to show that the word is derived from the horse-
men Tzannoi, who had vanished before the fourteenth century. But the Laz
call themselves 'Chani,' and the peoples of the coast, Greek and 'Chani,' were,
and are, popularly called 'Lazoi.' Hence Canik means the Pontos, or, more
particularly, west Pontos in which the Tiirkmen emirates emerged. Like the
term 'Rum,' it indicated a Christian land which had (here inconclusively)
passed into Muslim hands.47 It had a secondary, economic, meaning, for
Canik seems also to have served as a synonym for kzsla, winter pasture.48
For some purposes Canik was still part of the empire of Trebizond. In the
capital his subjects described their ruler as Grand Komnenos and Turks and
Armenians would call him tekfur.49 But in west Pontos the Grand Komnenos
became, in ibn Saddad's term, melik of Canik.5o In this respect, and in this
area alone, al Umari is right in concluding his important description of the
empire of Trebizond with the observation that it resembled the Turkish
46 D. E. Pitcher, An Historical Geography of the Ottoman Empire, from earliest times to the end of
the sixteenth century (Leiden, 1972), map viii. Terme and Ordu are 19th-century revivals.
4 Bryer, "Laz and Tzan," pt. 1, 174, 195; pt. 2, 163-68; C. Cahen, "Ibn Sai'd sur l'Asie Mineure
Seldjuqide," Ankara Universitesi FakWiltesi Tarih Arastirmalari Dergisi, 6 (1968), 49: "les montagnes
de DJANIK ... sont habit6es par un peuple indocile et rude, pratiquant la religion chr6tienne."
48 De Planhol, Fondements, 224, 236.
49 J. P. Fallmerayer, Original-Fragmente, Chroniken, Inschriften und anderes Materiale zur Geschichte
des Kaiserthums Trapezunt, pt. 2, AbhMiin, Hist.Kl., 4 (1) (1844), 106-7: Trapezuntine Armenian
inscription of 1413/14 (incorrectly computed to 1415 by Fallmerayer), referring to the emperors
of Trebizond as kyr and 'takhavor.'
S0 Cahen, "Quelques textes" (supra, note 2), 137.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
128 ANTHONY BRYER
principalities which surrounded it

must be illustrated in detail.
Sinope may be discounted. Its fourteenth-century history is baffling an
despite his fame (as the first recorded Turkish frogman), we do not kn
who its Gazi felebi was52-except that he was not the husband of Eudok
despoina of Sinope and daughter of Alexios II. But Sinope was a cors
emirate. The early Pontic Tiirkmen groups were, in origin, pastoral. Th
emirates of Trebizond in its guise as Canik are most conveniently tak
geographically, from west to east, in more or less antichronological order.
The most westerly is Limnia (later probably centered on 9argamba), in th
alluvial delta of the Iris. At its mouth there had always been a skala and sup
base-the classical Ancona and, I suggest, also the twelfth-century Kin
where John Komnenos grazed his cavalry horses in the winter solstice
1140, while campaigning against the Danigmendids. The skala finally becam
the Trapezuntine stronghold of Limnia, with a see and thirteen imper
fortresses; it figures on portulan maps until the sixteenth century.53 Lim
51 Al Umari, trans. Quatrembre (supra, note 25), 379-80: "Le royaume de Tr6bizonde, qui apparti
aux adorateurs de la croix, est un empire consid6rable, situ6 le long du rivage de la mer, sur un go
demi-circulaire, form6 par les eaux du Pont. Il se prolonge, d'occident en orient, sur la fronti6re
territoire occup6 par des Turcs, sur cette c6te que nous avons d6crite. Plac6 au midi du pays de Rou
il a ses deux extr6mit6s 6troites, tandis que son milieu pr6sente une largeur consid6rable. Il semble
que, des deux c6t6s, on a promen6 un compas en dehors du centre de la circonf6rence, de manier
imiter la figure d'un fruit de mirobolan. C'est sous cette forme que Belban le G6nois m'a trac6 la
carte de ce royaume. Suivant son r6cit, le souverain de Tr6bizonde est un prince grec descendu d
Constantin, fondateur de la ville de Constantinople. Ce monarque s'assied sur un tr6ne, port
couronne, a des revenus dignes d'un roi, une cour nombreuse, et jouit, aupres du pape, d'une hau
consid6ration. Lui et tous ses sujets se distinguent par une extreme beaut6; seulement, le pr
aujourd'hui r6gnant a, comme son pbre, sur les reins une excroissance mince, allong6e, pro6mine
qui a la longueur et la largeur d'un pouce, et qui ressemble a une sorte de queue. On voit, dans le m
pays, plusieurs personnes dont la conformation pr6sente cette particularit6. Lorsque Belban me f
ce r6cit, j'h6sitais ' le transcrire, attendu qu'il me paraissait peu vraisemblable; mais il me fut confir
par Behadur-Abouami, et ensuite, successivement, par trois hommes vou6s ' la vie religieuse. Alor
je n'ai pas h6sit6 a consigner le fait dans mon ouvrage, en en laissant la responsabilit6 a ceux don
je le tiens. D'ailleurs, la puissance de Dieu est infinie; il fait tout ce qu'il lui plait, et cr6e, quand i
veut, des objets inconnus aux hommes.
Suivant le t6moignage du meme Belban, les habitants du royaume de Tr6bizonde sont des homm
belliqueux et hardis. Leur pays est continuellement travers6 par des voyageurs qui se dirigent ve
la province de Krim, le d6sert de Kabdjak et les autres contr6es du nord. L'empire de Tr6bizonde
est plus vaste que la royaume des Kurdjes (la G6orgie), et plus important aux yeux des monarques
chr6tiens, attendu que la population est plus forte et plus brave. Le roi de Tr6bizonde porte, com
celui d'Armenie, le titre de Takafour. Il est d'une naissance plus illustre que l'empereur de G
actuellement r6gnant, et il s'attribue sur ce monarque une grande sup6riorit6. Les soldats qui com
posent son arm6e, quoique peu nombreux et mal 6quip6s, sont autant des h6ros, autant de lions r
doutables qui ne laissent jamais 6chapper leur proie. Du reste, ce pays ressemble, sous tous les rappo
aux principaut6s turques qui l'avoisinent. Tel est le r6cit de Belban le G6nois." This Egyptian compil
tion was made in the period 1342-48; Belban was a former Genoese slave: F. Taeschner, Al-Umaris
Bericht i4ber Anatolien in seinem Werhe (Leipzig, 1929). I have no comment on the tailed Grand
Komnenoi (a tradition otherwise unattested), but see H. Yule, The Book of Ser Marco Polo,
(London, 1871), 244; and S. Thompson, Motif-Index of Folk-Literature, III (Copenhagen, 195
138, no. F518.
52 H. A. R. Gibb, The Travels of Ibn Battfita, A.D. 1325-1354, II (Cambridge, 1962), 465-6
Cahen, Pre-Ottoman Turkey, 312, 321.
65 Papadopoulos-Kerameus, FHIT, 61; Arrian, 22; Miller, Itineraria Romana, col. 646; Nice
Acominatus (Choniates), Bonn ed. (1835), 45-46; Theodore Prodromos in PG, 133, cols. 1340-
Idrisi (supra, note 41), ed. Jaubert, 394, trans. Nedkov, 98-99; F. Chalandon, Les Comndne,
Jean II Comndne (1118-1143) et Manuel I Comndne (1143-1180) (Paris, 1912; New York, 1971

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
?1?14
4e~ ~ ~
lr' lt
2. Tzanicha (Canc
"Fortresses there will be, that rub shoulders wit
rr6KacTrpov of the akritic Greco-Laz Tzanichites
1. The Pontic Rain Forest. two late medieval painted churches and overlook
route and the silver mines of Thia (later Argyrop
"Forests there will be, where the red serpent can find no path ..." Dede
shane is in the middle distance and the Pontic Al
Korkut. Coniferous forest, receiving about 2,500 mm. of rain annually,
from the monastery of Sonmela at about 1,600 m., looking south to the
tree line at about 2,000 m., above which is the Parcharin of Larachanes
(Larhan yayla), leading to the Pontic Gates.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
3. The Summer Pastures at the Pontic Gates.
"Bylae": the principal summer pass on the Trebizond-Tabriz route, standing abov
Parcharin of Stavri (Istavri yayla), at about 2470 m., looking north.
4. A Winter Emirate: Limnia.
"We went down into Rum, we wintered, we wrought much good and evil; Spring came, w
back again, elhamdiililldh." Yunus Emre. Aerial view of the Iris (Yesil Irmak) River delta,
north from above (arsamba to the Black Sea, about 20 km. away. Taslikki*y (? Limni
the coast to the right (east), near the medieval mouth of the Iris, which now debouches to
(west).

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
36* """"""""
Trebiz nd 1204 38*** * ***3***** ****U
> * Seljuk 1214
*' nhope lKastam
Trebizo d c.1254
nu
anda
c. 1267 Suer ta
Ottoma~n
EPNi 1277 (toinop 16
1322) Otto Trebizorr
n 1461 eot
Trebizo d Gemo
Trikomi S a
5a /1 AQ o Matzouk ?Rhizain
Ainsosr Palaioma zouka G
S (t Canik 1330) raSmsn (c.1197)
imisso Athena
1 1O
to (tkC
A emi
1393)
o) Iolar emiritLmn
naion Om
tr aleodlu Kal - St. Bai I SPharos
E/P.E I 1277 A/ . r-1/Phadisan /L TrebizodO
Emirla / -Polemor Ker asous Ko alla St.ma
2 ) CAP J.O Kenc ina Soumaina
7ro Io 0) 12 Ophis
HAr6u)3 130P ZAS
o Havza 7mas 1
Taoanogula
OPrireoe
emirate B B
NromamKas P
(1313 a
n
( *
rd) Akkoguniu E)' lon

1301 X 'I Von i 3711~~
'Koloneia 0~ rl~i~
13 W fled Town %~iiii
M 'or MonasterjiF~riii :::1;
oM or MC aster* * H2 3
A XM
M ne,
*or ravan, route from
B attle c aTrobiizoimd ..i ::
n ... .......
C ravaa route f m TrebizondA
Kf~y o :,...... ~ .... .........~
5 M jor Battle i4sa

dppe r epr, a Ola) S -
The Empire o

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
should be at the ruined village of Taylikk

of it, Kinte, or Ancona among the fourt
for a combination of malaria and the wand
in turn. However, the most important aspe
and its great green river, which climbs int
Although Limnia was the last and lowlie
its bishops assumed the metropolitan right
The Grand Komnenoi visited the place re
district (perhaps a bandon) with an imp
place to incarcerate important rebels. But
imperial visits take on a certain consistenc
- from 22 September to late December 1
- from 19 December to soon after 25 De
- from soon after 6 January to early Mar
- from 6 December 1360 to about 20 Marc
- from the end of January to the end of
- and in October 1379.
During this visit Alexios III "took control of Limnia" and marr
daughter to a certain Taceddin pelebi, who had first sought a ma
alliance in June 1362. In 1384 comes a final reference to a bishop of
and administrator of Amaseia,55 and in 1386, when he was succeeded by
Altamur (? I), Panaretos entitles Taceddin pelebi emir of Limnia.
What had been happening? In the words of the poet Yunus Emre, sp
for Tiirkmens further west and earlier in the century:
We went down into Rum, we wintered, we wrought much good and
Spring came, we went back again, grace be to Allah.56
As they do today, the shepherds of the interior come down for

grazing in the Iris delta, much to the resentment of its fourteen villa
the six visits by Alexios III, all timed for the winter grazing season, I
that the Emperor was fighting a losing battle by trying to head off a
Tiirkmen movement over three decades, 1351-79, when Limnia be
Trapezuntine 'Plain of Jars.' Taceddin may have assumed it to be his te
in 1362, when he solicited a marriage alliance; Alexios emphatically reg
177, 178 note 1; A. Bryer, "The littoral of the Empire of Trebizond in two fourteenth-centu
lano maps," 'ApX.f6v-r., 24 (1961), 101-2. The problem, which has exercised historians of t
since Fallmerayer, Trapezunt (supra, note 15), 303, will be further explored in the forthcom
by Bryer and Winfield.
54 F. Miklosich and J. Miiller, Acta et Diplomata Graeca medii aevi sacra et pro/ana, I (V
1860), 69-71; Vryonis, Decline, 291, 318.
s5 Miklosich and Miiller, Acta et diplomata, II (Vienna, 1862), 64-66. Metropolitan Joseph o
was consecrated by Paul Tagaris, successive charlatan Orthodox patriarch of Jerusalem and
less genuine Latin patriarch of Constantinople, and was, not unnaturally, uneasy about the
of his ordination.
56 "ndik Rumu kisladhk, qok hayr u ?er isledik/U? bahar oldu, geri gogtiik elhamdiilillah." De
Planhol, Fondements, 224.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
130 ANTHONY BRYER
it as his as part of the eventual allian

although his daughter could still pat
of 1432 shows that Greek Limnian re
a high southerly village in Trikomia,
Pontos. Alexios III had lost. Or had
Taceddin into the beginnings of a t
tabs on the Tiirkmens. He may ha
Ottomans conquered Canik,58 the T
in the overspill, Taceddin's probable
by 1404 not only sharing territory u
Komnenos of Trebizond, but also a
mention of another Altamur, obviou
son or grandson, and perhaps also a r
Grand Mesazon, the last prime minist
Chalybia lay to the east, closer to
earlier. Panaretos states, significantl
in the 1280s, "all those choria becam
absence of Christian monuments th
the Tiirkmens harrassed before they
first named of the Emirogullarl of C
other Tiirkmens, the Chalybians ai
Oinaion, with its mountain road to
Jason. Perhaps Alexios III took adva
monastery on Cape Jason during th
hacz emir, in January 1357, but the E
Hagarene heads. May 1357 found th
pastures, and the emir had to postp
safe on the coast again for the winte
he broke into the Matzoukan valley-ba
damage. It took the Emperor only eig
tional, solution, for on 29 August
(the only princess then available) to
57 Laurent, "Deux chrysobulles" (supra, note 35
58 Mehmed Negri, Kitdb-z Cihan-niamd, eds. F
1949), 323. In 1415 Mustafa, brother of Sulta
Kreki6, Dubrovnik (Raguse) et le Levant au Moy
F. Thiriet, Rdgestes des ddlibdrations du Sdnat de
nos. 1563-64 of 15-18 January 1415.
59 Clavijo, trans. Le Strange, 109. W. Miller, T
75, makes him son of Taceddin and Eudokia.
60 There is confusion over the identity of the last one, or two, Grand Mesazontes. N. Iorga, Byzance
aprds Byzance (Bucarest, 1971), 57, names a Kabazites as final Grand Mesazon of Trebizond, citing
an enigmatic "Seconde chronique grecque," which I cannot trace. But Chalkokondyles, Bonn ed.
(1843), 496, states that in 1461 the Kabazites chief was fighting as Pansebastos in the Chaldian moun-
tains. Thus, even if Kabazites had held the office of mesazon before, David would have required
another to head his government in the beleaguered city. Ecthesis chronica and Chronicon A thenarum,
ed. S. P. Lambros (London, 1902), 26, also mentions Kabazites, but states that the last Grand Mesa-
zon was 'AkrarporoploS. Miller, Trebizond, 105, suggests that the name "doubtless disguises the Muslim
Artamir, son or grandson of Tadjeddin, emir of Limnia, the husband of Eudokia, and therefore a
cousin of David."

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
she left Trebizond, Theodora had the cha

practice from her sister Maria, six years m
then visiting the capital. The Chalybian
began. Oinaion seems to have been return
the honey-conglomerate building stone of T
tine hands until after 1404. In the winter
Alexios III and Panaretos paid a state visi
(60'-rT6Kaco-rpov), when "the Turks follow
At that moment, Alexios III was simultaneo
and melik of Canik. Alexios had made the h
owner, like his other fringe lords, the K
castles. His Tiirkmens were beginning to
the winter grazing season of late October 1
immense satisfaction of seeing two of his
Limnia and Stileyman (successor to the hac
castles and coastal lands, rather than o
Taceddin was killed, but Stileyman went
in 1396/97.61 Whether he actually took K
is a matter of conjecture. Eight years lat
that the then Trapezuntine frontier began
of Kerasous, and that it reached three days
of the monastery of the Pharos in Trebizon
as far as Kerasous. But Clavijo's paradox ma
symbiosis of Trapezuntines and Tiirkmens i
of the Eleousa on the island of Ares, oppos
against the Ottomans until 1468,63 while
Kale itself was 221 Christian, and only 31 M
same ambiguity in Chalybia, Clavijo's two o
pelebi was to speak of Cape Jason and
Vona) that "the inhabitants are for the m
are interspersed with well-cultivated Gr
known by the name of W6na Greeks and
perhaps a winter phenomenon. Three day
(Giresun), and past the lands of Altamu
situation in the former Chalybian capital, O
zuntine Melesianos (? Melissenos)66 family o
e1 H. H.imGiesecke,
mittelalters KleinasienDas
(i.e.,Werk
Bazm udes 'Azz(Leipzig,
Razm) ibn Ardaslr Astarab.df.
1940), 110; Eine Quelle
Siimer, Oguzlar, 328. zur Geschichte des Spdt-
62 Clavijo, trans. Le Strange, 109; ed. Estrada, 73; Schiltberger, trans. Telfer, 41-43; Miller, Trebi-
zond, 74-75.
63 J. Bordier, "Relation d'un voyage en Orient," ed. Chrysanthos (Philippides) Metropolitan of
Trebizond, 'ApX.TT6v-r., 6 (1935), 115.
64 G6kbilgin, "Trabzon," 333.
65 Efendi, Travels, trans. von Hammer (supra, note 14), II, 40.
66 Clavijo, trans. Le Strange, 108; ed. Estrada, 73, has "Meliseno" or "Melaseno," which Miller,
Trebizond, 75, identifies with "the great Greek family of Melissenos, well known for its pious foundations
on Mount Pelion.... " But they are otherwise unrecorded in Trebizond, while a MsFAsoIvos of Rhizaion
is attested in 1432: Laurent, "Deux chrysobulles," 265.105

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
132 ANTHONY BRYER
subject to Timur, held the town, i

Kale, in which a colony of about 3
appears to report that this precipito
tomb with inaccessible Byzantine
it is a good candidate for the hacz
are G6lk6y Kilise Kale, in the foo
(Polemonion) Kale, which encases a
was not disturbed when it becam
palace. Whatever the exact status
of Chalybia and Limnia who, by t
sharing the territory of Chalybia
when Mehmed II had to fight his w
zond on 15 August 1461--although it
as much from rainstorms as from
The process observable in Panare
(and many other agriculturalists')
been the Byzantine method of tami
the grant of pronoiai to Tiirkmens
tion. But in the Pontos one can see p
settled, albeit seaborne, Greek sta
end up by sharing the same territor
one must take them out of the sum
path to a 'home-castle' on the coast,
One must give them a title, not ju
make doubly sure, one must plug o
work, generation after generation.
But in two cases the Grand Komnen
?epni, Panaretos' "Tziapnides" who a
Tzannoi. Perhaps the Limnian and
Panaretos does not say so. The ?ep
Anatolia. One or more groups began
occasionally raiding down the vall
century. They moved slowly. They
In 1348 their first known leader
ill-fated coalition with the Turks
Akkoyunlu chief against Trebizond.
the Philabonites (Hargit) valley, the
Trebizond to virtually lose its Gree
ship with the Akkoyunlu (whose su
been on Cepni territory), but the
lands and finally, under Uzun Has
67 Telfer, Johann Schiltberger (supra, note
monuments," pt. 3, 248 note 1.
68 Asikpagazade, in R. F. Kreutel, Vom Hirtenz
69 Ibn Bibi, trans. Duda (supra, note 5), 32
pt. 1, 191-93.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
Komnenoi found the ?epni difficult to dea

alliances with the Akkoyunlu leaders an
prevented them from forming relationship
after Boz Dogan, the ?epni do not seem t
with which to intermarry. Alexios III was
important campaign at the end of the win
?epni along the lower and western reaches
the coast about 75 km. inland in all. He rav
before reaching a possibly Trapezuntine su
yayla) at 2919 m. above sea level. But the
valley became the ?epni ndhiyesi after 1
to have dealt gingerly with the area, alt
timars round Trabzon. Chalkokondyles rega
to Amastris (Amasra), as being infested wi
they roamed round Kerasous, making a fou
in the fifteenth century. From the Philab
passes of Cheriana by 1404, where Gree
complaining of being pestered by them in
not until after 1461 that they leapt across t
valleys to continue their slow drift east, m
and raiding down the valleys. There wer
century Lazistan and they troubled the Ch
until recent times, eventually reaching t
to them) in 1915. From thirteenth-cent
'Colchis' is a ?epni migration of 600 km. in
With the fourth and greatest of the pastoral
the Grand Komnenoi were outstandingly s
failed to make tame emirs out of them or
calls them Turks after (all commentator
Panaretos is our only near-contemporary so
tos, who last refers to the 'Amitiotai' by
died in, or just after, 1390.71 The 'Amitiota
of Amida until 1401 at the earliest. Panaretos' 'Amitiotai' were 300 km. and
five mountain ranges away from Amida. Cahen first pointed out the problem ;72
here is a possible solution.
The first apparent member of the dynasty to emerge in documentary evi-
dence is a 'Koustouganes,' who attacked Kerasous in 1301. Alexios II "dea-
dened his pride" and rebuilt the walls of Kerasous."3 'Koustouganes' appears
70 Chalkokondyles, Bonn ed., 65, 496; G6kbilgin, "Trabzon," 329-30 (in the early 16th century
the 9epni had 52 villages, 1,892 taxpayers, and 1,674 households); Clavijo, trans. Le Strange, 120
('Chapenies'); Ioannides, 'co-ropia (supra, note 32), 252-53; S. Athanasiades, 'lo-ropia Kal Aaoypacpia
T-rs lXacras, I (Salonica, 1967), 33-59; Siimer, Oguzlar, 327-35; Bryer, "Laz and Tzan," pt. 1, 191-93.
71 Panaretos, ed. Lampsides, 23.
72 Cahen, Pre-Ottoman Turkey, 363-64; idem, "Contribution a l'histoire du Diyar Bakr au quator-
zieme siecle," JA, 243 (1955), 91-92.
73 S. Sgouropoulos, ed. A. Papadopoulos-Kerameus, 'AvdAEK-ra 'IEpohCoXvpTrtKi XTrcaXvoXoyfaS, I (St.
Petersburg, 1894), 431-37.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
134 ANTHONY BRYER
to have attacked from the immedia

down to the sea and where, for over
Middle Ages; the stretch is near
1318, two new place names emerge;
suyukdy, by Piraziz, 18 km. west
'Omidia' between Pazar and St. Basil
'Omidia' can no longer be located, b
(now Bulancak). Is 'Omidia' where
very tentatively, that the remote orig
unlu federation that controlled eastern Anatolia and northern Persia under
its 'Little Turk' should be sought in a modest area just west of Kerasous at the
end of the thirteenth century.
The 'Amitiotai' were initially tiresome. There were summer skirmishes in
the pastures in August 1340 (when they were chased off), and in June 1343
(when they retaliated). By then the 'Amitiotai' had begun their ?epni relation-
ship, and in June 1348 the combined Tiirkmens established a potentially
dangerous alliance with the Turks of Erzincan and Bayburt. It failed, but the
Tiirkmens clearly needed detaching. Alexios III, therefore, took a step which
is now familiar. He was probably following a policy initiated with Sinope
(and, perhaps, Erzincan) by Alexios II, and he may well have been influenced
by the success of his ally and kinsman, John VI Kantakouzenos, in marrying
his daughter to Orhan, six years before.76 But it was the first time that a Byzan-
tine (let alone a Trapezuntine) princess had been offered a Tiirkmen, and it
profoundly affected the Oguz imagination. Ballad 6 of the Dede Korkut cycle
describes how Hanh hoca, the Akkoyunlu leader, sent his son Han Turali to
fetch an Amazon bride. Naturally, only the tekfur of Trabzon had such a
daughter and, naturally, Turali had to undergo three labors which would have
daunted Hercules before he could snatch his bride: to down a black bull, a
royal lion, and a vicious camel in the maydan of Trebizond. When Turali
unveiled himself, however, the Trapezuntine hatun had other ideas; she
74W. Tomaschek, Zur historischen Topographie von Kleinasien im Mittelalter, SBWien, Phil.-
Hist.Kl., 124 (8) (1891), 80; K. Kretschmer, Die italienischen Portolanen des Mittelalters (Berlin,
1969), 648; A. Delatte, Les portulans grecs (Liege-Paris, 1947), I, 238; II, 33; Bryer, "Littoral,"
105-6. Despite the fact that it appears as ?1 Mi?Sta in a late portolan (and even as NIKOp?8EIa in
another), Tomaschek's suggestion that it is in fact 6 Mij8ias [AlCpv] must be no more than a happy
notion, for the Greek definite article cannot be expected to be so firmly attached to a name (especially
in a confusing gender). Kretschmer's opinion that the place was in fact the Boz Tepe of Ordu is equally
farfetched. Balabanes argues improbably that Bulancak was originally nepavrL~dlv, a diminutive of
Pera. It is possible, I suggest instead, that Omidia could be the otherwise unidentified settlement of
'Appc8iov beach in the district of TO0 HouTrLUa, mentioned by Lazaropoulos in connection with the
events of 1222/23, and not heard of again, suggesting that it was lost to the Trapezuntines at an early
stage. G. Balabanes, H6Sev -r6 6vopa T~irS ovXaavrrLKS T-ro0 H6vrov, in Horl&ovTK caoXa, 3 (27) (1938),
108-11; Papadopoulos-Kerameus, FHIT, 122.
75 In discussion, Mme Ahrweiler proposed that the 'Amitiotai' derived from 'Apr-rT6S (Amisos).
This would be a possibility were it not for the fact that this form is otherwise unattested throughout
the history
which of theed.
Panaretos, city, which was
Lampsides, 69, known
75, usesincreasingly
exclusively. from the tenth century as 'Apiva6;, a form
76IrJne M61likoff-Sayar, Le destan d'Umur Pacha (Paris, 1954), 106-8; P. Lemerle, L'dmirat
d'Aydin, Byzance et l'Occident (Paris, 1957), 175, 221; M. Izeddin, "Notes sur les mariages princiers
en Orient au Moyen Age," JA, 257 (1969), 144-45.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
"was watching from the palace and she w

miaowed, she slavered like a sick calf. To the maidens at her side she said,
'If only God Most High would put mercy into my father's heart, if only he
would fix a brideprice and give me to this man! Alas that such a man should
perish at the hands of monsters!"' Turali disembowled the bull; there were
alarming scenes in the maydan when "The lion roared, and every single horse
in the square pissed blood"; and the Trapezuntines tried to nobble the camel.
Turali won his princess; they rode off. But the tekfur of Trabzon repented his
loss and sent his army to retrieve his daughter. She took matters in hand
and sliced up six hundred Trapezuntine warriors. Turali's manhood was
slighted; they duelled and the hatun shot a headless arrow "that sent the lice
in his hair scuttling down to his feet." They were reconciled and Turali and
the princess of Trebizond lived happily ever after."
For the historian, Turkish heroic poetry is more than usually tricky material.
But here is an almost certain (if rare) link, for it corresponds with (though
may not actually represent) Panaretos' laconic entry that Alexios III married
his sister Maria to Fahreddin Kutlug beg of the 'Amitiotai,' son of Turali, in
August 1352. Note the discrepancy of names; it is unlikely that Turali himself
had married a Trapezuntine and Panaretos is certainly to be preferred to
Dede Korkut. But the ballad is right in singling out the alliance for communal
memory, for it was itself successful and initiated an Akkoyunlu-Trapezuntine
understanding, which remained the cornerstone of the foreign policy of the
Grand Komnenoi for over a century. A program of state visits ensued after
1352. Maria visited Trebizond again in 1358; Alexios III was prepared to assist
Kutlug beg during the summer grazing of 1363; the emir visited Trebizond and
was much honored there in 1365; and the Emperor returned the visit in 1367.
Up to eleven princesses of Trebizond are known to have married Muslims,
eight of them Tiirkmens; indeed only five are known to have married Chris-
tians.78 By contrast, thirteen male Grand Komnenoi took Orthodox brides,
Trapezuntine, Byzantine, and Georgian--with the possible exception of the
parricidal John IV, who, rather unconvincingly, maintained that he had
married a Turk in the interests of piecemeal proselytization (she was perhaps
a concubine, for he already had a Georgian wife).'" Medieval Orthodox and
Muslim Tiirkmen custom is directly opposed on how marriages are arranged."8
77 Extracts from Dede Korkut (supra, note 14), trans. Lewis, 17-18, 117-32. See also M61ikoff,
"G6orgiens et Arm6niens," 18-28; M. Fahrettin Kirzioglu, Dede Korkut. Oguzndnelevi (Istanbul,
1952), 117-18; E. Rossi, II "Kitab-i Dede Qorqut," ST, 159 (Vatican City, 1952), 31-33, 180-93, 219.
78 Na, d. of Alexios I = Andronikos I Gidon; Anna, d. of Alexios III = Bagrat V-VI; Maria,
d. of Alexios IV = John VIII Palaiologos; ? Na, d. of Alexios IV = George Brankovi6; ? Na, d. of
Alexios IV = Niccol6 Crispo of Santorini.
9 Alexios I = Na, d. of Axouchos (Axuch); Manuel I = Rusudan; Andronikos I Gidon = Na,
d. of Alexios I; John II = Eudokia Palaiologine; Michael = Na, d. of Constantine Akropolites;
Alexios II = Na, d. of Beka Jaqeli; Basil = Eirene Palaiologine; Alexios III = Theodora Kanta-
kouzene; Manuel II = Eudokia (Giilhan), d. of David VI-VII and Anna Philanthropene; Alexios IV
= Theodora Kantakouzene; Alexander = Maria Gattilusio; John IV = Na, d. of Alexander the
Great; David = Maria of Gotthia and Helena Kantakouzene. Tafur, trans. Letts (supra, note 22), 131.
80 Islamic law speaks of a dowry (mahr) but no bride-price. Where, however, bride-price was a
local custom, it seems to supersede the dowry: e.g., the bride-price used by medieval Tiirkmens and
Mongols is today common throughout Anatolia. Christian-Muslim marriages are rare enough, even

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
136 ANTHONY BRYER
Orthodox princesses would certain

thirteen dowries; did eleven Muslim
bride-prices too? Could the Grand
worlds, twenty-four dowries and brid
It is possible that both systems o
certain that Trapezuntines and Tiirkm
lights. It looks as if Alexios III ced
hacs emir of Chalybia may also have m
Certainly, in the last and most fam
with the Akkoyunlu Uzun Hasan in
Trebizond itself. But Uzun Hasan h
protection against the Ottomans. At
It is difficult to make out if ther
which were advanced in similar cases
Brides remained Christian and were e
no stephanokrator (as when the fut
daughter of Bagrat V-VI in 1377),81
lands. Almost certainly the Orthodox
of the famous marriage of Orhan and
interesting to know why popular opi
first request for a bride in 1362; th
Eudokia, and only married her afte
religious objection, for there had bee
and the case of Eudokia was to show t
widow going on to marry a Byzantin
with Alexios' long struggle with T
about what Tiirkmens expected of su
brother-in-law, the hacs emir, as som
went to help his other brother-in-law
answer to the summons of a revers
an unhappy experience of serving Ot
ibn Battuta's vivid picture of the st
household (a sort of permanent Hig
in Cyprus and the Lebanon, to preclude a regula

but in the Caucasus coexistence of faiths may p
the dowry in favor of the bride-price (kalim) of
similar tendency among Pontic Greeks. As Busbe
position by the 16th century, having both the d
independence) and the bride-price (for lesser co
were anxious to provide princesses married to Ot
of Theodora Kantakouzene = Orhan (1346); t
and Maria/Mara of Serbia = Murad (1345). The G
guarantee the security of their princesses, but un
probably not have given dowries themselves. Se
Flanks of the Caucasus (Oxford, 1940), I, 207, 267
de Busbecq (Oxford, 1968), 28, 118-19; Ducas,
(Bucharest, 1958), 39, 59, 257, Bonn ed. (1834), 17
Rakintzakis and to Dr. David Kerr for discussion

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
protectress of Christians with her own establ

have been paralleled only by the case of Th
But what one side might regard as a High C
regard as a hostage. On a smaller scale, the pr
similar status in the Tiirkmen courts, but of
to have been an ulu hatun (principal wife),
enjoyed the considerable privileges of being th
had not been in the eleventh century, there m
against rulers of part-Christian origin, and of
with Trebizond only Uzun Hasan is known t
Orthodox rulers of Trebizond were unable
have felt that they were indebting local emirs
by subordinating the wives, local emirs may h
sort of mark of vassalage. It was probably l
is on the subject, for, so far as Alexios III w
justified the means. His alliances, and his di
on Boz Tepe, where he could receive his six Mu
worked. By the middle of his reign, the Tiirk
to the Grand Komnenos melik of Canik, and A
saving Chaldia and Cheriana, to the south, f
Bayburt.
But if the marriage issues were confused, both Trapezuntines and Tiirkmens
shared certain notions of dynastic and family structure. On the highest plane
was a system of dynastic government, what might be termed extended and
interlocking ruling families stretching from France to the kiriltays of Central
Asia.83 The Grand Komnenos did not quite make this level, but shared its
assumptions, for the very term 'Grand' Komnenos became more than a
name, or even an epithet: it was an entitlement to rule shared even by
daughters who married Tiirkmens, such as the Grand Komnene Maria; a
sort of Trapezuntine equivalent of the 'title' of the nameless hacs emir of
Chalybia.84 Similarly, the name Kantakouzenos was once sold as a sort of
entitlement to rule.85
On the next scale is the 'large family,' a grouping of families often ac-
knowledging a common and usually mythical ancestor. Pontic Greeks were
long used to this type of society on their eastern and southern marches,
beginning with the Bagratids of ispir and Bayburt, who eventually turned
Caucasian history into a sort of animated genealogy, where clan status,
82 tbn Battuta, trans. Gibb (supra, note 52), II, 488-89, 498-503.
83 See V. Vladimirtsov, Le rigime social des Mongols (Paris, 1948).
4 I am not here concerned with the origins of the title, on which there is an immense bibliography,
ending with D. Polemis, "A note on the origin of the title 'Miyas Kowqv6s',"Neo-Hellenika, 1 (1970),
18-23; 0. Lampsides, "Bessarions Zeugnis iiber den Titel Miyas Kovrlv6s," 'ApX.TT6vr., 30 (1971),
386-97; B. Hemmerdinger, "M-yas Kopvnv6s, Calque de Hohenstaufen," Byzantion, 40 (1970), 33-35
(perhaps the most bizarre and unconvincing of all explanations); and 0. Lampsides, "Miyas Kopivv6s-
Hohenstaufen," Byzantion, 40 (1970), 544-45.
85 D. M. Nicol, The Byzantine Family of Kantakouzenos (Cantacuzenus), ca. 1100-1460. A Geneal-
ogical and Prosopographical Study, DOS, XI (Washington, D.C., 1968), xi.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
138 ANTHONY BRYER
rather than office or land, was the

Akampsis valleys certainly understo
and Kabazitai shared it too. The ar
families,' may therefore be expecte
to affect local Trapezuntine Greek v
miracles of St. Eugenios, a Byzant
zuntine litigants apart, because th
Saint was energetically promoted
with St. Eugenios, 9th6nroAt Kiai lA
nation for the cohesion of Greek v
Tiirkmen model? Uspenskij went so
"had in mind the origin of one clan
must be dismissed, for the fact is t
names and the charters show no e
more than the usual Greek nucle
factors, the modest size of the goni
Of lesser marriages we know litt
there is no Tiirkmen source equi
reveal a number of Turkish surna
name; indeed the first two certain
soil do not come until 1432: a par
smallholder of Sourmaina called A
faith persons such as 'Echiseni,' dau
of 'Murti,' recorded by the Genoe
Phadisane (Vatiza, Fatsa), professe
and Christian given names may offe
Turks took their Christian parents'
The confrontation with the Tiirk
familiar elsewhere.93 Manpower, no
and stock was relatively expensive.
was 5 aspers; and on a larger one,
4-6 aspers; of a second-hand copp
aspers; and of a riding horse, 150
aged eleven, fetched 600 aspers, a
8" Papadopoulos-Kerameus, FHIT, 141.

87 Ibid., 51.
88 Acts of Vazelon, clI-cilI.
89 Only two people called Eugenios appear in
all. On the size of Byzantine rural families of t
rurale ' Byzance aux XIIIe, XIVe et XVe siec
Laiou, reported in DOP, 27 (1973), 328.
90 E.g., 'lcavvd~K 6 'A-rLovurr&; Baoi~Aos 6 T
wovAos ( ? akar); E6800pos 6 TLaucb'lrn (? Cam
MapouXa ' TLtXrntv6wovAos ( ? pelebi); Bai2stos
91 G.
92 Laurent, "DeuxActes
I. Bratianu, Chrysobulles,"
des notaires26816s,
gdnois 269181:
de Pdra MaxporMrs;
et de Cafa de 'AaS?,avrrK1&S.
la fin du treizibme sidcle (1281-1290)
(Bucharest, 1927), 172; idem, Commerce ginois (supra, note 12), 172-73.
93 Vryonis, Decline, 241, 257 note 706, 307-14.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
ransom of a sister of a hieromonk (age and

Tiirkmens was fixed at 850 aspers.94
The Tiirkmens may well have contributed
they needed manpower even more. As e
Panaretos never records the taking of Tiirk
of Vazelon provide ample evidence that t
The charters suggest that the scale of T
relatively small, was sometimes enough
economy. Through minor raids and inexpli
on the pastures in May, there was a steady
rupted widows and, in an extreme case,
a classic case as early as 1261, Maria Tzarch
to Vazelon monastery, for her salvation an
of those of her parents. The formula is co
reason: "I have five sons made prisoner; if
their shares; if they do not come back, thi
gardens and trees, land cultivated and fallo
The monastery kept the land she could no
The appearance of AIXPuicoTos ('refugee
name,97 and of an exiled bishop of Satal
fact that most Trapezuntine monasteries, e
were fortified," are telltale signs of Tiirk
whether the Tiirkmen economy influenc
pastoralists turned agriculturalists into t
likely that the Pontos was already practic
humant economy. Panaretos reveals that
terms for the system: XEtgatt& (kisla); and
94 Vazelon Acts, nos. 84 of 1263, 90 of the 13th centur
the English accounts in note 13 supra; Moretto Bon,
1408) (Venice, 1963), 11-12, 16.
95 Trigoliktos, in Vazelon Acts, no. 106 of the 13th cen
96 Vazelon Acts, no. 38 of 1261; Germaine Rouillard, L
1953), 165-66.
97 Vazelon Acts, no. 106 of the 13th century; Alice-Mar
(1289-1293; 1303-1309) and the Church," DOP, 27 (197
98 Vazelon Acts, no. 53 of 1256.
99 E.g., the Pharos (Laurent, "Deux Chrysobulles"), and the Hagia Sophia in Trebizond. Fortified
monasteries are a feature of Dede Korkut and the Melikdanifmendname; de Planhol, "Signification"
(supra, note 11), 235.
100 The Armenian equivalent for yayla was leri (cf. the bishopric and village of Leri in Chaldia);
G. Dum6zil, Contes Lazes (Paris, 1937), 23. There is a substantial bibliography on the Parcharin
(e.g., G. Bapheiades, 'EgoXis T~-r TpacrElo roS, in Xpov.TT6vr., 1 [9] [May, 1944], 186-88), but beyond
A. A. Papadopoulos' derivation from wrrapaXcptov ('kIrrtoplK6v AElK6V T s ITOVTrlKl AtaCh~KTOU, II [Athens,
1961], s.v.), there has been no discussion of the origin of the term, which I suggest may be derived
from the Paryadres-Barkar-Bul'ar Da' of the Pontic Alps: E. Honigmann, Die Ostgrenze des
byzantinischen Reiches (Brussels, 1935), 181 and note 5; V. Langlois, Collection des historiens anciens
et modernes d'Armdnie, I (Paris, 1869), 83; II (Paris, 1880), 82, 299; J. Laurent, Byzance et les turcs
seldjoucides (Nancy, 1913), 43; N. Asim, Celalijttin Harezemfah (Istanbul, 1934), 131; J. Markwart,
Siidarmenien und die Tigrisquellen (Vienna, 1930), 17*"-18*, 21*-23*, 230, 404, 411; Bryer, "Laz and
Tzan," pt. 1, 192 note 124. I am grateful to Dr. V. L. M6nage for discussions on the subject. Among
Greek agricultural terms which have passed into modern Oflu Turkish, Mr. Michael Meeker kindly
tells me of fo(r)man = the Oov&pta of Vazelon Acts, no. 62 of the 13th century (the haystacks on
"hatstands" peculiar to the Pontos).

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
140 ANTHONY BRYER
Summer pastures, some part of the

the charters. The scale of Trapezun
as the Tiirkmen, and most Trapez
within a day of the permanent wint
Matzouka (Palaiomatzouka), just be
Pontos, will illustrate how the dual
Taken almost due west-east, from M
section was :10
Name Function Height Distance from

Mt. Boudoxe
Mount Boudoxe Summer pasture below 2300 m. -
Koleles' Summer settlement 1500 m. 1500 m.
Tsimprika Permanent village 400 m. 4000 m.
River Prytanis Fields above 300 m. 4300 m.
'Imperial Highway' Fields above 400 m. 4800 m.
Giannakanton Permanent village 500 m. 5500 m.
Choumerixa Summer settlement 1700 m. 8500 m.
Mount Karakaban Summer pasture below 2400 m. 10500 m.
Not surprisingly, there was considerable penetration of Turkish names and

terms in Trebizond. Turbaned102 Trapezuntine soldiers "make use of the sword
and bow, the like of what arms the Turks employ, and they ride after the
fashion of these last"103 (with a short stirrup). Exceptionally in the Byzantine
world, hatun (XaTo'va)'04 became an acceptable Greek alternative for a despoina,
and by the fifteenth century the Grand Komnenos Alexander was locally
called iskender (xKav-rTptos).105 The tekfur would receive the ritual Easter
acclamations in the maydan (MaiT-vtv)1?6 and his delal (vUKToTcaXaios, night-
watchman) would patrol the bur9 (TovprTLios, castle);107 his protospatharios wa
the emir candar (aPvprLavraptos, captain of the guard);10s his akolouthos was the
'Horchi' (xovpTlfis, "the page who carries the imperial bow before his majes-
ty") ;19 and his chief falconer may have been an emir dogan (a~rpaTroavaKav-
101 This area will be described in detail, and places identified, in the forthcoming study by Bryer
and Winfield.
102 D. Talbot Rice, Haghia Sophia, fig. 121.
103 Clavijo, trans. Le Strange, 115.
104 Vazelon Acts, no. 166 of the 15th century; cf. G. Moravcsik, Byzantinoturcica, II, 2nd ed. (Berlin
1958), 343-44.
105 Chalkokondyles, Bonn ed., 437.
107 H. Gregoire, "Les veilleurs de nuit ' Tr6bizonde au XIVe sidcle," BZ, 18 (1909), 490-99.
108 J. Verpeaux, Pseudo-Kodinos, Traitd des Offices (Paris, 1966), 341-42, 348; Laurent, "Deu
Chrysobulles," 26154, 2671s5, 269202; Moravcsik, Byzantinoturcica, II, 68. I am grateful to Professo
Speros Vryonis, Jr., for discussion of the term.
109 A term whose etymology (not, apparently, Georgian or Turkish) eludes me. Clavijo, ed. Estrada
75-76; trans. Le Strange, 112; Verpeaux, Pseudo-Kodinos, 341-42, 345. Although the list published
by Verpeaux contains, exceptionally, the emir candar and "Vrcho," he insists that it does not represen
a Trapezuntine order of precedence; if it were, it would indeed raise difficulties, but Verpeaux over-
looks the fact that it is inserted in Cod. Marc. gr. 608 immediately after the unique MS of Panaretos,
and is in the same hand. That this is, therefore, a Trapezuntine list is a strong possibility.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
Tos).lo It is therefore tempting to find in the civil wars
1340's, in which the emir candars took a prominent part, at
Tiirkmen menace, a struggle between Tiirkmen and Trap
government. The politics are complex, but in fact none of th
social and political disturbances (which reflected those of
can be said to have had anything to do with the Tiirkme
from the possible exception of the Grand Mesazon Altam
career of the Genoese Jeronimo de Nigro (Grand Mesazo
Vestiarios, 1445-49),111 none of the 192 other known office
history of the empire of Trebizond have other than Pon
names-unless the famous George Amoiroutzes is given ances
court. Turkish (and Laz) influence upon Trapezuntine society
administration was unexpectedly superficial. Instead, Tr
care to keep up with developing forms of Byzantine gove
even more Roman than the Romaioi they called themselves. A
remarked, the Grand Komnenoi preserved "Hellenic manner
speech," and only intermarried with neighboring barbarians
trouble through the ravaging of their lands by them.""'4
Of the interplay of faiths there is even less to record, for t
one of what may be called the "Vryonis Rules" for other
which is that the missionary activities of the dervis orders c
on the heels of the collapse of the structure of the Ortho
the time that the dervises came, their zeal was long abated an
only one known Pontic tekke, at Tirebolu (Tripolis).x16 Whi
Trebizond lost its suffragans of the ninth-century limes far
Pontos is the only area which (in the face of all other Anato
actually had to create new bishoprics from the fourteen
burden of Wa~chter's study of the fate of Byzantine Anatol
litany of abandoned sees."' But in the Pontos new bish
there was Limnia (fl. 1317-84); the Chaldia of 1390 was t
late fifteenth century, to be refounded from about 1624;118
Roman Catholic see in partibus) is recorded in 1461;119 Op
briefly in the late fifteenth century;120 and finally the Mat
110 Vazelon Acts, no. 166 of the 15th century; but see Laurent, "Deux Chryso
KoqcAwEOV).
111 lorga, Notes et extraits (supra, note 26), I, 34; III, 246.
112 Listed in Bryer, doctoral thesis (supra, note 12), II, 171-96.
113 For example, the overhaul of the judicial structure in Constantinople in 1
imitated in Trebizond, with its Supreme Judges of the Imperial Court: Vazelo
of the 14th century; 133 of 1381; 134 of 1415.
114 Chalkokondyles, Bonn ed., 461-62.
115 Vryonis, Decline, 363-402.
116 V. Cuinet, La Turquie d'Asie, I (Paris, 1890), 55.
117 A. H. Wichter, Der Verfall des Griechenthums in Kleinasien im XIV. Jahrh
118 Papadopoulos-Kerameus, "Trapezountiaka," 679; Laurent, DHGE, X
Gelzer, Ungedruckte und ungeniigend veriffentlichte Texte der Notitiae episcopa
philol.Kl., 21 (1901), 635.
119 Laurent, DHGE, XII, cols. 632-33.
o20 Gelzer, loc. cit. in note 118.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
142 ANTHONY BRYER
was awarded the bishopric of Rhod

sheltered ancient and distinguished
1256; Amaseia, which ended up, vi
which ended up, via Oinaion (per
Ordu.122 It iS possible that the weight o
upon Tiirkmen. The earliest manuscript
for the abjuration of Islam and entry
tos.123 The Tiirkmens of the region w
they "worshipped the shah of Persi
suspects, twilit, enabling the later '
and out of them with ease. The Musl
until very recently, and the modern
the feast of the Koimesis will find h
animal sacrifices.
An obvious explanation for Trapezuntine integrity in the face of the Tiirk-
mens is that of numbers. Many pastoralists may have escaped enumeration
in them, but the defters of the 1520's show that even after some Ottoman
settlement, the Pontic population was still overwhelmingly Christian.125 Many
of the rules which had applied to the earlier Byzantine collapse in Anatoli
no longer held in 1461. The Ottoman conquest of Trapezuntines and Tiirk-
mens was swift and relatively painless. A sizeable proportion of monastic
lands were not sequestered, and the three major inland monasteries (Peristere-
ota, Soumela, and Vazelon) still retained forty-four of their villages in 1890.
When your landlord is an abbot, you think twice about apostasy, although
Soumela, at least, faced long-deserved Greek peasants' revolts in the nine
teenth century.126 Of suppressed monasteries, the Chrysokephalos cathedral'
lands became the Fatih Evkaf, and the Pharos lands (nearly fifty holdings
along 222 km. of coast from Gonia west) were widely transferred to the Evkaf
of Giilbahar-Maria of Doubera (Livera, now Yazlik, on the Soumelan estates),
who, in the last and most exalted of Trapezuntine-Muslim marriages was wif
of Sultan Bayezid II and mother of Sultan Selim. Both Evkaf survive; thei
archives should be interesting.
So the people of Halt, their Orthodox Church, and an exceptionally large
part of its economic basis survived, not initially threatened by Muslim
missionaries, conversion, or large-scale Turkish settlement. The balance does
not seem to have changed until the late seventeenth century, when a new
breed of coastal derebeys emerged, turning Trapezuntine Greeks into doulo-
paroikoi. Many fled to new nucleated settlements on the marginal lands
121 Chrysanthos, 'EKKXlicfa (supra, note 24), 683-89.
122 Bryer, Winfield, and Isaac, "Nineteenth-century monuments," pt. 4, 128-29.
123 Vryonis, Decline, 442 note 122.
124 F. Taeschener, "Mehmed Aschyqs Berichte iiber Tschepnis," ZDMG, 76, N.F., 1 (1922), 141,
282-84.
125 G6kbilgin's figures, "Trabzon" (supra, note 22), suggest a 91-94 percent Christian and a 9-6
percent Muslim population in the 1520's.
126 E. Th. Kyriakides, 'lo-ropia -rij wTap& Triv TparrELoovra Itp& a ao'AKiP j w1TrarpcxapxtKi oa-ravpowTrytConKai
povijS "riS 'YwpaCryfas Eor6Kov 'riS lovpEa& (Athens, 1898), 204-59.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
below the summer pastures, where they enc

more. For the first time there were neo-marty
during the breakdown of Ottoman authority in
Thereafter, the Anatolian rules are that Greek
before their faith, their faith with their Ri?tm
Islam. The Pontic experience was almost the op
as the prestige language in some Turkish ar
that the last vestiges of the medieval confr
Tiirkmens have vanished. The descendants of
remembered with favor) may well be sortin
Giresun, or picking tea in Rize; Greek Trapezu
the Exchange of Populations; Turkish Trapezun
have substituted American maize (sweetcorn) f
their transhumance in well-organized carav
with women, children, and long-haired cattle,
APPENDIX I: GREEKS AND TURKMENS IN PANARETOS
Much of the argument in this paper depends upon a close interpretation of the text
of Panaretos and upon the accuracy and logic of identifications of names in it (where
Moravcsik, Byzantinoturcica, II, has been widely consulted). A translation of the relevant
passages is therefore given below. References are to the page and line numbers of Lampsides'
edition of Cod. Marc. gr. 608, fols. 287r-312r (supra, note 22); Panaretos' dating system
is simplified and indiction numbers are given in Roman numeral capitals.
633 And Kaloioannes Komnenos [John II, 1280-97] ... died in Limnia on Friday 16 August,
X, 6805 [1297]. But during his reign the Turks seized Chalybia [Onye district] and launched
a great invasion [into Trebizond proper], so that all those places became uninhabited.
6315 The Emperor kyr Alexios [II, 1297-1330] set out against the Turks in September 6810
[1301]; he found KouorrouydvrS129 at Kerasous [Giresun] and overwhelmed him, slaughtering
many Turks.
6324 lTapidsT&CS [Bayram] seized the -rTpya [pergi, tented stalls]a30 in the evening of Tuesday
2 October 6822 [1313].
6326 In 6827 [1318/19] the Sinopitans started a great fire which entirely devastated the beauties
of the city [of Trebizond], both inside and outside.
127 Bryer, "Tourkokratia," 30-54.

12s J. Humlum, Zur Geographie des Maisbaus (Copenhagen, 1942), 29, 90; cf. Baddeley, Caucasus
(supra, note 80), I, 255.
12s Unique mention of the name, possibly 'kiigiik aga,' or even kiistah = 'insolent' (which Stephen
Sgouropoulos' infuriatingly vague, but contemporary, encomia claim he was). Moravcsik's tentative
identification with the Boz Do'an of 1348 is chronologically improbable. Janssens is mistaken in
following Miller in computing Panaretos' date to 1302. Papadopoulos-Kerameus, 'Av&SK-ra 'IEpoo-
Vplrt1Krnd T-raoXvooyias (supra, note 73), 431-37 (cf. Triantaphyllos Papatheodorides, 'Av4~KoTo o riXoI
Ireqpdvov t-ro XyovporroiAov, in 'ApX.fl6v-r., 19 [1954], 262-82); Moravcsik, Byzantinoturcica (supra,
note 104), II, 170; E. Janssens, Trdbizonde en Colchide (Brussels, 1969), 93; W. Miller, The Last
Greek Empire (of the Byzantine Era, 1204-1461) (London, 1926; ed. A. C. Bandy, Chicago, 1969), 33;
Fallmerayer, Original-Fragmente (note 49 supra), pt. 2, p. 15**; S. Fassoulakis, The Byzantine Family
of Raoul-Ral(l)es (Athens, 1973), 57 note 1.
130 Possibly tented market booths, such as Bessarion described in the Trebizond bazaar. I. T.
Pampoukes,
pkovos flov-rtaK&
'EyKcbptOV rrolKfica,ininNos
Ei TpaCTrLOVTCa, Xpov.J6vr., 1 (3) (November
~'EAA., 13 (1916), 1943),as74-75;
187; also published S. P. Lampros,
a monograph (Athens, Brjacra-
1916), 45.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
144 ANTHONY BRYER
6412 "apitdpr [Bayram] brought a large ar

now Hamsik6y],131 and a great number of
they even took many Turkish horses, on
6430 But on Friday 5 July 6844 [1336], ItX
of Tapap-ro'rns [Temiirtas, died 1327/281
at the palisade ('AXdvTraKaQS) of Saint K
through the favor of God, he was turned b
[Abdiirrahman], son of Pouvorr&dris [Ruste
6529 About August of that same year [1340
X&ptv) and plundered the 'APClTrWTat, taki
666 ... on Wednesday 4 July [1341], the
pushed back without battle and many Ch
burnt down, inside and outside, and ma
burnt to death. After this disaster the st
the Sudden Death (alvi8vtos $avaToS).
6717 In June 6851 [1343], the 'AplTricTQa c
686 In 6855 [1347], Saint Andreas [probably
6813 On 29 June, I, of that year [1348], nu
Trrcd [ahi Ayna beg] from Erzikain [Erz
Paipert [Bayburt], ToupaAiTrEK [Turali begl]
[Boz Dogan] and the Tliacrvi&sE [Cepni], a
and crestfallen, losing many Turks on the w
709 On 22 September of the same year, 68
stantinos Doranites, who was exercising
we came back again.
7020 In the same year, 6860 [1352], the Emp
away to marry XouTrhonrrS1s [Kutlug be
'ApatrLiTrat, in August.
7116 Around August of the same year, 6863
the modern vilayet of Gfimiishane], mar
[$iran-Ulusiran] and took it; at this tim
joined the imperial dominions.
712 On Friday 27 November, IX, 6864 [1
the Emperor [Alexios III, 1349-90] agains
ravaged, we laid siege, and we took priso
in a disorderly flight when a few Turks
slaughtered and destroyed then, and Ioa
and if the Lord had not been with us, I m
lost; but by the favor of God my horse was
we were free and reached Trebizond after t
726 On 19 December, X, 6865 [1356], we we
of Christmas at Kerasous [Giresun], and
Yasun Burunu] when fourteen Turks wer
and got to Trebizond in good health, hav
131 I have hitherto followed Chrysanthos an
with the church of the Archangels, Platana (A
epithet (which is, anyway, in the singular), an
likely that it can be identified with the Asoma
gift to Vazelon, in Act no. 104. Ballance, Brye
(supra, note 26), pt. 1, 258; Chrysanthos, 'EKKA
131a Moravcsik, Byzantinoturcica, II, s.v. the
in fact ,obanlis: Cahen, Pre-Ottoman Turkey, 3
(Cambridge, 1968), 409-15.
132 Unidentified in Moravcsik, but see Fallm
80: Iktidar = power; "Erkebdar" = "Grossk

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
7214 In May, X, 6865 [1357], the Emperor went up

an army and marched round the whole of the summ
7216 On Saturday 11 November, XI, 6866 [1357], the
of kyr Alexios [II, 1297-1330] the Grand Komneno
7219 On Monday 13 of that month [November], XI,
hacz emir], son of Uraip&6yrl [Bayram], invaded
Palaiomatzouka [HamsikSy] as far as Dikaisim
warriors, and ravaged and abducted many people,
preparedness of our guards.
7226 On Wednesday 22 August, XI, 6866 [1358], the E
sister of the Emperor, who was married to XouTX
came to Trebizond.
7229 On Wednesday 29 August, XI, 6866 [1358], kyra Theodora, daughter of the Emperor ky
Basileios [1332-40], left to marry the emir Xa-rTUpplst [the hacz emir], with kyr Basileio
Choupakas, the Scholaris [Protovestiarios, 1355-58], as bridal escort (vuvpoo-r6AoS).
733 In April, XIII, 6868 [1360], the Emperor went up to Chaldia and built The [fortress of the
Cuckoo (To3 Koinou); and Xo-rblca-raTiqrs [hoca Latif] moved up from Paipertios [Bayburt
bringing three hundred cavalry, on which occasion Ioannes Kabazites lost the Headship.
7314 On 6 December of that year 6869 [1360] ... the Emperor went to Limnia and stayed there
for three months and a half and came back again.
7324 On Friday 23 July, XIV, 6869 [1361], XOT-LcAiaT{)ls [hoca Latif],134 the ruler (KExyaki
from Paipertios [Bayburt], brought 400 selected soldiers and penetrated Matzouka [Maqka]
toward Lacharenes [Larhan, now Akarsu, at 1250 m.] and Chasdenichas [Gizenica, Upper
Chortokopin, Hortokobubala, Yukarihortokop, now YukarikSy]. The Matzoukaitai, on the
other hand, surprised some two hundred Turks and carried off the majority, together with
many arms and horses, and decapitated this Xo-rlaCaXTrrs [hoca Latif], and on the following
day marched in triumph with their heads through all Trebizond.
7331 On 13 December we went to Chalybia with the Emperor, to the home fortress (6omrrtT6
Kcxaorpov) of Xacvupptl [the hact emir], son of Haipulrs [Bayram]-or rather, he joined
us after reaching Kerasous [Giresun]. We went to Kerasous from Chalybia by land, and
the Emir Xa-rTLpptlS and the Turks followed us in an almost servile manner (plKpo' ESiv
SouAtlKcO): this was in XV, 6870 [1361].
743 In October, XV, 6870 [1361], 'AXXi AlvrrwdK [ahi Ayna Beg] descended from Erzigka
[Erzincan] and besieged Golachas [? Kolasa] castle for sixteen days with siege engines and
violent petards. But without God nothing can be achieved, and he retired shamefaced
and empty-handed ...
7412 (In March, 1362, the Imperial Family retired to Mesochaldia [Kouazi, Kovans] because
of the plague.) On its return from Chaldia the Imperial Family did not enter Trebizond
and the citadel, because of the raging violence of the Death-for it was June--but it
camped in Saint John the Sanctifier on Mount Minthron [Boz Tepe]; and then an envoy
came from the TLoalcxhari TcXrcrbarivTlr [9elebi Taceddin], seeking a marriage alliance. But
some people came almost to rebelling against the Emperor.
754 ... on 15 August [1363], we went to the aid of [the Emperor's] brother-in-law, XouTrhouwkn1
[Kutlug beg], son of ToupAhiS [Turali], but we did not help because the Turks got infected
133 It has hitherto been thought that this Eudokia was the widow of the Gazi felebi of Sinope
and subsequently took the veil as the nun Euphemia in the Theoskepastos, Trebizond; considerable
literature, based upon misunderstandings by Zambaur and Fallmerayer, has accumulated on th
subject. The problem will be discussed at length in Bryer's and Winfield's work (supra, note 6). Briefly
Eudokia cannot have been the widow of the Gazi Velebi, who died between 1324 and 1332 (his tomb
in Sinope bears no date), but may have been married to the Candarid, Adil Beg ibn Yakub (1345- ?),
who is known from coin inscriptions. Eudokia-Euphemia was a nun of St. Gregory of Nyssa, Trebizond
not of the Theoskepastos, and is most unlikely to have been a Grand Komnene. Eudokia's career is
therefore obscure before 1357, and unknown thereafter.
134 KayonS 'OAcrroq;ps, in N. Banescu, "Quelques morceaux in6dits d'Andr6as Libad6nus," Byzantis,
2 (1911-12), 390.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
146 ANTHONY BRYER
with the Sudden Death,135 and we marched

days.
761 On 14 July, III, 6873 [1365], the Emperor's brother-in-law, the emir Xov-rhou'rEKrg
[Kutlu' beg], visited this Fortunate City of Trebizond with his wife, kyra Maria 8Eo'rrotv6acx-r
[despoina hatun], the Grand Komnene, and met the Emperor and was received into the
palace, and after encamping his tents around Saint John the Sanctifier [Boz Tepe] for
some eight days, and being paid great honors, he departed in peace again.
767 In the following year the Emperor made an expedition to the summer pastures (HcapXaptv).
And we-being more than two thousand foot and cavalry altogether-marched from Spelia
[Ispela, now Ocakli, at 850 m., 3 km. east of Fikanoy yayla, 3 km. west of Hava, 7 km.
west of Maqka] to Phiano? [Fikanoy yayla, at 1919 m., 20 km. south of Akqaabat, 10 km.
west of Maqka] with him and on past Gantopedin and Marmara, we went by Saint Mer-
kourios to Achantakas ;3 and after spending four days with the emir, we returned in June,
V, 6875 [1367] ....
7617 [Later in summer, 1367] ... the Emperor went up into the summer pastures of Larachan
[Akarsu, at 1250 m.], to Limnion [? near the Kolatdag Pass, at 2400 m.], and then went
off as far as Chaldia.
7624 . . the raid which the &lrriTKa [azap, sipahi]137 boats made on the Araniotai [Ares Island,
Puga, Giresun Adasi] [before 19 July, 1368] ...
7631 In March of the same year 6876 [1368], Ri'TLiaac9Sxavr [Kili Arslan] invaded and besieged
the places belonging to us in Chaldia, and the Emperor then marched up there at the head
of his army.
773 On Epiphany, January, VII, 6877 [1369], Golachas [? Kolasa] was deceitfully captured
by the Turks, and because of this Chaldia was devastated; some of its inhabitants perished
in the fighting and some in the treacherous cave there.
777 Around the end of January of the same year, 6877 [1369], the Emperor sailed to Limnia
with a fine fleet, and returned after spending a period of four months.
7710 In May, VIII, 6878 [1370], the Emperor set out to the summer pastures (VapXapiv) in the
district of Marmara, with some few men. And on Tuesday the 21st of the same month,
they suddenly came upon a force of Turks-some five hundred cavalry and three hundred
135 D. M. Nicol, The Last Centuries of Byzantium, 1261-1453 (London, 1972), 225.
136 In the map attached to H. Kiepert, "Der Verbreitung der griechischen Sprache im Pontischen
Kiistengebirge," Zeitschrift der Gesellschaft fiir Erdkunde zu Berlin, 25 (4) (1890), 317-30, Phiande
pastures are placed too close to the western ridge of the Matzouka-Prytanis valley, and Gantopedin
is apparently equated with the Lavra castle (and fortified chapel) of Vazelon, near Hava (Papadopoulos-
Kerameus, FHIT, I, 119), although it should also be a pasture. The route of this expedition is impor-
tant because, if it can be worked out beyond Fikanoy yayla, it should lead to Kutlu' beg's camp.
One possibility is that Gantopedin represents the Haska yayla, at 2150 m., that Marmara is the yayla
of Mimera (now Erikli, attested in 1371 and 1432), that Saint Merkourios is in Trikomia (now the
Kalenima Dere) and that Achantakas (which is not the chandax of Saint Kerykos) is the Achantakas
attested in 1429 (Ahanda, now Kavakh, near the coast, 2 km. northwest of Akgaabat). This makes a
round tour of about 40 km., but leads straight into the densely populated Trapezuntine coastal
bandon of Trikomia, and hardly to an Akkoyunlu camp. It is more likely that from Fikanoy Alexios
followed the yayla watershed at the heads of the valleys onward past Tonya and the Horosdag,
southwest to Kiirtiin, Suma Kale, and the Harsit-Philabonites valley. From the pastures of Phianoe
this would have entailed a round trip of about 80 km. It would have brought him into the known
lands of the Cepni, a principal Akkoyunlu federate.
137 The term is usually taken to indicate specifically Ottoman pirates and hence the first Trapezuntine
encounter with the Osmanhs (in, e.g., Miller, Trebizond, 66), but at this date is much more likely
to represent Sinopitan or local Tiirkmen corsairs. In what is perhaps his most egregious flight of
scholarship, Gr6goire erected a Greco-Turkish emirate of Ares, but his proposed state (which would
have extended the entire 180 x 150 m. of the walled island) must, sadly, be dismissed, if only on
epigraphic grounds. H. Gr6goire, "Notes 6pigraphiques. X, Michel Comnene, 6mir d'Arane," Revue
de l'instruction publique en Belgique, 52 (1909), 12-17; F. Cumont, "Notes sur une inscription d'Ico-
nium," BZ, 4 (1895), 101-5; P. Wittek, "L'6pitaphe d'un ComnBne a Konia," Byzantion, 10 (1935),
505-15; and 12 (1937), 206-11; Cahen, Pre-Ottoman Turkey (note 1 supra), 210; Fallmerayer, Original-
Fragmente, pt. 2, p. 91.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
foot. The Emperor's following was about a hundred

battle and gained the victory by force of arms and p
some Hagarenic heads and their banner.
7729 On 13 January [1373], the Emperor set out again
forces were routed when a great deal of snow had fallen
them; one hundred and forty Christians were killed, som
more than half-died of the cold. ...
788 On Sunday 16 April, XII, 6882 [1374], Golachas [? Kolasa] was taken by the Chal
and restored to the Emperor's dominion; but the enemy recaptured it immediately.
7834 After much discussion and exchange of embassies between the Romaioi and the Mou
manoi-I mean between the Emperor and TaTrbaTrivl rTLicahalrs [Taceddin 9elebi]-t
Emperor set out on 14 August, II [1379], with two great warships and two boats, with
daughter, kyra Eudokia. And we sailed as far as Kerasous [Giresun]. But a message arr
from Trebizond that XXlaroLoaacSDxvfl [Kiliq Arslan] was about to invade the dist
around Trebizond. The Emperor left his daughter in Kerasous and returned to Trebiz
with the nobles and he strengthened the citadel and set the country in order. Around
end of September he set out and, collecting his daughter in Kerasous, went as far as Oinaio
where he met the 9elebi, and on 8 October, III, 6888 [1379], he betrothed his daugh
kyra Eudokia, to him; and at this time the Emperor took control of Limnia.
7912 In February the Emperor set out against the TlawrrviSES [Cepni] by land and sea.
about Sunday 4 March, III, 6888 [1380], he divided his army into two sections. He s
some six hundred footsoldiers away from Petroma [Bedrama Kale, on the east bank of
Harsit, 5 km. from the coast], and the Emperor himself took command of the cavalry
another very large party of footsoldiers, crossed the country of those who live all a
the upper course of the Philabonites River [Harsit Dere] up to their winter camps [XEqtlac
and he destroyed their tents and he slaughtered and he burnt and he set free many captiv
of ours, liberating t(pA'iKM'a [?],138 and he turned back and halted for a short whil
XSxapooTidCo'rs [? the coastal Biiytik Liman of modern Vakfikebir]. The six hundred
had set out from Petroma made a raid into Kotzauta [apparently around Kiirtiin and S
Kale] and massacred and ravaged and burned; and whenever they came to grips with
Turks who were pursuing them, as they fought their way down to the coast, many of
Turks fell. The Romaioi expected to meet the Emperor, but reached the shore first, fighti
hard and slaying as they went. When they got to Sthlabopiastes beach and did not find th
Emperor there, as they had arranged, they were a little inclined to let themselves be push
138 This is the only word in Panaretos which has presented serious difficulty. It is unaccentua
in Cod. Marc. gr. 608, fol. 309r, top. There are three possibilities: (1) that a&clptKa is intended,
Cod. Marc. gr. 608, fol. 309r, which is very legible, does not say so; (2) that it is a latinizing 'si
taneously,' which would fit grammatically, but not phonetically, and is an otherwise unatteste
neologism; and (3) that it is a place name. On geographical grounds I favor the yayla and now r
village of ~1ptKAj or 1CprKA ~ (now Siimiiklii), attested as Christian in 1717 and 1733 (when it
associated with Kiirtiin), and in 1779 (when Chrysanthos, its parish priest, venerated the Gosp
of Sebinkarahisar). The place lies 44 km. northeast of Sebinkarahisar and 9 km. southwest of Kiir
and the Harsit Dere and, therefore, fits as the extreme limit of Alexios' expedition of 1380 an
possible Greek settlement in the area then. Settlement need not have been continuous between
and 1717, for Simikle may have been recolonized from the Tsite (Cit) valley during the flight to
highlands of the late 17th century. One Tsite village was traditionally linked with Platana (Akg
bat) on the coast; elsewhere I have assumed that this entailed an 80 km. transhumance via Trebizo
but the Harsit valley and Simikle would have afforded a more convenient passage. I. T. Pampou
rovTriaK&, 2. ToO TCrpoTooEpaoG-o1. Kai TrpCTOVOvoTapiou MtiXa 7TOiO n'avapTrovu, Trrpi TC"v TiTs Tpa-rrlovo~vro
p3aaiMcov, T-rcv pEy6?Aov Kopvrlv~cv, 6Trcos Kai -r6Tr Kai -rW6ov iKao-roS P3aalaEav (1204-1426) (Athens,
1947), 25; Panaretos, ed. Lampsides, 44, 93; A. Papadopoulos-Kerameus, 'Ispoaovpl-rtKI Biio$19KrI,
IV (St. Petersburg, 1899; Brussels, 1963), 299; N. Bees, 'AguplEap-raT Kai EI-TOpylKaI auvvpolpai IovTricov
0wIrrp Toi0 0lavayiov T&~qov, in 'ApX.rH6vr., 14 (1949), 136; Bryer, Isaac, and Winfield, "Nineteenth-
century monuments," pt. 4, 177-78, 240; E. C. Colwell, The Four Gospels of Karahisar, I (Chicago,
1936), colophon no. 41; S. Papadopoulos, Atopscbi~S E -rT TparrLovv-rtaK0v XpoviK6v MitXat\h cavaptpov,
BNJbb, 6 (1928), 19; P. I. Melanophrydes, T6 KlovpTro0v, in HovrtaK& cI0A7a, 2 (12) (1937), 6-7; Fall-
merayer, Original-Fragmente, pt. 2, p. 99.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
148 ANTHONY BRYER
back and some forty-two Romaioi fell. Th

numbered over one hundred.139
8012 On Wednesday 24 October, X, 6895 [1386], TaTrLarivis [Taceddin] emir of Limnia [?ar-
samba district], son-in-law of the Emperor, moved against the other son-in-law of the
Emperor, son of Xa-rTLuvpplt [the hacs emir], called loUapa'XP''EK [Siileyman beg],140 of
Chalybia, with an army of twelve thousand men, invading Chalybia, when Taceddin was
the first to fall and he was cut to pieces and died there; and some three thousand of his
followers were slaughtered and the rest fled, abandoning their equipment-they lost seven
thousand horses and countless arms.
819 On Saturday 4 September, IV, 6904 [1395], the despoina kyra Eudokia141 the Grand Kom-
nene arrived from Constantinople at Saint Phokas [Kordyle, now Akgakale] .... She
brought with her some brides .... And they entered Trebizond on Sunday, the following
day, in a shower of rain ...
139 The geography of the campaign will be discussed in the forthcoming study by Bryer and Win-
field. Some identifications depend upon those of a somewhat similar campaign fought over the same
area by ?eyh Ciineyd of Erdebil and the Grand Komnenos John IV in the 1430s: Chalkokondyles,
Bonn ed., 464; W. Hinz, Uzun Hasan ve Seyh Ciineyd (Ankara, 1948), 19-22.
140 There are three problems here: (1) Janssens, following Miller, names XacrrLvpCpti as Haci Omar,
but -pvpts is not a Byzantine form for Omar and I agree with Fallmerayer and Izeddin that Xa-rLv~Cpt1S
is a name-title, hact emir; (2) Panaretos can be read to mean that Siileyman was the name of either
the hact emir, or of his son; Lampsides is alone in apparently assuming the former, which would
imply that the usually meticulous Panaretos gave a name to the hact emir only upon mentioning
him for the fifth time, leaving his son nameless; so I agree with all other commentators in regarding
the Siileyman who attacked Giresun in 1398/97 as son of the hact emir; and (3) there is the question
of Alexios III's relationship with this Stileyman. He and Taceddin are separately described as yappCp6s
of the Emperor. In Panaretos, ed. Lampsides, 754, y0Cxrrrp6s is used as 'brother-in-law,' but in all
other ascertainable cases (including that of Taceddin, with whom Siileyman is coupled), the chronicler
uses it as 'son-in-law.' The hact emir was indeed brother-in-law of Alexios III, and yapplp6s here might
only describe his son's indirect relationship with the Emperor. But Panaretos has more elaborate
(if rather vague) terms to describe relationships like this: e.g., 618 for what is actually 'first cousin
once removed on the father's side,' and 7817 for what is actually 'great-uncle of a sister-in-law.' So
I infer that Alexios III married an unknown daughter to emir Stileyman of Chalybia before 1386.
Janssens, Trdbizonde, 116; Miller, Trebizond, 60 (a daughter of Alexios III "marrying Hadji-Omar's
son, Suleiman Bey ..."); Panaretos, ed. Lampsides, 86; Moravcsik, Byzantinoturcica, II, 216-17,
285, 343 (where no transliteration is offered); Fallmerayer, Trapezunt (supra, note 15), 196; idem,
Original-Fragmente, pt. 2, p. 67; Stimer, Oguzlar, 328; Izeddin, "Notes sur les mariages princiers" (supra,
note 76), 156 note 72; C. Toumanoff, "On the Relationship between the Founder of the Empire
of Trebizond and the Georgian Queen Thamara," Speculum, 15 (1940), 299; Nicol, Kantakouzenos
(supra, note 85), 143-46.
141 Apparently the widow of the emir of Limnia. Eudokia's career may be reconstructed thus.
Her parents, Alexios III and Theodora Kantakouzene, were born on 5 October 1337 and ca. 1340
respectively. They were married after 8 October 1349 and had no marriageable offspring until after
29 August 1358 (or even June 1362, when part of the objection to Taceddin's demand for a bride may
have been that only an infant was available). It was an age of uncanonically early state marriages:
Eudokia's sister, Anna, was married to Bagrat V (VI) in June 1367 at the age of ten years and two
months, and the extreme case is that of Simonis, daughter of Andronikos II, married to Stephen
Uro' II Milyutin of Serbia in 1299 at the age of five. It is, therefore, impossible to say how old Eudokia
was when she married Taceddin of Limnia on 8 October 1379. After the death of her first husband on
24 October 1386 the opinionated, spirited, gossipy but unreliable Trapezuntine interpolator of Chalko-
kondyles maintains that she was intended for Manuel II Palaiologos of Constantinople, but that
his father, John V (d. 1391), took her instead. In fact Sphrantzes reveals that she probably married
Constantine Draga' (Dejanovi'), grandfather of the last emperor of Constantinople, Constantine XI,
who died in 1395, and cites the case as a precedent for a Byzantine ruler marrying a former member
of a Turkish harem. After her second widowhood Eudokia seems to have retired to Trebizond. Pana-
retos, ed. Lampsides, passim; Angeliki E. Laiou, Constantinople and the Latins. The Foreign Policy of
Andronicus II, 1282-1328 (Cambridge, Mass., 1972), 96; Chalkokondyles, Bonn ed., 81-82, ed. Dark6,
I, 75-76; G. Th. Zoras, XpovtK6V -TElp Trc To'pKcoV OVATd&VCov (Athens, 1958), 35; G. Sphrantzes,
Memorii, 1401-1477, ed. V. Grecu (Bucharest, 1966), 80, 358; R.-J. Loenertz, "Une erreur singulibre
de Laonic Chalcocandyle. Le pr6tendu second mariage de Jean V PalBologue," REB, 15 (1957),
176-81; V. Grecu, "Zu den Interpolationen in Geschichtswerk des Laonikos Chalkokondyles,"
BSHAcRoum, 27 (1946), 92-94; A. Bryer, "Pisanello and the Princess of Trebizond," Apollo, 76
(1962), 601-3; 0. Lampsides, IJpptlCKTa i T-r XpovivKMV MiXCA FlTTavapk-rov, in 'ApX.TT6vr., 23 (1959),
46-48; J. W. Barker, Manuel II Palaeologus (1391-1425). A Study in Late Byzantine Statesmanship
(New Brunswick, N. J., 1969), 474-77.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
GREEKS AND TURKMENS
NOTES TO APPENDIX II
142 OSMAN KARA ILtK ('Yuluk' or 'Siiliik,' the 'Black Leech') = Na, d. of Alexios III. Ducas,
Istoria Turco-Bizantind, ed. Grecu (supra, note 80), 163, 165.
143 MUTAHHARTEN (or Taharten) = Na, d. of either Alexios III or (less likely) Manuel III.
The third of three proposed matrimonial alliances between Trebizond and Erzincan, and the only
one certain enough to include in the genealogy. They are: (1) An ulu hatun and presbyterissa, who
died on 28 December 1342, recorded on a ceramic Greek epitaph from Erzincan. Cumont proposed
that she was "sans doute une princesse du sang des Comnbnes ou la fille de quelque grande famille
de leur entourage." The notion is attractive, for she would perhaps be a daughter of Alexios II, who
seems to have initiated the Trapezuntine sequence of marriage alliances with Muslim neighbors,
but it is curious that, if she were a Komnene, her name was not included on the epitaph-as was
the famous Komnenos of Konya. Perhaps it was on a second and now lost tile. (2) Unaware of this
inscription and of Mutahharten's alliance, Izeddin concludes from Brosset's summary translation of
Panaretos that Alexios III married a daughter to an emir of Erzincan, probably the ahi Ayna beg,
who invaded Trebizond in 1348 and 1361. There is no evidence for the supposition, nor can the ulu
hatun who died in 1342 be a wife of ahi Ayna beg, for she evidently outlived her husband to become
a presbyterissa (? wife of a priest, ? deaconess) in the Orthodox Church of Erzincan, which was then
undergoing travails), although how she managed to change her status is a matter of speculation, in
which the possibility that she had, all along, been the ulu hatun of a Greek priest, should not be ruled
out. (3) The ahi Ayna beg was succeeded in Erzincan by Mutahharten, under Burhan al-Din (poet,
kadz, and self-proclaimed Sultan of Eretna in 1381, who may, or may not, have died at the hands of
the Karakoyunlu in 1398). Sultan Bayezid captured and released Mutahharten's Trapezuntine wife
at his siege of Erzincan in 1401. Mutahharten was restored by Timur, but the Muslims of the largely
gavur city (as Ibn Battuta pointed out, its Armenian bishop enjoyed more than local influence) took
the opportunity to complain that "prince Taharten was even more favourable to [Christians] than
he was to them, seeing that he specially cherished the Christians. Further they asserted that the
Christian churches were superior in size to their mosques. On this Timur, it is said, sent for Taharten,
reporting to him the complaint of the Moslems, and in reply Taharten agreed that it was true he
especially favoured the Christians throughout his country for they brought him, he said, wealth by
their trade. For an answer Timur then gave an order, commanding that a certain priest of the Greek
church, who was regarded as a chief among the Christians, should be brought to him. Then by reason
of the hatred in which he, Timur, held the Greeks of Constantinople, as also the Genoese of Pera,
he forthwith ordered that this priest of the Christians should immediately change his faith and be-
come a Moslem, but to do this that man would by no means consent. Timur on the spot gave orders
that all the Christian population of Arzinjin should be put to the sword, but prince Taharten now
besought him to have mercy, offering if Timur would spare their lives to bring him a gift of 9,000
aspers. ... This sum Taharten having paid on behalf of the citizens Timur relented, but none the
less ordered
ruined castle that all the Christian
overlooking churches
Erzincan from be demolished
the muddy ... ."Not
Euphrates. Timur took possession
surprisingly, of theto
there seems now
have
been a continuous state of tension, economic and religious, in this Christian majority caravan city,
serving a Muslim majority area (which, on a smaller scale, was probably reflected in Bayburt). In
1314 three Roman Catholic missionaries were martyred in Erzincan maydan and in 1316-19 its Greek
bishop (not subject to Trebizond) complained of the dire state of his establishment. But for Christian
artisans and entrepreneurs there, things seem to have bettered with Mutahharten's rule. He died
before 1404 and "left no legitimate male issue to succeed him, and his wife was the daughter of the
Emperor of Trebizond." Clavijo, trans. Le Strange, 125, 130 ('Zaratan'); Askipasazade, in Kreutel,
Vom Hirtenzelt zur Hohen Pforte (supra, note 68), 107, 109, 113; Nesgr, eds. Unat and K6ymen (supra,
note 58), I, 334-35; Ibn Battuta, trans. Gibb (supra, note 52), II, 437; Cahen, Pre-Ottoman Turkey,
231, 286, 362-63; Golubovich, Biblioteca (supra, note 27), II, 64-68, 544; III, 183-84; M. Bihl, "De
duabus epistolis fratrum minorum Tartariae Aquilonaris An. 1323," AFrH, 16 (1923), 90; Miklosich
and Miiller, Acta et diplomata, I, 83; Izeddin, "Notes sur les mariages princiers" (supra, note 76), 151;
A. A. M. Bryer, "The fate of George Komnenos, ruler of Trebizond (1266-1280)," BZ, 66 (1973),
334 note 9, 347 note 76; F. Cumont, "Inscription de l'6poque des Comn6nes de Trebizonde," Mdlanges
Pirenne, I (Paris, 1926), 67-72; Giesecke, Quelle (= Bazm u Razm), 45-49, 66-69, 94-111.
144 ALI (ULU) BEG = Na, d. of Alexios IV. This marriage has a long and respectable history in
secondary sources, ending with Nicol, Kantakouzenos, 169. I have a ten-year-old note that it begins with
"D'Herbelot, 893," but Barthdlemrny d'Herbelot, Bibliothkque Orientale, ou Dictionaire Universel con-
tenant gdndralement tout ce qui regarde la connoissance des Peuples de l'Orient, 893, in the editions of
Paris, 1697; Maestricht, 1776; La Haye, 1777-79; and Paris, 1781, does not contain it. More to the
point, I am unable to find an original source to justify this oft-cited marriage, other than Chalkokondy-
les, Bonn ed., 462, which is far from explicit. The alliance must therefore be regarded as dubious.
Uzun Hasan's actual mother was, in any event, the Armenian Sara hatun: Asikpasazade, trans.
Kreutel, 225.
145 CIHANSAH = Na, d. of Alexios IV. Chalkokondyles, Bonn ed., 462.

5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00
AKKOY UNLU
Bayundur
TURALI
(fl. 1352)
FAHREDDIN KUTLUG BEG

(fl. 1352-67)
11
OSMAN KARA ILtYK -- Na4
(d. 1435)
ALl (ULU) BEG = ? Nal44

(1447-49)
UZUN HASAN = THEODOR

(1449-78) buried St. (
YAKUB
(d. 1478)
146 THEODORA = UZUN HASAN and I

in 1458, perhaps in Diyarbekir (her br
Harput, retired, after the debacle fol
buried in the Jacobite church of St. G
tine-Muslim marriage, in which Theo
the only alliance in which the terms c
docia' (as was rashly made out to Mehm
of Halanik and Sesera in the banda of Matz
equivalent of a 'bride-price' in that Uzun
which he failed to do in 1461; (3) Theodo
tine household, and act as protector of l
able (and probably unusual) right to infl
century Italian visitors describe Theodora
woman of that time, and throughout Pe
the condition that she might remain in
lady in the world, always remained a go
manner, which she attended with much dev
as ordered by our true religion" (technic
her Italian visitors); "there also came witl
high condition, who were always to remain
Greeks and many calogieri-the latter ke
that a niece and a nephew (George, b. 14
Muslim, recanted, and was given a wife by 'G
The story is dubious, but in 1463 Theodorn
nephew of David, last Grand Komnenos, t
Uzun Hasan's diplomatic overtures to Veni
in 1474-there were other connections with
its patron saint, John the New, and Ste
This content downloaded from who had married David Komnenos in 142(
5.27.45.11 on Sat, 13 May 2023 11:24:34 +00:00 foreign affairs with one Italian visitor, "
APPENDIX 11: (GENEALOGY OF THE MUSLIM MARRIAGES OF THE
References are to those marriages not
AKKOY UNLU CHALYBIA TREBIZOND

Bayundur Emiroglu Grand Komnenos
ALEXIOS II GK
(1297-1330)
BASIL GK
(1332-40)
TURALI BAYRAM
(fl. 1352) (fl. 1313-32)
FAHREDDIN KUTLUG BEG = MARIA HACI EMIR = THEODORA ALEXIOS III GK
(fl. 1352-67) (fl. 1357-61), (1349-90)
OSMAN KARA ILtYK - Na142 SULEYMAN = Na MUTAHHARTEN -- Na143 MANUEL
(d. 1435) (fl. 1386-96) Emir of Erzincan (1390-1417
(fl. 1381-1401)
ALEXIOS IV GK
(1417-29)
ALI (ULU) BEG = ? Na44 JOHN IV GK

(1447-49) (1429-58/60)
UZUN HASAN = THEODORA despoina hatun CtNEYD of Erdebil

(1449-78) buried St. George, Diyarbekir146 invader of Trebizond
1430s; d. 1460
St
YAKUB I = HAYDAR
'ELIEL' 'EZIEL' MAR(T)A I (
(d. 1478) Pontic Greek-speaking daughters. (1460-88) fi
Fled Harput, 1478, for Damascus, the Safavid
where still living in 1512.
146 THEODORA = UZUN HASAN and their offspring. Theodora w

1458, perhaps in Diyarbekir (her brother David acting as nymph
,rput, retired, after the debacle following Uzun Hasan's death
ried in the Jacobite church of St. George (Mar Jurjis) there. This
e-Muslim marriage, in which Theodora became the ulu hatun
. only alliance in which the terms can be reconstructed. They incl
cia' (as was rashly made out to Mehmed II), but of imperial demesn
Halanik and Sesera in the banda of Matzouka and Sourmaina, as O
livalent of a 'bride-price' in that Uzun Hasan made some sort of un
ich he failed to do in 1461; (3) Theodora's right to remain Christia
e household, and act as protector of local Christians; and (4), as it
Le (and probably unusual) right to influence Akkoyunlu foreign
itury Italian visitors describe Theodora as "very beautiful, being
man of that time, and throughout Persia was spread the fame of
- condition that she might remain in the Christian faith"; "this
ly in the world, always remained a good Christian and every day
.nner, which she attended with much devotion"; she kept "a chapl
ordered by our true religion" (technically, she and her father w
Italian visitors); "there also came with her many young maid
rh condition, who were always to remain with her"; Harput was "
eeks and many calogieri-the latter keep company with the s
Lt a niece and a nephew (George, b. 1461) of Theodora joined
Lslim, recanted, and was given a wife by 'Giurgiubei' (perhaps Con
e story is dubious, but in 1463 Theodora indeed made a disast
phew of David, last Grand Komnenos, to bring up at her court. S
un Hasan's diplomatic overtures to Venice in 1465-66 and to S
1474-there were other connections with Moldavia, for Trebizond
patron saint, John the New, and Stephen was married to Maria
.o had married David Komnenos in
This content downloaded from1426. Theodora and Uzun
eign affairs with 5.27.45.11
one Italian visitor,
on Sat, 13 May 2023 "even when both Their Maie
11:24:34 +00:00
THE MUSLIM MARRIAGES OF THE 'RINCESSES OF I REBIZOND
are to those marriages not explicit in Appendix I
TREBIZOND LIMNIA KARAKOYUNLU

Grand Komnenos Taceddinoglu
ALEXIOS II GK
(1297-1330)
BASIL GK EUDOKIA = ?ADIL BEG IBN YAKUB

(1332-40) Despoina of Sinope (1345-? 1357)
,ODORA ALEXIOS III GK BAYRAM HOCA

(1349-90) (d. 1380)
MIUTAHHARTEN - Na14
Emir of Erzincan (1390-1417) (fl. 1367-86) (1) (d. 1390)
(fl. 1381-1401) (2)
- ? KONSTAN
(d. 1395)
ALEXIOS IV GK ALTAMUR ( ? I)
(1417-29) (fl. 1386-1404)
? ALTAMUR (? II) KARA YUSUF

Grand Mesazon, 1461 (1388-1420)1
SI I
JOHN IV GK DAVID GK Na145 = CIHANSAH
(1429-58/60) (1458/60-61) (1435-67)
killed by Uzun Hasan
CtNEYD of Erdebil, ANNA147 = ZAGAN PASA

invader of Trebizond,
1430s; d. 1460
MAR(T)A = HAYDAR of Erdebil

(1460-88) from whom
the Safavids of Iran
think any other Mahommedan or Christian king ever granted, even to their neare
tomb in St. George, Diyarbekir (ruined in 1883 and now lost) was shown to an Italian
"meanly buried under a portico near the door of the church in the earth, and above
like a box, one cubit high and one wide and about three in length, built of bricks an
Yakub is apparently not identical with the Yakub who succeeded Uzun Hasan (
probably killed by one of Uzun Hasan's Muslim wives in 1478. In 1500 there was a
to recognize another son as pretender to Trebizond. A daughter, Mar(t)a, became m
conqueror of the Akkoyunlu and (from 1501) first Safavid Shah of Persia. Her two un
found their way to Damascus, where one Italian wrote: "I myself have often conve
in Trebizond Greek, which they learnt from Queen Despinacaton, their mother." H
met them in 1512; they recognized each other as relatives and spoke of their dynast
Trebizond over half a century before. Chalkokondyles, Bonn ed., 490, 497; Doukas,
F. Babinger, Mahomet II le Conqudrant et son temps, 1432-1481 (Paris, 1954), 229;
zon" (supra, note 22), 317, 320; Caterino Zeno, Iosafat Barbaro, Antonio Contarini,
in Persia,' Travels to Tana and Persia, Hakluyt Society, 1st ser., 49 (London, 18
18, 41 note 4, 43, 74 note 1, 146 note 6, 173-74, 178-79, 183; Iosafat Barbaro, Viagg
(supra, note 26), 48V; V. Laurent, "Le Vaticanus latinus 4789," REB, 9 (1951),
"Catherine ou Theodora ?", BZ, 22 (1913), 88-89; idem, Dans l'Orient Byzantin (Pari
(= "La Princesse de Trebizonde," Revue de Paris, 1 October 1912); V. Minorsky,
sidcle entre la Turquie et Venise," Orientalia Romana, 1 (1958), 99-117; Jorga, Notes
note 26), V (Paris-Bucharest, 1916), 328; Hinz, Uzun Hasan, 29-30, 57, 63, 80
Kur'anskis, "Une alliance probl6matique au XVe siecle. Le mariage de Valenza C
empereur de Trebizonde, a Niccolo Crispo, Seigneur de Santorin," 'ApX.T6vr.,
N. Iorga, Studii istorice asupra Chiliei si Cetitii-Albe (Bucharest, 1899), 121; A
Goths in the Crimea (Cambridge, Mass., 1936), 282; P. Nasturel, "Une pr6tendue oe
Tsamblak: 'Le Martyre de Saint Jean le Nouveau'," Actes du Premier Congrds
Etudes Balkaniques et Sud-Est Europdennes, VII (Sofia, 1971), 345-51.
147 ANNA = ZAGAN PASA. Chalkokondyles, Bonn ed., 527; Babinger, Mah
legend, she fled to the Thismountains south
content downloaded fromof Trebizond. A. Bryer, "Trebizond and Se
27 (1966), 5.27.45.11
38 noteon27; Bryer,
Sat, 13 May 2023Isaac, and
11:24:34 Winfield, "Nineteenth-century monuments
+00:00

234567

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

234567

Uploaded by

Copyright:

Available Formats

Greeks and Türkmens: The Pontic Exception

Author(s): Anthony Bryer

Stable URL: https://www.jstor.org/stable/1291371

This content downloaded from

This content downloaded from

The term 'Tiirkmens' defies definition; by it I intend no

This content downloaded from

This content downloaded from

of two great centralized and local A

* Ahrweiler, "Smyrne," 26-28, 137-38; D. M

This content downloaded from

side of the mountains,' a sort of Perateia, a

This content downloaded from

insidious threat of all, the classic

This content downloaded from

To the south of the Pontic Alps, the Armen

Son, in the place where you would go,

Between the summer pastures and Trebizond

13 This and items cited later are compiled from account

This content downloaded from

by forests .... The paths we had t

This content downloaded from

Date Choria Christians Muslims Christian %

This content downloaded from

impress the Akkoyunlu emir in 1

26 Two misconceptions, upon which de Planh

This content downloaded from

27 Duda, Ibn Bibi (supra, note 5), 64-68; A. Papadop(o)ulos-Kerameus, Fontes

This content downloaded from

agreeable cultural confusion.3s It m

This content downloaded from

Things began to change not so much aft

33 Hayton, Flos historiarum terre Orientis, in Recueil des

This content downloaded from

Fatih in 1461.39 To the southwe

39 G. Th. Kandilaptes, FEcoypa<PtK6v Kai I(Tr

This content downloaded from

Thus, the west Pontic-Canik Tiirkmen cent

This content downloaded from

principalities which surrounded it

This content downloaded from

This content downloaded from

4. A Winter Emirate: Limnia.

This content downloaded from

rd) Akkoguniu E)' lon

5 M jor Battle i4sa

This content downloaded from

should be at the ruined village of Taylikk

As they do today, the shepherds of the interior come down for

This content downloaded from

it as his as part of the eventual allian

This content downloaded from

she left Trebizond, Theodora had the cha

This content downloaded from

subject to Timur, held the town, i

This content downloaded from

Komnenoi found the ?epni difficult to dea

This content downloaded from

to have attacked from the immedia

This content downloaded from

"was watching from the palace and she w

This content downloaded from

Orthodox princesses would certain