The Qavars (Qabars) and their Role

in the Hungarian Tribal Federation

Sá n d o r Lá s z l ó Tóth

The seven Hungarian (Magyar) tribes called 'Hetumoger' (Seven Hungarians) had
been joined by a smaller foreign group, the so-called Qavars or Qabars before the
conquest of the Carpathian Basin at the end of the ninth century. Their history and
historical role will be examined in this study. I will use both names alternatively.
Only two sources preserved their name and some data concerning their histo­
ry. The earlier source is the Annals of Salzburg (Annales Iuvavenses Maximi), which
recorded at the year 881 an eclipse of the sun, then two wars (battles?) on the ter­
ritory of Ostmark (today: Austria) belonging to the Eastern Frankish Empire: "The
first war [was] with the Hungarians [cum Ungris] at Wenia. The second war [was]
with the Qabars [cum Cowans] at Culmite."1 This source referred to a raid of the
Hungarians (Magyars) and their ally, the Qabars/Qavars, who "fought on separa­
te fronts and in separate units" during the same campaign.2
The second source contains a brief and concise history "of the nation of the Ka-
baroi" in chapter 39 of De administrando imperio (henceforth: DAI) compiled from
different sources around 948-952 in the name and order of Emperor Constantine
VII (944-959).3 After the compulsory preliminary notes "it is worth knowing"4 this
chapter enumerates the following data: "The so-called Kabaroi were of the race of
the Chazars. Now, it fell out that a secession was made by them to their govern­
ment, and when a civil war broke out their first government prevailed, and some

1 "Sol obscuratus est a tercio usque ad sextam horam. Primum bellum cum Ungaris ad
Weniam, secundum bellum cum Cowaris ad Culmite." See for the Annals of Salzburg,
Continuatio Annalium luvavensium Maximorum, red. H. Bresslau, in Monumenta Ger-
maniae Historica. Scriptores XXX/2. Hannoverae-Lipsiae 1934, 742.; for the discovery
of this source see E. Klebel/'Eine neu aufgefundene Salzburger Geschichtsquelle,"
Mitteilungen der Gesellschaft fiir Salzburger Landeskunde, 1921; for its Hungarian trans­
lation see S. L. Tóth, A honfoglalás korának írott forrásai. [The Written Sources of the Age
of the Hungarian Conquest] Ed. Gy. Kristó. Szegedi Középkortörténeti Könyvtár 7.
Szeged 1995, (henceforth: HKÍF) 209; for the English rendering of the Latin text, see
Gy. Kristó, Hungarian History in the Ninth Century, Szeged 1996,150.
2 Kristó, Hungarian History, 150.
3 See for the analysis of De administrando imperio, J. B. Bury,"The treatise De administran­
do imperio." Byzantinische Zeitschrift 14 (1906), 517-577; Gy. Moravcsik, Byzantinoturci-
ca I— II, Berlin 1983, 2: 356-389; Constantine Porphyrogenitus, De administrando imperio.
Vol. 2, Commentary, ed. R. J. H. Jenkins, London 1962. (henceforth: Commentary)
4 Every new information is introduced by the phrases "it is worth knowing" (in Greek:
isteon oti or oti), cf. H. Grégoire, Le nom et l'origine de Hongrois. Zeitschrift der
Deutschen Morganlandischen Gesellschaft 91 (1937), 630-642.

3
 Sá n d o r Lá s z l ó Tó t h

of them were slain, but others escaped and came and settled with the Turks in the
land of the Pechenegs, and they made friends with one another, and were called
'Kabaroi'. And so to these Turks they taught also the tongue of the Chazars, and
to this day they have this same language, but they have also the other tongue of
the Turks. And because in wars they show themselves strongest and most valo­
rous of the eight clans, and are leaders in war, they have been promoted to be first
clans. There is one prince among them, I mean, among the three clans of the Kaba­
roi, who survives to this day."5 After this more detailed history of the Qavars, in
the next, 40th chapter, entitled "of the clans of the Kabaroi and the Turks" Empe­
ror Constantine briefly summarized the story of the Qabars, when he mentioned
that "the first is the aforesaid clan of the Kabaroi which split off the Chazars". Ha­
ving listed all the eight clans of the Qabars and Turks (Hungarians/Magyars) the
Byzantine emperor added, that "having thus combined with one another, the Ka­
baroi dwelt with the Turks in the land of the Pechenegs."6 Apart from these sour­
ces we have no direct data and evidence on the Qabars/Qavars, so we must base
our hypotheses and conclusions on the brief remark of the Annals of Salzburg and
the valuable compilation of Emperor Constantine, in whose work a whole chapter
is dedicated to the Qabar story.
The first problem to be solved is the exact form of the ethnonym and/or tribal
name Qabar and its origin and etymology (meaning). "The name Qabar is found
in Constantine Porphyrogenitus in the form Kalapot and Kafiápav (gen. pi.)."7 Em­
peror Constantine refers to these people with these phrases: 'so-called Kabaroi'
(Aeyópevoi KáfJapoi) or 'certain Kabaroi’ (Kafiapoi nvcg).8These vague expressions may
reflect uncertainty from the part of the Byzantine emperor or from the part of the
possible informers as regards to the meaning of the ethnonym.9 One further in­
stance of Qabar in the form of Kabeiroi (Kafeipot) might appear in Byzantine sourc­
es in connection with the revolt of Thomas the Slav against the Byzantine emper­
or Michael II (in 821-824).10This Kabeiroi ethnonym is on the list of Thomas allies,
most of these people enumerated were the inhabitants of the Caucasus region.
However, it seems very probable, that 'Kafleipoi' can be read 'Zafleipoi', so the real
allies of the Byzantines rebels must have been the Sabirs, not the Qabars, if any
people of the Caucasus played any role in this revolt at all.11 It is a radical, but from

5 For chapter 39 of DAI, see Constantine Porphyrogenitus, De administrando imperio.

Greek text edited by Gy. Moravcsik, English tr. R. J. H. Jenkins. Washington DC 1967,
(henceforth: DAI) 174-175; for an earlier English translation of the Greek text, see C.
A. Macartney, The Magyars in the Ninth Century. Cambridge 1930,231; for the summa­
ry interpretation of this chapter cf. Commentary, 149-150.
6 For chapter 40 on the Kabaroi, see DAI 174-175.; cf. Commentary 150.
7 P. B. Golden, Khazar Studies. A Historico-Philological Inquiry into the Origins of the
Khazars, 2 vols. Budapest, 1980,1:139.
8 See S. L. Tóth, "Kabarok (Kavarok) a 9. századi magyar törzsszövetségben." [Qabars
(Qavars) in the ninth century Hungarian tribal federation.] Századok 118 (1984), 105;
idem, herediától a Kárpát-medencéig. [From Levedia to the Carpathian Basin] Buda­
pest 2011,2 58.
9 Tóth, herediától, 58.
10 Theophanes Continuatus, Históriáé, ed. I. Bekker. Bonnae 1838. 55. cf. Golden, Khazar
Studies, 139, note 432.
11 Moravcsik, Byzantinoturcica II. 262-263; Golden, Khazar Studies, 139-140.

4
 T he Q a v a r s (Qa b a r s ) a n d t h e i r R o l e in t h e H u n g a r i a n . . .

linguistic and historical aspect a faulty hypothesis, which identified the Qabaroi
with the three Qangar tribes, belonging to the Pecheneg confederation.12 The Qa-
bar "ethnonym also figures in the Latin source (Annals of Salzburg) in the form:
Cowaris,"13 which can be rendered as Qovar or rather Qavar/Kavar.14 Hungari­
an place names, such as Kővár or Kovárc may also testify the Qovar/Kovar form
of the ethnonym of this dissident Khazar group.15 It is difficult to decide wheth­
er the Byzantine form, i.e. Qabar (Kábaroi) - which can reflect a Qavar ethnonym
as well - or the Latin version, i.e. Qovar/Qavar would better denote the ethnic
group, which joined the Hungarians. In historical research the Qabar form is more
frequently used,16 though both in earlier studies and mainly in Hungarian recent
research the Qavar ethnonym is preferred.17 According to a view, the "designa­
tion Khabar denoting the Khavars, like so many other names, is Hungarian histor­
ical convention. The Greek text reads as Khavar. Although in those days Khabar
might also have been transliterated by the Greeks in the same way, we learn from
the Latin written sources that the correct reading was indeed Khavar."18 The Qa­
bar form is considered the original Turkish form of the ethnonym, while the Latin
Qovar ('Cowaris') version "probably came from a Hungarian source,"19 as proved
by Hungarian place names. It is practically impossible to state, which of these ver­
sions, Qabar-Qavar or Qobar-Qovar, is used, which can be considered the original,
or rather with both the Qabar/Kabar and the Qavar/Kovar names this Khazar eth­
nic group were denoted.
The Qabar/Qavar name is clearly referred to an ethnic group, which in­
cludes three clans or rather tribes (yeveá), and was called by Constantine a na­
tion (éüvouq).20 So it must have been an ethnonym, not a tribal name, though the
two categories do not exclude each other. In the Annals o f Salzburg some peoples
were recorded, the Hungarians (Ungri) and the Qavars/Qovars. Similarly in DAI
two people are mentioned, the Turks (Tourkoi) and the Qabars (Kabaroi). In both
of these sources the Magyars (Hetumoger)21 are denoted by an ethnonym (Ungri,

12 Grégoire, Le nom, 637-640.

13 Golden, Khazar Studies, 139.
14 M. Gyóni, A magyar nyelv görög feljegyzéses szórványemlékei. [The sporadic relics of
Hungarian language survived in Greek sources], Budapest 1943, 61-62.
15 Kővár in Hont and Nyitra counties, Kovárc in the Nyitra valley, see Gy. Németh, A
honfoglaló magyarság kialakulása. [The Formation of the Landtaking Hungarians]
Budapest 1930, 235-236; Golden, Khazar Studies, 139.
16 See e.g. Kristó, Hungarian History, 150-152; Golden, Khazar Studies 139-141. etc.
17 Macartney, The Magyars, 112-124.; Gy. Györffy, "A kabar kérdés" [The Kabar Prob­
lem], in idem, A magyarság keleti elemei [The Oriental elements of the Hungarians]. Bu­
dapest 1990, 83-84; Tóth, Levediától 53-65; idem, A honfoglalástól az államalapításig. A
magyarság története a X. században. [From the Conquest to the Fundation of the State.
The history of the Magyars in the tenth century.] Budapest 2010, 136-140; A. Ró-
na-Tas-Á. Berta, West Old Turkic Loanwords in Hungarian, 2 vols. Wiesbaden 2011,1:35.
18 A. Róna-Tas, Hungarians and Europe in the Early Middle Ages. An Introduction to Early
Hungarian History. Budapest 1999, 347.
19 Golden, Khazar Studies, 139,140.
20 DAI 166-167.
21 Anonymus (P. dictus magister), the author of Gesta Ungarorum (written around
1208) used both names of the Magyars (Hungarii, Mogerii) for the people, and wrote
that "per ydioma alienigenarum Hungarii et in sua lingua propria Mogerii vocantur," and

5
 Sá n d o r Lá s z l ó Tó t h

Tourkoi), which is surely not their own name, but a name given by other people. So
it seems highly probable, that the Qabar/Qavar name was not the collective name
of the three tribes splitting off the Khazar Empire.22 However, it was supposed as
well, that the Qabar/Qavar ethnonym may have been their own name, reflecting
the name of one of their clans/tribes.23 It was assumed, that "one of the leading
clans of the Khavars must have been called Khavar, a name which the new group
consisting of three tribes commonly adopted."24 Turkish people were rather fre­
quently called by a combination of a numeral and and an ethnonym.25 The Hun­
garians/Magyars may have been denoted as 'Seven Magyars' (Hetumoger), there
was a hypothesis that the joining Khazar tribes were named 'Three Qabars/Qa-
vars' (in Hungarian: 'Haromkabar/Haromkavar').26 This assumption holds true
just in that case, if this ethnonym comes from a Hungarian source. The Hungari­
ans may have used this Three Qavars/Qovars denomination in connection with
the joining three tribes. The Hungarians may have called this dissident Khazar
group 'Khozar' as well referring to their former confederation, the Khazar Em­
pire.27 A late Hungarian source, Gesta Hungarorum written by Anonymus prob­
ably at the beginning of the thirteenth century mentioned that a territory in the
Carpathian Basin (between the Maros and Szamos rivers) was inhabited by people
called Cozar.28 Although the late Hungarian author could not be aware of the ex­
istence of a joining Khazar group, he may have preserved the memory of the Qa-
vars, called by the Hungarians/Magyars either Qovar or Khozar.
What could be the meaning of the ethnonym of the Qabars/Qavars? Most of
the researchers regarded it to be of Turkish origin.29 For a long time the most ac­
cepted theory was that the original meaning of the Qabar name was 'those who re­
volt,' 'rebel.'30 This etymology was criticized, and was emphasized, that qab- root
"exists only theoretically," so "etymologies based on this root are highly specu­
lative."31 Although in spite of criticism, the 'rebels' etymology was regarded "the

the 'Seven Magyars' for the leaders as well, "septem principles persone, qui Hetumoger
vocantur," cf. Scriptores rerum Hungaricarum tempore ducum regumque stirpis Ar-
padianae gestarum. Ed. E. Szentpétery. (henceforth: SRH) I. Budapestini 1937, 33,37.
22 Tóth, A kabarok, 106.
23 Tóth, Levediától 59.
24 Róna-Tas, Hungarians, 348.
25 With some examples as the Three Kharlukh (Üch Karlukh) or the Nine Oghuz
(Tokhuz Oghuz), Ten Oghur (On Oghur) etc. see Róna-Tas, Hungarians, 340-341.
26 See Gy. Kristó, Magyar honfoglalás, honfoglaló magyarok. [Hungarian Landtaking, Hun­
garians Settling in Homeland] Budapest 1996,57.
27 For the use of the earlier Khazar/Khozar name in case of the Qabars see Németh,
HMK 1930, 238.
28 SRH 1:49. "afluvioMorus usque adfluvium Zomus....et terram illám habitarentgentes qui
dicuntur Cozar."
29 Moravcsik, Byzantinoturcica II. 144; Commentary 149.
30 Á. Vámbéry, A magyarság keletkezése és gyarapodása. [The Formation and Development
of the Hungarians]. Budapest 1895.160. note 32; Németh, HMK 1930, 237.
31 See e. g. Golden, IGiazar Studies, 140.; accepted his criticism Gy. Németh, A honfoglaló
magyarság kialakulása. [The Formation of Landtaking Hungarians] Second, Enlar-
rged and Modified Edition, ed. Á. Berta. Budapest 1991, (henceforth: Németh, HMK
1991) 264.

6
 T he Q a v a r s (Qa b a r s ) a n d t h e i r R o l e in t h e H u n g a r i a n . ..

most likely" solution, because "this fits the actual events of their history."32 Later
the theory was modified a bit, and a Turkish qop- root was assumed ('to rise', 'to
revolt'), and the ethnonym Qovar would mean a command: 'rise, revolt.'33 In my
view this 'Rebel' etymology has some difficulties, and we cannot accept the pre­
conception, that the meaning of the ethnonym should reflect the history of peo­
ple, in this case the revolt of the Qabars/Qavars against the Khazar Empire. This
explanation of the ethnonym shows the Khazar aspect, the negative judgment on
the rebel tribes, who lost in the civil war and fled. From a historical perspective I
do not regard this etymology a likely interpretation. The Hungarians, the seven
Magyar tribes could not use this kind of ethnonym, and the joining three Khazar
tribes were probaly even less willing to accept such a 'nickname.'34 Another theory
assumes, that using a qap- root, Qabar "could then, conceivably be etymologyzed
as 'those who attack, the Attackers.'35 This etymology seems to me a more plau­
sible explanation, because the Qabars were characterized by Emperor Constan­
tine, that "in wars they show themselves strongest and most valorous of the eight
clans, and are leaders in war."36 This 'attackers' etymology fits their role inside the
Hungarian tribal federation, and may be derived from the joining Khazar tribes
as well.37 However, the likelihood of this etymology was judged less likely, than
the 'Rebel' explanation.38 According to a hypothesis the Qabar ethnonym or its
equivalent was given by the Hungarians to the newcomers and had originally the
meaning of 'nose', which expressed their foremost position in the battle order and
military system of the Hungarian tribes wearing similar names.39 The Qavar eth­
nonym was also deduced from the ethnonym Chwár, Chowár, people of Ch/w/
oresm, "a people of mixed stock and judaizing religion,"who joined the Chazars
after their expulsion from Choresm, and later revolted and joined the Hungarians,
and later appeared under the name Kaliz in the twelfth century.40 It was assumed
as well, that the Qavar ethnonym was of Hungarian origin, meaning 'mixture' (in
Hungarian: 'keverék') denoting that the dissident Qavars comprised different eth­
nic groups.41 According to a view, "if we were to find a related Turkic word, a pos­

32 Golden, Khazar Studies, 142.

33 Németh, HMK1991, 265.
34 Tóth, Levediától, 59.
35 Golden, Khazar Studies, 141.
36 DAI 174-175.
37 Tóth, A kabarok, 106; Tóth, Levediától 58; Kristó, Hungarian History, 150.
38 Golden, Khazar Studies, 140-141. - He preferred the criticized version ('Rebel') to his
own ('Attackers').
39 Á. Berta, "Magyarok a steppe országútján (Törzsek és törzsnevek)." [Hungarians on
the road of the steppe. Tribes and tribal names]. In: Árpád előtt és után. [Before Árpád
and after him] ed. Gy. Kristó-F. Makk. Szeged 1996,37.
40 H. Schönebaum, "Zur Kabarenfrage." in Aus der byzantinischen Arbeit der Deutschen
Demokratischen Republik. ed. I. Irmscher, Berlin 1957,142-146; for the identification of
Qavars with the Káliz, see M. Gyóni, Kalizok, kazárok, kabarok, magyarok. [Kaliz,
Khazars, Kabars, Hungarians] Magyar Nyelv 34 (1938), 86-96, 159-168; Commentary,
149.
41 Gy. Kristó-F. Makk-L. Szegfű, "Adatok korai helyneveink ismeretéhez," [Data to the
knowledge of early place names]. Acta Universitatis Szegediensis de Attila József nomina-
tae. Acta Historica 44 (1973), 12.

7
 Sá n d o r Lá s z l ó Tó t h

sibility would be kavir- 'to collect, assemble',42 and it is probable that the leading,
organizing clan's name was adopted.43 It is impossible to decide, which etymology
fits the best the Qabar/Qavar ethnonym.
It would be important to reveal the real cause of the Qabar revolt against the
Khazar government, since it can be considered the beginning of their own short
history. Our only source in this respect, Emperor Constantine just stated the facts,
that "a secession was made by them to their government, and when a civil war
broke out their first government prevailed, and some of them were slain, but others
escaped and settled with the Turks."44 Most frequently the secession and the civil
war were connected with religious strife in Khazaria, the conversion to Judaism in
historical research. It was supposed, that the "antagonism within the Khazar Em­
pire [...] was largely connected by the forcible conversion of the leading Khazar
strata to Judaism."45 According to a debated source, the letter of Qagan Joseph,
the first converted king was Bulan, but the faith was strengthened by his third
descendant, Obadiah.46 Although it is discussed, most scholars accepted, that di­
rectly or indirectly the conversion to Judaism or the activity of Obadiah to make
Judaism a kind of state religion could cause the breaking out of a civil war.47 It is a
generally assumed view, that not the more strict Karaim/Karaite strand (which ac­
cepted just the five books of Moses), but the orthodox Rabbinistic Judaism (Talmu­
dist) was adopted by the Khagan and the ruling circles of Khazaria.48 The Kazars'
conversion, which probably proceeded in more stages, was dated by the testimony
of later Hebrew sources at about 740-76049 or at 760-770,50 around 800 (based on
Masudi),51 or at 860-861 (using the legend of Cyrill).52 The Qabars were regarded
either pagans or Karaites, who revolted against the orthodox Talmudist govern­
ment.53 Another radical theory regarded the Qabars to be of Judaistic faith, who
had conflicts with the genuine pagan Khazars.54 The Qavars were considered Mus­
lims as well, or "at least under the command of Muslim military aristocracy." The
civil strife was connected to religious conflicts between the ruling class professing

42 Róna-Tas, Hungarians, 348.

43 Róna-Tas, Hungarians, 348. -"Popular etymological explanations came later, like for
example bodun kavradi 'brought the people together'.
44 DAI 174-175.
45 Róna-Tas, Hungarians, 329.
46 P. K. Kokovcov, levrejsko-hazarskoje perepiske v X. veke. Leningrad 1932, 80; Golden,
Khazar Studies, 133-134; Róna-Tas, Hungarians, 232.
47 M. I. Artamonov, Istoria Hazar. Leningrad 1962, 324-330; D. M. Dunlop, The History of
the Jewish Khazars. Princeton 1954,116-170,203; A. P. Novoselcev, Hazarskoe gosudarst-
vo i evo rol v istorii vostocnoj Evropy i Kavkaza. Moskva 1990,144-154.
48 Róna-Tas, Hungarians, 232.
49 Artamonov, Istoria Hazar, 324-334.
50 Schönebaum, Zur Kabarenfrage, 143-144.
51 Dunlop, History, 170; S. A. Pletnjowa, Die Chasaren. Mittelalterlichen Reich an Don und
Wolga. Leipzig 1978,128; Róna-Tas, Hungarians, 232.
52 Gyóni, Kalizok, 161-163; Györffy, A kabar kérdés, 86.
53 G. Vékony, "A kazár kérdés," [The Khazar problem] Életünk (1992:1-11), 915-938,
1034-1046.
54 S. P. Tolstov, Az ősi Khorezm [The ancient Khorezm] Budapest 1950, 226-233; Schöne­
baum, Zur Kabarenfrage, 143-144.; for the criticism of this "somewhat fantastic theo­
ries" Golden, Khazar Studies, 138.

8
 T he Q a v a r s (Qa b a r s ) a n d t h e i r R o l e in t h e H u n g a r i a n . . .

Judaism and the Muslims inside Khazaria and to the growing tension and milita­
ry clashes between Khazaria and the Muslim Caliphate in the 850s.55 In regards to
civil war in Khazaria ethnic conflicts were referred as well in research, supposing
that the Qabars were either Bulgars or Khoresmians, who revolted against the ru­
ling Khazars.56 It was assumed too, that "the actual secession of the Kavars to the
Magyars seem to be the result of [...] some contest between the army and the local
authorities." In this theory, Qavars were mercenary Turks in permanent service of
the Khagan, who had revolted and having been defeated "fled across the frontier"
to the Magyars.57 According to an interesting hypothesis the civil war was brought
about by the contradiction inside the Khazar double kingship, the conflicts betwe­
en the sacral Khagan and the effective ruler, the Beg/Sad. It was supposed that the
Qabars were the "partisans of the Khagan, whose power have been eroded by the
Beg, sought to reassert himself and lost."58 At last a hypothesis is to be mentioned,
which seems to be the closest to the historical reality, reflected in Constantine's re­
port. In these view "nomadic empires, such as the Khazar Empire, were compo­
sed of conglomerate forces, often centrifugal [...] The picture painted by the Em­
peror-Historian is that of a typical falling out of tribes within a nomadic state. The
losing side removes itself from the tribal union and joins another."59
Considering these hypotheses on the causes of the civil war, and the secession
of the Qabars, one should emphasize the fact, that Emperor Constantine did not
mention why the secession and the civil war had broken out. It is very likely, that if
"the civil war, which brought about the flight of the Qabars to the Hungarians [...]
was the result of a religious conflict [...] one of our source surely mentioned it."60
On the other hand "the possibility cannot ruled out" [...] that the Khavars, joining
the Magyars was not wholly independent from the conversion of the leading Kha­
zar strata to Jewish faith."61 However, Ibn Rusta remarked, that the great prince
(Khagan) and the j.sad and some viziers and leaders are willing to follow the Jew­
ish faith, and the others faith are similar to the Turks.62 It can be assumed that only
the leading strata, or rather part of it practiced a kind of Jewish faith, the Muslims

55 I. Boba, Nomads, Northmen and Slavs. Eastern Europe in the Ninth Century. The Hague-
Wisbaden 1967,115-117.
56 Bulgars, cf. I. Fodor, A magyar-bolgár-török kapcsolatok történeti hátteréről [On
the historical background of the Hungarian-Bulgar-Turkish relations.] in Bolgár ta­
nulmányok, ed. I. Darkó, Hajdú-Bihar Megyei Múzeumok Közleményei 37 (1980) 20;
Khorezmians/Kwarezmians. see Tolstov, Az ősi Khorezm, 226-233; Schönebaum, Zűr
Kabarenfrage, 143-144.
5/ Macartney, The Magyars, 133; Cs. Bálint, A honfoglaláskori lovastemetkezések.[The burial
with horse at the time of the Hungarian Conquest.] A Móra Ferenc Múzeum Évkönyvei
(1971:2), 100-102.
58 Originally the theory was invented by A. Krymskyi, and this unpublished theory was
taken over and referred by O. Pritsak, "Yowár und Káfkíp Kawár," Ural-Altaische
Jahrbücher 36 (1965) 383; for the criticism of this view see Golden, Khazar Studies, 138.
59 Golden, Khazar Studies, 135-136.
60 Golden, Khazar Studies, 135.
61 Róna-Tas, Hungarians, 348; to a similar view see Róna-Tas-Berta, West Old Turkic, 1,35.
62 For Ibn Rusta, M. Kmoskó, Mohamedán írók a steppe népeiről. Földrajzi irodalom 1/1.
[Muslin writers on the peoples of the steppe. Geographical literature] ed. I. Zimonyi,
Budapest 1997,204.
 Sá n d o r Lá s z l ó Tó t h

and Christians constituted small minorities, while the majority of the people of
Khazaria were pagans, following the ancient beliefs of the Turks.63 So considering
the structure of a nomadic empire the conversion of one part of the leading stra­
ta could not have so important implications as to provoke a civil war. The ethnic
cause of the revolt cannot be proved, since according to Constantine "the Kabaroi
were of the race (genea) of the Khazars."64 Although Khazaria comprised a lot of
people and different nations (Bulgars or Onogurs, Sabirs, Alans, Huns etc.), it is
impossible to determine, whether the Qabars had been just called Khazars earli­
er, because they belonged to the multi-ethnical Khazar Empire, or they were real­
ly 'genuine' Khazars.65 One should agree with the standpoint, that from the Byz­
antine account "it would appear that the Qabars were undistinquished from the
rest of the Khazars up to the time of the revolt."66 It is not likely at all, that the Qa­
bars could have represented a distinct tribe within the Khazar tribal union (named
Yubar/Yuwar).67 It cannot be proved either, that the Qabar revolted against the ef­
fective ruler, the Beg and served the interests of the Khagan. According to the ac­
count of Emperor Constantine, the rising directed against their 'leadership' (dpyq)
or 'first leadership' (npooxq apytj), by which phrase mainly the first dignitary of
Khazaria, the Khagan was denoted, and besides him the second dignitary, the ef­
fective ruler, the Bek/Beg (J.sad) was meant as well, because together they repre­
sented the Khazar nomad state.68 It was these leaders, by whom an envoy had been
sent to the Byzantine emperor Theophilos about 838 to ask him to make the fort of
Sharkel build for them.69
In my view the Qabar revolt must have had political aims, a typical revolt insi­
de a nomadic state against the leading tribes. Three Khazar tribes - later called Qa-
bars/Qavars - wanted to get the power from the tribes ruled by the Khagan and
the Beg, which would have transformed the Khaganate and its leadership. These
rebel tribes probably lived on the western parts of the Khaganate.70 This revolt led
to a civil war and in the ensuing battle(s) the rising tribes were defeated, the for­
ces of the Khagan and the Beg overcame. Some of the rebels were killed, but the
majority of them left Khazaria and settled with the Hungarians living between the
rivers Don and Danube (called Etelköz).71 The secession of the Qabars weakened
Khazaria and increased the military force of the Hungarians/Magyars.72
The next problem is the date of the Khazar civil war and the joining of the Qa­
bars to the Hungarians. The earliest date assumed is about 780 for some research­

63 Tóth, Levediától, 55.

64 DAI 174-175.
65 For the multi-ethnicity of Khazaria and its peoples cf. Z. V. Togan, "Völkerschaften
des Chazarenreiches im Neunten Jahrhundert," in Körösi Csorna Archívum, III. Buda­
pest 1940,40-76; D. Ludwig, Struktur und Gesellschaft des Chazaren Reiches im Licht des
schriftlichen Quellen. Münster 1982, 69-111.
66 Golden, Khazar Studies, 135; similarly Kristó, Hungarian history, 150.
67 Pritsak, Yowar, 381-382.
68 Golden, Khazar Studies, 138; Tóth, Levediától, 56.
69 DAI 180-181.
70 Boba, Nomads 117.
71 Kristó, Hungarian history, 150; Tóth, Levediától, 56.
72 Tóth, Levediától, 117; Boba, Nomads, 115-116.

10
 T he Q a v a r s (Qa b a r s ) a n d t h e i r R o l e in t h e H u n g a r i a n . . .

ers, who identified Chwärizmians with the Qavars.73 Since the identification is un­
certain, and the presence of the Hungarians in the region north of the Black Sea
cannot be proved by written sources before 830s, this theory is unacceptable. An­
other hypothesis put the Khazar civil war at around 800. This view is based on the
account of Masudi about the conversion of the Khagan at that date, and on the as­
sumed connection between the conversion and a possible reaction from the part of
the 'pagan' tribes.74 However, these connections and the Hungarians' presence on
the steppe at that time cannot be surely verified, so this date cannot be accepted ei­
ther. The secession of the Qabars was dated at the second decade of the ninth cen­
tury (810s) as well, because the old fortresses of Khazaria had been demolished
at around this time according to archeological evidence.7576 Most scholars put the
Khazar civil war at around 830, when for the request of the Khagan and Beg the
fortress Sharkel was built near river Don with Byzantine help.75 It was supposed,
that Sharkel was built against the attacks of the Hungarians, so the Qavars could
join the enemies of Khazaria.77 In this case the Qabars could settle together with
the Hungarians, who probably appeared or even could settle on their new home­
land, Etelköz (between the river Don and the Danube) at the end of the third de­
cade of the ninth century. However, the Qavars preserved their autonomy inside
the developing Hungarian tribal confederation, so it seems to me very likely, that
the joining of the three dissident Khazar tribes could have happened later or much
later.78 According to a theory, the secession of the Qavars happened around 850 or
at 854-855, when some sources refer to movements (resettlement of Khazar families
by the governor of Armenia on Arab territory) and conflicts on the Khazar-Arab
frontier (between the 'Sanariyah' and the Arab governor of Armenia) which may
have been in connection with the civil war in Khazaria.79 A hypothesis combined
this Qabar revolt with a Hungarian-Pecheneg war near the Volga, which resulted
the Hungarians' leave from this assumed Khazar homeland of theirs and their set­
tlement with their new allies, the Qavars in their new, common homeland, west of
river Don (Etelköz).80 In my view, these scattered and indirect references to some
turmoil at the Khazar-Arab frontier cannot be related to a large civil war inside
Khazaria, and cannot be connected with an assumed first Hungarian-Pecheneg
war on the eastern frontiers of the Khaganate.81 The Khazar civil war and the join-

73 Tolstov, Az ősi Khorezm, 233-234; Schönebaum, Zur Kabarenfrage, 144.

74 Róna-Tas, Hungarians, 232,348-349; Róna-Tas-Berta, West Old Turkic, 1; 35.
75 Mainly this theory was based on Arab dirhems, the last of them had the name of Ca­
liph Amin (809-812) on, see, A. Bartha, A IX-X. századi magyar társadalom. [The Hun­
garian society in the ninth-tenth century] Budapest 1968, 99; J. Harmatta, "A hon­
foglalás mai szemmel," [The Hungarian Conquest from a recent aspect] Magyar Nyelv
91 (1998), 144.
76 DAI 180-181.
77 Sz. Vajay, Der Eintritt des ungarischen Stämmebundes in die europäische Geschichte (862-
933). München 1968,11,87.
78 Tóth, Levediától, 57.
79 Boba, Nomads, 116-117; K. Czeglédy, Magyar őstörténeti tanulmányok. [Studies on Hun­
garian Prehistory]. Budapest 1985, 257.
80 Gy. Kristó, Levedi törzsszövetségétől Szent István államáig. [From the tribal federation of
Levedi to the state of St Stephen] Budapest 1980,113-114.
81 Tóth, Levediától, 58.

11
 Sá n d o r Lá s z l ó Tó t h

ing of the Khavars to the Hungarians were dated after 860 as well. One of the ar­
gument was that the legend of Constantine-Cyrill mentioned a religious debate
at the court of the Khagan about 861, and this was connected with the conversion
to the Jewish faith, leading to a revolt.82 Although this relation cannot be verified
either, I agree with dating the secession of the Qavars after 861/862. In my view
their joining can be dated between 862 (terminus post quern) and 881 (terminus ante
quam). The first date, 862 is the time when only the Hungarians (Ungri) were men­
tioned by Hincmar in Annales Bertiniani during a raid on the territory (probably
Ostmark) of Emperor Louis (the German), and in 881 both the Hungarians (Ungri)
and the Qavars (Coivari) participated on a common campaign in Ostmark. I regard
it very likely that between these dates, 861/862 and 881 the Khazar civil war and
the Qavars secession and joining to the Hungarians/Magyars took place.85
The organization and position of the three Khazar rebel tribes inside the Hun­
garian tribal confederation is quite interesting, and can be called a unique and
special one. The Qabars or Qavars are referred by Constantine in chapter 39 in the
title as 'nation' (ethnous), when he writes "Of the nation of the Kabaroi", and lat­
er he remarks, that there are "three clans of the Kabaroi."81 However, in contrast
of this statement, the Byzantine emperor refers to them in chapter 40 as the 'clan'
(genea) of the Qabars, the first among the eight clans (genea) of the Qabars and
the Turks/Hungarians.85 It is also mentioned in chapter 39 that the Kabaroi "have
been promoted to be the first clans" and there is "one prince among them."86 So
it seems that these joining Khazar tribes were regarded both as one nation (eth­
nous), consisting of three clans (genea) or rather tribes, and one clan (genea) headed
by one prince (arkhón).*7 The account of Constantine is best interpreted, that "the
Kabaroi were divided in three clans of their own, but in the Hungarian clan or­
ganization they formed one single clan, the first."88 This strange duality was ob­
served and explained differently. Most scholars supposed, that the Qabars were a
military auxiliary nation, led by one prince, and consisting of three tribes.89 It was
emphasized as well, that "the Qabars were not received into the tribal confedera­
cy system of the Hungarians" they "were not part of the Hungarian tribal confed­
eracy called Hetumoger" even in the middle of the tenth century. The Hungarians

82 Gy. Györffy, Tanulmányok a magyar állam eredetéről. A nemzetségtől a vármegyéig, a törz­

stől az országig. Kurszán és Kurszán vára. [Studies on the origin of the Hungarian state.
From the clan to the county, from the tribe to the country. Kurszán and his fort.] Bu­
dapest 1959, 50-51, 79; Györffy, A kabar kérdés, 86.
83 Tóth, Kabarok, 102; Tóth, Levediától, 58; I. Fodor questioned this hypothesis, arguing
that there is no later reference to the participation of the Qavars in Hungarian raids,
though they surely took part in these campaigns, see I. Fodor, "Kazárok és kabarok,"
[Khazars and Qabars] in Magyarrá lett keleti népek. [Oriental peoples, who were be­
coming Hungarians] Budapest 1988, 93-94.
84 DAI 174-175.
85 DAI 174-175.
86 DAI 174-175.
87 Tóth, Kabarok, 106; Tóth, Levediától, 60-61.
88 Commentary, 150.
89 Bartha, A IX-X. századi, 101,116; Gy. Györffy, Honfoglalás, megtelepedés és kalandozások.
[Conquest, settlement and raids] Budapest 1977, 135; H. Göckenjan, Hilfsvölker und
Grenzwachter im mittelalterlichen Ungam. Wiesbaden 1972, 35-36.

12

.
 T he Q a v a r s (Qa b a r s ) a n d t h e i r R o l e in t h e H u n g a r i a n . . .

"limited themselves to appointing one leader to lead their tribes."90 According to

a general view the joining Qabars constituted one tribe, the eighth tribe among
the Hungarian and Qabar tribes.91 A similar hypothesis assumed, that the three
Khazar tribes, or the remnants of tribes were united into one tribe by the Hungar­
ians.92 A radical hypothesis maintained, that it were not the Hungarians, who re­
organized the Qavars, but in fact it were the Qavars, who were sent by the Khaz­
ar ruler to lead the Hungarian tribal federation and reorganized the Hungarians.93
In contradiction of this view, it was emphasized, that "the three Kabar tribes were
united, were reorganized by the Hungarians, and accepted one leader who seems
to have ruled over them until the source DAI Chapter 39 was written."94 In an­
other version it was remarked that "the seven Magyar tribes were joined by three
Khazar tribes that unified to form one tribe" under the leadership of a Hungarian
prince, the Kharha.95 The Qavar clans (genea) were regarded as sub-tribes as well.96
I presume that this special "double structure" of the Qabars cannot be connect­
ed to their subjugation by the Hungarians, as a theory maintained, which accept­
ed the story of Anonymus about the defeat of the seven Cuman chiefs (associated
with the Qabars) by prince Álmos and their joining to the Hungarians (Hetumo-
ger).97 This late Hungarian source cannot be taken for granted, the Qabars would
be rather considered a joining people after their unsuccessful revolt in Khazaria.
It is more likely, that this special organization of the Qabars was due to the fact,
that they joined a developed tribal confederation of the Hungarians consisting of
seven tribes. It is very probable, that these three Khazar tribes were not sub-tribes,
they had not been united into a big tribe ('supertribe') or had not organized them-

90 Kristó, Hungarian history, 152-153.

91 Gy. Pauler, A magyar nemzet története Szent Istvánig. [The history of the Hungarian
nation till St Stephen], Budapest 1900,14; Németh, HMK 1930, 223, 234-235; I. Zichy,
Magyar őstörténet [Hungarian prehistory] Budapest 1939,15-16, 26; I. Dienes, A hon­
foglaló magyarok. [The Conquering Hungarians] Budapest 1978,3 9; Czeglédy, Magyar
őstörténeti tanulmányok, 118; Gy. Szabados, Magyar államalapítások a IX-XI. században.
[Hungarian state foundations in the ninth-eleventh century] Szeged 2011,197.
92 Kristó, Levedi törzsszövetségétől, 65,118, 454-457; K. Mesterházy, A nemzetségi szervezet
és az osztályviszonyok kialakulása a honfoglaló magyaroknál. Budapest 1980, 51-52; I.
Erdélyi, A magyar honfoglalás és előzményei. [The Hungarian Conquest and its anteced­
ents] Budapest 1986,44; J. Szűcs, A magyar nemzeti tudat kialakulása. [The Formation of
the Hungarian National Conscience] Budapest 1992, 287-288; Róna-Tas-Berta, West
Old Turkic, 2:1175.
93 For this view see J. Howard-Johnston, "Byzantine sources for Khazar history," in The
world of the Khazars. New perspectives. Selected papers for the Jerusalem 1999 International
Khazar Colloquium hosted by the Ben Zvi Institute, ed. P. B. Golden-H. Ben-Shammai-A.
Róna-Tas. Leiden-Boston 2007,190.
94 Róna-Tas-Berta, West Old Turkic, 2:1175.
95 Róna-Tas, Hungarians, 349,389.
96 Cf. E. Moór, A honfoglaló magyarság megtelepedése és a székelyek eredete. [The settlement
of the conquering Hungarians and the origin of the Seklers] Szeged 1944, 7, 31; Ró­
na-Tas, Hungarians, 351.
97 For the story of Anonymus, see SRH 1; 46-47; for the subjugation of the Qabars by
Álmos, cf. D. Dümmerth, Az Árpádok nyomában. Budapest 1977, 77, 97; Kristó, Levedi
törzsszövetségétől, 115; Gy. Németh, who identified the Cumans of Anonymus with the
Qabars of Constantine, maintained, that the tribes of the Qovars consisted of seven
clans, Németh, HMK 1991, 265-266.

13
 Sá n d o r Lá s z l ó Tó t h

selves into one tribe.98 First of all, there is not any example of a tribe, besides the case
of the Qavars, which would consist of three tribes. In the Hungarian tribal system,
only one tribe, the Kürtgyarmat contained two tribes, which were united or had
been united.991 assumed "that two allied people (ethnos) lived in the ninth-century
Hungarian area of settlement, the Hungarians and the Qabars. Both peoples had
a tribal confederacy type of organization, and they may have been independent
of one another. Therefore, we are inclined rather to speak of a Hungarian-Qabar
tribal confederacy with a seven-plus-three pattern of organization. The relation­
ship with the Qabars, who joined the Hungarians later, displayed a particular du­
ality: fighting together in war (subordination) and retaining the internal framework
of organization (autonomy)."1001 also emphasized, that in outside connections the
Hungarian (Hetumoger) and Qabar tribal federations constituted just one confeder­
ation, but in their inner relations they were aware of their different ethnicities and
languages. The three Khazar-Qabar tribes constituted a smaller federation inside
the Hungarian tribal federation.101 My theory of a smaller tribal confederation of
the Qabars were criticized and this concept "would prefer to place emphasis on
the relationship between the Hungarians and the Qabars [...] not on the alliance
and the Qabar autonomy, but on the system of sub/superordination and the Qa-
bar's status as a people of military auxiliaries."102 This critical view has accepted
that the structure of the Qabars remained intact, just a prince was appointed to
lead these tribes.103 An interesting model of the Hungarian tribal confederation in­
cluding the Qavars was set up based on the Turkic-type tribal federation (mainly
on the East Turks).104 According to this model, the Hétmagyar (Hetumoger) "con­
federation was a multiply differentiated system of tribes," described as 6 + 1 (+1)
+ 1 (3). "The tribal federation was headed by the Megyer tribe (Árpád's tribe)"
and six other Hungarian tribes (Nyék, the merging of Kürt and Gyarmat into Kürt­
gyarmat, Tarján, Jenő, Kér, Keszi) joined the leading tribe and made the second cir­
cle. "The third circle within this set-up was the Khavar tribe, led by one chieftain,
was divided into three subtribes."105 As regards to the place of the Qavars in this
model, it is remarked, that "the system of three tribes having one chieftain corre­
sponds perfectly to the organizational samples of Turkic tribes."106 In my view, the
organization of the Hungarian tribal federation was quite complex. It can be com­

98 Tóth, Levediától, 60-61.

99 For these examples (Qovars, Kürtgyarmat) see, Németh, HMK 1991. 37; Róna-Tas,
Hungarians, 348,350.
100 Cf. Tóth, Kabars, 112; the translation is by Gy. Novák, see. Kristó, Hungarian history,
153.
101 Tóth, Kabars, 107; idem, Levediától, 61.
102 Kristó, Hungarian history, 153.
103 Gy. Kristó, A magyar állam megszületése [The birth of the Hungarian state], Szeged
1995,125; idem, Hungarian history, 152.
104 Róna-Tas, Hungarians, 341.
105 Róna-Tas, Hungarians, 350-351; for the criticism of a Hungarian confederation of eight
tribes, see F. Makk, A korai magyar történelemről. [On the early Hungarian history)
in A turulmadártól a kettőskeresztig. Tanulmányok a magyarság régebbi történetéről. [From
the Bird Turul to the double cross. Studies on the earlier history of Magyars] Szeged
1998, 36.
106 Róna-Tas, Hungarians, 348.

14
 T he Q a v a r s (Qa b a r s ) a n d t h e i r R o Le in t h e H u n g a r i a n . . .

pared in a way to the Nine Oghuz (Tokhuz Oghuz) tribal confederation. This latter
federation contained nine tribes, but the last, the ninth tribe itself consisted of nine
tribes.107 According to Constantine's report and list of tribes, the Turks (Hungari­
ans/Magyars) had seven tribes, which he enumerated (Nyék, Megyer, Kürtgyar­
mat, Tarján, Jenő, Keszi), and later with the joining of the Qabars, who became the
first tribe(s), the Hungarian tribal confederation had eight tribes.108 In my opinion
this structure can be best described as l(3)+ 7.109 The Qabars/Qavars functioned as
a tribe inside the Hungarian/Magyar tribal federation, which cannot be called a
Hungarian-Qavar confederation, but remained Hetumoger in spite of the change in
the number of tribes. At the same time this first tribe, the Qabar/Qavar in fact con­
tained three Khazar tribes, which probably retained their names, identity. The po­
sition of the Qabars showed duality, they were regarded as a tribe inside a great­
er Magyar confederation, at the same time they retained their own organization,
which was a federation of three tribes. It means, that not just two federations ex­
isted side by side, but rather inside a great confederation existed a smaller feder­
ation, as part of it.
In case of the Qabars/Qavars we know just their common name, how they
were called by the Hungarians/Magyars and other people, but we do not have
any information about the name of the separate tribes. Scholars try to identify
the Qavars with people, who lived in the Carpathian Basin after the Hungarians/
Magyars conquered the territory about 895-900, and settled there. Scholars as­
sumed that the Seklers (in Hungarian: Székely) were either identical with the Ka-
vars or at least with part of them, one tribe. These theories were based on the fact
that the Seklers had some autonomy, seemed to be different from the Hungari­
ans, had a special organization, and played an important role in the military sys­
tem of medieval Hungary. The problem is that, the first reference to the Seklers
to be credited is from 1146, when together with the Pechenegs living in Hunga­
ry they fought as an advance guard of the Hungarian army.110 The people called
Kaliz have been identified with Kavars as well. This people of Muslim faith lived
mainly on the southern parts of the Hungarian Kingdom, had mainly military
role in the twelfth century.111 Although we cannot exclude the possibility of com­
plete or partial identity of Qavars with either the Seklers or with the Káliz, or with
both, it cannot be proved with sources.112 Generally it was assumed by scholars,
that the Qavars were a mixture of different people (Turkish and Iranian) and so
they connected them with different people living at the territory of Khazaria. So

107 Róna-Tas, Hungarians, 341.

108 DAI 174-175.
109 S. Tóth, "Kabarok és fekete magyarok," Acta Universitatis Szegediensis de Attila József
nominatae. Acta Historica 84 (1987), 25; idem, A honfoglalástól, 137.
110 For the theories on the origin of the Seklers, see e. g. Z. Kordé, A székely kérdés története.
[The history of the Sekler problem]. Székelyudvarhely 1991; Gy. Kristó, A székelyek
eredetéről. [On the origin of the Seklers], Szeged 1996; for the identification of Seklers
with Kavars see e. g. Moór, A honfoglaló magyarság, 34-36, 72-76; Györffy, Tanulmán­
yok, 45-46; the Seklers were part of me Qavars, see Kristó, A székelyek, 49-59.
111 For the identification of Káliz people with Qavars see Gyóni, Kalizok, 85-96,159-168;
Schönebaum, Zur Kabarenfrage, 142-146; Kristó, A székelyek, 36-47.
112 Tóth, Kabarok, 108; Tóth, Levediától, 61.

15
 Sá n d o r L á s z l ó Tó t h

the assumed names of the Khazar tribes were connected to different people by
scholars, who used later medieval Hungarian place names to find these names,
which would resemble to place names presumably derived from known Hungari­
an tribal names. These supposed Qavar tribal names besides the mentioned Széke­
ly and Káliz were Tárkány, Ság, Ladány, Oszlár-Eszlár, Bércéi, Varsány, Berény,
Örs etc.113 In my opinion the Qabar tribal names cannot be reconstructed.114
The common leader of the three Khazar/Qavar tribes is called prince (arkhón)
by Emperor Constantine. The dignity of arkhón denoted the prince/great prince of
the Hungarians (Árpád, his son, Liuntika and grandson, Falicsi) and the chieftains
of the tribes as well.115 Since the Qavars were regarded as a tribe inside the Hun­
garian tribal federation, but at the same time they had a federation-type organiza­
tion with three tribes, it is not easy to decide, whether the common leader of the
Qabars should be considered a tribal chieftain or a prince of a smaller tribal fed­
eration. According to a theory, the three tribes of Qavars had one chieftain, "who
was always supervised by a Magyar governor, the Kharha."116 This hypothesis in
fact combines the two possibilities, because it is uncertain, who can be regarded
the arkhón in this case, the "Kavar chieftain" or the "Hungarian supervisor", the
Kharha (karkhas), who was the third dignity in the Hungarian tribal federation.1171
would rather assume, that not a chieftain, but a prince ruled over the three Qavar
tribes.11819It is a long disputed question, what was the title of the Qavar leader, and
what clan he represented and who the Qavar arkhón was. As far as dignities are
concerned, the Qavar arkhón was identified with the great prince (kündü) of the
Hungarians (Árpád and his dynasty), the effective ruler and military command­
er (gyula), and the third prince, karchas (karha).u9 It was assumed as well, that the

113 For possible tribal names of the Qabars see, Cf. Györffy, Tanulmányok, 44-76; idem, A
kabar kérdés, 83-93; Gy. Györffy-Török Sándor, "Mi volt a neve a három kabar törzs­
nek?" [What were the names of the three Qabar tribes?] Századok 116 (1982), 986-1059;
Kristó-Makk-Szegfű, "Adatok korai helyneveink ismeretéhez," 12, 39-42; L. Kiss, A
honfoglalás és a letelepedés a földrajzi nevek tükrében. [The Conquest and the settle­
ment in the mirror of geographical names] Magyar Tudomány 1996/8, 968-969.
114 Tóth, Kabars, 108.
115 S. L. Tóth, Princes and dignitaries in the ninth-tenth century Magyar tribal federation.
Chronica 3 (2003), 27.
116 Róna-Tas, Hungarians, 348, 349.
117 For the karkhas or karha, see Tóth, Princes, 28, 36.
118 Tóth, Levediától, 62.
119 See comprehensively Tóth, Princes, 28-29; Kristó, Hungarians, 153; Árpád, the great
prince (kündü) was considered the Qavar arkhón, see, J. B. Bury, A History of the east­
ern Roman Empire. From the Fall of Irene to the Accession of Basil I. (A. D. 862-867). Lon­
don 1912, 426; R. Grousset, The Empire of the Steppes. A History of Central Asia. New
Brunswick 1970,178; L. Várady, "Revision des Ungam-Image von Konstantinos Por-
phyrogennetos," Byzantinische Zeitschrifi 90 (1989), 34-35; the eldest son of the ruling
prince governed the Qavars as their prince, e. g. Liuntika (son of Árpád), see J. Mar-
quart, Osteuropäische und ostasiatische Streifzüge. Leipzig 1903, 52, 522; the gyula was
the Qavar arkhón, see Macartney, The Magyars, 116; I. Uhrmann, "A gyula-dinasztia,
a kabarok és Szent István intelmei," Hadtörténelmi Közlemények 116 (2003), 267-366;
the karchas was the prince of the Qavars see, J. Deér, "Le probleme du chapitre 38 du
de Administrando imperio," Annuire de l,Institut de Philologie et d.Histoire Orientale et

16
 T he Q a v a r s (Qa b a r s ) a n d t h e i r R o l e in t h e H u n g a r i a n . . .

Qavars had their own-prince, and the Aba-clan hold this dignity.120 The election
of Árpád is described in Chapter 38 of DAI dealing with the seven Hungarian/
Magyar tribes, so the first dignity of kündü or great prince cannot be connected to
the Qavars.'21 The dignity of gyula, the commander-in-chief and judge may be as­
sumed as the leader of the Qavars considering their important role in wars em­
phasized by Emperor Constantine in Chapter 39. However, the name of the first
chieftain of the seven Hungarian leaders (emirs) in 942 on a raid may be identi­
fied as gyula, so probably the title of gyula cannot be associated with the arkhón of
the Qavars.122 Similarly, the name of Bulcsú, who was third prince, karchas at this
time, can be found on the list of Ibn Háyyán enumerating the seven Turk/Hun­
garian leaders. So the karchas belonged to the seven Hungarian tribes as well.123 So
one "cannot connect these dignities with specific tribes", but it is highly probable
that the kündü, gyula and karkhas belonged to Hungarian tribes, while the "Qavar
tribes had one prince of their own."1241 assume that maybe this Qavar prince wore
the highest Khazar title, Khagan, which is reflected in Hungarian chronicles in the
name of Kean (Keanus).125 So the three Qavar tribes may call their joint leader or
prince Kean/ Khagan.126 The last Qavar prince or Khagan/Kean may have been
Ajtony, defeated by King Stephen in the first decade of the eleventh century.127
The next problem concerning the Qavars is their role and position in the Hun­
garian confederation. One of the most characteristic feature of the Qavars is their
'warlike nature', as testified by Emperor Constantine, who remarked, that the Ka-
baroi "in wars they show themselves strongest and most valorous of the eight
clans, and are leaders in war, they have been promoted to be the first clans."128 It
is one the most debated phrase on Qavars, whose role and position is generally
overrated or underestimated. According to a view, "the nomadic race of Kavars,
true Turks, herdsmen and fighters, the driving force and the army of the nation"
led by their prince {gyula) undertook most campaigns and can be contrasted with
the more or less pacific and sedentary Finno-Ugric Magyars.129 This theory evi­
dently exaggerates the military role of the Qavars, attributing every campaigns

Slave 12 (1952), 102-110; L. Ligeti, A magyar nyelv török kapcsolatai a honfoglalás előtt és az
Árpád-korban. [The Turkish connections of the Hungarian Language before the Land­
taking and in the Arpadian Age] Budapest 1986.
120 For an independent Qavar ruling dinasty see B. Hóman- Gy. Szekfű, Magyar történet
[Hungarian history], I. Budapest 1935, 67-68; Kristó, Levedi törzsszövetségétől, 456-457;
Kristó, Hungarian history, 154.
121 Tóth, Princes, 29.
122 Tóth, Princes, 29; for the identification of the first name, 'T.x.x.la' with the title gyula
on the list of Ibn Háyyán, see Czeglédy, Magyar őstörténeti tanulmányok, 132,136; re­
cently, I. Eltér, Ibn Háyyán a kalandozó magyarokról. [Ibn Háyyán on the raiding Hun­
garians] Szeged 2009,71, 87,91-95.
123 Tóth, Princes, 28-29; Tóth, A honfoglalástól, 165.
124 Tóth, Princes, 29.
125 For the name (in fact title) Kean, see the fourteenth century chronicle composition or
Chronicon pictum, SRH 1: 281, 315-316; Anonymus, see SRH1: 48, 51, 86; Tóth, A hon­
foglalástól, 253.
126 Tóth, A honfoglalástól, 237,255.
127 Tóth, A honfoglalástól, 237,253-258.
128 DAI 174-175.
129 Macartney, The Magyars, 122.

17
 Sá n d o r Lá s z l ó Tó t h

to them, while excluding the Hungarian tribes from the raids and consequently
from booty. However, it is a well-known fact, that only one source, the Annals o f
Salzburg mentioned the Qavars together with the Hungarians on a raid in the ter­
ritory of Ostmark.130 Even at this time "they fought on separate fronts," because
the Hungarians skirmished or battled at Wenia (Vianna?) and the Qavars at Cul-
mite.131132 Other sources referred just to the Hungarians in connection with these
raids. Ibn Häyyän described the raid of 942 and noted that the Turks had sev­
en leaders (emirs), which may be related just to the seven Hungarian tribes. So it
can be concluded, that the Hungarian tribes participated in these campaigns, not
just the Qavars. At the same time, the Qavars had probably relevant part in these
raids; otherwise they could not show their braveness and valor. So it can be con­
cluded, that Qavars and Hungarians fought together as the members of the same
tribal federation. We know, that this joint fight was the duty of the eight tribes (in­
cluding the Qavars), as testified by Emperor Constantine: "these eight clans of the
Turks do not obey their particular princes, but have a joint agreement to fight to­
gether with all earnestness and zeal upon the rivers, wheresoever war breaks out
"i32 Although sometimes it is interpreted that the "tribal confederation no longer
had a military role in the middle of the 10th century,"133 this passage can be rath­
er related to wars of defense at rivers.134 It was rightly emphasized, that this data
"really referred to the joint military cooperation of the eight tribes (seven Magyar
tribes and one Qavar tribe) agreed on the confederation level revealing relevant
military functions of the tribal union."135 It seems probable, that these joint wars
were fought abroad as well, not only for the defense of the frontiers of the Hun­
garian confederation.136 The participation of Qavars were supposed besides the
881 raid in most campaigns and wars of the 9th century.137 It was assumed, that
they took part in the regular campaigns against the eastern Slavs in the second half
of the 9th century.138 According to a theory Qavars invaded Moravia as the allies
of King Arnulf in July 892 and might occupy the territory east of the Danube.139
The Qavars were connected with the campaign of 894, when Hungarian raided
the Transdanubian (Pannónia) territory as the allies of the Moravian prince, Sva-
topluk. It was supposed that these Qavar troops even settled on the Upper-Tisza
valley at the end of 894, before the arrival of prince Árpád in 895.140 It is possible
that the Qavars participated in the Bulgarian campaign of 894/894 as the allies of
the Byzantine Emperor Leon VI against Prince Simeon of Bulgaria.141 And at last it
130 MGH SS XXX/2, 742; HKÍF 209.
131 Kristó, Hungarians, 150.
132 DAI 178-179.
133 Róna-Tas, Hungarians, 380-380.
134 Tóth, A honfoglalástól, 168.
135 Makk, A Turulmadártól, 40.
136 Tóth, A honfoglalástól, 168.
137 For a summary of their participation in the ninth century campaigns, see Tóth, Qabars,
109-111.
138 Györffy, Tanulmányok, 68,123; Macartney, The Magyars, 122.
139 For this hypothesis see Kristó, Levedi törzsszövetségétől, 167-169.
140 For this theory see Györffy, Honfoglalás, megtelepedés, 129.
141 Marquart, Osteuropäische und ostasiatische Streifzüge, 52, 522; Macartney, The Magyars,
114,122; Györffy, Honfoglalás, megtelepedés, 128.

18
 T h e Q a v a r s (Qa b a r s ) a n d t h e i r R o l e in t h e H u n g a r i a n . . .

was assumed, that the Qavars as guards of frontiers fought against the invading
Pechenegs about 895.142 Although it is highly probable, that the Qavars could fight
together with the Hungarian tribes in most of these campaigns, their participation
can be proved just in case of the 881 campaign, where they raided the territory of
Ostmark, and fought two battles against the Eastern Franks. It seems likely that
the Qavars took part mainly in the western raids. What is surprising, that we have
no information at all about the participation of the Qavars in raids after the con­
quest of the Carpathian basin between 895-900. However, it can be assumed that
they kept on fighting together with the seven Hungarian tribes.
In connection with the role of the Qavars, we must analyze the passage cited
about their leading position in wars. Emperor Constantine emphasizes, that the
Qavars were promoted or elevated to be the first clans (i.e. tribes). He explains
this outstanding position of the Qavars with their strength and valor, and with the
fact that they were in front position in war, i. e. in battle. This passage was inter­
preted in two ways. According to one theory, this passage can be meant in the lit­
eral sense, i.e. the Qavars had a leading role, the first rank in the hierarchy of the
Hungarian tribal federation. So Árpád may have been the prince of the Qavars
and at the same time the prince of the whole Hungarian confederation.143 Accord­
ing to a similar view, the prince of the Qavars was the Gyula, the military leader
of the Hungarian tribal federation.144 Both of these hypotheses regarded the Qa­
vars the first tribes inside the tribal confederation. It may be stated, that the ma­
jor part of foreign scholars considered the Qabars a leading tribe or rather people,
who played an important role in campaigns.145 The other hypothesis represented
mainly by Hungarian scholars maintains that the Qavars should be regarded as
military auxiliaries, who provided the front and the rear guard, so their first po­
sition showed their place not in rank, but in the military organization.146 So it was
assumed by the majority of Hungarian scholars and some foreign researchers, that
the Qabars being the newest tribe of a nomadic confederation became a military
auxiliary, who had to fight in the front, and had to serve as a rear guard as well.147
It was emphasized that "this was hardly an honorary position (though Constan­
tine thought it so, thus added the phrase about the courage and bravery of the
Qabars) especially when viewed in the context of steppe tactics. The first line in
a cavalry charge must take considerable losses."148 It was remarked as well, that
"according to the rules of nomadic military organization, hosts were headed (and
142 Györffy, Honfoglalás, megtelepedés, 130.
143 Cf. Marquart, Osteropaische und ostasiatische Streifziige, 52; Bury, A History, 426; Grous-
set, The Empire, 178; Dunlop, The History of the Jewish Khnzars, 197-198; Várady, Revi­
sion, 34-35; a similar view regarded Levedi the prince of the Qabars, Pritsak, Yowar,
382-384.
144 Macartney, The Magyars, 116; G. Vékony, "Egy kazár felirat a Kárpát-medencében,"
Életünk (1987:4), 383; Uhrman, lulus rex, 287-342.
145 For a summary of views of foreign scholars on Qabars see, Uhrman, lulus Rex, 280.
146 For a summary of Hungarian scholars on Qabars see, Uhrman, lulus Rex, 273-278.
147 Németh, HMK1930.19,38,134-136,235; Németh, HMK1991.37-38,262; Györffy, Ta­
nulmányok, 46, 83; Artamonov, Istorija, 345; Göckenjan, Hilfsvölker, 35-36; Vajay, Der
Eintritt, 16; I. Fodor, Verecke híres útján. [On the famous road of Verecke]. Budapest
1975; 192-193; Fodor, Kazárok, 94-95; Kristó, Levedi, 115;
148 Golden, Khazar Studies, 137.

19
 Sá n d o r Lá s z l ó Tó t h

the rear guard also provided), not by the leading tribe but by the associated tribes
and thus the Qabars, in complete accordance with historical circumstances, should
be regarded as military auxiliaries."149 A similar view already quoted stated, that
"the last tribe(s) to join a confederation [...] were the first to be sent against the en­
emy" and the "first tribe refers to military position, not to a political rank."150 It is
very difficult to determine the position of the rank of the Qabars in the Hungarian
tribal federation. It was generally supposed, that Emperor Constantine misunder­
stood the informations on Qavars, did not know the Turkish military organization
or had been deceived by a Qavar informant.151 I am of the opinion, that we should
interpret the Byzantine source according to the text. There is a logical connection
between the fact, that the Qabars "show themselves strongest and most valorous"
and that "they are leaders in war". The word 'proexarkhein' may be interpreted as
'to go ahead, in front', but its probable meaning is 'to lead', in the political or mil­
itary sense.152 I think that the sentence is clearly referring to the military leading
role of the Qabars, which is testified by their promotion, or elevation to be the first
tribes. This consequent and logical statement is reinforced by the tribal list in the
beginning of chapter 40, where the Kabaroi are mentioned first.153 The Qabars may
serve as subordinate people, joining the last to the 'Seven Magyars' and forming
in battle the front guard or the first line. However, it is possible to assume the Qa­
bars as an allied people, constituting a smaller confederation inside the joint Hun-
garian-Qavar confederation, fought together and really led the joint troops under
their prince (Khagan/Kean).154 Both interpretations may be valid. It is sure, that
the Qavars were constituted not by fragments of tribes, and their historical role
must not be underestimated, they had important military role in the Hungarian
confederation.155
The cultural role of the Qavars must be emphasized too. They represented the
culture and customs of Khazaria, and had part in the formation of Hungarian lan­
guage. Emperor Constantine remarked, that the Kabaroi "to these Turks taught
also the tongue of the Chazars. and to this day they have this same language, but
they have also the other tongue."156 This passage clearly "testifies to the fact that
the Hungarians, owing to their association with the Kabaroi, became bilingual."157
Most scholars regarded the Qabars bilingual, but doubted, that this bilingual sta­
te may be referred to the Hungarian tribes as well. They rather attributed it to
the Hungarian upper strata due to the former Khazarian contacts.158 It was assu­

149 Kristó, Hungarian history, 153.

150 Róna-Tas, Hungarians, 348, 350; similarly Róna-Tas-Berta, West Old Turkic, 1175.
151 For the deception of the Byzantine emperor by Bulcsú, who was assumed the infor­
mant on Qavars, D. Dümmerth, Álmos, az áldozat. [Álmos, the victim] Budapest 1986,
28-32.
152 Uhrman, lulus rex. 280-281.
153 DAI 174-175.
154 For a similar interpretation see Tóth, Levediától, 64.
155 Uhrman, lulus rex, 285-287; - in his opinion the Gyula was the prince of the Qabars.
156 DAI 174-175.
157 Commentary, 150.
158 Fodor, Kazárok, 95; Gy. Györffy, Krónikáink és a magyar őstörténet. Régi kérdések-új
válaszok. Budapest 1993, 66.

20
 T he Q a v a r s (Qa b a r s ) a n d t h e i r R o l e in t h e H u n g a r i a n . . .

med, that the "Magyar tribal federation undoubtedly spoke Hungarian, and the
Magyarisation of Khavars occurred very quickly, while some groups of the Ma­
gyars were for a while bilingual, speaking Turkic as well."159 In another words
"the Hungarians also learnt the language of the Khazars in the late ninth century,
and some of them became bilingual for a time, but the Kavars began to assimila­
te to the Hungarians."160 At the same time, it is evident, that the Qavars had some
part on the doubted bilingual state of the Hungarian tribes, and the confedera­
tion. It is not doubted, that "Khazar loan-words were adopted in the Hungarian
language through contacts between the Qabars and the Hungarians."161 Generally
linguists has doubts concerning the importance of Qavars in their linguistic influ­
ence on Hungarians and on the Hungarian (Magyar) language. Generally it is as­
sumed that there was a long interval of Turkish influence on the Hungarians by
Turkish people, like the Bulgars of Volga and the Khazars, and the Qabars meant
only the last and least influence on the Magyar language. In this concept the Qa­
bars meant "the last contact with people that spoke a language of the Khazar Em­
pire."162 It is also supposed that "Hungarian copied the Turkic loanwords from
Oguric, which was spoken by both the Khazars and the Bulgars."163 According
to the latest researches, the "Hungarian language contains 384 words of West old
Turkic origin."164 Contrary to this linguistic aspect historians contributed much
more importance to the role of Qabars in this respect as well.165 According to a
view "since the coexistence of the Hungarians and the Qabars covered a long time,
in fact one or two centuries from the joining of the Qabars to their Magyarization,
the Qabars obviously must have exerted an intensive influence on the Hungarian
language."166 Considering the possibility, that the Hungarians had much shorter
connections with the Khazars, than was assumed earlier, and maybe it was restri­
cted to just some decades of the ninth century, while the interval of the relations
of the Hungarians with the Volga Bulghars cannot be determined exactly, we may
contribute more relevance with the Qabars seceded from Khazaria, who lived and
fought with the Hungarians/Magyars not only in the ninth century, but in the
next centuries in the Carpathian Basin as well.
The three Khazar tribes, called Qabars/Qavars had an interesting history. They
appeared on the historical scene with their revolt against the Khazar government.
Although this revolt failed, and the 'rebels' were defeated, with their joining to the
Hungarians/Magyars they were able to continue their own history. They preser­
ved their autonomy or part of it, they were regarded by the Hungarians as one 'tri­
be', although in their own tribal federation they retained the former tribes. They
had their own prince, probably called Khagan ('Kean'), preserved their Khazar

159 Róna-Tas, Hungarians, 389.

160 Róna-Tas-Berta, West old Turkic, 2:1175.
161 Ligeti, A magyar nyelv, 533.
162 Róna-Tas-Berta, West Old Turkic, 2:1175.
163 Róna-Tas-Berta, West Old Turkic, 2:1176.
164 Róna-Tas-Berta, West Old Turkic, 2:1176.
165 A. Bartha, "Türk-ősmagyar etnogenetikai kérdések," [Etnogenetic problems of
Turk-ancient Hungarians] in F. Tőkei, ed. Nomád társadalmak és államalakulatok (Ta­
nulmányok), Budapest 1983, 67-69; Tóth, Qabars, 112-113; Tóth, Levediától, 65.
166 Kristó, Hungarian history, 151.

21
 Sá n d o r L á s z l ó Tó t h

identity, had an ethnonym and had important military role in the Hungarian tribal
federation, participated in lot of campaigns, and proved to be brave 'allies' of the
'Seven Magyars.' Their historical importance cannot be doubted; they exerted po­
litical, military, cultural and linguistic influence in the Hungarian tribal confedera­
tion. And at the same time their language preserved in Turkic/Khazar loanwords
in Hungarian language can be regarded as the relics of the culture of the Khazar
Khaganate, which fell in 965.

22