§ 49 The Sarir 447
§ 49. The Sarīr.
Sarīr is no original name of the country, being only an abridgement of the title ṣāḥib al-sarīr "Master of the Throne" which the Arabs gave to the local king (on his golden throne sarīr, in our text takht, see I.R., 147, Mas'ūdī, Murūj, ii, 41, Iṣṭ., 223). I.R. and Gardīzī, 101, call the king respectively and . This detail suggested to Marquart and Barthold (Daghestān in EI) the identification of the Sarīr with the Avar people in Daghestan (on the middle one of the Qoy-su rivers of which the united waters, under the name of Sulaq, flow into the Caspian south of the Terek).  Balādhurī, p. 196 in fine, discussing the titles given by Anūshirvān to the princes of Daghestān says: "the khāqān of the mountain (khāqān al-jabal!) who is the Master of the Throne is called (variants: )". Thomas Artsruni (tenth century) mentions the people Aurhaz-k' () jointly with the Tsanar-k' (Ṣanār, § 36, 33.), cf. Marquart, o.c., 496. In the Ẓafar-nāma (written in A.D. 1424), i, 772, 777, 779, the Daghestan Avars are called Auhar, and this form is very much like the first element of the king's name in Balādhurī.  Mas'ūdī, ii, 41, relates a story of the descendance of the Master of the Throne from Bahrām Gūr, and abusively confers on him an additional title of Fīlān-shāh (v.i., § 50, 3c.). The connexion with the Master of the Throne seems to have been much appreciated for not only Mas'ūdī speaks of the matrimonial ties between the Alān and the Sarīr princely families but we know that Isḥaq b. Ismā'īl, ruler of Tiflis circa A.D. 830-53, was married to a daughter of this king, see Ṭabarī, iii, 1416, and Thomas Artsruni, transl. Brosset, pp. 143, 168. The important historical role of the Avar principality is shown by the fact that the Avar language is still a sort of lingua franca in Daghestan. This language, belonging to the local eastern Caucasian group has, however, nothing to do with the original language of the Avar invaders who first organized the principality of the Sarīr. See Map xi.
[The Avars were the great conquerors of the fifth century who about
A.D. 461-5 drove the Sabirs westwards. They are the people which the Chinese
call Jou-jan or Juan-juan and which, according to P. Pelliot,
was of Mongol race.  In A.D. 551 they were vanquished
by the Turks (T'u-chüeh)
1. The Avars are now an entirely Daghestanian people, but their name may reflect that of some former dynasty issued from the pre-Islamic , , see I.Kh., 92 and 119: al-Abar, cf. Marquart, Komanen, 74-5.
2. In Armenian Aurhazk', k' is the suffix of plural and the group rh, with usual metathesis, must stand for hr. In *Auhraz I consider z as an Iranian suffix of origin similar to that of Gur-z, Lak-z, Lāy-z (§ 36, 36.). Consequently the Armenian form *Auhra-z (easily < Auhar-z) is identical with later Persian Auhar ( > Avār) and in its light we may restore Balādhurī's title, which consists of a name + suffix z + suffix ān. As regards the form we may suppose that it is a result of several tentatives to transcribe the difficult name containing the elements usually confused in writing. One of the may be superfluous and with (or still better ) we approach the Armenian form. The tentative restoration would be *Auhar-z-ān-shāh.
3. Cf. T.-Pao, xxxii/4, 1936, p. 235.
448 Commentary § 49
and according to Theophylactos Simocatta fled partly to China ( = Tabghach = the Wei dynasty) and partly to the . On the contrary, the Avars who first came into touch with the Byzantines towards A.D. 557 only assumed the name of their famous predecessors whereas in reality they belonged to the (Uyghur) tribes , &c. Cf. Marquart, Die Chronologie d. alttürk. Inschriften, 94, Ērānšahr, 52-5, 84, &c., Streifzüge, passim, and Chavannes, Documents, 229-33. One should think then that the Avars who gave a dynasty to northern Daghestan belonged to the later "Pseudo-Avars", though the question needs still a special investigation. As regards the Mukri who were joined by a part of the true Avars they are usually identified with the Tunguz people Mu-chi (Wu-chi) or Mo-ho which lived north of Corea, cf. Herrmann, Historical Atlas of China, Cambridge, U.S.A., 1935, Map 30), but Marquart, Ērānšahr, p. 54, ult. was inclined to take them for the Merkit (see, however, his later Komanen, 88). In favour of this opinion one could quote a fact which seems to have escaped the notice of the scholars. According to Rashīd al-dīn, ed. Bérézine, vii, 90 (transl. v, 70) one of the four divisions of the or was called (*Auhar?), which name is, indeed, reminiscent of the title borne by the Sarīr king. This item does not, of course, constitute a proof for the "true-Avar"' origin of the rulers of Sarīr, who, following Theophyl. Simocatta's theory, could have usurped a name which did not strictly belong to them, but the resemblance of the names in both cases is a testimony for its original form. It must be finally remembered that the evidence for the distinction of the true Avars and Pseudo-Avars, depending chiefly on one Byzantine source, is rather frail and a complete disappearance after A.D. 552 of a warrior race would be unexpected.]
On the mountain of the Sarīr see § 5, 18. On the wrong idea of the sources of the Kur see § 6, 56. Our author makes the Sarīr border on the Rūm, the latter term evidently referring only to the Georgians (§ 42, 15.).
No parallels could be found for the story about the giant flies (pasha). Mas'ūdī, ii, 42, gives the name *Maghaṣun to the Alān capital (v.s.). Could not some popular etymology *Maghaṣ > Persian magas "fly" have served as a starting-point for a story reflecting some discontent with the neighbours? [?]
1. On the king's castle see also § 5, 18 c. Mas'ūdī, ii, 42, calls the seat of government (dār al-mulk) of the Sarīr (; d'Ohsson quotes the variant ). This place may correspond to the present centre of the Avar territory, Khunzakh (). See Map xi.
(perhaps repeated under § 48, 2.) corresponds to
in I.R., 147, and Gardīzī,
ii, 7, 39. Already d'Ohsson in his very good book Des peuples du Caucase,
Paris 1828, p. 19, wrote: "il me paraît hors de doute qu'il faut lire
puisque le pays ici désigné est encore habité par les Kaitacs ( dans Abul-Fidā
1. Abul-Fidā, p. 71, says that Jabal al-Qytq (perhaps for *Qabq ?) stretches from the coast of the Caspian sea near the town of Bāb al-abwāb in a southern
§ 49 The Sarir 449
Kātib-Chelebi)."  Marquart in the additions to his Streifzüge, 492, came practically to the same conclusion. The Qaytaq (speaking [now] a Daghestanian language) live on the rivers Gumri (*Humri?) and Darbakh, along the coast immediately north of Darband.  I.R., 147-8, quoting from some earlier source calls the king of Khyzān Ādharnarsē and says that he is equally respectful to the three religions of his subjects (Muslims, Jews, and Christians). He places Khyzān at 12 manzils "to the right"  of the Master of the Throne's castle which distance is certainly exaggerated, unless it is based on a confusion with a different Khyzān. According to Mas'ūdī the Jydān was formerly a part of the Khazarian principalities (dākhila fi jumlat mulūk al-Khazar); in his times the king, a Muslim said to be of the Qaḥṭān tribe, bore the hereditary name of . 
A certain difficulty results from some of Balādhurī's data on : p. 204, in the enumeration of Salmān's federates he mentions "Shakkī, Khyzān, Shirvān", &c., as if they were neighbours; p. 206, he says that Jarrāḥ after having attacked the people of ( *Ḥumrīn? cf. Balādhurī, 208) settled them in the "rustāq (sic) of Khyzān" which suggests a place in the rear [?]. Finally I.Kh., 124, in his identification of the places mentioned in the Qor'ānic legend of Moses, mentions "the village of Khyzān"  along with Bājarvān (south of the Araxes) and Shirvān. Between Baku and Qubba there exists on the Ata-chay a village and a district of Khizi which may be taken into consideration in connexion with the above hints. Its distance from the capital of the Sanr would roughly suit I.R.'s "12 manzils" (v.s.).
Mas'ūdī entirely separates Jydān from the Sarīr but I.R. and Gardīzī speak of Khyzān in the middle of the chapter on the Sarīr. This arrangement which evidently existed in the common source is responsible for the place which our author assigns to Kh.ndān in the present chapter. The Sarīrian "generals" living in the town may be only an echo of the original report on the king Ādharnarsē .
3. A parallel of
is found only in I.R., 148, who places
at a distance of 10 farsakhs from the town of Khyzān
and describes the cult of the tree practised by its inhabitants, cf. Marquart,
direction. Then quoting from al'-Azīzī (cf. note to §60, 4.) Abul-Fidā adds that this mountain is called Jabal al-alsun because some 300 languages are said to be spoken on it. Then he proceeds "and on its northern side are *Qaytaq who are a (special) race (jins); and on southern side the Lazgī who are also a (special) race”.
1. Jihān-nümā, pp. 401-2.
2. Marquart, o.c., p. 492, identifies Jydān with Majālis (on the left bank of the Buam). However, the usual residence of the Utsmi of the Qaytaq was at Bashl, or Barshli on the Humri river, cf. Klaproth, Tableau du Caucase, 1827, p. 139. D'Ohsson, o.c., p. 10, proposed the identification of this place to , where according to Balādhurī, p. 195, Anūshirvān met the khāqān of the Khazars. Consequently may be Bashli = [ = Armenian Varač'an].
3. It can mean only to the south-east for, p. 148. the Alān territory begins at 3 days distance “to the left" of the Sarīr.
4. The tltle may be Turkish, ct. the title Se-li-fa many times mentioned in Chinese sources, see Chavannes, Documents, Index.
5. Marquart, o.c., 492, was inclined to distinguish it from Mas'ūdi's Jydān.
450 Commentary §§ 49-50
only name I can think of is that of the river Rubas which waters the Tabarsarān district to the south-west of Darband. Following the coast north to south the districts come as follows: Qaytaq, Darband, Ṭabarsarān, but west of Darband the sources of the Darbakh (in Qaytaq) and the Rubas (in Ṭabarsarān) lie very close to one another. Therefore the mention of Rubas would not be unexpected after Qaytaq (*Khaytāq). could easily be restored as Rubkhas > Rubas. 
The mention of *Masqut lying farther south in
the same direction (§ 50, 3.) may be explained by our author's reading off his
map the whole series of names situated in one line. I.R., whose text is our sure
parallel, after the intermezzo on Khyzān
and Rnḥs, finishes
the chapter by mentioning a second castle belonging to the Master of the
Throne: "it is called
and has a strong position; in it is the treasury (bayt al-māl)
of the king and it was given to him by Anūshirvān".
This fortress is undoubtedly
mentioned in Mas'ūdī,
ii, 39, in the neighbourhood of Zirīgarān
and the Sarīr,
i.e. the present-day Qumiq situated on the southern branch of the Qoy-su
in the proximity of the Avar territory. 
1. The Tā'rīkh Bāb al-Abwāb quoted by A. Z. Validi, Azerbaycan, p. 39, spells the name but the name may have had an older form.
2. The territory of Qumiq
is occupied by the Lak who are also called Qazi-Qumuq (to be distinguished
from the Turkish Qumq
living further north-east towards the Caspian sea). Cf. Dirr, in Peterm.
Mitteil, 54, 1908, pp. 204-12.
[Back to Index]