Prehistory of Saladin

Vladimir Minorsky, Prehistory of Saladin

Saladin's Origins (A)

The famous biographer Ibn Khallikân (A.D. 1211-82) made a special inquiry into the history of Saladin's family1 and came to the following conclusion2: "Historians agree in stating that his father and family belonged to Duwîn, which is a small town situated at the farther extremity of Adharbayjan, in the direction of Arrân and the country of the Kurj (i.e. the Georgians). They were Kurds and belonged to the tribe of Rawâdiya (sic) which is a branch of the great tribe al-Hadâniya (read: *Hadhbâniya).

I was informed by a legist (faqîh) who was a native of Duwîn and never said anything of which he was not certain, that near the gate of the town lies a village called Ajdanaqân, all the inhabitants of which are Rawâdiya Kurds, and that the father of Salâh al-dîn was born there. Shâdî went to Baghdad with his two sons, Asad ad-dîn Shîrkûh and Najm al-dîn Ayyûb, and thence to Tikrît where they settled. Shâdî died there, and his tomb with a cupola (qubba) over it is within the town. I have carefully studied their genealogy but have not found any mention beyond Shâdî".3 Other sources say much less and only stress the fact that Saladin's father was born in Dvin.4

We shall examine one by one the questions raised by the passage of Ibn Khallikân.

1. The sure point in this important statement is Duwîn, i.e., the early Armenian capital Dvin, later one of the key-points of the Muslim domination in the Caucasus.5 The position of Dvin should be sought on the left bank of the river Garni (ancient Azat6 which flows into the Araxes to the east of the river Zanga (Hrazdan) on which Erivan is situated.

2. I have not seen it noticed that the native village of Saladin's ancestors Ajdanaqân should be identified with Azhdanakan which, some four centuries earlier, the Armenian historian Moses of Khoran places in the same neighbourhood.7 The passage occurs in the fantastic story of the king Tigran, who in alliance with Cyrus is said to have crushed the Medians (Arm. Mar < Persian Mâ&a). Tigran was first an ally of *astyages (whom Moses calls Azhdahak),8 and gave his sister Tigranuhi. Having unmasked the wiles of Azhdahak, Tigran killed him and sent Tigranuhi back to Armenia. As to the first wife of Azhdahak, called Anunysh, and ten thousand other prisoners, he settled them "beyond the eastern range of the great mountain (i.e., of Ararat) down to the confines of Golt'n, that is in Tambat, Oskiol, Dazhguynk' and in other palaces standing near the bank of the river (scil.Araxes), one of which called Vranjunik' is opposite the fortress of Naxçavan. Tigran settled Anuysh and her sons in a secure place where from stretch (the traces) of the slide of the great mountain, said to have been caused by a formidable earthquake ... As servants to Anuysh Tigran appointed the Medians (Mar) whom he settled at the foot of the said mountain." What is more important, Moses adds that the story was recorded in the "Historical Songs" which were sung in Golt'n; one of them told how, when Artashat was founded, Artavazd, son of Artashês, "went forth and among the Medians (Mar) built Marakert in the plain called Sharur".9

This text is most remarkable for the accuracy of its geographical indications. The great chasm on the northern face of the Ararat is that of Akhuri. Tambat is one of the high valleys of the Lesser Ararat where in 1905 I found an ancient town lying in ruins. Jula is the well-known frontier point of Julfa. On the northern bank of the Araxes, Sharur lies to the north-west of Nakhchavan,10 and Golt'n corresponds to the tract between Julfa and Ordûbâd. Azhdanakan11 lying at the head of the plain must be located near where the Garni river emerges from the hills, i.e., in the neighbourhood of Dvin.12

3. There is no doubt that the term Mar (Medians) refers to the Kurds.13 In the time of Moses of Khoren there were no Medians in existence, but even now the Kurds continue to occupy the slopes of the Ararat. In the curious Armenian manuscript containing samples of alphabets and languages, written some time before A.D. 1446, a prayer in Kurdish figures as a specimen of "the languages of the Medians (Mar)" and such a use of the term is still attested in dictionaries.14

When in 22/643 Habîb ibn Maslama arrived in Ardasât (*Artashat) he "crossed the river of the Kurds (nahr al-Akrâd) and descended into the plain of Dabîb (Dvin)", see balâdhurî, p. 200, Tabari, I, 2674. This stream can be only the Garni river, for according to a reliable Armenian source15 the Arshakid Khosrov II (A.D. 381-9) planted a wood "from the solid royal castle of Garni down to the plain of Metsamor, down to the hill called Dvin and situated to the north of the great city of Artashat, and down the river to the palace Tiknuni". As already said, towards the end of the 10th century, the great Arab geographer Muqaddasi, 377, remarked that the Christians formed the predominant element in Dvin16 but that the town was held by the Kurds (yadbituhu al-Akrâd).

The whole story of Tigran and Anuysh is a legendary superstructure over the positive fact that in the neighbourhood of Ajdanaqân on the territory between the Ararat, Dvin and Ordûbâd, there lived numerous Kurds, from times more ancient than has been usually assumed. Astyages (Ishtuvegu) was confused by Moses of Khoren with Azhdahak, and possibly the homonymy of Persian mâr ("snake") with Armenian Mar ("Median") also contributed to the imbroglio especially as some connection of the Kurds with Azhdahak was hinted at even in Persian epics.17

4. We come now to Saladin's clan Rawadi (rwâdî) which according to Ibn Khallikan's reliable informant was a sept (batn) of the tribe Hadhânî. Although this name is carefully spelt out in this form (hdhânî), the older parallel forms18 guarantee the reading *Hadhbânî (or Hadhabânî) . This name is derived from the old geographical term for the region of Irbil (Arbela), which is preserved in the name of the Nestorian diocese, Adiabene (Ha&ayyab).19 The Hadhbani had their summer quarters near Ushnû (I. Hauqal) but their activities are recorded in Salmas, Maragha and other places of Azerbayjan.

There is always some uncertainty about the names of the Kurdish tribes as the original tribal name is often replaced by the names of the outsiders who succeeded in taking the leadership of the tribe. This apprantly happened to some sections of the Hadhbânis, who came under the rule of the originally Arab families issued from Rawwâd al-Azdî. In my opinion (see above p. 123), the clan to which Saladin's family belonged was somehow connected with the Rawwâdî family whose name may have been simplified by non-Arabs into Rawâdî.

At this point we have to straighten out another kink in the Ayyubid chart. According to the historian of the Kurds, Sharaf al-dîn, who wrote in 1005/1596, the Ayyûbids belonged to the Ravanda Kurds of Dvin (Sharaf-nâma, p. 55). This term is not found in the earlier sources, although a tribe of a similar name (Ravand, Ravandî) exists nowadays on the western slopes of the frontier range separating Ushnû from Ravânduz. It is possible that Ravandi is only a later mutilation of the former Rawâdî, especially in view of the identity of their summer quarters with those of the old Hadhbânî (see above).20

And to sum up, there may have been successive waves of Kurdish migrations to the Araxes valley. The Kurds were there before the time of Moses of Khoren, but it is possible that after the Arab invasion some Hadhbânîs came to reinforce the original settlers. Finally, the existence of the Rawâdî clan among the Kurds of Dvin is likely to be explained by some particular connection between them and the Rawwâdî of Azarbayjan.

(B) / No historian has recorded the exact year in which Shâdhî ibn Marwân, whith his sons Ayyûb and Shîrkûh, left Dvin. We only know that they made for Baghdad at the invitation of Jamâl al-daula Mujâhid al-dîn Bihrûz b. 'Abdullâh al-Ghiyâthî. The latter had been Shâdhî's close friend in Dvin (?)21 and at the time in question was acting as shihna of 'Irâq. Bihrûz appointed Shâdhî to be dizdâr(commandant) of the castle of Tekrît on the Tigris. As Bihrûz was "the master of Tekrît" Shâdhî had probably not too long to wait for the post which, strictly speaking, was not a government appointment. He died in Tekrit and was succeeded by Ayyûb. The first definite date in the Ayyubid Odyssey was the service rendered by Ayyûb to the ruler of Mosul. In 526/1132 Zangî b. Aqsunqur led an expedition against Baghdad to support the candidature of the Seljukid Mas'ûd b. Muhammad b. Malik-shâh. In the battle which fought near Tekrît on the 2nd of March 1132 Zangi was defeated by the caliph's general Qaraja.22 His retreat was facilitated by Ayyûb who placed at his disposla boats for crossing the Tigris. This generosity to an enemy did not affect Ayyûb's appointment, apparently because of Bihrûz's link with Mas'ûd. Again in the following year 527/1133 ayyûb showed his independence by refusing to surrender to the former vazir, al-'Azîz, who was placed in his custody.23 Bihrûz, "the master of Tekrit", had himself to visit the fortress to carry out the orders (in 527/1132-3). The brothers were turned out of Tekrit only when Shîrkûh killed a man (an isfahsâlâr?) who was abusing a weeping woman. On the night of their departure, the tradition says, Saladin was born in Tekrit, in 532/1138. The family sought the protection of Zangi and the atabek gave them a fief. In 533/19 August 1130 Zangi captured Baalbek and put Ayyûb in charge of this fortress (mustahfiz). After the death of Zangi (14 September 1146) Ayyûb ceded the fortress to the army of Damascus (October 1146) and took service with the Bûrids. Meanwhile Shîrkûh joined Nûr al-dîn of Aleppo, and when the latter decided to dispossess the Bûrids, Shîrkûh took part in the negotiations with his brother and Ayyûb peacefully surrendered Damascus to Nûr al-dîn in 549/1154. Ayyûb kept Damascus on behalf of Nûr al-dîn and Shîrkûh received Hims.24 Soon after 550/1155 the Shaddâdid of Ani Shaddâd, see above p. 88, came as an exile to the residence of the former vassal of his family, Shîrkûh. The latter died an old man in 564/23 March 1169 and Ayyûb followed him in 568/16 August 1173.

The first certain date in the history of the Ayyubids is 1132 and we should remember that in 524/1130 the cruel Qurti b. Tughan-Arslan wrested Dvin from Fadlûn II [the Shaddadis] who lost his life in the fighting. As the Ayyubids are repeatedly called close dependents of the Shaddâdids, it is most probable that, after the death of the last energetic Shaddâdid and in the presence of a brutal usurper, the position of the family became intolerable and this is the most likely explanation of the flight of Shâdhî's family to Baghdad.25


  1. See his notices (translated by de Slane, 1842): I, 243-8, Najm al-dîn Ayyûb ibn Shâdî ibn Marwân; I, 626, Asad al-dîn Shîrkûh ibn Shâdî; III, 235: al-Malik al-'âdil Sayf al-dîn Abû Bakr Muhammad ibn Ayyûb; IV, 479-563: Salâh al-dîn Yûsuf ibn Ayyûb ibn Shâdî.
  2. See also Recueil des historiens des croisades, III, 399, with a translation by de Slane, IV, 479.
  3. Only as a curiosity Khallikân mentions the tree drawn up by an obliging nassâba, which goes up to Adam (hattâ intahâ ilâ âdam).
  4. See Bahâ al-dîn in Recueil, III, 6;I. Athîr, XI, 225 (very brief) and Abul-Fidâ (himself an Ayyûbid!), Annales Moslemici, ed. Reiske, III, 616 (nothing original).
  5. See above chapter 3.
  6. On the Russian 5-verst map three small villages bearing the name Devin are marked in this neighbourhood. Curiously enough one of them is called Dvin-Kurdakend. On the site of the ancient town see N.Y. Marr's archaeological reconnaissance near Ardashahr in Otchal Imperat. Arkheolog. Komissii za 1899, St. Petersburg 1902, pp. 91-4; also N.Y. Marr, "Ani", 1934, p. 17. Manandian, l.c., places Dvin near the village Artashar (whose site he distinguishes from the ancient Artashat, Artaxata, which lay nearer to the Araxes).In 1949 the government of Soviet Armenia decided to restore the ancient Dvin by uniting the villages standing on its presumed site. [The recent discovery of the ruins of the cathedral and the palace of the catholicos has permitted to pinpoint the site of Dvin, see K.G. Kafandarian, quoted in Vestnik Drevney Istorii, 1950, I, 151.]
  7. Moses of Khoren, I, ch. 30, Russian translation by N.O Emin, Moscow 1893, p. 44.
  8. Azhdahâk is a mythical Iranian name < Avestan Azhi Dahâka "the dragon Dahâka". This name has nothing to do with the name Astyages which in the cuneiform texts appears as Ishtuvegu.
  9. Hubschmann, Arm. Ortsnamen, p. 451, comments on this Marakert: "von den Modern gemacht?"? Sachlich unwarscheinlich". On the contrary, the overflowing of the Mar (Medians, Kurds) to the northern bank of Araxes was as natural in olden times as during the "Iranian intermezzo" (see above, p.123).
  10. Shah Ismâ'îl I defeated his Aq-qoyunlu enemeis in Sharur. This district lies between Dvin and Nakhchavan.
  11. Azhdahâ+ Arm. toponymic suffix -akan.
  12. The Nuzhat al-qulûb(A.D. 1340) mentions an Ajnân among the boroughs od Nakhchavan, adding that it is also called kârkhâna ("works") because it possesses a copper mine. This place, however, should not be confused with Ajdanaqân. Brosset in his translation of S. Orbelian, II, 63, quotes a passage from Alishan's Great Armenia, 1855, $171: "Melri (Meghri), to the south of Zangazor (between Ordubad, the Araxes and Akera river) contains the ancient canton of Balk', later called Ajanan and Kajunik'". This Ajanân is then the famous mining centre of Kapan (now Kafan).
  13. Minorsky, "L'origine des Kurdes", Travaux du XX-ene Congres des Orientalistes, Bruxelles 1938, pp. 143-52.
  14. A. Shanidze, "The newly discovered alphabet of the Caucasian Albaninas" (in Russian), Tbilisi 1938. Shanidze quotes the dictionary of Avetikian, Surmelian, Avgerian, Venice 1837: mar"a Kurd, a Shirvani". See "The Medians in ancient Armenia" (in Russian) in N.O. Emin's Izsledovaniya, Moscow 1896, pp. 122-32.
  15. Faustus of Byzantium, III, ch. 8.
  16. Muqaddasi states with regret: ma'a nublihi"despite the nobility (of the town)."
  17. Shâh-nâma, ed. Vullers, I, 36: The Kurds are the decendants of the children whom the cook Armâyîl saved from the ravenous snakes which grew out of the shoulders of Zohhâk
  18. Miskawayh, GMS, VII/6 (fascimile), p. 237 al-hdhâbânyt; Ibn Hauqal, 239 (ed. Kramers, 336): al-hdhbânyh.
  19. Ravânduz and Ushnû lie on the way from Irbil to Azarbayjan.
  20. The alternative would be two seek the origin of Ravandî in the castle Ravânduz, whose name has been connected by Sir H. Rawlinson with the ancient Orontes quoted in this region by Pliny, VI, 118. See JRGS, X, 1840, p. 23; cf. Marquart, Sudarmenien, 1930, 393-6, but in any case Rawwâdî (Rawâdî) cannot be derived from Orontes/Ravânduz.
  21. I. Khallikan, l.c., I, 243. "An amir" with whose wife he was improperly familiar had him casterated and he left to take service with Sultan Ghiyâth al-dîn Mas'ûd ans rose to be tutor to his sons. This must have happened even before Mas'ûd accession to the throne (528-47/1133-52). I. Khallikân, IV, 481, says that Bihrûz was a Greek slave and had a fair complexion. The links of Bihrûz with Transcaucasia appear also from Ibn al-Athîr, XI, 51: reporting on the earthquake which shook Ganja in 434/1139 he says that two sons of the lord of the country Qara-sunqur (*Aq-sunqur?) were killed and "the castle was ruined with Mujâhid al-dîn Bihrûz had there (hunâka)." Immediately after this record the author speaks of a canal (sikr) which Bihrûz built in the region of Nahrawânât.
  22. The famous memoirist Usâma ibn Munqidh was an eye witness of the battle but his account of it has not come down to us, see H. Derenbourg, "Vie d'Ousâma", pp. 144-6.
  23. Al-'Aziz was the uncle of the historian 'Imâd al-dîn who speaks with sympathy of Ayyûb ("a good Muslim") and of his brother Shîrkûh whose personal recollections he quotes, see al-Bunaârî, 154, 163, 167.
  24. See Wiet, Shirkuh in Enc. Isl.
  25. Samuel of Ani, Recueil, Doc. armeniens, I, 453, misunderstood the situation when he wrote that the cause of the exile was the poverty of the family and that in Tekrit Eyyub and "Shiraku" worked as water-carriers. He ends with the story of the prophetic dream of Ayyûb who see flames flaring up from his loins, etc. The pronunciation and the meaning of Shîrkûh is not quite certain: it might be "the mountain lion" and one might think of Shîr-gôh(Shêr-gûê) in Kurdish "having the lion's ear."

Source: V. Minorsky, Studies in Caucasian history, Cambridge University Press, 1957, pp. 124-132.