Navigation – Plan du site

AccueilNuméros54Acts of Protection in Early Islam...Sogdian Ḏimmī. Religious and Poli...

Acts of Protection in Early Islamicate Societies

Sogdian Ḏimmī. Religious and Political Protection in Early 8th Century Central Asia

Les Sogdiens ḏimmī-s, la protection religieuse et politique au début du viiie siècle en Asie centrale
أهل الذمة الصُغد، الحماية الدينية والسياسية في آسيا الوسطى في أوائل القرن الثامن الميلادي
Étienne de la Vaissière
p. 165-176

Résumés

On tente ici de comprendre les processus par lesquels les Sogdiens ont pu devenir ḏimmī‑s au cours de la première moitié du viiie siècle, à partir des données textuelles, tirées des récits de la conquête, et archéologiques – les destructions de temples – tout en proposant un parallèle avec la conquête simultanée du Sind.

Haut de page

Texte intégral

  • 1 Al-Ṭabarī, Tārīḫ al-rusul wa-l-mulūk, pp. 1507–1510; transl., XXV, pp. 46–48.

1The main political protection offered by Islam to non‑Muslims was the status of ḏimmī, with its correlate, the payment of the ǧizya. As early as 728, within one generation after the conquest of 705–712, the Sogdians were regarded as ḏimmī: they were paying the ǧizya, in unambiguous texts making use of the word in its precise meaning of religiously based capitation, in exchange for military protection.1 However, the status of ḏimmī and the ǧizya (with this meaning) are not mentioned in the treatises of surrender of Bukhara and Samarqand, as they are transmitted, and the aim of this article is to try to explain how they were nevertheless implemented. Two caveats should be made.

  • 2 See generally Crone, 2012. For the differences between Zoroastrianism in Iran and in Sogdiana, Shko (...)
  • 3 Linghu Defen, Zhoushu, 50.920.
  • 4 Shenkar, 2014, on the complex Sogdian Pantheon.

2First, if the dominant religion in Pre-Islamic Central Asia was Zoroastrianism, Zoroastrianism can mean a whole array of religious groups and practices.2 Visually speaking, for outsiders, Zoroastrianism in Iran and Central Asia differed markedly and Chinese travellers in Central Asia made use of two different names to describe them.3 In Central Asia, as seen from the paintings and statues discovered on the territory of Sogdiana, there was no ascetic cult of fire, in buildings without paintings or sculptures for fear of allowing the demon to be embodied in them: on the contrary, in Sogdiana the religion adopted from the 5th century onwards an extremely rich and abundant iconography based on Indian models.4 For Muslim witnesses, as for the Chinese, this religion could only be seen as polytheist and pagan. Moreover, it seems to have produced only a few books, and not used them in most rituals. It could in no way be counted a priori among the religions of the Book, endowed with a revealed Book, recognized by the Muslim statute of ḏimmī.

  • 5 Friedmann, 2003, pp. 72–76.

3Second, if we should be careful not to think that the corpus of the Muslim norms was already fixed and codified at the beginning of the 8th c., when Sogdiana was conquered, still the status of ḏimmī was firmly rooted in the practices of Muḥammad in Arabia. However, Zoroastrianism was not mentioned in the list of peoples with revealed Books, i.e. Christian, Jews and Sabeans. Muḥammad had to deal with Jewish and Christian Arabs and, if the global truth of the sources we have for this period is to be accepted, we know what he decided to do with them: after submission, exclude them from the military in exchange for protection and the payment of the ǧizya. But there were also some Persian Zoroastrians in Arabia. When Muḥammad conquered Bahrayn, he agreed to extend the same system of protection to the Zoroastrians there, as quasi‑ḏimmī. This pragmatic choice had great consequences when the Sassanian empire was conquered, but still it was only a precedent, not a full ḏimmī status: they were not usually regarded as ahl alKitāb because they had no Book, or have lost it.5

4Eventually, the practice of Sogdian religion, a not standard form of Zoroastrianism, itself not a standard protected religion, became protected. How?

5We should differentiate several stages.

6The first stage was the military conquest itself. The record, both textual and archaeological, of the first wave of invading Muslim armies in Central Asia in the last quarter of the 7th century and the first quarter of the 8th century is quite straightforward and simple: all the temples or cultural life we are aware of were destroyed or transformed into mosques, in the three sub‑regions of conquered Central Asia.

  1. As regard Bactria, the sources are very limited, but according to the geographers, the raid in 671 pillaged and destroyed the Nawbahār, the main monastery of Bactra. In a Chinese literary text describing a temple to the god of the Oxus on an island in the river, it is written that “recently a king of the Arabs, an unbeliever, entered the temple to destroy it.”6
  2. In Sogdiana, when Qutayba took Paykent, the troops pillaged the temple, took the idols of silver and melted them down, so that, wrote al‑Ṭabarī “from Paykand there came into the hands of the Muslims (booty) the like of which they had never acquired in Khurasan”.7 In nearby Bukhara, Narshakhī wrote that “Qutayba b. Muslim built a grand mosque inside the citadel of Bukhara in the year 94/712–713. That place (formerly) had been a temple”.8 When, in 712, Qutayba took Samarqand, according to al‑Ṭabarī, “Qutayba made peace with them in return for one hundred thousand slaves, the fire temples, and the adornments of the idols. He took receipt of that on the basis of which he had made peace with them, and he was brought the idols, which were despoiled and then placed before him; gathered together, they were like an enormous edifice. He ordered that they be burned”.9 In Panjikent, taken and pillaged in 722, we have archaeological proof of the destruction of the two temples: one of them was totally burnt, the other one abandoned.10 In between Samarqand and Panjikent, the rural temple of Jartepe was burned during this period, although in that case we do not have a precise date.11
  3. In Khorezm, after the conquest in 712, we do not have archaeological data but, according to the most famous son of the country, al‑Bīrūnī, Qutayba b. Muslim “has extinguished and ruined in every possible way all those who knew how to write and to read the Ḫwārizmī writing, who knew the history of the country and who studied their sciences” and “for after Qutayba b. Muslim al‑Bāhilī had killed their learned men and priests, and had burned their books and writings, they became entirely illiterate”.12
  • 13 Marshak, Raspopova, 1990, p. 155.

7There is a clear pattern here, all the main temples were at one point destroyed, the religious life of the Central Asian oasis was utterly destroyed in its main monumental and public forms. It was wartime, but it was also a cultural war, targeting specifically other religions. As far as we know, this was new in Central Asia. We do not have previous examples, for instance during the numerous Nomadic conquests, or the Chinese conquests, violent as they were, of a specific targeting of religious life. The shock should have been major. Not only the Gods have failed to protect the cities, but they were expelled from their own houses and burnt. Central Asian Zoroastrianism, very different from Iranian Zoroastrianism, never recovered and disappeared without any trace, up to the archaeologists of the 20th century. Only in local contexts, in the countryside or the mountains, could parts of this religion survive, in a non‑iconic form, as demonstrated by 20th c. ethnographic work.13

8We have nothing else. We do not know why, when and how ǧizya and ḏimmī status were implemented from Central Asian data during or after this first stage. The texts are silent.

9However, if we accept to make use of a wider corpus, there are some very clear texts, contemporary with the conquest of Central Asia but outside of it, which might be helpful as they describe the next stage, immediately after the conquest.

  • 14 Asif, 2016. I cannot see any incompatibility in the Chachnāmeh being a romance and making use of ol (...)
  • 15 Friedmann, 1984, p. 26.

10As is well known, al‑Ḥajjāj organized two conquests simultaneously in the east. One in Central Asia, sending Qutayba b. Muslim as governor of Khurāsān, and one in Sind in 711–712, sending Muḥammad b. al‑Qāsim. The conquest of Sind is described in a very detailed way in the Chachnāmeh, a 13th century Persian text of composite origin. It is globally speaking a 13th c. romance but, the important point for us, it made use of several lost sources and preserves most probably the data of al‑Madāʾinī, the main source both for Central Asia and for Sind.14 The data of the Chachnāmeh have been confirmed by external sources: a detailed reading of al‑Balāḏurī has validated most of the prosopography and the global time frame presented in it. As Johanan Friedmann wrote: “The author of the Chach Nama had the Arab historiographical tradition at his disposal. Other passages also seem to reflect the fact that this author, al‑Kūfī, used material hinted at in Arab historiography. The Futūḥ albuldān says, for instance, that ‘letters of al‑Ḥajjāj were received by Muḥammad (b. al‑Qāsim) and letters of Muḥammad (b. al‑Qāsim) were received by al‑Ḥajjāj every three days. Muḥammad (b. al‑Qāsim) used to describe the situation to al‑Ḥajjāj and to ask his opinion about the proper course of action’. This passage evidently refers to the extensive exchange of letters between al‑Ḥajjāj and Muḥammad b. al‑Qāsim incorporated in the Chach Nama.”15

  • 16 I have first proposed this parallel in my course Early Islamisation of Central Asia in 2012–2013 at (...)

11Al-Ḥajjāj was in touch in his very directive way both with Qutayba conquering Central Asia and Muḥammad b. al-Qāsim conquering Sind. The letters to Qutayba are lost, but the letters to Muḥammad b. al‑Qāsim are not. They do describe a policy of protection, and a quite precise one, immediately after the first destructive stage. If the only thing we see in Central Asia are looting, burning and destruction, it might be that these texts help us to understand what went on after them. It is entirely possible that al‑Ḥajjāj might have conducted a different policy in Central Asia and in Sind according to local circumstances, but still, the ideas and values governing his detailed instructions should be common to both.16

  • 17 Al-Balāḏurī, Futūḥ alBuldān, transl. Murgotten, II, p. 221.
  • 18 Al-Kūfī, The Chachnamah, transl. Fredunbeg, pp. 168–169.
  • 19 Al-Kūfī, The Chachnamah, transl. Fredunbeg, p. 169.

12According to the Chachnāmeh the Arab religious policy in Sind was a corollary to political submission and came after this submission. It seems that, as both Hinduism and Buddhism were literate religions possessing scriptures, it was not conceptually difficult for the Arabs to extend the Zoroastrian precedent to the non-Muslims of Sind and consider them quasi‑ḏimmī. The general religious policy of Muḥammad b. al‑Qāsim is noted by al‑Balāḏurī: “He conquered the city by treaty (ṣulḥ) with the condition that he would not kill them nor enter their temple (budd). And he said: ‘The budd will be considered similar to the churches of the Christians and Jews and the fire-temples of the Zoroastrians (maǧūs)’. He imposed the tribute (ḫarāǧ) on those in al‑Rur and built a mosque.”17 The Chachnāmeh applies a similar ruling to the Hindu. According to the Chachnāmeh the Buddhists monks requested to be able to rebuild their destroyed temples. Muḥammad b. al‑Qāsim submitted the request to al‑Ḥajjāj, and he received the following answer: “The letter of my dear cousin, Muḥammad b. al‑Qāsim has been received and the situation as outlined understood. With regard to the petition of the headmen (muqaddamān) of Brahmanabād concerning the building of temples (ʿimārat-i buddah), since they have submitted peacefully and have adhered to their status of ḏimma by remitting the amwāl to the capital, apart from this māl, there can be no just claim on them. Because when they have become ḏimmī, we have absolutely no further rights to their lives or property. Permission is hereby granted for them to worship their own deity. No one should be forbidden or prevented from, following his own faith. They can do as they will in their own homes.”18 Al‑Ḥajjāj did not answer on the precise question of the building of temples. But in Sind, Muḥammad b. al‑Qāsim interpreted the answer in a maximalist way, he granted permission for the building of temples, authorized religious mendicants, authorized public festivals and rituals and even deducted a contribution to the priests of 3 percent of the poll‑tax, as was the case in Pre‑Islamic, Hindu dominated, Sind.19

13The global scenario outlined for Sind, with superior textual sources but without archaeological ones, might be also useful for understanding the situation in Central Asia, without the orders of al‑Ḥajjāj but with archaeological data.

  • 20 Al-Balāḏurī, Futūḥ, transl. Murgotten, II, p. 217ff.

14The first stage is common to both Central Asia and Sind: a shock and awe tactic, extremely destructive for every single political entity trying to resist. At this stage, the invading armies destroyed the main temples. This was the case in Sind, with the destruction of the main sanctuaries during the siege of the towns. We have for Sind, in al‑Balāḏurī or in the Chachnāmeh, details on temples destroyed by catapults, idols melted, mosques built on former temples, exactly as in Central Asia.20 There is no distinction in the text between the political and the religious and it was also the case in Sogdiana.

  • 21 Shenkar, 2017, p. 198.
  • 22 La Vaissière, Riboud, 2003, p. 128 quoting the Beishi 97.3234: 有胡律,置於祅祠.
  • 23 Shenkar, 2017.

15As regard the political status of religion in Pre-Islamic Central Asia, if we know next to nothing on the Sogdian priests, the political status of the main Sogdian temples is clear from Sogdian and Chinese sources: at the beginning of the 8th c. the coinage in Panjikent was in the name of Nana the Lady of Panch, that is the main deity of the city.21 In Panjikent, the courtyards of the temples might have been the meeting point of the population, the only ones in the town. In Samarqand, a Chinese text is saying that the laws are deposited in the temples.22 In a recent article Mickael Shenkar has developed the idea that the Greek Polis might be a good model for the Sogdian cities, with a major role for the local temples and deities as central in the policy and identity of the cities.23 All the textual and archaeological data we have are linked with the first stage of the conquest. The destruction of the major temples in the centre of the cities, the symbol of the urban and oasis communities, left major traces in the historiography and archaeology. But it seems that the text of the letter of al‑Ḥajjāj to Muḥammad b. al‑Qāsim might explain a second stage in Sogdiana, a new pattern after political submission.

16The letter of al‑Ḥajjāj authorized the religious life on a private basis. In Sogdiana, the main temples in the middle of the towns were not rebuilt: this is the difference between the lenient Muḥammad b. al‑Qāsim and the, by all account, very harsh Qutayba b. Muslim. But we have proof of an accommodation of low‑level religious life under the Islamic umbrella in Central Asia.

  • 24 Narshakhī, transl Frye, 1954, p. 23.
  • 25 Grenet, Azarnouche, 2011, p. 160.

17On the longue durée the Zoroastrian festivals were held in Sogdiana: some of them were famous in the Bukhara oasis in the tenth century and were pertaining to the local form of Zoroastrianism, quite close to pure paganism in the eyes of the Arabs.24 Some Zoroastrian temples were built at the latest at the beginning of the 9th c., but on an Iranian Zoroastrian model, devoid of the exuberant polytheism of the Sogdian model, if we are to believe a small passage of the Persian Rivayat.25

18But much closer to the time of the conquest, it seems possible to demonstrate with archaeological data that the order of al‑Ḥajjāj might have indeed been the general policy implemented by the Arab power after the conquest.

  • 26 Berdimuradov, Samibaev, 1999, pp. 62–63. The authors consider two scenario, one in which the main t (...)

19In the rural setting of the Jartepe temple, quite different to the politically symbolic main urban temples, it seems that a small reconstruction, of limited size and short-lived, took place.26 Most of our data however come from Panjikent, as the town has been thoroughly studied. The town was partly destroyed in 722, laid abandoned during one generation and was rebuilt in the 740s when a general peace was agreed between the governor of Khurāsān, Naṣr b. Sayyār, and the Sogdians. The exiled elites came back and rebuilt the town. It was later again totally abandoned in the 770s.

  • 27 Shkoda, 2009, pp. 118–120; Grenet, Azarnouche, 2011, p. 164.
  • 28 Pavel Lur’e, personal communication. See provisionally Lur’e, Semenov, 2019, pp. 24–28. The excavat (...)
  • 29 Raspopova, 1990, p. 134.
  • 30 Marshak, Raspopova, 2012, p. 256

20When the town was rebuilt in the 740s, we still do not have any testimony to Muslim religious life in it. No mosque has been discovered, only barracks for Arab soldiers at the citadel. The two temples in Panjikent were not reconstructed after their burning and looting in 722, but it seems that at least a purification ritual of the ruins was implemented.27 Some religious life might have restarted in some small rooms just outside of the temple area, a small annex temple existing prior to the 722 fire and slightly renovated after that.28 The palaces and big houses of the nobility and the wealthy were rebuilt. However, there was a major difference in the planning of these houses and palaces. About 40% of the houses of the previous period, pre‑722, had reception halls splendidly decorated with mainly religious paintings. The gods were in full display, enthroned on the main wall, in what were essentially privately owned public spaces, everywhere in the town. During the pillage of the town in 722, M. Shenkar has demonstrated that these Reception halls were especially targeted.29 The paintings were defaced, and Arabic graffiti written on them.30

  • 31 Marshak, Raspopova, 2012, p. 258.

21This was no longer the case in the houses built in the 740s, even for very wealthy or noble citizens. The reception halls with the gods in full display disappear from Panjikent. The houses and the social life is now inward looking. The religious paintings still existed, but they receded to the inner spaces of the house. They could no longer be seen by casual visitors.31 The religious life has receded to private buildings. In other words, it seems that al‑Ḥajjāj’s letter to Muhammad b. al‑Qāsim exactly describes what went on in Sogdiana. The Sogdian religion was protected on a private basis, in the houses, precisely in accordance with al‑Ḥajjāj’s words.

  • 32 Marshak, Raspopova, 2003, ill. 109.
  • 33 La Vaissière, Riboud, 2003; Marshak, Raspopova, 2003, ill. 110–113.
  • 34 Grenet in La Vaissière, Riboud, 2003.

22We might even have traces of an attempt of a Sogdian religious answer to this policy. In sector XXVI in Panjikent, a house was burned in 722 and rebuilt after that, most probably in the 740s. On the first floor, before 722, a room (room nº 3) was entirely decorated with the usual religious iconography of sitting gods, hunting demons—we have only some limited fragments.32 The décor of the second stage was entirely different: a lively procession all around the room has dancers with masks and music in the streets of the town, to honour a big book, more than one meter high, that was carried and displayed in public33 (Fig. 1). This is not a Coran, as Sogdian gods are displayed on its cover. A god is also shown as he would go out of the book. He is clearly depicted as an emanation of the text, relying here on one of the epithet of the god Srosh in the Avesta, “whose body is the sacred speech”.34 This iconography is entirely foreign to whatever is known of Pre‑Islamic Sogdian iconography. Moreover, books were certainly not the strong side of Zoroastrianism or its Sogdian variant. We have only one single Zoroastrian manuscript fragment in Sogdian, none in Pahlavi, while we have dozens of Manichaeans or Christian manuscripts in Sogdian. In other words, instead of putting forward their usual pagan iconography, it is as if the Sogdians chose to display that they are faithful to a Religion of the Book. It was an attempt to accommodate and react to the conditions of the protection, to their fragile status of quasi‑ḏimmī.

Fig. 1. Panjikent XXVI-3, Procession of a Sacred Book, Marshak Rapopova, 2003, ill. 109.

Fig. 1. Panjikent XXVI-3, Procession of a Sacred Book, Marshak Rapopova, 2003, ill. 109.
  • 35 La Vaissière, 2010; La Vaissière, Marquis, 2013.
  • 36 Al-Ṭabarī, Tārīḫ alrusul, p. 1205, transl., XXIII, p. 154.
  • 37 Jan, Yang et al., 1984, p. 52.
  • 38 MacLean, 1984, p. 121–124.
  • 39 La Vaissière, 2010, pp. 526–527.
  • 40 Al-Ṭabarī, Tārīḫ alrusul, p. 1490, transl., XXV, p. 27.
  • 41 Gimaret, 1969, pp. 300–301.

23The successive patterns are quite clear. There is however one major exception, the fate of the Nawbahār in Bactra and more generally of Buddhism in Bactria. The Nawbahār was destroyed in 671, according to the pattern we see elsewhere. The second fitna then precluded any Muslim raid in Bactria for thirty years and the temple was rebuilt as a gigantic octagonal monastery.35 The key point is that it does not seem to have been fully destroyed again when the Muslims came back to Bactra at the beginning of the 8th century. We do not know at all what was its fate, as it is not mentioned in the futūḥ literature, but Nezāk tarqan, one of the major players in Bactrian political life was able to pray in it in 708.36 It should not have been totally destroyed again and might even be fully standing. Similarly, the Korean monk Hye Ch’o (Hui Chao) 慧超 who came back from India to China in 725–726 does not describe Buddhism in Bactria as destroyed. He wrote: “From Bamiyan, I travelled further north and after twenty days arrived at the country of Tokharestan. The capital city is called Bactra. At present, the place is guarded and oppressed by Arab forces. The original king was compelled to leave the capital, and he resided at Badakhshan, which is one month’s journey from the capital towards the east. It is also under the authority of the Arabs […] the king, the chiefs, and the common people respect the Three Jewels. There are many monasteries and monks. Hīnayāna Buddhism is practiced […] they do not profess any other religions.”37 Although Hye Ch’o might have not been in Bactra itself, and might be describing the situation further east, farther away from the Muslim encroachments, still he seems to describe a local Buddhism very much alive. The parallel with Sind is once more striking. As D. MacLean has demonstrated, the Buddhists there quickly aligned themselves with the invading Muslims.38 The policy of the master of the Nawbahār, the Barmak family, seems to be similar. Their collaboration with the Muslim power—lacking any military force to resist it—drew them the hostility of Nezāk tarqan, who had most of the family killed. This attitude left in the Muslim sources the legend of a very early conversion of the head of the family to Islam, based on a confusion of generations between the various Barmak.39 Nevertheless, we see the Barmak marrying with the daughter of an ally to the Muslim, the king of Chaghaniyan, rebuilding Balkh on the order of the Muslim governor Asad b. ʿAbdallāh in 725–72640, and eventually converting and siring the Abbasid vizier’s family. In Termez, some monks are known in the 720s.41 We do not know how long this policy protected the monks themselves, but the Arab governors, after the first shock of the conquest, were pragmatic enough not to weaken a key component of their local political network.

  • 42 The question of the rhythm of demise of Buddhism in Bactria cannot be dealt with the very limited d (...)

24At the root of the political attitude of the Muslim conquerors in front of the very Pagan religion of the Sogdians, there were plain orders of al‑Ḥajjāj, as quoted in the Chachnāmeh. The protection could be only secondary, after a first stage of political annihilation. The status of quasi-ḏimmī was recognized to them, once defeated. It was a decision from the centre. The archaeological data, beyond the eye-catching destruction of the main temples, does support the idea of an effective implementation of the orders of al‑Ḥajjāj. Pagan religion was accepted as an in-home private dealing. Under these very strict conditions, it was protected. The evolution in Bactria shared with Sogdiana the initial stage of political annihilation, but the inner balance of symbolic and military power in the province seems to have tilted the main representative of the Buddhists there, the Nawbahār’s Barmak, in favour of collaboration with the new power in the West. Protection, at least up to the 720s,42 was here the consequence of political pragmatism.

Haut de page

Bibliographie

Primary Sources

al-Balāḏurī, The Origins of the Islamic State: Being a Translation from the Arabic, accompanied with annotations, geographic and historic notes of the Kitāb Futūḥ alBuldān, 2 vols., Ph. Kh. Hitti, Fr. Cl. Murgotten (transl.), Columbia University Press, New York, 1916–1924.

al-Bīrūnī, The Chronology of Ancient Nations, C. E. Sachau (transl.), W. H. Allen, London, 1879.

al-Kūfī, ʿAlī ibn Ḥāmid, The Chachnamah, An Ancient History of Sind, Giving the Hindu Period Down to the Arab Conquest, M. K. Fredunbeg (transl.), Commissioner’s Press, Karachi, 1900.

Linghu Defen, Tang, Zhoushu, Zhonghua Shuju, Beijing, 1971.

Narshakhī, M., The history of Bukhara, R. Frye (transl.), Medieval Academy of America, Cambridge, 1954.

al-Ṭabarī, Tārīḫ al-rusul wa-l-mulūk, De Goeje (ed.), Brill, Leiden, 2010; English transl., E. Yarshater (dir.), The History of alṬabarī, 38 vols., State University of New York Press, New York, 1985.

Secondary Sources

Asif, M. A., A Book of Conquest. The Chachnama and Muslim Origins in South Asia, Harvard University Press, Cambridge, 2016.

Berdimuradov, A. È. And Samibaev, M. K., Khram DzhartepaII (k problemam kul’turnoĭ zhizni Sogda v IV-VIII vv.) (The temple of JarTepe II. On some problems of the cultural life of Sogd, 4th–8th cents.), Toshkent, 1999.

Bosworth, E., “Abū Ḥafṣ ʿUmar al‑Kirmānī and the Rise of the Barmakids”, Bulletin of the SOAS 57, 2, 1994, pp. 268–282.

Crone, P., The Nativist Prophets of Early Islamic Iran. Rural Revolt and Local Zoroastrianism, Cambridge University Press, Cambridge, 2012.

Drège, J.-P., Grenet, Fr., “Un temple de l’Oxus près de Takht‑i Snagin d’après un témoignage chinois du viiie siècle”, Studia Iranica 16, 1987, p. 117–121.

Friedmann, Y., “The Origins and Significance of the Chach Nāma”, in Y. Friedmann (ed.), Islam in Asia. Vol. I South Asia, The Magnes Press, Jerusalem, 1984, p. 23–37.

Friedmann, Y., Tolerance and Coercion in Islam. Interfaith Relations in the Muslim Tradition, Cambridge University Press, Cambridge, 2003.

Gimaret, D., “Bouddha et les bouddhistes dans la tradition musulmane”, Journal Asiatique 257, 1969, pp. 273–316.

Grenet, Fr., Azarnouche, S., “Where are the Sogdian Magi?”, Bulletin of the Asia Institute 21, 2011, pp. 159–177.00

Jan, Y., Yang, H. et al. (ed., transl.), Hye Ch’o Diary: A Memoir of the Pilgrimage to the Five Regions of India, Asian Humanities Press, Berkeley, 1984.

La Vaissière, Étienne de, “De Bactres à Balkh, par le Nowbahār”, Journal Asiatique 298, 2, 2010, pp. 517–533.

La Vaissière, Étienne de, “Asie centrale médiévale : la première islamisation”, Annuaire de l’EHESS. Comptesrendus des cours et conférences 2012–2013, 2013, pp. 191–192.

La Vaissière, Étienne de, Marquis, Ph., “Nouvelles recherches sur le paysage archéologique de Bactres”, Comptes rendus de l’Académie des inscriptions et belleslettres, 2013, pp. 1155–1171.

La Vaissière, Étienne de, Riboud, P., Grenet, Fr., “Les livres des Sogdiens (avec une note additionnelle par Frantz Grenet)”, Studia Iranica 32, 2003, 1, pp. 127–136.

Lur’e, P., Semenov, N., Materialy Pendzhikentskoj Arxeologicheskoj Èkspedicii, XXIII, Saint Petersburg, 2019.

MacLean, D., Religion and Society in Arab Sind, PhD Dissertation, McGill University Libraries, Montreal, 1984.

Marshak, B., Raspopova, V., “Wall Paintings from a House with a Granary. Panjikent, 1st Quarter of the Eighth Century A.D.”, Silk Road Art and Archaeology 1, 1990, pp. 122–176.

Marshak, B., Raspopova, V., Otchet o raskopkax gorodishcha drevenego Pendzhikenta v 2002 godu, Gosudarstvennyj Èrmitazh, Saint Petersburg, 2003.

Marshak (†), B., Raspopova, V., “Panjikent and the Arab Conquest”, Journal of Inner Asian Art and Archaeology 7, 2012, pp. 255–275.

Melikian-Chirvani, A. S., “L’évocation littéraire du bouddhisme dans l’Iran musulman”, in Le monde iranien et l’Islam. II. Sociétés et Cultures, Libraire Droz, Geneva, 1974.

Raspopova, V., Zhilishcha Pendzhikenta, Nauka, Leningrad, 1990.

Shenkar, M., Intangible Spirits and Graven Images: The Iconography of Deities in the PreIslamic Iranian World, Brill, Leiden, 2014.

Shenkar, M., “The Religion and the Pantheon of the Sogdians (5th–8th centuries CE) in Light of their Sociopolitical Structures”, Journal Asiatique 305, 2, 2017, pp. 191–209.

Shkoda, V. G., Pendzhikentskie khramy i problemy religii Sogda (V–VIII vv.), Gos. Ėrmitazh, Saint Petersburg, 2009.

Van Bladel, K., “The Bactrian Background of the Barmakids”, in Charles Burnett, Anna Akasoy and Ronit Yoeli‑Tlalim (eds.), Islam and Tibet: Interactions along the Musk Routes, Ashgate, Farnham, 2010, pp. 45–87.

Haut de page

Notes

1 Al-Ṭabarī, Tārīḫ al-rusul wa-l-mulūk, pp. 1507–1510; transl., XXV, pp. 46–48.

2 See generally Crone, 2012. For the differences between Zoroastrianism in Iran and in Sogdiana, Shkoda, 2009, pp. 10–25; Shenkar, 2017, p. 195.

3 Linghu Defen, Zhoushu, 50.920.

4 Shenkar, 2014, on the complex Sogdian Pantheon.

5 Friedmann, 2003, pp. 72–76.

6 Drège, Grenet, 1987.

7 Al-Ṭabarī, Tārīḫ alrusul walmulūk, p. 1189, transl., XXIII, p. 137.

8 Narshakhī, transl. Frye, 1954, p. 48.

9 Al-Ṭabarī, Tārīḫ alrusul walmulūk, p. 1246, transl., XXIII, p. 194.

10 Shkoda, 2009, pp. 48–49.

11 Berdimuradov, Samibaev, 1999, see below.

12 Al-Bīrūnī, The Chronology of Ancient Nations, transl. Sachau, pp. 42, 58.

13 Marshak, Raspopova, 1990, p. 155.

14 Asif, 2016. I cannot see any incompatibility in the Chachnāmeh being a romance and making use of old sources.

15 Friedmann, 1984, p. 26.

16 I have first proposed this parallel in my course Early Islamisation of Central Asia in 2012–2013 at the École des Hautes Études en Sciences Sociales, see the abstract in La Vaissière, 2013.

17 Al-Balāḏurī, Futūḥ alBuldān, transl. Murgotten, II, p. 221.

18 Al-Kūfī, The Chachnamah, transl. Fredunbeg, pp. 168–169.

19 Al-Kūfī, The Chachnamah, transl. Fredunbeg, p. 169.

20 Al-Balāḏurī, Futūḥ, transl. Murgotten, II, p. 217ff.

21 Shenkar, 2017, p. 198.

22 La Vaissière, Riboud, 2003, p. 128 quoting the Beishi 97.3234: 有胡律,置於祅祠.

23 Shenkar, 2017.

24 Narshakhī, transl Frye, 1954, p. 23.

25 Grenet, Azarnouche, 2011, p. 160.

26 Berdimuradov, Samibaev, 1999, pp. 62–63. The authors consider two scenario, one in which the main temple (phase 4) is destroyed during the first Muslim raids of the 670s, reconstructed on a small scale (phase 5) before Qutayba, and destroyed in the early 720s when nearby Waraghsar and Panjikent were specifically targeted. Or a destruction of the main temple in the early 720s, a reconstruction under Naṣr b. Sayyār and a final destruction under Abū Muslim.

27 Shkoda, 2009, pp. 118–120; Grenet, Azarnouche, 2011, p. 164.

28 Pavel Lur’e, personal communication. See provisionally Lur’e, Semenov, 2019, pp. 24–28. The excavation is still ongoing and the archaeological interpretation might change.

29 Raspopova, 1990, p. 134.

30 Marshak, Raspopova, 2012, p. 256

31 Marshak, Raspopova, 2012, p. 258.

32 Marshak, Raspopova, 2003, ill. 109.

33 La Vaissière, Riboud, 2003; Marshak, Raspopova, 2003, ill. 110–113.

34 Grenet in La Vaissière, Riboud, 2003.

35 La Vaissière, 2010; La Vaissière, Marquis, 2013.

36 Al-Ṭabarī, Tārīḫ alrusul, p. 1205, transl., XXIII, p. 154.

37 Jan, Yang et al., 1984, p. 52.

38 MacLean, 1984, p. 121–124.

39 La Vaissière, 2010, pp. 526–527.

40 Al-Ṭabarī, Tārīḫ alrusul, p. 1490, transl., XXV, p. 27.

41 Gimaret, 1969, pp. 300–301.

42 The question of the rhythm of demise of Buddhism in Bactria cannot be dealt with the very limited data available. It should be however pointed out that the witnesses quoted in the late 8th century Aḫbār al‑Barāmika wa‑faḍāʾiluhum saw a functional Nawbahār, as the depiction of the ceremonies there, transmitted by Ibn al‑Faqīh and Yāqūt, are very accurate. Bosworth, 1994, esp. pp. 269–271. It should be noted however that some rāwīs were very old and described precisely the conversion of the Barmak during Hisham’s reign (724–743). The ceremonies might have been seen during this period. Melikian‑Chirvani, 1974, pp. 12–18; Van Bladel, 2010, p. 66–67.

Haut de page

Table des illustrations

Titre Fig. 1. Panjikent XXVI-3, Procession of a Sacred Book, Marshak Rapopova, 2003, ill. 109.
URL http://journals.openedition.org/anisl/docannexe/image/7908/img-1.jpg
Fichier image/jpeg, 673k
Haut de page

Pour citer cet article

Référence papier

Étienne de la Vaissière, « Sogdian Ḏimmī. Religious and Political Protection in Early 8th Century Central Asia »Annales islamologiques, 54 | 2020, 165-176.

Référence électronique

Étienne de la Vaissière, « Sogdian Ḏimmī. Religious and Political Protection in Early 8th Century Central Asia »Annales islamologiques [En ligne], 54 | 2020, mis en ligne le 27 octobre 2021, consulté le 29 mars 2024. URL : http://journals.openedition.org/anisl/7908 ; DOI : https://doi.org/10.4000/anisl.7908

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search