Jump to ContentJump to Main Navigation
Afghanistan's IslamFrom Conversion to the Taliban$

Nile Green

Print publication date: 2016

Print ISBN-13: 9780520294134

Published to California Scholarship Online: September 2017

DOI: 10.1525/california/9780520294134.001.0001

Show Summary Details
Page of

PRINTED FROM CALIFORNIA SCHOLARSHIP ONLINE (www.california.universitypressscholarship.com). (c) Copyright University of California Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in CALSO for personal use (for details see http://california.universitypressscholarship.com/page/535/privacy-policy-and-legal-notice). Subscriber: null; date: 21 June 2018

The Beginnings of Islam in Afghanistan

The Beginnings of Islam in Afghanistan

Conquest, Acculturation, and Islamization

(p.41) 1 The Beginnings of Islam in Afghanistan
Afghanistan's Islam

Arezou Azad

University of California Press

Abstract and Keywords

Covering the period from 709 to 871, this chapter traces the initial conversion of Afghanistan from Zoroastrianism and Buddhism to Islam. Highlighting the differential developments in four regions of Afghanistan, it discusses the very earliest history of Afghan Islam both as a religion and as a political system in the form of a caliphate.  The chapter draws on under-utilized sources, such as fourth to eighth century Bactrian documents from Tukharistan and medieval Arabic and Persian histories of Balkh, Herat and Sistan. In so doing, it offers a paradigm shift in the way early Islam is understood by arguing that it did not arrive in Afghanistan as a finished product, but instead grew out of Afghanistan’s multi-religious context. Through fusions with Buddhism, Zoroastrianism, early Abrahamic traditions, and local cult practices, the Islam that resulted was less an Arab Islam that was imported wholesale than a patchwork of various cultural practices.

Keywords:   Afghanistan, Islam, Buddhism, Zoroastrianism, Conversion, Abbasid, Arabic, Persian, Bactrian

The conquest of Balkh in 708–9 marked the beginning of the Umayyad caliphate’s control over the lands that are today Afghanistan.1 Some of the people of Afghanistan rebelled against the new Damascus-based overlords. Others joined the militias that in 749 enabled the rival ‘Abbasid caliphs to take over from the Umayyads. By the ninth century, the city of Balkh was being canonized as the Dome of Islam and its Muslim intellectuals memorialized as saints with sanctuaries deeply intertwined with the Islamic identity of their city of burial. How could the Islamic caliphate become so firmly embedded in classical and late antique Afghanistan’s thousand-year-old civilization within the relatively short time span of a hundred and sixty years? What strategies did the Muslim conquerors use to establish their authority in Afghanistan and maintain an economically viable and politically sustainable engagement? These questions guide this chapter’s investigations, which serve to test the conversion models proposed by Nehemiah Levtzion, Richard Bulliet, and Richard M. Eaton for neighboring parts of the Islamic world.2

These three scholars have all emphasized that large-scale patterns of conversion cannot be explained by popular, unsubstantiated statements alluding to an immediate and violent conversion by the sword. Nor can softer methods, such as conversion through political or economic patronage (e.g., for tax reasons), provide the full answer. Rather, under the umbrella of Muslim rule, the coexistence of multiple religions in the early Islamic period initiated a centuries-long process of acculturation and adaptation of rituals and belief systems. As the following pages show, similar trends can be detected in early medieval Afghanistan.

This chapter is divided into four parts. After an overview of the geographic context and the sources for the study of early Islamic history in Afghanistan, there (p.42) follows a historical sketch of Afghanistan before the Muslim arrival and of the Islamic conquests and the transition to Muslim rule. The chapter then explores the cross-fertilization of the Muslim conquerors’ religious ideas with preexisting local beliefs in Afghanistan. Although the time period of a hundred and sixty years is broad, the state of research is still limited. For this reason, the discussions and conclusions of this chapter are necessarily tentative and anecdotal.

Contexts and Sources

In the eighth and ninth centuries, two successive caliphal dynasties—first the Umayyads, then the ‘Abbasids—ruled over parts of Afghanistan. The period covered in this chapter begins with the establishment of Umayyad rule in 709, when the Muslim armies conquered Balkh once and for all. It closes in 870 with the defacto ‘Abbasid retreat after the takeover of Balkh by the Saffarid dynasts of Sistan. So far, no detailed study of Afghanistan in this crucial period has been under-taken, partly because of scholarly neglect and partly because of the tense security situation in Afghanistan that has made access to primary sources difficult. The tradition of limiting the discipline of Afghan Studies to the history of the nation-state of Afghanistan created in the eighteenth and the nineteenth century has further restricted the study of early Islam in Afghanistan.

Yet writing the early Islamic history of the region that later became Afghanistan fills important gaps in its cultural history. Nor is it a teleological exercise, given that while the boundaries of the modern nation-state are relatively new, those of its four main regional components are not. The regions of Herat and the west; Qandahar and the south; Balkh (today’s Mazar-i Sharif) and the north; and Kabul and the east have all existed since late antiquity.3 In order to reconstruct the early Islamic history of Afghanistan, these regions need to be brought into one historical narrative. If we do so, Afghanistan obtains an early Islamic history that is based on scholarly evidence and embedded within the diachronic study of Afghanistan’s various versions of Islam that this volume provides.

It is pertinent to understand how medieval Islamic scholars described Afghanistan’s regions. They placed them within the provinces that they knew as Khurasan and Sistan. Khurasan was the Sasanian Empire’s eastern province (satrapy) and home to the cities of Balkh and Herat in what is today Afghanistan, as well as Nishapur in Iran and Merv in Turkmenistan.4 Each of these cities was an oasis town that was part of a larger district known by the same name. Our knowledge of Umayyad Khurasan is limited, largely because of the inadequate source base for Umayyad history in general.5 In The Nativist Prophets of Iran, Patricia Crone attempted to break the impasse in understanding Umayyad Khurasan by assessing the impact on the development of Islam of the rebellions and supposed heresies that burgeoned in the region during the Umayyad occupation. Through Elton Daniel’s monograph on (p.43) early ‘Abbasid Khurasan, things come into sharper focus after the ‘Abbasid revolution.6 Daniel inventoried the opposition movements that ensued after the treacherous murder in 755 of Abu Muslim al-Khurasani, the ‘Abbasid governor in Khurasan and previous leader of the covert subversion known as the da‘wa on behalf of the ‘Abbasids, at the hands of the second ‘Abbasid caliph, al-Mansur (r. 754–75).

To understand the effects of Umayyad rule on Afghanistan, historians are now turning to recently discovered documents that have been translated into English from Bactrian and Arabic by Nicholas Sims-Williams and Geoffrey Khan.7 The corpus includes some 250 documents dating from the fourth to the eighth century. Most, if not all, are believed to originate in the family archive of the local rulers of Rob, in Tukharistan. The corpus is marked by its bilingual nature, with documents written in Bactrian or Arabic following different documentary protocols and templates, sometimes within the same year. The Bactrian documents bring exciting new material to the study of Umayyad and very early ‘Abbasid Khurasan, providing a reality check on the standard sources for the study of early Islamic Afghanistan. The standard sources include the medieval Arabic political and geoadministrative accounts cited in this chapter by Ibn al-Faqih al-Hamadani (fl. 903), Ibn Khurradadbih (d. 911?), Qudama ibn Ja‘far (d. 922?), al-Tabari (d. 923), and Yaqut al-Rumi (d. 1229).8 Such purportedly universal chronicles and reports come with two main caveats, namely that they were written in the distant ‘Abbasid center, Baghdad, and that they were produced at least a century after the events that they recount. In their accounts of Khurasan, this inevitably leads to exoticisms, occasional lapses, and political biases (such as seeking to depict the Umayyads in a negative light to justify the ‘Abbasid takeover in 749).

Local histories provide another important check against these inaccuracies. The Faza’il-i Balkh, a local history of Balkh written by a certain Shaykh al-Islam al-Wa‘iz in the late twelfth century, nuances the teleological ‘Abbasid narratives of wholesale conversion.9 A recent discovery of a fragmentary twelfth-century history of Herat by ‘Abd al-Rahman Fami is another local source that will recalibrate our knowledge of Umayyad and early ‘Abbasid Afghanistan.10

After Khurasan, Sistan is the second medieval region to which a number of major Iranian and Afghan cities belonged. It included the Afghan cities and regions of Ghazna (modern-day Ghazni), Zarang, Bust, Qandahar, Kabul, Kabulistan, and Zabulistan (the lands between Ghazna and Kabul). Qandahar was better known in the medieval sources as al-Rukhkhaj and Zamindawar.11 Zarang already became an Umayyad base in 652–53, and it was from there that the Arabs undertook their eastward campaigns. An anonymous local history used mainly for the study of the Saffarid dynasty, the Tarikh-i Sistan (completed in 1062), also provides interesting details on the early years of Islam in Sistan. C. E. Bosworth’s monograph Sīstān under the Arabs, even fifty years after its publication, has not been surpassed in its extensive coverage of the subject.12

(p.44) Yet the local histories also come with their limitations, notably the obscurity of their authors and their poor manuscript-survival rates, which sometimes make it difficult to verify their accounts. For the historian in particular, another caveat is the authorial license established in the literary genres (faza’il, tabaqat, tarikh) from which the local histories grew, which allowed the authors of such texts to prioritize the ideal state of places in an Islamic imagination at the expense of factual accuracy.13 An example of these problems is the Faza’il-i Balkh, which was written in Arabic in 1214 (this original is now lost) and translated into Persian in 1278 (recensions of which survive): it has a spurious author and translator, and its surviving manuscripts date from three to seven centuries later. The main surviving local history of Badakhshan dates from as late as the eighteenth century.14 Local histories of Kabul date from even later, and there do not appear to be any surviving medieval local histories from the Qandahar region at all.15 Some of these gaps can be filled by using documentary and archaeological evidence, such as the Bactrian documents discussed earlier and the excavation reports of the Délégation Archéologique Française en Afghanistan (DAFA), the Italian archaeological mission in Ghazni, or the more recent discoveries at Mes Aynak.16

Before the Muslims Arrived

In order to test the Islamization process in Afghanistan, we must necessarily begin with the prelude and context for the Muslim conquests in Bactria, a vast province centered on northern Afghanistan and subjected to Sasanian viceregal control. Prior to Muslim rule, three out of the four Afghan regions—Balkh, Herat, and Sistan—were integrated within the Sasanian orbit. Balkh and Herat belonged to Khurasan, which was one of four Sasanian provinces governed from Merv by an ispahbadh, or Sasanian general. Effectively, this meant that the Sasanian footprint on Balkh and Herat was light. Practically, this translated into an engagement focused on the exaction of taxes and the policing of the frontiers with the Hephthalite domains in Central Asia and eastern and southern Afghanistan, and with the domains of the Chinese T’ang Empire. In economic terms, Balkh and Herat also served as Sasanian mint towns.

This light Sasanian footprint can be demonstrated rather easily in the region of Bactra (which was Arabicized to Balkh). The conquest accounts often stand out for the limited presence of Sasanian administrative and military machinery. In 705, the Umayyad general commanding Balkh’s final conquest, Qutayba ibn Muslim (d. 715 or 716), was met by the ispahbadh of Balkh and some local dignitaries. In 708–9, the ispahbadh was one of the local rulers to whom the Hephtalite rebel Nizak Tarkhan wrote when he was trying to unite against Qutayba the local aristocracy of Tukharistan (the subregion, north of the Hindu Kush, of which Balkh was the capital). The barmak, the leader of the Buddhist Naw Bahar (p.45)

The Beginnings of Islam in AfghanistanConquest, Acculturation, and Islamization

Figure 6. Site believed to be the Naw Bahar Temple of Balkh.

(Photograph © Arezou Azad)

monastery, also appears to have been one of the architects of this revolt; his wife was taken prisoner.17

The lack of any reference in the sources to a princely leader in Balkh at the time of Qutayba’s advances may be explained by the power of the barmak. The source of his power was the Naw Bahar, which was not only a Buddhist religious complex but also a hub for major landholding, agricultural, and revenue-generating enterprises that extended over two-thirds of the large Balkh oasis, an area of more than 72 square kilometers.18 The transmission of Sasanian origin myths in the sources on Balkh (notably the Zoroastrian legends that the prophet Zoroaster died there and that his patron, Gushtasp, had built the city), as well as the presence of Zoroastrian fire temples in the region, points to Sasanian cultural and religious influences.19 The archaeological remains at Chashma-yi Shafa‘, near Balkh, currently under investigation by the DAFA, may also support this argument.20

Looking beyond Balkh, in Tukharistan’s rural metropolises of Samangan and Rob, we find an even weaker Sasanian presence. While making frequent references to Sasanian taxes, the Bactrian documents from this era provide rather scant (p.46) notice of Zoroastrian or Buddhist deities. In this part of late Sasanian and early Islamic Tukharistan, the principal objects of worship were a set of local deities, such as Zhun, Wakhsh, and Kamird.21 Like that of pre-Islamic Balkh, the population of Tukharistan was diverse: people spoke and read many languages, including Bactrian, Turkic, Syriac, Pali, and Sanskrit.22 Since the Bactrian documents are discussed in more detail below, it is worth noting that Bactrian belonged to the Iranian language group and was written in Cyrillic script. It was a legacy of the conquests of Alexander the Great and the Seleucid Empire that emerged from them between the fourth and the first century B.C.E.

Once we reach the southern point of the region of Balkh, at Bamiyan, indications of a possible Sasanian presence diminish quickly. The Chinese Buddhist pilgrim Xuanzang traveled through the area in the 630s and described the two colossal Buddha statues at Bamiyan, which were probably built in the sixth or the seventh century.23 Late antique Bamiyan appears to have been ruled by a dynasty that was perhaps of Hephthalite origin but was certainly subject to the prince (yabghu) of the western Turks. The dynasty was still ruling in the first quarter of the eighth century, when it continued to profess Buddhism.24

With regard to Herat farther west, “Harev” (hryw) is listed in the inscription of the Sasanian ruler Shapur I (r. 239–70) at Ka’ba-yi Zardusht, near Persepolis, in Iran. The name “Hariy” (hr’y) is also mentioned in the Pahlavi (middle Persian) catalogue of the provincial capitals of the empire.25 Moreover, Sasanian seals and engraved gemstones have been found in and around Herat.26 Kushano-Bactrian coins have also been studied, revealing Sasanian fire temples and other iconography typical of Bactrian coins from this period.27 However, the Sasanian grip did not extend into Herat’s hinterland. There, the northern branch of the Hephthalites and their political successors continued to rule in the Ghurid mountains and river valleys well into the Islamic period.

The third area that was part of the Sasanian domains was Sistan, a region that was south of the Hindu Kush in Afghanistan (centered around Zarang and Nimruz) and in southwestern Iran (centered at Zahidan).28 The area was a shallow basin in which civilizations clustered around the Helmand riverine areas. Sistan—the Arabic form of the name—derives from the Middle Persian Sakastan, which is also mentioned in another inscription of the Sasanid ruler Shapur I at Naqsh-i Rustam, near Persepolis.29 When the Umayyad Muslim armies established their base at Zarang in 652/3, Zoroastrianism was well established, marked by the presence of a chief mobadh and hirbadh, while the major fire temple of Karkuya continued to function long after the Umayyad conquest.30 The Nestorian Church was represented with a bishopric that continued well into the Islamic period.31 However, there remained one part of the Sistan region that was not under Sasanian or Islamic control for two centuries, namely Qandahar (viz. al-Rukhkhaj and Zamindawar).

(p.47) This leads us to the fourth region, which was Kabul and Zabulistan, the latter lying between Kabul and the Kabul River Valley on the north and the territories around the confluence of the Helmand River and Arghandab River of Qandahar and based around the city of Ghazna. From the third century B.C.E. to the first century C.E., the region had formed part of the Greco-Bactrian realm. It was subsequently overrun by nomadic tribes from the north, including the Kushans, Kidarites, and Hephthalites. Buddhism flourished there, as did Indic cults. The Zunbils and Kabulshahs were persistent in their resistance to Muslim rule for some two centuries, until the Saffarids of Sistan temporarily subdued them in 870.32 This did not prevent the Muslim armies and traders from carrying out business, in which Kabul served as an entrepôt for the India trade. The region was not brought into the dar al-Islam (realm of Islam) until the Ghaznavid conquests of the eleventh century.

Conquest and the Transition to Caliphal Administration

Back in the seventh and eighth centuries, the Umayyad Muslim conquests had toppled the Sasanian Empire. The fall of the Sasanian administrative capital at Ctesiphon (20 miles southeast of modern Baghdad) in 637 and the assassination of the last Sasanian emperor, Yazdgird III, in 651 at Merv had a domino effect. Farther east, the Muslim conquests were by no means immediate. The population of Afghanistan went through a repetitive pattern of submission, rebellion, and resistance, and eventually final submission. Balkh was the first city to come under direct Umayyad control, after its conquest by the Umayyad general Qutayba ibn Muslim in 709. Having previously garrisoned twelve kilometers away at al-Baruqan, the Muslim soldiers and their mawla (clients of Arab tribesmen) moved into the city of Balkh in 724/5.33

The Umayyad dynasty ruled over Khurasan from its regional command center at Merv, following the example of the Sasanians. It was not the only continuation of Sasanian practice. The Barmakid family, who had previously run most of the district through the Buddhist Naw Bahar estate, were now proving to be useful local operators for the Umayyads. The caliphate was stretched far from its capital at Damascus, and needed to co-opt and eventually assimilate local power-holders. Among the most popular medieval Islamic conversion narratives is that of the formerly Buddhist Barmakid family. Islamic accounts describe the head barmak as a Muslim mawla who voluntarily traveled to Syria to declare his loyalty to the caliph Hisham ibn ‘Abd al-Malik (r. 723–42). The Barmakids’ influence increased after their support for the ‘Abbasid struggle against the Umayyads, leading to the ‘Abbasid takeover of the caliphate in 749.34 The next generation of Barmakids formed the immediate entourage of the caliph Harun al-Rashid (p.48) (r. 786–809) in the new caliphal capital, Baghdad. There, Yahya ibn Barmak (d. 805) and his sons Fadl (d. 808) and Ja‘far (d. 803) served as vizier and governor, respectively. In the tales of the Arabian Nights, the relationship between Harun al-Rashid and the Barmakids was turned into legend. The Barmakids had become an Afghan dynasty that permeated the highest echelons of caliphal power, both in the imperial center in Baghdad and in the vital province of Khurasan, where they participated in the creation of caliphal policies and practices. Yet they did not go native in Baghdad—at least not entirely. Rather, the Barmakid family used their contacts and influence in the western and eastern lands of the caliphate to bring together two previously disparate worlds. Yahya ibn Barmak (d. 805), for example, commissioned translations of Sanskrit texts and patronized a versified life of the Buddha.35

Other indigenous winners during the first hundred and sixty years of caliphal rule in Afghanistan were the local rulers of Rob, a town lying a hundred and thirty kilometers to the southeast of Balkh, in Tukharistan. The Arabic documents from the Bactrian corpus attest the caliphal tax administration in this rural area, which required land surveys for the calculation of land-tax assessments and the issuance of tax receipts. The Bactrian documents contain the very first reference to the caliphate’s kharaj tax on agricultural land and its produce, two decades before the next-earliest reference appears in the corpus of Egyptian papyri. As Geoffrey Khan has argued, the term kharaj stems from a Middle Iranian etymology, thus pointing to the Muslim adoption of elements of the pre-Islamic tax system of Khurasan.36 Influential landholders, who colluded with the Umayyads and early ‘Abbasids and became their mawlas (clients), reaped rewards. A case in point is the Kamird-far family, to whom the bilingual Bactrian document corpus probably belonged. The family managed large tracts of arable land and orchards for which they were consistently taxed.

But, as might be expected, not all local rulers or members of the old guard won out or supported the Umayyad project. Rival local elites vied for influence with the Muslim newcomers or used the new overlords to help settle old accounts. The khar of Rob, for example, showed Qutayba ibn Muslim an alternative access route to the fortress where his nemesis, the apostate-rebel Nizak Tarkhan, was hiding.37 Along with his family and supporters, Nizak eventually met with an inglorious end: the Umayyads slaughtered thousands of them and ate bread made from flour milled with the copious amounts of blood that flowed from the massacre.38

The Umayyads in Afghanistan faced opposition again after they reinstated the poll tax on new converts. Rebellions broke out in the Balkh, Herat, and Sistan regions. The uprising of a rebel named al-Harith ibn Surayj was particularly popular in the areas around Balkh, lasting from 734 to 746. The Baghdad-based chronicler al-Tabari (d. 923) reported that al-Harith followed the doctrine of the Murji’ite, who professed that faith alone was sufficient to be a Muslim.39 This attracted support from those converts who were told that their conversion was not valid unless (p.49) accompanied by ritual acts. Al-Harith challenged the selection process for the governorship of Balkh, directly confronting the governor Nasr ibn Sayyar (d. 748) and even the caliph about the matter. Almost immediately afterward, the Umayyads were challenged again by rebels with strong support in Khurasan. This time the rebellion was led by Abu Muslim and the ‘Abbasids, who also sought to bring the caliphate into line with Islamic principles. In 749, they succeeded in overthrowing the Umayyad caliphate and establishing their own in its place.

Moving westward, Herat was only of secondary importance to the Umayyads compared with other Khurasani cities like Merv, Nishapur, and Balkh. As a result, we read very little about Herat in the sources on this period, though Herat’s status would certainly change by the time of the Timurids, as shown in the chapters in this volume by Nushin Arbabzadah and Jürgen Paul. One popular account describes the rebel Ustadhsis leading a major opposition to the early ‘Abbasids from 767 to at least 770, during which he took control of Herat and Bushanj. Ustadhsis killed hundreds of Arab tribesmen and numerous ‘Abbasid leaders in the process. He had converted to the Zoroastrian reformist doctrines propounded by Bihafarid (d. ca. 748 or 749), which embraced practices and prohibitions inspired by Islam. Ustadhsis’s supporters are said to have numbered three hundred thousand.40 He eventually escaped to a fort in Badghis, and thereafter Khazim ibn Khuzayma, the general of the ‘Abbasid caliph al-Mansur (r. 754–75), defeated the rebels and arrested Ustadhsis.41

In the south, Sistan’s capital, Zarang, surrendered to the Muslim armies in 652/3. It subsequently became an important base from which Umayyad and early ‘Abbasid campaigns were undertaken in the east against the zunbils, local rulers of al-Rukhkhaj (Qandahar) and Zamindawar, as well as against the Kabulshah rulers of Kabul. Umayyad coins of the Arab-Sasanian type were minted in Zarang, and taxes were collected. The high tax burdens placed on the local population seem to have contributed to a rise in support for the Kharijites, who had fled from Kerman, in Iran, where they had earlier been defeated by the Umayyads.42 The Kharijites continued their militant campaigns in the small towns of Khurasan and Sistan of Afghanistan until the reign of the Saffarid dynasty between 861 and 1003. The vulnerability of the local population in Sistan and Bust to Kharijite attacks lent a raison d’être to the bands of local Muslim fighters known as ‘ayyarun. In time, these fighters gave rise to the Saffarid dynasty, which would go on to end non-Muslim control of Kabul.43

Control over Afghanistan brought financial gain to the Umayyads and ‘Abbasids alike. For example, according to Ibn Khurradadbih (d. ca. 911), in the financial year 826–27, Khurasan alone contributed 44.8 million dirhams of kharaj tax to the caliphal coffers.44 The loss of this tax revenue to the Saffarid dynasty is frequently cited as having precipitated the overall demise of the ‘Abbasid caliphate.45 Control over the cities of Afghanistan, such as Balkh, also enabled direct access into the (p.50) lucrative trade with India via the land routes of the so-called Silk Road. By tapping into existing mercantile networks, the caliphate linked up with the wealthy kafir (infidel, non-Muslim) lands of India and Central Asia, thus bringing such exotica as elephants from India and musk from Tibet to the Islamic world and beyond. The Muslim chroniclers tell us that such long-distance trade filled Afghan bazaars with horses, camels, dried fruit, wine, spices, sweets, and textiles.46

A New Religious Language

The transition to caliphal rule also brought the new religion of Islam to Afghanistan. So how did Islam embed itself in Afghan society? More specifically, who were its proponents, and how did they transmit their new religious knowledge? What was the rate of conversion to Islam, and what may have motivated people to convert? And what strategies may people have used to retain local practices within the syncretistic environment that the caliphate inadvertently created? The following section tackles these questions by considering the case of Bactria (that is, Balkh and Tukharistan). In Balkh, as elsewhere in the caliphate, the proponents of Islam were the Muslim scholars, or ‘ulama. Forty of the seventy ‘ulama of Balkh described in the Faza’il-i Balkh lived there during the first two centuries of Islam. They were an eclectic mix of Arabs and mawla converts who had come from all over the caliphate, from the Nile to the Oxus. The ‘ulama in the early days of Islam were not professional Muslims as such, since their pursuit of Islamic scholarship was largely a private vocation, or at best a part-time job. The forty Muslim scholars of Balkh, who are described in the Faza’il-i Balkh at variable lengths of between half a page and more than thirty pages each, studied and memorized the Quran and its exegeses, along with hundreds of Hadith. They did this while keeping their day jobs as merchants, landowners, and suchlike: to be an ‘alim, a religious scholar, required private wealth.47 This presents a great contrast with the lucrative later economy of Islam discussed in R. D. McChesney’s chapter in this volume, which reveals various options for earning a living through religion by the sixteenth century.

The knowledge acquired by Balkh’s ‘ulama was passed down through the scholarly generations known as silsilas (chains or lineages). The method of teaching was the suhba (disputation) carried out during a gathering (majlis) held in an informal setting such as a private home or a room in a mosque. This lack of educational formality gave access to Islamic learning to women, some of whom became scholars and teachers in their own right.48 The majlis gatherings provided the glue for the silsila lineages of students and disciples who continued the transmission of religious knowledge.49 This trajectory of Muslim education and professionalization shows that Balkh resembled other early Muslim religious centers farther west, most notably Damascus and Baghdad.50

(p.51) As a corollary of their religious knowledge, a large proportion of Balkh’s ‘ulama distinguished themselves as pious ascetics (zuhhad). Men such as Ibrahim ibn Adham Mansur (d. 777/8) and Shaqiq ibn Ibrahim al-Zahid al-Balkhi (d. 809/10) were especially celebrated for their asceticism. A zahid (ascetic) was someone who was unconcerned with the world, but not necessarily someone seeking mystical communion with God. The latter could possibly be a Sufi. However, the Faza’il-i Balkh does not use the term “Sufi.” It is only from the fifteenth century onward in the Sufi hagiographical literature discussed later in this volume, in the chapter by Jürgen Paul, that these early mystics were retrospectively institutionalized as the founding fathers of Sufism. They are still remembered as such today.

Balkh’s early Muslim scholars also included legal experts who advised and judged on the application of Shari‘a in people’s daily lives. Their remit was both broad and specific, covering a wide range of issues that included diet, personal hygiene, inheritance, property, and marriage rights. Despite the fact that they were appointed directly by the caliph, the early qazis (judges) of Balkh were powerful and independent. The qazis presented themselves as protectors of the general populace against the corruption and impunity of political rulers.51 Their confrontational stance against the political establishment may have led to a purported disbandment of the qazi establishment in Balkh in the ninth and tenth centuries. In its stead, the political authorities installed a system of courts known as mazalim, which specialized in the redress of grievances. By this means, complaints by members of the public would be heard and judged directly in public by the political leader, whether the governor or the vizier.52

While we can identify the proponents of early Islam, thanks to the written testimony left by the ‘ulama who were the main authors of our primary sources, it is much harder to assess the effects of Islam on the wider population. Richard Bulliet has estimated that by the mid-eleventh century this part of the caliphate acquired a 90-percent Muslim population.53 However, his method of using personal names as indications of conversion is arguably problematic. Moreover, such quantitative data do not tell us about the qualitative nature of events. What motivated people to convert, and under what circumstances did they change their religion? Here the Bactrian documents provide a significant piece in the puzzle, giving us a direct and unfiltered look into people’s daily lives from the Umayyad period to the first two decades of ‘Abbasid rule in the 770s. A study of the documents from both the Arabic and Bactrian language groups brings to life the case of Kamird-far (also known as Sa‘id) and Zeran over a five-year period from 750 to 755. It is to this case study that we now turn.

In Bactrian document BT I X, dated to the year 750 (E.B.D. 527), three out of four brothers (named Kamird-far, Wahran, and Mir) agreed to own homes and estates equally, and to “possess the woman [Bactrian zin] whose name [is] Zeran …, as (p.52) it is not necessary for us to destroy our house.”54 The document provides crucial evidence that in Rob, Tukharistan, fraternal polyandry was practiced until at least this late in the period of Islamic rule. The corpus of Bactrian documents ends here, and we cannot know how long the practice continued. But if we are to believe some of the Arabic heresiographical literature, or al-Biruni’s History of India (completed in 1030), then fraternal polyandry continued to be practiced until the first half of the eleventh century, and possibly longer.55 That the practice had an early precedent is absolutely clear from the very first document in the Bactrian corpus. This is a marriage contract between two brothers and one woman called Ralik, dated to the year 333 (110 E.B.D.). The contract emphasizes that this practice “is the established custom in the land,” meaning that it was already in existence before the mid-fourth century.56 The document also includes a justifying clause about “the need to keep the house together.” Thus, the practice of fraternal polyandry attested in the mid-eighth century may well be a continuation of an age-old custom from this part of Bactria. The impetus behind the practice was principally financial. Bactrian households were taxed in house units, which necessitated minimizing the inheritance into one line, thus limiting it to one wife shared by two or more brothers.57

However, the story does not end here. Arabic document 29 (dated to 755) may refer to Kamird-far, one of the three brothers who married Zeran. The reference here is to the marriage of a convert to Islam called Sa‘id (the Arabic name for Kamird-far) with a woman called Zeran, a manumitted slave who bore him four children. If this is indeed the same Kamird-far as in document BT I X, then we may well ask what led to this sudden shift from fraternal polyandry. The answer seems to lie again in the fiscal system that gave rise to fraternal polyandry in the first place: because the caliphate changed its house-based tax system into a tax on individuals. Once this Muslim tax system came into being, it was no longer necessary to share wives.58

The third question to be explored relates to the antiquity of early Islamic Afghan rituals, and specifically the extent of Islamic syncretism with preexisting Afghan religious practices. Even in the Muslim homeland of the Hijaz, syncretism has been the mainstay of Islam. For example, both Muslim and non-Muslim scholars agree that in the hajj Muhammad reinstituted a preexisting pilgrimage ritual. Several Western scholars also hold that the Prophet incorporated rituals from Arabian paganism, litholatry (that is, stone worship), and even Judaism.59 In Balkh, the situation was no different. The Faza’il-i Balkh makes reference to the veneration of multiple sacred sites, notably shrines built for the ‘ulama in Balkh between the eighth and the twelfth century. In the eyes of Shaykh al-Islam al-Wa‘iz, the author of the Faza’il-i Balkh, what made these shrines sacred was not only the saintly body that lay buried in them but the pre-Islamic antiquity of the places where they were built. The author attributed to these places a history that went back as far as the Old (p.53) Testament. All twenty-seven of the shrines to ‘ulama located within the city of Balkh were concentrated in five particular points in the city. One by one, Shaykh al-Islam al-Wa‘iz wove these sites into a narrative of sacred landscape. When superimposed on a map of medieval Balkh, the sites appear in a mandala-like constellation along the cardinal points of the compass and around a mound located at their spiritual (if not spatial) center.

It is not inconceivable that the Muslim sacred landscape of Balkh has a Buddhist past. Buddhism began to be institutionalized in the second century by the court of the Kushan king Kanishka I (r. ca. 127–ca. 140). Later, in the seventh and eighth centuries, Balkh’s landscape was described by Chinese pilgrims as being dotted with hundreds of Buddhist stupa shrines. Among Balkh’s Buddhist monasteries, the Nava Vihara was by far the largest. The Nava Vihara’s monks studied the religious teaching of the Shravakayana (Vehicle of Listeners), to which the Chinese pilgrim Xuanzang referred by the pejorative name of Hinayana (Lesser Vehicle).60 After the Islamic conquest, the monastery’s Sanskrit name, Nava Vihara (which means “New Monastery”), was Persianized to Naw Bahar and became the name-sake of one of Balkh’s three sacred gateways.

In the hypothetical but very possible scenario that, during the first century and a half of Islamic rule, Muslims adopted existing sacred sites and added to them, Balkh appears as a highly syncretic sacred landscape. In this view, the Buddhist past was melded with origin narratives from Zoroastrian and biblical sources. For example, the oldest of the five sacred sites in Balkh that al-Wa‘iz listed in his Faza’il-i Balkh is Gushtasp’s Mound (tall-i Gushtasp). He attributed to it the following biblical connection:

According to Anas ibn Malik … it is related by the Prophet (may God’s prayers and peace be upon him and his family) that Job the Forbearer (may God’s prayer’s be upon him) is at rest on Gushtasp’s Mound.61 At each gate there are seventy thousand angels praying for God’s mercy and praising and glorifying him, and the recompense for that will be bestowed upon the people of Balkh.62

While the Buddhist meaning of Gushtasp’s Mound was lost by the time al-Wa‘iz wrote this account, in the late twelfth century, its sacredness was retained. In Zoroastrian tradition, Gushtasp had been the royal patron of Zoroaster. Such a continuity of sacredness has many parallels elsewhere in the Islamic world. By comparison, when looking at dozens of Anatolian sanctuaries that were transferred from one religion to another, F. W. Hasluck concluded in the 1920s that narratives of sacredness perpetuate the idea that particular places—both natural and man-made—have a quality that outlives the vicissitudes of time.63 Naturally, Muslims did not adopt sites such as Gushtasp’s Mound without reconfiguring their symbolic meanings within Islamic language and imagery. But these sites nonetheless continued to be sacred and iconic. (p.54)

The Beginnings of Islam in AfghanistanConquest, Acculturation, and Islamization

Figure 7. The Naw Bahar, site of a fifth-through eighth-century Buddhist monastery, Balkh, 2009.

(Photograph © Arezou Azad)


This chapter’s study of Afghanistan’s early encounter with Islam has highlighted the fact that the historical development of Islam in the region varied between the areas to the north and the south of the Hindu Kush, and those between the east and west of the country. In the eighth and ninth centuries, the main interest of the Umayyads and ‘Abbasids lay in the north and southwest, in Balkh, Tukharistan, Sistan, and, to a lesser extent, Herat. In order to ensure these regions’ political and economic allegiance to the distant caliphal capitals, the agents of the caliphate co-opted Afghan elites, including the Barmakids of Balkh and the khars of Rob. Islamization progressed over centuries in a slow and adaptive process, during which time Afghanistan saw the continuation of old religious practices within an Islamic language. By the tenth century, Afghanistan’s Islamic scholars in the north and southwest had developed scholarly and professional religious specializations in much the same way as elsewhere in the caliphate. But in the south and southeast (that is, in Kabul, Zabulistan, and Qandahar), Islamization did not begin until the tenth and eleventh centuries. Even so, this did not hinder the caliphate from establishing firm trade links with such kafir lands so as to tap into the high-value India trade.

Much as in other parts of the caliphate, early Islam developed in a syncretic environment in the lands that we now call Afghanistan. But unlike anywhere else in the caliphate, in Afghanistan Buddhism fused with Zoroastrianism, early Abrahamic traditions, and local cult practices, as well as with Islam. By looking at this unique context, this chapter has tried to test the models proposed by Levtzion (p.55) and others. The conclusion must be not that Islam embedded itself in medieval Afghan society as a finished product but rather that the medieval people of Afghanistan gave early Islam some of its shape and color. For example, the early Murji’ite movement in Balkh (which according to the Faza’il-i Balkh was even known as Murjiabad, “Land of the Murji’ites”) enabled people to convert to Islam without abandoning their age-old religious practices and rituals. Meanwhile, the Muslim conquerors and their local clients developed power-sharing strategies that secured an economically viable and politically sustainable Muslim engagement in Afghanistan. From the very early stages of Islamic history, this cocktail of religious combinations led to multiple versions of Islam in Afghanistan, which form a fitting backdrop to the developments described in the following chapters.


(1.) This chapter uses the term “Afghanistan” to denote the territory that is today circumscribed within the boundaries of the Afghan nation-state. The term “Afghan” is used in the same vein, without ascribing to it any modern ethnic value.

(2.) See in particular Nehemiah Levtzion, “Towards a Comparative Study of Islamization,” in idem (ed.), Conversion to Islam (New York: Holmes & Meier, 1979). Levtzion distinguishes between the individual “reorientation of the soul” from the “adhesion,” which typically is a communal process entailing the “acceptance of new worships as useful supplements and not as substitutes” to what went before. See also Richard Bulliet, Conversion to Islam in the Medieval Period (Cambridge, Mass.: Harvard University Press, 1979); Richard M. Eaton, “Approaches to the Study of Conversion to Islam in India,” in Richard C. Martin (ed.), Approaches to Islam in Religious Studies (Tucson: University of Arizona Press, 1985); (p.264) and a more recent summary of the scholarship on Islamization in Arietta Papaconstantinou, “Introduction,” in Conversion in Late Antiquity: Christianity, Islam, and Beyond; Papers from the Andrew W. Mellon Foundation Sawyer Seminar, University of Oxford, 2009–2010 (Farnham: Ashgate, [2015]).

(3.) Peshawar and the Northwest Frontier Province constitute a fifth region, which Thomas Barfield has aptly referred to as Afghanistan’s “phantom limb.” It was handed to Pakistan when the British departed. See Thomas Barfield, Afghanistan: A Cultural and Political History (Princeton: Princeton University Press, 2010), pp. 47–54.

(4.) The name should not be confused with the modern province of Khurasan, in Iran. In Sasanian and early Islamic times, the province covered much of today’s former Soviet Central Asia and Afghanistan: C. E. Bosworth, “ Khurāsān,” Encyclopaedia of Islam, 2nd edition (henceforth EI2), vol. 5 (Leiden: Brill, 1986), pp. 55–59.

(5.) Some inroads are being made through the study of documents and rock graffiti. See, for example, Robert Hoyland, In God’s Path: The Arab Conquests and the Creation of an Islamic Empire (New York: Oxford University Press, 2015), pp. 100 ff.

(6.) Elton Daniel, The Political and Social History of Khurasan under Abbasid Rule, 747–820 (Minneapolis: Bibliotheca Islamica, 1979).

(7.) Nicholas Sims-Williams, Bactrian Documents from Northern Afghanistan, vol. 1, Legal and Economic Documents, Studies in the Khalili Collection 3,rev. ed. (Oxford: The Nour Foundation in association with Azimuth Editions and Oxford University Press, 2012 [2000, 2001]); Geoffrey Khan, Arabic Documents from Early Islamic Khurasan, Studies in the Khalili Collection 5 (London: Nour Foundation in association with Azimuth Editions, 2006); Nicholas Sims-Williams, Bactrian Documents from Northern Afghanistan, vol. 2, Letters and Buddhist Texts, Studies in the Khalili Collection 3 (London: The Nour Foundation in association with Azimuth Editions, 2007).

(8.) Ibn al-Faqih, Kitab al-Buldan, ed. M. J. de Goeje (Leiden: Brill, 1967); facsimile Mashhad manuscript in Fuat Sezgin (ed.), Collections of Geographical Works by Ibn al-Faqīh, Ibn Faḍlān, Abū Dulaf al-Khazrajī (Frankfurt a.M.: Institute for the History of Arabic-Islamic Science, 1987); Qudama ibn Ja‘far, Kitab al-kharaj, ed. M. J. de Geoje, part 6, Bibliotheca Geographicorum Arabicorum (Leiden: Brill, 1889); al-Tabari, Tarikh al-rusul wa-l-muluk, ed. M. J. de Goeje, in 15 vols. and 3 series (Leiden: Brill, 1879–1901); Yaqut al-Rumi, Mu’jam al-buldan, in 5 vols. (Beirut: Dar Sadir, 1955).

(9.) For further details, see Arezou Azad, “The Fada’il-i Balkh and Its Place in Islamic Historiography,” Iran 50 (2012), pp. 79–102. Some of the local histories have continuations up to the near-present.

(10.) ‘Abd al-Rahman Fami Harawi, Tarikh-i Harat, facsimile reprint (Tehran: Miras-i Maktub, 1387/2008). It also provides an important parallel source for the Tarikh-i Harat, by Sayfi Harawi, written some three hundred years later.

(11.) Guy Le Strange, Lands of the Eastern Caliphate (Cambridge: Cambridge University Press, 1930 [1905]), pp. 334–51.

(12.) C. E. Bosworth, Sīstān under the Arabs: From the Conquest to the Rise of the Ṣaffārids (30–250/651–864 (Rome: IsMEO, 1968).

(13.) For discussions of early Islamic local histories, see Richard Frye, “City Chronicles of Central Asia and Khurasan: Ta’rix-i Nisapur,” in Zeki Velidi Togan’a armạğan: Symbolae in honorem (p.265) Z. V. Togan (Istanbul: Maarif basĭmevi, 1950–55), pp. 405–20; Richard Bulliet, “City Histories in Medieval Iran,” Iranian Studies 3 (1968), pp. 104–9; A. K. S. Lambton, “Persian Local Histories: The Traditions behind Them and the Assumptions of Their Authors,” in Biancamaria Scarcia Amoretti and Lucia Rostagno (eds.), Yad-nama in memoria di Alessandro Bausani (Rome: Bardi, 1991); Charles Melville, “Persian Local Histories: Views from the Wings,” Iranian Studies 33 (2000), pp. 7–14.

(14.) Mirza Sang Muhammad Badakhshi, Tarikh-i Badakhshan, ed. Manuchihr Sutuda ([Tehran:] Mu’assasa-yi farhangi-yi jahangiri, 1367/1988); idem, Tarikh-i Badakhshan = Istorii͡a Badakhshana, ed. Aleksandr Nikolaevich Boldyrev (Leningrad: Izd-vo Leningrad-skogo un-ta, 1959).

(15.) Fayd Muhammad Katib Hazarah, The History of Afghanistan: Fayż Muḥammad Kātib Hazārah’s Sirāj al-tawārīkh, trans. R. D. McChesney and Mohammad Mehdi Khorrami, 3 vols. (Leiden: Brill, 2013).

(16.) On northern Afghanistan, see Roman Ghirshman, Les chionites-hephtalites, MDAFA series (Cairo: Imprimerie de l’Institut Français d’Archéologie Orientale, 1948); Jean-Claude Gardin, Céramiques de Bactres, MDAFA series 15 (Paris: Klincksieck, 1957); Marc Le Berre and Daniel Schlumberger, “Observations sur les remparts de Bactres,” in Bruno Dagens, Marc Le Berre, and Daniel Schlumberger (eds.), Monuments préislamiques d’Afghanistan, MDAFA series 19 (Paris: Klincksieck, 1964); and other MDAFA reports. On Ghazni: Anna Filigenzi and Roberta Giunta (eds.), Fifty Years of Research in the Heart of Eurasia: The IsIAO Italian Archaeological Mission in Afghanistan 1957–2007; Proceedings of the Symposium Held in the Istituto Italiano per l’Africa e l’Oriente, Rome, January 8th, 2008 (Rome: Istituto Italiano per l’Africa e l’Oriente, 2009). The currently active, major archaeological site of Mes Aynak in Logar district (spanning 4,000 ha) lies in the ancient territories of Kabulistan and has sanctuaries, artifacts, and Buddhist manuscripts dating from the Kushan and Kushano-Sasanian periods up to the eighth century A.D.: Omara Khan Massoudi, Mes Aynak: New Excavations in Afghanistan (Chicago: Serindia, 2011). An earlier archaeological project (active mainly in the 1970s and 1980s) at Ghazni itself, on the Buddhist site of Tepe Sardar, has unearthed a major sanctuary that functioned until at least the eighth or ninth century A.D., with more than two thousand artifacts, including Buddhist sculptures.

(17.) Al-Tabari (1879–1901; above, note 8), vol. 2, pp. 1181, 1206–7, 1219; Al-Wa‘iz, Fada’il-i Balkh, ed. ‘Abd al-Hayy Habibi (Tehran: Intisharat-i Bunyad-i Farhang-i Iran, 1350/1971), p. 34. Étienne de la Vaissière has recently argued that “the title barmak derives directly from paramaka, a title that broadly means ‘excellent’ or ‘superior’, without any need to go to the more distant pramukha [i.e., an important figure in the monastic hierarchy, as attested in Khotanese texts].” See Étienne de la Vaissière, “De Bactres à Balkh, par le Now Bahar,” Journal Asiatique 298.2 (2010), p. 531.

(18.) Ibn al-Faqih (1967; above, note 8), pp. 322–24; and facsimile ms (1987; above, note 8), fols. 321–24.

(20.) Roland Besenval and Philippe Marquis, “Le rêve accompli d’Alfred Foucher à Bactres: Nouvelles fouilles de la DAFA, 2002–2007,” Comptes Rendus des Séances de l’Académie (p.266) des Inscriptions et Belles-Lettres, 151.4 (2007), pp. 1847–74. Also relevant are the early Islamic accounts of the Naw Bahar that refer to the temple of the mughan (Magians: i.e., Zoroastrians). Such terms were used rather inaccurately by medieval Islamic scholars, like Ibn al-Faqih and al-Wa‘iz al-Balkhi, as catchall phrases for the followers of religions that were not ahl al-kitab (Christian or Jewish “people of the book”). Terms on Buddhism were imprecise and greatly corrupted. For an excellent excursus on this historiographical phenomenon, see Daniel Gimaret, “Bouddha et les bouddhistes dans la tradition musulmane,” Journal Asiatique 257 (1969), pp. 273–316.

(21.) “Wakhsh, the king of gods” (BT I O and U, dated 440 E.B.D./663 A.D. AND 490 E.B.D./713 A.D., respectively), a god called Ram-set (BT I P and Q, dated 446 E.B.D./669 A.D. AND 449 E.B.D./672 A.D., respectively), and “Kamird, the king of gods” (BT I T, dated 478 E.B.D./700 A.D.). (For E.B.D., see below, note 54.) All documents are published in Sims-Williams, Bactrian Documents from Northern Afghanistan, vol. 1. For a more detailed discussion, see Arezou Azad, “Living Happily Ever After: Fraternal Polyandry, Taxes and ‘the House’ in Early Islamic Bactria,” Bulletin of the School of Oriental and African Studies 79.1 (2016), pp. 33–56.

(22.) Nicholas Sims-Williams, “Bactrian Language,” Encylopaedia Iranica (henceforth EIr), vol. 3 (London: Routledge and Kegan Paul, 1989), pp. 344–49; idem, Bactrian Documents, vols. 1 and 2; Édouard Chavannes, Documents sur les Tou-Kiue (Turcs) occidentaux, receuillis et commentés, suivi de notes additionnelles … avec une carte (Paris: Adrien-Maisonneuve, 1942 [1903]); P. Y. Saeki, The Nestorian Monument in China (London: Society for Promoting Christian Knowledge, 1916 [rev. ed. Tokyo: The Maruzen Company, Ltd., 1937]). Jewish communities, too, have been identified through references to religious thinkers like Hiwi Balkhi (ninth-century). The recently discovered corpus of eleventh-century Judeo-Persian documents attributed to the region of Tukharistan and twelfth/thirteenth-century tombstones in Jam (Herat/Ghur region) point to a long heritage. See various works by Shaul Shaked, notably “Epigraphica Judaeo-Iranica,” in I. Ben Ami, N. A. Stillman, and S. Morag (eds.), Studies in Judaism and Islam Presented to S. D. Goitein (Jerusalem, 1981); “New Data on the Jews of Afghanistan in the Middle Ages,” Pe’amim 79 (1999), pp. 5–14 (in Hebrew); “A Note on ‘Hebrew-Script Tombstones from Jam, Afghanistan,’” Journal of Jewish Studies 61 (2010), pp. 305–7; and “Early Persian Documents from Khorasan,” Journal of Persianate Studies 6 (2013), pp. 153–62.

(23.) Xuanzang, Si-Yu-Ki: Buddhist Records of the Western World, ed. and trans. Samuel Beal, 2 vols. (London: K. Paul, Trench, Trübner and Co., 1906). For a recent study of the Buddhas of Bamiyan and a useful review of literature on the archaeological site, see Llewellyn Morgan, The Buddhas of Bamiyan (London: Profile Books, 2012), pp. 205–6.

(24.) ́ Edouard Chavannes (1942 [1903]; above, note 22), pp. 291–92; and J. Hackin, A. Godard, and Y. Godard, Les antiquités bouddhiques de Bamiyan (Paris: G. van Est, 1928). The Hephthalite connections are discussed by Ghirshman (1958; above, note 16).

(25.) Joseph Markwart, A Catalogue of the Provincial Capitals of Eranshahr (Rome: Pontificio Istituto Biblico, 1931): 11, 46.

(26.) H. Torrens, cited in W. J. Vogelsang, “Herat, ii: History, Pre-islamic Period,” EIr, vol. 12 (London: Routledge and Kegan Paul, 2004), pp. 205–6.

(27.) A. H. Dani and B. A. Litvinsky, “The Kushano-Sasanian Kingdom,” in B. A. Litvin-sky, Chang Kuan-ta, R. Shabani Samghabadi (eds.), History of Civilizations of Central Asia, vol. 3, The Crossroads of Civilizations, A.D. 250 to 750 (Paris: UNESCO, 1996).

(p.267) (28.) Sistan was already mentioned in the Shahnama: C. E. Bosworth, “Sīstān,” EI2, vol. 9 (Leiden: Brill, 1997), pp. 681–85; and The History of the Saffarids of Sistan and the Maliks of Nimruz (247/861 to 949/1542–3) (Costa Mesa: Mazda Publishers in association with Bibliotheca Persica, 1994), pp. 30–38.

(29.) Sakastan was the “Land of the Sakas,” referring to the Scythians, an Indo-European people who lived in the lands of what is now Afghanistan and northwestern India. One of the earlier designations of the region had been the Avestan “Land of the Haetumant”: i.e., Land of the Helmand River, appearing in the early Greek geographical sources as Erymandus. Bosworth (1997; above, note 28), p. 681.

(30.) C. E. Bosworth, Sīstān under the Arabs, from the Islamic Conquest to the Rise of the Ṣaffārids (30–250/651–864) (Rome: IsMEO, 1968), pp. 5–6.

(32.) C. E. Bosworth, “‘Ubaidallah ibn Abi Bakra and the ‘Army of Destruction’ in Zabulistan (79/698),” Der Islam 1 (1973), pp. 268–83.

(33.) The first date is given by al-Tabari (1879–1901 [above, note 8], vol. 2, part 3, pp. 1490–91); the second appears in Balkh’s local history: al-Wa‘iz (1350/1971; above, note 17), p. 35.

(34.) Hugh Kennedy, “The Barmakid Revolution in Islamic Government,” in Charles Melville (ed.), History and Literature in Iran: Persian and Islamic Studies in Honour of P. W. Avery, Pembroke Persian Papers 1 (London: University of Cambridge Centre of Middle East Studies, 1990); C. E. Bosworth, “Abū Ḥafṣ̣ ʿUmar al-Kirmānī and the Rise of the Barmakids,” Bulletin of the School of Oriental and African Studies 57.2 (1994), pp. 268–82; Kevin van Bladel, “The Bactrian Background of the Barmakids,” in Anna Akasoy, Charles Burnett, and Ronit Yoeli-Tlalim (eds.), Islam and Tibet. Interactions along the Musk Routes (Aldershot: Ashgate, 2011).

(35.) François de Blois, “On the Sources of the Barlaam Romance; or, How the Buddha Became a Christian Saint,” in Desmond Durkin-Meisterernst, Christiane Reck, and Dieter Weber (eds.), Literarische Stoffe und ihre Gestaltung in mitteliranischer Zeit (Wiesbaden: Dr. Ludwig Reichert, 2009).

(36.) Khan (2006; above, note 7), p. 96. The group of Arabic documents contains tax receipts issued by caliphal governors and financial agents; land survey reports; contracts of slave manumission; and dowry attestations. The Bactrian-language contracts set peace agreements between feuding parties, the purchase of land or goods, slave manumission, gifts, leases, declarations of trust (or impost?), loan receipts, and marriage.

(39.) Al-Tabari (1879–1901; above, note 8), vol. 2, p. 1575; Wilferd Madelung, “The Early Murji‘a in Khurasan and Transoxiana and the Spread of Hanafism,” Der Islam 59 (1982), pp. 33–35.

(40.) The Kharijites were the earliest religious sects of Islam, who formulated questions relative to the theory of the caliphate and to justification by faith or by works, and carried out continual insurrections until their virtual extinction in Iraq by the ‘Abbasids. G. Levi Della Vida, “Khāridjites,” EI2, vol. IV (Leiden: Brill, 1997), pp. 1074–77.

(41.) Daniel (1979; above, note 6), pp. 134, 137; Wolfgang Madelung,“Ustadhsis,” EI2, vol. 10 (Leiden: Brill, 2000), pp. 926–27; Elton Daniel, “Bihafarid b. Farwardin,” EI3 Online ; Maria Szuppe, “Herat, iii: History, Medieval Period,” EIr, vol. 12 (London: Routledge and Kegan Paul, 2004), pp. 206–11.

(43.) D. G. Tor, Violent Order: Religious Warfare, Chivalry, and the ‘Ayyār Phenomenon in the Medieval Islamic World (Würzburg: Ergon in Kommission, 2007); Bosworth (1997; above, note 28), p. 683. A local ‘ayyar leader in Sistan, Ya‘qub ibn Layth (d. 265/879), rose to become the founder of the Saffarid dynasty, which ruled over a vast empire for 150 years and included Sistan, Khurasan, Kerman, and Fars (in Iran) and Makran (in India).

(44.) The amount of 44.8 million dirhams was collected by the Tahirid governor Abu’l-‘Abbas ‘Abd Allah ibn Tahir for the ‘Abbasid treasury as kharaj in Khurasan and other provinces under his authority in 211–12/826–27. Rob and Samangan accounted for 12,600 dirhams, which indicates that they were relatively small in size. Balkh, on the other hand (together with Khuttalan and Sa’d Khurra and its mountains) accounted for 193,300 dirhams of kharaj. Abu al-Qasim ‘Ubayd Allah Ibn Khurradadhbih, Kitab al-Masalik al-Mamalik, ed. M. J. de Goeje, part 6, Bibliotheca Geographicorum Arabicorum (Leiden: Brill, 1889), pp. 24–28, 34–39; Qudama b. Ja‘far, Kharaj (1889; above, note 8), p. 190.

(45.) Other contributing factors were financial mismanagement of the ‘Abbasid treasury and the loss of control over the mamluk (slave-soldier) army. For detailed discussions, see Hugh Kennedy, “The Decline and Fall of the First Muslim Empire,” Der Islam 81 (2004), pp. 3–30; and David Waines, “The Third Century Internal Crisis of the Abbasids,” Journal of the Economic and Social History of the Orient 20 (1977), pp. 282–306.

(46.) Al-Wa‘iz (1350/1971; above, note 17), pp. 47–48; Le Strange (1930 [1905]; above, note 11), pp. 334–446; Ronit Yoeli-Tlalim, “Islam and Tibet: Cultural Interactions; An Introduction,” in Akasoy, Burnett, and Yoeli-Tlalim (2011; above, note 34).

(47.) For references to the wealth of Balkh’s ‘ulama, see al-Wa‘iz (1350/1971; above, note 17), pp. 64, 73 n. 3, 77, 215, 297, 347.

(48.) See, for example, Jonathan P. Berkey, “Women and Islamic Education in the Mamluk Period,” in Nikki R. Keddie and Beth Baron (eds.), Women in Middle Eastern History: Shifting Boundaries in Sex and Gender (New Haven: Yale University Press, 1991).

(49.) Two silsilas are listed in Arezou Azad, Sacred Landscape in Medieval Afghanistan: Revisiting the Fada’il-i Balkh (Oxford: Oxford University Press, 2013), annex II.

(50.) Professionalization through legal school (madhahib) and Islamic training institutions (madrasas) begins only in the tenth and the eleventh century, respectively. For a recent discussion, see Eyyup Said Kaya, “Continuity and Change in Islamic Law: the Concept of madhhab and the Dimensions of Legal Disagreement in Hanafi Scholarship of the Tenth Century,” in Peri Bearman, Rudolph Peters, and Frank Vogel (eds.), The Islamic School of Law (Cambridge, Mass.: Harvard University Press, 2005). Also George Makdisi, The Rise of Colleges: Institutions of Learning in Islam and the West (Edinburgh: Edinburgh University Press, 1981). The suhba stage usually followed four years of study during which the student was apprenticed as mu’id or repetitor of his master or as someone who made himself useful to younger students (mufid). After this learning period the student could obtain a license to teach law and to issue legal opinions (ijazat al-tadris wa-l-fatwa). See J. Pedersen (rev. George Makdisi), “Madrasa, 1: The Institution in the Arabic, Persian and Turkish Lands,” EI2, vol. 5 (Leoden: Brill, 1986), pp. 1123–24.

(51.) Adam Mez, The Renaissance of Islam (Patna: Jubilee Printing and Publishing House, 1937 [1922]); and Azad (2013; above, note 49), chapter 3.

(p.269) (52.) Al-Wa‘iz (1350/1971; above, note 17), pp. 209–10. On the use of the mazalim system at this time elsewhere in the caliphate, see Mathieu Tillier, “Qadis and the Political Use of the Mazalim Jurisdiction under the ‘Abbasids,” in Christian Lange and Maribel Fierro (eds.), Public Violence in Islamic Societies: Power, Discipline and the Construction of the Public Sphere, 7th–9th Centuries CE (Edinburgh: Edinburgh University Press, 2009).

(53.) Richard Bulliet, Conversion to Islam in the Medieval Period: An Essay in Quantitative History (Cambridge, Mass.: Harvard University Press, 1979), pp. 16–32 and Graph 3.

(54.) E.B.D. signifies the “era of the Bactrian documents.” The dating of this era is still debated with regard to a variance of about ten years. I follow François de Blois’s argument for the start in 223 A.D., which is also applied by Sims-Williams in the revised translation of the Bactrian documents: François de Blois, “Du nouveau sur la chronologie bactrienne post-hellénistique: L’ère de 223–4 ap. J.-C.,” Comptes Rendus de l’Académie des Inscriptions et Belles-Lettres, 2006 (2008), pp. 991–97.

(55.) Abu al-Rayhan al-Biruni, Alberuni’s India: An English Edition, with Notes by E. C. Sachau (London: Trübner & Co., 1910 [1888]), p. 108. Elton Daniel relegates references to fraternal polyandry to heresiographical stereotypes intended to besmudge the syncretic ghulat leaders, such as al-Muqanna’: Daniel (1979; above, note 6), p. 145.

(56.) The thesis forwarded by Kazuo Enoki, that only Hephthalites practiced fraternal polyandry in Tukharistan, has been corrected by Étienne de la Vaissière (the latter basing his commentary on the evidence from the Bactrian documents, which came to light only long after Enoki’s article). See Kazuo Enoki, “On the Nationality of the Ephtalites,” Memoirs of the Research Department of the Toyo Bunko 18 (1959), pp. 1–58; and Étienne de la Vaissière, “Is There a ‘Nationality’ of the Hephthalites?” Bulletin ofthe Asia Institute 17 (2003), pp. 119–32, especially pp. 119–21.

(57.) Tibet provides a useful comparandum. Melvyn Goldstein argues that Tibetan fraternal polyandry is the “lesser evil;” a compromise strategy, stimulated by the need to pool human resources to meet excessive activity requirements of living in a harsh environment at high altitudes, in a semiarid land with limited rainfall, and to discharge the obligation of high tax burdens. Goldstein argues that through fraternal polyandry landholdings maintain their economies of scale in relation to labor costs, and brothers share the property within a “stem family.” Melvyn Goldstein, “Stratification, Polyandry, and Family Structure in Central Tibet,” Southwestern Journal of Anthropology 27.1 (1971), pp. 64–74; idem, “Pahari and Tibetan Polyandry Revisited,” Ethnology 17.3 (1978), pp. 327–32.

(59.) Peter Webb, “The Hajj before Muhammad: Journeys to Mecca in Muslim Narratives of Pre-Islamic History,” in Venetia Porter and Liana Saif (eds.), The Hajj: Collected Essays, (London: British Museum, 2013).

(60.) This is a reference to early forms of Buddhism that must have continued in Balkh. Xuanzang (1906; above, note 23), vol. 1, p. 46. Later phases of Buddhism, notably the Mahayana and Vajrayana, do not preclude the use of early forms of Buddhism. For a more detailed discussion of Balkh’s sacred landscape, refer to my Sacred Landscape in Medieval Afghanistan (2013; above, note 49).

(61.) Anas ibn Malik, Abu Hamza (d. ca. 91–93/709–11), a servant of the Prophet Muhammad from his childhood and a prolific traditionist: A. J. Wensinck and J. Robson, “Anas b. Mālik,” EI2, vol. 1 (Leiden: Brill, 1960), pp. 482–83.

(p.270) (62.) The text uses four Quranic terms to refer to the angels’ prayers corresponding to the four gates of the city: istighfar, takbir, tahmid, and tahlil.

(63.) F. W. Hasluck, Christianity and Islam under the Sultans (Oxford: Clarendon Press, 1929), pp. 75–77.