Islamization and Acculturation at the Time of ʻUmar II (717-720)

I

The goal of this paper is to examine two unrelated processes that took place in early Islam: conversion to Islam and, concomitantly, Islamization and acculturation. I perceive conversion as the beginning of a long process of religious change that led to Islamization. Recently, Corisande Fenwick, has referred to Islamization as a much abused term which is taken as synonymous with conversion.¹See, Early Islamic North Africa, p. 129. I owe this reference to the kindness of one of the anonymous readers on behalf of Al-Qanṭara. I would argue that Islamization is an elusive term since we mostly see its end-result, while its stages remain obscure. How thorough Islamization was is another difficult question to answer since non-Muslim communities were a constant feature of the socio-religious landscape of the medieval Middle East, and submission to Muslim rule did not render them insignificant.

The methodology for the study of Islamization is still ill defined, and most studies are focused on the presumed indicators of the process such as name patterns. In an attempt to integrate sources and studies into a meaningful discussion, the first part of the paper consists of extensive discussion of studies on Islamization and the insights derived from this discussion are constantly referred to in parts two and three. Part two is a case study of Ifrīqiya (Tunisia and north-east Algeria) and ʻUmar II’s policies in the province, while part three examines the broader scope of the caliph’s policies on both conversion and acculturation. My discussion of acculturation relies on sources and is also informed by recent publications on the subject.

Before any discussion of methodology is attempted, the rites associated with conversion to Islam require an explanation. For example, the question of whether a conversion to Islam driven by opportunistic motives such as the wish to escape financial hardship or gain personal benefits can be considered as genuine was much debated in diverse medieval Arabic writings. The position attributed to the Prophet was that the pronouncement of the declaration of faith (There is no god except for God) was enough to accept the conversion of a pagan Arab. Christians and Jews were, however, also obliged to admit Muḥammad’s status as a Prophet. They had to pronounce the double declaration of faith: “There is no god except for God and Muḥammad is God’s Messenger”.²There is ample literature on the sincerity of conversion and its rites, and the most recent publications include Friedmann, “Conversion, Apostasy and Excommunication”, pp. 107-116; Chrysostomides, “There is no God but God’’, pp. 118-121; Cooperson, “An Early Conversion Story”, pp. 390-391. The eighth century Syriac Chronicle of Zuqnīn (pp. 327-328) describes a rite of conversion that took place in the region of Edessa (Urfa in Turkey), which involved both verbal renunciation of Christianity and acceptance of Islam that went beyond the formula “There is no god except for God and Muḥammad is God’s Messenger”. For the Chronicle of Zuqnīn, see Hoyland, Seeing Islam as Others Saw It, pp. 337-339. In social terms conversion obliged association with Arabs which brought converts to seek affiliation with Arab patrons (powerful individuals, families or tribes) in order to become their clients (mawlā pl. mawālī), while adopting their patronymics. The relation between clients and their Arab patrons was especially strong in the case of freedmen (i.e. slaves or captives who had converted to Islam and were set free by their masters). ³For the role of patronage in the conversion process, see Bulliet, “Conversion-Based Patronage”, pp. 246-263; Nawas, “A Client’s Client”, pp. 143-158. For the legal and social mechanisms that were instrumental for the inclusion of slave converts, especially female slaves, into the Jewish society of the Geniza period, see Perry, “Conversion as an Aspect of Master-Slave Relationships”, pp. 135-160.The social repercussions of the conversion process - the position of the mawālī in Muslim society during the seventh-eighth centuries- are frequently mentioned in the sources and extensively discussed by scholars.⁴In the Ifrīqiya context, the shifting policies towards the mawālī are illustrated by the events which took place in 102/720-721, during the governorship of Yazīd ibn Abī Muslim, and the treatment of the mawālī of his predecessor Musā ibn Nuṣayr, who served as his bodyguard (ḥaras). Yazīd, himself the mawlā of al-Ḥajjāj, the governor of Iraq, tattooed their arms and re-enslaved them. In response, they assassinated him. See al-Yaʻqūbī, vol. II, p. 376; al-Balādhurī, Futūḥ, p. 325. For the most recent scholar discussion of the mawālī in early Islam, see Urban, Conquered Populations, pp. 20-25, and ch. 3.

The inner logic of the conversion process, especially when well-known people were involved, obliged the act to be public. One of the most famous converts to Islam in tenth century Egypt was the Jew Yaʻqūb ibn Killis, who converted during the rule of Kāfūr (966-968) and became his vizier. The conversion took place at Kāfūr’s residence and, subsequently, a great crowd accompanied Ibn Killis to his house. The procession was witnessed by Sībawayh, a noted contemporary “holy fool”, who was highly critical of the event.⁵ Ibn Zūlāq, p. 44. For Sῑbawayh, see Tillier, “Un ‘Alceste musulman’”, pp. 117-139; Lev, “Aspects”, pp. 2-5. For a broader discussion of the conversion act, see Simonsohn, “Conversion, Apostasy”, pp. 212-213.

As important as the rites associated with conversion to Islam were, the meaning of adopting Islam must also be addressed. The attempt to elucidate this issue at the time of the Prophet and the caliphs of Medina (632-661) and Damascus (661-750) is, for example, at the heart of Harry Munt’s, Jack Tannous’ and Uriel Simonsohn’s discussions of the transformation from Arab paganism (jāhiliyya) to Islam. Islamization during those periods involved individuals and tribal groups and was motivated by convictions and a myriad of personal and political reasons. Munt has emphasized the political aspects of the Islamization process of Arabia during the Prophet’s lifetime, whose driving force was the desire to preserve property. As has been shown by Simonsohn, at the time of the Prophet, the conversion of certain individuals proved to be insincere, while Tannous has pointed out that the difference between genuine acceptance of Islam on a personal level versus instrumental mass conversion of various social groups had long term socio-religious consequences well into the Umayyad period.⁶ Munt, “What Did Conversion to Islam Mean?”, pp. 89-93; Tannous, The Making of the Medieval Middle East, pp. 263-278; Simonsohn, “Conversion, Apostasy”, pp. 204-205.

On the individual and collective levels, the knowledge of the precepts of the new religion remained rudimentary for many decades and was characterized by syncretism between paganism and Islam. The mechanism of propagating Islam is illustrated by a unique report recounted by the Damascus historian Abū Zurʻa (d. 894), which refers to majālis al-dhikr. The early meaning of the term alluded to sessions in which the question of what is or is not permissible was discussed. The range of issues that fell under these categories was wide and involved rituals (prayer and pilgrimage), personal status (marriage and divorce), and commercial transactions (sale and purchase).⁷See, vol. I, p. 420.

Although the rites associated with conversion to Islam were simple, the question of the sincerity of conversion continued to dominate the medieval discourse on conversion and is highlighted by the term muslimānī, which does not appear in medieval Arabic dictionaries. The standard approach has been to adopt Reinhart P. A. Dozy’s translation as “nouveau musulman”, which conveys well the basic meaning of the term.⁸ Supplément, vol. I, p. 677. This translation has been adopted by Tillier and Vanthieghem, “La rançon du serment”, pp. 55-56 (line 2, text and trans.). Elsewhere I have argued that the term refers to converts to Islam who showed no devotion to their new faith. Although such conduct was tolerated, a convert labeled as a muslimānī found it difficult to assimilate into Muslim society since the term had clear pejorative connotations.⁹ Lev, “Persecutions and Conversion”, pp. 88-89. The pejorative connotation of the term is also noted by Tillier and Vanthieghem, “La rançon du serment”, p. 63. For the Fatimid period, the most revealing text is that of al-Musabbiḥī (977-1029) who, in the obituaries of 1025, reports on the killing of a person referred to as “Abū Saʻd al-muslimānī, the secretary employed in the Office of the Army in Ramla”. In another report, al-Musabbiḥī refers to the death of a very wealthy person simply described as “al-ʻAdanī al-muslimānī,” meaning a person from Aden who lived in Fusṭāṭ. His address (the street and house) and family connections are fully reported. See Akhbār Miṣr, pp. 94, 109. The unmistakable impression is that a person labeled as a muslimānī could live and function in society but the term had an abusive meaning. See Akhbār Miṣr, pp. 37-38.

The discussion of the methodological issues involved in the conversion process was brought to the fore of the scholarly debate in 1950 by Daniel C. Dennett, who provided a summary of Julius Wellhausen’s findings (1902), and of subsequent studies. Dennett proposed adopting a regional approach to questions such as the status of land and taxation and pointed out the need for precision when terms such as kharāj (land tax) and poll-tax (jizya) are examined and translated. Both Wellhausen and Dennett also discussed the policies of ʻUmar II concerning taxation and conversion to Islam.¹⁰See, Conversion and the Poll Tax in Early Islam, pp. 3-11 (Dennett’ summary of earlier scholarship), pp. 11-13, 83-88 (Dennett’s views on the subject and the policies of ʻUmar II); Wellhausen, The Arab Kingdom and Its Fall, pp. 296-311.

One must bear in mind that since the publication of Dennett’s book a significant shift about how Islamization is studied has taken place. In 1979, Richard W. Bulliet suggested a completely different methodological approach to Islamization, claiming that: “…changes in name patterns accurately reflect the general course of religious conversion…” He went on to explain the quite complex structure of medieval Arabic names and suggested that the first appearance of a name drawn from the Arabic onomasticon serves as an indicator of conversion to Islam. Bulliet’s data are culled from the extensive medieval Arabic biographical literature, which mostly deals with people of the learned class and is presented in the form of graphs depicting the process of Islamization in various regions of the Middle East.¹¹See Conversion to Islam in the Medieval Period, pp. 4, 10, 19. The name categories that are considered as typically Arabic/Islamic are explained in ch. 6. For the genesis of the book, see Bulliet, “The Conversion Curve Revisited”, pp. 69-72. The notion that religious change is reflected by changes in naming patterns is corroborated by Arabic sources. The transition from Arab paganism to Islam was marked by the change of male and female first names, while conversion from Christianity to Islam involved the adoption of Arabic names with Islamic connotations. See Abū Zurʻa, vol. I, p. 453, vol. II, pp. 764-766; Al-Quḍāʻī, ʻUyūn al-Ma‘ārif, p. 129. The change of names and the registration of converts to Islam is also alluded to in the Chronicle of Zuqnīn. The reference to the registration must be understood as referring to the dīwān system in which mawālī converts were registered along with their Arabic patrons to receive payments from the state. See pp. 329-330. For the kunya Abū Muslim (agnomen, consisting of Abū followed by the name of a son) as a marker of converts to Islam, see Ibn Yūnus, Ta’rīkh, vol. II, pp. 236-237. Without ever alluding to Dennett, Bulliet makes clear that Islamization must be approached regionally.

In the context of Islamization in al-Andalus, Mayte Penelas has suggested a modification to Bulliet’s methodology by asserting that Islamization is represented by the name ʻAbd Allāh (servant of God) being assigned to the convert’s father and the appearance of the nisba (a part of the Arabic name pattern that signifies affiliation, ending in ī) al-Islāmī (meaning belonging to Islam). David J. Wasserstein has proved both assertions wrong, indicating how problematic the onomastic approach is.¹² Penelas, “Some Remarks on Conversion to Islam in al-Andalus”, pp. 193-200; Wasserstein, “Where Have all the Converts Gone?”, pp. 325-342. For a completely different approach to the Islamization of al-Andalus, which focuses on regional differences, see Aillet, “Islamisation et évolution du peuplement chrétien”, pp. 151-195. Alwyn Harrison, however, has argued that Bulliet’s approach is misunderstood since Bulliet’s graphs represent the process of conversion as derived from the biographical data and not the overall proportions of Muslim-non-Muslim populations. He has also shown that in the case of Egypt and Palestine the literary sources referring to Islamization and Bulliet’s graphs completely diverge.¹³See “Behind the Curve”, pp. 35-51, esp. pp. 38, 44-47.

It must be pointed out that the onomastic approach is a general methodological tool employed for the tracing of cultural and religious changes and is also extensively applied outside the field of Islamic studies. The most relevant case is perhaps the application of this methodology for the study of Egypt’s conversion from paganism to Christianity. Explicitly quoting Bulliet, Roger S. Bagnall has used the onomastic approach and postulated that there was a rapid Christianization of Egypt between 310 and 360 and that 90% Christianization was achieved by the end of the century.¹⁴See “Religious Conversion and Onomastic Change in Early Byzantine Egypt”, pp. 105-123. In 2013, Bagnall’s findings were re-examined and modified by Mark Depauw and Willy Clarysee and, according to them, overwhelming Christianization was achieved in about the mid-fifth century.¹⁵See “How Christian was Fourth Century Egypt?”, pp. 407-435. In a 2014 publication, David Frankfurter asked what, in the fourth century, constituted “being Christian” and what the demarcation lines between paganism and Christianity could have been.¹⁶See “Onomastic Statistics and the Christianization of Egypt”, pp. 284-289. In their reply Depauw and Clarysee dealt with the issues raised by Frankfurter and reaffirmed the validity of the onomastic approach.¹⁷See “Christian Onomastics: A Response to Frankfurter”, pp. 327-329. The wide application of the onomastic approach has also been demonstrated in Bulliet’s recent publication but the question about how his findings should be interpreted in the Islamic context remains.¹⁸See “The Conversion Curve Revisited”, pp. 69-79.

In a broader comparative framework, the discussion about what constitutes religious identity and the nature of denominational demarcation lines is also relevant in the Islamic context. Lev Weitz, for example has suggested to shift the discussion from conversion to Islam and the demographic ratio between Muslim and non-Muslim populations to “…other meaningful markers of social change…”, which reflected “…the rising influence of Islamic institutions and norms of social organization on non-Muslims, with the emphasis that such influences need not have led to conversion”.¹⁹See “Islamic Law on the Provincial Margins”, pp. 5-6. For the complex relations between acculturation and Islamization, see Picard, “Islamisation et arabisation”, pp. 55-56, who refers to literature about the Muslim West. Weitz’s remarks highlight the complex relations between acculturation and Islamization and the fact that these could have been parallel processes, while in other cases acculturation paved the way to Islamization.

As innovative as Bulliet’s attempt to write quantitative history of the conversion process is, the more traditional approach focusing on conversion stories maintains its allure. Biographical literature came to typify medieval Arabic historical writings, which offer many stories about men and women who converted to Islam during the Prophet’s lifetime. In some cases the political aspects of the conversion, or the so-called conversion, come to the fore, while other cases reveal a complex interplay between the old kinship relations and the newly acquired religious identities and freshly forged loyalties.²⁰ Stern, “The First Women Converts”, pp. 294, 296-297; Calasso, “Récits de conversion”, pp. 19-47; Hershkovits, “Kinship”, pp. 303-304, 310.

Dennett’s regional approach has its merits and in Egypt, for example, the process of Islamization stretched over several centuries. Its chronological turning points are discussed in a number of studies, which focus on the regional differences between the Fayyūm and the Delta.²¹For the Delta, see Lev, “Coptic Rebellions”, pp. 303-345; “The Fatimid Caliphs”, pp. 390-410. For the Fayyūm, see Rapoport, Rural Economy, Ch. 9; Gaubet and Mouton, Hommes et villages, pp. 237-246, 259-264. The importance of the regional context is also illuminated by the case of Palestine in which, in contrast to Egypt’s homogeneous population, each one of the socio-religious groups: the Samaritans of Samaria, the Melkites (the largest Christian community in Palestine) and the Jews, must be studied separately. In the context of the Islamization of the Melkite community Milka Levy-Rubin has drawn attention to the difference between Islamization and Arabization or, to put it differently, acculturation. Arabization/acculturation involved the acceptance of Arabic culture while adhering to separate religious identity and, in the context of the Melkites of Palestine, Arabization meant the adoption of the Arabic language, Arabic name forms and the Hijrī dating. Levy-Rubin argues that by 980 full Arabization was achieved, when Arabic replaced Aramaic in the reading of the Scriptures.²²See, “Arabization versus Islamization”, pp. 152-153; “New Evidence”, pp. 257-258; “Islamization of Space and People”, p. 370.

In the Egyptian context, however, Bulliet’s and Levy-Rubin’s remarks about the adoption of Arabic name forms are especially intriguing since the extensive epigraphic and documentary record pertaining to Egypt reflects wider social groups than those depicted by the biographical dictionaries of the learned class (ʻulamāʼ) studied by Bulliet. For example, the stelae of the Muslim cemetery of Aswān in Upper Egypt (eighth-tenth centuries) bear the names of people from diverse walks of life and, to some extent, Bulliet’s methodology works. Stela 59, for example, dated to 857, bears the name of Muḥammad ibn Yūhannis al-Khayyāt. Christian Décobert, who has studied those inscriptions, offers the following commentary on the stela:

Other inscriptions depict more complex naming patterns. Stela 216, for instance, dated to 886, bears the name of Muḥammad ibn ʻAbd Allāh ibn Fawtis. Décobert interprets the name ʻAbd Allāh as signifying mere adhesion to the new religion and assumes that the conversion process progressed from adhesion in the first generation to full Islamization in the second generation.²³See, “Sur lʻarabisation et lʼislamisation de lʼÉgypte”, pp. 278-279. It seems that Décobert’s interpretation of the evidence relies on two hidden assumptions: the first being that the name ʻAbd Allāh (servant of God) is a monotheistic name without clear Islamic connotations. The second assumption seems to be derived from Arthur D. Nock’s distinction between two forms of religious change: one that involved a complete reorientation of the soul of an individual (conversion) while the other form was marked by mere adhesion to the new religion.²⁴See, Conversion, pp. 1-16.

Certain names engraved on the tombstones of the Aswān cemetery, however, offer no clue to the religious identity of the deceased. Décobert provides two examples of such enigmatic names: Bilāl ibn Andriyā and Isḥāq ibn Ibrāhīm. While Bilāl is a name that could have been used by both Christians and Muslims, Isḥāq and Ibrāhīm are names derived from the Semitic monotheistic tradition.²⁵See, “Sur lʻarabisation et lʼislamisation de lʼÉgypte “, pp. 278-279. Décobert’s discussion indicates that name forms are difficult to interpret and the studies by Marie Legendre and Elinoar Bareket, which rely on different source material, elaborate and expand upon these conclusions. Papyri, like tombstones, reflect people from all walks of life and a great variety of name forms. Legendre offers a detailed discussion of Coptic and Arabic names depicted in papyri from the Fayyūm, which roughly can be dated to the second half of the eleventh century. 26% of the names used in those documents are unmistakably Muslim while 38% bear no clear indication of religious identity and could have been used by either a Muslim or by a “chrétien arabisé”. Biblical names offer no clue to the religious identity of the person and Legendre draws attention to the practice of giving two names to be used in different socio-religious settings.²⁶See “Perméabilité l inguistique”, pp. 326-330, esp. pp. 329-330 (for dating of those documents), pp. 348-351, esp. pp. 350-351 (for analyzing the meanings of the name forms).

In the context of the Jewish society of the Geniza period, the practice of giving two names to men is noted by Elinoar Bareket. The Hebrew name served a person in community life while the Arabic name facilitated blending into street life and maintaining social contacts outside the community. 48% of names depicted by the Geniza documents studied by Bareket are Biblical-Koranic names while 36% are Arabic names unrelated to the Koran that signify abstract positive qualities. One of the most significant observations made by Bareket concerns feminine name patterns of which 92% are Arabic names composed of the word sitt (mistress) following by attributes describing women as having authority over the house, family and other positive qualities and features. She asserts that female name patterns reflect the fact that women played no role in synagogue life and, therefore, had no need for a Hebrew name.²⁷See, “Note on Jewish Naming Pattern”, pp. 83-86. For broader discussion of women’s names, see Goitein, A Mediterranean Society, vol. III, pp. 314-319.

II

Although occasionally insightful, the limitations of the onomastic approach are serious and for the study of Islamization one must go back to the literary sources. The case study of Ifrīqiya involves several intertwined topics, which include the Byzantine background of the region, the impact of the Muslim conquest and the policies of the Umayyad caliphs, especially those of ʻUmar II. The Byzantine background of North Africa is well known and the reader can be referred, for example, to a recent collection of articles that deal with a wide range of pertinent issues.²⁸See North Africa under Byzantium and Islam. The same applies to the much discussed topic of the Muslim conquest of the region and one must bear in mind Walter E. Kaegi’s observation that: “A process of Islamization, which is not identical or synonymous with North Africa’s conquest, would last many centuries”.²⁹See “The Islamic Conquest and the Defense of Byzantine Africa”, p. 67. The significance of this observation is that it goes against a main theme found in Arabic sources, which postulates a direct link between conquest and Islamization.

The Islamization of Ifrīqiya has been examined by Allaoua Amara who, authoritatively has combined a discussion of both sources and studies. The main thrust of Amara’s arguments is that the sources depict the Islamization of Ifrīqiya as having taken place in two stages: a limited Islamization which occurred between 647 and 701 and a massive Islamization of “…les confédérations tribales pastorales du Maghreb central et du Sud de lʼIfrīqiya”, which took place between 701 and 718. The second phase is directly linked to the fiscal and religious policies of the Umayyad caliphs, especially those of ʻUmar II. Amara also draws attention to the late and retrospective nature of the sources, which have been “… élaborés dans des contextes idéologique et politique différents”.³⁰See “Lʼislamisation du Maghreb Central”, p. 112. For a more detailed discussion of the Islamization of the Kutāma of the southern Ifrīqiya, see Amara, “Peuplement et arabisation au Maghreb médiéval”, pp. 270-285.

The Islamic scene of North Africa was complex and variegated and Islamization could have meant embracing the Sunnī Islam of the ruling dynastiess - the Umayyad and Abbasid and, in the local context, the Aghlabids - as well as Khawārijī Islam. Equally complex was the ethnic-religious scene and any discussion of the Islamization of the Berbers invokes Michael Brett’s question: “Why single out the question of their (the Berbers) conversion to Islam as distinct from, say, the conversion of all North Africans, Latins, and Greeks? And, as a corollary to this, what were they converted from?”.³¹See “Conversion of the Berbers”, p. 189; “The Islamization of Egypt and North Africa”, pp. 15-16. As significant as these questions are the sources and literature exclusively focus on the Berbers.

Amara’s reference to limited Islamization occurring between 647 and 701 possibly alludes to the Kāhina rebellion, which took place between the late 690s and 703 and still poses a mystery. The rebellion involved the Berbers of the Aurès Mountains and the sources claim that when Kāhina, the female leader of the uprising, realized that the rebellion had failed she exhorted her sons to adopt Islam and thousands converted following the collapse of the rebellion. This description of the events is, for example, at the fore of al-Raqīq al-Qayrawānī’s account (d. 1033), which remains uncorroborated.³²See pp. 49-50, 51, 52, 61. These accounts refer to both the immediate Islamization in the wake of the rebellion and to later events that led to Islamization. For the most recent publication on the Kāhina rebellion, see Hendrickx, “Al-Kahina”, pp. 47-61. For al-Raqīq al-Qayrawānī’s writings embedded in late medieval historiography, see Amara, “Lʼislamisation du Maghreb central”, p. 109. For al-Raqīq al-Qayrawānī’s diplomatic mission to al-Ḥākim’s court in 998, see al-Maqrīzī, Kitāb al-Muqaffā, vol. I, pp. 256-257. Al-Maqrīzī also provides a list of his works and praises him as a historian. The account fits the broader context of the medieval discourse about the process of Islamization and, for what it is worth, the theme of “subjection and persecutions leading to Islamization” is a common explanatory trope used in medieval Coptic historiography.³³ History of the Patriarchs of the Egyptian Church, vol. II, part 1, pp. 4-6 (Arabic), pp. 6-12 (trans.); Lev, “Coptic Rebellions”, pp. 326-327. For more complex Coptic perceptions of Muslim rule, see Mikhail, From Byzantine to Islamic Egypt, pp. 61-78; Palombo, “The View from the Monasteries”, pp. 327-332. This broad conceptualization of Kāhina’s rebellion, however, says nothing about the validity of the claim that Islamization followed the failure of the rebellion.

Amara’s discussion of the second stage of the Islamization of Ifrīqiya is linked to the policies of ʻUmar II and for that we must turn to the writings of Abū Bakr al-Mālikī (d.1055) and his biographical dictionary of the ʻulamāʼ of Kairouan. ʻUmar II sent a delegation of ten tābiʻūn to Ifrīqiya with the explicit mission to teach law and religion to the people of the province. The term tābiʻūn refers to Muslims who had met the Companions of the Prophet and were thus considered as possessors of authoritative knowledge about the Prophet’s life, deeds and utterances. One gets the impression that the primary motive behind the mission was to enhance the Islamic identity of the Arab population (ahl Ifrīqiya) of the province through the personal example set by the members of the delegation and through their teaching of law and religion. Members of the delegation settled in Kairouan, and al-Mālikī is at pains to emphasize the beneficial aspects of their activity and their impact on the local scene.³⁴Some members of the delegation are praised for their personal virtues (al-Mālikī, vol. I, pp. 102, 106, 111) while others were jurists (al-Mālikī, vol. I, pp. 100, 110. 112). When the broader context of the Islamization process is considered, ʻulamāʼ served as proponents of Islam. For the discussion of Balkh and Tukharistan of the seventh-eight centuries, see Azad, “The Beginnings of Islam in Afghanistan”, pp. 50-52.

Reading between the lines of al-Mālikī’s account leaves one with the unmistakable impression that members of the delegation acted as individuals, not as a collective. Ismāʻīl ibn ʻUbayd al-Anṣārī, for example, became involved in the building of a mosque, which later came to be known as the al-Zaytūna mosque, which gained fame through North Africa as a center of Islamic learning. Another member of the delegation, Abū Masʻūd Saʻd ibn Masʻūd, participated in the writing of a pamphlet outlining the principles of faith (a type of writing known as ʻaqīda) at the behest of the governor Ḥanẓala ibn Ṣafwān al-Bakrī who, in 742, faced a Khawārijī rebellion.³⁵A-Mālikī, vol. I, pp. 102-103, 106-107. For the Khawārijī rebellion, see Savage, A Gateway to Hell, a Gateway to Paradise, pp. 43-44. Since the members of the delegation acted as individuals some of them stepped outside their mission, so to speak, and became active proselytizers. Ismāʻīl ibn ʻUbayd Allāh ibn Afram Abū l-Muhājir, a mawlā of Banū l-Makhzūmī, is describe, for example, as a religious person and ascetic who was sent to Ifrīqiya to judge people according to the Koran and the tradition of the Prophet. He fulfilled these expectations and showed great zeal in spreading Islam among the Berbers. Al-Mālikī’s biography of Ismāʻīl relies on quotations from three named sources: the Egyptian biographer Abū Saʻīd ibn Yūnus (894-958), the famous historian al-Ṭabarī (d. 923), and the renowned literato al-Tanūkhī (939-994). The quote attributed to al-Ṭabarī about Ismāʻīl states that “He was a good governor and a good emir” is not to be found in his History of the Prophets and Kings.³⁶Al-Mālikī, vol. I, p. 116. Other ninth-tenth century works, however, throw important light on Ismāʻīl’s appointment as governor. Al-Balādhurī (d. 892), for example, on the authority of al-Wāqidī (d. 832), reports about Ismāʻīl’s appointment as governor and praises his conduct. He states that Berbers converted to Islam during his tenure in office and presents both the caliph and the governor as the driving force behind the process. The caliph’s contribution is vaguely referred to as a letter sent to the Berbers that was read by the governor. The report ends with the sweeping statement that: “Islam prevailed in the Maghreb”.³⁷ Al-Balādhurī, Futūḥ, pp. 323-324.

A further survey of the sources reveals a wide range of opinions about ʻUmar II’s delegation to Ifrīqiya. Omitting any reference to the caliph’s role in the endeavor to spread Islam in North Africa, Khalīfa ibn Khayyāṭ (d. 855) praises Ismāʻīl’s conduct and states that the Berbers converted to Islam during his governorship.³⁸ Khalīfa b. Khayyāṭ, Ta’rīkh, vol. I, p. 330. For other governors of Ifrīqiya immediately prior to Ismāʻīl, see p. 324. North African historiography is in line with Khalīfa ibn Khayyāṭ’s account. Al-Raqīq al-Qayrawānī, for example, praises Ismāʻīl for the way he governed Ifrīqiya (100-101/718-719) and states that he was zealous in converting the Berbers to Islam.³⁹ Al-Raqīq al-Qayrawānī, Ta’rīkh, p. 61. Ibn ʻIdhārī (late thirteenth century-early fourteenth century) depicts Ismāʻīl as being an ardent proselytizer and educator who instructed the ahl Ifrīqiya about what was permissible and what was not. He, and the other members of ʻUmar II’s delegation, enlightened ahl Ifrīqiya with the injunction that wine was a forbidden beverage.⁴⁰See vol. I, p. 48.

In his biographical dictionary, Ibn Yūnus’s emphasis is, however, different. He states that:

Ibn al-Athīr (1160-1233) writes that Ismāʻīl and another person named al-Samḥ ibn Mālik al-Khawlānī were appointed as governors in 100/718-719, and points out that ʻUmar II was much impressed by Ismāʻīl’s conduct.⁴² Ibn Athīr, Kāmil, vol. IV, pp. 231, 246, which refers to their dismissal following ʻUmar II’s death.

As extensive as al-Mālikī’s biography of Ismāʻīl is, he fails to shed light on Ismāʻīl’s life before his arrival to Ifrīqiya. For that we must turn to Syrian historians such as Ibn ʻAsākir (1105-1176) and Ibn al-ʻAdīm (1192-1262). Ibn ʻAsākir’s account of Ismāʻīl’s life also explains his inclusion in the delegation sent to North Africa. Ibn ʻAsākir depicts Ismāʻīl (680-749) as a native of Damascus closely associated with the Umayyad caliphs. He served as a tutor of the sons of the caliphs ʻAbd al-Malik (685-705) and al-Walīd (705-715), and had access to ʻUmar II.⁴³ Ibn ‘Asākir, Ta’rīkh, vol. VIII, pp. 434, 436, 437, 441-442. Ibn Asākir extensively dwells upon Ismāʻīl’s role in the circulation of Prophetic ḥadīth and his involvement in jihād. Without offering any concrete details, Ibn ʻAsākir alludes to his participation in the summer raids on Byzantium and his stay in Beirut as a warrior of the holy war.⁴⁴ Ibn ‘Asākir, Ta’rīkh, vol. VIII, pp. 432, 435. Ibn ʻAsākir portrays Ismāʻīl as living an exemplary life, combining learning and militant piety and as a natural candidate for a delegation, which aims was to enhance Islamic values in a remote province. As informative as Ibn Asākir is about the Syrian phase of Ismāʻīl’s life, he has little to say about the Ifrīqiya phase. He quotes Khalīfa ibn Khayyāṭ’s account about Ismāʻīl’s good governorship and his zeal to proselytize.⁴⁵ Ibn ‘Asākir, Ta’rīkh, vol. VIII, p. 440. Ibn al-ʻAsākir’s account about Ismāʻīl is echoed by Ibn al-ʻAdīm (vol. IV, pp. 1695-1706).

Because of the close relations between Ifrīqiya and Egypt in early Islam, rich information about Ifrīqiya is to be found in the writings of al-Maqrīzī (1364-1442), the famous historian of Egypt and Cairo. Al-Maqrīzī also writes about the Islamization of North Africa but the information is second hand. He, for example, echoes the reports that the Islamization of the Berbers took place following the failure of Kāhina’s rebellion.⁴⁶ Al-Maqrīzī, Kitāb al-Muqaffā, vol. III, p. 282. Al-Maqrīzī’s biography of Ismāʻīl is, however, completely different from Ibn Yūnus’s short report or al-Mālikī’s fuller account. In addition, through Ismāʻīl’s biography, al-Maqrῑzῑ, sheds light on the political-administrative practices of the Umayyad period, especially the wafd (delegation) institution. Tax money sent to Damascus from the provinces was accompanied by a delegation of eminent people who had to swear that the money had been justly collected and constituted a surplus after local expenses such as paying the troops and their families had been met. The Treasury in Damascus was strict about accepting provincial tax money only after such oaths were taken. For example, when a delegation from Ifrīqiya arrived in Damascus during Sulaymān’s reign (715-717), two of its members,a certain Ismāʻīl ibn ʻUbayd [Allāh ?] and al-Samḥ ibn Mālik al-Khawlānī, refused to take the oath. Consequently, the caliph refused to accept the money.

However, two of al-Maqrīzī’s statements about Ismāʻīl’s life cannot be accepted. Al-Maqrīzī writes that Ismāʻīl lived (sakana) in Ifrīqiya. There is no reason to assume that he had lived in Ifrīqiya prior to his appointment as the governor of the province. Therefore, al-Maqrīzī’s vague allusion to Ismāʻīl’s arrival in Damascus with the wafd from Ifrīqiya must also be rejected. Al-Maqrīzī, nonetheless, offers one important detail about Ismāʻīl’s range of responsibilities as the governor of the province, which also included the collection of the kharāj (land-tax) and ṣadaqāt (i.e. the Islamic alms-tax, also known as zakat), levied on both rural and urban Muslim populations. Al-Maqrīzī praises Ismāʻīl for the way he executed his duties and repeats the remark that during his term in office the Berbers converted to Islam. Ismāʻīl’s governorship ended with the death of ʻUmar II.⁴⁷ Al-Maqrīzī, Kitāb al-Muqaffā, vol. II, pp. 125-126.

The question that must be asked is why Ismāʻīl was successful in converting the Berbers to Islam. The simplest answer would be that, as noted by al-Maqrīzī, he put ʻUmar II’s policies into practice and made use of the fiscal responsibilities vested in him to promote conversion. The caliph’s policy toward Islamization was, however, ambivalent and why his governor in Ifrīqiya was successful in converting the Berbers remains a puzzle.⁴⁸In the distant Bukhara, the governor Qutayba ibn Muslim (d. 715-716) pursued a policy of active Islamization by building a mosque and destroying Zoroastrian places of worship. He also handed out money to the poor for attending the Friday prayer. In the wider Iranian context, the case of Bukhara was exceptional. Kennedy, “The Coming of Islam to Bukhara”, pp. 88-90. Perhaps personal circumstances rather than state policies provide the key for understanding Ismāʻīl’s proselytizing zeal since he himself was a convert, a client of Banū l-Makhzūm. Leaving aside the question of Ismāʻīl’s efforts to convert the Berbers to Islam, ʻUmar II’s delegation to Ifrīqiya was not an isolated attempt on the part of the caliph who, for example, had sent Nāfiʻ on a similar mission to Egypt.

Ninth century sources are unclear about Nāfiʻ’s mission to Egypt. Ibn Saʻd (d. 845), for example, writes that he was sent to teach traditions (sunan). Nāfiʻ was a Persian freedman of ʻAbd Allāh, the son of the caliph ʻUmar I (634-644), and is described as his protégé (mawlā). He lived in Medina but met ʻUmar II in Damascus. Because of Nāfiʻ’s encounter with the caliph, Ibn ʻAsākir included Nāfiʻ in his biographical dictionary of the people connected with the history of Damascus. Although Ibn ʻAsākir was a twelfth century author, his narrative of the encounter sheds light on the mood of that formative period of Islam. On the authority of ʻUmar I’s son, Nāfiʻ related the following episode to ‘Umar II. When the Prophet inspected the troops before the Battle of Uḥud, he forbade the fourteen year old ʻAbd Allāh from participating in the battle but when ʻAbd Allāh was fifteen years old the Prophet allowed him to take part in the Battle of the Khandaq. Relying on the tale, which in the terminology of the period was considered as a ḥadīth, ʻUmar II ruled that the age threshold dividing between someone who is boy (ṣaghīr) from someone who is an adult (kabīr) should be fifteen. The caliph conveyed the instruction to his governors and ordered them to consider fourteen year olds as minors who were part of the family group.⁴⁹ Ibn ʻAsākir, Ta’rīkh, vol. LXI, p. 422. Although Nāfiʻ prominently figures in the ḥadīth literature, he remains an elusive personality. For a critical approach to Nāfiʻ’s eminence, see Juynboll, “Nāfiʻ, the Mawlā of Ibn ʻUmar”, pp. 217, 219-221. Juynboll is also skeptical about the historical feasibility of the purported isnād strand Nāfiʻ/Ibn ʼUmar/Prophet. See pp. 240-241. Ibn ʻAsākir provides no additional details about Nāfiʻ’s mission to Egypt, which is reported on the authority of Nāfiʻ himself who also asserted that ʻUmar II had appointed him to collect the ṣadaqāt taxes in Yemen.⁵⁰ Ibn Saʻd, Kitāb al-Ṭabaqāt, vol. VII, p. 423; Ibn ʻAsākir, Ta’rīkh, vol. LXI, pp. 424, 432, 434; Juynboll, “Nāfiʻ, the Mawlā of Ibn ʻUmar”, p. 217.

ʻUmar II’s involvement with the inculcation of Islamic values comes to the fore through al-Maqrīzī’s long quote from al-Kindī’s lost Kitāb al-Mawālī. The quote from al-Kindī (897-961), the famous historian of Egypt and its governors and cadis, begins with the statement that Yazīd ibn Abū Ḥabīb was the first who expressed his thoughts about what was permissible and forbidden in Fusṭāṭ (Miṣr). He was also one of three persons authorized by ʻUmar II to issue religious/legal opinions. The task was delegated to two mawālī (the aforementioned Yazīd and ʻUbayd Allāh ibn Abū Jaʻfar) and an Arab, Jaʻfar ibn Rabīʻ. The caliph dismissed a complaint raised by the Arabs, saying that it was not his fault that the mawālī exceeded the Arabs in this domain. To illustrate the prominent position held by these two mawālī in Egypt, one of al-Kindī’s main authorities (Ibn Qadīd) is quoted as saying that they were the first to offer the oath of allegiance to the new caliph and the people followed them.⁵¹ Al-Maqrīzī, Musawwadat Kitāb al-Mawāʻiẓ, p. 327. Al-Kindī’s Kitāb al-Mawālī was composed for the qāḍī Muḥammad ibn Badr (878-942), whose father was a convert to Islam and a mawlā. See al-Maqrīzī, Kitāb al-Muqaffā, vol. V, p. 425.

The notion that during the Umayyad period certain people in the provinces were recognized as being towering figures in the world of ʻilm is clearly borne out by al-Maqrīzī’s biography of the famous ḥadīth transmitter Ibn Shihāb al-Zuhrī (671-724). He is described as being the first who recorded and put down ʻilm, which must be understood as broadly referring to knowledge and particularly to religious knowledge or, in a narrow sense, to ḥadīth.⁵² Al-Maqrīzī, Kitāb al-Muqaffā, vol. VII, pp. 244, 245, 246. These accounts provide a broader context for ʻUmar II’s delegation to Ifrīqiya but the Islamization of the province remains unexplained, especially as the literature offers little information about the possible link between taxation and the Islamization of Ifrīqiya. Although Mohamed Hassen has stated that: “Lʼévolution des structures foncières est considérée comme un aspect fondamental de lʼislamisation de lʼIfrīqiya at de la Sicile au haut Moyen Âge”, the study fails to show a direct link between the two.⁵³See “Genèse et évolution du système foncier en Ifrīqiya”, p. 309. Amara has made the general observation that the lands of the Berbers who had converted to Islam were subjected to the tithe. See “Lʼorganisation foncière du Maghreb central”, p. 55. As a general observation, Hassen’s remark fits the drift of the sources, which perceive a linkage between taxation and Islamization. The notions that Berber converts to Islam enjoyed tax exemptions is emphasized by what is described as the misdeeds of ʻAmr ibn ʻUbayd Allāh al-Murādī, the governor of Tangier (Ṭanja) on behalf of ʻUbayd Allāh al-Ḥabḥāb (finance director in Egypt, 724-734, and governor of Ifrīqiya, 734-741). He increased the ṣadaqāt taxes levied from Muslims and re-imposed the khums (the fifth) on Berbers who had converted to Islam.⁵⁴ Ta’rīkh, p. 67. For Ṭanja, see Cornu, Atlas, p. 118, Map, XII.

Another motive that appears as an explanatory trope for the Islamization of North Africa goes back to the events of the Kāhina rebellion and postulates a link between a military victory and the subjection of people and territories to Islamization. The Islamization of the region known as Sūs al-Adnā, between Tangier and Fez (Fās), is explained in that way and dated to the governorship of Musā ibn Nuṣayr who was appointed in 705 and vanquished the Berbers who, subsequently, accepted Islam and devotedly adhered to their new faith. Musā also installed a group of 27 people in Tangier to teach the Berbers the Koran and law and instructed the local governor to supervise the educational mission.⁵⁵ Al-Raqīq al-Qayrawānī, Ta’rīkh, pp. 51-52. For the geography of the region, see Cornu, Atlas, pp. 111, 117, Map XII.

In late medieval Egyptian and North African historiography, the link between a military victory and Islamization became prominent and repetitive. Like al-Maqrῑzῑ the Mamlūk historian Ibn Taghrī Birdī (1409-1470), was well- informed about early Islamic North Africa and how its affairs were interwoven with the history of Egypt. He portrays the Islamization of Ifrīqiya as having taken place during the reign of ʻUthmān (644-656) and being related to the campaigns of Ibn Abī Sarḥ, but is somewhat contradictory about whether Islam indeed became deeply rooted among the Berbers. He is also vague about the geographical context of these events and refers to the Mediterranean coast (saḥl) and inland hilly country (jabal).⁵⁶ Ibn Taghrī Birdī, Nujūm, vol. I, pp. 79-80, 85. Ibn ʻIdhārī, who was better informed about the history of North Africa, is more cautious about the paradigm that Islamization came in the wake of military victories. He, for instance, attributes to ʻUqba ibn Nāfiʻ the observation that defeating the Christians was not enough since they reverted to their former religion once the Muslim armies had withdrawn. ʻUqba explained the foundation of Kairouan (670) as driven by the desire to establish a permanent stronghold for Islam in the region. Ibn ʻIdhārī summarizes ʻUqba’s exploits by stating that he was a good military leader and a good governor who was keen to spread Islam (daʻwa).⁵⁷ Ibn ‘Idhārī, Kitāb al-Bayān, vol. I, pp. 19, 31.

The question of the superficiality of conversion to Islam is another issue mentioned by Ibn ʻIdhārī and appears as a central motive in his account of the rebellion (685-688) of the Berber leader Kasīla (or Kusayla). ʻUqba ibn Nāfiʻ was warned by the former governor Abū l-Muhājir that Kasīla’s conversion was shallow but he mistreated him and brought about his rebellion. As informative as Ibn ʻIdhārī is about the history of North Africa, he provides no valuable information about the Islamization of the region. He, for example, states that the establishment of Muslim rule in Tangier in 704 brought about the Islamization of al-Maghrib al-Aqṣā.⁵⁸See vol. I, pp. 39, 43. Evidently, the topic was not his top priority and he was unable, or unwilling, to go beyond his sources. The unavoidable conclusion is that literary sources provide no satisfactory explanation for the Islamization of North Africa and whether there are other more promising lines of inquiry is beyond the scope of this paper.

I would argue that we are at the limit of what the sources and studies have to offer about the historical aspects of the Islamization of Ifrīqiya. For reasons that are beyond our grasp, medieval authors provide no information on the mechanism, the scope and the tempo of the process. They present the process as driven by conquest and taxation without establishing direct explanatory link between the two. Beyond the case study of Ifrīqiya, the theme “conquest, taxation and Islamization” surfaces in many contexts but it appears to be a literary motif rather than an explanatory model. Al-Balādhurī, for example, depicts ʻUmar II as having been an active proselytizer who invited the rulers of Transoxiana to embrace Islam but few had responded positively. In Khurāsān, where Umayyad rule was in the process of consolidation, he exempted the converts from the kharāj. The link between Islamization and taxation came to the fore during the reign of al-Hishām (724-743), who appointed Ashras ibn ʻAbd Allāh al-Sulamī on the finances of Khurāsān. He reissued the call to the rulers of Transoxiana to embrace Islam while, in Khurāsān, he exempted the converts from the jizya, which was an inducement that appealed to many. Consequently, the tax income (kharāj) dropped and he re-imposed taxes on fresh converts (musālima, the plural form of maslama). Al-Balādhurī, referring to al-Muʻtaṣim’s reign (833-842), states that Islam prevailed in Transoxiana.⁵⁹ Al-Balādhurī, Futūḥ, pp. 599-600, 602-603; El-Housseini, “The Umayyad Policy in Khorasan”, pp. 1-21, esp. p. 8; Athamina, “Taxation Reforms”, pp. 337-338.

Al-Ṭabarī provides a long account of ʻUmar II’s tax policies in Khurāsān. The governor of Khurāsān, al-Jarrāḥ ibn ʻAbd Allāh al-Ḥakamī, sent a delegation to Damascus and one of its members, a mawlā named Ṣāliḥ ibn Ṭarīf, told the caliph the bitter truth about the situation in the province. Mawālī and converts were discriminated against: the mawālī participated in raids but received no pay while the kharāj was levied from the converts. Furthermore, the governor publicly declared his uncompromising preference for the Arabs. The caliph was impressed with the messenger’s candidness and took measures to remedy the situation. He ordered the governor to lift the jizya from the converts and many converted in order to avoid the tax. The governor, however, began to inquiry into whether they had undergone circumcision, and wrote about this to the caliph who replied: “God sent Muḥammad as summoner (to Islam) and not as a circumciser”.⁶⁰ Al-Ṭabarī, Ta’rīkh, vol. VI, p. 559; trans. Powers, p. 83. Al-Yaʻqūbī’s version of these events (vol. II, p. 362) is terse and the governor’s advocacy of Arab superiority is alluded to by the term ʻaṣabīya (partisanship, zealotry). See also Urban, Conquered Populations, p. 176. For the wide range of views concerning circumcision in Islam, see Kister, “…And He was Born Circumcised…”, pp. 10-30.

In Ibn Saʻd’s version of these events the caliph was the one who set the Islamization process in Khurāsān in motion. He wrote to al-Jarrāḥ ibn ʻAbd Allāh al-Ḥakamī and ordered him to summon people to Islam and to cancel the jizya from converts. ʻUmar II also ruled that converts should receive what Muslims receive and be subjugated to what Muslims were subjugated. A Khurāsāni notable asked the caliph to impose circumcision on the converts but ʻUmar II refused. According to Ibn Saʻd, the caliph’s success with conversion was modest, only 4,000 people converted following the caliph’s letter. Ibn Saʻd also states that ʻUmar II ordered equal remuneration for the Arabs and the mawālī and provided special payment to freedmen i.e. slaves and captives who embraced Islam and were set free.⁶¹ Ibn Sa‘d, Kitāb al-Ṭabaqāt, vol. VII, pp. 366, 375. The Islamization of Transoxiana was far more complex than the accounts of al-Balādhurī and al-Ṭabarī depict and whether their remarks about Khurāsān offer any valuable insight into the Islamization of the province is beyond the scope of this paper.

The broader question that must be asked is whether the linkage between taxation (poll-tax) and Islamization, repetitively stated by the sources, withstands the test of evidence. A powerful argument for the link between the two has been made by Yossef Rapoport in the case of the Fayyūm. During the late Ayyubid period, the imposition of a flat rate of jizya of two dīnārs per person, irrespective of his circumstances, triggered the conversion process in the province.⁶² Rapoport, Rural Economy, pp. 215-219, 245, 248.The overall significance of Rapoport’s findings is not clear yet and has no necessarily bearing on the subsequent section of the paper, dealing with the policies of ʻUmar II.

III

The sources provide ample information on ʻUmar II’s policies, which have also attracted considerable scholarly attention. These policies are described as having been promulgated through a “Fiscal Rescript” and a “ghiyār edict”. A main source for the caliph’s policies is the Sīra (caliph’s biography) written by Ibn ʻAbd al-Ḥakam (d. 871), which consists of reports about his deeds and fragments of correspondence but not a “rescript” or an “edict”, in the strict technical sense of these terms. In some cases, the author informs his readers to whom ʻUmar II’s letters were addressed but many letters in the Sīra are presented in the following manner: “ʻUmar ibn ʻAbd al-ʻAzῑz wrote: From the servant of God, Commander of the Faithful, to ʻumarāʼ al-ajnād…”. On the same page of the printed text there is another fragment: “ʻUmar ibn ʻAbd al-ʻAzīz wrote: From the servant of God, Commander of the Faithful, to the ‘ummāl…”.⁶³ Ibn ʻAbd al-Ḥakam, Sīra, p. 91. Ibn ʻAbd al-Ḥakam’ sources are enigmatic and have not been studied. Some of his reports begin with phrases like: “he wrote about…”, and “he wrote to…”. See, for example, pp. 37, 38, 65, 68. For two additional short fragments addressed to umarāʼ al-ajnād, see Ibn Saʼd, vol. VII, pp. 349, 366. One of Ibn Saʻd’s informants stated that he was at a mosque in Kūfa when ʻUmar II’s letter was publicly read. See vol. VII, p. 365. Another of Ibn Saʻd’s informants stated that he read a risāla of ʻUmar II, which was sent to the governor Yazīd ibn al-Muhallab. See vol. VII, pp. 369-370. Tillier provides a list of recipients, including umarāʼ al-ajnād, that appear in the caliph’s letters, which were transmitted by ʻAbd al-Razzāq al-Ṣanʻānī in the Muṣannaf. See “Califes”, p. 172. The term ʻumarāʼ al-ajnād can be understood as meaning provincial governors, while ‘ummāl can indicate provincial fiscal administrators but the use of these terms is imprecise.

The question of how ʻUmar II’s letters were preserved and accessed by Ibn ʻAbd al-Ḥakam and other historians is an issue that must be discussed. The existence of archives in early and late medieval Muslim Middle East has been firmly established by modern scholarship. One might argue that the caliph’s letters were deposited in the archive of the Chancery in Damascus as well as in the archives of the provincial governors and, therefore, widely available. Furthermore, old archives were sold as scrap paper and the historians incorporated fragments of state documents in their writings.⁶⁴According to Bravmann archives had existed already by time of the caliph ʻUthmān. See “State Archives”, pp. 87-89. For Abbasid administrative archives, see van Berkel, “Reconstructing Archival Practices”, pp. 12-14. For archives and the sale of documents as scrap paper, see Sijpesteijn, “The Archival Mind”, pp. 163-167; “Multilingual Archives and Documents”, pp. 105-124. For an allusion to the archival practices of ʻUmar I, see Ibn ʻAbd al-Ḥakam, Futūḥ, p. 89. Khalīfa ibn Khayyāṭ’s lists of state officials, from 705, 717 and 734 contain references to kātib al-rasāʼil (the clerk responsible for the writing of letters). The Abbasid caliph al-Manṣūr added a new position that of kātib rasāʼil al-futūḥ (the clerk responsible for the writing of letters announcing conquests). Khalīfa b. Khayyāṭ, Ta’rīkh, vol. I, pp. 302, 325, 379. The usual tendency in medieval Muslim historiography was to credit ʻUmar I with the establishment of administrative offices. See, for example, al-Quḍāʼi, Taʼrīkh, p. 89; Kitāb al-Inbāʼ, p. 180. In this context, one must take notice of Wadād al-Qāḍī’s detailed study of the authenticity and transmission of the letters of ʻAbd al-Ḥamīd ibn Yaḥyā, the kātib of the caliph Marwān II (both of whom were killed in 750 in Egypt). She lists fourteen literary sources spanning six centuries (9^th-15^th) in which some of these letters have been preserved.⁶⁵ Al-Qāḍī, “Early Islamic State Letters”, pp. 277-317, esp. pp. 232-234. The main question is why ʻAbd al-Ḥamīd ibn Yaḥyā’s letters have been preserved and the answer seems to be that they were considered to have intrinsic value. Most likely, this explanation also applies to the fragmentary preservation of ʻUmar II’s letters in the Sīra, the aim being to extol the meritorious qualities of the caliph’s personality and policies and to depict him as a model Muslim ruler. In the Arabic sources, the caliph is presented as one who followed the deeds of ʻUmar ibn al-Khaṭṭāb (ʻUmar I 634-644). According to al-Balādhurī, the caliph asked Sālim ibn ʻAbd Allāh ibn ʻAmr to compose a biography of ʻUmar I for him and the designated author pointed out that ʻUmar I’s age and the people of his time bore no resemblance to ʻUmar II’s epoch. He assured the caliph that if he would be able to imitate ʻUmar I’s deeds, he would surpass him.⁶⁶ Al-Balādhurī, Ansāb, vol. VIII, p. 169. The motif that ʻUmar II faced greater challenges than ʻUmar I and, potentially, could surpass him is prominent in Ibn ʻAsākir’s biography of the Umayyad caliph. See Khalek, “Early Islamic History Reimagined”, pp. 447-448. The traditional depiction of ʻUmar II as an unusual caliph is echoed by modern scholarship but Patricia Crone and Martin Hinds have questioned the authenticity of his vast correspondence some of which they regard as “apocryphal” and “super-apocryphal”.⁶⁷ Crone, God’s Caliph, p. 77. For the authors critical approach to ʻUmar II’s rulings and his alleged imitation of ʻUmar I, see pp. 78-80. For ʻUmar II’s depiction as an exemplary caliph in line with the rightly guided caliphs of Medina, see Borrut, “Entre tradition et histoire”, pp. 329-345. I owe this reference to the kindness of one of the anonymous readers on behalf of Al-Qanṭara. Mathieu Tillier finds no reason to doubt the extent of ʻUmar II’s correspondence and its preservation because of his special status while the degree of ʻUmar II’s entanglement in legal matters is revealed by Tillier’ study of ʻAbd al-Razzāq al-Ṣanʻānī’s Muṣannaf, which is considered to be one of the earliest preserved compilations of mainly legal traditions dating from the eighth century. The caliph usually corresponded with governors but, in the case of Egypt, he wrote directly to the qāḍī.⁶⁸ Tillier, “Califes”, pp. 168-170, 172-175. For the text, see Motzki, “The Author and His Work”, p. 176; “The Muṣannaf of ʻAbd al-Razzāq al-Ṣanʻānī”, pp. 1-21. ʻUmar II’s Chancery is discussed by Elizabeth Urban, who has pointed out that the caliph employed eight scribes of whom six were mawālī. Apparently, he had dictated his letters.⁶⁹ Urban, Conquered Populations, p. 165.

Ibn ʻAbd al-Ḥakam and other historians refer to ʻUmar II’s letters not as documentary fragments but as reports (khabar pl. akhbār) whose transmission was modeled on the way ḥadīths were recorded, including the names of the informants responsible for each khabar - in other words with isnād strands for each khabar. It seems that Ibn ʻAbd al-Ḥakam’s knowledge of the letters was not derived from seeing the original texts but from hearing them narrated and, possibly, recorded in the way that ḥadīths were put down. This might explain why the context of ʻUmar II’s letters has been lost. The use of akhbār in historical writings was widespread and is extensively discussed in modern scholarship. It is important to bear in mind some of Stefan Leder’s remarks about this method and especially the fact that: “Akhbār in general are characterized by an intense and complex process of ‘reactive transmission’, including unadmitted authorship and fundamental reshaping of the ‘original” narrations’. Two other characteristics of the akhbār genre must be pointed out: typically akhbār lack context and in the medieval texts they are juxtaposed one to another.⁷⁰ Leder, “The Literary Use of the Khabar”, pp. 278-279.

One of al-Kindī’s historical writings, which did survive is the history of the Egyptian qāḍīs and it contains fragments of the correspondence between ʻUmar II and the qāḍī ʻIyāḍ ibn ʻUbayd Allāh. Six such fragments are quoted, including information on specific cases and the caliph’s instructions how they should be handled. Because of its length and complexity, however, one fragment must have been copied from the original letter kept in an archive or preserved in another way. The possibility that the letter was kept in the qāḍī’s archive in Fusṭāṭ must not be ruled out.⁷¹For the qādī’s archives, see Hallaq, “The Qāḍī’s Dīwān”, pp. 415-436; Tillier, “Le statut et la conservation des archives judiciaires”, pp. 263-277. The caliph’s letter begins with an introductory phrase: “In the name of God, the Merciful and Compassionate. From the servant of God, the Commander of the Faithful, to ʻIyāḍ ibn ʻUbayd Allāh”. The fragment ends with al-Kindī’s statement that the letter was written on Thursday morning 4 Dhū l-Ḥijja 99 (7 July 718)”.⁷² Al-Kindī, The Governors and Judges of Egypt, pp. 333-337, esp. pp. 336-337. For annotated French translation, see Tillier, Histoire des cadis Égyptiens, pp. 87-92.

The discussion about how the caliph’s letters were transmitted and Leder’s remarks about the akhbār genre must be taken into account when the fragment described by Hamilton A. R. Gibb as “The Fiscal Rescript of ʻUmar II”, is examined. In the printed edition of the Sīra the section in question begins in the following way: “ʻUmar ibn ʻAbd al-ʻAzīz wrote: From the servant of God, Commander of the Faithful, to the ʻummāl…”. While it can be argued that the text is subdivided into paragraphs beginning with introductory phrases, the coherence of the text is loose.⁷³ Ibn ʻAbd al-Ḥakam, Sīra, pp. 93-99. In Gibb’s translation the internal division of the text is marked by Roman numerals I to XX. The main question is what the text known as “The Fiscal Rescript” represents and how it was put together. Did it exist as one long letter written by the caliph in unknown circumstances or is it a collage of fragments from the caliph’s correspondence put together by Ibn ʻAbd al-Ḥakam? Even if one accepts Gibb’s assumption that it was an organic text, and not a sequence of juxtaposed akhbār, the question of whether it was a “Fiscal Rescript” remains.⁷⁴As has been pointed out by Crone and Hinds, in Roman law, rescript meant the emperor’s answer to a question addressed to him. See God’s Caliph, p. 46, n. 23, 47.

The so-called “Fiscal Rescript” is available in both English and French translations. The section of the text, immediately following the phrase: “ʻUmar ibn ʻAbd al-ʻAzīz wrote: From the servant of God, Commander of the Faithful, to the ʻummāl…”, was labelled by Gibb as the Preamble. It states that the nature of the prophetic mission was to bring a universal message of guidance and the religion of truth, and quotes Koran 9:34, 2:108 and 7:157. For the topic of Islamization, the most pertinent passage seems to be the following one: Paragraph II (Gibb’s pagination and translation):

The policy announced by the caliph can be summarized as follows: the converts are, or rather should be, exempted from the poll tax. Fiscally they should be treated on equal terms with Muslims although they have to fulfill certain conditions. It can be fairly said that the demands outweighed any tax exemption offered. One can argue that the caliph’s directives as to how converts should be treated shed no light on the motive behind the conversion process.⁷⁶Some of the information about ʻUmar II’s tax policies is ambiguous. The governor of Egypt, for example, received a letter from the caliph, stating that the threshold for taxing a Muslim should be 40 dīnārs, while for a Christian or a Jew it should be 20 dīnārs. The report can be understood as stating that the threshold for the Islamic alms-tax levied from Muslims was 40 dīnārs, while the threshold for the payment of the poll-tax was 20 dīnārs. The practical implications, if any, of this directive are difficult to interpret. See Ibn Yūnus, Ta’rīkh, vol. I, pp. 54-55. In another report, it is stated that ʻUmar II abolished custom duties and jizya. See Ibn Saʻd, vol. VII, p. 339.

Reports about ʻUmar II’s directives concerning how converts to Islam should be treated are also presented in Ibn ʻAbd al-Ḥakam’s Futūḥ Miṣr, a book about the conquest of Egypt and the Islamic west. These reports do not pretend to be fragments from the caliph’s correspondence but are rather summaries of his sayings presented through short chains of transmitters. One of these reports attributed to ʻUmar II states that the meaning of conversion is that the convert gains protection for his māl (money/property) but his arḍ (land) becomes fayʼ Allāh in favor of the Muslims. Another contradictory report (also attributed to ʻUmar II) states that when a community (qawm) reached a ṣulḥ agreement for the payment of the poll-tax and one of the parties has converted, his land and house remain in the community.⁷⁷ Ibn ʻAbd al-Ḥakam, Futūḥ, p. 154. The author also quotes other more detailed accounts, not directly on the authority of ʻUmar II, about the rights and obligations of the converts. The term ṣulḥ refers to the surrender agreements made during the Muslim conquest of the Middle East between Arab generals and the local population. Given the purpose of Futūḥ Miṣr, which discusses the link between the status of the lands conquered by the Muslims and taxation, the impression is that the use of the reports attributed to ʻUmar II was manipulative and thus their authenticity must be suspect. The author explicitly says that a letter of the caliph, supposedly claiming that the poll-tax of dead Copts should be imposed on their brothers, testifies to the fact that the caliph considered Egypt to be a land conquered by force.⁷⁸ Ibn ʻAbd al-Ḥakam, Futūḥ, pp. 154-155. In another report, it is explicitly stated that the caliph considered Egypt as a land conquered by force. See p. 90.

In contrast to Futūḥ Miṣr, the text of the Sīra can be perceived as reflecting internal Muslim discours about how to harmonize Koranic ethics with taxation and the contentious issue was the collection of the jizya from the converts. The fact that the issue was high on the caliph’s agenda is also attested to in the chronicle of Theophilus of Edessa, who was in the service of the caliph al-Mahdī (775-785). He explicitly states that many Christians converted but he paints a complex picture of the caliph’s religious policies, which can be described as a stiffening of the permeable religious boundaries between Islam and Christianity. According to Theophilus of Edessa, ʻUmar II forbade wine and fermented drinks to Muslims and declared that a Christian’s testimony against a Muslim was invalid.⁷⁹For the reconstruction of the text and translation, see Hoyland, Seeing Islam as Others Saw It, p. 654. From a doctrinal point of view, the acceptance of a non-Muslim’s testimony against Muslims contravened the notion of the exaltedness of Islam. For a broad discussion of the subject, see Friedmann, Tolerance and Coercion in Islam, pp. 34-38, esp. pp. 35-36. In Byzantine sources, which draw on Theophilus of Edessa, ʻUmar II is described as a persecutor of Christians who used tax exemptions as a tool to force them to convert to Islam.⁸⁰ The Chronicle of Theophanes Confessor, p. 550. For broader discussion of permeable religious boundaries and the attitudes of religious leaders, see Simonsohn, “The Pact of ʻUmar”, pp. 35-57.

A rich source for ʻUmar II’s policies is al-Balādhurī’s account of the caliph’s reign in Ansāb al-Ashrāf, which sheds light on many issues discussed so far. The text, for example, offers insights into the caliph’s perception of the rites of conversion and into the deeper meaning of embracing Islam. The caliph’s governor of Baṣra, ‘Adī ibn Arṭāt, wrote to ʻUmar II, asking for his opinion about the fact that many non-Muslims “…seek the protection of Islam being afraid of the jizya”. The caliph’s response begins with the statement: “God sent the Prophet as summoner (to Islam) and not as a tax collector”. The caliph also declared that a convert was entitled to what Muslims got (māl) and should be subjugated to what Muslims were subjugated to. On the one hand, the reference to māl can be understood as alluding to the payments rendered to Muslims by the dīwān system. On the other hand, one can also argue that by declaring that the converts should be subjugated to what Muslims were subjugated, the caliph was alluding to the levy of the compulsory alms tax imposed on the Muslims. The caliph sent very specific instructions about how his policies should be implemented: the governor was ordered to make sure that a person who had converted (aẓhara al-Islam) underwent circumcision and read a sūra from the Koran. Having done this, the jizya should be lifted from him.⁸¹ Al-Balādhurī, Ansāb, vol. VIII, p. 146.

Ibn Saʻd offers a different context for the caliph’s statement about the prophetic mission of Muḥammad according to which it was the governor of Egypt who informed the caliph that people were hastening to embrace Islam and stop paying jizya. In response the caliph stated that: “God sent the Prophet as summoner (to Islam) and not as a tax collector”, and declared that he was fine with the process.⁸²See vol. VII, p. 373.

In the source material, the caliph’s letters to the governors are presented as following a simple format, which included stating the conceptual (religious) underpinnings of his policies and specific instructions on what should be done. Another letter to an unnamed governor begins with the declaration that God honored his people with Islam and through Islam alleviated them from an ignominious existence. On the practical level, the caliph instructed the governors to make sure that a person who proclaimed Islam should declare: “There is no god but God and Muḥammad is His Prophet”. The convert also had to state that he believes in God, His angels and His messengers. Additional statements of belief were required from Christians and Jews. A Christian had to declare that: “ʻIsā is a servant of God, the word of God and His messenger”, while a Jew had to declare that: “ʻUzayr is a servant of God.” All converts were supposed to perform the prescribed number of daily prayers at the right times, to read at least the opening sūra of the Koran and to scrupulously perform the ablution. The governors were also instructed to check that a convert underwent circumcision. The fulfillment of all these requirements earned the convert exemption from the jizya.⁸³ Al-Balādhurī, Ansāb, vol. VIII, p. 162.

Al-Balādhurī’s reference to the caliph’s insistence on circumcision as a requirement for conversion invokes the question of the letter’s authenticity, especially as al-Ṭabarī attributes a completely different stand on the issue to the caliph. I see no reason to doubt that Ibn ʻAbd al-Ḥakam, al-Kindī, al-Balādhurῑ, al-Ṭabarī and others had knowledge of ʻUmar II’s correspondence but what they presented to their readers is an edited selection of the letters written by the caliph.⁸⁴The same can be said about Ibn Saʻd, who presents condensed short fragments of these letters. The editorial process was shaped by their perception of Islam, or what it should be, in terms of practice and beliefs. In some cases (to be discussed later) we are able to ascertain what ʻUmar II’s views on certain issues were. In other cases, the editorial process partially or completely distorted the caliph’s views on different issues and the case of circumcision belongs to this category. We do not know what the caliph’s thoughts on circumcision were and it can be argued that al-Balādhurῑ and al-Ṭabarī projected their own contradictory views about the subject on the caliph and edited his letters accordingly.⁸⁵In a broader historiographical perspective, al-Balādhurῑ and al-Ṭabarī reached different conclusions about the demise of the Umayyad regime. Although, in this case, they relied on a common source (al-Madāʼinῑ d. ca. 840). See Judd, “Narratives and Character Development”, pp. 209-227.

The caliph’s justification for the jizya is articulated in another letter sent to the governor of Baṣra, which begins with the statement that God imposed the jizya on those who erred and shunned Islam. The operative part of the letter, however, takes a completely different direction. The governor is instructed to pay attention to elderly non-Muslims and to provide them with nourishment from the Treasury. In another letter with identical instructions, the caliph justifies his directive by invoking the deeds of ʻUmar I. Allegedly ʻUmar I once saw an old non-Muslim beggar and told him: “We have not provided justice to you. When you were young we took from you the kharāj and in your old age we have forsaken you”. The caliph consequently gave orders to supply him with nourishment from the Treasury.⁸⁶ Al-Balādhurī, Ansāb, vol. VIII, pp. 147, 204.

One can argue that al-Balādhurī, like others, sheds no light on the mechanism of the conversion and what drove the process. The fragments quoted by al-Balādhurī do provide, however, significant insight into the acculturation process and the desire to blend into street life. Although a certain letter addressed to one of the ʻummāl lacks context, the caliph states that he was informed that many non-Muslims resemble Muslims in their clothes and turbans. The governor was instructed to strictly prevent this and was also ordered to make sure that non-Muslims shave the middle of their heads.⁸⁷ Al-Balādhurī, Ansāb, vol. VIII, p. 151; Yarbrough, “Origins of the ghiyār”, p. 117. One is inclined to conclude that acculturation was swifter than Islamization and was considered as a more threatening process. If indeed acculturation was the decisive social process that took place in the urban centers of the caliphate, one can understand why the caliph firmly objected to the appointment of non-Muslims officials. The combination of acculturation and the integration of non-Muslims into the administrative structure of the state was perceived as undermining Muslim predominance and identity.⁸⁸The caliph’s letter illustrates only one aspect of the acculturation process: the desire of the non-Muslims to blend into the culture and street life of the urban centers of the caliphate. At the other end of the process was the Islamic doctrine of lā tashabbū, which opposed imitation of non-Muslims by Muslims. It represented a wider conceptual drive to distinguish Islam from Christianity and Judaism. See, for example, Kister, “Lā Tashabbū”, pp. 321-371. For turbans as a marker of Muslim men, see Kister, “The Crowns of this Community”, pp. 217-245. For more recent publications, see Patel, “Whoever Imitates a People”, pp. 359-426; “The Islamic Treatises against Imitation”, pp. 597-639.

The question of whether ʻUmar II issued an edict ordering the dismissal of non-Muslim officials has been discussed by Luke Yarbrough. I would argue that the use of the term “edict” should be avoided and one has to closely examine Ibn ʻAbd al-Ḥakam’s text, which contains a letter, devoid of context, written by the caliph to one his ʻummāl. Referring to the Koran, the letter begins with a conceptual justification for the idea that non-Muslims should be kept in the position of al-dhull wa-l-ṣaghār (humiliation and abasement) and, therefore, non-Muslim officials should be dismissed.⁸⁹The word ṣaghār that became a technical term for humiliation of non-Muslims invokes the Koranic verse 9:29, which states the jizya should be paid out of hand in humiliation (ṣāghirūn). Strangely, this specific verse is never alluded in the letters referring to the payment of the jizya. The caliph also asked the addressee to report on the implementation of the instruction. Oddly, the text continues with a further set of instructions concerning what kind of saddles Christian men and women should use. This section of the text ends with a demand from the addressee to forward the instructions to his subordinate ʻummāl and to see that these regulations were strictly enforced.⁹⁰ Ibn ʻAbd al-Ḥakam, Sīra, pp. 158-159; Yarbrough, “Did ʻUmar b. ʻAbd al-ʻAzīz Issue an Edict”, p. 176. On purely stylistic grounds, the disjointed format of the text indicates that it is not an organic text but rather juxtaposition of fragments.

One of ʻUmar II’s letters reported by al-Balādhurī seems to present another, more cohesive version of the same letter in which he provides the names of two of his informants but offers no context. With no references to the Koran, the caliph states his philosophy about what the position of non-Muslims should be. The caliph’s thinking was dominated by the perception that policies must be in harmony with God’s deeds, and the insistence on dhull wa-l-ṣaghār is actually an act that creates harmony between God and human actions. The dismissal of non-Muslim officials and the imposition of distinctive clothing and riding on pack saddles are policies which follow God’s example.⁹¹ Al-Balādhurī, Ansāb, vol. VIII, p. 196; Yarbrough, “Did ʻUmar b. ʻAbd al-ʻAzīz Issue an Edict”, p. 177. The caliph’s letters as reported by both Ibn ʻAbd al-Ḥakam and al-Balādhurī allow conceptualization of the caliph’s policies but fall short on contextualization.⁹²See, for example, al-Balādhurī, Ansāb, vol. VIII, p. 164; Yarbrough, “Did ʻUmar b. ʻAbd al-ʻAzīz Issue an Edict”, p. 177. For ʻUmar II’s propensity to rely on the Koran, see Tillier, “Califes”, p. 166. In one case, however, the context is provided: the governor of Baṣra dismissed two officials and was specifically ordered to dismiss a third one.⁹³ Al-Balādhurī, Ansāb, vol. VIII, p. 164. For the identity of these officials, see Yarbrough, “Did ʻUmar b. ʻAbd al-ʻAzīz Issue an Edict”, pp. 178, 185-6. Non-Muslims and conversion figure prominently in ʻUmar II’s letters transmitted by ʻAbd al-Razzāq al-Ṣanʻānī in the Muṣannaf. See Tillier “Califes”, p. 168.

It can be argued that in the caliph’s thinking acculturation stood in stark contrast to the dhull wa-l-ṣaghār paradigm and the caliph’s policies on the subject were promulgated through the ghiyār edict, i.e. an edict, which specified what distinctive clothes the non-Muslims should wear, how they should ride and what shape their haircuts should be. A short letter reproduced by Ibn ʻAbd al-Ḥakam supposedly constitutes the ghiyār edict. The text is presented in the following way: “ʻUmar ibn ʻAbd al-ʻAzīz wrote to the regions of the land (āfāq)…”. In three lines of text, the instructions about clothing, riding and the hairdos of Christians are spelled out, and the text ends with the warning that no weapons should be found in their homes. In the printed edition of Ibn ʻAbd al-Ḥakam’s text, the letter is juxtaposed with the so-called edict concerning non-Muslim officials.⁹⁴ Ibn ʻAbd al-Ḥakam, Sīra, p. 159; Levy-Rubin, Non-Muslims in the Early Islamic Empire, p. 90; “ʻUmar II’s ghiyār Edict”, p. 161. Even by Ibn ʻAbd al-Ḥakam’ slack standards of textual transmission this fragment stands out as being a haphazardly put together collage of instructions taken from different letters.

Other fragments of ʻUmar II’s letters provide insights into the caliph’s broader vision of social relations and policies, which have not been studied so far.⁹⁵For accounts relating to legal/social justice, see Al-Balādhurī, Ansāb, vol. VIII, pp. 137, 140, 149, 177, 182-184. Although the caliph had a clear vision about the relations between Muslims and non-Muslims, it cannot be said that his world-view was dominated by this issue. As stern as the caliph’s views about Muslim superiority were, his world-view also accentuated the need for the just treatment of non-Muslims and a charitable attitude toward their needs. Two variants in al-Ṭabarī’s edited monumental History of the Prophets and Kings shed light on the issue and highlight the editorial process that the caliph’s letters underwent. According to one version of the text, ʻUmar II sent instructions that non-Muslim houses of worship, such as churches, synagogues and fire temples, and the pasture lands (thinna) belonging to churches and fire temples whose existence had been confirmed by ṣulḥ treaties, should not be tampered with. Another version of the text contains the addition that no new non-Muslim houses of worship should be built but omits the reference to the pasture lands of these institutions.⁹⁶ Al-Ṭabarī, (ed) Abū l-Faḍl Ibrāhīm, vol. VI, p. 572. The addition is to be found in De Goeje’s edition, Second Series, vol. III, pp. 1371-1372; trans. Powers, p. 101, who follows De Goeje’s edition.

A strong case for the basic authenticity of ṣulḥ agreements has been made by Milka Levy-Rubin.⁹⁷ Levy-Rubin. Non-Muslims in the Early Islamic Empire, Ch. 1. The two different versions of the text seem to represent two stages in the development of Muslim attitudes toward non-Muslim houses of worship. The first version appears to preserve ʻUmar II’s genuine letter which echoes the original stipulations of those agreements in which the local population asked to maintain their institutions and the endowed properties and the conquerors had agreed to these demands. The second version of the text represents evolving Muslim legal thinking about these agreements, which binding legality was reconfirmed but new stipulations were added. At the time of the conquest, the question of building new houses of worship was not an issue but it became later. The second version is an amended text, which reveals an unknown editorial hand at work and reflects ninth-tenth century developments in Islamic law. The obligation that non-Muslims should not build new houses of worship is, for example, attested to in Qāḍī ibn Zabr’ Shurūṭ al-Naṣāra. The author (870-940) was an Egyptian jurist and his booklet (the Stipulations [Concerning] the Christians) contains utterances attributed to the Prophet and ʻUmar I about the prohibition to build new non-Muslim houses of worship or to restore decaying ones.⁹⁸See (ed) Cohen, “What was the Pact of ʻUmar?”, pp. 133-136, 138, and passim. For the author and the text, see pp. 110-116. Significantly, the text contains four versions of the Pact of ʻUmar, each with a different chain of transmitters.

The assumed authenticity of the first version of al-Ṭabarī’s text is corroborated by Ibn Saʻd’s fragments of ʻUmar II’s letters and instructions. One of these reports states that the caliph allowed people to make bequests in favor of churches and permitted Christians and Jews to establish endowments.⁹⁹See vol. VII, p. 349. At the risk of making a broader generalization, it seems that the caliph valued social peace, fiscal justice and urged others to adhere to original practices that governed the relations between Muslims and non-Muslims. The caliph’s pragmatic approach, even in potentially inflammatory situations, can be discerned in the case of a person who had renounced Islam in Kūfa. Prior to writing to the caliph, the governor had asked people for an opinion on the matter and was advised to execute the offender. He also wrote to the caliph who instructed him to impose the jizya on the person and to leave him alone.¹⁰⁰ Al-Balādhurī, Ansāb, vol. VIII, p. 189.

Conclusions

The notion that ʻUmar II ruled through “rescripts” and “edicts” is not borne out by the sources. I have adopted Luke Yarbrough’s stand that the origin of the so-called ghiyār edict, or regulations, is intractable.¹⁰¹See “Origins of the ghiyār”, p. 113. Furthermore, I would also argue that the literary depiction of ʻUmar II’s short reign remains impermeable and that the format in which the story of his rule is cast is misleading. We are informed about his instructions and motives but left in the dark about whether his policies were ever implemented.

The caliph did write to governors but this does not mean that instructions sent to a certain governor had any wider relevance and were widely implemented. The caliphate was vast and diverse and many problems were particular and local. Regionalism should be taken into account but not exaggerated since Islamization and acculturation were sweeping processes taking place across the region. From the caliph’s point of view, these processes posed great challenges since they undermined Arab superiority. In Khurāsān there was a specific problem in which Arab predominance in the military and their privileges based on military service were jeopardized by the mawālī. Islamization undermined Arab predominance in Islam across the geographical divide and, because of ist broad implications, the process had to be regulated. It seems that the caliph’s dilemma was how to formulate a policy that was both in line with Islamic ethics and, at the same time, to control the process by setting a high entrance bar for the converts. The struggle against acculturation was an attempt to preserve Muslim political dominance by controlling the administration and barring non-Muslims from its ranks.¹⁰²The opposition to the employment of non-Muslims in the state administration typified the whole span of medieval Islam. Its origins can perhaps be traced to mid-to-late eighth century Kūfa. See Yarbrough, “Upholding God’s Rule”, pp. 57-59, 68-71. On both fronts (Islamization and acculturation) the caliph was fighting a lost cause. Eventually Islamization would create multi-ethnic Muslim society in which Arabic would hold prominent place. Acculturation, meaning the ability of non-Muslims to function in Muslim society and to take part in the affairs of the state, had many manifestations. This integration would never be smooth, sometimes revoked, and its boundaries would have to be repeatedly renegotiated.¹⁰³See, for example, the persecutions of the Abbasid caliph al-Mutawakkil (847-861), which involved the barring of non-Muslim children from Muslim primary education (kuttāb) and interdiction from teaching non-Muslim children in Arabic. Lev, “Coptic Rebellions”, pp. 327-328. For broader discussion of these events, see Levy-Rubin, Non-Muslims in the Early Islamic Empire, pp. 103-112. The history of acculturation and its manifestations is currently being written and, I would argue, that Weitz’ study of the use of Islamic courts by the Copts reflects a facet of the acculturation process, which has its parallels in the Jewish society of the Geniza period.¹⁰⁴See, for example, Rustow, “At the Limits of Communal Autonomy”, pp. 135-159; Zinger, “She Aims to Harass Him”, pp. 159-192.