Juristic discourses on non-Muslim viziers

How could non-Muslims possibly be legally appointed to the office of the vizierate? Early Muslim jurists differ on this issue. However, before attempting to answer this question, let us turn to how Muslim jurists have treated the issue of the vizierate. The earliest extensive treatment of the concept of vizierate from a legal point of view is al-Māwardī's al-Aḥkām al-sulṭāniyya, which was probably written for the caliph al-Qādir bi-Llāh (r. 991–1031) or for his successor al-Qā'im bi-Llāh (r. 1031–75).Footnote ⁴ His purpose was to give a legal exposition of the theory of government derived from the basis of theology and to set out the formal basis for government so that the ruler might fulfil the charge laid upon him. H. A. R. Gibb might be right to suggest that al-Māwardī's political theory was “merely the post eventum justification of the precedents” and “only rationalization of the history of community”,Footnote ⁵ yet his al-Aḥkām al-sulṭāniyya came to be accepted by the Muslim jurists of the eleventh century and later as the most authoritative exposition of Sunni political theory.

The structure of the vizierate is as follows: the caliph exercised his authority through the vizier, who in turn delegated his routine duties to various secretaries, each of whom controlled a dīwān or office in which a number of subordinate clerks were employed. Al-Māwardī distinguishes between two types of vizierate: the vizierate of delegation (wizāra al-tafwīḍ) and of execution (wizāra al-tanfīdh). The former implies that full powers were entrusted to the holder; such a vizier-in-charge is practically independent, exercising full power and authority, guided by his own ra'y (independent opinion) and ijtihād (judgement and decision), as long as he keeps the caliph fully informed of his actions. For this reason al-Māwardī argues that the qualities necessary for a vizier of delegation are the same as those required by the caliph himself, except for the matter of lineage (i.e. he did not have to be of Qurayshī descent).Footnote ⁶ In addition, he has to be competent in the domains of war and finance. The vizier was empowered to undertake almost all the tasks of the caliph, but he was not able to designate an heir apparent, nor could he dismiss officials whom the caliph had appointed, whereas the caliph could dismiss those appointed by the vizier.

The second kind of vizier merely executes the caliph's orders, with no independence or power to change anything his master has decided or commanded. As al-Māwardī puts it: “The vizier [of execution] is a mediator between the caliph and his subjects, carrying out his commands, executing his instructions, enacting what he decides and announcing any governmental appointment or military preparations of the armies”.Footnote ⁷ Since his functions are substantially limited, the qualities required of him are less stringent. Al-Māwardī emphasizes the qualities for this vizier: honesty, truthfulness, absence of avarice, lack of enmity or conflict with other people, good memory (to remember well the message of the caliph) and, finally, intelligence and sagacity. Al-Māwardī asserts that “a vizier of execution may be of the people of dhimma, although not a vizier of delegation”.Footnote ⁸ In other words, al-Māwardī allows public positions to be filled by non-Muslims if that position involves limited authority and no independent judgement. The reason he gives for not allowing the dhimmīs to be appointed viziers of delegation is pragmatic: the vizier of delegation enjoys wide political powers over governors, the army and the treasury. Otherwise, nothing could “prevent a dhimmī from being a vizier of execution”.Footnote ⁹

One may argue that in permitting dhimmīs to become viziers of execution, al-Māwardī is confirming a long-standing practice of caliphs and governors. Historical evidence shows that non-Muslims were employed in large numbers and high government positions under the Umayyads and the ʿAbbāsids, and they are known to have become particularly influential because of their economic power and wealth. However, this does not mean that all Muslim jurists agree with al-Māwardī's view. On the matter of whether a vizier of execution must be a Muslim, Muḥammad ibn Talḥah Abū Sālim (d. 1255) says: “Scholars differ on this issue. Imām Abū al-Ḥasan ʿAli ibn Ḥabīb al-Basrī from Iraq argues that it is permissible for the sultan to appoint dhimmīs as viziers. However, Imām al-Ḥaramayn Abū al-Maʿālī al-Juwaynī argued that it was forbidden. He considers the view of the Iraqi imām (al-Baṣrī) who allowed the appointment of dhimmīs to the position of vizier as a falsehood that will not be forgiven (ʿathra lan tuqāl)”Footnote ¹⁰

In his Ghiyāth al-umam fī iltiyāth al-ẓulam, Imām Ḥaramayn al-Juwaynī strongly opposes al-Māwardī's view. The vizier must be truthful (thiqa), al-Juwaynī argues, and “the dhimmī cannot be trusted in his actions, statements and in all his activities. His transmission is rejected, and so too his testimony about Muslims. How could his information and report for the imām of the Muslims be then accepted?”Footnote ¹¹ He puts forward a rational argument that, since the testimony of dhimmīs on trivial things is not accepted, a fortiori they should not be appointed viziers. Although he adheres to the same Shāfiʿī madhhab as al-Māwardī, al-Juwaynī claims that his view is the one that conforms with Shāfiʿī law and is supported by the texts of the Quran and Sunna. He cites two quranic verses to support his view: “O you who believe! Do not take Jews and Christians as friends; they are friends to one another” (Q.5:51) and “Do not take for your bosom-friends people who are not of your kind. They spare no effort to corrupt you” (Q.3:118). He also notes the prophetic ḥadīth: “I am free from every Muslim [who lives along side] with a mushrik (idolater). Their two fires shall not be within sight of each other”.Footnote ¹²

The Ḥanbalī jurist Abū Yaʿlā, who wrote a book with the same title as that of al-Māwardī (al-Aḥkām al-sulṭāniyya), also mentions a practice of the Companions of the Prophet that he believes supports the prohibition of non-Muslim viziers. It is reported that ʿAlī ibn Abī Ṭālib (r. 656–661) was asked whether it is permissible to employ a Jew or Christian in public offices such as kharāj. He responded: “Do not ask any help from them!”Footnote ¹³ ʿUmar ibn Khaṭṭāb (r. 634–644) is also reported to have opposed Abū Mūsā al-Ashʿarī (d. 664) when the latter employed a Christian secretary.Footnote ¹⁴ It is related that ʿUmar sat in the mosque at Medina while Abū Mūsā was in front of him presenting the accounts for Isfahan – written in a fair hand and exactly reckoned, so that all who saw admired them. “Whose writing is this?” asked ʿUmar. He said, “My secretary”. When asked to bring his secretary, he replied: “He cannot come into the mosque”. ʿUmar said: “Is he unclean then?” He said: “No. He is a Christian”. Then, ʿUmar gave Abū Mūsā a slap on the thigh – so hard that he said he thought his thigh was broken – and said, “Have you not read the command of God: ‘O you who believe, do not take Jews and Christians as friends; they are friends to one another?’” (Q.5:51). Abū Mūsā then said: “This very hour I will dismiss him and give him leave to return to Iran”.Footnote ¹⁵

The juristic debate on whether or not non-Muslims could be appointed to public office shows that, at least in the eleventh century, the emergence of powerful non-Muslim bureaucrats had become controversial. Even for al-Māwardī, the presence of powerful non-Muslims was a problem. Thus while he permits dhimmīs becoming viziers of execution, he warns that those dhimmī viziers seeking a higher rank should be stopped.Footnote ¹⁶ For reasons that will be discussed in the final section of this article, it was widespread for non-Muslims to hold public office. In fact, throughout the Umayyad and ʿAbbāsid empires, effective administration meant the employment of Christian, Jewish and, in Persia, Zoroastrian bureaucrats. The first Umayyad caliph Muʿāwiyya (r. 661–680) regularly employed non-Muslims in the service of the government. One of his Christian secretaries of finance is thought to have been Manṣūr ibn Sarjūn (Sergius) from a Byzantine Orthodox family.Footnote ¹⁷ Caliph Marwān (r. 683–685) appointed two Christian secretaries, Athanasius and Isaac. Later, Isaac became head of the governmental offices in Alexandria.Footnote ¹⁸ In short, until the end of the Umayyad caliphate, “it was seldom that no Christians were to be found in government services”.Footnote ¹⁹

Next I will discuss the two most important categories of governmental service attributed to dhimmīs during the ʿAbbāsid period: viziers and secretaries. Louis Cheikho provides a detailed list of Christians who served as viziers and secretaries.Footnote ²⁰ Here I would like to reconstruct the relationship between the caliphs and their non-Muslim viziers on the one hand, and between the caliphs and/or viziers and their non-Muslim secretaries on the other. I will argue that these non-Muslim officers received all the usual marks of honour, since they could be promoted to the highest office: the vizierate. In spite of legal restrictions as outlined by al-Māwardī, and more strictly by al-Juwaynī and Abū Yaʿlā, some dhimmīs were even able to hold what al-Māwardī calls “the vizierate of delegation”.

Historical realities in the ʿAbbāsid period

If the ʿAbbāsids claimed that their call for overthrowing the Umayyads was to organize the administration and regulation of the state according to more clearly defined Islamic standards, then why were non-Muslims so prominent within their inner circles of power? Hans Putmann, in his L'Eglise et L'Islam sous Timothée I (780–823), provides the first part of the answer. He argues that, even though there was a desire to establish a purely Islamic state, non-Muslims maintained their positions during the early ʿAbbāsid period because of their education and ability to handle the affairs of the state (e.g. record keeping). During the transition from the Umayyads to the ʿAbbāsids, non-Muslims within the government were able to keep administrative stability within the state.Footnote ²¹ The second part of the answer is to do with the fact that the opinions of independent jurists had no formal hold on the caliph's policy: their desire to implement restrictive measures was not enough to rid the government of all dhimmīs from positions of power and authority.Footnote ²²

Al-Manṣūr (r. 754–775) was probably the first ʿAbbāsid caliph to attempt to remove dhimmīs from his administration. He decided to remove Christians from the office of public treasury, but was later compelled to return them to their positions owing to their expertise. Some Muslims complained to him about the Christian secretaries; he therefore wrote to his provincial governors asking them to expel the dhimmīs from government offices and replace them with Muslims. However, one of his close colleagues, Shābib ibn Shaybah, suggested he should take back his order fearing that those officers would find a way to avenge him. Instead, he advised the caliph to replace the dhimmīs with the Muslims gradually.Footnote ²³ Later, rather than dismissing non-Muslims from public office, al-Manṣūr himself appointed a Jew, named Mūsā, as one of the two collectors of revenue.Footnote ²⁴

Al-Mutawakkil was perhaps the most dismissive in ordering his officials in the cities and provinces of the empire not to employ dhimmīs. In an edict of 850, he reiterated the ban on dhimmīs in state service previously pronounced by the Umayyad caliph ʿUmar ibn ʿAbd al-ʿAzīz (r. 717–720). That edict opened with the preamble: “It has become known to the Commander of the Faithful that men without judgment and discernment are seeking the help of dhimmīs in their work, adopting them as confidants in preference to Muslims and giving them authority over Muslim subjects”.Footnote ²⁵ I believe that al-Mutawakkil's edict was connected with his wider programme to impose more orthodox beliefs on the Shia and the Mutazilites. It is telling that al-Mutawakkil himself is reported to have employed Christians, such as Dulayl ibn Yaʿqūb.Footnote ²⁶ He employed a Christian architect (muhandis masīḥī) to build his Jaʿfarī palace.Footnote ²⁷ Christian sources also confirm that a number of Christian clerks and physicians served in his court.Footnote ²⁸

Another ʿAbbāsid caliph who sought to remove non-Muslims from public office was al-Muqtadir (r. 908–932). He attempted to limit their employment to positions of court physicians and tax collectors. He also ordered the employment of non-Muslims in menial occupations as a mark of their inferiority, simultaneously prohibiting their riding on ordinary saddles.Footnote ²⁹ It is reported, however, that al-Muqtadir himself had four Christian secretaries.Footnote ³⁰ His own influential vizier Ibn al-Furāt (d. 924) was known for his close relationships with Christian subjects: it is said he had nine Christian secretaries, some of whom had such close access that they were always invited to eat with him.Footnote ³¹ The historian al-Hilāl al-Ṣābī (d. 1056) records an interesting event in his Tārīkh al-wuzarā' about the rivalry between two leading viziers under al-Muqtadir's reign: Ibn al-Furāt and ʿAlī ibn ʿĪsā (d. 946). After his second term, Ibn al-Furāt was brought to the court and interrogated by ʿAlī ibn ʿĪsā. The latter said: “Didn't you fear God for appointing a Christian to the office of the Muslim army and causing the defenders of religion to kiss his hand and obey his command?” Ibn al-Furāt responded: “This is something I have not made up or innovated with. Al-Nāṣir li-dīn Allāh [the epithet of the vizier Abū Muḥammad al-Ḥasan al-Bāzūrī] appointed a Christian named Isrā'īl to take charge of the army, and al-Muʿtaḍid bi-Llāh appointed a Christian named Mālik ibn al-Walīd as his secretary”. When ʿAlī ibn ʿĪsā confronted him and said that both viziers were wrong in doing so, he responded by saying that even the previous caliphs had had Christian secretaries.Footnote ³²

Here we can see how Ibn al-Furāt justified his decision to employ Christians as secretaries and, more importantly, as commanders in the army. He refers to the involvement of dhimmīs in political life in the early period of Islamic history. When he was blamed for placing a Christian in command of the army, he defended himself against the charge of impiety by pleading the example of previous caliphs who had given office to Christians. So, it is true that some caliphs decided to exclude dhimmīs from public office, but in reality they were never really able to get rid of them, instead continuing to depend on their skills and competencies to run the government. Evidently the reason why the caliphs, from al-Manṣūr to al-Mutawakkil and al-Muqtadir, again and again made such pronouncements is in part because the attempt to remove non-Muslims from secretariat, public treasury and other state services did not meet with much success. As Antoine Fattal notes, “The implementation of these restrictive measures was practically zero”.Footnote ³³ In addition, I would argue, the opposition to dhimmī officials can be understood as a response to the pervasive phenomenon of non-Muslim involvement in the political life of medieval Islam. As a reaction to non-Muslims in influential positions, many Muslims complained to the caliphs that they were being ruled, in their own empire, by non-Muslims.

The first ʿAbbāsid caliph to employ a dhimmī vizier was al-Muʿtaṣim (r. 833–842). He was known for favouritism in appointing a large number of Christians to public office. There were two brothers, both Christians, of very high standing in the confidence of the caliph: one, Salmuyah, occupied the position of secretary of state, and no royal documents were valid until he had countersigned them. His brother, Ibrāhīm, was entrusted with the care of the privy seal, and was set over the Bayt al-Māl or public treasury, an office that, from the nature of the funds and their disposition, might have been expected to have been put into the hands of a Muslim. Al-Muʿtaṣim's relationship with Ibrāhīm was so close that the caliph “visited him in his sickness, and was overwhelmed with grief at his death, and on the day of the funeral ordered the body to be brought to the palace and the Christian rites performed there with great solemnity”.Footnote ³⁴

Al-Muʿtaṣim too had a Christian vizier, named Faḍl ibn Marwān ibn Māsarjis (d. 865). According to biographer Ibn Khāllikān (d. 1282), al-Muʿtaṣim invested him with the dignity of vizier, entrusting to him at the same time the administration of all his affairs.Footnote ³⁵ At first, Faḍl ibn Marwān joined one of al-Muʿtaṣim's secretaries called Yahyā al-Jurmuqānī, and then replaced him. He dealt with al-Muʿtaṣim's business affairs and wrote down, in Muʿtaṣim's name, what ever he himself desired, until al-Muʿtaṣim became caliph. From then on, Faḍl became the effective director of the caliphate: all government departments came under his control, and he stored up all the taxes collected.Footnote ³⁶ His power was unlimited, and he therefore acted as if he had more power than the caliph himself. At an audience which he held one day for the dispatch of public business, he noticed, among the memorials presented to him, a paper on which were inscribed these lines:

According to Dominique Sourdel, Faḍl ibn Marwān was the first example of the many secretaries of Iraqi Christian origin to serve the caliphs during the ninth century.Footnote ³⁸ It seems that Faḍl ibn Marwān's power became so extensive that al-Māwardī calls him “a vizier of delegation”. Indeed, it was due to his excessive power that the relationship between caliph al-Muʿtaṣim and Faḍl ibn Marwān came to an end. The historian Abū Jaʿfar al-Ṭabarī (d. 923) provides a lengthy discussion of the reason behind al-Muʿtaṣim's anger against Faḍl ibn Marwān to the extent that he finally imprisoned him and seized his wealth. Al-Ṭabarī relates that on several occasions, Faḍl ibn Marwān refused to do what al-Muʿtaṣim had ordered, until the latter was reminded by his close friend Ibrāhīm al-Ḥaftī: “Do you really think that you have now attained success? In reality, you have nothing of the caliphate except the mere name! By God, your commands do not go farther than your own ears! The real caliph is Faḍl ibn Marwān, who issues commands that are immediately put into execution”.Footnote ³⁹ Al-Muʿtaṣim was shocked, and he asked: “What command of mine has not been carried out?” Al-Ḥaftī responded: “Two months ago you commanded that I should be given such and such, but since that time he has not given me even a single grain of what you commanded”.Footnote ⁴⁰ Ibn Abī Dāwūd also testified that Faḍl ibn Marwān often turned down al-Muʿtaṣim's requests for a certain amount of money to spend. Al-Muʿtaṣim was so angry that he finally replaced him as his vizier with Muḥammad ibn Abd al-Mālik.

Under the Fātimid dynasty in Egypt, there was more than the usual degree of favour for non-Muslims. There are no known parallel examples of employment of non-Muslim viziers among contemporary Muslim regimes. When the Ismaili Fātimids came to power in Egypt, many Jews and Christians were employed in the administration, although not to the rank of the viziers. This development is understandable because non-Muslims were regarded as less of a threat to their heterodox teaching than were the orthodox Muslims with whom they had to deal. Dhimmīs were evidently regarded by most of the populace as being in league with Shiis. The Fatimid caliph al-Muʿizz (r. 953–975) had a Christian wife. Among his brothers-in-law were the Melchite (Greek Catholic) patriarchs of Alexandria and Jerusalem.Footnote ⁴¹ It was said that at his court nothing could be done without the help of some or other Jew. During the Fātimid era, non-Muslims were able to attain the highest office of the vizierate. It is reported that the Fātimid caliph Al-ʿAzīz bi-Llāh (r. 975–996) had two non-Muslim viziers: one a Christian named ʿĪsā ibn Nestorius, the other a Jew named Manasseh.Footnote ⁴² In all there were seven Christian and three Jewish viziers in the Fātimid period.Footnote ⁴³

Two viziers in particular of the Fātimid dynasty deserve more attention because of the nature of the power vested in them. The first is Yaʿqūb ibn Kīllis (d. 991) of Iraqi Jewish origin, who was the first vizier appointed by al-ʿAzīz; and the second is Bahram al-Armānī (d. 1140) of Armenian Christian origin, the vizier of al-Ḥāfiẓ (r. 1130–49). Ibn Kīllis is said to have been “the architect of Fātimid administration”.Footnote ⁴⁴ He first served as head of the bureau under caliph al-Muʿizz. Soon after his master passed away in 978, he was appointed by al-ʿAzīz, the son of al-Muʿizz, and honoured with the title “the most illustrious vizier (al-wazīr al-ajall).Footnote ⁴⁵ Ibn Kīllis’ power as a vizier was unlimited because of his personal ties as trusted adviser and friend of al-ʿAzīz.Footnote ⁴⁶ However, scholars differ on the nature of his conversion to Islam. Walter J. Fischel, as cited by Yaacov Lev, emphasizes the Jewish origin of Ibn Kīllis.Footnote ⁴⁷ Lev finds this view problematic because “it is impossible to overlook his total absorption not only in the Fātimid causes as a political movement but also with Ismailism as a creed”.Footnote ⁴⁸ However, there were non-Muslim viziers such as Faḍl ibn Marwān who were deeply devoted to the ʿAbbāsid cause. I would argue that the cause of controversy lies in the unsettled description of Ibn Kīllis' religious identity in the early sources. Ibn Khāllikān, for instance, appears confused about Ibn Kīllis' conversion to Islam. He notes that Ibn Kīllis “died in his [original] faith (māta ʿalā dīnihī) and his adherence to Islam was only outward show”, yet he also says, “He professed Islam, and it was true that he converted to it”.Footnote ⁴⁹ Whether or not he was truly a Muslim is a matter for debate, but it is clear that he frequently joined the Jewish community and helped many Jews attain positions in the Fātimid bureaucracy.Footnote ⁵⁰

Most scholars of the Fātimids agree that Ibn Kīllis' power as a vizier was such that, as in the case of Faḍl ibn Marwān during al-Muʿtaṣim's reign, many people thought that it was he who was in fact the ruler, not al-ʿAzīz. He often obstructed the judicial authority of ʿAlī ibn al-Nuʿmān, the chief qāḍī of the Fātimid caliphate. Not only had the chief qāḍī's sentences been subjected to his approval, but he also appointed a qāḍī of his own, independent of al-Nuʿmān. Ibn Kīllis is also reported to have controlled great military manpower of his own.Footnote ⁵¹ When Ibn Kīllis died in the year 991, al-ʿAzīz grieved for him very deeply and closed the dīwān for several days in his honour. He subsequently appointed as vizier a Christian called ʿĪsā ibn Nestorius.Footnote ⁵²

Another Fātimid vizier meriting attention is Bahram al-Armānī. He was “the only Christian vizier who ever received such titles as sayf al-Islām (sword of Islam), tāj al-dawla (crown of the state), and tāj al-khilāfa (crown of caliphate)”.Footnote ⁵³ For a non-Muslim and an alien of sorts to be given such titles was indicative of the caliph's desperate state of affairs. According to some sources, Bahram was not freely appointed by al-Ḥāfiẓ, but rather nominated by the army. He was a military man who, in becoming vizier, combined two jobs, giving him great power.Footnote ⁵⁴ We should keep in mind that Bahram came to power thirty-five years before the end of the Fātimid era, and therefore he had to face different kinds of discontent. He had full control over the realm and subdued every revolt against the state. In the eyes of caliph al-Ḥāfiẓ, however, the vizierate of Bahram had many positive aspects. A Christian Armenian could not hope to supplant the Fātimid caliphate, or to undermine the Ismaili character of the state. Of course, this reliance on Christian Armenians had its price; it contributed further towards the waning of the religious aura of the Fātimid regime.

If we use the categorization of the vizierate proposed by al-Māwardī, both Ibn Kīllis and Bahram al-Armānī can be classified as viziers of delegation (wuzarā' al-tafwīḍ). It is, therefore, difficult to accept P. J. Vatikiotis' assertion that “the function of minister in the early Fātimid caliphate was insignificant for doctrinal reasons primarily”. He mentions Yaʿqūb ibn Kīllis as only an “executive assistant of caliph al-ʿAzīz”.Footnote ⁵⁵ This is certainly contrary to the historical evidence of the Fātimid practices. Rather than being an “executive assistant of al-ʿAzīz”, Ibn Kīllis enjoyed unlimited powers, for he had the initiative in all affairs of government. In many instances, his word overruled that of the caliph himself. Perhaps, it was viziers like Ibn Kīllis who prompted the Ismaili theologian Ḥamīd al-Dīn al-Kirmānī (d. 1021) to limit the political power of viziers within the Fātimid bureaucratic system. Al-Kirmānī says: “There are no delegated powers or offices in the state, only executive functions subordinated to the imām's absolute theocratic authority”.Footnote ⁵⁶ As a devoted theologian, he did not, for obvious reasons, want to see the power of the imām being delegated, and eventually result in a weakened position for the imām. This would be similar to the goal of al-Māwardī who tried to restore the public duties of the imām, and such important duties should not be delegated to his non-Muslim viziers. Al-Māwardī attempted, by the device of two forms of vizierate, to keep within the framework of Islamic government officials and other holders of power who, in exercise of this, neglected the caliph.Footnote ⁵⁷

Some possible explanations

The question remains: why were non-Muslims needed to hold such important offices as vizier and secretary? No single answer will suffice. The circumstances under which non-Muslims were appointed as viziers varied from one case to another. However, we can look for common features surrounding the employment of non-Muslim viziers during ʿAbbāsid times that can help us understand why they were needed in the first place to fulfil the posts of the vizierate. It seems to me that the ʿAbbāsids recognized the need to maintain those dhimmīs already in the administration. For practical reasons, it would be difficult for the caliphs to allow Arab Muslims to run a vast empire with a complex system of administration. Because the existing administrative systems were kept in place along with their existing languages, there was a practical need for those who were fluent and capable of keeping records in those languages. In the former Byzantine provinces, for example, there was a need for personnel who could work in Greek, in the former Sassanid provinces there was a need for those fluent in Persian. In other words, because the Arabs ruled as a minority community over established societies with their own administrative systems, they had to establish methods to maintain order through the existing bureaucracy. Non-Muslims were thus in demand as professional state administrators who often rose to influential and important position.Footnote ⁵⁸

To make sense of this explanation, we shall consider closely how the institution of vizierate emerged in the early ʿAbbāsid period. Scholars agree that there was no actual vizierate during the Umayyad period. This institution was not known until the triumph of the ʿAbbāsids. Early Western scholarship seems to emphasize the Persian origin of this vizierate. As early as the second half of the nineteenth century, one encounters scholars such as Alfred von Kremer who argue that “this [wizāra] office appears to be of Persian origin, and came into existence for the first time under the ʿAbbāsids”.Footnote ⁵⁹ Franz Babinger takes it for granted that both the word and the institution were borrowed from the Sassanian empire.Footnote ⁶⁰ Philip K. Hitti also speaks of “… the vizier (wazīr), whose office was of Persian origin”.Footnote ⁶¹ Michael G. Morony proposes an interesting suggestion that this earlier scholarship on the vizierate might have been influenced by “the extravagant claims of writers such as the Niẓām al-Mulk (d. 1092) who derived the Islamic office of wazīr from a Sasanian origin”.Footnote ⁶²

S. D. Goitein refutes the assumption that the vizierate was borrowed, saying that “the Persian origin of the vizierate is far from an established fact”.Footnote ⁶³ For Goitein, the ʿAbbāsid vizierate developed gradually and its origin could be traced back to the Arabian tribal custom according to which a tribal chief would entrust the education of his son to a slave or freedman of the family. The primary development in ʿAbbāsid times, Goitein argues, was the early caliphs' practice of putting their sons and heirs under the supervision of an experienced man of affairs, who was always a personal dependant. Here we can see that the caliph, while retaining the system of family-rule brought from Arabia, initiated at the same time the rise of the vizierate by transferring an important feature of the household of the Arabic chieftain to the administration of the ʿAbbāsid caliphate. When, as a second development, the tutor to the heir apparent became vizier to the new caliph, the element of personal service remained fundamental to the character of the emerging institution, and this explains why its powers were never properly defined. It was only at the time of Harūn al-Rashīd (r. 786–809) that “the vizierate reached its apogee”.Footnote ⁶⁴ Based on this evolutionary approach, he concludes: “the vizierate was not borrowed by the Muslims as a fixed and well-defined institution from the Sassanians or anyone”, rather it developed gradually under the special historical circumstances at the end of the Umayyad and early ʿAbbāsid rules.Footnote ⁶⁵

In his extensive work on the ʿAbbāsid vizierate, Dominique Sourdel is sympathetic to Goitein's suggestion that the ʿAbbāsid vizierate must have developed gradually, and argues that it did not exist as a single, well-defined institution of government practically from the beginning of the ʿAbbāsid caliphate. However, he emphasizes the specifically administrative function of the vizierate, and therefore he is more willing than Goitein to see a connection with the Persian past.Footnote ⁶⁶ For Sourdel, the practice of attaching tutors to the caliphs' heir, although it may be of Arabian origin, was a secondary development intended to provide them in advance with the viziers they would need on becoming caliph. By looking closely at the professional secretarial background from which the viziers were mainly recruited, he concludes that the vizierate developed primarily as a response to the administrative needs of an increasingly sophisticated government. The fact that the ʿAbbāsid viziers, at least from the time of al-Manṣūr, belonged to the famous family of the Barmakids is seen by Sourdel as “the undeniable Iranianization of the ʿAbbāsid courts and administration”.Footnote ⁶⁷

It may be safe to say that, for the time being, the problem of the origin of the vizierate has not been solved. Even the early Muslim sources seem to advocate two different portraits of the early ʿAbbāsid vizierate. On the one hand, we have al-Ṭabarī's Tārīkh al-rusul wa al-mulūk, which provides the early development of the vizierate in somewhat evolutionary terms and shows continuity with the previous history of the term wazīr in the Islamic political context. On the other hand, we have al-Jahshiyārī's al-Wuzarā' wa al-kuttāb, which shows that the ʿAbbāsids adopted what was effectively a ready-made institution. Its early history is simply a succession of different viziers, with no institutional development. There is no sense of the ʿAbbāsids experimenting with an Islamic cultural resource. For Jahshiyārī (d. 942), the vizierate existed as a single, well-defined institution of government practically from the beginning of the ʿAbbāsid caliphate. R. A. Kimber is certainly correct when he asserts that while “Goitein's more evolutionary approach, with its emphasis on the vizierate to the heir apparent, is closer to al-Ṭabarī's version of events”, Sourdel's uniform approach is “no doubt influenced by al-Jahshiyārī”.Footnote ⁶⁸ It is worth noting that the picture of the vizierate that emerges from both al-Ṭabarī and al-Jahshiyārī is essentially administrative in nature, and this tends to weaken Goitein's argument for Arabian tribal origins.

Even if we accept that the vizierate developed gradually during the ʿAbbāsid period, as Goitein claims, we cannot deny the impact of the Sassanian tradition that largely shaped the function and practice of the vizierate. I would argue that the vizierate, whether it grew out of the office of the secretaries whose profession in later Umayyad courts had drawn much inspiration from the Sassanian practice and style, or was a reinvention of the early ʿAbbāsids, owed a great deal to Sassanian practice. The Persian presence is evident from the fact that the vizierate's beginning and period during which it flourished coincide with the rise of the Persian mawālī, their occupation of the vizierate for several successive generations, and the adoption of many Persian administrative measures. The Sassanian administrative system, which was regarded by the Arabs as the highest example of wise statecraft, served as the model.Footnote ⁶⁹ It is therefore understandable that Muslim jurists feel the need to provide Islamic guidelines so that this vizierate could still be framed within Islamic principles of government. This is also the reason why most, if not all, viziers and secretaries, were non-Arabs, either non-Muslims or converts who had been familiar with this imported tradition. Generally speaking, ʿAbbāsid viziers were recruited from the scribal class, or kuttāb, who were educated either in the Persian tradition or the Byzantine one, depending on which phase of the caliphate one examines.Footnote ⁷⁰

Now, why non-Muslims precisely? Because of their ability and experience non-Muslims were often kept as administrators, clerks, tax collectors, even as viziers and governors. Well-trained bureaucrats of Persian and Byzantine origin were in ever greater demand to administer an institution that was somehow alien to the Arab culture. Experts in finance and bureaucratic routine, they brought to their work the wisdom passed on by generations of experience. Christian subjects in particular had experience in the Byzantine administration, and their knowledge of several languages – Roman, Syriac, Persian and Armenian, in addition to Arabic – made it easier for them to benefit from the experience of other civilizations. In most cases the experiences and educational abilities of Christians made acceptance of their services a great benefit. Fātimid recruitment of viziers was no different. They found some dhimmīs to be highly qualified for the post of vizierate. As A. S. Tritton puts it, “The Fātimids attached great importance to the post of chief secretary, and chose their secretaries, both Muslims and dhimmīs, for their skill in writing”.Footnote ⁷¹

In his State and Society in Fātimid Egypt, Yaacov Lev argues that the persistence of Christian administrative personnel was the outcome of several factors. First, the agrarian administration was very complex and not easily mastered; Christians played an important role at the local level as well as at the central offices in the capital. Second, administrative knowledge was passed on by the officials to their families when fathers employed their sons, thus maintaining the family hold over the post. Finally, the large number of Christians in the administration also reflected the only partial Islamization of Egypt: this was the case prior to the Fātimids and did not change under their rule.Footnote ⁷² It can be argued that, like the ʿAbbāsids, the Fātimids were simply pragmatic. On the one hand, they were Ismaili Shiis facing a Sunni majority, and on the other, because of the social, economic, and demographic importance of Christians and Jews, the caliphs had to take care of them and employ them in government services.

We also find that some Muslim rulers were reluctant to employ Arab Muslims because of tribal attachments which made them prone to tribal rivalries. ʿUbaydullah ibn Ziyād (d. 686), the Umayyad governor of Iraq, said: “If I employed an Arab as tax collector and embezzled the land-tax and I punished him I risked antagonizing his tribe. If I fined him, deducting the fine from the pension of his clan, I did him harm. If I dropped the matter I would be wasting God's money… . I found therefore the dihqāns (Persian gentry) were more knowledgeable about tax-collection, more honest with their trust and and easier for me to call to account”.Footnote ⁷³ The ʿAbbāsid caliph al-Muʿtaḍid (r. 892–902) also expressed strong confidence in Christians and preferred them to others. During his reign, the governor of Anbar, ʿUmar ibn Yūsuf, was a Christian, and the caliph approved of the appointment on the ground that Christians were found to be competent.Footnote ⁷⁴ It is related that his vizier, ʿUbaydullah ibn Sulaymān, was reluctant to appoint Christians to public office. Al-Muʿtaḍid advised him as follows: “If you found a Christian suitable for any office, you should appoint him. A Christian is more truthful than the Jews because the Jews want the return of kingdom to them; he is also better than a Muslim as the latter is of the same religion as yours he desires to take over your position; he is also better than Zoroastrians because the kingdom is in their hand.”Footnote ⁷⁵

Concluding remarks

This study has sought to establish three main points. First, the concept of the vizierate developed by both Sunni and Ismaili scholars had not been fully implemented by the caliphs in the way they appointed their viziers. While the jurists either prohibited the employment of non-Muslim viziers or limited their power merely to executing the caliphs' commands, the caliphs delegated to their non-Muslim viziers the affairs of state administration (wizāra tafwīḍ) in their entirety. Instead of being bound by Islamic law, the caliphs made their decisions regarding the state administrative procedures based on contextual practicality, that is, what worked in a given area. Second, there is a disjunction between the legal conceptualization of the institution of the vizierate developed by the ʿulama and its implementation during ʿAbbāsid times. We have discussed the case of the ʿAbbāsid vizier Faḍl ibn Marwān and the Fātimid viziers Ibn Kīllis and Bahram al-Armānī to show the delicate interplay between theory and practice. Third, the employment of non-Arab viziers (either non-Muslim or converts) supports the idea that this institution of the vizierate was not known to Islam in its early history. It was not until the triumph of the ʿAbbāsids that the vizierate gradually became an integral part of the state administration through the impact of Persian culture. It was precisely because this institution was somewhat alien to the Arab culture the caliphs were compelled to consider non-Muslims who had the experience and skills to fulfil this important post.

This study also has significant implications for the discussion of the nature of state–religion relations in early Islam. The fact that some caliphs delegated their administrative powers to non-Muslim viziers confirms earlier findings about the separation of state and religion in early Islamic society. It is inconceivable that dhimmī viziers would govern and administer the day-to-day affairs of state in accordance with the sharīʿa. Scholars such as Ira M. Lapidus, Patricia Crone and others have argued that there had been a sharp conflict between the ʿulama and the ʿAbbāsid caliphs, and that the ʿulama would have nothing to do with the state, which eventually led to the separation of religion and state.Footnote ⁷⁶ The argument emphasizing the disharmonious relation between the ʿulama and the caliph comes under attack by Muḥammad Qasim Zaman who argues for “the deep involvement of the early ʿAbbāsids in the religious life of the times”.Footnote ⁷⁷ Yet we have seen that, by delegating their administrative power to their non-Muslim viziers, the caliphs failed to integrate such an institution into some kind of “Islamic” framework as defined by Muslim jurists. This was also the case with the Fātimids, who were the product of the Ismaili daʿwa. The Fātimids practised a form of separation of state and religion whereby government institutions, although parallel in organization to the daʿwa, functioned independently of it. At the same time the evidence presented in this study shows that the caliphs had lost their political power earlier than Ira M. Lapidus has suggested, that is, from the middle of the tenth century. This study demonstrates that during al-Muʿtaṣim's reign in the first half of the ninth century effective control of the Arab–Muslim empire had passed into the hand of Christian vizier Faḍl ibn Marwān.

This study also shows the extent to which non-Muslim dhimmīs were embedded in the early Muslim state.Footnote ⁷⁸ What makes this sense of dhimmī embeddedness in Muslim society possible is that early Muslim/non-Muslim relations were not as hostile as is sometimes supposed. In most cases, the Arab rulers did not interfere with the internal civil and religious administration of dhimmīs. This also explains why the peoples of the conquered provinces did not confine themselves simply to accepting the new regime, but in some cases actively assisted in its establishment. As Bernard Lewis notes, “The change from Byzantine to Arab rule seems to have been gradually welcomed by the subject peoples, who found the new yoke far lighter than the old, both in taxation and in other matters”.Footnote ⁷⁹ It is now more fashionable to argue that the Islam of earlier times seems to be more tolerant than that of later times, because in earlier periods Muslims were ready to learn and adopt anything valuable from the experiences of other communities.