Non-Muslims in the Umayyad Army (Prof. Donner)

Article

1. Introduction: The Muslim Jurists on the Participation of Non-Muslims in the Muslim Army
2. The corpus of Prophetic traditions deals with two main issues: did the Prophet actually allow non-Muslims to fight in his campaigns; and if so, did he compensate them, and in what manner? There are two contradictory answers to the first question in this corpus, one showing that the Prophet’s sunna was not to permit non-Muslims to participate in his expeditions, and one showing the opposite. Regarding non-participation, five traditions are cited:
   • a. ʿĀʾishaʾs ḥadīth from ʿUrwa: in its longest form, it says that when the Prophet headed to Badr, he was met at a named place by a man known for his courage and resourcefulness, and that made the Companions rejoice. The man told the Prophet he wanted to follow him and gain [spoils] with him. The Prophet asked him whether he believed in God and His Messenger. The man said no, whereupon the Prophet sent him back, saying: “I shall not seek the assistance of a polytheist” (fa-lan astaʿīna bi-mushrik) (there’s variants of this hadith, see Yarbrough, “I’ll not Accept Aid from a mušrik”). The man then met the Prophet at another named place, and the same dialogue went on between them. When they met at a third named place, the man said he believed in God and His Messenger, whereupon the Prophet told him: “Go ahead!” (fa-inṭaliq) (Muslim b. alḤajjāj, Ṣaḥīḥ, vol. 3, pp. 1449–50 (“al-jihād wa-lsiyar,” 1870); Saḥnūn, Mudawwana, vol. 2, p. 617. It occurs in shorter forms in Ibn Abī Shayba, Muṣannaf, vol. 6, p. 487 (“al-jihād,” 33162); al-Tirmidhī, Sunan, vol. 3, pp. 58–59 (“al-siyar,” 1601); Abū Dāwūd, Sunan, vol. 3, p. 173 (“al-jihād,” 2732); Ibn Ḥazm, Muḥallā, vol. 7, p. 335. See also al-Shāfiʿī, Umm, vol. 4, p. 261).
   (edited)
   • b. The ḥadīth from Saʿīd b. al-Mundhir says that, on his way out to Uḥud, the Prophet looked back when he was at a named place and saw a fine squadron (katība ḥasnāʾ). He asked about its people and was told they were ʿAbd Allāh b. Ubayy b. Salūl and his clients of the Jews. He asked whether they had converted to Islam and was told they had not. He said: “We do not seek the assistance of unbelievers [in fighting] against polytheists” (fa-innā lā nastaʿīnu bi-l-kuffār ʿalā al-mushrikīn) (Ibn Abī Shayba, Muṣannaf, vol. 6, p. 487).
   • c. The ḥadīth from al-Zuhrī says that the Anṣār asked the Prophet at Uḥud whether to seek the assistance of their allies from the Jews. The Prophet said: “We do not need them” (lā ḥājata lanā fīhim) (Saḥnūn, Mudawwana, vol. 2, pp. 617–18).
   • d. The ḥadīth of Khubayb b. Yasāf, on the authority of his grandson, says that the Prophet went out toward an unnamed place (yurīdu wajhan) and the narrator, Khubayb, and another man from his people came to him and said: [Is it conceivable] that our people should attend a battle (mashhadan) while we do not? The Prophet asked them whether they had become Muslims (aslamtumā), and they said they had not. He thus said: “We do not seek the assistance of polytheists [in fighting] against polytheists (fa-innā lā nastaʿīnu bi-l-mushrikīn ʿalā al-mushrikīn). The narrator then said that Khubayb and his fellow tribesman later converted and participated in battle with the Prophet (Ibn Abī Shayba, Muṣannaf, vol. 6, p. 487).
   • e. A related ḥadīth from Abū Hurayra says that the Prophet said, in a longer tradition: “Booty has not been made licit for anyone before us” (fa-lam taḥilla al-ghanāʾim liaḥadin min qablinā). This has been interpreted to mean that no one other than Mulsims are allowed to take booty (Muslim b. al-Ḥajjāj, Ṣaḥīḥ, vol. 3, p. 1367 (“al-jihād wa-al-siyar,” 1747); Ibn Ḥazm, Muḥallā, vol. 7, p. 335).

1. These proof texts were convincing for two major scholars, Saḥnūn and Ibn Ḥazm, and thus they represent the positions of the Mālikī and Ẓāhirī schools. The ḥadīth corpus that shows the Prophet accepting the participation of non-Muslims with Muslims in battle consists again of five sunnas:
   • a. The ḥadīth from al-Zuhrī, narrated via four different channels, says that some Jews used to raid with the Prophet, and he would give them and the Muslims the same shares of the booty (kāna yahūd yaghzūna maʿa al-nabī fa-yushim lahum ka-sihām al-muslimīn) (ʿAbd al-Razzāq, Muṣannaf, vol. 5, pp. 188–89; Ibn Abī Shayba, Muṣannaf, vol. 6, p. 487; al-Tirmidhī, Sunan, vol. 3, p. 59; Ibn Ḥazm, al-Muḥallā, vol. 7, p. 334. See also al-Bayhaqī, al-Sunan al-kubrā, vol. 9, pp. 53, 92).
   • b. The ḥadīth from al-Wāqidī says that the Prophet took along with him to Khaybar ten Jews from Medina and gave them and the Muslims the same shares of the booty.9 This possibly deals with the same occasion spelled out less clearly in the ḥadīth from Abū Mūsā al-Ashʿarī, which says: “I came with a party of the Ashʿarīn to the Prophet at Khaybar and he gave us shares like those who conquered it” (Al-Tirmidhī, Sunan, vol. 3, p. 59).
   • c. The ḥadīth from Ibn ʿAbbās that the Prophet sought the assistance of the Banū Qaynuqāʿ (of the Jews) in an unidentified battle; he gave them a little but did not give them shares (i.e., as he did to the Muslims; istaʿāna rasūl Allāh bi-yahūd Qaynuqāʿ faraḍakha lahum wa-lam yushim) (Abū Yūsuf, Radd, p. 40; al-Bayhaqī, al-Sunan alṣughrā, vol. 3, p. 393).
   • d. It is narrated that Ṣafwān b. Umayya participated in the battle of Ḥunayn after the conquest of Mecca while he was still a polytheist (Al-Shāfiʿī, Umm, vol. 4, pp. 167, 261. This is confirmed by the sīra and the Companions’ tradition; see Guillaume, trans., The Life of Muhammad, pp. 567, 569; Ibn Ḥajar al-ʿAsqalānī, al-Iṣāba fī tamyīz al-ṣaḥāba, vol. 2, p. 187).
   • e. A related ḥadīth from Abū Hurayra says that the Prophet and Abū Bakr hired a man from the tribe of al-Diʾl as a skilled guide (khirrīt) when that man was still a polytheist. They gave him their two camels and made an appointment with him to meet them at the Thawr cave (ghār) three days later. He did, and went with them, together with another man, along the coastal road (Al-Bukhārī, Ṣaḥīḥ, vol. 3, pp. 181–82; Ibn Ḥazm, Muḥallā, vol. 7, p. 335).

This body of sunnas convinced all of the early jurists, except for those mentioned above, that the Prophet condoned the participation of polytheists/unbelievers in his campaigns. This was the position of al-Shaʿbī, Qatāda, Abū Ḥanīfa, al-Awzāʿī, Sufyān al-Thawrī, Abū Yūsuf, Muḥammad b. al-Ḥasan al-Shaybānī, al-Shāfiʿī (Ibn Ḥazm, Muḥallā, vol. 7, p. 334. For specific jurists, see Abū Yūsuf, Radd, p. 39; idem, Kharāj, p. 391; al-Shaybānī, Siyar, with the commentary of al-Sarakhsī, vol. 2, pp. 680–82; vol. 3, pp. 995–1000; al-Shāfiʿī, Umm, vol. 4, pp. 166–67, 261, 269–70; Abū ʿUbayd, Kharāj, p. 431), Abū ʿUbayd al-Qāsim b. Sallām, and probably Mālik personally (Saḥnūn, Mudawwana, vol. 2, p. 617). Even the jurists who were against this participation could not but see that there was a point to it. Thus, Saḥnūn had the following caveat: “unless they were sailors (nawātiyya) or servants (khadam); there I find no objection.” And Ibn Ḥazm dwelt at length on the Prophetic ḥadīths favoring participation, finding them compelling (Ibn Ḥazm, Muḥallā, vol. 7, pp. 334–35), allowing for the hiring of guides (Ibn Ḥazm, Muḥallā, vol. 7, p. 335), and giving the impression that he would have accepted them had they not violated his strict rules about which ḥadīth qualifies as valid. As for the discrepancy between the two sets of Prophetic ḥadīths, one prohibiting and one permitting the participation of non-Muslims with the Muslims in battle, only al-Shāfiʿī seemed to care to comment on it, attributing it to abrogation (naskh): the prohibition reports came only from the very early period (Badr, Uḥud), whereas the permission ones came from thereafter (Al-Shāfiʿī, Umm, vol. 4, p. 261).

In general, though, the jurists favoring the participation of non-Muslims in the Muslim armies based their legal opinion not solely on the Prophet’s sunna but also on the practice of the early leading Muslims. Here also we find a large body of reports, all of which speak in favor of participation. Indeed, one of the things that seemed to sway Ibn Ḥazm almost to the point of accepting the participation of non-Muslims is this corpus; see his Muḥallā, vol. 7, p. 334, particularly his comment on al-Shaʿbī and Saʿd b. Mālik. These “leading Muslims,” as I have called them, were either commanders in battle or “people in charge,” meaning either commanders or caliphs.

1. • a. It is related that Saʿd b. Mālik (= Saʿd b. Abī Waqqāṣ) led a campaign in which there were Jews. After the battle, he gave them a little of the booty (raḍakha lahum).20 This report elicited the following comment from Ibn Ḥazm: “We know of no one of the Companions who went against Saʿd in this [matter].” (Ibn Ḥazm, Muḥallā, vol. 7, p. 334).
   • b. It is related that Salmān b. Rabīʿa al-Bāhilī, one of the conquerors of Khurāsān, invaded Balanjar in one of his campaigns. There he sought the assistance of people from the polytheists to fight against the polytheists, saying: “Let the enemies of God fight the enemies of God (li-yaḥmil aʿdāʾ Allāh ʿalā aʿdāʾ Allāh).” (Ibn Abī Shayba, Muṣannaf, vol. 2, p. 487; Ibn Ḥazm, Muḥallā, vol. 7, p. 334).
   • c. It is narrated on the authority of al-Awzāʿī that the basis for his position is not only that the Prophet gave shares (of the booty) to those who fought with him from the Jews, but also that the people in charge of the Muslims (wulāt al-muslimīn) after the Prophet gave shares to the Jews and Magians (majūs) whose assistance they sought against their enemies (Abū Yūsuf, Radd, p. 39).
   • d. Finally, and perhaps most importantly, we have a saying of al-Shaʿbī’s which is broader and more detailed than the previous reports. In its most expanded form, it says that Jābir [b. ʿAbd Allāh al-Anṣārī] asked al-Shaʿbī about his opinion regarding the Muslims leading a campaign in which there were People of the Book. He said: “The imāms whom I have witnessed (adraktu al-aʾimmata), those knowledgeable in the law and those not knowledgeable in it (al-faqīh minhum wa-ghayr al-faqīh), conducted campaigns in which there were ahl al-dhimma. [After the battle,] the imāms would give them a portion [of the booty] (yaqsimūna lahum), and would remove their tribute (jizya); this is a good spoil (nafl) for them.”24 Qatāda is also reported to have transmitted a similar saying.25 Though he disagreed with the legal position ensuing from this report, Ibn Ḥazm could not but evaluate it positively; “al-Shaʿbī was born at the beginning of ʿAlī’s days and he lived long enough to witness Companions after ʿAlī” (Ibn Ḥazm, Muḥallā, vol. 7, p. 334).

The Christian Arab Tribes

It is important to state right from the start that one must be careful when one approaches the historical reports that deal with Christian Arab tribes and not assume that the discussion is always about Christians when mention is made of the Taghlib, Iyād, al-Namir b. Qāsiṭ, or other Arab tribes who were known to have been Christian at the beginning of the conquests, and who converted only very slowly to Islam later on. We can also judge by context that the Arabs being discussed were Christians. Examples of such contexts are when the sources mention that the Muslims requested the jizya from these Arabs (see al-Ṭabarī, Tārīkh, ed. Ibrāhīm, vol. 3, p. 464 / ed. de Goeje, vol. I, pp. 2044–45), or when they identify a time and place for a battle in which Islam did not exist prior to the conquest of that place at that time (Donner, Early Islamic Conquests, p. 198, and p. 335 n. 152), or when they clarify the Christianity of the participating Arabs in a certain battle when discussing the subsequent battle that resulted from the first ((Tārīkh, vol. 3, pp. 353–54/vol. I, pp. 2029–31), or when they clarify the Christianity of the participating Arabs in a certain battle when discussing the subsequent battle that resulted from the first.

It is well known that the Christian Arab tribes in Syria, Iraq, and northern Mesopotamia (al-Jazīra) played an important part in the early conquests, sometimes siding with the Byzantines or the Sasanians, and others with the Muslims before they converted to Islam. It is also well known that, already during the caliphate of ʿUmar b. al-Khaṭṭāb (r. 13–23/634–644), the Christian Arabs, in particular the Taghlib, were granted a special tax status conducive to winning them over to the side of the Muslims. The Islamic sources are clear that the Muslims observed Christian Arab tribes fighting alongside the Byzantine (al-Ṭabarī, Tārīkh, vol. 3, p. 410/vol. I, 2114–15), and the Sasanians, and that they fought them as enemies, and that they sometimes dealt harshly with them (al-Ṭabarī, Tārīkh, vol. 3, pp. 376–77/vol. I, 2062–64). We see this in al-Ṭabarī’s main reports on the battles of al-Walaja and Ullays in southern Iraq in 12/633; the Muslims were then fighting under the leadership of Khālid b. al-Walīd (al-Ṭabarī, Tārīkh, vol. 3, pp. 353–58/vol. I, 2029–36).

According to the Islamic sources, it was during these raids that the Muslims began to recruit guides and spies from the local Arab Christians in order to assist them in their wars in lands unknown to them. That some of the recruitment was undertaken intentionally is clear from al-Ṭabarī’s saying that, after the battle of al-Thiny in 12/633, Khālid b. al-Walīd put Suwayd b. Muqarrin in charge of taxation and recruiting spies (wa-aqāma li-ʿaduwwihi yatajassas al-akhbār) (Al-Ṭabarī, Tārīkh, vol. 3, p. 352/vol. I, p. 2029). In the same year, Khālid b. al-Walīd built into the peace agreement with the chiefs of al-Ḥīra a condition that the Ḥīrites “be spies for him” (ʿalā an yakūnū lahu ʿuyūnan). A man from the Taghlib, presumably a Christian, who was taken prisoner in one of the early raids of Iraq, is reported to have bought his freedom from the Muslims by volunteering to guide them to a place where a group of the (presumably Christian) Rabīʿa resided; he did so and got his freedom (Al-Balādhurī, Futūḥ al-buldān, p. 248).

1. Indeed, several of the accounts we have in the sources about the Muslims’ raids in Iraq in the early phase of the conquests are reported as having been propelled by volunteering almost enthusiastic local Arab (presumably Christian) guides, for no immediate reward. The accounts of the battle of al-Khanāfis in 13/634 are filled with such information. A report has it that after al-Muthannā b. Ḥāritha was through with the battle of al-Buwayb, two presumably Christian men, one from al-Anbār and the other from al-Ḥīra, importuned him, each desiring to guide him to rich marketplaces, the first to al-Khanāfis and the second to Baghdād [= Ctesiphon]. Al-Muthannā raided and plundered al-Khanāfis on its market day; it was guarded by two cavalry units from the Rabīʿa and the Quḍāʿa, both of whom seem to have been Christian, since the name of the Quḍāʿī leader is not Arabic (Rūmānūs b. Wabara). The man from al-Ḥīra guided al-Muthannā to a village frequented by the merchants of Ctesiphon and other cities and in which a great deal of money and goods exchanged hands. The guide then gave him detailed instructions as to how to proceed toward al-Anbār. There he should seek the assistance of some dihqāns in assigning guides (adillāʾ). Al-Muthannā did according to the information provided by the Ḥīrite and took guides who went with him on the last leg of five or six parasangs until he reached his destination, where he gained a great deal of valuable spoils. He returned and with him the guides, across deserts and canals, until they arrived in al-Anbār, where the dihqāns received them with hospitality and joy.
2. But did the Christian Arabs, while still Christian, fight with the Muslims in the latters’ campaigns against Byzantium other than from their own lands in Syria, Iraq, and Mesopotamia? The answer is uncertain, for we hear of only one, rather unusual, such case; it concerns a poet who went out with the expedition of Maslama b. ʿAbd al-Malik to al-Ṭiwāna in 85/704. This is the Taghlibī al-Nuʿmān b. Najwān, better known as Aʿshā Banī Taghlib (Aʿshā Banī Taghlib, see Ibn al-ʿAdīm, Zubdat alḥalab fī tārīkh ḥalab, vol. 8, pp. 3619–20).

1. The Non-Arab Non-Muslims and the Muslim Army
2. a. Who Were Those Non-Arab Non-Muslims? The Islamic sources mention several non-Arab groups who, while non-Muslim, served in one way or another with the Muslim army. Since some of these groups eventually did convert to Islam at some point, the information about them could be hazy and perhaps difficult to ascertain. Starting with the east, one such group is the Asāwira, the formerly elite cavalry unit of the Sasanian army. One such group is the Asāwira, the formerly elite cavalry unit of the Sasanian army.68 We first hear about them in the conquest of al-Madāʾin in ca. 17/638, that is, two or three years after the battle of al-Qādisiyya;69 at that point, the sources have it that the caliph ʿUmar b. al-Khaṭṭāb wrote to the leaders of the conquests in Iraq to “seek the assistance of whomever they needed from the Asāwira,” and that they should “drop their tribute” (wa-yaḍaʿū ʿanhum al-jizāʾ).70 The mere mention of the jizya indicates that the Asāwira had not yet converted at the conquest of al-Madāʾin, and that at least some of them had fallen under Muslim rule. Did any of the Asāwira then fight with the Muslims in return for being exempted from paying the jizya? The sources are silent on this matter. Some years later, however, shortly before the conquest of Tustar, the Asāwira reappear, now offering Abū Mūsā al-Ashʿarī to convert to Islam and join his forces, which they did, participating in the siege of Tustar.71 One would think, then, that the Asāwira who fought alongside the Muslim army at the latest in 20/640 were Muslims. And yet, many years later, in 64/683, the leader of the Asāwira still has a Persian name: Māh Afrīdhūn.
3. In Syria, the Islamic sources are clear in indicating that two groups, the Jarājima (Mardaites) and the Anbāṭ, assisted the Muslim armies while remaining Christian.85 Another group identified by confession as an ally who collaborated with the Muslims was the Samaritans.86 There is a passing reference in the sources to Slavs (al-Ṣaqāliba) helping the conquering Muslims.87 Other unidentified local groups in northern Syria and Mesopotamia are also referred to.88 In the northeast frontier region, and more so in the east, the Islamic sources provide indirect evidence to the participation of all kinds of non-Arab and non-Muslim peoples in Muslim military activities.89 This can be gleaned from the peace agreements that the Muslims concluded in these regions, where one of the conditions of the agreements is that the locals should provide military assistance to the Muslims upon demand, sometimes in return for dropping the jizya required of them.90 The key text that is relevant here is the one describing the negotiations that took place between the “king” of al-Bāb and the commander of the Muslim forces, ʿAbd al-Raḥmān b. Rabīʿa. These ended with a peace agreement, sanctioned, it is reported, by no less than the caliph ʿUmar b. al-Khaṭṭāb; it stipulated that there be, among other things, potential military assistance to the Muslims from the people of al-Bāb.
4. The text then continues, “and it became common practice (sunna) regarding those fighting the enemy from the polytheists (al-mushrikīn): […] if they were called upon to fight (yustanfarū), the tribute (jizya) of that year would be removed from them.”91 Such was the case with the Armenians,92 the people of Jurjān and Dahistān,93 Ādharbayjān,94 and Marw al-Rūdh.95 Other conditions of cooperation were established also in the peace agreements of Mūqān,96 Iṣfahān,97 and al-Rayy.98 In the northeast, too, help to the Muslims is reported to have come from Soghdia; but here there is also a report about an individual,99 not a group, as in the case of an Armenian. Finally, another group that was mostly non-Muslim (and certainly non-Arab) and yet fought with the Muslims is the slaves (mamālīk; also ʿabīd),101 of which the Muslims must have captured enormous numbers during the conquests.
5. Islamization of the Berbers It began in the form of the personal zeal of some Muslim officials, then became a government policy, the critical events separating the two phases being the decisive victory of the Muslims over the Byzantines in the battle of Carthage in 73/692 and over the formidable Berber rebel al-Kāhina (the Sorceress) in ca. 82/701.114 ʿUqba b. Nāfi al-Fihrī, the governor of Ifrīqiya in 49–55/669–674 and 62–63/681–683, represents the first, personal phase. He is said to have established at least three mosques in various parts of the Maghrib, and to have left there “some of his associates to teach the Qurʾān and Islam, among them Shākir of the ribāṭ.” As a result, most of the Maṣmūda tribe of the Berbers in the farthest Maghrib converted voluntarily to Islam (aslamū ṭawʿan).115 After that, the sources start talking about more systematic efforts at Islamization,116 so that the year 85/704 is said to be the year in which “the conversion of the people of the farthest Maghrib was completed.”117 The Islamization efforts continued and took in 100/718 an even more official form, when the highest Islamic authority, the Umayyad caliph ʿUmar II, sent a group of Muslim religious scholars (whose names have been preserved) to Ifrīqiya and the Maghrib in what amounts to a missionary operation.118 The result was that the rest of the Berbers of Ifrīqiya converted.119 All these efforts, the sources say, led various Berber tribes of the Maghrib to convert to Islam.120 That did not prevent some Berber uprisings, some quite successful,121 including the “defection” of some Berber groups to Muslim sectarian Khārijism.
6. b. How Did Non-Arab Non-Muslims Come to Have a Relation with the Muslim Army? The Islamic sources suggest two ways. The first is the voluntary movement to the Muslims’ camp by various groups. This is how they portray the actions of the Asāwira, the Daylam, the Zuṭṭ, the Sayābija, and the Indighār, as we have seen. The second is related to the Islamic state’s policy to provide enticements for the non-Muslims to join the Muslims. This could be done on an individual basis, as we saw in ʿUmar b. al-Khaṭṭāb’s first policy statement about the Asāwira, namely to remove the tribute requirement from them if their aid was needed by the Muslims. It can also be seen in the case of the village of Arabissos: ʿUmar decreed that, in order to secure the cooperation of its population, “give them two ewes in the place of one, two cows in the place of one, and two things in the place of every other thing.”148 More frequently, though, the Islamic sources put those policy-related enticements in the form of formal peace agreements, treaties that specify what the contracting parties require of each other. We have seen above how this policy was enacted with regard to many districts in the northeast and east. But the sources also mention that such treaties were also concluded with the Jarājima149 and the Anbāṭ of Syria.150 These concessions given by the Muslims to the local peoples varied from place to place, although they always included safety and protection of life, children, and property; in some cases they included provisions and promise of military assistance, of non-enforcement of conversion, and of exemption from the jizya.
7. There is, perhaps, a third way in which the Muslim state tried to reinforce its army through the assistance of local populations, except that the Islamic sources do not link it with army activities. By that I mean the transfer of entire groups and populations and their settlement in frontier areas or coastal cities. The sources report several such resettlements. In 42/662, Muʿāwiya moved a group (qawm) of Persians from Baalbeck, Ḥimṣ, and Antioch to Tyre and Acre, of the coast of Jordan, very possibly for maritime defense purposes.152 Around the same year, he moved a group of the Asāwira from Baṣra and Kūfa and a group of Persians from Baalbek to Ḥimṣ and Antioch.153 In 49/669 or 50/670, he moved a group of the Zuṭṭ and Sayābija from Baṣra to Antioch.154 Al-Walīd I moved to Antioch a group of the Zuṭṭ and Sindīs who were brought to al-Ḥajjāj in Iraq by Muḥammad b. al-Qāsim, the conqueror of India; al-Ḥajjāj then sent them to al-Walīd.