² Shahrastānī, , Kitāb al-Milal wa-’l-Nihal, ed. Cureton, (London, 1846) (hereafter MN[SH]).Google Scholar

³ Most recently, the early theological epistles examined by van Ess in a number of publications and Cook in Early Muslim dogma (Cambridge, 1981).Google Scholar

⁴ Watt puts these problems clearly, and suggests some sensible rules of thumb, in The formative period of Islamic thought (Edinburgh, 1973), 1–6.Google Scholar

⁵ Makdisi, , ‘Ash‘arī and the Ash‘arites in Islamic religious history’ (Part 1), SI, 17, 1962, 41Google Scholar; and cf. Bernand, , ‘Le Kitāb al-Radd ‘Ala ‘1-Bida‘ d'Abu MutṭT’ Makḥul al-Nasafī’, Annales islamologiques, 16, 1980, 45Google Scholar.

⁶ While historical and textual concerns are sometimes combined successfully (e.g. Wadād, al- Qādī'sal-Kaisaniyyafi ’l-Ta’rikh wa-’l-Adab [Beirut, 1974])Google Scholar, it often happens that the critical end of things is given only perfunctory attention. The best example of what can be accomplished by focusing entirely on textual questions is Madelung, , ‘Bemerkungen zur imāmitischen Firaq- Literatur’, Der Islam, 42, 1967, 37–52Google Scholar [= Religious Schools and Sects in Medieval Islam, London, 1985, XV]Google Scholar.

⁷ I have explored some of the possibilities here in connexion with eastern Hanafite heresiography in my unpublished thesis, Studies in Islamic heresiography: the Khawārij in two Firaq traditions (Princeton, 1989)Google Scholar. There and here I use the term ‘standard’ to designate the Ash‘arite and Mu‘tazilite works commonly exploited by scholars.

⁸ On the tafarruq of the Khawārij see Wilkinson, , ‘The early development of the IbadI movement in Basra’, Studies in the first century of Islamic society (Southern Illinois University, 1982), 132Google Scholar; and Pampus, , Über die Rolle der Hdrigiya imfruhen Islam (Wiesbaden, 1980), 76–7Google Scholar.

⁹ El (2nd ed.), Azāriḳa, s.v..Google Scholar

¹⁰ Ed. Ritter, H., 2 vols., Istanbul (1929–1933).Google Scholar

¹¹ Ash‘arī does not even bother to cite the famous seventy-three sects ḥadīth, long the standard frontispiece for heresiographical works.

¹² Ritter, , ‘Philologika 111: Muhammedanische Häresiographen’, Der Islam, 18, 1929, 38.Google Scholar

¹³ Madelung, , ‘The Shī‘ite and Khārijite contribution to pre-Ash‘arite Kalam’, Islamic philosophical theology, ed. Morewedge, P. (Albany, N.Y., 1979), 127.Google Scholar

¹⁴ Karābisi's work was widely known as the primary source for information on the Khawārij and other early sects (mu‘awwal al-mutakallimīn ft ma‘rifat madhāhib al-khawārij wa-sā'ir ahl al-ahwā') (SubkYī, Ṭabaqāt al-Shāfi‘iyya [Beirut n.d.], 1:252.6; Baghdādī, Uṣul al-Dīn [Beirut, 1401], 208.15–. 16).

¹⁵ This sectioning of the text is entirely subjective, of course, and reflects my own reading of the way the account has been constructed.

¹⁶ Ash‘arī, 86.7-.9. Cf. the parallel in Balkhī's Maqālāt, cited in Nashwān, al-Ḥimyari, al-Ḥūr alin (Tehran, 1973), 177.20–178.1Google Scholar.

¹⁷ This is clear from the wording elsewhere in the tradition. Cf., for example, Baghdādī, , al-Farq Baina ‘I-Firaq (ed. Badr, M., Cairo, 1323/1905), 63.1–6Google Scholar, where the first Muḥakkima (in contrast to the Azāriqa) are said to have avoided the takfīr of quietists who otherwise share their views (idhā kāna ‘aid ra'yihim); and Qalhātī, , Kitāb al-Kashf wa-'l-Bayān, (Arabic MS, Br. Mus. Or. 2606), fol. 197a.7Google Scholar: tabarra'a [Nāfi‘] min al-qā‘id wa-law kāna ‘ārifan li-amrihi tābi‘an li-madhhabihi. (The published edition of this work is unavailable to me.) And the first-century Ibādī polemicist Sālim b. Dhakwān states that the Azāriqa dissociate even from bedouin who sought affiliation with them (Hinds xerox, 173.4–.6).

¹⁸ Ash‘arī, 87.2–5.

¹⁹ Ash‘arī, 87.5. The justification is less defensive in the Farq, 63.17: ‘This is something we continue to do despite them’ (hādhā shai' ma zilnā dūnahum). The tight polemic in 1/3 is broken up by a passage of a very different sort, to be discussed below.

²⁰ Dogma, 98. I would add that the apparent redundancy between two elements of Ash‘arī's trinity (barā'a from and takfīr of those who remain at home) is suggestive of an originally Ibāḍī polemical concern. It is the Ibadls, and not the Sunnis, who like to make such distinctions. See above, note 17, for the distinction betweenquietist sympathizers and outright opponents implicit in the rhetoric. This would seem to be an Ibāḍī and not a Sunnī polemical move.

²¹ Ash‘arī, 87.5-.8. Balkī's parallel includes women as well as children (Ḥūr, 178.1–.2).

²² 86.3–.6. For Balkhī as Ash‘arī's source here, see below, note 28.

²³ Ash‘arī, 89.10–. 12.

²⁴ Baghdādl, , al-Milal_ wa-’l- Niḥal (hereafter MN[B\) ed. ‘Nader, A., Beirut, 1970), 63.16–64.1Google Scholar, reading ra'aw for za‘amū at 64.1; Ibn, al-Dā‘ī, Tabṣirat al-‘Awāmm (hereafter TA) (ed. ‘Iqbal, A., Tehran, 1313 A.S.), 39.7–.8Google Scholar.

²⁵ Ash‘arī, 87.6: wa-inna ’l-dār dār kufr ya‘nūna dār mukhālifīhim.

²⁶ A manuscript of Balkhī's Maqālāt al-Islāmiyyīn was discovered by Fu'ād Sayyid nearly 40 years ago in a private Yemeni collection. Sayyid, edited and published only the section relating to the Mu'tazila apud Faḍl al-I'tizāl wa- Tabaqdt al-Mu‘tazila (Tunis, 1974), 63 ffGoogle Scholar. The manuscript does not appear to have been microfilmed by the Egyptian research mission of which Sayyid was a member, and Sayyid's own copy presumably remains in private hands. For Balkhī's Khārijite material, we have thus to make do with the extracts preserved in Nashwān al-Ḥimyarī's Ḥūr. While there is some question about Balkhī's precise date of composition, I assume him to be slightly earlier than Ash‘arī, based on the comments of Ḥājjī, Khalīfa (Keṣf el-zunun [Istanbul, 1943]Google Scholar, 1782.11], and of Balkhī, himself (Faḍl al-I‘tizal, 55).Google Scholar

²⁷ See above, note 16.

²⁸ Both centre on what the sect in question ajma‘ū ‘alā (Balkhī, , Faḍl, 63.1–64.13Google Scholar; Ash‘arī, 86.1–.6). Note also the appearance in Ash‘arī of Balkhī's wa-hum mukhtalifūn. The general introductory passage in Ash‘arī is in fact explicitly ascribed to Balkhī in Farq, 55.8–.56.8; the two Najdite exemptions from the wider Khārijite ijmā‘ noted in Ash‘arī's introduction appear, however, to be the author's own insertions. It is also possible that these exemptions are to be attributed to the slightly earlier writer Zurqān: a similar exemption of the Najdiyya from the Khārijite ijma‘ appears in Nawbakhtī's, Firaq al-Shī‘a (ed. Ritter, , Istanbul, 1931), 10.5–.7Google Scholar, and the passage immediately precedes a statement of Zurqān's on a related subject.

²⁹ In addition to the similarities in their introductory passages, both include toward the end a geographical passage listing the areas where the sect in question is strongest: al-kuwar allatī ghalaba ‘alaihā ‘l-i‘tizāl/‘l-khārijiyya (Balkhī, , Faḍl, 108.1–114.5Google Scholar; Ash‘arī, 128.5–.8). Balkhī includes a similar section in his Khārijite chapter extracted in Ḥūr, 202.15–203.1. The geographical material is accompanied in both Balkhī and Ash‘arī by a parallel tasmiya passage: wa-’l-sabab alladhī la-hu summiyat al-mu‘lazila bi-’l-i‘tizal/summū ’l-khawārij (Balkhī, , Faḍl, 115.1–.15Google Scholar; Ash‘arī, 127.12–128.4). For Balkhī's Khārijite tasmiya passage, see Ḥur, 200.14–201.5. The appearance of these passages in both the Mu‘tazilite and Khārijite chapters of Balkhī suggests that they were important structural features of Balkhī's presentation which Ash‘arī borrowed for his own Khārijite chapter.

³⁰ Ash‘arī, 88.5–.89J.

³¹ At 89.15, Najda is met by a nafar min ahl ‘askar Nāfi’; earlier, at 89.2, Nāfi‘’s disapproval of taqiyya was supported by ahl ‘askarihi illā nafaran yasīran. At 90.1, these splitters tell Najda of their having dissociated from (bari'ū min) Nāfi‘, which echoes the Azāriqa's treatment of the ahl al-taqiyya at 89.3

³² Section 6 begins (at 88.5) ‘wa-kāna sabab al-ikhtilāf alladhī aḥdathahu Nāfi‘…,’ echoing Section l's ‘wa-awwal man aḥdatha ‘l-khilāf bainahum Nāfi‘ b. al-Azraq wa-'lladhī aḥdathahu …’ (86.7–.8). The phrasing also appears in the opening of the Najdiyya section (90.1), further strengthening the link between the two sections posited in the previous note.

³³ See Ash‘arī, 111.6–.12, and 112.3–.6. The sale of slave girls to opponents (110.1–. 12) may be a related issue.

³⁴ The group takes its name from a woman called Umm Najrān (= Ash‘arī's Yemenite woman?) who marries ‘among her people’ in Basra, but then secretly takes a husband from among her Khārijite co-religionists. When the first husband appears and compels her return, the Khawārij disagree about her status (in Malatī, , Kitāb al-Radd wa-'l-Tanbīh, ed. Dedering, S. [Istanbul, 1936], 137.3–.6)Google Scholar; cf. the similar division in Nāfi‘’s camp at Ash‘arī, 89.2.

³⁵ See Schwartz, , Die Anfänge der Ibāḍīlen in Nordafrika (Wiesbaden, 1983), 24, 29.Google Scholar

³⁶ For the Ibāḍī use of the term ‘Khārijite’ to refer to their extremist brethren, see Qalhātī, fol. 198a.3; and Sīrat Sālim, 193.5-.7.

³⁷ Ash‘arī, 86.9–87.2.

³⁸ For the former, see Tabarī, , Ta'rīkh al-Rusul wa-'l-Mulūk, ed. de Goeje, M. J. et al. , (Leiden, 1879–1901)Google Scholar, ser. ii, 1018.3 ft. He figures in inter-Azraqite schisms at a later date, opposing Qaṭarī b. al-Fujā'ā's leadership in the year 77. Baghdādī adds an ‘Abd Rabbih al-Ṣaghīr to the list (Farq, 63.11–. 12), a name which appears later in Ash‘arī's account (at 87.15, in connexion with the movement against Qatarī). I have not been able to identify Ibn al-Waḍīn, but he may be a scribal error for Abū Ismā'īl al-Baṭīhī, an Azraqite splitter mentioned by Ibn Ḥazm (see below, note 78).

³⁹ The similarity is even more striking in the parallel in the Farq, where Baghdādī ends the Muhakkima 'I-Ṻlā affiliation passage (Ash‘arī's Section 3) with the same phrase he has already used to close the Ibn al-Waḍīn issue (Ash‘arīs Section 2): wa-akfara [Nāfi‘] man yukhālifuhu ba‘da dhālika [Farq, 63.15–.17). But Baghdādī may simply be tidying things up.

⁴⁰ A good deal of such material is appended to the end of Ash‘aī's Khārijite chapter (128.9–131.6).

⁴¹ Ash‘arī, 87.9–88.4. See EI (2nd ed.), s.v. Kaṭarī b. al-Fudjā'a.

⁴² al-Tanbīh wa-’l-Radd, 41.9. The ‘Amr b. Fatāt mentioned here as founding the otherwise unknown ‘Amriyya sub-sect of the Azāriqa is presumably identical with the ‘Amr al-Qanā given in Ash‘arī's Section 5.

⁴³ Ash‘arī, 89.3–9. It is separated from the preceding passages by aḥdathū ashyā’, possibly indicating the use of a new source here.

⁴⁴ The scripturalist leanings of some early Khārijites have been noticed before. See Hawting, , ‘The significance of the slogan la hukma illā lillāh and the references to the ḥudūd in the traditions about the Fitna and the murder of ‘Uthmān’, BSOAS, XLI, 3, 1978, esp. 460–1Google Scholar and the sources cited at note 37; and Cook, , ‘‘Anan and Islam: the origins of Karaite scripturalism’, JSAI, 9, 1987, 169–72Google Scholar.

⁴⁵ Ash‘arī, 89.3. See Schacht, , The origins of Muhammadan jurisprudence (Oxford, 1950), 73–4Google Scholar; and cf. Burton, , The collection of the Qur'ān (Cambridge, 1977), 68–104, for one possible historical context.Google Scholar

⁴⁶ istaḥallū khafr al-amāna. It is in this passage only that the Azāriqa clearly label their opponents mushrikūn rather than kuffār (and thus depart from what would become the Ibāḍī position). Sālim (174.6.–7) makes the same charge (istiḥlāl akl al-amānāt). The refusal to abide by the letter of Q. 4:58 (at least in the case of opponents) goes badly with the sect's scripturalist tendencies.

⁴⁷ Mā kaffa aḥad yadahu ‘an al-qital mudh anzala 'llāhu ‘azza wa-jalla 'l-basṭ. Cf. Q. 5:11 and 5:28.

⁴⁸ There is no mention of them, for example, in Baghdādī (Farq, 64.1–.5; and MN[B], 64.1–.6, the latter usually close to Ash‘arī) or in Ibn al-Dā‘ī (TA, 39.8–.11).

⁴⁹ Farq, 75.15–.18: wa-qad kaffū aidiyahum ‘an al-qitāl li-faqdihim man yaṣluḥu li-'l-imāma minhum.

⁵⁰ The imāma doctrine here must accordingly be from a source behind both Baghdādī and Ash‘arī. It would follow that Madelung is unjustified in holding Baghdādī himself responsible for this (possibly inaccurate) explanation of Khalafite, quietism (Religious trends in early Islamic Iran [Albany, N.Y., 1988], p. 66, n. 54)Google Scholar.

⁵¹ ‘Khārijite thought in the Umayyad period’, Der Islam, 36, 1971, 221Google Scholar. Watt's pro-strictness reading here accords with his attempt to defend the Khawārij against charges of backsliding and laxity, and seems to have been influenced by a corrupt passage in Shahrastānī's Najdiyya account. Shahrastānī's statement (92.5) that Najda was ‘rough on people in imposing the penalty for intoxication’ (ghallaẓa ‘alā ‘l-nās fī ḥadd al-khamr taghlīẓan shadīdan), quoted by Watt here, is clearly a corruption of Ash‘arī's charge that Najda ignored the penalty in question (Ash‘arī, 91.12: ‘atlala ḥadd al-khamr; cf. Farq, 68.14, where the sense is unmistakable: asqaṭa hadd al-khamr). Watt appears to have changed his reading somewhat in a later work, where he clearly understands the Azăriqa as guaranteeing Paradise to everyone in their camp (even sinners): ‘We bear witness by God that of those professing Islam in the camp (dār al-hijra) all are approved by God’ (Formative period, 24).

⁵² Sālim, 174.9, where the sect has God forgiving even the adulterer and thief among them (this passage is presumably what led Cook [Dogma, p. 96] to see only laxity in Ash‘arī); and cf. Qalhātī, fol. 199b.–.8, where the Azăriqa grant even the habitual sinner membership in the ahl al-janna as long as he resides in their camp. This ahl al-janna doctrine is in Ash‘arī held by an anonymous Khārijite sect (119.8–.10), but elsewhere in the heresiographical tradition is associated with the mysterious Bid‘iyya sect of the Khawārij (Ḥūr, 178.6–.7; Ibn, ‘Abbād, K. al-Kashf'an Manāhij Aṣnāf al-Khawārij, ed. Danishpazhuh, M., Nashrayi Dānishkada-yi Adabiyyāt-i Tabrīz, 2 [1347sh], 157.23Google Scholar; Abū, 'l-Ma‘alī, Bayān al-Adyān, ed. H., Raḍī [Tehran 1342sh], 48, 3–.4Google Scholar; Khwārazmī, , Liber Mafātīḥ al-Olūm, ed. G. van, Vloten [Leiden 1895], 25.3]Google Scholar. The Bid‘iyya, as we shall see, are given a dubious Azraqite pedigree by some heresiographers.

⁵³ Mubarrad, , al- Kāmilfi ’l-Adab, ed. Ibrāhīm, M. A. (Cairo, n.d.), 3:285.1, 306.2-.4Google Scholar; Ḥatim, Abūal-Rāzī, , Kitāb al-Zina, vol. 3, published apud Sāmarrā'ī, al-Ghuluww wa-'l-Firaq al-Ghāliya fī 'l- Ḥaḍāra 'l-Islāmiyya (Baghdad, 1392/1972), 284.4-. 12; TA, 38.19Google Scholar; ‘b, Abbās. al-Saksakī, Manṣūr, al-Burhdn fī Ma'rifat ‘Aqā'id AM al-Adyān (Ar. MS Cairo, Dar, Kalam 578)Google Scholar, fol. 3a.6. The relationship between these texts will be discussed below.

⁵⁴ MN(B), 64.1-.5. Cf. also TA, 39.11.

⁵⁵ Shahrastānī is the only heresiographer to underline the scripturalist context here (MN[SH], 90.14–. 16).

⁵⁶ The Prophet's sunna is invoked against those who would follow only the ẓāhir al-Qur'ān in Shāfi'i, al-Umm (Cairo, n.d.), 6:130.1–131.11;Google Scholar and cf. Schacht, , Origins, 107Google Scholar.

⁵⁷ 173.3-.4. Sālim's use of the term sunna here, presupposing as it seems to a sunna documented with prophetic ḥadīth, might be an interpolation (cf. Dogma, 100). But the charge itself (i.e. rejection of the common practice of the community) could still be earlyGoogle Scholar.

⁵⁸ There is no mention of these questions in either the Kāmil of Mubarrad or al-Rāzī's, AbAbū ḤātimKitāb al-Zīna. We shall see below that both of these (as well as texts dependent on them) are to be kept apart from Ash'arī and his successorsGoogle Scholar.

⁵⁹ Published by van Ess in Friihe mu'tazilitische Häresiographie: Zwei Werke des Nāši’ al-Akbar igest. 293H) (Beirut, 1971)Google Scholar. For a revised dating and attribution of the text, see Madelung, , ‘Frühe mu'tazilitische Häresiographie: das Kitāb al-Uṣl des Ğa'far b. Ḥarb?’ Der Islam, 57, 1980, 220–36Google Scholar [= Religious schools and sects, vi].

⁶⁰ The term miḥna is used in the account (69.1), but with reference to the troubles caused by Ibn al-Zubair rather than to the Azraqite practice of examination.

⁶¹ Pseudo-Nāshi’, 69.4-.5. The phrase appears also in Nawbakhtī, , Firaq al-Shī'a, 64.3-.4Google Scholar; and Ḥazm, Ibn, al-Fiṣal fī ‘l-Milal wa-'l-Ahwā’ wa-'l-Niḥal (5 vols., Cairo, n.d.), 144.28Google Scholar. It appears in Ash'ari (129.2), but not in a narrowly Azraqite context.

⁶² Pseudo-Nāshi', 69.3–.4.

⁶³ The former is normally part of the ‘Ajārida-Tha'aliba network, and is not considered an Azraqite group elsewhere. The latter is frequently an orphan group, although some relationship with the Azāriqa is implied by Ash‘arī; see below, note 67.

⁶⁴ The earliest reference to such a group is in the Sīrat Sālim, and there entirely in a polemical context; the author says nothing of t he sect's beliefs, but simply calls on them to act in accordance with the sunna of the Prophet (193.15). Cook identifies pseudo-Nāshi”s Bid'iyya (69.15–70.6) with Salim's sect, based on the former's rejection of the sunna's five daily prayers (Dogma, 92). A twoprayers doctrine need not be behind Sālim's criticism of the Bid'iyya, though. As we shall see, there are any number of scripturalist positions ascribed to the Azāriqa (and its sub-sects) that might have done j u s t as well.

⁶⁵ Pseudo-Nāshi’ 69.11–70.6.

⁶⁶ Presumably the five daily prayers are not a requirement even given the unanimous agreement of Muslims, since in this case the ijmā’ goes beyond the Qur'ānic language. A parallel to the first point is listed for the Najdiyya in Ash‘arī, 125.11–12, MN(SH), 92.16–.18, and Nawbakhtī, , Firaq al-Shī'a, 10.8-.9Google Scholar. In all three places the scripturalist implications are less prominent than in pseudo- Nashi’, and the doctrine looks to be as much a rejection of imāma as sunna.

⁶⁷ Ash'arī's single reference to them (at 126.14) is apart from his main Kharijite account. This probably argues for a different source. (Note also that Ash‘arī, contrary to his custom, actually cites a source here, although an anonymous one [ḥakā ḥākin].) The significance of this will become apparent below.

⁶⁸ See below, note 78.

⁶⁹ See, for example, Umm, 5:5.5–6.10, where the issue is put explicitly as one between those who accept and those who reject tradition as a source of law. Cf. also Ricks, ‘Kinship bars to marriage in Jewish and Islamic law’, Studies in Islamic and Judaic Traditions (Atlanta, 1986), 131Google Scholar.

⁷⁰ On the status of the ‘floating fish’, see Cook, , ‘Early Islamic dietary law℉, JSAI, 7, 1986, 246–7Google Scholar.

⁷¹ His phrasing conflates the normally distinct issues of rajm and qadhf. ibtal rajm man zanā wahuwa muḥṣan (Fiṣal, 5:144.23-.24). The heresiographers everywhere else use muḥṣan only in the context of slander (and not stoning).

⁷² Fiṣsal, 4:144.24. The Khārijite position is also given in Sarakhsī, Kitāb al-Mabsūt (Cairo, 1324), 9:133.17–134.3; and Hazm, Ibn, Kitāb al-Muḥallā (Cairo, 1347), 11:357.10-.20Google Scholar. The Khawārij take the aidiyahumā of Q. 5:38 quite naturally as referring to the entire arm, while the jurists are forced to establish that the cutting be done only at the wrist.

⁷³ wa-awjabū ‘alā ‘l-ḥā'iḍ al-ṣalāt wa-'l-ṣiyām fī haiḍihā wa-qāla ba'ḍuhum lā wa-lākin taqḍī ‘l- ṣalāt idhā ṭahurat kamā taqḍī ‘l-ṣiyām (Fiṣal, 4:144.23–.24). For the general Sunnī position on qaḍā’ al-ṣiyām, see Khiraqī, , Mukhtasar (Damascus, 1384), 60.6–8Google Scholar; Muḥallā, 2:175.10–.12 (reading wa-lā for wa-qad); Umm, 2:104.9–.13, and 1:60.1–.2.

⁷⁴ See the discussion in Umm, 1:59.17ȓ60.1. The tradition quoted here (in which ‘Ā'isha is prohibited from making the ṭawāf) is one of the standard proof-texts demonstrating the unfitness of the ḥā'iḍ. It is worth noting that the Ahmadiyya today reject many of the traditional restrictions placed on the ḥā'iḍ, and justify their views on scripturalist lines (Ali, M. Mohammed, The religion of Islam, 392 fGoogle Scholar; cited in Bousquet, , ‘La puretee rituelle en Islam’, RHR, 138, 1950, p. 62, n. 5Google Scholar.

⁷⁵ Muḥallā, 2:162.11. The group permits prayer, fasting, circumambulation of the Ka'ba, and sexual intercourse. Even the last does not explicitly come under the prohibition of Q. 2:222, although it is hard to imagine an interpretation that would permit it. The departure from ijmā’ mentioned here is elsewhere associated with the Baṭīḥiyya sub-sect of the Azāriqa (see below, note 78).

⁷⁶ For example, Abū Dāwūd, Sunan (4 vols., Beirut, n.d.), 1:68.16–69.3Google Scholar; Aḥmad b., Hanbal, Musnad(6 vols., Beirut, n.d.), 6:231.25–232.2, and 6:94.19-.21Google Scholar; Bukhari, , Saḥīḥ (al-ḥaiḍ, Bab 20). ‘Ā'isha insists that while the Prophet was alive ‘fa-nu'maru bi-qaḍā’ al-ṣawm wa-lā nu'maru bi-qadā al-ṣalātGoogle Scholar’.

⁷⁷ The people of Ḥarūrā’ (here, said to be a place in Bahrain) are notorious for asking nettlesome questions (kāna ahluhu…muta'annitln fī ‘l-su'āl); in this case, they inquire about the Sunni inconsistency in requiring qadā al-ṣawm but not qaḍṣ al-ṣalāt (al-Firaq al-Muftariqa Baina Ahlal-Zaigh wa-'l-Zandaqa, ed. Kutluay, Y. [Ankara, 1961], 10.14–.19)Google Scholar.

⁷⁸ Fiṣal, 4:144.18-.21; cf. Saksakī, , Burhān, fol. 5a. 11–.14Google Scholar (corrupted to Matīkhiyya), and Yāfi'ī, Marham al-'Ilal al-Mu'ḍila (anon, abridgement), Arabic MS Berlin, We. 1819, Ahlwardt 2806, fol. 63b.l 1–.14 (corrupted to Matbakhiyya). Ibn Ḥazm names as founder of the group a certain Abū Ismā'īl al-Baṭīḥī, who, he tells us, was originally an Azraqite before outdoing them in extremism (ghalā –an sd'ir al-Azāriqa wa-zāda ‘alaihim). He considers the sect outside the ijmā‘ al-umma (Fiṣal, 2:89.9–10). This charge goes nicely with the label Bid'iyya (applied to the other Azraqite sub-sect), and with pseudo-Nāshi‘’s third point, mentioned above. It is also brought against an anonymous sub-sect of the Azāriqa elsewhere by Ibn Ḥazm (see above, note 75). This suggests that the distinction between the Bid'iyya and Batīhiyya was never very sharp; the two may well have been little more than convenient labels created to organize the legal material gathered on the Azāriqa.

⁷⁹ Fiṣal, 4A44.17–.20.

⁸⁰ The acceptance of jizya from them is based entirely on the Prophet's practice and the sīra of the first two Caliphs; cf. the traditions cited in Abū, Yūsuf, Kitāb al-Kharāj (Cairo, 1397), 139–42; and El (2nd ed.)Google Scholar, s.v. Madjūs. The ambiguity of the Qur'ānic passage at the same time allows the community to place the Majus in the category of idolators (rather than ahl al-kitāb) as far as munākaḥa and dhaba'ih go. The whole question may have been an object of dispute within the Azariqa. Ibn Hazm points out that the sect murders Muslims while leaving Christians, Jews, and Majus in safety (4:144.28–.29). There is some support for this in the Sīrat Sālim (175.2), where another Kharijite sect, the Najdiyya, are said to honour the covenants of –their people’ (i.e. ordinary Muslims) towards the dhimmṣs, even while considering ‘their people’ to be polytheists. Cf. Pampus, Über die Rolle, 86.

⁸¹ Possibly, we have here an allusion to a specific feature of the khuṭba rejected by the sect. The most polemical feature of the legal discussion concerning the festival sermon is the insistence that it be performed after the prayer (in constrast to the usual practice on Fridays); the sermon is not among the stipulations (sharā'iṭ) of the ṣalāt al- 'īd, and can in fact be omitted without invalidating the prayer. (See, for example, Sarakhsī, Mabsūt, 2:37.10–38.4; Umm, 1:239.14–.15; Rushd, Ibn, Bidāyal al-Mujtahid [Cairo, 1386], 221.12-.14)Google Scholar. The lawyers generally hold an Umayyad responsible for altering the Prophet's practice and placing the khuṭba before the ṣalāt in an effort to expand their own influence. Whatever the original practice, might a Khārijite opposition to the festival khuṭba have arisen in an anti-Umayyad context?

⁸² For the jurists’ discussion of the ashhur ma'lūmāt, see, for example, Muḥallā, 7:65.10–68.16 and 69.17–25; and Ibn Rushd, Bidāya, 1:334. Note that pseudo-Nāshi’ (69.8–.11) associates a hajj doctrine with his Khazimiyya sub-sect of the Azariqa: they prohibit pilgrimage as long as taqiyya is operative.

⁸³ Ash‘;arī, 126.14–15.

⁸⁴ See above, note 54.

⁸⁵ Lewinstein, , Studies, 70–4Google Scholar. With very minor changes in wording, Baghdādī's presentation is reproduced by Isfarā'inī, , al-Tabṣīr fī 'l-Dīn wa-Tamyīz al-Firqa 'l-Nājiya ‘an al-Firaq al-Hālikln (Cairo, 1359/1940), 29.11–30.14Google Scholar.

⁸⁶ MN(SH), 89.8–91.4.

⁸⁷ MN(SH), 90.18–.20. The other writers who refer to the doctrine are all clearly dependent on Shahrastānī; namely, Ibn, al-Murtaḍā, al-Bahr al-Zakhkhār (Muqaddimat Kilāb al-Baḥr al- Zakhkhār [Beirut, n.d.]), 48.11Google Scholar; Iji, Mawāqif (Cairo, n.d.), 324.15; Kirmāni, , Dhail Kitāb Sharḥ al- Mawāqif(= al-Firaq al-Islāmiyya) (Baghdad, 1973), 66.4–.5Google Scholar.

⁸⁸ Pseudo-Nāshi’, 70.1–.2.

⁸⁹ MN{SH), 90.17–.18: inna 'l-laqiyya ghair jā'iza fī qawl wa-lā fī ‘amal; cf. Shahrastānī's description of the Najadāt's taqiyya doctrine (cited from Balkhī) at 92.15–. 16.

⁹⁰ Qalhātī, fol. 202b.16–203a.1 (reading taqiyya for baqiyya, and with reference to the Ṣufriyya).

⁹¹ Qalhātī's, own dates are unknown. For reasons which are unclear to me, Wilkinson calls the Kashfa. seventh-century [hijrī] work based on early sources (‘The early development of the Ibādī movement in Basra’, Studies in the first century of Islamic society [Southern Illinois University, 1982], p. 242, n. 8), while Cook (working from the chain of transmission given at the end of the British Museum MS) tentatively places the author in the fifth century (Dogma, 234).Google Scholar

⁹² Lewinstein, , Studies, pp. 116, 150, n. 82Google Scholar. Ironically, Rieu thought that Shahrastānī might have been behind Qalhātī, (Supplement to the Catalogue of the Arabic manuscripts in the British Museum [London, 1894], 122)Google Scholar. The nature of the parallels, however, points to Shahrastānī as the dependent writer: as I argue in my thesis, the language is usually more at home in the Ibādī text.

⁹³ I am leaving aside here the eastern Ḥanafite material studied in Part 2 of my thesis.

⁹⁴ Zīna, 284.8–.10. The wording matches Kāmil, 3:184.15–. 16.

⁹⁵ Zīna, 284.13–. 15.

⁹⁶ Zīna, 284.11; Kāmil, 3:285.1.

⁹⁷ Zīna, 284.4–.6 (reading tukaffiru for taqbalu, and min ahl maqālātihim for min ghair, etc.), where it has fallen into the Baihasiyya section. Cf. also Kāmil, 3:306.2–.4.

⁹⁸ Zīna, 284.11–. 12; Kāmil, 3:285.2 (the miḥna point which follows in the Kāmil has fallen out of the Zīna version). Cf. Sālim, 172.3, for the coupling of inheritance and marriage. Qalhātī (fol. 197a.8) has all three prohibitions together. See Pampus, , Über die Rolle, 78Google Scholar. Cook's heresiographical parallel to Sālim on these points is the Kāmil (Dogma, p. 198, n. 58), but Mubarrad is atypical.

⁹⁹ Zīna, 284.12–.13; Kāmil, 3:285.3–.4.

¹⁰⁰ Zīna, 284.12–.13; Kāmil, 3:285.3–.4 has wa-'l-qa‘ada bi-manzilatihim (= manzilat kuffār al- ‘arab). The parallel in Sālim runs anzalahum bi-manāzil ‘abadat al-awthān (172.2.–.3).

¹⁰¹ Abū Hātim is mentioned at 146.17, and 152.2. The material from the Zīna is inserted at TA 38.18–39.2. (The Kdmil is clearly not Ibn al-Dā‘īs source, as the minor modifications made by Abū Ḥātim appear also in the TA.

¹⁰² The only source actually mentioned is a lost Kitāb al-Firaq of a certain Abū Muḥammad, composed during the Caliphate of al-Muqtafī (530–55). See Ritter, ‘Häresiographen’, 47.

¹⁰³ Burhān, fol. 1 3a.3–.7. This matches Zīna, 284.8–. 12, both in language and content.

¹⁰⁴ Burhān, fol. 3a.7–.9. They not only murder children, but they kill the blind, the crippled, and the old. They also throw children into pots of boiling curd. This and the material in the previous note find their way into the mukhtasar of Yāfi‘’s Marham cited above, note 78.

¹⁰⁵ Burhān, fol. 3a.9–3a.14; cf. Fiṣal, 5:30.10–.16.

¹⁰⁶ Qalhātī, fol. 197a.8.

¹⁰⁷ Nāfi‘ was the first to break with the ‘people of rectitude’ (as Qalhātī terms the proto-Ibāḍī moderates); the sect held hijra to be the primary religious duty of the true Muslim; the sect dissociated from those refusing to undertake a hijra; and the sect took ordinary Muslims to be polytheists (Qalhātī, fol. 197a.4–.7).

¹⁰⁸ The wording of the entire passage is in fact parallel. Qalhātī, fol. 197a.6: wa-sabā ahl al-qibla wa-ghanima amwālahum [for amālahum] wa-sabā dharāriyahum; Sālim, 172.6–.7: wa-'stahallū sabya qawmihim wa-'stinkāḥa nisā'ihim wa-khumusa amwālihim wa-qatla dharārīhim. Cf. also Nawbakhtī, Firaq al-Shī‘a, 64.7, where we have the phrase ra'aw sabya 'l-nisā' wa-qatl al-aṭfāl; just above this passage, we have qatl ahl al-qibla wa-akhdh amwālihim. Interestingly, Malaṭī (Tanbth, 41.9–. 10) preserves what appears to be a mirror image of Qalhātī's wording, which suggests that the enslavement charge in Sālim and QalhātIī might have enjoyed more heresiographical circulation than is apparent from the extant material: they do not spill Muslim blood, take Muslim property as booty, or enslave Muslim children (lā yarawna ihrāq dimā’, al-muslimīn wa-lā ghanma amwālihim wa-lā sabya qawmihim).

¹⁰⁹ Unfortunately, the (unpublished?) Risāla fī Bayān al-Firaq al-Ibādiyya al-Sitta wa-Ghairihā by the early sixth-century Ibādī scholar Abū ‘Amr g‘Uthman b. Khalīfa al-Ṣūfī is unavailable to me. The manuscript is cited by Schwartz (Anfänge, 311), and Ennami refers both to a manuscript and to a printed version of the text lacking place and date of publication (Studies in Ibāḍism [University of Libya, 1392], p. 186, note 189).