¹ Locke, John, An Essay Concerning Human Understanding (1690), ed. Yolton, John W. (London, 1961)Google Scholar, 4.17.20; cf. Vilhelm Hansen, H., ‘Locke and Whately on the Argumentum ad Ignorantiam’, Philosophy and Rhetoric, 31.1 (1998): 55–63Google Scholar. On the argument from ignorance and its recent history, see Walton, Douglas, Arguments from Ignorance (University Park, PA, 1996)Google Scholar.

² That said, I have found nothing to suggest that Locke's identification of this form of argument was inspired by this earlier discussion, although he was influenced by other aspects of medieval Arabic philosophy. See, for instance, Russell, Gül A., ‘The impact of the Philosophus Autodidactus: Pocockes, John Locke, and the Society of Friends’, in Russell, Gül A. (ed.), The ‘Arabick’ Interest of the Natural Philosophers in Seventeenth-Century England (Leiden and Boston, 1994), pp. 224–65Google Scholar.

³ Shihadeh, Ayman, ‘From al-Ghazālī to al-Rāzī: 6th/12th century developments in Muslim philosophical theology’, Arabic Sciences and Philosophy, 15.1 (2005): 141–79CrossRefGoogle Scholar, at p. 165.

⁴ See n. 110 below.

⁵ The reconstruction of the historical roots of this debate is hampered by the paucity of early sources and of relevant materials transmitted in later ones. Post-Rāzian treatments of the argument from ignorance are based closely on the sources examined here.

⁶ This overview is based on a number of primary sources, including: Fūrak, Ibn, Mujarrad maqālāt al-Shaykh Abī al-Ḥasan al-Ashʿarī, ed. Gimaret, Daniel (Beirut, 1987)Google Scholar, pp. 286 ff.; al-Bāqillānī, Kitāb al-Tamhīd, ed. McCarthy, Richard J. (Beirut, 1957)Google Scholar, pp. 11 ff.; al-Juwaynī, al-Shāmil fī uṣūl al-dīn, ed. Frank, Richard M. (Tehran, 1981)Google Scholar, pp. 60 ff.; al-Jabbār, ʿAbd, al-Mughnī fī abwāb al-tawḥīd wa-al-ʿadl, 12. al-Naẓar wa-al-maʿārif, ed. Madkūr, Ibrāhīm (Cairo, 1965)Google Scholar; Mattawayh, Ibn, al-Tadhkira fī aḥkām al-jawāhir wa-al-aʿrāḍ, ed. Gimaret, Daniel, 2 vols. (Cairo, 2009), vol. 2, pp. 579 ff.Google Scholar

⁷ For the philosophers, ‘dalīl’ generally refers to a proof, or an argument, rather than evidence. This, as we shall see further below, is the sense in which al-Rāzī, under the influence of the philosophers, often uses the term.

⁸ On the epistemology and logic of classical kalām, see also: van Ess, Josef, ‘The logical structure of Islamic theology’, in von Grunebaum, Gustave E. (ed.), Logic in Classical Islamic Culture; Giorgio Levi Della Vida Conferences (Los Angeles, 1967), pp. 21–50Google Scholar; id., Die Erkenntnislehre des ʿAḍudaddīn al-Īcī: Übersetzung und Kommentar des ersten Buches seiner Mawāqif (Wiesbaden, 1966)Google Scholar (al-Ījī, of course, is a post-classical theologian); Bernand, Marie, Le problème de la connaissance d'après le Muġnī du cadi ʿAbd al-Ǧabbār (Algiers, 1982)Google Scholar; Perler, Dominik and Rudolph, Ulrich (eds.), Logik und Theologie: das Organon im arabischen und im lateinischen Mittelalter (Leiden and Boston, 2005)Google Scholar.

⁹ For instance, al-Malāḥimī, Rukn al-Dīn, al-Muʿtamad fī uṣūl al-dīn, ed. Madelung, Wilferd (Tehran and Berlin, 2012)Google Scholar, pp. 183; 233; 253.

¹⁰ The term ‘argumentum ad ignorantiam’ was coined by Locke. For the purposes of our present study, I would have preferred the more neutral label, ‘the argument from the absence of evidence’, had it not been so cumbersome. The ‘argument from non-evidence’ is ambiguous and can be misleading. So in the absence of a good alternative, I shall use the conventional ‘argument from ignorance’, despite its shortcomings.

¹¹ See p. 184 below.

¹² Taqī al-Dīn al-ʿUjālī, al-Kāmil fī al-istiqṣāʾ fī-mā balaghanā min kalām al-qudamāʾ, ed. al-Sayyid al-Shāhid (Cairo, 1999)Google Scholar, p. 322. More on this source in § 3.4 below.

¹³ For instance, al-Juwaynī, al-Talkhīṣ fī uṣūl al-fiqh, ed. al-Nībālī, ʿAbdallāh J. and al-ʿUmarī, Shubbayr A., 3 vols. (Beirut, 1996), vol. 1, pp. 119–20Google Scholar. Following him, his student Abū al-Qāsim al-Anṣārī dedicates a short section to the problem ‘whether or not the non-existence [of evidence] can be viewed as evidence’ (al-Ghunya fī al-kalām, ed. al-Hādī, Muṣṭafā Ḥ. ʿAbd, 2 vols. [Cairo, 2010], vol. 1, p. 522Google Scholar [I check this edition against the only extant manuscript copy, MS Istanbul, Topkapı Sarayı Müzesi Kütüphanesi, Ahmet III, 1916]; id., Sharḥ al-Irshād, MS Princeton, University Library, Yahuda, 634, f. 59a).

¹⁴ al-Rāzī, Fakhr al-Dīn, Nihāyat al-ʿuqūl fī dirāyat al-uṣūl (MS Istanbul, Topkapı Sarayı Müzesi Kütüphanesi, Ahmet III, 1874)Google Scholar, f. 5a ff.

¹⁵ Al-Anṣārī, Ghunya, vol. 2, p. 930; cf. Ḍiyāʾ al-Dīn al-Makkī, Nihāyat al-marām fī dirāyat al-kalām (MS Hyderabad, Andhra Pradesh Oriental Manuscript Library, Kalām 13Google Scholar; published as Fakhr al-Dīn al-Rāzī's Father, Ḍiyāʾ al-Dīn al-Makkī, Nihāyat al-marām fī dirāyat al-kalām: Facsimile of the Autograph Manuscript of Vol. II, intro. Shihadeh, Ayman [Berlin and Tehran, 2013])Google Scholar, f. 81a.

¹⁶ ʿAbd al-Jabbār, al-Mughnī fī abwāb al-tawḥīd wa-al-ʿadl, 7. Khalq al-Qurʾān, ed. al-Abyārī, Ibrāhīm (Cairo, 1961)Google Scholar, p. 14; al-Zamakhsharī, al-Minhāj fī uṣūl al-dīn, ed. Schmidtke, Sabine (Stuttgart, 1997)Google Scholar, p. 55.

¹⁷ Al-Juwaynī, al-Shāmil fī uṣūl al-dīn, ed. al-Nashshār, ʿAlī S. (Alexandria, 1969)Google Scholar, p. 387 (reading yadullu instead of lā yadullu, ln. 8).

¹⁸ Al-Malāḥimī, Muʿtamad, p. 282; al-Zamakhsharī, Minhāj, p. 53; al-Rāzī, Nihāyat al-ʿuqūl, f. 5a.

¹⁹ Al-Juwaynī, Shāmil (ed. Frank), p. 76.

²⁰ Al-Malāḥimī, Muʿtamad, pp. 256 ff.

²¹ For instance, Ibn Mattawayh, Tadhkira, vol. 2, p. 721.

²² Misbeliefs are not, strictly speaking, impediments, although they can prevent us from arriving at knowledge by means of reflection. It is assumed that reflection should start from sound convictions and sound evidence.

²³ Al-Zamakhsharī, Minhāj, p. 53.

²⁴ Al-Malāḥimī, Muʿtamad, pp. 584–6; al-Juwaynī, Shāmil (ed. al-Nashshār), p. 387.

²⁵ Ibn Mattawayh, Tadhkira, vol. 1, p. 131; al-ʿUjālī, Kāmil, pp. 322 ff.

²⁶ Cited in: al-Rāzī, Muḥaṣṣal afkār al-mutaqaddimīn wa-al-mutaʾakhkhirīn min al-ḥukamāʾ wa-al-mutakallimīn, ed. Atay, Hüseyin (Cairo, 1991)Google Scholar, p. 438; cf. id., Maʿālim uṣūl al-dīn, ed. Saʿd, Ṭāha ʿAbd al-Raʾūf (Cairo, n.d.)Google Scholar, p. 66.

²⁷ Al-Zamakhsharī, Minhāj, p. 55; al-ʿUjālī, Kāmil, pp. 322 ff. ʿAbd al-Jabbār (Mughnī, vol. 7, pp. 14 ff.) uses the argument to prove that divine speech is not an entitative determinant, contra the Ashʿarīs.

²⁸ ʿAbd al-Jabbār, Mughnī, vol. 7, pp. 14–15.

²⁹ Abū al-Ḥusayn al-Baṣrī, Taṣaffuḥ al-adilla, The extant parts introduced and edited by Madelung, Wilferd and Schmidtke, Sabine (Wiesbaden, 2006), pp. 13–14Google Scholar.

³⁰ Al-Malāḥimī, Muʿtamad, pp. 256–60.

³¹ For instance, ʿAbd al-Jabbār, Mughnī, vol. 12, p. 14; al-Malāḥimī, Muʿtamad, pp. 183; 233; Ibn Mattawayh, Tadhkira, vol. 2, p. 721.

³² That ‘jahālāt’ (always, in this context, in the plural) is the plural of ‘jahl’ (rather than ‘jahāla’, as it may first seem) is evident, for instance, in Ibn Mattawayh, Tadhkira, vol. 2, pp. 635–6. Both ‘jahl’ and ‘ʿilm’ also have a generic sense and can serve as verbal nouns, in which cases they cannot be pluralised.

³³ For instance, al-Tahānawī, Muḥammad ʿAlī, Kashshāf iṣṭilāḥāt al-funūn, ed. Daḥrūj, ʿAlī, 2 vols. (Beirut, 1996), vol. 1, pp. 599–600Google Scholar.

³⁴ Theologians sometimes use the pejorative label ‘ahl al-jahālāt’ to refer to exponents of belief systems that abound with such misbeliefs (jahālāt ₂), whether they are based, for instance, on scripture (in cases where scriptural evidence, according to kalām theologians, cannot be adduced), dreams, mystical experience, popular superstition or received tradition. There are also references in earlier theological sources, particularly in al-Jāḥiẓ's (d. 255/869) Kitāb al-Ḥayawān, to so-called ‘exponents of the jahālāt’ (aṣḥāb al-jahālāt). In this case, I believe that both senses, jahālāt ₁ and jahālāt ₂, are intended. There is no space to discuss this here; however see both Patricia Crone's ‘Al-Jāḥiẓ on Aṣḥāb al-Jahālāt and the Jahmiyya’ (in Burnett, Charles et al. [eds.], Medieval Arabic Thought: Essays in Honour of Fritz Zimmermann [London, 2012], pp. 27–39Google Scholar), and ʿAbd al-Jabbār's account (Mughnī, vol. 12, pp. 47 ff.) of al-Jāḥiẓ's criticism of aṣḥāb al-jahālāt, which is probably based on a dedicated text in which the latter responds to this group.

³⁵ On Abū al-Ḥusayn and his school, see ‘Abū l-Ḥusayn al-Baṣrī’, EI3.

³⁶ ʿAbd al-Jabbār, Mughnī, vol. 7, p. 14.

³⁷ This is a commentary on ʿAbd al-Jabbār's own al-Jumal wa-al-ʿuqūd, a book on theology (al-Ḥākim al-Jishumī, Sharḥ al-ʿuyūn [fragments], in Sayyid, Fuʾād [ed.], Faḍl al-iʿtizāl wa-ṭabaqāt al-muʿtazila [Tunis, 1974], pp. 363–93Google Scholar, at p. 369). Neither work is extant.

³⁸ Cited in: al-Malāḥimī, Muʿtamad, p. 257; cf. ʿAbd al-Jabbār, Mughnī, vol. 7, p. 14.

³⁹ Cited in: al-Malāḥimī, Muʿtamad, p. 257; and Abū al-Ḥusayn al-Baṣrī, Taṣaffuḥ, p. 13; cf. ʿAbd al-Jabbār, Mughnī, vol. 7, p. 14.

⁴⁰ Cited in: al-Malāḥimī, Muʿtamad, p. 257; cf. ʿAbd al-Jabbār, Mughnī, vol. 7, p. 14.

⁴¹ Reading ʿibārāt instead of ʿibādāt (ʿAbd al-Jabbār, Mughnī, vol. 7, p. 14, ln. 16).

⁴² See ʿAbd al-Jabbār, Mughnī, vol. 12, pp. 41 ff. On scepticism in general, see van Ess, Josef, ‘Skepticism in Islamic religious thought’, al-Abḥāth, 21.1 (1968): 1–18Google Scholar.

⁴³ Al-Malāḥimī, Muʿtamad, p. 257.

⁴⁴ Al-Malāḥimī, Muʿtamad, p. 256.

⁴⁵ Abū al-Ḥusayn al-Baṣrī, Taṣaffuḥ, p. 13. Neither the extant fragments of the Taṣaffuḥ nor his work on the theory of jurisprudence al-Muʿtamad fī uṣūl al-fiqh (ed. Hamidullah, Muhammad et al. , 2 vols. [Damascus, 1965])Google Scholar, nor his Kitāb al-Qiyās al-sharʿī (appended to the published Muʿtamad, vol. 2, pp. 1029–50) contain a substantial discussion of the argument from ignorance other than an argument cited from ʿAbd al-Jabbār (see n. 29 above).

⁴⁶ al-aṣl fī hādhihi al-dalāla huwa anna mā lā yuʿlamu bi-nafsihi, wa-lā yaṣiḥḥu an yakūna ilā al-ʿilm bihi ṭarīq, fa-inna tajwīz thubūtihi yuʾaddī ilā al-jahālāt, wa-mā addā ilayhā fa-innahu yajibu nafyuhu.

⁴⁷ Al-Malāḥimī, Muʿtamad, p. 256. The description ‘maʿlūm bi-nafsihi’ is identical to ‘badīhī’ (cf. Muʿtamad, p. 273).

⁴⁸ Al-Malāḥimī, Muʿtamad, p. 256.

⁴⁹ Al-Malāḥimī, Muʿtamad, pp. 256–7.

⁵⁰ Al-Malāḥimī, Muʿtamad, p. 257.

⁵¹ See p. 211 below.

⁵² Al-Malāḥimī, Muʿtamad, p. 258.

⁵³ Al-Malāḥimī, Muʿtamad, pp. 258–9.

⁵⁴ Al-Malāḥimī, Muʿtamad, p. 257. A brief version of this objection against the argument from ignorance is advanced by al-Rāzī (Maʿālim, p. 66), although it is unclear whether he derived it from al-Malāḥimī or an earlier source.

⁵⁵ This obligation, for the Muʿtazila, would be an aspect of God's duty to assist human beings (luṭf) (on which, see ‘Favour, Divine’, EI3). On Muʿtazilī ethical theory, see most recently my ‘Theories of ethical value in kalām: a new interpretation’, in Schmidtke, Sabine (ed.), Oxford Handbook of Islamic Theology (Oxford, forthcoming in 2013)Google Scholar.

⁵⁶ Al-Malāḥimī, Muʿtamad, p. 258.

⁵⁷ On the classical kalām view that angels are bodies, see, for instance, my ‘Classical Ashʿarī anthropology: body, life and spirit’, The Muslim World, 102.3–4 (2012): 433–77CrossRefGoogle Scholar.

⁵⁸ Al-Malāḥimī, Muʿtamad, pp. 259–60.

⁵⁹ Al-Malāḥimī, Muʿtamad, pp. 259–60.

⁶⁰ Both positions are discussed in: al-Rāzī, Muḥaṣṣal, pp. 437–9.

⁶¹ This discussion occurs in one of the extant parts of the work: al-Juwaynī, Shāmil (ed. Frank), pp. 67–8. Frank's edition only contains parts of the book that do not appear in the earlier editions published by ʿAlī S. al-Nashshār and Helmut Klopfer. Cf. al-Juwaynī (Ibn al-Amīr), al-Kāmil fī ikhtiṣār al-Shāmil (henceforth Ikhtiṣār), ed. al-Munʿim, Jamāl ʿA. ʿAbd, 2 vols. (Cairo, 2010), vol. 1, pp. 288–9Google Scholar (I constantly check this very poor edition against the only extant manuscript copy, MS Istanbul, Topkapı Sarayı Müzesi Kütüphanesi, Ahmet III, 1322); al-Anṣārī, Ghunya, vol. 1, pp. 520–2; id., Sharḥ al-Irshād, ff. 58a–59a.

⁶² On investigation and disjunction, see, for instance: van Ess, Die Erkenntnislehre des ‘Aḍudaddīn al-Īcī, pp. 394 ff.; Hallaq, Wael, ‘The development of logical structure in Sunni legal theory’, Der Islam, 64.1 (1987): 42–67CrossRefGoogle Scholar, at pp. 60–1; id., ‘Logic, formal arguments and formalization of arguments in Sunnī jurisprudence’, Arabica, 37.3 (1990): 315–58CrossRefGoogle Scholar; al-Zarkashī, Badr al-Dīn, al-Baḥr al-muḥīṭ fī uṣūl al-fiqh, ed. Ghudda, ʿAbd al-Sattār Abū, 6 vols. (Kuwait, 1992), vol. 5, pp. 222–30Google Scholar.

⁶³ Al-Juwaynī, Shāmil (ed. Frank), pp. 66–7.

⁶⁴ Al-Juwaynī, Shāmil (ed. Frank), p. 67.

⁶⁵ Al-Juwaynī, Shāmil (ed. Frank), p. 67. This, incidentally, illustrates al-Juwaynī's close familiarity with ʿAbd al-Jabbār's work.

⁶⁶ Al-Anṣārī, Ghunya, vol. 1, p. 521; cf. id., Sharḥ al-Irshād, f. 58a.

⁶⁷ Al-Juwaynī, Shāmil (ed. Frank), pp. 67–8.

⁶⁸ Reading al-ṭarīqa al-thāniya instead of al-thālitha (ln. 11).

⁶⁹ Reading al-majāl instead of al-muḥāl (ln. 20).

⁷⁰ Al-Juwaynī, al-Burhān fī uṣūl al-fiqh, ed. al-Dīb, ʿAbd al-ʿAẓīm, 2 vols. (Doha, 1399 AH), vol. 1, p. 131Google Scholar.

⁷¹ Al-Juwaynī, Burhān, vol. 2, p. 815.

⁷² Al-Juwaynī, Burhān, vol. 2, p. 816.

⁷³ Al-Juwaynī (Ibn al-Amīr), Ikhtiṣār, vol. 2, pp. 675–7.

⁷⁴ Al-Anṣārī, Ghunya, vol. 2, pp. 929–30. The discussion is later reproduced by al-Anṣārī's student, Ḍiyāʾ al-Dīn al-Makkī (Nihāyat al-marām, ff. 81a–b).

⁷⁵ I use ‘ad hominem argument’, not in the relatively recent, pejorative sense of attacking an opponent's personal background and motives instead of responding to his views and arguments, but in the pre-modern sense, which is still evident in Locke's Essay (4.17.21): ‘… to press a man with consequences drawn from his own principles or concessions. This is already known under the name of argumentum ad hominem’. On this, see, for instance, Hamblin, C. L., Fallacies (London, 1970)Google Scholar; Nuchelmans, Gabriël, ‘On the fourfold root of the Argumentum ad Hominem’, in Krabbe, Erik C. W. et al. (eds.), Empirical Logic and Public Debate: Essays in Honour of Else M. Barth (Amsterdam, 1993), pp. 37–47Google Scholar.

⁷⁶ Al-Anṣārī, Ghunya, vol. 2, p. 929; cf. al-Juwaynī (Ibn al-Amīr), Ikhtiṣār, vol. 2, p. 675.

⁷⁷ Al-Juwaynī (Ibn al-Amīr), Ikhtiṣār, vol. 2, p. 675; al-Anṣārī, Ghunya, vol. 2, p. 929.

⁷⁸ Al-Juwaynī (Ibn al-Amīr), Ikhtiṣār, vol. 2, pp. 675–6; cf. id., Burhān, vol. 1, pp. 142–5.

⁷⁹ Al-Juwaynī (Ibn al-Amīr), Ikhtiṣār, vol. 2, p. 676. In the later Burhān, he defends a different position on this problem of unknown accident classes, which is of little relevance to our present purposes and deserves a separate treatment.

⁸⁰ Al-Juwaynī, Burhān, vol. 1, p. 145.

⁸¹ Al-Anṣārī, Ghunya, vol. 2, pp. 929–30; cf. al-Makkī, Nihāyat al-marām, f. 81a; al-Juwaynī (Ibn al-Amīr), Ikhtiṣār, vol. 2, p. 677 (abridged).

⁸² Used in literary contexts, especially in relation to poetry, ‘muʿāraḍā’ refers to the emulation in style and quality of another person's work. The same sense in intended in discussions of prophetic miracles, particularly the inimitability of the Qurʾān (for sources, see n. 90 below). In other contexts, the expression has the sense ‘contradiction, contrariety’, which we shall encounter a little later.

⁸³ Reading, with al-Makkī, fa-qāla instead of fa-qālū.

⁸⁴ Al-Anṣārī, Ghunya, vol. 2, p. 30; cf. al-Makkī, Nihāyat al-marām, ff. 81a–b; the passage is not transmitted in Ibn al-Amīr's abridgement.

⁸⁵ Al-Anṣārī, as we have seen, refers to the book with the title al-Taqrīb wa-al-taqrīr, which I have not found elsewhere.

⁸⁶ Al-Bāqillānī, al-Taqrīb wa-al-irshād, ed. al-Ḥamīd, ʿAbdZunayd, ʿA. Abū, 3 vols. (Beirut, 1998)Google Scholar. The three books are listed in al-Qāḍī ʿIyāḍ's biographical entry on al-Bāqillānī (Tartīb al-madārik wa-taqrīb al-masālik, ed. Aʿrāb, Saʿīd A. et al. , 8 vols. [Rabat, 1981–1983], vol. 7, pp. 69–70Google Scholar).

⁸⁷ See n. 13 above.

⁸⁸ Al-Bāqillānī, Taqrīb, vol. 1, p. 202. By contrast, in the Tamhīd (pp. 13–14) he does not mention that evidence could be non-existent.

⁸⁹ Al-Juwaynī, Talkhīṣ, vol. 1, pp. 119–20.

⁹⁰ Al-Bāqillānī, Taqrīb, vol. 1, pp. 202–3; al-Juwaynī, Talkhīṣ, vol. 1, p. 119. On al-Bāqillānī's argument that the absence of emulation (muʿāraḍa) for a prophetic miracle is evidence for a prophet's veracity, see his Tamhīd (pp. 141 ff.) and, of course, Iʿjāz al-Qurʾān, a book devoted to the miraculousness of the Qurʾān. It should be noted, however, that this is an argument from silence that arguably has little in common with the arguments from ignorance employed in physical and metaphysical contexts. Al-Bāqillānī is careful to explain the import of the absence of any successful attempts to emulate the Qurʾān by the Prophet's contemporaries, and even addresses the counter-argument that such attempts may have occurred but were not transmitted (Tamhīd, pp. 147–8; Iʿjāz al-Qurʾān, ed. Ṣaqr, Aḥmad [Cairo, 1954]Google Scholar, pp. 33 ff.). Another argument from silence, encountered in jurisprudence, is that the occurrence of consensus (ijmāʿ) on a certain view at a given time is confirmed by the absence of reports of dissenting views (on this, see van Ess, Die Erkenntnislehre des ʿAḍudaddīn al-Īcī, p. 379).

⁹¹ Al-Jurjānī, Sharḥ al-Mawāqif, ed. al-Naʿsānī, Muḥammad B., 8 vols. (Cairo, 1907), vol. 5, p. 12.Google Scholar

⁹² Al-Juwaynī, Shāmil (ed. al-Nashshār), p. 387.

⁹³ Al-Malāḥimī, Muʿtamad, pp. 584–6.

⁹⁴ The argument from ignorance, hence, is not used in this context in: Shashdīw, Mankdīm, Sharḥ al-Uṣūl al-khamsa, ed. ʿUthmān, ʿAbd al-Karīm (Cairo, 1965)Google Scholar, pp. 277 ff. Al-Malāḥimī notes that ʿAbd al-Jabbār fails to substantiate his claim that the argument from the unknowns does not apply to the notion of a second pre-eternal being.

⁹⁵ See p. 194 above.

⁹⁶ Al-Juwaynī, Shāmil (ed. al-Nashshār), p. 387.

⁹⁷ Al-Juwaynī, Shāmil (ed. Frank), p. 69.

⁹⁸ Only real causes (so-called ʿilla ʿaqliyya) are intended here, to the exclusion of scriptural causes (ʿilla samʿiyya), which are employed in jurisprudence. On whether convertibility is a condition of the latter, see al-Juwaynī, Burhān, vol. 2, pp. 835 ff. The expression ‘ʿaqlī’ here is used in the sense ‘apprehended independently by the mind’, rather than ‘pertaining to the mind per se’, and contrasts with ‘samʿī’ or ‘naqlī’, ‘taught by scripture’.

⁹⁹ On the convertibility of causes and effects, see al-Juwaynī, Shāmil (ed. al-Nashshār), pp. 660 ff.

¹⁰⁰ Al-Juwaynī, Talkhīṣ, vol. 3, p. 223.

¹⁰¹ Al-Juwaynī, Shāmil (ed. Frank), p. 70.

¹⁰² Al-Juwaynī, Shāmil (ed. Frank), pp. 69–70.

¹⁰³ Al-Juwaynī, Burhān, vol. 1, p. 138.

¹⁰⁴ Al-Juwaynī, Burhān, vol. 2, pp. 848–9.

¹⁰⁵ Al-Juwaynī, Burhān, vol. 2, pp. 849–50; vol. 1, pp. 139–40.

¹⁰⁶ Al-Juwaynī, Burhān, vol. 2, p. 847.

¹⁰⁷ ʿAbdallāh, Abūal-Māzarī, Muḥammad ibn ʿAlī, Īḍāḥ al-maḥṣūl min Burhān al-uṣūl, ed. al-Ṭālibī, ʿAmmār (Beirut, 2001)Google Scholar, p. 116.

¹⁰⁸ Al-Rāzī was well-versed in al-Juwaynī's works. As a young student, he reportedly learned the Shāmil by heart (al-Ṣafadī, Khalīl ibn Aybak, al-Wāfī bi-al-wafayāt, ed. Dedering, Sven et al. , 29 vols. [Wiesbaden, 1931–2004], vol. 4, p. 249Google Scholar); and his father, and first teacher, Ḍiyāʾ al-Dīn al-Makkī was the student of al-Anṣārī, a prominent student of al-Juwaynī (see my introduction to al-Makkī's Nihāyat al-marām, and forthcoming article, ‘Al-Rāzī's earliest kalām work’, in Gregor Schwarb and Lukas Muehlethaler [eds.], Theological Rationalism in Medieval Islam: New Sources and Perspectives [Leuven, 2013]).

¹⁰⁹ I have argued elsewhere that al-Ghazālī's advocacy of logic in some contexts and genres does not mark the introduction of Aristotelian logic into kalām (Shihadeh, ‘From al-Ghazālī to al-Rāzī’, pp. 142 ff.). Although the subject is too complex to revisit here, I would briefly propose that this shift should be defined by at least some of the following four conditions: (a) crucially, a robust criticism of earlier kalām logic and epistemology, (b) an explicit endorsement, and ideally a description in outline, of an alternative logical system, and (c) a degree of commitment to it in practise. The last two features (d) need to be evident in an authoritative general kalām summa. Unlike al-Rāzī's Nihāyat al-ʿuqūl which fulfils all these conditions, al-Ghazālī's only substantial theological summa al-Iqtiṣād fī al-iʿtiqād fulfils none, except perhaps to an extremely limited extent. On the subject, see also: Rudolph, Ulrich, ‘Die Neubewertung der Logik durch al-Ġazālī’, in Perler, Dominik and Rudolph, Ulrich (eds.), Logik und Theologie: das Organon im arabischen und im lateinischen Mittelalter (Leiden and Boston, 2005), pp. 73–97Google Scholar.

¹¹⁰ For instance, al-Āmidī, Sayf al-Dīn, Abkār al-afkār fī uṣūl al-dīn, ed. al-Mahdī, Aḥmad M., 5 vols. (Cairo, 2002), vol. 1, pp. 208–10Google Scholar; al-Jurjānī, Sharḥ al-Mawāqif (with ʿAḍud al-Dīn al-Ījī's Mawāqif, and super-commentaries by Ḥasan Chalabī and ʿAbd al-Ḥakīm al-Siyālkūtī), vol. 2, pp. 20–7; cf. van Ess, Die Erkenntnislehre des ʿAḍudaddīn al-Īcī, pp. 376–80; see also § 3.4 below. Al-Rāzī also criticises the argument from ignorance briefly in other works (see n. 26 above).

¹¹¹ Al-Rāzī, Nihāyat al-ʿuqūl, ff. 2a–b; 4b–5a. On why I render ‘dalīl’ here as ‘argument’, see n. 7, above.

¹¹² Al-Rāzī, Nihāyat al-ʿuqūl, ff. 5a–10b; cf. Shihadeh, ‘From al-Ghazālī to al-Rāzī’, pp. 165–8.

¹¹³ On this type of argument, see n. 75, above.

¹¹⁴ Al-Rāzī, Nihāyat al-ʿuqūl, f. 5b.

¹¹⁵ It is well-established, both in medieval sources and in recent studies, that al-Rāzī engaged closely with the works of al-Malāḥimī and Abū al-Ḥusayn al-Baṣrī.

¹¹⁶ On this, see Shihadeh, ‘From al-Ghazālī to al-Rāzī’, p. 169.

¹¹⁷ Al-Rāzī, Nihāyat al-ʿuqūl, f. 5a.

¹¹⁸ See n. 128 below.

¹¹⁹ Al-Rāzī, Nihāyat al-ʿuqūl, f. 5b.

¹²⁰ Al-Rāzī, al-Tafsīr al-kabīr, 32 vols. in 16 (Beirut, 1981), vol. 2, pp. 30–1Google Scholar.

¹²¹ Al-Rāzī, Nihāyat al-ʿuqūl, ff. 5a–b.

¹²² See pp. 185–6 above.

¹²³ Al-Rāzī, Nihāyat al-ʿuqūl, ff. 5b–6a.

¹²⁴ See p. 186 above.

¹²⁵ Al-Rāzī, Nihāyat al-ʿuqūl, f. 6a.

¹²⁶ Al-Rāzī, Nihāyat al-ʿuqūl, f. 5b (argument); f. 6a (response).

¹²⁷ See p. 194 above.

¹²⁸ Following his response to the evidence adduced for the major premise in the argument from ignorance, al-Rāzī (Nihāyat al-ʿuqūl, ff. 6a–b) goes on to offer a counter-argument against this premise (component iv of the discussion; see p. 208 above). He argues that if the absence of evidence for affirmation constitutes evidence for negation then the absence of evidence for negation must constitute evidence for affirmation. If neither type of evidence is known for X, then either X must be at once affirmed and negated, which is absurd, or neither affirmed nor negated. The latter can only be conceivable in the availability of a third option, namely the suspension of judgement (tawaqquf).

¹²⁹ Al-Rāzī, Nihāyat al-ʿuqūl, f. 9b.

¹³⁰ See pp. 202–3 above.

¹³¹ Al-Rāzī, Nihāyat al-ʿuqūl, ff. 9b–10a.

¹³² Al-Rāzī, Nihāyat al-ʿuqūl, f. 7b.

¹³³ For publication details, see n. 12 above. The editor of the Kāmil identifies the author as Mukhtār ibn Maḥmūd al-Najrānī (or al-Baḥrānī), a seventh/thirteenth-century Muʿtazilī (see his introduction, pp. 10–44; cf. Elsayed Elshahed, Das Problem der transzendenten sinnlichen Wahrnehmung in der spätmu‘tazilitischen Erkenntnistheorie nach der Darstellung des Taqīaddīn an-Nağrānī [Berlin, 1983]). More recently, Hasan Ansari (in a note published on www.kateban.com in 2011) has identified him as a sixth/twelfth-century figure by the name of Ṣāʿid ibn Aḥmad.

¹³⁴ Al-ʿUjālī, Kāmil, pp. 59–60.

¹³⁵ Al-ʿUjālī, Kāmil, pp. 322–34.

¹³⁶ Al-ʿUjālī, Kāmil, pp. 322–3.

¹³⁷ Al-ʿUjālī, Kāmil, p. 60.

¹³⁸ Al-ʿUjālī, Kāmil, p. 323.

¹³⁹ Al-ʿUjālī, Kāmil, p. 323; cf. p. 208 above.

¹⁴⁰ Al-ʿUjālī, Kāmil, p. 323; cf. pp. 208–9 above.

¹⁴¹ Al-ʿUjālī, Kāmil, p. 324; cf. pp. 210–11 above.

¹⁴² Al-ʿUjālī, Kāmil, pp. 324–5; cf. pp. 182–3 above.

¹⁴³ Al-ʿUjālī, Kāmil, p. 325; cf. p. 212 above.

¹⁴⁴ Al-ʿUjālī, Kāmil, p. 325; cf. pp. 188–90 (objections 3 and 4) above.

¹⁴⁵ Al-ʿUjālī, Kāmil, pp. 325–6; cf. pp. 178–9 and 188–90 (objections 3 and 4) above.

¹⁴⁶ Al-ʿUjālī, Kāmil, pp. 326–7; cf. p. 209 above.

¹⁴⁷ Al-ʿUjālī, Kāmil, pp. 327–8; cf. pp. 189–90 (objection 5) above.

¹⁴⁸ Al-ʿUjālī, Kāmil, p. 328.

¹⁴⁹ Al-ʿUjālī, Kāmil, pp. 328–30; cf. pp. 210–11 above. That al-Rāzī's response here is grounded in, and references, his discussion of syllogistic logic in an earlier section constitutes further evidence that the Nihāya is a source for the Kāmil.

¹⁵⁰ Al-ʿUjālī, Kāmil, p. 330; cf. pp. 182–3 and 210 above.

¹⁵¹ Al-ʿUjālī, Kāmil, p. 330; cf. pp. 212–13 above.

¹⁵² Al-ʿUjālī, Kāmil, pp. 331–4; cf. n. 128 above.

¹⁵³ Khaldūn, Ibn, al-Muqaddima, ed. al-Shaddādī, ʿAbd al-Salām, 5 vols. (Casablanca, 2005), vol. 3, pp. 34–6Google Scholar; 95–7. The discussion in the section on logic is absent in Ibn Khaldūn's earlier version of the work, and hence in some published editions (e.g. al-Muqaddima [Cairo, 1327 AH], pp. 548–9; however, see pp. 519–20 for improvements on the 2005 edition in the section on kalām).

¹⁵⁴ Ibn Fūrak, Mujarrad, p. 304.

¹⁵⁵ This is partly connected to their views on uncritical imitation (taqlīd). On this, see Frank, Richard M., ‘Knowledge and taqlîd: the foundations of religious belief in classical Ashʿarism’, Journal of the American Oriental Society, 109 (1989): 37–62CrossRefGoogle Scholar.

¹⁵⁶ Al-Juwaynī, Shāmil (ed. al-Nashshār), p. 212.

¹⁵⁷ I am grateful to the two anonymous reviewers and to Dr Gregor Schwarb for their invaluable comments on this article.