Falãsifa. The Chapters on God s Knowledge in Khojazãda s and ʿAlãʾ al-dìn s Studies on al-ghazãlì s Tahãfut al- Introduction

Transcription

1 The Chapters on God s Knowledge in Khojazãda s and ʿAlãʾ al-dìn s Studies on al-ghazãlì s Tahãfut al- Falãsifa Lambertus W. C. (Eric) van Lit Introduction McGill University The Incoherence of the Philosophers Institute of Islamic Studies (Tahāfut al-falāsifa) is one of the Montreal, Quebec/Canada most important texts in the corpus of al-ghazālī (d. 505/1111), who tries to show in this book that even if one takes over the premises and vocabulary of the philosophers [and what is meant here are the Peripatetics such as al-fārābī (d. 339/950) and Ibn Sīnā (d. 428/1037)], the conclusions they put forward are untenable. al-ghazālī organized the book around twenty topics, of which he selected three that were in his view diametrically opposed to religious doctrine (in the other cases, the philosophers opinion was wrong but could be pardoned). These three are the ideas of the philosophers that the universe was not created but is pre-eternal, that there is no bodily resurrection, and that God does not know particulars inasmuch as they are particulars. Although the Incoherence is nowadays held to be one of the most important products of medieval Islamic philosophy, we know of few direct responses to this text, so the scope of its explicit influence is hard to establish. There is the refutation by Ibn Rushd (d. 595/1198) under the appealing title The Incoherence of the Incoherence (Tahāfut al-tahāfut), afterwards translated into Latin (Destructio Destructionum), which has received much scholarly attention from modern Western scholars. 1 Yet the importance and influence of Ibn 1 al-ghazālī s Incoherence has been subject to many studies by modern scholars. One of the first was Boer, Tjitze de, Die Widersprüche der Philosophie nach al-gazzālī und ihr Ausgleich durch Ibn Rošd (Inauguraldissertation), K. J. Trübner, Strassburg 1894, 56 ff.,

2 Rushd s book lies within the Scholastic tradition, not within the subsequent development of Islamic philosophy. However, Ottoman scholars from the fifteenth and sixteenth centuries did occupy themselves with al-ghazālī s book. The first two works, by Khojazāda (d. 894/1488) and ʿAlāʾ al-dīn al-ṭūsī (d. 886/1482), were written on the order of Sultan Mehmed II as an intellectual duel. 2 The text of the winner, Khojazāda, became the source of a handful of super-commentaries. This later Ottoman commentary tradition has so far received only a passing mention by modern-day scholars, for example Louis Gardet. He writes in the Encyclopaedia of Islam that the work of the Turk Khojazāde (9 th /15 th century) [...] sought to refute the Tahāfut al-tahāfut of Ibn Rushd. 3 As is clear from this, the full context of Khojazāda s text (as one side of an intellectual challenge and source of super-commentaries) was not understood. More importantly, the reference to Ibn Rushd is incorrect, as it will become clear from this paper that there is no evidence that the Ottoman scholars knew about Ibn Rushd s text. To begin to understand the nature and scope of these Ottoman commentaries, this paper will look into the issue of God s knowledge as it is presented in the two treatises of Khojazāda and ʿAlāʾ al-dīn. First, an introduction to the philosophical issues will be provided. Then we will move on to discuss the exposition of Khojazāda and ʿAlāʾ al-dīn al-ṭūsī on it and discuss some arguments they put forward in detail. As al-ghazālī (and the Ottoman scholars following him) divides the issue into three different discussions ( On God s knowledge of other things, On God s self-knowledge, and On God s knowledge of particulars ), we will likewise follow this order. I. The Problem and Its History From the Qurʾān and the ḥadīth literature it is clear that God knows human beings and all other things. Examples are abundant, but among the most 2 3 another important contribution is Bergh, Simon van den (trans.), Averroes Tahāfut al- Tahāfut (translated from the Arabic with introduction and notes), Luzac & Company Ltd., London 1969, I, 255 ff., and II, 144 ff. Evidently, previous scholarship studied al-ghazālī s book almost solely in the light of its commentary by Ibn Rushd. Kātib Chalabī, Ḥājī Khalīfa Muṣṭafá ibn ʿAbd Allāh, Kashf al-ẓunūn ʿan Asāmī l-kutub wa l-funūn, Dār al-kutub al-ʿilmiyya, Beirut 1992, I, 513 [= entry on Tahāfut al-falāsifa]. Gardet, Louis, ʿIlm al-kalām, EI 2, III, 1149 a.

3 cited verses are He is knower of all things 4, and... Not an atom s weight, or less than that or greater, escapeth Him in the heavens or in the earth. 5 The mention of God s knowledge may need some explanation. One way of understanding it is that the God of scripture is a caring God (al-raḥmān), one who personally intervenes in the affairs of humanity and who will decide on a human s fate after death based on the person s actions here on earth. Knowledge of our affairs is therefore of the utmost importance. In the Christian tradition, philosophical discussions around the notion of God s knowledge would revolve for the most part around the seemingly incompatibility of His omniscience with other characteristics of Him (such as His omnipotence) and with human free will. 6 If the Christian doctrine of salvation through the cross requires people to decide for themselves whether to accept or refuse Christianity, is God capable of knowing what someone will decide? If so, that person turned out to have no real say in his affairs after all, for his actions are already determined by what God knows will happen. If not, then God s knowledge is not all encompassing, which goes against the standard view of God Translations are from Pickthall, Marmaduke William (trans.), The Meaning of the Glorious Qur ān, Muslim World League, New York See Q 2:29, Q 2:231, Q 2:282, Q 4:32, Q 4:176, Q 5:97, Q 6:101, Q 8:75, Q 9:115, Q 24:35, Q24:64, Q 29:62, Q 33:40, Q 33:54, Q 42:12, Q 48:26, Q 49:16, Q 57:3, Q 58:7, Q 64:11. E.g. cited by Taftāzānī, Saʿd al-dīn Masʿūd ibn ʿUmar ibn ʿAbd Allāh, Sharḥ al-maqāṣid, ʿĀlam al-kutub, Beirut 1989, IV, 118 [= maqṣad 5, faṣl 3, mabḥath 3]. Q 34:3, slighty different in Q 10:61. E.g. cited by Ghazālī, Abū Ḥāmid Muḥammad ibn Muḥammad, Iḥyāʾ ʿUlūm al-dīn, Dār al-kutub al-ʿilmiyya, Beirut 2004, 92 [= Kitāb Qawāʿid al-ʿaqāʾid, faṣl 1]; Ibn Sīnā, Abū ʿAlī Ḥusayn ibn ʿAbd Allāh ibn ʿAlī (Avicenna), The Metaphysics of The Healing (translated, introduced and annotated by Michael E. Marmura), Brigham Young University Press, Provo, Utah 2005, 288 [= al-ilāhiyyāt min al- Shifāʾ, Bk. 8, Ch. 6]; Naṣīr al-dīn al-ṭūsī, Abū Jaʿfar Muḥammad ibn Muḥammad ibn Ḥasan, Sharḥ al-ishārāt wa l-tanbīhāt (along with Ibn Sīnā s al-ishārāt wa l-tanbīhāt, ed. S. Dunyā), Dār al-maʿārif, Cairo 1957, III-IV, 716 [= namaṭ 7, faṣl 17]; Taftāzānī, Sharḥ al-maqāṣid, IV, 118 [= maqṣad 5, faṣl 3, mabḥath 3]. See e.g. Craig, William Lane, The Problem of Divine Foreknowledge and Future Contingents from Aristotle to Suarez, E. J. Brill, Leiden 1988, or Rudavsky, Tamar. (ed.), Divine Omniscience and Omnipotence in Medieval Philosophy: Islamic, Jewish and Christian Perspectives, D. Reidel Publishing, Dordrecht That this issue is not a problem for ancient times alone may be clear from the fact that to this date the compatibility of omniscience and free will for humans or other attributes of

4 I.1. Philosophers Among the philosophers, from Antiquity on, the discussions attempted to infuse Aristotelian epistemology with a Neoplatonic cosmology. The discussions on epistemology led to the idea that knowledge is properly called intellection if it is demonstrative, which in turn means it is based solely on universal truths which always hold true. 8 Neoplatonic cosmology added to the discussion the emphatic distinction between form and matter, 9 which led to the belief that God and the celestial entities are abstract from matter and therefore called intellects. This last remark is in fact an important development in Islamic philosophy. While Plotinus would argue that God is above thinking, the Arabic translations of Plotinus do not show this but rather normalize God s position as the One (beyond all other beings and therefore unique and incomparable to other beings) to the level of intellects (as one of many beings abstract from matter). 10 This normalization led the way for philosophers to describe and argue about God in a broader sense, applying notions to God that were before only applied to beings other than God. Specifically for knowledge, it meant that it was quite easy to argue that God held intellectual knowledge, for intellection could be seen as the acquisition of the form of the thing we want to know, which means the forms must be abstracted from the thing we want to know. Between God (an abstract thing) and all the abstract forms lies no impediment, and therefore God has full intellectual knowledge. A remaining issue is however what it means to have intellectual knowledge. If intellectual knowledge is intrinsically related to universals, it suggests that other types of knowledge deal with particulars (such as sense perception). This differentiation is already present in Aristotle, who explains in his Prior Analytics: Things are prior and more familiar in two ways; for it is not the same to be prior by nature and prior in relation to us, nor to be more familiar God is still receiving attention by philosophers and theologians. Cf. e.g. Kenny, Anthony, The God of the Philosophers, Clarendon Press, Oxford A very good exposition of this, together with an analysis of its reception by al-fārābī, Yaḥyá ibn c Adī and Abū l-faraj ibn al-ṭayyib can be found in Adamson, Peter, Knowledge of Universals and Particulars in the Baghdad School, Documenti e Studi sulla Tradizione Filosofica Medievale 18 (2007), E.g. Plotinus, Enneads, V.9.12 D Ancona, Cristina, Divine and Human Knowledge in the Plotiniana Arabica, The Perennial Tradition of Neoplatonism (ed. J. J. Cleary), Leuven University Press, Leuven 1997,

5 and more familiar to us. I call prior and more familiar in relation to us what is nearer to perception, prior and more familiar simpliciter what is further away. What is most universal is furthest away, and the particulars are nearest; and these are opposite to each other. 11 It is in this sense that Aristotle could claim that God is thought thinking itself. 12 Many different ideas were provided on the question whether God can have those other types too and if not, whether this jeopardizes the characterization of God as omniscient. Thus, the discussion on God s knowledge focused for the most part on the more specific discussion on God s knowledge of particulars. Ibn Sīnā If we allow ourselves to jump to the factual starting point for understanding al-ghazālī s, Khojazāda s and ʿAlāʾ al-dīn s texts, we have to look at the ideas of Ibn Sīnā on the topic. Ibn Sīnā s solution to this problem influenced many scholars after him. Ibn Sīnā argued that God knows particulars in a universal way, devoid of bodily perception or temporality. 13 To understand this more fully, we have to touch on some points of Ibn Sīnā s argumentation. Firstly, he argues for God s knowledge of other things by use of God s self-knowledge. In his Book of Scientific Knowledge (Dānishnāma-i ʿAlāʾī) he first notes that part of what makes God Himself is that he is the existentiator (hastī dah) of things according to the order in which they exist. 14 Knowing Himself, God therefore also must have knowledge of Himself being this existentiator. Knowledge of what is existentiated must then be included, for otherwise He would not know exactly for what He is an existentiator. Here Ibn Sīnā must draw a sharp difference with the usual way we acquire knowledge; God s knowledge does not depend on the different things he b34-72a5, Aristotle, The Complete Works of Aristotle (ed. by J. Barnes), Revised Oxford Translation, Princeton University Press, Princeton 1984, b20, Aristotle, The Complete Works of Aristotle, Ibn Sīnā, The Metaphysics of The Healing, 288 [= al-ilāhiyyāt min al-shifāʾ, Bk. 8, Ch. 6], Morewedge, Parviz, The Metaphysica of Avicenna (Ibn Sīnā): A Critical Translation- Commentary and Analysis of the Fundamental Arguments in Avicenna s Metaphysica in the Dānish nāma-i ʿAlāʾī (The Book of Scientific Knowledge), Routledge & Kegan Paul, London 1973, 65 [= Dānishnāma-i ʿAlāʾī, Ch. 32], Ibn Sīnā, al-ishārāt wa l-tanbīhāt, III- IV, 717 [= namaṭ 7, faṣl 18]. Morewedge, P., The Metaphysica of Avicenna, 61. Cf. Ibn Sīnā, al-ishārāt wa l-tanbīhāt, III-IV, 709 [= namaṭ 7, faṣl 15].

6 existentiates. On the contrary, as Ibn Sīnā puts it, its knowledge is the cause for the existence of all things. 15 In the Healing (al-shifāʾ) he expands this argument and explains that dependency means that... either His essence would be constituted by what He intellectually apprehends [...] or it would accidentally occur to Him to intellectually apprehend. 16 Either option is unacceptable, for God is not without reason named the Necessary Existent throughout Ibn Sīnā s Healing. For him, the concept of God is most strongly associated with the notion of complete necessity and therefore a complete negation of accidentality and dependency. It is thus by knowing Himself that He knows the cause of all other things and, by extension, therefore knows all these effects too. What makes him coin the phrase knowledge of particulars in a universal way is the consideration of God s knowledge of corruptible things. In his Healing he claims: When corruptibles are intellectually apprehended (ʿuqilat) in terms of the quiddity denuded [of matter] and the things that attach to it that are not individualized, they are not intellectually apprehended inasmuch as they are corruptible. If apprehended (udrikat) inasmuch as they are connected with matter and the accidents of matter, with a [particular] time and individuation, they would not be intellectually apprehended but would be sensed or imagined. 17 It thus seems that Ibn Sīnā on the one hand wants to distinguish between intellection and apprehension, while, on the other hand, he wants to maintain that this does not mean that God misses out on objects of knowledge. Interpretations as to what exactly Ibn Sīnā means and what the implications are for such a theory are many and varied and we will not occupy ourselves with all of them. 18 Instead, we will focus on ʿAlāʾ al-dīn s and Morewedge, P. The Metaphysica of Avicenna, 61. Ibn Sīnā, The Metaphysics of The Healing, 287. Translation adapted. Ibn Sīnā, The Metaphysics of The Healing, 287. Cf. Ibn Sīnā, al-ishārāt wa l-tanbīhāt, III- IV, [= namaṭ 7, faṣl 18]. For some modern interpretations on this see e.g. Marmura, M. E., Some Aspects of Avicenna s Theory of God s Knowledge of Particulars, Journal of the American Oriental Society 82 (1962), ; Adamson, P., On Knowledge of Particulars, Proceedings of the Aristotelian Society, CV/1 (2005), , and Belo, Catarina, Averroes on God s Knowledge of Particulars, Journal of Islamic Studies, XVII/2 (2006), (the first 11 pages are an exposition of Ibn Sīnā s theory).

7 Khojazāda s study of al-ghazālī s challenge to Ibn Sīnā s theory of God s knowledge from his famous Tahāfut al-falāsifa. 19 With this background information we can now turn towards the texts themselves and see how exactly God s knowledge is an issue. The discussion is divided into three parts which reflect Ibn Sīnā s composition of the problem. First, God s knowledge in an unconditional way is discussed. Then God s knowledge of Himself is discussed (as this plays an important role in this theory of God s knowledge). Lastly, God s knowledge of particulars is discussed. II. God s Knowledge of Other Things In this first of three chapters, al-ghazālī s text first puts forward al- Ghazālī s own view on the matter, namely, that God creates all things willingly. With will comes knowledge, and therefore God knows all things. He explains that because the philosophers deny will from God, they have to argue for it in a different way. al-ghazālī proposes two main theses the philosophers hold. The first is that because God is an abstract thing, He is a pure intellect, and a pure intellect has all other abstract things laid bare to it. He therefore knows all intelligibles (maʿqūlāt). The second argument is that even if God does not have a will, His activity still implies knowledge of the effects. A third, minor argument is an argument from perfection (which states that whatever is perfect can properly be ascribed to God). It argues that if other beings than God (like us) know things, then it is proper to say that God should know them too. We will see that these three arguments (the argument from being abstract, the argument from causality, and the argument from nobility) will resurface in the texts of 19 I have made use of: Ghazālī, transl. by M. E. Marmura, The Incoherence of the Philosophers: Tahāfut al-falāsifa, Brigham Young University Press, Provo, Utah 1997, [= Ch ]. Ṭūsī, ʿAlā al-dīn ʿAlī ibn Muḥammad, Tahāfut al-falāsifa (ed. by Riḍā Saʿāda), al-dār al-ʿālamiyya, Beirut 1981, [= Ch ]; older print edition is: Ṭūsī, ʿAlāʾ al-dīn, Kitāb al-dhakhīra, Maṭbaʿat Majlis Dāʾirat al-maʿārif al-niẓāmiyya, Hyderabad 1899, Khojazāda s text is printed in: Khojazāda, Tahāfut al-falāsifa, al-maṭbaʿat al-aʿlāmiyya, Egypt / , [= faṣl [sic], faṣl 13, and faṣl 14 ], and in: Ghazālī, Ibn Rushd & Khojazāda, Tahāfut al-falāsifa, 2 vols., printed at the expense of Muṣṭafá al-bābī al-ḥalabī and his brothers in Egypt (1321/1903), II, [= faṣl [sic], faṣl 13, and faṣl 14 ]. Nevertheless, I made use of several manuscripts: Millet Kütüphanesi (Millet Library, Istanbul), Feyzullah Efendi, MS 1182; Beyazıt Devlet Kütüphanesi (Beyazıt Devlet Library, Istanbul), MS 1990; and Süleymaniye Kütüphanesi (Süleymaniye Library, Istanbul), Şehid Ali Paşa, MS 1583 (= probably autograph).

8 Khojazāda and ʿAlāʾ al-dīn al-ṭūsī. They constitute the core of the discussion, on which the other two chapters are built. For al-ghazālī, the first argument (from being abstract) revolves mostly around what intellect means. It cannot mean something that intellectually apprehends all other things, for that is the desired conclusion and can thus not be part of the presuppositions. Of course, the usual response of the philosophers is that the condition of intellection is being free from matter. Only matter is the impediment to intellection. Against this, al-ghazālī responds that there might be another impediment next to matter; it is not proven that matter carries this impeding force alone. The second argument (from causality) is countered by distinguishing between a voluntary act and a natural act. Knowledge is only implied by the first, while al-ghazālī argues that the philosophers can only maintain that God s activity is of the second category. Even if knowledge of the effect is accepted, God would only know the first effect. This is because knowledge of the effect doesn t necessitate knowledge of the effect s effect, al-ghazālī argues. For example, one can knowingly push a stone from the top of a hill, but this doesn t imply knowledge that the stone will hit and break another one at the bottom of the hill. The minor argument (from nobility) is refuted ad absurdum. At first, al- Ghazālī elaborates the idea of the philosophers that sometimes absence of something (such as perception by bodily organs) is in fact more perfect than the presence of it. He then pushes it to the limit by arguing that likewise it would be more perfect for God not to know other things, as they are merely imperfect beings and there would therefore be no reason for God to know them. II.1. The Ottoman Scholars A common difference of all three chapters of Khojazāda and ʿAlāʾ al- Dīn al-ṭūsī against al-ghazālī s text is the formalization of the philosophical arguments. Although formally the texts of the Ottoman scholars are studies (if not commentaries) on al-ghazālī s text, they incorporate more material from the post-ghazālian period than from al-ghazālī s text itself. This is justified by a greater clarity of expression and a philosophically more rigorous argumentation. In fact, while al-ghazālī is merely out to show the incoherence of the philosophical argumentation, the Ottoman scholars seem to leave room open for one or the other argument.

9 Another striking commonality is the organization and presentation of the philosophers arguments for God s knowledge and the subsequent discussion of it. Both texts run almost exactly the same in order of arguments and frequently bring up the same counter-arguments. However, they do still follow the agenda of al-ghazālī s exposition quite strictly over the span of the three chapters, 20 though with some minor modifications. In this first chapter, ʿAlāʾ al-dīn, especially, makes some editorial decisions. He postpones large parts of the second argument and the complete third argument to the second chapter. Khojazāda doesn t do this. He covers all three arguments in this chapter and then builds the two subsequent chapters in extension to this order. Unlike ʿAlāʾ al-dīn, he also closes the first chapter with a critical evaluation of al-ghazālī s text, and includes in the third chapter again some critical notes on al-ghazālī s text. ʿAlāʾ al-dīn s text lacks a direct critical evaluation of al- Ghazālī s arguments. II.1.1. The First Argument For the first argument, both scholars put forward an exceedingly more sophisticated explanation than al-ghazālī. In argumentation, the scholars hardly differ, although they do offer different objections. For elaborating the philosophical argument, Khojazāda cites (without mention of the source) 21 al- Iṣfahānī s exposition of it from his commentary on al-bayḍāwī s Ṭawāliʿ al- Anwār min Maṭāliʿ al-anẓār. 22 It seems probable that Khojazāda also read Fakhr al-dīn al-rāzī s commentary on Ibn Sīnā s Ishārāt, as the few lines he For example, considering the reliance they show on al-jurjānī s commentary on al-ījī s Kitāb al-mawāqif, they also could have followed the order of al-jurjānī s passage on God s knowledge. Using but not mentioning a source is something he repeatedly does in these chapters. This may have been common practice in his time. For a similar practice from a contemporary, Cf. ʿAbd al-raḥmān Jāmī, Abū l-barakāt Nūr al-dīn Mullā ʿAbd al-raḥmān ibn Aḥmad ibn Muḥammad al-shīrāzī, The Precious Pearl: al-jāmī s al-durrah al-fākhirah with the Commentary of ʿAbd al-ghafūr al-lārī (translated with an introduction by Nicholas L. Heer), State University of New York Press, Albany Iṣfahānī, Abū l-thanāʾ Shams al-dīn Maḥmūd ibn ʿAbd al-raḥmān ibn Aḥmad, Maṭāliʿ al- Anẓār ʿalá Ṭawāliʿ al-anwār, Hulusi Efendi Matbaası, Istanbul 1305/ , [= kitāb 2, bāb 2, faṣl 1, mabḥath 2], cf. Calverley, Edwin E. & Pollock, James W. (ed. and trans), Nature, Man and God in Medieval Islam: ʿAbd Allah Baydawi s Text, Tawaliʿ al- Anwar min Mataliʿ al-anzar, along with Mahmud Isfahani s Commentary, Mataliʿ al- Anzar, Sharh Tawaliʿ al-anwar, Brill, Leiden 2002, II,

10 writes which are not taken from al-iṣfahānī contain a citation (again without mention) of Fakhr al-dīn s text. 23 It should be noted that al-iṣfahānī s text seems to have set the standard long before Khojazāda, as bits and pieces of the passage Khojazāda is citing (and which are not to be found directly in Fakhr al- Dīn s text) can also be found in, for example, al-jurjānī s commentary on al-ījī s Mawāqif, and al-taftāzānī s commentary on his own Maqāṣid. 24 In short, the argument plays on the differences between intellect (ʿaql), intellectual apprehender (ʿāqil) and intelligible (maʿqūl). First it is argued that all abstract things are intelligibles. Second, if an intelligible has self-subsistence, it is an intellectual apprehender for all other intelligibles. By conclusion, since God is abstract and has self-subsistence, He knows all intelligibles. To string these points together, they use a definition for knowledge as the connection (muqārana) of an intelligible to an abstract, self-subsisting thing. It is therefore sufficient to prove that such a connection occurs to all intelligibles. For this, two intelligibles are first considered. Any two intelligibles can always be conceived together, that is to say, be thought of in one proposition (for all x and y, x is y or x is not y is always true). This conceiving together constitutes a connection in the intellect between the two intelligibles. Obviously the connection in the intellect between the two intelligibles is more specific than the connection in general between the two intelligibles. Thus, by establishing the possibility of the connection in the intellect between any two intelligibles, one proves the possibility in general between any two intelligibles. Now we see why the definition of intellection is said to be the connection of an intelligible to an abstract, self-subsisting thing. For if one of the two intelligibles turns out to be self-subsistent, such a connection can be actualized, using the self-subsistent thing as a locus for the other intelligible to enter into as a state. But if this is possible, this intellection Rāzī, Abū ʿAbd Allāh Fakhr al-dīn Muḥammad ibn ʿUmar ibn Ḥusayn, Sharḥ al-ishārāt wa l-tanbīhāt (ed. ʿAlī Ridā Najafzāda), Anjuman-i Āthār wa Mafākhir-i Farhangī, Tehran 1383/2004, II, 299 [= namaṭ 3, faṣl 19]. The sentence under discussion reads: ن ر إ إ ج ن ا ذ ن, see Iṣfahānī, ; Ījī, Abū l-faḍl ʿAḍud al-dīn ʿAbd al- Raḥmān ibn Aḥmad & Jurjānī, Abū l-ḥasan al-sayyid al-sharīf ʿAlī ibn Muḥammad ibn ʿAlī, Kitāb al-mawāqif & Sharḥ al-mawāqif (ed. ʿAbd al-raḥmān ʿUmayra), Dār al-jīl, Beirut 1997, II, 695 [= mawqif 4, marṣad 4, maqṣad 3, 6 th ]; Taftāzānī, Sharḥ al-maqāṣid, IV, 114 [= maqṣad 5, faṣl 3, mabḥath 3].

11 must occur eternally. For if it would only occur after some time it would constitute a change to the intelligible, which is contrary to its being as change is connected with matter and its concomitants, whereas an intelligible is abstract from matter. In conclusion, as God is an abstract, self-subsisting thing, all of this applies to God and God therefore knows all intelligibles. After the exposition of the philosophers argument, the two Ottoman scholars both discuss an extensive list of objections to the argument. Although they make mention of al-ghazālī s argument that it is unclear why matter should be the only impediment, their main train of thought consists in trying to attack the argumentation that when a connection between two intelligibles is possible in an intellect, it must be possible in general. Both scholars argue that such a connection can still be conditioned on the mental existence of the intelligibles. As Khojazāda puts it, an intelligible quiddity is not absolutely abstract, for it is still predicated with (malḥūqa) mental existence, so it is possible that that is a condition for the possibility of a connection. Khojazāda relates this argument back to what he sees as Ibn Sīnā s view of intellection being the acquisition (ḥuṣūl) of a form to an intellect. He mentions that this conclusion is incorrect according to most philosophers but doesn t go into detail on the view of these philosophers. It seems that what Khojazāda is relating here is the idea (most prominently worked out and defended by Fakhr al-dīn al-rāzī) that knowledge is not the imprinting of a form, but rather a relation (or an attribute holding a relation). 25 II.1.2. The Second Argument The second argument is again argued for by both Ottoman scholars in a very similar fashion. As with the first argument, ʿAlāʾ al-dīn rephrases the argument to such an extent that it is hard to say exactly which sources he used. However Khojazāda s argumentation can be found in the writings of al-sayyid al-sharīf and Quṭb al-dīn al-rāzī. 26 The argumentation is in two steps, the first one dealing with God s self-knowledge and the second one dealing with the connection of cause and effect in terms of knowledge. 25 c Alāʾ al-dīn eludes to this concept of knowledge on numerous occasions in this part of his study on al-ghazālī s Tahāfut and it therefore seems reasonable to consider this to be his own point of view. 26 Ījī & Jurjānī, Kitāb al-mawāqif & Sharḥ al-mawāqif, III, 102 [= mawqif 5, marṣad 4, maqṣad 3, baḥth 1].

12 In this argument, a slightly different definition of knowledge is used. Here it is not said to be the connection but the the presence (ḥuḍūr) of an abstract thing to an abstract, self-subsisting thing. As God is an abstract, selfsubsisting thing, He is therefore both the first and the second term of the definition; He is present to Himself. Furthermore, knowledge of the cause implies knowledge of the effect. As God is the cause for all other things and has knowledge of Himself, He knows all other things. Two major objections are presented by the two Ottoman scholars. First it is argued that this presence must be between two distinctly different things. Philosophers may argue that a conceptual difference (being an intelligible and being an intellectual apprehender) is enough, but as Khojazāda puts it, drawing from al-jurjānī this is only sufficient in terms of conceptuality, not in terms of the thing itself (nafs al-amr). And the thing itself was under discussion, so the response is not good enough. 27 ʿAlāʾ al-dīn includes a similar objection, albeit in a less negative fashion. He argues that a conceptual difference would be enough for his definition of intellection that is, Fakhr al-dīn s definition of knowledge as a relation, but would not be enough for the definition his philosophical opponents use, as they place knowledge in the Aristotelian category of quality. Second, they argue (Khojazāda heavily relying on Quṭb al-dīn) that knowledge of the cause does not necessarily imply knowledge of all effects, that is, of the effects of the effects. ʿAlāʾ al-dīn argues for this by stating that it is not the normal use of the concept of knowledge to include the effects of the effects ad infinitum. Khojazāda gives a different argumentation. He points to the difference between mental and external existence, and proposes that although causality between two things in the external world may be necessary (that is, if the cause exists, the effect must also exist), this does not imply that a similar causality exists between their forms. His argument is thus based on the differences between the two modes of existence, external and mental Ījī & Jurjānī, Kitāb al-mawāqif & Sharḥ al-mawāqif, III, 105 [= mawqif 5, marṣad 4, maqṣad 3, baḥth 2, 1 st ]. Rāzī, Abū ʿAbd Allāh Quṭb al-dīn Muḥammad ibn Muḥammad al-taḥtānī, al-ilāhiyyāt min al-muḥākamāt bayna Sharḥay al-ishārāt (along with Mīrzā Ḥabīb Allāh al-fāḍil al- Bāghnawī s Ḥāshiyat al-muḥākamāt, ed. Muḥammad Majīd Hādīzāda), Mīrāth-i Maktūb, Tehran 1381/2002, 392 [= namaṭ 7, faṣl 15].

13 ʿAlāʾ al-dīn al-ṭūsī proposes one more objection, which is that even if this argument is accepted, it would follow that God knows particulars, not universals, for the effects of God s activity are particulars. But the philosophers deny this to God s knowledge, so an inconsistency arises. Khojazāda offers a similar argument, though not as direct as ʿAlāʾ al-dīn and moreover not here but in the third chapter. There he comments that excluding particulars from God s knowledge is in contradiction with what the second argument tries to establish. He explains that complete knowledge of the particularity of the cause (khuṣūṣīyyat al-ʿilla) requires complete knowledge of the particularity of the effect. It is questionable how well this objection holds. If Ibn Sīnā and his philosophical followers hold that God knows particulars in a universal way, this objection makes it sound as though this means that God knows only universals. But as Ibn Sīnā relies for the most part exactly on this argument from causality for God s knowledge, it seems that Ibn Sīnā would agree with the two Ottoman scholars that this argument implies that God has knowledge of particulars. Indeed, for Ibn Sīnā it seems to be the very function of this argument to do establish that. II.1.3. The Third Argument The way Khojazāda and ʿAlāʾ al-dīn present the third argument isn t very different from the way al-ghazālī explained it. This argument from perfection states that whatever is perfect can properly be ascribed to God. This is because being a perfection means that it is better to be applicable than not to an existing thing. It is further argued that if other beings than God (like us) know things, it is proper to say that God should know them too (otherwise these things would be more perfect than God). Further, whatever is properly possible for God is necessarily so. If not so, a contingent aspect would be present to God, which would constitute a multiplicity, which is impossible. Khojazāda raises three distinct objections. ʿAlāʾ al-dīn makes similar objections, though only distinguishing two objections. Khojazāda s first objection is that their definition of absolute perfection is wrong. It is not that it is better to be applicable than not, but it is perfect in all aspects (of a thing), not being imperfect in some other sense (i.e. not constituting an imperfection in some sense for the thing it is ascribed to). ʿAlāʾ al-dīn points to something similar. He argues that absence of knowledge in an absolute sense doesn t need to be ignorance. Only in regards to something will it be ignorance.

14 The second objection is that a contingent aspect that is, the part of the argument where it is argued that whatever can be applicable to God must be applicable to God is ambiguous. In reference to God s essence it is indeed impossible, according to Khojazāda, but a contingent aspect is conceivable when it refers to His accidents. This objection is related to the third objection raised by Khojazāda. This third objection is that likewise a multiplicity which would arise if some part of God would be merely possible is ambiguous. Again, in regard to His essence it is indeed deemed impossible by Khojazāda but in regard to conceptualizations and aspects it is possible. ʿAlāʾ al-dīn creates an argument that closely resembles the second and third objection of Khojazāda. He argues that if the philosophers want this argument to be valid they should also admit sight, hearing, and the like as attributes of God. As the philosophers don t do this, an inconsistency occurs. 29 II.1.4. Khojazāda s Critique on al-ghazālī Khojazāda ends the chapter with a critique of al-ghazālī. For example, where al-ghazālī claims that according to the philosophers God acts out of compulsion, he argues that: The argument that the emanation of the world is by God s nature and compulsion (bi-ṭabʿ wa l-iḍṭirār), not by way of will and volition (bi- ṭarīq al-irāda wa l-ikhtiyār), is not as it should be according to the philosophers. They don t say that God s activity is as a forced activity as is proper to those who have a nature and a body. Rather, they argue that God has power (qādir), meaning that if He wishes to, He does, and if He does not wish to, He does not. However the wish of the action is inherent to His essence and absence of the wish is impossible. Likewise he argues that the philosophers, in the second argument, don t say that God s knowledge is owing to His knowledge of His effective causation, but His knowledge of the complete cause. 30 al-ghazālī thereby did not represent the argument of the philosophers correctly. As an effect of this, the example al-ghazālī gives (of knowingly pushing a stone down the hill which according to al-ghazālī does not imply knowledge of the stone hitting This argument can also be found in al-ghazālī s text, in the second chapter. The complete cause (al- c illa al-tāmma) probably refers to the combination of all four Aristotelian causes, consisting next to the effective cause (al- c illa al-fā c iliyya) of the material cause, formal cause, and final cause. Cf. Jurjānī, al-ta c rīfāt, Dār al-kutub alc Ilmiyya, Beirut 2000, 157 [= #1243]

15 another stone at the bottom of the hill) is wrong because not all causes were taken into account, just the effective cause. If one were to take all aspects into account (such as the slope of the hill or the amount of force with which the stone was pushed) one would be able to know that the stone was to hit another stone and break it. It is important to note that Khojazāda s critique seems primarily undertaken from the perspective of the philosophers. It is not obvious whether Khojazāda accepts himself these objections to al-ghazālī s text as true. To show that al-ghazālī did not represent the philosophers well and to show that al- Ghazālī s counter-arguments are rhetorical seems to be the primary reason for Khojazāda to relate these objections to al-ghazālī s text. III. God s Knowledge of Himself As we saw already in the second argument for God s knowledge, God s self-knowledge plays a vital part in the discussion on God s knowledge. Therefore, the second chapter 31 is completely devoted to it. In this chapter, al- Ghazālī proposes three arguments for the denial of the philosophers that God knows Himself. The first argument is that if will is disallowed for God, knowledge can t be argued for either. If his activity is a forced activity, such as the sun emitting light, self-knowledge is not implied at all. Secondly, he adds that being an abstract thing doesn t guarantee that it is an intelligible, as was already discussed in the previous chapter. Lastly, an argument from nobility doesn t demonstrate God s self-knowledge either. As a counter-example al- Ghazālī gives the division of things into seeing or blind. Clearly, seeing is nobler, and should by analogy of the argument about knowledge be applied to God. However, the philosophers insist that God does not see, and therefore al- Ghazālī claims that philosophers can likewise not claim God s knowledge using the argument from nobility (we came across this argument as an objection raised by ʿAlāʾ al-dīn against the third argument for God s knowledge). III.1. The Ottoman Scholars Khojazāda and ʿAlāʾ al-dīn al-ṭūsī do not follow al-ghazālī s text here, but rather discuss philosophical arguments why God should have selfknowledge. As before, Khojazāda s main source is al-jurjānī s commentary on 31 Equal to chapter 12 in al-ghazālī s and ʿAlāʾ al-dīn s text and equal to chapter 14 in Khojazāda s text.

16 al-ījī s Mawāqif, 32 although the influence of Fakhr al-dīn s commentary on the Ishārāt is also textually evident. 33 He proposes two arguments, albeit in a very brief fashion and with little critical evaluation of them. This he leaves to the first and third chapter. ʿAlāʾ al-dīn includes the same two arguments, but treats them in a fuller fashion and also includes the argument from nobility in this chapter (applying it both to knowledge in general and self-knowledge). III.1.1. The First Argument The first argument is connected with the first argument from the previous discussion. It relates that since God knows other things, it is therefore even more likely that He knows Himself. This is because He is able to know that He knows the other thing, and this knowledge (the proposition I know x ) includes the self (the I ). Since the comprehension of a relationship is dependent on understanding the two related items, knowing the proposition I know x assumes knowledge of the self. To bring the argument home, the familiar argument for God s necessity is used to argue that if knowledge of the self is possible for God, it must be actualized. Khojazāda is not eager to respond to this argument, merely directing the reader back to the first chapter. ʿAlāʾ al-dīn al-ṭūsī, having refrained from treating some argumentation proper to the first chapter, is more eager to respond. He thinks it is a circular argument, and should therefore be disregarded. What exactly the circularity consists of, he does not go into. III.1.2. The Second Argument The other argument was already discussed before; since God is abstract, His essence is present to Himself. Since intellection is the presence of an abstract thing to an abstract, self-subsisting thing, God knows His own essence. Khojazāda here explains that there is an apparent contradiction between the two arguments for self-knowledge. This is because in the first argument it is first established that He knows other things, and from this it is concluded that He knows Himself. However, in the second argument it is first established that He knows Himself, and from here it is argued that He knows other things. Pertaining to this second argument, ʿAlāʾ al-dīn introduces three Ījī & Jurjānī, Kitāb al-mawāqif & Sharḥ al-mawāqif, III, 102. Rāzī, Fakhr al-dīn, Sharḥ al-ishārāt, II, 305 [= namaṭ 3, faṣl 19].

17 counterarguments. First he states that knowledge is of the category of quality, while presence is a relation. Secondly he repeats the claim from the previous chapter, namely that a relation needs two distinctly different things. Thirdly, he accuses the philosophers of formerly stating that knowledge has only mental existence, while here it is taken to have independent existence. He ties this to the difference between knowledge by occurrence (ʿilm ḥuṣūlī) and knowledge by presence (ʿilm ḥuḍūrī). Only a very short passage of ʿAlāʾ al-dīn can be traced back directly to al-jurjānī, 34 but it seems that Fakhr al-dīn al-rāzī s Mabāḥith was also a source. 35 IV. God s Knowledge of Particular Things The issue in this last chapter is on how God knows and how this mode delimits the number of things He knows. al-ghazālī takes Ibn Sīnā s position to be that God doesn t know particulars which are divisible in terms of temporal division into past, present, and future. al-ghazālī proposes the example of the eclipse the classic example of Ibn Sīnā s exposition, found in all three accounts of God s knowledge to be a good way to understand this theory. This example tells us that God can have all knowledge there is to know about the theory of an eclipse. He knows that the sun, moon, and earth must be in such-and-such a composition; He knows that after so-many hours, days, or years it will happen again, etc. But He doesn t know the actual occurrence of the eclipse; He doesn t know it is happening if an eclipse occurs (put it differently, He knows how an eclipse happens but He does not know how this eclipse happens), for when the eclipse has ceased to be two things could happen. Either He now knows that it had happened, which means His knowledge changed which is impossible for the immutable Necessary of Existence, or He still knows that it is happening, and this knowledge is now said to be ignorance, which is again something contrary to the nature of God. al-ghazālī continues that for Ibn Sīnā, God likewise doesn t know about particulars inasmuch as they are divisible in terms of matter and space. He would only know humans (for example) in a universal sense, not their particular attributes. Various serious objections rise because of such an idea. An Ījī & Jurjānī, Kitāb al-mawāqif & Sharḥ al-mawāqif, II, 695. Much of the same arguments are present in al-rāzī s text. Cf. Rāzī, Fakhr al-dīn, al- Mabāḥith al-mashriqiyya fī ʿIlm al-ilāhiyyāt wa l-ṭabīʿiyyāt (ed. Muḥammad al-muʿtaṣim bi llāh al-baghdādī), Dār al-kitāb al-ʿarabī, Beirut 1990, 461 and 494.

18 example al-ghazālī introduces is the belief or unbelief of Zayd: God would know that human beings could either believe or not believe, but he wouldn t know Zayd s particular choice. Nor would He know that Muḥammad was a prophet, knowing only about prophethood in general. This all comes from the inference that if He does not know the individual, He does not know his states and acts. 36 Khojazāda explicitly addresses these objections. He considers these objections to be very weak. In his view, God s knowledge specifies the difference of each individual from the rest and He therefore also knows about our actions and intentions. al-ghazālī continues about states (sing. ḥāl) in general, to be able to understand what kind of state knowledge is. The first is a pure relation which, if changed, does not change the essence it is related to. Only the related object changes, such as your being on the right or left of something. The second kind is an attribute holding a relation, such as ability to move. The one possessing it is always able to move, regardless of whether there is something to move. Actually moving something entails a change only in the relation, not in the essence. Lastly, there is the state which involves change in the essence if the related thing changes. This is the kind of state Ibn Sīnā relates to knowledge. This division is a brief and informal summary of what Ibn Sīnā alluded to in the Ishārāt, chapter seven, remarks nineteen and twenty. Naṣīr al-dīn al-ṭūsī, in his super-commentary on the Ishārāt, proposes the following division: 1. Attributes which are fixed in the described thing and which do not require a relation to something else (e.g. having a colour) 2. Attributes which require a relationship to something else and are not fixed in the described thing (e.g. being left or right of something) 3. Attributes which are both fixed and require a relationship: a. Those who don t change because of the change of the thing they are related to, because they are primarily related to a universal (e.g. holding power over something) b. Those who do change because of the change of the thing they are related to, because they are primarily related to a particular (e.g. holding knowledge over something) Ghazālī, Tahāfut al-falāsifa, 136. Ibn Sīnā & Naṣīr al-dīn al-ṭūsī, al-ishārāt & Sharḥ al-ishārāt, III-IV, [namaṭ 7, faṣl 19].

19 Note that al-ghazālī did not include the first type, nor did he explain the third type the way al-ṭūsī proposes. For now, however, we should see this merely as a further development refined by al-ṭūsī, rather than a shortcoming by al-ghazālī. If we study al-ghazālī s objections, we notice they are twofold: The first is that change might only occur in the relation. Knowledge that something will be is the cause for the knowledge that something is when the moment of its effectuation has arrived. Likewise, knowledge that something was will be caused by the earlier knowledge that something is. All three knowledge items are already contained in one prior knowledge, individually activated when its proper moment is there. 38 This is not accepted by Ibn Sīnā, for he argues that knowledge is of the kind of 3b; change in the object known will change the knowledge and this will change the knower. To this, al-ghazālī brings in the argument that likewise the many intelligibles would constitute a multiplicity to God, while a multiplicity is denied to God by the philosophers. So it would be coherent to deny both multiple intelligibles and temporal knowledge or allow both. Secondly, al-ghazālī argues that change in itself could very well be ascribed to God. He questions the philosophers asking why this wouldn t be permitted for knowledge and gives himself a possible answer. If temporal knowledge were admitted, it would be created either by God or by something else. The philosophers argue that a temporal thing cannot come forth from an eternal thing, but God also can t be dependent on something else for His perfection. Neither option is therefore satisfactory. al-ghazālī then argues that both options are actually plausible. The first is so because the celestial spheres are known to be eternal yet generate temporal events (their uniform and perpetual motion), so God could do likewise. The second is possible because God s knowledge of particulars could be thought of as the perception of the eye of a coloured object. In this example, it is not proper to say that the coloured object is the cause of perception; the eye is always seeing and when all impediments between the eye and the object are gone, it therefore instantly perceives. As a last remark al-ghazālī notes that we should keep in mind that God is the cause for the particulars and so, if we say that He depends on them in His knowledge, it is as if we were saying that He causes His knowledge through intermediaries. 38 Khojazāda relates this view back to the Muʿtazila.

20 IV.1. The Ottoman Scholars Khojazāda organizes the problems raised according to the arguments mentioned in the first chapter, primarily styling his arguments on al-jurjānī. 39 The first part, tied up with the first argument from the first chapter, is therefore concerned with a discussion on knowledge of particulars with a shape or abstract, temporally changing or eternal. Before he puts forward the philosophical difficulties, he introduces the chapter by expanding on the view of Quṭb al-dīn al-rāzī (again without mention). By doing this, he gives the impression that this is the solution to which he wants to commit himself, although he does not make this as explicit as we would maybe want it to be. In short, this view argues that knowledge in a universal way means that God knows each individual according to his or her unique bundle of universals that is, its characteristics. This means furthermore that all are known in their respective moments, without this knowledge being qualified by past, present, and future. 40 IV.1.1. The First Argument Coming back to the main train of thought, Khojazāda divides the problem into material shape vs. abstract, and changing vs. eternal. ʿAlāʾ al-dīn does this too, although reversing the order. In the first problem the philosophers argue that to know a particular that has a material shape, the apprehender needs a bodily organ. ʿAlāʾ al-dīn states it slightly differently. He states that the imprinting of something that has a magnitude requires a receptor with a magnitude. By rephrasing it like this, he shows what is exactly at stake: the definition of knowledge. Khojazāda makes the objection that this is not necessary if knowledge is a relation or an attribute holding a relation. Again, a response to this is that in this case knowledge may not occur before the occurrence of the object of knowledge. At this point a discussion on knowledge of future contingents could occur but emphatically does not. Instead Khojazāda replies that a relation is not dependent on the existence of the two related things, but on the distinction between them. A non-existent thing will have as much distinction as an existent one. Although Khojazāda does not go into the issue of future particulars, one Ījī & Jurjānī, Kitāb al-mawāqif & Sharḥ al-mawāqif, III, [mawqif 5, marṣad 4, maqṣad 3, baḥth 2, 5 th ]. Rāzī, Quṭb al-dīn, al-muḥākamāt, [= namaṭ 7, faṣl 18].