An Ottoman Commentary Tradition on Ghazālī’s Tahāfut al-falāsifa. Preliminary Observations*

Ghazālī’s “The incoherence of the philosophers” spurred a counter-commentary by Ibn Rushd, as is well known. Up to ten texts from Ottoman scholars also purport to be commentaries on the Tahāfut, constituting a commentary tradition that has been neglected by scholars. The first two commentators, Khojazāda (d. 1488) and ʿAlāʾ al-Dīn Ṭūsī (d. 1482), are not line-by-line exegetes ofGhazālī, but rather update the discussions that Ghazālī broached to the level of knowledge available to them. Khojazāda was favored by the Ottomans, but ʿAlāʾ al-Dīn’s content, methodology and argumentation style proves to be just as, if not more, interesting for us.


Introduction
It is well known that Ghazālī's Tahāfut al-falāsifa (The Incoherence of the philosophers) received a counter-commentary from the hand of Ibn Rushd, aptly called Tahāfut al-Tahāfut (The Incoherence of the Incoherence). This text, in turn, received a super-commentary from Agostino Nifo (d. 1545). We may * I wish to thank the MacMillan Center at Yale University, and the Netherlands Institute in Turkey, whose support made this paper possible. L.W.C. van Lit,MacMillan Center,34 Hillhouse Avenue, New Haven 06511, lambertus.vanlit@yale.edu.
Oriens 43 (2015) 368-413 speak in this case of a mini commentary tradition, connecting Baghdad, to Andalusia, to Italy. What has remained less well known is that around the same time as Nifo wrote his commentary, Ghazālī's Tahāfut inspired a small commentary tradition of up to ten texts, which had nothing to do with Ibn Rushd's book, in the Ottoman Empire, stretching from the 15th century until the 18th century. In this article, I sketch the contours of this commentary tradition and give a general assessment of the two commentaries that started this commentary tradition, those by ʿAlāʾ al-Dīn Ṭūsī (d. 1482) and Khojazāda (d. 1488). ʿAlāʾ al-Dīn Ṭūsī's and Khojazāda's books were written concurrently, as Sultan Mehmed ii had invited both of them to write such a study. In their time, Khojazāda's book was deemed the better of the two, and wielded its influence in the centuries after as some scholars wrote glosses on Khojazāda's text. For us, both books have interesting aspects to them worth studying. Most important, they are witnesses to an interesting period of Ottoman scholarship, namely, the time of Mehmed ii and his conquest of Constantinople. The Sultan's patronage ensured the final rise and establishment of an Ottoman intellectual milieu. The texts are 'commentaries' in an odd way; they do not comment word-by-word on Ghazālī's text, as will be explained and analyzed later on. Only little attention has so far been paid to these texts, and as a result only ʿAlāʾ al-Dīn Ṭūsī's text is edited and available.1 For Khojazāda's text the situation is more complicated and a critical edition is a still a desideratum.2 Studying the Tahāfuts ʿAlāʾ al-Dīn Ṭūsī's and Khojazāda's books have not remained completely unnoticed in modern scholarship. To begin with, I find it useful to string together 1 ʿAlāʾ al-Dīn Ṭūsī, Tahāfut al-falāsifa, ed. by R. Saʿāda (Beirut: al-Dār al-ʿālamiyya, 1981). There is a reprint from Dār al-fikr al- Lubnānī, 1990, with different pagination. I will refer to the al-Dār al-ʿālamiyya print. 2 I will cite multiple sources, including a manuscript, as the printed version is not a critical edition and difficult to read. The manuscript was chosen for three reasons: 1) it may be old if we are to trust the date, 2) it has been collated, 3) I find it to be written in a most elegant hand. For citations from the chapters on God's knowledge I also refer in this paper to my own semicritical edition in my ma thesis. Khojazāda, Tahāfut al-falāsifa (Cairo: al-Maṭbaʿa al-iʿlāmiyya, 1303/1884; Khojazāda, Tahāfut al-falāsifa, ms Beyazid Veliyyüddin 1990 (Istanbul:  previous scholarship on them, to show how a misinterpretation has managed to persist until this very day, namely, that Khojazāda was the only Ottoman commentator on the Tahāfut and that he commented not on Ghazālī's text but on Ibn Rushd's commentary. Modern interest in the Tahāfut and its commentaries starts with explorations of Ghazālī's Tahāfut by a number of 19th century scholars such as Schmölders,Gosche,and Munk.3 They were aware of it through Ibn Rushd's commentary, of which Renan makes frequent use to reconstruct Ibn Rushd's system of thought.4 After this prelude, it is the year 1302/1884 that marks the real beginning of modern interest in the Tahāfut and its commentaries, including the Ottoman reception. In that year, Ghazālī's Tahāfut al-falāsifa, together with Ibn Rushd's Tahāfut al-Tahāfut, and also Khojazāda's Tahāfut al-falāsifa was printed in Cairo.5 It was quickly followed by two subsequent printings (of the same editions) in 1319/1901and 1321/1903.6 D.B. Macdonald notes (in 1899) that the edition was put together "apparently from an earlier Constantinople edition,"7 and although Bouyges notes that this was confirmed to him by "plusieurs cheikhs de Constantinople,"8 it seems that no one has been able to provide evidence for it. The influence the Cairo printing had on scholarship can hardly be underestimated. Apparently the first to make use of it was Tjitze de Boer, who in his 1894 doctoral dissertation paraphrased and explained the entire Tahāfut.9 Though in one instance, at the end of the book, 1303 is printed as the year of publication. 6 The 1321 publication has Khojazāda's text on the margins, not as a standalone text. It further specifies that it was "printed at the expense of Muṣṭafá al-Bābī al-Ḥalabī and his brothers in Egypt." All three are mentioned by Bouyges in; M. Bouyges A few years later, in 1913, a German paraphrase appeared by Max Horten.10 From here on, more and more scholarly works start to appear, of which one of the highlights is Van den Bergh's English translation of Ibn  The familiarity of Ghazālī's book and Ibn Rushd's response among scholars of Islamic philosophy was thereby firmly established, to the point that it received generous attention in introductory books on the history of Islamic philosophy.12 All the while though, only passing mention of Khojazāda's book (let alone ʿAlāʾ al-Dīn's book) was made. Whenever it did receive attention, these mentions were possibly too obscure to reach a large audience. One of the earliest mentions of Khojazāda's (and ʿAlāʾ al-Dīn's) book was due to Flügel, in his catalogue of manuscripts of the Vienna court library (Hofbibliothek). He listed an ms of Khojazāda's work as "Tahāfut li-Khojazāda."13 Citing Ḥajjī Khalīfa's Kashf al-ẓunūn, he gave an accurate (albeit short) description of the debate between Khojazāda ("Chodschazâda") and ʿAlāʾ al-Dīn, and commented further that Khojazāda's goal was to take and expand Ghazālī's critique of the philosophers. He also referred to Ibn Kamāl Pāshā's commentary, which is included in a Viennese ms containing several works of Ibn Kamāl Pāshā. He listed this commentary under the apt title "Risāla fī l-jawāb 'ammā katabahu Khojazāda fī Tahāfut al-ḥukamā' ."14 A couple of years later Steinschneider referred to this in an entry on Ghazālī's Tahāfut. 15 Brockelmann, in his gal, called Khojazāda's book "chiefly a critique of Ghazālī and the philosophers."16 Unfortunately, he listed ʿAlāʾ al-Dīn's book as "Kitāb al-Dhakhīra (Dhukhr) fī l-Muḥākama bayna acknowledging this ("Im jahre 1302 d.H. (1884/5 d) erschienen in Cairo zwei schriften zusammengedruckt …"), and closes his book by suggesting more than 60 corrections to the edition. l-Ghazālī wa-Ibn Rushd (Tahāfut ʿalá l-Tahāfut)." This is some years later repeated in Laknawī's Fawāʾid al-bahīya and many years later repeated by Khayr al-Dīn al-Ziriklī in his al-Aʿlām and in  Add to that Horten's statement in the introduction of his Hauptlehren that Khojazāda and ʿAlāʾ al-Dīn Ṭūsī were asked to write a study to decide between Ghazālī and Ibn Rushd,18 and we see that in a matter of fifty years from Flügel to Horten, the narrative has been severely distorted. At this point, commentaries after Khojazāda are disregarded and the two earliest Ottoman commentators are said to have commented both on Ghazālī and Ibn Rushd, instead of only on Ghazālī.
Bouyges, writing a couple of years later, got the story of the two Ottoman scholars right. In the introduction to his edition of Ibn Rushd's Tahāfut al-Tahāfut he discussed the possibility of Khojazāda's and ʿAlāʾ al-Dīn's texts as sources for establishing a correct version of Ibn Rushd's text. He however noticed that no such possibility exists, as both texts only related back to Ghazālī's text, not Ibn Rushd's text. "Cette opinion, qui dramatise l' histoire, n' est pas suffisamment fondée en ce qui concerne Averroès,"19 as Bouyges noted. He also noted that there are glosses on Khojazāda's work, which were equally without any reference to Ibn Rushd. However, recourse to the earlier (wrong) story was made by Louis Gardet. He wrote in his entry on "ʿIlm al-Kalām" in the Encyclopaedia of Islam, New Edition that "the work of the Turk Khojazāde (9th/15th century) […] sought to refute the Tahāfut al-Tahāfut of Ibn Rushd."20 Besides this, no mention is made in the ei2 of other Ottoman commentaries on Ghazālī's Tahāfut.
on the subject."21 Though we may appreciate his mention of "two great theologians," thereby not forgetting ʿAlāʾ al-Dīn Ṭūsī, it is unclear whether "the subject" refers back to Inalcik's earlier statement concerning "the famous controversy between Ghazālī and Ibn Rushd" or to "the relationship between religion and philosophy." A doctoral dissertation completed at the Sorbonne, 1972, had as its objective to edit and analyze Khojazāda's text.22  The general narrative of this account does not need clarification. All later commentaries seem to have been, at least in form, responses to Khojazāda and in this story we find that it was deemed the better of the two initial books.
I would like to point out a couple of things we can take away from this passage. Perhaps most important, we may note that this story assumes that philosophy, or intellectual thought in general, was not considered to have received a fatal blow by Ghazālī as has sometimes been opined by modern scholars. That the discussion between Ghazālī and 'the philosophers' needed to be examined, and that for this examination not one but two intellectuals were commissioned, shows that the discussion was far from settled and both sides were considered credible.39 Further, towards the anthropology of Islamic intellectual history, we may note that this passage gives us insight into book production. These books came to be because they were commissioned, and were written in about half a year. This may strike us as a remarkably short period of time, given that they are both sizable texts, but it finds its explanation in a consideration of their style of formulating their arguments, which consists to a great extent of appropriations of earlier texts.40 A second remark of an anthropological nature is that both parties received the same monetary compensation and the only aspect in which the winner received more than the other was the respect accrued by the gift of the robe. 39 It seems that it was not even clear what Ghazālī's position exactly was concerning 'philosophy' . For example, the great polymath Suyūṭī, an Egyptian contemporary of Khojazāda and ʿAlāʾ al-Dīn Ṭūsī, changed his opinion on Ghazālī's opinion on logic twice, first asserting that Ghazālī favored logic, then proposing that Ghazālī saw practical use in logic, and finally concluding that Ghazālī  Further, such respect was apparently deemed of such importance that it may have played a role in ʿAlāʾ al-Dīn Ṭūsī's graceless exit from the Ottoman intellectual scene. This aspect of patronage brings us to the historical context in which this debate took place. Tāsh Kubrī Zāda reports that Khojazāda wrote this book in a madrasa in Constantinople,41 and given his entire report on Khojazāda this may have happened around the year 865/1460, several years after the conquest of Constantinople in 1453. After the conquest, soldiers were told they could loot the city for three days, except for the buildings themselves, which would fall to the Sultan.42 As a result, many of the buildings were instantly repurposed, which was convenient for an expanding power that soon turned into a full-fledged empire.43 Mehmed ii's patronage of the arts and sciences, a hallmark of his life,44 rose to a new level after the conquest. A total of eight churches were turned into mosque/madrasa complexes, and both ʿAlāʾ al-Dīn Ṭūsī and Khojazāda were installed as professors. Among these churches were the Hagia Sophia church, the Pantokrator church, the Christ Acataliptos church, the church of St The sudden creation of many more madrasas went hand in hand with a sudden rise in the number of intellectuals. Ṭāsh Kubrī Zada collected in his al-Shaqāʾiq al-Nuʿmānīya bio-bibliographical information on scholars, doctors, and mystics, and placed them in rubrics according to roughly the era of the sultan in which they were active. If we extract some basic information from that book, we can clearly see how intellectual activity took off in the 15th century, under Mehmed ii. The following graphic shows the number of entries per sultan (date of ascension between brackets) until Mehmed ii:47

46
This is implied by Tāsh Kubrī Zāda, al-Shaqāʾiq, 90 (lammā banā al-madāris al-thamān hunāka naqala l-tadrīs minhā ilayhā). 47 The division between mullas and shaykhs is a fine line and what is presented here is merely to give an overall impression. For example, in Ṭāsh Kubrī Zāda's al-Shaqāʾiq mention is made of 'al-Shaykh Ramaḍān' (p. 47), but we have counted him here as a mulla since we read that he became judge of the army (qaḍī bi-l-ʿaskar) and the veneration reads rawwaḥa Allāh rūhahu instead of qaddasa Allāh sirrahu, the latter being the usual one Ṭāsh Kubrī Zāda applies to sufi shaykhs. With 89 persons mentioned, the era of Mehmed ii towers above previous eras. If we compare just the number of scholars, then all eras before Mehmed ii combined held 69 scholars, against 62 in Mehmed ii's time. We should also keep in mind that most of the scholars from Murad ii's time were still active under Mehmed ii. To further catalyze scholarly activity, Mehmed ii 1) commissioned copies of famous philosophical and scientific texts,48 2) ordered new texts to be written, and 3) staged debates.49 I have here brought together different aspects of Mehmed ii's patronage of intellectual activity, implying that they are elements of a thoughtfully orchestrated master plan on behalf of Mehmed ii (and his inner circle), to ensure a boom in scholarly activities, effectively initiating an Ottoman intellectual tradition. I have done so in a relatively simple manner, and to what extent this thesis holds true or needs redressing awaits further investigation.
It is however against this background, Mehmed ii's efforts to create a new scholarly tradition, which I wish to place ʿAlāʾ al-Dīn Ṭūsī and Khojazāda. They both were appointed by Mehmed ii in various capacities, being in charge of madrasas and receiving salaries. Judging by their positions, they were in fact among the top scholars Mehmed ii had at his disposal. Arguably, Mehmed ii's plan to have ʿAlāʾ al-Dīn Ṭūsī and Khojazāda both write a commentary on a famous philosophical text such as Ghazālī's Tahāfut combines all three justmentioned elements of his patronage into one. latter always to mullas. I have put him under mullas only because he appears in the middle of a list of mullas.

The Legacy of the First Commentaries
The continuing interest in both commentaries is not only evident from the many manuscript copies available,50 and signs of use on them,51 but also from the production of a number of independently circulating 'commentaries' on them. I have found references to eight different authors, but only of three of them I have so far been able to establish that the work is extant. From the references and from the texts of the three extant works, it seems that all of them styled their works on Khojazāda's commentary. The first one, Ḥāshiya ʿalá l-Tahāfut, is by Ibn Kamāl Pāshā (d. 940/1534), one of the most famous intellectuals of Ottoman history. That this important intellectual busied himself with the Tahāfut commentaries is in itself an argument in favor of their importance. His text is a set of glosses, citing a few words from Khojazāda's text and then arguing on his own concerning the topic raised in Khojazāda's text. These glosses simply start at some point in Khojazāda's introduction, and end in the fifteenth chapter,52 though some manuscripts already end the glosses at the second chapter.53 An edition was apparently created, but never published.54 The glosses follow the text of Khojazāda quite closely. Sometimes the glosses follow each other up one sentence after the other, at other times there is a considerable gap in between. As he only cites the first few words of Khojazāda's text, Ibn Kamāl Pāshā's text is not usable without a copy of Khojazāda's text. One important note to make is that he cites ʿAlāʾ al-Dīn Ṭūsī's commentary, to which he refers as al-Dhukhr.55 He does so only rarely, but the few cases are significant enough in showing that Ibn Kamāl Pāshā was aware of both texts and found it useful to juxtapose them in some cases. He evinces no awareness of the existence of Ibn Rushd's commentary. Qarābāghī (d. 942/1535), working at about the same time as Ibn Kamāl Pāshā, also wrote a gloss on Khojazāda's text. In principle the text seems to be untitled. When considering the introduction and the three chapters on God's knowledge-which I will use in the last part of this paper-this proves to be true. The rest of the introduction in Uskudārī's text is more or less an abridgment of Khojazāda's introduction, and the three chapters on God's knowledge are constructed by copying certain parts of Khojazāda's text, with only very light editing to glue the parts together. Whereas the three chapters run in Khojazāda's book to about 6.400 words, Uskudārī's reduces this to only 2.700. As far as I can tell, no mention is made of ʿAlāʾ al-Dīn Ṭūsī's work.
So we see that Khojazāda's text received special attention from Ottoman scholars. As witness to interest in the Tahāfut commentaries outside of Ana-

Preliminary Analysis of the Tahāfuts
I call these texts commentaries, but they do not take the form of line-by-line commentaries. Ibn Rushd's commentary on the Tahāfut, conversely, contains word-for-word the entire text it is commenting on and refuting. This is perhaps what we would expect when we think of a commentary, a sharḥ, but I think it is meaningful to avail ourselves of a wider definition of 'commentary' , one defined as a text which evidently relies in structure on another text, and shows intentional textual correspondence exactly in those places which define the structure and composition of both texts. This way we can speak of a 'commentary tradition' constituted by all texts that show such structural textual correspondence. I would argue that this definition works especially well for the post-Avicennan Islamic intellectual discourse, as many later texts structurally rely on an older text, without this meaning that such later texts are servile to the older text.58 Such is the case for our Ottoman texts. As we shall see, their introductions expressly relate their books to Ghazālī's book, and the chapter headings undeniably prove that the Ottoman scholars are basing their book on Ghazālī's book. To some extent, this is also repeated on the smaller scale of the inner structure of the chapters, as we shall see. Only when we zoom in to the level of paragraphs and sentences, do we find little to no textual correspondence between any of the two Ottoman texts and Ghazālī's Tahāfut.
In terms of their content, these commentaries do not provide explanatory notes on Ghazālī's text, but rather bring the discussion of the topics that Ghazālī mentions to the standards of their own time, as we shall see later. Asad Ahmed observed that commentators regularly picked and chose which 58 The reader is reminded that for the better part of the previous century, the scholarly consensus regarding late-medieval Islamic intellectual discourse was to dismiss it as servile and unoriginal. parts of the source text they commented on and which parts they did not, and that these choices were not dictated by the needs of the source text but by the discourse contemporary to the commentator.59 This is exactly how the Ottoman commentaries on the Tahāfut are most usefully seen. Note that Ghazālī's chapters number six and seventeen have no chapter heading in the text, but this is what is given in Ghazālī's introduction. The heading in the text is very long for Ghazālī's chapters number sixteen, eighteen, nineteen, and twenty; what I have given here is what is stated in Ghazālī's introduction.

Comparing the Structure of the Tahāfuts
If we compare ʿAlāʾ al-Dīn's chapter headings and Ghazālī's chapter headings, we notice that ʿAlāʾ al-Dīn stays close to the subjects Ghazālī chooses. However, he does take the liberty to reword them in a more positive way. For example, while Ghazālī makes frequent use of words like 'invalidation' (ibṭāl) and 'inability' (taʿjīz), ʿAlāʾ al-Dīn does not follow this. Either he uses 'establishment' (ithbāt) instead or, more frequently, he uses a sentence structure such as 'on the proof of whether or not x is y' ( fī bayān anna … am lā). Sometimes his wording is more precise, such as in the case of chapter seven, where Ghazālī raises the question whether God can be divided into genus and differentia. Here ʿAlāʾ al-Dīn labels the discussion whether or not God can be composed of 'intelligible parts' (ajzāʾ ʿaqlīya). The most interesting difference is in chapter ten. In Ghazālī's text it is a very short chapter. ʿAlāʾ al-Dīn takes this opportunity to stray from Ghazālī's topic and discuss the concept of knowledge, as a prelude to the next three chapters on God's knowledge.
Comparing Khojazāda's chapter headings with Ghazālī's, we notice a closer resemblance than that between ʿAlāʾ al-Dīn and Ghazālī. The wording, including the use of ibṭāl and taʿjīz, stays close to Ghazālī's original. The only real difference is a rearrangement and addition of some chapters. This is already apparent with the first chapter, which, although dealing with a problem that is connected with the next chapters, is not attested as such in Ghazālī's book. Furthermore, Khojazāda expanded chapters three and five of Ghazālī's text (in Khojazāda's count chapters four and seven), with an additional chapter to further clarify the issues at stake. His additions, therefore, do not appear to introduce new material but instead offer a different systematization. Lastly, just as ʿAlāʾ al-Dīn deviated from Ghazālī in chapter ten, Khojazāda also deviated from him but in a more radical way. He simply did away with the chapter altogether and proceeded directly to chapter eleven (in his count chapter thirteen). Khojazāda therefore added three chapters and did away with one, leaving him with two extra chapters in total. The assertion in Kashf al-ẓunūn that Khojazāda's study contains two more chapters than Ghazālī's original is therefore not the whole story.60 60 Ḥajjī Khalīfa, Kashf al-ẓunūn, vol. 1, 513. This was taken over by Saʿāda,20. van lit Oriens 43 (2015) 368-413

Comparing the Introductions
The introductions to the commentaries of the two Ottoman intellectuals provide us further evidence of the difference in attitude they had. Of the two, it is ʿAlāʾ al-Dīn Ṭūsī's that for us may be more interesting. Khojazāda tried to give a concise introduction, merely giving an idea how and why the book was composed, and what the rest of the book would be about. ʿAlāʾ al-Dīn does this too, but goes beyond it by providing a more precise idea of how he thinks reason and revelation complement each other. In a sense he provides a philosophy of religion in his introduction. He actually departs in a rather radical way from Ghazālī, asserting that disagreement on the issue of God's knowledge of particulars does not constitute unbelief (kufr).61 Khojazāda also departs from Ghazālī's position, but retreats to a more conservative position. He states that his intention is first to relate what the philosophers think and what Ghazālī's reply is, and then to refute both of them (ubṭiluhā). Khojazāda begins his text by saying that reason (ʿaql) and revelation (naql) are congruous, and that searching for knowledge is commendable.
Khojazāda goes on to explain how the sciences are divided up. He follows a fairly standard approach, which is close to Ibn Sīnā's division;62 there is practical and theoretical knowledge and theoretical knowledge can further be divided into three subdivisions. The highest one is Metaphysics, the middle one is the Mathematical sciences and the lowest one is Physics. He states that: Only two categories of them are connected to our aim of refutation in this epistle, namely Physics and Metaphysics. [This is] because opposition to what is established by the principles of revelation (al-qawāʿid al-sharʿiyya) and the dogmas of religion is restricted to them, whereas the Middle Science-geometry and arithmetic-is not attached to revelation at all.63 61 Ghazālī, it may be reminded, famously proposed that the philosophers' denial of three ideas-the creation of the world, God's knowledge of particulars, and bodily resurrection-constitutes unbelief. Ghazālī This expresses a position more conservative than Ghazālī's, who wrote at the beginning of his book that he wanted to show "the contradiction of their word in matters relating to metaphysics,"64 not metaphysics and physics. It is true that the Tahāfut does contain a couple of chapters that are grouped under the heading 'Physics' (al-Ṭabīʿiyyāt), but in the introduction to these chapters Ghazālī makes clear that "the religious law does not require disputing them nor denying them, except in places we will mention."65 Further, when we consider what Ghazālī chose to discuss under this heading (causality and the immaterial soul), we see that these are in essence metaphysical problems. Khojazāda's blanket condemnation of Physics in his introduction is therefore well beyond Ghazālī's original intentions (at least as stated in the introduction to the Tahāfut).
ʿAlāʾ al-Dīn, on the other hand, did not want to restrict himself merely to proving the philosophers' error. "Rather," he adds, "we will import some of what they have erred in, in the evidence, and if the claim is true, to prove those from numerous aspects."66 This is because according to him some of the claims of the philosophers are true, even though they may not have argued for them correctly. He concludes this by saying that "this exaggeration in their viewpoints is purely because of blind following (mujarrad taqlīd), not because of correct proof."67 In short, it is his mission to sort out opinion from truth, showing that whenever reason and revelation seem to go against each other, at least one of the two is in fact merely an opinion. Citing cases for which the philosophers have merely followed earlier philosophers blindly and did not give a correct proof for an idea that is in itself correct, he will show the correct proof. This idea was also present in Ghazālī's introduction,68 but was less prominent.
ʿAlāʾ al-Dīn's introduction is divided into two parts: a preface (tamhīd) and an introduction (muqaddima). He discusses the happiness (saʿāda) of humanity's knowledge of God and His perfections, and how this can be achieved by reflection on His creations (al-tafakkur fi maṣnūʿātihi). He then discusses the role of the philosophers in this, and how some of what they do can be good, while some of it falls short. He then explains the reason why he wrote the treatise: he received an order from the Sultan to study the Tahāfut al-falāsifa, and to write what he thought of the two methods, Ghazālī's method versus the method 64 Ghazālī of the philosophers, "by way of expanding and favoring (min jihāt al-taḍʿīf wa-ltarjīḥ), and objecting and correcting."69 ʿAlāʾ al-Dīn explains that he will not be blindly following (bi-ṭarīq al-taqlīd) Ghazālī, although he will adopt Ghazālī's outline and method. He concludes the preface by promising only to put forward what he deems correct and understandable, followed by an invocation to God to help him with this. His final words in the preface concern the book as a whole. He affirms it is written in twenty chapters, "like the original." As can be seen from the table, this is true in terms of the chapter count, but it is slightly misleading with regard to chapter ten. ʿAlāʾ al-Dīn explains the aim of his book by giving a primer on the conflict between reason and revelation. He regards the variety of our different faculties of perception as something we should be thankful for, but at the same time we need to remain aware that they have their limits. In ʿAlāʾ al-Dīn's words: "they fail to reach their fulfillment, for we cannot see everything there is to see, nor can we hear everything there is to hear … [etc.]"70 This is the same for all other faculties, including the intellect; there will always be a bit of doubt left, no matter how deeply we think about an issue. Religion is there to help us in cases of doubt. In his view, the difference between philosophy and religion is threefold. This threefold division is directly taken over from Ghazālī's 'second introduction' to the Tahāfut.
The first part concerns the application of certain concepts to God, such as 'substance' ( jawhar), when it is taken to mean 'that which is self-sufficient' (qāʾim bi-nafsihi), for this concept is only properly applied to possible existents.71 According to ʿAlāʾ al-Dīn, who drew directly from Ghazālī's book, this difference is merely of a terminological type and does not constitute a real difference in meaning (maʿná). He therefore did not discuss it.
Secondly, there are those philosophical judgments that are built upon decisive evidence and which go against the apparent sense (ẓawāhir) of the Religious Law. ʿAlāʾ al-Dīn gives astronomy as an example, just as Ghazālī had done.72 Religious conceptions about the sphericity of the heavens and the earth, the composition of the heavenly bodies, their motions etc., go against scientific judgments. While it is clear that these scientific judgments are based upon mathematics and observations, it would be absurd to suggest that two contradictory judgments are both backed up by decisive evidence. The solution ʿAlāʾ al-Dīn proposes is that "the gate of interpretation of the apparent senses 69 ʿAlāʾ al-Dīn Ṭūsī, Tahāfut, 61. 70 ʿAlāʾ al-Dīn Ṭūsī, Tahāfut, 65. 71 As opposed to God who is the 'Necessary of Existence' (wājib al-wujūd). 72 Ghazālī, The Incoherence, 7.
Oriens 43 (2015) 368-413 is open."73 In this way, a real contradiction does not arise and therefore it will also not be the topic of discussion. ʿAlāʾ al-Dīn is in his phrasing more firm and direct about the primacy of reason over the apparent meaning of revelation than Ghazālī. For example, Ghazālī does not say immediately this category is about rational judgments that go against the apparent meaning of revelation, but simply talks of rational judgments. He also does not say that the gate of interpretation is open. Lastly, there are philosophical judgments that are not backed up by decisive evidence. As Ghazālī states: The third part is one where the dispute pertains to one of the principles of religion, such as upholding the doctrine of the world's origination and of the [positive] attributes of the Creator, [or] demonstrating the resurrection of bodies, all of which [the philosophers] have denied. It is in this topic and its likes, not any other, that one must show the falsity of their doctrine.74 Although ʿAlāʾ al-Dīn agrees that it is this subject that ought to be discussed, he distinguishes two subcategories: The first is their judgment that leads them to unbelief due to its conflict with what has been decisively established by the Lawgiver, such as the judgments on the eternity of the world, or the rejection of bodily resurrection. Their arguments for these two conclusions and their likes, whatever they may bring to it, are weak, whereas the proofs of the Law are decisive on them.
The second is their judgment that does not lead them to unbelief due to the absence of the Law's decisive arguments for the opposite, such as their denial of real attributes to God Most High, supposing that confirming them would deny [divine] Unity. The letter of the Law, even though its apparent indication indicates its confirmation, is susceptible to interpretation, just as the statements are interpretable that point to the confirmation of a To show how different ʿAlāʾ al-Dīn's sub-categorization is, it is worth repeating here that Ghazālī did not shy away from pushing his idea to its logical conclusion, stating at the end of his Tahāfut that: Pronouncing them infidels is necessary in three questions.
[…] The second is their statement that God's knowledge does not encompass the temporal particulars among individual [existents].76 In other words, arguing over what is and what is not in agreement with religion is not just an intellectual discussion, but ought to have legal ramifications. For Ghazālī, pronouncing them unbelievers (takfīr) is necessary on all three issues-denial of creation of the world, denial of bodily resurrection, and denial of God's knowledge of particulars. ʿAlāʾ al-Dīn clearly disagrees on this point with Ghazālī. This was clear enough for Ibn Kamāl Pāshā, who points it out by saying that "the author of al-Dhukhr did not include in his allegation that the philosophers have denied the real attributes of God."77 Assuming 'God's knowledge' under the umbrella of 'attributes of God' ,78 ʿAlāʾ al-Dīn sets it aside as a subcategory different from the discussion on the eternity or creation of the world and the discussion on bodily resurrection. He even adds that some religious scholars agree with the philosophers on the attributes of God. To ʿAlāʾ al-Dīn, then, discussion about the attributes of God should not be silenced by a legal ruling of unbelief.
For a fair understanding, however, I should point to the conclusions of both Ghazālī's and ʿAlāʾ al-Dīn's books. There ʿAlāʾ al-Dīn openly pronounced the philosophers to be unbelievers (nukaffiruhum), and not only because their denial of the creation of the world and bodily resurrection, but also because of their denial of God's knowledge of particulars.79 Ghazālī, on the flip side, asserted in his conclusion that divine attributes is indeed a topic on which philosophers should not be judged to be unbelievers: "As regards questions other than these three, such as their treatment of the divine attributes and their belief in divine unity entailed therein, their doctrine is close to that of the Muʿtazila."80 76 Ghazālī Nevertheless, ʿAlāʾ al-Dīn was following Ghazālī's introduction quite closely, and his efforts to change the most crucial part of it, the spelling out of those subjects that cause philosophers to be unbelievers, remains significant.

Comparing a Chapter on God's Knowledge
In light of these comments on their introductions, a comparative examination of a chapter on God's knowledge will yield a better understanding of all three books. This will at the same time help us understand their method of writing. Our main means of understanding these chapters will be to draw up structured summaries, which not only give a succinct rendering of the argumentation, but also lay bare the highly structured nature of these writings.
Already from a cursory glance over these summaries we notice that the Ottoman texts are far more advanced and detailed than Ghazālī. I would suggest that although Ghazālī's text was used as a guide to structure their texts, it was deemed too simple to engage much with. The discourse as a whole had moved far beyond the level it was at in the eleventh century, when Ghazālī was writing, and both Khojazāda and ʿAlāʾ al-Dīn Ṭūsī seem determined not to turn back the clock on these discussions but instead bring the discussions of the problems that Ghazālī includes in his Tahāfut to the standards of the fifteenth century.
The most eye-catching exception is the inclusion of certain passages from Ghazālī's text by Khojazāda, at the end of the chapter. Khojazāda cites in bullet point style one or two lines from Ghazālī's text and then proceeds to criticize it in a rather unforgiving style. Khojazāda's stiff criticism of Ghazālī is further evidence for the development of the Islamic discourse in between the eleventh and the fifteenth century. Judging from the introductions, ʿAlāʾ al-Dīn seemed more appreciative of philosophy whereas Khojazāda showed a greater determination to refute philosophy's use in matters pertaining to religious dogma. Ghazālī's objective in the Tahāfut is exactly to show the weakness of philosophy when it meddles in religious matters. We would therefore assume that Khojazāda is more amicable towards Ghazālī's position than ʿAlāʾ al-Dīn Ṭūsī. What I think Khojazāda's lashing out against Ghazālī shows, is not that Khojazāda is actually on the side of the philosophers, but merely that he wants to persuade his readers that they ought not to revert to the level of discussion of Ghazālī's time, but that they ought to rise to the level of discussion of his own (Khojazāda's) time.
From the structured outlines it becomes clear that both Khojazāda and ʿAlāʾ al-Dīn Ṭūsī wrote highly detailed studies with many layers of objections van lit Oriens 43 (2015) 368-413 and responses. In this labyrinth of argumentation, the texts become less and less a straightforward argumentation for one or the other position, and more and more an exploration of all positions, weighing their strengths and their weaknesses. As a result, both chapters from the Ottoman scholars read not so much as a theological discussion on what God knows, but more as a philosophical discussion on what knowledge is, with God being merely a test case. Put differently, one of the main strengths of the two Ottoman texts is that their authors updated the discussions that they find in Ghazālī's Tahāfut to the level of knowledge that was available to them.81 Elsewhere I have spent considerable effort to understand the construction of the two Ottoman texts on a paragraph and sentence level,82 and what I found was that large stretches of their texts are appropriations of texts written by scholars within a century before them. To mention the most frequently used scholars we have Iṣfahānī (d. 1348), Ījī (d. 1355), Quṭb al-Dīn Rāzī (d. 1364), and Jurjānī (d. 1413). A notable outlier is Fakhr al-Dīn Rāzī (d. 1209), of whom various writings are utilized by the Ottoman scholars and who lived not one but three centuries before them. Thus, beyond their immediate predecessors, they found in Fakhr al-Dīn Rāzī (not Ghazālī) an intellectual whose writings they considered classics, worthy of incorporating in their own texts. Their contemporaries who wrote important works already before them, such as Mollā Fanāri (d. 1430) and ʿAlī Qushjī (d. 1474), seem to have not been utilized directly as a source, as far as I have been able to establish for the chapters on God's knowledge.
The following is a structured summary of Ghazālī's chapter on God's knowledge (the first chapter out of three).

Structure of Ghazālī's Chapter Eleven83
-Muslims say: That which is willed must necessarily be known to the willer.
Everything is known to Him because all [things] are willed by Him. But philosophers do not agree on will so neither on knowledge. 81 Ayman Shihadeh makes a similar comment in his analysis of Khojazāda's sixth chapter, saying " [it] appears to be essentially a critical update of the corresponding discussion in al-Ghazālī's work." Cf. Shihadeh, "Khojazāda on al-Ghazālī's Criticism …", 146. 82 Van Lit, "The Chapters on God's Knowledge …"; Van Lit, "Two Ottoman Intellectuals …". 83 Ghazālī, The Incoherence, 125-30.
1. Proof: God is abstract, i.e. an intellect, and thus knows all intelligibles. a. Objection: What is meant by intellect? It cannot mean 'that which intellectually apprehends others' for that is the conclusion sought after and cannot be accepted as a premise. If it means 'that which intellectually apprehends itself' , then why would it follow that it apprehends others? i. Response: because the impediment is matter, of which intelligibles are devoid.

Proof:
The world is an act of God, therefore He knows it.
a. Objection: Acting is of two kinds: voluntary and natural. Only with voluntary acts is knowledge necessary. Since philosophers argue that the world follows necessarily from His essence (involuntarily), they cannot claim that He knows. i. Response: It is exactly His knowledge of the universe that causes it to come to be.
(1) Objection: This is not what the philosophers argue for. b. Objection: Only one thing is emanating from God so only one thing is known to Him. c. Objection: In fact, if He only knows Himself, then other things are better than Him for they know, their selves, Him, and others.
(1) Addition: The only solution is to admit to the temporal creation of the World. i. Response: Knowledge makes perfect, and God does not need to become perfect so He does not need knowledge.
(1) Objection: This makes knowledge into a deficiency.
Ghazālī's chapter is easy enough to summarize. He first provides a short statement on what, in his view, is the majority opinion among Muslim scholars on the issue of God's knowledge. The crucial element here is God's volition, which relates this problem back to the problem which receives the most attention in the Tahāfut, namely that of the creation or eternity of the world. According to Ghazālī, Islamic dogma requires one to accept that the world was created at a certain moment, by God's volition. Ghazālī argues here that since any voluntary act requires knowledge of the effect, God knows the world and everything in it. From his studies of philosophical writings, he concludes that philosophers propose two ways of arguing that God has knowledge. In the main, he argues against the first by saying that it is not a water-tight argument and therefore does not necessarily arrive at the desired conclusion, and he argues against the van lit Oriens 43 (2015) 368-413 second by pointing out that it relies on whether God has volition or not. It may be noted that near the end of both proofs the structured style of writing breaks down a bit.
It should further be noted that some of the objections brought forth against the second proof seem also objections against the majority opinion stated at the beginning of the chapter. Indeed, when we also consider the other two chapters on God's knowledge, it is clear that Ghazālī does not unequivocally accept this opinion. Instead, he tinkers with the concept of knowledge. He isolates the changing nature of knowledge in the relational aspect knowledge has. He argues that unwanted consequences (namely, a God susceptible to change) only occur when we think the "relation in knowledge is the reality of the essence of knowledge."84 In that case, a changing relation "necessitates a change in the essence of knowledge."85 Instead, Ghazālī relegates the role of relation in knowledge outside of the essence of knowledge to what he calls "pure relation" (iḍāfa maḥḍa).86 In such a case, change only happens in the relation, but not in the subject. Ghazālī argues that "He knows things by one knowledge in the eternal past and future, [His] state never changing."87 Although he does not say so himself, it seems to me that this is to be understood from the consideration that God is outside of time and therefore has an equal relation to every point in time.88 Relational terms such as 'will' , 'is' , and 'has' are, as such, taken out of the essence of knowledge. For example, because of God's equal relation to every moment in time, the three propositions "Zayd will arrive tomorrow," "Zayd is arriving at the moment," and "Zayd has arrived yesterday" are part of the same, one knowledge of Zayd's arriving on a specific date. Put differently, only for us who are trapped inside the spatial and temporal cosmos do relative references (here/there, yesterday/tomorrow, left/right, etc.) make sense; for God they simply do not apply.
When we move on to Khojazāda's chapter, the following structured summary can be drawn up: 84 Ghazālī

Structure of Khojazāda's Chapter Thirteen89
1. Proof: God is abstract from matter and its concomitants. It is therefore suitable to be an intelligible. All intelligibles are suitable to be an intellectual apprehender, if they are self-subsistent. i. Reason: Intellection is the apprehension of a thing as it is abstract from all material aspects. If a thing is already abstract, it will already be intelligible. ii. Reason: If it is an intelligible, it can be apprehended together with another intelligible. And if it is self-subsistent it can intellectually apprehend the other intelligible itself.
(1) Minor: For if it is an intelligible, it is possible to judge it together with another (in an intellect). (a) Example: As for example with existence, or unity, etc.
(2) Major: This judging together constitutes a connection between the two, i.e., a connection between two states in a locus.
(3) If one of the two things is self-subsistent, the connection can occur independently (from an intellect). For a connection taken absolutely is more general than a connection occurring in an intellect. (4) This event is then the occurrence of a state in a locus, for the self-subsistent thing will be the locus. (5) Thus the self-subsistent thing will intellect the other, for this is the definition of intellection. (6) This is necessary, for otherwise it would only occur anew, and occurrence implies matter (but an abstract thing is not connected to matter). a. Objection to minor: The statement that matter is the only impediment is in need of a proof. b. Objection to major: It does not follow that a connection needs to occur if one of them has external existence. Its mental existence is a condition for the suitability of the connection, and mental and external existence differ. A proof for the one is not a proof for the other. c. Objection: If mental existence would be a condition for the connection taken absolutely, a circular argument would occur. i. Response: Mental existence as a condition means that the connection taken absolutely between two intelligibles in an intellect is con- ditioned on the mental existence of both of them in the intellect. This does not involve a connection between the abstract thing and the intellectual apprehender.
(1) Objection: But then a substance can become an accident. ii. Response: There is external existence and mental existence. The specific relation in a thinker can rely on mental existence without thereby being a circular argument.
(1) Objection: Perhaps it may not rely on it, and it could still be the case that it appears together. iii.Response: The relation could simply be a concomitant of the quiddity of the abstract thing. If not, the possibility of relation either occurs 1) with, 2) after, or 3) before the mental relationship. The first two are false, because possibility needs to be before actual occurrence. The third remains, which must mean it is a concomitant of the quiddity.
(1) Objection: But it is difficult to conceive quiddity in itself, as we think of it in terms of its mental existence, so the relation relying on mental existence cannot be ruled out. d. Objection: The process of knowledge used here is the occurrence of a Form, but the philosophers deny that this is how knowledge works for God, so this argumentation cannot be applied to God. i. Response: But Ibn Sīnā uses it in al-Ishārāt. e. Objection: There are other objections too, but it would be too much to mention all. 2. Proof: God is an abstract, self-subsisting thing. He is therefore present to Himself without obscurity. He therefore knows Himself, as this is the definition of intellection. His essence is a cause for all other things. Knowledge of the cause implies knowledge of the effect. Therefore, He knows all other things. i. Other proof: He knows His essence, and His essence is a principle for something other than Him. Because He knows He is a principle, He must also know what He is a principle for. By extension, He also knows the effects of this effect, and therefore knows all things. a. Objection: To be present to something means to stand in a relation, and this can only occur when the two things are distinctly different. i. Response: The difference is conceptual, which is sufficient.
(1) Objection: But this is only sufficient in terms of conceptuality, not in terms of the things itself. And we were discussing the thing itself. b. Objection: The concept of intellection is wrong. It could also be a relational state.
Oriens 43 (2015) 368-413 c. Objection: Knowledge of the cause does not imply knowledge of the effect.
(1) If they mean: Knowledge of the cause inasmuch as its specific essence necessitates knowledge of the effect-as is used in the first exposition-then a proof is lacking.
(2) If they mean: 'Knowledge of the cause inasmuch as it is a principle and cause for the effect' is a necessity for knowledge of the effect, this is wrong. For this is dependent on knowledge of the effect, so it cannot be a necessity for it. (3) If they mean: 'Knowledge of the cause inasmuch as it is a cause for the effect' is a requirement for knowledge of the effect (not a necessity)-as is used in the second exposition-; a thing being a principle is impossible to know itself. Being a principle is relational, different from the specific essence. Yet they do not say that the intellection of that relation implies the intellection of the effect. i. Response: Just as a perfect cause of a specific essence necessitates a specific effect, so knowledge of its reality necessitates knowledge of the effect. With the assumption that intellection is as such, 'being a principle' is present to God Himself, and therefore He has knowledge of all others.
(1) Objection: We know that the external cause itself is necessitating the external effect itself, but we do not know that their forms have a similar relation. d. Objection: We don't accept that 'being a principle' is present to Him.
i. For 'being a principle' does not have existence like the real attributes or external things. ii. It also does not have mental existence. Otherwise all conceptual and negative attributes we have would be known to us. This is not the case. 3. Proof: Some contemporary scholars argue that knowledge is an absolute perfection. And all absolute perfections should hold for God. i. Minor: An absolute perfection is perfect in all regards; knowledge qualifies for this. ii. Major: Absolute perfection for things existing means by applying it that the thing does not lack in this regard. It is necessary to apply such things to God.
(1) Reason: All not-impossible things are necessary to Him, for possibility makes Him less than necessary, and God is the Necessary of Existence. a. Objection: The definition of absolute perfection is wrong. It should be be perfect unrestricted from any sense or aspect. i. If they mean: A contingent aspect in regards to its existence in itself, it is impossible. ii. If they mean: A contingent aspect in regards to its accidents, it is possible. c. Objection: What they say about 'a multiplicity would occur' is ambiguous.
i. If they mean: In regard to His essence, it is impossible.
ii. If they mean: In regard to conceptions and aspects, it is possible. -Remark: The last two proofs involve knowledge of all things, the first proof only some things. -Citations and paraphrases from Ghazālī's Tahāfut. Khojazāda objects: -Objection: The way Ghazālī sets out the proof is not in accordance with how the philosophers think about it. -Objection: Ghazālī leaves room for the second proof, by omitting some of the premises (that were refuted by Khojazāda). -Objection: The philosophers do not think that God's activity is compulsory. They rather say that whatever He wants, it happens, and whatever He does not want, it does not happen. The problem is that this volition is necessarily part of His essence, and not superadded to His knowledge. Because it is not superadded they did not prove His knowledge of His volition. -Objection: Ghazālī's statement that 'knowledge of the effects of the effect is not admitted' is true according to the philosophers. But they do not argue for His knowledge of His causality, but rather for knowledge of the complete cause. (Ghazālī's counter-arguments are therefore useless). -Objection: Ghazālī's example of the stone proves nothing, for the complete cause is not known to the mover and so the complete movement of the stone will not be known to the mover. For example, another contribution to the motion of the stone is the nature of the stone. If one wants to take exception to their proof, it should be in their premise that 'knowledge of the cause necessitates knowledge of the effect' .
Khojazāda's text has followed Ghazālī's order only loosely. He chose not to begin with the Islamic point of view, but has clearly a much more in-depth discussion of the two arguments. He adds another argument, namely the argument from nobility. Ghazālī was not completely silent about this as he speaks about it at the end of his chapter, though he fits it in his structure in an entirely different way and in a much less developed state. Khojazāda finishes his chapter with a discussion of Ghazālī's text. As mentioned before, Khojazāda and ʿAlāʾ al-Dīn Ṭūsī present a rather dense, belabored study of this discussion on how God is able to know (especially contingent particulars). As such, it is not immediately clear what their own points of view are. For Khojazāda, I find two statements to be most clear. The first comes at the end of his Chapter Fourteen (on God's knowledge of Himself). There he says: The result is that they sometimes first establish that it is necessary that God is knowing of something other than Him, consequently establishing that it follows from Him being knowing of something other than Him, that He is knowing of Himself, as is done in the first argument. And sometimes they turn the matter around, first establishing that He is knowing of Himself, consequently establishing that it follows from Him being knowing of Himself, that He is knowing of something other than Him, as is done in the second argument.90 Placed at the end of the chapter, introduced by "the result is" (wa-l-ḥāṣil annahum), this passage is clearly meant as a summary of his foregoing reasoning.
One could say that this shows that his main purpose was to bring to the surface the incoherence of the philosophers' argumentation. Another way of looking at it is that he wishes to show the large number of mutually exclusive epistemological inquiries. What Khojazāda believes to be the correct position on this issue is unclear to me. With the absence of a clear, unambiguous statement we need to resort to circumstantial evidence. Doing so, I would propose that Khojazāda favors the concept of knowledge as a bundle of universals, as argued for before him by Quṭb al-Dīn Rāzī.91 Two reasons indicate this might be Khojazāda's preferred view. First, he describes this concept of knowledge at the very beginning of the last of the three chapters, somewhat unattached to the rest of the chapter. Second, he does not propose any direct objection against this view. This type of knowledge is described as follows: This does not mean that one knows its universal quiddity only, but that one knows too that the universal quiddity is characterized by universal 90 Khojazāda, Tahāfut  (1) All abstract things are devoid of material relations and therefore can be intellectually apprehended.
(2) All intelligibles can be intellectually apprehended together with another. (a) Example: For example judgments of something together with 'possibility' , 'existent' , etc.
(3) 'Intellectually apprehended together' means the two intelligibles are joined together in an intellect. (4) Thus joining them together is possible in an absolute sense.
(a) Reason: It is either dependent on the relation in an absolute sense, or on the relation in an intellect.
(i) Relation dependent on relation in the intellect is circular. (ii) Relation dependent on relation in an absolute sense is possible. (5) If one of the two has external existence, being self-subsistent, it will intellectually apprehend the connected thing, for this is the very definition of intellection. (a) Reason: Only in the case of self-subsisting things can a connection take place independent of an intellect. (6) It will therefore definitely and always intellect it. For otherwise new things would occur to it, while occurrence anew is conditioned on matter. a. Objection: The proof for God being abstract has been treated of before. b. Objection: 'That every abstract thing intellects' is not correct.
i. Reason: It is not self-evident that matter is the only impediment.
(1) Example: Humans do not know God, even though He is without matter. ii. Reason: It need not be possible for every intelligible to be intellectually apprehended together with another.
(1) If they mean: All other things, then a proof for the absence of the impossibility of the connection is missing. It is impossible for them to go through all the possibilities and show it is correct to intellectually apprehend them together.
(2) If they mean: Everything taken as a whole, it is correct. But they do not say this, for this is not their objective. iii. Reason: Even if they can be intellectually apprehended together, a connection between them is not implied. It relies on their definition of knowledge, which is not admitted. iv. Reason: It is not admitted that this connection could occur absolutely. Not everything is conditioned in such a way that it is either necessary or impossible. All contingents are contingent by themselves.
(1) Possibilities: 1) A connection between two states in a locus.
2) A connection between a state and a locus. 3) A connection between a locus and a state. (a) Example: 2) An accident.
3) The species of substances. (2) All three are different, and one can act as the other's condition. v.
Reason: From the occurrence of a connection in the mental world a connection in the external world is not admitted. vi. Vice versa, intellection is not admitted only because a connection in the external world occurs. vii. Reason: Intellectual forms are equally not able to exist on their own so they either equally well know each other or equally do not know each other. The first is impossible and the second is what we sought.
(1) Response: Equal in some regard, but surely not in all. So maybe other regards make one likelier to know the other. (a) Example: Concerning speed and movement. viii. Though, surely, abstract things are able to apprehend each other. c. Final remark: They give many flawed counter-arguments, which we will not discuss here, as not to make it too long. 2. Proof: If He knows Himself, He knows others. But He knows Himself, so He knows others. i. Reason: 'He knows Himself' because of the proof to be explained in the next chapter. ii. Reason: 'He knows others' because He is the cause for all other things, and knowledge of the cause implies knowledge of the effect. a. Objection: They argue that only universals are part of His knowledge, yet changing particulars are present in the chain of His causality and universals are not. b. Objection: Knowledge of the cause does not imply knowledge of the effect. Then all effects, and all effects of all effects etc., ought to be included. This is not the normal use of knowledge. i. Response: What is meant is perfect knowledge, that is, knowing all aspects that contribute to causality. From this, knowledge of the effect necessarily occurs. God knows His essence perfectly, so He knows His effects.
(1) Objection: There is no proper argument for God's perfect knowledge of Himself.
Oriens 43 (2015) 368-413 c. Objection: The proof that God knows Himself is not correct, as will be explained in the next chapter.
In the case of this chapter on God's knowledge, ʿAlāʾ al-Dīn Ṭūsī follows Ghazālī's structure quite closely, first saying what the Islamic point of view is, then discussing the argument from God's being abstract, then discussing the argument from God's being a cause. To this, ʿAlāʾ al-Dīn only adds an introduction. ʿAlāʾ al-Dīn does not forget about the argument from nobility which Khojazāda discusses, but postpones it until the next chapter. ʿAlāʾ al-Dīn, like Khojazāda, offers an in-depth discussion of a variety of positions. In the chapters on God's knowledge he does not systematically rule out all but one option. What his own position is becomes clear from the beginning of the last of the three chapters on God's knowledge: For us, knowledge is the relation itself (nafs iḍāfa) between the knower and the known, or an attribute holding a relation. However it may be, there is no change in the same manner as mentioned except that relation itself.94 Indeed, to leave no doubt about it, at the very end of the third chapter, concluding the entire discussion on God's knowledge, he reiterates: For according to us knowledge is merely a relation or an attribute holding a relation. The consequences that they speak of only happen in the relation.95 Clearly then, ʿAlāʾ al-Dīn Ṭūsī is much more explicit about his own opinion than Khojazāda. ʿAlāʾ al-Dīn holds the position that knowledge is a relation, which can be traced as far back as Fakhr al-Dīn Rāzī.96 Any changes that happen to the object of knowledge only have an effect within that relation, but do not reach the subject of knowledge.

Conclusion
Not only Ibn Rushd and Agostino Nifo were inspired to structure their writings around Ghazālī's Tahāfut al-falāsifa, but also in the Ottoman Empire there were a number of intellectuals who wrote commentaries on Ghazālī's book. A few references to Khojazāda's commentary have been made in previous scholarship, leading us to believe that only Khojazāda wrote a commentary. We can in fact find references to up to ten authors. The two commentaries that set this practice off were the ones by Khojazāda and ʿAlāʾ al-Dīn Ṭūsī. They wrote their books upon request by the sultan, who seems to have been trying to establish a new, Ottoman intellectual discourse. Khojazāda's book was judged to be the better of the two, which may explain why most subsequent intellectuals focused their attention on this work. When giving the two earliest Ottoman commentaries a closer look, we notice that in chapter-structure they follow Ghazālī's book quite closely. Khojazāda added three chapters and omitted one. ʿAlāʾ al-Dīn Ṭūsī changed the tone of the chapter titles into a more positive, optimistic one. Rather than using words such as 'refuting' (ibṭāl) he prefers words such as 'explaining' (bayān). This is also reflected in their introductions. Both authors stay relatively close to the message of Ghazālī's introduction. On one crucial aspect, though, ʿAlāʾ al-Dīn Ṭūsī adapts Ghazālī's position, allowing for more leeway when it comes to God's attributes (including God's knowledge).
When, however, we look at the actual content of the chapters, we notice that they are not commentaries in the usual sense of the word. They do not cite Ghazālī's text in full, but only loosely follow its order of topics. They are simply not so much concerned with Ghazālī's actual text. In fact, whenever Khojazāda does cite Ghazālī, it is for the sole purpose of refuting him. Previous scholars have suggested that Khojazāda included Ibn Rushd's commentary on the Tahāfut in his discussions, but this does not seem the case. Indeed, not one of the Ottoman commentaries show traces of Ibn Rushd's commentary. Seeing that Khojazāda and ʿAlāʾ al-Dīn Ṭūsī make mostly use of sources which they most probably read during their time as students (texts by the likes of Jurjānī and Taftāzānī), I have interpreted their method as an attempt to update the discussions that Ghazālī broached to the level of knowledge available to them. The discourse of the fifteenth century had moved far beyond the level of depth and detail than it was in Ghazālī's time; Khojazāda's and ʿAlāʾ al-Dīn Ṭūsī's texts are a clear testament to this.
Similarly, the discussions that Khojazāda and ʿAlāʾ al-Dīn Ṭūsī bring forth are perhaps not so much about the theological question of what God knows, but more about the philosophical question of what knowledge is. Khojazāda seems unwilling to give a definitive answer to this question, treating pros and cons of various positions without showing personal appreciation for any. He may have favored the concept of knowledge as a bundle of universals which, as long as the bundle is big enough, only applies to one particular thing. ʿAlāʾ al-Dīn Ṭūsī is a lot clearer on his view: he holds that knowledge is the mere relation between knower and known. However, as I concluded elsewhere,97 Khojazāda includes more material in his chapters and gives a philosophically solid impression, and perhaps we ought to find in this aspect the reason for the Ottomans' preference for his text above ʿAlāʾ al-Dīn's.
My analysis of their method and intention relies on only the chapters that deal with God's knowledge, and is therefore tentative at best. Much work remains to be done; Khojazāda's text is not even properly edited yet. Close readings of other parts of their books is another clear desideratum. And even then we have only explored the first two texts of a larger phenomenon. I hope to have given at least some help for those brave enough to take on these challenges.