[In the following excerpt, Marmura assesses the importance of al-Ghazālī's Tahāfut al-falāsifa, explains its purpose and chief arguments, and examines some of the critical responses it generated.]

I

Al-Ghazālī's Tahāfut al-falāsifa (The Incoherence of the Philosophers) marks a turning point in the intellectual and religious history of medieval Islam. It brought to a head a conflict between Islamic speculative theology (kalām) and philosophy (falsafa) as it undertook to refute twenty philosophical doctrines. Seventeen are condemned as heretical innovations, three as totally opposed to Islamic belief, and those upholding them as outright infidels. Not that the philosophers it condemned were atheists—far from it. Their entire philosophical system rested on affirming the existence of God, from whom all other existents emanated. But, according to the Islamic philosophers, these existents emanated as the necessary consequence of the divine essence. As al-Ghazālī saw it, this meant that God produces the world by necessity in the same way that an inanimate object like the sun was said to produce its light by its very nature—by its essence, necessarily. It meant for him the denial of the divine attributes of life, will, power, and knowledge. Denuded of these attributes, he maintained, the God of the philosophers was not the God of the Qur’ān. At issue was not the question of God's existence, but the nature of the godhead.

The Tahāfut certainly put Islamic philosophy on the defensive in a way that it had never been before. Paradoxically, however, it also served to make it better known in the Islamic world. It brought to the fore the conflict between philosophy and more traditional Islamic belief. But perhaps more to the point, in order to refute the Islamic philosophers, al-Ghazālī had to explain them. He explained them so clearly and well that he rendered philosophical ideas accessible to nonphilosophers. Inadvertently, so to speak, the Tahāfut helped spread philosophical ideas, as it also set a new tradition in kalām. After al-Ghazālī, no Islamic theologian worth his salt avoided detailed discussion of the philosophical theories al-Ghazālī had criticized. Kalām thereafter became, as it had never been before, thoroughly involved with the theories of the falāsifa.

The Tahāfut also marks a high point in the history of medieval Arabic thought because of its intellectual caliber. Although its motivation is religious and theological, it makes its case through closely argued criticisms that are ultimately philosophical. A logical critique, largely of the emanative metaphysics, casual theory, and psychology of Avicenna (Ibn Sīnā, d. 1037), it is incisive and thorough. It is true that theological criticism of philosophy was not entirely new in medieval Islam: one does encounter prior to al-Ghazālī kalām criticisms of philosophical ideas. But one does not encounter anything like the comprehensive, sustained critique of the Tahāfut—a work entirely devoted to refuting the philosophers. Whatever it failings—some of these shown by the answer Averroës (Ibn Rushd, d. 1198) gave to it in his Tahāfut al-Tahāfut (The Incoherence of the Incoherence)—it remains a brilliant, incisive critique.

II

Abu Hāmid Muḥammad Ibn Muḥammad al-Tusī al-Ghazālī, perhaps the best known of medieval Islam's religious intellectuals, was trained as an Islamic lawyer (faqīh) and theologian (mutakallim) and became a noted Islamic mystic (ṣūfī). He was born in 1058 in the city or district of Tūs, in northeast Persia. He studied in madāris, religious colleges that focused on the teaching of Islamic law, first in Tūs, then for a short period in Jurjān on the Caspian Sea, and then in 1077 at a major madrasa in Nīshāpūr. There he was taught by Imām al-Haramayn al-Juwaynī (d. 1085), a noted lawyer of the school of al-Shāfi‘ī (d. 820) and the then leading theologian of the school of al-Ash‘arī (d. 935). In law, al-Ghazālī was a Shāfi‘ite; in kalām, he was an Ash‘arite.

Ash‘arism, by the eleventh century, was becoming the dominant school of kalām. It subscribed to a metaphysics of transient atoms and accidents, from which material bodies are composed. It regarded all temporal existents as the direct creation of God, decreed by His eternal attribute of will and enacted by His attribute of power. What humans habitually regard as sequences of natural causes and effects are in reality concomitant events whose constant association is arbitrarily decreed by the divine will. Between created things, there is no necessary causal connection—indeed, no causal interaction at all. God is the sole cause: all events are His direct creation. There is no inherent necessity in the uniformity of nature. Hence, when at certain times in history God interrupts this uniformity by creating on behalf of a prophet or holy man a miracle, no contradiction ensues. In his works of kalām, al-Ghazālī ardently defended this atomist-occasionalist doctrine on logical and epistemological grounds.

For some six years after the death of al-Juwaynī, al-Ghazālī spent much of his time at the court-camp of Nizām al-Mulk (d. 1092), the vizier of the Seljuk sultans, but seems to have also taught in Nīshāpūr. He became known as a distinguished scholar and author of works on Islamic law. In 1091, at the invitation of Nizām al-Mulk, he became the professor of law at the Nizāmiyya in Baghdād. This was the most prestigious of a number of madāris instituted by Nizām al-Mulk (hence their name Nizāmiyyas) in various eastern Islamic cities for the teaching of Islamic law according to the school of al-Shāfi‘ī. These colleges were intended in part to train scholars to counter the religious propaganda of the rulers of Egypt, the Fāṭimid caliphs. For in the eleventh century the Islamic world was divided, with two opposing caliphates—the “orthodox” Sunnī Abbasid caliphate in Baghdād, and the Shī‘ite Fāṭimid caliphate in Cairo. The caliph in Baghdād, who wielded moral and religious authority rather than actual political power, stood as a symbol of Sunnī Islam. Real power rested with the Seljuk Turks, nomadic warriors who had occupied Baghdād in 1055. But the Seljuks had converted to Islam in its Sunnī form, and their power was legitimized by the Sunnī Abbasid caliph. There was hence an Abbasid-Seljuk establishment, and al-Ghazālī's appointment at the Nizāmiyya of Baghdād made him part of it. Significantly, one of works he wrote during this period was Faḍā’iḥ al-bāṭiniyya (Scandals of the Esoterics), a critique of the esoteric (bāṭini) doctrine of the Ismā‘īlī Fāṭimids. This work was also entitled Al-Mustazhirī, after the Abbasid caliph, al-Mustazhirī, who had asked al-Ghazālī to write a refutation of Ismā‘īlī doctrine.

It was during this period, which extended from 1091 to 1095, that al-Ghazālī wrote his Tahāfut and three other works closely related to it. The first of these was Maqāṣid al-falāsifa (The Aims of the Philosophers), an exposition in Arabic that closely follows Avicenna's Persian work, Dānesh nāmeh ‘Alālī (The Book of Science Dedicated to ‘Alā’ al-Dawla). In the introduction of this work and at its conclusion, al-Ghazālī states that he wrote this work of exposition to explain the philosophers' theories as a prelude to his refuting them in the Tahāfut. (Strangely enough, in the Tahāfut there is never any mention of the Maqāṣid al-falāsifa, nor any allusion to it.) The second work, Mi‘yār al-‘ilm (The Standard for Knowledge), is an exposition of Avicennan logic, the most comprehensive of such expositions that al-Ghazālī wrote. This logic, for al-Ghazālī, was philosophically neutral, no more than a tool for knowledge, differing from the logic used by the theologians only in its vocabulary and greater elaboration and refinement. He urged his fellow theologians and lawyers to adopt it. The Mi‘yār was written expressly as an appendix to the Tahāfut. For, as al-Ghazālī proclaimed in introducing his Tahāfut, he would be using the very logic of the philosophers in refuting them. The third work is a sequel to the Tahāfut. This is his Al-iqtiṣād fī al-i‘tiqād (Moderation in Belief), an exposition of Ash‘arite theology. In the Tahāfut al-Ghazālī intended to refute and negate; in the Iqtiṣād, to build and affirm what he declared to be true doctrine, a point to which I will shortly return.

Probably around the time of his move to Baghdād, al-Ghazālī underwent a period of skepticism. As he recorded in his autobiography, written a few years before his death, he examined the various sciences he had studied but found that they did not yield certainty. Nor could he trust the senses, which, he maintained, deceive us. The faculty of sight, he wrote, “would look at the star and would see it small, the size of a dinar, but then astronomical proofs would show that it is greater in magnitude than the earth.” Distrust of the senses, he then relates, extended itself to reason. He began to doubt the basis of all reason, the self-evident truths of logic. For two months, he states, he remained in this “illness,” until in His mercy God restored to him his faith in reason.

In 1095, al-Ghazālī underwent another spiritual crisis that changed the course of his life. This came to a head in July when for a period of time he lost his ability to speak. Part of the reason, he stated in his autobiography, was that he came to realize that his motivation in pursuing his career was worldly glory, rather than genuine religious impulse. But he also hinted at a dissatisfaction with the purely doctrinal and intellectual approaches to religion. These, he maintained, bypassed the heart of the matter, that which is directly experiential in religion, the dhawq, a ṣūfī term that literally means “taste.” He had read the works of the Islamic mystics and become convinced that their path was the one that led to true knowledge. He made the decision to forsake his career and follow their path.

After making arrangements for his family, he left Baghdād and went first to Damascus, where he secluded himself in the minaret of its great mosque. He then went to Jerusalem, where again he secluded himself in the Dome of the Rock. He then traveled to Hebron, to Madina, and to Mecca. For some eleven years he lived the life of asceticism, pursuing the mystic's way. It was also during this period that he composed his magnum opus, his Iḥyā’ ‘ulūm al-dīn (The Revivification of the Sciences of Religion). In this work, as well as other shorter treatises he wrote, he strove to reconcile traditional Islamic beliefs with ṣūfī teaching. This involved his reinterpretation of what ṣūfīs declared to be the ultimate mystical experience: “annihilation” (al-fanā’) in the divine essence, a declaration that for the more traditional Muslim violated the fundamental Islamic concept of divine transcendence. For al-Ghazālī, the end of the mystical experience is proximity (qurb) to the divine attributes, which in Ash‘arite dogma are “additional” to the divine essence. The divine essence, at least in this world, remains for al-Ghazālī beyond any human experience, although he adheres to the Ash‘arite doctrine that God can be “seen” in the hereafter. He further suggests that the mystical experience of “annihilation” consists in seeing nothing in existence except the unity of all things and hence losing experience of oneself. In the Iḥyā’ he also sought a synthesis between Islamic theological principles, the Aristotelian doctrine of the mean, and the virtues expounded by the ṣūfīs, the highest of which is the love of God.

In 1106 al-Ghazālī returned to teaching, first in Nīshāpūr and then in Tūs, until his death in 1111. His writings during this period included theological and mystical works, his autobiography, and a major book in Islamic law, Al-mustaṣfā min uṣūl al-dīn (The Choice Essentials of the Principles of Religion). Needless to say, all the works he wrote after he left Baghdād, which include the voluminous Iḥyā’, are basic for understanding the religious views of this remarkable thinker and are certainly not without intrinsic philosophical interest and value. But speaking strictly from the point of view of the history of philosophy, the pivotal work remains his detailed critique of the Islamic philosophers, his Tahāfut.

III

In the Tahāfut, al-Ghazālī singles out for his criticism al-Fārābī (d. 950) and Avicenna (d. 1037) as the two most reliable Islamic exponents of Aristotle's philosophy. It should be stressed, however, that while these two philosophers were Aristotelian, they were also Neoplatonists who had formulated two closely related but quite distinct emanative schemes. There are, moreover, differences between these two thinkers, not only in their emanative schemes, but also in their theories of the soul, epistemologies, and eschatologies. At the same time, however, there is overlap in their ideas, so that many of al-Ghazālī's criticisms apply to both.

The main criticisms of al-Ghazālī, however, have Avicenna's philosophy as their direct target. Thus, to give only a few concrete examples, the third discussion includes a detailed critique of Avicenna's triadic emanative scheme, not the dyadic scheme of al-Fārābī. The doctrine that God knows only universals, or, rather, particulars “in a universal way,” criticized and rejected in the thirteenth discussion, is a distinctly Avicennan theory. Again, al-Ghazālī devotes the last three discussions of the Tahāfut to a detailed critique of Avicenna's theory of an immaterial soul that denies bodily resurrection, not the theory of al-Fārābī. There are differences between the psychological theories of these two philosophers that include a marked difference between their eschatologies. Both maintain that it is only the immaterial soul that is immortal. But while Avicenna maintains that all human souls are immortal (living a life of bliss or misery in the hereafter, depending on their performance in this life), al-Fārābī in his extant writings confines immortality to the few.

The Tahāfut divides into two parts. The first, consisting of the first through sixteenth discussions, is devoted to metaphysical questions; the second, containing the seventeenth through twentieth discussions, covers the natural sciences. Two of the philosophical theories that al-Ghazālī condemns as utterly irreligious (not merely heretical innovations) are discussed in the metaphysical part. These are the theory of a pre-eternal world and the theory that God knows only the universal characteristics of particulars. The third doctrine condemned as irreligious—namely, the Avicennan doctrine of the human soul that denies bodily resurrection—belongs to the second part. It is debated in the eighteenth through twentieth discussions but more specifically in the twentieth discussion. This second part on natural science begins (the seventeenth discussion) with al-Ghazālī's famous critique of causality and concludes with the lengthy discussion of Avicenna's psychology. In including psychology as part of natural science, al-Ghazālī follows the practice of the Islamic philosophers, who in turn follow Aristotle.

The theory of the world's pre-eternity debated in the first discussion is the longest in the Tahāfut. At the heart of this debate is the question of the nature of divine causality. As al-Ghazālī explains it, the philosophers maintain that the world is the necessitated effect of an eternally necessitating cause and hence must be eternal. At issue here is the question of whether God acts by the necessity of His nature or voluntarily. For al-Ghazālī, the doctrine of an eternal world means the denial of the divine attribute of will. The philosophers must demonstrate the impossibility of a world created in time by an eternal will, but he tries to show that they fail. At most, their theory of a pre-eternal world has not been demonstrated. It also leads to absurd consequences, he argues. Al-Ghazālī affirms that the world and time were created together at a finite moment in the past through the choice of the eternal divine will.

The second “irreligious doctrine” debated in the thirteenth discussion is Avicenna's theory that God knows particulars only “in a universal way.” It should perhaps be said here that Avicenna makes a distinction between celestial and terrestrial particulars. (The distinction is implicit but is not explicitly discussed in al-Ghazālī's otherwise masterly exposition of Avicenna's theory). For Avicenna, a celestial particular like the sun, unlike a terrestrial particular, represents the only member of its species. As such, God knows that one sun exists and knows its universal qualities. In this sense, one might be able to argue that God knows the particular sun. But the concern is with the terrestrial world, where the particular is not the only member of its species. More specifically, the concern is with the individual human and the individual human act. These, as al-Ghazālī pointedly shows, cannot in Avicenna's system be known by God individually. This theory, he argues, has not been demonstrated and plainly contradicts Qur’ānic assertions about divine omniscience.

The case is similar with Avicenna's doctrine of the soul that denies bodily resurrection. Al-Ghazālī argues in detail that the theory of the soul's immateriality, on which Avicenna's denial is based, has not been demonstrated. He then argues that even if one were to concede that the soul is immaterial, bodily resurrection would still be possible. The language of the Qur’ān affirming bodily resurrection, he points out, is explicit and must be accepted literally, not metaphorically. The interpretation of scriptural language as only metaphor is incumbent if, and only if, scriptural assertions are demonstrated—in the strictest sense of demonstration—to be impossible. Otherwise they must be accepted in their literal sense. (Elsewhere in his mystical writings al-Ghazālī insists that these statements have a deeper metaphorical and symbolic sense, beyond the literal. But this deeper sense must be based on their literal acceptance.) This criterion of demonstrability underlies the whole argument of the Tahāfut.

IV

Al-Ghazālī explains the purpose for his writing the Tahāfut in a religious preface and four short introductions. These relatively brief statements are extremely important for understanding the intention of this work. The religious preface reveals a “proximate cause” for his writing the book, as he inveighs in it against certain pseudo-intellectuals of his time. These, he says in effect, have been so impressed by such “high-sounding names such as ‘Socrates,’ ‘Hippocrates,’ ‘Plato,’ ‘Aristotle,’ and their likes” that they have become mere imitators of such philosophers and their followers, without having any real knowledge of their thought. Moreover, they have used the example of philosophers to rationalize their own disregard for the rituals and obligations imposed by the religious law, opting, in effect, for unbelief (kufr). He thus has undertaken to write this book, he states, to show “the incoherence of [the philosophers'] beliefs and the contradiction of their metaphysical statements, relating at the same time their doctrine as it actually is, so as to make it clear to those who embrace unbelief in God through imitation that all the significant thinkers, past and present, agree in believing in God and the last day.” In a tone of accommodation, adopted perhaps to stress the point that the “imitators” of the philosophers have totally misunderstood them, he states that his differences with the philosophers “reduce to matters of detail extraneous to those two pivotal points.” These “matters of detail,” however, turn out to be quite fundamental: they include the three philosophical doctrines he condemns as utterly irreligious, whose supporters, as he declares, should be punishable by death.

In the introductions that follow, he makes a number of basic points. His quarrel, he states, is not with the philosophers' mathematics, astronomical sciences, or logic, but only with those of their theories that contravene the principles of religion. His task, he further states, is not to defend any specific theological doctrine. On the contrary, in refuting the philosophers he will use against them arguments of various Islamic theological schools. His task, he explains, is simply to refute the philosophers, to show that, contrary to their claims, their theories contradicting religious principles have not been demonstrated; they have failed to fulfill the conditions for demonstration which they themselves had set down in their logical works. His assertion that the Tahāfut is intended only to refute is repeated at the conclusion of his critique of the philosophers' four proofs for the world's pre-eternity. He writes:

We have not endeavored to defend a particular doctrine, and as such we have not departed from the objective of this book. We will not argue exhaustively for the doctrine of the temporal origination [of the world], since our purpose is to refute their claim of knowing [its] pre-eternity.

As regards the true doctrine, we will write a book concerning it after completing this one—if success, God willing, comes to our aid—and will name it The Principles of Belief. We will engage in it in affirmation, just as we have devoted ourselves in this book to destruction.

Al-Ghazālī thus makes it quite clear that the Tahāfut is intended only to refute, not to defend any specific theological doctrine. He also tells us that he will write a sequel to the Tahāfut in which he will affirm true doctrine. These two points call for comment.

V

In his famous response to al-Ghazālī, his Tahāfut al-Tahāfut (The Incoherence of the Incoherence), Averroës repeatedly refers to al-Ghazālī's arguments as Ash‘arite. And it is true that more often than not al-Ghazālī argues from an Ash‘arite theological base and affirms Ash‘arite theological positions. But this is his prerogative. When he does this, he does it as part of his endeavor to refute. He does not develop a theological system in the Tahāfut. In one of his later works, Jawāhir al-Qur’ān (The Gems of the Qur’ān) he mentions the Tahāfut with two Ash‘arite works as books intended to defend the faith. But while intended to defend the faith, the Tahāfut, strictly speaking, is not an Ash‘arite “manual,” a systematically argued presentation of Ash‘arite doctrine. Moreover, al-Ghazālī sometimes defends non-Ash‘arite positions in this work. He defends them, to be sure, as I will shortly indicate, for the sake of argument, as a means for refuting the philosophers. Still, he defends them. Before turning to this point, however, a word is necessary about the title of the work he wrote intended as a sequel to the Tahāfut, the work mentioned in the quotation above. The title he gives is Qawā‘id al-‘aqā’id (The Principles of Belief). An Ash‘arite work bearing this title constitutes one of the books of al-Ghazālī's Iḥyā’. But the work that best fulfills the purpose stated in his Tahāfut is another Ash‘arite work—namely, Al-iqtiṣād fī al-i‘tiqād (Moderation in Belief). It is not that the two works do not complement each other; but the Iqtiṣād, written in Baghdād shortly after the writing of the Tahāfut and before the Iḥyā’, refers directly to the Tahāfut and is closer to it in spirit and in terms of the issues with which it deals. Significantly, in the Iqtiṣād, al-Ghazālī states that its concern is with qawā‘id al-‘aqā’id, “the principles of belief.” Perhaps more to the point, his statement that he will write a book in which he will engage “in affirmation” rather than “destruction” occurs, as already noted, at the end of the critique of the philosophers' first proof for the pre-eternity of the world. This statement comes after his assertion: “We will not argue exhaustively for the temporal origination [of the world] since our purpose is to refute their claim of knowing [its] pre-eternity.” Such exhaustive argument will be included in the book of affirmation he intends to write. The question of the world's origin, however, is not discussed exhaustively in the Qawā‘id, although one of the Tahāfut's arguments against infinite events in the past is repeated. The discussion in the Iqtiṣād is by far the more detailed. In it al-Ghazālī repeats his arguments in the Tahāfut for the temporal creation of the world much more fully, although admittedly he does not add much to them. It is the Iqtiṣād, not the Qawā‘id, that is the sequel to the Tahāfut.

This returns the discussion to the question of al-Ghazālī's defense of non-Ash‘arite positions. The two most important instances occur in his discussions of causality and of the doctrine of the soul's immateriality. I will begin with the causal question.

There are abundant statements in the Tahāfut indicating that its author subscribes to the Ash‘arite occasionalist doctrine that confines all causal action to God. Nonetheless, in the seventeenth discussion, in which al-Ghazālī argues for the possibility of certain types of miracles whose possibility is rejected by the Islamic philosophers, he defends two different causal theories. He begins with his famous declaration and defense of the Ash‘arite causal theory: “The connection between what is habitually believed to be a cause and what is habitually believed to be an effect is not necessary, according to us.” The connection between such events, he then states, “is due to God's prior decree, who creates them side by side, not to its being necessary in itself, incapable of separation.” Observation, he then argues, shows only concomitance—never any necessary causal connection. In the course of debating this question, however, al-Ghazālī puts into the mouth of his opponents an objection to this denial of necessary causal connection. According to this objection, such a denial leads inevitably to absurdities, to a confused, haphazard, irregular course of events. Al-Ghazālī answers quite emphatically that such a consequence does not follow. For God in His goodness has ordained a regular natural pattern (which in itself is not necessary but is inevitable because it is decreed by God) and has created in humanity knowledge that this uniformity will continue and will only be interrupted when a miracle takes place—an event also decreed and created by God. The opponent's contention that absurd consequences would result from the denial of causal necessity in things, al-Ghazālī declares, is hence an exercise in “sheer vilification” (tashnī‘maḥḍ). To avoid being subjected to such vilification, however, al-Ghazālī proposes another possible causal theory that would still allow the possibility of those miracles the philosophers deem impossible.

This second causal theory proposed by al-Ghazālī is perhaps best described as a modified Aristotelian theory. Created things would have causal efficacy, provided one maintains that the divine act remains voluntary, not necessitated by the divine essence. Moreover, according to this second theory, divine power is such that it can intervene in the natural order, creating new causal conditions that produce the miracle. Now al-Ghazālī elaborates this theory and insists that it is also possible. Hence he holds that there are two possible causal theories that allow the possibility of miracles. But while he holds both to be possible, he does not state that they are compossible. For they are mutually exclusive.

To which, then, of the two theories does al-Ghazālī actually subscribe? It is here that the Iqtisād provides the answer. For in it al-Ghazālī affirms without any equivocation the Ash‘arite causal theory. Divine power is pervasive and is the direct cause of each and every created existent and each and every temporal event. Inanimate things have no causal power (a point he also asserts more than once in the Tahāfut). Power belongs only to the animate. But this is a power which God “creates” in humanity. It is created power. Created power, however, has no causal efficacy. It is created simultaneously with the event human beings ordinarily, but erroneously, regard as the effect of their created power. This effect is also the direct creation of divine power. (This is an expression of the Ash‘arite doctrine of acquisition [kasb] and finds corroboration in the Qawā‘id). Human power and the effect associated with it are both the simultaneous creations of God. Causal action belongs exclusively to divine power. This necessarily negates the second causal theory, the modified Aristotelian theory, introduced in the Tahāfut. Whatever other reason there might be for introducing it in the Tahāfut, it is clearly introduced there for the sake of argument, to demonstrate that even if one allows a measure of causal efficacy in things, one can still allow the possibility (denied by the philosophers) of certain kinds of miracles.

Turning to the question of the immateriality of the human soul, al-Ghazālī, as mentioned earlier, devotes an entire section (the eighteenth discussion) to it. He offers a detailed refutation of the proofs Avicenna had given for its immateriality, arguing that none of these proofs (ten in number) has demonstrated such an immateriality. The doctrine of an immaterial soul is the basis for Avicenna's insistence that there is no bodily resurrection. In the twentieth discussion, devoted to the question of bodily resurrection, however, al-Ghazālī does not choose to defend the Ash‘arite doctrine of the human soul that denies its immateriality. Instead, he strives to show that even if one accepts a doctrine of an immaterial soul, one can maintain the possibility of bodily resurrection. Moreover, he states that a doctrine of an immaterial soul need not be inconsistent with Islamic teaching, that it has support in the Qur’ān. His presentation of the possibility of an immaterial soul is so persuasive in the Tahāfut that one is prone to believe that this is the doctrine to which al-Ghazālī subscribes. In the Iqtiṣād, however, he defends an Ash‘arite doctrine of a material human soul and states quite explicitly that in the Tahāfut he defended a doctrine of an immaterial soul only for the sake of argument, in order to refute the philosophers.

These are the two notable instances where he defends a doctrine to which he does not actually subscribe. It is also clear that he defends them for the sake of argument, as a means to refute the philosophers. In this he is not inconsistent with his declarations that the Tahāfut's primary aim is to refute.

VI

The term tahāfut has been variously translated—for example, as “destructio” by the Latins, “inconsistency,” “disintegration,” “collapse,” as well as “incoherence,” by modern scholars. A common meaning is “collapse,” or “collapsing,” sometimes with the nuance of rushing headlong and crowding to fall into disaster, into hellfire. It also is used to convey the idea of rushing and swarming into combat. The term also relates to haft, discourse that is not well thought out, that is unintelligible, incoherent. M. Bouyges, in the introduction to his edition of the text, gives a succinct discussion of the ways this term has been translated in English and other languages. He chose “incoherence” as perhaps conveying best what al-Ghazālī meant by the term tahāfut. His reasons for this choice are quite convincing and hence I have followed him in the translation of the term.

There have been two English translations of al-Ghazālī's Tahāfut. The first, published in 1954, is that of S. Van Den Bergh in his translation of Averroës's answer to al-Ghazālī—namely, the Tahāfut al-Tahāfut (The Incoherence of the Incoherence)—which embodies most of al-Ghazālī's text. This translation (2 vols., London, 1954), with its copious notes, is a veritable tour de force and is certainly a major contribution to the study of both al-Ghazālī and Averroës. At the same time, however, it is not free from serious errors. The second translation, by S. Kamali (Lahore, 1958), has for years served as a main introduction of this work to the English reader. Kamali's English version has the merit of conveying much of the argumentative flavor of the original—this in itself being a considerable achievement. This version, however, also has its share of inaccuracies and at times is more of a paraphrase than a translation.

A text sometimes poses difficulties for the translator, not because its author is a bad writer, but because its author is a good one. Al-Ghazālī is a master of Arabic prose. His style, however, is very personal and highly idiomatic; it carries with it nuances that are difficult to recapture in a translation. As such, the difficulties it often poses are not so much due to lack of clarity. For the most part, his presentation of complex and subtle arguments is remarkable for its clarity. But there are also lapses. Ambiguities do occur. And there are times when what is stated is so condensed that its intention is not immediately clear. There is also an ambiguity, frequently encountered in Arabic, relating to the use of pronouns, where the referents of these pronouns are not always immediately obvious. Hence, in translating the text, there is an ever-pressing need for clarification. In places where lengthy explanations are needed, I have placed these in notes. But there are numerous places where lengthy explanations are not needed. To place these in notes imposes unnecessary interruptions to the flow of the argument. For this reason, to clarify issues, I have made extensive use of square brackets. These include words or sentences which are not explicitly stated in the text but are implicit therein.

The Arabic text of the Tahāfut is based on the edition of Maurice Bouyges (Beirut, 1927). I have, however, introduced new paragraphing, dictated by sense and flow of argument and determined also by the manner in which I thought it was best to paragraph the translation. To help the reader, I have also added considerably to the sparse punctuation of the original edition, and also to the sparse gemination marks, the shaddas, over the letters. On the whole, I have followed the readings of Bouyges adopted in the body of the texts, but there are departures. I have chosen other manuscript readings given in Bouyges's apparatus criticus, the choice dictated by grammar and sense. These departures from Bouyges's edition have been placed between square brackets, with a note indicating the reading in the body of Bouyges's text. Some changes—and these are few—dictated by language and sense, are my own. These are placed in angular brackets, with notes indicating the reading in Bouyges's text.