Rationalism is one of the main issues which tends to be raised in interpreting Averroes' thought. Whoever tries to define Averroes as a rationalist, makes him basically a follower of the Greek heritage (and so, in the end, a "Western" scholar). Indeed, Averroes himself offered the opportunity for this assumption because of his attempt to harmonize Islamic truths with Western philosophical (i.e. Aristotelian and Platonic) outlooks. When we consider his works, however, we come across an indelible Islamic framework Averroes is not able to pass over even when strictly philosophical questions are concerned.

            Averroes' rationalism has been widely discussed before today. I already pointed out my dissatisfaction with an uncompromising rationalistic solution like that of L. Gauthier, M. Cruz Hernández or Martínez Lorca - and I shall not repeat my arguments here (1). Even Muhammad Abid al-Jabiri's enthusiastic defence of a progressive modern Averroes has in a sense missed the point. Al-Jabiri is completely right in showing Averroes' proper position within the Almohad ideology and its reformistic political programme (2). However, to put Averroes in the mainstream of Zahirism as though he were a follower of Ibn Hazm seems to me, on the whole, not always convincingly supported, especially considering his role as a Malikite jurist and qadi and his care in trying to legalize philosophy, that is to make philosophy available in a religious context (3). On the other side, al-Jabiri's contention that Averroes' rationalism lies mainly in his refusal to translate the known into the unknown (4) is rather simplistic, even though explicit hints are to be found in the Tahafut at-Tahafut. For instance when Averroes argues that the theological definition of God is not self-evident, unless we move from the known to unknown (5).

            I believe that the Tahafut at-Tahafut is a useful pointer to Averroes' Islamic commitment. It stands with Fasl al-Maqal and Kashf ‘an Manahij al-Adillah as the pivotal constituent of an entirely Muslim philosophy. My aim is neither to undervalue nor to undermine the Averroistic concept of rationality, but rather to make evident his failure in offering a holistic interpretation of philosophy solely from the point of view of rationality. Accordingly, the specifically Islamic concept of reason - founded on a legal and dialectical rather than a metaphysical and apodeictical basis - should be better appreciated. Our approach here will be more limited, though.

            It is well known that some perplexities arise about the translation of the title of the Tahafut at-Tahafut. In the Middle Ages, the Latin translator Calo Calonymos rendered it as Destructio destructionum, but destructio does not capture exactly the nuances of tahafut. To be sure, Incoherence of the Incoherence is better. Neither al-Ghazali nor Averroes wished to demonstrate alternative theses with respect to their adversaries (al-Ghazali vs. Avicenna; Averroes vs. al-Ghazali); but rather they wished to point out the incapacity of a disputant to support his own position. This implies that dialectic and not apodeictic demonstration is the true method of the Tahafut at-tahafut. Averroes himself says that the discourses of the Tahafut at-Tahafut are not fully demonstrative, but they only hints at the truth [pp. 47, 60, 69...].

            This methodological perspective does shift, intentionally, the debate from an onto-epistemic to a linguistic level. This happens or because of a view like that of Wittgenstein in the Tractatus logico-philosophicus, that we must be silent when we are not able to speak about a particular topic; or because of the arrangement of Being into a linguistic level. I mean that, in the latter case, it would be possible to work out the unavoidable aporias of metaphysics on a level of formal concordance between reality, mind and language. It is assumed here - but not proved of course - that the first solution, that nearer to Wittgenstein, is al-Ghazali's: it implies the final defeat of philosophy and a turn towards mystical fideism. The second solution is that of Averroes, who stresses the concordance less between thought and Being, than between Being and its predication, that is its linguistic formalization. So, against Wittgenstein, we can speak of everything and everything which we are speaking of - in obedience to the rules of logic - has a correspondence in reality.

            These premises, in my view, allow a better understanding of the structure of both the treatises (the Incoherence and the Incoherence of the Incoherence). We could choose as an exemplification of this thesis many topics, but let us concentrate on the well known question of the eternity of the world.

            In order to explain the philosophical doctrines concerning eternity ["First discussion" in Van den Bergh's translation, pp. 4-10 Bouyges' edition], al-Ghazali argues, on behalf of the philosophers, that the contingent cannot stem from the eternal and that, if the world was born in time, it would have needed a determining cause which sometimes did act, but sometimes did not. This latter option, with regard to an all-powerful and ever-creating God, would be impossible. Averroes disputes the philosophical character of al-Ghazali's discourse saying that: 1) the term "possible" is equivocal; 2) there are different kinds of possibility, and 3) each of these has different determining causes. Averroes disputes his adversary not on the ground of substance, but on the ground of form. Al-Ghazali is loose in his language, applying the same terms indifferently to man and God. On the contrary, these terms are not interchangeable if the subjects which they refer to are unlike each other. Also the term "eternal" is used in different ways, so that al-Ghazali's argument does not work, largely on the ground that it misinterprets issues of meaning.          

            Al-Ghazali goes on without offering any alternative "demonstration". He confines himself to wondering rhetorically why the philosophers claim not to believe that the world has been produced by a divine will, so that before it did not exist and after it came into existence. Answering this question, Averroes launches into a long philippic arguing that an effect can have a delay in relation to the will, but not in relation to the act; that al-Ghazali's suggestions do not rule out the issue whether acts give rise to changes in the Agent - provided that the eternal Agent is unchangeable -; and that, however, all is difficult to explain ( ‘asir al-bayan)...! In any case, we are not able to consider as "will" something that eternally distinguishes between two contraries (specifically, existence and non-existence) because the will naturally gives up after having distinguished between some choices: so, al-Ghazali does not respect the meaning of "will", considering divine will and human will interchangeable on a semantic plane. The matter is of especial interest to Averroes who is convinced that we can know the substance of a thing through its definition: definition is not only a mere linguistic factor, but an aspect of Being (6). Then, if a theologian retorts that the eternal divine will does not give up after the act of willing, Averroes, speaking as a philosopher, would argue that the religious Law (Shari‘ah) interprets equivocally and indiscriminately "will" either as a natural or a voluntary choice. This shows the existence of intermediate positions between two contraries or extremes. We realize that the religious Law is able to address any sort of man, even though the philosophers have at their disposal  apodeictic demonstration in order to get rid of any possible linguistic trap.

            It is worth defining this point more precisely. The problem of voluntary or non-voluntary action is faced again in the so-called third discussion [pp. 148-156]. Averroes denies that the antropomorphic idea of a willing, powerful, conscious God is an obvious definition; rather, it requires a demonstration. The First Agent transcends the attribute of will because someone who wills is in want of the object of the willing, while God cannot lack anything. Nor could God's action be natural, because a natural action is thoughtless, while God is an Agent because of His knowledge. If we distinguish between a voluntary and a non-voluntary act, we have no need to underline - as al-Ghazali does - that the former involves science and the latter does not. For action means to transform non-existence into existence, so that such an action does not need science. Averroes maintains that the true definition of agent is "someone who does transform non-existence into existence", from potentiality to actuality. An Agent working like an emanating deity does not necessarily need to know what he produces.

            We can distinguish between a voluntary and a non-voluntary action only if we consider a true act - the passage from non-existence to existence -, not a metaphorical one. Averroes quotes here [p. 156] the Qur’an (18:77): “a wall which wanted to fall to pieces”, to argue that this kind of action is only metaphorical because it does not involve a passage from non-existence to existence. The theologians seem here to be right in referring this kind of action directly to God and not to the wall itself; but Averroes argues that “the opponents of theologians might reverse the argument against them and say [...] that "voluntary act" is a metaphor” [p. 158], just because “No one ever says "He saw with his eye, and he saw without his eye" in the belief that this is a division of sight; we only say "He saw with his eye" to emphasize the fact that real sight is meant, and to exclude the metaphorical sense of "sight"”. Accordingly, we could conclude that creation is for Averroes only a metaphor, as it is a metaphor to describe the real Agent as a knowing or willing Being.

            The analysis of eternity follows in the Tahafut's "First Discussion" at [pp. 34-41]. Al-Ghazali is persuaded that philosophers, discussing the divine will, are trying to submit reality to dialectic by means of the removal of the the same reality on a linguistic-conceptual plane. Certainly, “by permission of the Divine Law (Shari’ah)”, we are allowed to use the word "will", but the issue is more substantial for the theologians: does there exist something, like an attribute, which is defined as the "will" of God? If so, it cannot but consist in the capacity of distinguishing between two similar things, and this due to its freedom. Averroes agrees that the word "will" can be used instead of another one, “by permission of the Divine Law”, to solve the clumsy device of ascribing to God a specific human attribute: what is, in fact, substantially the "will" of God? For Averroes, however, the problem is not the choice between two similars, but between two opposites and contraries, i.e. being and non-being.

            Al-Ghazali makes God resemble a hungry man who has in front of him two dates and who freely chooses to eat one of them leaving the other. Averroes retorts that the example is erroneous: the man in the story does not choose between two similar things, but rather he is facing a dilemma: to take one of the dates or to take none; in short he can have his fill or he can die of starvation. As for God, the alternative is not between the creation of this world and the creation of another world similar to the first, but between creation and inactivity. Since that inactivity is inconceivable for God, we must deduce that the world is eternal. For al-Ghazali, the eternity of the world would involve God's impossibility to act; for Averroes, if the world is merely contingent, the infinite power of God would be denied.

            Anyway, al-Ghazali is wrong in his example of the hungry man, because hunger needs a constraint: if God is like that man, it would mean that his choice between one world and another compossible one has been compelled by an external factor which leads Him to make a distinction. But Averroes, on the other hand, is wrong in mistaking two similar things for opposites: from a theological point of view, God has in front of Him a plurality of equivalent choices, not the horn of a dilemma. Displacing the problem on the level of a clear-cut alternative between opposites, Averroes is right in arguing that the will in se cannot mean a choice between two homologous facts: if I am thirsty, to drink one sort of beer rather than another of a different brand is irrilevant. Anyway, he betrayes al-Ghazali's fiction; and al-Ghazali, on his part, is right in assuming that, if God is omnipotent, His will does not concern two contraries, but two equivalents, because both possibilities are feasible for Him.

            Averroes' objection shows how much he was loathe to recognize a plural Being and an open epistemology. Discussing the relation between movement and time, we appreciate that, in al-Ghazali's view, time exists only in relation with timed realities, i.e. things connected in a temporal order: it is not an in se category. Imagination, he says, is unable to “imagine the beginning of a thing without something preceding it, and this "before" of which the imagination cannot rid itself is regarded as a really existent thing, namely time” [p. 72]. Accordingly, imagination is unable to imagine a space without other space beyond it. It happens because, in al-Ghazali's view, the ontological structure of the world is escaping from a stiff gnoseological determination, so that it is potentially changeable. As Nietzsche would have put it, truth is curvilinear. On the contrary, Averroes replies that truth is absolutely stright because our mind cannot but adhere to reality as it is, translating it into positive science. Averroes says that “the necessary connexion of movement and time is real and time is something the soul (dhihn) constructs in movement” [p. 74]. It is impossible to convert the necessary into the possible. Averroes charges al-Ghazali with confusion between time and space, a mistake all “who do not start their inquiry in a proper scientific order” [p. 78] easily fall into. The scientific order does correspond with the factual order of the universe: how we speak of the universe agrees exactly with what - that same universe - we are speaking of; so that, while al-Ghazali is able to say that "above" and "below" are relative (mudaf), Averroes repeats that there are natural directions in consequence of absolute weight and light [pp. 81-82].

            Al-Ghazali is ready to accept a relativistic perspective: if a reasoning is universally valid, everyone would agree with it, but we see that it does not happen thus with respect to God and world. Averroes contends that truth and reality do not stand opposed to demonstration and cannot be denied (and vice versa what cannot be refuted, exists) [pp. 13-15]. Again: Averroes replies to al-Ghazali's statement that, from a theological point of view and considering God's omnipotence, there are infinite possibilities (imkanat la nihayah laha), arguing that possibilities reside in the necessary properties of the things, namely possibility does not exist in itself [pp. 31-33].

            The debate about possibility is one of the most intriguing epistemological points [pp. 92 ss]. For al-Ghazali possibility is present in existence, it is a characteristic element of the  being of things, not a presupposition - a priority - of the becoming process. So, we must consider it as a maqdur, something subdued to God's arbitrary action and intervention. Al-Ghazali means that even God cannot reverse logical impossibility. He is perfectly able to make possible in some situations things which are impossible in other situations, and vice versa. Thus, apart from any logical restriction, possibility and impossibility are peculiar variables of the ontological meaningless of creatures in relation to God who stays infinitely beyond them: for possibility is like contingency [p. 98] and, moreover, the Intellect, in its judgements, is by no means obliged to admit that a possible entity must exist [p. 104].

            Averroes thinks that possibility is merely the state of non-existence; it is the premise allowing non-being to become being; in a clear context, Averroes links possibility to takawwun [p. 105], a very meaningful term I translated as “farsi essere” (“becoming being”) (7). After that, possibility becomes necessity, necessity of existence. To deny the priority of possibility does mean to deny necessity, because actuality, i.e. factual existence, is complementary to possibility. Averroes enunciates precisely the principle of plenitude: “the possible is the contrary of the impossible without there existing a middle term, and, if a thing is not possible before its existence, then it is necessarily impossible” [p. 94]. Meanwhile, however, he fluctuates between the logical and the ontological. Averroes agrees that an agent (and God in particular) cannot do what is logically impossible; but denies definitely that the impossible can become, in some occurrence, existent. So, “if it is assumed [by the theologians] that the world was impossible [to be made] (mumtani‘) for an infinite time before its production, the consequence is that, when it was produced, it changed over from [logical] impossibility (istihalah) to [existing] possibility (imkan)” [p. 94-95]. In Averroes' mind, this transference from one plane to another is intolerable: what is logically impossible will never become a possible existent ready to convert itself in actuality, and so become necessary.

            Averroes acknowledges that al-Ghazali views an infinite possibility of existence depending upon God's will, but retorts that this assumption would generate a consecutive chain of factually numerical infinite worlds -  not logically contemporary worlds. This is an absurd conclusion in Averroes' mind [pp. 98-99].

            As we can see, the dialogue is between the deaf. Al-Ghazali's aim is to suggest that philosophy "is not able to...": to demonstrate its propositions, to provide certainty, to know God and his hidden wisdom. Averroes replies that, on the contrary, philosophy is fully able to, mainly because of the greater accuracy of its language; and at the linguistic level, we cannot refute or even discuss the reality and cogency of Being. Everything depends upon an "ostensive" truth of knowledge, because, as Averros said in the Fasl al-Maqal, “assent to a thing as a result of an indication [of it] arising in the soul is something compulsory, not voluntary” (8). He does mean that assent (tasdiq) compelled by demonstration (burhan or dalil) is the standard by which philosophy can reach an inner self-verification.

            Averroes however charges al-Ghazali with deception for having claimed that self-known things have actually another meaning, because intelligibles are known as such because of their obviousness [pp. 30-31].

            Averroes thinks that there is a special correspondence between subject and object and he does repeat this belief in several passages of the Tahafut al-Tahafut: for instance, “all true intellectual concepts need a thing outside the soul, for truth, as it has been defined, is the agreement of what is in the soul with what is outside the soul” [p. 103]. Again: “as to the existent which has the meaning of the "true", all the categories participate in it in the same way, and the existent which has the meaning of the "true" is something in the mind, namely that a thing is outside the soul in conformity with what it is inside the soul” [p. 304].

            Averroes says that, if the existent is also true, existence will correspond to essence [p. 400]: in other words, if discourse and logic speak rightly of reality, there will be no split between what a thing is and how it is. So, language can describe adequately the structure of the world. The universe is eternal either from the point of view of rational demonstration or from the point of view of factual ontological reality. Eternity is the objective reality of the universe and, moreover, is a definition to express in a term the complex process of everlasting production [p. 162]. This  everlasting production is in fact the true relation between the Creator and the world, even though “the philosophers only call the world eternal to safeguard themselves against the word "product" in the sense of "a thing produced after a state of non-existence, from something, and in time"” [p. 162]. It is a conclusion repulsive from an orthodox religious point of view, but Averroes is comfortably sure that religion and philosophy are not at odds.

            The problem now is the following: does all what we explained happen because “our language is flexible enough to capture [the] diversity of view”, so that

            “when [Averroes] tries to reconcile apparently contradictory views his strategy is to argue that all these views are acceptable as different aspects of one thing. [...] In his tentative remarks on language Averroes suggests that this conflict comes down to a stress upon different aspects of one thing, namely, the way the world really is” (9)?

            Or must we reduce Being at a linguistic level in order to by-pass the inherent philosophical difficulties concerning the nature of that Being?

            Averroes is really frightened by the possibility of breaking off the linkage between Being and thought, a linkage which he is not able to demonstrate apodeictically even though he is sure of its occurrence; he tries to demonstrate that language is functional in order to translate one into the other. There are different kinds of method to attain this end. The Tahafut's solution has been sketched in order to show the prevalence of language over metaphysics. Another interesting solution is offered by Averroes in the Fasl al-Maqal, a treatise strictly linked to the Tahafut.

            We shall consider the same problem with the eternity of the world. Is the world created or eternal? The question is raised and keenly faced in the Fasl and we shall not repeat in detail Averroes' arguments. He concludes that we can define the universe as both created and uncreated: created in relation to God who is a being ontologically superior; uncreated in relation to time because universe is not preceded by time even though is brought into existence by something (10).

            The epistemological equivalence of the two qualities and the two propositions ('p' and 'non-p') from the point of view of truth's unicity is a linguistic deception. As long as Averroes cannot demonstrate that the world is created, he tries to suggest that eternity is tantamount to creation. As long as he cannot demonstrate the ontological reality either of eternity or of creation, he succeeds in showing that they are similar from a linguistic point of view.

            I suggest that a subtle jeu linguistique underlies this argument and that it can suggest the ontological unreliability of philosophical contention (11). Perhaps, I did not adequately stress that Averroes' discourse in the Fasl al-Maqal pointed at the general consensus (ijma‘) more than at demonstration in itself, as de Libera argues discussing the part of the treatise concerning the eternity of the world and concluding that “les philosophes, s'ils existent, appartiennent aux ‘ulama_ dont l'accord est indispensable à l'établissement d'un consensus” (12). For, as far as the primary principles of faith are concerned, the ijma‘ of philosophers and ‘ulama_ as well must command that the obvious sense (zahir) of the Holy Text must be retained. Now, even though the reality of a linguistic approach to metaphysics has to be kept as the main interpretative philosophical key of Averroes' thought, the previous remark is useful to establish an even more Islamic concept of Averroes' main position.

            Is being something of real or is it merely a linguistic convention? Undoubtedly, Averroes thought that Being is an absolute reality. Philosophy, however, is not able to prove this assertion: al-Ghazali is right in saying that philosophy is too ambitious and, in the end, ineffective. In my view, Averroes was completely aware of the difficulty. His endeavour to save philosophy is sincere, but rationalism is mocked: at the end, we need faith. For Averroes, the Holy Qur’an can provide the key to a revealed - and so unquestionable - foundation of logic and science, even because language, in the Qur’anic perspective, is not conventional, but rather, being created by God, it is an absolute tool for describing reality. I venture to suggest that the Qur’an does represent a complete system of language whose result - in quite Heideggerian terms - is to open the real meaning of Being.

            If it was not so, we would not be able to justify the use of the Holy Text in the Tahafut at-Tahafut. For instance, when Qur’an 21:22 (“If there were in heaven and earth gods beside God, both would surely have been corrupted”) is evoked to support that only One principle rules the multiplicity of events [p. 177]; or when Qur’an 41:11 (“He inspired every Heaven with its bidding”) is evoked to support the ordered structure of universe and of the moving Intelligences [p. 186 and 191]. The perfect functioning of universe is reflected in the providential care of God who reveals in the Qur’an: “And the sun and the moon and the stars are subjected to His bidding” (16:12) (13).

            The route we followed regarding the eternity of the world could be repeated again, for instance regarding the relation between multiplicity and the One or all the emanationistic system. Our purpose has been only to suggest in a viable way that Averroes' rationalism is more difficult to grasp than is generally believed; and that we must approach his work keeping in mind that he is first of all a Muslim thinker.



 (1) L. Gauthier, Ibn Rochd, P.U.F., Paris, 1948; M. Cruz Hernández, Averroes. Vida, Obra, Pensamiento, Influencia, Caja de Ahorros de Cordoba, Cordoba, 1986; A. Martínez Lorca, Al Encuentro de Averroes, Trotta, Madrid, 1993. I discussed Averroes' rationalism in both the Introductions to Averroè, Il Trattato Decisivo sull'accordo della religione con la filosofia (Kitab Fasl al-Maqal), Rizzoli, Milano, 1994, and Averroè, L'Incoerenza dell'incoerenza dei filosofi (Tahafut at-Tahafut), U.T.E.T., Torino, 1997.

 (2) See D. Urvoy, Ibn Rushd (Averroes), Routledge, London and New York, 1991.

 (3) M. ‘A. al-Jabiri, Bunyah al-_Aql al-‘Arabi, Markaz Dirasat al-Wahdah al-‘Arabiyyah, Beirut, 1992.

 (4) M. Abed al-Jabri, Introduction à la Critique de la Raison Arabe, La Découverte, Paris, 1994.

 (5) Ibn Rushd Tahafut at-Tahafut, texte arabe etabli par M. Bouyges, Dar al-Mashreq, Beirut, 1992, p. 220; and cfr. p. 425 for a meaningful quotation: “If the distance between eternity and non-eternity is greater than that between the various species, how then is it possibile to apply a judgement about the empirical world to the invisible: for those two are opposite extremes? And when you have understood the sense of the attributes which exist in the visible world and those which exist in the invisible world, it will be clear to you that through the ambiguity of the terms they are so equivocal that they do not permit a transference from the visible to the invisible” (English translation by S. Van den Bergh, Averroes' Tahafut at-Tahafut (The Incoherence of the Incoherence), Luzac, London, 1954, p. 256. I remember that it exists a my own Italian translation, L'Incoerenza dell'Incoerenza dei filosofi, already quoted in note 1. From this moment onward, I will quote in the text between square brackets the number of pages of Bouyges' edition. The English translation is always by Van den Bergh.

 (6) See the Commentary on Metaphysics, Book IV, Comm. 13.

 (7) My translation, p. 155. Van den Bergh simply: “becoming” (p. 62).

 (8) G. Hourani, Averroes on the Harmony of Religion and Philosophy, Luzac, London, 1976, p. 57.

 (9) O. Leaman, Averroes and his Philosophy, Clarendon Press, Oxford, 1988, p. 196.

 (10) Hourani, Averroes on the Harmony of Religion and Philosophy, p. 55.

 (11) In the Introduction to Il Trattato Decisivo (see note 1).

 (12) A. De Libera, "Introduction" to Averroès, Discours Décisif, Flammarion, Paris, 1996, p. 39.

 (13) English translation by N. Dawood, The Koran, Penguin, Harmondsworth, 1976.


E-mail: webmaster - Sign the Guest Book - Ibn Rushd Home

Page Information: