The Classical Islamic Arguments for the Existence of GodNovember 1, 2007 at 4:45 pm | Posted in University of Philosophy, X-Files | Comments Off on The Classical Islamic Arguments for the Existence of God
After Wensinck’s brilliant study, a fresh examination of the argument for the existence of God in Islam might appear impertinent. Some justification for the present discussion, however, may be found in the fact that some of the material on which this study is based was not available to Wensinck, when his monograph appeared in 1936, and in the slightly different interpretation of certain relevant data here attempted.
The systematic examination of the proofs of the existence of God should be preceded by a legitimate enquiry: Is the demonstration of God’s existence possible at all? In the Latin scholastic treatises of the Middle Ages, as for example in the Summa Theologica of St. Thomas Aquinas (d. 1274) this enquiry figures as the prelude to the demonstration of God’s existence proper. Although Wensinck has discussed some aspects of the problem of knowledge (erkenntnislehre) in his celebrated Muslim Creed, he does not touch upon this particular aspect of the problem in his monograph, except incidentally, as, for example, in connection with Al-Ghazali’s attitude to the question of God’s existence. But this question, it would seem, requires a fuller treatment than is accorded it in that parenthesis.
In his two little tracts; Fasl al-Maqal and al-Kashf ‘an Manahij al-Adillah, Ibn Rushd (d. 1198) raises this question in a systematic way. In the former tract, he is concerned with a wider problem: viz. Whether the philosophical method tallies with the teaching of revelation or not – to which he replies in the affirmative. “for if the aim of philosophy,” he writes, “is nothing other than the consideration of existing things and their examination, in so far as they manifest the Creator –viz. in so far as they created objects… revelation (al-shar‘) definitely enjoins the consideration of existing things and commends it” – a thesis which he supports by a wealth of Quranic quotations. When he returns to this question at the beginning of Al-Khasf, he distinguishes between three schools of thought on the specific problem of God’s existence: The Literalist who reject rational argument altogether and claim that God’s existence can be known by means of authority (al-sam‘) only. The Ash‘arites (with whom he includes the Mu‘tazilites) who admit the possibility of a rational demonstration of the existence of God from the concepts of temporality (huduth) or contingency (jawaz), as we will see later and finally the Sufis who claim that we apprehend God directly but “whose method,” as Ibn Rushd observes, “is not speculative at all” and which, even if its validity is conceded, is not common to all men.
The earliest systematic discussion of the problem of knowledge (erkenntnis) as a prelude to theological discussions which has come down to us is found in Al-Baghdadi’s (d. 1037) Usul al-Din. It is possible that Al-Baghdadi continues a more ancient tradition, initiated by the Mu‘tazilite doctors of the 9th century, as their preoccupation with such abstract questions as notions (ma‘ani), science (‘ilm), etc. suggests. But it is significant that al-Baqilani (d. 1013), who is credited by some ancient authorities with having refined the methods of Kalam, does not dwell on this question at any length in the opening chapter of his Tamhid.
The introductory chapter of Usul, to which Wensinck has drawn attention and discussed in some length in The Muslim Creed, is thus of considerable importance for the understanding of the Islamic approach to the question of knowledge or science.
We cannot dwell at length here on Baghdadi’s analysis of the divisions of knowledge (‘ilm), its presuppositions, the conditions of its validity, etc. which are genuinely reminiscent of Kant and the subsequent schools of modern epistemology. On the particular issue with which we are here concerned, it should be noted that Al-Baghdadi defines demonstrative knowledge “by means of reason” and instances “the knowledge of the temporality of the world, the eternity of its Maker, his unity, his attributes, his justice, his wisdom and the possibility (jawaz) of religious obligations (taklif),” etc. In further expounding the objects of knowledge, as distinct from the objects of revelation (al-shar‘), he states that the Ash‘arites (ashabuna) hold that reason is capable of proving the temporality of the world and the unity of its Maker, etc. as well as the admissibility in reason (jawaz) of what is possible and the inadmissibility of what is impossible, but adds significantly that religious obligations or prohibitions arising therefrom are not known by reason but only by revelation. Hence were one to arrive at knowledge of God, the creator of the universe, etc. prior to revelation by means of natural light of reason he would be “a believing monotheist” but he would not thereby deserve any particular reward; so that if God were to reward him in the life-to-come, such reward would be an act of divine grace. The Mu‘tazilah, on the other hand, argue that man was capable of discriminating between good and evil, prior to revelation, and was in proportion deserving of punishment and reward in the life to come.
Now it is patent that despite this distinction between the two aspects of our knowledge of God by means of reason: the one entailing reward and punishment, the other not, both the Mu‘tazilah and the Ash‘arites were in agreement, as Ibn Rushd remarks, on the actual demonstrability of God’s existence. What they differed on was simply the moral or religious implications of such knowledge: the Ash‘arites holding that punishment and reward are conditional upon the “advent of the law,” the Mu‘tazilah making them independent of the explicit dictates of the law.
Prior to the rise of the Mu‘tazilah, who initiated the whole current of scholastic theology (kalam) in Islam, of course, the question of the demonstrability of God’s existence, like the remaining questions of rational theology, could hardly arise. The early jurists and theologians, such as Malik b. Anas (d. 795) and his followers were content with a theological knowledge rooted in Scripture. Like the Sufis, who believed that God could be apprehended directly, these Traditionalist sought the ground of their belief in God in a non-rational sphere: that of revelation or authority. Thus neither for Traditionalism nor for Sufism was a proof of the existence of God necessary at all, since the existence of God was given directly either in Scripture, according to the former, or in the mystical process of direct apprehension, according to the latter.
If the argument from causality (cosmological or aitiological argument), initiated by Aristotle and developed by his followers throughout the centuries, is rightly regarded as the classical argument for the existence of God in the West, the argument a novitate mundi (dalil al-huduth), of which the argument a contingenti mundi, (dalil al-jawaz) is a mere variant, can be safely asserted to represent the classical argument for the existence of God in Islam. The Aristotelian argument, which rested upon the concept of causality, was never viewed with favor in the Muslim world, not even by the great representatives of Arab Aristotelianism: Avicenna (d. 1037) and Averroes (d. 1198). The former laid special emphasis on the argument from contingency in a manner which definitely influenced the later Mutakallims; the later showed definite predilection for the teleological argument (dalil al-‘inayah) which had a basis in the Qur’an, and was of a more compelling nature than the other arguments, according to him.
The main reason why the cosmological argument was thus rejected out of hand by both the philosophers and the theologians was the fact that the concept of causality upon which it rested had been exposed to doubt since the beginning of Kalam. Al-Ghazali (d. 1111) continuing a long tradition of speculation on this theme, repudiates the validity of the causal principle in Question 17 of his famous Tahafut on the ground that the alleged necessity of this principle is a mere illusion; because it is unwarranted inference, based on observation from the correlation of events. Observation, however shows simply that the alleged effect happens alongside the cause rather through it (cum se non per se: ‘indahu la bihi) and accordingly, such a correlation is not logically necessary but is rather the outcome of a correlation is not logically necessary but is rather the outcome of mere psychological disposition or habit.
It is clear from the foregoing that Wensinck’s statement that the argument a novitate mundi is ‘analogous’ to the Aristotelian-Thomist proof ex parte motus et ex ratione cause efficientis is rather surprising, since the very validity of the causal principle is challenged by the Mutakallims. Moreover, the Aristotelian argument presupposes the cardinal metaphysical distinction between potentiality and actuality (which the Mutakallims also rejected, substituting for it the duality of substance and accidents); and is further independent, as Maimonides (d. 1204) and Aquinas (d. 1272) both recognized, from the thesis of the beginning of the world (round which the argument of the Mutakallims centers as we are going to see). Instead, Aristotle’s casual argument for the existence of the Unmoved Mover grew logically and naturally from the Aristotelian thesis of the eternity of motion in an eternal universe.
The Traditional argument of Kalam presupposes a preliminary thesis upon which the theological treatises place a considerable emphasis: the thesis of the newness or temporality of the universe (al-huduth). This circumstance explains the vehemence with which the opposite thesis of an eternal universe is combated by the advocates of Orthodoxy. Ibn Hazm, the Zahiri jurist and heresiographer, who died in 1064, employs this as the principle on the basis of which he distinguishes between the orthodox or heterodox sects. Muslims or non-Muslim. Al-Ghazali, as is well-known, devoted the first question of his Tahafut to a refutation of the thesis of eternity, which he consider the most pernicious thesis of the philosophers.
The general procedure of the Mutakallims in proving the temporality of the universe considered in showing that the world, which they defined as everything other than God was composed of atoms and accidents. Now the accidents (singular ‘arad) they argued, cannot endure for two instants of time, but are continually created by God who creates or annihilates them at will. Al-Bailani (d. 1013) who appears to follow the lead of Al-Ash‘ari in this respect, actually defines the accident as entities “the duration of which is impossible … and which cease to exist in the second instant of their coming to be.” Similarly, the atoms (sing. al-juz’) in which the accidents inhere are continually created by God and endure simply by reason of the accident of duration (baqa’) which God creates in them. But insofar as this accident of duration, like the other accidents, is itself perishable, the whole world of atoms and accidents is in a state of continuous generation and corruption.
Although the argument for the temporality of the universe form the temporality of its component parts is the favorite argument of the Ash‘arite doctors, it is by no means the only argument of Islamic scholasticism. Unfortunately we are in no position, owing to the scantiness of our sources, to reconstruct the reasoning of the Mu‘tazilite doctors on this question; nevertheless there is good reason to suppose that Al-Ash‘ari and his successors simply inherited the methods of argument, on this and allied subjects, which the Mu‘tazilah had initiated.
As an instance of the interest of the Mutakallims in the thesis of a temporal universe, we might examine here at some length the five arguments for the beginning of the world which Ibn Hazm, the great Zahiri theologian (d. 1064) advances in his Fisal; especially since Ibn Hazm appears to be the first Muslim theologian to have attempted a refutation of the eternity of the world, on the one hand, and a proof of its temporality, on the other, with any completeness. The biographer of Al-Ash‘ari, Ibn ‘Asakir (d. 571 A.H.), reports that Al-Ash‘ari wrote a treatise called Kitab al-Fusul, in refutation of the Materialists and the ‘philosophers,’ who professed the eternity of the universe, which as far as I am aware, is the earliest scholastic treatise dealing with the question of eternity in a systematic way, our sources record. Despite the statement of Al-Shahrastani that Al-Ash‘ari preferred the negative method of refutation (al-ibtal), as distinct from the method of positive proof, it is reasonable to assume that like Ibn Hazm, Al-Razi and others, he coupled the former with the latter species of argument.
Ibn Hazm’s first proof of the temporality of the universe rests on the premise that the accidents and substances (sing. Shakhs) composing the universe are finite and that time, which he conceives as consisting of transient moments, is finite also. In proving the finitude of these three terms: accident, substance and time, Ibn Hazm does not resort to the traditional method of the Mutakallims already mentioned, but maintains that the finitude of substance is evident from the finitude of its dimensions, that of accidents from the finitude of substances in which they inhere and the finitude of time from the transitoriness of the moments composing it. The second proof involves the Aristotelian dictum that everything in act is finite. The universe exists in act and is numerically determinate, therefore it is finite.
In the third argument he resorts to the process of reductio ad absurdum. The thesis of an infinite time, which the eternity of the universe, implies, involves the following absurdities: (a) Since infinity cannot be increased, all the time that will elapse would add nothing to the time elapsed hitherto. (b) The revolutions of a planet (e.g. Saturn) which revolves once every thirty years would be equal to the revolutions of the Upper Heaven, which amount to some 11,000 revolutions during the same period since one infinity is not greater than another. (c) The time elapsed since the beginning of time till the Hijrah (622 A.D.) and the time elapsed since the beginning till our day would be equal.
In the fourth and fifth arguments, he argues that, were the universe without beginning and without end, it would be impossible to determine it in number or in nature and consequently we could not speak of first, second, or third, in speaking of existing things. But this is contradicted by the fact that we can number things and refer to the first and last things. Hence the universe must have a beginning (awwal).
We cannot dwell longer on Ibn Hazm’s discussion of this cardinal theme and the manner in which he resolves the many objections to his arguments. But it is worth noting that most of the arguments of the later doctors such as Al-Ghazali and Al-Razi (d. 1209) are found here in an embryonic, though sometimes confused, state. This circumstance would appear to strengthen the view expressed by Maimonides (d. 1204), the great Jewish philosopher, that the Mutakallims were influenced in these arguments by John Philoponus (d. 568), author of De aeternitate mundi, a refutation of Proclus’s argument for the eternity of the universe – since it would imply that the Mutakallims from Ibn Hazm downwards were drawing on some common source.
With the temporality of the world as a premise, the Mutakallims proceeded to prove that the world being created (hadith) must necessarily have a Creator (muhdith), by recourse to the so-called “principle of determination.” In its barest form, this principle meant that since prior to the existence of the universe it was equally possible for it to be or not-to-be, a determinant (murajjih) whereby the possibility of a being could prevail over the possibility of not-being was required; and this ‘determinant’ – they argued – was God. Al-Baqilani (d. 1013),* who belonged to the second generation of Ash‘arite doctors and who is credited with refining the methods of Kalam, sums up this argument in succinct way. The world being temporal (hadith), he writes, it must of necessity have a Maker and Fashioner (muhdith wa musawwir), “just as writing must have a writer, a picture must have a painter and building a builder.” To this argument, however, he adds two others in which the ‘middle term’ differs but which reveal the same dialectical structure. In the first, he maintains that the priority of certain things over others presupposes an “Agent who made them prior” (muqaddiman qaddamahu) since priority does not belong by nature to a pair of equals; and this “determinant of priority” is God. In the second, he introduces a concept of contingency (jawaz) and argues that things in themselves are capable of receiving various ‘forms’ or qualities. The fact that existing things are endowed with certain determined ‘forms’ presupposes a ‘determinant’ who has determined that they should receive these ‘forms’ and no others; and this ‘determinant’ is God.
The element common to these three arguments, it will be noticed, is the “principle of determination” which they all invoke. Only the first argument, however, presupposes in addition the beginning of the world or its temporality. As to the third, it constitutes the basis of the argument a contingentia mundi (dalil al-jawaz) which was later developed by Al-Juwayni (d. 1086) as Averroes states in Al-Khasf, in a treatise which has not come to us, Al-Risalat al-Nizamiyyah. This proof, as Wensinck rightly observes, is affiliated to Ibn Sina (d. 1037) who seems to follow the lead of Al-Farabi (d. 950) in his respect, as Madkour has shown in his monograph on Al-Farabi.
In his major treatise, Al-Irshad, Al-Juwayni sets forth the more popular argument from temporality or huduth. “if the temporality of the world (hadath) is established and if it is established that (the world) has a beginning (muftatah al-wujud), since the temporal can equally exist or not exist … reason requires that (the world) must have a determinant (mukhassis) who determined its actual existence.”
Al-Baghdadi’s argument, as expounded in Usul ‘al-Din, differs little from that of either Al-Baqilani or Al-Juwayni. All rest, as we have seen, on the thesis that the world consists of atoms and accidents which have no subsistent being in themselves since they cannot endure for two moments of time. What, we might ask, is the extent of their debt to ‘Abul-Hasan al-Ash‘ari? The publication recently of Kitab al-Luma‘ enables us to give a provisional answer to this question, pending the discovery of fresh material.
The arguments of Al-Ash‘ari in this treatise has a distinct Quranic ring. It has nothing of the dialectical stringency of the later arguments and rests on the observation of the ‘phases’ of man’s growth from “a drop of water, to a leech to an embryo,” which the Qur’an has rendered classical. In so far as it is impossible for man himself to cause this change in his condition (tahawwul), the author argues, it is necessary that an “Agent should have transformed him from one phase to the other and disposed him according to his actual state;” for it impossible that this should happen without an agent of transformation, and by analogy the whole universe requires such an “agent of transformation.” This terse argument is of course in keeping with the nature of Al-Luma‘, as an introductory treatise, but confirms nevertheless the view that Ash‘arite Kalam was not fully developed by the beginning of the 10th century so that the authors of this period in general were content with purely rhetorical arguments based on the Qur’an or the Traditions. It is only with Al-Baqilani that a rigorous application of syllogistic methods of proof begins to make its appearance. But even here as we have seen no attempt at an elaborate analysis of the logical concepts involved is made.
The later history of Kalam reflects greater refinement in employing the technique of argument and a greater subtlety in handling logical concepts. Ibn Khaldun distinguishes between modern and the ancient stages in the development of Kalam and assigns the credit for introducing the ‘method of he moderns’ to Al-Ghazali. whether the credit for initiating this new ‘philosophical’ stage in the development of Kalam rightly belongs to Al-Ghazali or some earlier theologian, as Al-Juwayni or Al-Baqilani, is a controversial issue. It is certain, however, that this stage, as we have seen, is subsequent to Al-Ash‘ari’s time, and belongs to the latter half of the 10th century.
Al-Ghazali’s major contribution to the discussion of the problem at issue was twofold. In the first place, he brought out in a very forcible way the radical opposition between the teaching of Islam and the Aristotelian conception of a universe developing itself eternally and everlastingly; and in the second place, he gave added point to the arguments already advanced by the Mutakallims, by amplifying and perfecting them. Wensinck’s stress on the bipolarity in the thought of Al-Ghazali, the mystic, and Al-Ghazali, the theologian, is perfectly justified. Nevertheless it is only in Al-Ghazali as a Mutakallim and in his version of the argument a novitate mundi the we are interested here. The most succinct statement of this argument is found in Kitab al-Iqtisad fi’l-‘Itiqad, which he invokes, in the traditional manner of the Ash‘arites, the “principle of determination.” The syllogism runs as follows: Everything temporal (hadith) must have a cause. The world is temporal. Therefore the world must have a cause. By hadith, Al-Ghazali tells us, he means “what did not previously exist and then began to exist.” Prior to its existence, this ‘temporal world’ was ‘possible’ (mumkin) i.e. “could equally exist and not exist.” To tilt the balance in favor of existence a ‘determinant’ (murajjih) was necessary – since otherwise this ‘possible’ universe would have always remained in a state of not-being.
It would seem, considering the devastating attack which Al-Ghazali levels against the concept of causality in Question 17 of Al-Tahafut, flagrant contradiction. Al-Ghazali, however, explains in the same passage that by cause here he simply means a ‘determinant’ (i.e. murajjih) and consequently the apparent contradiction vanishes. Owing to its Aristotelian associations, this term was never in vogue among the Mutakallims. The earliest systematic refutation of the concept of causality as implicit in the doctrine of Tawallud (or production), of which I am aware, is found in Usul al-Din of Al-Baghdadi, who died in 1037, and which bears a striking resemblance to the more elaborate refutation of Al-Tahafut. Nevertheless, theologians of the later period are not entirely averse to the use of the term cause in this special sense of determinant. For instance, Fakhr al-Din al-Razi (d. 1209) one of the subtlest theologians of Islam, employs this term and its synonym ‘illah repeatedly in his exposition of the scholastic proofs for the existence of God.
We might examine here Al-Razi’s exposition of the traditional proofs for the existence of God as outlined in Kitab al-Arba‘in, especially since this is one of the fullest expositions which our classical sources record, and one which Wensinck does not seem to have consulted in his important monograph.
Al- Razi sums up the proofs of the existence of God under four arguments. The argument from the possibility (Imkan) of the universe to the existence of a necessary being (wajib al-wujud), Creator thereof.
The argument from the possibility of the qualities of the universe to the necessity of a Determinant of the form, characteristics, and locus of bodies composing it, who is not Himself a body.
The argument form the temporality of substances and bodies to the existence of a Maker thereof. (p. 86) and finally, The argument from the temporality of qualities of the universe to the existence of an intelligent Designer who disposes things according to His power and will. (p. 91)
It will appear from this brief analysis that these four arguments resolve themselves – as Al-Razi himself points out in the preface to his discussion (p. 67) – into two: the argument from temporality (huduth) and that from possibility (imkan). The root-concept in the former proof is the concept of time; viz. the fact that the world has had a beginning in time or in Al-Razi’s words, the fact that, before its existence, the world was in a state of not-being (al-‘adam). The root-concept in the latter proof is the concept of contingency (jawaz or imkan); viz. the fact that the world, considered singly as in argument (1), or as a whole as in argument (2), could have been otherwise. Al- Razi, like the rest of the Mutakallims, however, does not distinguish sharply between these two distinct proofs, as Ibn Sina justly remarks, and is on that account liable to some confusion. Al-Razi, for instance, defines the ‘temporal’ (al-muhdath) in his third argument as “that whose being in itself is contingent” which he further describes “as that whose essence is equally susceptible of not-being and of being,” which he adds significantly, “is the precise meaning of the possible.”
We might overlook this point an dwell on the similarities between these two distinct arguments. In the first place, whether we argue from contingency for from temporality, a necessary Being distinct from the series of sensible things (p. 70) must be posited as a Determinant of the being of the universe, on the one hand, and of the particular mode or being proper to it, on the other. This in fact is the point of distinction between the two concepts round which these two arguments center. For the argument a novitate mundi presupposes as we have seen, that prior to its existence the being and the not-being of the universe were equally possible, no account being take of the mode of being proper to this universe as in the argument a contingentia mundi.
In the second place, the positing of a Necessary Being outside the series of temporal beings flows logically from the impossibility of the regress ad infinitum. That is why Al-Razi, more conscious of the importance of this circumstance than the earlier theologians, devotes a lengthy discussion to the refutation of the two concepts of circularity (al-daur) and the regressus ad infinitum (al-tasalsul). Although he summarizes what appears to be the traditional argument against circularity, viz. that if two possible things were said to cause each other, each would precede the other an consequently itself, which is absurd, – Al-Razi proposes a different argument which he states thus: “The effect (ma‘lul) requires the cause. Now if each of two (possible) agents was the effect of the other, each of them would require the other and accordingly each would require what requires itself. Therefore, each would require itself, which is absurd.”(p.81).
In refuting the regressus ad infinitum, Al-Razi begins by laying down as a postulate that it is necessary that the cause should exist actually at the time of the existence of the effect, or else the latter would be capable of existing by itself – i.e., independently of the agency of the cause – which contradicts our original postulate.
1. If so, then the regression of the series of causes an effects to infinity would entail that the whole series existed simultaneously. Now the whole series is either necessary in itself or possible in itself. The former alternative is absurd because “a whole requires each of its parts and each of these parts is possible in itself and that which requires the possible in itself is a fortiori possible in itself too.” Consequently, the whole series is possible in itself and requires a necessary determinant (mu’aththir) distinct from itself, and this determinant is the Necessary Being.
2. If the whole series is contingent or possible-in-itself as we have seen, and if every possible-in-itself must have a determinant, this determinant is either (a) the whole series itself, (b) something pertaining to it, or (c) something outside it. (a) is absurd since it entails that the series determines itself. (b) is also absurd because it entails the member of the series, which was assumed to determine it, was also its own cause or of the cause of its cause. The former is absurd for the same reason as above (viz.: that a thing cannot be its own cause); the latter because it involves us in the impasse of circularity. Hence the determinant must be something outside the series as in (c). But what lies outside the series of possibles must be necessary-in-itself, which is the Necessary Being.
3. Let us imagine a portion of the series of effects extending from the last effect (L) to infinity and consisting of five segments. Let us next imagine another portion extending from the fifth segment to infinity.
Now if we compare the first portion (A) with the second (B) then they would either be equal –which implies that the whole is greater than its parts –or that one is greater than the other; so that the shorter portion (B) would be finite, since it is shorter by four units; and the longer would be finite also, since it exceeds the former by four units. Consequently, the ascending (tasa‘ud) series of causes and effects would have an extremity and a starting-point which is contrary to the statement that it is infinite. Not if that extremity is possible-in-itself, then it would require another determinant and thus would not be the extremity; if, on the other hand, it is necessary-in-itself, then we would have proved our case.
With Al-Razi, we might safely state, the ‘golden Period’ in the history of Kalam comes to an end. The merit of this subtle theologian is that he reintroduced into scholastic discussions certain.