Vol. XV No.

AL-ASH'AE~~S OLOGICAL DETERMINISM AND THE THE SENSES OF 'CAN'

In this paper I argue that ~ l - A s h ' a g(873- -935) was a Theological Determinist whose position on free will and human responsibility was marred by his failure to distinguish between two senses of 'can' (yastati'ul, or, as he and his fellow Mutakallims might say, =man's having power over his actions". This is not to say that had made the distinction in question then his position if a l - ~ s h ' a s would have been an adequate or acceptable view of free will and responsibility. My main point is that the logic of his position, as he explains it to us in al-Luma', strongly hints a t a distinction which alAsh'ari does not make. Consequently, his position appears not only somewhat inelegant, but also subject to a serious objection which might have been easily avoided. The interest which I have in looking a t the way al-Ash'azdeals with 'can' is not just to show one respect in which his system stands to be improved. For I think that the distinction which al-Ash'agfails to make has intrinsic interest in that it can be used to yield a new perspective on the nature and extent of the disagreement which existed between the Ash'arites and the Mu'tazilites on the question of.free will and human responsibility. In the second section of this paper I explain what I mean by 'Determinism' in general, and Theological Determinism' in particular. In the third section I first explain how the distinction between different senses of 'can' comes into play in the discussion of the relation between Determinism and free will. Then I discuss the way al-Ash'arydeals with the concepts of 'can' and 'cannot'. Here i t will become evident that al-Ash'ags position strongly hints a t a distinction between two senses of 'can', a distinction which he is unable to make on account of his view of the relation between human power and i t objects. Finally, I compare al-Ash'ags position to the one held by the

Hamdard Islarnicw

40

Vol XV No. 1

Mu'tazilite thinker a & - $ l Q 'Abd alJabbEr (d. 1025) and I suggest that the positions of these two representatives of the Ash'arite and Mu'tazilite thought on free will and responsibility were not so much opposed to each other as merely different. This result should, I think, invite us to re-evaluate the nature and extent of the disagreement between the Ash'arites and the Mu'tazilites over the free will question.

The term 'Determinism' has had much currency in modern and recent discussions of the philosophy of physical science as well as the free will problem. Different philosophers define the term in somewhat different ways but, on the whole, the different definitions seem to come down to this. A Determinist is someone who believes that for everything that happens, including human actions, there are (i) antecedent conditions, which may be thought of as causes in the broad sense of the word, and (ii) laws of nature which, in the final analysis, say that under such and such (antecedent) conditions, such and such a result and not another, is bound to occur. This understanding of Determinism is well-reflected in Alvin Goldman's definition of a determined event as one that is 'deducible from some aet of antecedent conditions and laws of nature,"' as well as G.E.M. Anscombe's defmition of deterministic laws as 'laws such that always, given initial conditions and the laws, a unique result is determined.I t is by no means easy to find convincing examples from the realm of human affairs which would illustrate the determinist position. Still, for the purpose of fixing ideas, we will let the following serve as an example of how things would look if Determinism were true. Let the antecedent conditions be such facts as the fact that Smith is very thirsty, that there is a glass of water which he can see in front of him, that there is nothing to prevent him from drinking it - e.g., his hands are free, and he can move them. As to the laws involved in this case, they are even more difficult to specify. But according to the determinist picture they will register the biological and psychological determinants of behaviour, and will, in the fmal analysis, say that under such antecedent conditions as we mentioned above, a man will drink the glass of water. Thus Smith's act of drinking the water is determined by his antecedent conditions and the laws ofmature. The relevance of Determinism thus understood to human freedom is not difficult to see. For if human actions are thus determined, then given one's antecedent conditions which, together with the laws of nature, logically imply what one will (or will not) do, i t follows that no one can do other than what one actually does.g But if this is so, i t is then argued, it follows that no one is free. For to be free means having a possibility of doing as well as a possibility of not doing, which is something that Determinism rules out by restricting one's

Hamdard Islamicus

41

Vol. XV No. 1

I
I

possibilities to just one. In this way it is inferred that Determinism and free will are incompatible: if we are free, then Determinisim is false, and if Determinism is true, then there is no such thing as freedom. A l - ~ s h ' a s course, had no idea of man's actions being a t some of moment determined by his (and the world's) state a t the previous moment, together with the laws of nature. A notion like this would undoubtedly have seemed profoundly wrong and blasphemous to him. But still, he was a Determinist in the sense which interests those who are concerned with free will. In order to clarify the sense in which a l - ~ s h ' a zwas a Determinist, we need to distinguish between two kinds of Determinism: Scientific and Theological.' Scientific Determinism, which we briefly explained above, is a naturalistic, science-inspired kind of Determinism. Where ' is a human action, Scientific Determinism says that the P (natural) antecedent conditions (including, of course, the state of the agent) in conjunction with the laws of nature, necessitate (make necessary) either the occurrence of 'P, its non-occurrence. or Now, Theological Determinism ought to say something like this - that human actions are somehow necessitated. Otherwise we should really have no right to call it Determinism and expect to be understood. But how, according to Theological Determinism, are human actions to be made necessary? What would make them so? There is, I think, one explication of Theological Determinism which can answer the question. (It is a little artificial h c a u s e it is tailored to correspond to Scientific Determinism, but other than this, I think i t accords well with the intuitive understanding of the idea that human actions and other events are determined by God. Where ' is, again, a human action, replace Scientific P Determinism's idea of antecedent conditions with the idea of God's will, or decree. Furthermore, replace the laws of nature by one supreme law (of Divine Nature, if you wish) which says: For any event (E), if God wills, or decrees that Y E ) be the case, then %' comes to be. Theolgoical Determinsim may thus be briefly formulated a s the doctrine that everything which happens, including human actions, has been willed, or decreed, to happen by God, Who is such that whatever He wills, or decrees, comes to be. This formulation easily yields the conclusion that Zayd's unbelief (like Smith's drinking in the example mentioned above in connection with Scientific Determinism) i s something which had to happen, given God's will that i t should happen, and the necessary effectiveness of God's acts of will. The above explication of Theological Determinism is rather schematic, in that i t can be elaborated in more than one way. For it does not say whether God realizes His will directly, by bringing the event into being without mediation, or if He realizes His will in an indirect manner, by creating conditions which in their turn cause the

Hamdnrd Islamicus

42

Vol XV No. 1

event in question to happen. Both ways of elaborating Theological Determinism are equally deterministic leading, a s they do, to the result that whatever happens, including human actions, could not have failed to happen. The only difference is that the second way of elaborating Theological Determinism may be thought to place some efficacy in the things which God uses a s instruments for realizing His will. Now we can proceed to ascertain how al-Ash'asstands in relation to Theological Determinism. That al-Ash'a6 was a Determinist is clearly revealed in his discussion, in al-Luma', of the case of the unbeliever whom God charges with the duty to believe. An imaginary opponent asks al-Ash'as if he (the unbeliever, that is) can believe. Al-~sh'a; answers by saying that if the unbeliever could believe then he would b e l i e ~ eThe clear implication, therefore, is that, the .~ unbeliever cannot believe. And since believing is other than what the unbeliever does, we may safely infer that, according to al-Ash'ari the unbeliever cannot do other than what he actually does, namely, notbelieve. But why cannot the unbeliever believe? Al-Ash'ari has to offer an explanation of this, if his position is to be more than a bare assertion of Determinism. But, unfortunately, al-Ash'aas (apparent) explanation of the fact of unbelief seems to be rather vacuous. For, in the present context, al-Ash'as goes on to say that the unbeliever cannot believe because, or given that, "he leaves lbeliefl undone and is occupied with its contrary (li-tarkihi wa inshighiilihi bi-diddihi)."6 As R.M. Frank expresses it in hii analysis of this part ofal-Luma', 'the actuality of unbelief in the unbeliever excludes the pos$bility of the simultaneous actuality of belief in him." ' Well, of course it does. And it should, for otherwise the elementary logical Principle of Non-Contradiction would be violated. Is this, therefore, what al-~sh'ari"s Determinism amounts to: Zayd cannot believe (that is, his unbelief is a necessary fact) because he is busy not-believing? Surely, this is a trivial thesis which no one can deny, but which says nothing and explains nothing. Of course, one cannot sit, given that one is standing, and one cannot stop, given that one is running. But this does not explain why one is standing, or by one is running. In fact, if we understand by 'sitting' not-standing, and by 'stopping' not-running, then all we do here is to repeat ourselves in a completely uninformative manner. Let us therefore ask again why, according to al-Ash'ar: an unbeliever cannot believe. A better answer, I think, can be based on al-Ash'ari's Creed, which is to be found in his MaqZlZt al-IslZmiyyin. In the Creed, al-Ash'dmakes it clear that (i) Everything that happens has been willed by God. With Ah1 al-IthbZt ("the People of Truthn), alAsh'a6afirms that "There is nothing on earth, good or bad, except what God wills." What this means, of course, is that if something exists, then God willed it. If God wills something, then i t comes to be. This, and more, al-Ash'an' expresses by saying (%th the generality

Hamdnrd Islamicus

43

Vol. XV No. 1

of the Muslims"), that W h a t God wills is, and what He does not will, is not.m In these terms we can, I think, best explain why, according to al-Ash'ac the unbeliever cannot believe. Even if we say initially that the unbeliever cannot believe because he is busy not-believing, the fact remains that he is occupied in this way because God willed, or decreed, that he be so occupied. In short, what underlies al-Ash'ass claim that the unbeliever cannot believe is none other than his Theological Determinism. To complete our sketch of a l - ~ s h ' a a s Determinism we only need to ask whether, according to al-Ash'ar';, God realizes his will directly, without mediation, or (at least sometimes) by creating things, or conditions which in their turn, cause the events which are willed by God. I think that, on the whole, R.M. Frank is right in his belief that al-Ash'agintended to allow for some kind of human c a u ~ a l i t y ? ~ What is being referred to here, of course, is al-Ash'ags notion of generated power, the idea that God creates in man a power to act, which, in tum, 'causes" those acts for which man is responsible. This created power to act we may perhaps add other conditions or accompaniments of actior, which al-Ash'az seems to recognize if only implicitly, e.g., the bodily member (the limb, al-jasha) required for the action" and, in the case of % o l ~ n t a r i l y ? ~ Such considerations should, I think, make it reasonable to believe that an elaboration of al-Aah'ari's Determinism can have room for conditions, or causes, by which the fulfillment of God's will is mediated. But still, this does not mean that we should attribute to al-Aah'afi any willingness to accept secondary causality, as Frank seems to be inclined to doJ3 In fact, there seems to be a vast difference between the way humans cause their actions according to al-Ash'ac and the way they would cause them, if secondary causality, as it is normally understood, were operative. The difference may be briefly explained in terms of an analogy which will also serve to make clearer the workings of al-Ash'ags brand of Determinism. Consider an ordinary metal wire which does not normally, or naturally deflect the compass needle. But sometimes it will do just that - as when someone happens to pass an electric current in the wire, which becomes then magnetized. The effect which the ordinary metal wire has on the compass needle must not be regarded as an instance of secondary causality, because i t is not something which the piece of wire does normally, or naturally (by its nature). Rather, is purely accidental, depending a s it does on the accidental circumstance of someone passing an electric current through it. The situation is patently different in those cases which may fairly be taken to illustrate the operation of secondary causality: e.g., the fire which naturally bums a piece of wood, or food which naturally appeases hunger. But al-Ash'arTdoes hot believe that man by nature is able to act

Hamdard Islamicus

44

Vol. XV No. 1

(qiidirun bi-tab'ihi)." In this way man is like the ordinary piece of metal wire which is not naturally able to deflect the compass needle. Sometimes man will act, but this happens only when God creates in him the power to act. He is then momentarily able to act, just like the piece of wire which is momentarily magnetized. Humans might indeed cause their acts according to al-Ash'ari, but their causality is not secondary causality. Rather, it is the causality of tool, or an instrument which acts, not in accordance with a nature that it has, but in accordance with the purposes of its user. I11 Determinism, as explained earlier, seems a t first sight to conflict with free will. For (it is usually thought) one is not really free with respect to doing something unless one can do as well as refrain from doing it. Determinism, on the other hand, implies that what happens has to happen. Consequently, one cannot fail to do what one does. Still, with regard to the scientific variety of Determinism, there is a well-established school of thought that goes back to Hobbes according to which Determinism is compatible with freed~rn.'~ the In more recent expressions of this compatibilist view, attention is usually focussed on the analysis of 'can' as it occurs in freedom-expressing statements such as "He can (i.e. he is free to) do this n and 'He did that, but he wuld have done otherwise (i.e. he was free to do otherwise)". The aim is to show that what 'can' means here is not inconistent with the act's being determined or necessitated. The classical modern analysis of 'can' along these lines is to be found in G.E. Moore. This is what he has to say about 'could', which, of course, applies to 'can' with slight modifications. e What is the sense of the word 'could' in which i t is so certain that we o h n could have done what we did not do? What, for instance, is the sense in which I could have walked a mile in twenty minutes this morning though I did not? m e r e is one suggestion which is very obvious, namely, that what I mean is simply, after all, that I could if I had chosen ... or "I should if I had chosen"." Surely, nothing here is incompatible with th belief that one's antecedent conditions (together with the laws of nature) may be such as to determine or necessitate, one to choose and do 'A', or to choose and do otherwise. Despite the fact that one's action is determined, it may still be true that were one to choose 'A', then he would do 'A', and were one to choose not to do 'A', then one would not. He therefore acts freely, even if his action is determined. The fact that someone's action (or desire or choice) is determined (or necessitated) does not rob him of his freedom, according to the above analysis of'can'. What would rob him of his freedom is, rather,

Vol. XV No. 1 anything that makes it no longer the case that Yf he chooses, then he does and if he chooses not to, then he does not". And this might be such things as physical constraint or illness. The effect of the latter may be illustrated by means of one of KalEm's favorite examples. Consider the action of the person who, shaking with fever, moves his hand. His 'action", according to the above analysis, is not free: for i t is not the case that if he chooses to move his hand, then he moves it, and if Ke chooses not to mvoe his hand, then he does not move it. Clearly, he often wishes not to (and chooses not to) move his hand, but he nevertheless does. The compatibilist position explained above promises to accommodate both Determinism and freedom by distinguishing between two senses of 'can'. This is what makes this position of interest to us, as we try to understand the logic which underlies al-Ash'ass position. For al-Ash'aKis a compatibilist of sorts, or a t least he sets out to be one. He insists on Determinism in no equivocal terms, but he nevertheless wants to hold onto a meaningful notion of human responsibility, which (to our minds, a t least) presupposes freedom of the will. However, before we go on to ascertain how al-Ash'ag seeks to accommodate both Determinism and responsibility, we need to distinguish and label the two different senses of 'can' upon which the modem wmpatibilist position rests. Firstly, there is what we may call %he causal-deterministic sense of 'can'." This is the sense which we negate when we say, for example, of a glass of water that is exposed to extreme heat, that i t cannot stay cold. We mean that, given the water's antecedent conditions (which include exposure to heat), and the laws of nature, the water has to become hot - it cannot stay cold. If Determinism is true with regard to human actions also, then we cannot, in the present sense of 'can', do except what we actually do. Secondly, there is what we may call the 'can' of freedom. This is the 'can' which the compatibilist claims is in question when we affirm that someone is free to do something, meaning that he can do it, as well as refrain from doing it. This 'can', according to the Moore suggests. compatibilist, is analyzable along the lines which G.E. In this sense of 'can', we still can do something (or its opposite) even if we cannot, in the causal-deterministic sense of 'cad, do except what we in fact do. Only impotence occasioned by such things as illness or physical constraintcan make it true that we cannot, in the present sense of 'can', do something or other. Having distinguished the two senses of 'can' in terms of which the modem compatibilist seeks to reconcile freedom and Determinism, proposes to find room for let us now go on to see how a l - ~ s h ' a z Detrminism as well as human responsibility. In the part of al-Luma' which deals with istita'at, (capacity), al-Ash'a; is a t pains t o

Hamdard Islamicus

Vol XV No. 1

distinguish between two ways of understanding 'cannot' as i t occurs in statements as 'Zayd cannot believe." The crucial passage reads thus:
Q.

A Q.
A Q.

A.

Has not God charged the unbeliever with the #dutyof believing? Yes Then the unbeliever can believe. If he could believe, he would believe. Then God enjoins on him an obligation which he cannot fulfil. This is a statement which involves two matters. If' you mean by your words that he cannot believe because of his impotence ('ajz) to do so - no. But if you mean that he cannot believe because he omits to do so and is occupied with the contrary of belief - yes.I8

Thus when confronted with the opponent's challenge, which, in effect, says "According to you the unbeliever cannot believe", Al-Ash'aX responds by offering two different interpretations of the challenge. Interpreted one way, the opponent's challenge calls for the response 'No", that is, 'No, it is not true that the believer cannot believe". Interpreted another way, i t calls for the response Yes", that is, "Yes, gui it is true that the unbeliever cannot believe." Clearly, al-Aswasintends to distin sh here between two ways of taking 'cannot'. One way to take i t would be to think that Zayd is incapable of believing CcTju). Let us call this 'cannot' the cannot of impotence and abbreviate i t to 'cannot-i'. Thus in his first response to says, in effect, "It is not true that the opponent's challenge a l - ~ s h ' a g the unbeliever cannot-i believe." The other way of taking 'cannot' is to say that, given the unbeliever's condition of leaving belief undone (tark) and occupation with the opposite of believing, he cannot believe. Let us call this 'cannot' the cannot of tark, and abbreviate i t to 'cannot-t'. Thus in his second response to the opponent's challenge, alAsh'ari, in effect, says: It is true that the unbeliever cannot-t believe." In sum, then, al-Ash'ag asserts that the unbeliever cannot-t believe, while he denies that the unbeliever cannot-i believe. But what is 'cannot-t' and what is 'cannot-i? I think i t will be fair to understand them in terms of the two senses of 'can' which we encountered earlier in connection with the modem compatibilist position. That is to say, I propose to take 'cannot-i' as the negation of th 'can' of freedom, and 'cannot-t' as the negation of the causaldeterministic 'can1. That al-Ash'a6's 'cannot-i' may be understood as the negation of th 'can' of freedom which we explained earlier is made reasonable by the examples which al-Ash'an'mentions of cases where God would not impose duties on man. Not long after distinguishing between two ways in which the unbeliever cannot believe, al-Ash'as goes on to

Hamdard Islamicus

47

Vol. XV No. 1

say, or rather imply, that God would not impose a duty on man when the bodily member (Limb) required for that duty is missing (ghiyab aljaF%a).lg Similarly, the duty to pay alms is suspended when one has no money or wealth." What stands out when we consider these two examples and others in their spirit (e.g., a l - ~ s h ' a g w i l lagree that God does not impose duty to fight on one who is ill or who is in chains) is the following fact. In each case it is not true that if one chooses, or wills to perform the action in question, then one performs it, and if one chooses not to, or wills not to, then one does not. And this is precisely what the negation of the 'can' of freedom amounts to. Like al-Ash'ass 'cannot-?, the negation of the 'can' of freedom is incompatible with the assignment of responsibility. With regards to al-Ash'ari7s 'cannot-t', the situation is a little more complicated. For what al-Ash'agsays here is simply that someone cannot believe because he is not-believing. But as we suggested earlier, i t is not satisfactory to think that al-Ash'ds Determinism reduces to a trivial logical argument based on the premise that one cannot believe and not-believe a t the same time. but when we consider that the person's actual unbelief (which cannot co-exist with belie0 is something which was willed by God, Who is such that whatever He wills has to come to be, the matter begins to look more interesting. For then alAsh'ass position on the believer who cannott believe appears rather similar to the position of the Compatibilist who says that Smith cannot but drink the water which is placed in front of him, given his extreme thirst, and the laws of nature. In both cases the person cannot do except what he does (not believe, drink the water), not because of incapacity or constraint (as in the case a person who is subjected to brain-washing, or who is being force-fed), but because the antecedent conditions, together with certain "laws", make it impossible for something different to happen. If we are right to understand al-Ash'ass 'cannot-i' and 'cannott' in the above indicated manner, then we have some reason for thinking that al-Ash'an'must have seen in the distinction between two senses of 'cannot' a plausible method for reconciling the rival claims of Determinism and responsibility. On the one hand, Determinism is safeguarded by insisting that one cannot do except what one actually does. But, on the other hand, the fact that we cannot act differently does not abolish responsibility. This happens only when one 'cannot' (la yastaFu) in a different sense of 'cannot' - the sense of impotence. And since i t is not true that the unbeliever cannot, in this latter sense, believe (for he is not 'Giz),it follows that he is responsible for his unbelief. Thus there is room for both Determinism g d responsibility. Al-Ash'ari's compromise may sound very plausible, or, a t least, as plausible as Compatibilism in general is. But al-Ash'ari's compromise has problems of its own. The main difficulty is that he

Hamdard Islamicus

48

Vol XV No. 1

is logically required to distinguish between two senses of 'can', a requirement which al-Ash'ari does not, and, in fact, cannot meet. Let us take up the first point first. That al-Ash'ass position requires him to distinguish between two senses of 'can' follows from the fact that he wishes to distinguish between two senses of %annot'. He wishes to say that there are two ways one can take a statement like "Zayd cannot believe." But surely, if the statement "Zayd cannot believe" can signify in two different ways, then so must the statement "Zayd can believe." Furthermore, it is easy to specify what these two significations are. Firstly, we may take the statement "Zayd can believen as a denial of (Theological) Determinism. If Zayd can, in this sense, believe then i t is no longer the case that Zayd's antecedent conditions, together with the "lawsn, make i t impossible for him to do otherwise. (It is immaterial how one might come to take a position of denying Theological Determinism. One might think that God's willing is not necessarily effective, or, more plausibly, that i t does not extend to all human actions). Call this sense of 'can' 'can-t'. To say that Zayd can-t believe is, then, incompatible with (Theological) Determinism. The second way of taking the statement uZayd can believen is to take it not as a denial of Determinism, but as a denial of impotence. To say that Zayd can, in this sense, believe, is to say that he is sane, that he heard the call to believe, or simply that there is nothing which constrains him to not-believe. Call this sense of 'can' 'can-i'. To say that Zayd can-i believe is then incompatible with impotence. But it is not incompatible with Determinism. One can be a Determinist and still believe that one can-i do other than what one does. But al-~sh'a; does not permit himself to say that there is a sense in which the unbeliever can believe. His view is that the power (istiti'at) which the unbeliever has, in power over unbelief only - i t does not extend to both the act and its opposite. This is how his view unfolds. Immediately after distinguishing between 'cannot-i' and 'cannott', al-Ash'an'puts in the mouth of his opponent a question about God's imposing on the believer a duty which he, the unbeliever, is incapable 'jz ~ of doing ( i i ,impotent) because he leaves it undone ( t ~ r k ) .Of' course, from al-Ash'ass standpoint, the question represents a confusion between 'cannoti' and 'cannot-t', and so he hastens to set matters straight by explaining the difference between impotent and tark. Impotence, unlike tark, extends to both the act and its opposite. Thus we must not describe one who cannot-i stand as one who can sit. But it is not like this with the unbeliever who cannot-t believe. For he may be legitimately described a s one who can not-believe, or even disbelieve. Hence it must be incorrect to speak of someone who leaves something undone (tark) as someone who is "impotent". Having explained the difference between tark and impotence in I his way al-Ash'ac by a natural progression of thought, moves on to answer the question of whether qudraf (power), in analogy with '42,

Hamdnrd Islamisus

49

Vol.

XV No.

might also extend to both the act and its opposite. He clearly states that qudrat, unlike 'qir, does not extend to pairs of opposite acts. Two arguments are offered why this is so. One of them, the first, rests on unduly assimilating 'awn (assistance, helping, aiding) to qudrat, and may he easily disposed of. The second argument however, rests on alAsh'ass concept of the nature of the relation between power and its objects. I t is ultimately this latter which stands in the way of alAsh'arTs being able to distinguish between two senses of 'can'. The first argument may be re-stated thus. If we were to say that power over something was also power over its opposite, in analogy with 'ajz, then we must say that 'awn over something was also 'awn over its opposite. But this is not so. Therefore power over something is not power over its o p p ~ s i t e . ~ This argument is not convincing. It rests on the assumption that qudrat and 'awn are similar, and that they stand together in opposition to 'ajz. But this by no means the case. 'Awn over something means (the process of) lending assistance, or help-giving. As such, it stands in opposition not to 'ajz,as al-~sh'asseems think, but to obstruction to or obstacle-laying. Qudrat is opposed to 'ajz, and 'awn to obstruction. I t may indeed he that 'awn is not over something and its opposite. But then the same thing applies to obstruction: obstruction is not obstruction in something and its opposite. There is nothing here to show why qudrat, like impotence, does not extend to pairs of opposite acts. So much for al-Ash'ass first argument. His second argument is more effective, given his conception of the necessitating nature of power. Here the substance of the argument seems to he that if the unbeliever can, or has power to, believe then he will believe. For (originated) power (qudrat rnuwatha), a s al-Ash'as understands it, is such that, when it exists, so must that which it is power for.= Hence if he who actually does not believe could believe, then he would. And then he will be believing and not-believing all a t once.24 Thus al-Ash'acwill not say that the unbeliever can believe. If he had a different conception of human power, or if he had found a way to say that the unbeliever can believe, then he would have been well on his way to making a distinction between two different senses of 'can'. For given the distinction between 'cannot-t' and 'cannot-?, the statement T h e unbeliever can believe" can only mean that he is capable of believing, that is, he is sound of mind, he has heard the call to believe, and there is nothing to constrain him to not-believe. Or, more briefly, he can-i believe. The statement does not, and cannot, mean that i t is possible for the unbeliever to believe despite God's will in this regard. In other words, i t does not, and cannot, mean that the unbeliever can-t believe. As matters stand, al-Ash'aas conception of power prevents him from saying that the unbeliever can, in any way believe. This not only burdens a l - ~ s h ' a z with the charge that he is a Jabrite in disguise,

Hamdnrd Islamicus

50

Vol. XV No. 1

but i t also robs his view of a certain degree of elegance that i t might have otherwise had. For having already distinguished between two senses of 'cannot', nothing would have been more logical than to distinguish between two corresponding senses of 'can'. But that was not to be. Al-Ash'aas emphasis on Divine Omnipotence and Determinism, and the secondary status he assigns to Divine justice and human responsibility, seem to have prevented him from taking this further step.

Whether the compatibilist position explained earlier in 111 is ultimately found to be a satisfactory view of free will or not, one thing, at any rate, is certain: i t is by no means obvious that Determinism is incompatible with free will. The same thing should apply to Theological Determinism also, for, in essence, this latter is not profoundly different from Scientific Determinism. Be that as it may, al-Ash'az is not able to say that man can do other than what he does. And as the ability to do otherwise seems, at least on reflection, to be an essential part in any serious advocacy of free will and human responsibility, it is not surprising to find that alAsh'arTs Mu'tazilite opponents make much of this point in their criticism of the Ash'ass kasb (acquisition) theory. One such Mu'tazilite thinker is al-QSdT'Abd alJabb5r. It will be instructive for us to examine his view, for this will show not only that 'Abd alJabb5r's treatment of the word 'can' was in some ways even less satisfactory than the one offered by al-Ash'az, but also that the Mu'tazilite and Ash'arite positions on free will may not have been so much opposed to each other a s merely different, emphasizing different senses of 'can'. Let us take up 'Abd al-JabbZr's criticism of kasb first. In al~ u h he compares acts which occur by way of kasb to acts which one s is constrained to do (muljZ), going a s far as to illustrate this by means of an example of a man who falls down to the ground when he is pushed from high above. The similarity, to 'Abd al-Jabbar's mind, is that an act which occurs by way of kasb occurs necessaiily, so that one cannot avoid it, or do otherwise, any more than a person who is falling down can do other than fall towards the ground. In both cases there is no room for praise or blame (human responsibility) or choice. In 'Abd a l J a b b 3 s words, this is how matters stand with kasb. According to you, an act which occurs by way of kasb cannot be avoided. For it is not possible that God should bring into being this act and the power [to perform this act] without this act becoming necessarily an acquisition. He [the agent1 is in the position of one who is constrained (al-ma+miil al-mulja'): blame and praise cease to apply, and he is like one who is pushed from

Hamdard Islamicus

51

Vol. XV No. 1

a high place. There would be choice if it were the case that, given the power [to do], one could acquire or not." It is obvious what 'Abd alJabbZr9sattention is primarily focussed on here. To be deserving of praise and blame, and for there to be genuine choice in what one does, all this requires, according to 'Abd al-JabbZr, that one be able to do or refrain from doing. But since an acquired act is necessitated (God causes it by giving us the power to perform it), it follows that the act is one which cannot but be performed. Earlier we saw how al-Ash'az tries to deal with this matter. Essentially, he says that, in a sense, one cannot do except what.one does, but, in another sense, this is not so. One expects al-Ash'az to go on to acknowledge that, in this other sense, one can do other that what one does. But al-Ash'ac apparently, cannot bring himself to do this. So he stops short. Burdened with the claim that one cannot do other than what one does, he makes an easy target for those who think that such a position cannot accommodate genuine human responsibility. Abd alJabbZr's position, on the other hand, like that of the generality of the Mu'tazilah, is that man's power over his actions extends to both the action and its opposite. And because the Mu'tazilah in this way oppose the Jabrites and the kasb theorists, it is often thought that they must be regarded as the representatives of the Libertarian position in Islam. J.R.T.M. Peters (for example), commenting on 'Abd alJabbZr's view of human ability, says: By describing human ability in this way [ability hasthe possibility of two opposite acts1 'Abd alJabbZr lays the basis for his doctrine that man really has free choice in his acting and that he is really responsible for his acts.z6 But before we pronounce 'Abd al-JabbZr an advocate of free will it is only fair to determine how he conceived of this human ability which is supposed to extend to pairs of opposite acts. For our discussion so far has made it clear that it is possible to be a Determinist and to claim that one who performed a certain act could have performed its opposite. Thus if it turns out that 'Abd alJabb5r was a Determinist, then the gap which separates his position from that of a l - ~ s h ' a z w i l l be narrowed. For al-Ash'a; was a Determinist who did not really stand to lose much by extending ability to opposite acts. And the net result of all of this will be to weaken the ground on which 'Abd alJabbZr and the Mu'tazilah are distinguished by the advocacy of free will. Al-Ash'ar: and his followers can easily join the club bysanying to its logical conclusion a certain distinction which al-Ash'ari himself makes. Let us therefore begin by considering how 'Abd alJabbZr conceived of this human ability which is supposed to extend to pairs

Hamdard Islamicus

52

Vol. XV No. 1

of opposite acts. It seems to me that he thought of human ability as a disposition. In other words, to say of someone that h e is able (qiidir) is like saying of a piece of metal that it is magnetic. To say of a piece of metal that i t is magnetic is not to say that it is currently engaged in attracting iron filings. It may be doing this, or i t may not: what matters is that i t has a disposition to do so. I t is constituted in such a way that it can do so. And it is somewhat like this with one who is able (qiiddir): it may be that h e is engaged in performing a certain act, or it may not. In either case his ability stands indifferently to both alternatives. As 'Abd al-JabbEr expresses it, "One who is able (qiidir) may be in such conditions that he does not choose to act, or he may be in such conditions that he inevitably chooses to act".* In both cases h e is able - whether or not he acts. His ability, therefore, is a disposition which may be realized under certain conditions (e.g., when there is a strong motive to act) or which remain unrealized for the time being. 'Abd al-Jabhiir's dispositional understanding of ability is also evident in connection with his belief that God is 'able"." God is able not only in the general sense of the word, but also in specific ways e.g., He is able to perform unjust acts. With regard to God's general ability, 'Abd al-JabbZr does not hesitate to describe God as being "strong" (qaww3, for the meaning of qawwTand qZdu, he says, is the ~ a m e . ~ ~ A n d without saying that 'strong'is a term that describes it goes a disposition. For to say of someone that he is strong is not to say that he is doing this or that a t the moment, but only that he can, or that h e would, given the appropriate conditions. With regard to God's specific ability to perform unjust acts, 'Abd a l J a b b z r argues that unjust acts are of the same species as just acts, and hence God Who can perform just acts can perform unjust ones." But, of course to say that God ean perform unjust acts is not to say that He is currently performing them, or even that He ever will. I t only means that He can, or would, if He were to will to. Again, 'Abd aldabbzr's ability seems to be a disposition. Again distinguished between two senses of 'rannot' in connection with Scientific Determinism, a s well a s al-~sh'arps .Theological Determinism, we naturally want to know what manner of 'can' 'Abd al-JabbZr speaks of, when h e says about both man and God that they can perform a certain act or its opposite. Is it the causal-deterministic 'can' whose negation is required by the belief in Determinism? Or is it the freedom 'can' whose affirmation is required by the belief in free will and responsibility? Evidently, it is the freedom 'can', the 'can' which is incompatible with incapacity. (Incapacity may in fact be defined as the condition of one in whom the ability disposition does not exist). But what about the causal-deterministic 'can'? Does it receive any recognition from 'Abd al-JabbZr? The remarkable fact is that 'Abd al-JabbZr's position appears to require a causal-deterministic sense of 'can', which he is

Hamdnrd Islamicus

Vol. XV No. 1

as unwilling to admit a s al-Ash's; is unwilling to admit a freedomrelated sense of 'can'. Let us begin with the first point, that is, the claim that 'Abd alJabbZr's position requires a sense of'can' in which one cannot do that which he is not doing, There are many places where 'Abd al-fabbzr presents himself as a Determinist of some kind. Thus in Volume VIII of al-Mughnl in a chapter that deals with the connection between ability, actions, and motives (dnwZ'3, he says: Similarly, the dn'l' (motive), if it is unopposed (infarida) necessitates (yaqtadi) the action inevitably ( l i maGlah). For this reason God strengthens the motives of the mukallaf (he who is subject to religious duties and commands) by promise of reward and threat of punishment. This would not make sense if it were not the case that these [the promise and threat] necessitated the doing on the part of the mukallaf when h e is free from other motives. Yet some mukallafs do not choose to worship because there are other motives which oppose [the motive to worship], such as immediate desires, and the like.31 Here 'Abd alJabb5r seems to be a Psychological Determinist who believes that our actions (or choices) are necessitated by our motives and desires. Should one object by saying that there are times when one has a very strong motive, or desire, to do something, and yet does not do it, 'Abd alJabbZr would answer, in the manner of the modem Determinist, by making reference to opposing motives and desires. It is true that there are many places where 'Abd alJabb2r says that a motive is not necessitating (miij5) like the 'illah (cause).3zBut it seems that what this means is that the necessitation occasioned by the 'illah is one which the 'illah does per se, and which cannot be defeated. Motives, on the other hand, can be opposed and defeated by the other motives. But none of this makes it false to say that motives, when unopposed, do necessitate. Another place where 'Abd al-fabbzr exhibits a determinist line of thought is a chapter that deals with the question of whether something can be an object of power (maqdlr) to two possessors of power (qiidirayn). 'Abd alJabbEr wants to argue that we have no reason to connect Zayd's action with a being other than Zayd (e.g., God). This he does by drawing an analogy between Zayd's action which must happen when ,Zayd is willing, intending, and unrestrained, and someone's becoming %nowledgeable" ('alTrn), which also has to happen when the 'illah of knowledge comes to be present in him. (Just as we do not need to assume additional causes in the latter case, so we do not need to in the former case). This is how 'Abd alJabbZr presents the matter:

Hamdard Islamicus

54

Vol. XV No. 1

If this [that it is not right to account for someone's W i g knowing by invoking additional causes, when he is knowing by virtue of the 'illah of knowledge which i s present in him1 is proved, and it is furthermore true that Zayd's action has to happen fyajibu wuqiikhu) when Zayd is intending, willing, and unrestrained, so that his intention in this case ... is like (bimanzilati) the causes which necessitate attributes, then it follows that, just as we do not relate the effect of the 'illah to anything but the 'illah, we should not relate the occurrence of the act to any being but this intending being IZaydl."
Thus it would appear that 'Abd alJabbSr sometimes speaks as a Determinist." As such, h e must recognize that there is a sense in which one (sometimes) cannot do except what one does. This will be, of course, the causal-deterministic sense of 'can, which we negate when we say that, given the determining conditions, and the laws, one cannot but perform (or omit) a certain act. But i s 'Abd al-JabbZr willing to recognize such a sense of 'can'? It does not seem so. That he is averse to recognizing a causaldeterministic sense of 'can' is evident from the way he deals with an objection which h e puts in the mouth of a n imaginary opponent. The objection is made in Part 2 of Volume VI of al-Mughnx in a chapter that deals with the point that he who has knowledge of the nature of that which is evil, together with knowledge that he does not need to perform it (ghiiiahu hnhu), does not in any way choose to perform it. The opponent seems td be saying that the assertion that: He who has knowledge of the nature of that which is evil, together with knowledge that he does not need to perform it, does not choose to perform it, nor is it possible for him Lto choose to perform it1 contradicts the statement that He has power over it fqzdirun 'alayht) (i.e., he can perform it)." Consider how 'Abd alJabbXr tackles the objection, completely by-passing the causal-deterministic sense of 'can' which we are a h r : One who is able may be in such conditions that h e does not choose to act, or he may be in such conditions that he inevitably chooses to act. But just as the fact that h e who is constrained fmulja) to act necessarily acts, does not contradict his being able, so it is with the necessity of Chis1 not doing that which is evil, when h e has knowledge that it is evil together with knowledge that he does not need to perform it."

Hamdnrd Islamicus

55

Vol XV No. 1

As an example of someone who is mu&?to do something, consider the case of someone who is in danger of starving to death. He chooses to eat a dead animal, and he does. Now, surely, there is a sense of 'can' in which it would be right to say that the starving man cannot but do what he does, that is, eat the dead animal. And if this is so, then i t should be correct to say that the starving man is not able to refrain from eating. But 'Abd alJabb-r does not seem to see this. He says that being mulja7 does not contradict beingqiidir. And, of course, this is true when we take q a i r in the dispositional sense. For in this sense q d i r is opposed to incapucitated or being impotent. Clearly, one who is rnulji' may very well be able to eat as well as refrain from eating. For his limbs and general wndition may be such that it is not the case that his hands reach for the food, or fail to reach for the food, against his will, a s may be the case of someone who is subject to involuntary motion, or involuntary immobility. Thus we admit that being mulja' to do 'A' does not contradict its being the case that one can, in one sense of 'can', do 'A' as well as refrain from doing 'A'. But still, we want to insist that there is another sense of 'can' in which one who is mulja' to do 'A' cannot but do 'A'. Such a sense is what we want to associate with 'Abd a l - J a b b 2 s own statements about the unopposed da'rwhich necessitates the action inevitably (hi mahZlahJ, and Zayd's action which has to happen (yajibu wuqii'uhuJ when Zayd is willing, intending, and unrestrained. For clearly, if Zayd's action has to happen, then there is a sense of 'can' in which Zayd cannot but act. This will be true even if Zayd is sound of body and mind, so that in another, dispositional sense of 'can', he can refrain from acting. Thus it would appear that 'Abd a1J a b b s r understands 'can' only in its dispositional, freedom-related sense. Of the causal-deterministic sense of 'can', he shows no sign of recognition. If this understanding of 'Abd alJabbZr on the matter of 'can' is correct, then the contrast between his position and that of al-Ash'az begins to appear to be a matter of emphasizing different things, rather than a matter of conflicting principles. For his part, 'AM alJabbZr places emphasis on the idea that man can act or refrain from acting. This, he believes is required by the facts of Divine justice and human responsibility. As it turns out, the 'can' which he employs here is the dispositional 'can' which must be withheld only from those who are incapacitated (impotent) in some way or other. But this sense of 'can' is not one that al-Ash'aG cannot live with. In fact, he is logically required to recognize it when he denies that the unbeliever cannot (in the sense of incapacity) believe. For his own part, al-Ash'az emphasizes the idea that man's actions are determined, that man cannot do except what he does. But is this something that 'Abd alJabbiir needs to deny? No. In fact, there is evidence to suggest that he was a Psychological Determinist of sorts, and as such, he must recognize that there is a sense of 'can'

Hamdord Zslarnicus

56

Vol. XV No. 1

in which one (at least sometimes) cannot do except what one does. What is, therefore, the disagreement between 'Abd a1Jabb;r and al-Ash'as? The former says:' a M n can perform the act or its opposite", while the latter appears to deny this. But, if I am right, 'Abd alJabbZr and al-Ash'a: are not using 'can' in the same sense. One of them thinks in terms of the dispositional, freedom-related 'can', while the otherthinks in terms of the causal-deterministic 'can'. What we have here is not so much disagreement as diuergeme: al~ s h ' a r T 'Abd al-Jabbsr are simply talking about different things. and One consequence of this, I believe, is to weaken the ground for distinguishing 'Abd alJabb5r against al-AshCaXby the advocacy of free will. Another consequence should extend beyond al-Ash'az and 'Abd al-Jabbiir to the schools which they represented. We need to make more precise what the advocacy of free will amounted to in the debate between the Ash'arites and the Mu'tazilites. It may turn out that the relation between these two schools was one of divergence rather than contradiction.

NOTES and REFERENCES

Alvin Galdman, "Actions. Predictions and h k s of Life". in Reason a n d 91. Responsibility ed. .I. Feinberg (California: Dicksan Publishing Company, 1 7 ) o. 371. r G.E.M. Anscombe, "Causality a n d Determinism", i n Causation a n d Counlerfnctuals., ed. E. Sosa (Oxford: Oxford University Press, 1975), p. 65. Cf. R. Taylor, ' D e t e b n i s m and the Theory of Ageney", in Delerminism a n d Freedom, ed. S. Hmk (New York: Collier Books, 1961), p. 224. Here Taylor defmes Determinism aa the thesis that 'whatever o m m , -m under conditions given which nothing else muld occur.' The phrase Theistic Determinism' is used without definition by both M. Cook and WM. Watt. Far Cmk's usage see his Early Muslim Doetrim (Cambridge: Cambridge University Press, 1981), p. 150; far Watt's usage see his FHPWill and Predestination in Early Islam (London: Luzac & Co. Ltd., 19481,p. 96. M. Rahf('The Qur'an and Free Will", Muslim World, LX,1970, p. 205)also apeaks of Determinism used in a thenlogical as opposed ta seientifie sense. Thua there are p-denta for speaking of a Theistic or, as I shall say, Theological kind of Determinism. But we must explain how this kind of Determinism is related ta the more familiar kind of Determinism. And related it must be: for it can hardly be acrident that the Free Will problem arises in bath acientiiic and theological mntexts.Al-Ash'ari, The Theology of a l - A s h h c ed. R.J. McCarthy (Beirut: The Catholic Press, 1953),p. 82 (Arabic text p. 58, 1 1 ) See also note 18 on the translation .8. chosen here. Ibid.., p. 83 (Arabic text, p. 59, 1 2 . .) R.M. Frank,'The Structure of Created Causality Acmrdiog ta al-A&*, Studia Islamic* vol. XXV, (1966).p. 63. AI-~sh'ar?; MaqZlLt al-Islamiyyin, ed. H. Ritter Weisbaden: Franz Steiner Verlag), p. 291, 1 4 .. Ibid., 1.6. Frank, 'The Structure", pL 14. The Thedogy of d-Ash'ari, p. 84 (Arabic text, p. 60, 11.5-7).
~

~mdard Islamicus

57

Vol.

XV

No.

The Theolrgy d d-Ashhrg p. 59 (Arabic text, p. 41, 11.16-11). Frank, The Stmdure", p. 25. The Theolrgy of d - A s h h K p. 76 (Arabic text, p. 54, 11.7-9). See Hobhs, F.GB. Woodbridge (New York: Charles Scribner'~ Sons, 19301, p. 207. Here liberty' is defined as the absence of all impediment to adion.' As far as thjs definition is concerned, an action may be determined as well as free. For clearly, the abaence of impediments does not entail the denial af determinism. G B . Mwre, Tree Will", in Determinism, Free Will, and Humon responsibility, ed. G. Dworkin (Englewwd Clills: Prentice Hall Ine., 1970), p. 136. This name is suggested by G.EM. Anscornbe's description of this sense of 'can'. See her 'SoR Determinism", in The Collected Papers of G.E.M. Ansmmbe Metaphysics and the Philnsophy of Mind, (Oxlord: Basil Blackwell, 1981), p. 172. The Theology of NAshhrF pp. 82-93 (Arabic text, p. 58, 1.17 - p. 59, 1.5). I have lollawed McCerthy's translation, except lor ~referrinp'cannot' to 'is incapable' as s translation of 'la )asfati". Nothlng of importance depends on my verb-lor-verb translation, but it will facilitate seeing Al-Ash'ari's distinction between two senses of 'fannot'. The Theologyof d-Ashhri, p. 84 (Arabic text, p. 60, 11.5-7). Ibid., p. 85 (Arabic text, p. 60, 11.8-11). Ibid., p. 83 (Arabic text, p. 59, 1.3). Ibid., p. 83 (Arabic text, p. 59, 11.7-18). Ibid.. o. 77 (Arabic text. o. 54. 11.11-18). 1bid.i i1.79 (Arabic text; 56; 11.35). . d-Muhi!, val. I, ed. J.J. Houbin (Beirut: The Catholic Press, 1962), p. 432, 11.1in J.R.T.M. Petem, God's Created Speech, (Leiden: E.J. Brill, 1976). p. 203. ~ ~ - M u g h nvol. V , Pt. 1, ed. A F . al-Ahwani (Egypt, 19621, p. 185, 11.13-14. i, I This is relevant to the present point on-account ol t h e fact that 'able' is applicable to both God and man. No Lashbih (nnthrnpamorphism) is involved here, 'Abd aldabbzr claims, because God is able per se, while man is 'able by virtue a l a 'illah (cause, that is). See 01-MughnT vol. V, ed. M.M. Khudairi (Egypt, 1958). p. 205, 11.6-10. Ibid., p. 2-ffi, 11.2-3. d-Mughni, vol. VI, Pt. 1 p. 129, 11.7-8. , Ibid.. p. 18, 11.26. Ibid., p. 1_88, 1.13; also p. 193, 11.54. Al-Mughd, vol. VIII, ed. T. al-Taml( g p ,?), p. 126, 11.10-14.(italics added Eyt lor emphasis). The interp&etion prmented here of the mnnection between motives and actions is diflerent fmm the one which R.M. Frank prments in his 'The Autonomy of the Human Agent in the Teaching of 'Abd aldabbar", Le Museon, Val. 95, 1982. Aemrding to Frank, T o use a contemprary form a l expression, motivations amrding to the Basrian Mu'tazilah, operate, not as causes strictly speaking, but as reasons lor which the agent acts or does not act intentionally." (p. 348) But is obvious that reasons cannot (at least sometimes) be regarded as a soecies of cause? Why does Jones take his medicine? Beeause he desirerr to i(et well, and h e beliewa t h a t by taking hi. medicine he will get well. Thus the reamn lor, the cause of hie action is the desire to aet well m d the beliel about the means; Like causes, reasons are used in explanations, m d they are regularly asnociated with their 'ell"'. Far a discussion of the idea that reasons are a species of cause, D. Davidsan, .Actions, Reasons, and Causes", in the Nature ofHuman-Action. ed. M. Brand Glenview: Scott, Foreman m d Company, 1970). A-Mughni, Val. VI, R. 2, ed. I. Madkur (Egypt, ?), p. 185, 11.11-13, Ibid, pp. 185, 1.14-186, 1.3. The same point i. made in another mnnection: see p. 137, 11.6-7.

i.