The aim of this book is to show the different degrees of assent and conviction attained by the assertions in The Incoherence of the Philosophers, and to prove that the greater part has not reached the degree of evidence and of truth.

THE FIRST DISCUSSION

Concerning the Eternity of the World

Ghazali, speaking of the philosophers’ proofs for the eternity of the world, says:

Let us restrict ourselves in this chapter to those proofs that make an impression on the mind.

This chapter contains four proofs.

THE FIRST PROOF The philosophers say: It is impossible that the temporal should proceed from the absolutely Eternal. For it is clear if we assume the Eternal existing without, for instance, the world proceeding from Him, then, at a certain moment, the world beginning to proceed from Him-that it did not proceed before, because there was no determining principle for its existence, but its existence was pure possibility. When the world begins in time, a new determinant either does or does not arise. If it does not, the world will stay in the same state of pure possibility as before; if a new determinant does arise, the same question can be asked about this new determinant, why it determines now, and not before, and either we shall have an infinite regress or we shall arrive at a principle determining eternally.

I say: This argument is in the highest degree dialectical and does not reach the pitch of demonstrative proof. For its premisses are common notions, and common notions approach the equivocal, whereas demonstrative premisses are concerned with things proper to the same genus.

For the term ‘possible’ is used in an equivocal way of the possible that happens more often than not, of the possible that happens less often than not, and of the possible with equal chances of happening, and these three types of the possible do not seem to have the same need for a new determining principle. For the possible that happens more often than not is frequently believed to have its determining principle in itself, not outside, as is the case with the possible which has equal chances of happening and not happening. Further, the possible resides sometimes in the agent, i. e. the possibility of acting, and sometimes in the patient, i. e. the possibility of receiving, and it does not seem that the necessity for a determining principle is the same in both cases. For it is well known that the possible in the patient needs a new determinant from the outside; this can be perceived by the senses in artificial things and in many natural things too, although in regard to natural things there is a doubt, for in most natural things the principle of their change forms part of them. Therefore it is believed of many natural things that they move themselves, and it is by no means self-evident that everything that is moved has a mover and that there is nothing that moves itself.; But all this needs to be examined, and the old philosophers have therefore done so. As concerns the possible in the agent, however, in many cases it is believed that it can be actualized without an external principle, for the transition in the agent from inactivity to activity is often regarded as not being a change which requires a principle; e. g. the transition in the geometer from non-geometrizing to geometrizing, or in the teacher from non-teaching to teaching.

Further, those changes which are regarded as needing a principle of change can sometimes be changes in substance, sometimes in quality, or in quantity, or in place.

In addition, ‘eternal’ is predicated by many of the eternal-by-itself and the eternal-through-another. According to some, it is permissible to admit certain changes in the Eternal, for instance a new volition in the Eternal, according to the Karramites, and the possibility of generation and corruption which the ancients attribute to primary matter, although it is eternal. Equally, new concepts are admitted in the possible intellect although, according to most authors, it is eternal. But there are also changes which are inadmissible, especially according to certain ancients, though not according to others.

Then there is the agent who acts of his will and the agent which acts by nature, and the manner of actualization of the possible act is not the same for both agents, i. e. so far as the need for a new determinant is concerned. Further, is this division into two agents complete, or does demonstration lead to an agent which resembles neither the natural agent nor the voluntary agent of human experience?

All these are multifarious and difficult questions which need, each of them, a special examination, both in themselves and in regard to the opinions the ancients held about them. To treat what is in reality a plurality of questions as one problem is one of the well known seven sophisms, and a mistake in one of these principles becomes a great error by the end of the examination of reality.

Ghazali says:

There are two objections to this. The first objection is to say: why do you deny the theory of those who say that the world has been created by an eternal will which has decreed its existence in the time in which it exists; that its non-existence lasts until the moment it ceases and that its existence begins from the moment it begins; that its existence was not willed before and therefore did not happen, and that at the exact moment it began it was willed by an eternal will and therefore began? What is the objection to this theory and what is absurd in it?

I say:

This argument is sophistical: although it is not allowable for him to admit the possibility of the actual effect being delayed after the actual cause, and in a voluntary agent, after the decision to act, he regards it as possible that the effect should be delayed after the will of the agent. It is possible that the effect should be delayed after the will of the agent, but its being delayed after the actual cause is impossible, and equally impossible is its being delayed after a voluntary agent’s decision to act. The difficulty is thus unchanged, for he must of necessity draw one of these two conclusions: either that the act of the agent does not imply in him a change which itself would need an external principle of change, or that there are changes which arise by themselves, without the necessity of an agent in whom they occur and who causes them, and that therefore there are changes possible in the Eternal without an agent who causes them. And his adversaries insist on these two very points: ( 1 ) that the act of the agent necessarily implies a change and that each change has a principle which causes it; (2) that the Eternal cannot change in any way. But all this is difficult to prove.

The Ash’arites are forced to assume either a first agent or a first act of this agent, for they cannot admit that the disposition of the agent, relative to the effect, when he acts is the same as his disposition, when he does not act. This implies therefore a new disposition or a new relation, and this necessarily either in the agent, or in the effect, or in both? But in this case, if we posit as a principle that for each new disposition there is an agent, this new disposition in the first agent will either need another agent, and then this first agent was not the first and was not on his own account sufficient for the act but needed another, or the agent of the disposition which is the condition of the agent’s act will be identical with the agent of the act. Then this act which we regarded as being the first act arising out of him will not be the first, but his act producing the disposition which is the condition of the effect will be anterior to the act producing the effect. This, you see, is a necessary consequence, unless one allows that new dispositions may arise in the agents without a cause. But this is absurd, unless one believes that there are things which happen at haphazard and by themselves, a theory of the old philosophers who denied the agent,; the falsehood of which is self-evident.

In Ghazali’s objection there is a confusion. For our expressions ‘eternal will’ and ‘temporal will’ are equivocal, indeed contrary. For the empirical will is a faculty which possesses the possibility of doing equally one of two contraries and then of receiving equally one of the two contraries willed. For the will is the desire of the agent towards action. When the agent acts, the desire ceases and the thing willed happens, and this desire and this act are equally related to both the contraries. But when one says: ‘There is a Wilier who wills eternally one of two contraries in Himself’, the definition of the will is abandoned, for we have transferred its nature from the possible to the necessary. If it is objected that in an eternal will the will does not cease through the presence of the object willed, for as an eternal will has no beginning there is no moment in it which is specially determined for the realization of the object willed, we answer: this is not obvious, unless we say that demonstrative proof leads to the existence of an agent endowed with a power which is neither voluntary nor natural, which, however, the Divine Law calls ‘will’, in the same way as demonstrative proof leads to middle terms between things which seemed at first sight to be contrary, without being really so, as when we speak of an existence which is neither inside nor outside the world.

Ghazali answers, on behalf of the philosophers:

The philosophers say: This is clearly impossible, for everything that happens is necessitated and has its cause, and as it is impossible that there should be an effect without a necessitating principle and a cause, so it is impossible that there should exist a cause of which the effect is delayed, when all the conditions of its necessitating, its causes and elements are completely fulfilled. On the contrary, the existence of the effect, when the cause is realized with all its conditions, is necessary, and its delay is just as impossible as an effect without cause. Before the existence of the world there existed a Wilier, a will, and its relation to the thing willed. No new wilier arose, nor a new will, nor a new relation to the will-for all this is change; how then could a new object of will arise, and what prevented its arising before? The condition of the new production did not distinguish itself from the condition of the non-production in any way, in any mode, in any relation-on the contrary, everything remained as it was before. At one moment the object of will did not exist, everything remained as it was before, and then the object of will existed. Is not this a perfectly absurd theory?

I say:

This is perfectly clear, except for one who denies one of the premisses we have laid down previously. But Ghazali passes from this proof to an example based upon convention,’ and through this he confuses this defence of the philosophers.

Ghazali says:

This kind of impossibility is found not only in the necessary and essential cause and effect but also in the accidental and conventional. If a man pronounces the formula of divorce against his wife without the divorce becoming irrevocable immediately, one does not imagine that it will become so later. For he made the formula through convention and usage a cause of the judgement, and we do not believe that the effect can be delayed, except when the divorce depends on an ulterior event, e. g. on the arrival of tomorrow or on someone’s entering the house, for then the divorce does not take place at once, but only when tomorrow arrives or someone enters the house; in this case the man made the formula a cause only in conjunction with an ulterior event. But as this event, the coming of tomorrow and someone’s entering the house, is not yet actual, the effect is delayed until this future event is realized. The effect only takes place when a new event, i. e. entering the house or the arrival of tomorrow, has actually happened. Even if a man wanted to delay the effect after the formula, without making it dependent on an ulterior event, this would be regarded as impossible, although it is he himself who lays down the convention and fixes its modalities. If thus in conventional matters such a delay is incomprehensible and inadmissible, how can we admit it in essential, rational, and necessary causal relations? In respect of our conduct and our voluntary actions, there is a delay in actual volition only when there is some obstacle. When there is actual volition and actual power and the obstacles are eliminated, a delay in the object willed is inadmissible.; A delay in the object willed is imaginable only in decision, for decision is not sufficient for the existence of the act; the decision to write does not produce the writing, if it is not, as a new fact, accompanied by an act of volition, i. e. an impulse in the man which presents itself at the moment of the act. If there is thus an analogy between the eternal Will and our will to act, a delay of the object willed is inadmissible, unless through an obstacle, and an antecedent existence of the volition is equally inadmissible, for I cannot will to get up tomorrow except by way of decision. If, however, the eternal Will is analogous to our decision, it does not suffice to produce the thing decided upon, but the act of creation must be accompanied by a new act of volition, and this brings us again to the idea of a change. But then we have the same difficulty all over again. Why does this impulse or volition or will or whatever you choose to call it happen just now and not before? There remain, then, only these alternatives: either something happening without a cause, or an infinite regress. This is the upshot of the discussion: There is a cause the conditions of which are all completely fulfilled, but notwithstanding this the effect is delayed and is not realized during a period to the beginning of which imagination cannot attain and for which thousands of years would mean no diminution; then suddenly, without the addition of any new fact, and without the realization of any new condition, this effect comes into existence and is produced. And this is absurd.

I say:

This example of divorce based on convention seems to strengthen the argument of the philosophers, but in reality it weakens it. For it enables the Ash’arites to say: In the same way as the actual divorce is delayed after the formula of divorce till the moment when the condition of someone’s entering the house, or any other, is fulfilled, so the realization of the world can be delayed after God’s act of creation until the condition is fulfilled on which this realization depends, i. e. the moment when God willed it. But conventional things do not behave like rational. The Literalists, comparing these conventional things to rational, say: This divorce is not binding and does not become effective through the realization of the condition which is posterior to the pronouncement of the divorce by the divorcer, since it would be a divorce which became effective without connexion with the act of the divorcer. But in this matter there is no relation between the concept drawn from the nature of things and that which is artificial and conventional.

Then Ghazali says, on behalf of the Ash’arites:

The answer is: Do you recognize the impossibility of connecting the eternal Will with the temporal production of anything, through the necessity of intuitive thought or through a logical deduction, or-to use your own logical terminology-do you recognize the clash between these two concepts through a middle term or without a middle term? If you claim a middle term-and this is the deductive method-you will have to produce it, and if you assert that you know this through the necessity of thought, why do your adversaries not share this intuition with you? For the party which believes in the creation of the world in time through an eternal Will includes so many persons that no country can contain them and no number enumerate them, and they certainly do not contradict the logically minded out of obstinacy, while knowing better in their hearts. A proof according to the rules of logic must be produced to show this impossibility, as in all your arguments up till now there is only a presumption of impossibility and a comparison with our decision and our will; and this is false, for the eternal Will does not resemble temporal volitions, and a pure presumption of impossibility will not suffice without proof.

I say:

This argument is one of those which have only a very feeble persuasive power. It amounts to saying that one who claims the impossibility of delay in an effect, when its cause with all its conditions is realized, must assert that he knows this either by a syllogism or from first principles; if through a syllogism, he must produce it-but there is none; if from first principles, it must be known to all, adversaries and others alike. But this argument is mistaken, for it is not a condition of objective truth that it should be known to all. That anything should be held by all does not imply anything more than its being a common notion, just as the existence of a common notion does not imply objective truth.

Ghazali answers on behalf of the Ash’arites:

If it is said, ‘We know by the necessity of thought that, when all its conditions are fulfilled, a cause without effect is inadmissible and that to admit it is an affront to the necessity of thought,’ we answer: what is the difference between you and your adversaries, when they say to you, ‘We know by the necessity of thought the impossibility of a theory which affirms that one single being knows all the universals, without this knowledge forming a plurality in its essence or adding anything to it, and without this plurality of things known implying a plurality in the knowledge’? For this is your theory of God, which according to us and our science is quite absurd. You, however, say there is no analogy between eternal and temporal knowledge. Some of you acknowledge the impossibility involved, and say that God knows only Himself and that He is the knower, the knowledge and the known, and that the three are one. One might object: The unity of the knowledge, the knower, and the known is clearly an impossibility, for to suppose the Creator of the world ignorant of His own work is necessarily absurd, and the Eternal-who is far too high to be reached by your words and the words of any heretics-could, if He knows only Himself, never know His work.

I say

This amounts to saying that the theologians do not gratuitously and without proof deny the admitted impossibility of a delay between the effect and its cause, but base themselves on an argument which leads them to believe in the temporal creation of the world, and that they therefore act in the same way as the philosophers, who only deny the well-known necessary plurality of knowledge and known, so far as it concerns their unity in God, because of a demonstration which, according to them, leads them to their theory about Him. And that this is still more true of those philosophers who deny it to be necessary that God should know His own work, affirming that He knows only Himself. This assertion belongs to the class of assertions whose contrary is equally false., For there exists no proof which refutes anything that is evidently true, and universally acknowledged. Anything that can be refuted by a demonstrative proof is only supposed to be true, not really true.] Therefore, if it is absolutely and evidently true that knowledge and known form a plurality, both in the visible and in the invisible world, we can be sure that the philosophers cannot have a proof of this unity in God; but if the theory of the plurality of knowledge and known is only a supposition, then it is possible for the philosophers to have a proof. Equally, if it is absolutely true that the effect of a cause cannot be delayed after the causation and the Ash’arites claim that they can advance a proof to deny it, then we can be absolutely sure that they cannot have such a proof. If there is a controversy about questions like this, the final criterion rests with the sound understanding’ which does not base itself on prejudice and passion, when it probes according to the signs and rules by which truth and mere opinion are logically distinguished. Likewise, if two people dispute about a sentence and one says that it is poetry, the other that it is prose, the final judgment rests with the ‘sound understanding’ which can distinguish poetry from prose, and with the science of prosody. And as, in the case of metre, the denial of him who denies it does not interfere with its perception by him who perceives it, so the denial of a truth by a contradictor does not trouble the conviction of the men to whom it is evident.

This whole argument is extremely inept and weak, and Ghazali ought not to have filled his book with such talk if he intended to convince the learned.

And drawing consequences which are irrelevant and beside the point, Ghazali goes on to say:

But the consequences of this argument cannot be overcome. And we say to them: How will you refute your adversaries, when they say the eternity of the world is impossible, for it implies an infinite number and an infinity of unifies for the spherical revolutions, although they can be divided by six, by four, and by two.’ For the sphere of the sun revolves in one year, the sphere of Saturn in thirty years, and so Saturn’s revolution is a thirtieth and Jupiter’s revolution-for Jupiter revolves in twelve years-a twelfth of the sun’s revolution. But the number of revolutions of Saturn has the same infinity as the revolutions of the sun, although they are in a proportion of one to thirty and even the infinity of the sphere of the fixed stars which turns round once in thirty-six thousand years is the same as the daily revolution which the sun performs in twenty-four hours. If now your adversary says that this is plainly impossible, in what does your argument differ from his? And suppose it is asked: Are the numbers of these revolutions even or uneven or both even and uneven or neither even nor uneven? If you answer, both even and uneven, or neither even nor uneven, you say what is evidently absurd. If, however, you say ‘even’ or ‘uneven’, even and uneven become uneven and even by the addition of one unit and how could infinity be one unit short? You must, therefore, draw the conclusion that they are neither even nor uneven.

I say:

This too is a sophistical argument. It amounts to saying: In the same way as you are unable to refute our argument for the creation of the world in time, that if it were eternal, its revolutions would be neither even nor uneven, so we cannot refute your theory that the effect of an agent whose conditions to act are always fulfilled cannot be delayed. This argument aims only at creating and establishing a ; doubt, which is one of the sophist’s objectives.

But you, reader of this book, you have already heard the arguments of the philosophers to establish the eternity of the world and the refutation of the Ash’arites. Now hear the proofs of the Ash’arites for their refutation and hear the arguments of the philosophers to refute those proofs in the wording of Ghazali!

[Here, in the Arabic text, the last passage of Ghazali, which previously was given only in an abbreviated form, is repeated in full.]

I say:

This is in brief that, if you imagine two circular movements in one and the same finite time and imagine then a limited part of these movements in one and the same finite time, the proportion between the parts of these two circular movements and between their wholes will be the same. For instance, if the circular movement of Saturn in t the period which we call a year is a thirtieth of the circular movement of the sun in this period, and you imagine the whole of the circular movements of the sun in proportion to the whole of the circular movements of Saturn in one and the same period, necessarily the proportion between their wholes and between their parts will be the same. If, however, there is no proportion between two movements in their totality, because they are both potential, i. e. they have neither beginning nor end but there exists a proportion between the parts, because they are both actual, then the proportion between the wholes is not necessarily the same as the proportion between the parts-although many think so, basing their proof on this prejudice -for there is no proportion between two magnitudes or quantities which are both taken to be infinite. When, therefore, the ancients believed that, for instance, the totality of the movements of the sun and of Saturn had neither beginning nor end, there could be no proportion between them, for this would have implied the finitude of both these totalities, just as this is implied for the parts of both. This is self-evident. Our adversaries believe that, when a proportion of more and less exists between parts, this proportion holds good also for the totalities, but this is only binding when the totalities are finite. For where there is no end there is neither ‘more’ nor ‘less’. The admission in such a case of the proportion of more and less brings with it another absurd consequence, namely that one infinite could be greater than another. This is only absurd when one supposes two things actually infinite, for then a proportion does exist between them. When, however, one imagines things potentially infinite, there exists no proportion at all. This is the right answer to this question, not what Ghazali says in the name of the philosophers.

And through this are solved all the difficulties which beset our adversaries on this question, of which the greatest is that which they habitually formulate in this way: If the movements in the past are infinite, then no movement in the actual present can take place, unless an infinite number of preceding movements is terminated., This is true, and acknowledged by the philosophers, once granted that the anterior movement is the condition for the posterior movement’s taking place, i. e. once granted that the existence of one single movement implies an infinite number of causes. But no philosopher allows the existence of an infinite number of causes, as accepted by the materialists, for this would imply the existence of an effect without cause and a motion without mover. But when the existence of an eternal prime mover had been proved, whose act cannot be posterior to his being, it followed that there could as little be a beginning for his act as for his being; otherwise his act would be possible, not necessary, and he would not be a first principle.’ The acts of an agent who has no beginning have a beginning as little as his existence, and therefore it follows necessarily that no preceding act of his is the condition for the existence of a later, for neither of them is an agent by itself and their sequence is accidental. An accidental infinite, not an essential infinite, is admitted by the philosophers; nay, this type of infinite is in fact a necessary consequence of the existence of an eternal first principle., And this is not only true for successive or continuous movements and the like, but even where the earlier is regarded as the cause of the later, for instance the man who engenders a man like himself. For it is necessary that the series of temporal productions of one individual man by another should lead upwards to an eternal agent, for whom there is no beginning either of his existence or of his production of man out of man. The production of one man by another ad infinitum is accidental, whereas the relation of before and after in it is essential. The agent who has no beginning either for his existence or for those acts of his which he performs without an instrument, has no first instrument either to perform those acts of his without beginning which by their nature need an instrument .

But since the theologians mistook the accidental for the essential, they denied this eternal agent; the solution of their problem was difficult and they believed this proof to be stringent. But this theory of the philosophers is clear, and their first master Aristotle has explained that, if motion were produced by motion, or element by element, motion and element could not exists For this type of infinite the philosophers admit neither a beginning nor an end, and therefore one can never say of anything in this series that it has ended or has begun, not even in the past, for everything that has an end must have begun and what does not begin does not end. This can also be understood from the fact that beginning and end are correlatives. Therefore one who affirms that there is no end of the celestial revolutions in the future cannot logically ascribe a beginning to them, for what has a beginning has an end and what has no end has no beginning, and the same relation exists between first and last; i. e. what has a first term has also a last term, and what has no first term has no last term, and there is in reality neither end nor beginning for any part of a series that has no last term, and what has no beginning for any of its parts has no end for any of them either. When, therefore, the theologians ask the philosophers if the movements which precede the present one are ended, their answer is negative, for their assumption that they have no beginning implies their endlessness. The opinion of the theologians that the philosophers admit their end is erroneous, for they do not admit an end for what has no beginning.’ It will be clear to you that neither the arguments of the theologians for the temporal creation of the world of which Ghazali speaks, nor the arguments of the philosophers which he includes and describes in his book, suffice to reach absolute evidence or afford stringent proof. And this is what we have tried to show in this book. The best answer one can give to him who asks where in the past is the starting-point of His acts, is: The starting-point of His acts is at the starting-point of His existence; for neither of them has a beginning.

And here is the passage of Ghazali in which he sets forth the defence of the philosophers against the argument built on the difference in speed of the celestial spheres, and his refutation of their argument.

Ghazali says:

If one says, ‘The error in your argument consists in your considering those circular movements as an aggregate of units, but those movements have no real existence, for the past is no more and the future not yet; “aggregate” means units existing in the present, but in this case there is no existence.’

Then he says to refute this:

We answer: Number can be divided into even and uneven; there is no third possibility, whether for the numbered permanent reality, or for the numbered passing event. Therefore whatever number we imagine, we must believe it to be even or uneven, whether we regard it as existent or non-existent; and if the thing numbered vanishes from existence, our judgement of its being even or uneven does not vanish or change.

I say:

This is the end of his argument. But this argument-that the numbered thing must be judged as even or uneven, whether it exists or not-is only valid so far as it concerns external things or things in the soul that have a beginning and an end. For of the number which exists only potentially, i. e. which has neither beginning nor end, it cannot truly be said that it is even or uneven, or that it begins or ends; it happens neither in the past nor in the future, for what exists potentially falls under the law of non-existence. This is what the philosophers meant when they said that the circular movements of the past and the future are non-existent. The upshot of this question is: Everything that is called a limited aggregate with a beginning and an end is so called either because it has a beginning and end in the world exterior to the soul, or because it is inside, not outside, the soul. Every totality, actual and limited in the past, whether inside or outside the soul, is necessarily either even or uneven. But an unlimited aggregate existing outside the soul cannot be other than limited so far as it is represented in the soul, for the soul cannot represent unlimited existence. Therefore also this unlimited aggregate, as being limited in the soul, can be called even or uneven; in so far, however, as it exists outside the soul, it can be called neither even nor uneven. Equally, past aggregates which are considered to exist potentially outside the soul, i. e. which have no beginning, cannot be called even or uneven unless they are looked upon as actual, i. e. as having beginning and end. No motion possesses totality or forms an aggregate, i. e. is provided with a beginning or an end, except in so far as it is in the soul, as is the case with time.’ And it follows from the nature of circular movement that it is neither even nor uneven except as represented in the soul. The cause of this mistake is that it was believed that, when something possesses a certain quality in the soul, it must possess this quality also outside the soul, and, since anything that has happened in the past can only be represented in the soul as finite, it was thought that everything that has happened in the past must also be finite outside the soul. And as the circular movements of the future are regarded by the imagination as infinite, for it represents them as a sequence of part after part, Plato and the Ash’arites believed that they might be infinite, but this is simply a judgement based on imagination, not on proof. Therefore those who believe-as many theologians have done-that, if the world is supposed to have begun, it must have an end, are truer to their principles and show more consistency.

Ghazali says after this:

And we say moreover to the philosophers: According to your principles it is not absurd that there should be actual units, qualitatively differentiated, which are infinite in number; I am thinking of human souls, separated through death from their bodies. These are therefore realities that can neither be called even nor uneven. How will you refute the man who affirms that this is necessarily absurd in the same way as you claim the connexion between an eternal will and a temporal creation to be necessarily absurd? This theory about souls is that which Avicenna accented. and it is perhaps Aristotle’s.

I say:

This argument is extremely weak. It says, in brief, ‘You philosophers need not refute our assertion that what is a logical necessity for you is not necessary, as you consider things possible which your adversaries consider impossible by the necessity of thought. That is to say, just as you consider things possible which your adversaries consider impossible, so you consider things necessary which your adversaries do not consider so. And you cannot bring a criterion for judging the two claims.’ It has already been shown in the science of logic that this is a weak rhetorical or sophistical kind of argument., The answer is that what we claim to be necessarily true is objectively true, whereas what you claim as necessarily absurd is not as you claim it to be. For this there is no other criterion than immediate intuitive apprehension, just as, when one man claims that a line is rhythmical and another denies it, the criterion is the intuition of the sound understanding.

As for the thesis of a numerical plurality of immaterial souls, this is not a theory acknowledged by the philosophers, for they regard matter as the cause of numerical plurality and form as the cause of congruity in numerical plurality. And that there should be a numerical plurality without matter, having one unique form, is impossible. For in its description one individual can only be distinguished from another accidentally, as there is often another individual who participates in this descriptions but only through their matter do individuals differ in reality. And also this: the impossibility of an actual infinite is an acknowledged axiom in philosophical theory, equally valid for material and immaterial things. We do not know of any one who makes a distinction here between the spatial and the non-spatial, with the single exception of Avicenna. I do not know of any other philosopher who affirms this, it does not correspond with any of their principles and it makes no sense, for the philosophers deny the existence of an actual infinite equally for material and for immaterial things, as it would imply that one infinite could be greater than another. Perhaps Avicenna wanted only to satisfy the masses, telling them what they were accustomed to hear about the soul. But this theory is far from satisfactory. For if there were an actual infinite and it were divided in two, the part would equal the whole; e. g. if there were a line or a number actually infinite in both directions and it were divided in two, both the parts and the whole would be actually infinite; and this is absurd. All this is simply the consequence of the admission of an actual and not potential infinite.

Ghazali says:

If it is said, ‘The truth lies with Plato’s theory of one eternal soul which is only divided in bodies and returns after its separation from them to its original unity’, we answer: This theory is still worse, more objectionable and more apt to be regarded as contrary to the necessity of thought. For we say that the soul of Zaid is either identical with the soul of Amr or different from it; but their identity would mean something absurd, for everyone is conscious of his own identity and knows that he is not another, and, were they identical, their knowledge, which is an essential quality of their souls and enters into all the relations into which their souls enter, would be identical too. If you say their soul is unique and only divided through its association with bodies, we answer that the division of a unity which has no measurable volume is absurd by the necessity of thought. And how could the one become two, and indeed a thousand, and then return to its unity? This can be understood of things which have volume and quantity, like the water of the sea which is distributed into brooks and rivers and flows then back again into the sea, but how can that which has no quantity be divided? We seek to show by all this that the philosophers cannot shake the conviction of their adversaries that the eternal Will is connected with temporal creation, except by claiming its absurdity by the necessity of thought, and that therefore they are in no way different from the theologians who make the same claim against the philosophical doctrines opposed to theirs. And out of this there is no issue.

I say:

Zaid and Amr are numerically different, but identical in form. If, for example, the soul of Zaid were numerically different from the soul of Amr in the way Zaid is numerically different from Amr, the soul of Zaid and the soul of Amr would be numerically two, but one in their form, and the soul would possess another soul. The necessary conclusion is therefore that the soul of Zaid and the soul of Amr are identical in their form. An identical form inheres in a numerical, i. e. a divisible, multiplicity, only through the multiplicity of matter. If then the soul does not die when the body dies, or if it possesses an immortal element, it must, when it has left the bodies, form a numerical unity. But this is not the place to go deeper into this subject.

His argument against Plato is sophistical. It says in short that the soul of Zaid is either identical with the soul of Amr or different from it; but that the soul of Zaid is not identical with the soul of Amr and that therefore it is different from it. But ‘different’ is an equivocal term, and ‘identity’ too is predicated of a number of things which are also called ‘different’. The souls of Zaid and Amr are one in one sense and many in another; we might say, one in relation to their form, many in relation to their substratum. His remark that division can only be imagined of the quantitative is partially false; it is true of essential division, but not of accidental division, i. e. of those things which can be divided, because they exist in the essentially divisible. The essentially divisible is, for example, body; accidental division is, for instance, the division of whiteness, when the bodies in which it is present are divided, and in this way the forms and the soul are accidentally divisible, i. e. through the division of the substrate. The soul is closely similar to light: light is divided by the division of illuminated bodies, and is unified when the bodies are annihilated, and this same relation holds between soul and bodies. To advance such sophistical arguments is dishonest, for it may be supposed that he is not a man to have overlooked the points mentioned. What he said, he said only to flatter the masses of his times, but how far removed is such an attitude from the character of those who seek to set forth the truth! But perhaps the man may be forgiven on account of the time and place in which he lived; and indeed he only proceeded in his books in a tentative way.

And as these arguments carry no evidence whatsoever, Ghazali says:

We want to show by all this that the philosophers cannot shake the conviction of their adversaries that the eternal Will is connected with temporal creation, by claiming its absurdity by the necessity of thought, and that therefore they do not distinguish themselves from the theologians, who make the same claim against the philosophical doctrines opposed to theirs. And out of this there is no issue.

I say:

When someone denies a truth of which it is absolutely certain that it is such-and-such, there exists no argument by which we can come to an understanding with him; for every argument is based on known premisses about which both adversaries agree. When each point advanced is denied by the adversary, discussion with him becomes impossible, but such people stand outside the pale of humanity and have to be educated. But for him who denies an evident truth, t because of a difficulty which presents itself to him there is a remedy, i. e. the solution of this difficulty. He who does not understand evident truth, because he is lacking in intelligence, cannot be taught anything, nor can he be educated. It is like trying to make the blind imagine colours or know their existence.

Ghazali says:

The philosophers may object: This argument (that the present has been preceded by an infinite past) can be turned against you, for God before the creation of the world was able to create it, say, one year or two years before He did, and there is no limit to His power; but He seemed to have patience and did not create. Then He created. Now, the duration of His inactivity is either finite or infinite. If you say finite, the existence of the Creator becomes finite; if you say infinite, a duration in which there is an infinite number of possibilities receives its termination. We answer: Duration and time are, according to us, created, but we shall explain the real answer to this question when we reply to the second proof of the philosophers.

I say:

Most people who accept a temporal creation of the world believe time to have been created with it. Therefore his assertion that the duration of His inactivity was either limited or unlimited is untrue. For what has no beginning does not finish or end. And the opponent does not admit that the inactivity has any duration at all. What one has to ask them about the consequences of their theory is: Is it possible, when the creation of time is admitted, that the term of its beginning may lie beyond the real time in which we live? If they answer that it is not possible, they posit a limited extension beyond which the Creator cannot pass, and this is, in their view, shocking and absurd. If, however, they concede that its possible beginning may lie beyond the moment of its created term, it may further be asked if there may not lie another term beyond this second. If they answer in the affirmative-and they cannot do otherwise-it will be said: Then we shall have here a possible creation of an infinite number of durations, and you will be forced to admit-according to your argument about the spherical revolutions-that their termination is a condition for the real age which exists since them. If you say what is infinite does not finish, the arguments you use about the spherical revolutions against your opponents your opponents will use against you on the subject of the possibility of created durations. If it is objected that the difference between those two cases is that these infinite possibilities belong to extensions which do not become actual, whereas the spherical revolutions do become actual, the answer is that the possibilities of things belong to their necessary accidents and that it does not make any difference, according to the philosophers, if they precede these things or are simultaneous with them, for of necessity they are the dispositions of things. If, then, it is impossible that before the existence of the present spherical revolution there should have been infinite spherical revolutions, the existence of infinite possible revolutions is equally impossible. If one wants to avoid these consequences, one can say that the age of the world is a definite quantity and cannot be longer or shorter than it is, in conformity with the philosophical doctrine about the size of the world. Therefore these arguments are not stringent, and the safest way for him who accepts the temporal creation of the world is to regard time as of a definite extension and not to admit a possibility which precedes the possible; and to regard also the spatial extension of the world as finite. Only, spatial extension forms a simultaneous whole; not so time.

Ghazali expounds a certain kind of argument attributed to the philosophers on this subject against the theologians when they denied that the impossibility of delay in the Creator’s act after His existence is known by primitive intuition:

How will you defend yourselves, theologians, against the philosophers, when they drop this argument, based on the necessity of thought, and prove the eternity of the world in this way, saying that times are equivalent so far as the possibility that the Divine Will should attach itself to them is concerned, for what differentiates a given time from an earlier or a later time? And it is not absurd to believe that the earlier or the later might be chosen when on the contrary you theologians say about white, black, movement, and rest that the white is realized through the eternal Will although its substrate accepts equally black and white. Why, then, does the eternal Will attach itself to the white rather than to the black, and what differentiates one of the two possibles from the other for connexion with the eternal Will? But we philosophers know by the necessity of thought that one thing does not distinguish itself from a similar except by a differentiating principle, for if not, it would be possible that the world should come into existence, having the possibility both of existing and of not existing, and that the side of existence, although it has the same possibility as the side of non-existence, should be differentiated without a differentiating principle. If you answer that the Will of God is the differentiating principle, then one has to inquire what differentiates the Will, i. e. the reason why it has been differentiated in such or such way. And if you answer: One does not inquire after the motives of the Eternal, well, let the world then be eternal, and let us not inquire after its Creator and its cause, since one does not inquire after the motives of the Eternal! If it is regarded as possible that the Eternal should differentiate one of the two possibles by chance, it will be an extreme absurdity to say that the world is differentiated in differentiated forms which might just as well be otherwise, and one might then say that this has happened by chance in the same way as you say that the Divine Will has differentiated one time rather than another or one form rather than another by chance. If you say that such a question is irrelevant, because it refers to anything God can will or decide, we answer that this question is quite relevant, for it concerns any time and is pertinent for our opponents to any decision God takes.

We answer: The world exists, in the way it exists, in its time, with its qualities, and in its space, by the Divine Will and will is a quality which has the faculty of differentiating one thing from another,’ and if it had not this faculty, power in itself would suffice But, since power is equally related to two contraries’ and a differentiating principle is needed to differentiate one thing from a similar, it is said that the Eternal possesses besides His power a quality which can differentiate between two similars. And to ask why will differentiates one of two similars is like asking why knowledge must comprehend the knowable, and the answer is that ‘knowledge’ is the term for a quality which has just this nature. And in the same way, ‘will’ is the term for a quality the nature or rather the essence of which is to differentiate one thing from another.

The philosophers may object: The assumption of a quality the nature of which is to differentiate one thing from a similar one is something incomprehensible, nay even contradictory, for ‘similar’ means not to be differentiated, and ‘differentiated’ means not similar. And it must not be believed that two blacks in two substrates are similar in every way, since the one is in one place and the other in another, and this causes a distinction; nor are two blacks at two times in one substrate absolutely similar, since they are separated in time, and how could they therefore be similar in every way? When we say of two blacks that they are similar, we mean that they are similar in blackness, in their special relation to it-not absolutely. Certainly, if the substrate and the time were one without any distinction, one could not speak any more of two blacks or of any duality at all. This proves that the term ‘Divine Will’ is derived from our will, and one does not imagine that through our will two similar things can be differentiated.’ On the contrary, if someone who is thirsty has before him two cups of water, similar in everything in respect to his aim, it will not be possible for him to take either of them. No, he can only take the one he thinks more beautiful or lighter or nearer to his right hand, if he is right-handed, or act from some such reason, hidden or known. Without this the differentiation of the one from the other cannot be imagined.

I say:

The summary of what Ghazali relates in this section of the proofs of the philosophers for the impossibility of a temporal proceeding from an eternal agent is that in God there cannot be a will. The philosophers could only arrive at this argument after granting to their opponents that all opposites-opposites in time, b like anterior and posterior, as well as those in quality, like white and black-are equivalent in relation to the eternal Will. And also non-existence and existence are, according to the theologians, equivalent in relation to the Divine Will. And having granted their opponents this premiss, although they did not acknowledge its truth, they said to them: It is of the nature of will that it cannot give preponderance to one thing rather than to a similar one, except through a differentiating principle and a cause which only exist in one of these two similar things; if not, one of the two would happen by chance-and the philosophers argued for the sake of discussion, as if they had conceded that, if the Eternal had a will, a temporal could proceed from an eternal. As the theologians were unable to give a satisfactory answer, they took refuge in the theory that the eternal Will is a quality the nature of which is to differentiate between two similar things, without there being for God a differentiating principle which inclines Him to one of two similar acts; that the eternal Will is thus a quality like warmth which gives heat or like knowledge which comprehends the knowable. But their opponents, the philosophers, answered: It is impossible that this should happen, for two similar things are equivalent for the wilier, and his action can only attach itself to the one rather than to the other through their being dissimilar, i. e. through one’s having a quality the other has not. When, however, they are similar in every way and when for God there is no differentiating principle at all, His will will attach itself to both of them indifferently and, when this is the case-His will being the cause of His act-the act will not attach itself to the one rather than to the other, it will attach itself either to the two contrary actions simultaneously or to neither of them at all, and both cases are absurd. The philosophers, therefore, began their argument, as if they had it granted to them that all things were equivalent in relation to the First Agent, and they forced them to admit that there must be for God a differentiating principle which precedes Him, which is absurd. When the theologians answered that will is a quality the nature of which is to differentiate the similar from the similar, in so far as it is similar, the philosophers objected that this is not understood or meant by the idea of will. They therefore appear to reject the principle which they granted them in the beginning.’ This is in short the content of this section. It waves the argument from the original question to the problem of the will; to shift one’s ground, however, is an act of sophistry.

Ghazali answers in defence of the theological doctrine of the Divine Will:

There are two objections: First, as to your affirmation that you cannot imagine this, do you know it by the necessity of thought or through deduction? You can claim neither the one nor the other. Your comparison with our will is a bad analogy, which resembles that employed on the question of God’s knowledge. Now God’s knowledge is different from ours in several ways which we acknowledge. Therefore it is not absurd to admit a difference in the will. Your affirmation is like saying that an essence existing neither outside nor inside the world, neither continuous with the world nor separated from it, cannot be understood, because we cannot understand this according to our human measure; the right answer is that it is the fault of your imagination, for rational proof has led the learned to accept its truth. How, then, will you refute those who say that rational proof has led to establishing in God a quality the nature of which is to differentiate between two similar things? And, if the word ‘will’ does not apply, call it by another name, for let us not quibble about words! We only use the term ‘will’ by permission of the Divine Law. It may be objected that by its conventional meaning ‘will’ designates that which has desire, and God has no desire, but we are concerned here with a question not of words but of fact. Besides, we do not even with respect to our human will concede that this cannot be imagined. Suppose two similar dates in front of a man who has a strong desire for them, but who is unable to take them both. Surely he will take one of them through a quality in him the nature of which is to differentiate between two similar things. All the distinguishing qualities you have mentioned, like beauty or nearness or facility in taking, we can assume to be absent, but still the possibility of the taking remains. You can choose between two answers: either you merely say that an equivalence in respect to his desire cannot be imagined-but this is a silly answer, for to assume it is indeed possible or you say that if an equivalence is assumed, the man will remain for ever hungry and perplexed, looking at the dates without taking one of them, and without a power to choose or to will, distinct from his desire. And this again is one of those absurdities which are recognized by the necessity of thought. Everyone, therefore, who studies, in the human and the divine, the real working of the act of choice, must necessarily admit a quality the nature of which is to differentiate between two similar things.

I say:

This objection can be summarized in two parts: In the first Ghazali concedes that the human will is such that it is unable to differentiate one thing from a similar one, in so far as it is similar, but that a rational proof forces us to accept the existence of such a quality in the First Agent. To believe that such a quality cannot exist would be like believing that there cannot exist a being who is neither inside nor outside the world. According to this reasoning, will, which is attributed to the First Agent and to man, is predicated in an equivocal way, like knowledge and other qualities which exist in the Eternal in a different way from that in which they exist in the temporal, and it is only through the prescription of the Divine Law that we speak of the Divine Will. It is clear that this objection cannot have anything more than a dialectical value. For a proof that could demonstrate the existence of such a quality, i. e. a principle determining the existence of one thing rather than that of a similar, would have to assume things willed that are similar; things willed are, however, not similar, but on the contrary opposite, for all opposites can be reduced to the opposition of being and not being, which is the extreme form of opposition; and opposition is the contrary of similarity. The assumption of the theologians that the things to which the will attaches itself are similar is a false one, and we shall speak of it later. If they say: we affirm only that they are similar in relation to the First Wilier, who in His holiness is too exalted to possess desires, and it is through desires that two similar things are actually differentiated, we answer: as to the desires whose realization contributes to the perfection of the essence of the wilier, as happens with our desires, through which our will attaches itself to the things willed-those desires are impossible in God, for the will which acts in this way is a longing for perfection when there is an imperfection in the essence of the wilier; but as to the desires which belong to the essence of the things willed, nothing new comes to the wilier from their realization. It comes exclusively to the thing willed, for instance, when a thing passes into existence from non-existence, for it cannot be doubted that existence is better for it than non-existence. It is in this second way that the Primal Will is related to the existing things, for it chooses for them eternally the better of two opposites, and this essentially and primally. This is the first part of the objection contained in this argument.

In the second part he no longer concedes that this quality cannot exist in the human will, but tries to prove that there is also in us, in the face of similar things, a will which distinguishes one from the other; of this he gives examples. For instance, it is assumed that in front of a man there are two dates, similar in every way, and it is supposed that he cannot take them both at the same time. It is supposed that no special attraction need be imagined for him in either of them, and that nevertheless he will of necessity distinguish one of them by taking it. But this is an error. For, when one supposes such a thing, and a wilier whom necessity prompts to eat or to take the date, then it is by no means a matter of distinguishing between two similar things when, in this condition, he takes one of the two dates. It is nothing but the admission of an equivalence of two similar things; for whichever of the two dates he may take, his aim will be attained and his desire satisfied. His will attaches itself therefore merely to the distinction between the fact of taking one of them and the fact of leaving them altogether; it attaches itself by no means to the act of taking one definite date and distinguishing this act from the act of leaving the other (that is to say, when it is assumed that the desires for the two are equal); he does not prefer the act of taking the one to the act of taking the other, but he prefers the act of taking one of the two, whichever it may be, and he gives a preference to the act of taking over the act of leaving.’ This is self-evident. For distinguishing one from the other means giving a preference to the one over the other, and one cannot give a preponderance to one of two similar things in so far as it is similar to the other-although in their existence as individuals they are not similar since each of two individuals is different from the other by reason of a quality exclusive to it. If, therefore, we assume that the will attaches itself to that special character of one of them, then it can be imagined that the will attaches to the.-one rather than the other because of the element of difference existing in both. But then the will does not attach itself to two similar objects, in so far as they are similar. This is, in short, the meaning of Ghazali’s first objection. Then he gives his second objection against those who deny the existence of a quality, distinguishing two similar objects from one another.

Ghazali says:

The second objection is that we say: You in your system also are unable to do without a principle differentiating between two equals, for the world exists in virtue of a cause which has produced it in its peculiar shape out of a number of possible distinct shapes which are equivalent; why, then, has this cause differentiated some of them? If to distinguish two similar things is impossible, it is irrelevant whether this concerns the act of God, natural causality, or the logical necessity of ideas. Perhaps you will say: the universal order of the world could not be different from what it is; if the world were smaller or bigger than it actually is, this order would not be perfect, and the same may be asserted of the number of spheres and of stars. And perhaps you will say: The big differs from the small and the many from the few, in so far as they are the object of the will, and therefore they are not similar but different; but human power is too feeble to perceive the modes of Divine Wisdom in its determination of the measures and qualities of things; only in some of them can His wisdom be perceived, as in the obliquity of the ecliptic in relation to the equator, and in the wise contrivance of the apogee and the eccentric sphere.’ In most cases, however, the secret is not revealed, but the differences are known, and it is not impossible that a thing should be distinguished from another, because the order of the world depends on it; but certainly the times are absolutely indifferent in relation to the world’s possibility and its order, and it cannot be claimed that, if the world were created one moment later or earlier, this order could not be imagined; and this indifference is known by the necessity of thought.-But then we answer: Although we can employ the same reasoning against your argument in the matter of different times, for it might be said that God created the world at the time most propitious for its creation, we shall not limit ourselves to this refutation, but shall assume, according to your own principle, a differentiation in two points about which there can be no disagreement: (1) the difference in the direction of spherical movement; (2) the definite place of the poles in relation to the ecliptic in spherical movement. The proof of the statement relating to the poles is that heaven is a globe, moving on two poles, as on two immovable points, whereas the globe of heaven is homogeneous and simple, especially the highest sphere, the ninth, which possesses no stars at all, and these two spheres move on two poles, the north and the south. We now say: of all the opposite points, which are infinite, according to you philosophers, there is no pair one could not imagine as poles. Why then have the two points of the north and south pole been fixed upon as poles and as immovable; and why does the ecliptic not pass through these two poles, so that the poles would become the opposite points of the ecliptic? And if wisdom is shown in the size and shape of heaven, what then distinguishes the place of the poles from others, so that they are fixed upon to serve as poles, to the exclusion of all the other parts and points? And yet all the points are similar, and all parts of the globe are equivalent. And to this there is no answer.

One might say: Perhaps the spot in which the point of the poles is, is distinguished from other points by a special quality, in relation to its being the place of the poles and to its being at rest, for it does not seem to change its place or space or position or whatever one wishes to call it; and all the other spots of the sphere by turning change their position in relation to the earth and the other spheres and only the poles are at rest; perhaps this spot was more apt to be at rest than the others. We answer: If you say so, you explain the fact through a natural differentiation of the parts of the first sphere; the sphere, then, ceases to be homogeneous, and this is in contradiction with your principle, for one of the proofs by which you prove the necessity of the globular shape of heaven, is that its nature is simple, homogeneous, and without differentiation, and the simplest shape is the globe; for the quadrangle and the hexagon and other figures demand a salience and a differentiation of the angles,’ and this happens only when its simple nature is added to. But although this supposition of yours is in contradiction with your own theory, it does not break the strength of your opponents’ argument; the question about this special quality still holds good, namely, can those other parts accept this quality or not? If the answer is in the affirmative, why then is this quality limited to a few only of those homogeneous parts? If the answer is negative, we reply: the other parts, in so far as they constitute bodies, receiving the form of bodies, are homogeneous of necessity, and there is no justification for attributing this special quality to this spot exclusively on account of its being a part of a body and a part of heaven, for the other parts of heaven participate in this qualification. Therefore its differentiation must rest on a decision by God, or on a quality whose nature consists in differentiating between two similars. Therefore, just as among philosophers the theory is upheld that all times are equivalent in regard to the creation of the world, their opponents are justified in claiming that the parts of heaven are equivalent for the reception of the quality through which stability in position becomes more appropriate than a change of position. And out of this there is no issue.

I say:

This means in brief that the philosophers must acknowledge that there is a quality in the Creator of the world which differentiates between two similars, for it seems that the world might have had another shape and another quantity than it actually has, for it might have been bigger or smaller. Those different possibilities are, therefore, equivalent in regard to the determination of the existence of the world. On the other hand, if the philosophers say that the world can have only one special shape, the special quantity of its bodies and the special number of them it actually has, and that this equivalence of possibilities can only be imagined in relation to the times of temporal creation-since for God no moment is more suitable than another for its creation-they may be told that it is possible to answer this by saying that the creation of the world happened at its most propitious moment. But we, the theologians say, want to show the philosophers two equivalent things of which they cannot affirm that there exists any difference between them; the first is the particular direction of the spherical movement and the second the particular position of the poles, relative to the spheres; for any pair whatever of opposite points, united by a line which passes through the centre of the sphere, might constitute the poles. But the differentiation of these two points, exclusive of all other points which might just as well be the poles of this identical sphere cannot happen except by a quality differentiating between two similar objects. If the philosophers assert that it is not true that any other place on the sphere might be the seat for these poles, they will be told: such an assertion implies that the parts of the spheres are not homogeneous and yet you have often said that the sphere is of a simple nature and therefore has a simple form, viz. the spherical. And again, if the philosophers affirm that there are spots on the sphere which are not homogeneous, it will be asked how these spots came to be of a heterogeneous nature; is it because they are a body or because they are a celestial body? But the absence of homogeneity cannot be explained in this way. Therefore-Ghazali says just as among philosophers the theory is upheld that all times are equivalent in regard to the creation of the world, the theologians are justified in claiming that the parts of heaven are equivalent in regard to their serving as poles, and that the poles do not seem differentiated from the other points through a special position or through their being in an immovable place, exclusive of all other places.

This then in short is the objection; it is, however, a rhetorical one, for many things which by demonstration can be found to be necessary seem at first sight merely possible.’ The philosophers’ answer is that they assert that they have proved that the world is composed of five bodies: a body neither heavy nor light, i. e. the revolving spherical body of heaven and four other bodies, two of which are earth, absolutely heavy, which is the centre of the revolving spherical body, and fire, absolutely light, which is seated in the extremity of the revolving sphere; nearest to earth is water, which is heavy relatively to air, light relatively to earth; next to water comes air, which is light relatively to water, heavy relatively to fire. The reason why earth is absolutely heavy is that it is farthest away from the circular movement, and therefore it is the fixed centre of the revolving body; the reason why fire is absolutely light is that it is nearest to the revolving sphere; the intermediate bodies are both heavy and light, because they are in the middle between the two extremes, i. e. the farthest point and the nearest. If there were not a revolving body, surely there would be neither heavy nor light by nature, and neither high nor low by nature, and this whether absolutely or relatively; and the bodies would not differ by nature in the way in which, for instance, earth moves by nature to its specific place and fire moves by nature to another place, and equally so the intermediary bodies. And the world is only finite, because of the spherical body, and this because of the essential and natural finiteness of the spherical body, as one single plane circumscribes it.’ Rectilinear bodies are not essentially finite, as they allow of an increase and decrease; they are only finite because they are in the middle of a body that admits neither increase nor decrease, and is therefore essentially finite. And, therefore, the body circumscribing the world cannot but be spherical, as otherwise the bodies would either have to end in other bodies, and we should have an infinite regress, or they would end in empty space, and the impossibility of both suppositions has been demonstrated. He who understands this knows that every possible world imaginable can only consist of these bodies, and that bodies have to be either circular-and then they are neither heavy nor light-or rectilinear-and then they are either heavy or light, i. e. either fire or earth or the intermediate bodies; that these bodies have to be either revolving, or surrounded by a revolving periphery, for each body either moves from, towards, or round the centre; that by the movements of the heavenly bodies to the right and to the left all bodies are constituted and all that is produced from opposites is generated; and that through these movements the individuals of these four bodies never cease being in a continual production and corruption. Indeed, if a single one of these movements should cease, the order and proportion of this universe would disappear, for it is clear that this order must necessarily depend on the actual number of these movements-for if this were smaller or greater, either the order would be disturbed, or there would be another order-and that the number of these movements is as it is, either through its necessity for the existence of this sublunary world, or because it is the best .

Do not ask here for a proof for all this, but if you are interested in science, look for its proof, where you can find it. Here, however, listen to theories which are more convincing than those of the theologians and which, even if they do not bring you complete proof, will give your mind an inclination to lead you to proof through scientific speculation. You should imagine that each heavenly sphere is a living being, in so far as it possesses a body of a definite measure and shape and moves itself in definite directions, not at random. Anything of this nature is necessarily a living being; i. e. when we see a body of a definite quality and quantity move itself in space, in a definite direction, not at random, through its own power, not through an exterior cause, and move in opposite directions at the same time, we are absolutely sure that it is a living being, and we said only ‘not through an exterior cause’ because iron moves towards a magnet when the magnet is brought to it from the outside-and besides, iron moves to a magnet from any direction whatever., The heavenly bodies, therefore, possess places which are poles by nature, and these bodies cannot have their poles in other places, just as earthly animals have particular organs in particular parts of their bodies for particular actions, and cannot have them in other places, e. g. the organs of locomotion, which are located in definite parts. The poles represent the organs of locomotion in animals of spherical form, and the only difference in this respect between spherical and non-spherical animals is that in the latter these organs differ in both shape and power, whereas in the former they only differ in power. For this reason it has been thought on first sight that they do not differ at all, and that the poles could be in any two points on the sphere. And just as it would be ridiculous to say that a certain movement in a certain species of earthly animal could be in any part whatever of its body, or in that part where it is in another species, because this movement has been localized in each species in the place where it conforms most to its nature, or in the only place where this animal can perform the movement, so it stands with the differentiation in the heavenly bodies for the place of their poles. For the heavenly bodies are not one species and numerically many, but they form a plurality in species, like the plurality of different individuals of animals where there is only one individual in the species.

Exactly the same answer can be given to the question why the heavens move in different directions: that, because they are animals, they must move in definite directions, like right and left, before and behind, which are directions determined by the movements of animals, and the only difference between the movements of earthly animals and those of heavenly bodies is that in the different animals these movements are different in shape and in power, whereas in the heavenly animals they only differ in power. And it is for this reason that Aristotle thinks that heaven possesses the directions of right and left, before and behind, high and low. The diversity of the heavenly bodies in the direction of their movements rests on their diversity of species, and the fact that this difference in the directions of their movements forms the specific differentia of their species is something proper to them. Imagine the first heaven as one identical animal whose nature obliges it-either by necessity or because it is for the best-to move with all its parts in one movement from east to west. The other spheres are obliged by their nature to have the opposite movement. The direction which the body of the universe is compelled to follow through its nature is the best one, because its body is the best of bodies and the best among the moving bodies must also have the best direction. All this is explained here in this tentative way, but is proved apodictically in its proper place. This is also the manifest sense of the Divine Words, ‘There is no changing the words of God’, and ‘There is no altering the creation of God’. If you want to be an educated man, proceeding by proof, you should look for the proof of this in its proper place.

Now if you have understood all this, it will not be difficult for you to see the faults in Ghazali’s arguments here about the equivalence of the two opposite movements in relation to each heavenly body and to the sublunary world. On first thoughts it might be imagined that the movement from east to west might also belong to other spheres besides the first, and that the first sphere might equally well move from west to east. You might as well say that the crab could be imagined as having the same direction of movement as man. But, as a matter of fact, such a thought will not occur to you about men and crabs, because of their difference in shape, whereas it might occur to you about the heavenly spheres, since they agree in shape. He who contemplates a product of art does not perceive its wisdom if he does not perceive the wisdom of the intention embodied in it, and the effect intended. And if he does not understand its wisdom, he may well imagine that this object might have any form, any quantity, any configuration of its parts, and any composition whatever. This is the case with the theologians in regard to the body of the heavens, but all such opinions are superficial. He who has such beliefs about products of art understands neither the work nor the artist, and this holds also in respect of the works of God’s creation. Understand this principle, and do not judge the works of God’s creation hastily and superficially-so that you may not become one of those about whom the Qur’an says: ‘Say, shall we inform you of those who lose most by their works, those who erred in their endeavour after the life of this world and who think they are doing good deeds?’ May God make us perspicacious and lift from us the veils of ignorance; indeed He is the bounteous, the generous! To contemplate the various actions of the heavenly bodies is like contemplating the kingdom of heaven, which Abraham contemplated, according to the words of the Qur’an: ‘Thus did we show Abraham the kingdom of heaven and of the earth, that he should be of those who are sure.’ And let us now relate Ghazali’s argument about the movements.

Ghazali says:

The second point in this argument concerns the special direction of the movement of the spheres which move partially from east to west, partially in the opposite direction, whereas the equivalence of the directions in relation to their cause is exactly the same as the equivalence of the times. If it is said: If the universe revolved in only one direction, there would never be a difference in the configuration of the stars, and such relations of the stars as their being in trine, in sextile, and in conjunction would, never arise, but the universe would remain in one unique position without any change; the difference of these relations, however, is the principle of all production in the world-we answer: Our argument does not concern the difference in direction of movement; no, we concede that the highest sphere moves from east to west and the spheres beneath it in the opposite direction, but everything that happens in this way would happen equally if the reverse took place, i. e. if the highest sphere moved from west to east and the lower spheres in the opposite direction. For all the same differences in configuration would arise just as well. Granted that these movements are circular and in opposite directions, both directions are equivalent; why then is the one distinguished from the other, which is similar to it?’ If it is said: as the two directions are opposed and contrary, how can they be similar?-we answer: this is like saying ‘since before and after are opposed in the existing world, how could it be claimed that they are equivalent?’ Still, it is asserted by you philosophers that the equivalence of times, so far as the possibility of their realization and any purpose one might imagine in their realization is concerned, is an evident fact. Now, we regard it as equally evident that spaces, positions, situations, and directions are equivalent so far as concerns their receiving movement and any purpose that might be connected with it. If therefore the philosophers are allowed to claim that notwithstanding this equivalence they are different, their opponents are fully justified in claiming the same in regard to the times.

I say:

From what I have said previously, the speciousness of this argument and the way in which it has to be answered will not be obscure to you. All this is the work of one who does not understand the exalted natures of the heavenly bodies and their acts of wisdom for the sake of which they have been created, and who compares God’s knowledge with the knowledge of ignorant man.

Ghazali says:

If it is said: as the two directions are opposed and contrary, how can they be similar?-we answer: this is like saying ‘since before and after in the existing world are opposed, how could it be claimed that they are equivalent?’ Still, it is asserted by you philosophers that the equivalence of times so far as the possibility of their realization, and any purpose one might imagine in their realization is concerned, is an evident fact. Now, we regard it as equally evident that spaces, positions, situations, and directions are equivalent so far as concerns their receiving the movement and any purpose that might be connected with it.

I say:

The falsehood of this is self-evident. Even if one should admit that the possibilities of man’s existence and non-existence are equivalent in the matter out of which he has been created, and that this is a proof for the existence of a determining principle which prefers his existence to his non-existence, still it cannot be imagined that the possibilities of seeing and not seeing are equivalent in the eye. Thus no one can claim that the opposite directions are equivalent, although he may claim that the substratum for both is indifferent, and that therefore out of both directions similar actions result. And the same holds good for before and after: they are not equivalent, in so far as this event is earlier and that event later; they can only be claimed to be equivalent so far as their possibility of existence is concerned. But the whole assumption is wrong: for essential opposites also need essentially opposite substrata and a unique substratum giving rise to opposite acts at one and the same time is an impossibility. The philosophers do not believe that the possibilities of a thing’s existence and of its non-existence are equivalent at one and the same time; no, the time of the possibility of its existence is different from the time of the possibility of its non-existence, time for them is the condition for the production of what is produced, and for the corruption of what perishes. If the time for the possibility of the existence of a thing and the time for the possibility of its non-existence were the same, that is to say in its proximate matter, its existence would be vitiated, because of the possibility of its non-existence, and the possibility of its existence and of its non-existence would be dependent only on the agent, not on the substratum.

Thus he who tries to prove the existence of an agent in this way gives only persuasive, dialectical arguments, not apodictic proof. It is believed that Farabi and Avicenna followed this line to establish that every act must have an agent, but it is not a proof of the ancient philosophers, and both of them merely took it over from the theologians of our religion. In relation, however, to the temporal creation of the world-for him who believes in it-before and after cannot even be imagined, for before and after in time can only be imagined in relation to the present moment, and as, according to the theologians, there was before the creation of the world no time, how could there be imagined something preceding the moment when the world was created? A definite moment cannot be assigned for the creation of the world, for either time did not exist before it, or there was an infinite time, and in neither case could a definite time be fixed to which the Divine could attach itself. Therefore it would be more suitable to call this book ‘Incoherence’ without qualification rather than ‘The Incoherence of the Philosophers’, for the only profit it gives the reader is to make him incoherent.

Ghazali says:

If, therefore, the philosophers are allowed to claim that, notwithstanding this equivalence, they are different, their opponents are fully justified in claiming the same in regard to times.

I say:

He wants to say: If the philosophers are justified in claiming a difference in the direction of movement, the theologians have the right to assert a difference in times, notwithstanding their belief in their equivalence. This is only a verbal argument, and does not refer to the facts themselves, even if one admits an analogy between the opposite directions and the different times, but this is often objected to, because there is no analogy between this difference in times and directions. Our adversary, however, is forced to admit that there is an analogy between them, because they are both claimed to be different, and both to be equivalent! These, therefore, are one and all only dialectical arguments.

Ghazali says:

The second objection against the basis of their argument is that the philosophers are told: ‘You regard the creation of a temporal being by an eternal as impossible, but you have to acknowledge it too, for there are new events happening in the world and they have causes. It is absurd to think that these events lead to other events ad infinitum, and no intelligent person can believe such a thing. If such a thing were possible, you need not acknowledge a creator and establish a necessary being on whom possible existences depend. If, however, there is a limit for those events in which their sequence ends, this limit will be the eternal and then indubitably you too acknowledge the principle that a temporal can proceed from an eternal being.’

I say:

If the philosophers had introduced the eternal being into reality from the side of the temporal by this kind of argument, i. e. if they had admitted that the temporal, in so far as temporal, proceeds from an eternal being, there would be no possibility of their avoiding the difficulty in this problem. But you must understand that the philosophers permit the existence of a temporal which comes out of a temporal being ad infinitum in an accidental way, when this is repeated in a limited and finite matter-when, for instance, the corruption of one of two things becomes the necessary condition for the existence of the other. For instance, according to the philosophers it is necessary that man should be produced from man on condition that the anterior man perishes so as to become the matter for the production of a third. For instance, we must imagine two men of whom the first produces the second from the matter of a man who perishes; when the second becomes a man himself, the first perishes, then the second man produces a third man out of the matter of the first, and then the second perishes and the third produces out of his matter a fourth, and so we can imagine in two matters an activity continuing ad infinitum, without any impossibility arising. And this happens as long as the agent lasts, for if this agent has neither beginning nor end for his existence, the activity has neither beginning nor end for its existence, as it has been explained before. And in the same way you may imagine this happening in them in the past: When a man exists, there must before him have been a man who produced him and a man who perished, and before this second man a man who produced him and a man who perished, for everything that is produced in this way is, when it depends on an eternal agent, of a circular nature in which no actual totality can be reached. If, on the other hand, a man were produced from another man out of infinite matters, or there were an infinite addition of them, there would be an impossibility, for then there could arise an infinite matter and there could be an infinite whole. For if a finite whole existed to which things were added ad infinitum without any corruption taking place in it, an infinite whole could come into existence, as Aristotle proved in his Physics. For this reason the ancients introduce an eternal absolutely unchanging being, having in mind not temporal beings, proceeding from him in so far as they are temporal, but beings proceeding from him as being eternal generically, and they hold that this infinite series is the necessary consequence of an eternal agent, for the temporal needs for its own existence only a temporal cause. Now there are two reasons why the ancients introduce the existence of an eternal numerically unique being which does not suffer any change. The first is that they discovered that this revolving being is eternal, for they discovered that the present individual is produced through the corruption of its predecessor and that the corruption of this previous individual implies the production of the one that follows it, and that it is necessary that this everlasting change should proceed from an eternal mover and an eternal moved body, which does not change in its substance, but which changes only in place so far as concerns its parts, and approaches certain of the transitory things and recedes from certain of them, and this is the cause of the corruption of one half of them and the production of the other half. And this heavenly body is the being that changes in place only, not in any of the other kinds of change, and is through its temporal activities the cause of all things temporal; and because of the continuity of its activities which have neither beginning nor end, it proceeds from a cause which has neither beginning nor end. The second reason why they introduce an eternal being absolutely without body and matter is that they found that all the kinds of movement depend on spatial movement, and that spatial movement depends on a being moved essentially by a prime mover, absolutely unmoved, both essentially and accidentally, for otherwise there would exist at the same time an infinite number of moved movers, and this is impossible. And it is necessary that this first mover should be eternal, or else it would not be the first. Every movement, therefore, depends on this mover and its setting in motion essentially, not accidentally. And this mover exists simultaneously with each thing moved, at the time of its motion, for a mover existing before the thing moved-such as a man producing a man-sets only in motion accidentally, not essentially; but the mover who is the condition of man’s existence from the beginning of his production till its end, or rather from the beginning of his existence till its end, is the prime mover. And likewise his existence is the condition for the existence of all beings and the preservation- of heaven and earth and all that is between them. All this is not proved here apodictically, but only in the way we follow here and which is in any case more plausible for an impartial reader than the arguments of our opponents.

If this is clear to you, you certainly are in no need of the subterfuge by which Ghazali in his argument against the philosophers tries to conciliate them with their adversaries in this matter; indeed these artifices will not do, for if you have not understood how the philosophers introduce an eternal being into reality, you have not understood how they settle the difficulty of the rise of the temporal out of the eternal; they do that, as we said, either through the medium of a being eternal in its essence but generable and corruptible in its particular movements, not, however, in its universal circular movement, or through the medium of what is generically eternal-i. e. has neither beginning nor end-in its acts.

Ghazali answers in the name of the philosophers:

The philosophers may say, ‘we do not consider it impossible that any temporal being, whatever it may be, should proceed from an eternal being, but we regard it as impossible that the first temporal should proceed from the eternal, as the mode of its procession does not differ from that which precedes it, either in a greater inclination towards existence or through the presence of some particular time, or through an instrument, condition, nature, accident, or any cause whatever which might produce a new mode. If this therefore is not the first temporal, it will be possible that it should proceed from the eternal, when another thing proceeds from it, because of the disposition of the receiving substratum, or because the time was propitious or for any other reason.

Having given this reply on the part of the philosophers, Ghazali answers it:

This question about the actualization of the disposition, whether of the time and of any new condition which arises in it, still holds good, and we must either come to an infinite regress or arrive at an eternal being out of which a first temporal being proceeds.

I say:

This question is the same question all over again as he asked the philosophers first,’ and this is the same kind of conclusion as he made them draw then, namely that a temporal proceeds from an eternal, and having given as their answer something which does not correspond with the question, i. e. that it is possible that a temporal being should proceed from the Eternal without there being a first temporal being, he turns the same question against them again. The correct answer to this question was given above: the temporal proceeds from the First Eternal, not in so far as it is temporal but in so far as it is eternal, i. e. through being eternal generically, though temporal in its parts. For according to the philosophers an eternal being out of which a temporal being proceeds essentially’ is not the First Eternal, but its acts, according to them, depend on the First Eternal; i. e. the actualization of the condition for activity of the eternal, which is not the First Eternal, depends on the First Eternal in the same way as the temporal products depend on the First Eternal and this is a dependence based on the universal, not on individuals.

After this Ghazali introduces an answer of the philosophers, in one of the forms in which this theory can be represented, which amounts to this: A temporal being proceeding from an eternal can only be represented by means of a circular movement which resembles the eternal by not having beginning or end and which resembles the temporal in so far as each part of it is transient, so that this movement through the generation of its parts is the principle of temporal things, and through the eternity of its totality the activity of the eternal.

Then Ghazali argues against this view, according to which in the opinion of the philosophers the temporal proceeds from the First Eternal, and says to them:

Is this circular movement temporal or eternal? If it is eternal, how does it become the principle for temporal things? And if it is temporal, it will need another temporal being and we shall have an infinite regress. And when you say that it partially resembles the eternal, partially the temporal, for it resembles the eternal in so far as it is permanent and the temporal in so far as it arises anew, we answer: Is it the principle of temporal things, because of its permanence, or because of its arising anew? In the former case, how can a temporal proceed from something because of its permanence? And in the latter case, what arises anew will need a cause for its arising anew, and we have an infinite regress.

I say:

This argument is sophistical. The temporal does not proceed from it in so far as it is eternal, but in so far as it is temporal; it does not need, however, for its arising anew a cause arising anew, for its arising anew is not a new fact, but is an eternal act, i. e. an act without

o beginning or end. Therefore its agent must be an eternal agent, for an eternal act has an eternal agent, and a temporal act a temporal agent. Only through the eternal element in it can it be understood that movement has neither beginning nor end, and this is meant by its permanence, for movement itself is not permanent, but changing.

And since Ghazali knew this, he said:

In order to elude this consequence the philosophers have a kind of artifice which we will expose briefly.