Tahafut al-Falasifa

(Incoherence of the Philosophers)


Abu Hamid al-Ghazali (1058-1111 CE)

Translated into English from Urdu Translation by Sabih Ahmad Kamali


Problem VII
Refutation of their thesis that it is impossible that something should share a genus with God, being separated from Him by differentia; and that the intellectual division into genus and differentia is inapplicable to Him


THEY have agreed on this point, and based on it the contention that, since nothing shares with Him a generic factor, nothing can be different from Him by virtue of a differential one. Hence He cannot be defined. For definition consists of genus and differentia; and that wherein there is no composition cannot be defined, as this is a kind of composition.


And they assert

One's saying: "He is comparable to the first effect in respect of His being an existent; a substance, and a cause of others; whereas, obviously, He is distinguished from the first effect in respect of something else" does not show a cosharing in genus, but only in a common inseparable accident. As Logic shows, the difference between a genus and an inseparable accident has a basis in reality, even though the two may not differ in regard to inseparable­ness. For the essential genus is that common property which is given in answer to the question: What is it? It enters into the quiddity of the defined thin, and is a constituent of its being. For instance, man s being a living being enters into man's quiddity — i.e., animality: hence it is the genus. On the other hand, his being a born or a created thing is an inseparable relation which never leaves him, but which does not enter into his quiddity — even though it is an inseparable accident common to all men. The way in which these things are known in Logic is indisputable.

Existence never enters into the quiddity of things, but is related to quiddity — either as an inseparable accident which never leaves the thing (e.g., the existence of heaven), or as something which happens after it had not been (e.g., the existence of things which have a beginning in time). Therefore, cosharing in existence is not cosharing in genus.

As regards His sharing with other causes in the character of being the cause of other things, it is again the cosharing of a relation which, however inseparable, does not enter into quiddity, Neither principleness, nor existence, constitutes His being; but each is inseparably related to it after the constitution of being by the parts of quiddity. If either is coshared, it will be but the cosharing of an inseparable accident whose inseparableness is subordinate to being. It will not be the cosharing of genus. This is the reason why things can be defined only by reference to their constituents. If they are defined by reference to their inseparable accidents, it is only a description in order to mark something off, not to depict its reality. For instance, in the definition of a triangle it is not said that all its angles are equal to two right angles, although it is an accident inseparably related to every triangle. On the contrary, it is said that a triangle is a figure enclosed by three sides.

The same holds of the cosharing of the state of being a substance. What being a substance means is that something is an existent not-in-subject. An existent is not a genus. Nor by the relation of something negative — i.e., not-in-subject — does it become a constituting genus. Nay, even if the positive aspect is related to it — i.e., when it is called an existent in a subject — it will not become a genus per accidens. This is so, because he who knows substance through its definition (viz., that it is an existent not in subject), which serves the purpose of a description of it, does not know anything about its being, except that it may, or may not, be in a subject. When in order to describe substance we say that it is an existent not-in-subject, the meaning is that it is a reality which, when found, is found not-in-subject. We do not mean that actually it is an existent at the time of definition It follows that cosharing of it is not cosharing of genus. Only the cosharing of the constituents of quiddity is the cosharing of genus, which consequently requires distinction by virtue of differentia. And God has not quiddity other than necessary existence. Hence necessary existence is the real nature, the quiddity-in-itself which belongs to Him, and to no one else. And because necessity of existence belongs to Him alone and to no one else, no one can be cosharer with Him. Therefore, no one can be differentiated from Him by a specific difference. Therefore, He is indefinable.

(This much was necessary by way of the exposition of their doctrine. Now, we proceed to criticise it from two points of view: first that of a questioner, and the second that of one who refutes.)


In the questioning, it may be said:

This, then, is the summary of the doctrine. Now, how do you know the impossibility of this, in regard to God — so as to base upon it the denial of duality (cf. your position that the second necessary being must coshare with Him something, being differentiated in respect of another; and that he who has one thing to be shared and another not to be shared is composite, whereas the composite — as an attribute of God — is impossible) ?


Let us say:

Whence comes to you the knowledge that this kind of composition is impossible? There is no proof of it, except what you were reported to say in denial of the attributes — namely, that the composition of genus and differentia produces an aggregate of parts: that if the existence of a part or the whole is valid independently of the other, then it will be necessary of existence, as set over against the other; and that if the existence of the parts is not valid independently of the aggregate, nor that of the aggregate independently of the parts, then each will be an effect or a dependent.

We have considered this argument in connection with the attributes, and have shown — in connection with the question of the termination of the series of causes — that this is not impossible. What can be rationally proved is only the termination of regress. As regards those grand things which have been invented by them so as to form inseparable attributes of the necessary being, there is no argument for them. If necessary being means what they explain it to be — namely, that there is no composition in it, and that, therefore, it does not depend on anyone other than itself for its constitution — then there is no argument to prove necessary being. What there is an argument for is only the termina­tion of regress. With this (argument) we have already dealt in connection with the attributes.

Our contention is all the more valid in this section. For to divide something into genus and differentia is not like dividing a bearer of attributes into essence and attribute. An attribute is other than the essence, and the essence is other than the attribute. But the species is not other than the genus in all respects. When we speak of the species, we only mean the genus plus an additional factor. Thus, when we speak of Man, we mean Animal plus the additional factor of rationality. To ask whether man-ness can be independent of animality is like asking whether man­ness can be independent of itself, when something else is added to it. It is, therefore, further removed from plurality than an attribute and its bearer can be.

Why should it be impossible for the series of causes and effects to stop at two causes — one being the cause of the heavens, and the other that of the elements: or one the cause of the heavens, and the other that of all bodies? Why should it be impossible for the two causes conceptually to be separate and distinct — as, for instance, between redness and heat even in the same place there is conceptual distinction without our supposing any genus — differentia composition in redness so as to make it analysable? Nay, if there is plurality in it, this sort of plurality will not be repugnant to the unity of essence. So what is the reason why this should be impossible in the case of causes? And this shows how they fail to defend their denial of the possibility of two gods or creators.


If it is said:

This is impossible, because that wherein the two beings differ must be present in each necessary being — if it is a condition for the necessity of existence. But then there will be no difference between the two. On the contrary, if it is not a condition, then, insofar as necessary existence can do without all that is no condition for the necessity of existence, necessity of existence will be completed without it.


we will answer:

This is the same as what you had to say in connection with the attributes. We have already dealt with it. In all this the source of confusion is the term ' necessary being.' This term should, therefore, be discarded. We do not concede that rational arguments prove the necessary being — it the term does not mean an eternal existent who has no efficient cause. But if this meaning is adopted, then the term 'necessary being' should be dismissed, and you must prove that in an existent which has no cause or agent, number or composition is impossible. But that just cannot be proved.

There remains their question whether the necessary being's uncausedness is conditioned by that which is supposed to be common to the two necessary beings. This is silly, for we have shown that something which is uncaused is not caused to be so, wherefore its condition should be sought out. It is like one's asking whether blackness is a condition for colour's being colour: and, if it were, why should redness be colour. To this question, the answer will be: As regards its reality — viz., the reality of colouredness as realised in the Intellect — neither blackness nor redness is a condition. But as far as its existence is concerned, each one may be a condition, although not the only one. That is, no genus can possibly exist without some difference. Similar­ly, therefore, he who affirms two causes, thereby putting an end to the causal series, might say: They are distinguish­able from each other by differences one of which is of neces­sity a condition for existence — not in an exhaustive way.


If it is said

This is possible in the case of colour, for colour has an existence related to quiddity or additional to it. But it is not applicable to the necessary being; for such being has nothing but the necessity of existence. There is no quiddity to which existence could be related. As the differentia of blackness or redness is no condition for colouredness, qua colouredness, but only a condition for the existence of colouredness which is caused, similarly it is proper that the differentia of the necessary being should not be a condition for the necessity of existence. For necessary existence is to the Necessary Being what colouredness is to colour — unlike the existence of a coloured thing which is related to colouredness.


we will say:

This cannot be allowed. The necessary being must have a reality to which existence is attributed. The point will be explained in the next problem. The philosophers' assertion that the necessary being has no quiddity but existence is unintelligible. The upshot of the whole discussion is that they base the denial of quality on that of genus-differentia composition. And this they base on the denial of quiddity as set over against existence. So when we have destroyed the last one, which is the foundation of foundation, the whole structure will collapse. For verily, it is a shaky structure like the spider's web.


The Second Method: Objection

We say: Even if existence, substantiality, or being a principle is no genus (for none is given in answer to the question: What is it?), still you consider God to be a pure intelligence, as all other intelligences (who are the secondary principles of existence, and whom the philosophers also call Angels — i.e., the effects of the First Cause) are pure intelligences divested of Matter. So this reality will include God and His first effect. For the first effect, too, is simple, having no composition in it, except by way of its inseparable accidents. The two will, therefore, stand on par, inasmuch as each is intelligence divested of Matter. And this is a generic reality, for being pure intelligence is not one of the inseparable accidents of being, but the very quiddity. So this quiddity will be common to God and all the intelligences. Now, if God is not distinguished from the intelligences by something else, you will have conceived a duality without mutual distinction. But if He is distinguishable, then that which causes distinction must be other than that which gives God and the intelligences their common character of being intelligent. And the cosharing of this character will be the cosharing of the generic reality. For, according to those who believe in God's knowledge of what is other than Himself, God has self-knowledge and the knowledge of the Other­by virtue of His being intelligence divested of Matter. And the first effect — viz., the first intelligence, whom God has already caused forth — does share this character with God. The argument for this is that the intelligences, which are the primary effects, are different species sharing the charac­ter of being intelligences, but separated from each other by certain differentiae. And God will share with all of them the character of being an intelligence. So here the philosophers have — two alternatives: either the rule laid down by them will be broken, or they should revert to the position that being an intelligence does not constitute the essence of God. And for them each alternative is impossible.


Table of Contents