hence his rejection of the third point. Interesting as these topics are, we shall not treat them at length; rather we shall focus on one of al-Ghazzālī's most important philosophical theses discussed in Tehāfut al-Falāsifah, i.e., his views on the theory of causation. Specifically, we shall discuss his refutation of the concept of a necessary causal nexus and note the striking parallel of his views to those of the British philosopher David Hume (1711-1776).

In the last part of Tahāfut al-Falāsifah, al-Ghazzālī embarked on the task, interalia, to refute those philosophers who advocated the concept of a necessary causal nexus in natural events. His discussion of causality commences with the following:

Customarily the association (connection) between that which is believed to be a cause and that which is believed to be an effect is not necessary for us. But consider any two things whereby neither of which is the other. And the affirmation of one of them does not guarantee the other, and its denial does not guarantee the denial of the other. And thus the existence of one does not necessarily depend on the existence of the other; neither on the non-existence of the other. Take any two things such as the quenching of thirst and drinking, satiety and eating, burning and contact with fire, light and the appearance of the sun, death and decapitation, cure and medicine, evacuation and the use of a laxative, and so forth for all the empirical connections in medicine, or astronomy, or arts or crafts.[14]

This passage clearly reveals al-Ghazzālī's basic point that when two things invariably follow each other, nothing can prove or demonstrate that one is the cause of the other. In other words, a necessary causal connection is neither logically nor empirically demonstrated. So far it would seem that al-Ghazzālī is as much an empiricist as the British philosopher, Hume. To wit, compare the above passage with the following:

They are still more frivolous, who say, that every effect must have a cause, because 'tis imply'd in the very idea of effect. Every effect necessarily pre-supposes a cause; effect being a relative term, of which cause is the correlative. But this does not prove, that every being must be preceded by a cause; no more than it follows, because every husband must have a wife, and therefore every man must be marry'd. The true state of the question is, whether every object, which begins to exist, must owe its existence to a cause; and this I assert neither to be intuitively nor demonstratively certain...[15]

Just as al-Ghazzālī asserted that co-existence does not indicate causation, so also did Hume. Because two events invariably follow each other, there is no necessary justification from experience for claiming that one is or is not the cause of the other. Let us consider one of al-Ghazzālī's examples in order to further reveal the similarity between him and Hume. Contact with fire burns cotton. al-Ghazzālī admitted that it is quite conceivable that the contact might occur without the burning or that the cotton might be changed into ashes without coming into contact with fire.[16] What he inferred from such an example was that sense observation does not disclose that fire necessarily causes the burning of cotton. As he put it:

They [philosophers] have no other proof than observation of the occurrence of burning when there is contact with fire. And that observation proves only that one is with the other, but does not prove that it is by it and that it has no other cause than it.[17]

So far al-Ghazzālī and Hume are in fundamental agreement concerning their denial of necessity to empirical events. To both the concept of a necessary causal nexus is not given in experience. all that we actually perceive when we look at the external world are events followed by other events with no necessary connection between them. The necessity that the concept of causality implies is nowhere to be demonstrated by experience. However, it is with respect to the explanation of how one arrives at the idea of "necessary connection" that both adopted different views. Al-Ghazzālī contended that events are necessarily connected in nature because of God's will. As he explained it:

The connection of these things [i.e., empirical events] is a result of God's power, which preceded their existence. If one follows another, it is because God created them in that manner, and not because this connection is necessary in itself and cannot be disjoined. He has the power to create satiety without hunger, and death without decapitation, to prolong life after decapitation, and so on with respect to all concomitant things.[18]

Al-Ghazzālī's contention with respect to why things in nature arc connected differed from Hume's explanation. To Hume, the belief in causality was found to be a matter of custom, habit, instinct, expectation. Specifically, all sequences do invariably occur, but these sequences, if repeated again and again, will produce in the mind a strong disposition to connect the events that form the sequence. Thus Hume's answer boiled down to the assertion that necessary connection is no more than psychological, i.e., the way in which we think. As he put it, "Upon the whole, necessary is something that exists in the mind, not in objects: nor is it possible for us even to form the most distant idea of it, considered as a quality in bodies."[19]

With respect to al-Ghazzāli's analysis of causation there is first the empiricist element which, as I attempted to indicate, closely parallels that of Hume. Second, there is the metaphysical element which al-Ghazzālī on the one hand invokes and Hume on the other rejects.

Further examination of Tah āfut seems to indicate al-Ghazzāli's closeness to Hume. After al-Ghazzālī denied that there is any necessary connection between cause and effect, he entertained the possibility of whether a book could turn into an animal, a slave boy into a dog or a stone into gold[20]. If we denied the necessary dependence of effects on causes and maintained that an effect is to be attributed to the will of its creator and, furthermore, that that will has no definite pattern, but that it may vary and be quite arbitrary, then one would be led to entertain the possibility of the above examples. al-Ghazzālī's answer to that was:

...God has created in us the knowledge that he will not do all these possible things ... but they are possible in that they may or may not happen, and the persistence of habit firmly fixes their occurrence in our minds in accordance with a previous habit in a fixed impression.[21]

This passage is quite revealing in its statemeent that God does not permit, for example, a book to be turned into an animal, or a slave boy into a dog. We do not expect such phenomena to occur, because we are not accustomed to seeing such things. Clearly there are over-toes of Hume's analysis here too. As human beings, we are creatures of habit and tend to depend upon past events to justify future events. Yet, whereas al-Ghazzālī speaks about the orderly course of nature as being the habit of the divine will,[22] Hume believes that the justification of the orderliness of nature is psycho-logical, the way we think, and therefore has nothing to do with God. Thus al-Ghazzālī's position is again similar to Hume's except that he introduced God as part of his argument.[23]

Tahāfut has proved of permaent interest and influence, not just in the Islāmic world but also in the Medieval Latin West. The following points are of importance:

1. It revived the religious atmosphere by saving orthodox Islām. al-Ghazzālī's refutation of the philosophers' views were for the most part based on religious explanations, believing, as he did, that philosophic theory cannot be the basis of religious beliefs; revelation alone can discover the essentials of truth. Thus, in opposing the positions of the philosophers, he was able to sub-ordinate philosophy to theology.

2.  Tahāfut was one of the first attempts to separate Islāmic philosophy from Greek philosophy. Indeed, al-Ghazzālī was one of the first Muslim philosophers to fundamentally deviate from Greek philosophy.

3.  Al-Ghazzālī's attack on the concept of the causal necessary nexus was classical. He was the first thinker to embark on a systematic refutation of the notion of necessary connection in nature. It is quite conceivable, however, that al-Ghazzālī, in his attack on the concept of the causal necessary nexus, could have been influenced by the Greek skeptics, especially those of the Pyrrhonian school).[24]

4.   al-Ghazzālī's arguments in support of the theory of creation exnihilo, God's cognizance of particulars, and the resurrection of the bodies were widely received in the Islāmic world. When Tahāfut was translated into Latin, it was adopted by many scholastic philosophers.[25]

III. AL-GHAZZĀLĪ'S PLACE IN LOGIC

Now, we shall turn to al-Ghazzālī's significance in the development of logic in Medieval Islām. We indicated earlier that he had a keen interest in logic. although he rejected most of Ibn Sīnā's metaphysical and physical theses, al-Ghazzālī  was a staunch follower of Ibn Sins in the field of logic. al-Ghazzālī desire to accept logic stemmed primarily from the lack of any doctrinal content in it that might conflict with orthodox Islām. This indeed is the criterion that he adopted in his intellectual life. In other words, if the contents of a particular subject matter were not in conflict with religion, it was useful and hence should be utilized by the community. On the other hand, if the contents were in conflict with religion, then it was harmful and hence should be avoided by the community. It is in this sense then that logic passed al-Ghazzālī scrutiny.

Because he found logic theologically useful and acceptable, al-Ghazzālī encouraged its study. arguing that it was doctrinally neutral, that it was merely "a tool of knowledge," he wrote logical treatises for his fellow theologians, urging them to master the tools of logic in order to be better equipped in rebutting the opponents' doctrines. The most comprehensive logic book which al-Ghazzālī appended to Tahāfut was Meyār al-'Ilm. One of the motives for writing this book was, as he tells us, to explain the technical, logical vocabulary of the philosophers to his fellow theologians, thus enabling them to have a better understanding of the arguments delineated in Tahāfut. In Mihākk al-Nazar, al-Ghazzālī offered a shorter version of Mi`yār, still hoping that logic would be utilized by his fellow theologians. In AI-Qusṭās al-Mustaqīm, he applied syllogistic reasoning to theological questions.

It should be noted in this connection that all of al-Ghazzālī's treatises on logic faithfully followed the ideas of his predecessors, particularly Ibn Sīnā. although al-Ghazzālī was not original in his treatment of logic, he was still credited as being the first Islamic Medieval philosopher who encouraged and utilized logic in discussion of theological questions. Indeed the link between theology and logic proved to be a decisive factor in the survival of logical studies in Medieval Islām, for logic came to be accepted by theologians and others as an important tool in the shaping of the Islāmic basic curriculum. The following structure was developed: (a) Qur'ān, (b) arabic Language, (c) Logic, (d) Theology and Religious Law.[26] The slogan "man tamantaqa tazandaga" (he who uses logic commits heresy) was no longer seriously entertained in the Islāmic community. Once again "Persia was to become the undisputed center of logical studies in the Arabic speaking world.[27]

although the detachment of logic from philosophy proved to be of significance, it did not advance the development of logic. as Professor Nicholas Rescher so aptly stated:

In making its peace with theology, as it was ultimately to succeed in doing, Arabic logic did not get off without paying a price. By breaking away from philosophy and becoming an accepted part of advanced religious instruction, the study of logic became sterile and stylized — a matter of memorizing handbooks rather than mastering a living discipline.[28]

Professor Rescher's comment is historically correct. But be that as it may, the logical treatises that the Muslim philosophers wrote, especially those of al-Fārābī, Ibn Sīnā and al-Ghazzālī, were of profound historical significance in the development of logic. Though Islamic logic was Aristotelian in its general outline, a closer examination of the logical treatises of the Muslim philosophers reveals that it was wider in scope and subject matter. No one is certain about the actual influences on Islāmic logic, but it may be safe to suggest that logic as propounded by Muslim logicians tended for the most part toward assimilation of Aistotle's Organon which was available to the Muslim scholars through the expertise of various commentators and the works of the Stoics. If this claim is correct, one may consider Muslim logicians as synthesizers of aristotle's logic as well as that of the Stoics. In order to substantiate this claim, let us consider how the two significant Aabic terms taṣawwur (concept) and (taṣdīq (assent) as well as hamliyyah (categorical, attributive, non-conditional) and sharṭiyyah (conditional) propositions were utilized by Muslim logicians.

(I) Concept and Assent

The origin of these two terms is not known; yet every Islāmic logic textbook of the Medieval period begins by stating that all knowledge is either of concept or assent (judgment). The concept, roughly speaking, is the realization of essence without judging it either affirmatively or negatively; and the way of apprehending it is by means of a definition (ḥadd). Assent (judgment) is the concept which is judged; that is to say, it is the ascription (predication) of one thing of another affirmatively or negatively; and the way of apprehending it is by means of a demonstration (burhān).[29]

The above indicates the bifurcation of logic into two parts: one dealing with concepts, the other with assent (judgment). It would indeed be historically helpful to locate the underlying reasons for this division. One explanation would be to claim that the desire on the part of Islāmic logicians to accommodate Aristotle's logic as well as the works of the Stoic logicians led them to such a division. If one were to claim that the part of logic dealing with concept is equivalent to the logic of terms and that of assent equivalent to the logic of propositions, and, furthermore, if one were to agree with Lukasiewicz's contention.[30] that Aristotelian logic was a logic of terms, while Stoic logic was a logic of propositions (judgments), then this would lend some credibility — not just to the view that is being suggested for the division of logic into concept and assent, but also to the view that Islāmic logicians tended to synthesize the best of Aristotle's logic and Stoic logic.

(II) Conditional and Non-Conditional Propositions

A non-conditional (categorical) proposition, according to Islāmic logicians, is one which affirms a predicate to a subject, or denies it of a subject. For instance: Zayd is a writer; Zayd is not a writer. In connection with the discussion of categorical propositions and syllogisms, Islāmic logicians did not substantially differ from the Aristotelian tradition, as evidenced in their discussions of the kinds of categorical propositions, the three figures, categorical syllogisms, etc. On the other hand, in their treatment of conditional propositions, Islāmic logicians deviated from the Aristotelian tradition. Conditional propositions consist of two parts, each part including a proposition. Starting with al-Fārābī and possibly earlier, Islāmic logicians discussed two kinds of conditional propositions: (a) hypothetical conjunctive (al-sharṭiyyah al-muttaṣilah), and (b) hypothetical disjunctive (al-sharṭiyyah al-munfaṣilah). The paradigm examples of (a) and (b) are respectively: "If the sun has risen, it is day." "Either the world is created or eternal." It should be pointed out that though conditional propositions were known to Aristotle, they were, however, never explored by him.[31] However, Islāmic logicians went beyond Aristotle's promisory note. One indeed finds detailed accounts of coditional statements in the writings of Islāmic Medieval philosophers.[32]

It is interesting to note in this connection that Islāmic logicians were quite familiar with valid as well as invalid conditional syllogisms. In Maqāṣid, al-Ghazzālī cited the following paradigm examples of hypothetical conjunctive arguments:[33]

A) I. If the world is created, then it has a creator.

 The world is created

2. Therefore, it has a creator

Letting "A" stand for "The world is created," and "B" for "The world has a creator," the argument form of (A) is:

 If A, then B

A     

\  B

B) I. If the world is created, then it has a creator.

   There is no creator

2. Therefore, the world is not created.

The argument form of (B) is:

If A, then B

Not B         

\      Not A

C)     1. If the world is created, then it has a creator 2.

 The world is not created

2. Therefore, there is no creator[34]

The argument form of (C) is:

If A, then B

Not A                  

\      Not B

D)     1. If the world is created, then it has a creator 2.

2. There is a creator                

  Therefore, the world is created[35]

The argument form of (D) is:

If A, then B

B      

\      A

Obviously (A) and (B) are valid arguments. They respectively correspond to Modus Ponens and Modus Tollens of the Latin logicians. (C) and (D) are invalid arguments. (C) commits the fallacy of denying the antecedent, while (D) commits the fallacy of affirming the consequent.

Islāmic logicians were also familiar with disjunctive syllogisms. In Maqāṣid, al-Ghazzālī reports the following as paradigm examples of disjunctive arguments:

The hypothetical [disjunctive] syllogism is when you say "Either the world is created or eternal." From this kind of hypothetical disjunctive [sentence], there results four exceptive syllogisms. For you say that it is created and hence not eternal; but it is not created and hence eternal; but it is eternal and hence not created; but it is not eternal and hence created.[36]

Letting "A" stand for "the world is created," and "B" for "the world is eternal," then the above passage contains the following argument forms respectively:

A)  Either A or B

A          

\ NotB

B)   Either A or B

Not A

\ B

C) Either A or B

B         

\ NotA

D)Either A or B

Not B

\ A

These four argument forms are all considered valid provided the logical operator "either...or" is understood in the exclusive sense. If the inclusive[37] sense is used, (A) and (B) would then be invalid argument forms.

The preceding discussion indicates al-Ghazzālī's familiarity with formal logic, Admittedly nothing is original in al-Ghazzālī's discussion of logic; he closely follows Ibn Sīnā. Al-Ghazzālī's aim, as mentioned earlier, was to render logic acceptable and palatable to his fellow theologians. Hence his significance in logic does not lie in introducing any new techniques to the development of logic. However, credit should go to al-Ghazzālī for his insistence that logic, since it does not conflict with the articles of faith, should be employed as a tool of inquiry.

IV. CONCLUSION

In conclusion, it may be mentioned that much needs to be done with respect to al-Ghazzālī's Maqāṣid. So far the book has not received the scholarly attention that it should. Much light could be shed on al-Ghazzālī's predecessors if this book were available in English. We have suggested that a worthwile study would consist in a textual comparison of views of al-Fārābī and Ibn Sīnā with those expounded by al-Ghazzālī in Maqāṣid in order to locate philosophical misinterpretations, if any, on the part of al-Ghazzālī. In other words, it behoves us to find out how accurate, philosophically speaking, al-Ghazzālī was in reporting the doctrines of the philosophers. In Tahāfut one encounters a brilliant systematic refutation of the views of Ibn Sīnā. The translation of the book into English by S. A. Kamālī has enabled the western scholars to study and appreciate the philosophic acumen embodied in the book. Finally, with respect to al-Ghazzālī's contribution to logic, much credit should go to him for the support which he gave to its study. This enabled his fellow theologians to use logic as a tool in their intellectual endeavors.

Notes and References