Alhassanain(p) Network for Heritage and Islamic Thought

CHAPTER III: PROBLEMS OF LOGIC

What is the object of logic, and what is its relation to philosophy? This had become the subject of some dispute among the Greeks of the post-classical period. Aristotle himself was not clear on the point, and had been inclined to consider logic as a creative art (téchne); he could not very well classify it as one of the theoretical or practical sciences. The Stoics after him contended that logic was actually a part of philosophy; while the Peripatetics maintained that it was merely an instrument of thought. Alexander of Aphrodisias, between the second and third century, was the first to call it an organon (instrument) of the sciences; and it is after him that the logical words of Aristotle became known as the Organon. The Platonists, taking a middle course, said that it was both a part of philosophy and an instrument of the sciences.

Both views are reflected in the conception of the Islamic philosophers, but not regarded as being of any great importance. The subject had been entirely new to them, and its methods and applications seemed almost revolutionary. The deductive method of reasoning from general premises which had now reached them, was seized upon with great enthusiasm and led them into fields as yet unexplored. They were therefore principally occupied with the use of logic in their reasoning, and did not worry overmuch about how to classify it. It had focused their attention on Aristotle as the owner of logic, though some Christian and Muslim theologians took strong exception to it. The Islamic philosophers became acquainted almost simultaneously with the Arabic renderings of the Aristotelian Organon and various commentaries by Peripatetic, Neo-Platonic and Stoic authors who had raised the question of the use and purpose of logic. They could not therefore avoid taking some part in the controversy, more especially since they had taken upon themselves the task of justifying the whole subject and defending it against its detractors. Kindi, of whose works not all have survived, seems silent on this matter; he speaks of the eight books which included thePoetica and the Rhetorica as the logicals (al-Mantiqiyyat). Farabi calls logic an art in his classification; and takes no part in the dispute, at least in any of his published writings. In theEpistles we find some reflection of the point at issue. There, probably under Stoic influence, logic is classified as one of the four species of “true philosophy”; and is also spoken of as “the scales of philosophy” and as “the tool of the philosopher” which conforms to the Peripatetic conception.

Avicenna is fully aware of the problem but avoids taking sides. He insists in the Shifathat the entire dispute is irrelevant, and that there is no contradiction between considering it a part of philosophy and an instrument of it. He adopts the term instrument which he knew came from Alexander, and refers to logic as the instrumental science. But having considered it a science in one place, he calls it an art in another; while in Persian following the Epistles, he names it “the science of the scales”. He thus follows Boethius, called the last of the Romans and the first of the Scholastics, who maintains that logic is both a science and an instrument of science.

Aristotle had never used the term logic in its modern sense; nor is it quite clear who it was that first gave it that sense. It has been contended that the credit must go to the Stoics, and we know that the term already occurs in Chrysippus. Cicero employs it, but only to mean dialectics. By the time of Alexander of Aphrodisias and Galen, it is in current use in the form of the Greek logiké. The Arabic term mantiq we find in the fragments of the translation of the Metaphysica being used more than once as the equivalent of the Greekdialectiké and also, in some passages, of logiké. The rendering is that of Ustath who, as has already been observed, was one of the early pre-Hunain translators. It may be thought, therefore, that he was the man who chose the word that he supposed had never had that connotation in the Arabic language, only to find that even before him Ibn al-Muqaffa had given it that same new sense in one of his literary works; and also in that short paraphrase of Aristotelian logic of which mention has already been made. Arab purists never approved of this neologism, and the subject of logic was never to the taste of the theologians whether Christian or Muslim. Cases are recorded where in their heated discussions with logicians, they poured ridicule on the choice of the word, even though linguistically it is perfectly justifed.

Kindi’s definition of logic has not come down to us in a clear form. Farabi says “the art of logic gives in general the principles whose purpose it is to help the intelligence forward, and to lead man to the path of correct [thought], and to the truth the relation of the art of logic to the intelligence and the intelligibles is as the relation of the art of syntax to language and words”. For Aristotle also logic was primarily a matter of right thinking and secondarily of correct speaking. The authors of the Epistles maintained that “the sciences of logic are of two kinds, linguistic and philosophical; the linguistic is such as the art of syntax . and the logic of judgments is of different branches, among which is the art of reasoning, and of dialectics, and of sophistics”. The logic of language, they thought, should be mastered before the logic of philosophy, for “it is incumbent upon him who desires to theorize in philosophical logic, to be first trained in the science of syntax”.

Avicenna’s definitions are numerous and somewhat varied. In one place he says, “logic is that science in which may be seen the state of knowing the unknown by the known; that which it is that is in truth, and that which it is that is near the truth, and that which it is that is false; and the different varieties of each”. In another place he states that logic '”is for the intelligence a guarding instrument against error, in what we conceive and give assent to; and it is that which leads to true belief by giving the reasons and methods of arriving at it”. In still another he remarks, “thus logic is a science from which is learnt the modes of passing from matters determined in the human thought, to matters to be determined; and the state of these matters, and the number and varieties wherein the order and the form of the transposition lead to correctness, and the varieties wherein it is otherwise”.

The logic of Avicenna has not yet been properly studied. Nor would the effort prove fruitful unless the logic of the Commentators of Aristotle had first been carefully examined. No such study of the original Greek has yet been made; for the purposes of the present inquiry it would be even more important to study the Arabic version, for only then could the contributions of Avicenna be placed in their historical setting, and their originality, if any, definitely determined. Even the most superficial acquaintance with Islamic logic reveals the fact that although Aristotelian in general outline, it goes much farther in scope and subject-matter. Many have suspected that the additions are derived from Stoic sources; but there were Peripatetic and Neo-Platonic influences as well, Furthermore, it remains to be seen whether such additions as are indisputably Stoic reached them directly or through the various commentators of whom there were so many in the Hellenistic age. One author makes mention of the “fourth figure” in syllogisms, which as has been shown, was not introduced by Galen, but by some unknown logician after the fifth century; and Avicenna, well aware of the Stoic attempt to reduce the Aristotelian categories, speaks of “those who took pains to make some of these enter into others, and to limit them to categories of fewer number; among them those who made the categories four”. In fact throughout the Shifa he differentiates between what he calls “the first teaching”, meaning the Aristotelian, and later teachings; and significantly adds that “philosophy, where it is according to the Peripatetics, and where according to the Stoics, is not to be referred to with absolute synonimity”

But by far the most conclusive evidence is in the field of terminology. The vocabulary ofAvicenna abounds in logical terms for which there are no equivalents in the translations of the Organon, and which correspond very well with such Stoic terms as have survived. Although our knowledge of Stoic logic is very limited, and all a priori attempts to equate Avicennian terms with those used by the Stoics are to be discouraged as dangerous, the correspondence is sometimes so close as to give some measure of certainty. Nevertheless, we have the testimony of Ibn Taimiya “that Avicenna and his followers dissented from the ancients in a number of their logical statements and in various other things”.

The Islamic Falasifa did know of Zeno and Chrysippus and also Diogenes, but it is difficult to say to what extent they were acquainted with their works. Farabi has frequent references to Zeno the great, and Zeno the small, as he calls them. In one source-book there is mention of a group who are associated with the science of Aristotle, and they are those who are called and known as the men of the shaded place, and they are the spirituals, which clearly points to the Stoics. Nevertheless, it is far more likely that Stoic logic reached Avicenna not directly but by way of Peripatetic and Neo-Platonic commentators. Among these were Galen, whose work on logic we know to have been translated into Arabic and widely read; Alexander, for whom Avicenna expresses much appreciation and who in his refutation of the Stoics had discussed much of their logic; Ammonius, the noted disciple of Proclus and the author of various commentaries on Aristotelian logic; Porphyry, whose commentary was almost a textbook in its Arabic rendering and was sometimes called by its Greek name of Eisagoge (Introduction) or by the Arabic equivalent of Al-Madkhal. This was considered a necessary introduction to logic and some supposed it actually a part of the Organon; and finally John Philoponus of Alexandria, commonly called the Grammarian. It is from these, besides the works of Aristotle, that Avicenna must have derived most of his knowledge of Greek logic.

Lukasiewicz was among the first to demonstrate that whereas the Aristotelian was a logic of classes, the Stoic was one of propositions. But towards the close of the Greek period in the history of logic, the two had already merged; and while the Arabs had the whole of theOrganon before them, and may have had a translation of the Stoic works, this particular amalgam of the two which developed in the late Hellenistic age influenced them greatly. With this in mind it may be claimed that the logic of Avicenna really combines the two, not by a mechanical super-imposition of one on the other, but via a critical assessment of the two doctrines, with a good measure of simplification and perfecting on his part. Simplification was desirable for one whose conception of the subject was practical: logic, as a tool for correct thinking, was to be made sharp and effective. In point of fact a distinctive feature of Avicenna’s entire philosophy is that he shows himself perfectly ready to accept, to discard, to modify and to augment without the least hesitation. Avicenna does not go as far as Russell in dismissing all the Aristotelian categories, and even the word “category”, as meaningless, but he does not mind stating that at least one of them means nothing to him; and on the other hand he asserts in his Physics that we need not necessarily postulate only ten genera of being, for other categories may be added, including one of motion. In the case of the hypothetical syllogism, which, as Alexander and John Philoponus testify, was first discussed by Theophrastus and Eudemus and later developed by the Stoics, Avicenna, ignoring the original sources, simplifies the matter almost out of all recognition.

Avicenna had discussed logic in some fifteen different works, but judging from what has survived, they differ somewhat in form and in content. In the Shifa, mistakenly translated by the Latins as Sufficientia, as well as in the abridged version called Najat (Deliverance), he may be considered more Aristotelian in approach and to some extent in subject-matter. In later books such as al-Isharat wa al-Tanbihat (The Directives and Remarks), in his PersianDanish-Nameh (Book of Knowledge), and in the fragment called Mantiq al-Mashriqiyyin(Logic of the Orientals), he is inclined to deviate from Aristotle. It should not be supposed that the deviation is very marked, but there is certainly an attempt to think over the problems independently. The Logic of the Orientals has become the subject of much controversy; both title and contents have been interpreted in various ways. The latest and the most plausible theory is that it formed part of a much larger book, which we know Avicenna had written, which was entitled al-Hikmat al-Mashriqiyya (The Philosophy of the Orientals) and in which he had expressed his own mature views towards the end of his life. It is contended that he called it Oriental so as to contrast it with the servile Aristotelianism of some Christian philosophers in Baghdad who were to him Occidentals. “We do not worry”, he says, “o show a departure from those philosophers enamored of the Peripatetics who imagine that God did not guide any except themselves”

This attitude is best expressed in what is supposed to be his last work on logic. “That we may put down some statements on what men of investigation have disagreed upon we do not worry about any departure that may appear on our part from what the expounders of the books of the Greeks have been occupied with, either out of oversight or lack of understanding it became easy for us to comprehend what they said, when we first took up that subject. And it is not improbable that certain sciences may have reached us from elsewhere than the side of the Greeks we then compared all these with that variety of science which the Greeks call logic - and it is not improbable that it may have a different name among the Orientals and because those who occupy themselves with science are extremely proud of the Peripatetics we disliked to dissent from and oppose the public and we overlooked what they were struggling with and if we venture to oppose them, it is in things in which we can no more show patience they consider that looking deep into matters is a heresy, and that opposing what is widely accepted is a departure from the right path and we did not compile this book except for ourselves, I mean for those who take the same position as ourselves. And as to the common people who engage in such things, we gave them in the book of the Shifa what is even too much for them and beyond their requirements”. This passage is provocative. What is the source other than the Greek from which, he says, certain sciences may have reached us; and what is the name the Orientals gave to logic different from that of the Greeks? Is he referring to Indian thought, or Middle Persian writings, or what had developed in his own part of the world? In spite of innumerable theories, no satisfactory answer has yet been found. In any case the vague and fragmentary parts that have reached us of this work hardly fulfill the promise that he gives.

Having defined logic, Avicenna, like the Stoics, begins with a brief discussion of the theory of knowledge. All knowledge, according to Aristotle, starts from particulars, and every belief comes by way of a syllogism. For Farabi “the knowledge of a thing could be through the rational faculty, and it could be through the imaginative faculty, and it could be through the senses”. For Avicenna “all knowledge and cognition is either a concept or an assent, and the concept is the first knowledge and is acquired through definition and what follows the same method, such as our conception of the quiddity of man. And assent is acquired through syllogism and what follows the same method, such as our assent that for everything there is a beginning. Thus definition and syllogism are twin tools with which are acquired the knowledgeables that are known and which through thought become known”.

The origin of these two terms and their Greek equivalents in particular have “baffled modern scholarship for over a century”. Some have tried to attribute them to Sextus Empiricus. They could just as well be attributed to Chrysippus. Actually the terms of Avicenna and to some extent the concept, can be traced back to Arabic translations of theOrganon. But the Stoics, with their well-known interest in language, altered the terms and developed the thought, and it may be presumed, though there is no direct evidence, that it was through some commentator that it reached Avicenna. Among the Falasifa it is first found in Farabi, but in a highly suspect treatise which may be actually by his successor. After Avicenna it becomes the introductory statement of almost every manual on logic whether in Arabic or Persian.

Again he says that all knowledge is either the concept of some particular notion that has meaning, or an assent to it. There could be a concept without an assent, and all assents and concepts are either acquired as a result of some investigation or they are a priori. It may be observed that he regards concepts and assents as the primary sources and correlates them with what he takes to be the fundamentals of logic, viz. definition and syllogism. But there are matters to which we give our assent without the intermediary of syllogistic reasoning. There are sense data “which are matters to which the sense causes assent”, and empirical data “which are matters to which the sense in association with syllogistic reasoning causes assent”. And mere are transmitted data “which are matters to which the transmission of news causes assent”. And there are the accepted data “which are matters to which the word of the person in whose truthfulness there is confidence causes assent; this is either because of a heavenly injunction in his favor, or because of an opinion and effective thought by which he has distinguished himself”. And there are imagined data “which are opinions in which the faculty of the imagination necessitates a belief”. And there are generally widespread data “which are propositions and opinions, famous and praiseworthy, to which the evidence of everybody or of the majority or the evidence of the learned or of most of them, causes assent”. And there are presumed data. And there are imaginative data “which are propositions not stated to obtain assent of any kind, but to imagine something to be something else”' And there are a priori data “which are premises and propusitions originating in man by way of his intellectual faculty without any cause except itself to necessitate its assent”. Moreover the current practice has been to call what leads to the required concept an expository discourse; definitions, descriptions and similar statements are of this kind, and to call what leads to the required assent a proof, and proofs are of three varieties, syllogism, induction and analogy.

Avicenna pays much attention throughout to definition, and considers it of fundamental importance; but before taking up that subject he realizes the necessity of specifying the terms and determining their meaning, because there is a certain relation between the vocable and its connotation; and states affecting the vocables may also affect what they designate. There are three ways, he points out, in which a vocable signifies the meaning for which it stands. One is by way of complete accord between the two, another is by way of implication, and yet another is by way of concomitance.

The vocable could be singular or composite, and the composite may be a complete or an incomplete discourse. The vocable could also be particular or universal; and every universal could be essential or accidental. It may be noted that some of the terms used here are shared by Arabic grammar; and the problem thus arises; did Greek logic have any influence on the development of Arabic grammar, which was systematized and established rather late in the history of the language? This is a moot question on which opinion is divided. In our view there is very little evidence in favor of this theory, though some scholars have held to it tenaciously.

On predication, Avicenna says that every predicate may be either constitutive or concomitant or accidental. Aristotle had discussed the predicates in the Topica and had there specified that they were definition, genus, property, and accident, with differentia as a subdivision, thus making them five in all. Porphyry in his Eisagoge, “losing sight of the principle on which the division was made”, replaced definition by species and maintained that the predicables were genus, species, differentia, property, and accident. This was for him an unusual departure from Aristotle which proved rather confusing to his successors who had thought of him as a faithful interpreter of the Stagirite, though he eventually lost that position after the bitter attacks of Avicenna and his scornful reference to his works. The Eisagoge had been translated into Arabic, and this division of the predicables had been accepted by some logicians of Baghdad, though there occurs a curious classification into six:genus, species, individual, differentia, property and accident, probably under Stoic influence.

Avicenna accepts the five predicables, but not Porphyry’s definition in every case. “Do not pay any attention” he says, “to what the author of the Eisagoge has to say on the descriptive definition of the genus by the species”. Avicenna is opposed to this because he himself distinguishes between natural genus and logical genus. Natural genus is equivalent to the actual essence of a thing in answer to the question “What is it?”, such as animality; logical genus on the other hand is what is added to natural genus in order to give it universality, for logic is a subject that treats of universals. And in this connection he dubs Porphyry “the master of bluff and misrepresentation”, whereas Alexander he had called “the accomplished of the latter ones”, and Themistius, “he who polished his phrases on the books of the first teacher [i.e. Aristotle]”.

Modern logicians share Avicenna’s view on this point and take exception to Porphyry’s definition of the genus. Again Porphyry had divided accident into separable and inseparable, which modern logicians consider impermissible, because “if a singular term be the subject, it is confused; if a general, self-contradictory”; and Avicenna says “do not worry that anaccident be separable or inseparable”. He then proceeds on his descriptive definitions. “Agenus may be descriptively defined as a universal predicated of things of different essences in answer to the question what is it? A differentia may be descriptively defined as a universal predicated of a thing in answer to the question which thing is it? in its substance. And species may be descriptively defined in either of two meanings: first as a universal predicated of things that do not differ except in number in answer to the question what is it? and in the second meaning as a universal to which, as to others, the genus is given as predicate, an essential and primary predication. And property may be descriptively defined as a universal predicated of what is, under one essence only, an attribute that is not essential. And the general accident may be descriptively defined as a universal predicated of what is under one essence, and also of others, an attribute that is not essential.

Just as Aristotelian metaphysics was to become sadly confused with Neo-platonic thought through the translation of the so-called Theology of Aristotle, to the utter confusion of Islamic philosophers, so here we find Aristotelian logic becoming intermingled with that of his followers and also with Stoic logic either directly or through the perplexing disquisitions of the commentators. Galen, whose extant Institutio Logica has been vehemently denounced as spurious and equally vehemently proclaimed authentic, was among those who transmitted this combination. As to Chrysippus, of whom it was said “if gods have logic, this must be Chrysippian”, there is no sufficient evidence that the Falasifa, and Avicenna in particular, had direct knowledge of his work.

With regard to definition, which Avicenna discusses in a number of places and at great length, he states that it is not something that can be obtained through division, which we know to have been the method suggested by Plato. Nor is it possible to reach an adequate definition through demonstration; and even induction must be ruled out since it does not give conclusive knowledge and cannot therefore be of much help. Definition can only be attained through a combination of the above, based on the individuals that are indivisible. In attempting a definition, philosophers do not seek differentiation even though that may follow. What they seek is the reality of a thing and its essence. For this reason there is really no definition for what has no existence: there could only be a statement explaining the name. Where definition is confined to the cause, it is called the principle of demonstration; and where it is confined to the caused or effect, it is then called the consequence of demonstration. The complete definition combines these two together with the genus. Like Aristotle, Avicenna defines a definition as “a phrase signifying the essence of a thing”. And in Persian he repeats that the purpose of a definition is the recognition of the actual essence of that thing, and differentiation is something that follows by itself. It is to be remembered that the authors of the Epistles before him had stated that differentiation was an actual element and a part of every definition; and Averroes after him asserts that all definitions are composed of two natural parts, genus and differentia.

From definition Avicenna turns his attention to the second source of knowledge which is assent, obtainable through syllogistic reasoning. But actually he continually reverts to the subject of definition, particularly descriptive definition. A proposition he defines as “every discourse in which there is a relation between two things in such manner that a true or false judgment follows”. It is known that the Stoics also considered a proposition to be either true or false; they believed that Aristotle held that propositions about future contingencies were neither true nor false. Avicenna adds that “as with interrogation, supplication, expectation, request, surprise and the like, the person who expresses them is not told that he is truthful or untruthful except accidentally”.

Like the Stoics, Avicenna divides propositions into atomic and molecular; the latter being compounded out of the former by a conjunction or connective. The molecular is then divided into “the categorical, the hypothetical conjunctive and the hypothetical disjunctive - a classification which has its Stoic counterpart.

The hypothetical proposition was already known to Aristotle though he does not seem to have explored it. Theophrastus is supposed to have studied it, but only to a limited extent. It is therefore impossible to state with any certainty the source from which it reached Avicenna. The similarity of his approach to that of the Stoics, however, is very close, and like them he devotes much attention to it. Yet he does not stop there and goes on to discuss a number of other propositions such as the singular, the particular, the indefinite, the limited or quantified, the modal, the absolute, and various others for not all of which it is possible to find an equivalent in Aristotelian logic or those Stoic writings that have reached us. One proposition which he definitely claims to be his own, is what he calls “the existential", and this he explains in detail in the Shifa. It arises from the fact that the copula does not exist in the Arabic language, and this was a complication of which Avicenna was well aware and to which he frequently refers. To remedy this linguistic obstacle, various equivalents had been used in different contexts, and among them was the verb to exist. It was from this root and for this purpose that he formed his existential proposition. And Ibn Tumlus testifies to that and explains that it was called existential because it signifies existence without having anything in common with the idea of necessity or contingence. Avicenna, of course, was not the source of Boethius who centuries earlier had discussed these matters in his De Syllogismis Categoricis and De Syllogismis Hypotheticis. These works which had an undoubted influence on mediaeval logic stem from Neo-Platonic and Stoic writings which Boethius had imbibed in Rome.

A review of the conditional proposition leads to the theory of consequence, a notion which, as the fundamental conception of formal logic, played an important role in all Arabic and Persian as well as Western mediaeval systems, and continues to occupy contemporary logicians. Whether the doctrine can or cannot be traced farther back than the Stoic and Megarian school, as described by Sextus Empiricus and Diogenes Laertius, it is the case that the Arabic terms for antecedent and consequence are not to be found in the translation of the Organon, and must therefore have entered the language through some other source. This could have been through Stoic writings directly, in which we find the Greek equivalents, or through the works of some of the commentators of Aristotle. It is in Avicenna that the terms are first defined, and successors like Suhrawardi and Ibn Tumlus only copy him. He states that just as the categorical has two parts, a subject and a predicate, the conditional also has two parts. In the hypothetical conjunctive proposition there are two and only two parts or clauses; one is the antecedent and the other the consequent. The antecedent is that to which the condition is bound, and the consequent is that which constitutes the answer. In the disjunctive, however, there could be one or many consequents to the antecedent. So that the difference between antecedent and consequent and subject and predicate is that subject and predicate could be replaced by a simple term, whereas antecedent and consequent could not because each is in itself a proposition.

Another set of terms for which there are no Aristotelian equivalents, and which must have therefore entered Arabic from some other - probably Stoic - source, are those used for a conclusive and an inconclusive proposition. But in his definition of a thing (pragma) which so occupied the Stoics and led to so much discussion, Avicenna follows “the owner of logic” as stated in the De Interpretatione. “A thing is either an existing entity; or a form derived from it existing in the imagination or in the mind or a sound signifying the form or a writing signifying the sound” .

These examples go to show that Avicenna is no servile imitator of any school, but thinks over every question independently and with an open mind. Another illustration of this attitude occurs in connection with his examination of absolute propositions. “There are two views with regard to the absolute proposition”, he says, “the view of Theophrastus and Themistius and others; and that of Alexander and a number of the accomplished ones”. And after giving their viewpoint, he adds what he supposes may have been the original conception of Aristotle himself. And he finally concludes with the remark that “we do not occupy ourselves with showing preference for either the Themistian or the Alexandrian viewpoint; we would rather consider judgments concerning the absolute in both manners”.

There are three procedures for proving something. One is syllogism, the second is induction and what accompanies it, and the third is analogy and what accompanies it. In agreement with Aristotle in the Analytica Priora, Avicenna says, “a syllogism is a statement composed of statements from which, when made, another different statement by itself and not by accident, follows necessarily”; and syllogisms are perfect or imperfect. It is in his division of the kinds of syllogism that he differs from Aristotle. In all his works without exception (and therefore it could not be a late development in his system), he says that syllogisms are of two kinds, by combination and by exclusion; and in one passage he claims that this division is according to what we verified ourselves. The origin of this division, if indeed it has any outside Avicenna’s own mind, is not known. (Aristotle in the Topica had divided syllogisms into the demonstrative, the dialectical, and the sophistic. Galen divides syllogisms into the hypothetical, the categorical and the relative.) It may well be a case of Avicennian simplification; but the terms that he has employed are difficult to translate correctly. The attempt of a modern author to equate them with the categorical and the hypothetical is not satisfactory. They are definitely not of Aristotelian origin. The termiqtiran (combination) does indeed occur in an Arabic translation of a fragment by Themistius without any explanation, however. Ghazali says “the categorical syllogism is sometimes called the iqtiraniy syllogism and sometimes the ostensive” but he seems confused himself. Avicenna states in Persian that “an iqtiraniy syllogism is that in which two premisses are brought together, having one term in common and the other different; then there necessarily follows from them another proposition which is composed of those two terms which were not in common between them” every body is formed, and everything that is formed is created, hence it necessarily follows that every body is created”. All this is simple, and in Arabic he adds that “iqtiraniy syllogisms could be formed from pure categoricals, or from pure hypothetical, or from the two combined”. What is to be resolved is the origin of the name. This is of Stoic origin and is a literal translation of the Greek yoke which had a vague and general sense in Aristotle, but which became a technical term with the Stoics. The word as used by Aristotle in the Organon had been translated into Arabic as izdiwaj. On the other hand the Arab iqtiran had been used by the translators to render other Aristotelian terms in the Sophistica and the De Interpretatione. The equivalence of the Avicennianiqtiran with the Stoic yoke becomes evident from the statement of various authors before and after him; and on the Stoic side by some fragments that have survived.

The istinhnaiy (by exclusion) syllogism is more difficult to identify by association with any particular Aristotelian or Stoic term. He explains that it “is composed of two premises, one conditional and the otherwith an ecthesis or exclusion of either of the two parts; and it could possibly be categorical or hypothetical; and it is this which is called the excluded”. And again the syllogism by combination is different from the syllogism by exclusion in that one of the two extremes of what is wanted is found in the former actually, and is not found in the latter except potentially”.

Aristotle had divided the syllogistic modes into three figures; and all throughout his logical works we have not seen Avicenna make any mention of the fourth figure. But the fact that it had been introduced into Islamic logic through some external source - possibly Galen - is shown by its use in Qazwini, as we have already noted, and also in Tusi.

The Stoics, we are told, distinguished between “true” and “the truth”; and the same distinction is found in Avicenna who calls the first sadiq and the second sidq. This corresponds with his differentiation between haqq and haqiqa which go back to Aristotle himself and are to be found in the translations of his Organon. Farabi had said that “the truth of a thing is the existence particular to that thing”. Avicenna stated that “the truth of a thing is that particularity of its existence which is proven of it”; and Suhrawardi, after repeating the definition of Avicenna, adds that truth is a mental consideration; which corresponds with the Stoic doctrine that it was a simple and incorporeal notion.

An argument, according to the Stoics, was a statement composed of premisses and a conclusion. With their zeal for linguistic innovation, they had changed the terms of Aristotle into those of their own; but the Arabic equivalents of both the Aristotelian and Stoic remained the same; and we find them used by Avicenna also as premiss and conclusion. It is, however, in his enumeration of the different varieties of premisses that we find him going beyond anything said by Aristotle; and it is difficult to determine whether the varieties were his own or taken from some other source. He mentions as many as thirteen.

The doctrine of the Quantification of the Predicate is not of Aristotelian origin, and the Arabic term sur standing for quantification is not to be found in the translations of theOrganon. kindi uses the term rather vaguely; the authors of the Epistles have more to say on the subject and distinguish two forms of predication: the general and the particular; it may be presumed that Farabi too dealt with it, though it does not appear in any of the works so far published. In Avicenna it is discussed at length and all his successors follow him in stressing that there are two forms. Considering that this doctrine had already a long history in the post-classical period, before it was invented anew by Hamilton and Jevons; and that in the opinion of some modern logicians there can be no truth in it, it is interesting to speculate on the sources from which it entered Arabic and Persian logic. Avicenna says sur is the term which signifies the quantity of limitation, like all and not one and some and not all; and a lexicographer explains that “a proposition that comprises the sur is called quantified and limited and it is either general or particular”.

Aristotle’s distinctions of modality are four, viz. the possible, the contingent, the impossible and the necessary. This is confirmed in the commentary of Ammonius, who is said to have been the first to use the term tropos in that sense. Modern scholars have argued with some justice that actually the contingent and the possible are practically indistinguishable in Aristotle. In any case we find Avicenna saying “the modalities are three, necessary, which denotes permanence of existence, impossible, which denotes permanence of non-existence, and possible, which denotes neither permanence of existence nor of non-existence”. This division into three rather than four is copied by his successors as far away as Andalusia. This might suggest that unlike Aristotle, Avicenna does not differentiate between the possible and the contingent; but in fact he does differentiate between the two notions, contrary to what some have supposed. The confusion is only due to terminology. The Aristotelian term for contingency has been translated differently in different passages. Avicenna, who had no access to the original Greek, seems to have preferred the termmumkin for both notions, specifying at the same time, in Persian and at much greater length and clarity in Arabic, that it had a twofold connotation comprising possibility and contingency. He even coins Persian abstract terms for these concepts.

His definition of the contingent as that “judgment which in the negative or the affirmative is not necessary”, hardly differs from that of Aristotle. But in his lengthy explanations he contrasts the ordinary and the special senses of the term munkin and he distinguishes between what is binding and what is necessary. In fact the notions of possibility and contingency are of fundamental importance to him, and extend far beyond logic to the field of metaphysics, which is the pivot of his entire philosophy. Philo had defined the necessary as “that which being true, is in its very nature not susceptible of falsehood”.

Avicenna ends his logical treatises in the traditional way with a discussion of the different fallacies, and in close correspondence with the Sophistics of Aristotle. But even before arriving at that, he takes up the problem of the Petitio Principii. It is generally thought that this problem first appears in the Prior Analytics, but the Arabic terms as used by Avicenna are slightly different from those of the actual translations, and may therefore have come to him by way of some commentary and not from the Aristotelian texts direct. There is a passing mention of it in the Epistles. Avicenna, however, devotes more attention to it, even though he is inclined to consider it a fallacy. In the Shifa he speaks of “the petitio principii that is included among the genus [of those things] that it has not been possible to prove”; while in the shorter works like the Najat, he refers to the matter with an explanation and without specifying whether it is a correct method of reasoning. In the writings of his successors and certain lexicographers, it seems to be accepted as a valid way of reasoning.

The question whether Avicenna was a nominalist or realist is not easy to resolve, and his position not always very clear. But he maintains that a definition is either according to the name or according to the essence; and that which is according to the name is a detailed discourse signifying what is understood by the name for the person who uses it; and that which is according to the essence is a detailed discourse making known the essence through its quiddity”; thus he accepts the conceptions of both nominalism and realism, and may therefore be considered a conceptualist. This is confirmed by his statement in the Shifa that “the logical science its subject was the secondary intelligible meanings that are based on the primary intelligible meanings”; and this conceptualism is the attitude of many modern logicians.

The Aristotelian Organon with its sometimes conflicting accretions in the form of treatises of hellenistic origin had produced a hybrid mixture of extraordinary complexity and of diverse traditions, Megarian, Stoic, Peripatetic and Neo-Platonic. The genius of Avicenna consisted in his careful selection of the fundamental principles from what he called “the first teaching”; in his discriminating acceptance of some of the later additions and modifications; and finally in his critical reconstruction of a system which he considered valid and adequate. Furthermore he can claim the credit of having set the direction of development - if there was to be any - for those who were to follow, along the path that he had opened. When the logical works of his successors are examined, it is seen that they had hardly anything to add. Even among the Andalusian philosophers who were hignly critical of him, such as Averroes with his sterile Anristotelianism, or Ibn Tumlus with his avowed preference for Farabi, there is nothing worthy of note.

The only person to challenge his philosophy effectively, and attack his logic, and even try to change some of its terms, was Ghazali. But the measure of his success, as far as logic was concerned, is reflected in the disparaging remarks of Ibn Tumlus. The arguments of Ibn Taimiyya, one of the most able and accomplished theologians, was directed against Greek logic in general. Nevertheless interest in the subject continued until it became an essential part of the curriculum in all seminaries. One person who attempted alterations and the development of what he called a logic of his own was Suhrawardi, the mystic author of the “illuminative” philosophy, not with any notable results, however.

In the long vista of Arabic and Persian logic, early authors tended to give the place of honor to Farabi, but until more of his works come to light we are in no position to judge his full contribution. After him Avicenna stands supreme. His influence dominates every single book on the subject in either of the two languages. The line extends directly to mediaeval times; and we find Albertus Magnus saying: 'Quae ex logicis doctrinis arabum in latinum transtulit Avendar israelita Philosophus et maxime de logica Avicennae.'