Philosophy East and West

In my Oriens article on Avicenna's empiricism (Oriens 40 [2012]: 391–436), I present what Avicenna calls the principles of syllogism, which are the different types of propositions that form the irreducible and axiomatic starting points of syllogisms and definitions. As Avicenna states both explicitly and implicitly in numerous passages that I cite, these are all based on experience. Two of these are the primary propositions (awwaliyyāt) and those with built-in syllogisms (muqaddamāt fiṭriyyat al-qiyās), literally, "premises of fiṭra syllogisms," fiṭra being the natural operation of the intellect—thus, "premises whose syllogisms are constructed by the natural operation of the intellect." In his "Note" on my article, Mohammad Saleh Zarepour disagrees with me and claims that, according to Avicenna, these two propositions are not based on experience but are a priori.

To build his case, Zarepour engages in an extensive discussion of the notion of a priori, in particular that of Kant. This seems odd, for in the entire article I never discuss the concept, and the very few times that I use the term "a priori" it is never as the focus of the main argument; in particular I never use either the concept or the term in section V (pp. 404–410), where I deal with those propositions whose analysis Zarepour disputes. The reason I do not use it is because Avicenna himself never uses a term that could be so translated in his epistemological discussions: it does not appear in Avicenna, nor do I present it as appearing in Avicenna. The concepts of apriority and a-posteriority, if one were to insist, are treated by Avicenna in totally different terms through his description of the constitution of the [End Page 833] intellect—it is completely potential upon birth, and hence there are no innate ideas—and through his empiricism, by making "the senses," to quote him, "the means by which the human soul comes into possession of different kinds of knowledge" (L13 in the article). So whence Zarepour's preoccupation with a priori?

I do use the term once at the very end of the Abstract at the beginning of the article (p. 392), in the statement that the intellect "has no innate or a priori contents," where it clearly has a meaning equivalent to "innate" (disjunctive "or" equating the terms on either side), which is the basic dictionary sense of the word: "existing in the mind prior to and independent of experience; contrasted with 'a posteriori' (empirical)."¹ Beyond this, nowhere in the texts I present, but also elsewhere, does Avicenna indicate that there is a time or instance between birth and a child's first experiences or after his first experiences when the intellect would form, develop, or acquire any ideas independently of experience. Hence for Avicenna there are no innate ideas, and none, even after birth, independent of experience, and thus for Avicenna "a priori" in that sense, were he to use the expression, is tantamount to innate. Zarepour imputes this statement to me; he says, "[Gutas] thinks that the notion of a priority is identical to the notions of innateness and pre-givenness." But this is Avicenna's position, as it can be determined: if he says that the intellect upon birth is purely potential and that ideas are developed only on the basis of experience, we can only say, if we have to use "a priori" terminology, that a-priority is identical to innateness.

However, "a priori" as used at the very end of my Abstract in that article is not the focus of attention or its subject, as is obvious to anyone who reads the whole passage and the article. But Zarepour picks on this passage, which he then splices into one half of a sentence in my footnote 41 to create an artificial thesis, foisted on me. He writes: "[Gutas] thinks that primary data and fiṭrīyāt… both… are non-a priori and 'analytic, in Kantian terms'." Apart from the fact that there is no such sentence by me and I never use the term non-a priori, the last four words, which are mine, occur in a footnote (41), where I merely suggest, with proper conditional circumspection, that "it appears that both kinds of propositions would be analytic, in Kantian terms." By this I do not mean to say that Avicenna prefigures Kant—I do not say that even for Locke, with whom the parallels are much more striking—but merely to give an idea of the nature of these propositions, and particularly, given my interest in drawing parallels with Locke, to indicate, in the same footnote, which Zarepour ignores, that they share the same qualities as those propositions that Locke calls "trifling" or "verbal." The reference to analytic propositions and Kant, the only place where I mention him, is not a thesis of mine, let alone a major thesis that I wish to put forth. Some of these propositions are certainly analytic and "verbal," as Locke says, and arguably all of them are. But even if the [End Page 834] analogy were not totally appropriate it has no bearing on what Avicenna says and on my exposition of it; it only suggests that further study may be needed on Locke's and Kant's similar concepts to establish whether they share common features, if one were interested in that.

This splicing together of two independent statements of mine in different semantic contexts in order to concoct an artificial thesis, allegedly mine, and set the problem in Kantian terms was rather unnecessary on Zarepour's part;² all he had to do was say "I disagree with Gutas because according to Kant, such propositions are 'synthetic a priori in the Kantian sense'," and then proceed to his main point about "the a priority of justifications" (see below), avoiding the irrelevant, for Avicenna, discussion of a priori.

Having set the problem artificially as an issue about a priori in Kantian terms, though, Zarepour argues for his claim by offering no evidence whatsoever from Avicenna himself (despite his claim that he has taken "a deeper look at Avicenna's writings on fiṭrīyāt"). He only repeats—in a footnote, no less (no. 29)—a statement by Avicenna in my article, which he completely misreads, as follows (L7, p. 406 in the article; here I add the Arabic text):

One mode of intellects is that it is "potential in all aspects, like the human intellect [i.e., the material intellect], because the intelligibles are potentially in it except the primary [intelligibles,] which come about in it as he grows up" ([anḥā'u l-'uqūli]… minhā mā yakūnu bi-l-quwwati min kulli wajhin ka-l'uqūli l-insāniyyati fa-inna l-ma'qūlāti fīhā bi-l-quwwati illā l-awā'ila fa-innahā taḥṣulu fīhā ba'da tara'ru'ihī).

Avicenna is referring here to the two ways in which different intelligibles are acquired, something he does elsewhere (e.g., in L8, pp. 406–407 in my article): the primary intelligibles come about in the intellect naturally, in the way just described, through the natural operation of the intellect on data and concepts received from experience as the child grows up; this happens at the stage of the intellect Avicenna calls dispositional (bi-l-malaka). The other intelligibles, which are potentially in it as he says, come about after acquisition, that is, through syllogistic means by hitting upon the middle term in syllogisms; this happens at the stages of actual and acquired intellect. Strangely, Zarepour reads the sentence with the emphasis on "potentially in it" disregarding the meaning of "potentially," which is intended to explain the phrase "potential in all aspects" in the main clause of the sentence: an intellect that potentially has intelligibles in it actually has none; a glass that potentially has water in it actually is empty. Zarepour is completely unaware that if, on his reading, the intelligibles in the phrase "the intelligibles are potentially in it" were actually in it, this would mean that according to Avicenna the intelligibles that are acquired at the end of a syllogistic process are actually in it beforehand! [End Page 835]

Avicenna thus being discounted as a possible source for understanding what Avicenna says, Zarepour brings in, as his sole argument, Kant's theory of "the a priority of justifications," which means that the justification or acknowledgment of the truth of a proposition ("judgment" as Zarepour says) "is not based on empirical evidence that cannot be grasped without having some experience of the external world"; the concepts that make up the proposition may be based on experience, but the justification of the proposition is not, and hence the proposition itself is a priori. Zarepour says that "Assenting to the truth of fiṭriyāt is internal to the intellect and does not depend on the empirical information we grasp through our senses. So it seems that, contrary to what Gutas suggests, according to Avicenna fiṭrīyāt are justified in an a priori way." And again: "It is very difficult to deny that Avicennian primary data and fiṭrīyāt are a priori. These principles, though possibly constituted from non-innate concepts, are a priori, since they have a priori justification."

The problem is that in Avicenna's system there is no stage of "justification" or judgment between the concepts derived from experience and the formation of both the primary propositions and those with their syllogism built-in; the process is automatic: as soon as the concepts are derived from experience, primary notions and propositions with built-in syllogisms "come about" or are formed. To start with the latter, Avicenna gives as an example the proposition "four is even," which, if stated as the built-in syllogism, would have a middle term "divisible in two equal parts"; thus, the syllogism would be "four is divisible in two equal parts, whatever is divisible in two equal parts is even, therefore four is even." Here, however, the middle term is not an a priori concept that causes or justifies the conclusion, but is part of both the minor and major terms, for this is essentially a tautology: in the minor, "four" is the species of the genus "divisible in two equal parts" and thus included in it (a "verbal" proposition as Locke says in my note 41); and in the major, "divisible in two equal parts" is the definition of "even," and thus completely identical with it, tautologically. So the intellect does not have to do any operation to reach the conclusion; it is automatic, because the conclusion is already built into the terms themselves as perceived by the natural operation of the intellect, its fiṭra, which is what Avicenna means when he says, "whoever understands 'four' and understands 'even', there appears to him that four is even." Thus, there is no middle term outside the major and minor terms that can cause the conclusion and thus require an operation of justification at some point. And the absence of any stage of justification or judgment to reach the conclusion is also indicated when Avicenna says [L9–11] that for such propositions it is impossible for the intellect to doubt them or not accept them. Propositions with built-in syllogisms are thus based only on experience.

Next, regarding the primary propositions, about which Avicenna writes more frequently, this is how he describes their formation: [End Page 836]

The awwaliyyāt, primary concepts, come about (taḥdutu) in humans through their intellective faculty (quwwa 'aqliyya) without any cause that would make their acknowledgment necessary other than themselves and the thing (ma'nā) which makes them (al-jā'il lahā) propositions, which is the cogitative faculty (almufakkira) which combines simple [concepts, basā'iṭ] by way of affirmation and negation; so when the simple concepts (al-basā'iṭ min al-ma'ānī) come about (ḥadatat) in humans with the help of sensation and [the internal sense of] the imagery [ḫ ayāl—which stores in the brain the forms acquired from the senses] or in some other way, and then the combinative cogitative faculty compiles them (allafathā), the mind must acknowledge their truth unmediatedly (ibtidā'an, ab initio) without any other reason and without being conscious of having just acquired this but rather thinking that he had always known it.³

As can be seen here, Avicenna talks about the primary notions as "coming about" in the mind automatically as the intellect or cogitative faculty combines (jama'a) and compiles (allafa) previous concepts to make propositions. There is no mention of any justification or judgment here—as a matter of fact, what he says precludes any judgment: he says that the intellect acknowledges the truth (taṣdīq) of these primary propositions. Now taṣdīq is something that the intellect does when it works with syllogisms: there are the minor and major terms, and once one recognizes the cause of the conclusion, the middle term, then one acknowledges its truth. But for the primary notions there is no middle term, for its existence is precluded: there is no "cause [or: middle term] that would make their acknowledgment necessary other than themselves and the thing which makes them propositions, which is the cogitative faculty," as he says above.

Further down in the passage, what I quote in L6 (p. 406), gives further details about the natural operation of the intellect with regard to primary concepts:

An example of that [a primary intelligible] is [the proposition] that "the whole is greater than the part." One does not come into possession of this either from sense perception (ḥiss) or induction (istiqrā') or anything else. To be sure, it is possible that sense perception provides him with a concept (taṣawwur) of "the whole," "the greater," and "the part," but acknowledging the truth of this proposition comes from his nature (jibilla)….

In forming concepts, sense perception and imagination (taḫ ayyul) assist the intellect because sense perception presents to the [internal sense of] imagery (ḫ ayāl) things in a mixture, and imagery [presents them] to the intellect. The intellect then discriminates among them, breaks them down into parts [i.e., categories], takes up each one of the concepts individually, and arranges [in order] the most particular and the most general, and the essential and the accidental. Thereupon there are impressed on the intellect, in [a process of] concept formation, the primary notions (al-ma'ānī al-ūlā), and then definitions are composed out of them. [End Page 837]

The part I italicize says precisely how these propositions are formed: through the natural operation of the intellect without any justification or judgment on the basis of an (a priori) concept being involved. The rest of the passage provides details about many operations of the intellect, like sorting concepts, arranging them according to what is essential and what accidental, et cetera. This is the function of the intellect, what it is meant to do, just as the other faculties of the soul each have their proper functions: the eye sees and the ear hears, the imagination in the central ventricle of the brain combines and separates forms, images, and notions, and memory stores notions and ideas in the rear ventricle of the brain, et cetera. These functions of the intellect—the sorting, the arranging, the concluding, et cetera—are not ideas or intelligibles in the intellect, innate or not, that help it do the thinking or justification, or upon which it operates, but the very nature of its operation; the function of a faculty is not the same as its objects. Without these functions the intellect would not be intellect except homonymously, as Aristotle would say. In Kant's model, though (as Zarepour presents it and as I understand Zarepour), a priori justification is not part of the operation of the intellect and hence, we can conclude, could be removed from it without the intellect ceasing to be intellect.

Thus, in Kant's system and terms of reference, primary propositions and the propositions with built-in syllogisms may be "a priori," as Zarepour says Kant understands the concept, but in Avicenna's system and terms of reference they are experience-based and not innate (with the concept "a priori," not used by Avicenna, being meaningful in Avicenna's terms only if it means "independent of experience"). So in essence, it would seem that what Avicenna calls the natural operation of the intellect, its fiṭra, Kant would call "a priori justification," something which Zarepour fails to realize. Thus "the a priority of justifications" may allow Kant to introduce a priori concepts in the intellect, as Zarepour claims, but for Avicenna the fiṭra concept safeguards the experience basis of all knowledge, as I present it in my article. This does not raise the issue of who is right and who is wrong (whose resolution would require a supreme philosopher-arbiter—and who would that be?), but it does make it abundantly clear that in the study of the history of philosophy one cannot explain one philosopher through another because the terms of discourse and referents are never exactly the same in the two philosophers. For in that case one could turn the tables on Kant and say that, from Avicenna's point of view, Kant is mistaken to claim non-experience–based notions in the intellect by introducing a meaningless "a priority of justifications" concept because there is no such procedure of justification involved, as these propositions are formed automatically through the natural operation of the intellect.

On this point I may indulge in some minor criticism of Zarepour's procedure, as I would disapprove of his taking sides and even, perhaps, of his trying to play the part of the supreme philosopher-arbiter. For the upshot of his article is that he is actually arguing with Kant against Avicenna in the [End Page 838] guise of arguing against my presentation of Avicenna's theories—and sometimes not so much in the guise as openly. For example, Zarepour writes: "Coming-about-in-mind is not the same notion as non-a priority." "Coming-about" reflects Avicenna's usage of ḥaṣala, in my translation, to express the appearance of primary notions in the intellect: they come about in it, ḥaṣala fīhi, or they occur to it/him (ḥaṣala lahū). Now "non-a priority" is not a concept that Avicenna uses, so we have no way to understand what Zarepour is talking about other than to assume that "non-a priority" means "coming from experience," in which case Zarepour's statement would be "Coming-about-in-mind is not the same notion as coming from experience." In this case it is clear that Zarepour disagrees with Avicenna and disapproves of his use of "coming about," hḥaṣala, because Avicenna says precisely this (L8 in my article): "As for the primary notions (al-awā'il) that come about for him / occur to him (taḥṣulu lahū), they come to pass from induction (istiqrā'), which is testing and proving (tajriba), and from Experience (šahāda)."⁴ I don't know how much more explicitly one can say that primary propositions do indeed come only from experience—he even insists that induction itself is also experience—and that they are not innate, and hence not a priori in the sense of independent from experience, the only meaning of the term that would make sense in Avicenna's context. But Zarepour apparently finds fault with Avicenna and disapproves of the very words Avicenna uses, and thinks that he can interpret them correctly by using Kant's analytic categories. Historians of philosophy should not be in the business of judging philosophers, but understanding them.

And now a final, and oft-repeated,⁵ cri-de-coeur appeal of mine (and Avicenna's, I would think, were he alive) to read Avicenna through Avicenna. Among other qualities that distinguish Avicenna, he also has the unfortunate and undeserved distinction of being the most tendentiously interpreted and misinterpreted philosopher—even Aristotle, whose thought was variously transformed and re-interpreted throughout the centuries (as Sorabji aptly puts it), was not interpreted in so many widely divergent ways (I don't think anyone claimed him to be a mystic). To be sure, Avicenna's thought is deep, complex, meticulous, and of course the mark of genius; but it is also clear and not obscurantist. And yet it has been his fate to be read from early on in the Islamic world as an Ismā'īlī, and then a Twelver Shi'ite, an Ash'arite, a sufi, an illuminationist (išrāqī), the chief exponent of 'irfān, and certainly in some other obscure ways; and in European scholarship he has been read as a mystic and "oriental" philosopher, a theosophist, a Marxist, and now as a Kantian. Some of these interpretations are certainly ideological: everybody wanted a piece of Avicenna, for claiming affinities to, or seeming to reject his thought, enhanced one's credibility and stature; others are just as certainly based on ignorance, sensationalism, even selfpromotion.⁶ Whatever the case, the historically accurate and honest way to read Avicenna would be through Avicenna himself and in this case state, as [End Page 839] he does, that primary propositions and those with built-in syllogisms are experience-based.