Influence of Late-Antique (ca. 200-800 A.D.) Prolegomena to Aristotle’s Categories on Arabic Doctrines of the Subject Matter of Logic: Alfarabi (d. 950 A.D.), Baghdad Peripatetics, Avicenna (d. 1037 A.D.)[1]
1 Introduction
Avicenna was the first to hold that the subject matter of logic is secondary intelligibles (al-ma‘qūlāt al-ṯāniya),[2] also called “second intentions.”[3] The current literature suggests that Avicenna developed this doctrine in opposition to doctrines like those outlined by Alfarabi.[4] Abdelhamid Sabra and Kwame Gyekye have observed that Alfarabi uses the idea of secondary intelligibles in his commentary on De Interpretatione. Fritz Zimmermann has noted that Alfarabi makes use of this idea in the first book of the Book of Letters (Kitāb al-Ḥurūf), where Alfarabi discusses the subject matter of logic in some detail.[5] More recently, Stephen Menn has called Alfarabi the “inventor” of the expression,[6] saying that for Alfarabi a “second intention is a concept applying to concepts, so something that is predicated of thoughts or ‘intelligibles’ in the soul rather than directly of external things,” a typical example of which is being-a-predicate.[7] This characterization of Alfarabi’s understanding of secondary intelligibles is similar to Avicenna’s. Nevertheless, there are important differences between how the two employ the concept of second intelligibles. In section 2, I will show that Alfarabi says that primary intelligibles not secondary intelligibles are the subject matter of logic. In coming to this conclusion, Alfarabi is influenced by debates among Aristotle’s ancient commentators about the purpose (skopos) of Aristotle’s Categories.
In section 3 I consider how the Baghdad Peripatetics Yaḥyā ibn ‘Adī (d. 974 A.D.) and Abū al-Faraǧ ‘Abdallāh ibn al-Ṭayyib (d. 1043 A.D.) treated the question of the purpose and subject matter of logic. I show that both authors discuss this set of issues in terms that are indebted to the ancient commentaries on the Categories. In a short treatise entitled Exposition on the Difference between Philosophical Logic and Arabic Grammar (Tabyīn al-Faṣl bayna Ṣinā‘atay al-Manṭiq al-Falsafī wa-l-Naḥw al-‘Arabī) Ibn ‘Adī is motivated less by the desire to proffer a novel doctrine of logic’s subject matter and purpose than by the need to distinguish logic from grammar.[8] Ibn ‘Adī’s defense of his views and his rebuttal of the grammarians indicate that even in the last quarter of the tenth-century the Mattā ibn Yūnus-Ṣirāfī debate over the prerogatives of logic and grammar remained important. Ibn al-Ṭayyib discusses logic’s subject matter and purpose in his commentaries on the Categories and the Eisagōgē in terms that are wholly traditional.[9] In the Exposition Ibn ‘Adī does not show that he is aware of Alfarabi’s philosophical innovations about logic’s subject matter in Letters. The same is true of Ibn al-Ṭayyib in his commentary on the Categories. However, the influence of Alfarabi’s discussion of the subject matter of logic in Letters on Ibn al-Ṭayyib’s commentary on the Eisagōgē is noticeable. Ultimately, Ibn al-Ṭayyib’s and Ibn ‘Adī’s arguments and conclusions do not depart greatly from Aristotle’s ancient commentators. With respect to the question of logic’s subject matter and purpose, the influence exerted by these Baghdad Peripatetics on Avicenna was minimal.
In section 4, I show how deeply Avicenna is indebted to Alfarabi in how he conceives of secondary intelligibles. However, against Alfarabi he insists that secondary intelligibles are the subject matter of logic, not primary intelligibles. Avicenna uses Alfarabi’s doctrine of secondary intelligibles as a means for departing from late-antique thinking about what logic is about, and its relation to the rest of philosophy. Despite Avicenna’s appropriation of Alfarabi’s idea, he remains highly critical of Alfarabi’s use of it in developing his doctrine of the subject matter of logic in Letters. In part Avicenna is motivated to depart from Alfarabi because he is interested in establishing logic as a branch of the philosophical sciences. In particular, Avicenna wants to fit logic into the Alexandrian division of the sciences as part of theoretical philosophy (falsafa naẓarīya), rather than insisting with most Peripatetics, Aristotle’s Neoplatonist commentators, Alfarabi, and the Baghdad Peripatetics that logic is a tool rather than a sub-discipline of philosophy.[10]
2 Alfarabi and Aristotle’s Ancient Commentators on the Purpose and Subject Matter of the Categories and Logic
Alfarabi’s discussion of logic’s subject matter arises out of the prolegomena tradition that was ubiquitous among pagan and Christian authors and commentators.[11] In this regard, Aristotle’s Categories was no exception.[12] Before beginning their lemmatic commentary, commentators would go through the motions of settling preliminary questions (kephalaia) about the Categories: its purpose, authenticity, the reason for its title, its place vis-à-vis the rest of the Organon and the rest of Aristotle’s philosophy, etc.[13] The question to examine here is this. How does a discussion in Aristotle’s late antique Neoplatonic commentators about the purpose (skopos) of the Categories get taken up by Alfarabi and transformed into a discussion about the subject matter of logic?
In order to move toward an answer to this question, let us take Simplicius’ (d. 560 A.D.) commentary on the Categories.[14] Before moving to the lemmatic portion of his commentary, Simplicius addresses the question of what the purpose (skopos) of the Categories is. He briefly surveys the different answers given to this question, though he eventually settles on the Porphyrian solution as the correct one.[15] He presents Porphyry’s (d. 305 A.D.) solution both as a culmination of three earlier views about the purpose of the Categories and as a synthesis of all three of them. To the question “What is the Categories about?” all parties were in agreement, according to Simplicius, that the Categories is “about ten simple things, which, since they are most universal, they are called ‘genera.’”[16] Beyond this preliminary agreement, the parties stood at odds. One held that it is about words or expressions (phōnai), another that it is about things (ta onta, or pragmata), and the last that the Categories is about notions (noēmata).
[Text 1] Now some say that they are about words [phōnai], and that the goal is about simple words, and that it is the first part of logic. Just as the first part of the book on propositions [i.e. De Interpretatione] is about composite words, but not about realities [pragmata], so this [book], being about the parts of the proposition, would be about words […]
Others, however, do not accept this goal. It does not, they say, pertain to the philosopher to theorize about words, but rather to the grammarians, who investigate their modifications, configurations, and changes in word-endings, as well as their proper usages and their types. They say the goal is about the very beings which are signified by words; and that these are what is said [to legomenon] […]
In opposition to these considerations too, however, is [the fact that] the present book is a part of the study of logic, whereas to occupy oneself with beings qua beings is to engage in that philosophy which is metaphysical, and in general, primary […]
Others say that the goal is neither about significant words nor about signified realities, but rather about simple notions [noēmata]. For if, they say, the discussion in the [Categories] is about the ten genera, and the latter are posterior and conceptual, then the discussion is about notions […] These people, however, should have considered that to speak about notions qua notions does not pertain to the study of logic, but rather to that of the soul. Of these people, each one had an imperfect grasp of the goal […][17]
Simplicius, with most of Aristotle’s later Neoplatonist commentators,[18] adds a fourth view (the correct one in his eyes) saying that there is a sense in which the Categories is about all of them. Simplicius concludes that it “is about simple, primary words which signify the primary and most generic beings by means of simple, primary notions.”[19]
There are two important conclusions that we might draw with respect to how this debate about the purpose of the Categories might have been adapted by Alfarabi’s for a discussion of the subject matter of logic in general. The first is that despite Simplicius’ insistence that intelligibles (noēmata) play a role in this story about the aim of the Categories, it is not clear what he means when he says that the categories qua expressions (phōnai) signify things (ta pragmata, ta onta) by means of intelligibles (noēmata). Earlier in his commentary, Simplicius appears to be wrestling with how to fit intelligibles into the mechanics of how words signify things.[20] In order to do so, he appeals to a Platonic-Aristotelian theory of cognition, which stipulates that at the moment of cognition the knower, act of knowing, and object of knowledge are one, and, thus, share the same nature.
[Text 2] For neither are significant expressions wholly separate from the nature of beings, nor are beings detached from the names which are naturally suited to signify them. Nor, finally, are notions extraneous to the nature of the other two; for these three things were previously one and became differentiated later. For Intellect, being identical with realities and with intellection, possesses as one both beings and the notions of them by virtue of its undifferentiated unity; and there [sc. in the intelligible word] there is no need for language.[21]
Simplicius seems to be trying to justify or at least give a clear sense to the role he claims intelligibles play in this economy of signification. Yet, his reasoning is less than satisfactory for several reasons. Firstly, it falls prey to the charge of psychologism that he leveled against the party that contended that the categories are intelligibles. Second, while this justification may explain why notions share a nature with the beings that are signified by categories, it obscures the relation between terms and intelligibles. Thirdly, Simplicius’ reasoning does not address the question of signification vis-à-vis intelligibles at all. All he has argued is that intelligibles and beings share a nature at some point in their history. Thus, in spite of his efforts, Simplicius’ commentary leaves the precise details about the place of notions in words’ signification of things an open question for later commentators to work out.
Alexander of Aphrodisias (fl. 200 A.D.) also introduced intelligibles into his discussion of the skopos of the Categories. Simplicius calls Alexander’s solution among the “more complete,”[22] and quotes from Alexander’s (lost) commentary at some length.[23]
[Text 3] This book is the beginning [arkhē] of the study of logic since “the proposition [ho logos] is significant because its primary elements are significant. Now, Aristotle wants to show what the notions [noēmata] that are signified by the primary and simple elements are. In order to do this, he carries out a division of being [to on] not into particular individuals (for these are uncircumscribable [aperilēpta][24] and unknowable, owing to their multiplicity and the fact that they undergo various changes). Rather, he divides [being] into the highest genera, the ten genera presented here, which Aristotle calls “categories,” since they are the most generic, and since they do not themselves serve as the substrate for any other thing but are predicated of others. Thus, the aim is to examine the simple and most generic elements of a proposition that signify simple notions about simple realities.[25]
Like the solution Simplicius prefers, the categories are verbal expressions, but unlike the Porphyrian solution, Alexander says that verbal expressions signify notions (noēmata), not things directly. First, these verbal expressions do not signify just any notion that comes to the mind, but those that are about things outside the mind. Second, since these verbal expressions are predicates of the highest genera, the notions they signify are not of concrete objects, which only naturally stand in the subject term position. Instead, these verbal expressions signify intelligibles about things outside the mind about which scientific knowledge is possible (in the strong way required by Aristotle at beginning of the Posterior Antalyics[26]). Thus, according to Alexander these categories cannot, inter alia, signify notions about concrete, individual objects due to the fact that they are subject to change, can only ever serve as the substrate for properties, and are so numerous as to lack a unifying principle. In order to determine what kinds of thing these intelligibles can be about, it might be useful to consider an example. The verbal expressions “human” and “man” signify intelligibles in the mind that are about realities outside it. In this case, the realities happen to be quiddities (viz. the things you speak about when someone asks “What is X?”, where X is a particular individual). As a consequence, the verbal expressions “human” and “man” belong in the category substance. The verbal expressions “three”, “red”, “slave” are similar but they do not so obviously accord with Alexander’s scheme. They are each predicate expressions that signify intelligibles in the mind that correspond to some state of affairs outside it. In this instance, the predicate “three,” which serves as a response to the question “How many are X?”, the predicate “red” to the question “How is X?”, and the predicate “slave” to the question “What relation does X have (with respect to Y)?”. All of these predicates may be said about individual Xs, which generally are subject to change, are contingent, and are too multifarious, as individual beings, to be organized according to a principle. There cannot, therefore, be any scientific knowledge about X as an individual.[27] However, in each of the example questions mentioned above, we are asking about (1) a real state of affairs, which holds of an X that may or may not be eternal, unchanging and necessary. Nevertheless, regardless of whether X itself can be the object of scientific knowledge, (2) there are intelligibles red, three, and slave with a unitary and constant meaning that do circumscribe states of affairs outside the mind. Thus, (3) the predicate terms “red”, “three”, and “slave” belong to the categories quality (or “How?”), quantity (or “How many?”), and relation insofar as they signify notions that circumscribe such states of affairs. We may, thus, reasonably construe Alexander’s claim that the aim of the Categories is to examine those terms that fulfill conditions (1) – (3).
Whether or not the above is a convincing interpretation of Alexander, I think it is clear that Simplicius’ claim that his and Porphry’s solution to the question of the purpose of the Categories is shared and even anticipated by Alexander is disingenuous. Notions play no role at all in Porphyry’s discussion of this issue in his catechism commentary on Aristotle’s Categories, since according to this account categories are terms directly signifying things. Notions occupy an important place in Alexander’s discussion on the other hand, since according to his account category terms signify notions. Alexander’s idea that words signify notions rather than things found vocal support in Ammonius (d. ca. 517-26). In his commentary on the Categories he rehearses the traditional division of the debate about the purpose of the book into three parties, those who say it is about words, those who say it is about things and those who say it is about notions (noēmata).[28] When he was writing his own Categories-commentary, Simplicius does not mention Ammonius in connection with the “noetic thesis.” This is surprising, since Ammonius is one of the few commentators to come down clearly on the side of the noetic thesis. That is not to say that Ammonius is uncritical.[29] However, notions play a much greater role in Ammonius’ thinking about what Aristotle wanted to accomplish in the Categories.
[Text 4] Now, the Philosopher’s purpose here is to discuss the expressions that signify things by means of notions [dialabein peri phōnōn sēmainousōn pragmata dia mesōn noēmatōn]. From the statement that one of the things that has been mentioned [viz. words, things, notions] is the purpose [of the Categories], it is necessary to infer the remaining two, as we will show. For one speaks about notions when one speaks words, for there is no discussion of non-significant expressions by philosophers. It is clear that these significant expressions signify things by means of notions, for saying a certain thing makes known that one has a concept about it first [legōn ti pragma tēn peri autou proton ennoian echōn tote auto phēsin]. Thus, of necessity [when one speaks about any one of words, things, or notions] one speaks about all three.[30]
Ammonius does not say that the purpose of Categories is to discuss notions. Indeed, he eventually comes to accept the basic Porphyrian formula that Aristotle’s aim was to discuss expressions that signify things.[31] However, Ammonius allots a mediating role to notions in way that is absent from Porphyry if not from Alexander. Futhermore, Ammonius’ insistence—which is more clearly stated than any other commentator—that words, things, and notions are ultimately interchangeable in the context of Aristotle’s Categories was to play an important role in how Alfarabi conceived of the subject matter of logic.[32]
Despite the differences between these thinkers, there is an important similarity that should be highlighted. All parties reduce the question of the purpose of the Categories to a question of determining what exactly categories are. The first party argues that since categories are words, the purpose of the Categories is to study words; the second that since categories are things, the purpose of the Categories is to study things; the third that since categories are notions, the purpose of the Categories is to study notions; the fourth argues that since categories are words that signify things outside the soul, the purpose of the Categories is to study words that signify things outside the soul; the Alexandrian account argues that since categories are terms that signify notions circumscribing states of affairs outside the mind, the purpose of the Categories is the study of expressions answering to that description. Indeed, Porphyry makes this way of addressing the problem of the purpose of the Categories explicit in his commentary, saying that the question of the Categories purpose can only be determined when the subject matter of the Categories has been properly determined.[33] The most important outcome of this debate, at least for the history of Arabic philosophy and logic, is not what Aristotle’s commentators had primarily intended this debate to settle, i.e. the purpose of the Categories. Rather, I suggest that it is plausible to conclude that what Arabic logicians such as Alfarabi took away from this debate are ideas about what the nature of Aristotle’s categories is, and that any adequate answer to this question must involve giving an account of the relation between words, things, and notions.