Yaʿqūb al-Qirqisānī on Human Intellect, Legal Inference, and the Meaning of the Aristotelian Syllogism

In: The Journal of Jewish Thought and Philosophy
View More View Less
  • 1 Ariel University
Full Access


In the fourth treatise of his legal-theological work Kitāb al-Anwār wa-al-Marāqib, Yaʿqūb al-Qirqisānī analyzes a criticism of the Aristotelian syllogism and its epistemological foundations. Qirqisānī defends Aristotelian logic by quoting a passage from an unknown commentary on Aristotle in which the Aristotelian theory of syllogism is explicated. This paper focuses on the historical, theological, and philosophical meanings of the criticism of the syllogism in Qirqisānī’s discussion and analyzes his interpretation of the syllogism as a source of knowledge that should be applied in the realm of legal reasoning and in the interpretation of biblical law.


In the fourth treatise of his legal-theological work Kitāb al-Anwār wa-al-Marāqib, Yaʿqūb al-Qirqisānī analyzes a criticism of the Aristotelian syllogism and its epistemological foundations. Qirqisānī defends Aristotelian logic by quoting a passage from an unknown commentary on Aristotle in which the Aristotelian theory of syllogism is explicated. This paper focuses on the historical, theological, and philosophical meanings of the criticism of the syllogism in Qirqisānī’s discussion and analyzes his interpretation of the syllogism as a source of knowledge that should be applied in the realm of legal reasoning and in the interpretation of biblical law.

1 Introduction

The fourth treatise in the legal-theological work of the Karaite scholar Yaʿqūb al-Qirqisānī (first half of the tenth century), Kitāb al-Anwār wa-al-Marāqib (Book of Luminaries and Observatories), is titled: “On the methods and causes through which [a person] acquires knowledge of the commandments, and analysis of the ways of the [scriptural] verses and the methods of inference, [brought] in 68 chapters.”1 The treatise’s scope is actually wider. It discusses principles of epistemology; methods of logical, judicial, and talmudic inference; ways of biblical exegesis and halakhic methodology; principles of formulating philosophical arguments; and participation in theological disputes.2 In the greater part of the fourth treatise, the characteristics of legal inference (qiyās) are examined, from many and variegated perspectives—a treatment that expresses the Karaite leaning towards rationalism3 and the Karaite emphasis on applying human reasoning in biblical exegesis as a religious obligation.4

Within the framework of his discussion of legal inference, al-Qirqisānī deals with a theological (kalāmic) criticism that was raised against the Aristotelian syllogism. Qirqisānī defends the need for syllogistic reasoning and demonstrates its fruitfulness in enabling humans to recognize previously unknown knowledge.

The essence of the present study is an analysis of al-Qirqisānī’s discussion of the Aristotelian syllogism. In order to provide some context for this discussion, I would like to commence by demonstrating the importance of the Aristotelian syllogism in Qirqisānī’s legal thought.

2 The Aristotelian Syllogism in Qirqisānī’s View of Legal Inference

One of the central points of controversy between the Rabbanites and the Karaites is the legitimacy of inference in the realm of the law, as a means of obtaining knowledge of the divine commandments. This issue is of central importance in the controversy between Saʿadya Gaʾon and Qirqisānī, who were among the conspicuous representatives of the two contending camps.5

In the first chapters of the fourth treatise of Kitāb al-Anwār, Qirqisānī analyzes a number of logical-epistemological principles that serve as criteria for recognizing truth. These principles have their root in kalāmic thought, but most of them can be traced back to ancient Greek or Hellenistic sources (i.e., the writings of Aristotle and Peripatetic and Stoic texts).6 From Qirqisānī’s point of view, as it appears from the legal-exegetical context in which he writes, the importance of these Hellenic-Arab principles is their application in the domain of law and the exegesis of the biblical commandments.

The epistemological principles Qirqisānī discusses include different versions of the Aristotelian syllogism. Qirqisānī clarifies the fruitfulness of using syllogism in the legal exegesis of Scripture and in extending the halakha to circumstances not explicitly mentioned in the Torah’s verses. Thus, for example, says Qirqisānī:

And among the obvious7 [cases] in the commandments, the words of the book [the Torah]: “you shall not eat any fat or any blood” (Lev 3:17)—this is a premise. And it is stated: “its fat, the entire tail fat” (v. 9)—this is another premise. And since all fat was prohibited, and afterwards he informed that the tail fat is fat, it follows8 that it is forbidden to eat the tail fat. And in this manner the sacrifices of the wicked have been forbidden by a group [of scholars], as it is stated: “You shall not eat any abomination” (Deut 14:3), and elsewhere it is stated: “the offering of the wicked is an abomination” (Prov 21:27). Thus, when it is said “You shall not eat any abomination,” and afterwards we are informed that the sacrifices of the wicked are an abomination, it follows from these two premises that the sacrifices of the wicked may not be eaten. And there are more similar cases; each has been mentioned in its place.9

What makes possible the construction and articulation of the syllogism—both in the domain of science as well as in the domain of law—is the identification of the middle term, the term shared by the syllogism’s two premises, which explicates the necessary connection between the subject and the predicate of the inferred conclusion.10 According to Qirqisānī, the middle term of the legal syllogism is the rationale of the precept, that is to say, the inner logic by which God determined the commandment. For example, in the text quoted above, “being fat” is the middle term through which we know that “tail fat” is forbidden. Likewise, “being an abomination” is the middle term through which we know that the sacrifices of the wicked may not be eaten.11

The seventh chapter of the fourth treatise of Kitāb al-Anwār discusses the fact that every commandment has a cause, a rationale (ʿillah), which is either clear and explicit, or has to be learned through one of the ways of inference, or it is completely hidden from us.12 The methodological implications of this chapter are clear: it is incumbent upon us to find, to the extent that this is possible, the rationale of every single commandment, and to apply it by broadening its validity to additional cases where this rationale is present:

Every [thing] that has been prohibited has this [kind of] rationale found in it, [that is to say] God abhors it. But we have to investigate the nature of the abhorrence and the prohibition. And since this [thing] comes about as we have explained, and we shall know the rationale of the commandment and its purpose, whether through [the testimony of] an [explicit] verse or otherwise, then it is incumbent upon us to apply this rationale in cases where it is applicable.13

This way of formulating the question constitutes a characteristic application of the Muʿtazilite method, according to which good and evil are not simply expressions of God’s arbitrary will but objective entities that can be recognized by human reason.14 According to Qirqisānī, the human intellect is also capable of discerning the reasons that led God to command or to prohibit, and to apply these reasons in cases not explicitly dealt with by the revealed sources.15

Given the central place of the syllogistic procedure in Qirqisānī’s legal theory, it is clear that in his mind the Aristotelian syllogism is a valid and useful mode of inference. As will be shown, Qirqisānī takes pains to explicate the logical basis of the syllogism and thereby to refute critics who deny its utility.

3 Criticism of the Aristotelian Syllogism: between Qirqisānī and Ibn Taymiyya

The title of the fifth chapter of the fourth treatise of Kitāb al-Anwār is “Concerning the speech about the premises and the conclusions.”16 According to this title, the subject of the chapter should be understood as a discussion of the relations between the premises and the conclusion of the syllogism. At the beginning of the chapter Qirqisānī cites an argument of “one of the theologians” (a scholar of kalām), who criticizes the Aristotelian syllogism. Qirqisānī himself does not accept the objections of the theologian, and he discusses them only in order to refute them. These are the words of Qirqisānī:

This issue, which concerns the derivation of the truths and the knowledge of the demonstration (al-burhān), belongs to the determinations of logic. It [i.e., the topic of syllogism] has already been condemned by one of the theologians (baʿḍ al-mutakallimīn). He was of the opinion that it does not generate (la yustakhraju) unknown knowledge. He said: If we claim that “every human being is a living being” and “every living being is a substance”—both of these premises being true—and then infer, on their basis, that “every human being is a substance”—then nothing that we did not [already] know was concluded and we have not learned anything of which we were [previously] ignorant, since our knowledge that a human being is a substance preceded (mutaqaddimah) [the syllogism], and that which does not teach anything unknown cannot serve as proof or demonstration. And since this is so, it has been verified that this matter [i.e., the syllogism] is of no use.17

During the tenth century, criticisms were voiced against Aristotelian logic,18 seemingly within the framework of the general religious opposition to Aristotelian philosophy. It should be noted that Qirqisānī cites these words of “the theologian” in a chapter whose aim is to discuss “the premises and conclusions,” and that the theologian’s criticism, at least in Qirqisānī’s view, is not that the syllogism is useless because we know the content of the conclusion (“every human being is a substance”) without the syllogism. Rather, the theologian’s criticism concerns the relation between the syllogism’s premises and its conclusion. The theologian’s statement that “since our knowledge that a human being is a substance preceded (mutaqaddimah) [the syllogism]” means that our knowledge of the conclusion not only stems from a source of knowledge that is not related to the syllogism, but it also “precedes” the syllogism—we know the conclusion before we undertake the logical procedure of inferring, and hence the logical procedure is useless. This is the meaning of Qirqisānī’s statement that the theologian argues that the syllogism “does not generate (la yustakhraju) unknown knowledge”—namely, according to the theologian the procedure of istikhrāj (inferring) becomes useless. All this means that, at least according to Qirqisānī, the theologian’s claim is that the syllogism’s conclusion is implicitly contained in its premises, and hence the syllogism is useless.19

The argument of the theologian cited by Qirqisānī is reminiscent of the criticism leveled at the Aristotelian syllogism by the skeptic Sextus Empiricus in the second century CE. Sextus argued that the major premise of the syllogism is necessarily based on induction, meaning that a single particular case can refute the truth of the premise if its characteristics deviate from the rule.20 In Sextus’s opinion, if the major premise is based on an incomplete induction (that is, the examination of only a sample of cases and not all of them), then the major premise cannot be considered valid, and any conclusion concerning a particular case not already included in the major premise must in any case be invalid. On the other hand, if a complete induction has been carried out (i.e., all the relevant particular cases have been examined), then the conclusion of the syllogism is necessarily found already in the induction on which the major premise is based. Thus, according to Sextus, the premise “every human being is a living being” was determined only after we have verified and seen that “Socrates is a living being” (and also Plato, and Dion, and so on). That is to say, in his opinion, any meaningful syllogism is necessarily a vicious circle, a petitio principii: the syllogism’s conclusion is already included in its major premise.21

In modern times, a criticism similar to that of Sextus has been put forth by several scholars,22 among whom the most famous is John Stuart Mill. Mill’s criticism is based on the perception that a general proposition is a result of an inductive inference. For example, the generalization “All men are mortal” must be based upon empirical observation of individual cases—“John is mortal,” “Thomas is mortal,” and so on—a process that culminates in the induction “All men are mortal.”23 Hence, according to Mill, “it must be granted that in every syllogism, considered as an argument to prove the conclusion, there is a petitio principii.”24

Although it might seem surprising, a similar line of thought is found in al-Fārābī’s Kitāb al-qiyās al-ṣaghīr (Small Book on the Syllogism).25 Al-Fārābī cannot be considered a critic of Aristotelian logic, and he in no way cooperated with scholars of kalām who wished to attack Aristotelian philosophy. However al-Fārābī, like Sextus, recognizes a circularity in the Aristotelian syllogism. In his discussion of the syllogism, al-Fārābī clarifies the difference between complete and incomplete induction,26 and he goes on to claim that induction might be used in order to construct the major premise of the syllogism. When the induction is incomplete, the major premise remains unjustified. On the other hand, when the induction is complete, the conclusion of the syllogism is a priori known already at the inductive stage, when the major premise is established. In this case, the syllogism is superfluous and also circular.27

It should be stressed, however, that al-Fārābī did not put forth this argument as a general criticism of the meaning of the Aristotelian syllogism. He presented it within the framework of his discussion of induction, and did not take it up in his discussion of syllogism. From the words of al-Fārābī we learn that it is indeed possible to use induction in order to construct the major premise of the syllogism. However, this is a unique case, since in the view of al-Fārābī, the Aristotelian syllogism is not always based on induction. Al-Fārābī sums up his discussion of the issue in the following words: “It has been demonstrated that induction (al-istiqrāʾ, lit. “investigation”) cannot be used to verify a fact in order that it may serve as a premise of a syllogism, that its purpose is to affirm or to negate its predicate to one of the things contained in its subject.”28 This criticism concerns induction only. Al-Fārābī explicitly states that the premises of a syllogism are based either on another syllogism or on one of the four types of knowledge, categorized by source: received knowledge (maqbūlah); well-known (or: commonly accepted) knowledge (mashhūrah); sensory knowledge (maḥsūsah); and knowledge that is intellectual by nature (maʿqūlah bi-al-ṭibʿ).29 Obviously, as long as a syllogism is not based on empirical induction, criticism like that voiced by Sextus and Mill is not valid. Al-Fārābī did not subscribe to an epistemology in which knowledge is based exclusively on empirical examination of particular cases, and neither did Qirqisānī (as attested by the epistemological principles that he presents in the beginning of the fourth treatise of Kitāb al-Anwār).30

It should be noticed that the criticism expressed by Qirqisānī takes a slightly different path than the one followed by Sextus Empiricus. Like Sextus, the theologian cited by Qirqisānī does note that knowledge of the syllogism’s conclusion precedes knowledge of its premises. But his focus is on the other side of the coin—on the fact that syllogism does not generate unknown knowledge.31

A similar argument is expressed by Ibn Taymiyya (d. 1328 CE), who was perhaps the Muslim scholar best known for his arguments against the Aristotelian syllogism. According to Ibn Taymiyya, a syllogism does not discover unknown knowledge.32

Qirqisānī, for his part, rejects the criticism expressed by the theologian and defends both the validity of the syllogism and its utility. Qirqisānī’s standpoint is thus opposite to that of Ibn Taymiyya. The views of Qirqisānī and Ibn Taymiyya share a common point of departure—namely, the similarity between syllogism and analogy—but the two scholars analyze this resemblance in different ways. Ibn Taymiyya analyzes the epistemological similarity (from the point of view of the ways of gaining knowledge) between syllogism and analogy. In his view, inasmuch as syllogistic reasoning includes the drawing of conclusions, its major premise is not truly general, since in order to be general it has to be based on complete induction, which is practically impossible (and if it were possible, the syllogism would not generate new knowledge). The major premise of the syllogism is therefore based on incomplete induction.33 According to Ibn Taymiyya, in the syllogism “all human beings are mortal; Socrates is a human being; Socrates is mortal,” the major premise is based on the examination of a limited number of human beings, that is, on the empirical determination that they are mortals. From this perspective, it appears that the conclusion concerning Socrates’s mortality is based on the analogy between Socrates and the particular human beings who have been observed, who all share the characteristic of being human beings.34 Wael Hallaq notes that Ibn Taymiyya, as a scholar of Muslim jurisprudence, appreciated the force of legal analogy (qiyās), and in attacking Aristotelian logic Ibn Taymiyya claims that a syllogism does not generate knowledge that is more certain than the knowledge arrived at through an analogy. In Ibn Taymiyya’s view, the analogy is preferable to the syllogism, because an analogy makes explicit at least one of the particulars on which an inference is based, which is not the case in a syllogism.35 Of course, all this is opposed to the Aristotelian standpoint, which sees things the other way around. In a syllogism, according to the Peripatetics, if the premises are true then the conclusion is necessarily also true, but in an analogy, even if the premises are true the conclusion is not necessarily so. Therefore, in their view, the syllogism is epistemologically superior to analogy.

As demonstrated earlier, Qirqisānī also analyzes the epistemological resemblance between analogy as it is employed in juridical-halakhic inference and the syllogism. However, while Ibn Taymiyya used this resemblance in order to attack philosophy and the epistemological superiority of the Aristotelian syllogism, and in effect to defend religious jurisprudence, Qirqisānī defends the syllogism and claims that it generates new knowledge. Thus, Qirqisānī is of the opinion that legal inference (qiyās) is based on Aristotelian logic. In short, the resemblance of the legal-halakhic analogy to the syllogism leads Qirqisānī to endorse Aristotelian logic, while it causes Ibn Taymiyya to reject that very same logic.

4 Defending Aristotelian Syllogism and Explaining Its Epistemological Function

In order to understand the philosophical meaning of the Aristotelian syllogism, and indeed Qirqisānī’s Peripatetic background of his defense of the syllogism, let us consider how Aristotle’s schoolmen interpreted their master’s theory. From the point of view of the Peripatetics, the syllogism is a valid and useful mode of argumentation. Alexander of Aphrodisias, in his commentary on Aristotle’s Analytica Priora, discusses Aristotle’s claim that the syllogism is an argument (logos) in which certain things are laid down and something else necessarily follows from them.36 Alexander explains that Aristotle was right in claiming that in a syllogism the conclusion has to be different from the premises, for if the conclusion is identical to the premises, the syllogism is of no value: “It is useless—it destroys any syllogistic utility—to infer what is agreed and supposed.”37 A similar approach is found in the fragments of the Hebrew translation of Themistius’s commentary on the Analytica Priora (or of what is considered to be Themistius’s commentary on this book).38 There Themistius says:

How excellent is his [i.e., Aristotle’s] saying “something other than they.” For if the sentence39 that is derived is itself among the sentences that are assumed, there is no benefit40 in such syllogism, since then we shall not acquire anything hidden from what is manifest.41 Now, the syllogism is everywhere a tool we employ to receive benefit.42

According to Themistius, if the conclusion were identical with knowledge already contained in the premises, the syllogism would be useless. This is why, he explains, Aristotle stressed that the conclusion is “something other than” the premises.

Yet, the Peripatetics still needed to explain in what way the syllogism’s conclusion is “something other than” its premises. What is the relation between the syllogism’s conclusion and its premises and how the syllogism functions as a mode of inference? An answer to these questions can be found in Qirqisānī’s answer to the claims of the theologian. Qirqisānī’s answer can also serve as an answer to Ibn Taymiyya, and to objectors like Sextus Empiricus and John Stuart Mill.

Immediately after he voices a short criticism of the words of the theologian, Qirqisānī cites a long passage from “one of the commentators” (baʿḍ al-mufassirīn)43 that contains an explanation of the logic at the base of the Aristotelian syllogism. From the words of Qirqisānī it appears that, at least in his view, the commentator’s position constitutes an answer to the claim of the theologian, who rejected syllogism as a generator of new knowledge and claimed it to be epistemologically superfluous. Qirqisānī writes: “[O]ur only purpose [in citing the commentator] has been to [clarify] the aforementioned reason of premises and conclusions, and that it is possible through them [or: by means of them] to draw conclusions concerning concealed, unknown matters.”44

According to this anonymous commentator, statements belonging to the categories of “[t]he imperative, the petition, the question and the request—do not announce anything.”45 Therefore the syllogism is not based on these kinds of statements. It is based only on “the fifth mode—the announcement46 […] where the decision and the judgment are found.”47 Hence, the concern of logic is logos apophantikos (or, in al-Fārābī’s terminology, qawl jāzim,48 what medieval Hebrew logic literature calls ʾomer gozer [lit. “a saying that cuts” or “judges”]),49 that is, a statement structured as a judgment concerning the nature of reality. This stems from the fact that the syllogism does not deal with the types of utterances that serve man in expressing himself in general, but with statements that derive unknown knowledge about reality (even if it is only a possible reality) from what is already known. And for the sake of the process of deriving knowledge from knowledge, one should rely on indicative propositions.

The words of the commentator cited by Qirqisānī make it clear that syllogism does not only sum up an inductive process (if this were the case then the process of inference would be superfluous); it also reveals hitherto unknown knowledge. In the words of the commentator: “And [Aristotle] found logic [al-maṭiq] to be the instrument that assists people in obtaining the knowledge they needed to obtain, and the proof for matters they wished to prove, so that they used it to know the unknown.”50

The commentator explains the importance of the two propositions joined by a middle term (the concept that connects the two premises of the syllogism), as well as the fact that from the moment the premises are determined, also the conclusion is determined. As the commentator says: “[E]ven though he [the one who formulates a syllogism] did not express it [the conclusion, explicitly]. [This is because] the two propositions supported each other at the moment they were joined.”51

The commentator’s interpretation of Aristotle’s syllogistics can be understood as a development of Aristotle’s notion in Analytica Priora 2.21, 67a35–38. According to Aristotle, a person may know that all mules are infertile but at the same time think that a certain beast is capable of begetting even though he knows that this beast is a mule, as long as he has not yet joined the two premises “all mules are infertile” and “that beast is a mule.” Knowledge of the two premises—of each one separately—cannot serve as the basis for the necessity of the conclusion. This can be explained in the following way: A person who knows that “all mules are barren” and also knows that “this beast is a mule” might well think that the statement “this beast can give birth” is true, that is to say, he does not know that the connection between the subject and the predicate in the statement “this beast is barren” is a necessary connection whose contradiction (“this beast can give birth”) is necessarily false. The joining of the two premises teaches that there is a necessary connection between the subject and the predicate in the syllogism’s conclusion.52

Moreover, according to Ernst Kapp, “the practical aim of the Prior Analytics, which Aristotle pretty strongly stresses, is not to show how one should advance to a conclusion from given premises, but rather to show how to find the premises necessary to demonstrate a given conclusion.”53 The words of the commentator cited by Qirqasānī are in line with this analysis of the Aristotelian syllogism:

If someone wishes to verify (ithbāt) the suitability of the nutriments to the temperaments,54 and says: “every [thing] that strengthens something—suits it.” Consequently he attaches to this proposition another proposition with a common denominator, and says: “the nutriments strengthen the temperaments”—the determination of these two explicit propositions, which are neither concealed nor refutable, verifies (thabāt) what he sought to prove—the suitability of the nutriments to the temperaments, even though he did not express it [i.e., the conclusion, explicitly]. [This is because] the two propositions supported each other at the moment they were joined.55

In other words, the aim of the syllogism is “to verify” (ithbāt) a given fact, or knowledge. That is to say, to prove why the nutriments are suitable to the temperaments, and by that to determine that it is indeed so. The syllogistic verification is thus: The nutriments are compatible with the temperaments, because the nutriments strengthen the temperaments, since it is known that every matter that strengthens something is compatible with it. This is how one “verifies what he sought to prove—the suitability of the nutriments to the temperaments”—meaning, the verification of an already known fact, or item of knowledge. According to this approach, the Aristotelian theory of inference does not, indeed, claim that the syllogism expresses knowledge not included in the premises. It suggests a way of disclosing the logical-metaphysical structure of reality, the structure by which the subject and predicate of a scientific claim (in our case, “the nutriments are compatible with the temperaments”) are necessarily connected (through the middle term—in our case, “whatever strengthens a thing”).

Kapp56 bases himself on the definition of the syllogism and analysis of its function in Aristotle’s Topica. The framework there is dialectic, meaning that the inquirer presents someone with a question formulated according to the pattern “Is B attributed to A, or not?” (meaning: “Is it true, or is it not true, that A is B?”), and the responder is expected to answer yes or no. Dependent on the standpoint of the responder, the inquirer suggests propositions—premises—that the responder is supposed to accept, and from which it follows that the responder’s standpoint is faulty. That is to say, in the context of a disputation, a syllogism is a discussion where certain claims are assumed as premises (which the inquirer suggests and the responder accepts), and from them something else follows—a proposition that contradicts the belief of the responder. But the inquirer knew this proposition from the outset, and the syllogism is only providing the basis for the already known proposition.

A similar approach can be found in al-Fārābī’s presentation of the syllogism in his Kitāb al-qiyās (Book on Syllogism; = Kitāb al-madkhal ʾilā al-qiyās [Book of Introduction to the Syllogism]), chapter 7:

The syllogism (wa-al-qiyās) is a statement in which at least two premises are set down,57 so that if they are joined together, something ensues from them, substantially and not accidentally, that necessarily is different from themselves. […] And the syllogism is constructed upon a certain conclusion (maṭlūb) that precedes [it], which he afterwards strives to prove through syllogism. And the conclusion is a part that contradicts [another part], and the two of them are connected [to each other] through the particle (lit. “letter”) of distinction, while there is added to both of them the particle of the question of factuality. And the particle of distinction is the particle “or” (aw) or its equivalent, and the particle of the question of factuality is the particle “is it true that” (hal)58 or its equivalent, like when we say: Is it true that every material substance moves, or does not every material substance move? And the conclusion may also be called “inquiry.”59

These words of al-Fārābī, where he characterizes the syllogism, express the disputational background of the Topica: the thing sought for (the conclusion) can be one of two mutually exclusive possibilities (either A, or not-A), and it is presented as a question (Is it A, or is it not-A?). This is the process that opens the dialectic of the Topica, and according to al-Fārābī this is the basis of the fact that the conclusion of the syllogism is known even before its premises have been formulated. This “dialectical” approach to the syllogism, found in al-Fārābī (and in Kapp), explains also the method of the unknown commentator: the proposition that is claimed in the syllogism’s conclusion indeed precedes the syllogism’s premises (hence, the theologian’s criticism is not valid). The syllogism’s function is to verify the conclusion, to prove it and determine it.60

From this point of view, the difference between syllogism on the one hand, and analogy and induction on the other, stands out, since in the case of the latter two, the new knowledge contained in the conclusion is an addition to the knowledge contained in the two premises. Let us, for example, examine a case of analogy discussed in Maimonides’s Treatise on Logic:

Major premise: The heavens are material, and a wall is material.

Minor premise: The wall is created through labor.

Conclusion: Also the heavens must have been created through labor.61

It is obvious that in this analogy the conclusion is based on an inference beyond the mere joining of the two premises. On the other hand, the conclusion of the Aristotelian syllogism consists in the statement that is expressed in the very joining of the two premises. It seems that it was to this explanation given by the commentator that Qirqisānī sought to draw our attention, because it contains an answer to the criticism of the theologian. Indirectly, this vindication of the effectiveness and necessity of the syllogism also constitutes a defense of Aristotelian logic in general.

As noted earlier, Qirqisānī applies the syllogistic method to the law and to legal exegesis of the Torah.62 This is why it was so important to Qirqisānī to show that the syllogism generates new knowledge and to explain the inherent logic of this type of argumentation. In other words, Qirqisānī’s main motivation for defending the necessity and usefulness of the Aristotelian syllogism derives from his legal outlook: the incentive to apply the human intellect and judgment in the realm of divine law. In the course of clarifying his legal outlook, Qirqisānī succeeded in explaining the meaning of the Aristotelian syllogism and refuting theological standpoints hostile to it.

Qirqisānī’s methods, concepts, and theological worldview are for the most part those of the kalām, but in his defense of the basic tool of human reason—logic, the very means by which, according to Qirqisānī, we are obligated by God to interpret the law and to expound it—Qirqisānī reveals himself to be a faylasūf (an Aristotelian philosopher). He defends Aristotelian logic, and more generally, he defends the scientific method of reason.

5 Appendix: Yaʿqūb al-Qirqisānī, Kitāb al-Anwār w al-Marāqib, Treatise 4, Chapter 5, pars. 1–6 (Nemoy, 354–357)

(354) The Fifth Chapter: Concerning the speech about the premises and the conclusions.63

This issue, which concerns the derivation64 of the truths and the knowledge of the demonstration,65 is of66 the determinations of logic.67 It has already been condemned68 by one of the theologians.69 He was of the opinion that it does not generate unknown knowledge. He said: If we claim that “every human being is a living being” and “every living being is a substance”—both of these premises being true—and then infer,70 on their basis, that “every human being is a substance”—then nothing that we did not [already] know was concluded,71 and we have not learned anything of which we were [previously] ignorant,72 since our knowledge that a human being is a substance preceded;73 and that which does not teach anything unknown cannot serve as proof 74 or demonstration.75 And since this is so, it has been verified76 that this matter [the syllogism] is of no use.77

[But] the person who holds these claims is mistaken in what he said, insofar as he related78 only (355) to a particular aspect79 of the matter, and [on this basis] he sought to weaken it in general.80 And those who said that, as presented by him,81 only used it82 as an example meant to verify their opinion.83

2. Let us now quote the words of one of the commentators84 concerning the reasons why the master of logic [Aristotle] saw fit to determine this matter.85 He said: [That] which aroused Aristotle to receive the logic which he founded, [is] that he saw that people determined [something]86 in their intellects [a thing]87 which is defective and in need of being perfected.88 And he saw that the matters are of two kinds: revealed and concealed. And that [kind] whose knowledge must be sought89 is that which is concealed. And with regard to what is revealed, [the very fact of] its being revealed is what makes it unnecessary to inquire concerning it and to seek its proof. And [Aristotle] knew that the only means to obtain knowledge concerning concealed matters is to bring evidence from revealed [matters]; that is, if by coincidence the matter is [both] concealed and there is [also] no revealed evidence concerning it, then the quest for knowledge about it shall cease. And [Aristotle] found logic90 to be the instrument91 that assists people in obtaining the knowledge they needed to obtain, and the proof for matters they wished to prove, as they used it to know the unknown, and to join separate [matters], and to draw remote [matters] nearer, and to distinguish by it between [matters] in agreement from matters in dispute.

3. And [Aristotle] found that the modes of speech are five: [one is]92 the imperative, like one who says: Do this and that; [another one]93 is the question, like someone who says: Where did you come from?; [another one] is the petition;94 and [another one] is the request, like one who says: Lend me money; and [another one] is the announcement, like one who says: This and that happened.95 And he found that four of these modes—the imperative, the petition,96 the question, and the request—do not announce anything.97 And he found that the fifth mode—the announcement—is where98 the decision and the judgment are found.99 And he found that the speech of decision100 can be confirmation, as when someone claims: So-and-so is writing; or denial, like when someone claims: So-and-so is not writing. And he named the confirmation “affirmative,” insofar as it affirms101 one thing to another; and he named the denial “negative,” since it negates one thing with respect to another.

4. After that, [Aristotle] observed and found that every proposition has two parts, the two extremes that encompass it.102 One of them is the name,103 and the other that which is carried by the name.104 And he called the name (356) “the subject105 extreme,” and that which is carried by the name he called “the predicate106 extreme.” And he found that the single proposition does not announce anything beyond itself and does not add knowledge to anyone, even if [the proposition] is completely true and clear.107 [For example], if someone says: All fire is hot—this proposition does not teach anything apart from the heat of fire, and there is no doubt that the one who hears it either knows that, or does not know that.108 And if he knows that, then he has obtained no [new] knowledge [from this proposition]; and if he does not know, then he should not accept this, and no one of his companions should force him to accept it without verifying it through proof.109

5. And [Aristotle] knew that the evidence brought concerning the concealed matter can only be completed110 by bringing clear and true, non-refutable, evidences, which when verified prove the hidden matter about which they testify.111 And when he found that a single proposition does not announce nor signify [anything] beyond itself, he realized that it112 does not suffice for this.113 And [Aristotle] contemplated the different propositions with no common denominator between them114 and found them to have the status of a single proposition, in the sense that [both of them] are of little avail,115 and equally in distance from announcing anything beyond themselves.116 [An example of this is] if someone says: “All fire is hot,” and “All snow is cold”—each one of these two propositions stands alone, without a common denominator with the other one,117 and does not testify about anything beyond itself.

6. And when [Aristotle] found neither the single proposition, nor the different propositions without a common denominator between them, to be sufficient, what remained to analyze was the attached propositions. And when he contemplated and analyzed these, he found what he had looked for118 and [found] peace in trust.119 [For example], if someone wishes to verify the suitability of the nutriments to the temperaments120 and says: “Every [thing] that strengthens something—suits it.” Consequently he attaches to this proposition another proposition with a common denominator,121 and says: “the nutriments strengthen the temperaments”—the determination of these two explicit propositions, which are neither concealed nor refutable, verifies what he sought to prove—the suitability of the nutriments to the temperaments, even though he did not express it [explicitly].122 [This is because] the two propositions supported each other at the moment they were joined, the attachment of the two [consisting in the fact that] one of them predicated suitability of everything which strengthens, while the other predicated strengthening-of-temperaments of the nutriments.

Afterwards [Aristotle] divided the attached propositions into categories,123 (357) on which he commented and elaborated, but there is no need for us to mention this here, [since] our only purpose124 has been to [clarify] the aforementioned reason of premises and conclusions, and that it is possible through them125 to draw conclusions concerning concealed, unknown matters.


I would like to thank Prof. Warren Z. Harvey, Prof. Haggai Ben-Shammai, Prof. Charles H. Manekin, Dr Tony Street, and Dr Joep Lameer for their useful comments. Unless otherwise noted, the original Arabic texts quoted below have been translated by me, and bracketed phrases in the translations indicate my own additions, which are intended to facilitate the reader’s understanding of the text.


L. Nemoy, ed., Kitāb al-Anwār wal-Marāqib (New York: Alexander Kohut Memorial Foundation, 1940), 343. Henceforth cited as Nemoy.


Georges (Yehuda Aryeh) Vajda invested much energy in his analysis of the major part of the fourth treatise. For his French translation (accompanied by critical comments) of the majority of the chapters of the fourth treatise, see Georges Vajda, “Études sur Qirqisānī,” Revue des Études Juives 108 (1948): 63–91; 120 (1961): 211–257; 122 (1963): 7–74. The chapter treated in the present study is discussed only briefly in the first article in this series (henceforth cited as Vajda 1948). Vajda did not translate most of the text that is analyzed here; for my translation, see the appendix.


On this leaning, which was extremely dominant among the Karaite “mourners of Zion” but not exclusive to them, see Hayim H. Ben-Sasson, On Jewish History in the Middle Ages [Hebrew] (Tel Aviv: Am Oved, 1962), 156–171; idem, “The Karaite Community of Jerusalem in the Tenth-Eleventh Centuries” [Hebrew], Shalem: Studies in the History of the Jews in Eretz Israel 2 (1976): 1–18. On the individualistic-rationalistic exegetical method of the Karaites and their rejection of tradition and the religious establishment of the geonic academies, see Meira Polliack, “The Emergence of Karaite Bible Exegesis” [Hebrew], Sefunot 22, n. s. 7 (1999): 299–311.


On Qirqisānī’s rationalistic approach, see Haggai Ben-Shammai, “The Doctrine of Religious Thought of Abu Yusuf Yaʿqub al-Qirqisani and Yefet ben ʿEli” [Hebrew] (PhD thesis, Hebrew University, 1977), part 1, pp. 8–35; Daniel J. Lasker, From Judah Hadasi to Elijah Bashyatchi: Studies in Late Medieval Karaite Philosophy (Leiden: Brill, 2008), 18–22. For an analysis of the words of Qirqisānī on the obligation of rational analysis and the use of inference, see also Georges Vajda, “Études sur Qirqisānī,” Revue des Études Juives 107 (1946–1947): 52–98.


Many a scholar has analyzed different aspects of this controversy while pointing out parallels in the Muslim world. For a renewed discussion of this issue, see Daniel Lasker, “The Use of Reason in Rabbanite and Karaite Legal Exegesis of the Bible.” Lasker’s article is due to be published in the near future. I thank Prof. Lasker for allowing me to read his article before its publication. For further discussion, see Joseph E. David, “Legal Comparability and Cultural Identity: The Case of Legal Reasoning in Jewish and Islamic Traditions,” Electronic Journal of Comparative Law 14 (2010): 1–24; Y. Erder, The Karaite Mourners of Zion and the Qumran Scrolls: On the History of an Alternative to Rabbinic Judaism (Turnhout: Brepols, 2018), 50–58; Moshe Zucker, “Fragments from Rav Saadya Gaon’s Commentary to the Pentateuch from MSS” [Hebrew], Sura 2 (1955–1956): 313–355; idem, “Fragments of the Kitāb taḥṣīl al-sharāʾiʿ al-samāʿīyyah” [Hebrew], Tarbitz 41 (1971–1972): 373–410.


For an analysis of Qirqisānī’s epistemological principles, see Aviram Ravitsky, “Logic and Karaite Legal Methodology: Hebrew Translation of Qirqisānī’s Kitāb al-Anwār wa-al-Marāqib, section 4, chapters 1–8 with Introduction and Critical Notes” [Hebrew], AJS Review 41 (2017): 2–8.


wa-min jalīlihi: See the doubts expressed by Vajda about the translation of this phrase (Vajda 1948, 69 n. 14), and cf. Joshua Blau, A Dictionary of Mediaeval Judaeo-Arabic Texts, ed. Yechiel Kara (Jerusalem: The Academy of the Hebrew Language and The Israel Academy of Sciences and Humanities, 2006), 91. Qirqisānī means that these are obvious examples in which a syllogistic method must be applied.


al-natījah: Lit. “the conclusion” or “the consequence.”


Nemoy, 358, par. 9.


The Aristotelian syllogism is an argument that is constructed of two premises and a conclusion. The premises contain three terms, one of which is found in both premises—the middle term, which forms the connection between the two premises—and the conclusion contains the other two terms as its subject and predicate. For example: All men are mortal (major premise); Socrates is a man (minor premise; “being a man” is the middle term); hence, Socrates (subject of the conclusion) is mortal (predicate of the conclusion). The major premise contains the predicate of the conclusion; the minor premise contains the subject of the conclusion.


For discussion of the function of the middle term in the legal thought of Qirqisānī, see Aviram Ravitsky, “Saʿadya Gaʾon and Yaʿqūb al-Qirqisānī on the Logical Structure of the Rational and Traditional Laws: Logic and Kalām in the Karaite-Rabbanite Controversy” [Hebrew], Tarbitz 84 (2016): 177–181.


Nemoy, 363–364, par. 4.




See George F. Hourani, Islamic Rationalism: The Ethics of Abd al-Jabbar (Oxford: Clarendon Press, 1971); idem, Reason and Tradition in Islamic Ethics (Cambridge: Cambridge University Press, 1985), 57–66; Majid Fakhry, Ethical Theories in Islam (Leiden: Brill, 1991), 31–45.


This is also the Muʿtazilite view concerning the function of the ʿillah; see Ahmad Hasan, “The Legal Cause in Islamic Jurisprudence: An Analysis of ʿillat al-ḥukm,” Islamic Studies 19 (1980): 247–270, esp. 250–251.


Nemoy, 354.


Ibid., par. 1.


On the criticism of logic in the medieval Muslim world, see Wael B. Hallaq, Ibn Taymiyya against the Greek Logicians (Oxford: Clarendon Press, 1993), 32–34.


As will be shown below, Qirqisānī’s quotation of “the commentator” is supposed to be an answer to the theologian’s criticism. The commentator, as will be explained in the article’s next section, does not argue that we know the proposition expressed in the syllogism’s conclusion only by its premises, and that this is why the syllogism is useful. Rather, the commentator claims that in an Aristotelian syllogism (or rather in an Aristotelian demonstration), the premises show the necessity of the conclusion—the necessity of the connection between the conclusion’s subject and its predicate. Hence, the theologian’s criticism should be understood also as denying the usefulness of the relation between the syllogism’s premises and its conclusion, and not just as an assertion that we know the conclusion without a syllogism. For an analysis of the opposition to the syllogism expressed by the theologian in the context of skeptical trends in Muslim theology of the Middle Ages, see Josef van Ess, “Skepticism in Islamic Religious Thought,” Al-Abḥāth 21 (1968): 13.


For example, says Sextus, most animals move their lower jaw, but the premise “all animals move their lower jaw” is false, since the crocodile moves its upper jaw. See Sextus Empiricus, Pyrroneion Hypotyposeon, book 2, ch. 14, in Sextus Empiricus, Outlines of Pyrrhonism, trans. R. G. Bury (London: Heinemann; Cambridge, MA: Harvard University Press, 1961), 277.


Ibid., 276–279. Sextus also criticizes the inferences of implication, conjunction, and disjunction in a way that resembles the basic structure of his criticism of the Aristotelian syllogism; see 278–283. On Sextus’s criticism of the syllogism, cf. Lenn E. Goodman, In Defense of Truth: A Pluralistic Approach (Amherst, NY: Humanity Books, 2001), 27–51.


See Arthur N. Prior, “Skeptical Criticisms of Syllogistic Reasoning,” The Encyclopedia of Philosophy, 8 vols. (New York: Collier Macmillan, 1972), 5:41–43.


John Stuart Mill, System of Logic: Ratiocination and Inductive (London: Longmans, Green and Co., 1941), 122, cf. 128. According to Mill: “All which man can observe are individual cases. From these all general truths must be drawn” (122, and see the discussion in 119–137). Therefore, in Mill’s opinion, syllogism is, in reality, an induction from many particulars to an additional particular (or particulars), and not from the general to the particular. See ibid., 122, 128.


Ibid., 120.


For the Arabic source, see Mubahat Türker, “Fārābī’nin Bazi Mantik Eserleri,” Ankara Üniversitesi Dil ve Tarih Coğraya: Fakültesi Dergisi (Revue de la Faculté de Langues, d’Histoire, et de Géographie de l’Université d’Ankara) 16 (1958): 244–286. For a Hebrew translation of Moses b. Samuel ibn Tibbon, see MS Paris, National Library, Heb. 917 (microfilm no. F 30335), 184a–210b. Nicholas Rescher thought that this work is al-Fārābī’s short commentary on Aristotle’s Analytica Priora. See Nicholas Rescher, Al-Fārābī’s Short Commentary on Aristotle’s “Prior Analytics” (Pittsburgh: University of Pittsburgh Press, 1963). But Abdelhamid Sabra showed that this work of al-Fārābī is an independent work dedicated to demonstrating that the different kinds of argument, such as induction, analogy, exemplification, and so on, can all be formulated in a syllogistic structure. See Abdelhamid I. Sabra, review of Al-Fārābī’s Short Commentary on Aristotle’s “Prior Analytics,” by Nicholas Rescher, Journal of the American Oriental Society 85 (1965): 241–242. For an analysis of this work by al-Fārābī, see also Kwame Gyekye, “Al-Fārābī on the Logic of the Arguments of the Muslim Philosophical Theologians,” Journal of the History of Philosophy 27 (1989): 135–143.


See Türker, “Fārābī’nin Bazi Mantik Eserleri,” 265 (ET: Rescher, Al-Fārābī’s Short Commentary, 90).


Hence, says al-Fārābī, if we wish to claim that “all motion takes place in time,” in order to prove, for example, that swimming takes place in time, we have to investigate the different types of motion (like walking, flying, etc.), in order to see whether they take place in time. If our investigation included swimming also, then we have recognized empirically that swimming takes place in time, and any inference “proving” this would be superfluous and circular. If, on the other hand, our investigation had not included swimming, the major premise “all motion takes place in time” would have been unjustified and therefore unfit to serve as a major premise of a syllogism. See Türker, “Fārābī’nin Bazi Mantik Eserleri,” 264–266, esp. 64–265 (ET: Rescher, Al-Fārābī’s Short Commentary, 88–92, esp. 90–92), and see also Gyekye, “Al-Fārābī on the Logic of the Arguments,” 140.


Türker, “Fārābī’nin Bazi Mantik Eserleri,” 266 (for an alternative English translation, see Rescher, Al-Fārābī’s Short Commentary, 92). What al-Fārābī means to say is that it is impossible to verify through induction, for example, the claim “every movement takes place in time,” so that it may function as the major premise in the syllogism: every movement takes place in time; swimming is a movement; hence it follows that swimming takes place in time. In this syllogism, swimming is “one of the things that are contained in the subject” of the premise (movement), and the syllogism affirms the predicate of the premise (“takes place in time”) with respect to swimming.


Türker, “Fārābī’nin Bazi Mantik Eserleri,” 250 (ET: Rescher, Al-Fārābī’s Short Commentary, 58–59). On these four sources of knowledge, see Deborah L. Black, Logic and Aristotle’s Rhetoric and Poetics in Medieval Arabic Philosophy (Leiden: Brill, 1990), 94–101, and see further Douglas M. Dunlop, “Al-Fārābī’s Introductory Sections on Logic,” Islamic Quarterly 2 (1955): 267 (ET: 275–276).


See n. 6 above. The stance taken by al-Fārābī is probably the key to understanding what Qirqisānī states immediately after his presentation of the words of the theologian. Qirqisānī says: “[But] the person who holds these claims is mistaken in what he said, insofar as he related only to a particular aspect [or: a part] of the matter, and [on this basis] he sought to weaken it in general [or: to weaken the whole].” See Nemoy, 354–355, par. 1. Qirqisānī’s intention is not clear from the context, but if we interpret it in light of al-Fārābī’s opinion, then his intention seems to be that the syllogism might indeed be based on induction, and such a syllogism would have to deal with skeptical criticism. But this is only a “part,” a unique case, which does not testify about “the general rule,” that is to say, about all kinds of syllogism. On the ways in which Qirqisānī refutes the argument of the theologian, see below.


While Aristotle’s Organon was available in the Arab world in the time of Qirqisānī, the writings and ideas of Sextus Empiricus were almost unknown in the Arab world, and they were certainly not known through Sextus’s own writings. For this reason alone it is difficult to claim that Sextus’s standpoint is the historical-textual source for the opinion of the theologian or for those of al-Fārābī (concerning induction). However, according to Josef van Ess, it is plausible to assume that Hellenistic skeptic influences infiltrated the Muslim world through unknown paths: “[The arguments of Hellenistic skeptics] entered in a more or less underground way—in the same way many items of Stoic logic for instance entered into kalām and into uṣūl al-fiqh” (van Ess, “Skepticism in Islamic Religious Thought,” 3, and see further 11). On the difficulty of determining the paths whereby knowledge was transferred from the Hellenistic to the Arab world, and for a criticism of the approach reflected by van Ess, see Dimitri Gutas, “Pre-Plotinian Philosophy in Arabic (Other than Platonism and Aristotelianism): A Review of the Sources,” Aufstieg und Niedergang der Römischen Welt 36 (1994), 4939–4973. On the question of the impact of the writings of the Hellenistic skeptics, and especially the ideas of Pyrrho, on the Arab world, see Michael Cook, Early Muslim Dogma (Cambridge: Cambridge University Press, 1981), 44–47. According to Harry Wolfson, the eleventh, twelfth, and thirteenth teachings dealt with by Saʿadya in his Kitāb al-Amānāt wa-al-iʿtiqādāt (Book of Beliefs and Opinions), treatise 1, express three Greek theories (those of Protagoras, Pyrrho, and Carneades), each of which represents a more or less skeptical approach. See Saʿadya Gaʾon, Ha-nivḥar be-ʾemunut u-ve-deʿot, ed. Yofef Qafiḥ (New York: Sura, Yeshiva University, 1970), 68–72; Harry A. Wolfson, Repercussions of the Kalam in Jewish Philosophy (Cambridge, MA: Harvard University Press, 1979), 160–162; and see further Joseph Stern, The Matter and Form of Maimonides’ “Guide” (Cambridge, MA: Harvard University Press, 2013), 146–147 n. 25.


For an analysis of Ibn Taymiyya’s principal objection to logic, see C. A. Qadir, “An Early Islamic Critique of Aristotelian Logic: Ibn Taimiyyah,” International Philosophical Quarterly 8 (1968): 511–512; Hallaq, Ibn Taymiyya, 14–34.


See ibid., 28–31, 33–34.


This understanding of the syllogism is similar to that of Mill. See n. 23 above.


Hallaq, Ibn Taymiyya, 35–37.


See Aristotle, Analytica Priora 1.1, 24b19–20; Topica 1.1, 100a25–26.


Alexander of Aphrodisias on Aristotle, “Prior Analytics” 1.1–7, trans. Jonathan Barnes, Susanne Bobzien, Kevin Flannery, S.J., and Katerina Ierodiakonou (London: Duckworth; Ithaca: Cornell University Press, 1991), 66. An example of a useless argumentation provided by Alexander is: If this is a day, then there is light. But this is a day, ergo: this is a day (ibid.). For a similar explanation, see Aram Topchyan, David the Invincible, Commentary on Aristotle’s “Prior Analytics”: Old Armenian Text with an English Translation, Introduction, and Notes (Leiden: Brill, 2010), 99.


These fragments were preserved in the philosophical anthology composed by Todros Todrosi in the fourteenth century. See Shalom Rosenberg and Charles H. Manekin, “Japheth in the Tents of Shem: Themistius’ Commentary on the Analytica Priora” [Hebrew], in Shlomo Pines Jubilee Volume, ed. Moshe Idel et al., 2 vols., Jerusalem Studies in Jewish Thought 9 (Jerusalem: Mandel Institute for Jewish Studies, 1990), 2:267–274. The commentary was translated into Arabic in the tenth century by Abu ʿUthmān Saʿīd ben Yaʿaqub ha-Damaski (ibid., 268).


In the Hebrew source: ha-davar. In footnote 39 to the English translation (see n. 42 below), the authors add “lit.: ‘thing.’”


In the Hebrew source: ein toʿelet.


In the Hebrew source: ki lo niqneh mehem davar neʿelam midevarim geluyim.


For the Hebrew source, see Rosenberg and Manekin, “Japheth in the Tents of Shem,” 269. For an English translation, see Shalom Rosenberg and Charles Manekin, “Themistius on Modal Logic: Excerpts from a Commentary on the Prior Analytics Attributed to Themistius,” Jerusalem Studies in Arabic and Islam 11 (1988): 93–94. See ibid., 85, 90, for a discussion of whether the attribution of these Hebrew fragments to Themistius’s lost commentary on the Prior Analytics is correct.


Nemoy, 355, par. 2. Vajda wrote, in the notes to his translation of Qirqisānī, that he was unable to identify Qirqisānī’s source. Vajda 1948, 68 n. 12. It is possible that the source of Qirqisānī’s quotation is found in one of the commentaries on Aristotle’s Analytica Priora, but this question is still in need of further clarification, primarily because many of the commentaries (especially those written in Arabic) that might have been available to Qirqisānī have been lost. On the commentaries on the Analytica Priora (that may have been available to Qirqisānī), see Rescher, Al-Fārābī’s Short Commentary, 31–34; Joep Lameer, Al-Fārābī and Aristotelian Syllogistics (Leiden: Brill, 1994), 5–9.


Nemoy, 357, par. 6.


Ibid., 355, par. 3.


Below, Qirqisānī calls “the announcement” kalām al-faṣl (the speech of decision).


Meaning that the decision and the judgment concerning truth and falsehood are relevant only with regard to the fifth mode. The quote is taken from the fifth chapter (Nemoy, 355, par. 3).


See Al-Farabi’s Commentary and Short Treatise on Aristotle’s “De Interpretatione,” trans. F. W. Zimmermann (London: Oxford University Press, 1981), 43 n. 4.


See Shalom Rosenberg, “Logic and Ontology in Jewish Philosophy in the 14th Century” [Hebrew] (PhD thesis, Hebrew University, 1973), 174–175, 187–189.


Nemoy, 355, par. 2.


Ibid., 356, par. 6.


To use another Aristotelian example (Analytica Posteriora, 2.16–17, 98a35–98b24): the proposition “vine leaves are deciduous” is explained and proved through the premises “vine leaves are broad” and “all broad leaves are deciduous.” The connection between vine leaves and deciduousness, namely the middle term in this syllogism is, therefore, the breadth of the leaves. In his discussion, Aristotle explains that “deciduousness” is the solidifying of sap at the point where the leaf joins the stalk, and that this solidifying takes place in broad leaves. This is why deciduousness is connected to the breadth of the leaves. For an analysis of Aristotle’s words in the context of his scientific approach, see Robert J. Hankinson, “Philosophy of Science,” in The Cambridge Companion to Aristotle (Cambridge: Cambridge University Press, 1995), 111–113.


Ernst Kapp, “Syllogistic,” in Articles on Aristotle, vol. 1, ed. J. Barnes et al. (London: Duckworth, 1975), 39.


See n. 120 below.


Nemoy, 356, par. 6.


Kapp, “Syllogistic,” 39–41.


Lit. “in which more than one thing is set down.”


Lit. “whether.”


Rafīq al-ʿAjam, Al-mantiq ʿind al-Fārābī, vol. 2 (Beirut: Dar al-mashriq, 1986), 19.


Cf. n. 109 below.


See Israel Efros, “Maimonides’ Treatise on Logic,” Proceedings of the American Academy for Jewish Research 8 (1938): 46. For the Arabic original, see idem, “Maimonides’ Arabic Treatise on Logic,” Proceedings of the American Academy for Jewish Research 34 (1966): 20 (Judeo-Arabic part).


Among modern scholars of Talmudic methodology it is possible to find the view that the syllogism does not generate new knowledge. These scholars emphasize the difference between inferential forms such as analogy and induction, on the one hand, which generate new knowledge, and syllogism on the other hand, and prefer to identify the Talmudic methods of inference not with syllogism, but with different kinds of analogy and induction. For an analysis of these trends, see Aviram Ravitsky, Logic and Talmudic Methodology: The Application of Aristotelian Logic in the Commentaries on the Methods of Jewish Legal Inference [Hebrew] (Jerusalem: Magnes Press, 2009), 17–18. Adolph Schwarz set out to examine the Talmudic a fortiori principle (qal va-ḥomer) as an Aristotelian syllogism. See Adolf Schwarz, Der hermeneutische Syllogismus in der talmudischen Literatur (Karlsruhe: Bielefeld, 1901); Hebrew version: Aryeh Schwarz, Midat Qal va-Ḥomer ba-Sifrut ha-Talmudit, trans. M. Berkowitz (Warsaw: Yavne, 1905). Although Qirqisānī, in his remarks on the syllogism, does not deal directly with the a fortiori principle, he understands the essence of halakhic methodology as being syllogistic, and in this his orientation agrees with that of Schwarz.


That is, the premises and the conclusions of the syllogism.


istikhrāj: Lit. “to take out, to extract.”


al-burhān: The demonstrative syllogism.


Or: “belongs to.”


That is, the syllogistic deduction belongs to the science of logic, and see the translation in Vajda 1948, 67–68. In the literature dealing with logic, the term burhān indicates a demonstrative syllogism, a demonstration that teaches scientific knowledge and whose conditions were discussed by Aristotle in the Analytica Posteriora. See, for example, Efros, “Maimonides’ Arabic Treatise on Logic,” 22. As is clear from the context below, Qirqisānī does not discuss here the content-related aspects of the burhān, the fact that it analyzes the scientific status of reality, but rather its formal aspect, which is the meaning of the relation between the conclusion and the premises.


That is, he found a flaw in the syllogistic structure of the burhān.


baʿḍ al-mutakallimīn: One of the scholars of the kalām.


fā-nantiju: Nemoy’s vocalization, assigning the verb to the first conjugation (faʿala). But in my opinion it is more plausible to read it as a form of the fourth conjugation (afʿala): fa-nuntiju.


fā-lam yachruj: It is also possible to read here a passive form: fā-lam yuchraj, lit. “was not taken out, extracted.”


Jāhilīn: I.e., of which we have no knowledge.


mutaqaddimah: I.e., preceded the syllogism, or the premises.


dalīl: I.e., what refers to, or serves as a guide to, knowledge or fact.




fa-qad ṣaḥḥa: “It was correct, true.”


laysa mi-mmā yaḥtāju ilayhi: Lit. “it is not something that one needs.”


Tanāwala: Lit. “he was engaged.”


Or: “a part.”


Or: “to weaken the whole.” For a possible explanation of these words, see n. 30 above.


That is, according to what was mentioned by that theologian. In Nemoy’s edition the text reads: ʿalā mā dhakarūhu, “according to that which they [the logicians] mentioned.” My translation follows the correction in Vajda 1948, 68 n. 11.


I.e., what they had said.


Lit. “in order to verify what they aimed at.” The last sentence in Qirqisānī’s words is not entirely clear. My translation follows Vajda 1948, 68. If my interpretation of Qirqisānī’s words (n. 30 above) is correct, his intention in the last sentence might be expressed thus: even if the theologian is right in claiming that the specific inference discussed by him is based on induction, and that it is true that the conclusion of this particular syllogism (“every human being is a substance”) is known prior to the premises and serves as their basis (and in the opinion of the commentator whose words will be brought below, which is also the opinion of Qirqisānī himself, the theologian is wrong in this), this still does not refute the necessity and usefulness of the syllogism in general, since the premises of the syllogism are not always based on induction. According to Qirqisānī, those theologians who criticized the syllogism used only one example (which concerns only one type of syllogism) in order to “verify” (meaning in this context “to prove rhetorically”) their true intention, and that is a general refutation of Aristotelian logic (or philosophy). Nevertheless, as mentioned above, it has to be admitted that Qirqisānī’s words are not entirely clear.


baʿḍ al-mufassirīn.


hadhā al-qawl: Lit. “this saying, that speech.”


The parentheses are original in Nemoy’s edition.


The addition is in Nemoy’s edition.


Lit. “[something which is] in a place of defection, and is in need for an increase, or complement.”


Lit. “that which needs to ask its knowledge.”


al-maṭiq: The art of logic.


al-ʾālah: It is possible that the term reflects the notion of logic as “organon.”


minhā: Lit. “from them.”


wa-minhā: Lit. “and from them.”


Nemoy points out a manuscript variant that may be translated as “the vow” (it is also possible to translate “the warning”).


The Arabic terms employed by Qirqisānī are al-ʾamr (the imperative), al-suʾāl (the question), al-nidāʾ (the petition), al-masʾalah (the request), and al-khabr (the announcement). The logical distinction between a statement that can be judged in terms of truth and falsehood and a statement that cannot be judged in these terms belongs to Aristotle. See De Interpretatione, 4.17a, 3–7. Aristotle clarifies there that the subject of the De Interpretatione is only indicative statements that are true or false, but not the rest of the types of statements that resemble prayer in so far as they cannot be judged as true or false. The different types of statements have been analyzed by medieval grammarians and logicians, and the distinctions of the commentator brought here are also found in the writings of Qirqisānī’s contemporaries, namely al-Fārābī and Saʿadya Gaʾon. See Al-Farabi’s Commentary and Short Treatise on Aristotle’s “De Interpretatione,” 43–44, 226–227 (and see the addition to p. 10, n. 5, on p. 254); Saʿadya Gaʾon, Ha-Egron: Kitāb ʾuṣūl al-Shiʿr al-ʿibrānī, ed. Neḥemya Allony (Jerusalem: The Academy of the Hebrew Language, 1969), part 1, pp. 154–155, and see also 76–77. And see further Rabbi Jonah ibn Janaḥ, Sefer ha-Riqmah (Kitāb al-lumaʿ), trans. Rabbi Judah ibn Tibbon, ed. Michael Wilensky, 2nd ed., vol. 1 (Jerusalem: The Academy of the Hebrew Language, 1964), 35.


Nemoy points out a manuscript variant that may be translated as “the vow” (it is also possible to translate it as “the warning”).


That is, they do not constitute an indicative description of any objective condition; they are not statements.


Thus according to the variant reading cited by Nemoy in his notes. I prefer this reading over the reading in Nemoy’s text itself, according to which the translation should be “the fifth mode […] is the decision and the judgment.”


That is, judgment and decision concerning truth and falsehood are relevant only to the fifth mode.


kalām al-faṣl: A reference to the aforementioned “announcement.”


In the sense that it binds one thing to another, that is, it links the subject and the predicate in a positive connection. The opposite is true concerning the negative proposition.


The idea he intends to convey is that the proposition is placed between its two extremes, which, so to speak, encompass it.




mā yuḥmal.






Lit.: The utmost of truthfulness and clarity.


Alternative translation: “that the listener knows this” (i.e., that fire is hot), or (lit.): “he will be ignorant [concerning this fact].”


It is possible that these words express the dialectics of the Topica, where the syllogism is presented as a kind of argumentation in which the inquirer sets forth the premises, which the responder then accepts as being true, and which prove above all doubt that the proposition of the inquirer is true while the proposition of the responder is wrong. But it is also possible that what is reflected here is not the method of the Athenian Academy, but rather the kalāmic method of disputation and the ilzām technique, namely, forcing the opponent to accept the truth of a standpoint that is opposed to his own opinion. On dialectics in the Athenian academy, see Gilbert Ryle, “Dialectic in the Academy,” in Aristotle on Dialectic, ed. G. E. L. Owen (Oxford: Clarendon Press, 1968), 69–79, and cf. Paul Slomkowski, Aristotle’s Topics (Leiden: Brill, 1997), 12. On the polemic style of arguing that shaped the kalām, see Josef van Ess, “The Logical Structure of Islamic Theology,” in Logic in Classical Islamic Culture, ed. G. E. von Grunebaum (Wiesbaden: Harrassowitz, 1970), 21–50.


lā yaṣiḥḥu. It is also possible to translate: “it will not be true.”


aladhī stushhida bihā ʿalayhi: I have understood the verb (ʾustushhida) in the passive voice (lit.: “on which it has been testified upon”), but translated it in the active voice.


The one proposition.


“This”: I.e., what is more than the one proposition, the hidden thing that is in need of proof. The Arabic source reads: “It (the hidden thing that is in need of proof), cannot be sufficed by one proposition.” I altered the order of the sentence in my translation.


ghayr al-mutaʿāwinah: Lit.: “which do not support each other.”


fī qilat al-ghanāʾ. It is also possible to translate the phrase thus: “to be in lack of sufficiency,” the sense being that it does not suffice in order to generate unknown knowledge.


That is, two propositions that are not linked through a middle term resemble a single proposition in that no conclusion can be inferred from either of them, and in that both of them are far from announcing any knowledge beyond themselves.


ghayr muʿāwinah: Lit.: “without supporting another.”


bi-al-bughyah: It is also possible to translate the phrase thus: “he found the object, the goal.”


Lit.: “and found rest in it, on [being] secured.” That is, by observing the joined propositions Aristotle reached the goal that he sought for, seeing that this supplies reliable information that satisfies him.


According to the theory of the four biles (humors), nutrition has an influence on the relations between the biles and their temperaments, and therefore also on health and sickness, which depend on them. On the theory of the four biles, see Noga Arikha, Passions and Tempers: A History of the Humours (New York: Ecco, 2007), and see also Rabbi Moses ben Maimon, (Medical) Aphorisms of Moses, translated into Hebrew by R. Nathan Hameathi, ed. Suessmann Muntner (Jerusalem: Mosad ha-Rav Kook, 1961), treatise 2, pp. 32–38. For examples of the influence of foodstuffs on the biles, see ch. 20, pp. 227–242.


Lit., “that supports the other.”


I.e., even though that “someone” did not explicitly express the conclusion that he desired to verify, it is verified through the very joining of the two propositions.


The quotation from the words of “the commentator” appears to end here.


In citing the commentator.


I.e., by means of the premises.

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 263 37 0
Full Text Views 108 30 2
PDF Views & Downloads 50 22 6