Hostname: page-component-7d684dbfc8-dh8xm Total loading time: 0 Render date: 2023-09-30T17:01:56.068Z Has data issue: false Feature Flags: { "corePageComponentGetUserInfoFromSharedSession": true, "coreDisableEcommerce": false, "coreDisableSocialShare": false, "coreDisableEcommerceForArticlePurchase": false, "coreDisableEcommerceForBookPurchase": false, "coreDisableEcommerceForElementPurchase": false, "coreUseNewShare": true, "useRatesEcommerce": true } hasContentIssue false


Published online by Cambridge University Press:  02 February 2016

Yehuda Halper*
Department of Jewish Philosophy, Jacobovitz Building (1002), Bar Ilan University, Ramat Gan, Israel


While Averroes’ work is often considered to represent the culmination of the method of Aristotelian demonstration in Arabic philosophy, a short passage of his Long Commentary on Aristotle's Metaphysics Γ.2 emphasizes the prominence of dialectic and calls for a re-examination of dialectic and demonstration in Averroes’ philosophical works. In this passage Averroes describes dialectic as an acceptable form of philosophy and the dialectician as a kind of scientist. In putting dialectic and demonstration on an equal, or nearly equal footing, Averroes seems to go against his own account of the dialectical and demonstrative classes of people in the Decisive Treatise. Moreover, this interpretation of Metaphysics Γ.2 also contradicts Averroes’ explanation of the same passage in the Middle Commentary on the Metaphysics as well as Aristotle's own description of dialectic throughout the Metaphysics. That is, in the Long Commentary on the Metaphysics, Averroes departs from his earlier views, and describes dialectic as a necessary part of metaphysics, even though the centrality of dialectic argumentation could call into question the entire project of metaphysics and consequently of the sciences whose demonstrations rely on metaphysical ground, i.e., all sciences. Averroes does not emphasize this view, but its presence is nevertheless unambiguous.


Si l’œuvre d'Averroès est souvent considérée comme étant le sommet de la méthode démonstrative aristotélicienne dans la philosophie arabe, un bref passage de son Grand commentaire à Métaphysique Γ.2 d'Aristote souligne l'importance primordiale de la dialectique et suggère d'examiner à nouveaux frais le statut de la dialectique et de la démonstration dans l’œuvre philosophique d'Averroès. Dans ce passage, Averroès décrit la dialectique comme une forme acceptable de philosophie et le dialecticien comme un certain type de scientifique. En mettant la dialectique et la démonstration à un niveau égal ou presque, Averroès semble aller contre sa propre conception d'une distinction entre des types d'hommes dits “dialectiques” et “démonstratifs”, qu'il développe dans son Traité décisif. En outre, cette interprétation de Métaphysique Γ.2 contredit aussi son explication du même passage dans le Commentaire moyen sur la Métaphysique, et la description donnée par Aristote lui-même de la dialectique tout au long de la Métaphysique, à savoir que, dans le Grand commentaire à Métaphysique, il s’écarte de ses vues antérieures en décrivant la dialectique comme une partie nécessaire de la métaphysique, même si la centralité de l'argumentation dialectique pourrait remettre en question le projet de la métaphysique tout entier et, par là, celui des sciences dont les démonstrations s'appuient sur le terrain métaphysique, c'est-à-dire toutes les sciences. Averroès ne souligne pas cette vue, mais sa présence n'en est pas moins ambiguë.

Research Article
Copyright © Cambridge University Press 2016 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)


1 See, e.g., Endress, Gerhard, “The language of demonstration: translating science and the formation of terminology in Arabic philosophy and science,” Early Science and Medicine, 7 (2002): 231–54CrossRefGoogle Scholar and Gutas, Dimitri, “Certainty, doubt, error: comments on the epistemological foundations of medieval Arabic science,” Early Science and Medicine, 7 (2002): 276–89CrossRefGoogle Scholar. Roshdi Rashed has explored the role of mathematical demonstrations on grounding philosophy and science in, e.g., Between philosophy and mathematics: examples of interactions in classical Islam,” Islam and Science, 3 (2005): 153–65Google Scholar. Another view emphasizes that natural science and philosophy were not as well founded in the Islamic world as the principles and science of fiqh and that metaphysics in particular was not well grounded in a consistent epistemology. See, e.g., Gustave E. von Grunebaum, “Relations of philosophy and science: a general view,” in George Hourani (ed.), Essays on Islamic Philosophy and Science (Albany, 1975), pp. 1–4.

2 This oft mentioned division appears twice in the Decisive Treatise Determining the Connection between the Law and Wisdom, at paragraphs 11 and 44 (ff.). See Averroes, Decisive Treatise and Epistle Dedicatory, trans. Charles Butterworth (Provo, Utah, 2001), pp. 8 and 26 ff.

3 E.g., ibid., paragraph 47, p. 27.

4 The passage in questions is found in Averroes, Tafsir ma ba‘d at-tabi‘at, ed. Maurice Bouyges (Beirut, 1938–42), vol. 1, pp. 324–30. It is commenting on Aristotle, Metaphysics, 1004b16–26.

5 Averroes’ Middle Commentary survives on in two medieval Hebrew translations, edited in Il Commento medio di Averroe alla Metafisica di Aristotele nella tradizione ebraica: Edizione delle versioni ebraiche medievali di Zeraḥyah Ḥen e di Qalonymos ben Qalonymos con introduzione storica e filologica (Averroes's Middle Commentary on Aristotle's Metaphysics in the Hebrew Tradition. Edition of the Medieval Hebrew versions by Zeraḥyah Ḥen and Qalonymos ben Qalonymos, together with a historical and philological introduction), by Mauro Zonta (Pavia, 2011), vol. 2. The passage in question occurs on pp. 34–5 of tom. 1 of vol. 2.

6 Aristotle's explicit mentions of dialectics or dialecticians in the Metaphysics are rather few: Greek words with that root appear only 18 times. 8 of these mentions occur in Book Γ (6 appearing in the passage in question [1004b16–26], the other two at 1007a20 and 1012b7); 2 in Book A (987b32 and 989b33); 1 in Book B (995b23); 4 in Book K (1061b8, 1062b11, 1063b11, and 1066b10); and 2 in Book M (1078a29 and b25). I discuss the more significant of these in section II below. Of course, this does not rule out the possibility that other passages of the Metaphysics have dialectic in mind.

7 Usṭāth's translation is preserved as lemmata to the Arabic text of Averroes’ Long Commentary on the Metaphysics. See Tafsir ma ba‘d at-tabi‘at, ed. Bouyges, pp. 324–5, Lemma 5, paragraphs e-l.

8 Richard Taylor, “Averroes on the Sharia of the philosophers,” in Richard Taylor and Irfan Omar (eds.), The Judeo-Christian Islamic Heritage: Philosophical and Theological Perspectives (Milwaukee, 2012), pp. 283–304.

9 This account of assent has a parallel in Abū Naṣr al-Fārābī's Book of Demonstration, where he describes three forms of assent: certitude, which he relates to demonstration, dialectical assent and rhetorical assent. The latter two are inferior approximations of the first. See Al-Manṭiq ʿinda al-Fārābī, ed. Rafīq al-ʿAjam and Majid Fakhry (Beirut, 1986–87), vol. 4, chap. 1, p. 20. However, in other places al-Fārābī distinguishes between absolute certainty and other less than absolute forms of certainty. See Black, Deborah, “Knowledge (ʿilm) and certitude (yaqīn) in al-Fārābī's epistemology,” Arabic Sciences and Philosophy, 16 (2006): 1145CrossRefGoogle Scholar, esp. 36 ff., where she discusses the Book of Demonstration. On the goal of science as “the attainment of perfect conceptualization and assent” in Ibn Bājja, a major influence on Averroes, and the roots of this view of assent in al-Fārābī, see Forcada, Miquel, “Ibn Bājja on taṣawwur and taṣdīq: science and psychology,” Arabic Sciences and Philosophy, 24 (2014): 103–26CrossRefGoogle Scholar.

10 Reason and Revelation in the Middle Ages (New York, 1938), p. 43.

11 Cf., e.g., Taylor, Richard, “‘Truth does not contradict truth’: Averroes and the unity of truth,” Topoi, 19 (2000): 316CrossRefGoogle Scholar and “Averroes on the Sharīʿah of the philosophers,” in Richard Taylor and Irfan Omar (eds.), The Judeo-Christian-Islamic Heritage: Philosophical and Theological Perspectives (Milwaukee, 2012), pp. 283–304, esp. 285–91. See also Thérèse-Anne Druart, “Averroes on God's knowledge of being qua being,” in Paul Lockey (ed.), Studies in Thomistic Theology (Houston, 1995), pp. 175–205. Majid Fakhry ends his Averroes (Ibn Rushd): His Life Works and Influence (Oxford, 2001; repr. 2008), by connecting several statements in the Decisive Treatise to statements in Aristotle's Metaphysics Λ (p. 164).

12 On the legal form of the Decisive Treatise as a fatwā, see Heller-Roazen, DanielPhilosophy before the Law: Averroes's Decisive Treatise,” Critical Inquiry, 32 (2006): 412–42Google Scholar, esp. 419 ff. Heller-Roazen notes that the legal syllogism is determined by Averroes to be obligatory (p. 422) and that there is a kind of “unity of demonstration and legal doctrine” (p. 426), yet it is less clear that the whole of philosophy is deemed obligatory (this is true even if the whole of philosophy is contemplating existing things using the most complete demonstration, as at Decisive Treatise, pp. 2–3, paragraph 3). On the one hand, Heller-Roazen seems to suggest that Averroes assimilates philosophy to logic, implying that demonstration is an adequate description of philosophy (e.g., pp. 432–3, esp. n. 35). Yet, he also tells us that “wisdom,” which presumably includes philosophy, is beyond the grasp of the Law: “Wisdom as such remains outside the purview of the Treatise, and it therefore remains possible that the juridical tract has not truly seized hold of the pagan science” (p. 434).

13 Averroes, Decisive Treatise and Epistle Dedicatory, p. 1. On the ambiguity of what is meant by “philosophy” (falsafa) in Averroes’ Decisive Treatise, see Heller-Roazen “Philosophy before the Law,” pp. 434–6. See also previous note.

14 Cf. Fārābī, “A Treatise about the intentions of the Metaphysics”: “As for universal science, it reflects on what is common to all existents, such as being and unity … Also on the origin common to all existents, and that is the thing which should be referred to by the name of God, may He be exalted and may His names be sanctified” (unpublished translation by Aryeh Leo Motzkin). Nevertheless, it is not clear to what extent metaphysics, in Averroes’ view, sets out to study God rather than to study truth as such, which arrives, indirectly, at a study of God toward the end of the Metaphysics (Book Λ). See Harvey, Steven, “The quiddity of philosophy according to Averroes and Falaquera, a Muslim philosopher and his Jewish interpreter,” Miscellanea Mediaevalia, 26 (1998): 904–7Google Scholar. Harvey compares the definition of philosophy according to the Decisive Treatise with how Averroes proceeds philosophically in his commentaries.

15 Averroes, Decisive Treatise and Epistle Dedicatory, p. 26:

16 Paragraph 11, Butterworth trans., p. 8.

17 Paragraph 44. See Butterworth trans. p. 26.

18 Note that Averroes’ statements in the rest of paragraph 44 that the demonstrative classes are so by nature and art and that the dialectical classes are so by nature or habit in context imply only that the Law recognizes them to be such.

19 Aristotle, Metaphysics. 1004b16–26. Hippocrates Apostle trans. (Bloomington, 1970), p. 57. Translation modified.

20 It is also possible that Aristotle alludes here to the discussion of capacity in Plato's account of discussing being in the Sophist. Note also that Aristotle's contention that sophists and dialecticians put on the same appearance as the philosopher may refer to the beginning of the Sophist, where Socrates says that philosophers sometimes put on the appearance of statesmen, sophists, or madmen (216c-d). If Aristotle is introducing the dialectician into this group of appearance makers, we must wonder how the dialectician differs from the sophist, philosopher, statesman, or madman.

21 Hugh Tredennick (Loeb Classical Library [Cambridge, Mass. and London,1968], p. 157) translates the term, πειραστικὴ, “treats as an exercise,” which he contrasts with the “trying to understand” (γνωριστική) of the philosophers. William D. Ross (in The Complete Works of Aristotle, ed. Jonathan Barnes [Princeton, 1984], vol. 2, p. 1586) gives “merely critical” for πειραστικὴ, similarly implying a kind of futility to the efforts of the dialectician. However, Ross gives a more nuanced look at the word πειραστικὴ in his edition of the Metaphysics (London, 1924), where he notes the difficulty of deciding what the term might mean (see pp. 260–1). Sachs (Santa Fe, NM, 2002) translates πειραστικὴ as “tentative,” i.e., “tentative about those things which philosophy seeks to know.” I have used the word “striving” to translate πειραστικὴ because it reflects the word's root of πειράω, to strive or try. Moreover, “striving” works with all of the other meanings sketched above, while preserving the ambiguity of what precisely the dialectician strives for.

22 See Tafsir ma ba‘d at-tabi‘at, p. 325, Lemma 5, paragraphs e-l. Note also that of the other places where Aristotle mentions dialectic or dialecticians (in Books A, B, Γ, K and M), we only have access to two in Arabic translation (those in Books B and Γ). Two of these places, in Books K and M, probably were translated into Arabic since Averroes comments on them in his Middle Commentary on the Metaphysics. The text from Book A happens to be missing at that point.

23 While it is not entirely clear why Usṭāth uses the word “labor” (taʿab) here, it seems to convey the sense of “project” or “intellectual endeavor” in much the same way the sciences in Arabic are often referred to as “crafts” or “arts” (ṣināʿāt)

24 For a discussion of the issues facing 9th and 10th century translators of Greek philosophy and Usṭāth's place among these translators see Gerhard Endress, “The circle of al-Kindi: early Arabic translations from the Greek and the rise of Islamic philosophy,” in Gerhard Endress and Remke Kruk (eds.), The Ancient Tradition in Christian and Islamic Hellenism: Studies on the Transmission of Greek Philosophy and Sciences (Leiden, 1997), pp. 43–76. See also the sources mentioned in note 1.

25 Since the Arabic of the Middle Commentary on the Metaphysics is not extant, the text printed here is from the Hebrew translation of Qalonimos b. Qalonimos, printed in Averroes, Il Commento medio, ed. Zonta, vol. 2, tom. 1, pp. 34–5. Another translation by Zeraḥyah Ḥen, which does not differ in any significant way from the Qalonimos translation of this passage, is included in Zonta's edition. Both translations strove to be as literal as possible and can therefore be relied upon for our purposes here. On the reliability and usefulness of the Hebrew translations of Averroes, especially those of Qalonimos, see Steven Harvey, “Are the medieval Hebrew translations of Averroes’ Commentaries on Aristotle still of value and worth editing?” in Aafke M. I. van Oppenraay (ed.), The Letter before the Spirit: the Importance of Text Editions for the Study of the Reception of Aristotle (Leiden and Boston, 2012), pp. 95–210.

26 Or: always appear to be the same. Zeraḥyah’s translation has no equivalent of לעולם.

27 Or: or [philosophy is different from] the imagined science.

28 Both Qalonimos’ translation (tigboret) and Zeraḥyah's translation (niṣaḥon) have roughly the same meaning.

29 Cf. Averroes’ Short Commentary on Aristotle's Topics, par. 21 in Averroes’ Three Short Commentaries on Aristotle's “Topics,” “Rhetoric,” and “Poetics,” ed. Charles Butterworth (Albany, 1977), pp. 54–5, where Averroes refers to two disputants training their dialectical abilities and compares such training to fencing. Averroes then ends this Short Commentary by stating, “Training like this seems unnecessary for the perfection of the demonstrative arts. But if it were, without a doubt, it would be from the standpoint of the most excellent.” This statement certainly reinforces the distinction between dialectics and demonstration and seems to accord with Averroes’ view of dialectics in the Middle Commentary on the Metaphysics.

30 There is, however, no outside evidence for this claim. On possible sources available to Qalonimos and Zeraḥyah see Zonta, Il Commento Medio, vol. 1, pp. 45 ff. and the secondary sources cited therein.

31 This is not, though, entirely clear. Accordingly, I suspect there may be a transmission problem here in the Hebrew.

32 See Bouyges, Tafsir, Comm. 5, paragraphs e-l, pp. 328–30. The paragraph letters in this paragraph correspond to the superscript letters in Usṭāth's translation quoted above.

33 In the context of this article, I do not differentiate between the Arabic terms هوية and موجود, both which I translate “being.” On some of the difficulties in translating these terms see my Revision and standardization of Hebrew philosophical terminology in the fourteenth century: the example of Averroes’ Long Commentary on Aristotle’s Metaphysics Δ,” Aleph, 13 (2013): 95137CrossRefGoogle Scholar.

34 The Hebrew ha-ḥaqirah (“examining”) is generally used to translate al-fakhṣ.

35 On “attributes” as a possible meaning of maʿna, see Frank, Richard, “Al-ma‘nà: some reflections on the technical meanings of the term in the Kalâm and its use in the Physics of Mu‘ammar,” Journal of the American Oriental Society, 87 (1967): 248–59CrossRefGoogle Scholar. I discuss Averroes’ use of maʿnā for a form that is “about” something or in relation (bi-nisba) to something in my “Averroes on intentionality and the human experience of the natural world,” in Ahmed Alwishah and Josh Hayes (eds.), Aristotle and the Arabic Tradition (Cambridge, 2015), pp. 164–76.

36 See, e.g., Posterior Analytics I 7, where Aristotle argues that demonstrations made of one genus cannot be applied to another genus. If dialectics concerns one genus, “absolute being,” then there is a science of that one being. Of course, it would presumably be studied through demonstrations, not dialectical argument. Averroes’ definition of science at this point in the Long Commentary does not, however, seem to rely on demonstration. One may well ask, then, if a non-demonstrative, dialectical science such as that which emerges from the ensuing paragraphs, would still be confined to examining one genus.

37 This would seem to conform to Fārābī's differentiation of the “true philosopher” from false, vain, and counterfeit philosophers at the end of the Attainment of Happiness, paragraphs 57–64 (see Muhsin Mahdi's Translation, Alfarabi's Philosophy of Plato and Aristotle [Ithaca, NY, 1969], pp. 46–50). Among the many (perhaps not directly philosophic) jobs assigned to the true philosopher (including ruler, legislator, and imām) is that of making images and persuasive arguments (par. 59). That is, the philosopher seems to take on the jobs of poet and sophist. According to Fārābī, “the images and persuasive arguments are intended for others,” i.e., for non-philosophers, “whereas, so far as he is concerned these things are certain” (ibid.). That is, “true philosophy” encompasses arguments that are not strictly speaking philosophical, though they may refer to or relate to things that are grasped by the philosopher. For Averroes, in this vein, “true philosophy” would refer to demonstration, while other activities of the philosopher, including dialectics, would be philosophy only by extension. However, in neither Averroes nor Fārābī, are “persuasive arguments” clearly the domain of the dialectician rather than of the sophist.

38 Cf. Aristotle, Topics 101b3–4: dialectic “holds the path toward the principles of all inquiries.”

39 Averroes, Tafsir, ed. Bouyges, Book B, C1.a, pp. 166–7.

40 At least, this seems to be the argument at the beginning of Metaphysics Γ.2. Aristotle speaks of one principle in relation to which (πρὸς μίαν ἀρχήν, 1003b5-6) things are said to be. Cf. Averroes, Middle and Long Commentaries ad loc.

41 “This is Aristotle's approach in all of the sciences…” Tafsir, p. 167.

42 Such critiques occur in, e.g., Metaphysics A and De Anima A.

43 Averroes’ reference to “this book” (hādhā al-kitāb) implies that he has in mind the Metaphysics as a whole. Averroes consistently calls each of the “books” that make up the Metaphysics a “treatise” maqāla.

44 Tafsir, p. 167.

45 Averroes says, “And because it is the nature of dialectical statements to go before demonstrative statements, Aristotle saw fit to make this kind of theorization a separate part of this science, placed at the front of [the Metaphysics], just as he did regarding the terms that indicate the meanings (al-maʿānī) that are taken up in this science. These are the terms that generally concern all the meanings that are taken up in this science. He made this a separate part at the beginning; not like he did with the other sciences. I mean, he expounds the terms that indicate each of the meanings that are sought in those sciences in the theorization about each thing sought … However, in this science he placed the examination of all the terms in one treatise [viz. Book Δ] not to distinguish the equivocal terms from each other, but to enumerate the meanings about which this science is to theorize.” Tafsir, pp. 167–8.

46 “This treatise (i.e., Book Γ) includes a discussion of … the way of theorizing in this science … and of the solution to the questions included in it.” Tafsir, p. 297. Presumably, these “questions” include the questions mentioned in Book B.

47 Tafsir, pp. 302–3.

48 Tafsir, p. 308, par. u.

49 Tafsir, pp. 302 ff., esp. par. d, k, and x.

50 Charles Butterworth questions whether demonstrations can ever be made. Regarding the short commentaries on the logical Organon, Butterworth understands Averroes to believe that “the art of demonstration appears more to be something yet to be attained than it does an art that has been perfected.” This conclusion, though, is not based on anything Averroes says, but rather on Averroes’ silence about how to reach the first principles of demonstrations. See “Finding first principles, possibility or impasse?”, in Rüdiger Arnzen and Jörn Thielmann (eds.), Words, Texts and Concepts Cruising the Mediterranean Sea (Leuven, Paris, and Dudley, MA, 2004), pp. 211–22, esp. 220–2. The passage I bring here from Metaphysics Γ raises the question of the dialectical grounding of demonstration directly and, while it does not answer the question, does not require argumentum ex silentio.

51 Apostle translation, p. 55.

52 Il Commento Medio, vol. 2, tom. 1, p. 31 (155b22). Both translations have the same expression here; both obscure the “philosopher” in favor of “philosophy.”

53 Averroes uses the general term for knowledge, maʿrifa, not the term for scientific knowledge, ʿilm.

54 Tafsir, p. 309, par. x.

55 Cf. Shem Tob ibn Falaquera's description of studying psychology so as to know oneself and thereby to attain knowledge of God in his Book of the Soul, quoted in Raphael Jospe, Torah and Sophia: The Life and Thought of Shem to Ibn Falaquera (Cincinnati, 1988), p. 321, see also pp. 321–50.

56 Tafsir, p. 7, par. f-g, translated by Adamson, Peter in “Yaḥyā ibn ͑Adī and Averroes on Metaphysics Alpha Elatton,” Documenti e studi sulla tradizione filosofica medieval, 21 (2010): 343–74Google Scholar.

57 Compare this statement with Averroes’ description of Aristotle in his Prooemium to the Long Commentary on the Physics: “[Aristotle] is the one who originated … the art of logic, natural science, and divine science, and it is he who completed them. … No one who has come after him to this our time – and this is close to fifteen hundred years later – has been able to add a word worthy of attention to what he said.” Translated in Harvey, Steven, “The Hebrew translation of Averroes’ Prooemium to his Long Commentary on Aristotle's Physics,” Proceedings of the American Academy for Jewish Research 52 (1985): 5584, p. 83CrossRefGoogle Scholar. Averroes’ statement of Aristotle's accomplishment is considerably stronger in the Long Commentary on the Physics, than in the Long Commentary on the Metaphysics. Even though the statement in the former appears to include metaphysics (“the divine science”), it is in the latter statement that Averroes suggests human limitations.