Author's note: Readers familiar with the contemporary secondary literature may wonder why thisarticle—focused so on the review essay cited in n. 1, below—does not appear in the journal that published the original review essay. The principal reason is that the Journal of the American Oriental Society allows no replies, even though the editors permitted the review essay to conclude by demanding one. Another is that, although speaking to issues raised there, I go beyond them to broader questions of interest to scholars generally. I am very grateful to the editorial referees at IJMES for the professional, thoughtful reading given my manuscript and for their helpful suggestions.

¹ See Gutas, Dimitri, “On Translating Averroes' Commentaries,” Journal of the American Oriental Society 110 (1990): 93, column b (henceforth 93b)CrossRefGoogle Scholar. The emphasis is in the original.

² See “On Translating,” 93b–94a; again, the emphasis is in the original. The reference is to Averroes' Middle Commentary on Aristotle's Poetics, trans, (with intro. and notes) Butterworth, Charles E. (Princeton, N.J.: Princeton University Press, 1986), xGoogle Scholar. In context, the passage within single quotation marks reads:

In the introduction, I explore some of the questions that need to be addressed for a more accurate grasp of Averroes' argument and elucidate the basic themes of his interpretation. For example, aware that Averroes considers the art of poetry to be part of the art of logic, I try to explain why he adopts such a position and search for corroboration of it in Aristotle's own writing.

³ See “Introduction,” Averroes' Middle Commentary on Aristotle's Poetics, 49Google Scholar; also ibid., 14 and the references to Averroes' text. The citation that follows is from “On Translating,” 94a.

⁴ For indications of how Averroes diverges from Aristotle when it suits his purposes, see Averroes' Middle Commentaries on Aristotle's Categories and De Interpretatione, trans, (with notes and intro.) Butterworth, Charles E. (Princeton, N.J.: Princeton University Press, 1983), 91–92Google Scholar. That such innovations were consciously used by philosophers within the medieval Arabic tradition for various doctrinal reasons is now generally accepted among most scholars. Galston, Miriam S., for example, shows in a recently published article that al-Farabi does not portray Aristotle as saying rhetoric and poetry belong to the art of logic in his Philosophy of Aristotle even though he attributes this notion to Aristotle in his Enumeration of the SciencesGoogle Scholar; see “Al-Farabi et la logique aristotélicienne dans la philosophie islamique,” in Aristote aujourdhui, ed. Sinaceur, M. A. (Paris: Erès, 1988), 202–6, 208–10Google Scholar. Another article relevant in this regard is Druart's, Thérèse-Anne “Al-Farabi and Emanationism,” in Studies in Medieval Philosophy, ed. Wippel, John F. (Washington, D.C.: The Catholic University of America Press, 1987), 42–43Google Scholar. See also Black, Deborah L., Logic and Aristotle's Rhetoric and Poetics in Medieval Arabic Philosophy (Leiden: E. J. Brill, 1990), 1–16, esp. 8–13Google Scholar.

⁵ See Dahiyat, Ismail M., Avicenna's Commentary on the Poetics of Aristotle, A Critical Study with an Annotated Translation of the Text (Leiden: E. J. Brill, 1974), 18Google Scholar; see also ibid., 20 and 28. A review of the work that examines these questions at greater length is to be found in The Middle East Journal 30 (1976): 576–77Google Scholar.

⁶ “On Translating,” 93a; for what follows, see ibid., 94b–97b and 95a.

⁷ The original texts are to be found in Averroes, Talkhīṣ Kitāb al-Shiʿr, ed. Butterworth, Charles E. and Haridi, Ahmad (Cairo: GEBO, 1986), par. 20Google Scholar; and Yunus al-Qunnaʿi, Abū Bishr Mattā Ibn, Kitāb Arisṭūaṭālīs ft al-Shiʿr, in Arisṭūṭālīs Fann al-Shiʿr, ed. Badawī, ʿAbd al-Raḥmān (Cairo: Maktabat al-Nahḍa al-Miṣriyya, 1953), 96:24–27Google Scholar.

⁸ The conjectural text differs from Abu Bishr Matta's only in the last six words. For Abu Bishr Matta's wa tunaqql wa tunẓifu [sic, “On Translating,” 95a; but the Arabic should be read as tunaẓẓifu, there being no fourth form of the verb] alladhīna yanfaʿilūna, the critic proposes that Averroes understood wa yanqā wa yanẓufu alladhīna yafʿalūna (96a). As he translates the texts, this corresponds in English to Abu Bishr Matta's “and it [sc. the craft of poetry of praise] purifies and cleanses those who suffer] being understood as “while those who act are pure and clean.” Readers of Arabic will note that the critic needs to read tunaẓẓifu in order to translate “cleanses” causitively and defies all rules of grammar in his translation of the text he presents as Averroes' understanding of Mattā.

For Abu Bishr Matta's yanfaʿilūna, the critic also suggests reading yafʿalūna and rendering that as “[those who] act.” The text would make better sense and not need to be emended were yanfaʿilūna simply rendered in its more obvious meaning as “[those who] are affected,” that is, affected by the passions of pity or compassion and fear.

The “apparently slightly modified form” of Abu Bishr Matta's text, as it “was understood by [Averroes],” is proposed here by the critic to help him make sense of Averroes' commentary on this passage: wa dhālika bi-mā yukhayyal fi al-fāḍilīna min al-naqāʿi wa al-naẓāfa, which he translates as “through the images of purity and cleanliness which are evoked [as existing] in virtuous men” (94b; the brackets are his addition). A direct and forceful version, one that gives a sense of poetry's unique power to move those who hear it by giving an affective imitation of what others do, is: “it does this by imitating the purity and immaculateness of the virtuous”; see Averroes' Middle Commentary on Aristotle's Poetics, par. 20.

⁹ Though the critic refers to Dahiyat's translation here, as well as to the accompanying footnotes, he apparently prefers his own rendition of the passage; see “On Translating,” 96b, n. 18.

¹⁰ For example, S. A. Bonebakker and J. C. Bürgel; see ibid., 99a.

¹¹ ibid., 92.

¹² For the critic's translation, see ibid., 94b; the square brackets are his. The other version is from par. 20 of Averroes' Middle Commentary on Aristotle's Poetics.

¹³ Although “mercy” is intended to indicate how Averroes reads Aristotle from a Muslimperspective, nothing warrants such a prejudgment. Nor does it advance our understanding of the text.

¹⁴ See “On Translating,” 97b. For the purposes of his subsequent remarks, the critic reproduces the passage in the following manner, acknowledging the marks of emphasis as his own:

Aristotle and Averroes infer the constituent parts of poetry[!] from their understanding of what tragedy or[!] eulogy is supposed to do, namely, to represent a complete or whole virtuous[!] action in speech that is both metrical and harmonious.

The sentence in question occurs on p. 20 of Averroes' Middle Commentary of Aristotle's Poetics.

¹⁵ “On Translating,” 97b; emphasis in the original. For what follows, see ibid., 94b and also n. 7.

¹⁶ See Lucas, D. W., Aristotle, Poetics: Introduction, Commentary, and Appendices (Oxford: Clarendon Press, 1968), 96Google Scholar. As the context suggests, Lucas understands mimēsis here as “imitation”; prattontes can be rendered as “doers of actions.”

¹⁷ See ibid., 118. The translation is mine. There is no argument about translating aretē as virtue or excellence. Although a simple craft is at issue, the linkage of aretē and ho spoudaios is important. Clearly, ho spoudaios means the good man—one excellent, virtuous, or skilled in the important human tasks. And so the question turns on whether spoudaios as an adjective relating to human actions can be understood as referring to such concerns or merely to mean something ponderous. Here, at any rate, itis a bit thick to claim that a “serious shoe” results from the “shoemaking excellence” exercised by “a good shoemaker.”

¹⁸ See Nicomachean Ethics 1166^a13; again, the translation is my own. The context emphasizes that the good man (ho spoudaios) is the man having both practical wisdom or intellectual excellence for action and moral virtue. In lieu of this passage, Lucas cites Nicomachean Ethics 1177^a3 where Aristotle uses spoudaios to illustrate virtue (aretē) in action:

The life of happiness is thought to be what is according to virtue. It comes about by what is good [spoudēs], not by what is childish. We say that good things [to spoudaia] are better than funny and childish things and that the activity of a part or of a man is better the more it is in accord with goodness [spoudaioteran].

Clearly, each of these occurrences of spoudaios could be translated as “serious” only if “serious” were understood as “lofty” or “noble,” that is, as something virtuous.

¹⁹ For a fuller explanation of these thoughts and indication of how they might be applied, see Butterworth, Charles E., “Review of F. W. Zimmermann, Al-Farabi's Commentary and Short Treatise on Aristotle's ‘De Interpretation,’” The Muslim World 78 (1988): 149–50Google Scholar; and “An Account of Recent Scholarship in Medieval Philosophy,” Interpretation 16 (1988): 87–97Google Scholar. Consider also the following remarks from the preface to the translation of Averroes' Middle Commentary on Aristotle's Poetics, xii–xiii:

I strive here for both literalness and eloquence, as is only fitting for a text on the art of poetry, but I have willingly sacrificed eloquence for literalness when the choice appeared inevitable. To the extent feasible, I have used the same English word for the same Arabic word throughout the translation and have always noted significant exceptions to this rule …. My goal has been to present a readable yet faithful English translation of Averroes' treatise. In keeping with that goal, I have alerted the reader to problems via footnotes while avoiding interpretative translations that conceal the problems without resolving them.

²⁰ Gutas, Dimitri, “Review of Charles Généquand, Ibn Rushd's Metaphysics: A Translation, with Introduction, oflbn Rushd's Commentary on Aristotle's Metaphysics, Book Lam,” Der Islam 64 (1987): 123Google Scholar.

²¹ See “On Translating,” p. 92; also 93a and 98a.

²² See ibid., 99a. At 98a, n. 21, the critic refers to Heinrichs, Wolfhart, Arabische Dichtung undgriechische Poetik:Ḥāzim al-Qarṭāǧannī's Grundlegung der Poetik mit Hilfe aristotelischer Begriffe, Beiruter Texte und Studien, Vol. 8 (Beirut: Imprimerie Catholique, 1969), 149–54Google Scholar, in urging that takhyīl be translated as Vorstellungsevokation. Also useful for this issue is Heinrichs's, “Die antikeVerkn ü pfung von Phantasia und Dichtung bei den Arabern,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 128 (1978): 252–98Google Scholar.

It should be noted, however, that the salient question here is not the use the translator did or did not make of the secondary literature, but what takhyīl (and thus mimēsis) means and what single word best renders its meaning. Citations from other scholars, no matter how numerous, are not equivalent to reasoned argument.

²³ See Averroes' Middle Commentary on Aristotle's Poetics, 63, n. 18. The reference to paragraph 2 in the first sentence of the footnote, a typographical error, should be to paragraph 3.

²⁴ In this respect, D. W. Lucas, Aristotle, Poetics, app. I, “Mimesis” is especially informative. Acknowledging that “the stock translation of mimēsis is ‘imitation’” (258), Lucas nonetheless goes on to urge (259) that:

The word mimesis has an extraordinary width of meaning, which makes it difficult to discover just what the Greeks had in mind when they used the word to describe what it is that the poet and artist do. To translate it we need in different contexts “imitate,” “represent”, “indicate,” “suggest,” “express.” All of these can be referred to the single notion of making or doing something which resembles something else.

Then, having traced the different ways mimēsis is used in 4th-century Greek writings and considered various other possible translations, Lucas pays special attention to Koller's, HermannMimesis in der Antike (Bern: A. Francke, 1954)Google Scholar. Arguing that the term mimēsis eventually goes back to the ritual dancer, mimos. Koller urges that the “primary meaning of mimeisthai is not ‘copy’ or ‘imitate’ but ‘give expression’” (see Lucas, Aristotle, Poetics, 270–71).

Although he does not provide page references, Lucas seems to have Roller's pp. 46, 104–6, 110, and esp. 39 in mind. Roller's contention is that to translate mim“0113;sis as “Nachahmung” (“imitation”) is too narrow and sometimes even misleading, especially with respect to dance; see ibid., 18 and 210. He urges, instead, that mim ēsis be rendered as “Darstellung” (“expression” or, even better, “representation”). In a schematic diagram of the development of the terms mimēsis and mimēisthai on p. 120, Roller urges that “Darstellung” best captures the theoretical use of the term, whereas “Nachahmung” better captures the way it is used in everyday speech.

Lucas's final judgment, based on his examination of the various texts cited by Roller in defense of his argument, is that in most instances “it cannot be said that the conventional meaning ‘imitate’ is impossible.” Indeed, he is willing only to say “it must be granted that there are passages where Roller's rendering is neater” (Lucas, Aristotle, Poetics, 271). That judgment, applied to the case at hand, makes eminently good sense: there are instances where another rendering—even “evoking images”—would be “neater.” On the whole, however, and it is the whole that is at issue, “imitation” captures Averroes' meaning much more clearly. And it provides an important link with Aristotle's Poetics, the text on which he is commenting. For example, even though it is implied in “imitation,” “evoking images” would simply not work when rendering takhyīl in par. 77 of Averroes' Middle Commentary on Aristotle's Poetics.

²⁵ See al-Fārābī, , Risāla fi Qawānīn Ṣināʿat al-Shuʿarāʿ, in Arisṭūṭālīs, Fann al-Shiʿr, ed. Badawī, ʿAbd al-Raḥmān, 150:3–9, 15–16, 151:7–8, 15, 155:10–14, 156:1–3, and 158:2–3;Google Scholar also al-Fārābī, , Kiātab al-Shiʿr, ed. Mahdi, Muhsin, in ShiƬr 12 (1959) 91:4–6, 92:3–4, 12–17, 93:7–10 ffGoogle Scholar. These are the two works on which Heinrichs grounds his first discussion of takhyīl in Arabische Dichtung and from which he departs when he seeks to determine how al-Farabi uses this term and its cognates by examining virtually all of the “second teacher's” writings in the article “Die antike Verknupfung.”Google Scholar

The reason for the different use of muḥākā and takhyīl by al-Farabi and Averroes may go back to Abu Bishr Matta's frequent rendition of mimēsis by both muḥākā and tashbīh (see Heinrichs, Arabische Dichtung, 121, 146) or arise from the different understanding each of these philosophers formed of Aristotle's text. Whatever the reason, it is clear that al-Farabi's terminology cannot be used to explain Averroes without first considering how Averroes understands poetry—somewhat as the translator sought to do in footnotes like the one just mentioned.

Though the difference between the two philosophers is the decisive issue here, it should also be noted that Heinrichs is unable to preserve his novel terminology in translating passages from al-Farabi—especially when he has to broaden his net to include taṣawwur (see “Die antike Verknüpfung,” 283–84, 288, 290 and nn. 112 and 114, 294). And, as he acknowledges, not all of the texts support this interpretation of al-Farabi's terminology (see ibid., 285–86, nn. 101–4, esp. n. 104 and the two lines coming before those cited).

²⁶ See Averroes' Middle Commentary on Aristotle's Poetics, 76, n. 19; and “On Translating,” 98a.

The reference is to the text in par. 22 of the edition and translation.

²⁷ See Butterworth, trans., Averroes' Middle Commentaries on Aristotle's Categories and De Interpretation, xx. Pascal's assessment of the human condition, with its implications for learning, seems especially apt in this context: “notre état véritable … nous rend incapables de savoir certainement et d'ignorer absolument”; see Pensées (Paris: Gamier Freres, 1958), 90, no. 72Google Scholar. So, too, does Nietzsche's reflection in his letter to Overbeck of 23 February 1887: “Zuletzt geht mein Misstrauen jetzt bis zur Frage, ob Geschichte überhaupt möglich ist?”

²⁸ See “On Translating,” 98b, and Cantarino's, V.Arabic Poetics in the Golden Age: Selection of Texts accompanied by a Preliminary Study, Studies in Arabic Literature, Supplements to the Journal of Arabic Literature, Vol. IV (Leiden: E. J. Brill, 1975)Google Scholar. Although he does not cite the subtitle of Cantarino's work, the critic makes a special point of mentioning pp. 70–99 for “a discussion of Averroes' work in the context of the development of Aristotelian poetics in Arabic” and pp. 177–90 for the translation of the text.

Despite a lengthy denunciation of the translator for not making more of an earlier article on Averroes and poetry by Cantarino, the critic fails to note that the discussion on pp. 70–99 of the Arabic Poetics in the Golden Age volume is simply a reworking of the article; see Cantarino, V., “Averroes on Poetry,” in Islam and its Cultural Divergence: Studies in Honor of Gustave E. von Grunebaum, ed. Tikku, Girdhari L. (Urbana: University of Illinois Press, 1971), 10–26Google Scholar; also “On Translating,” 99a–100 and nn. 22– 23. The following remarks about Cantarino's translation in the Arabic Poetics in the Golden Age volume explain the translator's complaint in the preface (ix) that Cantarino's “imaginative rewriting of Averroes' phrases to fit his own preconceptions has gone unchallenged for more than fifteen years.”

In passing, it should be noted that Heinrichs's appreciation of the merit of Cantarino's work is more in keeping with that of the translator than that of the critic; see “Die antike Verkniipfung,” 263, n. 31; 264, n. 37Google Scholar.

²⁹ See Arabic Poetics in the Golden Age, 184Google Scholar; the translation moves abruptly from what Cantarino calls “Section II” (Averroes' Middle Commentary on Aristotle's Poetics, chap. 3, i.e., pars. 13–19Google Scholar) to what he calls “Section IV” (ibid., chap. 5, i.e., pars. 33–47).

³⁰ In the “Averroes on Poetry” article, Cantarino displays the same kind of hesitancy; see pp. 14–18Google Scholar.

³¹ The Arabic passage is: fi al-umür al-irādiyya—aʿnī al-ḥusna wa al-qabīḥa. See Averroes, Talkhīṣ Kitāb al-Shiʿr and Averroes' Middle Commentary on Aristotle's Poetics, par. 3, with Cantarino, Arabic Poetics, 177. Again, the “Averroes on Poetry” article (14, 18–19) shows a similar confusion about these terms. The moral judgment inherent in these terms does not escape Heinrichs; see Arabische Dichtung, 161, n. 4.

³² See Averroes' Middle Commentary on Aristotle's Poetics, par. 8Google Scholar, with Cantarino, , Arabic Poetics, 180Google Scholar.

³³ See Cantarino, , Arabic Poetics, 83–84, 92Google Scholar with Averroes' Middle Commentary on Aristotle's Poetics, par. 65; see idem, “Averroes on Poetry,” 20.

³⁴ In contrast to such unwarranted assertiveness, it is instructive to consider the observation of an older, more thoughtful philologist that “in der Wissenschaft haben die Uberzeugungen kein Biirgerrecht, so sagt man mit guten Grunde”; see Nietzsche, F., Fröhliche Wissenschaft, Aphorism 344Google Scholar. See also Jenseits von Gut und Böse, Aphorisms 224 and 227; and Genealogie der Moral, III.23Google Scholar.

³⁵ The Latin phrase is: “et dum criminationes falsas contemnimus refutare, videamur crimen agnoscere.”