Abstract
Neither in the Guide of the Perplexed nor in his other philosophical and legal writings does Maimonides lay out a systematic theory of language. In this respect the subject of language differs from other subjects such as prophecy which Maimonides does treat in explicit systematic detail and, in this respect, Maimonides also differs from his Muslim counterparts, especially Al-Farabi and Avicenna, who present elaborate accounts of language. At the same time, it hardly needs to be repeated that the philosophical problems and conceptual confusions to which language gives rise are topics that occupy Maimonides throughout his writings. The absence of an explicitly worked-out theory of language should not, then, be necessarily understood to mean that Maimonides was not working with a theory of language.
For comments on earlier drafts of this paper, I wish to thank Moshe Halbertal, Warren Zev Harvey, Joel Kraemer, Yoram Navon, A.l. Sabra, Bernard Septimus, Sarah Stroumsa, Raymond Weiss, and especially Jonathan Malino. A slightly earlier version of the major parts of Parts III and VI appears in my “Logical Syntax as a Key to a Secret of the Guide of the Perplexed” (in Heb.), Ivyun 38 (April 1989): 137–166.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
Reference
This assumption would have been standard given the medieval philosophical curriculum in which logic, as the tool of philosophy, was studied before philosophy proper. Note also that the study of logic meant the study of a closed set of fixed texts: the various works of Aristotle and their Hellenistic and Arabic commentaries.
Guide, Dedication, 3; all references to the Guide of the Perplexed are to the S. Pines translation (Chicago, University of Chicago Press, 1963). In-text parenthetic references are to Part, Chapter, and Page.
See A. Marx, Jewish Quarterly Review (N.S. XXV): 378–80; translated in Pines’ Introduction to his translation of the Guide, ix.
BT Rosh Hashanah 32a; BT Sanhedrin I9a.
See Gershom Scholem, “The Meaning of the Torah in Jewish Mysticism,’ in his On the Kabbalah and Its Symbolism (New York, Schocken, 1965) and his ”The Name of God and the Linguistic Theory of the Kabbalah,“ Diogenes, no. 79–80 (1972), 57–80, 164–94.
Harry Austryn Wolfson, “The Veracity of Scripture from Philo to Spinoza,” reprinted in his Religious Philosophy: A Group of Essays (Cambridge, Harvard University Press, 1961), pp. 217–245.
Judah Halevi, Ku_ari 1:55; for further discussion of this objection, see below, Part IV.
Ibid. IV:25; cf. also Nahmanides, Commentary on the Torah, c. on Gen. ii:19.
On Halevi and earlier rabbinic divine language theories, see now Bernard Septimus, Maimonides on Language, forthcoming, especially Part Ill and references therein; and Josef Stern, “Language,” in Arthur A. Cohen and Paul Mendes-Flohr, eds., Contemporary Jewish Religious Thought (New York, Scribner’s, 1987), 543–551.
On doctrines of the pre-or untreated Word in medieval Judaism, see Harry Austryn Wolfson, Repercussions of the Kalam in Jewish Philosophy (Cambridge, Harvard University Press, 1979), p. 111f.
In 1:66, among other passages, Maimonides goes still further, identifying the divine and the natural (as opposed to the artifactual). Thus to say that the Torah is “created,” “the work of God,” “the writing of God,” or “the word of God” (1:66:160) is to say that “its existence is natural and not artificial”; (I:66:160). Compare also Guide 111:32:525 and, for the sense in which the Law is divine, namely, natural, 11:39, 11:40 and 111:27; see especially 11:40:382 on the relation between the natural and the conventional in Law. See also below, Part IV, on the parallel between law and language in Al-Farabi. Note that Maimonides uses terms of language and speech, e.g. ‘speaking’ or ‘saying,’ to express the social and political perfection of the Law because he takes the primary function of (external) language and speech to be socio-political; see e.g. his remark that “speech may only be addressed to an existent that receives the command in question” (I:65:159).
Note Maimonides’ shift from the name of the scriptural book, Bereshit/Genesis, to the rabbinic term, ma‘aseh bereshit, which he interprets as (Aristotelian) natural science.
In the same vein, Maimonides explains that, wherever Scripture states in Genesis that “God named something thus, this [is] in order to differentiate between the particular notion envisaged and the notion equally signified by the term” (11:30:351). I.e. the literary device of divine naming refers to natural processes of differentiation. On this passage, compare Shalom Rosenberg, “The Interpretation of the Torah in the Guide,” (Heb.) Jerusalem Studies in Jewish Thought 1/1 (Î981):85–157, pp. 131–2.
Maimonides, Treatise on the Art of Logic, ch. XIV, edited and introduced by Israel Efros, “Maimonides’ Arabic Treatise on Logic,” Proceedings of the American Academy of Jewish Research 34 (1966); Israel Efros, trans., “Maimonides’ Treatise on Logic,” Proceedings of the American Academy of Jewish Research (1938):1–65. (Except where noted I have relied on Efros’ English translation which, it should be noted, was made before the full Arabic and Judaeo-Arabic text was available.) The passage cited in the text appears almost verbatim in “Al-Farabi’s Introductory Riscilah on Logic,” ed. and trans. by D.M. Dunlop, The Islamic Quarterly, 111:4 (January 1957): 224235.
See his discussion of the Platonic, Epicurean, and Aristotelian uses of these terms in “The Veracity of Scripture.”
Apart from these three questions, Maimonides discusses other questions which bear on divine predication concerning the manner of an expression’s signification - whether it signifies univocally, equivocally, amphibolously, metaphorically, etc. - in both the Logic and Part I of the Guide. Apart from the theory of negative attributes, I shall not discuss these topics in this paper. For discussions, see, e.g. H.A. Wolfson, “The Aristotelian Predicables and Maimonides’ Division of Attributes,” in his Studies in the History of Philosophy and Religion, ed. I. Twersky and G.H. Williams (Cambridge, Mass., Harvard University Press, 1977), pp. 161–195; Shalom Rosenberg, “Signification of Names in Medieval Jewish Logic” (Heb.), Iyyun 27, 1976–7: 105–144; and now Arthur Hyman, “Maimonides on Religious Language,” in Norbert M. Samuelson, ed., Studies in Jewish Philosophy (Lanhem, MD, University Press of America, 1987). For a valuable overview of all these issues, see Shalom Rosenberg’s seminal dissertation, Logic and Ontology in Jewish Philosophy in the Fourteenth Century (unpublished Hebrew University Ph.D. thesis, 1973), especially pp. 121–130 and 164–197.
As will become clear in our subsequent discussion, the medieval use of the term ‘logic’ included considerably more than what we would nowadays classify under it, and more too than the Aristotelian science of the syllogism. “Theory of meaning” comes closer to capturing its true scope, though ‘meaning’ should include the content of thought more generally rather than linguistic meaning in a narrow sense only.
I owe the quoted phrase to A.I. Sabra, “Avicenna on the Subject Matter of Logic,” Journal of Philosophy (Nov. 1980): 746–764; p. 747.
See Al-Farabi, Enumeration of the Sciences; presumably Maimonides follows Al-Farabi on this, although he nowhere puts it in writing. Compare his negative comment in Guide I:67:162 about the knowledge of his contemporaries of “the science of our language,” i.e. the grammar of (biblical) Hebrew, and his criticism of the grammarian Menahem b. Saruk in his Commentary on the Mishnah, Terumah 1:1 (ed. and trans., Joseph Kafih, Jerusalem, Mossad Harav Kook, 1968), vol. 1, p. 167.
For further discussion, see Sabra, ibid.; Muhsin Mandi, “Science, Philosophy, and Religion in Alfarabi’s Enumeration of the Sciences,” in J.E. Murdoch and E.D. Sylla, eds, The Cultural Context of Medieval Learning (Dordrecht-Holland, Reidel, 1975), 113–147, especially 118–124; F.W. Zimmermann (trans., with Introduction and Notes), Al-Farabi’s Commentary and Short Treatise on Aristotle’s De Interpretatione (London, Oxford University Press, 1981), p. cxviii ff.
On the controversy between the grammarians and logicians, see D.S. Margoliouth, “The Discussion between Abu Bishr Matta and Abu Said al-Sirafi on the Merits of Logic and Grammar,” The Journal of the Royal Asiatic Society, 1905, pp. 79–129; Muhsin Mandi, “Language and Logic in Classical Islam,” in G.E. von Grunebaum, ed., Logic in Classical Islamic Culture (Weisbaden, 1970): 51–83; Zimmermann, p. cxxiiff; Joel Kraemer, Humanism in the Renaissance of Islam: The Cultural Revival during the Buyid Age (Leiden, Brill, 1986), pp. 104–116; idem., Philosophy in the Renaissance of Islam: Abu Sulayman al-Sijistani and his Circle (Leiden, Brill, 1986), pp. 14–165; and Kwame Gyekye, “Al-Farabi on the Logic of the Arguments of the Muslim Philosophical Theologians,” Journal of the History of Philosophy, XXVII,1 (January 1989): 135143.
See also the beginning of Logic, Ch. XIII for further differences between grammar and logic.
For this approach to interpreting Maimonides, I am very much indebted to Zimmermann’s interpretation of Al-Farabi; cf. his Introduction, pp. xxxviii-xlviii.
Cf. Dunlop, “Al-Farabi’s Introductory Risàlah on Logic,” for the original on which Maimonides’ virtually verbatim paraphrase is based.
Cf. Maimonides’ Shemonah Peragim, Ch. H.
Al-Farabi, Ihsa’al-’ulûm [Enumeration of the Sciences], ‘Uthmân Amin, ed. (Cairo: Maktabat alKhanji, 1931), pp. 17/18; cited in Sabra, p. 762.
On the notion of a rule, see Mandi, “Science, Philosophy, and Religion in Alfarabi’s Enumeration of the Sciences,” p. 120.
Cf. Dunlop, “Introductory Risalah,” p. 232f.
Cf. Zimmermann, pp. xliiif and xlvi-xlviii.
Cf. Ihsà’n al-‘ulùm 54.2, p. 23; cited in Dunlop, “Introductory Risàlah,” p. 226, n. 3. and in Zimmermann, p. xliii.
See Zimmermann, pp. xxiv-xlviii.
Cf. Zimmermann, xliii; Sabra, 748. It should be emphasized that Al-Farabi himself never describes his conception of logic in these terms, i.e. as a universal or mental grammar.
On Maimonides’ use of the phrase “rules common to all languages,” see also his reference in his Commentary on the Mishnah, Terumah 1:1 to “the perfect scholars who discourse concerning general matters common to all languages” (vol. 1, p. 167), a reference presumably to Al-Farabi. 4 Cf. Solomon Munk’s comment in his French translation Le Guide des Égarés (Paris, 1856), 3 vol., vol. 1, p. 237, n. 1.
Cf. I:31:67; 1:53:119.
I1:8:432; the choice of name’Adamites’ for humans in this context alludes to the dictum, “The Torah speaks in the language of the sons of man [henei adam],” which is repeatedly used by Maimonides to describe the limitations of scriptural language. On Maimonides’ use of this dictum, see my “Language” and further below.
Elsewhere, i.e. where there is independent evidence that Maimonides held a restricted view of human knowledge, such qualifying claims might indeed be read in that limiting sense. Cf. e.g., Guide II1:51, 53.
See Shlomo Pines, “The Limits of Human Knowledge according to Al-Farabi, Ibn Bajja and Maimonides,” in I. Twersky, ed., Studies in Medieval Jewish History and Literature, vol. 1 (Cambridge: Harvard University Press, 1979); “Lés Limites de la Métaphysique selon Al-Farabi, ibn Bajja et Maimonides; Source et Antitheses de ces Doctrines chez Alexandre d’Aphrodise et Chez Themistius,” Micellanea Mediaevalia, 13:1 (1981): 211–25; and “The Philosophical Purport of Maimonides’ Halachic Works and the Purport of the Guide of the Perplexed,” in S. Pines and Y. Yovel, eds., Maimonides and Philosophy (Dordrecht-Boston, 1986). For an impressive though not entirely successful response to Pines’ thesis, see now Alexander Altmann’s “Maimonides on the Intellect and the Scope of Metaphysics,” in his Von der mittelalterlichen _ur modernen Aufklarung: Studien sur jüdischen Geistesgeschichte (Tubingen, West Germany, J.C.B. Mohr (Paul Siebeck), 1987, pp. 60–129, especially pp. 72–76 for an excellent discussion of Maimonides’ theory of intellection against its Islamic background; also Arthur Hyman, “Demonstrative, Dialectical, and Sophistic Arguments in the Philosophy of Moses Maimonides,” in Eric Ormsby, ed., Moses Maimonides and His Time (Washington, D.C., Catholic University of America Press, 1988); Barry S. Kogan, “’What Can We Know and When Can We Know It?’ Maimonides on the Active Intelligence and Human Cognition,” ibid.; Joel L. Kraemer, “Maimonides on Aristotle and Scientific Method,” ibid., and my “Skeptical Themes in the Guide of the Perplexed,” ms. The intellectual limitations in question, it should be added, bear only on our knowledge and representation of metaphysical truths; there may, however, exist additional limitations with respect to non-metaphysical human knowledge which do not have the same representational consequences. For examples of these other general incapacities of the intellect, which also arise for subjects other than metaphysics, see Guide I1I:49:605, 111:23, and Commentary on the Mishnah Eiruvin 1, v.
See Sabra, pp. 746ff., who argues that Avicenna also held this Platonic view. Note, however, that, while the conception of science is Platonic, the reason for lack of subjectmatter is Aristotelian.
Cf. Zimmermann, pp. xxxix; and C. on De. Im., 84.9–15 (p. 78), where Al-Farabi criticizes earlier commentators for “making logic an investigation into the nature and mode of being or existence of existing things [i.e. what is signified by expressions].” Although this view of logic was widely held among the Aristotelians, it was not universally accepted. Avicenna, it has been argued, claimed that logic has its own conceptual subjectmatter, the so-called “secondary intelligible concepts”; on this, see Sabra, pp. 751–757.
To be sure, there was precedent for including these topics in a logic, after Porphyry and the canonization of the Isagoge and Categories within the Organon. For further discussion, see now Raymond L. Weiss, “On the Scope of Maimonides’ Logic, Or, What Joseph Knew,” in R. Link-Salinger et al., eds., A Straight Path: Studies in Medieval Philosophy and Culture: Essays in Honor of Arthur Hyman (Washington D.C., Catholic University Press, 1988), 255–265; p. 256. My thanks to Professor Weiss for kindly providing me with a prepublication copy of his paper.
See Guide 11I:18:474, where he interjects, “as you know,” alluding to the Logic. Cf. also ad. loc. Munk’s comment in his translation, vol. 3, p. 137, n. 1, and vol. 1, p. 185, n. 2. In Vivas’es medieval Hebrew translation of the Logic, ch. X begins “When we see with our mind the universal,” which, as Efros notes, may “reflect an anti-realistic view of universals” (p. 51, n. 20). Of course, anti-realism about universals was the standard Peripatetic position, hardly unique to Maimonides.
See also Logic, Ch. XI, where Maimonides writes that he has digressed to discuss the distinctions between the potential and the actual, the essential and the accidental, the conventional and the natural, and the universal and particular because “the philosophers say that anyone who cannot [draw these distinctions] is unfit to reason,” i.e. engage in rational discourse or scientific inquiry. For a different interpretation of this passage, see Weiss, “On the Scope of Maimonides’ Logic,” p. 257f.
Cratylus 436b. For conflicting views of the - at best small amount of - direct knowledge Al-Farabi might have had of Platonic texts, see Franz Rosenthal, “On the Knowledge of Plato’s Philosophy in the Islamic World,” Islamic Culture 14 (194), pp. 387–422; and Zimmermann, lxxx, n. 3; p. 42, n. I. The brief summary of the Cratylus in the “Philosophy of Plato,” iii;7.0–7.10, is hardly evidence of direct familiarity with the dialogue; cf. Muhsin Mandi, trans. and ed., Alfarahi’s Philosophy of Plato and Aristotle (Glencoe, Ill., Free Press, 1962), p. 56.
Cf. De Int. 16a 5–8.
This Arabic term is also sometimes used equivocally to signify the extra-mental referent - an individual or state of’ affairs - and the relation of reference. More often, however, Al-Farabi uses it in place of the original maqûl where the latter would be the direct translation of the Greek. Cf. Zimmermann, p. 11, n. 2, and xlii, n. 1. I am indebted to Zimmermann (personal correspondence) for clarification of this point.
Zimmermann (personal communication) points out, however, that the “thought-componentsignified by words can be of different kinds: sense data [mahsasat], images [mutakhayyilat], and concepts [ma‘qûlat]; all of these should, obviously, be distinguished from the objects sensed, imaged, and conceived. Al-Farabi himself explains that Aristotle uses the phrase ”‘traces in the soul’ rather than ‘thoughts’ because he means to cover all that arises in the soul after the sense-objects have withdrawn from the senses,“ including, in addition to thoughts proper, ”images of sense-objects“ that were actually perceived and (composite) images that the imagination ”invents by combining images“ (C. 24.17–21). It is not clear to me whether Al-Farabi means to include all of these mental entities under his own subsequent use of the term ‘thoughts’ [ma’galat] or whether he uses this term in a narrow sense.
The significance of this point will emerge in the final section of the paper.
In an important paper which came to my attention only as I was completing this paper, Norman Kretzmann points out that Aristotle in fact uses two different Greek terms to describe two apparently different relations between spoken sounds and the mental trace or impression: ‘symbol’ (6646Xa) and ‘sign’ (64eïa). Subsequent translators and commentators beginning in the Middle Ages, however, translated or interpreted both terms in one way and, in Kretzmann’s opinion, thereby obscured the true meaning of the text. Among these was Ishaq ibn Hunayn, whose translation Al-Farabi used. This matter requires more discussion but reasons of space prohibit it here. Cf. Norman Kretzmann, “Aristotle on Spoken Sound Significant by Convention,” in J. Corcoran, ed., Ancient Logic and its Modern Interpretations (Dordrecht-Holland, Reidel, 1974), pp. 3–21.
For Al-Farabi the distinction between signification in the first and second place is not limited to speech; thus written inscriptions signify in the first place the corresponding speech or vocal utterance and in the second place the thought expressed. It is not obvious, then, that Al-Farabi meant to account specifically for semantic signification by means of this distinction. There is also no textual evidence that he would have generalized the idea to include signification in the third place, etc.
It is not clear whether Al-Farabi thought that the universality of the thought/referent correlation and non-universality of the speech/thought correlation was simply evidence for the naturalness of the one and conventionality of the other, or that there is a stronger conceptual relation between the two sets of notions.
Ed., Muhsin Mandi (Beirut, 1970). My exposition of the Book of Letters is based on an unpublished preliminary translation of Part II by Prof. Mandi. I wish to thank Prof. Mandi for permission to consult his translation. All responsibility for the interpretation and exposition that follows is mine.
Cf. Con De Int. 27.15–20.
In the Book of Letters, Part II, Section XXI, par. 122, Al-Farabi also proposes that the more moderate the natural make-up, or physiology, of a people, the greater their acumen and knowledge, and the more their words will imitate or resemble their associated meanings and the extra-mental existents which they designate. This, he emphasizes, is by virtue of their natural make-up and not by design. That is, at least for certain aspects of signification, Al-Farabi seems to believe that there are naturalistic causal explanations that are independent of, or in addition to the role of, human legislative decisions. This Al-Farabian idea is, in turn, adopted by Maimonides, in his work of medical aphorisms, Pirkei Moshe, in response to Galen’s claims for the superiority of the Greek language over all others. Using Al-Farabi’s idea to justify a rather liberal re-reading of Galen, Maimonides claims that Galen meant to defend the superiority not only of Greek, but of all languages spoken by people in temperate climates, including Hebrew and Arabic (!). See Pirkei Moshe Birefuah, ed., Sussmann Muntner (Jerusalem, Mossad Harav Kook, 1959), Treatise XXV, par. 56–58; English translation: The Medical Aphorisms of Moses Maimonides, 2 vols., trans. and eds., Fred Rosner and Sussman Muntner (New York, Yeshiva University Press, 1970), vol. II, pp. 201–3. For further discussion, see also Septimus, “Maimonides on Language,” pp. 79–87; Shalom Rosenberg and Charles H. Manekin, “Philosophical Observations on Maimonides’ Critique of Galen,” Koroth, vol. 9, Special Issue 1988; and below on Maimonides’ discussion of Babel in Guide 111:50.
See Kuzari 1:55, and above Part I. Cf. also Wolfson, “The Veracity of Scripture,” p. 232.
For an influential contemporary account of linguistic convention of the sort Al-Farabi describes at the first stage, see David Lewis, Convention (Cambridge, Harvard University Press, 1969). For a vivid informal description of the type of view of convention under attack, see also W.V. Quine’s Forward to this volume.
E.g. Stephanus; cf. Zimmermann’s note 3. p. 11.
C . 29.20–24; cf. also C. 50–51.8.
See Zimmermann p. 42, nn. 1, 2, 4 on the source of Al-Farabi’s discussion of these views.
It should be noted that, while it is clear what is meant by the “composition” of a sentence, namely, its syntactic structure, Al-Farabi does not make clear what exactly he means by the structure, or composition, of a state of affairs, nor does he explain how the one resembles or imitates the other. The latter question, it should be added, is a general difficulty that has infected correspondence theories of truth since their inception. i.e. what is the relevant notion of correspondence?
If the “composition,” or structure/form, at issue here is the logical rather than grammatical structure of the sentence - which, as we explained in the previous section, is the structure of the thought signified by the sentence - the following question arises: Why doesn’t the structure of the thought naturally signify the structure of the corresponding state of affairs, as we would generally expect of thought/object signification? In response. I suggest that this reflects the fact, mentioned above, that our only access to, or knowledge of, thought is through language. Hence, while the structure of the thought - the structure of what is expressed - naturally signifies the structure of the corresponding state of affairs, the way in which the thought is expressed in language - how the structure of the thought is expressed - only conventionally signifies the structure of the corresponding state of affairs. So long as the correspondence is systematic and principled, there should then be no problem using the sentences as “substitutes” for the thoughts.
See, for example, Saul Kripke, Naming and Necessity (Cambridge, Mass., Harvard Univ. Press, 1980).
See Gottlob Frege, “On Sense and Meaning,” in P.T. Geach and Max Black, eds., Translations from the Philosophical Writings of Gottlob Frege, Third Edition (Totowa, New Jersey, Rowman and Littlefield, 1980).
For reasons of space, I will not discuss the heavenly spheres and separate intelligences in this paper; whether or not the problems to be discussed bear on them requires further investigation. 5 On these senses of the term’ Adam,’see Guide 1:14. On ’Adam as ruler, see Guide 1:2. It is unclear whether Maimonides would agree with Al-Farabi’s identification of the namegiver with the lawgiver. Although there are certain analogies between the Law and language - such as, Maimonides’ claim that “in all languages rules merely conform to the majority of cases” (1:67:162) and that the Law “is directed only toward the things that occur in the majority of cases” (III:34:534) - there are also significant disanalogies between them, most important, that the Law is said to be divine (i.e. natural) while language is not. The question of their relation deserves more attention than space allows here.
Guide 111:8:435; on wudi’at,compare 1:61:149 and ad locum. ft. 8.
On the identification of the divine and the natural, see above n. 11 and Guide 1:39, 1:66, I1:40, and I1L27.
Cf. above ft. 54 and the references therein to Al-Farabi’s Book of Letters and to Maimonides’ Pirkei Moshe. On the story of Babel, see 111:50:614. Note that Maimonides takes the existence of dissimilar human languages as prima facie evidence against the scriptural claim that the world was created in time and that all of humanity is descended from one human creation; i.e. he takes the diversity of natural languages to be prima facie evidence that language is natural and the world eternal. Why this should be so is not entirely clear.
To state that the “notion lacks existence” is not to say something about the existence or ontological status of the notion itself - which is mental - but to say that the notion is not instantiated by any extra-mental existent; for a related point concerning essences, see Alexander Altmann, “Essence and Existence in Maimonides,” reprinted in his Studies in Religious Philosophy and Mysticism (Ithaca, Cornell University Press, 1969), pp. 108–127; p. 1 1 1 f.
Although Maimonides nowhere explicitly addresses this question, his universe of discourse, the set of objects which can be signified by terms in his language, seems to include both existent and non-existent objects. That is, the objects which one’s terms or notions can signify may not actually but only possibly exist; indeed they may even include impossible objects, e.g. a deity whose essence is absolutely simple and in which subsist multiple notions, which is an entity Maimonides describes as “absolute nonexistence” (I:61:147). ‘Existence; correlatively, is (in contemporary terminology) a predicate which will apply to an object depending on whether it in fact exists. To the best of my knowledge, Maimonides nowhere indicates what criterion he has for object-hood.
Here I have altered Maimonides’ own example which has the person apprehending in the notion of a man his animality but not his rationality, because it hardly seems plausible that one could signify a man with so weak a representation, even if the attribute in question is part of its definition. Unless Maimonides thinks that just any attributional particularization is sufficient to determine the signification of the associated term, the necessary thought-component must be richer than Maimonides’ own example would suggest. (On particularization as constitutive of attribution, see I:58:134–5). None of these important details is, it should be noted, made clear in Maimonides’ account. Similarly, it remains unclear in what exact way the first and second cases differ, although the general contours of Maimonides’ distinction are clear enough. How, for example, does Maimonides’ elephant representation, which fails to represent or signify an elephant, differ from his example of a man representation, which none the less does signify man? In the text I have called the elephant representation “incoherent” but I have not attempted to make this notion of (in)coherence clear. It is narrower than logical incoherence but broader than physical impossibility; one would hope for some intermediate notion of demonstrative or rational incoherence but this requires more investigation than present space allows; for a start, however, see Maimonides’ criticisms of the kaldm notion of admissability in 1:73:206–212 and III:15.
On Maimonides’ other examples, ‘anqa mughrib’ and ‘centaur,’ see Al-Farabi’s use of the same terms, in C. on De Int. 28.13–29.
This view of reference might raise one last question for the contemporary reader. Suppose beliefs are composed of such representations, and beliefs are the bearers of truth. Suppose also that the individual does not have a false belief about God (or, in another of Maimonides’ examples, elephants) because he has no belief about God but rather a belief about some imaginary nonexistent entity — call it Ged. Does he then have a true (or, for that matter, a false) belief about Ged (since his representation does, by hypothesis, signify Ged and attributes some property to it)? Or does his imaginary belief lack all truth-value? Although he gives no argument for his answer, Maimonides does seem to state that such a belief is false — and false because the thing signified does not exist. Whether this is because every sentence carries an existential presupposition requires further investigation.
Because designation involves signification by nature, the competent speaker of language must, therefore, know the natures, or at least distinguishing attributes, of the entities to which he wishes to refer. This may or may not be a plausible description of our actual practices of signification, but it may be another reason — as I alluded at the end of Part IV — why Maimonides includes the subject of language in his exposition of ma aseh bereshit.
Throughout this passage, Maimonides makes use of the two classical Arabic terms of logic, tasawwur and tasdiq (Heb: tziyur and ’imut). Literally, tasawwur is the act of grasping or receiving a form in the mind, the act of conception or representation; tasdiq, the act of taking or believing something to be true (sadiq), the act of affirmation or verification. But the same terms also come to serve for the objects of those mental acts. Properly speaking, truth and falsity, and therefore belief, have only to do with tasdiq. In this role, the object of tasdiq corresponds to the modern notion of a proposition, although this proposition-like object of tasdiq should also (unlike certain modern conceptions of the proposition) be distinguished from the sentence uttered in the act of affirmation. On the tasawwur/tasdiq distinction, see the classic article by H.A. Wolfson, “The Terms Tasawwur and Tasdiq in Arabic Philosophy and their Greek, Latin, and Hebrew Equivalents,”’ repr. in his Studies in the History of Philosophy and Religion, eds., I. Twersky and G.H. Williams (Cambridge, Harvard U. Press, 1973), pp. 478–492; and Sabra, p. 758ff. (and references therein).
Here, as well as in 1:50 and in 11:5 (in connection with the spheres), Maimonides cites as prooftext Ps. 4:5. For further examples of the priority of internal over external speech, see Guide 1:51:113f; 1.60:144; 11:12:279–80; 1II:8:432–3; and Mishneh Torah Hilkhot Debt, 2:4: “Similarly, in matters of Torah and wisdom [hokhmah] one’s words should be few and his thoughts [inyaneihem] many.” ’was the standard translation of ma’na’ in medieval philosophical Hebrew.
See Abraham Nuriel, “The Torah Speaks According to the Language of the Sons of Man,” (Heb.) in M. Hallamish and A. Kasher, eds., Religion and Language: General and Jewish Philosophical Essays (Tel Aviv, University Publishing Projects, 1981). Nuriel does not, however, note Maimonides’ shift in 1:57:133 in his truncated use of the phrase “according to the language of the sons of man (ke-leshon benei ‘adam),” from signifying the language of the multitude to signifying the language of the species of humanity — the species distinguished by its faculty of speech or reason — namely, the language of intellectual apprehension and thought.
Altmann, “Maimonides on the Intellect,” p. 120.
See H.A. Wolfson, in (among other places) his “Maimonides on Negative Attributes,” rep. in his Studies in the History of Philosophy and Religion, vol. II, pp. 195–230. Although Wolfson was the first to sharply distinguish between the “metaphysical” and “logical” problems of divine attributes, I disagree with his interpretation of the “logical” problem. On Wolfson’s interpretation, see now Altmann’s critical discussion, in “Maimonides on the Intellect,” pp. 122–4. Furthermore, it should be noted that Wolfson’s solution to his “logical” problem would be more accurately described as a pragmatic rather than logical solution. According to Wolfson, despite the fact that all predicative statements about God are tautologies, they can none the less be “useful,” i.e. pragmatically informative, in contexts where they serve as an answer to a challenge, e.g. where they “emphasize… the negation of something which otherwise might be considered as admissable of God… As the answer to the challenge that A is B, it is necessary to affirm that A is A in order to emphasize that A is not B” (p. 204, my emphasis).
That Maimonides saw the difference between the two questions is clear from a number of remarks. The existence of certain divine attributes - and the problem of compositionality which they raise - is said to be “consider[ed] to be cognized by the intellect” independently of whether those attributes are or are not found in “the text of the word of a prophet” (1:55:121) and, therefore, independently of any exegetical need. Maimonides also envisions a messianic age when the two questions will coincide, when people will invoke God by one and only one name which will correspond to the unique notion or representation that can be intellectually apprehended about God (1:61:148–9). During the present age, however, the two questions, the exegetical and the normative, diverge.
See Seymour Feldman, “A Scholastic Misinterpretation of Maimonides’ Doctrine of Divine Attributes,” Journal of Jewish Studies, vol. 19 (1968): 23–39.
See Altmann, “Essence and Existence in Maimonides.”
Fazlur Rahman, “Ibn Sinâ s Theory of the God-World Relationship,” in Bernard McGinn and David Burrell, eds., God and Creation (Notre Dame, Ind., University of Notre Dame Press, 1990), gip. 38–52, p. 38.
Altmann, “Maimonides on the Intellect,” p. 119, n. 75.
See below for implications of this interpretation for Maimonides’ Avicennean argument for the existence of a necessarily existent being in Guide 1I:1. Despite the differences in argument, nothing I have said conflicts with the claim that the source of Maimonides’ distinction in 1:57 is Avicenna. On the Mu‘tazilite sources of the formulae such as “living but not through life,” see H.A. Wolfson, Repercussions,pp. 73–4.
The illicit causal presuppositions of compositionality are Maimonides’ central arguments for the unity of God in II:1:250–1.
Compare Maimonides’ argument for the unity of God in Mishneh Torah, Book of Knowledge, “Laws concerning the Foundations of the Law,” 1:7.
Compare Maimonides’ argument against divine corporeality in II:1:251–2.
For the example of the wall, cf. Al-Farabi, C. on De Im. 38.9, Zimmermann’s appended note on p. 259, and Maimonides’ Logic, ch. 11. On categorical negation, see also Al-Farabi’s discussion in the C. 38.1–39.18 and 107.9–125.24 and the medieval Hebrew commentaries on the Guide,ad. loc., of Moses of Narbonne, (Be’ur, Ed., J. Goldenthal, 1832), p. 10; Shem Tob ibn Shem Tob, Efodi, and Asher Crescas (the last three all reprinted in the standard edition of the Ibn Tibbon Hebrew translation of the Guide). The medieval Hebrew term for categorical (as opposed to particular) negation was shelilah mushlahat or kellalit (lit.: absolute or general negation). For further discussion, see Wolfson, “Negative Attributes,” op. cit.
C . on De Int. 125.11–23. According to Al-Farabi, “a noun… becomes indefinite when the negative particle, i.e. the particle ‘not,’ is linked with it in such a way that together the two words assume the shape of a single expression” (Short Treatise on Aristotle’s De Interpretatione, 39 ff.). Such nouns, he emphasizes, are not negative phrases because, like single expressions, they have affirmative significations, namely, kinds of privation. On this, see further C. on De Int., 32.1–8, 38.1–39.18.
ere is, I would propose, additional circumstantial evidence that Maimonides may have had this Al-Farabian passage in mind while writing 1:58. Following the passage quoted in the text, Al-Farabi goes on to illustrate the indefinite noun with the example of the indefinite nouns ‘not-heavy’ and ‘not-light’ as applied to heavenly bodies. Following this example, he writes enigmatically: “What we have said is not without problems. But to say more than this would be to go beyond the limits of logical inquiry. It must suffice to have pointed out as much as this” (C. on De Int. 125.1523). This last example is especially intriguing because Maimonides’ discussion of categorical denial, which he illustrates with the same example as Al-Farabi, the unseeing wall, is also followed by a digression addressed to “you who read this Treatise with speculative intent.” In this digression he takes up the limitations of human knowledge with respect to celestial phenomena and its implications for the meaning of our affirmative descriptions of the heavens. Adopting the same example as Al-Farabi, Maimonides says, “we are unable to predicate of [the heavens] any attributes except in terms whose meaning is not completely understood… Accordingly we say that the heavens are neither light nor heavy… and we make other negations of this kind. All this is due to our ignorance with regard to that matter” (I:58; cf. also 1I:22). Both in the case of the heavens and in the case of God, Maimonides gives a solution different from Al-Farabi’s. However, since there is no other apparent reason why Maimonides should introduce this digression where he does, it would seem most plausible that it was Al-Farabi’s juxtaposition of the deity with the heavens that influenced Maimonides’ decision in Guide I:58 also to juxtapose the two.
In fact, even this way of solving the problem remains incomplete: only two of the eight predicates discussed in this passage are given explicit action interpretations (although not all of them strictly speaking involve privations, e.g. multiplicity or being uncaused). Furthermore, the passage vacillates between more straightforward action description (like that found in I:54) and neo-platonic emanationist description. The relation between these two requires further investigation.
Maimonides explicitly describes the problem in terms of affirmative attributes in order, I think, to emphasize the superiority of negative attributes to affirmative ones as a solution to the exegetical problem of attributes. For obvious reasons, Maimonides openly plays down the normative question throughout the last three chapters I:58–60.
I am indebted here to Moshe Habertal for sharpening this issue in my mind.
Cf. also I:60:145 according to which even the bare distinction between substratum and adjunct implies “polytheism.” “For every substratum bearing things is undoubtedly, according to its definition, a duality, even if it be one in its existence. For the notion of the substratum is different from that of the adjunct borne by it.” And this, in turn, conflicts with “the demonstration of the impossibility of composition in Him… and, to go even further, the demonstration of His absolute simplicity” (I:60:145). Here also the composition is structural rather than a matter of different autonomous parts compounded together.
On possible Ismaili influences on Maimonides’ conception of negative attributes reflected in this passage, see Shlomo Pines, “Shiite terms and conceptions in Judah Halevi’s Kuzari,” Jerusalem Studies in Arabic and Islam II (1980): 165–251, Appendix VI, p. 240–243; idem., Translator’s Introduction, pp. xlvff.
Cf. Al-Farabi, C. on De Im’. 38.14–39.3.
Cf. also Altmann, “Essence and Existence in Maimonides,” p. 110f.
On the dominating negative influence of the imagination on the mental representation of immaterial entities in general, see Guide I1:12:279–80.
See Al-Farabi’s Short Treatise, 52–53, pp. 230, especially n. 4, and Zimmermann’s discussion in his Introduction, p. xxiv-xxxviii; especially xxxviii. Maimonides’ Arabic term for ‘derived’ is the same as Al-Farabi’s term al-ism al-mushtagq which he uses for ‘derived terms’ and which Zimmermann explains is the “paronym” of the Greek school tradition. Although the term ‘ishtigaq’ [derivation] is borrowed from Arab grammar and philology where it means etymology (which may be reflected in Maimonides’ first sense of ‘derived’), in the logical tradition it comes to designate derivative nouns like adjectives and participles.
Shlomo Pines, “Dieu et l’être selon Maimonide: Exégèse d’Exode 3, 14 et Doctrine Connexe,” in A. de Libera et E. Zum Brunn, eds., Celui qui Est: Interprétations Juives et Chrétiennes d’Exode 3, 14 (Collection Patremaines, Editions du Cerf, Paris 1986), pp. 15–24; p. 17 and notes ad loc.; idem. “Lés textes arabes dits plotiniens et le courant ‘porphyrien’ dans le néoplatonisme grec,” in Le NéoPlatonisme (Paris, 1971): 303–317; reprinted in Collected Works of Shlomo Pines (Jerusalem, Magnes, 1986), vol. II. (My special thanks to Sarah Stroumsa and Zev Harvey for bringing these papers to my attention.) On ’anniyya,see also F. Rosenthal, Orientalia, 21 (1952): 474–479, and Joel Kraemer’s discussion in the works cited in n. 21. For an excellent summary and evaluation of the scholarly debate over ’anniyya in connection with the claims of this paper, see now W.Z. Harvey and S. Harvey, “A Note on the Arabic Term ’Anniyyal’Aniyya/’Inniyya” (in Heb.), Iyyun (April 1989): 167–71. On Maimonides’ own sensitivity to the problems of translating into Hebrew the subtlety different terms for existence in Arabic, see his revealing comment in Guide I:63 that “there is no difference whether you say ‘was’ or ‘existed’ in Hebrew” (my emphasis), implying, of course, that there is all the difference philosophically between the two terms.
For Munk’s view, see Le guide des égarés, vol. 1, p. 241, n. 2; Baladi expresses his opinion in the transcribed oral discussion of Pines’ paper “Lés textes arabes,” appended to its published version, p. 316. It should be noted that, despite his view expressed in the above papers, Pines himself translates ’anniyya in some of the above passages in this very way.
For D’Alverny, see also the discussion appended to Pines’ paper, “Lés textes arabes”; for Kafih, see his Hebrew translation of the Guide (Jerusalem, Mossad HaRav Kook, 1972), vol. 1, pp. 141–2, 144.
For Vajda, see also the oral discussion following Pines’ “Lés textes arabes,” p. 317. Al-Farabi’s explication is found both in his Kitdb al-Alfaz (ed., M. Mandi, p. 45) and in his Kitdb al-Harûf (ed., M. Mandi, p. 61). It should also be noted that, according to Al-Farabi’s interpretation, the expression in question should be vocalized ’inniyyah rather than ’anniyyah.
See Harvey and Harvey, op. cit. None the less they argue for the Vajda/Al-Farabi explication, in large part because of Maimonides’ high regard for Al-Farabi’s logical works.
For this view, and discussion of it, see the papers of Pines, Altmann, Hyman, Kogan, Kraemer, and this author, cited earlier in n. 38.
On the “greatest proof’ from the motion of the heavens, see also Maimonides’ skeptical (and apparently contradictory) claim in 11:24:327 that ”even the general conclusion that may be drawn from [the heavens], namely, that they prove the existence of their Mover, is a matter the knowledge of which cannot be reached by human intellects.“ On these apparently contradictory passages, see Pines, ”Translator’s Introduction,“ pp. cxii—cxiv, and Kraemer, ”Maimonides on Aristotle and Scientific Method,“ op. cit. pp. 79fî, who also emphasizes the dialectical character of Maimonides’ argument. Throughout these passages on the ”proof“ from the motion of the heavens, note too that Maimonides uses the Arabic term dal7, which refers to persuasive (usually dialectical but sometimes also rhetorical) arguments that lead to accepted opinions, rather than the term burhdn which he reserves for demonstrations that lead to certain knowledge. A full evaluation of Maimonides’ view of the relation between the Necessasry Existent and First Mover must take all these differences of argumentative style and terminology into account. Finally, compare also Maimonides’ apparently contradictory discussions of the role of the imagination and intellect in deciding questions of possibility and necessity, in Guide 1:73 and 111:15, a contradiction first pointed out by Emil Fackenheim, ”The Possibility of the Universe in Al-Farabi, Ibn Sina, and Maimonides,“ reprinted in Arthur Hyman, ed., Essays in Medieval Jewish and Islamic Philosophy (New York, Ktav, 1977), pp. 303–334; p. 324, n. 61, which has not, to my knowledge, been satisfactorily explained to date.
Note added in proof (November 1999): Since submission of this paper in 1988, in a number of late publications and lectures I have elaborated, refined, and qualified the ‘skeptical’ interpretation of Maimonides’ philosophy first broached here in the argument for the impossibility of propositional knowledge about the deity based on the syntactic problem of divine attributes. On the difference between subjects with respect to which our knowledge is limited (e.g. cosmology and astronomy) and those about which it is humanly impossible to have knowledge (e.g. the deity), see my “Maimonides on the Growth of Knowledge and the Limitations of the Intellect,” to appear in Tony Levy, ed., Maimonide: Traditions philosophiques et scientifiques médievales arabe, hébraique,latine. For the implications of this argument from divine attributes for Maimonides’ conception of intellectual happiness and spiritual worship, for broader epistemological and ethical consequences of his view of matter and the body, and for additional examples of skeptical arguments in the Guide, see my “Maimonides in the Skeptical Tradition,” ms. and The Matter and Form of Maimonides’ Guide (Cambridge, Harvard University Press, forthcoming 2002). In these writings, I also elaborate the role of the imagination in Maimonides’ theory of internal representations and revise details in my reconstruction of the argument in 1:50–63. On one implication of Maimonides’ skepticism for his conception of prophecy and law, see also my Problems and Parables of Law (Albany, SUNY Press, 1998): 128–133. I should also mention two important articles that have subsequently appeared that address the issues raised in this paper and are critical of my position: Herbert Davidson, “Maimonides on Metaphysical Knowledge,” Maimonidean Studies, ed. A. Hyman (New York, Ktav, vol. 3, 1992–3): 49–103 and Alfred Ivry, “The Logical and Scientific Premises of Maimonides’ Thought,” in Perspectives on Jewish Thought and Mysticism, eds. A.L. Ivry, E.R.W. Wolfson and A. Arkush (Amsterdam, Harwood, 1998): 63–97. Lack of space precluded responses to their criticisms in this essay, but I address both in the aforementioned articles and book.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2000 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Stern, J. (2000). Maimonides on Language and the Science of Language. In: Cohen, R.S., Levine, H. (eds) Maimonides and the Sciences. Boston Studies in the Philosophy of Science, vol 211. Springer, Dordrecht. https://doi.org/10.1007/978-94-017-2128-8_10
Download citation
DOI: https://doi.org/10.1007/978-94-017-2128-8_10
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-5348-0
Online ISBN: 978-94-017-2128-8
eBook Packages: Springer Book Archive