1 Beiser, Frederick, The Fate of Reason (Cambridge, MA: Harvard University Press, 1987), p. 38.Google Scholar In what follows, I rely heavily on Beiser's well-documented account of the relationship between Hamann and Kant. T. C. Williams argues that Kant would also have been well aware of the issues involved via Herder's 1772 “Prize Essay” (“Über den Ursprung der Sprache”), both through Kant's personal relationship with Herder as well as the attendant publicity the essay generated; see his Kant's Philosophy of Language (Lewiston, NY: Edwin Mellen Press, 1993), p. 67.Google Scholar In contrast, Butts, Robert argues, in “The Grammar of Reason: Hamann's Challenge to Kant,” Synthese, 75 (1988): 251CrossRefGoogle Scholar, that Kant was “almost surely unaware” of Hamann's, as well as Herder's, critical responses to the first Critique, although he also suggests that Kant may simply have been “disinclined to answer” them. Butts's valuable article focuses on issues, e.g., the role of the sensus communis and Kant's theory of symbols, that go beyond the scope of this article, and cannot be discussed here; his brief “Kant's Schematism as Semantical Rules” (in Kant Studies Today, edited by Beck, Lewis White [LaSalle: Open Court, 1969], pp. 290–300)Google Scholar, arguing, with Sellars, that Kant is providing the “epistemological grammar” for judging, adopts a perspective not unlike that which is contained here

2 Surber, Jere Paul, Language and German Idealism: Fichte's Linguistic Philosophy (Atlantic Highlands, NJ: Humanities Press, 1996), p. 4.Google Scholar

3 Aarsleff, Hans, From Locke to Saussure (Minneapolis: University of Minnesota Press, 1982), p. 11;Google ScholarMcCumber, John, The Company of Words: Hegel, Language, and Systematic Philosophy (Evanston, IL: Northwestern University Press, 1993), pp. 290, 394 n.68;Google Scholar and Butts, “The Grammar of Reason.”

4 In addition to Williams (Kant's Philosophy of Language), an exception is Markis, Dimitrios, “Das Problem der Sprache bei Kant,” in Dimensionen der Sprache in der Philosophic des deutschen Idealismus, edited by Scheer, B. and Wohlfart, G. (Würzburg: Konigshausen & Neumann, 1982), pp. 110–54.Google Scholar

5 Eco, Umberto, The Search for the Perfect Language, translated by Fentress, James (Oxford: Blackwell, 1995);Google Scholar Eco does not mention Kant in his text. In an earlier draft, I argued at some length that Kant can be seen as making a significant, if overlooked, contribution to this tradition, by comparing some of his views with those of the Modistae, such as Thomas of Erfurt and Siger of Courtrai. I have had to eliminate that discussion here for reasons of space.

6 Of course, at least some of the questions involved go back as early as Plato's Cratylus; see the reference to Benfey, T.'s 1866 “Über die Aufgabe des platonischen Dialogs Kratylos,” in W K. C. Guthrie, A History of Greek Philosophy, Vol. 5 (Cambridge: Cambridge University Press, 1986), p. 31.Google Scholar In addition to Surber (Language and German Idealism), some of this history is traced by Lars Fr. H. Svendsen in his dissertation, “Kant's Critical Hermeneutics: An Interpretation and Defense of the Schematism” (University of Oslo, 1999); I am indebted to him both for sending me his published and unpublished work, and for extensive discussions of the topics involved here.

7 Williams, Kant's Philosophy of Language, p. 47.

8 Ibid., p. 170.

9 Ibid., p. 73.

10 Hamann, Johann Georg, “Metakritik über den Purismum der Vernunft,” in Sämtliche Werke, Vol. 3 (Vienna: Hamann Verlag, 1949–1957)Google Scholar: “das einzige erste und letzte Organon und Kriterion der Vernunft, ohne ein ander Creditiv als Ueberlieferung und Usum,” p. 284 (Hamann's emphasis.) All references to Hamann are to this work, and will be given parenthetically in the text.

11 “Was die Transcendentalphilosophie metagrabolosirt”; R. G. Smith identifies the source of Hamann's neologism as Rabelais, from naraux [vain], γραΦειυ [to write], and βoλζειυ [to plumb], glossing it in general as “something like nonsensical investigations.” See Smith, R. G., Johann Georg Hamann, 1730-1788 (London: Collins, 1960), p. 220.Google Scholar As Beck, L. W. notes in his Early German Philosophy (Cambridge, MA: Harvard University Press, 1969), p. 379 n.43Google Scholar, W. M. Alexander translates the corresponding adjective as “meta-obscure.”

12 This is not to suggest that this “tension” originated in the debate between Hamann and Kant; it could be argued one can find a similar issue at stake between Parmenides and Heraclitus. Hamann's objection simply is useful in couching the discussion specifically in terms of language.

13 I am not suggesting that this debate reduces to one between relativism and absolutism. Without arguing for it here, however, I think each is often seen as the “endpoint,” or logical consequence, to which the adherents of one view see the others at risk of implying, if not succumbing to. I have discussed this issue, relative to some of the tensions that exist between “Enlightenment” and “Standpoint” feminism, in “Kant and Feminism,” Kant Studien, 90 (2000): 322–53.Google Scholar Recently, Nagel, Thomas (The Last Word [Oxford: Oxford University Press, 1996])Google Scholar, McDowell, John (Mind and World [Cambridge: Harvard University Press, 1996])Google Scholar, and BonJour, Laurence (In Defense of Pure Reason [New York: Cambridge University Press, 1998])Google Scholar, among others, have renewed the debate over the need for universal principles—of logic, reason, ethics, and epistemology, among them—in contrast to the somewhat more fashionable emphasis on “situatedness” or “contextualism.”

14 As Beiser (The Fate of Reason, p. 38) notes, “Although the 'Metakritik' was not published until 1800, it exercised a considerable subterranean influence. Hamann sent a copy to Herder, who in turn sent one to Jacobi. Through Herder and Jacobi, some of the ideas of the ‘Metakritik’ became common post-Kantian currency.” Surber agrees: “Although Hamann's Metakritik was not published until 1800, it appears to have circulated fairly extensively and had able advocates in the persons of Herder and Jacobi” (Language and German Idealism, p. 11).

15 If one excludes the introductory material and the important, but relatively neglected, Transcendental Doctrine of Method, the remaining text consists of the two radically disproportionate halves of the Transcendental Doctrine of Elements: the Transcendental Aesthetic (pp. 19-49) and the Transcendental Natural Language 47 Logic (pp. 50-704) [in the pagination of the first (“A”) edition of 1781.] In Kants Logikvorlesungen als neuer Schlüssel zur Architektonik der Kritik der reinen Vernunft (Stuttgart: Frommann Verlag, 1994)Google Scholar, Elfriede Conrad has argued in impressive detail “daβ Kant die Logik dem systematsichen Aufbau der KrV als Leitfaden zugrunde gelegt hat” (p. 12), indicating that Kant's understanding of logic is considerably more sophisticated and complex than the view standardly attributed to him.

16 Vleeschauwer, H. J. De, La déduction transcendentale dans l'oeuvre de Kant, Vol. 3 (Paris: Leroux, 1934–1937), p. 101Google Scholar, and Reich, Klaus, Die Vollständigkeit der kantischen Urteilstafel (Ph.D. dissertation, University of Rostock, 1932), p. 27.Google Scholar A key passage for understanding Kant's strategy is his distinction between “principle” and “theorem” at A737=B765.

17 “Welches sind die wirklichen Fortschritte, die die Metaphysik seit Leibnitzens und Wolffs Zeiten in Deutschland gemacht hat?,” Academy Edition, Vol. 20 (Ak. XX), p. 260. All references to Kant's work other than the Critique of Pure Reason are to this standard edition (Berlin and Leipzig: de Gruyter et al., 1902-)Google Scholar, indicated by “Ak.,” volume number, and an abbreviated title. References to the first Critique are to the standard first-edition (“A”) and second-edition (“B”) pagination, using Raymund Schmidt's Philosophische Bibliothek edition (Hamburg: Meiner, 1956), and are given parenthetically. While I generally follow Kemp Smith's translation of the Critique of Pure Reason, I have on occasion altered it. One relevant point should at least be noted here: Kemp Smith translates both “Gegenstand” and “Objekt” (and their derivatives) as “object,” although the terms are not co-extensive in Kant's own usage. This issue, obviously enough, causes systematic difficulties in any account of Kant's treatment of “objectivity"; for some discussion, see Rousset, Bernard, La doctrine kantienne de l'objectivité (Paris: Vrin, 1967), pp. 294ff.Google Scholar All other translations, unless specified, are my own.

18 Kant, Prolegomena, Ak. IV, p. 323. As we will see below, Surber takes this passage as central to his understanding of Kant's approach to language, and I will argue that he misinterprets it.

19 Logik Jäsche, Ak. IX, pp. 12-13; Kant's emphasis. In spite of this remark, Terry Boswell indicates, correctly I think, one reason the conviction that Kant was entirely unconcerned with language persists. The Jäsche logic, as the only set of Kant's logic lectures translated into English, until very recently, has played a disproportionate role in determining Kant's views on logic. As Boswell observes, “For example, in his lectures on logic Kant made remarks about the nature and use of language which Jasche did not include in his edition” (“On the Textual Authenticity of Kant's Logic,” History and Philosophy of Logic, 9 [1988]: 201).Google Scholar

20 See Svendsen, “Kant's Critical Hermeneutics,” for some discussion, including the citation of an important text (Logik Busolt [Ak. XXIV, p. 609]) where Kant discusses this distinction.

21 Thompson, Manley, “On A Priori Truth,” Journal of Philosophy, 78 (1981): 471 n.8.CrossRefGoogle Scholar

22 Svendsen, “Kant's Critical Hermeneutics.”

23 “Reflexion 1592,” Ak. XVI, p. 28. The editor of these notes, Erich Adickes, is unsure of the dating of this note, but places it between 1760 at the earliest and 1775 at the latest.

24 “Logik Phillipi,” Ak. XXIV. 1, p. 316; see also, “Reflexion 1620,” Ak. XVI, p. 39 where Kant observes, among other things, that Moliere's bourgeois gentleman not only discovers he has been speaking prose, but grammatically, his entire life.

25 “Logik Pölitz,” Ak. XXIV.2, p. 502 (delivered in 1789).

26 “Logik Busolt,” Ak. XXIV.2, p. 609 (delivered in 1790); cf. “Metaphysik Mongrovius," Ak. XXIX, p. 1046 (delivered in 1783).

27 “Logik Phillipi,” Ak. XXIV. 1, p. 339. Cf. Wittgenstein, L., Philosophical Investigations (New York: Macmillan, 1958)Google Scholar: “Die Erklarungen haben irgendwo ein Ende” (§1). Additionally, Kant might reply that any objection to his approach, to be communicable, must satisfy the minimal logical conditions he has articulated. Wittgenstein and Kant both share, as well, a commitment to the notion that grammatical and logic questions “have the character of depth” (see Investigations, §111). There remains, of course, the alternative of scepticism; as Laurence BonJour has pointed out, “There are certain versions of skepticism which are so deep and thoroughgoing that it is impossible in principle to refute them” (The Structure of Empirical Knowledge [Cambridge, MA: Harvard University Press, 1985], p. 194)Google Scholar. Presumably, this is not an alternative to which Hamann would be attracted.

28 The qualification is, of course, that introduced by Donald Davidson, calling into question the very coherence of conceptual schemes in his “On the Very Idea of a Conceptual Scheme,” reprinted in Inquiries into Truth & Interpretation (Oxford: Oxford University Press, 1984), pp. 183–98.Google Scholar While Davidson's refusal to embrace the a priori in any strong sense distinguishes his position from Kant's, there are strong similarities to the conclusion Davidson draws in this article and Kant's own views. The question, which goes beyond the scope of this discussion, is whether in attributing agency to another, we must attribute some set of minimal logical constraints on meaning and communicability, and, if so, whether they can be satisfactorily identified.

29 See Prolegomena, Ak. IV, p. 352, where Kant distinguishes “Schranken” from “Grenzen.”

30 Melnick, Arthur, Kant's Analogies of Experience (Chicago: University of Chicago Press, 1973), p. 156.Google Scholar

31 Frege, Gottlob, “Über Sinn und Bedeutung,” Zeitschrift für Philosophic und philosophische Kritik, 100 (1892): 29.Google Scholar

32 Kant, Prolegomena, Ak. IV, p. 373n.

33 Kuehn, Manfred, Scottish Common Sense Philosophy in Germany 1768-1800 (Kingston, ON: McGill-Queen's University Press, 1987).Google Scholar Kant emphasizes throughout his critical writings, but particularly in the Lectures on Logic, that he is concerned with a healthy (gesund) human understanding, which both reinforces his insistence on the normativity involved in outlining its concepts and principles and contains an implicit nod to Rousseau, whom Kant claims taught him to honour humanity. See also Kant's discussion of the concomitant need for a sensus communis in the Critique of Judgement, Ak. V, pp. 238-39 (§21) and pp. 293-96 (§40).

34 The Persius quotation is from the translation by Ramsey, G. G., Persius and Juvenal (London: Heinemann, 1918), Vol. 4, p. 52;Google Scholar the original reads “Tecum habita, et nor is quam sit tibi curta supellex.” The citation from Kant's Trdume eines Geistersehers, erldutert durch Trdume der Metaphysikis from Ak. II, p. 369.

35 Tonelli, Giorgio, “Kant und die Antiken Skeptiker,” in Studien zu Kants philosophischer Entwicklung, edited by Henrich, D. and Tonelli, G. (Hildesheim: G. Olms, 1967), p. 96.Google Scholar

36 Prauss, Gerald, “Zum Wahrheitsproblem bei Kant,” Kant-Studien, 60 (1969): 181CrossRefGoogle Scholar, and Prauss, Gerold, Erscheinung bei Kant (Berlin: De Gruyter, 1971), pp. 275ff.CrossRefGoogle Scholar

37 This way of putting it raises complications that, while relevant, must be ignored here. I will simply note that, strictly speaking, one should speak in terms of a concept being employed inconsistently in a judgement; this allows the kind of distinction noted above, between “thinking a contradiction” and “thinking it as such” (Thompson, “On A Priori Truth, p. 471 n.8). This strategy may also provide a Kantian-flavoured response to paraconsistent logic, as developed by Graham Priest, Richard Routley, and others (see Priest, G., Routley, R., and Norman, J., Paraconsistent Logic: Essays on the Inconsistent [Munich: Philosophia Verlag, 1989])Google Scholar. Although I cannot pursue it here, my guess is that Kant views the “rules” grounded and developed within thepractical philosophy in a similar way, as providing transcendental conditions for moral agency and action. I would further argue that this general strategy can be found in Rousseau, and that in this sense, at least, Kant's debt to Rousseau is strategic, going well beyond the influence standardly restricted to Rousseau's role in the development of Kant's practical philosophy, but informing the very approach that Kant adopts—which is fundamentally based on the notion of “self-legislation”—throughout the Critical philosophy.

38 Thompson, “On A Priori Truth,” p. 464.

39 I have gone into some detail on how Kant's conception of the subject as radically subjective in its imposition of objective conditions for thought and cognition in “Kant's Dialectical Subject and the Bias Paradox,” Contemporary Philosophy, 19, 1-2 (January-April 1997): 15–21.Google Scholar

40 A clear statement of the normativity claim is given in Frege, G.'s notes on logic, collected in Schriften zur Logik und Sprachphilosophie, edited by Gabriel, G. (Hamburg: Felix Meiner Verlag, 1978), p. 38Google Scholar: “Wie die Ethik kann man auch die Logik eine normative Wissenschaft nennen. Wie muβ ich denken, um das Ziel, die Wahrheit, zu erreichen?” The cited remark in the text is from Frege, The Foundations of Arithmetic, translated by Austin, J. L. (Oxford: Blackwell, 1978), p. iii.Google Scholar The original reads, “Das Denken ist im Wesentlichen überall dasselbe: es kommen nicht je nach dem Gegenstande verschiedene Arten von Denkgesetzen in Betracht.”

41 Tennant, Neil, “Logic and Its Place in Nature,” in Kant and Contemporary Epistemology, edited by Parrini, Paolo (Dordrecht: Kluwer, 1994), p. 103.Google Scholar

42 Surber, Language and German Idealism, p. 111.

43 Ibid., p. 29 n.8.

44 Surber, J. P., “J. G. Fichte and the ‘Scientific’ Reconstruction of Grammar,” in New Perspectives on Fichte, edited by Breazeale, Daniel and Rockmore, Thomas (Atlantic Highlands, NJ: Humanities Press, 1996), p. 64.Google Scholar

45 Surber, Language and German Idealism, p. 25. The Prolegomena text is cited as well in Surber, “J. G. Fichte and the 'Scientific' Reconstruction of Grammar”; cf. Surber, J. P., “The Historical and Systematic Place of Fichte's Reflections on Language,” in Fichte: Historical Contexts/Contemporary Controversies, edited by Breazeale, Daniel and Rockmore, Thomas (Atlantic Highlands, NJ: Humanities Press, 1994), pp. 113–27.Google Scholar

46 “Logik Dohna-Wundlacken,” Ak. XXIV.2, p. 693 (delivered in 1792). As Svendsen emphasizes (Kant's Critical Hermeneutics”), Kant is interested in a universal grammar, but not, as was Leibniz and many others, in a universal language.

47 “Reflexion 1622,” Ak. XVI, p. 42.

48 “Philosophical Encyclopedia,” Ak. XXIX. 1, p. 13. The distinction here is, I think, similar to that given in John Wilkins's Essays towards a Real Character and a Philosophical Language (1668), where a contrast is drawn between a “natural” grammar and an “instituted”” or “particular” grammar. This same distinction runs throughout the long tradition of the philosophy of grammar and is similar to that drawn by some medieval philosophers (e.g., Aquinas), as well as by Peirce, between a logica utens and a logica docens.

49 Aarsleff, From Locke to Saussure, p. 278.

50 “MuthmaBlicher Anfang der Menschengeschichte,” Ak. VIII; see especially pp. 110-11.

51 Horstmann, Rolf-Peter, “Die metaphysische Deduktion in Kants 'Kritik der reinen Vernunft,'” in Probleme der “Kritik der Reinen Vernunft,”, edited by Tuschling, Burkhard (Berlin: de Gruyter, 1984), pp. 15–33.Google Scholar

52 The range of sympathy extends from Reich's Die Vollständigkeit der kantischen Urteilstafel and Wolff, Michael's Die Vollständigkeit der kantischen Urteilstafel (Frankfurt a.M.: Klostermann, 1995)CrossRefGoogle Scholar, through Krüger, Lorenze's “Wollte Kant die Vollständigkeit seiner Urteilstafel beweisen?” Kant Studien 59 (1968): 333–56CrossRefGoogle Scholar, to Schopenhauer, whose imagery Jonathan Bennett adopts in writing that Kant's “favoured dozen” judgement-forms “serve throughout the Critique only as a Procrustean bed on which he hacks and wrenches his philosophical insights into a grotesque 'system'” (Kant's Analytic [Cambridge: Cambridge University Press, 1966], p. 89).CrossRefGoogle Scholar

53 Among those who use the term “derivation” are Young, J. Michael, in the Introduction to his translation and edition of Kant's Lectures on Logic (Cambridge: Cambridge University Press, 1992), p. xvi;Google Scholar Butts, “The Grammar of Reason,” p. 256; Shamoon, Alan, Kant's Logic (Ph.D. dissertation, Columbia University, 1981), p. 1;Google ScholarWolff, Robert Paul, Kant's Theory of Mental Activity (Gloucester, MA: Peter Smith, 1973), p. 62Google Scholar; and Strawson, P. F., The Bounds of Sense (London: Methuen, 1966), p. 82.Google ScholarPippin, Robert refers to the transition from the table of judgements to the table of categories as a “derivation of sorts” (Kant's Theory of Form [New Haven, CT: Yale University Press, 1982], p. 90).Google Scholar These examples could easily be multiplied. Grimms, 's Wörterbuch (Leipzig: S. Hirzel, 1885), Vol. 6, pp. 736–37Google Scholar, notes that “Leitfaden” was used as an ordinary term by Lessing in 1751, as well as by another author Kant was familiar with, Christoph Wieland; it was also used by both F. Bacon and Leibniz. Its common context was in describing the thread of Ariadne, used in helping Theseus escape from the Labyrinth after slaying the Minotaur. While it may be ironic that Kant's Leitfaden seems rather to lead us into a labyrinth, it was a common enough expression, and one which he continues to use, as in the Critique of Judgement (e.g., Ak.V, p. 353); cf. “Logik Phillipi,” Ak. XXIV. 1, p. 389.

54 “Logik Blomberg,” Ak. XXIV, p. 27 (Immanuel Kant: Lectures on Logic, edited and translated by Young, J. Michael [Cambridge: Cambridge University Press, 1992], p. 16;Google Scholar Young points out that this lecture series could not have been given earlier than 1770, and suggests the early 1770s as a date; pp. xiii-xiv).

55 Strawson, P. F., “Sensibility, Understanding, and the Doctrine of Synthesis: Comments on Henrich and Guyer,” in Kant's Transcendental Deductions, edited by Förster, Eckart (Stanford: Stanford University Press, 1989), p. 71.Google Scholar Although I obviously cannot pursue the point here, there is a further ambiguity in Kant's references to “logic,” in terms of whether he means the specific formal apparatus of syllogistic, or intends something more general, similar to what became known as the “laws of thought.” This point could be developed in terms of the distinction between logica docens and logica utens (see n. 48 above), or in terms of the point Locke, John makes in his An Essay Concerning Human Understanding, edited by Nidditch, P. H. (Oxford: Clarendon Press, 1975)Google Scholar, Bk. IV, chap. XVII, §4: “But God has not been so sparing to Men to make them barely two-legged Creatures, and left it to Aristotle to make them rational.”

56 In the original, of course, “Idee,” as a noun, is always capitalized; the point is obscured by Kemp Smith, who sometimes capitalizes “Idea,” and sometimes does not, in his translation.

57 A section of this paper was presented at the Kentucky Conference on Foreign Languages at the University of Kentucky, where Professors Daniel Breazeale and Wolfgang Natter offered some valuable remarks. I would like to thank two anonymous readers for this journal for some invaluable suggestions, as well as Robert Batterman, Jere Paul Surber, William Taschek, and, especially, Lars Frederik Händler Svendsen, for insightful comments.