Is the Private Language Argument a Transcendental Argument?
Is the Private Language Argument a Transcendental Argument?


In this paper, starting from Hacker´s reading of the PLA as an extension and elaboration of the Kantian dictum that intuitions without concepts are blind (1972, 216), which suggests a similarity of the PLA with transcendental arguments, the PLA is compared with the Kantian models of such arguments: the Transcendental Deduction of the Categories and the Refutation of Idealism. In this comparison some dissimilarities between the PLA and those models are singled out. A transcendental reading along Hacker´s lines turns out to be objectionable. It also leaves the status of the linguistic ‘I’ undecided. Based on an assessment of the relevance of the PLA a proposal is made as what to do instead.

Table of contents

    1 Comparisons between Kant´s critique of pure reason and Wittgenstein´s critique of language, which became current in analytic philosophy (Cf. e.g Hacker 1972, 30.) seem not far-fetched in view of the impetus for the destruction of dogmatic metaphysics both philosophers share. Their relevance would gain though by an elaboration of their dissimilarities rather than by just stressing similarities.

    An example of the former approach, Weinert (1983, 412) contrasts the tools both critics of metaphysics employ: the distinction between analytic and synthetic judgments, and the description of the logic of language-use and in particular the ´Argument from Epistemic Operators´ (Wittgenstein 1961, sects. 6.5, 6.51; 1958, §§ 246, 247, 251, 303; references in the form of paragraph-numbers are to the latter text) respectively. The analysis of the employment of those tools in anti-metaphysical arguments -and Weinert treats the Private Language Argument (PLA) as such (427-429)- enables to uncover underlying assumptions, e.g. the implicit assumption in Kant of a prior conceptual relation between concepts to which his notion of analysis is to be applied (430-431) and Wittgenstein´s explicit doctrine that “ordinary language is alright” (434).

    An example of the latter approach, Hacker suggests an outright similarity in scope of the PLA and those arguments:

    One can, without undue caricature, conceive of Wittgenstein's purpose in the private language argument as being an endeavour to extend and elaborate the Kantian dictum that intuitions without concepts are blind. (Hacker 1972, 216.)

    Due to its disregard of contextual differences, the illuminative value of that assessment is dubious. As Kitcher contends, those differences are so pervasive:

    that it is an interpretive and philosophical mistake to try to force an alliance between what are, in fact, deeply opposed camps. (...). (2000, 35.)

    And indeed, in what sense one could understand such a treatment of Kant´s dictum (see: 1990, B75) is unclear.

    Kitcher´s criticises the conflation of Kant´s transcendental ´I´ with Wittgenstein´s linguistic ´I´ that prevails in the “British tradition of Kant interpretation from Strawson on”, and mentions, among others, Hacker as a target (2000, 35). In this paper I extend Kitcher´s criticism to a ´Kantian´ reception of the PLA as a transcendental argument (TA), taking Hacker´s qualification as a starting point (section 2). Adopting Weinert’s approach, I will single out some characteristics of the PLA that suggest an ‘unkantian’ reception. An assessment of its relevance can then be made. (Section 3).

    2 As a follow-up of Strawson´s (1959) and (1966), the debate on TA´s facilitated the reception the of the PLA as such an argument (Wilkerson 1975, 108; recently Stern 1999, 3), thereby attributing to its Kantian flavour. For the Transcendental deduction (TD) of the categories and the Refutation of Idealism (RI) are generally regarded as models for TA´s (Brueckner 1983, 552; Stern 1999, 2).

    Under a general denomination, TA´s claim “that one thing (X) is a necessary condition for the possibility of something else (Y), so that (...) the latter cannot obtain without the former” (Stern 2000, 3; Wilkerson, 1975, 102). This does not mark off anything special about the argument itself, but only says something about the propositions it contains. (Gram, 1971, 15; Grayling 1985, 95.) Ample discussion has produced various distinctions between Kantian and revisionist accounts of TA´s (Stern 1999, 9), concerning the nature of the subject matter of (Y) (1999, 3). In ´weaker´ TA´s (Y) stands for the necessary features of a conceptual scheme which makes possible a given structure of experience (Schaper 1972, 102). ‘Weightier’ TA´s (Grayling, 1985, 83) establish necessary conditions for the possibility of experience (Stern 1999, 3). This distinction coincides with Kant´s between regressive and progressive arguments (1993, 26, n). The former can function in refutational arguments to demonstrate the incoherence of a sceptical claim with the necessary assumptions underlying that claim. (Kant 1990, A388-389.) Thus, in the RI, for example, it is argued from the assumption of empirical self-consciousness, the empirical ´I´ of Descartes´ Cogito. (Kant 1990, B275), to the conditions of its possibility, outer objects. The latter argue from a principle to its principiatum. Thus, in the TD, it is argued from an a priori principle (the transcendental unity of the apperception, or transcendental ‘I’), concerning the subject of possible experience, the conditions of which it has to establish, to its principiatum, those conditions, i.e. the categories. Aschenberg (1982, 262, 284), regards only weightier TA´s as transcendental. They, like Kant´s TD, establish basic principles for a conceptual scheme “without which we could not think coherently about experience at all” (Schaper 1972, 102). Accordingly, Kantian would, for Brueckner, be the argument which shows that “the existence of physical objects of a general character is a condition for the possibility of self-conscious experience”, as he beliefs both the TD and RI do (1983, 552). Hintikka requires the conclusion of the Kantian variant to be an assertion concerning the epistemic proces involved (1972, 275). This marks it of from any inference from a successful use of concepts a priori to its presuppositions and a fortiori from arguments which establish the logical presuppositions of something being the case, that Schaper reckons among the general variant (1972, 277, 101).

    This survey shows that the premises of a TA generally may contain synthetic and analytic, a priori and a posteriori, grammatical (Grayling 1985, 109) and empirical propositions. Consequently progressive as well as regressive arguments fall into its scope. Moreover, it shows that the genuine transcendentality of revisionist TA´s is controversial. Finally, that the dialectical distinction between the TD and the RI is often unaccounted for. In view of all this, a reception of the PLA as a TA seems unobvious. Besides, the use of the Argument from Epistemic Operators in the PLA to vitiate a private linguist´s claim concerning the epistemic privacy of sensations -e.g. §248 asserts the non-epistemic character of that claim- indicates already at the outset pace Hacker (1972, 271) the immoderation of a transcendental reading of the PLA, taking into account the primarily epistemological connotation of ´transcendental´ (Kant 1990, B25). Moreover, it turns out to be equivocal.

    This is illustrated by Hacker´s suggestion of a similarity between the PLA and the TD when he cites Kant´s emphasis on the necessary co-operation of the Understanding and Sensibility for the objective validity of judgments (1972, 216) on the one hand, and his suggestion of a similarity withe the RI when he presents the PLA as an argument demonstrating the incoherence of scepticism about the existence of the external world, on the other (215).

    If one would want to project the PLA on the Kantian models, the RI seems the most obvious candidate, judged by its dialectical similarity with the PLA, both being refutations of a sceptical claim. Even if the success of the PLA at refuting the private linguist can be contested (see: Stroud 2000, 78) it could be, in a qualification Bennett coined for the Kantian prototype (1966, 203), taken as a gesture towards such a refutation. Precisely due to this similarity however, the PLA as an undogmatic procedure, merely clarifies that the private linguist´s claim stems from an inconsistency in his use of words (cf. Glock, 1991, 85). As a refutation it would not entail the possibility of interpersonal communicability, let alone, by contraposition of its result, the claim that language is ´inevitably public´. Hintikkaet al. notice that this claim is one of Wittgenstein´s basic assumptions but that he does not use it as a premise in the argument (1986, 261-262, 265). To his observation that his argument would “not be nearly as radical and as sweeping as it in reality” (264) if he nevertheless had done so has to be added that the refutational character of the PLA prevents such use.

    Due to its exclusive orientation on the logic of rules, the Kripkean PLA (Kripke 1982, passim) suggests that it were merely dealing with questions of interpersonal communicability of S-words. As §257 shows however, its greater import is that it deals with the logic of language-games, that are constitutive for name-object relations and hence with the ´inevitably public´ nature of ´grammar.´ (Hintikka et al. 1986, 201, 262.) It remains to be seen whether the PLA purports to show the constitutive force of language-games. In view of this, it is questionable whether its projection on a Kantian TD-model, can convince of its radicality. For the PLA to be a deduction in an analogous Kantian sense it would have to be an argument that would establish a principle of significance (cf. Strawson 1966, 16) to warrant the objective validity of the S-words the private linguist writes in her diary. This principle should obviate the appeal to a Restriktionsthese (cf. Prauss 1974, 174) for that would presuppose the adoption of transcendental idealism (cf. Aschenberg 1982, 61). Therefore, it should principiate the conditions that would make possible that the concepts falling under that principle, would exhaust the universe of sensible discourse, thus leaving no room for concepts that would be meaningful, without having objective reality blosse Gedankenformen (Kant 1990, B148). To paraphrase Strawson: if the private linguist would use a concept that would not fall within the scope of the principle, she would not merely be saying what she does not know, she would not really know what she were saying (1966, 16), and, one should add, she would “just emit an inarticulate sound” (§ 261). But, for all it does, the PLA is not concerned with a justification of the use of S-words: §§267, 289. Wittgenstein´s descriptions of language-games in §§249, 261, 290, 293, 300 are not meant to be exhaustive; that would contradict the notion of language-game as part of a form of life (§23, cited in Hintikka et. Al. (1986, 218); cf. Stroud´s (2000, 78) reading of §258): the positive claims of the PLA are of a much more ´familiar´, even ´mundane´ nature.

    3 With all its incontestable exegetical merits, its equivocity corroborates an extension of Kitcher´s criticism to a projection of the PLA on a Kantian model. It is bound to oscillate between the contemporary aliases of the empirical ´I´ and the transcendental ´I´ thus leaving the status of the linguistic `I´ undecided. This approach would, anachronistically, allow of a reconstruction via Strawson of the PLA as a TD and vice versa, as is offered in Stevenson (1982, 334). Rorty´s appreciation of “Strawson's only good transcendental argument for the 'necessary' character of material object concepts”, i.e. their TD in his (1959), as a PLA (1971,13), and Bennett´s rephrasing of the RI as a ´realism argument´ for the claim that selfconsciousness requires outer experience, the PLA being a version of that argument (1966, 203 ff.), may serve as further illustrations of the systematic intricacies of this approach.

    Its preoccupation with subjectivity and reciprocally, with its objective counterpart forms an implicit plea for postneokantianism. To use a Kuhnian, obviously un-Wittgensteinian concept, that move would be a regression. For, unlike Kant, the PLA does neither deal with Cartesian metaphysics (Hintikkaet al. 1986, 250) nor with Cartesian epistemology. It rather shows e.g. in §§269, 278 the redundancy of the ´transcendental´ subject, as well as of the metaphysical private object, sensation, most notoriously in §§272, 293 and 299.

    By contrast, Weinert´s observation in Wittgenstein of a trend towards expliciteness (1983, 423, 434-435) indicates a progressive problemshift. As such the emergence of a sharper awareness of the linguistic counterpart of the Paradox of Transcendental Knowledge, the intrinsic link between the unknowability of things considered independently of our epistemic procedures and the conceptual framework they utilize, and their unknowability, is also noteworthy. Wittgenstein´s awareness of this latter paradox, the mutual dependence of linguistic relativity and the ineffability of semantics (Hintikkaet al. 1986, 5, is expressed in his calling the sensation itself “not a something, but not a nothing either!” (§304). Wittgenstein does not merely offer a Kripkean sceptical solution (1982, 95) for this paradox. Instead, he proposes a ‘radical break’ with the idea that language always functions as a means to convey thoughts. More than welcoming Kant´s ´cognitive theoretical analyses´ to give direction to cognitive sciences, as Kitcher proposes (2000, 61), contemporary philosophers should study the consequences of this break.


    1. Aschenberg, R. 1982 Sprachanalyse und Transzendentalphilosophie, Stuttgart: Klett-Cotta.
    2. Bennett, J. 1966 Kant's Analytic, Cambridge: Cambridge University Press.
    3. Brueckner, A.L. 1983 “Transcendental arguments”, Noûs, 17, 551-575.
    4. Glock, H.J. 1991 “Philosophical Investigations section 128: ´theses in philosophy´ and undogmatic procedure”, in Arrington, and Glock, H.-J. (eds.), Wittgenstein´s Philosophical Investigations. Text and context, London: Routledge, 69-88.
    5. Gram, M.S. 1971 “Transcendental Arguments”, Noûs, 5, 15-26.
    6. Grayling, A.C. 1985 The refutation of skepticism, Illinois: La Salle.
    7. Hacker, P.M. 1972 Insight and Illusion. Wittgenstein on Philosophy and the Metaphysics of Experience, Oxford: Clarendon Press.
    8. Hintikka, J. 1972 “Transcendental Arguments: genuine and spurious”, Noûs, 6, 274-281.
    9. Hintikka, M. B. and J. 1986 Investigating Wittgenstein, Oxford: Basil Blackwell.
    10. Kant, I. 1990 Kritik der reinen Vernunft (1st ed.: 1781 =A and 2nd ed. = B), Hamburg: Felix Meiner Verlag.
    11. Kant, I. 1993 Prolegomena zu einer jeden künftigen Metaphysik, die als Wissenschaft wird auftreten können (1783), Hamburg: Felix Meiner Verlag.
    12. Kripke, S.A. 1982 Wittgenstein on rules and private language. An elementary exposition, Oxford: Basil Blackwell.
    13. Niquet, M. 1991 Transzendentale Argumente. Kant, Strawson und die sinnkritische Aporetik der Detranszendentalisierung, Frankfurt am Main.
    14. Prauss, G. 1974 Kant und das Problem der Dinge an sich, Bonn.
    15. Rorty, R. 1971 “Verificationism and transcendental arguments”, Noûs, 5, 3-14.
    16. Schaper, E. 1972 “Arguing transcendentally”, Kant-Studien, 63, 101-116.
    17. Stern, R. 1999 Introduction in Stern, R. (ed.) Transcendental arguments. Problems and prospects. (Mind association occasional series), Oxford: Oxford University Press, 1-11.
    18. Stevenson, L. 1982 “Wittgenstein's transcendental deduction and Kant's private language argument” Kant-Studien, 73, nr. 3, 321-337.
    19. Strawson, P. F. 1959 Individuals. An essay in descriptive metaphysics, London/New York: Routledge.
    20. Strawson, P. F. 1966 The bounds of sense. An essay on Kant's Critique of pure reason. (8th ed. 1993), London/New York: Routledge.
    21. Stroud, B. 2000 “Wittgenstein´s ´Treatment´ of the Quest for ´a language which describes my inner experiences and which only I myself can understand´ (1983)”, in Meaning, understanding and practice. Philosophical essays, Oxford: Oxford University Press, 67-80.
    22. Weinert, F. 1983 “Ways of criticizing metaphysics: Kant and Wittgenstein”, Kant-Studien, 74, 412-436.
    23. Wilkerson, T. E. 1975 „Transcendental arguments revisited“, Kant-Studien, 66, 102-115.
    24. Wittgenstein, L.W. 1958 Philosophische Untersuchungen - Philosophical Investigations (3rd ed.1968), Oxford: Basil Blackwell.
    25. Wittgenstein, L.W. 1961 Tractatus Logico-Philosophicus, London: Routledge & Kegan Paul.
    Karel Mom. Date: XML TEI markup by WAB (Rune J. Falch, Heinz W. Krüger, Alois Pichler, Deirdre C.P. Smith) 2011-13. Last change 18.12.2013.
    This page is made available under the Creative Commons General Public License "Attribution, Non-Commercial, Share-Alike", version 3.0 (CCPL BY-NC-SA)


    • There are currently no refbacks.