From the ALWS archives: A selection of papers from the International Wittgenstein Symposia in Kirchberg am Wechsel

1. Introduction

In the introduction of Individuals Strawson expounds his idea of descriptive metaphysics (Strawson 1959 9 ff.; cf. A834 = B862). The subtitle of Individuals, “An essay in descriptive metaphysics”, indicates that Strawson is concerned with an elaboration of this idea. In this respect, Strawson's metaphysics is meant to be similar to Kant's (B24), which, with Aristotle's, equally is called descriptive. Strawson's subsequent The Bounds of Sense is the outcome of his decision “that (he) must try to get to grips with the work [CPR, k.m.] as a whole” (Strawson 2003 8). This development could, therefore, arguably be appealed to in making a case for Strawson's Kantianism.

This book, though, has a multifarious purpose. For it is a compound of Strawson's polemic intention to “give decisive reasons for rejecting some parts [of CPR, k.m.] altogether” and his reconstructive intention to “indicate (...) how the arguments and conclusions of other parts might be so modified or reconstructed as to be made more acceptable (...)” (Strawson 1966 11). Unsurprisingly, therefore, Strawson's alleged Kantianism became a matter–and probably also a source–of much controversy. A prompt settlement of this stubborn controversy by just qualifying it as a paradigm of analytic Kantianism is unlikely. For if it is to be taken as “a distinctly analytic interpretation, defence, and elaboration of Kant's ideas” (Glock 2003 16 f.), this nomenclature can hardly satisfy the sceptic about its distinctive method: connective analysis. Assessing Strawson's Kantianism might, however, clarify its systematic and historical position.

This paper attempts to do so, taking Strawson's exposition of his project as point of departure. I will argue that Strawson's project shares a similarity with the later Wittgenstein; its Kantian remnants, however, prohibit it to team up with the full potential of Wittgenstein's linguistic analysis.

2. Descriptive metaphysics

Departing from Strawson's definition of 'metaphysics' in The Bounds of Sense, descriptive metaphysics is the “description of the limiting framework of what we can conceive of or make intelligible to ourselves as a possible general structure of experience” (Strawson 1966 15). This definition might seem ambiguity-ridden, as it relies upon what is meant by 'description'. It thus comprises both the “important and interesting philosophical undertaking” (Strawson 1966 15) of an inquiry, which “is concerned with describing and clarifying the concepts we employ in discourse about ourselves and about the world, and in elucidating their relationships–their forms of relative priority, dependency and interdependency” (Hacker 2003 49), and its outcome.

Hacker observes that Strawson's project is continuous with traditional metaphysics in its quest for the “most general forms of connectedness that permeate our conceptual scheme (...)”, while it departs from it, insofar as it yields “insight only (...) into the forms and structures of our thoughts about reality” (Hacker 2003 62; my emph.). This suggests a dissociation of the epistemological element of the description–the aforementioned philosophical inquiry–from the ontological status of the description–the result of that inquiry: “the concepts we employ etc.” It is questionable, however, how this inquiry can be said to be continuous with the tradition even if its result differs categorically from the traditional one: “the (knowledge of) the primary causes and principles” (cf. Aristotle 1989 A 981b 26 – 982a 20).

If such a dissociation is defensible at all, it is improbable in Strawson's case. For, however Strawson's strategy to “develop a conception of the a priori in which pure intuition play(s) no role” justifies his assignment to the semantic tradition (cf. Coffa 1991 22), the shift from ontology to semantics that looms in Strawson's project has not cleared away all remnants of the Kantian convertibility of the epistemic and metaphysical conditions of experience (cf. Aschenberg 1978 335). Hence, the ambiguity of the definition of descriptive metaphysics somehow is inevitable. This is because the way theoretical assumptions about the subject-matter of this inquiry–the conceptual framework to be described–are intertwined with its method, without this intertwining being given due account. For however Strawson acknowledges “having been subtly and in part consciously influenced by it [CPR, k.m.] in (his) own independent thinking about metaphysics and epistemology (in Individuals)” (Strawson 2003 8), Kant's epistemological considerations regarding the possibility of metaphysics as a philosophical discipline (cf. Kant 1993 §40 ff.), are not parallelled in Strawson. As one critic puts it, Strawson's “novel merger of the virtues of cautious and piecemeal analysis with techniques of bold generalizations and systematic theorizing concerning the character of 'our conceptual scheme' (...) results (...) from a failure to attend sufficiently to the character and implications of their interconnection” (Glouberman 1976 364).

I will illustrate this by the way Strawson demarcates descriptive metaphysics, and, corollarilly, distinguishes its method from reductive analysis. It is on the interface of both components, indeed, that the aforementioned intertwining of theoretical and methodical aspects appears. This intertwining, and the ambiguity that goes with it, constitute a line of justification for Strawson's project, and thus provide, in a way, its prolegomenon. An attempt to construe a counter to relevant objections against it, with recourse to this 'prolegomenon' might show this.

To start with the demarcation, Strawson speaks of a “limiting framework of necessary general features of experience” (Strawson 1966 15; my emph.). To grasp the meaning of this phrase–and of its variants that occur throughout the book–I recall that Strawson demarcates descriptive metaphysics which is “content to describe the actual structure of our thought about the world”, from revisionary metaphysics “which is concerned to produce a better structure” (Strawson 1959 9). I will call this demarcation: demarcation₁. Strawson also draws a demarcation line (demarcation₂) between “gramatically permissible description(s) of a possible kind of experience”, which we could imagine, and a subclass of those: “truly intelligible descriptions” (Strawson 1966 15).

A first objection concerns demarcation₂. As Davidson observes, demarcation₂ assumes that many imagined worlds are seen from the same point of view. Thus, Strawson's Conceptual Invariance Thesis states that ‘our conceptual scheme’ “is constant over time and between different languages” (Haack 1979 361). Hence, demarcation₂ requires a linguistic dualism of concept and content, insofar it supposes that a fixed system of concepts is used to describe alternative universes. It thus rests on the idea of a distinction between theory and language; mistakenly, though, for meaning is contaminated by theory (Davidson 1984 187 f.).

The effectiveness of this objection could be weakened, however, by differently from Davidson, emphasising the methodological, rather than the semantical aspect of demarcation₂. Such a reading could be based on the assumption that demarcation₂ can be mapped on demarcation₁. Demarcation₁ should thus be understood in the aforementioned methodical sense of a particular type of inquiry. In this sense, revisionary metaphysics pertains to regimentations of our ordinary 'discourse about ourselves and the world' (cf. Hacker 2003 49); descriptive metaphysics to the employment of connective analysis as the method of analysis of this discourse. Strawson's methodological exposition of his connective model of analysis, which sets its apart from the reductive or atomist model (Strawson 1985 31 f.), viz. Quine's ontological reductionism, viewed as a consequence of his regimentation of ordinary concepts, and Moore's linguistic reductionism, which overlooks the (inter)dependency of concepts (Strawson 1985 59, 43) supports this reading.

To take full recourse to what has been labelled Strawson's prolegomenon to descriptive metaphysics, this reading should be supplemented by a reading that emphasises the semantical aspect of demarcation₂, and which is consonant with a plausible reading of demarcation₁. Here, some semantical assumptions of Strawson's logical theory can serve as point of departure. Among these assumptions, which are recurrent in Strawson's work, from On Referring onwards, and which are explicitly stated in Individuals but form implicitely in The Bounds of Sense a heuristic framework for the interpretation and reconstruction of CPR, are: (i) “the central importance of the subject-predicate distinction; (ii) “the role of particulars as objects of reference”; (iii) the conceptual “priority of particulars over universals” (Haack 1979 362). These assumptions warrant a semantic reading of demarcation₂ which is consonant with demarcation₁. It is on their account that Strawson's logical theory renders the modality of a priori necessity to the (inter)connections that make up the significance, in Strawson's sense of intelligibility, of the conceptual scheme which it is intended to describe. As such, e.g. assuptions (i) and (ii) jointly warrant in Strawson the objectivity of referring expressions in a subject-predicate sentence similarly as does the category of substance in Kant (cf. B129). By force of these assumptions, though Strawson does not state this explicitly, demarcation₂, of course, echoes Kant's distinction between the use of the categories in mere thinking and their application to intuitions “by which a thing is given” (B146). In this respect Strawson's logic has a transcendental aspect at it. For, although Strawson states e.g. that the performance of the referential task of certain linguistic expressions “requires no transcendental explanation”, it is precisely by the use of uniquely referring expressions that “existential assertions may be implied” (Strawson 1950 335, 343).

Interestingly, though, Strawson shares his emphasis on the use of language with Wittgenstein (Wittgenstein 1958 §43), as does his suspicion of logical regimentation of ordinary language (cf. Strawson 1950 344; Wittgenstein 1958 §38). However, unlike in Wittgenstein, where 'use' extends to the grammar of expressions within a language game (cf. Hintikka 1973 55), in Strawson it seems to be restricted to their pragmatic, performative aspects.

A second objection is derived from Stroud and concerns the argument for the a priority of some basic concepts. The criterion to decide whether a concept is basic is whether it answers the demands of Strawson's austere issue (PS_S) of Kant's so-called Principle of Significance (PS_K). By this epistemic principle Kant distinguishes representations which are informed by intuition from empty representations (cf. B75, B150). PS_S “forbids, as empty, the employment of any concept for which no empirical conditions of application could possibly be specified” (Strawson 1966 192). It can be noticed that PS_S resembles the neo-positivist verificationist criterion of meaning, because it likewise is semantic in scope, as it pertains to concepts belonging to a conceptual framework. As such its employment in descriptive metaphysics could resuscitate the objection Stroud raised against the argument, in Individuals (Strawson 1959 38 ff.) for our knowledge of the existence of objective particulars. Stroud argues that insofar as its soundness requires the introduction of a verification principle, as he thinks it does, it is superfluous as a transcendental argument against epistemic scepticism about the existence of objective particulars (Stroud 1968 247).

However, although PS_S might resemble the verificationist criterion of meaning, its method of application, unlike the (neo-)positivist method (cf. Coffa 1991 327) is not merely verificationist in the sense Stroud would be inclined to take it, but rather transcendental. For it is applied in a test to establish the a priori status of some concepts within a conceptual framework, rather than their meaning. The scheme of this test is reductive, as opposed to deductive, and, as such, is not logically valid. It argues from a conditional assertion and its known consequent to its unknown antecedent (Bocheński 1954 101, 102 f.). The test, as Strawson conceives it, consists in the sequential performance of this scheme of reasoning in its progressive and its regressive variants, which Bocheński calls verification and explanation respectively. First, the admissibility of a concept as basic is (progressively) verified, by testing if it answers the demands of PS_S. Then its a priority is explained (regressively) by “fram(ing) a counterfactual antecedent from which we could derive (...)” the consequent that “we should have no use for this concept (Strawson 1966 115).

The degree of universality and necessity of the a priori concepts that pass this test is a function of the epistemic use of the powers of our imagination to ‘frame counterfactuals’. In contrast, there is no such appeal to such powers in Wittgenstein, where the words “I can’t imagine the opposite” e.g. of knowing to feel my pain, merely is a defence against a grammatical proposition being presented in the form of an empirical proposition (Wittgenstein 1958 §251). Therefore, if I can agree with Hacker's assessment of Strawson's descriptive metaphysics as being metaphysics only in an attenuated sense, “just more grammar”, that is, “in Wittgenstein's extended sense of the term” (Hacker 2003 54, 59), it is only with the first part of it.

3. Conclusion

To conclude, I have shown that a defence of Strawson’s project of descriptive metaphysics can draw upon an ambiguity in Strawson’s exposition of it, by mobilizing its transcendental tendencies. This is because this ambiguity is due to the unreflective intertwining of theory and method in this exposition and the unconscious denial of a contamination of meaning by theory that goes with it; hence Strawson’s Kantianism.

Strawson’s descriptive metaphysics, as it concerns ‘our conceptual scheme’ about the world, involves a shift of focus from language to the world (cf. Strawson 1950 328f.), unlike Wittgenstein’s descriptive analysis of language, which remains within language games. Due to this digression, Strawson's analysis does not share the full potential of Wittgenstein's analysis.^*