Theory-Change And The Logic Of Enquiry: New Bearings In Philosophy Of Science

Review of Metaphysics, The, Sept, 1999 by Christopher Norris

ANGLO-AMERICAN PHILOSOPHY OF SCIENCE has tended to define itself squarely against the kinds of so-called metaphysical approaches that have characterized so-called continental philosophy in the line of descent from Husserl. Indeed, Husserl's project of phenomenological enquiry was the target of criticism by Frege--and later by Gilbert Ryle--which pretty much set the agenda for subsequent debate.(1) That project seemed to them some form of argument that reveals his basically psychologistic approach, one that purported to address issues of truth, validity, rational warrant, and so forth, but which fell far short of the logical rigor attained by thinkers in the analytic tradition. Thus Husserl might claim--like Descartes and Kant before him--to be raising questions about the a priori forms of human knowledge and experience, forms that were given or necessarily presupposed in every possible act of cognition.(2) Moreover, he might claim to have advanced beyond Kant in distinguishing more clearly between formal and transcendental logic, or judgments whose necessity follows from the ground rules of this or that logically binding system of thought, and judgments that result from a rigorous reflection on the genesis and structure of human understanding in general.(3) However these claims counted for little with Husserl's critics in the other, that is, post-Fregean analytical camp. What they chiefly objected to in Husserl's project was the approach via thoughts and ideas in the mind of some perceiving or reasoning subject, even though Husserl was very often at pains to reject any merely empirical or psychologistic construal of his claims. To their way of thinking, all this talk about transcendental truth- and validity-conditions was just another variant of the bad old Cartesian-Kantian retreat to consciousness as the last court of appeal in epistemological matters. Only by rejecting that entire line of thought--that is to say, by adopting a strictly analytical or logico-semantic approach--could philosophy at last break free of its attachment to naive, subject-centered, or metaphysical notions of meaning and truth.

More recently some analytic philosophers, Michael Dummett among them, have begun to question this doxastic account and to offer a more nuanced appraisal of the differences between Frege and Husserl.(4) Nevertheless there is still a marked tendency to suppose that, wherever such differences show up, it will always be a matter of Frege's having pointed the best or most rigorous and logically adequate way forward, and of Husserl's having slipped--despite all his strenuous protests to the contrary--into some form of argument that unwittingly reveals his residual Cartesian premises. Now it seems to me that these belated, rather grudging gestures of atonement are partly a result of certain emergent problems within the analytic tradition, in particular those raised by Quine in his famous attack on the two last dogmas of Carnap-style logical empiricism.(5) That is to say, they have to do with the impossibility, as Quine sees it, of maintaining any version of the logical-empiricist dichotomy between matters of fact and truths of reason, or synthetic and analytic orders of statement. This distinction was crucial to the logical-empiricist program and also, in a somewhat less doctrinaire form, to the whole enterprise of analytical philosophy in the wake of Frege and Russell. Whether Quine actually succeeded in giving it the coup de grace is still very much an open question, not least, as I shall argue, in view of his failure to offer any convincing alternative approach to issues of truth, meaning, and interpretation.

So there is a case for taking a different view of the grounds for dispute between the two philosophical traditions. This revisionist approach would go much further than Dummett in acknowledging both the importance of Husserl's project and the various problems that analytic philosophy has encountered in its attempt to enforce the Fregean veto on talk of consciousness, intentionality, ideas, eidetic structures, the noetic/noematic distinction, and so forth. In so doing it would raise very different objections to logical empiricism from those put forward by Quine and other thinkers who continue to endorse that veto. After all, Quine is insistent that philosophy of language should adopt a strictly extensionalist semantics which finds no room for such opaque or intensional items as beliefs or propositional attitudes. Thus, for instance, while it is jointly true that "W. V. Quine is a Harvard philosopher and author of the essay "Two Dogmas of Empiricism,'" there is no equivalent conjunctive truth value which attaches to that statement when preceded by the phrase "Mary believes that." For she might as it happens be in error about various facts, among them his academic calling, or his institutional affiliation, or his having written that particular essay, or perhaps about the name of whomever it was that answered to one or more of those identifying descriptions. (Maybe she thinks it was written by Hilary Putnam or some other well-known Harvard philosopher.) In short, such statements fail to meet the test of substitutivity salva veritate for terms--such as "Quine" and "the author of `Two Dogmas of Empiricism'"--which pick out an identical referent and are therefore extensionally equivalent. Hence also his argument against the possibility of quantifying into modal contexts, a practice which--on Quine's submission--creates all manner of insoluble problems, such as the breakdown of logical equivalence between "9 is necessarily greater than 7" and "The number of planets is necessarily greater than 7."(6)