Hume and Nonlogical Necessity

95. HUME AND NONLOGI CAL NECESSITY Flew substantially agrees with our main themes in Causal Powers, though he thinks our discussion should begin closer to the work of Hume himself. What to reply? Needless to say, we find no fault in his agreement with us and are happy for his support. The answer to the second part of his response is far more complicated and requires a bit of background from Causal Powers. While contemporary Humeans differ in important details, they share certain commitments which constitute what we call the Humean tradition. This tradition and contemporary refinements of it are the targets of our criticisms and provide the crucial foil for presenting our own positive theory of natural necessity. Two important strands of the tradition, as we characterized it, are the In-principle (No-contradiction) argument and the argument against the possibility of a priori knowledge of matters of fact. (i) Since nature may change its course, 'C and not E' is not self-inconsistent. To the Humean the phrase "change in the course of nature" seems intuitively clear; to us it seems utterly opaque and itself in great need of analysis. We show that there are two possible interpretations of the phrase, one of which is selfinconsistent while the other, though coherent, does not establish the Humean conclusion that there is no necessary connection between matters of fact, (ii) The Humean assumes that 'p is contingent' and 'p is a posteriori . ' and 'p is necessary' and 'p is a priori, ' are materially equivalent, and then argues easily that since all causal propositions are a posteriori none of them can be a priori and so necessary. We offer numerous arguments to break up the assumed equivalencies and thus defeat the conclusion 96. of the Humean argument. We show that while all scientific propositions are indeed known a posteriori it does not follow that there are no justifiable propositions about natural kinds. Philosophers in other traditions have made a similar point in showing that "metaphysical necessity" does not entail "epistemic necessity." Indeed, in contrast to many contemporary Humeans , we are complete fallibilists and even reject the notion that there is a self-contained given which is incorrigible. In portraying various strands of "the Humean tradition" in Causal Powers, we were, of course, imputing these views to Hume himself. How could it be otherwise? In referring to the Humean tradition instead of to Hume we meant only to indicate (as do contemporary Humeans) that we were concerned only with what we take to be still very much alive in Hume's work. Hence it is puzzling to read Flew' s remark that we relate our own contributions to Hume "only most distantly." That we and contemporary Humeans are always correct in what we ascribe to Hume and in what we take to be still philosophically alive in his work may be open to question, but since Flew does not challenge any specific ascription there is no alleged misinterpretation to defend ourselves against. Flew suggests that we should show Hume to be wrong at the outset — convict him, so to speak, on the grounds of the inconsistencies in his own texts. He refers to the famous definition of 'cause' in the Inquiry, where Hume gives a factual rendition of 'cause' and then innocently equates it with a counterfactual counterpart. C. J. Ducasse, for many years an ardent critic of Hume texts, made this point nicely in 1924 in his splendid book Causation and the Types of Necessity. After quoting Hume's definition in the Inquiry Ducasse remarked: This statement, incidentally, is immediately followed by: "Or in other words where, if the first object had not been, the second never had 97. existed." But this, far from being only a different way of saying precisely the same thing, is on the contrary a precise way of saying a very different thing. For Hume very definitely maintains elsewhere that there is never any "if-ness" objectively observable at all, but only a "that-ness" constantly repeated in experience. So that only the first statement is consistent with his empiricism, and even in that first statement the words "are followed" should in strictness be replaced by...