Variations on a thesis: intuitionism and computability (Q1102263)

The bulk of this paper (Variation I) is in the tradition of academic philosophy about hackneyed isms, and thus outside the scope of this journal. But for readers of good will it has interest provided they are not carried away by its majestic prose. Thus, on p. 560, where computability is dubbed ``the central issue of intuitionistic foundations'', such readers will supply specifics: What difference does it make, if any, when the computations considered operate on meanings and other socalled mental, and not only on formal, objects? pedantically: difference on the basis of what is known of those mental objects (and operations on them). Superficially, knowledge - implicit in the laws - of mere propositional logic is enough; Gödel's translation into modal terms uses \(\square (\square p\to p)\), where \(\square\) means provability, and this is false for formal derivability (including a familiar minimum). But the readers in question will think (for themselves) of another translation, say *, where \((p\to q)^*\) is \((p^*\to q^*)\wedge \square (p^*\to q^*)\). The modal properties required here are indeed satisfied by suitable formal derivabilities. Incidentally, the propositional laws valid under Gödel's translation for formal derivability are set out by \textit{A. Visser} [Stud. Logica 40, 155-175 (1981; Zbl 0469.03012)]. Kleene's (q-) realizability is of course another interpretation, for which the specific issue above is left open by all currently formulated intuitionistic knowledge (of mental objects). Without the particular good will postulated above, such generalities will be viewed as below the threshold of any informed discussion of the issue in question; as were Brouwer's tirades, who, contrary to footnote 14 on p. 577, never cited specific ``knowledge of a non-trivial part of mathematics'' in this connection, but relied on isms and exploration of his innermost consciousness. Variation II (pp. 561-576) begins with the impeccable reminder that the potential of intuitionistic mathematics - tacitly, of the author's kind, in terms of recursiveness rather than continuity (choice sequences, jazzed up in sheaves and the like) - lies in ``new, characteristically intuitionistic'' theorems. But the particular evidence for that potential, on pp. 562-576, is simply embarrassing; compared either to ordinary analysis, say, in Newton's days or even to the other kind of intuitionistic mathematics. What pp. 562-576, which are elegant, sometimes ingenious and of course coherent, show convincingly is that these properties are not enough by a long chalk for what the trade coyly calls good mathematical taste \((= judicious\) selection). On a less ethereal plane - and contrary to 1.-14 and 1.-13 on p. 561, the restriction, in the tradition of purity of method, to - intuitionism has not proved nearly as effective for knowledge of finite and computable objects as the alternative: of embedding them in - suitable, not just big (set-theoretic) - non-computable structures and exploiting properties of the latter. Of course, philistines have been morally certain of some conclusion of this kind from a rough acquaintance with intuitionism for a long time; to be compared with scepticism about a perpetuum mobile long before more precise work on conservation laws for energy and mass (separately, and their correction by \(e=mc^ 2\) in circumstances hard to realize on earth). Viewed this way a third variation, not considered by the author, concerns a comparably precise analysis of the philistine conclusion.