This is a preview of subscription content, log in via an institution to check access.
Notes
Instead of talking of the truth value of sentences containing pragmatic contexts we may as well talk of the pragmatic value of the nonpragmatic sentence-like parts of those sentences. Thus “trivial” or “nontrivial” would be two different pragmatic values.
This solution has been suggested by Langford (The Philosophy of G. E. Moore, p. 326), Church (Journal of Symbolic Logic, 1946, involving Frege's highly dubious distinction between sense and reference), Carnap (Meaning and Necessity, pp. 63f). Cf. also Pap (Analytische Erkenntnistheorie, pp. 231f) whose position is, however, not very clear.
A similar consequence can be derived if we look a bit more closely at solution (b): Assume that, although ∼S′(A = A, A = D) (where “BC” has been contracted into “D”) nevertheless S′(A = A, A = E). Then "A = E" expresses what we might call an analysis of order ('), an analysis' of “A.” If this analysis is to be non-trivial we have to assume that ∼-S"(A = A, A = E), and S" not = S′ not = S not = S". For S = S" would imply ambiguity of analyses of the first order and S" = S′ or S′= S would involve a contradiction. In this way we generate a series of synonymies, S, S′, S" … Assume now that ∼S1 (A = A, A = P) but that there is no Q such that S1 (A = A, A = Q). In this case S1(X,Y) if and only if “X” and “Y” are two tokens of the same type. But this implies that ∼S1 (A = A, A = P) (P being such that S1−1(A = A, A = P)) asserts only a notational difference between “A = A” and “A = P,” i.e., we obtain the result which follows from our solution if we are not prepared to admit either that analyses may be ambiguous, or that there may be infinitely many concepts of synonymy.
It seems to me that Wittgenstein has emphasized just this point when saying that “If one tried to advancetheses in philosophy, it would never be possible to question them, because everybody would agree to them” (Philosophical Investigations, p. 128). This, indeed, applies to a philosophy which “leaves everything as it is” (loc. cit., p. 124) and this is overlooked by philosophers who want to have it both ways, i.e., who want to have a philosophy which is purely descriptive (of either language, or of pure phenomena, or of ideas in the Platonic heaven: analytic philosophy, phenomenology, Aristotle's theory of real definitions) and which at the same time leads to discoveries, extends the knowledge of its subject matter, etc. It is to be admitted that many so-called analytic philosophers, when analyzing, do change the language they use and do improve it (cf. Russell's “analysis” of number). But theideal of pure analysis which only “exhibits” (Koerner's expression) the rules of the underlying language is still prevalent among so many philosophers that it must be criticized.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Feyerabend, P.K. A note on the paradox of analysis. Philos Stud 7, 92–96 (1956). https://doi.org/10.1007/BF02221761
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF02221761