¹ For a further discussion of the concept of a law of nature, with particular attention to the “problem of induction,” i.e. the problem of justifying the adoption of a material move or material auxiliary position into our language, see below, sections 57–72.

² Hector Castaneda, The Logical Structure of Moral Reasoning, a Ph.D. thesis submitted to the Faculty of the Graduate School at the University of Minnesota, April 1954.

³ Just as the term “bishop,” which occurs in the language of both Texas and ordinary chess, can be correctly said to have a common meaning—indeed, to mean the bishop role, embodied in the one case by pieces of wood, and in the other by, say, Chevrolets, and which Frenchmen would refer to as le role de l'evecque—so “α,” on the above assumptions, can correctly be said to mean a certain linguistic role, a role which is embodied in different linguistic materials,—in English by the sound redd, and in German by the sound roat. For a discussion of linguistic roles thus conceived, see my “Quotation Marks, Sentences and Propositions,” Philosophy and Phenomenological Research, 10, 1950, pp. 515–525; also “The Identity of Linguistic Expressions and the Paradox of Analysis,” Philosophical Studies, 1, 1950, pp. 24–31.

⁴ For a more elaborate discussion of semantical statements and the disastrous consequences to philosophy of assimilating them to relation statements, see my “Is there a Synthetic A Priori?” Philosophy of Science, 20, 1953, pp. 121–138, especially pp. 134 ff.

⁵ In a footnote to page 195 of a paper on “Particulars,” Philosophy and Phenomenological Research, 13, 1952, I wrote, “If, as I am claiming, the sentences which formulate what we regard as the laws of the world in which we live are true ex vi terminorum, then how can it be rational to abandon such a sentence? What role could observational evidence play in the “establishing” of sentences which are to be true ex vi terminorum?

“The inductive establishing of laws is misconceived if it is regarded as a process of supplementing observation sentences formulated in a language whose basic conceptual meanings are plucked from “data” and immune from revision “Hume's Principle”). The rationality of “induction” is, rather, the rationality of adopting that framework of material rules of inference (meanings—even for observation predicates) and, within this framework, those (sketchy) statements of unobserved matters of fact (world picture) which together give maximum probability to our observation utterances interpreted as sentences in the system. Only if we do this do we adopt (and this is, of course, an analytic proposition) that world picture which is “most probable on the basis of our observations.”