From the act of judging to the sentence. The problem of truth bearers from Bolzano to Tarski (Q850995)

This is a posthumously published book by Artur Rojszczak, edited by Jan Woleński. Its central problem, ``the truth bearer problem'', concerns what category of item the predicate ``true'' is predicated of. [The question whether truth \textit{is} predicable in some definite sense, like in Tarski's semantic questionings, is not considered in the book.] Interesting historical studies, mainly centered around the Lvov-Warsaw School, illuminate early insights into truth bearers, from acts of judgment to Tarski's alleged view of sentences as truth bearers in Tarski 1956/1933 [\textit{A. Tarski}, ``The concept of truth in formalized languages'' (Polish), Travaux Soc. Sc. Varsovie, Cl. III, No. 34 (1933; JFM 58.1001.01); English translation in: \textit{A. Tarski}, Logic, semantics, meta-mathematics. Papers from 1923 to 1938, 152--278 (1956; Zbl 0075.00702); see also the German version in Stud. Philos. 1, 261--405 (1935; Zbl 0013.28903)]. Examples of the considered philosophical questions are: ``How far does the choice of truth bearer affect the conceptual framework for the theory of truth?''; ``Does the choice of truth bearer exclude some possibilities in developing a theory of truth?'' Concerning Tarski, the author writes (page 205): ``The bearer of truth, according to Tarski's definition of truth as given in his 1933 work, is the sentence. It is of a sentence in a given language of which the metalinguistic predicate `true' is affirmed.'' [Yes, provided the metapredicate is realizable!] But on page 248 of Tarski 1933/1956 we find doubts concerning `the language of the general theory of classes'. Tarski writes: ``Should we succeed in constructing in the metalanguage a correct definition of truth, then the metalanguage -- with reference to the above interpretation -- would acquire that universal character which was the primary source of the semantical antinomies in colloquial language''. Although the author quotes a plurality of definitions of ``sentence'' from Tarski 1933/1956, and tries to unify them in a single definition (page 211), he seems to be of the impression that Tarski in his 1933 work has a unique definition of truth. Yet Tarski here provides several attempted construals of definitions of \textit{true sentence} for various languages [notice that Tarski conceives of `language' to include intended deductive system]. These are not explained in the book and nothing is said of Tarski's central impossibility results for constructing an adequate definition of true sentence [or for the impossibility of capturing truth as a strictly predicative concept]. In fact, in Tarski 1933/1956 three languages are considered, colloquial language and two formal languages. For colloquial language Tarski argues impossibility of ``a consistent use of the expression `true sentence' which is in harmony with the laws of logic and the spirit of everyday language''. For the the ``language of the calculus of classes'', Tarski claims success for his truth definition (in terms of satisfaction of the theorems by given interpretation structure). However, the satisfaction relation is here impredicative; cf. [\textit{S. Feferman}, Ann. Pure Appl. Logic 126, No. 1--3, 5--13 (2004; Zbl 1055.03003)]. Again, for the language of ``the general theory of classes'', Tarski argues impossibility for constructing an adequate definition of truth (cf. quote above). In the reviewer's understanding the view of \textit{sentences as bearer} of semantic concepts, like truth, is tied with the \textit{presupposition} that we write and read in a (universal) language \textit{that allows describability} (i.e., in sentences) of the semantic concepts. Or, when using predicate languages, allows \textit{predicability}. Hence we associate `\textit{sentence as truth bearer}' with `\textit{truth as predicable concept}'. In this perspective the reviewer would not find it likely that Tarski (with his impossibility results concerning formal definability of `true sentence') would consider truth as (strictly) bearable by sentences, or as (formally) predicable. The reviewer finds the author's willingness to attribute the truth bearer terminology to various authors, even if these themselves don't use it, noteworthy. By further example, on page 23 the author refers to Russell as giving a separate treatment of the question of the bearers of truth in his book [The problems of philosophy. London: Williams and Norgate; New York: Henry Holt and Co. (1912; JFM 43.0095.01)]. Yet Russell never uses the terminology of truth bearer -- but speaks of truth and falsity as \textit{extrinsic properties}. Truth and falsehood of a belief always depends upon something which lies outside the belief itself. The author's historical expositions of \textit{judgements} are on several philosophical levels. Concerning Brentano's and Kotarbinski's reisms, the author writes: (page 108) ``Now there are two ways of correctly predicating truth of two different kinds of things, i.e. of a judger and of linguistic things such as utterances or inscriptions. A true sentence is the utterance of a truly judging person. Are these two ways of speaking about truth both valid and interchangeable?'' \dots ``We say of sentences that they are true or false with regard to their relation to the truth and falsity of the judgments of which they are expressions, i.e., actually due to the relations to the judger.'' In this context, the author might have attached to Tarski's attempt to grasp the intentions in the classical conception of truth (`true-corresponding with reality') in his ``Convention T'', judged a general characterization of adequate formal definitions of `true sentence'. Allowing him to judge truly (by logical reasoning) that truth in sufficiently rich languages is not formally predicable of its sentences. In the reviewer's experience, the difficulties of conceiving semantic concepts (like truth) predicatively naturally lead into attempts to conceive semantic concepts non-predicatively, for example by \textit{holistic comprehension} rather than predicative; cf. [\textit{L. Löfgren}, Found. Sci. 9, No. 2, 135--189 (2004; Zbl 1056.03006)]. Holistic language is holistic in a sense of not being autonomously fragmentable. It does not allow its own strict fragmentation into syntax, semantics, pragmatics. However, there is a tension between describability and interpretability which can be weighed, for example towards high describability with a small but nonvanishing interpretation residue, or correspondingly towards high interpretability. This is compatible with Tarski's nondefinability statements. Yet holistic language is understandable by holistic comprehension. The reviewer has found the author's expositions of early philosophical semantic insights interesting in the holistic perspective. For example, the definition (page 211) of `sentence', a traditionally syntactic object, which appears somehow nonfragmentable from its semantic functionings in language. And Ossowski's conception of `sentence' as `semantic product' of second order, and of language. Surprisingly, however, there seems to be less interest in the concept of language itself in the book, whereas in the holistic perspective, it is the phenomenon of language and its natural evolution that is the primary source of our communicable knowledge and its creation in language.