The Semantic Theory of Knowledge

by Adam Olech (Author)
Monographs 296 Pages
Series: Modernity in Question, Volume 12


The aim of this book is the analysis of Kazimierz Ajdukiewicz’s meta-epistemological project of the semantic theory of knowledge and its implementations to solve certain traditional epistemological problems and their metaphysical consequences. This project claims that cognitive problems need to be approached from the perspective of language. One of the results of this analysis is the thesis that the philosophical-linguistic legitimisation for the meta-epistemological project is the philosophy of Edmund Husserl from his Logical Investigations. This is the philosophy that makes it possible to speak reasonably of a close relation between thinking and language and provides thereby the legitimisation for this project.

Table Of Content

  • Cover
  • Title Page
  • Copyright Page
  • About the author
  • About the book
  • Citability of the eBook
  • Contents
  • Introduction
  • I Historical and Philosophical Context of the Proposal of the Semantic Theory of Knowledge
  • II Analytic Approach to Philosophical Problems
  • III Presentation of Philosophical Problems in Nominalistic Form
  • IV On Method in Philosophy. Meta-Philosophy of Jan Łukasiewicz
  • V Meta-Epistemological Proposal of the Semantic Theory of Knowledge and its Analysis
  • VI Introduction to Formal Theory of Knowledge
  • VII Fundamental Epistemological Opposition
  • VIII Roman Suszko’s Formal Theory of Knowledge
  • IX Jan Woleński’s Formal Theory of Knowledge
  • X Was the Semantic Theory of Knowledge Inspired by the Results of Tarski’s Investigations?
  • XI Semantic Theory of Knowledge and the Problem of Truth Carriers
  • Conclusion
  • References
  • Index of names


The name “semantic theory of knowledge” was introduced by Kazimierz Ajdukiewicz. He used it for the first time in Problemat transcendentalnego idealizmu w sformułowaniu semantycznym [The Problem of Transcendental Idealism in Semantic Formulation], in which he wrote that the semantic theory of knowledge is such a way of pursuing the theory of knowledge that approaches

[…] the problems of knowledge deliberately from the perspective of language understood as a system of expressions equipped in meaning. It formulates its propositions in such way that they pertain to expressions, i.e. sentences and terms, however always treated as sentences and terms of a certain language that equips them in specific meaning.1

The aforementioned treatise is composed of two parts: a lecture explicating what semantic theory of knowledge is, which de facto is a meta-epistemological credo of this philosopher, and its implementation in the form of an analysis of Heinrich Rickert’s transcendental idealism. I separate the lecture from the entire treatise, since it enjoys a characteristic meta-epistemological autonomy, and I describe it as “a meta-epistemological proposal of the semantic theory of knowledge”. This proposal and its implementations– apart from the aforementioned one there were also others, which will be discussed further on – are one of the most profound accomplishments of the 20th century analytic philosophy.

I devote this monograph to two topics: an analysis of this proposal, which I consider in the context of semiotic and epistemological views of other representatives of the Lvov-Warsaw school, and secondly, an analysis of those works of Roman Suszko and Jan Woleński that are inspired by Ajdukiewicz’s implementations of this proposal, and indirectly – although the authors do not admit this – also by the proposal itself.2

The works of Suszko and Woleński that I have analyzed are classified as belonging to the formal theory of knowledge, since the tools they employ are at the same time semantic-logical as well as formal tools. On the other hand, Ajdukiewicz’s works in which he pursues his proposal are classified as simply belonging to the semantic theory of knowledge, since his semantic-epistemological analyses, although they rely on results obtained in logic, do without formal tools characteristic for Suszko’s and Woleński’s analyses and the symbolism connected therewith.

I wish to once again emphasize that when I discuss the semantic theory of knowledge I distinguish between the aforementioned proposal and its ←7 | 8→implementation, i.e. application thereof in analysis of traditional philosophical problems. The mere proposal, if it were to be defined volume-wise, is contained on two pages of the aforementioned treatise, and is de facto an introduction thereto. I consider it to be of such profound importance, and I focus my deliberations around these two pages, since I do not find in the entire philosophical analytic literature an equally convincing argumentation in support of the thesis that cognitive problems need to be approached from the perspective of language. Choosing this short proposal of semantic epistemology as the object of a considerable part of this monograph, I necessarily exceed the scope of Ajdukiewicz’s philosophical views, and I also address those problems which this philosopher did not handle, and which, in my opinion, should have been addressed. The present monograph, even when I am analyzing Ajdukiewicz’s proposal, is not only a comprehending analysis of this proposal, but also an interpretation thereof, for which only the author of the interpretation is responsible, and if this is the case, then potential transgressions connected therewith are the sole responsibility of the author of the interpretation and not the author of the interpreted text.

Those Ajdukiewicz’s papers that semiotically analyze traditional philosophical problems, in particular those whose subject is the dispute between epistemological and metaphysical idealism and epistemological and metaphysical realism are the implementations of the abovementioned meta-epistemological proposal. If an epistemological analysis of these problems were described as a “semantic theory of knowledge”, and indeed – as already mentioned – this is the name used in this respect, then the proposal of the semantic theory of knowledge could be described as “an answer to the question, whether a semantic theory of knowledge is possible, and if yes, under what conditions”?3

In the present work I do not exhaustively discuss semantic analyses of philosophical problems, although they are broadly commented upon, but I above all focus on the proposal itself whose implementation makes these semantic analyses possible, i.e. makes them meta-epistemologically legitimate, because I believe that this proposal has supra-individual character: it touches upon a fundamental problem of the relation between language and knowledge and solves it sufficiently universally for the solution to provide enough space not only for the philosopher who is the author of this proposal, but also for other philosophers who do not entirely agree with the philosophy of the author, or even philosophers who differ from him radically.

Nonetheless, when I discuss this proposal, I keep in mind its author’s directival theory of language which perceives idealisingly understood common language as a deductive system.4 And I take this fact, i.e. this theory, into consideration, although an obligation to do so does not follow from the proposal itself. If one ←8 | 9→however does take it into consideration, then – provided that the cognitive subject is determined in the categories of a linguistic subject5 – the cognitive subject may be perceived as a deductive system. And this makes it possible to draw epistemological conclusions from results arrived at in meta-mathematics with respect to cognitive capabilities of the cognitive subject.

Jan Woleński notes, not referring thereby to Ajdukiewicz’s theory of knowledge, that if a language contains the sign of negation, it cannot be perceived as a deductive system, since such language contains two mutually contradictory sentences, which – in accordance with the assumption of classical logic – cannot be true in any model.6 It is impossible not to agree with this observation, albeit not unconditionally, which means it is impossible to leave it without a comment, which has been especially elaborated in this monograph. At this point I would only like to say – defending Ajdukiewicz at the same time against such possible accusation – that Ajdukiewicz, understanding intensionally interpreted language as a deductive system, understands it both in a syntactic-pragmatic and in an assertive manner. A deductive system understood that way does not exists without a subject – the subject is its constitutive component, since the subject is to such an extent embedded into this system that the system «contains» it. Its cognitive activity (understood as cognitively within the strict meaning of this word) is limited by its linguistic activity, in other words: it is contained in its linguistic activity, and this means that every cognitive act of the cognitive subject, to deserve to be called cognitive, must at the same time be a linguistic act. I write about this on many pages of this monograph, in particular in chapter seven hereof. I am of the opinion that Ajdukiewicz’s firm believe that language is a deductive system is – with such understanding of the deductive system – identical to the proposition that a deductive system is a set of sentences affirmed by the subject and a set of consequences thereof; after all we may assume that the subject embedded in the language acts rationally, i.e. it does not affirm mutually contradictory sentences and that it disposes at least of such knowledge (proportionally understood knowledge) that is generated by the directives of direct and indirect rules of sentence affirmation applicable in the language, in which the subject is «embedded», i.e. by the axiological and deductive directives. This question, the question of understanding the language as a deductive system, will be constantly remerging throughout this monograph. I would like the reader to constantly keep in mind the above clarification; it will become clearer in the light of further deliberations.

After this clarification I return to Ajdukiewicz’s meta-epistemological proposal that is of interest to me. It is possible to accept this proposal, disregarding at the same time the directival theory of language, and then this proposal, becoming ←9 | 10→more general, states equally reasonably that if knowledge is to be definite in content,7 then concepts and judgements/propositions8 need to be treated as meanings of expressions, i.e. it is necessary to approach knowledge from the perspective of language. Since, if we speak of knowledge in the psychological sense (cognition), then these concepts and judgements are psychological meanings of expressions, but if we speak of knowledge in the logical sense, then these concepts and propositions are meanings in the logical sense – and the significance of the propositions presented only in this paragraph for the entire analytical philosophy, if they are valid, and they are in case of this philosopher, is difficult to underestimate.

The semantic theory of knowledge analyzes traditional philosophical problems approaching them from the perspective of language and does so in a manner that makes it possible to apply logical tools, which is particularly important for a semantic epistemologist. Thus, the proposal of the semantic theory of knowledge is an answer to the question whether it is possible to approach knowledge from the language perspective, and if it is, how to do it, as knowledge understood in the traditional manner is composed of the cognitive subject’s cognitive acts and their results. In other words, the proposal of the semantic theory of knowledge is an answer to the question, whether it is possible to translate a cognitive act and its result into a linguistic act and result, and it if its, how to do it, for this translation to be a translation as to the scope, and at the same a translation which perceives the language as a logically analyzable medium.9

I therefore treat the question about the philosophy of language as interesting, or even as indispensable for the proposal of the semantic theory of knowledge, since the philosophy of language is fundamental for the discussed proposal, as well as for every meta-philosophical declaration characteristic for an analytical philosopher proclaiming that instead of concepts and propositions it is possible and advisable to speak of expressions, i.e. of names and sentences, whose meanings these concepts and propositions are. When he presented his proposal, Ajdukiewicz did not legitimize it from the point of view of philosophy of language and to be precise from the point of view of the philosophy of language and knowledge understood as cognitive acts and results connected with them, i.e. as cognitive creations. And I find this question, the question of philosophical-linguistic legitimization of the meta-epistemological proposal, to be of key importance, which is reflected in this monograph.

If we however, pursuing the semantic theory of knowledge, abstracted from the real cognitive subject – from his cognitive acts and their results, from the rules of thinking this subject applies in its reasoning, etc., and if we abstracted from the objects cognitive acts of the real cognitive subject are directed at, i.e. if we only considered well-structured language and properly determined rules ←10 | 11→of creating language expressions and the rules of transforming them, and also if speaking of language understood this way, we understood it as an extensionally interpreted creation, then we could treat the meta-epistemological credo, i.e. what I describe as “the proposal of the semantic theory of knowledge” as redundant. In such case it would be sufficient to ascertain that the traditional theory of knowledge deals with the real epistemological opposition, i.e. the opposition: real subject of knowledge – object of knowledge, whereas we deal with the abstract opposition: language – model of language, which is the object of the investigations of logical semantics. Speaking of this opposition we could assume that it is an opposition analogous to the real opposition, since – and this would be our second assumption – if knowledge is to be of inter-subjective character, then it should be linguistically expressible, since only such knowledge in fact deserves to be called “knowledge”.

This approach to knowledge – generally speaking – is presented by the representatives of the formal-logic theory of knowledge, i.e. by Roman Suszko and Jan Woleński; I say “generally”, since the first segment of the semantic opposition: language – model of language contains a constructed subject of knowledge that – which is obvious in this situation – is of abstract character, and therefore whose «acts of consciousness» and «acts of self-consciousness» are names, predicates and sentences respectively of the objective language and the meta-language super-structured thereupon, and further whose rules of thinking are the rules of constructing expressions and the rules of transformation thereof, etc.

If we however assumed that this meta-philosophical credo, as already mentioned, is redundant, and if we, investigating the abstract semantic opposition language – model of language, wanted to draw from our investigations conclusions concerning the real epistemological opposition: subject of knowledge – object of knowledge, then we would need to reconsider redundancy of this assumption, since this credo describes the relations between cognitive acts and results and language, and therefore we should demonstrate these relations. The need to demonstrate these relations is also forced by the valid assumption that these two oppositions are analogous. But apart from demonstration of these relations we should also – if we want to draw the aforementioned conclusions– assume a standpoint with respect to further relations, namely those existing between formalized languages and the natural language. And if we adopt this standpoint, then we are forced to assume a standpoint in the dispute between reconstructivism and descriptivism, i.e. in the dispute concerning the possibility to apply logical tools to analyze and solve philosophical problems.10

I have contended that I have included into my deliberations the philosophy of language underlying Ajdukiewicz’s meta-epistemological proposal. I do hope ←11 | 12→however that the findings made in the course of these deliberations will attain a supra-individual value and will apply to the foundations of every philosophical declaration stating that speaking of concepts and propositions is replaceable with speaking of expressions of language, as – and this is important – first such replace ability makes it possible to apply logical measures in analyses of knowledge. Therefore, if Suszko’s and Woleński’s formal theory of knowledge grew out of Ajdukiewicz’s semantic theory of knowledge, then I have decided to include into my deliberations the possible philosophy of language underlying their manner of pursuing the theory of knowledge; I say “possible”, since similarly to Ajdukiewicz these philosophers do not explicitly refer to this issue.11 And I believe that a relevant philosophy of language in this respect is the philosophy of Edmund Husserl presented in the second volume of his Logical Investigations,12 which – similarly to the solutions contained in both volumes of Logical Investigations – influenced the philosophical convictions of many representatives of the Lvov-Warsaw School, if we were to limit ourselves only to those representatives of the Polish 21st century philosophy, and not to mention influence of this philosophy on the most distinguished student of Husserl, i.e. Roman Ingarden. This is the philosophy that makes it possible to reasonably speak of a close relation between thinking and language, which I wish to particularly emphasize, and provides thereby a philosophical-linguistic legitimization for the meta-epistemological proposal. And by legitimizing this proposal, it makes it possible to demonstrate the relation between the traditional theory of knowledge and the semantic theory of knowledge; both proposed by Ajdukiewicz, as well as the formal semantic theory of knowledge of Suszko and Woleński.


ISBN (Hardcover)
Publication date
2020 (January)
Logic of language Cognition Epistemology Metaphysics Explication Intentionality
Berlin, Bern, Bruxelles, New York, Oxford, Warszawa, Wien, 2020. 296 pp.

Biographical notes

Adam Olech (Author)

Adam Olech graduated at the Philosophical-Historical Faculty of the Jagiellonian University in Cracow, from which he next obtained his PhD. He holds a habilitation from the Nicolaus Copernicus University in Toruń and is now a professor at the Jan Długosz University in Częstochowa. His academic interests are analytical philosophy and logical semiotics.


Title: The Semantic Theory of Knowledge