Odwoływalność (cancellability) jest powszechnie uznawana za swoistą własność implikatur konwersacyjnych, czasami bywa jednak rozciągana także na inne implikacje pragmatyczne. W pracy zostają postawione i rozważone dwa następujące pytania: 1) czy wszystkie implikatury konwersacyjne są odwoływalne? 2) czy odwoływalne są też inne implikacje pragmatyczne, w szczególności presupozycja i eksplikatura? Przedstawiam argumenty prowadzące do twierdzącej odpowiedzi na pierwsze pytanie oraz przeczącej odpowiedzi (odwoływalność eksplicytna) na drugie.

In the paper I present the famous argument between Peter F. Strawson and Bertrand Russell on definite descriptions. I do not go into details of the two rival solutions to the problem of definite descriptions. Instead I present the controversy against the background of two traditions within analytic philosophy, i.e. the philosophy of natural language (Strawson) and the philosophy of ideal language (Russell). In consequence, the aim of this paper is to sketch the principal features of the two traditions and to indicate their influence on the argument. In the first paragraph I discuss Russell’s theory of descriptions and present it as a result of dramatic changes that he had made in his philosophy before he finally presented them in On Denoting in 1905. The second paragraph deals with the two traditions within analytic philosophy after the linguistic turn and underlines the role of Strawson in the philosophy of natural language. In the third paragraph I analyze in detail Strawson’s arguments against the theory of descriptions and I focus on some details that are usually omitted in standard presentations. The fourth paragraph discusses Russell’s response to Strawson’s objections, i.e. the counter-arguments formulated from the standpoint of philosophy of ideal language. I end with some suggestions about how to reconcile both approaches.

The article consists of two parts. In the first one (introductory) I recall—following Edmund Husserl, Stanisław Ossowski and Adam Schaff—the main formulations of the “principle of transparency of the sign.” In these formulations it is usually said about (1) the transparency of the sign regarding objects denoted by the sign (denoted, designated and/or named), or (2) the transparency of the sign regarding its meaning (respectively, events, states of affairs and facts designated by the sign). However, as Husserl pointed out, one can also speak about (3) the transparency of the sign in relations to the activities and mental states of the sign’s users (senders and recipients). After all, only due to the transparency of the sign understood in this way, it is possible for people to communicate with each other, thus the sign can also has an expressive and communicative function. In turn, the second part of the article (essential) contains a reconstruction of the Leon Koj’s approach; Koj gave a consistently formalized form to the theory of sign based on the principle of transparency— the form of an axiomatized logical system (using Quine's formalism from his Mathematical Logic). One of Koj's main goals was also to indicate the close relationship between semantics and pragmatics, and even the primacy of pragmatics over semantics. Formal-logical tools have also shown that the theory of sign based on the principle of transparency neither contravene The Law of Non-Contradiction (at least in its psychological formulation), nor contain or imply semantic antinomies such us antinomy of the liar. Because it is a theory easily negotiable with Alfred Tarski’s theory of language levels.