The originality of Ajdukiewicz’s approach is to have put forward that the syntactical calculus should be carried out in taking into account the semantical well-formedness of the sentence. Ajdukiewicz writes in his 1935 paper which lays the foundations for his categorial syntax: “It [the problem of syntactic connexion] is concerned with the specification of the conditions under which a word pattern, constituted of meaningful words, forms an expression which itself has a unified meaning (constituted to be sure, by the meaning of the single words belonging to it). A word pattern of this kind is syntactically connected” (1935/1978, p. 18). This choice of a joint calculus of grammaticality and semantical well-formedness was hailed as particularly innovative (Bach 1988, p. 31-32), since for the first time it did justice to the fact that any speakers assesses simultaneously the grammaticality and the meaning of the statements read or heard. To carry out this syntactico-semantical approach, Ajdukiewicz had two successive sources of inspiration: Husserl’s Fourth Logical Research in the thirties and Frege’s theory of meaning in the sixties. The aim of this chapter is to retrace Ajdukiewicz’s semantical path by highlighting why these two sources of philosophical inspiration did not allow him to achieve his objective. To show this, we will start by comparing Ajdukiewicz’s syntactical and semantical approach with that of his contemporary Rudolf Carnap1 who, in the wake of Bertrand Russell’s theory of types, had an opposite theoretical position by considering that syntax has to operate on meaningless words. This study, which will show that Ajdukiewicz and Carnap reverse their semantical points of view when they apply their syntax to natural languages, will allow us to understand why Ajdukiewicz looked for another solution by drawing inspiration from Frege’s “Sinn und Bedeutung”. The analysis of the new conception of meaning propounded by Ajdukiewicz in 1960 will, however, show that it results from a misuse of the Fregean theory. If using philosophical theories to develop a semantics turns out to be a difficult task, Ajdukiewicz’s objective would be nevertheless rediscovered and successfully developed in Richard Montague’s categorial grammar (1970).