This article examines the traditional classification of words into parts of speech, highlighting the limitations of approaches based solely on morphosyntactic criteria. In this context, the necessity of a functional-semantic perspective is argued, one grounded in the principles of functional grammar, where meaning and function become essential criteria for the lexico-grammatical categorization of words. The four major notional parts of speech - noun, verb, adjective, and adverb - are analysed as domains structured around the core-periphery dichotomy: the core includes subclasses of words with stable grammatical functions and semantically motivated syntactic roles, while the periphery comprises units with ambiguous status or multiple functions, thus emphasizing the fluidity of these categories. This approach highlights the dynamic and context-dependent nature of parts of speech, whose classification may vary according to the specific conditions of communication.
Ion Bărbuţă (Fri,) studied this question.