In a 1997 paper, Violi wondered whether semantics could be non–cogni-tive. This question is still significant since, in the last decades, cognitive semantics has been thought of as the more suitable paradigm to investi-gate the profound “mystery of meaning” precisely because of its criticism of a presumed “non–cognitive” (i.e. “autonomist”) linguistic semantics. However, within the cognitive framework, the disapproval of linguistic semantics often appears unfocused, being addressed to a variety of theo-ries and authors gathered under unspecified labels. Bringing into play the quandary of the “autonomy”, Violi’s question should be recalled to dis-cuss the meaning of “cognitive” when referred to semantics. Such a re-flection seems, in fact, essential as to elucidate why some theorists believe that cognitive semantics is not so “cognitive” as it claims, and why one may find even “more cognitive” those approaches which lay on a philo-sophical question about language, such as the one outlined by Eugenio Coseriu. With this aim, in the first section, comparing the Chomskyan paradigm with the cognitive theories of the second generation, I will briefly discuss what is meant by “cognitive”. In the second section, I will instead investigate Eugenio Coseriu’s monist conception of language and the ruthless criticism he moved to cognitive semantics, enlightening some of the theoretical standstills of a “non–autonomist” semantic approach.