True problems will never disappear-science loves them too much. Romance Linguists, for instance, have been fascinated by the parallelism in the evolution of different Romance languages ever since the birth of their science; this parallelism has so much more appeal, because it is essentially of a typological or structural order, without always implicating the etymological identity of the linguistic material. Whereas diachronic morphosyntax has accepted this descriptive (and explanatory) challenge, semantics, specially lexical semantics, have for the most part remained reluctant. The cognitive orientation could though stir up this state of things.