Currently Reading "Meanings as Species" by Mark Richard, some provoking bits:
"...Of course languages, lexicons, and individual words—like species—evolve. As one population ‘reproduces’ its language in the next, new words arise, old words have their meanings changed, phonology shifts, grammatical rules may be modified, and so on. As is the case with species, some of this evolution will be quite gradual: successive generations are able to fluently communicate with one another, just as successive generations in a population lineage are (counterfactually and in principle) able to interbreed, have fertile progeny, share a system for recognizing mates, etc. As is the case with species, even when abutting generations enjoy the sort of cohesion that would drive the observer to classify them as speakers of a single language, the soritical ways of linguistic change, given enough time, will lead to a diachronic lack of cohesion so great that no one will say that ancestor and descendent populations speak the same language, even though the languages they use are related by descent."
"...Whether we stick with this sort of terminology or not, we should agree that Quine is correct to think that the notion of analyticity is of little to no use in the study of language. But it certainly doesn’t follow from this that the notion of meaning can’t bear any explanatory load, any more than it follows from the fact—biological species do not have essences; none of the small changes that might lead, when summed over time, to speciation are themselves intrinsically changes that separate one species from another— that the notion of species cannot bear any explanatory weight. The notions of word meaning, concept, and (public) language are no worse off because of Quine’s and allied arguments and observations than is the notion of species because of Darwinian arguments and observations that speciation is a historical process and that it is folly to think that species have anything like essences."
"...Our talk about meaning, like our talk about species, tracks something that is event- like, more process than product. Our talk about meaning, like our talk about species, tends to be cast in terms that are more appropriate to something that is not event-like: thus the attraction of the views that species have some sort of essence, and that a word’s meaning can be identified once and for all with a definition or a Fregean sense or something of the sort. Because of the apparent lack of fit between what our talk about meanings and species tracks and the conceptual box that talk creates, we might at the end of the day decide that rather radical conceptual engineering is called for: we might even recommend dropping talk about species or meanings in favor of talk about populations related by descent or lineages of lexicons linked by various relations of communication. To do so in the biological case is not to suggest that species talk does not track a real phenomenon, or that the claims and generalizations biologists make in speaking about species are empty or unverifiable or false. Ditto, for the linguistic case."