The state/event distinction in language and beyond

According to long-standing linguistic observations, languages of the world seem to systematically distinguish between states and events. Differences can be observed at the lexical level, between predicates of the type be tall and predicates such as walk (Vendler 1957, Dowty 1979); also at the grammatical level, where temporal and aspectual morphology seem to be sensitive to the state/event division too, manifested in the general unwillingness of states to take progressive aspect, or the remarkable parallelisms in the behaviour of (lexical) stative predicates and the Perfect (Katz 2003). At the level of discourse, events and states contrast in their ability to contribute to the narrative progression, and when it comes to language processing tasks, there is evidence that states are easier to process than events (Brennan and Pylkännen 2010). In this context, my talk has two aims: one is to evaluate the semantic status of the Perfect as a derived state (after Parsons 1990) and its contribution to discourse representation, in contrast with underived states. The empirical weight of the proposal comes from my own fieldwork in areas of contact between Spanish and Galician. Once we have examined how these distinctions work within the language system, a second goal is to place the discussion at the interface between language and cognition and to ask a more general and difficult question, which is: is there a distinction between events and states in our mental architecture, independent of language? In other words, are states and events cognitively real? Can a state, as linguists understand it, be a computational primitive?