Change of state: from the BECOME operator to the mereotopology of events

Louis McNally

Universitat Pompeu Fabra

Lattimore 513

How can and should we model the semantics of change of state verbs? When I first read Dowty (1979), I thought nothing could be more straightforward, more precise, and therefore more appropriate than his truth-conditionally defined BECOME operator. Over the years, however, I have seen that the semantics of change of state verbs is more complex than I thought. In particular, I have come to appreciate the subtle but important difference between analyses whose central focus is to capture truth conditions associated with verbs vs. those whose goal is to capture the mereotopological properties of events described by verbs (see, e.g., Casati and Varzi 1999 on mereotopology applied to language in general; see Piñón 1997 for one of the few explicit applications of mereotopology in verb semantics).

In this talk, I explain what I have come to understand as the main differences between the simply truth-conditional and the mereotopological ways of thinking about change of state verbs. In the latter category I would put, for example, the work of Pustejovsky (1991) and Williams (2015), although they do not themselves use the term “mereotopology.” The fact that the two approaches are not incompatible in principle, and that mereotopology has a less established tradition in semantic theory, has obscured these differences. I will discuss how adopting a mereotopological perspective has helped me think in new ways about verb semantics, illustrating with examples from Marín and McNally (2011) and McNally and Spalek (to appear). Finally, I will make some provocative suggestions concerning the relation between work in verb syntax and these two perspectives on verb meaning.