In Verbs: Aspect and Causal Structure (2012), I presented a two-dimensional analysis of aspectual semantics. In that book, I argue for a two-dimensional model of aspectual meaning, in which aspect is defined as how events unfold over time. The unfolding is modeled by a temporal dimension and a qualitative dimension which represents the changes (if any) that occur over the course of the event.

A second, equally important characteristic of this analysis of aspect is that a verb may be construed as having different aspectual types depending on what tense-aspect construction it occurs in. This process results from an interaction between constructional meaning and lexical meaning. Thus, verbs do not belong to a single lexical type. Instead, they have an aspectual potential to be construed in a certain range of aspectual types.

There is a great variety in the aspectual potential (range of aspectual construals) of different predicates. There are nevertheless common patterns. In Verbs, I used a multidimensional scaling analysis of the aspectual potential of 44 English predicates and their Japanese translation equivalents to construct a model of pairs of related aspectual construals.

One can go beyond this analysis and look at verb classes with more than two potential aspectual construals. In this talk, I argue that by doing so, we can characterize the manner vs. result verb distinction of Levin and Rappaport Hovav, the verb-framing vs. satellite-framing distinction of Talmy, and other typological differences in the verbal lexicon, in terms of different aspectual potentials of verbs and verb classes across languages.