抄録
The present study is concerned with the proper representation of events, and argues for a new model of event representation, which is referred to as the “parallel event structure (PES).” The PES plots time on the horizontal dimension, and has two or more tiers aligned on the vertical dimension, each of which encodes a certain type of event (e.g. action/change). This multi-tiered structure represents three types of information on the two dimensions: it mainly represents (i) force-dynamic information (the exertion of energy and its effect) vertically, (ii) simple aspectual information (the temporal evolution of each subevent) horizontally, and (iii) complex aspectual information (two or more subevents’ parallel unfolding over time) two-dimensionally. It also represents the flow of time and the unfolding of (the subevents of) an event in terms of scale, thereby capturing their homomorphism as well as parallelism. This article argues that the PES is preferable to alternative models of event representation in terms of descriptive power and uniformity/coherence, showing that it clearly delineates various types of events from one another while capturing their commonalities in a uniform and coherent way. Empirical evidence is adduced primarily from verbal and adverbial expressions in English. Theoretical issues revolving around event structure are also discussed, which include force-dynamics, aspect, telicity, and scale structure.
