Skip to main content
Abstract

Aspectual classes as lexically-conditioned predictors of aspectual choice

Author
  • Laurestine Bradford

Abstract

I investigate a distribution-based characterization of lexical aspectual classes.

The grammatical aspect of a verb is morphology which reflects either an internal perspective or an external perspective on the time course of an event. Progressive takes an internal perspective, while perfective takes an external one. However, not all verbs are felicitous in all aspects. For example, verbs denoting static situations usually sound worse in the progressive. It has long been theorized that each verb in a language has an aspectual class which captures something about the temporal shape of the described event and thereby explains its compatibilities with different grammatical aspects.

I investigate the idea that aspectual class is precisely the lexical information which contributes to grammatical aspect choice. Therefore, it should be detectable by statistically computing each lexical item's contribution to aspectual choice. I fit a Bayesian mixed-effects logistic regression predicting aspectual choice using lexical item as a predictor. I then used the fit weights for each lexical item to characterize its lexical aspectual class. I discuss the relationship between these patterns and the theory of aspect. In particular, I found that stative verbs and a subclass of incremental theme verbs both show distinctive aspectual behaviour.

Keywords: aspect, semantics, verb, aktionsart, aspectual class, lexical aspect, bayesian, logistic model

How to Cite:

Bradford, L., (2025) “Aspectual classes as lexically-conditioned predictors of aspectual choice”, Society for Computation in Linguistics 8(1): 37. doi: https://doi.org/10.7275/scil.3168

17 Views

7 Downloads

Published on
2025-06-14

Peer Reviewed