This paper explores whether Pre-trained Large Language Models (PLLMs) like GPT-4 can grasp profound linguistic insights into language phenomena such as Aspectual Coercion through interaction with Microsoft’s Copilot, which integrates GPT-4. Firstly, we examined Copilot’s understanding of the co-occurrence constraints of the aspectual verb “begin” and the complex-type noun “book” using the classic illustration of Aspectual Coercion, “begin the book.” Secondly, we verified Copilot’s awareness of both the default interpretation of “begin the book” with no specific context and the contextually preferred interpretation. Ultimately, Copilot provided appropriate responses regarding potential interpretations of “begin the book” based on its distributional properties and context-dependent preferred interpretations. However, it did not furnish sophisticated explanations concerning these interpretations from a linguistic theoretical perspective. On the other hand, by offering diverse interpretations grounded in distributional properties, language models like GPT-4 demonstrated their potential contribution to the refinement of linguistic theories. Furthermore, we suggested the feasibility of employing Language Models to construct language resources associated with language phenomena including Aspectual Coercion.

In this paper, we present semi-automatic annotation of the Event Structure Frames to synsets of English verbs in WordNet. The Event Structure Frame is a sub-eventual structure frame which combines event structure (lexical aspect) with argument structure represented by semantic roles and opposition structure which represents the presupposed and entailed sub-events of a matrix event. Our annotation work is done semi-automatically by GESL-based automatic annotation and manual error-correction. GESL is an automatic annotation tool of the Event Structure Frame to verbs in a sentence. We apply GESL to the example sentence given for each synset of a verb in WordNet. We expect that our work will make WordNet much more useful for any NLP and its applications which require lexical semantic information of English verbs.