A Scalable Tool for Measuring Manner and Result Verbs in Developmental Language Research

2026-05-15

AI Summary

This study addresses the scarcity of large-scale annotated resources in developmental linguistics, which hinders efficient measurement of verb semantic classes—specifically, manner and result. Integrating linguistic theory with large language models, the authors employ prompt engineering to automatically generate sentence-level annotations on the MASC and InterCorp corpora, extending VerbNet coverage to 436 verb classes. They further train a context-aware classifier based on RoBERTa to distinguish between manner and result verbs. This approach achieves, for the first time, large-scale automatic annotation and identification of these semantic categories, attaining an average accuracy of 89.6% across three expert-annotated test sets. The resulting framework offers a scalable and reusable tool for verb semantics research.

Abstract

Manner and result verbs encode different aspects of event structure and have been discussed in developmental work as a potentially informative distinction for studying early verb learning. However, this distinction remains difficult to measure at scale because large annotated resources for manner and result classification are not currently available. We present a computational approach for identifying manner and result verbs in sentence context. Using linguistically informed prompts, we generate sentence-level annotations with large language models over data drawn from MASC and InterCorp, extending coverage from previously annotated portions of VerbNet to 436 classes. We then train a RoBERTa-based classifier on these annotations and evaluate it on three held-out gold-standard datasets, including previously annotated items and a new expert-annotated set. Across these evaluations, the model shows promising performance, with average accuracy up to 89.6%. We present this work as a scalable measurement tool that can support future research on verb semantics in developmental and other language datasets, while noting that further validation is needed for borderline cases, mixed manner/result verbs, and downstream developmental applications.

Problem

• manner verbs
• result verbs
• developmental language research
• verb semantics
• scalable measurement

Innovation

• manner verbs
• result verbs
• large language models
• RoBERTa classifier
• scalable annotation