Investigating the Effects of Valence, Arousal, Concreteness, and Humor on Words Unique to Singapore English

TL;DR

This paper studies how emotional and semantic features affect processing of unique words in Singapore English.

Contribution

It introduces new norms for valence, arousal, concreteness, and humor for Singapore Colloquial English words.

Findings

Valence, arousal, and concreteness improved lexical decision performance beyond word frequency and orthographic similarity.

Positive, arousing, and concrete Singapore English words were recognized faster and more accurately.

Human and model ratings aligned closely for valence, arousal, and concreteness but diverged for humor norms.

Abstract

Singapore English is a dialect of English spoken by individuals living in Singapore, whose colloquial form (i.e., Singapore Colloquial English) contains unique lexical items not found in dominant dialects of English. The absence of these items from the lexicon of dominant English dialects indicates that lexical-semantic and affective norms central to psycholinguistic research do not exist for these Singapore English concepts, and it is unclear what is the specific influence of these effects when processing Singapore Colloquial English words. The present paper describes the development of valence, arousal, concreteness, and humor norms for a core vocabulary list of approximately 300 words and concepts, via human ratings and probing a Large Language Model, and evaluates the contribution of these norms to account for lexical processing performance in a visual lexical decision task. Results indicated that valence, arousal, and concreteness explained additional variance over and above orthographic similarity and word frequency in the visual lexical decision task. Specifically, Singapore English words that were more positively valenced, highly arousing, and more concrete, were responded to more rapidly and accurately. In addition, although there was generally a high convergence of valence, arousal, and concreteness ratings across human raters and the Large Language Model, humor norms were much less closely aligned. Overall, this paper provides a case study of how psycholinguistic research can be extended to diverse, understudied dialects of English, and showcases how doing so offers an opportunity for psycholinguistics to examine the importance of various lexical-semantic and affective measures to quantify lexical information in colloquial, informal language.

Figures