Sweetness or Mouthfeel: A corpus-based study of the conceptualization of taste

Yin Zhong, Chu-ren Huang (Corresponding Author)

Research output: Journal article publicationJournal articleAcademic researchpeer-review


The sensory lexicon plays a pivotal role in bridging our cognitive system to the physical world. In this role, it has been the focus of recent interdisciplinary investigations on cognition, language, culture, and their interactions. Recent studies on linguistic synesthesia and sensory modality exclusivity showed unequivocally that cross-modality usages of sensory words are the norm rather than the exception. Given the dominance of cross-modality uses, the null hypothesis that the five senses are separate but equal modules merits a closer examination. In this paper, we focus on the gustatory quality of sweetness because of its universal appeal as well as the well-attested cultural influence on the gustatory lexicon. Based on an analysis of online food reviews containing descriptions of desserts, we show that mouthfeel, a multisensory concept, is strongly preferred over sweetness. Mouthfeel is associated with words from all the sensory domains, including both sensory and abstract (e.g., mental state) concepts. The highly non-exclusive characteristic of gustatory sensation suggests that it might be the most connected sensory modality, and the cross-modality expressions indicating personal preferences further imply the subjective propensity of the gustatory sense. Our study adds to the existing literature the interrelationship among sensory modalities through language use, and further sheds light on the interactions between language, cognition, and culture.
Original languageEnglish
Pages (from-to)359-387
JournalLinguistic Research
Issue number3
Publication statusPublished - Dec 2020


  • sensory lexicon
  • taste
  • mouthfeel
  • linguistic synesthesia
  • online food review


Dive into the research topics of 'Sweetness or Mouthfeel: A corpus-based study of the conceptualization of taste'. Together they form a unique fingerprint.

Cite this