Iconicity Ratings Across the Japanese Lexicon: A Comparative Study with English

Abstract

Iconicity is a perceived resemblance between form and meaning. For example, “X” can represent “a cross” because its form contains an intersection of lines. In spoken language, iconicity is understood as a word “sounding like what it means.” Recent studies on English and Spanish use rating tasks to see which words native speakers consider to be iconic. Perry et al. (2015) show that, in English, onomatopoeia are rated highest, followed by adjectives/verbs > nouns > function words. The current study replicates the same rating task for Japanese but, owing to additional variables tested, yields more nuanced findings. To be consistent with Perry et al. (2015), first we tested for word class only and found the same hierarchical pattern for Japanese that Perry et al. found for English. Given the stratified nature of the Japanese lexicon, next we tested for interactional effects of word class plus lexical stratum: ideophonic (mimetic), foreign-loan, Sino-Japanese, and Yamato (native). Our results show that Japanese speakers are more likely to rate a word as iconic if it is an ideophone or if it falls into the Yamato stratum. Finally, the Yamato stratum contains words of historically iconic etymology but are synchronically prosaic and no longer morphologically or phonologically resemble members of the ideophonic stratum. We conducted a separate test that showed Yamato words of historically iconic etymology are not rated significantly higher. For comparative purposes, we recoded and reanalyzed the English ratings from Perry et al. (2015), and found that neither lexical strata (Germanic, Latinate, French) nor historically iconic etymology (Germanic words only) had a significant effect. Taken together, our results show that, word class aside, Japanese speakers are much more likely to rate a word as being iconic if it is (1) an outright ideophone > (2) a Yamato prosaic word > or (3) a non-Yamato prosaic word. Japanese speakers are thus sensitive to three factors when assigning ratings: (a) the theoretical definition of iconic status, i.e., Is it an ideophone or not?, (b) nativeness, i.e., Is it borrowed from another language or not?, and (c) ability for word class to convey sensory information, i.e., Is it a sensory word or not? Given our reanalysis of Perry et al. (2015), English speakers are sensitive to (a) and (c). With all these factors in mind, it seems unlikely that such ratings are any indication of iconicity as pervasive throughout an entire lexicon. More realistically, we propose the following explanation: Speakers are aware of theoretical iconicity and its special relationship to sensory information. When pressed, speakers extrapolate this awareness to other members of the lexicon and instead rate according to what they know about a word’s relationship to sensory information. At the same time, language- and culture-specific factors, like (meta-linguistic awareness of) lexical stratification, mitigate these ratings.