DIFFERENTIAL CONTRIBUTIONS OF THE SPECTRO-TEMPORAL AND VOCAL CHARACTERISTICS OF AUDITORY PSEUDOWORDS TO MULTIPLE SOUND-SYMBOLIC MAPPINGS

ABSTRACT In spoken language, iconic (sound-symbolic) words are those whose sounds convey their meaning. Iconicity is widespread in natural languages, whether signed or spoken, but its instantiation across different domains of meaning has not been systematically studied. Here, participants rated a set of 537 auditory pseudowords on opposing dimensions of eight different sound-symbolic domains: shape (rounded-pointed), roughness (smooth-rough), hardness (hard-soft), weight (light-heavy), size (small-big), brightness (bright-dark), arousal (calming-exciting), and valence (good-bad). Ratings showed cross-domain relationships, some mirroring those between corresponding physical domains, e.g. size and weight ratings were associated, reflecting a physical size-weight relationship, while others involved figurative relationships, e.g. bright/dark mapped onto good/bad, respectively. Using four separate multiple regression analyses, we found that the phonetic categories, phonemic associations, and acoustic associations at both the whole-item and segmental levels, formed unique sets with characteristic feature weightings for each meaning domain studied. For the majority (12 of 16 dimensions), the phonemic regression accounted for the most variance in the ratings, followed by the phonetic category, segmental acoustic and whole-item acoustic regressions. We conclude that pseudoword iconicity is present across a range of meaning domains, with domain-specific patterns of linguistic and acoustic properties. Linguistic characterizations best captured judgments of iconicity across domains, suggesting that these properties represent the bundles of acoustic, articulatory, and abstract linguistic factors that may underlie iconicity in spoken language.