| The dissertation explores the question of segmental phonetic symbolism—the attribution of qualities such as brightness and largeness to speech sounds. An experiment investigates attributions for the point vowels [i], [a] and [u], and the mid central vowel [v], paired stimuli being specified in [hV1, hV2] form. A questionnaire asking Which of the two stimuli is more X? for adjectives bright and dark, active and passive, large and small (representing brilliance, energy, and size dimensions, respectively) elicited judgments from speakers of American English, Korean, Japanese and Spanish.; These judgment data are analyzed in terms of three basic questions. First is whether evidence exists of a consistent phonosymbolic response, the claim here turning on the extent of statistically significant concurrence about the stimulus preferred in any pair-comparison, within language groups. This is answered with a strong affirmative. Second is whether phonosymbolism should be ascribed a universal or language-specific basis, as determined by agreements/disagreements across language groups about particular stimulus rankings (e.g., brightest to least bright). Findings suggest a predominant universal basis, tuned in special circumstances by a language- or culture-specific influence. Third is which vowel-features (phonetic or phonemic) are likely to have underpinned the stimulus rankings. The dissertation proposes several features plausibly triggering phonosymbolism, and suggests mechanisms for their joint application (feature harmony, equivalence, and conflict). |
