To investigate relations, if any, between phonetic symbolism and learning, a paired-associate learning experiment was conducted with two stimulus variations (complete vs. incomplete figures) and three response variations [repetitive vs. nonrepetitive same (8) vs. nonrepetitive different (D) disyllables, where 8 and D were defined in terms of the first vowel of the repetitive group). Disyllables produced significant effects in all aspects of data. Effects of figures were complex: First-trial data (approximating word matching) indicated suggestive differences, with significant Figure by Disyllable interactions; but overall performance produced insignificant differences between figures. Both figures and disyllables significantly interacted with learning trials. The advantage of the complete-repetitive combination was observed only early in learning. and that of the incomplete-nonrepetitive combination only late in learning. The second sound of disyllables produced significant effects on learning. Phonetic symbolism research may be facilitated by taking learning processes into account.Ever since Plato, phonetic symbolism, i.e., an intrinsic correspondence between sounds (physical properties) and meanings of words, has been of considerable interest (for reviews see Taylor, 1963, e.g.). Yet, evidence beyond anecdotal or cursory observations in this area has been largely limited to studying individual words: matching words against words or nonsense syllables against nonsense syllables (subjective phonetic symbolism) and examining the lexicon of languages in terms of denotata (objective phonetic symgolism). Recently, however, De Vito and Civikly (1972) approached the issue differently and found repetition of sound within disyllables (e.g., LOLO) is reliably associated wth visual representations of homogeneity (vis-a-vis heterogeneity) and completeness (vis-a-vis incompleteness).No matter whether phonetic symbolism may be assumed to be universal (e .g., Miron, 1961) or language specific (e .g., Taylor & Taylor, 1962, 1965, provided that phonetic symbolism is based on languag{~, such symbolism is likely to be learned, along with the language itself. This is particularly true when phonetic symbolism is assumed to be language specific. Taylor and Taylor (i 965), e.g., assumed the identity of subjective and objective phonetic symbolism; the latter, no doubt, involved some sort of learning. On the other hand, phonetic symbolism might be demonstrable with nonsense syllables or foreign words (e.g., Newman, 1933;Sapir, 1929). If so, the learning of phonetic symbolism is minimaL The latter evidence in a sense, however, may lead to an ultimate rejection of the existence The present study is supported, in part, by a Biomedical Research Grant (69175) from NIH via Tulane University to C. Izawa. The authors are grateful to Janice Rauch for her assistance with data reduction and preparing figures. Requests for reprints should be sent to Chizuko Izawa, Department of Psychology, Tulane University, New Orleans, Louisiana 70118.
475of ph...