Semantic judgments of compressed monosyllables: evidence for phonetic symbolism

Both frequency and speed of utterance have been implicated in studies of phonetic symbolism. Therefore, these two variables were manipulated independently. Three monosyllables were recorded and distorted by either increasing or decreasing frequency, holding speed constant, and by increasing or decreasing speed, holding frequency constant. The subjects were 15 college students who rated the resulting 15 stimuli (3 monosyllables X fast speed-normal frequency, slow speed-normal frequency, normal speed-high frequency, normal speed-low frequency, and normal speed-normal frequency) using 15 bipolar adjective scales, chosen on the basis of previous semantic differential and phonetic symbolism research. Five separate factor analyses were applied to the data. It was found that the normal speed-low frequency, slow speed-normal frequency, and normal speed-normal frequency stimuli generated approximately the same factors, while the fast speed-normal frequency and normal speed-high frequency stimuli generated factors similar to each other but with some notable differences. Separate analyses of variance were applied to the bipolar adjectives using speed, frequency, and vowels as main effects. The 15 analyses showed that subjects did judge the monosyllables as different on the basis of both speed and frequency. The implications of this study for phonetic symbolism research are discussed.