Measuring the fricative differences within and between languages

We do not know the relevant parameters for comparing fricatives in different languages. In an attempt to determine how to characterize fricatives, five speakers each of Amharic, Egyptian Arabic, English, Hebrew, Hopi, Japanese, Korean, Navaho, Papago, Pima, Polish, Yoruba, and Zuni produced fricatives in a sentence frame in the environments i-i, a-a, u-u. A 25-ms section of each fricative was subjected to a critical-band analysis (within the frequencies 0–10 kHz). The resulting auditory spectra were then compared for fricatives within and between the languages, by means of PARAFAC and other statistical techniques. These techniques enabled us to determine the extent to which fricatives which have been described as the same in different languages are, in fact, the same. [This work was supported in part by NSF, and by the Institute of American Cultures in collaboration with the American Indian Studies Center at UCLA.]