A survey of research in the history of the "h-aspire" in French. It is suggested that a synchronic rule of h-deletion never existed and a synchronic rule of insertion existed only in Old French. The evolution of liaison is compared to that of "h-aspire." Questions are raised for further study. (AMH)

Analysis of how word-final "m" in Spanish is pronounced in various distributional situations. The major tendency is for it to become an "n" except before "p" and "b" when it is pronounced as bilabial sonorant "m." (Text is in Spanish.) (TL)

Stop consonants after initial /s/ are standardly spelled as the unvoiced stops /p/, /t/, and /k/. Phonetically, however, they are similar to the voiced stops /b/, /d/, and /g/. Research suggests that many young children make consistent, reasonable, but unconventional, judgments about sounds and English spelling. This paper considers the case of…

Three models were constructed to predict pronunciation responses: final consonant, type-token, and invariant principal response. Six synthetic words were constructed according to contextual and word-position constraints. The final consonant model was superior. (Author/CP)

This paper examines the distinction between primary and secondary articulations of consonants. It shows that the description and classification of speech sounds should not be based on physical parameters alone. Some essential distinctions reveal themselves only inferentially through the relation of a sound to others in the language system.…

In polysyllabic segments, the composition of medial clusters is a reflection of morpheme structure constraints as they apply to initial and final groups of syllables. Also, medial clusters ideally overlap, i.e., have simultaneous membership in both the preceding and following syllable segments. (DB)

Sentences were read by six informants to determine the presence or absence of /n/ in /nth/ sequences. The sentences contained seven different levels of juncture with /nth/ occurring in word final position, intervocalically, and across word boundaries, among other places. Dental coarticulation was not hindered by most junctures. (SC)

Reasons are given for the idea that Hueyapan (Morelos) Nahuatl has an underlying velar nasal with lip rounding phoneme which never has the surface reflex of a rounded velar nasal allophone, but occurs phonetically as a velar nasal allophone or a labial dental voiced allophone or disappears. (SCC)

This article discusses the problem of identical sounds which represent different phonemes. A distinction between basic distinctive features is rejected in favor of a theory of phonological fluctuation. (Text is in French.) (AM)

This article describes an experiment in which the following hypothesis was tested and supported: that tension serves to reinforce the phoneme's articulatory characteristic, whether it be closure or openness; and that audibility interferes with perception of tension. (Text is in French.) (AM)