demisyllabic (adj.)

A term used in ACOUSTIC PHONETICS for an approach to the analysis of SYLLABLES which recognizes an acoustically identifiable unit between the syllable and the PHONE. A syllable is decomposed into two elements: a syllabic core and an optional affix (or affixes). The core is then analyzed, on the basis of a SPECTROGRAPHIC or waveform display, into an initial and final demisyllable, corresponding respectively to the initial part of the VOWEL and the mid-to-final part of the vowel. Because there are fewer demisyllables than syllables (in English, for example, the demisyllabic inventory is about a fifth of the syllabic), and because they display a degree of independence and acoustic stability which facilitates CONCATENATION, this kind of analysis has had applications in SPEECH SYNTHESIS and automatic SPEECH RECOGNITION