Vowel-Vowel Production: The Distinctive Region Model (DRM) and Vowel Harmony

Abstract

According to the distinctive region model (DRM), sequencing one vowel to another involves one, two, or three commands or phonetic movements corresponding to specific deformations of the vocal tract. Criteria pertaining to communication theory require that these movements, or gestures, be of a small number, efficient (small deformations imply large formant variations), and simple (e.g. rectilinear movements). For a given language, the vowel-vowel transitions involving one movement should then be statistically dominant. This hypothesis is verified using a large corpus of French.

Harmony versus the OCP: Vowel and Consonant Cooccurrence in the Lexicon

“…Limiting to these three allows diphone classification into four categories, according to whether the change from the first vowel to the second involves a change of none of the variables, one, two, or three. Carré et al (1995) find that vowel-vowel transitions involving change of none or one variable are overrepresented (72% versus the expected percentage of 59%), while those involving change to two or three are underrepresented. The effect heightens for more frequent vowels.…”

“…French. A more quantitatively rigorous study is that of Carré et al (1995), which draws on the Sankoff-Cedergren corpus of phonemically-transcribed s poken Montreal French. The corpus contains data from 105 speakers, yielding approximately 250,000 vowel-vowel diphones.…”

“…Thus the data generated reflects transition probabilities in the speech stream as a whole, rather than within individual lexical items. Unlike Tamaoka et al (2004), Carré et al (1995) take into account the relative frequencies of the individual vowel tokens by decomposing the table of cooccurrences into adjusted residuals, in which the sign of the resulting metric indicates either over-( positive) or under-(negative) representation of a given diphone.…”

“…Carré et al (1995) attribute this tendency to cognitive factors, hypothesizing that making fewer changes in articulatory parameters puts less of a strain on the memory. However, their fi ndings are still not strictly comparable to those of the consonant studies.…”

Prediction of vowel systems using a deductive approach

“…Indeed, the vowel systems can be divided into two main classes: one with the central vowel / /, and the other without, and they are obtained by the successive addition of a new vowel [1]; the vowels can be classified into: labials or non-labials, nasals or non-nasals, ATR, retroflexes, diphthongs, etc. ; the vowels can be classified according to their complexity [3], [4], [5]. The first vowels of any systems are initially differentiated by their places and degrees of tongue constriction, then by the use of a distinctive labial or nasal gesture, etc.…”