Pairwise lingual differences in tongue shape-location

Abstract

Vowel and consonant phoneme inventories have been structured into multidimensional systems using a relatively consistent range of factors since the 19th century. For consonants these include primary place of articulation, degree of constriction, and voicing. For vowels these include the dimensions close-open, front-back, and rounded vs. unrounded. Secondary dimensions encode airstream, source characteristics or other aspects. Detailed impressionistic and instrumental research can be used to group phonetically-similar phonemes into natural classes (e.g. “alveolars” or “close-front vowels”). Phonological allophony, morphophonemic alternation and patterns of phonetic variation within and across speakers support more phonetically-heterogenous (aka “abstract”) classes. But if two contrasting phonemes are phonetically almost identical for some phonetic property, such as lingual shape and place, it is straightforward to argue that they share whichever phonological features are used in the analysis of that property. For example, if two vowels contrast, but their midsagittal tongue profiles are identical (or close enough), then we would be highly likely to analyse them with the same lingual features. For a specific example, consider two vowels [i] and [y]. They would probably both be phonologized as /i/ and /y/, i.e. both /+HIGH/ and /+FRONT/, and contrast in /ROUND/.