In a Greenbergian perspective grammatical systems of tense, aspect and modality have their sources in the processes that create them through grammaticalization, and the similarity of these processes across languages accounts for similarities and differences in synchronic states. Applying this perspective to phoneme inventories, we expect the process of sound change to create, change or obliterate phonemes. Thus the similarity of sound change across language would account for the similarities and differences among phoneme inventories in the languages of the world. This hypothesis is tested using Lindblom and Maddieson’s 1988 analysis of basic, elaborated and complex phoneme inventories, and a database of phonetic processes (standing in for sound changes) in a genetically balanced sample of 82 languages (Bybee and Easterday 2019). We find strong evidence that elaborated and complex Cs are created by common sound changes. What are regarded as ‘basic consonants’, however, form two distinct categories: those with attested sources in sound change and those which are rarely created anew by sound change. Paradoxically, the most common Cs in inventories—the (voiceless) stops—are often the input to sound change but rarely occur as its outcome, raising the question of why it is that most languages have them in their phoneme inventories.
February 27, 2020 - 11:21am