On the purposes of phonological phenomena: phonetics, parsing and/or acquisition?

The first part of this paper investigates two hypotheses that have been put forward about the purpose of phonological phenomena, namely the mainstream view that the properties of the articulatory or auditory system motivate the phonology (the Phonetic Hypothesis), and Kaye's proposal that the phonology helps the hearer parse the continuous input string into distinct cognitive units (the Parsing Hypothesis); ultimately, both the Phonetic Hypothesis and Kaye's version of the Parsing Hypothesis will be rejected, even though a more scientific version of the latter will be offered. This will lead to the conclusion that various phonological phenomena cannot adequately be explained by either hypothesis. The second part of the paper will subsequently introduce the Acquisition Hypothesis, i.e. the claim that certain phonological phenomena (specifically, vowel harmony in Turkic languages) disambiguate acoustic percepts, such that a child acquiring some language can come to a unique decision about what kind of vowel system s/he is dealing with. In this way, phonological phenomena narrow down the analytical options the child has and make many otherwise unacquirable vowel (and, presumably, consonant) systems acquirable. Southern African Linguistics and Applied Language Studies 2005, 23(4): 405–42