| Summary: | <p>This thesis investigates conditioned sound changes of Proto-Indo-European labialized velars *kw, *gw, and *gwh to labials, coronals, and velars in Ancient Greek. Acoustic and perceptual experiments in British English and Western Zapotec provide typological evidence to inform conclusions regarding the phonetic inception of the sound changes.</p>
<p>Before back vowels and consonants PIE labialized velars became labials in Ancient Greek. Two competing hypotheses have been proposed for these sound changes: perceptual confusion, in which e.g. *kw developed to /p/ via misperception due to acoustic similarity between the two stops (Ohala, 1989, 1993); and incremental articulatory change, in which e.g. *kw gradually developed to /kp/ and then /p/ (Whatmough, 1937; Garrett & Johnson, 2013). The acoustic studies here found little acoustic or perceptual similarity between /kw/ and /p/ in any vocalic environment, indicating that the latter explanation is more plausible.</p>
<p>Before front vowels PIE labialized velars became coronals. Philologists have proposed developmental pathways involving processes of incremental palatalization. This thesis finds evidence for coarticulatory fronting of the velar articulation of labialized velars in front vowel environments, but the effect is language-specific and its auditory prominence depends upon the durational overlap of the primary and secondary articulations. Increased auditory prominence of coarticulatory fronting in mid vowel environments may explain divergent developments of *gwh in Greek to labials before front vowels and to coronals before front mid vowels.</p>
<p>Before the back round vowel /u/ the PIE labialized velars lost contrastive labialization in Greek. The acoustic and perceptual experiments support a perceptual reanalysis account, in which contrastive secondary labialization of the velar is reinterpreted as coarticulatory rounding adjacent to a rounded vowel.</p>
|
|---|