Date of Award
Doctor of Philosophy (PhD)
Modern High German final obstruent devoicing is usually thought to descend from Middle German devoicing without any chronological break, despite the fact that the graphic representation of final devoicing ceased in the Early Modern period. However, an alternative account holds that the spelling change reflects the actual loss of the devoicing rule, and that therefore the modern rule has an independent origin. In particular, apocope of final schwa has been suggested as the cause of the loss of devoicing in Early Modern German.
According to this theory, loss of devoicing occurred because schwa apocope rendered the devoicing rule opaque, and hence hard to learn. If true, we expect to see some evidence for opaque devoicing during the period that apocope was in progress. In accordance with this prediction, we found a statistically significant correlation between apocope and absence of final devoicing in a number of German texts of the 14th and 15th centuries. After the 15th century, devoicing is lost across the board, which correlates with the completion of schwa apocope and the loss of the opaque devoicing rule. This confirms our theoretical predictions. If apocope had not rendered devoicing opaque, we would have to conclude that Early Modern German schwa apocope was an instance of rule insertion. However, the structural description of neither apocope nor devoicing leads us to expect insertion. Instead, Modern German final devoicing appears to be an instance of rule re-affirmation, which entails that the devoicing rule, though opaque, remained productive in some dialects.
Gress-Wright, Jonathan, "Opacity and Transparency in Phonological Change" (2010). Publicly accessible Penn Dissertations. Paper 266.