UC Berkeley

This dissertation investigates three sound changes in the early history of Germanic with an approach grounded in phonetics. Historical phonology has traditionally focused on the articulatory aspects of change (e.g., Hoenigswald 1960; King 1969). However, more recent work in phonetics on sound change has emphasized the acoustic and auditory aspects of sound change, alongside the articulatory (e.g., Beddor 2009; Blevins 2004; Ohala 1981). The present work has two goals: first, to advance the state of research on the sound changes in question; and second, to show how the findings of modern laboratory phonetics can complement the study of historical phonology.

In Chapter 2, I review past approaches to sound change, as well as more recent work in phonetics. In Chapter 3, I consider OHG i-umlaut, a longstanding problem in the field. Although umlaut-type changes are common in Germanic, and other types of vowel harmony are widely attested in diverse languages, I show that by attending to all the relevant phonetic factors, the change can profitably be reanalyzed, despite the vast literature surrounding it. I conclude that OHG i-umlaut was a type of hypo-correction, and that the phonological conditions in the late OHG period, coupled with individual variation in coarticulation, conspired to form the necessary circumstances for the change to occur. In Chapter 4, I investigate the raising of nasalized mid vowels in the early Germanic dialects. This change has been well documented in the literature, but there have been very few attempts to actually explain why nasals condition the change. By surveying the articulatory, acoustic, and perceptual properties of vowel nasalization, I show that the change is best understood as an instance of hyper-correction. This model predicts the observed changes, as well as some of the variation in the change’s conditioning environment among the dialects. In Chapter 5, I evaluate two different interpretations of the orthographic sequence in Gothic. I argue that Gothic represents [ŋɡ] in all instances, regardless of etymology. The literature on this subject has generally rejected such a view, because it presupposes an earlier change of *[ɡɡ] > [ŋɡ], which many scholars have viewed as implausible. By evaluating [ɡɡ] in the light of the aerodynamic voicing constraint relative to the structure of Gothic phonology, I conclude that such a change was in fact highly plausible, thereby strengthening the argument for the single interpretation of as [ŋɡ].

Each of the analyses of the individual sound changes stands on its own, but also serves a larger theoretical goal: to demonstrate the value of the study of both phonetic production and phonetic perception to historical phonology. In each of the chapters I identify how phonetics can help solve a more general type of problem, not just the specific sound change investigated in that chapter. OHG i-umlaut, discussed in Chapter 3, exemplifies sound changes in which the basic mechanism is already well understood. The raising of nasalized vowels, discussed in Chapter 4, is an instance of a sound change where the environment has been identified, but the conditioning is not well understood. The issue of Gothic discussed in Chapter 5 raises the question of phonetic plausibility.

In Chapter 6, I review the analyses of the preceding chapters, and outline possible avenues of further research.