UCLA

This dissertation examines the phenomenon of Proto-Indo-European schwebeablaut (German Schwebeablaut “floating vowel gradation”), whereby a number of reconstructed forms are observed to alternate in their root shape between CeRC (termed State I) and CReC (termed State II). This mechanism of Proto-Indo-European (PIE) root ablaut has long been tacitly accepted (in one form or another) by scholars; however, the only comprehensive treatment has been Anttila (1969), and the matter therefore merits a thorough review. This dissertation reviews material from the daughter languages considered to be evidence for schwebeablaut by using some of the same techniques employed in the work of Anttila, only in an updated fashion. A large majority of the remaining cases are explained without requiring recourse to schwebeablaut, while several more difficult forms are discussed in individual chapters. The second part of the dissertation examines a unique subtype of roots extended by an s-formant that exhibit the following alternation: CeRC : CReC-s. It is argued that the descriptive schwebeablaut inherent in these formations may legitimately be traced to PIE. An Optimality Theoretic account for this phonologically motivated metathesis is offered, which likewise eliminates the need to invoke schwebeablaut (as classically defined).