Murray and Vennemann, (1983) have argued in favor of (A) *- - over (B) *- - for Proto-Germanic on the basis of word division in Gothic manuscripts and meter in older verse. They note that (A) is consistent with the Stressed Syllable Law (SSL), which favors bimoric rhymes, but that (A) violates Hooper's Law (HL), which favors codas weaker than onsets. They cite three phenomena as remedies: gemination, vowel lengthening and glide strengthening. This paper disputes their analysis and cites the role of the Simple (metrical) Foot as evidence in favor of (B), which satisfies HL but violates the SSL. The three phenomena are interpreted as remedies to this violation.*

Previous studies on the development of PGmc. */ga-/ have been inadequate due, in part, to a misunderstanding or underestimation of the roles of word accent and syntactic accent. Purported evidence for stressed */ga-/ in Proto-Germanic or later stages is dubious and can be refuted convincingly. This article suggests that PGmc. */ga-/ never received primary word stress because it had become a clitic. Such a hypothesis allows for an unproblematic phonological derivation from Proto-Indo-European and is a logical premise for interpreting subsequent developments in the Germanic languages.

This article demonstrates that at least one reflex of the oral-formulaic compositional device called ‘the theme of the Hero on the Beach’ is to be found in the Heliand. It is suggested that its presence can enhance the interpretation of another incident in that poem. Further, the author considers the implications of its use in this text for our understanding of the nature of indigenous Germanic poetry prior to the conversion to Christianity.

The long-term practice of translating within a social context where both a dominant and receding language are spoken promotes linguistic convergence: speakers alter grammatical structures of the receding language to match the structures of the dominant language. During individual acts of translating, speakers often resort to two enabling strategies: the speaker attempts structures which are no longer regularly used in the community and thereby promotes the maintenance of conservative norms; or the speaker uses the language being translated as a model and thereby contributes toward convergence. In both cases, translation biases language forms and produces an inaccurate picture of the language skills current in the speech community. Evidence is provided from translated and free discourse in two Pennsylvania German communities: among nonsectarians, where Pennsylvania German is dying, and among Amish and Men-nonite sectarians, where Pennsylvania German continues to be used for daily discourse.