Barrack (1991) argues against the reconstruction of PGmc. syllable structure proposed in Murray and Vennemann (1983), namely, . According to Murray and Vennemann, this reconstruction is compatible with characteristics of early Germanic verse as well as Gothic word divisions and provides a basis for a coherent interpretation of major phonological changes in early Germanic; namely Sievers's Law, West Germanic gemination, North Germanic resyllabication, and glide strengthening in Gothic. Focusing on these same points, Barrack attempts to reinterpret them as providing evidence for . In this response, I demonstrate that Barrack fails on all counts. Most importantly, he fails to provide a coherent, well-motivated interpretation of the pertinent phonological changes on the basis of his assumed .

Standard movement theory stipulates that any X"-phrase that is moved must move to the specifier position (SPEC) of C". In this paper, however, it is argued on the basis of coordination phenomena, verb-final structure, and diachronic evidence that, in relative clauses and indirect questions, the moved pronoun or w-element functions more as a true complementizer than as a fronted X"-phrase and thus should move to the head position C. Movement to C is thus determined not by the phrase-structure level of the moved category, but by feature compatibility.

Several years ago we started work on a new etymological dictionary of the English language and published a short announcement of our plans (Liberman and Mitchell 1988). Now the contours of the dictionary have become clear enough for us to begin writing our first entries and to introduce our project to a wider range of scholars.