Published online by Cambridge University Press: 01 September 2010
Gothic scribes divided words at the ends of lines following basic principles of syllabification except in the case of glide-final clusters, where the division appeared immediately before the glide (-C/G-). Pierce (2006) accepts this practice as evidence that such sequences were heterosyllabified, -C.G-. He rejects arguments that such breaks occur because morpheme boundaries (#) normally precede the glides, -VC0C/#GV-, and that hence these clusters were actually tauto-syllabified: -VC0.C#GV-. Pierce counters that (a) the Law of Initials forbids such tautosyllabification; (b) morphology cannot account for similar division in forms evincing Verschärfung or /-ngw-/, where there are no morpheme boundaries; (c) his opponents are inconsistent because they ignore the role of morphology in the division of stop + liquid clusters; (d) evidence from the other Germanic dialects cor-roborates his position; (e) if -VC0CGV- were truly tautosyllabified, one would expect occasionally to find erroneous divisions exhibiting this, -VC0/.C#GV-. In this rejoinder, I argue that Pierce's arguments rest on false premises.