tgpedersen wrote:
> Jysk has o: > uó > wo, 'Ole' (name) > 'Wolle'
> which shows you can get all the way without using a term such
> as 'prop glide'.
It _is_ a prop glide if it appears only in syllable onsets
(word-initially or in hiatus). Of course o > uo may be a case of
diphthongisation, but in such cases the onglide will appear also in
other contexts. For example, in some Polish dialects inherited /o/ is
realised as a rising diphthong, [uo] or even [ue], in all positions,
whereas older /O/ from lengthened *a: is realised as plain [o]. But even
dialects without such diphthongisations insert a glide before an
_initial_ /o/.
Piotr