Re: OT: Evolution of dialects (was Re: Tirelat and related dialects)

Benct Philip Jonsson wrote:

> Maybe the problematic status of Tirelat (IIRC)
> stress and length which you, Herman, mentioned
> some weeks back is because the dialects which
> formed the base of the standard language had
> actually lost length, but it was erratically
> reintroduced from erratic spellings where the old
> graphies for long vowels had partly come to be
> used to mark stress?
I think it's possible that words like "katee" (arch) may have had short
vowels originally, but stressed vowels in final position were lengthened
in the standard dialect: /ka'te/ -> /ka'te:/ . In that case, the
dialects that retain the pronunciation /ka'te/ are preserving the
original form of the word. Also, the "epenthetic" vowels in numerous
words correspond with regular unstressed vowels in other dialects. But
other explanations are possible. Stress may have become phonemic when
the distinction between long and short vowels was lost. In any case,
it's possible that a "vowel length" mark may have been interpreted as a
"stress" mark or vice versa.