--- In
cybalist@yahoogroups.com, "alex" <alxmoeller@...> wrote:
> Richard Wordingham wrote:
> Question:
> was this probable example an example to show me there is _indeed_
an x >
> f ?( even if with the status of "probable")
There is no doubt about the change of syllable-final Old English "h"
to Modern English "ugh" after a back vowel. Sometimes it is
realised as /f/, sometime as mere vowel length or a diphthong. Your
challenge was to find a case of k > f. The only doubt lies in the
descent from PIE, though OE _ru:h_ 'rough', Lithuanian
_rauka_ 'wrinkle', Sanskrit _ru:kSas_ 'rough' seems to be a good
correspondence. I'm not sure that their PIE root (*reuHk ?) belongs
in Pokorny's reu-2, but that's another issue.
> So , the specialist in phonology will argue now that no, there is
no pt
> > ft but ct which a) ct > xt > t and b) ct > xt > ft and c) from
this
> "ft" we have an "pt"?
pt > ft and ft > pt both occur, even though the former seems to be
commoner. One can also find unconditioned f > pH, e.g. in the
development of Shan, where it is part of a 3-way merger of *f, *v
and *pH.
Richard.