--- In
cybalist@yahoogroups.com, "Glen Gordon" <glengordon01@...>
wrote:
>
> Jens:
> > The thematic vowel itself
> > alternates e/o ~ u ~ i. The Caland business alternates *-re/o-
~ -u- ~
> > -i-. Is this unrelated?
[Glen:]
> It would seem to me it isn't related. It looks like nothing more
than
> someone trying to find a pattern where there is none.
>
> [...] I feel uneasy at inventing a new
> soundrule every time there's an alternation. It seems more like
> something analogical in this case.
I have heard this many times. Contrary to what is implied I am aware
of the dangers pointed out in such constant admonitions and take
pains to keep a clear head, and I would appreciate if critics could
do the same. However, the rules of the game are not as simple as is
often implied. If an alternation is observed, the first thing to be
looked for ought to be a phonetic rule, for that is the vulnerable
position. If you posit a wrong phonetic rule it is liable to be
discovered that it does not work. But if you ascribe an allomorphy
to older expressions of a lost functional difference, you take the
matter into a lawless territory where mistakes are much harder to
detect. Functional differences are not amenable to algebraic
analysis, phonetic alternations are. That's why I concentrate on
phonetic solutions, and that in turn is why I have come up with more
phonetic than functional suggestions in my time. That is also why
some of the phonetic hypotheses (and none of the others) have had to
fall in the face of decisive evidence.
That said, it is not really true that I invent a sound law every
time I sense an alternation. I watch out for it, yes, but I do not
put forward a theory about it as long as it is completely ad hoc.
But I suggest it, and with decreasing reservation, as indications
for it accumulate and it gets ad illud and ad aliud and ad-what-not
also.
So, sure, most of my results are phonetic in nature, but that is
just because that is what I mostly look for. You don't catch rabbits
if you go fishing.
I wonder if you have any basis for the last sentence, "It seems more
like something analogical in this case." Could you explain how the
Caland alternations could possibly be analogical, and state the
reasons you have for believing that such a possibility is true?
Jens