Re: [tied] Nominative Loss. A strengthened theory?

On Sat, 17 Apr 2004 enlil@... wrote:

> Jens:
> > Then why does Brugmann bracket the r of *p&té:(r) and *k^uo:(n)? And
> > why does Pokorny do the same? Because they saw no good reason? You
> > are styling ignorance as a virtue.
>
> Brugmann and Pokorny do a lot of things that they shouldn't have like
> forget to reconstruct laryngeals (oops!) but they can be forgiven
> because they lived in a different time with less knowledge than
> we have now.

This is preposterous. Not reconstructing laryngeals was no fault of
Brugmann's, for the theory had not been sufficiently developed and
discussed to be ready for a handbook. Even half a century later that had
not really changed, so Pokorny made the deliberate choice of not noting
them. His preface contains an explicit message to that effect so that any
user is at liberty to read the reconstructions as he thinks they should
be. And neither Brugmann nor Pokorny is at fault in the matter we are
addressing now because you asked me about it.

> However, the absence of the final consonant of the stem here would
> seem to not matter much because if this optional absence is produced
> by sandhi conditions. Then it has no bearing on our topic which was
> concerned only with these words in _isolation_.

> In isolation, the words don't drop *r or *n which is why we can
> reconstruct them in the first place! It merely drops the nominative
> in *-s. You are bringing in this new issue but it's only clouding
> the search for a solution.

That is an unmotivated choice. While it seems rather obvious that the
wordforms are best preserved before a following vowel, and less so before
a following consonant, nothing really indicates what they looked like
before zero. The branches simply make different choices. This may be a
silly thing to introduce (actually I don't think it was), but I did it on
your request.

> Perhaps at best you're trying to explore how *-s might have
> disappeared because of sandhi but I'm at a loss to know how this
> might occur? Is that it, or does your mention of sandhi have no
> logical relationship to any of this at all?

No, I was trying to explain why I reconstruct the PIE forms as I did.

> >> If it were sandhi then, it still doesn't mean that the stem lacks
> >> these sonants in the nominative, but rather that in colloquial
> >> speech they were omitted based on a context and rules external to
> >> the stem as it would be found in isolation.
> >
> > Pray reveal to us the basis of your insight.
>
> It may be as simple as: before another consonant, one would have
> said *pxte:, instead of *pxte:r before a word beginning with a
> vowel or in isolation. This then would say nothing about sound
> changes per se or how the word changes in isolation, but rather
> about how a word interacts in a larger sentence between other
> words.

Yes, that's what I said and had to repeat when you protested against it.
Only, we do not know which variant belonged in pausa.


[...]

> You haven't established how sandhi relates to this, so until you
> do, why introduce trivia to confuse the topic?
>
> PS: */_Nu_ so ?so:nts estne/? :)

No, he is not truthful either - he asked me about it, and now he won't
remember.

Jens