Re: Re[2]: [tied] Re: Marked nominative

On Sat, 17 May 2003, Glen Gordon wrote:

>
> In reference to George Knysh's post on Occam's Razor, a "natural inference"
> is not a _necessary_ inference. While *o seems to indicate voicing, the *o
> can be from various sources that could be equally valid. The contrast of
> *-s and *-d in pronominals however suggests strongly that the origin of
> the nominative in *-s, by way of phonetics and grammatical function, derives
> from the demonstrative stems *so-, for otherwise we remain stumped on
> where *-d comes from... yet it is from *to- quite undeniably.
>
> To reconstruct an otherwise non-existent phoneme **z ignores a more
> unavoidable inference of the nominative's origin in favour of an unnecessary
> and highly assumptive inference that creates more problems than it solves.

Hey, don't think your own thery is all that good either. Though it could
be right by sheer luck, it does look rotten on the surface:

If the nom.sg. *-s and the pronominal nom.-acc.ntr.sg. *-d are in origin
identical with the stems of the forms *so and *tod, the lack of system
leaps to the eye:

If the whole form *so is identical with the nominative marker and so
fulfils that function too, it is understandable that there is no further
"ending" sitting on the end of the form.

But if the *-d of *to-s is identical with the *to- part, why are both
present in this form?

And, if *so and *to-s are not structured in parallel fashion, why are we
to believe an analysis so flawed, even to the point of taking it as the
point of departure for the analysis of the rest of the system and its
phonological elucidation?

Where have I caused problems and you none?

Jens