Re: [tied] Re: Syncope

On Tue, 30 Mar 2004 enlil@... wrote:

> Jens opposes:
> > That is simply not true. QUANTITAtive ablaut has indeed
> > worked here.
>
> Where? In *suxnu-? In *wlkWo-? In *muhs- "mouse" too?
> Damned if I can see it.

In the mophological grouop we were talking about, the causative. The
accent is on the suffix *-éye-, so the root vowel is deleted, and the
prefix-turned-infix therefore left to serve as syllabic peak:

**O-men-éye- > *mOen-éye- > *mOnéye- > IE *monéye-;

and with a Narten-type root vocalism:

**O-swe:p-éye- > *swOe:p-éye- > *swOepéye- > *swOépeye- > *swOépye- >
*swoépye- > IE *swó:pie-.

All of this obeys the rules of the gradation ablaut we already have.

> It's obvious by the more unusual
> zeroing of _accented_ syllables that the general pattern
> of quantitative ablaut does not apply to these words. The
> question is why.

Of the forms quoted this applies only to *wl'kWo-s, and actually only the
accent is odd. What if it is not original? Then there is no problem and no
interesting information to be gained from it. As it goes on, I get the
feeling that that is in fact what you are saying. Am I right?

>
> Well, either these words originally operated under normal
> ablaut and then deviated by a more recent rule or rules,
> or they are recent already and this process of quantitative
> ablaut no longer applies as strictly as it evidently
> did before, based on patterns like *kwon-/*kun- that we
> see almost everywhere else.

What we see almost everywhere else is in actuality lack of alternation,
presenting sometimes unchanging full grade, other times unchaning
zero-grade. There is no profound mysticism to be found in that.

>
> If they are more recent, we don't need to surmise about
> a hypothetical rule that _may_ (but very possibly may not)
> have swayed these forms from the normal ablaut pattern.
> Assuming first thing without basis that they must be
> sufficiently ancient forms that once operated under ablaut
> is more pointless conjecture because what we immediately
> see is that they do _not_.
>
> Rather, since we can see as the system evolves that
> zeroed verb roots were becoming productive elements in
> derivative words, irregardless of accent, we can simply
> explain these forms during a "postAblaut" stage. (Though,
> calling it a postAblaut stage might be a misnomer since
> ablaut still operates here, but to a looser degree.)
>
> Now, taking *wlkWo- for example, even Mr Burrow mentions
> the painfully obvious in The Sanskrit Language: This is
> just a nominal derivative of an adjective, explaining
> away the accented zero syllable by accent alternation.
>
> Adjectives have accent on the thematic vowel but the mere
> alternation of accent came to nominalize these adjectives.

Sure.

> The accent alternation is the direct result of Acrostatic
> Regularization during a mid-Late IE stage.

No, it appears to have been an expression element in its own right. It is
inherent in the system that you can change the part of speech by changing
the accent. If you do that before the ablaut had worked, you get ablaut
variants, if you wait you get accent variants without covariation in the
vocalism.

> Acrostatic
> Regularization is the cause of Stage II Phonotactics
> where zeroed syllables were now allowed in the first
> syllable of strong forms of stems.

If this is puffed-up style for "after ablaut", I am sure you are right. I
have never seen it explained as anything else.

> However, because
> the accent is _on_ the zeroed syllable, the form is
> purely Stage III, the last layer of Late IE before
> Reconstructed IE. The accent was placed on that first
> syllable to make it a noun. So *wlkWo- doesn't follow
> ablaut because it operates under these younger morphological
> rules that violate it.

Bingo. But there are much more than three stages in the prehistory of IE.

> With *muhs-, either there is a sound rule causing *ou > *u,
> let's say, or the word is formed with a zeroed syllable
> because it is also a deverbal noun, yanked perhaps from
> an aorist form where this zeroing came to occur in Late
> IE.

It looks like a trivial case of paradigm levelling. Weak cases should have
*muHs- anyway, and by levelling the whole paradigm got it.

> In fact, this "derived-noun-from-a-zeroed-verb-root" idea
> works very well for all these more aberrant stems and
> all allowable in the morphology known to IE.

Yes, for this is in reality quite a trivial matter.

Jens