Re: Negation

On 2008-12-05 21:25, Arnaud Fournet wrote:

> Which is not inherited ?
> Why this restriction ?

Some words (admittedly, not _many_, but certainly _some_, especially of
"expressive" origin) are coined from scratch (e.g. <google> or <zoom>).
Others are "without etymology" and we can't know if they are inherited
or not.

> This sounds a little bit spurious to me.
>
> For example, considering the fact that j(e) is a bound form in modern French
> is it a new word when compared to LAtin ego, which was a free form ?
> Is this a phonetic or a morphological change ?

Formally, <j(e)> reflects <ego> through purely phonetic development. The
fact that its grammatical status has changed doesn't affect this
ancestor/descendant relation. <knight> is the modern descendant of OE
cniht (note the loss of two consonants), though its meaning has changed.

> And what about l'outre "otter" > la loutre ?
> is loutre a new word when compared to outre ?

Not that it matters, but Fr. loutre comes from lat. lu:tra. Of course
there are cases when new phonetic material is added to a word through
univerbation, reanalysis or the like, but see below.

> Russian gorod from grad is one syllable longer, is it a new word ?

Russ. gorod is from PSl. *gordU, not from "grad". "The same" word,
historically, and yes, East Slavic pleophony results in phonetic gain
rather than loss. So what? For every process like that one can cite a
dozen reductions. On the _average_, Russian words are shorter than their
Proto-Slavic ancestors.

> This is a counter-example to phonetic changes implying shorter forms, is it
> not ?

I have never said that _all_ phonetic change consists in loss. It's a
long-term tendency, not an iron-clad rule, but it's strong enough to
predominate in the history of every language. A statistically
significant tendency cannot be falsified by quoting a few deviations
from it.

Piotr