Re: PIE **fr- (was: Reasons)

From: Richard Wordingham
Message: 34583
Date: 2004-10-10

--- In cybalist@yahoogroups.com, Sean Whalen <stlatos@...> wrote:
>
> --- Richard Wordingham
> <richard.wordingham@...> wrote:
>
> >
> > --- In cybalist@yahoogroups.com, Sean Whalen
> > <stlatos@...> wrote:
> >
> > > Many languages lack sr- or str-.
>
> > Since for
> > other
> > > reasons I added f, there's no reason not to think
> > that
> > > fr- existed.

> However, I don't think there is a
> gap. Mainly based on words like *sri:g- "cold" that if
> from **friXg- would be analyzable as onomatopoeia and
> possible links to words in Proto-Uralic and
> Proto-Semitic for which I'm working out regular sound
> shifts.

That would be 'comparative' evidence. Good luck!

> > Siamese doesn't have initial fl- or fr-, though it
> > has p(h)l-, p(h)r-
> > , thr- and k(h)r-.
>
> So, no sr- or any other cluster with a fricative?

Not in Siamese, unless you want to count the spellings <h> +
resonant, which probably were once clusters like Old English hn-, hl-
, hr-, hw-, but by the time of writing may merely have indicated
voicelessness like Welsh <mh>, <nh>, <ngh>, <rh>, and now, 700 years
later, merely help record the tone of a syllable. siamese has three
fricatives, /s/, /h/, and /f/. <sr-> and <çr-> occur in writing,
but the <r> comes from foreign loans (chiefly Sanskirt) and is
silent.

If you roll back to Proto-Tai, Li reconstructs the following Proto-
Tai initial clusters:

*hm (> Siamese /m/)
*pl (> Siamese /pl/)
*pr (> Siamese /t/)
*pHl/r (> Siamese /pH/)
*bl (> Siamese /pHl/)
*br (> Siamese /pHr/)
*?bl/r (> Siamese /d/)
*ml/r (> Siamese /m/, /l/, /mal/)
*fr (> Siamese /pH/) - reconstruction 'uncertain' and 'unreliable'.
*vr (> Siamese /pHr/) - reconstruction 'uncertain'.

*hn (> Siamese /n/)
*tl (> Siamese /t/)
*tr (> Siamese /t/)
*tHl (> Siamese /tH/)
*tHr (> Siamese /h/)
*dl (> Siamese /l/)
*dr (> Siamese /r/)
*?dl/r (> Siamese /d/) (Proto-Kam-Tai rather than Proto-Tai if you
exclude Saek from Tai, as Ethnologue does.)
*nl/r (> Siamese /n/)

*hl (> Siamese /l/)
*hr (> Siamese /r/)
*hñ (> Siamese /j/)

*hN (> Siamese /N/) - /N/ = velar nasal
*kl (> Siamese /kl/)
*kr (> Siamese /kr/, /kl/)
*kHl (> Siamese /kH/)
*kHr (> Siamese /kH/)
*gl (> Siamese /kHl/)
*gr (> Siamese /kHr/)
*Nl/r (> Siamese /N/) (one example, but backed up
by /Nklau/ 'sesame' in Gelao, a Kra language of Guizhou. Ostapirat
reconstructs l-N- for the Proto-Kra initial.)
*xr (> Siamese /h/)

*kw (> Siamese /kw/)
*kHw (> Siamese /kHw/)
*gw (> Siamese /kHw/)
*Nw (> Siamese /N/, /w/)
*xw (> Siamese /khw/)
*Gw (> Siamese /khw/) (/G/ = voiced velar fricative)
You could add the preglottalised consonants *?b (> Siamese /b/), *?d
(> Siamese /d/) and *?j (> Siamese /j/, but spelt <'><j> in 4
words). The 'palatals' are unsplittable hushing affricates -
Siamese has no hushing fricatives.

> Well, that would be pretty interesting. Are there
> any compound words that might have formed fL- but were
> prevented?

Bear in mind that Siamese is fundamentally monosyllabic, even though
it does have quite a few sesquisyllabic words and Sanskrit/Pali-
derive polysyllables. Totally fused compounds are pretty rare in
Siamese - the only ones I know of are /kHrai_M/ 'who'
< /kHon_M/ '(classifier for) person'+ /rai_M/ 'what (adj.)?' and the
colloquial /siip/ (not sure of the tone - I even seem to have
misheard the vowel length!) '20' from /yii_F sip_L/ '20',
literally 'two tens'. For incipient compounds, you might expect to
see sesquisyllables starting fa-, but I couldn't find any in my
dictionary.

Speaking as a moderator, if we're going to have much discussion of
non-Nostratic analogies, we should probably move to Phonet. One
drawback is that membership requests wouldn't be approved until
Piotr made some time at work, hopefully on Monday.

Richard.