--- In
cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:
> On Thu, 08 Apr 2004 10:43:48 -0500, Crystal Odenkirk
> <crys@...> wrote:
>
> >I'll have to sift through the archives then, perhaps tomorrow or
this
> >weekend, and see what I can come up with. Maybe, since I'm
compiling it for
> >myself anyway, I'll post it here at some point to make sure I've
got them
> >right and then put it in the "Files". There are a lot of archives
though so
> >I'm sure it'll take a while...
> >
> >So far there's really only one symbol that I'm unfamiliar with as
a sound
> >though. What does the 2 represent when it's within an IE word?
And am I
> >correct that &=schwa?
>
> The inevntory of PIE sounds is not that large:
>
> Stops:
> *p, *b (rare), *bh. Piotr writes *bH, to mark the fact that
> <h> is here not a separate phoneme, but serves to modify <b>
> [*bH is aspirated b]. I write *bh because there's usually
> no confusion possible (there is no phoneme /h/).
> Similarly:
> dentals: *t *d *dH
> velars/uvulars: *k *g *gH
> palatals/velars: *k^ *g^ *g^H (some people use *k', *g',
> *g'h)
> labio-velars: *kW, *gW, *gWH (or is it *gHW?). W is here a
> labialization modifier, and there _is_ the possibility of
> confusion with /w/).
>
> Fricatives:
> *s
> *h1 *h2 *h3 (laryngeals). Also written H1, H2, H3.
> Unspecified laryngeal *H.
There is a potential confusion here between aspiration and a
laryngeal. That's why the commoner notation here is to use "H" for
aspiration, and "h" for laryngeal. A syllabic laryngeal is written
as "&". "h" or "&" is followed by a digit to distinguish the
different laryngeals.
For languages other than PIE, we also write "&" for schwa, because
words containing "@" are mangled in the archives to hide e-mail
addresses from spammers.
To associate the laryngeals with vowels, number across the bottom of
the (compressed) vowel trapezium:
e a o
1 2 3
There isn't actually any e-colouring; /e/ is the easily coloured
vowel. When the laryngeals vocalise in Greek, we do get /h1/
> /e/, /h2/ > /a/ and /h3/ > /o/, except that /h1/ may be coloured
by a neighbouring vowel, as in *h1dont- > odont-, but edont- in some
dialects.
/h2/ converted a neighbouring /e/ to /a/, so whether one
writes /h2e/ and /eh2/ or /h2a/ and /ah2/ is largely a matter of
taste. An argument for consistently writing /h2a/ is that that is
correct even if the vowel was originally /a/.
Miguel's statement that PIE didn't have /h/ is misleading. On
typological grounds, it ought to have had [h], and as all words seem
to have begun with consonants, it should also have had [?]. It is
quite likely that their reflexes have merged so completely that all
we can reconstruct is *h1 for the two of them.
Glen Gordon has been in the habit of writing the laryngeals as he
things they were pronounced, and so writing /?/ for /h1/, /x/
for /h2/ and /xw/ for /h3/. For pre-PIE he also writes the stops as
he thinks they were phonetically, so writing /t:/ for /d/ and /d/
for /dH/. I prefer to treat the way of writing PIE as a spelling
system, and I think I have actually written <d> to mean 'what is
conventionally reconstructed as /d/'. Slashes (//) delimit
(sequences of) phonemes, square brackets ([]) delimit (sequences of)
phones and angle brackets (<>) delimit spelling. There have been
pleas to use double quotes (") instead of angle brackets because
some browsers misinterpret them as HTML directives. </soapbox> :)
Richard.