Re: 'Vocalic Theory'

--- In cybalist@yahoogroups.com, "Patrick Ryan" <proto-language@...>
wrote:

> From: "Richard Wordingham" <richard@...>
>> Patrick Ryan wrote:

>> > From: "Miguel Carrasquer Vidal" <miguelc@>
>> > > On Sat, 29 Mar 2008 19:48:44 -0000, "etherman23" wrote:

>> > >> If we have, for example, *H2eH3 is the
>> > >> outcome o: or a:? Does the initial laryngeal color first
>> > >> or does the
>> > >> final color first? Or is there some other rule?

>> > > It looks as if *h3 > *h2 > *h1.

>> ... '*h3 > *h2 > *h1' clearly means that the effect of
>> *h3 takes precedence over the effect of *h2, which in turns takes
>> precedence over the effect of *h1. I think the latter just means that
>> *h2eh1 > *h2ah1 > a: (last stage only as laryngeals are lost), as
>> opposed to *h2eh1 > e: with lengthening preventing colouring.

> ... if I interpret it as you see it, it is simply ridiculous.

> Why should a backing effect trump a centralizing effect?

> Explain that phonologically if you can.

Or does *h2 merely inhibit a negatively conditioned fronting? Perhaps
it is more like *e ([&]) becoming [e] except in certain environments
(next to *h2, to some dialect-dependent extent next to pure velars).

> As for poor *H1, what effect does it have other than to prevent hiatus?

> Winning the coloring bee over *H1 is a no-contest.

Which made me wonder why *h1 was mentioned. However it may have been
mentioned because its (post-PIE?) lengthening effect does not bleed
the colouring effects.

> And, as I pointed out to Miguel, that length in vowels prevents

coloring is

> phonologically repudiated by the _fact_ that 'emphatics' in Arabic

produced

> 'colored' allophones of all vowels, long and short.

The PIE colouring resistance is probably an instance of a
language-specific feature - just like Dutch umlaut resistance, and the
fact that glottalisation colours vowels in some languages but not in
others. (Arabic emphatics are not glottalised, but their cognates are
in some Semitic languages.)

Richard.