Underlying Circumflex in Greek

> -----Original Message-----
> From: Miguel Carrasquer [mailto:mcv@...]
> Sent: Tuesday, October 08, 2002 1:51 PM
> To: cybalist@yahoogroups.com
> Subject: Re: [tied] Re: Does Saussure's Law Apply Synchronically to
> Lithuanian?

Responds to: http://groups.yahoo.com/group/cybalist/message/16069>

>
> On Tue, 08 Oct 2002 10:48:45 -0000, "Richard Wordingham" wrote:
>
> >--- In cybalist@..., Miguel Carrasquer <mcv@...> wrote:
> >
> >One could ask what an 'underlying circumflex' is in Greek(?) -
> >http://groups.yahoo.com/group/cybalist/message/15812 :)
>
> Well, either historically circumflex (at a time when Greek may have
> distinguished pitches in a unaccented syllable), or as you say:
>
> >An 'underlyingly circumflex' syllable would be one that would be
> >circumflex if it were to receive the accent, e.g. the first

syllable

> >of krau~jas, which has nominative plural kraujai~.

krau~jas is (historically at least) composed of four morphemes: krau~
+ j + a + s: 'krau~' is the root; 'j' is a suffix; 'a' is the
thematic vowel; 's' is the flexion. We need terms for some of the
aggregations:

a + s is the 'ending'.
krau~ + j + a is the 'stem'.
What is krau~ + j? As I can't remember a better term, I will call it
the 'extended root'.

In Lithuanian, long, potentially stressable nuclei in morphemes have
to be marked as acute or circumflex, for example the 'extended roots'
with mobile stress (e.g. làng/láng versus krau~j) and the endings to
which stress may move. Is there anything parallel to this in
Classical Greek? I can't think of any examples, in which case the
Greek distinction of acute and circumflex is purely a surface
phenomenon.

What is the evidence for the difference between acute and circumflex
in Greek ever being anything other than a matter of where the accent
is placed?

Richard.