Re: USAGE: VOT and the status of /r/

Tristan wrote:
> Tom Weir:
NB:  My family name is <Wier>. :)

> > Marcos wrote:
> > > What's the phonemic status of trailing -r in non-rhotic
> > > English?  Is it considered an underlying phoneme whose phonetic
> > > realization is a modification of the preceding vowel (or none
> > > at all, in the case of schwa), or merely a graphical convention
> > > used to indicate in writing which of two vowel phonemes should
> > > be selected?  (At least in those cases where English spelling
> > > has some bearing on pronunciation.)
> >
> > With respect, I must disagree with Tristan.  In most nonrhotic
> > dialects, the /r/ is really there underlyingly.  One can prove
> > this by how it will show up as a regular English [r] in liaison
> > situations, as when the following word begins with a vowel.  The
> > only catch is that a lot of nonrhotic dialects, such as some in
> > Britain and New England, have an epenthetic linking-r where
> > historically no such /r/ existed*.  To my knowledge, nonrhotic
> > varieties of the American South do not have that linking-r.
>
> I wasn't talking about varieties of the American South, but only
> my English.
Okay, but the question was asking about nonrhotic dialects in
general, of which Australian dialects are only a part. Indeed,
the number of speakers of nonrhotic English dialects in the
United States, though a small minority there, exceeds the
entire population of Australia, and then that's not including
the 50 or so million people in the UK who speak nonrhotic
dialects.

> So I really don't see why you'd say it was underlyingly there
> in the only English I'm capable of commenting on :)
You might be right about your own dialect, although I'm pretty
sure I've heard Australians who don't always have intrusive
linking-r for hiatus between words.

> > *(This fact is not expected if one holds to a theory of phonology
> > such as Optimality Theory where the quality of epenthetic segments
> > should fall out from general markedness constraints.)
>
> I'm not sure what any of that means :)
In OT, the quality of epenthetic segments is supposed to be a
result of constraint interaction, and is not written directly
into the grammar. So, e.g. if one encounters an instance of
hiatus, and the other principles of the grammar say it's better
to epenthesize a consonant than to delete one of the vowels,
the choice between [j], [t], [w], [h], etc. is supposed to be
decided by how highly that language ranks the constraint
favoring stops versus glides, or labials versus coronals,
etc.  But since the set of constraints is not symmetric
(i.e., there are some constraints militating against retroflexes,
but not necessarily any militating against coronals), one
does not expect to see a marked segment like [r] ever arising
through constraint interaction.  And yet it's there, and apparently
in your dialect thorough-goingly so.

 =========================================================================
Thomas Wier	       "I find it useful to meet my subjects personally,
Dept. of Linguistics    because our secret police don't get it right
University of Chicago   half the time." -- octogenarian Sheikh Zayed of
1010 E. 59th Street     Abu Dhabi, to a French reporter.
Chicago, IL 60637

Reply