Newsgroups: alt.usage.english,sci.lang
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!scramble.lm.com!news.math.psu.edu!news.cse.psu.edu!uwm.edu!cs.utexas.edu!news.tamu.edu!newshost.comco.com!newsfeed.concentric.net!vyzynz!nielsen!news2.usa1.com!dmn1-68.usa1.com!user
From: aaron_j._dinkin@fourd.com (Aaron J. Dinkin)
Subject: Re: Allophones vs. Phonemes (was: ae)
Message-ID: <aaron_j._dinkin-2906962236180001@dmn1-68.usa1.com>
Date: Sat, 29 Jun 1996 22:36:18 -0500
References: <31c8ba59.28879345@news.Leiden.NL.net> <31CC0FC3.296E@atlcom.net> <4qs3o1$fqe@gerry.cc.keele.ac.uk> <4qs55l$pof@bone.think.com> <aaron_j._dinkin-2606961905410001@wakma1-23.usa1.com> <ALDERSON.96Jun28141810@netcom10.netcom.com>
Lines: 27

In article <ALDERSON.96Jun28141810@netcom10.netcom.com>,
alderson@netcom.com wrote:

> In article <aaron_j._dinkin-2606961905410001@wakma1-23.usa1.com>
> aaron_j._dinkin@fourd.com (Aaron J. Dinkin) writes:
> 
> >Similarly, although your dialect forbids [r] at the end of a word, /r/ still
> >exists phonemically: you would say ['siz@r Iz], not ['siz@ Iz] for "Caesar
> >is"; therefore the /r/ has not been totally eliminated.
> 
> Now *here* is where we *really* disagree.  The very essence of dialects is
> that they can differ in phoneme inventory, lexical representation, syntax,
> and so on.  So it is extremely likely that in the dialect in question the
> lexical form is /si:z@/, *not* /si:z@r/, and there is a rule (not a process)
> that hiatus in the context /-@#V/ is avoided by means of an inserted /r/.
> Evidence for this is the fact that not only "Caesar" and "father" but also
> "sofa" get this form of hiatus avoidance.

My apologies; I thought "intrusive /r/" (in contrast with "linking /r/")
was frowned upon - i.e., that ['siz@r Iz] was standard but ['s@Uf@r Iz]
was not. I hereby retract my above accusation, since it was based on
ignorance.

(Relish the moment; I'm usually far more arrogant.)

Aaron J. Dinkin
Dr. Whom

