LINGUIST List 2.807

Thu 21 Nov 1991

Disc: Hiatus

Editor for this issue: <>

Directory

1. Jim Scobbie, Re: 2.796 Queries: hiatus; Spanish; Italian; MAC; pidgins; pronouns
2. Nancy L. Dray, a apple (cf. 2.790)
3. Logical Language Group, hiatus

Message 1: Re: 2.796 Queries: hiatus; Spanish; Italian; MAC; pidgins; pronouns

Geoffrey Russom <EL403015brownvm.brown.edu> says:
>It seems to me that if the vowel in "a, the" is schwa, only a glottal
>stop could prevent hiatus in your cited phrases "a apple, the apple."
>Does anyone know of other possibilities...
The schwa-final words "a" "the" and "to" do not, in my experience, take
part in r-sandhi. An underlying /r/ analysis must posit a different
representation for these nonalternating schwa than it does for the
"-er" suffix in "manager" and the schwa in "India". (I analyse the
long vowels /scripta:/ of "ma", "mar" and /openo:/ of "sore", "saw"
as the same thing also, [-high] for concreteness.) Now this seems
a bit of a problem, and perhaps it is. But note that these words that
do not deal with the ill-formed hiatus of [nonhigh-vowel][vowel]
by the realisation of an onset [r] each has a pre-vocalic allomorph.
a /a/ - an /an/ (or //-/n/)
the /dh/ - the /dhi:/
to /t/ - to /tu:/
So does the prevocalic allomorph bleed the r-sandhi environment? Well,
for a-an I think it fair to say it does. "A apple" is strongly deviant
I think. But the "to eat" /t?i:t/ and "the eagle" /dh?i:gl/
versions seem to be ok. The allomorphic pronunciations
tuwi:t			(with, to answer geoffrey's question, a /w/ glide)
dhiyi:gl		(a /j/ glide)
seem to be being lost.
(I may be getting mixed up here by those accents which have strong onset
glides after high vowels in addition to the strong onset glides
after nonhigh vowels (r-sandhi), but I'm fairly sure the use of /w/ and
/y/ above is independent of these accents.)
Of course, this predicts that the allomorphs "to" /tu:/ and "the" /dhi:/
can appear before pause, although r-sandhi cannot.
Th[i:] ... apple
*fo[r] ... Andy
I think this is correct.
It means supposing that the availability of an allomorph takes precedence
over the appearance of some phonology. I take it this is no problem.
Otherwise it means supposing that to/the have underlying glottal glide,
or, better, being unstressed vowels in function words, have different
vowels than word-final vowels do (a polysystemic analysis).
--
James M. Scobbie: Dept of Linguistics, Stanford University, CA 94305-2150

Re Russom's comment (2.790) on glottal stop between, e.g., "a"
(pronounced schwa) and "apple" to prevent hiatus:
Another alternative is to have intrusive "h"--e.g., "a happle."
If I recall correctly, this is attested in at least one variety of
Newfoundland English. It may be in the same area where you "drop
your H in 'olyrood and pick it up in H'Avondale," but I seem to
recall there being several versions of h-dropping/adding, so I
won't try to be any more specific (though I'd be happy to hear from
anyone who can be...). I'm sure there are also other dialects of
English that have "a happle" rather than "an apple" or "a ?apple".
NLD

>Date: Wed, 13 Nov 91 09:47:32 EST
>From: Geoffrey Russom <EL403015brownvm.brown.edu>
>Subject: A Apple
>
>It seems to me that if the vowel in "a, the" is schwa, only a glottal
>stop could prevent hiatus in your cited phrases "a apple, the apple."
>Does anyone know of other possibilities (not necessarily restricted to
>those viable in English)? In general, what SORT of sound is introduced
>to prevent hiatus? One thinks of liquids, nasals, and glides first, but
>of course there's the glottal stop as perhaps the unmarked
>hiatus-preventer....
Lojban prevents hiatus between vowels of the same word using a devoiced
breathy 'glide' (represented by apostrophe) to prevent glottal stop
(represented by period), which is phonemic with pause as forcing a word
break. When both vowels are the same, the effect is an [h] sound, when
they differ, Lojban gains an effective 3 way contrast between diphthongs
/ei/, disyllable /e'i/ and word-break /e.i/. In some words (names and
borrowings), voiced glides are permitted as a fourth contrast.
The contrasts work quite well, but I have to admit that we invented it
out of whole cloth. Is something similar with an /h/-like sound found
in any natural language? natural languages?
----
lojbab = Bob LeChevalier, President, The Logical Language Group, Inc.
 2904 Beau Lane, Fairfax VA 22031-1303 USA
 703-385-0273
 lojbabgrebyn.com