"Connecticut accent" in the Times

[Alice Faber]

Arnold M. Zwicky said:
>On Sep 9, 2004, at 2:10 PM, Robert Wachal wrote:
>
>>  I believe that most of us have a glottally released /t/ rather than a
>>  glottal stop in the cases cited below.
>
>i most emphatically do not have a glottally released /t/ in "Britain",
>"button", etc.  but it only now (after decades) occurs to me that some
>people might actually have a glottally released /t/, which i've just
>been hearing as my [?]-released-into-syllabic-[n].  (to confound things
>further, Wells, Accents of English, reports on a british "glottal
>reinforcement", or *pre*glottalization, in various sets of words,
>including the "button" type: [?t].)
>
>i see now that Kreidler, the Pronunciation of English (1989), is cagey
>about these allophones.  p. 112: "The realization of /t/ may be a
>glottal stop [?] or a simultaneous -- or nearly simultaneous -- double
>closure, the vocal cords coming together while the apex makes contact
>with the alveolar ridge, [?t] or [t?]."  the environment before
>syllabic [n] is one of the relevant environments.  (kreidler recognizes
>that different dialects have different realizations in different
>environments.)
>
>so there are several dimensions of possible variation here -- the
>nature of the stop gesture(s), and also (as alice faber suggests) the
>nature of the following syllabic.
>
>facts, we need facts.

And facts are hard to come by, at least facts that can be arranged in
a coherent picture.

When I was researching the topic (for an LSA paper a number of years
ago), a number of things became clear:

(1) There are several prosodically-different phenomena in English
that might be labeled "glottalization" or "glottal replacement".

(2) There are gross regional differences, so that what's described
for various varieties of British English is very different from
what's described for American English. One such difference is whether
whatever-it-is happens to all syllable-final voiceless stops,
regardless of place, or whether it only happens to /t/.

(3) The specific allophones described by different observers differ,
so that it's hard to know whether [?t] observed in Tyneside English
is at all like [?t] observed in New England. (After an oral
presentation at NYU, one attendee assured me that they're not.)

(4) The prosodic and social conditions in which whatever-it-is occurs
differ across dialects.

(5) Now what I've observed in the northeastern US:
        a. whatever-it-is affects predominantly but not exclusively /t/
        b. at least for some speakers, it's fed by syllable-final
devoicing, so I have notes from a Knicks basketball broadcast a
number of years ago in which the name (Dennis) Rodman was regularly
pronounced [ra?(t)mn]
        c. whatever-it-is is favored in thematic/new/contrastive
contexts (as opposed to old information); from the same broadcast,
Nets was [nE?(t)s] in scores, e.g. Nets 31 Knicks 27 (in an era when
the Knicks were good and the Nets weren't, so such a score was an
upset-in-the-making). And, for a non-alveolar, Buck Williams was
[bV?k] (V=wedge), in contrast to HERB Williams.
        d. whatever-it-is is favored when the following syllable
begins with a resonant
        e. there are huge differences among speakers in their overall
rate of glottalization

At the same LSA meeting at which I presented my data, Janet
Pierrehumbert described an ingenious study looking at oral-laryngeal
coordination in instances such as "oatmeal" contrasted with, on the
one hand "oatcakes" and on the other hand "cornmeal" (that is,
looking at the effects of points c and d above) that was utterly
foiled by the effects of point e. In my study, the fixed and boring
elicitation phrase "Say ___ again" provided most favored contexts for
whatever-it-is, yet, in accord with point e, 2 out of 5 subjects
produced appreciable numbers of glottal-type tokens.

I have *got* to get back to working on this.

--
 =============================================================================
Alice Faber                                             faber at haskins.yale.edu
Haskins Laboratories                                  tel: (203) 865-6163 x258
New Haven, CT 06511 USA                                     fax (203) 865-8963