back to the dialects of NJ

[Arnold M. Zwicky]

On Oct 20, 2004, at 5:49 PM, Dale Coye wrote:

> To be more specific about what I was referring to in the Trenton
> area--I have only my impressions from asking students who do this
> where they're from--in words like "written, Trenton" is a "real"
> glottal stop--like the Cockney in "letter" with no tongue contact
> anywhere.  The more usual US reflex of this /t/ before syllabic /n/
> has a nasal release, with the tongue tip in its usual alveolar
> position--someone please correct me if I'm wrong.  There may be an
> accompanying glottal stop.

at the risk of re-starting the Great Connecticut Glottal Perplexity,
the plain glottal stop (released into the syllabic /n/)  is the
pronunciation that was reported there, including the allusion to
cockney.  it is also, as i pointed out then, the pronunciation *i*
have, which has never been commented on (in the u.s.).

please, please, don't post to say that you have a glottal(ized)
something or other that isn't just a glottal stop; i understand (now)
that lots of americans, maybe most of them, have such pronunciations.

there is still a perceptual puzzle here, though.  the acoustic
differences between these various productions (preglottalized alveolar,
glottalized alveolar, postglottalized alveolar, and glottal stop)
before syllabic /n/ are pretty subtle, and i'll bet that most people
have different variants according to speed of speech, casualness, hypo-
or hyper-articulation, prosodic concomitants, etc.  yet there are
groups of listeners who report reliably detecting one of these variants
in the productions of certain other groups; this variant serves, in
fact, as a marker of membership in these other groups.

my call for facts was a call for information about the actual range of
productions of *all* the people involved in these events and for
information about the detection abilities of the listeners.

i asked because things might be as described by the listeners, or
something more complex might be going on.  it's possible that the
listeners are treating *any* occurrence of the glottal-stop variant,
even if such occurrences are infrequent, as diagnostic.  it's even
possible that the listeners are doing this when they themselves have
some glottal-stop variants (but perhaps not as frequently as speakers
in the other groups).  it's even possible that the groups in question
do not differ in their use of this variant, and that the effect is
entirely a matter of selective attention on the part of the listeners.

arnold