Re: Why does this vowel change its length?

From: Jonathan Jordan (jonathan.jordan_at_sheffield.ac.uk)
Date: 08/18/04 

Date: Wed, 18 Aug 2004 18:09:03 +0100

"Arnold Zwicky" <zwicky@Turing.Stanford.EDU> wrote in message
news:cfu5tc$a66$1@news.Stanford.EDU...
> in article <ed2a6a11.0408151953.497a071e@posting.google.com>,
> A Gwilliam <devnull@southernskies.co.uk> wrote:
>
> >I've noticed that, at least for me, the vowel /&/ is long in some
> >words but short in others. I even have a minimal pair, although
it
> >seems unlikely to me that this offers proof of two separate
phonemes:
>
> >Short: bad[e] (verbal form), cad, dad, fad, had, pad, rad, tad
>
> >Long: bad (not good), lad, mad, sad
>
> >For what it's worth, "badly", "madly", and "sadly" all have long
> >vowels also, but "laddish" has a short one.
>
> >Do other people make this distinction? And can anyone see any
sort
> >of a pattern here?
>
> i don't recall having seen this particular pattern before, but you
> could see if it's mentioned in the British Isles volume of John
> Wells's _Accents of English_.

Something similar is mentioned, on p288, as a possible feature of some
RP and "near-RP".

> what it does remind me of is the split of /&/ into what charles
> ferguson labels /&/ and /&:/ (well, his symbols are ash and ash with
a
> macron over it, which i've asciied here) along the eastern seaboard
> of the u.s., from boston to washington. the reference is:
>
> Ferguson, Charles A. 1975. 'Short a' in Philadelphia English.
> M. Estellie Smith (ed.), Studies in linguistics in honor of George
> L. Trager (The Hague: Mouton), 259-74.
>
> (this follows up on a 1930 article by trager on this very
phenomenon.)
>
> the phonetic realization of the distinction varies from place to
place
> (and time to time and social group to social group), and the sets of
> affected words also vary.
>
> the core of the affected vocabulary is words with historical /&/
> followed by syllabic-final /m n/ or /f T s/, which have /&:/; this
is
> essentially the set in which british english has /a:/ rather than
/&/.
> the phenomenon is, from a historical point of view, pretty clearly
> contextual lengthening.

Not all British English, of course. (Speaking as someone from
northern England, where we don't have the long vowel in the likes of
"bath", "grass" and "chance".)

> the interesting stuff is the exceptions to this, and the situation
> before /d/. ferguson reports /&:/ in monosyllabic adjectives ending
> in /d/, except for "sad". (i have /&/ in "glad" as well.)
>
> in any case, the distribution that gwilliam reports is not quite
like
> any i've seen (though i agree that "bade" has /&/ and "bad" /&:/),
but
> it's very close to the u.s. eastern seaboard stuff before /d/: /&/
is
> the default, with /&:/ in most monosyllabic adjectives (which are
also
> adjectives of emotion). ("lad" with a /&:/ is the surprise here.
> maybe "vernacular affectively tinged monosyllables" is the class in
> question.)

The Wells description also focuses on adjectives with following /d/ as
the words that get the longer vowel.

I don't have anything like this, although I'm pretty sure my accent
qualifies for Wells's "near-RP"; for me all the words in question have
a short [a], like the vast majority of words in Wells's TRAP and BATH
classes.

Jonathan