8.1176, Disc: British <a>

[The LINGUIST List]

LINGUIST List:  Vol-8-1176. Thu Aug 14 1997. ISSN: 1068-4875.

Subject: 8.1176, Disc: British <a>

Moderators: Anthony Rodrigues Aristar: Texas A&M U. <aristar at linguistlist.org>
            Helen Dry: Eastern Michigan U. <hdry at linguistlist.org>
            T. Daniel Seely: Eastern Michigan U. <seely at linguistlist.org>

Review Editor:     Andrew Carnie <carnie at linguistlist.org>

Associate Editors: Ljuba Veselinova <ljuba at linguistlist.org>
                   Ann Dizdar <ann at linguistlist.org>
Assistant Editor:  Martin Jacobsen <marty at linguistlist.org>

Software development: John H. Remmers <remmers at emunix.emich.edu>
                      Zhiping Zheng <zzheng at online.emich.edu>

Home Page:  http://linguistlist.org/


Editor for this issue: Martin Jacobsen <marty at linguistlist.org>

=================================Directory=================================

1)
Date:  Tue, 12 Aug 1997 15:26:57 +0000 (GMT)
From:  "James L. Fidelholtz" <jfidel at cen.buap.mx>
Subject:  Re: 8.1173, Disc: British <a>

2)
Date:  Tue, 12 Aug 1997 23:10:20 -0700 (PDT)
From:  bwald at HUMnet.UCLA.EDU (benji wald)
Subject:  Re: 8.1173, Disc: British <a>

-------------------------------- Message 1 -------------------------------

Date:  Tue, 12 Aug 1997 15:26:57 +0000 (GMT)
From:  "James L. Fidelholtz" <jfidel at cen.buap.mx>
Subject:  Re: 8.1173, Disc: British <a>

Dear LINGUISTs:

	Some of the discussion so far has confused two separate
issues: (1) the historical development within English of [ae] into
[a]; and (2) the greater and more rapid tendency of British English in
general to 'anglicize' the pronunciation, especially of orthographic
<a>.  The reality of the second phenomenon was impressed upon me at
the beginning of the Sandinista revolution (middle late 70s), when I
was in Europe listening to the BBC.  The announcer reported on the
happenings in 'Man[agw@], Nicar[aegyuw@]', which I ascribed to
'Nicaragua' being a known country name, and therefore it was
anglicized, while 'Managua' was (at that time) at least relatively
unknown, and therefore still pronounced 'foreignly'.  In contrast,
American news broadcasters (or me either) would never anglicize
'Nicaragua', always giving it the [a] pronunciation (but with the
appropriate shwas in the unstressed syllables).

	Jim

James L. Fidelholtz           e-mail: jfidel at cen.buap.mx
A'rea de Ciencias del Lenguaje
Instituto de Ciencias Sociales y Humanidades
Beneme'rita Universidad Auto'noma de Puebla, Me'xico


-------------------------------- Message 2 -------------------------------

Date:  Tue, 12 Aug 1997 23:10:20 -0700 (PDT)
From:  bwald at HUMnet.UCLA.EDU (benji wald)
Subject:  Re: 8.1173, Disc: British <a>

I have followed with increasing interest the discussion of
nativisation of names (proper nouns from non-English sources) and of
English <a> as a historical entity.

I got interested enough to think that there are two sets of issues
here that have a likely but problematic relation to each other.  ONE
is that as a historical entity English <a> has some interesting and
even unique properties in comparison with other English vowels --
certainly for relevant British and American dialects, which is most of
the American ones but only the Southern British ones, including the RP
"standard".  TWO is that the unique nature of this historical entity
is most visible in how new loans from other languages are adopted into
English -- or even repeated as foreign words.  And TWO is most evident
with proper names like Jacques Chirac, which may even have a short
life span in spoken English, as the entities referred to recede from
public interest and memory.  But NOTE, JAcques is a long-lived and
stable word across dialects, ChirAc is not.  Perhaps surprisingly in
the American context, the "a" in Viet NAm almost immediately spawned
variant pronunciations, one with "front" or "back" "a", and it has
remained that way ever since.  It is remiinicent of such older words
as "pajama/s" ("pyjama/s" for you Britishers), where some dialects
have "front" and others "back" "a", and a few more like it.

Evidentally, the concern in the discussion is that there should be
some predictability, based on an understanding of the English vowel
system, about how an unrounded low vowel from another language would
be adopted as an English word.  Yet we find in the cases under
discussion that there is no precise predictability to whether such a
word will be adopted with a "front" or "back" low vowel, or how it
will stabilise over time (if it does).  This leads to consideration of
how the English low vowel (or is there more than one?) fits into the
English phonological systems of the various dialects (among other
issues I will mention).

In his 12 Aug message Roger Lass noted the following on the difference
in treatment of <a> as a historical entity between relevant current
dialects of English:

Lengthening [of <a>] increases during the 17th-early 18th century, and
by around the 1740s (e.g. in Mather Flint's Prononciation de la
language Angloise, 1740) there is variable lowering of the lengthened
[ae:] to [a:], partly lexically determined. This continues throughout
the 18th

I would prefer to speak of variable BACKING rather than "lowering"
(even though "lowering" might be phonetically correct for the phase
Roger is referring to) to call attention to the basic problem of <a>
throughout the history of English as one of whether it, or some of its
allophones, will go "front" or "back" (in some periods back enough to
become rounded and merge with "open o", short as in "got" or long as
in "caught", "talk", "drawer").


In the case involving recent and current nativisation of words having
in the source language a phonetically low usually mid, occasionally
back, but hardly ever FRONT low vowel, the choices remain LOW.
However, as we recognise that whatever the LOWEST vowel in English has
been throughout its history it has constantly spawned BACK AND FRONT
allophones (sometimes merging with other phonemes), we note further
polarisation of FRONT or BACK in pronunciation even as the spelling
remains.  Thus, "day" and such words have a FRONT mid vowel in most
dialects, while "talk" and "caught" have vowels so BACK that they are
even rounded in many dialects, certainly the British ones (and
stigmatised in NYC among some other East Coast dialects in a HIGH BACK
position).  As the set "dray", "drag", "draw" illustrates, the same
historical <a> vowel (in earliest English) has assumed a variety of
phonetic positions, even within individual dialects.

In the history of English, but esp in the so-called modern period,
many splits in <a> have occurred, producing back variants alongside
the more conservative front ones, some with relatively reliable
phonological conditioning, e.g., the split that allows only BACK
variant before "r" in CLOSED syllables, as in "part" (originally a
French word), vs. only FRONT variants before "r" in OPEN syllables, as
in "carry" (except following w-, as in "warrant", which has variant
BACK pronunciations, rounded and unrounded), OR the one which
distinguished the BACK vowel of "talk", "caught", etc.  from the
(originally) FRONT one of "half", "glass", "bath", etc.  Whatever the
phonetic realisation of these two classes, they remain distinct in all
American and British dialects.  In some US dialects, both of these
word classes are also distinct from the one with "palm", "calm", etc.,
where dialects vary in the phonetic realisation of the latter class,
and even as to whether or not the "l" is pronounced.  Quite precise
and intricate phonetic conditioning is involved in these cases.
Similarly, for the split following /w-/ "wax", "wag", "swam" etc
VS. "wad", "swamp", "swat", "squash" etc (vs. "water", "war",
"warrior")

The lexical determination Roger refers to implies that the process, as
it affected the difference between American and S British dialects, is
not one of "classic" sound change, but of lexical diffusion of the
sound change LOW FRONT > LOW BACK.  It starts in one phonetic context
and proceeds to others -- but NOT on a phonetic basis alone.
Meanwhile, new words come into the language, and if they have a low
vowel (usually SPELLED "a" in English in such words) they get caught
up in the lexical diffusion process, but the principles by which they
do are not well understood.

For all affected dialects it is not clear why the low vowel in
non-English words should be adopted into English as FRONT rather than
BACK variants, IF the vowel is BACK in the source language, AND IF the
variety of English already has a LOW BACK vowel.  And they all do, at
least in the word "father" (vs. "lather" with a front vowel), and also
"swap", "swat", etc.  It is certainly reasonable to assume (more than
in many other cases where the same device is attempted) that the
SPELLING of the adopted word with "a" (as in the source language) is
the source of interpretation of the vowel as FRONT in English.  This
presupposes that such pronunciations arise NOT from hearing how the
word is pronounced in the source language, but from extending the
front pronunciation from familiar words spelled with "a" to unfamiliar
words.

There remains the problem of noting that "a" as a spelling also has a
BACK pronunciation, as the result of historical processes such as
those mentioned above.  In that case it still remains to be seen
whether the notion that "a" with a FRONT vowel is an "unmarked"
spelling pronunciation, such that English speakers have a sense that
words spelled "a" but pronounced with a back vowel are the "marked"
case, subject to certain phonological conditions that do not
automatically apply to unfamiliar words, even if there is a suspicion
on the part of the reader that they represent another language.  AND
there still remains the problem of determining why with some
unfamiliar words a BACK vowel survives, and in others a FRONT vowel,
e.g., "JAcques" vs. "ChirAc" or "Viet NAm (for those with the
problematic pronunciation).

The solution is complicated.  After all, "Jacques" with a BACK vowel
is also "Jack" with a front vowel.  No doubt the preservation of an
initial sound "zh" in "Jacques" is involved with the fidelity to the
stressed vowel as well, as in the case of such words as "garage" and
"barrage", etc. for American speakers, where final "zh" is not more
characteristic of the pronunciation of Germanic words than is initial
"zh".  S Brit "garage" with shift of stress to the first vowel falls
in with the predictability of a FRONT (rather than BACK) lowEST vowel
before /r/ in open syllables, cf.  "marry" (EXCEPT following /w-/ as
in "quarrel" if this does not merge with short "o" as in S Brit
"sorry").  For S Brit "garage" no non-phonological explanation is
necessary, and change of (French) final /zh/ to /j/ (so that "garage"
has the same final unstressed syllable as "storAge", "porridge" etc.)
is irrelevant to the realisation of the French stressed low back
vowel.

In S Brit and American English words with stressed LOW BACK vowel
sources, the stressed vowel seems to be unstable and tending to FRONT
in the long run.  For whatever reason, a following /m/ seems to afford
to the most resistance to a FRONT pronunciation, e.g., pajama,
(l)lama, pastrami, salami.  But maybe there is a correlation between
dialects that keep the back vowels in words like these and dialects
that have the same vowel in the historical word class "calm", "palm"
etc.  My NYC dialect has the same vowel in both sets of words, and
also contrasts "calmer" with "comma", where "comma" is FRONTER (in
central low position) than "calmer", BUT I have no CLOSED syllable
words with the vowel of "comma", e.g., "bomb" has the vowel of "calm"
NOT "comma".  In the meantime I split the "-rama" words, so that
"cinerAma" (now extinct) has the stressed BACK vowel of "comma" (NOT
"calmer") but "panorama" (variably) has the stressed FRONT vowel of
"hammer".  This complexity has to do with whether short "o" also gets
involved.  It does not in Southern British dialects (except following
/w-/), since the short "o", as in "bomb", "pot", "robin", "sorry" etc.
remains rounded, but it does in many American dialects where short "o"
is LOWERED and UNROUNDED in such words.

With regard to lexical diffusion, the broad "a" sound, comprising a
sub-set of historical <a> words, is a feature of Southern British (and
descendant) dialects which is very striking to all Americans,
including Bostonians, who have a similar set of the broad "a" words,
but pronounce them as FRONTED, not BACKED.  Yet at least one <a> word
is "broad a" in ALL American dialects, FATHER -- in contrast to
LATHER.  LATHER remains front, and in most American dialects has the
same phonetic vowel as "hat" etc. ( adjusting for whatever PHONETIC
influence the following fricative has on the quality or length of its
stressed "a").  Lexical diffusion was already operating with <a> words
BEFORE British and American dialects became differentiated with
respect to <a>.

To appreciate the influence of phonetic context on variation in <a>
words, consider in particular the effect of a following /l/, e.g., the
variant pronunciations of "falcon" and "almond", either with the vowel
of "talk/caught" or the vowel of "scaLp" or "half", i.e., with a FRONT
or BACK vowel.  Also, note in "falcon" but not "almond" the BACK vowel
may round (whether the "l" is still pronounced or not is a separate
matter, "a" is the spelling for ALL <a> words in English, even
"pajama", "pastrami", etc.).

I wonder how stable an English word like "safari" is with a back
vowel.  I already noted that <a> before /r/ in an open syllable
historically shifts to FRONT.  However, that is only true in the
absence of a morpheme boundary.  Thus, "starry" like "star" has a BACK
vowel, and in most dialects of US English (but not Brit or Canadian
inter alia) rhymes with "sorry".  Is "safari" stable because it rhymes
with "starry"? -- even though there is no morpheme boundary following
the /r/ -- OR can we expect dialects in which it has a front vowel,
like "carry"?  Does it matter if "sorry" rhymes with "starry", in
favoring maintenance of the BACK pronunciation, since in such dialects
"sorry", or at least "borrow" have no morpheme boundary but the same
vowel as "safari"?  If the origin of the fronting is SPELLING with
"a", it shouldn't matter.  So there may be an empirical issue here --
which is always a good thing to have.  (Of course, "starry" is stable,
since there is an association with "star", i.e., a morpheme boundary
-  and we would expect that if the comparative of "far" was
regularised to "far-er" etc, it would also maintain a BACK
pronunciation.)  And does anybody pronounce "sonAta", "cantAta",
"toccAta" with a front vowel?  Why (not)?  -- Benji

---------------------------------------------------------------------------
LINGUIST List: Vol-8-1176