* * * * * * * * * * * * * * * * * * * * * * * *
LINGUIST List logo Eastern Michigan University Wayne State University *
* People & Organizations * Jobs * Calls & Conferences * Publications * Language Resources * Text & Computer Tools * Teaching & Learning * Mailing Lists * Search *
* *
LINGUIST List 19.3236

Mon Oct 27 2008

Sum: Intrusive Liquids in English

Editor for this issue: Dan Parker <danlinguistlist.org>


To post to LINGUIST, use our convenient web form at http://linguistlist.org/LL/posttolinguist.html.
Directory
        1.    Katalin Balogné Bérces, Intrusive Liquids in English


Message 1: Intrusive Liquids in English
Date: 27-Oct-2008
From: Katalin Balogné Bérces <bbkatiyahoo.com>
Subject: Intrusive Liquids in English
E-mail this message to a friend

Query for this summary posted in LINGUIST Issue: 19.3043
Dear Linguists,

I got the most comments in connection with Cockney and its pronunciation of
''Paul arrived''. It seems generally agreed upon that it's not possible for
/l/ vocalization (if it applies at all) of ''Paul'' to trigger R-sandhi in
''Paul arrived'', as /l/ does not delete altogether but turns into a /w/ or
/U/ (a (lax) high back (rounded or unrounded) vowel), which does not
trigger linking/intrusive-R. (But see below for the ''tickle him''
example.) In another possible pronunciation the /l/ does not vocalize and
surfaces as a (linking) /l/, in which case, of course, no R-sandhi is
expected again. A. F. Gupta explains this as the result of the fact that in
L-vocalising dialects the phoneme /l/ is still present (although notice
that it is exactly in the intrusive-L dialects I'm asking about that the
underlying status of the /l/ becomes debatable, similarly to that of the
/r/ in intrusive-R accents). Chris Lucas points out that in ''Paul
arrived'' there may be some kind of longer-distance environment effects,
too: the closely following /r/ may also support L-linking. Alex Bellem also
says that this example is perhaps not the best because of the closely
following /r/ in ''arrived'', in addition, in fast / non-careful speech the
first vowel of ''arrived'' would probably elide so that the /w/ at the end
of ''Paul'' and the now adjacent /r/ would be harder to tell apart. She
then refers to accents of British English in which /r/ sounds more like /w/
(e.g. the TV presenter Jonathan Ross, who gets called Jonathan Woss because
his /r/ is fairly close to /w/) -- for such speakers, and certainly in fast
/ very casual speech, syllable-final /r/ and /l/ (vocalized to /w/) are
pretty-well merged anyway.

Unfortunately I didn't receive any data about other accents of English
which are both non-rhotic and L-vocalizing, especially ones where there's
some merger of vowels before syllable-final /l/ and /r/ (like my lame
example with ''Paul'' and ''law'' above), but I suspect that once an accent
is non-rhotic, it develops intrusive-R first and then no intrusive-L is
able to emerge. I think (and this is not a novel idea) that the primary
drive behind the emergence of intrusive liquids in English dialects is
hiatus filling. The choice of the hiatus filler is determined by the first
term of the hiatus, and glides are used to cover the vowel space
accordingly. In all accents of English, the high area of the vowel space is
covered (high front glide /j/, high back glide /w/), and in most non-rhotic
accents /r/ is used as a kind of ''third glide'' to cover the non-high area
(= linking/intrusive-R). In certain (= L-vocalizing) rhotic accents /l/ is
used as the ''third glide'' to cover the non-high area (= linking/intrusive-L).

What happens in accents which are both non-rhotic and L-vocalizing is that
historically, R-dropping (together with Linking/Intrusive-R) precedes
L-vocalization, so the non-high area is already covered by /r/. Word-final
/l/ can NOT be replaced with an /r/ in sandhi in cases like ''Paul
arrived'' (by analogy to ''law and order''), because either the /l/ remains
in the underlying representation and is pronounced as a (clear/light)
Linking-L; or it is lexicalized as a high back offglide in /aw/ (esp. in
broad Cockney), and as such it triggers hiatus filling with /w/.
Up to date, I've found one exception: Christian Uffmann (2008: 8) shows
that in younger SE British English, syllabic /l/ vocalizes as a non-high
(''lax'') /U/ and triggers R-intrusion:
/tsIkUrIm/ 'tickle him'. (Cf. the fact that even in other
clear/dark-L-systems syllabic /l/ is always dark, it never links to the
following morpheme!)

In fact, in connection with R-sandhi, two opposite processes seem to be
present in the accents of English: on the one hand, with the emergence of
''new'' non-high vowels (esp. resulting from the smoothing of the
diphthongs in ''now'', ''fair'', ''rear'', ''pure'') R-sandhi spreads to
new environments; on the other hand, however, in certain accents/registers
R-sandhi is slowly receding: as also remarked by Mark Jones and Philip Carr
in their replies, R-sandhi appears to be disappearing in favour of a
glottal hiatus marker, more and more frequently even linking-R is avoided.

P.S. A piece of data coming from Chris Lucas, which may or may not be
related (you are invited to decide): in London/Cockney, the monosyllabic
short form of the name Sharon (cf. Bill < William) is pronounced
identically to the word ''shall'', ending in the same glide.

Reference:
Uffmann, Christian (2008) 'Incursions of the idiosyncratic' as faithfulness
optimization. Handout of paper presented at the 16th Manchester Phonology
Meeting, 22-24 May 2008.
Linguistic Field(s): Phonology

Read more issues|LINGUIST home page|Top of issue




Please report any bad links or misclassified data

LINGUIST Homepage | Read LINGUIST | Contact us

NSF Logo

While the LINGUIST List makes every effort to ensure the linguistic relevance of sites listed
on its pages, it cannot vouch for their contents.