LINGUIST List 10.1908

Sat Dec 11 1999

Sum: for Query 10.1767 IPA Handbook/French Section

Editor for this issue: Naomi Ogasawara <>


  1. Neil Coffey, Sum: French section of the Handbook of the IPA

Message 1: Sum: French section of the Handbook of the IPA

Date: Fri, 10 Dec 1999 16:51:58 -0500
From: Neil Coffey <>
Subject: Sum: French section of the Handbook of the IPA

- -----

A couple of weeks ago, I posted a request for feedback on
a number of issues relating to the section on French in the
Handbook of the IPA. Several people have replied both with
feedback and a request for a summary, which I give below.
My thanks to all those who replied, and in particular to
the following for their very detailed responses:

 {a} Roger Billerey <>
 {b} Aurelien Max <>
 {c} Chantal Rittaud-Hutinet <>
 {d} Joaquim de Carvalho <>
 {e} Marc Bavant <>
 {f} Johannes Reese <>
 {g} Douglas Walker <>

In the summary below, letters in braces refer to respondents
who specifically advocated a particular view; note that
their absence does not necessarily mean that they uphold
a contrary view.

- -------

(1) There was large, though (surprisingly?) not unanimous,
 agreement that the nasalised vowel in e.g. 'matin' is
 closer in timbre to [ae~] than [E~] {acdg}, though
 all comments referred to perception/formant structure
 rather than articulation. One younger speaker made
 the point that the tongue seemed _lower_ in the mouth
 for the first vowel of 'intention' than for 'attention'
 {b}; an older speaker made the point that if there
 existed the word 'ettention' (with [E-]), it would
 be closer to 'intention' than 'attention' {e}.

 Two (non-native) respondents wondered if a [a~]/[E~]
 difference in timbre might operate depending on
 morphological position of the vowel {f}.

 One respondent recommended Hansen (1998), referenced
 below {g}. I have it on order, and will summarise
 any relevant information.

 As it stands, then, it seems questionable whether
 the Handbook's claim that "the vowel [E~] is produced
 with tongue and lip position very similar to its oral
 counterpart [E]" is likely to be true for most
 speakers, and even if it were, it seems unclear
 whether this fact is something that represents the
 way speakers perceptualise the two vowels and/or
 something that should be represented in the

(2) All those who commented specifically agreed that
 the transcription of 'abeille' as [abej] is
 probably simply wrong, and that [e] does not
 occur in a closed syllable[1].

(3) All those who spefically commented agreed that
 the second schwa in 'serait regarde' [sR-]
 is (almost) certainly a typo {abcdeg}.

(4) Similarly for the anomolous transcription of
 the [o~] vowel in 'renonca', it was attributed to
 either a typo {acde} or something unknown and
 at best questionable {bg}.

(5) All those who specifically commented could see
 no reason for the lengthened [i] in "ils sont
 tombe's" {bcefg}.

 On the subject of [a], there was more diversity,
 though all agreed that the transcription given
 in the Handbook is probably wrong. Several
 respondents argued for [aa] rather than [a:],
 apparently on perceptual/articulatory rather
 than purely theoretical grounds {bcef}.

- -------------------------

A couple of other issues were raised in passing which
I'd like to mention here and give my opinion on.

Firstly, it has been pointed out that the transcription
given is somewhat inconsistent in its treatment of
enchainement/liaison consonants ("quant ils ont"
[kA~t ilzO~] but "voyageur echauffe'" [vwAja3oe Re]).
Although I would disagree that the consonant should
necessarily be transcribed on the beginning of the
following word (as insisted upon by many textbooks/
teachers), since (a) it is questionable whether the
consonant can be uniquely identified with a single
syllable position at a particular stage in the
derivation; and (b) the transcription is not on the
whole giving information about syllabification in any
case, the difference in transcription does here seem
somewhat anomalous. Particularly as it is specifically
[R] which a priori seems more likely to show some surface
effect on the end of the first word in, say, hesitation
"voyageur... echauffe'".

Secondly, the matter of vowel lengthening before
certain final consonants has been raised. For example,
the final word of the passage transcribed, 'fort',
is given without a length mark on the vowel.
Traditionally, it is assumed that vowels are
lengthened before tone-unit-final closed syllables
whose final consonantal position is filled entirely
by voiced fricative(s) (and that certain vowels are
prone to lengthening in any tone-unit-final closed
syllable). I have personally been unconvinced for
some time that this assumption is actually borne out
by the speech patterns of many (younger?) speakers,
either simply because the lengthening doesn't take
place to the same degree or with the same consistency
as has traditionally been assumed, or else because a
process of enchainement moves the fricative (or final
consonant) from the end of the tone unit to the
beginning of the next (this wouldn't necessarily
vie with the phonological constraint mentioned, but
it would raise questions as to why this enchainement
across tune unit boundaries was taking place).

Any comments on these two issues would be welcomed.

Finally, in light of many of the issues raised in
this discussion, I would be keen to carry out some
informal spectrographic analyses at some point
to try and concretise some of the suggestions/
suspicions. I would be interested in hearing from
anybody who might be able to assist in making
recordings of native speakers. I stress that at
the moment, I envisage this on a fairly informal
basis (e.g. recordings made with a reasonable-quality
sound card and microphone and e-mailed to me as a
WAV file would be fine). If anybody thinks they
might be willing to help out, then perhaps they
could e-mail me. If I can get enough recordings,
then I'll be happy to collate/analyse.

Neil Coffey
10 December 1999

- ---

[1] A possible exception that springs to mind is the
 loanword '(e-)mail', which seems to allow both
 [e] and [E]. It would be interesting to see whether
 speakers admitted [e] in closed syllable as a
 plausible pronunciation of other loans in cases
 where the normal pronunciation avoided it (cf.
 "Game Gear", normally pronounced [gEm3iR] or
 "Game Boy", normally pronounced [gambOj]).

- --------

Hansen, 1998, 'Les voyelles nasales du francais parisien
 moderne', Museum Tusculanum Press, Univ. of

Neil Coffey WWW: Fax: +44 870 0553 662
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue