Gagne de la cryptomonnaie GRATUITE en 5 clics et aide institut numérique à propager la connaissance universitaire >> CLIQUEZ ICI <<

1.2.1. Segmental difficulties and recurrent errors

Non classé

As is specified in Avery and Ehrlich (1992), many English words were borrowed from French
after the Norman Conquest. Still today, the two languages share many vocabulary items, at
least orthographically. As regards pronunciation, the difficulty encountered by French
speakers is noticeable, and it partly originates in too great an influence of spelling (Burgess &
Spencer, 2000). This idea is confirmed by Hodges’ (2006: 4) statement: “French EFL students
of novice proficiency often see words with the same spelling in their native language and
assume that the pronunciation, stresses and even meaning are the same”. Segmentals thus
have an obviously important role when it comes to EFL pronunciation teaching, all the more
as very few common features are to be found between the French and the English phonetic
realizations, even of the most used phonemes like /t/, /l/, and /e/ (Birdsong, 2003). Although
teachers emphasize the productions of vowels more than consonants – possibly because the
former constitute the nucleus of a syllable –, both have equally visible differences with the
French sounds, and the errors made by learners are as significant. Hodges (2006) goes even
further and affirms that the problem French speakers have with the pronunciation of English
sounds is such that many French-speaking English teachers never acquire some typically
English phonemes. Consequently, they transmit incorrect pronunciation to their students.
In this subsection, the only segmental errors by French EFL learners that are considered as
relevant are those that might lead to unintelligibility. After a descriptive account of learners’
difficulties and production errors at the level of consonants and vowels, the problem of
minimal pairs, i.e. where clashes occur the most because of segmental errors, is worth being
discussed, since it is one of the few consequences of the errors, if not the main one.


At the phonemic level, English and French consonants do not seem to differ very much, and
misproductions do not overly affect communication. That may explain why EFL teachers
prefer to focus on English vowel sounds, which have more evident differences with French.
Phonetically, however, English and French consonants are almost systematically articulated
differently even as far as commonly used phonemes are concerned. For example, the
phonetic system of RP (Received Pronunciation) English distinguishes between clear [l],
which occurs in syllable-onsets, and dark or velarized [ɫ], which occurs in syllable-codas.
Despite the presence of the same phoneme /l/, the French system does not have that
allophonic distinction, so L2 learners only use a clear . The phonemes /t/ and /d/, both
present in the French and English inventories, too, correspond to different phonetic
realizations in the two languages – they are usually alveolar plosives in English, and dental
plosives in French (Birdsong, 2003; Mortreux, 2008). Furthermore, if they are produced as
dental plosives in English, they might be misheard as (/θ/-/ð/), just as they are realized in
Irish English ([t̪]-[d̪]). In the same way, final plosives, or stops, are articulated differently in
English and French, as is pointed out in Flege (1992: 568): “[…] French learners of English
might give greater weight to release burst cues in word-final stops than native speakers of
English because French stops, unlike English final stops, are usually produced with an
audible release burst”. In other words, English final plosives are unreleased, or incomplete,
as opposed to French ones.

All these phonetic differences among common phonemes do not really affect intelligibility
and communication – the allophonic distinction between clear and dark , for one, is
also absent in some varieties of English. Native speakers will therefore be quite tolerant to
such misproduction of phonetic sound qualities (Lemmens, 2010). The production of
normally silent consonants also seem to have little significance – it is not because a French
learner says /ˈwɔːlk/ instead of /ˈwɔːk/ that a native English speaker will be totally at sea. The
context plays a crucial role in such situations. Other difficulties for French EFL learners,
triggering off more significant errors and sometimes grammatical mistakes (e.g. the nonpronunciation
of the third person inflection -s) are noteworthy. One of the few that can be
mentioned is the (non-)realization of the glottal fricative /h/. While this phoneme does not
exist in French apart from interjections, Hodges (2006) notes that French learners frequently
fail to pronounce it where it is present, and yet they have a tendency to place one before a
word beginning with a vowel. As a result, the sentence I’m happy /aɪm ˈhæpi/ regularly
becomes /hajm ˈapi/. Such errors can cause clashes in minimal pairs, e.g. heart vs. art, hair vs.
air, hi vs. I (see below for the importance of minimal pairs in intelligibility). According to
Roach (2009), the lack of aspirated plosives [pʰ], [tʰ], [kʰ] in languages other than English is
another factor that affects intelligibility. In English, syllable-initial voiceless plosives /p/, /t/
and /k/ in stressed position contrast with syllable-initial voiced plosives /b/, /d/ and /g/
through aspiration, and hardly through voicing, which is why the former will be misheard as
the latter if aspiration disappears. The French language having no aspirated plosives, pack
might be understood as back by a native English speaker if the French learner has not been
made aware of that feature of English phonology and keeps an unaspirated /p/.

The /θ/-/ð/ pair is one of the best-known and systematic instances of production difficulty
for French speakers, as well as for many other foreign speakers (O’Connor, 2002). These
dental fricatives do not exist in the French phonemic inventory – with the obvious exception
of people with lisps –, and they are replaced, or “equated” (Flege, 1992), by /s/-/z/, or more
rarely /f/-/v/ (Herry-Bénit, 2011) and /t/-/d/. O’Connor (2002: 5) explains: “[t] is a good
substitute because it preserves the mellowness, or lack of stridency, of /θ/, while [s] preserves
the continuancy of /θ/”. Yet, the /s/-/z/ substitution is the one that should be avoided the
most, because it provokes unequivocal clashes in minimal pairs (thin vs. sin), whereas the
labiodental fricatives /f/-/v/ may still be assimilated with the Cockney English accent, and the
dental plosives /t/-/d/ with Irish English, for example(2). Similarly, the post-alveolar
approximant [ɹ] sounds typically English to French ears, and it can take many years for an
adult French speaker to acquire it (Hodges, 2006: 10). Usually, learners replace it by either
their own [ʁ] or some kind of /w/, so that rain [ˈɹeɪn] might be understood as wane [ˈweɪn].
This leads us to the occurrence of that phoneme and the problem of rhoticity; due to the
influence of spelling, French learners pronounce the English in all contexts, as in many
accents of English for that matter. Notwithstanding, pronunciation teaching is often based on
RP English, a non-rhotic variety of English, and learners get a mixed accent. For instance, the
word better is produced as /ˈbetər/, a mixed RP /ˈbetə/ and General American (GA) /ˈbet̬ər/.
The consequence is a lack of coherence and merely a stronger foreign accent, though, which
does not necessarily affect intelligibility (Nakashima, 2006).

Finally, French EFL learners come across difficulties with the syllabic consonants /l̩/
and /n̩/, all the more as they are the consequence of the typically English rhythm (cf. 1.2.2.
and 1.2.3.). The most common production error is the insertion of a full vowel, very close to
the French phoneme /oe/, and not reduced enough to be identified with the otherwise
correct /ə/. Apart from a stronger foreign accent to the native speaker’s ears, this last point on
syllabic consonants does not actually lead to utter unintelligibility or misunderstanding, but
it gives an insight into the problem of vowel production and reduction.


The English language makes a distinction between “pure vowels”, or monophthongs, on the
one hand (i.e. lax vowels: /æ, e, ɪ, ɒ, ə, ʊ, ʌ/, and tense vowels except diphthongs: /ɑ:, i:, ɜ:,
u:, ɔː/), and diphthongs on the other, which are all tense vowels: /əʊ, ɔɪ, ɪə, ʊə, aʊ, eɪ, eə, aɪ/(3).
As far as production skills only are concerned, diphthongs are not so problematic for EFL
learners, since they are just glides from one vowel sound to another, and misproductions are
mostly due to spelling influence. The real problem concerns pure vowels, i.e. the distinction
between lax vowels and tense vowels that are not diphthongs, commonly (albeit not rightly:
Roach, 2009) referred to as “short” and “long” vowels respectively. On the contrary, French
only has a single type of vowel, and vowel duration is the same for all them. The two figures
below show English and French vowels. The principal characteristics of vowel articulation
are found, such as the places of articulation: backness (front vs. back), and height, or aperture
(close vs. open). Lip roundedness is only relevant in the French vowel diagram because it is
the only distinctive feature of some phonemes, while in English, the front/back distinction
already implies unroundedness/roundedness respectively, except for unrounded /ʌ/
(Deschamps et al., 2004). Thus, where there are two vowels in Figure 2, the one on the left is
unrounded, and the one on the right is rounded. French nasal vowels are not represented:

Figure 1 experimental research into the acquisition of english rhythm and prosody by french learners
Figure 1: English monophthongs, adapted from Roach (2009)

Figure 2 experimental research into the acquisition of english rhythm and prosody by french learners
Figure 2: French oral vowels, adapted from Deschamps et al. (2004)

As is shown by these two figures, despite the frequent use of a unique phoneme in the two
systems, most vowels are articulated differently in English and in French. For example, the
one phoneme /e/ is much closer, or higher, in French than in English, which evinces the fact
that the French word bête is by no means homonymous with the English word bet (not to
mention the phonetic realizations of /b/ and /t/). It is thus acknowledged that it is the
phonetic realizations that are a great source of problem for French EFL learners (Mortreux,
2008), and the latter are not conscious of the differences.

The unavoidable production errors are well-known; the lax/tense distinction is neutralized,
and two English vowels become one French vowel. Mortreux (idem) carried out an analysis
of the recurrent errors made by French learners by transcribing the recordings of two French
students’ productions of English, and using questionnaires to phonetics teachers. To quote
just a few examples, the /æ/-/ɑ:/ pair is replaced by the French phoneme /a/; /ɪ/-/i:/ become /i/;
/ɒ/-/ɔ:/ become /ɔ/ (Herry-Bénit, 2011). As a consequence, one pronunciation has at least two
possible corresponding words, creating confusion in minimal pairs: live and leave are both
pronounced /ˈliv/. Contrary to what French learners usually think, this substitution will be
more likely to be understood as leave – as is shown in Figures 1 and 2, the French /i/ is much
closer to the English /i:/ than to /ɪ/, which itself would be “better substituted” by a French /e/.
In other words, instead of using the unique pronunciation /ˈliv/ for both live and leave, a
substitution of live with French /ˈlev/ would turn out to be better understood by native
speakers of English, provided that the rest of the utterance is grammatically correct. Collins
and Mees (2008) explain that these two English phonemes are heard as if they were
allophones of the one French phoneme by French speakers(4), and that learners must learn to
make contrasts. That may account for the use of minimal pair drills by English teachers in
France. However, as is rightly noted by Brown (1995), the teaching of phonemic contrasts
through minimal pairs has some shortcomings, and intelligibility is not necessarily affected
by such neutralizations as the example of live/leave.

In most, if not all, cases implying a possible confusion in a minimal pair, such as
pronouncing leave instead of live, Brown (idem: 171) notes that the context plays a significant
role, as it enables the hearer to disambiguate the item. That is why he firmly believes that
mispronunciation involving a minimal pair does not lead to unintelligibility. The fact that the
verbs live and leave are followed by different prepositions and often occur in different
grammatical tenses or aspects (I live in London vs. I’m leaving for London) is decisive and seems
to preclude cases of misunderstanding or ambiguity. Similarly, Nakashima (2006) uses the
example of Japanese EFL/ESL speakers, who would most likely substitute the English /r/
with /l/ in the sentence I would like to eat rice. If a Japanese speaker says lice instead of rice, the
utterance is still understandable since Japanese people are not used to eating lice.
Nonetheless, the author also points to the chance that a native English hearer who is not
aware of the Japanese culture, might actually think that it is lice the non-native speaker is
talking about. Lemmens (2010) takes up the live-leave pair, which he asserts can lead to
misunderstanding even though the prepositions are different. If a French speaker says he
“leaves” in London instead of he lives in London, an English speaker might conclude that the
French speaker meant he leaves for London. In other words, it cannot be ascertained that a
native speaker will mentally correct live/leave, but perhaps they will mentally correct the
following preposition. Such an example implies that minimal pairs can lead to
misunderstanding, but also to grammatical mistakes.

The impact of misproduction of English segments on intelligibility and communication
often involves two members of a minimal pair. One of the few other consequences, also true
of suprasegmental misproductions, is foreign-accentedness, which, as was said above, does
not systematically causes unintelligibility. However, this seemingly unimportant detail can
prove to be an impediment to communication. An EFL or ESL speaker’s having a strong
foreign accent might lead to the native English speaker’s simply abandoning communication
by dint of accumulating mental corrections (Lemmens, 2010). Furthermore, it is the cause for
many stereotypes (Mennen, 2006; Vergun, 2006), be they good or bad.

This account of some segmental difficulties for French EFL learners and the impact on
production and intelligibility has deliberately overlooked the widespread problem of the
phoneme called “schwa”. In fact, it corresponds to the phenomenon of vocalic reduction,
itself being a consequence of the rhythm of English and the stressed/unstressed alternation
(Huart, 2002). That is why Brown (1995) classifies vowel reduction and the schwa among
suprasegmental features. The following subsections thereby deal with suprasegmental
difficulties for French learners, including intonation – and particularly tonicity –, stress,
vowel reduction, and rhythm. The stress-timing/syllable-timing typology of languages is also
looked at, for it is a basis for the understanding of the difference between the English and
French prosodic systems.

2 Also note the colloquial use of da for the, even in phrase-initial position.
3 Some phoneticians specify that /.:, i:, .:, u:, ../ are in fact slightly diphthongized, which amounts
4 For a more detailed explanation, see 2.2.1. and e.g. Flege’s Speech Learning Model.

Page suivante : 1.2.2. Suprasegmental difficulties and recurrent errors