The major claim of this work was that the importance of prosody in the acquisition of
English as a Foreign Language by French learners is as strong as, indeed stronger than, the
importance of segmental features such as phonemes. By extension, our view was that
intelligibility and the welfare of communication greatly depended on the realization – and
thereby perception – of L2 suprasegmental features. In the scope of this experimental
research, only read speech as produced by French EFL learners was tested, and future
studies are still needed.
The first hypothesis that we put forward as regards the results of the experiment was that
the French speakers who received a training on L2 prosodic features would be better
evaluated by the native English speakers and expert than the group with segmental focus
(Hypothesis 1). As a matter of fact, the hypothesis was not confirmed. The two groups turned
out to have obtained very similar results: Group A has obtained a mean score of 3.63 out of 7,
and Group B has obtained 3.64. The extreme closeness between these two scores is quite
telling. As far as read speech capacities are concerned, a prosodic training does not
necessarily help French learners improve their pronunciation and accent more than a
segmental training does. Judging by the three listeners’ scores of the post-training
productions, intelligibility does not depend on segmentals only, nor does it depend on
suprasegmentals only: (British speaker:) 3.37 vs. 3.3; (American speaker:) 4.29 vs. 4.24;
(expert:) 3.24 vs. 3.37. Consequently, it seems that segments and prosody are of equal
importance, and both aspects of English phonology should be taken into account in EFL
pronunciation teaching.
While we hypothesized that Group A might be better at word reading after the segmental
training, and Group B would be better at phrase reading after the prosodic training
(Hypothesis 2), the analysis of the mean scores is not very conclusive. Yet, one can indeed
observe a slight advantage of Group A in words (3.68 vs. 3.6), and a slight advantage of
Group B in phrases (3.58 vs. 3.67), somehow confirming Hypothesis 2. However, the gaps
between the two mean scores are so small – i.e. 0.08 for words and 0.09 for phrases – that it is
wiser to conclude that the two groups are at the same level for both word reading and phrase
reading.
Despite the unexpected results of the experiment, it is interesting to note the spontaneous
reactions and perplexity of two of the three listener-judges during the rating task(1). Both
Judge 2 (the American speaker) and Judge 3 (the expert in English phonology) pointed out
that a wrong realization of some prosodic patterns prevented them from giving a better score
to some of the productions. The perfect pronunciation of segments in these specific cases was
not enough to get full score. Judge 2’s comment concerned the bad rhythm of a sentence,
despite the perfect realization of the phonemes. Judge 3’s was about the stress pattern of a
disyllabic word, and the rising intonation of an isolated word – the expert in English
phonology remarked that a rising tone was not the default tone for such an isolated word,
and a falling tone would have made the score higher. These comments made by two of the
three listeners seem to highlight the importance of prosody in spoken English, even though
the overall results of the experiment do not confirm that idea.
Our findings are at odds with the previous comparative experiments that investigated the
effect of prosody with respect to that of segments. Both the experiments carried out by
Derwing, Munro and Wiebe (1998) on the one hand, and Missaglia (1999) on the other,
revealed that a prosody-centred training has a better effect on L2 learners’ skills than a
segment-based training, whatever the L1 of the subjects. This discrepancy between their
findings and ours may be explained by the fact that our experiment only investigated read
speech, and the number of French-speaking subjects and listener-judges was rather limited.
Also, many other extra-linguistic factors played a role in the subjects’ productions, such as
their being tired or bored. More comparative experiments on the role of prosody are
necessary, as the importance of the latter in intelligibility is still to be objectively proved. The
conclusion of the between-groups design is only a first step to a closer examination of
prosodic features in the production and perception of English as a Foreign Language by
French speakers.
1 It must be borne in mind that the rating task was individual. The three raters never met to carry out
the task, and they were not aware of the aim of the experiment, either.