Throughout this work, the main objective has been to enhance the role of prosody in L2
phonology acquisition, and its contribution to intelligibility. In the theoretical section
(Chapter 1), we saw that, interestingly enough, French teachers of English as a Foreign
Language tend to focus on phonemes when it comes to pronunciation teaching, thereby
overlooking prosodic features in spite of the great differences between the French system
and the English system. For that matter, when one talks about the pronunciation of a foreign
language, what immediately comes to one’s mind is L2 phonemes and their differences with
those of the L1. Yet, as was seen in Chapter 2, the importance of prosody in both first
language and second/foreign language acquisition is acknowledged by many researchers,
and sometimes teachers even (Burgess & Spencer, 2000), and it is hardly understandable that
learners should not be aware – or made aware – of this aspect of the English phonological
system. French speakers thus do not realize that prosody is essential to communication, since
its function is not the same in their mother tongue and in the target language. Our
hypothesis has been that the suprasegmental features of English actually contribute to
speech intelligibility and foreign-accentedness more than segmental features do, contrary to
what non-native speakers usually believe.
The pilot experiment that we conducted on the production of English by French EFL
learners and the whole protocol that we developed, described in Chapter 3, were initially
designed to allow future comparative experiments on the acquisition of English segments
and suprasegments by French learners, both at the perception and the production levels. The
results given in Chapter 4, however, do not support the hypothesis that a prosody-based
training has more impact on French learners’ production skills, according to the judgements
of native speakers and experts. The between-groups analysis of the post-training scores has
revealed that both a segmental training and suprasegmental training equally help the learner
improve his/her pronunciation, at least as far as read production is concerned. The withingroups
design has shown that a prosodic training does not help learners improve more than
a segmental training. Notwithstanding, one may challenge the very significance and
conclusiveness of the results. The shortness of the trainings, as well as many other external
factors, possibly prevented the French speakers from assimilating the training on L2 prosody
and its oral practice properly. Moreover, the twenty stimuli that were used cannot account
for a whole L2 phonological system, nor can they evince every segmental and prosodic
problem and difference with the source language. While it is necessary to improve teaching
methods and materials, experiments on L2 acquisition also need improving, developing, and
expanding. This pilot experiment is a first step towards more elaborate experiments.
In order to examine carefully and objectively the contribution of prosody to intelligibility
and foreign-accentedness with respect to that of segments, and to address the issue of the
link between prosodic accuracy and segmental accuracy as put forward by Birdsong (2003),
the ideal method would be to compare French EFL learners’ productions of perfect prosody
on the one hand, with perfect productions of segments only, on the other. However, this
utopian design would be of no avail or interest. Discovering the importance of prosody in L2
phonology acquisition is supposed to contribute to the field of EFL teaching and didactics,
and actual French learners of English, be they in primary school, secondary school, or at
university, all have different levels and capacities of understanding a lesson. It would
consequently be pointless to train some French subjects to produce accurate L2 prosody and
ignore segments altogether, and conversely to train other subjects to produce perfect L2
phonemes and phones only; the findings would never be comparable and applicable to real
situations. Our experiment has served to underline that teaching English prosody to
speakers of such a syllable-timed language as French turns out to be a hard task, even if
much oral practice is done, and future experiments with similar trainings should take that
into consideration.
As is pointed out in Busà (2008: 118): “Because, in speech, segmentals and
suprasegmentals overlap and contribute to each other in many important ways, in
pronunciation classes they should be taught together rather than separately”. The teaching
and learning of the pronunciation of a foreign language equally depends on segments and
suprasegments, despite the central claim of this work. The mean scores that were obtained
by the two groups in our experiment seem to be in keeping with that, as the segmental group
and the suprasegmental group proved to be at the same level after their respective trainings.
Nonetheless, future experiments should provide more evidence of the roles of prosody and
segments in communication.
Unavoidable extra-linguistic factors constitute limitations to any experimental study. As
far as this study is concerned, one can mention the subjects’ and the listeners’ personal
timetable constraints, and the restricted durations of the trainings, which prevented the
subjects from having enough time to practise and assimilate the training correctly. As a
conclusion, more research should be done in the field of L2 acquisition of suprasegments and
segments by French EFL learners, so that future findings can be broadened and applied to
the field of teaching and didactics.