The creation of this pilot experiment now serves as a basis for future comparative studies on
the acquisition of English segments and suprasegments by French learners. Given the
limitations of this study, more elaborate experiments investigating the role of prosody in
intelligibility, foreign-accentedness, and communication in general, are required.
As was reminded several times throughout this work, our focus was production, and more
specifically read production capacities of French EFL learners. That is why the results of the
experiment cannot – must not – be generalized to other capacities such as perception and
spontaneous speech, and the link between prosodic accuracy and overall pronunciation is
still to be analyzed more closely. In order to bring strong evidence of the importance of
suprasegmental features vis-à-vis segmental features, future research must test read speech
and spontaneous speech capacities as well as perceptual capacities of French learners before
and after longer trainings – prosody-based or segment-based. The examination of all these
capacities separately and then together will lead to a safer conclusion as to the importance of
suprasegmental features in L2 acquisition in general. More particularly, testing spontaneous
speech capacities after either a prosodic training or a segmental training will be relevant,
because it is only in natural, extemporaneous conversation that the importance of the
prosodic structure of a language appears (cf. the experiment conducted by Derwing, Munro
& Wiebe, 1998). As regards perception, hypotheses are open as to whether or not a prosodic
training helps French learners improve their perceptual skills more than a segmental training
does. A future experiment should indeed resort to a pre-training evaluation and a posttraining
evaluation of French subjects’ capacities to perceive and understand English, both in
a between-groups design and a within-groups design. Besides, subjects’ overall capacities
should also be checked in the long term, i.e. some time after the post-training tests, so as to
gauge the efficiency of prosodic instructions on French speakers and possible subsequent
application to EFL class situations.
The number of participants in such a comparative study of L2 segments and prosody is
significant. Even though ten French speakers and three listeners were used for this pilot
experiment, more subjects and judges should be involved in the future. Concerning
production, having a control group of native English speakers do the recordings could be
interesting; they would act as distractors among all the productions, and it would be possible
to see if raters actually give them the highest scores. The reliability of the subjective
evaluations would be tested, as well. Similarly, a control group of French EFL learners
receiving no training at all could increase the validity of the experiment and prove the
efficiency of the trainings. Furthermore, whether subjective evaluations – i.e. listeners’
judgements – are done by ten or twenty native speakers, they should be completed by
objective evaluations of the productions through fine acoustic analyses via professional
software such as Praat.
The link between the findings of comparative experiments on the one hand and the field of
EFL teaching in France on the other, could be strengthened if the students of a whole form
were used as subjects. If feasible, testing the effect of a prosodic training on students who are
at the same academic level – e.g. a whole form of students in last year of general secondary
education (Terminale) – would be most appropriate to investigate the acquisition of English
prosody by typical French EFL learners. Accordingly, the division of a form into three groups
– control, segmental, and prosodic – would be better for subsequent applications of the
findings to the domain of didactics, all the more as a typical form consists of students with
various personal capacities of acquiring a language. The two trainings could take place over
several sessions, similarly to Derwing, Munro and Wiebe’s (1998) experiment, so that the
constraints of the durations of the trainings would not come up, and the resemblance with a
typical pronunciation lesson in an actual EFL class situation would be kept.
Thus, the pilot experiment that we have carried out opens the way to many other
possibilities to analyze the contributions of L2 segmentals and suprasegmentals to
intelligibility, foreign-accentedness, and the welfare of communication. Despite the
pioneering results of this study, the role and importance of suprasegmentals in the
acquisition of English phonology are not lessened, and they still need to be investigated in
the future.