Graphics Reference
In-Depth Information
used by the speaker. It has been shown more recently that co-speech
gestures also participate in the conveying of semantic information, and
that they play a role in the organization of discourse by speakers. At
last, much like what happens in the verbal and vocal modalities, they
reveal something of the interpersonal relationship between participants
to an interaction. It would, however, be simplistic to suggest that
in any utterance, exactly the same information is conveyed in the
three modalities at the same time. It cannot be expected therefore
that when information is repeated by a participant to an interaction,
all of the information will be copied. Rather, the participant is more
liable to copy different pieces of the message: part of what was said
(semantic information) and/or part of its format (prosodic and gestural
information). And since the main role of repetition, as seen in previous
studies described in Section 3 of this chapter, is to help participants to
an interaction achieve some sort of convergence, it is to be expected
that depending on the amount of information repeated by a participant,
the degree of convergence will be lower or higher.
In order to test this, we analyzed some examples with a focus
on gesture repetition. The examples were drawn from the Corpus of
Interactional Data (CID) recorded at Aix en Provence. It comprises a
series of video recordings of unprepared dialogs in French which were
transcribed and annotated in several linguistic domains, including
gesture for part of the corpus.
The examples confirmed results from previous studies showing
that gesture repetition does not have to be strictly identical to be
considered as repetition and that it is rather what makes the semantics
of the gesture (namely the type and direction of movement, general
hand shape) which has to be copied, whereas other features are not
strictly necessary in the repeat (gesture speed or gesture space for
instance). These may be considered as variable features of the gesture.
It became apparent as well that although the copy goes towards a
reduction of the model in most cases, it sometimes happens that the
copy is an improved version of the model, both in terms of length and
structure. Gesture repetition may be used to accompany a confirmation
request on the part of one of the participants, and therefore as a means
to achieving a convergence which is not yet there. In some cases, when
the speaker repeats a gesture whereas the prosodic pattern and the
verbal message are in contradiction with what was said by the other
participant, the gesture repetition may be seen as a means to fake
convergence. This reveals how important convergence is to participants
in an interaction. It also reveals that, although co-speech gesture
is sometimes considered as forming a single idea unit with speech
(McNeill, 1992), there must be some kind in independence between
Search WWH ::
Custom Search