The Situated Multimodal Facets of Human Communication - Coverbal Synchrony in Human-Machine Interaction - page 188

The audio/video analysis of the recorded data proved that
synchrony between holds and speech pauses exists independently
of the tasks and the language, with variations that can typically be
attributed to the language and cultural specificity. American English
speakers equally distribute the time for holds and speech pauses
over their fluent speech duration, while holds in Italian speakers are
in general shorter than speech pauses. Also, American English holds
overlap in similar amounts both with empty and filled pauses, whereas
the percentage of Italian holds overlapping with empty pauses is
significantly lower than that overlapping with filled pauses and Italian
speakers pause less than Americans (Esposito et al., 2002).
Further support to the above data was found in Italian narrative
discourse data from children (9 years old ± 3 months females) and
adults (two males and two females, average age 28 years old ± 3
years) who participated in a similar type of elicitation (Esposito and
Marinaro, 2007). Both adults and children were native speakers of
Italian. The two groups of speakers produced a similar distribution
of hold and speech pause overlaps. An ANOVA performed on the
data collected from each group using hold and speech pause rates
as within subject variables showed that children's holds and speech
pauses were equally distributed along their narrations ( F 1,10 = 1.09,
ρ = 0.32) while adults used holds significantly more frequently ( F 1,6 =
11.38, ρ = 0.01) than speech pauses. Rates were computed as the ratios
between the number of holds and/or speech pauses over the length
of the subject's narrations measured in seconds.
The differences between hold and speech pause rates found
in adults and not in children may be explained by the fact that
sophisticated utterances are a prerogative of the adult communication
behavior. Adults exploit holds also as a strategy for structuring the
discourse in absence of speech pauses, whereas children, being less
skilled in assembling bodily and verbal information, tend to attribute
the same functions to both holds and speech pauses.
Does this synchrony develop during language acquisition? In order
to answer this question, a further study was conducted with three new
groups of children (Esposito and Esposito, 2011):
- 8 females, of 9 ± 3 months years old;
- 5 males, and 5 females of 5 ± 3 months years old;
- 3 males, and 3 females of 3 ± 3 months years old.
As in the abovementioned studies (Section 3), the children told
the story of the “Sylvester and Tweety” cartoon and their complete
narrations were analyzed. The lengths of the narrations were different