evolved from primate lip smacks and other mandibular cyclicities asso-
ciated with vocal production.Thus far,however,no study has explored
whether nonhuman primates exhibit asymmetries during vocal
articulation.
In humans,Graves and colleagues (Graves,Goodglass,and Landis
1982;Graves and Landis 1985;Graves and Potter 1988;Graves and
Landis 1990;Graves,Strauss,and Wada 1990) demonstrated that during
speech production,the right side of the mouth opens wider than the left.
Moreover,in aphasics with damage to the left hemisphere,bias for the
right side of the mouth is observed for spontaneous speech,repetition,
and word list generation,whereas bias for the left side of the mouth
is observed for serial speech (counting to ten) and singing (familiar
rhymes). This difference suggests that when an automatic motor
sequence is enlisted for vocal production,the right hemisphere is domi-
nant.In contrast,even aphasics show left hemisphere dominance for
nonautomatic vocal articulations,specifically those involving speech
articulation.Studies such as these in humans are now critically needed
for nonhuman primates.Given our increasing knowledge of primate
vocal communication,including its function,acoustic architecture,and
mechanisms underlying its production (Jürgens 1990;Cheney and Sey-
farth 1990;Snowdon 1990;Hauser,Evans,and Marler 1993;Hauser
1993b,Hauser and Schön Ybarra 1994),we are in an excellent position
to examine hemispheric biases underlying the production of species-
typical vocalizations.
In parallel with our analyses of facial expressions (Hauser 1993a),
assessment of articulatory gestures was derived from two measures.
First,for each vocalization (acoustics and visual articulation captured
on video),we scored whether or not one side of the mouth opened or
shut before the other—a timing measure.An articulation was scored as
asymmetric if one side of the mouth started or ended the articulation at
least one frame earlier than the other side.Second,we scored,frame by
frame,which side of the mouth was open wider at the start of articula-
tion as well as the midpoint of the call.Specifically,a frame was digitized
and the mouth divided down the middle,and the number of pixels on
the right and left sides were derived.For both the timing clips and the
digitized frames,half of the exemplars were flipped in the horizontal
plane so that observers were blind with regard to the subject’s original
orientation.The end product of this analysis was an overall assessment
of articulatory asymmetry and its production time course.Below,I focus
on results from the timing measure.
The first set of analyses focused on three call types:coos,screams,and
grunts.Coos and grunts are produced by lip protrusion and an open
mouth,whereas screams are produced by lip retraction.Results (figure96 Marc D.Hauser