With the help of kinematic analysis, the temporal organization of the compl
ex daily activity 'drinking from a bottle with a glass' was described in de
tail. The analysis focused on the sequential action structure, the prehensi
le acts, and the bimanual coordination as well as on the effect of differen
t instruction modalities on these parameters to explore the underlying repr
esentation for this complex action. Movements of the two arms were recorded
in three-dimensional space with the help of an optoelectronic device in 12
normal subjects under four conditions: (1) action pantomime after verbal i
nstruction; (2) action imitation after observation of the action performed
by the experimenter without the objects; (3) action pantomime while seeing,
but not touching the objects; and finally (4) action execution with object
s.
Despite high execution variability, the temporal structure of the action co
uld be precisely described by the relative duration and peak velocity of ac
tion segments, by the MGA-object size-correlation, and by linear regression
analysis between the onsets of functionally related action segments. A sim
ilar structure of the action as characterized by these kinematic parameters
was retained across different instruction modalities. Only when the action
was executed with the objects, the interval between the movement onsets of
either hand and the peak velocity of the manipulative acts were reduced, w
hile no change was observed across the other three instruction modalities.
This stability of the temporal structure suggests the existence: of a Level
in the representation of an action where all the modalities converge. (C)
2000 Elsevier Science Ltd. Ail rights reserved.