非表示:
キーワード:
-
要旨:
oint-actions are an integral part of everyday human life. It is often critical
in everyday situations that joint-act
ions are recognized quickly as when one
is driving on a road and has to recognize children playing with a ball on a
sidewalk to detect a possible danger. Surprisingly relatively little is known
about the speed of joint-action recogn
ition. Here we investigated how fast
joint-actions can be recognized on
three levels of detail (detection,
categorization, and identification). We assessed the speed of joint-action
recognition by comparing the speed of j
oint-action recognition with object
recognition, which is known to be fast (e.g. Thorpe et al., 1996). In a series
of experiments we present
ed static images of object
s and joint-actions at
varying presentation times to partici
pants and measured their detection,
categorization, and identification pe
rformance. We find that presentation
times of less than 80 ms allowed joint-action recognition to be highly
accurate (79%) in all three recognition
tasks. Interestingly for some joint-
actions we found identification to be as fast as object identification. Overall
it seems that the speed of detecting
and identifying joint-actions and
objects are comparable. This poses a challenge to the notion that humans
employ time consuming inferential processes in the recognition of joint-
action (“theory of mind”). Moreover we
find that the speed of joint-action
and object categorization differ significantly suggesting that objects and
joint-actions are processed early on
by different perceptual processes.
