ABSTRACT
David McNeill points out that two different viewpoints appear in the gestures people perform during narratives: observer viewpoint and character viewpoint. An observer viewpoint gesture 'excludes the speaker's body from the gesture space and his hands play the part of the character as a whole'. McNeill's research has suggested that character viewpoint gestures are strongly associated with verbs that take a grammatical object. The authors uses eight broad semantic categories to break the meaning down into its parts, namely identity, description of action, shape, size, movement, direction, speed and relative position, to determine what individual types of information the participants received from gestures. Identity semantic category reflects whether or not the participant correctly specified the main entity associated with the iconic gesture. Movement category reflects whether or not the participant correctly specified whether the main entity was moving or not, it was therefore a straightforward dichotomous category.
