On the relation between affordance and representation of the agent's effector

The present study aimed to determine whether the representation of object aVordances requires speciWcation of the eVector potentially inter- acting with the object: speciWcally, in this study, vision of the interacting hand. In Experiment 1 we used an apparatus by which a fruit to be reached and grasped was identiWed by word reading, whereas another (interfering) fruit was visually perceived at the same location as the target. The apparatus allowed visual presentation of the agent’s interacting hand or pre- vented it. When visually presented, the hand was per- ceived as still at the start position even when it moved to grasp the fruit. An interference eVect on the grasp congruent with the distractor size was observed only when the hand was visible. In Experiment 2, interfer- ence was observed also when a hand diVerent from the agent’s own was visually presented. In both Experi- ments 1 and 2 the visible fruit interfered with the arm’s reach, but the eVect was independent of its size and less dependent on the visually-presented hand. A control experiment (Experiment 3) enabled comparison of the interference of visual stimuli on targets identiWed by word reading (Experiments 1 and 2) with that of objects identiWed by word reading on visually-pre- sented targets (Experiment 3). The interference induced by visual stimuli was stronger than the inter- ference induced by objects identiWed by words (i.e. aVordances evoked by visual stimuli were stronger than aVordances evoked by semantics). Taken together, the results of the present study suggest that the speciWcation of the agent’s eVector is necessary for the elicitation of aVordances. However, the elicitation of these aVordances was observed for interactions between object and hand (grasp), rather than for inter- actions between object and arm (reach). Finally, our data conWrm the inXuence of semantics on the control of arm movements, though less strong than that due to visual input.