Motor activation in language processing : effects of handedness, experience, and planning
Beveridge, Madeleine Edith Louisa
Embodied Cognition accounts propose that motor activation contributes to semantic representations in action language (Fischer & Zwaan, 2008). However, the nature of this activation remains largely unspecified: in particular, which processes result in relevant activation? Long-term motor experience (e.g., the comprehender’s dominant hand), short-term motor experience (e.g., the hand the comprehender has recently used), and action planning (e.g., the hand the comprehender is planning to use) are all potential candidates. This thesis uses a range of psycholinguistic methods (e.g., timed sentence-picture matching, two-alternative forced-choice sentence-picture matching, spoken sensibility judgements) to distinguish between these possibilities. A first set of experiments investigated how comprehenders’ handedness affects their interpretation of sentences describing manual actions (e.g., I am slicing the tomato). Participants matched sentences of actions to pictures of that action. The Body-Specificity Hypothesis (Casasanto, 2009; Willems, Hagoort, & Casasanto, 2010) predicts that right-handed and left-handed comprehenders will interpret manual action sentences differently, according to whether they would perform that action with their right or their left hand. However, we found that comprehenders appear to interpret manual action sentences according to the hand they use to respond to the task, and not the hand they would typically use to perform manual actions. In addition, this effect was stronger for first-person than third-person sentences, implying that the effect of motor activation is moderated by linguistic context. A second set of experiments used the same paradigm but manipulated at what point comprehenders knew which hand they would use to respond to the sentences: during sentence processing, or after sentence processing was complete. We replicated the finding that comprehenders interpret manual action sentences according to their response hand, and that this effect was stronger for first- than for third-person sentences; but only when comprehenders knew their response hand during sentence processing. In both sets of experiments, there was no effect of whether the picture of the action was presented from an egocentric or allocentric perspective, implying that action sentences are encoded for what effector (in this case, hand) will be used in the action, but not necessarily from what perspective the action will occur. A third set of experiments investigated the existence of a causal role of action planning-based activation on sentence processing. Many studies have shown an effect of language processing on action execution (e.g., Glenberg & Kaschak, 2002; Glenberg et al., 2008), but a fully embodied theory of language also predicts an effect of motor activation on language processing. Here, right-handed participants made spoken judgements about sentences while planning an action with their right or left hand that matched or did not match the action described in the sentence. An effect of response hand on accuracy was found when the task required participants to explicitly judge the congruency of sentence and the action they were preparing, but not otherwise. These results corroborate recent research suggesting that activation of embodied lexical representations may be goal-driven rather than an automatic aspect of language processing (Hoedemaker & Gordon, 2013). Overall, the experiments presented in this thesis suggest a possible role for planning-based motor activation in sentence processing, in line with embodied approaches; however, the results challenge strong accounts of embodiment by suggesting that the effect of planning-based activation is not automatic, and is moderated by linguistic context and task demands.