Something in the way she moves me: morphology and motion of observed goal-directed and pantomimed actions

Abstract

Observing other actors excites our own system (Fadiga, Fogassi, Pavesi, & Rizzolatti, 1995) and alters behavioural output (Kilner, Paulignan, & Blakemore, 2003; Ocampo & Kritikos, 2010). The features of the actor that drive these alterations, however, are unclear. In the present study, we focused on the appearance of the actor versus the intrinsic biological motion (BM) of her action, and the category of action-goal directed as compared with action-nongoal directed (pantomimed).In ideomotor priming paradigms, observed actions trigger corresponding actions in the observer. Participants typically view videos of a hand lifting the middle or index finger, and respond by making the same (congruent) or opposite (incongruent) lift. In the crucial comparison, symbols (shapes or numbers) determine responses. Response times (RTs) are faster to fingers than to symbols (Brass, Bekkering, Wohlschlager, & Prinz, 2000; Jonas et al., 2007). Moreover, watching humans alters the spatial components of observers' actions. Kilner et al. (2003) had participants make horizontal or vertical movements with their arms while watching either a human or a robot make congruent or incongruent actions. Participants' trajectories were more variable during incongruent than congruent actions, but only when watching the human actor. Kilner et al. argued that this motor contagion is specific to watching another human, and that the human visuomotor system processes human and robot actions differently.The main impetus for the present study was that appearance is conflated with BM: We attempt to disambiguate the two. BM is the movement pattern of animate beings from degraded signals, typically point lights placed at joints (Blake & Shiffrar, 2007; Johansson, 1973). Observers can distinguish BM from arbitrary but coherent motion (Hiris, Krebeck, Edmonds, & Stout, 2005) and can extract identity, gender and affect (Pollick, Fidopiastis, & Braden, 2001; Pollick, Paterson, Bruderlin, & Sanford, 2001). Observers recognize their own motion most accurately, then that of their friends, whereas stranger recognition is at chance (Loula, Prasad, Harber, & Shiffrar, 2005). When displays are inverted, selfas well as other recognition is at chance. Thus, and visual experience of the actor are required for recognition, but not low-level information about form and motion (Loula et al., 2005). We cannot assume the same influence on the output of the observer, however.Most studies showing an advantage in action execution when watching human actors have used nongoal-directed actions (e.g., Brass et al., 2000; Kilner et al., 2003). Importantly, pantomimes have longer movement duration and smaller maximum grasp than do goal-directed actions (Goodale, Jakobson, & Keilor, 1994). Goodale et al. (1994) postulated that goal-directed movements are computed online, whereas pantomimes are based on stored representations of objects. Thus, ideomotor apraxics have difficulty pantomiming meaningful actions, which is ameliorated when they are able to use the object (Buxbaum, Sirigu, Schwartz, & Klatzky, 2003). Conversely, the pantomimed reaches of the visual agnosic patient DF to familiar objects were intact because she had a premorbid stored representation of these. She could not pantomime to novel objects, however, because she did not have stored representations and could not construct a visual one (Goodale et al., 1994). Thus, when asking whether motion or appearance of the actor has an impact on the actions of observers, we also need to specify whether they are goal directed or pantomimed.Using motion capture software, we recorded human pantomimed or goal-directed actions, also exporting them as a point displays (PD) of identical BM parameters. Participants made congruent and incongruent responses to the real or the PD stimuli. Observing the real actor should benefit performance if the visuomotor system processes human appearance preferentially, with faster responses and changes in the grasp aperture than in the PD or no-actor baseline. …