Interactor's body shape does not affect visuo-motor interference effects during motor coordination

Highlights

• Visuo-motor interference is found in online motor coordination.

• The presence of the interaction partner's body shape does not affect visuo-motor interference.

• Biological kinematics trigger the automatic tendency to imitate observed incongruent movements independently of body-shape.

Abstract

The biological-tuning of the Action Observation Network is highly debated. A current open question relates to whether the morphological appearance (body shape) and/or the biological motion of the observed agent triggers action simulation processes. Motor simulation of the partner's action is critical for motor interactions, where two partners coordinate their actions in space and time. It supports interpersonal alignment and facilitates online coordination. However, motor simulation also leads to visuo-motor interference effects when people are required to coordinate with complementary actions, i.e. incongruent movements as compared to the observed ones. Movement kinematics of interactive partners allows us to capture their automatic tendency to simulate and imitate the partner's complementary movements. In an online reach-to-grasp task, we investigated whether visuo-motor interference effects, visible in the kinematics of complementary movements, are modulated by the visual presence of the interactor's body shape. We asked participants to interact with 1) a humanoid agent with a human-like body shape and with real human, biological, movement kinematics, or 2) a non-humanoid agent, which did not resemble the human body-shape but moved with the same real kinematics. Through the combination of inferential and Bayesian statistics, the results show no effect of interactor's body shape on visuo-motor interference in reaching and grasping kinematics during online motor coordination. We discuss the results and propose that the kinematics of the observed movements, during motor interactions, might be the key factor for visuo-motor interference to take place independently from the morphological appearance of the partner. This is particularly relevant in a technological society that constantly asks humans to interact with artificial agents.

Introduction

Interpersonal motor interactions are fundamental forms of social encounters. At a performance level, even during the simplest on-line realistic interaction we need to accurately perceive, predict, and respond to others' movements to efficiently coordinate with them. Although we also coordinate with non-biological events, we frequently need to interact with our conspecifics, and the brain seems to have developed a specific network to control our behavior while we interact with other individuals (Era, Boukarras, & Candidi, 2019; Kokal, Gazzola, & Keysers, 2009; Newman-Norlund, van Schie, van Zuijlen, & Bekkering, 2007).

Electrophysiological and imaging studies in monkeys and humans have demonstrated that the visual system dedicates a specialized region in the posterior part of the Superior Temporal Sulcus (pSTS) to the processing of “biological motion”, i.e. motion patterns generated by biological agents (Puce & Perrett, 2003). Furthermore, studies with stroke patients found that lesions to pSTS impair the detection of biological motion (Saygin, 2007). pSTS receives inputs from lower-level visual areas and forwards these signals to parietal and premotor regions (Schippers & Keysers, 2011). Functional and anatomical connections between pSTS and parietal and premotor regions are implied in the transformation of visual information into motor coordinates (Sasaki, Kochiyama, Sugiura, Tanabe, & Sadato, 2012). Importantly, the pSTS also responds to the presentation of static body images implying movements (Giese & Poggio, 2003; Jellema, Maassen, & Perrett, 2004; Perrett, Xiao, Barraclough, Keysers, & Oram, 2009) and inhibition of its activity affects the strength of motor responses associated to implied action observation (Arfeller et al., 2013; Avenanti, Annella, Candidi, Urgesi, & Aglioti, 2012). Evidence from non-invasive brain stimulation (Avenanti et al., 2012) and brain damaged patients (Urgesi, Candidi, & Avenanti, 2014) shows that a temporo-parieto-frontal network, including pSTS, is involved in the perception and understanding of others' actions, possibly via simulative-like mechanisms (Keysers & Gazzola, 2014; Keysers & Perrett, 2004; Kilner, 2011).

Perceiving, simulating and predicting others' actions is fundamental when we need to coordinate with them in order to perform on-line joint actions, i.e. interactions that are directed to the achievement of a shared goal (Candidi, Sacheli, & Aglioti, 2015; Sebanz, Bekkering, & Knoblich, 2006). During on-line joint actions, however, the same neural resources that are activated to predictively represent the goal of others' actions are also needed for individual action planning and control (Kokal et al., 2009).

The tendency to simulate the observed movements in order to predict their fate (Aglioti, Cesari, Romani, & Urgesi, 2008), while also performing an action, generates facilitatory and interferent visuo-motor effects: the former during the execution of imitative movements, the latter in the case of complementary movements (Brass, Bekkering, & Prinz, 2001; Brass, Bekkering, Wohlschläger, & Prinz, 2000; Craighero, Bello, Fadiga, & Rizzolatti, 2002). Measuring movement kinematics can capture the behavioural effects of these visuo-motor transformations during online imitation and complementation of observed actions (D'Ausilio, Bartoli, & Maffongelli, 2015; Kilner, Paulignan, & Blakemore, 2003; Sacheli, Christensen, et al., 2015; Sacheli, Tidoni, Pavone, Aglioti, & Candidi, 2013). Visuo-motor interference on movement kinematics, for example, results in a higher variance of the executed movements when participants observe orthogonal vs imitative movements to those they execute (Kilner et al., 2003).

The degree to which movement kinematics and physical appearance of a stimulus that performs an action trigger motor simulation processes, and thus modulate our tendency to imitate the observed trajectories, is still a matter of debate. Previous studies show that physical appearance of a body and the kinematics of the observed movements are able to modulate behavioural (Kilner, Hamilton, & Blakemore, 2007; Kilner et al., 2003), neurophysiological (Agosta, Battelli, & Casile, 2016), electrophysiological (Meirovitch, Harris, Dayan, Arieli, & Flash, 2015; Urgen, Plank, Ishiguro, Poizner, & Saygin, 2013) and functional (Dayan et al., 2007; Gazzola, Rizzolatti, Wicker, & Keysers, 2007; Tai, Scherfler, Brooks, Sawamoto, & Castiello, 2004) indexes of motor simulation. However, the role of the physical appearance of a body has not yet been investigated during realistic, online, interpersonal motor coordination.

We tested whether visuo-motor interference effects (that is, the automatic tendency to imitate other's movement when we perform an action), present when individuals interact with a human (Candidi, Curioni, Donnarumma, Sacheli, & Pezzulo, 2015; Era, Candidi, Gandolfo, Sacheli, & Aglioti, 2018; Sacheli et al., 2013) or a virtual partner with a realistic body appearance (Candidi et al., 2017; Era, Aglioti, Mancusi, & Candidi, 2018; Moreau, Candidi, Era, Tieri, & Aglioti, 2018; Sacheli, Candidi, Era, & Aglioti, 2015; Sacheli, Christensen, et al., 2015), also emerge when participants coordinate with a stimulus that conveys no visual information of the form of the body, i.e. a set of dots connected by lines (Fig. 1), but moves with the very same kinematic pattern of a real person. We measured human motion kinematics while participants were reaching and grasping an object in synchrony with two stimuli and were asked to perform same or complementary movements to them. In line with previous work (Sacheli, Candidi, et al., 2015; Sacheli, Christensen, et al., 2015), we expected visuo-motor interference to emerge in the reaching component (wrist height) and in the pre-shaping component (grip aperture) of the joint grasping movements. To investigate the role of body shape information in the emergence of visuo-motor interference during motor interactions, we manipulated the physical appearance of the stimuli so that one would resemble a humanoid body shape and the other a non-humanoid shape while their biological motion kinematics were unchanged.