Investigating the role of sensorimotor versus contextual cues in the sense of joint agency: a human-human and human-robot study
Abstract
Sense of Joint Agency (SoJA), is the feeling of control experienced by humans for their own, as well as their partner's actions, when acting in joint action with others. SoJA is ubiquitous in human-human interaction. Therefore, it is both interesting as well as relevant to understand the factors that affect the formation of SoJA both in human-human and in human-robot interaction. On the one hand, previous work suggests that sensorimotor cues may be the main determinant of “lower-level” implicit SoJA in human-human joint action. On the other hand, recent work shows that contextual factors, such as perceived intentionality, can impact the formation of implicit SoJA with a humanoid robot. In the current study, we aimed to investigate, using behavioral and electroencephalography (EEG) measures, whether endowing a humanoid robot with a precisely human sensorimotor pattern would be sufficient to elicit SoJA with the robot, even without manipulating the perceived intentionality of the robot. Participants completed a joint task with another human, and with a robot that was controlled by another human, thus endowing the robot with a precisely human sensorimotor repertoire. Importantly, participants completed two sessions with this controlled robot. In one session, they were (factually) told that the robot was controlled by another human. In another session, they were told that the robot was pre-programmed. We expected that participants may attribute less intentionality to the robot they believed was pre-programmed. In either session, participants' perception regarding the robot's intentionality was not explicitly manipulated. Interval estimates for self- and other-generated actions were used to estimate SoJA. In addition, we also measured participants' readiness potential (RP) for self- and partner actions, and their N100 responses for self- and partner-generated sensory outcomes in the task. The results show that temporal estimates, and ERPs, did not differ between self- and partner-generated action-outcome conditions, in both human-human and human-robot sessions. Thus, these results suggest that endowing a humanoid robot with a human sensorimotor pattern may be sufficient to elicit SoJA in joint action with the robot, regardless of the intentionality attributed to the robot. Furthermore, an exploratory spectral analysis of movement-related beta activity suggested that people may nevertheless disengage earlier from the joint action when interacting with a robot partner, as compared to a human partner. Together, our results contribute to understanding the mechanisms that underlie the emergence of SoJA in joint action, as well as the extent to which artificial agents, such as humanoid robots, can be integrated into our teamwork as “full” interaction partners.
Related
- A Unified Neurocomputational Framework for Closed-Loop Motor Control and Sense of Agency
- Virtual Reality Simulation of Prosthesis embodiment: A Multimodal Investigation Examining Gaze Behaviour and Sense of agency
- From eco-affective narrowing to agency: acceptance, externalization, and efficacy in climate-related cognition
- Sensorimotor learning compensates for irreducible perceptual bias
- Does Pre-Sensorimotor Brain Activity Modulate TMS-Evoked Potentials?
BibTeX
@article{navare2026investigating,
title = {{Investigating the role of sensorimotor versus contextual cues in the sense of joint agency: a human-human and human-robot study}},
author = {Uma Prashant Navare and Veerle Hobbelink and Francesca Ciardo and Kyveli Kompatsiari and Davide de Tommaso and Agnieszka Wykowska},
journal = {Neuropsychologia},
year = {2026},
month = oct,
doi = {10.1016/j.neuropsychologia.2026.109372},
url = {https://hal.science/hal-05501432v1},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - Investigating the role of sensorimotor versus contextual cues in the sense of joint agency: a human-human and human-robot study
AU - Uma Prashant Navare
AU - Veerle Hobbelink
AU - Francesca Ciardo
AU - Kyveli Kompatsiari
AU - Davide de Tommaso
AU - Agnieszka Wykowska
PY - 2026
DA - 2026/10/13
JO - Neuropsychologia
DO - 10.1016/j.neuropsychologia.2026.109372
UR - https://hal.science/hal-05501432v1
AB - Sense of Joint Agency (SoJA), is the feeling of control experienced by humans for their own, as well as their partner's actions, when acting in joint action with others. SoJA is ubiquitous in human-human interaction. Therefore, it is both interesting as well as relevant to understand the factors that affect the formation of SoJA both in human-human and in human-robot interaction. On the one hand, previous work suggests that sensorimotor cues may be the main determinant of “lower-level” implicit SoJA in human-human joint action. On the other hand, recent work shows that contextual factors, such as perceived intentionality, can impact the formation of implicit SoJA with a humanoid robot. In the current study, we aimed to investigate, using behavioral and electroencephalography (EEG) measures, whether endowing a humanoid robot with a precisely human sensorimotor pattern would be sufficient to elicit SoJA with the robot, even without manipulating the perceived intentionality of the robot. Participants completed a joint task with another human, and with a robot that was controlled by another human, thus endowing the robot with a precisely human sensorimotor repertoire. Importantly, participants completed two sessions with this controlled robot. In one session, they were (factually) told that the robot was controlled by another human. In another session, they were told that the robot was pre-programmed. We expected that participants may attribute less intentionality to the robot they believed was pre-programmed. In either session, participants' perception regarding the robot's intentionality was not explicitly manipulated. Interval estimates for self- and other-generated actions were used to estimate SoJA. In addition, we also measured participants' readiness potential (RP) for self- and partner actions, and their N100 responses for self- and partner-generated sensory outcomes in the task. The results show that temporal estimates, and ERPs, did not differ between self- and partner-generated action-outcome conditions, in both human-human and human-robot sessions. Thus, these results suggest that endowing a humanoid robot with a human sensorimotor pattern may be sufficient to elicit SoJA in joint action with the robot, regardless of the intentionality attributed to the robot. Furthermore, an exploratory spectral analysis of movement-related beta activity suggested that people may nevertheless disengage earlier from the joint action when interacting with a robot partner, as compared to a human partner. Together, our results contribute to understanding the mechanisms that underlie the emergence of SoJA in joint action, as well as the extent to which artificial agents, such as humanoid robots, can be integrated into our teamwork as “full” interaction partners.
SN - 0028-3932
SN - 1873-3514
ER -