Institut des Systèmes Intelligents
et de Robotique


Sorbonne Universite



Tremplin CARNOT Interfaces



Holding hands may be more effective in group decision making than talking to each other


Human-machine kinesthetic communication. © Cyril FRÉSILLON, ISIR, CNRS Photothèque

Communicating is part of our daily lives, but do we really know what communicating is? We are naturally tempted to think about speech. We can easily guess that we can communicate through gestures like in sign language, or more generally when our gestures, postures and expressions complement speech. But do we communicate when we hold someone's hand?
In the act of holding hands, there is the movement that resembles sign language. There is touch, which allows us to perceive fine surface sensations such as roughness or temperature. And there is kinaesthesia, which makes it possible to feel efforts, whether or not they provoke movement.
Ludovic Saint-Bauzel, Lecturer at the Sorbonne University and member of the ISIR (1), is convinced that holding hands is a means of communication and more particularly the kinesthetic dimension (i.e. movement and force). He has been exploring this idea for several years. He has already shown that this channel makes it possible to negotiate a decision and that this negotiation can be imitated via an algorithm reproducing the human process to achieve the same decision-making efficiency. However, analyses on implicit sensation show that the virtual partner is still felt differently from a human partner (a study designed with Ouriel Grynszpan, Bruno Berberian and Elisabeth Pacherie).
To better understand how the implicit decision works, he then decided to explore the notion of shared trust. Indeed, effective group decisions require that group members communicate their levels of trust. Previous research has shown this for explicit communication channels, such as verbal communication, as a source for sharing trust levels and negotiating optimal decisions. The potential for implicit communication to enable optimal decision making could pave the way for negotiation technologies using rapid sensorimotor communication to solve group problems. The question of the kinesthetic channel as a communication channel is studied via a robotic device capable of transmitting forces precisely, and the position between two partners has been developed to address this hypothesis. It is then possible to study the importance and efficiency of kinesthetic communication in decision making.
The study carried out on thirty-six participants, associated in dyads (couples of individuals), shows that kinesthetic communication enables group decisions to be optimised and surpasses the individual precision of the participants. Indeed, pairs solve a task of perceptual discrimination more precisely than their best individual members, and five times faster than using explicit (verbal) communication. Computational analysis indicates that the kinesthetic channel (sensorimotor communication) allows sharing confidence levels when choosing a group, although they are never made explicit. The latter results therefore show that we are able to implicitly transmit trust by force-sharing when making decisions related to perceptual group discrimination, as well as with discussion, but more quickly.

The setup of the study carried out on 36 participants. © Lucas ROCHE, ISIR


The fruit of this work, a Franco-Italian collaboration, is presented to you in a document that has just been published in Nature Scientific Report. "Haptic communication optimises joint decisions and affords implicit confidence sharing", by Giovanni Pezzulo (2), Lucas Roche (3) and Ludovic Saint-Bauzel (3).




Referent contact : Ludovic Saint-Bauzel ;



(1) The Institute for Intelligent Systems and Robotics (Isir) is a Joint Research Unit (UMR7222) under the supervision of the Sorbonne University, the National Centre for Scientific Research (CNRS) and Inserm (ERL-U1150). This multidisciplinary research laboratory brings together researchers and teacher-researchers from different disciplines in robotics, life sciences and medical sciences.
(2) Institute of Cognitive Sciences and Technologies, National Research Council, Rome, Italy.
(3) Institute of Intelligent Systems and Robotics, Sorbonne University, Paris, France