Taking two to Tango: fMRI analysis of improvised joint action with physical contact Léa A. S. Chauvigné ,
Michel Belyk,
Steven Brown
Abstract
Many forms of joint action involve physical coupling between the participants, such as when moving a sofa together or dancing a tango.
We report the results of a novel two-person functional MRI study in which trained couple dancers engaged in bimanual contact with an experimenter standing next to the bore of the magnet, and in which the two alternated between being the leader and the follower of joint improvised movements. Leading showed a general pattern of self-orientation, being associated with brain areas involved in motor planning, navigation, sequencing, action monitoring, and error correction. In contrast, following showed a far more sensory, externally-oriented pattern, revealing areas involved in somatosensation, proprioception, motion tracking, social cognition, and outcome monitoring. We also had participants perform a “mutual” condition in which the movement patterns were pre-learned and the roles were symmetric, thereby minimizing any tendency toward either leading or following. The mutual condition showed greater activity in brain areas involved in mentalizing and social reward than did leading or following. Finally, the analysis of improvisation revealed the dual importance of motor-planning and working-memory areas. We discuss these results in terms of theories of both joint action and improvisation.
Figures
Fig 1. Experimental design.
(A) This figure shows the contact between the participant and experimenter during the partnered conditions. They performed bimanual movements of the wrist and fingers in all three planes of motion, with principal contact occurring at the inner surfaces of the fingers. The participant’s hands were always palm-up below the experimenter’s hands. (B) The tasks were organized according to a 2 x 2 scheme, where one variable was partnership (partnered vs. solo tasks) and the other was improvisation (improvised vs. pre-learned movement patterns). There were five movement conditions and a baseline condition of Rest (not shown in the figure). The tasks of interest were the partnered conditions of Leading, Following, and Mutual. As a control for partnering, we had participants perform similar motor tasks, but on their own (Solo and Alone). Regarding the improvisation variable, the movements during Leading, Following and Solo were improvised. During the two non-improvised conditions, namely Mutual (partnered) and Alone (non-partnered), participants performed pre-learned movement sequences. Note that the Following condition did not tap into improvisational mechanisms of production on the part of the participant, as indicated by the jagged line for Following in the figure. Hence, the Following condition was excluded in the analysis of the main effect of improvisation
Fig 2. Main effect of partnering.
Partnering was examined by contrasting the partnered conditions with the non-partnered conditions (Leading + Following + Mutual > Solo + Alone), with results reported at p < 0.05, FDR corrected, with a cluster threshold k = 20. The results in Figs 2–4 are registered onto a Talairach-normalized anatomical template MRI (the Colin brain). The Talairach z coordinate is shown below each slice. The left side of the slice is the left side of the brain. Abbreviations: MCC: middle cingulate cortex; mPFC: medial prefrontal cortex; pSTS: posterior superior temporal sulcus; S1: primary somatosensory cortex; S2: secondary somatosensory cortex; TPJ: temporo-parietal junction.
https://doi.org/10.1371/journal.pone.0191098.g002
Fig 3. Specificity for Leading, Following, and Mutual.
Neural specificity for Leading, Following and Mutual is shown, after removing both basic motoric effects (through subtraction of the non-partnered conditions Solo + Alone) and partnering effects, as seen in a conjunction of the three partnership contrasts: [Leading > non-partnered conditions] ∩ [Following > non-partnered conditions] ∩ [Mutual> non-partnered conditions]. The role-specific activations are color-coded as follows: Leading (red): [Leading > Non-partnered conditions] > [Partnership Conjunction]; Following (blue): [Following > Non-partnered conditions] > [Partnership Conjunction]; and Mutual (yellow): [Mutual> Non-partnered conditions] > [Partnership Conjunction]. The results are p < 0.005 uncorrected, with a cluster threshold k = 20. Abbreviations: aIPL: anterior inferior parietal lobule; MCC: middle cingulate cortex; MT+/V5, motion area of the middle temporal region; mPFC: medial prefrontal cortex; NA: nucleus accumbens; pSTS: posterior superior temporal sulcus; S1: primary somatosensory cortex; S2: secondary somatosensory cortex; TPJ: temporo-parietal junction.
https://doi.org/10.1371/journal.pone.0191098.g003
Conclusions
Using a novel two-person fMRI scanning arrangement, we elucidated for the first time neural differences between the motor-driven task of being a leader and the sensory-driven task of being a follower during a situation of joint improvisation with direct haptic contact. The results shed light not only on the complementary features of leading and following, but on the neural basis of improvisation as well. We found that performing partnered hand movements activated somatosensory as well as social networks. Leading such movements principally activated a motor network involved in motor planning, spatial navigation, and monitoring self-initiated action. In a complementary fashion, haptically following partnered movements engaged areas that monitor externally-triggered action as well as sensory-oriented areas that process somatosensation, motion perception, and the perception of dynamic social stimuli. In contrast to the asymmetry of leading and following, engagement in a more symmetric and mutual interaction increased activity in mentalizing areas and regions involved in social reward. We observed that dance-like improvisation engaged a similar network to musical improvisation or random sequence generation. Moreover, we were able to dissociate a network devoted to improvisation–such as would be engaged in internal sequence generation, decision making, and willed action–from a network involved in sequence variability and movement complexity.
Haptic contact has been a neglected topic in the neuroscience of social interaction. Our study unites haptic contact with the topic of joint action, and by doing so highlights the importance not only of social touch but of the reciprocal exchange of forces necessary for joint cooperative actions of all types.
More:
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0191098
https://www.facebook.com/mariabuenosaires