Neuronal bases of structural coherence in contemporary dance observation

Asaf Bachrach, Corinne Jola, Christophe PallierNeuroImage 2016 · DOI

TL;DR

This study investigates how the brain processes the structural coherence of contemporary dance, moving beyond isolated movements to understand how sequences are interpreted. Using fMRI on expert dancers, they found that perceiving coherent dance structures engages brain regions associated with both action observation and higher-level cognitive functions like mentalizing and language processing, suggesting shared neural mechanisms for processing structure across different domains.

What Problem Does This Paper Address?

The field of action observation has extensively studied how the brain processes individual movements, often focusing on the "action observation network" (AON). However, a significant gap exists in understanding how we interpret sequences of movements, particularly in complex, non-verbal domains like contemporary dance. Unlike language, dance lacks a formal, universally agreed-upon "syntax," making it challenging to study its structural coherence. This paper aims to bridge this gap by exploring the neural underpinnings of interpreting the structural coherence of dance choreographies, questioning whether mechanisms similar to those used for language and music processing are involved.

Methods

The researchers employed a functional magnetic resonance imaging (fMRI) study with 22 professional contemporary dancers. They created video stimuli of dance excerpts that were parametrically manipulated for structural coherence. This was achieved by segmenting 16-second dance videos into 2-second snippets and then recombining them. The most coherent condition (c8) used snippets from a single, continuous dance phrase, while less coherent conditions (c4, c2, c1) involved concatenating snippets from different phrases, progressively disrupting the original structure. To control for low-level visual discontinuities introduced by scrambling, artificial "blanks" and "jitters" were overlaid on the video transitions in all conditions. Participants watched these videos and their brain activity was recorded. A whole-brain analysis was conducted to identify regions activated by dance observation and regions whose activity parametrically increased or decreased with coherence. Additionally, a region-of-interest (ROI) analysis was performed, focusing on areas previously implicated in processing linguistic structure.

Datasets

Not discussed in the paper. The study utilized custom-created video stimuli of contemporary dance excerpts.

Key Insights and Ideas

  • Dance Coherence Engages Beyond the AON: Interpreting structured dance sequences involves more than just the canonical action observation network. Regions associated with mentalizing (theory of mind) and language processing are also recruited.
  • Shared Neural Substrates for Structure: The study provides evidence for overlapping neural mechanisms used to process structural coherence in dance, language, and potentially music. This suggests that the brain might employ general-purpose mechanisms for parsing sequential information, regardless of the domain.
  • dmPFC and Protagonist Perspective: The dorsomedial prefrontal cortex (dmPFC) showed increased activity with higher dance coherence. This is linked to the idea of constructing a "protagonist perspective," inferring the intentions and motivations of the dancer, a function also observed in narrative comprehension.
  • Superior Parietal Lobe and Dance Syntax: The superior parietal lobe's increased activity with coherence is proposed to be involved in representing the "syntax" of dance – the spatial and temporal organization of the body's movement, akin to how language syntax organizes words.
  • IFG and pSTS as Cross-Domain Structure Processors: Regions in the inferior frontal gyrus (IFG) and posterior superior temporal sulcus (pSTS), known for linguistic syntax processing, also responded to structural coherence in dance, reinforcing the idea of shared structural processing mechanisms.

Key Results

  • Whole-Brain Analysis: Identified visual, somatosensory, motor, medial, and lateral frontal regions activated during dance observation.
  • Positive Coherence Effect: Significant increases in activation with higher dance coherence were found in:
  • Left dorsomedial prefrontal cortex (dmPFC)
    • Left superior parietal region
  • Negative Coherence Effect (Deactivation): Decreased activation with higher coherence (i.e., stronger response to less coherent stimuli) was observed in:
  • Occipital cortex
    • Ventral temporal lobes (lingual and fusiform gyri)
  • ROI Analysis (Linguistic Overlap): In regions previously identified for linguistic structure processing, significant effects of dance coherence were found in:
  • Pars orbitalis of the left inferior frontal gyrus (IFG)
    • Pars triangularis of the left IFG
    • Left posterior superior temporal sulcus (pSTS)
  • Task Performance: Participants correctly identified speeded-up, inverted video frames 95% of the time with minimal false alarms, indicating engagement with the task.

Limitations and Future Work

  • Stimulus Construction Confound: While efforts were made to control for low-level visual discontinuities introduced by scrambling, the authors acknowledge that these might still influence results. Future work could explore this further.
  • Cross-Domain Comparison: The study used separate experiments and participant groups for dance and language ROI analyses. The authors suggest that a more direct within-participant comparison of neural responses to structural manipulations across language, music, and dance is needed to solidify the findings of shared mechanisms.
  • Generalizability of dmPFC Role: The proposed role of dmPFC in protagonist perspective is speculative and could be further investigated in other complex, intentional action observation tasks.
  • Dance Syntax Formalization: The lack of a formal "dance syntax" makes direct comparisons to linguistic syntax challenging. Future research could explore ways to formalize dance structure to enable more direct neural mapping.

Summary updated

BibTeX
@article{bachrach2016neuronal,
  author = {Asaf Bachrach and Corinne Jola and Christophe Pallier},
  title = {{Neuronal bases of structural coherence in contemporary dance observation}},
  journal = {NeuroImage},
  year = {2016},
  doi = {10.1016/j.neuroimage.2015.08.072},
}
Figure