Grand Unified Theories of the Brain Need Better Understanding of Behavior: The Two-Tiered Emergence of Function

Abstract

Over the last few decades, neuroscience, and various associated disciples, has expanded enormously in terms of output, tools, methods, concepts, and large-scale projects. In spite of these developments, the principles underlying brain function and behavior are yet only partially understood. We claim that brain functioning requires the elucidation of the rules associated with all possible task realizations, rather than targeting the activity underlying a specific realization. A first step into that direction was taken by approaches focusing on dynamical structures underlying task performances, as exemplified by Coordination Dynamics. The latter was originally based on experimental demonstrations of functional synergies/coordinative structures in complex movements like speech and interlimb coordination. Its theoretical foundation owes much to Hermann Haken's Synergetics, which provides a formalism, through which the degrees of freedom associated with highdimensional systems may be effectively reduced to one or a few functional variables. Synergetics' dimension reduction, however, is generally valid in the vicinity of phase transitions, a limiting factor in the framework's domain of explanation. The recent development of Structured Flows on Manifolds (SFM) is less generic than Synergetics, but allows the employment to a potentially broader range of applications. Following novel theoretical work on the onset, propagation, and offset of epileptic seizures, we expand the SFM framework, and propose that the emergent two-tiered fast-slow dynamics may be a basic mathematical organization underlying the architecture of brain and behavior dynamics. Finally, along a few examples, we illustrate how this framework allows for the incorporation of notions cardinal to ecological psychology.

Dynamic Oscillations Evoked by Subcallosal Cingulate Deep Brain Stimulation

“…Direct analyzes of DO dynamics are needed to fully extract signals that can test this hypothesis more fully. Further work constraining SINDy to known neural structures, like validated neural mass models, may be necessary to gain more direct insight into the neural components underlying the observed DO (Abbott et al, 2016 ; Jirsa et al, 2019 ).…”

Semiology and Epileptic Networks

“…More generally in neuroscience, understanding the brain-behavior relationship may require a move toward more behaviorally driven studies. 44,45 Technological advances could allow quantification of semiology by automated methods (eg, video analysis, movement quantification). 15,46 This ability could facilitate research efforts based on multimodal modelization of epilepsy, 7,19 which are currently not powered to take account of behavioral data.…”

Linking fast and slow: The case for generative models

“…The framework of structured flows on manifolds (SFM) adopts this procedure. SFM explains how slow low-dimensional variables, such as movement parameters, could be instantiated by the collective behaviour of a large number of fast variables, such as networks of neurons ( V. Jirsa & Sheheitli, 2022 ; V. K. Jirsa, McIntosh, & Huys, 2019 ; Pillai & Jirsa, 2017 ). In SFM, the fast components of the flow drive neuronal activity to rapidly evolve into synchronization modes (functional modes) that correspond to points on the manifold.…”