Why do we do what we do? Motivation can be defined as the process that forms goals and transforms these into action. A goal can be reduced to a situation with anticipated positive (hedonic) value: earning money, performing well, being loved, etc... However, goal achievement typically involves investing effort, which is aversive. Hence, goal-directed behaviour necessarily trades incentive values with effort costs. This raises three basic questions:
(i) how does the brain computes net value (e.g., from prospective benefits and costs)?
(ii) how do psychological (e.g., cognitive biases) and/or biological (e.g., limited neural firing range) constraints influence value computations?
(iii) how do value computations determine behavioural outputs (choices, allocated effort, etc...)?
We start with the premise that most cognitive processes (e.g., value computations) can be traced back in the structure of macro-scale brain activity, as measured with modern neuroimaging and/or electrophysiology techniques. Typically, value computations involve many regions in the brain (from midbrain nuclei to basal ganglia, to limbic and prefrontal cortices), whose contribution to behaviour may depend upon the context (e.g., the specific goal that the brain is pursuing). In other terms, value computations are distributed over large-scale brain circuits, whose intimate connectivity structure determines information processing pathways. Importantly, this structure is plastic and under the influence of phasic and tonic changes in neuromodulatory activity. When altered, these regulatory processes may yield maladaptive behaviour.
Understanding the mechanics of motivational processes from the multimodal observation of brain activity and behaviour thus requires relating processes of value computation to the neurobiology of brain networks in a quantitative manner. To do this, I rely on formal mathematical theories that I borrow from diverse academic fields, such as Artificial Intelligence, Control Engineering and Statistical Physics.
My current research projects include:
Identifying the biological constraints on value computations in the orbitofrontal cortex (OFC) using fMRI/electrophysiology and artificial neural network models
Decomposing the metacognitive control of decisions using optogenetics in mice (collab. with E. Burguiere) and Markov decision processes.
Investigating attribution biases in the metacognitive regulation of mental effort using fMRI and Bayesian learning models.
Questioning the role of different kinds of resource constraints on mental fatigue using fMRI and optimal control theory (with applications to burnout and depression).
Understanding the motivational control of attentional resources using EEG neurofeedback (collab. with J. Mattout) and decision theory (with applications to ADHD).