Mike Gilbert

Mike Gilbert

School of Psychology
Doctoral Researcher

Contact details

Address
School of Psychology
University of Birmingham
Edgbaston
Birmingham
B15 2TT
UK

PhD title A Theory of the Cerebellum
Supervisor Professor Chris Miall

The basic circuit wiring of the cerebellum is well preserved across species. There is a tendency for research to be in silos with the result that it is being underused. We pull together cross-disciplinary research into a proposal of the principles underpinning modular circuit wiring. The scope is to answer the question, how do circuits derive output from input? It is a convention for theory to begin with a proposal for the function of a brain region or structure, model it with an algorithm, and only then consider how it might be implemented physiologically (evidence  last). We use instead a bottom up approach, and a wide evidence base, assembling the pieces into a 4-D jigsaw, and using modelling to go beyond the data and to test predictions.

Qualifications

  • LLB 1st Class Hons (University of Westminster)
  • Legal Practice Course (a professional postgraduate qualification)

Biography

I practiced as a corporate lawyer for 14 years before coming to neuropsychology. To improvise a route into neuroscience I developed the ideas which underpin part of the theory, created and posted them on a website, took it down when it led nowhere, but in the meantime began a dialogue with Chris. I have been a speaker at St John’s College, Oxford at an international cross-disciplinary workshop on internal representations of space, in addition to giving internal talks. I am a member of the BNA.

Research

Research group

Prism Lab

Work

Most theories of the cerebellum are heavily influenced by Marr. Marr proposed (Marr’s 3 levels of analysis, applying to modelling generally) that the form of a model should be an algorithm, with evidence relegated to the final, ‘implementation’ level, after modelling.

Studies of cerebellar physiology and synaptic plasticity have created a pool of evidence with the asset that cerebellar circuits are modular, so that most research has relevance for most circuits. We have reversed Marr’s approach, putting evidence first and using modelling to draw conclusions about function (as opposed to starting with a proposed function and using modelling to fix the problem, and then checking it against the evidence).

Other activities

Walking in Yorkshire. Novel and song writing.