July 29, 2004 | David F. Coppedge

Your Brain Learned Physics and Calculus Before You Did

Tilt your head to the right while moving to the left.  The neurons in your brain just solved Newton’s equations of motion, and performed complex vector calculus equations almost instantaneously.  That’s what four neurologists Washington University of Medicine (St. Louis, MO) essentially claimed in Nature July 29,1 describing how your brain interprets the information coming from multiple sensory inputs.
The title of their article says it: “Neurons compute internal models of the physical laws of motion.”  The article is filled with equations that the neurons have to solve correctly to help you determine whether the motion you feel means you are moving left or right, tilting, or a combination of the two.  The signals come from the otoliths in your inner ear (see 10/10/2003 headline) and from the fluid in the semicircular canals.  What if these inputs give contradictory information?  The net vectors of the inputs could cancel each other out, or sum up to give a wrong impression.  The scientists mapped out the equations that would have to be solved to distinguish between the components of translational and gravitational motion, regardless of phase, and then experimented on monkeys while watching the activity of the brain.  They found that the way neurons fire in response to the stimuli match predictions of how the information would have to be parsed to fit the terms of the equation.  In conclusion, they state:

These results illustrate a direct correlation between cell firing rates and the equations of motion, as applied to movement in a gravitational environment and the physics of the external world.  A neural basis for an internal model representation of the relationship between the physical environment and either the sensory detectors or the motor apparatus has only recently begun to be explored.  Here we have shown evidence that, in support of theoretical predictions, subcortical neural populations might provide a distributed solution to the inertial motion detection problem.

1Angelaki, Shaikh, Green and Dickman, “Neurons compute internal models of the physical laws of motion,” Nature 430, 560 – 564 (29 July 2004); doi:10.1038/nature02754.

As expected, this paper makes no attempt to explain how such a system could have evolved.  The more detail provided in a research paper about the workings of a biological system, the less apt are the authors to attribute it to time and chance.