Darwinian aspects of generating novel structured movement programs and choosing between alternatives.

 William H. Calvin
University of Washington
Department of Psychiatry & Behavioral Sciences
Box 351800
Seattle, Washington 98195-1800 USA
WCalvin@U.Washington.edu
faculty.washington.edu/wcalvin  

In my 1996 book, The Cerebral Code, I explored the consequences of a Hebbian cell-assembly, a minimal spatiotemporal firing pattern (“code”) able to clone itself in adjacent cortex via recurrent excitatory cortical connections of the “express-train” variety.  And when such a “plainchant choir” was sufficiently large, able to establish a distant metastasis of the identical code despite the usually incoherent corticocortical connections; this allows compounding codes by superposition.  A code could easily represent a collection of attributes constituting a concept, and it could be movement-related as easily as sensory.  It seems particularly suitable for higher-order concepts such as a string of movements (say, characteristic posture and finger movement sequence - or speaking a phrase).  Because the “choirs” can compete offline for cortical workspace with variant choirs, a Darwinian shaping up of quality can emerge, successive generations reflecting the current sensory and neurohumoral environment plus memorized environments of life so far.  As such, this provides a model for how novel movement sequences could be shaped up to match observed sequences.  I will also discuss tree-like structuring of concepts or movements (“I think I saw him leave to go home” is four sentences nested like Russian dolls) along the lines that Derek Bickerton  and I discuss in our new book, Lingua ex Machina: Reconciling Darwin and Chomsky with the Human Brain.  This provides a view of how movement subroutines could be separately shaped up and merged.