[Comp-neuro] Re: puzzlement

Axel HUTT axel.hutt at loria.fr
Thu Aug 14 12:45:35 CEST 2008


On Wed, 2008-08-13 at 13:26 -0400, Bill Lytton wrote:
> >  as a physicist by education but working in neuroscience, I am really puzzled about the
> >  approach to start from the single neuron to understand information processing.
> 
> I am puzzled by your puzzlement
> 
> There is great complexity in a neuron that directly effects its I/O
> -- the internal quantum complexity of a neutral particle doesn't much effect its billiard-ball
> like behavior; the complexity of H2O does start to effect its interactions which makes aqueous
> solutions difficult to model and to reduce cleanly to an analytic form
> 
> the brain is often compared to a computer, and sometimes, as implicitly by Jim, to a modern
> aircraft -- complexity arising from the connection of complex components.  Individual
> neuron types differ in their complexity, perhaps lying somewhere between an op-amp (a granule
> cell) and an 8086 CPU (2e4 transistors) -- a Purkinje cell
> 
> there are papers by a phys-type that lays this out in a relatively formal way
>   Csete, ME and Doyle, JC, Science 295:1664-1669, 2002 
>   Carlson, JM and Doyle, JC, PNAS    99:2538-2545, 2002
> I recall that they also compare brains to airplanes

I am fully aware of the structure of complex systems (sub-systems of
particles interact and thus create other sub-systems which also interact
and so on). My point is that todays neuroscience mixes spatial and
 temporal scales and aims to find a model description of all scales 
in one model (if you want a Theory Of Everything). In my opinion, this 
approach is reasonable but not practical, since the neural systems under
 study are mostly too complex for us. 
Considering your airplane example (which is a good one indeed), we have 
to understand the mechanism of buoyancy, the actuation of turbines, the 
control of the electronic circuits and so on. If we understand the major
mechanism of each part, we can combine the parts together and control 
the aircraft. Now taking a look at the physics basics of each part, 
we find a macroscopic model for fluids (Bernoullis law for the
 buoyancy), thermodynamics for the turbines ans so on but no detailed 
model for molecular interactions. The aircraft does fly because we know
 the macroscopic properties of each part (and engineers indeed construct
 it with a large variance of the material properties to ensure that it
really works..).
The same with the brain: for instance Hubel and Wiesel found out as ones
of the first that the visual cortex exhibits patchy connections which 
are strongly related to the orientation tuning in the visual field. In 
this finding, it was sufficient to know that the spatial distribution
of the population firing rate exhibits a periodic structure while the
 detailed activity of each neuron was not important. This macroscopic
finding (spatial patchy connection) is an important contribution to 
the understanding of the brain. Another good example is the Ermentrout-
Cowan explanation of visual hallucinations: today we know the underlying
model (the celebrated Wilson-Cowan equations) considers a simple
 rate-coding mechanism and neglects synaptic activity. But, hey, it is 
strinking that this simple model allows for the description of such 
a complicated phenomenon as visual hallucinations. Hence may the
 Ermentrout-Cowan model reflects a basic mechanism.

I am sorry if this post is too off-topic, I just wanted to clarify 
my point. Eventually, my point is some kind of warning of the tyranny 
of the idea of a All-in-one approach. IMHO the history of physics, 
chemistry AND biology (everything started by classifying animals and 
plants to finally gain deeper insight into finer structures) shows that 
first we need to model the macroscopic structures to eventually approach
finer scales. Research in the last centuries have not not succeeded by
 considering all experimental details at once.

Best regards


Axel




More information about the Comp-neuro mailing list