Memes

HARRY R. ERWIN herwin at mason1.gmu.edu
Wed Jun 29 07:16:43 EST 1994


Has anyone considered the neuropil as possibly a suitable substrate for
Richard Dawkins's memes? The analogy with the double helix is loose, since
the neuropil is at least two-dimensional, but the genotype or template of
a meme might be identified with an activation network, with the phenotypes
then corresponding to pattern(s) of activation tracing to the original
template. The neuropil's ability to sustain active activation networks
over extended periods while also storing and adapting the associated
template pattern are properties to be expected of the meme substrate. This
leads me to an image of a 'soliton ecology' with 'soliton organisms'
making up the mind. Natural selection would result in at least some of
these memes corresponding to objects in the outside world and interacting
naturally (modulo some transformation between the external and inner
worlds). 

My reason for suggesting this has to do with my earlier work on n-person
non-sero-sum games with information collection. Those games, which underly
much of social behavior, are highly chaotic. The most effective way of
playing them appears to be to use personality simulations to predict the
behavior and interactions of the other players in the game and then act
based on those predictions. Those personality simulations would be natural
examples of activated memes. The open-ended nature of the game with a
variable number of players leads to a requirement for a single meme
template to be able to be the basis for multiple activated memes. 

I'm able to deliberately dissociate for short periods while instantiating
a personality simulation for a significant other person. I can then feed
it simulated sensory and situational data to help me predict that person's
behavior and motivations. (In a sense, reading his or her mind.) I assume
other people can do the same. This is the experience underlying my 
proposal here.

I think the connection to gestalt psychology is obvious, although this 
approach is eventually reductionistic. The question of what experiments 
would be appropriate is interesting.

--
Harry Erwin
Internet: herwin at gmu.edu 
Working on Katchalsky nets....



More information about the Neur-sci mailing list