Message-Id: <Version.32.19990801170657.00f18610@pop3.htcomp.net>
Date: Sun, 01 Aug 1999 18:00:42 -0500
To: memetics@mmu.ac.uk
From: "Mark M. Mills" <mmills@htcomp.net>
Subject: Biophysics of memetics
Those interested in the neurology of memetic processes might want to look
into Koch's 'Biophysics of Computation, Information Processing in Single
Neurons.' (1999) Koch teaches at Cal Tech. The book covers a wide
variety of cellular processes capable of performing data processing
functions. In particular, it describes evidence of non-linear signal
processing and molecular flip/flop units capable of changing state within
periods of under 1 msec.
In terms of the Lynch-meme definition, the book suggests our memetic
substrate is the cellular membrane itself. The primary unit of
Lynch-memetic sequences could be autophosphorylating kinases (proteins)
found on cell membranes in synapses. Autophosphorylating kinases are the
molecular flip/flop switches I mentioned above. Thus, the meme would be
sequences of on/off switches. For those uncomfortable with a binary
memetics, the book makes it clear that though binary, the wide variety of
switching-element refresh periods produce analog responses. 13 different
switch mechanisms are discussed.
The book does not attempt to answer many obvious questions. It only
reports on experimental evidence relating to single neuron data processing
abilities. Thus, the reader is left with a number of capabilities, but no
'system.' It is clear that scientists can measure data transfer rates in
terms of 'bits,' but there are no comments on any 'address' schemes that
allow particular memories to be accessed.
On the other hand, it seems clear the author sees every neuron as an
independent computer.
The book closes with 'Strategic questions or how to find a topic for a PhD
thesis.' The 7 questions can be paraphrased as:
1. We can observe multiplication of signals, how is it implemented at the
cellular level?
2. What are the principle, intrinsic noise sources in the nervous system?
3. To what extent do considerations involving the energy metabolism play a
role in constraining the style of neuronal computation?
4. What is the function of the apical dendrite, the most prominent
component of many pyramidal cells in the cortex?
5. What type of learning rules could help pattern dendritic tree morphology
and the spatial distribution of voltage-dependent membrane conductances?
6. What is the function of dendritic trees?
7. Can you construct canonical single-cell models that are more faithful to
biology than the linear threshold models in usage today?
Mark
===============================================================
This was distributed via the memetics list associated with the
Journal of Memetics - Evolutionary Models of Information Transmission
For information about the journal and the list (e.g. unsubscribing)
see: http://www.cpm.mmu.ac.uk/jom-emit