From: Douglas P. Wilson (dp-wilson@shaw.ca)
Date: Mon Apr 01 2002 - 10:44:44 MST
Resend of bounced message:
----- Original Message -----
From: "Douglas P. Wilson" <dp-wilson@shaw.ca>
To: <virus@lucifer.com>
Sent: Monday, April 01, 2002 6:16 AM
Subject: Re: Re:virus: noise, superposition,linearity vs. \"Butterfly
Effect\" of a virus
> [rhinoceros] A lot of work you have done at your site!
> I hope I will find some time to read everything.
Thank you. I will be pleased if you read anything from my very large
number of pages.
My argument on linearity requires careful formulation and I hope you will
forgive me for finding fault with your restatement of it:
> Just one point. I think we cannot just say that *social animals are*
> (almost) linear because their social behavior is not chaotic
Careful statement of this argument must include the amount of feedback in
the system of interacting social animals, since it is the presence of both
feedback and non-linearity that causes chaos in the systems systems studied
by mathematicians
> and *biological viruses are* (extremely) nonlinear because their
reproduction and spread are chaotic
The effects of a biological virus and its reproduction can so greatly exceed
the cause that they are clearly non-linear, probably non-linear in various
ways, but without much than can be described as feedback there is nothing
that I see of as chaotic, in the sense given above. Artificial virus-like
software and more complicated memetic organisms could communicate and use
feedback, which could result in chaotic behaviour if reproductive effects
were out of all proportion to their causes.
Your comment on a need to use different mathematical models for these
different aspects may be correct, but analysis is often reductionistic on
purpose and using a single model may help, for an initial analysis -- a
pragmatic kind of Occam's Razor.
> Something can be nonlinear (chaotic) at one level and linear at a higher
level.
Only where lots of feedback is understood can we equate nonlinearity and
chaos, but except for that quibble I can agree with this statement.
> ... I just visualized the memeplexes in the human brain as raster images
> of patterns formed by synaptic paths, and I speculated that overexposure
> to many unimportant memes would floodfill areas of the image and would
> make it harder to recognize the patterns.
There is a great non-linearity behind raster images, which cut off at high
and low intensity levels, often those represented by the numbers 0 (dark, no
intensity) and 255 (bright, high intensity). Raster images represented by
these small unsigned integer values are like film which can be overexposed
(in the other sense of that term) by too much light. In either case it is
the non-linearity of the medium which is the problem.
dpw
This archive was generated by hypermail 2.1.5 : Wed Sep 25 2002 - 13:28:45 MDT