IUBio

Modelling the human brain by modelling its evolutionary emergence

mat mats_trash at hotmail.com
Fri Feb 22 12:35:43 EST 2002


> 
> You are certainly not stupid, but you did know that already. I will try to
> keep it as short as
> possible. Your first problem could be answered as follows:
> 1  Evolution is not entirely chaotic. There is an ordering principle
> involved as well, namely
>    the selection criterion 'Survival of the Fittest'

This is, in my view, wrong.  You seem to be suggesting that survival
of the fittest is a rule within the framework of evolution whereas in
fact it *is* evolution.  Survival of the fittest is the objective
emergent property we see due to the interaction of random mutations in
the genome and the unpredictable environment.  To assert that there is
a univeral rule for deciding which animals survive is incorrect.  To
correctly re-model survival of the fittest you would have to know
about the detail of the environment, which brings us back to the
problem in my previous post.  Don't get me wrong here - its perfectly
possible to set a model of an evolving system, and survival of the
fittest will emerge, based on what 'fittest' actually means within
that environmental context.  However, to correctly re-model the
evolution of the brain is impossible IMHO.

> 2  A major modelling problem is therefore: 'What is the result of this
> ordering principle for
>    the build-up procedure for the 'life form design', as it will become
> stored in DNA?

Why do you insist of saying things in such strange manners?  In every
sentence you seemingly introduce a new term that is previously
undefined eg build-up procedure and 'life form design'

> 3  This, very important, aspect is discussed at length in chapter 1 of
> 'Biography of Man'.
>    Below I will try to sketch a very short summary.
> 4  The test in the (hostile) environment, i.e. the struggle for life, will
> determine whether or
>    not any adaptation is favourable for that survival. This aspect must
> furnish the unifying
>    concept that will transform the results of the successful trials of the
> tinkering procedure of
>    evolution into a ‘neat' information handling architecture.

Again, what is this unifying concept you introduce here.  I'm sure you
could write this thesis in a much clearer and explicit manner, which
might help you get your ideas across.

> 5  The evolutionary analyses show that the evolutionary tinkerer made, via
> the survival test,
>    successive adaptations in the transformation of sensory observations into
> motor acts,
>    leading to a continuous improvement, at least for successful adaptations,
> in some virtual
>    image, created in the brain, that 'mirrored' the observed environment and
> the
>    individual's, and other's, position and (intended) actions in that
> environment. 

Can you really not write it out better than this?

> This is the
>    major, in fact the only, 'design strategy' of evolution. 

Really? How do you know this?  haven't you already agreed that
evolutioon has no design intended or strategy to follow?

> 6  This gradually improved ‘mirroring' of outside reality in some
> representation/image in the
>    brain is an absolute necessity for coping in a favourable way with any
> challenge that the
>    environment may pose to the individual. It is therefore no wonder that
> evolution's ‘trial
>    and error' resulted in such a virtual image representation.
> 7  This virtual image is shown to have evolved, in successive life forms,
> from simple, reflex-
>    like, structures, via separate egocentric- and allocentric images into a
> viewpoint
>    independent true virtual image, that became in Homo sapiens and, later,
> in Homo sapiens
>    sapiens further extended into purely abstract representations, that have
> lost any direct
>    relationship with material outside reality, but nevertheless were
> extremely important for
>    survival via tool use conceptions, planning, foresight, recalls from
> expanding know-how-
>    and experience databases, etc.
> 8  The successfullness of these adaptations is judged by the primary
> selection criterion in
>    evolution:
> 
> 'Does the modification/extension result in less fertile offspring?'
> 
That is not the 'selection criterion' of evolution, rather it is
whether an animal is disadvantaged so as not to be as likely to
reproduce.  Thats nothing to do with fertility, and if evolution had
only worked through altering fertility then none of what you describe
and what plainly has evolved, would have.  If mutations only worked on
fertility then the brain wouldn't be here at all.

>    If yes, the modification will disappear; otherwise it will remain and
> contribute to the
>    multiple variations in life forms, and if successful, will become even
> dominant. It results
>    in the evolutionary arbitration strategy that:
>    8a  gives, in general, preference to social instinct norms above
> individual instincts (care
>        for the individual), at least within the individual's social group
> (take primarily care of
>        the weaker members of the social group, such as children, older and
> handicapped),

That is highly speculative.  Many highly evolved animals do not form
societal relationships.


>        and:
>    8b  avoids, if possible, unnecessary deadly (dangerous) struggle.
>    This strategy promotes the number of fertile offspring.
> 9  These observations justify the statement that evolution constructed not
> only some
>    hierarchic structure but applied that structure also to one well-defined
> information entity,

>From where does this hierarchy come?

>    the virtual image and the response to the challenge(s) it contained. The
> struggle for
>    survival ensured that this information entity, and the resulting
> response, retained its well-
>    defined character as a gradually improving 'mirroring' of outside reality
> and of the
>    individual's, and other's, position in that environment.

Evolution does not 'ensure' the continued presence of anything.  If a
mutation arose to get rid of your 'virtual image' in favour of running
twice as fast then in many species it would win out.

> Your second problem does not exist, as far as I can see. The modelling
> involves the gradual
> upgrading of the virtual image in successive life forms and an analysis,
> with several,
> unavoidably somewhat speculative, assumptions, on the implementation of that
> virtual imagein the neural structure. We keep away from (sub)cellular
> levels; information handling is the
> primary issue. In this aspect of the modelling procedure we make use of the
> various available
> 'aiming points', as these can be derived, not only from Homo sapiens sapiens
> but also from
> much older life forms. A publication like that of Winson, 1990; 'The meaning
> of dreams';
> Scientific American November 1990 pages 42 through 48; See also: Scientific
> American,
> volume 276, special issue, 1997, pages 58 through 67, has played an
> important role in this
> 'aiming procedure'. We found that this gradual neural implementation
> procedure can also be
> characterized by a hierarchic and recursive development procedure. As a
> result, the immense
> complexity of the brain becomes much easier to describe and understand. It
> is (afterwards) no
> wonder that evolution concluded that the neural emergence procedure should
> be a simple
> recursive process; seen from a control viewpoint for this 'design
> information' handling task, it
> could not have been differently. The resulting brain emergence process is
> described in detail
> in chapter 5 of 'Biography of Man', the 'aiming points' in section 5/3; the
> neural
> implementation in section 5/4.

Look from all of what you've said I still do not have a clear idea of
what your starting point is going to be, what you actually want to
artifically evolve (what the heck are virtual images?), nor any clue
of how you intend to actually implement it.  And I don't mean in broad
terms.  Set out some explicit details.
> 
> Your third problem is, I think, resolved, or at least answered, already by
> the above
> observations. The resulting modelling is in fact the additional info, that
> is at present lacking in
> the genome project. 

After all, the genome project has discovered the
> existence of a very big
> book, written in 3 character words (the triplets), each built up from only
> four different
> characters (A, C, T, G) and written in a language we do not know.

Since when was the genetic code not known?  All triplet codes
correspond to amino acid residues or stop codons.  The language is
known.  Maybe we can't yet perfectly predict tertiary protein
structure but its not in the too distant future.
It is a
> very difficult version
> of the 'Rosetta Stone Problem'. The resulting modelling can be seen as a
> first step in
> unravelling the language problem for that Rosetta Stone.
> 
> I hope I have kept it sufficiently short. I look forward to further critical
> comments and I thank
> you for the stimulating discussion.
> Regards,
> Frans van der Walle

As far as I can tell you want to model some sort of cognitvie
architecture and its development under certain simulated envrionmental
conditions.  However, why do you think this will give you any sort of
information as to how the genome gives rise to 'real' cognitive
processes within the brain?  Its a huge gap from genome to neural
function and your project as outlined will do nothing to close it.




More information about the Neur-sci mailing list

Send comments to us at biosci-help [At] net.bio.net