Another Thursday Mull

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Another Thursday Mull

Nick Thompson
Thoughts on what cellular automata have to tell us about development.

Dear Friamers,

        I fell across Wolfram in Steve’s Office and abducted it.   At the risk of
creating a GIANT windsucking all around FRIAMLAND, I have to admit I had
never read or even SEEN the book.   I had read a review which suggested
that the first six pages were the best and I have read those pages and am
having thoughts about why the book so excites people.  It provides an
instant model for the nature/nurture problem.  

        So, for argument purposes, allow me to define a WS = a Wolfram Space,
which is simply an R by C grid  with one square blackened in the first row
(1, n/2)  and squares blackened in subsequent rows in accordance with some
simple rule.   Here is an eleven by x wolfram space where blackened squares
are represented by x’s.  .

oooooxooooo
ooooxoxoooo
oooxoxoxooo
ooxoxoxoxoo
oxoxoxoxoxo
xoxoxoxoxox

this is the ws which is given by the set of eight rules which may be
summarized as “a cell is black if either of its neighbors was black on the
step before, …and white if both its neighbors were white on the step before.

The eight rules in this case are

xxx xxo xox xoo oxx oxo oox ooo
  x     x     x     x     x     o     x     o  

OK SO! (For those of you who love this stuff, God will reward you patience
as you have watched me, a neophyte, make a dog’s breakfast of it.  Thank
you from both of us.)

Let’s map a developmental space (DS) onto a WS  as follows. Each Row will
represent an instant in development  of some trait with the first row
representing  the first cell division and the last row representing the
form of the trait in the adult organism.  Ok, so let’s say  the 6th row on
a the above WS represents the final form of some pattern in an organism. We
might think of “xoxoxoxoxo” as being alternating bands of a salamander’s
tail.  Three ways of thinking about the nature/nurture problem can be
represented as three ways of understanding the columns in this DS.  

GENETIC ATOMISM.  Each column represents a series of gene products produced
by a single gene.  If asked why the adult has a particular pattern, the
GENETIC ATOMIST will say “because there is a geenfur each element in the
pattern.”

GENETIC HOLISM  Each Column represents the impact of a single genetic
instruction dictating the pattern at that stage of development.  Asked why
the adult has a particular pattern the GENETIC HOLIST will answer, “because
there is a geenfur the pattern.”

THE  DEVELOPMENTAL SYSTEMS THEORY (aka EVO-DEVO):  Each  column represents
a location at that stage of development homologous with the location in the
adult.  (Don’t hold me to this idea; it’s not the important part)  Each of
the eight RULES is a gene.  Asked why the adult has a particular pattern,
the DST theorist will say, “because there are eight genes determining the
relation between the state of affairs at each homologous location and the
state of affairs at the same location in the next developmental stage.  

What I like about this metaphor is that it gives me a sense of why
development is so.......MYSTERIOUS.  It is mysterious for the same reasons
that the patterns formed by cellular automata are mysterious.  

Quoting pp 39-40 from Wolfram,

“The pictures in the previous section plainly show that it takes only very
simple rules to  produce highly complex behavior.  Yet a first this
[conclusion] may seem almost impossible to believe.  For it goes against
some of the most basic intuition[s] about the way things normally work.  
For our everyday experience has led us to expect that an object that looks
complicated must have been constructed in a complicated way.   But…such an
assumption can be completely wrong.  For the patterns we saw are in effect
built according to very simple plans.  … Yet what emerges from these plans
shows an immense level of complexity.”  

How is may the contribution of the environment be represented in this
metaphor.  Recall that each row in a WS (=DS) represents a stage in
development.  According to DEVELOPMENTAL systems theory’s approach to the
nature/nurture problem, the form of the organism at each next stage of
development is the result of the interaction of the organism (including its
genes and gene products and the results of all previous environmental
impacts) with the environment at the present stage of development.   Thus
we can understand the role of the environment to be to perturb the
automatic progression of the stages in the DS by changing squares and thus
toggling the organism into a new developmental pattern.  For instance a
single environmental event, the changing of the 6,2 location from o to x
changes our salamander from a creature with a striped tail to one with a
solid black tail.  

oooooxooooo
ooooxXxoooo
oooxxxxxooo
ooxxxxxxxoo
oxxxxxxxxxo
xxxxxxxxxxx

This example gives new life to what is meant by environment/gene
interaction.  Let us imagine for a moment two populations of salamanders,
genetically identical, but whose environments differ ONLY in the presence
or absence of the environmental factor that transforms the 2,6 location
from o to x.  Imagine further that  tail pattern is connected to a specific
mate recognition mechanism, such that striped salamanders will mate only
with striped salamanders, etc.  In such a system a SINGLE environmental
impact could result in reproductive isolation and speciation.  

One implication of this idea is that if, as Wolfram repeatedly asserts,
biological systems must work this way, then biological systems must be
extraordinarily buffered against environmental variance.  In fact, I would
be on the look out for a class of cellular automata that are self
repairing.  

I can year you all shouting at me “HOORAY.  THE OLD GUY GOT IT”  If so,
thanks again for your patience.  You will be rewarded in heaven.
Unfortunately, I shall probably not be there it see the presentation.

All the best,

Nick





Nicholas S. Thompson
Professor of Psychology and Ethology
Clark University
[hidden email]
http://home.earthlink.net/~nickthompson/
 [hidden email]

-------------- next part --------------
Dear Friamers,

        I fell across Wolfram in Steve’s Office and abducted it.   At the risk of creating a GIANT windsucking all around FRIAMLAND, I have to admit I had never read or even SEEN the book.   I had read a review which suggested that the first six pages were the best and I have read those pages and am having thoughts about why the book so excites people.  It provides an instant model for the nature/nurture problem.  

        So, for argument purposes, allow me to define a WS = a Wolfram Space, which is simply an R by C grid  with one square blackened in the first row (1, n/2)  and squares blackened in subsequent rows in accordance with some simple rule.   Here is an eleven by x wolfram space where blackened squares are represented by x’s.  .

oooooxooooo
ooooxoxoooo
oooxoxoxooo
ooxoxoxoxoo
oxoxoxoxoxo
xoxoxoxoxox

this is the ws which is given by the set of eight rules which may be summarized as “a cell is black if either of its neighbors was black on the step before, …and white if both its neighbors were white on the step before.

The eight rules in this case are

xxx xxo xox xoo oxx oxo oox ooo
  x     x     x     x     x     o     x     o  

OK SO! (For those of you who love this stuff, God will reward you patience as you have watched me, a neophyte, make a dog’s breakfast of it.  Thank you from both of us.)

Let’s map a developmental space (DS) onto a WS  as follows. Each Row will represent an instant in development  of some trait with the first row representing  the first cell division and the last row representing the form of the trait in the adult organism.  Ok, so let’s say  the 6th row on a the above WS represents the final form of some pattern in an organism. We might think of “xoxoxoxoxo” as being alternating bands of a salamander’s tail.  Three ways of thinking about the nature/nurture problem can be represented as three ways of understanding the columns in this DS.  

GENETIC ATOMISM.  Each column represents a series of gene products produced by a single gene.  If asked why the adult has a particular pattern, the GENETIC ATOMIST will say “because there is a geenfur each element in the pattern.”

GENETIC HOLISM  Each Column represents the impact of a single genetic instruction dictating the pattern at that stage of development.  Asked why the adult has a particular pattern the GENETIC HOLIST will answer, “because there is a geenfur the pattern.”

THE  DEVELOPMENTAL SYSTEMS THEORY (aka EVO-DEVO):  Each  column represents a location at that stage of development homologous with the location in the adult.  (Don’t hold me to this idea; it’s not the important part)  Each of the eight RULES is a gene.  Asked why the adult has a particular pattern, the DST theorist will say, “because there are eight genes determining the relation between the state of affairs at each homologous location and the state of affairs at the same location in the next developmental stage.  

What I like about this metaphor is that it gives me a sense of why development is so …. so……. so………………..MYSTERIOUS.  It is mysterious for the same reasons that the patterns formed by cellular automata are mysterious.  

Quoting pp 39-40 from Wolfram,

“The pictures in the previous section plainly show that it takes only very simple rules to  produce highly complex behavior.  Yet a first this [conclusion] may seem almost impossible to believe.  For it goes against some of the most basic intuition[s] about the way things normally work.  ¶ For our everyday experience has led us to expect that an object that looks complicated must have been constructed in a complicated way.  ¶ But…such an assumption can be completely wrong.  For the patterns we saw are in effect built according to very simple plans.  … Yet what emerges from these plans shows an immense level of complexity.”  

How is may the contribution of the environment be represented in this metaphor.  Recall that each row in a WS (=DS) represents a stage in development.  According to DEVELOPMENTAL systems theory’s approach to the nature/nurture problem, the form of the organism at each next stage of development is the result of the interaction of the organism (including its genes and gene products and the results of all previous environmental impacts) with the environment at the present stage of development.   Thus we can understand the role of the environment to be to perturb the automatic progression of the stages in the DS by changing squares and thus toggling the organism into a new developmental pattern.  For instance a single environmental event, the changing of the 6,2 location from o to x changes our salamander from a creature with a striped tail to one with a solid black tail.  

oooooxooooo
ooooxXxoooo
oooxxxxxooo
ooxxxxxxxoo
oxxxxxxxxxo
xxxxxxxxxxx

This example gives new life to what is meant by environment/gene interaction.  Let us imagine for a moment two populations of salamanders, genetically identical, but whose environments differ ONLY in the presence or absence of the environmental factor that transforms the 2,6 location from o to x.  Imagine further that  tail pattern is connected to a specific mate recognition mechanism, such that striped salamanders will mate only with striped salamanders, etc.  In such a system a SINGLE environmental impact could result in reproductive isolation and speciation.  

One implication of this idea is that if, as Wolfram repeatedly asserts, biological systems must work this way, then biological systems must be extraordinarily buffered against environmental variance.  In fact, I would be on the look out for a class of cellular automata that are self repairing.  

I can year you all shouting at me “HOORAY.  THE OLD GUY GOT IT”  If so, thanks again for your patience.  You will be rewarded in heaven.  Unfortunately, I shall probably not be there it see the presentation.

All the best,

Nick