Abbott is not alone; Evolutionary errata II

[From Bruce Abbott (970812.0720 EST)]

Cleaning up some loose ends . . .

Rick Marken (970803.0900)]

Yesterday I discovered that Bruce Abbott is not the only one
who believes that Darwinian natural selection is a closed-loop
control process.

Bruce Abbott does not believe that Darwinian evolution is a closed-loop
control process, and never did.

Rick Marken believes that Bruce Abbott sees it this way, but Rick Marken is
jumping to conclusions.

Bruce Abbott believes that Darwinian evolution is a closed-loop, negative
feedback process.

The same goes for Skinnerian reinforcement.

Rick Marken (970805.0832)]

OK. One more time:

         1.0 1.0 1.0
genotypes--->phenotypes--> fitness <---disturbances
   ^ | 1.0
   > v
   -----------------------survival
                 1.0

I would like to enquire of Rick how he arrived at the fictional gain values
all around this loop. It looks to me as though he started with the
assumption that this is a positive feedback loop with gain of 1.0 and worked
backward from there to provide a set of numbers that would make it come out
this way.

The diagram mixes quantitative and qualitative variables. How does the
feedback from "survival" to "genotypes" affect "genotypes"? Where in this
diagram do mutations enter? Are these and changes in environmental
conditions lumped together under the heading of "disturbances"? Also, this
diagram appears to assert a one-to-one correspondence between genotypes and
phenotypes, and between phenotypes and fitness. Is that what you meant to do?

Regards,

Bruce

[From Bruce gregory 9970812.0930 EDT)]

Bruce Abbott (970812.0720 EST)

>Rick Marken (970803.0900)]

>Yesterday I discovered that Bruce Abbott is not the only one
>who believes that Darwinian natural selection is a closed-loop
>control process.

Bruce Abbott does not believe that Darwinian evolution is a closed-loop
control process, and never did.

Rick Marken believes that Bruce Abbott sees it this way, but Rick Marken is
jumping to conclusions.

Bruce Abbott believes that Darwinian evolution is a closed-loop, negative
feedback process.

There seems to be some misunderstanding. CSGnet is not an
old time revival meeting where crying out, "I believe..." gains
one entrance into heaven. Personally, I am uninterested in what
Bruce Abbott believes. A model of Darwinian evolution
_demonstrating_ negative feedback _would_ be interesting,
however.

Bruce Gregory (who _tries_ to keep his beliefs to himself on
the assumption they are not all that interesting to others)

[From Bruce Abbott (970812.1005 EST)]

Bruce gregory 9970812.0930 EDT) --

Bruce Abbott (970812.0720 EST)

>Rick Marken (970803.0900)

>Yesterday I discovered that Bruce Abbott is not the only one
>who believes that Darwinian natural selection is a closed-loop
>control process.

Bruce Abbott does not believe that Darwinian evolution is a closed-loop
control process, and never did.

Rick Marken believes that Bruce Abbott sees it this way, but Rick Marken is
jumping to conclusions.

Bruce Abbott believes that Darwinian evolution is a closed-loop, negative
feedback process.

There seems to be some misunderstanding. CSGnet is not an
old time revival meeting where crying out, "I believe..." gains
one entrance into heaven. Personally, I am uninterested in what
Bruce Abbott believes. A model of Darwinian evolution
_demonstrating_ negative feedback _would_ be interesting,
however.

Why didn't you complain to Rick Marken that you are personally uninterested
in what HE believes I believe? Do I detect a bit of bias here?

When one's views are mischaracterized on CSGnet, I think one has the right
to provide a correction, whether you personally like to hear it or not. As
for a model _demonstrating_ negative feedback, are you really _asking_ me to
provide such a model? Last we heard from you, you had already rejected it,
apriori.

Your logic went as follows, as I recall: You couldn't see how it could
involve negative feedback, therefore one of us (I infer you meant me) must
not understand negative feedback. This from a man who claimed in another
post that he likes to try to view an idea from another man's perspective
before making a decision. Sure you do. You never even _asked_ for my
reasoning. Your mind was already made up.

As for my providing my reasoning, if you will recall, this would be a waste
of my time: Rick Marken has asserted that I would just be making it come out
the way I wanted it to, so it would mean nothing.

Bruce Gregory (who _tries_ to keep his beliefs to himself on
the assumption they are not all that interesting to others)

And fails. You have already asserted (without proof) that Darwinian
evolution does not involve negative feedback. That's a belief, and yet you
posted it on CSGnet. Let's not castigate others for what you yourself do.

Regards,

Bruce

[From Rick Marken (970812.0755)]

Bruce Abbott (970812.0720 EST)

Bruce Abbott does not believe that Darwinian evolution is a
closed-loop control process, and never did...Bruce Abbott
believes that Darwinian evolution is a closed-loop, negative
feedback process...The same goes for Skinnerian reinforcement.

A closed-loop negative feedback process _is_ a control process.
Just solve the system equations and you'll see that this is
the case (or just see p.144-145 of LCS I). When the loop gain
in a negative feedback loop is relatively high (and in living
systems it is typically _very_ high_), the input (perceptual)
variable is made to match an explicit or tacit reference
specification for that variable.

If Darwinian evolution and Skinnerian reinforcement really are
negative feedback processes, then they are processes that bring
a perceptual representation of some environmental state of
affairs into a match with a fixed or variable reference state
and maintain it in that state against disturbance; that is, they
_control_. If Darwinians and Skinnerians really knew that their
models were negative feedback models then they would be testing
to determine whether the variables controlled by the models are
the same as those controlled by the modeled organisms -- that is,
they would be testing for controlled variables.

As far as I know, neither Darwinians nor Skinnerians have been
or are testing for controlled variables. This means either that
1) their model is really not a negative feedback process or
2) they don't understand how their own models works.

Me:

         1.0 1.0 1.0
genotypes--->phenotypes--> fitness <---disturbances
   ^ | 1.0
   > v
   -----------------------survival
                 1.0

Bruce:

I would like to enquire of Rick how he arrived at the fictional
gain values all around this loop.

I guessed. But I'm not married to this model. I was just trying
to show the _kind_ of systems analysis I would like to see in
order to determine whether Darwinian selection is, indeed, a
negative feedback process. You are the one who says that Darwinian
selection _is_ a negative feedback process so I think you should
post the equations that show that there is a net negative loop
gain in the Darwinian model.

Best

Rick

ยทยทยท

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

[From Bruce Gregory (970812.1320 EDT)]

Bruce Abbott (970812.1005 EST)

Your logic went as follows, as I recall: You couldn't see how it could
involve negative feedback, therefore one of us (I infer you meant me) must
not understand negative feedback.

I meant what I said.

This from a man who claimed in another
post that he likes to try to view an idea from another man's perspective
before making a decision. Sure you do. You never even _asked_ for my
reasoning. Your mind was already made up.

You win.

>Bruce Gregory (who _tries_ to keep his beliefs to himself on
>the assumption they are not all that interesting to others)

And fails. You have already asserted (without proof) that Darwinian
evolution does not involve negative feedback. That's a belief, and yet you
posted it on CSGnet. Let's not castigate others for what you yourself do.

Just for the record. I said

>I cannot construct a model of
>natural selection that embodies negative feedback. If you had
>told me that a seive is an example of negative feedback, I
>would understand how you are able to do so.

This does not constitute an assertion that Darwinian evolution
does not involve negative feedback as far as I am concerned. I
see we again differ.

Bruce