Mutations Or Natural Selection

in article [email hidden], Representative Trantis at
[email hidden] wrote on 10/9/03 6:18 PM:

Quoted message said:

There is a debate in evolutionary theory as to what is the most important factor in evolution.

1) The mutations which occur.

2) The fitness of such mutations.

This is described well in Richard Dawkins' 'Climbing Mount Improbable', in which he states that it
is possible that natural selection would love to allow a creature to develop in a particular way,
but can't because the appropriate mutations never arose. He also states that it is possible that
mutations aren't the limiting factor, rather it is that, for whatever reason, natural selection
doesn't 'keep' those mutations. He hints that he has an open mind on this one but leans slightly
towards number 2.

Quoted message said:

that it is in the interest of the other to develop a counter strategy. It's nearly always there
and the books tend to rattle off the appropriate

Quoted message said:

appropriate counter instincts are oftern there, can't help thinking that this strongly hints
towards the latter of the two in the debate, (The side Dawkins slightly favours).

Thoughts?

My thought is that I think you have posed a bad question. For example, mutation is completely
essential for evolution, so it is hard for me to imagine that ANYTHING is a more important factor in
evolution. I also think that evolution could not persist without the influence of a deterministic,
adaptive sub-process like natural selection. I don't know what it would mean to say that one
essential component is more or less important than another.

Guy

Representative Trantis <[email hidden]> wrote or quoted:

Quoted message said:

There is a debate in evolutionary theory as to what is the most important factor in evolution.

1) The mutations which occur.

2) The fitness of such mutations.

This is described well in Richard Dawkins' 'Climbing Mount Improbable', in which he states that it
is possible that natural selection would love to allow a creature to develop in a particular way,
but can't because the appropriate mutations never arose. He also states that it is possible that
mutations aren't the limiting factor, rather it is that, for whatever reason, natural selection
doesn't 'keep' those mutations. He hints that he has an open mind on this one but leans slightly
towards number 2.

C. M. I. - Chapter 6, I believe.

There /is/ a selection vs mutation debate - but that's more to do with the neutral theory - which is
not what Dawkins is talking about here.

For one thing, the "anti-selectionist" he describes has more than just mutation up his sleeve. In
particular there is also self-organisation.

Dawkins' chapter is about the ease with which one organisms can adapt new forms in response to
selection. He's discussing whether intermediate forms are probable or improbable. He's trying to
decide - if you selected for a flying pig - how easy it would be to get one.

It's a bit of a different debate to the mutation vs selection debate - which is usually discussed in
the context of Kimura's neutral theory of population genetics.

Neutral theory claimed that most evolutionary change is the result of genetic drift acting on
neutral alleles.

More recent experiments seem to have established that selection still has a substantial role to play
- e.g. see:

``The neutralist, the fly and the selectionist''

- timtyler.orgpmidOpen ↗

Selected loci - such as disease resistance genes - drag other linked parts of the genome along with
them - resulting in "substantial" levels of change being caused by selection.
--
__________
|im |yler timtyler.orgtimtyler.orgOpen ↗ [email hidden] Remove lock to reply.

In Guy Hoelzer said:

My thought is that I think you have posed a bad question. For example, mutation is completely
essential for evolution, so it is hard for me to imagine that ANYTHING is a more important factor
in evolution. I also think that evolution could not persist without the influence of a
deterministic, adaptive sub-process like natural selection. I don't know what it would mean to say
that one essential component is more or less important than another.

I think it's meaningful. Is copious variation maintained in populations by selection, always present
to supply whatever selection needs, or does the absence of variation and the appearance of new
mutations bias the direction of evolution? It could be argued that these are the issues that define
neo-Darwinism. Arlin Stolzfus' posts are useful, for example:
groups.google.comgroupsOpen ↗
groups.google.comgroupsOpen ↗
groups.google.comgroupsOpen ↗

Paul

Guy Hoelzer <[email hidden]> wrote in message news:<[email hidden]>...

Quoted message said:

in article [email hidden], Representative Trantis at [email hidden] wrote on 10/9/03
6:18 PM:

Quoted message said:

There is a debate in evolutionary theory as to what is the most important factor in evolution.

1) The mutations which occur.

2) The fitness of such mutations.

This is described well in Richard Dawkins' 'Climbing Mount Improbable', in which he states that
it is possible that natural selection would love to allow a creature to develop in a particular
way, but can't because the appropriate mutations never arose. He also states that it is possible
that mutations aren't the limiting factor, rather it is that, for whatever reason, natural
selection doesn't 'keep' those mutations. He hints that he has an open mind on this one but
leans slightly towards number 2.

Quoted message said:

Quoted message said:

that it is in the interest of the other to develop a counter strategy. It's nearly always there
and the books tend to rattle off the appropriate

Quoted message said:

Quoted message said:

appropriate counter instincts are oftern there, can't help thinking that this strongly hints
towards the latter of the two in the debate, (The side Dawkins slightly favours).

Thoughts?

My thought is that I think you have posed a bad question. For example, mutation is completely
essential for evolution, so it is hard for me to imagine that ANYTHING is a more important factor
in evolution. I also think that evolution could not persist without the influence of a
deterministic, adaptive sub-process like natural selection. I don't know what it would mean to say
that one essential component is more or less important than another.

Put another way, even the most strongly Darwinian arguments of human (and other) evolution tend to
restrict their strategy set to the realm of the physically possible. A genetic variant male who
could snap his fingers and have every female of his species impregnated with his offspring would
very quickly dominate the gene pool, but noone believes that such a gene is likely to arise any time
soon. Implicit in selective arguments are constraints, and among these constraints are the
availability of a genetic basis for hypothetical traits.

in article [email hidden], Paul Gallagher at
[email hidden] wrote on 10/11/03 9:03 AM:

Quoted message said:

In Guy Hoelzer said:

My thought is that I think you have posed a bad question. For example, mutation is completely
essential for evolution, so it is hard for me to imagine that ANYTHING is a more important factor
in evolution. I also think that evolution could not persist without the influence of a
deterministic, adaptive sub-process like natural selection. I don't know what it would mean to
say that one essential component is more or less important than another.

I think it's meaningful. Is copious variation maintained in populations by selection, always
present to supply whatever selection needs, or does the absence of variation and the appearance of
new mutations bias the direction of evolution? It could be argued that these are the issues that
define neo-Darwinism. Arlin Stolzfus' posts are useful, for example:
groups.google.comgroupsOpen ↗
groups.google.comgroupsOpen ↗
groups.google.comgroupsOpen ↗

These are perfectly good questions, but I don't see how they help to distinguish the relative
importances of essential factors. I am adopting a functionalist stance, and I think that all
functional systems require a balance between stochastic (creative) and deterministic (adaptive)
factors. To me saying something like "selection is more important to evolution than mutation" is
like saying that my circulatory system is more important to me than my digestive system.

Cheers,

Guy

Guy Hoelzer said:

in article [email hidden], Paul Gallagher at [email hidden] wrote on 10/11/03
9:03 AM:

Quoted message said:

In Guy Hoelzer said:

My thought is that I think you have posed a bad question. For example, mutation is completely
essential for evolution, so it is hard for me to imagine that ANYTHING is a more important
factor in evolution. I also think that evolution could not persist without the influence of a
deterministic, adaptive sub-process like natural selection. I don't know what it would mean to
say that one essential component is more or less important than another.

I think it's meaningful. Is copious variation maintained in populations by selection, always
present to supply whatever selection needs, or does the absence of variation and the appearance
of new mutations bias the direction of evolution? It could be argued that these are the issues
that define neo-Darwinism. Arlin Stolzfus' posts are useful, for example:
groups.google.comgroupsOpen ↗
groups.google.comgroupsOpen ↗
groups.google.comgroupsOpen ↗

These are perfectly good questions, but I don't see how they help to distinguish the relative
importances of essential factors. I am adopting a functionalist stance, and I think that all
functional systems require a balance between stochastic (creative) and deterministic (adaptive)
factors. To me saying something like "selection is more important to evolution than mutation" is
like saying that my circulatory system is more important to me than my digestive system.

Evolution is what it is. It seems to me a mistake to say there is this "thing", evolution, and ask
what is most important to it the way we might ask what is a critical path in a manufacturing
process. If it turns out in a particular case that there is a strong selective bias in favour of
some trait or gene, and that in another there is a lot of drift and stochastic variation, then both
of these cases are evolution. So too is stabilising selection, punctuated equilibria, and all the
other facets of the dynamics of evolution. I fail to see how any of these things can be "more
important" than another simply because at some point in any evolutionary process, it has happened. I
must be missing the point, though.

--
John Wilkins wilkins.id.au For long you live and high you fly, and smiles you'll give and tears
you'll cry and all you touch and all you see is all your life will ever be

In Guy Hoelzer said:

These are perfectly good questions, but I don't see how they help to distinguish the relative
importances of essential factors. I am adopting a functionalist stance, and I think that all
functional systems require a balance between stochastic (creative) and deterministic (adaptive)
factors. To me saying something like "selection is more important to evolution than mutation" is
like saying that my circulatory system is more important to me than my digestive system.

They are both peculiar questions -- they sound like value judgements, but that just means you need
to ask better questions.

Maybe this will help: "We need a conceptual framework that makes sense of what we already know, and
models that flesh out the implications of what we might already believe."
groups.google.comgroupsOpen ↗

It's meaningful to quantify what effects mutation and other mechanisms they have in actual
populations. (The importance you ascribe to these quantities is up to you.)

Apart from the issues that Arlin Stolzfus raised, what if someone were to propose this scenario:
point mutations and selection are all there is: point mutations appear and are immediately fixed or
eliminated by selection. Would you consider it uninteresting whether this model is or is not
correct? Obviously you're familiar with recombination, gene flow, hybridization, horizontal
transfer, different kinds of mutations, etc. Don't you think it's interesting to study to what
extent they give rise to new variation, and whether selection often creates and maintains variation?
Even if your sole interest is population response to selection, it should be clear that what kinds
and amounts of variation that are present in populations affect selection.

But back to Arlin Stolzfus' posts, the point is that many models do not properly represent the role
of the origin of variation: by implication, mutation is "less important." One thing you shouldn't
miss is that mutation itself can be a "deterministic factor," a cause of directional change. Also,
it should be obvious there are limitless "deterministic factors" that are not selection.

Do propensities in the origin of variation influence the process of evolution? You'll find people
who say no, and people who say yes. Stating that "all functional systems require a balance between
stochastic (mutation) and deterministic (selection) factors" doesn't even address the question. And
is it even true? I don't see why a predominantly deterministic system can't be functional and why a
"balance" is necessary. And what is the "functionalist stance"?

Many people have made a lot of effort to quantify the relative influence of different evolutionary
mechanisms in creating and maintaining variation, and it's of course possible to create models in
which different mechanisms contribute in different degrees to evolution.

Paul

in article [email hidden], John Wilkins at
[email hidden] wrote on 10/13/03 7:34 PM:

Quoted message said:

Guy Hoelzer said:

in article [email hidden], Paul Gallagher at [email hidden] wrote on 10/11/03
9:03 AM:

Quoted message said:

In <[email hidden]> Guy Hoelzer <[email hidden]> writes:

> My thought is that I think you have posed a bad question. For example, mutation is completely
> essential for evolution, so it is hard for me to imagine that ANYTHING is a more important
> factor in evolution. I also think that evolution could not persist without the influence of a
> deterministic, adaptive sub-process like natural selection. I don't know what it would mean to
> say that one essential component is more or less important than another.

I think it's meaningful. Is copious variation maintained in populations by selection, always
present to supply whatever selection needs, or does the absence of variation and the appearance
of new mutations bias the direction of evolution? It could be argued that these are the issues
that define neo-Darwinism. Arlin Stolzfus' posts are useful, for example:
groups.google.comgroupsOpen ↗
groups.google.comgroupsOpen ↗
groups.google.comgroupsOpen ↗

These are perfectly good questions, but I don't see how they help to distinguish the relative
importances of essential factors. I am adopting a functionalist stance, and I think that all
functional systems require a balance between stochastic (creative) and deterministic (adaptive)
factors. To me saying something like "selection is more important to evolution than mutation" is
like saying that my circulatory system is more important to me than my digestive system.

Evolution is what it is.

Indeed. I wouldn't try to argue with that.

Quoted message said:

It seems to me a mistake to say there is this "thing", evolution, and ask what is most important
to it the way we might ask what is a critical path in a manufacturing process.

Hmm. This seems like a deep difference between our views. I see a universe filled with entities and
processes, and these are often entangled because entities are always products of processes. By this
I mean more than products of past processes. I think that all physical entities are the products of
past and current processes, even if the current processes are merely slowing the rate of decay
(e.g., the tension of forces that hold a rock together). You may have noticed that I am fond of
pointing to logical connections between evolution as a process (a single processes that can also be
decomposed into sub-processes like natural selection, drift, and mutation) and weather systems as
processes. The main reason I like to do this is to emphasize that evolution is a physically
manifested process like a storm. I argue, for example, that a convection cell in the atmosphere is
both a process and a "thing" with agency (a cause of effects), and I think that evolution is in fact
a similar sort of "thing". I think that evolution is not separable from the biosphere in the same
sense that convection is separable from a convection cell.

Quoted message said:

If it turns out in a particular case that there is a strong selective bias in favour of some trait
or gene, and that in another there is a lot of drift and stochastic variation, then both of these
cases are evolution. So too is stabilising selection, punctuated equilibria, and all the other
facets of the dynamics of evolution. I fail to see how any of these things can be "more important"
than another simply because at some point in any evolutionary process, it has happened. I must be
missing the point, though.

That was my original point, exactly! You didn't copy the post that I was responding to above. That
post raised the question of which process, mutation or selection, was more important in evolution. I
argued that this was a bad question.

Cheers,

Guy

in article [email hidden], Paul Gallagher at
[email hidden] wrote on 10/14/03 8:16 AM:

Quoted message said:

In Guy Hoelzer said:

These are perfectly good questions, but I don't see how they help to distinguish the relative
importances of essential factors. I am adopting a functionalist stance, and I think that all
functional systems require a balance between stochastic (creative) and deterministic (adaptive)
factors. To me saying something like "selection is more important to evolution than mutation" is
like saying that my circulatory system is more important to me than my digestive system.

They are both peculiar questions -- they sound like value judgements, but that just means you need
to ask better questions.

Yup.

Quoted message said:

Maybe this will help: "We need a conceptual framework that makes sense of what we already know,
and models that flesh out the implications of what we might already believe."
groups.google.comgroupsOpen ↗

This is a very nice quote.

Quoted message said:

It's meaningful to quantify what effects mutation and other mechanisms they have in actual
populations. (The importance you ascribe to these quantities is up to you.)

Apart from the issues that Arlin Stolzfus raised, what if someone were to propose this scenario:
point mutations and selection are all there is: point mutations appear and are immediately fixed
or eliminated by selection. Would you consider it uninteresting whether this model is or is not
correct? Obviously you're familiar with recombination, gene flow, hybridization, horizontal
transfer, different kinds of mutations, etc. Don't you think it's interesting to study to what
extent they give rise to new variation, and whether selection often creates and maintains
variation? Even if your sole interest is population response to selection, it should be clear that
what kinds and amounts of variation that are present in populations affect selection.

But back to Arlin Stolzfus' posts, the point is that many models do not properly represent
the role of the origin of variation: by implication, mutation is "less important." One thing
you shouldn't miss is that mutation itself can be a "deterministic factor," a cause of
directional change.

This is a good point. I often try to emphasize, for example, that random (with regard to fitness)
mutation tends to decrease fitness in well-designed (adaptive) systems.

Quoted message said:

Also, it should be obvious there are limitless "deterministic factors" that are not selection.

I guess, but I don't see the point. I didn't explain myself very thoroughly, so I can understand how
I might have been misinterpreted.

Natural selection, like all mechanisms of self-organization, are both deterministic and lead to
adaptive change in the structure of the system. So, what I was trying to say is that all such
systems must balance stochastic (both creative and destructive) factors with adaptive (function-
improving, deterministic, and codifying) factors.

Quoted message said:

Do propensities in the origin of variation influence the process of evolution? You'll find people
who say no, and people who say yes. Stating that "all functional systems require a balance between
stochastic (mutation) and deterministic (selection) factors" doesn't even address the question.

True, but why should it?

Quoted message said:

And is it even true? I don't see why a predominantly deterministic system can't be functional and
why a "balance" is necessary. And what is the "functionalist stance"?

I recommend reading one of Stan Salthe's books for a thorough explanation of this view. To summarize
the issue, all complex adaptive systems (biological and otherwise) are thought to undergo the same
sort of life-history transitions we are familiar with from the study of organisms. Salthe
categorizes these transitions in three stages: the developmental stage, the mature stage, and the
senescent stage. The relative influences of stochastic vs. deterministic factors change across the
life history, with monotonically increasing degrees of determinism causing canalization and reduced
potential for adaptive response to environmental novelty. Senescence (of an organism, species,
weather system, ...) makes systems increasingly vulnerable to breaking (death, extinction, ...).

If you are familiar with the variety of theoretical approaches to understanding dynamical systems
you might identify chaos with domination of stochastic influences and equilibrium with the
domination of deterministic influences. Neither kind of system can persist for long in nature.
Permanently chaotic systems cannot respond adaptively to the external environment, because they
cannot incorporate the external information into structural responses. Permanently deterministic
systems can't break out of their equilibrium or generate sufficiently novel responses to novelties
in the environment. Per Bak developed a theory that all complex adaptive systems find their way to
the "edge of chaos" where they fluctuate between chaotic and deterministic realms, achieving much
greater potential for adaptive change. More recent work suggests that the necessary balance is
struck in the process of self-organization independently of the state of the system (it need not be
poised at the edge of chaos).

The "functionalist stance" holds a teleological view of all dynamical complex adaptive systems. The
theoretical rationale for natural teleology (real, not imagined teleology in natural systems) is
based on the second law of thermodynamics (some, like Stuart Kaufmann are arguing for a fourth law
to cover natural teleology). Ultimately, all such systems become manifested because they increase
the rate of entropy increase at the scale of the universe as a whole, without exception.

Quoted message said:

Many people have made a lot of effort to quantify the relative influence of different evolutionary
mechanisms in creating and maintaining variation, and it's of course possible to create models in
which different mechanisms contribute in different degrees to evolution.

Sure, but if your models don't have a proper balance they won't actually predict persistent
evolution, which I think we can all agree is what actually happens. Eternal equilibrium of structure
is not evolution. My favorite way of checking whether your model actually predicts persistent
evolution is to run it as a numerical computation. I predict that any such exploration of an
putative evolutionary model will stop evolving (equilibrium) or die out rather quickly.

Cheers,

Guy

Guy Hoelzer said:

Quoted message said:

It seems to me a mistake to say there is this "thing", evolution, and ask what is most important
to it the way we might ask what is a critical path in a manufacturing process.

Hmm. This seems like a deep difference between our views. I see a universe filled with entities
and processes, and these are often entangled because entities are always products of processes. By
this I mean more than products of past processes. I think that all physical entities are the
products of past and current processes, even if the current processes are merely slowing the rate
of decay (e.g., the tension of forces that hold a rock together). You may have noticed that I am
fond of pointing to logical connections between evolution as a process (a single processes that
can also be decomposed into sub-processes like natural selection, drift, and mutation) and weather
systems as processes. The main reason I like to do this is to emphasize that evolution is a
physically manifested process like a storm. I argue, for example, that a convection cell in the
atmosphere is both a process and a "thing" with agency (a cause of effects), and I think that
evolution is in fact a similar sort of "thing". I think that evolution is not separable from the
biosphere in the same sense that convection is separable from a convection cell.

For a deep division we agree extraordinarily.

But is "weather" a thing? If so, does it include tectonic processes? Vulcanism? Bolide impacts? All
these things impact (sorry :-) on weather processes, so why not. How about Melankovitch cycles <sp?>
My point is that something like "weather" is a convention name we give to a range of phenomena that
suit us to collect under one term. It is not a "thing" or "process" in and of itself. Evolution is
like that. So properties of that non-thing are empty, or they are properties of particular things
and processes we would do better to investigate less generally and not go looking for
universalisations like "adaptation is the most important aspect of evolution", etc.

Hull somewhere says there are exceptions to every rule in biology (including that one).

--
John Wilkins wilkins.id.au For long you live and high you fly, and smiles you'll give and tears
you'll cry and all you touch and all you see is all your life will ever be

In Guy Hoelzer said:

Quoted message said:

implication, mutation is "less important." One thing you shouldn't miss is that mutation itself
can be a "deterministic factor," a cause of directional change.

Quoted message said:

This is a good point. I often try to emphasize, for example, that random (with regard to fitness)
mutation tends to decrease fitness in well-designed (adaptive) systems.

I was thinking in particular of mutation bias, rather than mutation load.

Quoted message said:

Natural selection, like all mechanisms of self-organization, are both deterministic and lead to
adaptive change in the structure of the system. So, what I was trying to say is that all such
systems must balance stochastic (both creative and destructive) factors with adaptive (function-
improving, deterministic, and codifying) factors.

That seems reasonable. I did go off on a tangent.

But I still think it's reasonable to ask whether, so to speak, appropriate variation will be
available for selection to produce a better adapted phenotype -- which was Representative Trantis'
original question, based on Dawkins' "Climbing Mount Improbable." I think this is related to many
issues of interest: if the widespread assumption of abundant, small and continuous, non-directed
variation is wrong, it makes a difference, and it invalidates many models of evolutionary change
that presume the sufficiency of selectionist explanations.

Dawkins may phrase the question badly, and I think I may have echoed his errors. For example, here
is Arlin Stolzfus again:
[T]he metaphor of constraints is usually based on the implicit notion of omnipotent selection,
though the idea can scarcely be defended in any given case because the problem space is not
defined clearly enough. In some interpretations it leads to unfathomable contradictions, like
the idea that evolution would somehow go wherever selection would take it if it were not for
that constraining process of variation!

The rest of his post is here:
groups.google.comgroupsOpen ↗

Also, see: groups.google.comgroupsOpen ↗

Paul

in article [email hidden], John Wilkins at
[email hidden] wrote on 10/14/03 10:27 PM:

Quoted message said:

Guy Hoelzer said:

Quoted message said:

It seems to me a mistake to say there is this "thing", evolution, and ask what is most important
to it the way we might ask what is a critical path in a manufacturing process.

Hmm. This seems like a deep difference between our views. I see a universe filled with entities
and processes, and these are often entangled because entities are always products of processes.
By this I mean more than products of past processes. I think that all physical entities are the
products of past and current processes, even if the current processes are merely slowing the rate
of decay (e.g., the tension of forces that hold a rock together). You may have noticed that I am
fond of pointing to logical connections between evolution as a process (a single processes that
can also be decomposed into sub-processes like natural selection, drift, and mutation) and
weather systems as processes. The main reason I like to do this is to emphasize that evolution is
a physically manifested process like a storm. I argue, for example, that a convection cell in the
atmosphere is both a process and a "thing" with agency (a cause of effects), and I think that
evolution is in fact a similar sort of "thing". I think that evolution is not separable from the
biosphere in the same sense that convection is separable from a convection cell.

For a deep division we agree extraordinarily.

But is "weather" a thing?

I would say not, but I would say that a particular storm is a thing. I see the biosphere as a storm
that has persisted for billions of years.

Quoted message said:

If so, does it include tectonic processes? Vulcanism? Bolide impacts? All these things impact
(sorry :-) on weather processes, so why not.

Again, I would say not. I see tectonics as part of a geological process that is functionally
uncoupled from the weather. One thing that reveals the uncoupled nature of these processes is that
the operate on very different time scales. Note that functionally decoupled processes can interact,
just as two different organisms (which are functionally decoupled entities) can interact, and such
interactions among entities can contribute to the functional organization of some particular
functional process/entity at a larger scale (e.g., consider the GAIA hypothesis).

Quoted message said:

How about Melankovitch cycles <sp?> My point is that something like "weather" is a convention name
we give to a range of phenomena that suit us to collect under one term.

You could be right about weather as a general category. I am much more comfortable thinking about
individual storms as process/entities that have individuated and have agency. I am open to
considering the global weather in this way (see below), but it is not as clear to me that global
weather is a single, functional process/entity.

Quoted message said:

It is not a "thing" or "process" in and of itself. Evolution is like that. So properties of that
non-thing are empty, or they are properties of particular things and processes we would do better
to investigate less generally and not go looking for universalisations like "adaptation is the
most important aspect of evolution", etc.

Your position is certainly arguable. I would not dismiss it. However, my view is different, and I am
convinced that my view is also not so easily dismissed. It may be that the functional coherence of
evolution comes and goes, much like the functional coherence of weather systems. We wouldn't say
that the dissipation of a convection cell is tantamount to the end of weather. There are top-down
global weather effects caused by interactions among weather systems and the cohesion of the
atmosphere that maintain wide-scale distributions of weather patterns across regions of the globe.
Evolution is likely to work in a similar way, IMHO. Evolution storms (most likely coevolution
storms) can spring up in a rush of coordinated self-organization, which eventually run there course
and dissipate. I think that there are also top-down effects that coordinate such local activities
(e.g., maintenance of food web structures, tendencies toward niche filling, ...), which is basically
the GAIA idea. So, my model has evolution as a mix of chaos (e.g., the effects of mutation and
drift) and self-organization, which can manifest locally in time and space at several different
scales simultaneously.

The good news is that I don't think the difference in our views is just a matter of interpretation
or airy fairy ideas. Critical evaluation of agency is something that science is potentially good at,
because agency is about functional coherence (self-sustaining dynamics with positive feedbacks) and
external effects that should be measurable. I am optimistic that we will develop a science of
assigning agency in the near future (a decade or two?). Such a science would be able to assess, for
example, whether global weather is or is not a "thing".

Quoted message said:

Hull somewhere says there are exceptions to every rule in biology (including that one).

Meaning that there are rules without exceptions? This has essentially been proven by a number
of physicists, three of whom won the Nobel prize, with the development of renormalization
group analysis.

Cheers,

Guy

Paul Gallagher <[email hidden]> wrote in
:"]news:[email hidden]:

Quoted message said:

But I still think it's reasonable to ask whether, so to speak, appropriate variation will be
available for selection to produce a better adapted phenotype -- which was Representative Trantis'
original question, based on Dawkins' "Climbing Mount Improbable." I think this is related to many
issues of interest: if the widespread assumption of abundant, small and continuous, non-directed
variation is wrong, it makes a difference, and it invalidates many models of evolutionary change
that presume the sufficiency of selectionist explanations.

There was an interesting snippet in the November 2003 Discover that described research indicating
that there are constraints on where mutations occur, based on a comparison of the mouse and human
genome. Unfortunately the article was too short to follow the details of the research, and I have
not yet seen a reference elsewhere. But I think this speaks to your point (and one that Arlin used
to make when he was still contributing to the newsgroup) that the details of the biochemistry will
become more important as we refine our models of evolution.

Yours,

Bill Morse

Guy Hoelzer said:

in article [email hidden], John Wilkins at [email hidden] wrote on
10/14/03 10:27 PM:

Quoted message said:

Guy Hoelzer said:

> It seems to me a mistake to say there is this "thing", evolution, and ask what is most
> important to it the way we might ask what is a critical path in a manufacturing process.

Hmm. This seems like a deep difference between our views. I see a universe filled with entities
and processes, and these are often entangled because entities are always products of processes.
By this I mean more than products of past processes. I think that all physical entities are the
products of past and current processes, even if the current processes are merely slowing the
rate of decay (e.g., the tension of forces that hold a rock together). You may have noticed
that I am fond of pointing to logical connections between evolution as a process (a single
processes that can also be decomposed into sub-processes like natural selection, drift, and
mutation) and weather systems as processes. The main reason I like to do this is to emphasize
that evolution is a physically manifested process like a storm. I argue, for example, that a
convection cell in the atmosphere is both a process and a "thing" with agency (a cause of
effects), and I think that evolution is in fact a similar sort of "thing". I think that
evolution is not separable from the biosphere in the same sense that convection is separable
from a convection cell.

For a deep division we agree extraordinarily.

But is "weather" a thing?

I would say not, but I would say that a particular storm is a thing. I see the biosphere as a
storm that has persisted for billions of years.

Well in one way you might be right - it is a phenomenal individual - that is to say, we could locate
it in time and space and say "look at what the surface of that planet is doing" (had we but time
enough and space). But my question was alittle more specific - is it a single cohesive individual?
Is it, in short, a single system? Is it one process?

Now for both weather and evolution, I think it is a series of occasionally coupled systems, and
mostly decoupled. While what happens in the Deccan Traps or Krakatoa or Chixulub or the breakup
of Gondwana *affect* evolution and the weather, the latter are just that - effects. So far as I
know, none of the geological or astronomical phenomena are caused by either the weather or the
evolution of organisms. So I would *not* say that everything that happens *in* weather or
evolution is part of it.

But even within those two domains - while some aspects of weather affect other aspects - many things
happen in weather than dampen out in larger scales, and so a dustdevil in central Australia and a
snowstorm in Antarctica are decoupled. Likewise, the evolution of a fungus in the Amazon and the
evolution of a bee in Europe are decoupled (until they do meet and affect each other's environment).
So I would also not say that all evolution or all weather is a single cohesive syste.

Hence I would not say that there has to be a property common to all instances of evolution or
weather, because there is no property-bearer, no single "thing".

Quoted message said:

Quoted message said:

If so, does it include tectonic processes? Vulcanism? Bolide impacts? All these things impact
(sorry :-) on weather processes, so why not.

Again, I would say not. I see tectonics as part of a geological process that is functionally
uncoupled from the weather. One thing that reveals the uncoupled nature of these processes is that
the operate on very different time scales. Note that functionally decoupled processes can
interact, just as two different organisms (which are functionally decoupled entities) can
interact, and such interactions among entities can contribute to the functional organization of
some particular functional process/entity at a larger scale (e.g., consider the GAIA hypothesis).

But now it is open to me to say that weather *does* affect geology, through erosion and deposition.
Where do you draw that line?

Quoted message said:

Quoted message said:

How about Melankovitch cycles <sp?> My point is that something like "weather" is a convention
name we give to a range of phenomena that suit us to collect under one term.

You could be right about weather as a general category. I am much more comfortable thinking about
individual storms as process/entities that have individuated and have agency. I am open to
considering the global weather in this way (see below), but it is not as clear to me that global
weather is a single, functional process/entity.

Then why claim the same for the formally analogous process of evolution? Saying that things
evolve and things climatically and atmospherically change doesn't imply that there is an
"evolving" nor a "weathering", so why would we expect to find that there is a "normal" rate or
mode of evolution anyway?

I believe that we have always given into the entification (as Henry Plotkin once called it) of
evolution, and we suffer from that. It causes teleology - we expect to find norms and functions in
things that are entities (cohesive individuals).

Quoted message said:

Quoted message said:

It is not a "thing" or "process" in and of itself. Evolution is like that. So properties of that
non-thing are empty, or they are properties of particular things and processes we would do
better to investigate less generally and not go looking for universalisations like "adaptation
is the most important aspect of evolution", etc.

Your position is certainly arguable. I would not dismiss it. However, my view is different, and I
am convinced that my view is also not so easily dismissed. It may be that the functional coherence
of evolution comes and goes, much like the functional coherence of weather systems. We wouldn't
say that the dissipation of a convection cell is tantamount to the end of weather. There are top-
down global weather effects caused by interactions among weather systems and the cohesion of the
atmosphere that maintain wide-scale distributions of weather patterns across regions of the globe.
Evolution is likely to work in a similar way, IMHO. Evolution storms (most likely coevolution
storms) can spring up in a rush of coordinated self-organization, which eventually run there
course and dissipate. I think that there are also top-down effects that coordinate such local
activities
(e.g., maintenance of food web structures, tendencies toward niche filling, ...), which is
basically the GAIA idea. So, my model has evolution as a mix of chaos (e.g., the effects of
mutation and drift) and self-organization, which can manifest locally in time and space at
several different scales simultaneously.

Gaia is not a counterexample to my view, so long as it can be empirically demonstrated. After all,
if Hamilton entertained it, then it is eminently entertainable. But it has to be shown to occur. If
there is top-down causation, then there is, but I'm from Australia - show me. Hamilton's idea that
algal spores form clouds to disseminate the spores is a Gaianesque model - but one that can be
tested. The idea that ther eis "Evolution" because there must be some top-down causation is argument
by assertion unless we get specifics. It seems to me that evolutionary processes will be much more
interesting if we can find out how they *actually* work; more likely there will be a mix of coupled,
uncoupled, level-restricted and multilevel processes, and I want to know what they are, not what we
think they will be.

Hamilton, W. D. , and T. M. Lenton. 1998. Spora and Gaia - how microbes fly with their clouds.
Ethology Ecology and Evolution 10 (1):1-16.

Quoted message said:

The good news is that I don't think the difference in our views is just a matter of interpretation
or airy fairy ideas. Critical evaluation of agency is something that science is potentially good
at, because agency is about functional coherence (self-sustaining dynamics with positive
feedbacks) and external effects that should be measurable. I am optimistic that we will develop a
science of assigning agency in the near future (a decade or two?). Such a science would be able to
assess, for example, whether global weather is or is not a "thing".

I oppose agency in all cases that do not involve agents. With intentions. And internal models of the
world. The temptation to project our own agency on the world is as old as flok religion.

Quoted message said:

Quoted message said:

Hull somewhere says there are exceptions to every rule in biology (including that one).

Meaning that there are rules without exceptions? This has essentially been proven by a number of
physicists, three of whom won the Nobel prize, with the development of renormalization group
analysis.

Exceptional.

--
John Wilkins wilkins.id.au For long you live and high you fly, and smiles you'll give and tears
you'll cry and all you touch and all you see is all your life will ever be

In William Morse said:

There was an interesting snippet in the November 2003 Discover that described research indicating
that there are constraints on where mutations occur, based on a comparison of the mouse and human
genome. Unfortunately the article was too short to follow the details of the research, and I have
not yet seen a reference elsewhere. But I think this speaks to your point (and one that Arlin used
to make when he was still contributing to the newsgroup) that the details of the biochemistry will
become more important as we refine our models of evolution.

That was published in PNAS (June 24, 2003) 100(13): 7672-7677.

The abstract is online: pnas.org7672Open ↗

Also, see: calit2.net10 10 pevzner.htmlOpen ↗

It would be nice to have Arlin Stolzfus back to explain it!

Paul

in article [email hidden], John Wilkins at
[email hidden] wrote on 10/17/03 8:33 AM:

Quoted message said:

Guy Hoelzer said:

in article [email hidden], John Wilkins at [email hidden] wrote on
10/14/03 10:27 PM:

Quoted message said:

Guy Hoelzer <[email hidden]> wrote:

>> It seems to me a mistake to say there is this "thing", evolution, and ask what is most
>> important to it the way we might ask what is a critical path in a manufacturing process.
>
> Hmm. This seems like a deep difference between our views. I see a universe filled with entities
> and processes, and these are often entangled because entities are always products of processes.
> By this I mean more than products of past processes. I think that all physical entities are the
> products of past and current processes, even if the current processes are merely slowing the
> rate of decay (e.g., the tension of forces that hold a rock together). You may have noticed
> that I am fond of pointing to logical connections between evolution as a process (a single
> processes that can also be decomposed into sub-processes like natural selection, drift, and
> mutation) and weather systems as processes. The main reason I like to do this is to emphasize
> that evolution is a physically manifested process like a storm. I argue, for example, that a
> convection cell in the atmosphere is both a process and a "thing" with agency (a cause of
> effects), and I think that evolution is in fact a similar sort of "thing". I think that
> evolution is not separable from the biosphere in the same sense that convection is separable
> from a convection cell.

For a deep division we agree extraordinarily.

But is "weather" a thing?

I would say not, but I would say that a particular storm is a thing. I see the biosphere as a
storm that has persisted for billions of years.

Well in one way you might be right - it is a phenomenal individual - that is to say, we could
locate it in time and space and say "look at what the surface of that planet is doing" (had we but
time enough and space). But my question was alittle more specific - is it a single cohesive
individual? Is it, in short, a single system? Is it one process?

That is my position. Individual tornados, hurricane, and convection cells are each single systems in
the same sense that you and I are single systems.

Quoted message said:

Now for both weather and evolution, I think it is a series of occasionally coupled systems, and
mostly decoupled.

I think I agree, but I if I am interpreting your comment correctly, I would also argue that all
coherent systems can be statistically decomposed in this way. You, for example, can be decomposed
into organ systems, organs, or cells, each of which exhibits much more tightly coupled dynamics of
its subcomponents than the same type of subcomponents exhibit at the scale of you as a whole. IMHO,
this does not make you a mere epiphenomenon giving the mere illusion of coherent function. Your
intelligent posts reveal otherwise.

Quoted message said:

While what happens in the Deccan Traps or Krakatoa or Chixulub or the breakup of Gondwana
*affect* evolution and the weather, the latter are just that - effects. So far as I know, none
of the geological or astronomical phenomena are caused by either the weather or the evolution
of organisms. So I would *not* say that everything that happens *in* weather or evolution is
part of it.

I think I agree here, although I am certain you would find that effects go both ways if you examined
things in sufficient detail. I would agree that the flow of effects is highly asymmetrical in these
comparisons.

Quoted message said:

But even within those two domains - while some aspects of weather affect other aspects - many
things happen in weather than dampen out in larger scales, and so a dustdevil in central Australia
and a snowstorm in Antarctica are decoupled.

Exactly. This supports my contention that each of these dynamical, self-supporting (containing
internal, positive feedbacks) structures in the atmosphere are individuated things. This conclusion
does not IMHO preclude the possibility of a coherent global weather system, too.

Quoted message said:

Likewise, the evolution of a fungus in the Amazon and the evolution of a bee in Europe are
decoupled (until they do meet and affect each other's environment). So I would also not say that
all evolution or all weather is a single cohesive syste.

I agree with your premise, but not your conclusion. My logic is consistent with the dynamical
networks paradigm (e.g., small world graphs) and hierarchy theory sensu Stan Salthe, if you are
familiar with them. Graph theory and hierarchy theory are both largely about nested entities.

Quoted message said:

Hence I would not say that there has to be a property common to all instances of evolution or
weather, because there is no property-bearer, no single "thing".

I agree that there "would not [have] to be a property common to all instances of evolution or
weather," but disagree as to whether or not there is. I would caution you here to avoid the common
mistake (IMHO) of failing to recognize the dynamical and coherent nature of systems that function on
very different time and spatial scales than those of our existence. For example, most scientists
would quickly accept the notion that a rock is not a dynamical system. In fact, it is, although it
is in a senescent stage of degradation. Its structure, however, is maintained by the tension caused
by attractive and repulsive forces among its component particles. The time scale of functional
activity in evolution is far slower than the time scales at which we exist, which can obfuscate its
functional coherence, if I am right that it exists.

Quoted message said:

Quoted message said:

Quoted message said:

If so, does it include tectonic processes? Vulcanism? Bolide impacts? All these things impact
(sorry :-) on weather processes, so why not.

Again, I would say not. I see tectonics as part of a geological process that is functionally
uncoupled from the weather. One thing that reveals the uncoupled nature of these processes is
that the operate on very different time scales. Note that functionally decoupled processes can
interact, just as two different organisms (which are functionally decoupled entities) can
interact, and such interactions among entities can contribute to the functional organization of
some particular functional process/entity at a larger scale (e.g., consider the GAIA hypothesis).

But now it is open to me to say that weather *does* affect geology, through erosion and
deposition. Where do you draw that line?

All real systems are open, hence effects enter from the outside as perturbations and effects leak
out to perturb other systems. Showing some form of effective interaction between agents (note that I
don't mean to imply anything about intentionality) does not undermine their existence as
functionally coherent systems. Nature draws the line by causing systemic organization and
compartmentalization, meaning that there is a clear (but never absolute) distinction between the
extent of "particle" interaction within a system compared with between systems. I view it as perhaps
the most fundamental role of all empirical scientists to identify the natural bounds of real systems
before they target someTHING for study. This can be easy and familiar (e.g., the bounds of an
individual mammal) or hard and abstract (e.g., the bounds of an ecosystem).

Quoted message said:

Quoted message said:

Quoted message said:

How about Melankovitch cycles <sp?> My point is that something like "weather" is a convention
name we give to a range of phenomena that suit us to collect under one term.

You could be right about weather as a general category. I am much more comfortable thinking about
individual storms as process/entities that have individuated and have agency. I am open to
considering the global weather in this way (see below), but it is not as clear to me that global
weather is a single, functional process/entity.

Then why claim the same for the formally analogous process of evolution?

Perhaps I have overstated my case to make it. I would not argue that my position is so irrefutable
or sufficiently tested that it should be universally considered scientific "fact" at this point.
Consider it a hypothesis, which I think is highly likely to be a good one for both atmosphere and
biosphere. I would also argue that the validity of my hypothesis for evolution does not rise or fall
on its validity for weather, although I do think that these contexts seem to be formally the same.

Quoted message said:

Saying that things evolve and things climatically and atmospherically change doesn't imply that
there is an "evolving" nor a "weathering", so why would we expect to find that there is a "normal"
rate or mode of evolution anyway?

I would not expect a "normal" rate or mode of evolution any more than I would expect a "normal" rate
or mode of weather. Similarly, I would not expect, and we do not see, a "normal" rate or mode of
biological existence. However, I would expect universal "metabolic" scaling properties to hold for
global weather and the biosphere, in the same why folks like Brown and West are exploring for
organisms and ecosystems.

Quoted message said:

I believe that we have always given into the entification (as Henry Plotkin once called it) of
evolution, and we suffer from that. It causes teleology - we expect to find norms and functions in
things that are entities (cohesive individuals).

That¹s funny, because I think we have shackled ourselves as scientists with notions like Skinnerian
behaviorism and avoidance of anything that smacks of anthropomorphism. In fact, I am hoping for a
resurgence of the vary things you feel we suffer from. I argue, for example, that the second law of
thermodynamics implies the existence of natural teleology; that is, nature itself is teleological,
and we must accept teleological explanations of natural phenomena if we are to understand them
correctly. I understand why the "shackles" were introduced to science, which I attribute to the
potential for generating uncritical and unconstrained inferences. If I am right about natural
teleology it will be important to figure out a way to embrace teleological explanations without
opening Pandora's box. I like the potential for the framework of thermodynamics to constrain the
kinds of teleological hypotheses and inferences we can accept.

Quoted message said:

Quoted message said:

Quoted message said:

It is not a "thing" or "process" in and of itself. Evolution is like that. So properties of that
non-thing are empty, or they are properties of particular things and processes we would do
better to investigate less generally and not go looking for universalisations like "adaptation
is the most important aspect of evolution", etc.

Your position is certainly arguable. I would not dismiss it. However, my view is different, and I
am convinced that my view is also not so easily dismissed. It may be that the functional
coherence of evolution comes and goes, much like the functional coherence of weather systems. We
wouldn't say that the dissipation of a convection cell is tantamount to the end of weather. There
are top-down global weather effects caused by interactions among weather systems and the cohesion
of the atmosphere that maintain wide-scale distributions of weather patterns across regions of
the globe. Evolution is likely to work in a similar way, IMHO. Evolution storms (most likely
coevolution storms) can spring up in a rush of coordinated self-organization, which eventually
run there course and dissipate. I think that there are also top-down effects that coordinate such
local activities
(e.g., maintenance of food web structures, tendencies toward niche filling, ...), which is
basically the GAIA idea. So, my model has evolution as a mix of chaos (e.g., the effects of
mutation and drift) and self-organization, which can manifest locally in time and space at
several different scales simultaneously.

Gaia is not a counterexample to my view, so long as it can be empirically demonstrated. After all,
if Hamilton entertained it, then it is eminently entertainable. But it has to be shown to occur.
If there is top-down causation, then there is, but I'm from Australia - show me.

Do you accept the notion of top-down causation (or more likely pattern-forming constraints that
facilitate global functioning) in general, or are you skeptical that they exist? What do you think
of the daisy world model?

Quoted message said:

Hamilton's idea that algal spores form clouds to disseminate the spores is a Gaianesque model -
but one that can be tested. The idea that ther eis "Evolution" because there must be some top-down
causation is argument by assertion unless we get specifics.

I have never argued that there "must be some top-down causation" for the biosphere or any other
putative system. I do argue that there "must be some top-down causation" for a real system to exist.
The question, as you have argued above, is whether or not the putative system is a real one.

Quoted message said:

It seems to me that evolutionary processes will be much more interesting if we can find out how
they *actually* work; more likely there will be a mix of coupled, uncoupled, level-restricted and
multilevel processes, and I want to know what they are, not what we think they will be.

OK, but I am sure that you recognize finding out what they are depends heavily on what we think
they will be.

Quoted message said:

Hamilton, W. D. , and T. M. Lenton. 1998. Spora and Gaia - how microbes fly with their clouds.
Ethology Ecology and Evolution 10 (1):1-16.

Quoted message said:

The good news is that I don't think the difference in our views is just a matter of
interpretation or airy fairy ideas. Critical evaluation of agency is something that science is
potentially good at, because agency is about functional coherence (self-sustaining dynamics with
positive feedbacks) and external effects that should be measurable. I am optimistic that we will
develop a science of assigning agency in the near future (a decade or two?). Such a science would
be able to assess, for example, whether global weather is or is not a "thing".

I oppose agency in all cases that do not involve agents. With intentions. And internal models of
the world. The temptation to project our own agency on the world is as old as flok religion.

We have quite different semantics regarding agency, so we will have to be careful in using this
word. I don't know of anything resembling a synonym for my use of agency, but it seems like
intentionality is close to a synonym for your use. Am I correct?

Cheers,

Guy

Guy Hoelzer said:

in article [email hidden], John Wilkins at [email hidden] wrote on
10/17/03 8:33 AM:

Quoted message said:

Guy Hoelzer said:

in article [email hidden], John Wilkins at [email hidden] wrote on
10/14/03 10:27 PM:

> Guy Hoelzer <[email hidden]> wrote:
>
>>> It seems to me a mistake to say there is this "thing", evolution, and ask what is most
>>> important to it the way we might ask what is a critical path in a manufacturing process.
>>
>> Hmm. This seems like a deep difference between our views. I see a universe filled with
>> entities and processes, and these are often entangled because entities are always products of
>> processes. By this I mean more than products of past processes. I think that all physical
>> entities are the products of past and current processes, even if the current processes are
>> merely slowing the rate of decay (e.g., the tension of forces that hold a rock together). You
>> may have noticed that I am fond of pointing to logical connections between evolution as a
>> process (a single processes that can also be decomposed into sub-processes like natural
>> selection, drift, and mutation) and weather systems as processes. The main reason I like to
>> do this is to emphasize that evolution is a physically manifested process like a storm. I
>> argue, for example, that a convection cell in the atmosphere is both a process and a "thing"
>> with agency (a cause of effects), and I think that evolution is in fact a similar sort of
>> "thing". I think that evolution is not separable from the biosphere in the same sense that
>> convection is separable from a convection cell.
>
> For a deep division we agree extraordinarily.
>
> But is "weather" a thing?

I would say not, but I would say that a particular storm is a thing. I see the biosphere as a
storm that has persisted for billions of years.

Well in one way you might be right - it is a phenomenal individual - that is to say, we could
locate it in time and space and say "look at what the surface of that planet is doing" (had we
but time enough and space). But my question was alittle more specific - is it a single cohesive
individual? Is it, in short, a single system? Is it one process?

That is my position. Individual tornados, hurricane, and convection cells are each single systems
in the same sense that you and I are single systems.

But I am not a single system in some ways. I am dependent on the massive number of endosymbionts I
support. I require a constant environment with a ready source of food. I must live, because it is an
autapomorphy of my species, in society.

Quoted message said:

Quoted message said:

Now for both weather and evolution, I think it is a series of occasionally coupled systems, and
mostly decoupled.

I think I agree, but I if I am interpreting your comment correctly, I would also argue that all
coherent systems can be statistically decomposed in this way. You, for example, can be decomposed
into organ systems, organs, or cells, each of which exhibits much more tightly coupled dynamics of
its subcomponents than the same type of subcomponents exhibit at the scale of you as a whole.
IMHO, this does not make you a mere epiphenomenon giving the mere illusion of coherent function.
Your intelligent posts reveal otherwise.

Flatterer. But you still don't tell me how we can decompose systems such that I can tell if
something, some subsystem, is a "thing" or not.

This is an old and somewhat sophomoric (soporific?) problem in philosophy. We are asking, in effect,
how far does an explanation need to go. If we draw a line at some point (say, my skin or my local
ecosystem, and so on) we are entitled to ask, why there? If it turns out that we draw the line there
for convenience' sake, then we have only asserted some fact about us as observers.

My objection to treating evolution, or weather, as a single thing is that there is none of the
cohesion required to make either of them a functional or cohesive individual. We draw those lines
because there is a reason, a theoretical reason, to do so. I am merely saying that while a tornado
might be a cohesive individual, since it is a dissipative structure, weather is not. Likewise, while
we might treat the speciation of a mammal species into two as a single process (and can ask sensibly
what the mode, frequency and mechanisms were in each case), we can't ask such questions of
"evolution as a whole".

Quoted message said:

Quoted message said:

While what happens in the Deccan Traps or Krakatoa or Chixulub or the breakup of Gondwana
*affect* evolution and the weather, the latter are just that - effects. So far as I know, none
of the geological or astronomical phenomena are caused by either the weather or the evolution of
organisms. So I would *not* say that everything that happens *in* weather or evolution is part
of it.

I think I agree here, although I am certain you would find that effects go both ways if you
examined things in sufficient detail. I would agree that the flow of effects is highly
asymmetrical in these comparisons.

To be less critical and more constructive for a moment (i.e., not to behave as a philosopher), allow
me to suggest that what you are looking for *is* dissipation - maintenance of some phenomenal
structure through the input and output of energy and matter. Dissipative structures form coulds and
cuckoos, and we treat such structures and their dynamics as those of individuals.

[There's a logical sense of "individual", too, which basically means anything that is
spatiotemorally limited. Hence, the set of all things touched by John Wilkins is an individual in
this sense.]

Quoted message said:

Quoted message said:

But even within those two domains - while some aspects of weather affect other aspects - many
things happen in weather than dampen out in larger scales, and so a dustdevil in central
Australia and a snowstorm in Antarctica are decoupled.

Exactly. This supports my contention that each of these dynamical, self-supporting (containing
internal, positive feedbacks) structures in the atmosphere are individuated things. This
conclusion does not IMHO preclude the possibility of a coherent global weather system, too.

Not in theory, no.

Quoted message said:

Quoted message said:

Likewise, the evolution of a fungus in the Amazon and the evolution of a bee in Europe are
decoupled (until they do meet and affect each other's environment). So I would also not say that
all evolution or all weather is a single cohesive syste.

I agree with your premise, but not your conclusion. My logic is consistent with the dynamical
networks paradigm (e.g., small world graphs) and hierarchy theory sensu Stan Salthe, if you are
familiar with them. Graph theory and hierarchy theory are both largely about nested entities.

I have a bit of familiarity here. I'm still not convinced, though, that this will
revolutionise anythign, but hey, it's worth a try. At worst, like Brooks and Wiley, you get
talked about for a while.

Quoted message said:

Quoted message said:

Hence I would not say that there has to be a property common to all instances of evolution or
weather, because there is no property-bearer, no single "thing".

I agree that there "would not [have] to be a property common to all instances of evolution or
weather," but disagree as to whether or not there is. I would caution you here to avoid the common
mistake (IMHO) of failing to recognize the dynamical and coherent nature of systems that function
on very different time and spatial scales than those of our existence. For example, most
scientists would quickly accept the notion that a rock is not a dynamical system. In fact, it is,
although it is in a senescent stage of degradation. Its structure, however, is maintained by the
tension caused by attractive and repulsive forces among its component particles. The time scale of
functional activity in evolution is far slower than the time scales at which we exist, which can
obfuscate its functional coherence, if I am right that it exists.

I don't believe that I do make that mistake here. But that doesn't justify claiming that any
collection of things, whether atoms in a rock or organisms in an ecosystem, qualifies as a "property-
bearer" in the theoretical sense.

Quoted message said:

Quoted message said:

Quoted message said:

> If so, does it include tectonic processes? Vulcanism? Bolide impacts? All these things impact
> (sorry :-) on weather processes, so why not.

Again, I would say not. I see tectonics as part of a geological process that is functionally
uncoupled from the weather. One thing that reveals the uncoupled nature of these processes is
that the operate on very different time scales. Note that functionally decoupled processes can
interact, just as two different organisms (which are functionally decoupled entities) can
interact, and such interactions among entities can contribute to the functional organization of
some particular functional process/entity at a larger scale (e.g., consider the GAIA
hypothesis).

But now it is open to me to say that weather *does* affect geology, through erosion and
deposition. Where do you draw that line?

All real systems are open, hence effects enter from the outside as perturbations and effects leak
out to perturb other systems. Showing some form of effective interaction between agents (note that
I don't mean to imply anything about intentionality) does not undermine their existence as
functionally coherent systems. Nature draws the line by causing systemic organization and
compartmentalization, meaning that there is a clear (but never absolute) distinction between the
extent of "particle" interaction within a system compared with between systems. I view it as
perhaps the most fundamental role of all empirical scientists to identify the natural bounds of
real systems before they target someTHING for study. This can be easy and familiar (e.g., the
bounds of an individual mammal) or hard and abstract (e.g., the bounds of an ecosystem).

And it can be wrong. Since nobody starts tabula rasa, everyone is in the business of taking prior
classifications and refining them in concert with theory and observation. Hennig called it, in
another context, reciprocal illumination. One of the things that *can* happen in science is that you
end up dissolving the apparently coherent individual type you began with.

Quoted message said:

Quoted message said:

Quoted message said:

> How about Melankovitch cycles <sp?> My point is that something like "weather" is a convention
> name we give to a range of phenomena that suit us to collect under one term.

You could be right about weather as a general category. I am much more comfortable thinking
about individual storms as process/entities that have individuated and have agency. I am open
to considering the global weather in this way (see below), but it is not as clear to me that
global weather is a single, functional process/entity.

Then why claim the same for the formally analogous process of evolution?

Perhaps I have overstated my case to make it. I would not argue that my position is so
irrefutable or sufficiently tested that it should be universally considered scientific "fact" at
this point. Consider it a hypothesis, which I think is highly likely to be a good one for both
atmosphere and biosphere. I would also argue that the validity of my hypothesis for evolution
does not rise or fall on its validity for weather, although I do think that these contexts seem
to be formally the same.

OK, it is an empirical claim. How can you go about testing it?

Quoted message said:

Quoted message said:

Saying that things evolve and things climatically and atmospherically change doesn't imply that
there is an "evolving" nor a "weathering", so why would we expect to find that there is a
"normal" rate or mode of evolution anyway?

I would not expect a "normal" rate or mode of evolution any more than I would expect a "normal"
rate or mode of weather. Similarly, I would not expect, and we do not see, a "normal" rate or mode
of biological existence. However, I would expect universal "metabolic" scaling properties to hold
for global weather and the biosphere, in the same why folks like Brown and West are exploring for
organisms and ecosystems.

In my continuing role as gadfly, allow me to ask whether it is not actually true that weather has a
"normal" mode on earth. We tend not to get the supersonic roaring forties that occur on Jupiter or
the Sun. We also have no helium ice. So we can say there is a "normal range".

Quoted message said:

Quoted message said:

I believe that we have always given into the entification (as Henry Plotkin once called it) of
evolution, and we suffer from that. It causes teleology - we expect to find norms and functions
in things that are entities (cohesive individuals).

That's funny, because I think we have shackled ourselves as scientists with notions like
Skinnerian behaviorism and avoidance of anything that smacks of anthropomorphism. In fact, I am
hoping for a resurgence of the vary things you feel we suffer from. I argue, for example, that the
second law of thermodynamics implies the existence of natural teleology; that is, nature itself is
teleological, and we must accept teleological explanations of natural phenomena if we are to
understand them correctly. I understand why the "shackles" were introduced to science, which I
attribute to the potential for generating uncritical and unconstrained inferences. If I am right
about natural teleology it will be important to figure out a way to embrace teleological
explanations without opening Pandora's box. I like the potential for the framework of
thermodynamics to constrain the kinds of teleological hypotheses and inferences we can accept.

I believe that we name general terms on the basis of our culturally encountered exemplars, and of
course we encounter goal-directed systems all the time. We are they. But we also over-generalise
from our experience, and so we tend to "see" goal-directedness wherever we look. We need to hold
this tendency in check, and ensure that our applications of general terms are trulyt appropriate in
each case. Teleology in biology is almost always inappropriate.

Quoted message said:

Quoted message said:

Quoted message said:

> It is not a "thing" or "process" in and of itself. Evolution is like that. So properties of
> that non-thing are empty, or they are properties of particular things and processes we would
> do better to investigate less generally and not go looking for universalisations like
> "adaptation is the most important aspect of evolution", etc.

Your position is certainly arguable. I would not dismiss it. However, my view is different, and
I am convinced that my view is also not so easily dismissed. It may be that the functional
coherence of evolution comes and goes, much like the functional coherence of weather systems.
We wouldn't say that the dissipation of a convection cell is tantamount to the end of weather.
There are top-down global weather effects caused by interactions among weather systems and the
cohesion of the atmosphere that maintain wide-scale distributions of weather patterns across
regions of the globe. Evolution is likely to work in a similar way, IMHO. Evolution storms
(most likely coevolution storms) can spring up in a rush of coordinated self-organization,
which eventually run there course and dissipate. I think that there are also top-down effects
that coordinate such local activities (e.g., maintenance of food web structures, tendencies
toward niche filling, ...), which is basically the GAIA idea. So, my model has evolution as a
mix of chaos
(e.g., the effects of mutation and drift) and self-organization, which can manifest locally in
time and space at several different scales simultaneously.

Gaia is not a counterexample to my view, so long as it can be empirically demonstrated. After
all, if Hamilton entertained it, then it is eminently entertainable. But it has to be shown to
occur. If there is top-down causation, then there is, but I'm from Australia - show me.

Do you accept the notion of top-down causation (or more likely pattern-forming constraints that
facilitate global functioning) in general, or are you skeptical that they exist? What do you think
of the daisy world model?

Yes, yes, and it's interesting but inconclusive. Abstract models abound
- the issue is what ones apply and why them and not all the others. Recall the Raup shellspace of
the 60s. Recall that although the algorithms defined a space, it was in fact only partially
occupied. *That* is the interesting thing - why only those regions? Daisyworld defines the formal
outcomes of a series of assumptions, but we cannot therefore assume it models the actual world.
And what parts of it match and what don't is the interesting part of it all.

Quoted message said:

Quoted message said:

Hamilton's idea that algal spores form clouds to disseminate the spores is a Gaianesque model -
but one that can be tested. The idea that ther eis "Evolution" because there must be some top-
down causation is argument by assertion unless we get specifics.

I have never argued that there "must be some top-down causation" for the biosphere or any
other putative system. I do argue that there "must be some top-down causation" for a real
system to exist. The question, as you have argued above, is whether or not the putative system
is a real one.

Quoted message said:

It seems to me that evolutionary processes will be much more interesting if we can find out how
they *actually* work; more likely there will be a mix of coupled, uncoupled, level-restricted
and multilevel processes, and I want to know what they are, not what we think they will be.

OK, but I am sure that you recognize finding out what they are depends heavily on what we think
they will be.

Not entirely. We can be surprised. I don't hold with this newfangled social construction of reality
stuff. I'm not even sure I like much of the theory-dependence of observation stuff either. Colour me
positivistic if you like.

Quoted message said:

Quoted message said:

Hamilton, W. D. , and T. M. Lenton. 1998. Spora and Gaia - how microbes fly with their clouds.
Ethology Ecology and Evolution 10 (1):1-16.

Quoted message said:

The good news is that I don't think the difference in our views is just a matter of
interpretation or airy fairy ideas. Critical evaluation of agency is something that science is
potentially good at, because agency is about functional coherence (self-sustaining dynamics
with positive feedbacks) and external effects that should be measurable. I am optimistic that
we will develop a science of assigning agency in the near future (a decade or two?). Such a
science would be able to assess, for example, whether global weather is or is not a "thing".

I oppose agency in all cases that do not involve agents. With intentions. And internal models of
the world. The temptation to project our own agency on the world is as old as flok religion.

We have quite different semantics regarding agency, so we will have to be careful in using this
word. I don't know of anything resembling a synonym for my use of agency, but it seems like
intentionality is close to a synonym for your use. Am I correct?

Almost. That's for another time.

Quoted message said:

Cheers,

Guy

--
John Wilkins wilkins.id.au For long you live and high you fly, and smiles you'll give and tears
you'll cry and all you touch and all you see is all your life will ever be

in article [email hidden], John Wilkins at
[email hidden] wrote on 10/22/03 8:23 AM:

Quoted message said:

Guy Hoelzer said:

in article [email hidden], John Wilkins at [email hidden] wrote on
10/17/03 8:33 AM:

Quoted message said:

Guy Hoelzer <[email hidden]> wrote:

> in article [email hidden], John Wilkins at [email hidden] wrote on
> 10/14/03 10:27 PM:
>
>> Guy Hoelzer <[email hidden]> wrote:
>>
>>>> It seems to me a mistake to say there is this "thing", evolution, and ask what is most
>>>> important to it the way we might ask what is a critical path in a manufacturing process.
>>>
>>> Hmm. This seems like a deep difference between our views. I see a universe filled with
>>> entities and processes, and these are often entangled because entities are always products of
>>> processes. By this I mean more than products of past processes. I think that all physical
>>> entities are the products of past and current processes, even if the current processes are
>>> merely slowing the rate of decay (e.g., the tension of forces that hold a rock together). You
>>> may have noticed that I am fond of pointing to logical connections between evolution as a
>>> process (a single processes that can also be decomposed into sub-processes like natural
>>> selection, drift, and mutation) and weather systems as processes. The main reason I like to
>>> do this is to emphasize that evolution is a physically manifested process like a storm. I
>>> argue, for example, that a convection cell in the atmosphere is both a process and a "thing"
>>> with agency (a cause of effects), and I think that evolution is in fact a similar sort of
>>> "thing". I think that evolution is not separable from the biosphere in the same sense that
>>> convection is separable from a convection cell.
>>
>> For a deep division we agree extraordinarily.
>>
>> But is "weather" a thing?
>
> I would say not, but I would say that a particular storm is a thing. I see the biosphere as a
> storm that has persisted for billions of years.

Well in one way you might be right - it is a phenomenal individual - that is to say, we could
locate it in time and space and say "look at what the surface of that planet is doing" (had we
but time enough and space). But my question was alittle more specific - is it a single cohesive
individual? Is it, in short, a single system? Is it one process?

That is my position. Individual tornados, hurricane, and convection cells are each single systems
in the same sense that you and I are single systems.

But I am not a single system in some ways. I am dependent on the massive number of endosymbionts I
support. I require a constant environment with a ready source of food. I must live, because it is
an autapomorphy of my species, in society.

I agree with all of this, so I would have started with the word "And" rather than "But". All natural
systems are open and communicate with the external environment. They must all be externally fueled
and release energy/matter/information back to the outside. They can also be perturbed by outside
influences and cause perturbations to other systems.

Regarding the "John Wilkins" system, you hinted at an excellent point. Those endosymbionts are
essential components of the "John Wilkins" system. They are part of the system, rather than being
external to it. Most systems that we consider, probably all biological systems, are composed of
elements that are themselves complex dynamical systems. Hierarchy theory (sensu Stan Salthe) has
much to say on this point. It seems to me that there is also no general reason to expect that
particular systems must be restricted to participating as elements in only one higher order system.
Many humans, for example, participate in several institutional social systems (e.g., corporations,
families, churches, ...), as well as biological and physical systems. Conflicts of interest
regarding the behavior of "particles" participating in multiple higher order systems would
inevitably arise, but I don't see why this would prevent such "divided loyalties" from occurring.

Quoted message said:

Quoted message said:

Quoted message said:

Now for both weather and evolution, I think it is a series of occasionally coupled systems, and
mostly decoupled.

I think I agree, but I if I am interpreting your comment correctly, I would also argue that all
coherent systems can be statistically decomposed in this way. You, for example, can be decomposed
into organ systems, organs, or cells, each of which exhibits much more tightly coupled dynamics
of its subcomponents than the same type of subcomponents exhibit at the scale of you as a whole.
IMHO, this does not make you a mere epiphenomenon giving the mere illusion of coherent function.
Your intelligent posts reveal otherwise.

Flatterer. But you still don't tell me how we can decompose systems such that I can tell if
something, some subsystem, is a "thing" or not.

This is an old and somewhat sophomoric (soporific?) problem in philosophy. We are asking, in
effect, how far does an explanation need to go. If we draw a line at some point (say, my skin or
my local ecosystem, and so on) we are entitled to ask, why there? If it turns out that we draw the
line there for convenience' sake, then we have only asserted some fact about us as observers.

This is a key point, and I agree that we have yet to determine necessary and sufficient conditions
of "systemhood". Skeptics will continue to play an important role in the growth of complex systems
science until we know those conditions. At the moment I would argue that this set of conditions will
include positive feedback loops that sustain structure and function, and internal constraints that
restrict the potential for component behaviors detrimental to the whole. Structural and functional
dynamics must be driven primarily by internal interactions among components that are merely fueled
by external sources. For example, energy/matter must enter the system from the outside, but
information that drives structural patterning resulting from extraction of work from the fuel comes
primarily from from interactions within the system (e.g., entrainment). This does not prohibit the
imposition of pattern from the outside (e.g., if you were hit hard with an open waffle iron you
would have a pattern imposed), but I can't think of an asystematic way to generate patterns other
than by external imposition. Ultimately, I think the necessary and sufficient conditions will come
from thermodynamics (sensu Prigogine). All systems extract work from their fuel with which they
build and repair their structure, which sustains the temporal efficiency of the system in breaking
down gradients and maximizes the rate of entropy increase at the scale of the whole universe.

Quoted message said:

My objection to treating evolution, or weather, as a single thing is that there is none of the
cohesion required to make either of them a functional or cohesive individual. We draw those lines
because there is a reason, a theoretical reason, to do so. I am merely saying that while a tornado
might be a cohesive individual, since it is a dissipative structure, weather is not. Likewise,
while we might treat the speciation of a mammal species into two as a single process (and can ask
sensibly what the mode, frequency and mechanisms were in each case), we can't ask such questions
of "evolution as a whole".

Maybe we can, although we haven't tended to do so. This is, I think, the GAIA proposition. It may
be, for example, that it is very hard to see or imagine the coherent operation of higher order
systems of which we are a part. We may cherish the idea of free will so much that we blind
ourselves to the top-down effects of higher order systems that guide us as compliant parts of some
larger whole, unless the larger wholes spring from our intentional actions (e.g., many social
organizations). We can see the whole when we intend to be part of a team, but we need guidelines
(as you suggested) to help us identify coherent systems that are generally outside of our
perceptive range.

Quoted message said:

Quoted message said:

Quoted message said:

While what happens in the Deccan Traps or Krakatoa or Chixulub or the breakup of Gondwana
*affect* evolution and the weather, the latter are just that - effects. So far as I know, none
of the geological or astronomical phenomena are caused by either the weather or the evolution of
organisms. So I would *not* say that everything that happens *in* weather or evolution is part
of it.

I think I agree here, although I am certain you would find that effects go both ways if you
examined things in sufficient detail. I would agree that the flow of effects is highly
asymmetrical in these comparisons.

To be less critical and more constructive for a moment (i.e., not to behave as a philosopher),
allow me to suggest that what you are looking for *is* dissipation - maintenance of some
phenomenal structure through the input and output of energy and matter. Dissipative structures
form coulds and cuckoos, and we treat such structures and their dynamics as those of individuals.

Yup - this is the Prigogine terminology derived from a thermodynamic perspective, and it
works for me.

Quoted message said:

[There's a logical sense of "individual", too, which basically means anything that is
spatiotemorally limited. Hence, the set of all things touched by John Wilkins is an individual in
this sense.]

Right. I tend to think in spatio-temporal terms, too, but this may not be sufficient in all cases.
For example, I have recently had an interesting conversation with a physicist colleague of mine
(Eric Smith from the Santa Fe Institute and Los Alamos) on a kind of boundary that is discrete, but
not in space or time. This is the nature of energy as it falls into the visible spectrum as opposed
to the thermal microwave radiation. The idea is that photons can be decomposed into components (I
don't think we have names for them) that exist in the thermal microwave state (a state of higher
entropy) after decoupling. I would not call visible light and thermal microwave energy individuals,
but then my imagination has its limits.

Quoted message said:

Quoted message said:

Quoted message said:

But even within those two domains - while some aspects of weather affect other aspects - many
things happen in weather than dampen out in larger scales, and so a dustdevil in central
Australia and a snowstorm in Antarctica are decoupled.

Sure. There isn't much detectable interaction between of one of your liver cells and one of your
hair cells either. Whole systems are often held together by indirect interactions (e.g.,
butterfly effects).

Quoted message said:

Quoted message said:

Exactly. This supports my contention that each of these dynamical, self-supporting (containing
internal, positive feedbacks) structures in the atmosphere are individuated things. This
conclusion does not IMHO preclude the possibility of a coherent global weather system, too.

Not in theory, no.

Quoted message said:

Quoted message said:

Likewise, the evolution of a fungus in the Amazon and the evolution of a bee in Europe are
decoupled (until they do meet and affect each other's environment). So I would also not say that
all evolution or all weather is a single cohesive syste.

I agree with your premise, but not your conclusion. My logic is consistent with the dynamical
networks paradigm (e.g., small world graphs) and hierarchy theory sensu Stan Salthe, if you are
familiar with them. Graph theory and hierarchy theory are both largely about nested entities.

I have a bit of familiarity here. I'm still not convinced, though, that this will revolutionise
anythign, but hey, it's worth a try. At worst, like Brooks and Wiley, you get talked about for
a while.

I wouldn't bury Brooks and Wiley yet. That book may have been written before its time, and might
achieve greater importance in the future. You mentioned the book by Depew and Weber (Darwinism
Evolving) previously, which I think is a better exploration of these ideas. I would argue, for
example, that the paper (see below) laying out the functional thermodynamic view of ecosystems is
just beginning to have a strong influence.

Schneider, E. D. & Kay, J. J. 1994 Life as a manifestation of the second law of thermodynamics.
Math. Comp. Modeling 19, 25-48.

Quoted message said:

Quoted message said:

Quoted message said:

Hence I would not say that there has to be a property common to all instances of evolution or
weather, because there is no property-bearer, no single "thing".

I agree that there "would not [have] to be a property common to all instances of evolution or
weather," but disagree as to whether or not there is. I would caution you here to avoid the
common mistake (IMHO) of failing to recognize the dynamical and coherent nature of systems that
function on very different time and spatial scales than those of our existence. For example, most
scientists would quickly accept the notion that a rock is not a dynamical system. In fact, it is,
although it is in a senescent stage of degradation. Its structure, however, is maintained by the
tension caused by attractive and repulsive forces among its component particles. The time scale
of functional activity in evolution is far slower than the time scales at which we exist, which
can obfuscate its functional coherence, if I am right that it exists.

I don't believe that I do make that mistake here. But that doesn't justify claiming that any
collection of things, whether atoms in a rock or organisms in an ecosystem, qualifies as a "property-
bearer" in the theoretical sense.

I agree. I tried to explain why a rock fits this claim, but I strongly argue against the idea that
any collection of things would do so. I think that progress in biology (especially ecology and
evolution) has frequently been limited by failures to make exactly this distinction leading to
studies of collections that do not constitute systems.

Quoted message said:

Quoted message said:

Quoted message said:

>> If so, does it include tectonic processes? Vulcanism? Bolide impacts? All these things impact
>> (sorry :-) on weather processes, so why not.
>
> Again, I would say not. I see tectonics as part of a geological process that is functionally
> uncoupled from the weather. One thing that reveals the uncoupled nature of these processes is
> that the operate on very different time scales. Note that functionally decoupled processes can
> interact, just as two different organisms (which are functionally decoupled entities) can
> interact, and such interactions among entities can contribute to the functional organization of
> some particular functional process/entity at a larger scale (e.g., consider the GAIA
> hypothesis).

But now it is open to me to say that weather *does* affect geology, through erosion and
deposition. Where do you draw that line?

All real systems are open, hence effects enter from the outside as perturbations and effects leak
out to perturb other systems. Showing some form of effective interaction between agents (note
that I don't mean to imply anything about intentionality) does not undermine their existence as
functionally coherent systems. Nature draws the line by causing systemic organization and
compartmentalization, meaning that there is a clear (but never absolute) distinction between the
extent of "particle" interaction within a system compared with between systems. I view it as
perhaps the most fundamental role of all empirical scientists to identify the natural bounds of
real systems before they target someTHING for study. This can be easy and familiar (e.g., the
bounds of an individual mammal) or hard and abstract (e.g., the bounds of an ecosystem).

And it can be wrong. Since nobody starts tabula rasa, everyone is in the business of taking prior
classifications and refining them in concert with theory and observation. Hennig called it, in
another context, reciprocal illumination. One of the things that *can* happen in science is that
you end up dissolving the apparently coherent individual type you began with.

I agree. Wouldn't it be nice to have those necessary and sufficient conditions to examine and
justify the objects of our study?

Quoted message said:

Quoted message said:

Quoted message said:

>> How about Melankovitch cycles <sp?> My point is that something like "weather" is a convention
>> name we give to a range of phenomena that suit us to collect under one term.
>
> You could be right about weather as a general category. I am much more comfortable thinking
> about individual storms as process/entities that have individuated and have agency. I am open
> to considering the global weather in this way (see below), but it is not as clear to me that
> global weather is a single, functional process/entity.

Then why claim the same for the formally analogous process of evolution?

Perhaps I have overstated my case to make it. I would not argue that my position is so
irrefutable or sufficiently tested that it should be universally considered scientific "fact" at
this point. Consider it a hypothesis, which I think is highly likely to be a good one for both
atmosphere and biosphere. I would also argue that the validity of my hypothesis for evolution
does not rise or fall on its validity for weather, although I do think that these contexts seem
to be formally the same.

OK, it is an empirical claim. How can you go about testing it?

My approach of choice at the moment, in the absence of those necessary and sufficient conditions, is
to model the putative system computationally. If you can't get the computational model to generate
emergent structures (self-organize) or behaviors different from what you would expect from bottom-up
only causation, then you either don't have a system or you have missed some key component(s). At
least, at that point I would argue that you have failed to demonstrate that the collection has the
potential to operate as a system. If you do encounter the hallmarks of systematic activity, I think
the computational framework also provides you with a platform for exploring the generic tendencies
(inducible; tendencies about which you can be inductive) of that class of system.

Quoted message said:

Quoted message said:

Quoted message said:

Saying that things evolve and things climatically and atmospherically change doesn't imply that
there is an "evolving" nor a "weathering", so why would we expect to find that there is a
"normal" rate or mode of evolution anyway?

I would not expect a "normal" rate or mode of evolution any more than I would expect a "normal"
rate or mode of weather. Similarly, I would not expect, and we do not see, a "normal" rate or
mode of biological existence. However, I would expect universal "metabolic" scaling properties to
hold for global weather and the biosphere, in the same why folks like Brown and West are
exploring for organisms and ecosystems.

In my continuing role as gadfly, allow me to ask whether it is not actually true that weather has
a "normal" mode on earth. We tend not to get the supersonic roaring forties that occur on Jupiter
or the Sun. We also have no helium ice. So we can say there is a "normal range".

It seems fine to me to say there is a "normal range", although the power laws so often observed for
systems regarding the frequency distribution of event sizes leads me to expect that catastrophic
events outside the "normal range" are likely to happen surprisingly often. In other words, I
wouldn't be surprised to find surprisingly unusual climatological events.

Quoted message said:

Quoted message said:

Quoted message said:

I believe that we have always given into the entification (as Henry Plotkin once called it) of
evolution, and we suffer from that. It causes teleology - we expect to find norms and functions
in things that are entities (cohesive individuals).

That's funny, because I think we have shackled ourselves as scientists with notions like
Skinnerian behaviorism and avoidance of anything that smacks of anthropomorphism. In fact, I am
hoping for a resurgence of the vary things you feel we suffer from. I argue, for example, that
the second law of thermodynamics implies the existence of natural teleology; that is, nature
itself is teleological, and we must accept teleological explanations of natural phenomena if we
are to understand them correctly. I understand why the "shackles" were introduced to science,
which I attribute to the potential for generating uncritical and unconstrained inferences. If I
am right about natural teleology it will be important to figure out a way to embrace teleological
explanations without opening Pandora's box. I like the potential for the framework of
thermodynamics to constrain the kinds of teleological hypotheses and inferences we can accept.

I believe that we name general terms on the basis of our culturally encountered exemplars, and of
course we encounter goal-directed systems all the time. We are they. But we also over-generalise
from our experience, and so we tend to "see" goal-directedness wherever we look. We need to hold
this tendency in check, and ensure that our applications of general terms are trulyt appropriate
in each case. Teleology in biology is almost always inappropriate.

I think we agree in principle, but emphasize different sides of the same coin.

Quoted message said:

Quoted message said:

Quoted message said:

>> It is not a "thing" or "process" in and of itself. Evolution is like that. So properties of
>> that non-thing are empty, or they are properties of particular things and processes we would
>> do better to investigate less generally and not go looking for universalisations like
>> "adaptation is the most important aspect of evolution", etc.
>
> Your position is certainly arguable. I would not dismiss it. However, my view is different, and
> I am convinced that my view is also not so easily dismissed. It may be that the functional
> coherence of evolution comes and goes, much like the functional coherence of weather systems.
> We wouldn't say that the dissipation of a convection cell is tantamount to the end of weather.
> There are top-down global weather effects caused by interactions among weather systems and the
> cohesion of the atmosphere that maintain wide-scale distributions of weather patterns across
> regions of the globe. Evolution is likely to work in a similar way, IMHO. Evolution storms
> (most likely coevolution storms) can spring up in a rush of coordinated self-organization,
> which eventually run there course and dissipate. I think that there are also top-down effects
> that coordinate such local activities (e.g., maintenance of food web structures, tendencies
> toward niche filling, ...), which is basically the GAIA idea. So, my model has evolution as a
> mix of chaos
> (e.g., the effects of mutation and drift) and self-organization, which can manifest locally in
> time and space at several different scales simultaneously.

Gaia is not a counterexample to my view, so long as it can be empirically demonstrated. After
all, if Hamilton entertained it, then it is eminently entertainable. But it has to be shown to
occur. If there is top-down causation, then there is, but I'm from Australia - show me.

Do you accept the notion of top-down causation (or more likely pattern-forming constraints that
facilitate global functioning) in general, or are you skeptical that they exist? What do you
think of the daisy world model?

Yes, yes, and it's interesting but inconclusive. Abstract models abound
- the issue is what ones apply and why them and not all the others. Recall the Raup shellspace of
the 60s. Recall that although the algorithms defined a space, it was in fact only partially
occupied. *That* is the interesting thing - why only those regions? Daisyworld defines the
formal outcomes of a series of assumptions, but we cannot therefore assume it models the actual
world. And what parts of it match and what don't is the interesting part of it all.

Yes, but it is also important to realize that our observations of manifested reality, and perhaps
all of manifested reality, is only a subset of the possible. And it may be an internally correlated
subset, which results in only particular regions of possibility space being occupied. Asking about
the idiosyncratic (historical) reasons for a system's failure to explore the whole space is quite
different from asking whether we have correctly identified the limits to the possibility space.
There is a new movement in the statistics community to develop "possibility theory" as opposed to
"probability theory", which I think is based on this distinction. I would argue that to understand
the nature of a system you want to know about its possibility space, rather than the limits of that
space that have actually been explored.

Quoted message said:

Quoted message said:

Quoted message said:

Hamilton's idea that algal spores form clouds to disseminate the spores is a Gaianesque model -
but one that can be tested. The idea that ther eis "Evolution" because there must be some top-
down causation is argument by assertion unless we get specifics.

I have never argued that there "must be some top-down causation" for the biosphere or any
other putative system. I do argue that there "must be some top-down causation" for a real
system to exist. The question, as you have argued above, is whether or not the putative system
is a real one.

Quoted message said:

It seems to me that evolutionary processes will be much more interesting if we can find out how
they *actually* work; more likely there will be a mix of coupled, uncoupled, level-restricted
and multilevel processes, and I want to know what they are, not what we think they will be.

OK, but I am sure that you recognize finding out what they are depends heavily on what we think
they will be.

Not entirely. We can be surprised. I don't hold with this newfangled social construction of
reality stuff. I'm not even sure I like much of the theory-dependence of observation stuff either.
Colour me positivistic if you like.

I don't like the notion of "individual realities" either. This is antithetical to science, IMHO.

I would only add that I am interested in both the fundamental nature of evolution (based on its
systematic tendencies) and its historical manifestation. To use Hennig's terminology as you did, I
see these endeavors as "reciprocally illuminating".

Regards,

Guy

Guy Hoelzer <[email hidden]> wrote in
:"]news:[email hidden]:

Quoted message said:

in article [email hidden], John Wilkins at [email hidden] wrote on
10/22/03 8:23 AM:

Quoted message said:

Quoted message said:

Guy Hoelzer <[email hidden]> wrote:

Quoted message said:

Quoted message said:

Quoted message said:

That is my position. Individual tornados, hurricane, and convection cells are each single
systems in the same sense that you and I are single systems.

Quoted message said:

Quoted message said:

But I am not a single system in some ways. I am dependent on the massive number of endosymbionts
I support. I require a constant environment with a ready source of food. I must live, because it
is an autapomorphy of my species, in society.

Quoted message said:

I agree with all of this, so I would have started with the word "And" rather than "But". All
natural systems are open and communicate with the external environment. They must all be
externally fueled and release energy/matter/information back to the outside. They can also be
perturbed by outside influences and cause perturbations to other systems.

Regarding the "John Wilkins" system, you hinted at an excellent point. Those endosymbionts are
essential components of the "John Wilkins" system. They are part of the system, rather than being
external to it. Most systems that we consider, probably all biological systems, are composed of
elements that are themselves complex dynamical systems. Hierarchy theory (sensu Stan Salthe) has
much to say on this point. It seems to me that there is also no general reason to expect that
particular systems must be restricted to participating as elements in only one higher order
system. Many humans, for example, participate in several institutional social systems (e.g.,
corporations, families, churches, ...), as well as biological and physical systems. Conflicts of
interest regarding the behavior of "particles" participating in multiple higher order systems
would inevitably arise, but I don't see why this would prevent such "divided loyalties" from
occurring.

Which gets to the point of what the division is between entities. This is ultimately meaningless, it
is a result of the filter we impose on the world as observer. Now we have to impose the filter, we
have to entify, because we cannot hold all that there is in our heads. So we create a model, which
loses information but allows us to manipulate it. But the meaning comes from the model. It is
intrinsic to the system only in the accuracy to which the system can be modeled by a simpler system.

Quoted message said:

Quoted message said:

This is an old and somewhat sophomoric (soporific?) problem in philosophy. We are asking, in
effect, how far does an explanation need to go. If we draw a line at some point (say, my skin or
my local ecosystem, and so on) we are entitled to ask, why there? If it turns out that we draw
the line there for convenience' sake, then we have only asserted some fact about us as observers.

Quoted message said:

This is a key point, and I agree that we have yet to determine necessary and sufficient conditions
of "systemhood". Skeptics will continue to play an important role in the growth of complex systems
science until we know those conditions. At the moment I would argue that this set of conditions
will include positive feedback loops that sustain structure and function, and internal constraints
that restrict the potential for component behaviors detrimental to the whole. Structural and
functional dynamics must be driven primarily by internal interactions among components that are
merely fueled by external sources. For example, energy/matter must enter the system from the
outside, but information that drives structural patterning resulting from extraction of work from
the fuel comes primarily from from interactions within the system (e.g., entrainment). This does
not prohibit the imposition of pattern from the outside (e.g., if you were hit hard with an open
waffle iron you would have a pattern imposed), but I can't think of an asystematic way to generate
patterns other than by external imposition. Ultimately, I think the necessary and sufficient
conditions will come from thermodynamics (sensu Prigogine). All systems extract work from their
fuel with which they build and repair their structure, which sustains the temporal efficiency of
the system in breaking down gradients and maximizes the rate of entropy increase at the scale of
the whole universe.

I don't think you have answered John's objection. I would answer that the line is drawn where we
obtain the maximum accuracy of prediction with the minimum required information. From an energetic
standpoint, the line is drawn at areas of sharp gradients - which is why it is easy to identify
tornados and humans as individual systems.

Quoted message said:

Quoted message said:

My objection to treating evolution, or weather, as a single thing is that there is none of the
cohesion required to make either of them a functional or cohesive individual. We draw those lines
because there is a reason, a theoretical reason, to do so. I am merely saying that while a
tornado might be a cohesive individual, since it is a dissipative structure, weather is not.
Likewise, while we might treat the speciation of a mammal species into two as a single process
(and can ask sensibly what the mode, frequency and mechanisms were in each case), we can't ask
such questions of "evolution as a whole".

Maybe we can, although we haven't tended to do so. This is, I think, the GAIA proposition. It may
be, for example, that it is very hard to see or imagine the coherent operation of higher order
systems of which we are a part. We may cherish the idea of free will so much that we blind
ourselves to the top-down effects of higher order systems that guide us as compliant parts of some
larger whole, unless the larger wholes spring from our intentional actions (e.g., many social
organizations). We can see the whole when we intend to be part of a team, but we need guidelines
(as you suggested) to help us identify coherent systems that are generally outside of our
perceptive range.

Even though on every other Tuesday I am willing to consider GAIA, I think it is a separate issue
than free will, and in fact may be a similar process to free will, with both being emergent
properties of complex systems. And to argue that GAIA guides evolution to produce GAIA would seem to
imply a weird combination of the Cosmological and Ontological arguments. I tend to agree with John
that evolution for the most part is better envisioned as a historical result of a series of
processes than as a process itself. I can still sympathize with the thought that it is hard (and may
be impossible) for humans to envision systems of significantly higher order than ourselves (or even
to understand our own consciousness). But if evolution is to be considered a process, I would think
the most that could be said is that it is a process of searching the available phenotype space (a la
Dennett). On the other hand, I can sympathize with the thought that it is hard for humans to
envision systems of higher order than ourselves - and without envisioning them, it is hard to see
the kind of gradients that would allow us to classify them. It is generally easier to recognize
ecosystem boundaries from the air than from the ground

(snip)

Quoted message said:

Quoted message said:

Quoted message said:

> Hence I would not say that there has to be a property common to all instances of evolution or
> weather, because there is no property-bearer, no single "thing".

Quoted message said:

Quoted message said:

Quoted message said:

I agree that there "would not [have] to be a property common to all instances of evolution or
weather," but disagree as to whether or not there is. I would caution you here to avoid the
common mistake (IMHO) of failing to recognize the dynamical and coherent nature of systems that
function on very different time and spatial scales than those of our existence. For example,
most scientists would quickly accept the notion that a rock is not a dynamical system. In fact,
it is, although it is in a senescent stage of degradation. Its structure, however, is maintained
by the tension caused by attractive and repulsive forces among its component particles. The time
scale of functional activity in evolution is far slower than the time scales at which we exist,
which can obfuscate its functional coherence, if I am right that it exists.

Quoted message said:

Quoted message said:

I don't believe that I do make that mistake here. But that doesn't justify claiming that any
collection of things, whether atoms in a rock or organisms in an ecosystem, qualifies as a "property-
bearer" in the theoretical sense.

Quoted message said:

I agree. I tried to explain why a rock fits this claim, but I strongly argue against the idea that
any collection of things would do so. I think that progress in biology (especially ecology and
evolution) has frequently been limited by failures to make exactly this distinction leading to
studies of collections that do not constitute systems.

To play the devil's advocate, why can't I claim that the list of objects in a scavenger hunt all
share the property that they are the object of the hunt, and may further share some obscure symbolic
connection in the mind of the person who drew up the list? I don't claim that this is a useful
"entity", but I also think that it could be hard to distinguish it from some other collections that
many would consider a "system", such as the local highway department.(They certainly often qualify
as a dissipative structure in my book :-) Ultimately the test of calling a collection of things a
system has to rest on its usefulness as a predictive tool. If calling "weather" and "evolution"
systems leads to a better understanding of either or both, it is valid. If it leads to unfounded
teleological explanations it is a mistake.

Yours,

Bill Morse