Re: Replicating What?

From: Guy Hoelzer (hoelzer_at_unr.edu)
Date: 09/21/04 

Date: Tue, 21 Sep 2004 00:35:02 +0000 (UTC)

Hi Tony,

in article cikgvt$2u3e$1@darwin.ediacara.org, Anthony Cerrato at
tcerrato@optonline.net wrote on 9/19/04 10:51 AM:

> "Guy Hoelzer" <hoelzer@unr.edu> wrote in message
> news:cif7vb$19pn$1@darwin.ediacara.org...
>> in article cidugl$tul$1@darwin.ediacara.org, Anthony
> Cerrato at
>> tcerrato@optonline.net wrote on 9/16/04 10:59 PM:
>>
>>> "Guy Hoelzer" <hoelzer@unr.edu> wrote in message
>>> news:ciaksj$311n$1@darwin.ediacara.org...
>>>> in article ci9ngq$2o58$1@darwin.ediacara.org, Anthony
>>>> Cerrato at tcerrato@optonline.net wrote on 9/15/04 8:35
> AM:
>
> I have excerpted in the form of dialog for easier replies.
>
> Guy:
> ....the vast majority of the information (structure) in the
> adult phenotype is generated through the process of
> self-organization during development. ...
>
> Tony:
> This is true.
>
> Guy:
> .... It is not strictly in the genome or any other sort of
> pre-existing template.
>
> Tony:
> This is not strictly true, although it might be considered a
> semantic issue in reconciling a holistic vs. reductionist
> viewpoint.

I think it is "strictly true," but I know that there are great semantic
difficulties generated in such a major change in paradigm.
 
> Guy:
> ....there is nothing like a "blueprint" used at all in organismal
> development...
>
> ....Consider, for example, that your body generates over 100 million different
> antibodies at any one moment, given a genome with about 30,000 genes. Of
> course the genome would have to encode everything else about the organism,
> too, if it were really a blueprint.your body generates over 100 million
> different antibodies at any one moment, given a genome with about 30,000
> genes.
>
> Tony:
> The "blueprint" is the information which tells the body how
> to _develop_ the organized information you speak of! It of
> course does not tell the body how to react to all
> contingencies, but provides the appropriate response
> programs to do so.

I agree that it provides some "appropriate response programs." I expect
that most of these result in canalization, which protects the developmental
outcome by dampening the effects of perturbations. This is pretty far from
what blueprints do.
 
> I think you are entwined in a semantic black hole here--yes,
> there are self-organizing systems, but you seem to imply
> that they arise out of thin air.

Well, there is the weather.

> It even sounds as if they are, themselves, somewhat anthropomorphic. In fact,
> they are purely the results of physical laws.

Indeed they are. I would never say that they are "anthropomorphic," but
they are often described as teleological because they serve the imperatives
of thermodynamics, or as "complex adaptive systems" (CAS) because they
behave "selfishly."
 
> To generate any such system (or subsystem in the body)
> requires that the proper conditions be set up in advance for
> it to arise--this is the information which is encoded in the
> genome.

I agree!

> Your example of antibodies is simply a straw man
> since the ability to generate such antibodies, as well as
> the myriad biochemical subsystems in the body and
> neurological brain organization (amongst many others), must
> indeed be hard-wired into the genome.

I think it is far from a "straw man." It is a concrete example of how great
diversity, complexity, and an adaptively flexible system is economically
encoded by the genome. Why would you expect that other aspects of the
genome would eschew the economic efficiencies made possible by natural
tendencies for self-organization?

Parts of the organism's cells (e.g., nucleic acids) are indeed
"blueprinted," or more precisely "templated." I am assuming that this is
essentially what you mean by "hard-wired." However, I doubt that there is
ANY empirical or theoretical support for your claim that anything above the
molecular level is hard-wired. This seems like a soft metaphor that has
risen to mythical status.

> This encoded
> information then works in conjunction with ever-present
> physical and chemical laws which obtain to complete the
> subcontract project of the moment.
>
> The utilization of the actions of these subsystems is indeed
> honed by natural selection over time, but there is a
> primitive original blueprint as to how the systems will
> originate. You may call this self-organization if you like,
> (it is, just as, in a sense, the universe is
> self-organizing) but just to do so adds little extra to
> understanding of such subsystems--i.e., simply studying the
> detailed function of the subsystems (purely as subsystems)
> and how they interact, alone provides the same information.

I think that the purely reductionistic approach generally yield different
information from the complex systems approach because you focus on different
issues and ask different kinds of questions. I think it certainly yields a
different (more well-balanced, comprehensive and mechanistic) understanding
of the process.

> The disagreement here may simply be one of a reductionist
> viewpoint vs. a holistic one. (Though it may not seem so
> here, my view is generally that both are needed to clarify
> different aspects of a problem.) However, the question you
> must answer is, what is the information that allows systems
> to "self-organize" and where does it come from. It of
> necessity is encoded somewhere, unless you wish to resort to
> some mystical magic which supercedes natural physical laws.
> :))

You know me better than that. Clearly some of the information for
establishing the conditions necessary for organismal self-organization
(development) resides in the structure of the genome, and further
information resides outside of the genome in the zygote, and outside of the
developing organism altogether. My view is that self-organization results
in the creation of information (new structure), which is where most of the
information contained in the adult comes from. There is nothing mystical
about the existence of mechanisms of pattern formation.
 
> Guy:
> My gut feeling is that your genome doesn't contain enough information to
> encode your eyes, let alone the rest of you. It merely puts into place lots of
> different kinds of components that participate in the dynamics of a growing
> complex system, and the information is generated through the interactions of
> those components. I also expect that a substantial fraction of the necessary
> components (perhaps the vast majority?) come from sources other than the
> genome.
>
> Tony:
> Of course I agree the bulk of the components come from
> external sources--the whole program of growth is based on
> outside air and water, amongst many things. I don't think
> it's correct to say though that the genome "merely" puts the
> components in place. Are there not a panoply of
> time-sequential steps/reactions in all these processes which
> require exact timing in specificity and order, i.e., a
> "program"?

Absolutely. This is a big reason why the genome must be relegated to
seeding self-organization, rather than directly encoding the details of such
temporally delicate processes. IMHO the temporal sequencing of critical
events is better left to self-organization, which allows event sequences to
unfold naturally along the path of least resistance. It is only where the
fitness of the adult requires a different path that natural selection would
lead to evolution of obstructive mechanisms with parts encoded in the
genome. I admit that we have no empirical evidence at the moment concerning
how much of development is actively steered away from self-organizing
tendencies compared with seeding the developmental process and merely
allowing it to follow the path of least resistance. As I said before, it is
my gut feeling that the vast majority of the information built into a
developing organism comes from sources outside the genome.

> There are also myriad chemical
> intermediates which must be generated in part by the program
> in intermediate reaction steps. This is no different than
> building a house: all the materials are obtained externally
> and the construction (albeit, including that of the systems
> which will arise or react later, e.g. the burglar alarm) is
> done by workers according to a primary blueprint.

I have no problem with the fact that construction from blueprints requires
consumption. The issue at hand is about the nature and source of
information used during development.
 
> Guy:
> If you find my view to be implausible, then I would ask you why natural
> selection would result in the unnecessary encoding of information in the
> genome, which would be relatively inflexible in the face of unpredictable
> environmental variation, when it would be far more efficient to take advantage
> of the opportunities for automatic information generation through physical
> tendencies for self-organization? I expect that fighting those tendencies
> through rigid and expansive genomic encoding would be highly maladaptive.
>
> Tony:
> I do not find your view implausible, only excessive.

Fair enough. It could be. I think it can be hard to tell when developing
an argument that is so different in perspective from both the mainstream and
one's own background training.
 
> I am not sure at all what your point is here though--natural
> selection has indeed encoded fixed and inflexible
> information in the genome!? And it does take advantage in
> your sense of self-organization, although these processes
> are based in simple (or hard) physical laws utilized in the
> genomic programmed time sequences. I think it's obvious the
> genome is not fighting these tendencies, which would indeed
> be maladaptive (actually it would be fatal.) It is in fact,
> executing them as planned to large extent. To have a fully
> flexible genome would only result in a lump of goo. It is
> however fortuitous that the vagaries of the universe include
> effects which cause random mutations which may (or not), at
> times, improve on the "rigid" genomic plan (i.e., it's not
> _that_ rigid.) In the house analogy this is equivalent to
> the bumbling workman who deviates from the blueprint via
> blunders...but who only very rarely might improve on it. :)

I am working with the subtle distinctions made by Camazine et al.(2001)
between classes of external (to the developing individual, in this context)
sources of information. They distinguish "blueprints" from "recipes" from
"templates" from "leaders." They define a blueprint as specifying the
desired outcome of construction, as opposed to a recipe, which specifies
"how" something is to be built. They go on to explain that recipes allow
for the incorporation of improvements along the way (e.g., taste testing as
you are cooking), while blueprints do not. These semantic distinctions may
seem trivial, but I find them helpful in communicating more precisely about
these issues relating to information sources for physical processes.
 
> Guy:
> I agree that metaphors of the "blueprint" variety always fall apart as we
> learn more. The hope is that we move on to metaphors that are closer to
> Truth, even though we can't know exactly where Truth lies. I hope the view I
> articulated above seems less metaphorical and more mechanistic, although you
> (and many others, I'm sure) might remain skeptical of the idea of a
> "self-organizing system," which represents a mechanistic approach to
> understanding the balance between bottom-up and top-down (e.g., entrainment)
> forces in determining structural dynamics.
>
> Tony
> What I said was _no_ analogy is perfect. All usually fall
> apart as we learn more. You imply that any view which does
> not solely espouse the idea of self-organization is less
> mechanistic and more metaphorical. This certainly is not
> true--just the opposite perhaps.

I did not mean to imply this, and I would not make this argument. Our
current level of scientific understanding of natural phenomena has been
driven primarily by the reductionistic paradigm since the time of Newton.
This paradigm has been extraordinarily useful, perhaps in part because the
scientific understanding of nature relied (of necessity, given the state of
technology) so heavily on holism prior to that time. IMHO, the balance has
swung too long and far in the other direction, so that the greatest
opportunities for growth in scientific understanding now lies in studying
the roles of top-down effects, to bring our theories closer to the balance
struck by natural systems. We should begin to better appreciate top-down
effects and study their mechanisms.

> As I noted above, both
> bottom-up (reductionist) and top-down (holistic) views have
> their uses, and both may be applicable under various
> circumstances.

Yup.

Cheers,

Guy

Relevant Pages

  • Re: Replicating What?
    ... given a genome with about ... The "blueprint" is the information which tells the body how ... You may call this self-organization if you like, ... information to encode your eyes, let alone the rest of you. ...
    (sci.bio.evolution)
  • Self-Organization of Complex Systems
    ... against the very idea of evolutionary biology. ... "And the role of genome as part of a much more complex ... called "Self-Organization in Biological Systems." ... physics or statistics are usually criticized as reductionistic. ...
    (sci.bio.evolution)
  • Re: Whose version of evolution is correct?
    ... testable mechanisms causing genuinely novel ... There is no chemical barrier to any organism evolving into ... if they use the same materials in their genome. ... have to be something in place to prevent said transformations, ...
    (talk.origins)