I was pondering a related question just before these posts, so I'll try to
tie it in. My thoughts were related to the Aristotelian causes and the
definition that Don often quotes from Rosen for an organism, i.e., a
natural (complex) system that is "closed to efficient cause."
This is related in the following way, if I understand these causes and what
we mean by closure: Efficient cause, in this case, I take to mean the
metabolic and repair agents, or "functional components" of the system. In
other words, whatever performs the functions of mapping "inputs" to
"outputs" and maintaining the operability of the system. What we include in
this mapping, however, seems to me to depend on our opinion about the
organism and what we consider its essential functions. Rosen's mathematics
seem to allow us to lump conceptually everything (because his formalism
addresses "why" not "how," we can sweepingly say that all possible "how"
answers are included in a "why" formalism; just as the mechanists have done
the opposite), and thus the statement about closure can be made with
certainty. But I question this certainty. As Francis points out, the
separation of organism and environment is largely a metaphor - one that has
definite limits to its usefulness. To what extent can we truly say that a
cow is closed to efficient cause? Do we consider microorganisms in the
stomach to be part of the cow and do they not perform important metabolic
functions? Do we not consider social functions to be essential to the
survival and maintenance of an organism, for example preening among birds,
grooming among primates, etc.? Are these functions contained in the
organism under this definition? Or are they part of the environment and
thus not "closed" within the organism's definition? So, I'm sure Don will
kill me, but I question the absolute nature of this closure, which seems
similar to the problem of separation of environment and organism. This
leads me to the question about evolution.
If, for the moment, we grant me that "closure" of causes is often a
relative matter, let's look at the other Aristotelian causes too. Are
organisms closed to material cause? We would generally say no. Clearly
inputs of energy and raw materials from the environment are required. But
also organisms manufacture materials actually used for construction from
these raw materials, so it is partly closed to material production. Again,
what is an open system depends on the system boundaries; but I'm
comfortable with the general "no" answer here because of conservation of
energy and matter laws, which means that wherever one defines the system,
unless it is the entire universe, it must be open to some inputs and
outputs. Yet this is similar to the problem with efficient cause - both
depend on where one draws the boundary and we get different answers to the
question because of where we are comfortable drawing the boundary.
Now let's ask about the other kinds of causes. Are organisms closed to
formal cause? That refers to their "blueprint" or design. Obviously we
could focus on the DNA and claim there is some closure. But again,
traditional evolution gives the environment the exclusive role in
selection, so is the environment really the designer even though the
organism carries and replicates the design? Here we must consider time
scale. The evolved organism was clearly open to environmental selection
during its evolution, but today, as it exists now, does it not operate as
if it were largely closed to its own design? Is that not why extinction
occurs (because the closed design doesn't fit the environment)? Now in my
own speculations about "autevolution" I suggest another kind of formal
closure in the ability for organisms to modify their own selective forces
through functional decisions. That allows consideration of functional
design rather than just physical design. It's another wrinkle we can take
up or leave out for now. Then there is the matter of final cause. Again, in
my own views, organisms have subjective experience and evolve through a
complimentarity between their self-defined "purpose" or "function" (meaning
function in life or in the ecosystem, not to be confused with the
functional components discussed under efficient cause) and their
evolved/selected form, resulting from interaction with the environment
(including self-referential loops). Hence, purpose modifies behavior, which
modifies the environment and selection, which modifies form through
selection, which returns to influence purpose - functional (purposeful) and
physical design closure over time and evolution. For my money, organisms
are partly closed to all causes, but it's a relative thing related to where
we draw the line between organism and environment and in which dimensions
we look.
I have to agree with Don, Alexei, and (I think) Francis that
"self-organizing" is a concept that is frought with difficulties such as
described above. What is a "self?" Where do we draw the line? As a minimum
we can identify the four Aristotelian causal levels as different ways for
making the distinction. The kind of "self" I'm thinking about in
autevolution is really at the level of formal and final cause -
self-defined purpose involving final cause to some degree; and life
strategy / function in the ecosystem involving formal cause, where I'm
thinking of the mental blueprint for the organism as opposed to a strictly
physical blueprint. So we have to be clear about the context in which we
are discussing "self-organizing," "self-defining." and
"self-constructing," etc. This seems related to the question of Varela
regarding the "topological boundary" for autopoietic systems.
Also, now that Louis has chimed in on this, I'll remark that his
definitions are state-dependent; whereas Rosen's central thesis was
functional. Does it make any sense to define "self-organizing" systems only
in terms of their states and physical components? I accept that both form
and function are real and important characteristics of organisms and living
systems. Functionally self-organizing (as in organism strategies, as well
as social systems or "enterprises") should be equally valid, as Maturana
argues. But I understand that Varela excludes this because it doesn't meet
physical definitions of component production. Who says components of an
organism are only physical? Or that the physical components are more
important than the functional ones? Again, we have to define the "self"
before discussing what kinds of organization we are interested in, and in
any case I don't think we find neat boundaries. I get more the impression
of a relative "self" differentiated out of a continuum.
A more direct answer to Bruce's question: The term "organization" certainly
involves multiple levels of causes that change over time with feedbacks
(including anticipatory feedbacks), hence evolution MUST be involved in
some way. Otherwise we have to say that crystals are self-organizing, or
for that matter, that everything is; or we would have to identify evolution
simply with change, and that's not its important meaning.
I hope I am using these concepts correctly in thinking about this, but in
any case, please excuse my particular form of insanity.
Yours
P.S. I like what Peter Medawar once said: "For a biologist, the alternative
to thinking in terms of evolution is not to think."
-----------------------------------------------
John J. Kineman, Physical Scientist/Ecologist
National Geophysical Data Center
325 Broadway E/GC1 (3100 Marine St. Rm: A-152)
Boulder, Colorado 80303 USA
(303) 497-6900 (phone)
(303) 497-6513 (fax)
jjk@ngdc.noaa.gov (email)