Abstract. The prevailing common assumptions about how nature
behaves have their origins in the early Enlightenment. The notion
of emergence does not sit comfortably within this framework.
Emergence appears virtually impossible within a world determined
by ineluctable and unwavering natural laws. But the variety and combinations
inherent in living systems render physical laws indeterminate.
The study of ecological dynamics suggests that processes rather
than laws are what accounts for most order seen in the living realm.
As a consequence, there are aspects of ecological dynamics that violate
each of the Newtonian postulates. The dynamics of ecosystems
suggest a smaller set of rational assumptions through which to view
nature—an “ecological metaphysic.” Emergence appears as a rare but
wholly natural phenomenon within the new rational platform. In
addition, several apparent conflicts between science and theism that
arose under the Newtonian framework simply vanish under the new
perspective.
Keywords: causality; chance; Darwinism; determinism; dialectic;
ecology; emergence; evolution; free will; indeterminacy; materialism;
metaphysics; naturalism; Newtonianism
n
of random events upon a configuration of constraints that results in a
nonrandom but indeterminate outcome. A useful example of a simple arRobert
E. Ulanowicz 951
tificial process is Polya’s Urn (Cohen 1976). This exercise begins with a
collection of many red and blue balls and an urn that initially contains one
red ball and one blue ball. The urn is shaken and a ball is drawn blindly
from it. If the ball is the blue one, it is returned to the urn along with
another blue ball from the reserve collection. The urn is shaken and another
draw is made. If a red ball drawn, it and another red ball are likewise
returned to the urn, etc. A first question arising is whether, after a long
sequence of such draws and additions, the ratio of red to blue balls would
converge to a precise limit. It is rather easy to demonstrate that after many
draws, the ratio does converge to a constant, say 0.46967135. A further
question would be what would happen if the urn were emptied and the
starting configuration were recreated. Would a subsequent series of draws
converge to the same limit as the first? It is easy to demonstrate that it will
not. The second time it might converge to 0.81427465. After continued
repetitions of the process, one eventually discovers that the ratio of balls is
progressively constrained by the series of draws that have already occurred.
It likewise becomes clear that the limiting ratio for any series of draws and
replacements can be any fraction between zero and one.
Before going on, it is very useful to note three features possessed by even
the artificial and simplistic Polya process:
1. It involves chance.
2. It involves self-reference.
3. The history of draws is crucial to any particular series.
As an aside, it also is helpful to point out that a mechanical law is a limiting
form of a process. That is, if a process converges to a mechanical-like
behavior (as the Polya process does on those occasions when it approaches
a limit near the extremes zero or one), its behavior becomes indistinguishable
from the action of a law. Generally speaking, however, processes remain
indeterminate in their outcomes. Popper (1990) likewise suggested
that physical forces could be considered limiting forms of more general
interactions, which he called “propensities.” In his lexicon, a propensity was
the tendency for a certain event to occur in a particular context. With a
law, every time A happens, B is “forced” to follow. More generally, however,
when A happens, B usually follows, but not each and every time.
Polya’s Urn is but a hypothetical process, and its constraints are imposed
from without. Bateson (1972), however, provided a clue to where natural
processes might of their own impart order to affairs. He noted that the
outcome of random noise acting upon a feedback circuit is generally nonrandom.
Such bias is especially characteristic of one particular form of
feedback—autocatalysis (Ulanowicz 1986; 1997; Kauffman 1995). By
autocatalysis here is meant any instance of positive feedback wherein the
direct effect of every link on its downstream neighbor is positive.
In chemistry, where reactants are simple and fixed, autocatalysis may be
regarded as simply another mechanism. As soon as one or a few participants
are able to undergo small, incremental alterations in response to stochastic
events, the picture can change dramatically.
It is ontic chance, or
more precisely radical contingency, that makes real change possible.
1. Radical Contingency: Nature in its complexity is rife with singular
events. Most do not upset prevailing regularities, but on rare occasions
one can carry a system into a wholly different mode of emergent
behavior.
The second feature enables systems to maintain their integrities and grow.
At its root lies autocatalytic action, which is a particular form of
2. Self-Influence: A process in nature, via its interaction with other natural
processes, can influence itself.
Third, as Darwin long ago inferred, the system must retain some record of
its past configurations. That is, it must possess a
3. History: The range of self-influence is constrained by the culmination
of past changes as recorded in the system configurations. Such
configurations can be static material forms, as are the genomes of
Darwinian theory, but they could as well inhere in the topologies of
interacting processes.
Starting with these three postulates, one may deduce in logicodeductive
fashion most of the key organic behaviors exhibited by ensemble living
systems, such as ecosystems, immune systems
No comments:
Post a Comment