These are excerpts and elaborations from my book "The Nature of Consciousness"
|
Universal Selectionism The thesis of the US
psychologist Gary Cziko is that there is a universal process of Darwinian evolution that
is responsible for knowledge at all levels, and not only at the biological
level. Cziko relates "knowledge"
to the fitness of living beings, to their adaptation to the environment.
Knowledge as the product of the interaction with the environment (and as the
necessary cause for survival in that environment) implies that all knowledge is
created through a Darwinian process of blind variation coupled with
environmental selection. Cziko observes that there are really
two kinds of fitness: living beings are adapted to their environment when they
are born, and living beings are capable of adapting to changes in their
environment during their lifetime. A theory of fitness has to deal with both
forms of fitness, the one that has been shaped over the centuries and become
part of a species' identity and the other that is shaped over an individual's
lifetime and becomes part of the individual's identity ("instinct"
and "learning"). The Austrian ethologist
Konrad Lorenz saw instinct as having been shaped by blind variation and natural
selection: it is "knowledge" acquired over millennia that (in modern
terms) is now encoded in the genome of a species. However, the same behavior
does not yield the same outcome unless the environment remains exactly the
same; which, in general, it doesn't. William James noted that a living being is
capable of achieving consistent goals using (slightly) different behaviors.
Individuals can "adapt" to circumstances. Examples of
"ontogenetic" adaptation are the muscles that get bigger the more
they are used and the immune system, that "learns" what antibodies to
make based on which ones are "used" to fight antigens. The immune
system is particularly effective in its job. However, it operates on an
absolutely blind basis: it creates all the time a lot of different kinds of
antibodies hoping that, when attacked, at least one will work against the
invader. It is the diversity of its army of soldiers (and the fact that they
are permanently available) that makes it effective in fighting the enemy, not a
careful training of each soldier and a timely deployment of them. It's
diversity and continuity that matter. And they are due to a process of blind
variation, not to a process of careful engineering. (Another key feature for
the proper functioning of the immune system is, of course, that it produces
only antibodies that destroy antigens and no antibodies that destroy body
cells; in other words, it is capable of distinguishing self from nonself). The "blindness" of
selectionist processes turns out to be an advantage in other ways as well.
There are several examples of "functional shifts", i.e. of parts that
evolved for a purpose but then ended up being used for a different purpose,
simply because it worked (what the US biologist Stephen Jay Gould termed "exaptation").
The current use of an organ or behavior does not necessarily explain its
origin. It may well be that it originated for a different function. The other major example of
"ontogenetic" adaptation is the brain that is shaped not only by the
genes but also by experience. Interaction with the environment
"selects" which synapses are useful and eliminates the ones that are
useless. Drawing from Konrad Lorenz's evolutionary epistemology of 1941, Cziko views knowledge as adaptation
of the brain to the environment. A-priori knowledge (the innate knowledge of
entities such as time and space) is the product of the biological evolution of
the human brain. During our lifetime the senses provides us with true
information about the environment because they have been selected over the
millennia based on their usefulness. Therefore the world is not an illusion,
and we know it by adapting to what it is. Cziko observes that
"fitness" has to do with "purpose": there is fit when the
structure of an organism serves a purposeful function. Animal behavior is
purposeful and changes the environment that operates on the animal's behavior.
Stimuli influence responses, but responses also influence stimuli. This is
William Powers' "perceptual control
theory", according to which behavior controls perception as much as
perception determines behavior. A control system is as blind as the immune
system that creates an army of antibodies. Nonetheless, a control system
exhibits a behavior that appears to be "purposeful". It is, in turn,
"controlled" by higher-level control systems. An organism is
ultimately a hierarchy of control systems, each of which senses something in
the environment and tries to control it. Instinctive behavior is the result of
the interaction between control systems that have internal goals. Each control
system must have survival value if it is still part of an organism. In a sense,
there is no learning: there is just the blind functioning of a network of
control systems. In another sense, that "is" precisely what we call
"learning": a control system at work. When something changes in the
environment, the control system senses it and needs to restore its internal goal.
It does so by triggering random responses and rewarding the ones that move it
closer to its goal. A hierarchy of control systems can create the illusion of
learning and of intelligence (as in Valentino Breitenberg's progressively complex robots). Cziko argues that there is a
Darwinian selection not of behaviors but of control systems. Knowledge is the
result of a hierarchy of selectionist processes, starting with the biological
one studied by Darwin. "Learning" is
necessary because animals need to adapt to changing environments. The brain and
the immune system allow animals to find food and to fight lethal viruses.
Humans have also developed a higher form of "learning" that consists
in cultural knowledge. Cziko shows that it too obeys a
process of blind variation and selective retention. Even when we learn
something from somebody else we are simply interacting with the environment
(the "somebody else") and fine-tuning our knowledge based on that
interaction. Both technological and scientific evolution, for example, are due
to a number of different "trial and error" processes. Cziko applies "universal
selectionism" to a number of different fields. The most interesting is
language. Following the US linguists Elizabeth Bates and Brian MacWhinney ("Competition, variation,
and language learning", 1987), Cziko shows that a selectionist approach
can well complement Chomsky's nativism to explain how children learn language. There might be
innate linguistic skills in the brain (selected over the millennia by
evolution) but children learn a language by the same "trial and
error" process that Nature employs everywhere. Children try words and
sentences and reinforce the ones that "work", just like the brain and
the immune system try synapses and antibodies and reinforce the ones that work
(or, better, the ones that work are reinforced by the positive outcome). Back to the beginning of the chapter "The Evolution of Life: Of Designers and Design" | Back to the index of all chapters |
|
|
|
|