Of Gods and Homunculi
Roger D. Carlson
Lebanon Valley College
DETERMINISM. Determinism is a necessary assumption in any
science, and as scientists, psychologists must adhere to the
assumption. The search for causes, order, lawfulness implies
that there be determinants to discover and that such
determinants can become known or at least known about. A major
problem in psychology is the devisal of deterministic models and
theories that are inclusive--that account for all of the data
rather than accounting for some and then excluding some as being
inexplicable by the model. Traditional apologists for practices
such as the latter have suggested that psychology is a young
science and being so young, it cannot be expected to have
discovered a complete compendium of determinants, nor can it be
expected to understand the complexity of the interrelationships
which may or must exist between determinants. Others have
subscribed to a kind of "partial" determinism; a determinism
which includes a little bit of "wiggling room" or "free will".
Unfortunately the latter frequently becomes an excuse for
ignorance. Because of the absolutistic nature of the necessity
for a subscription to a paradigm of inquiry if one is serious
about his or her science, such a double standard of
philosophical subscription is not at all satisfactory. Either
the lawfulness of the universe can be discovered or it cannot,
and if one is a true scientist and internally consistent in
one's assumptions, one has to remain sceptical about the
capriciousness of any event in nature. So traditionally the
quest has been in the science of psychology, for more and more
complete models and theories about the phenomena which are
observed in order to close in on left over within-subject error
variance or "wiggling room".
CHANGE, NOVELTY, INVENTIVENESS, EVOLUTION, AND CULTURE. A
critical omission that theoreticians in psychology regularly
skip over is that of the determinants for initiating CHANGE.
Too often our models account for a relatively steady state in
the organism but fail to include an adequate account for
development. By "change" I am not using the trivial sense of
the word at the molecular level; for example, attitude change,
change due to changes in environmental stimuli, and so forth.
But rather I am refering to more profound change--the kind of
permanent evolutionary change that we see with
history--historical change, if you will.
In development, one must account for newness and change of
the the display of the human being in some way. If we conceive
as the person as a mechanism, then in order to get something new
out of this mechanism, one must have a way of explaining what
the source of that newness is.
In a finite processing system, the only way it seems that
one can talk about "processing" which will lead to new output in
e.g. the generation of linguistic output, creative output, or
developmental change, is through the notion of some sort of
random event generator. Otherwise the only sensible prediction
is that if one puts something finite in one is only going to get
some permutation or combination of that which has gone in.
And if you and I cannot predict the creative act or product
of the future, then how can either of us study the determinants
of that creative act? Except on a post hoc basis? How can one
predict what one will say without a knowledge of what has been
said in the past? The study of habits and probabilities only
reflects the study of the non-creative aspects of human
existance rather than that which is truly creative.
Take Piaget's model for example, as has been recently
(particularly clarified by U. Neisser in his book, COGNITION AND
REALITY.) In order to accommodate already formed schemata to
meet the assimilated input, one must have the assimilated input.
Of course, the true nature of the input from the environment,
and that which has been incorporated already within a given
individual, is unpredictable.
One serious question that we must seriously put to rest
before proceding further is: Have we ever changed through
history? Or are the things which we take to be historical
changes simply illusions in our impoverished intellects which
are too limited to note that what we take for change are simply
permutations and combinations of the past?
A serious philosophical and empirical question then, is:
Does history or development ever repeat itself? It seems
difficult to answer this in the affirmative because of the
impossibility of specifying WHEN the "elements" of human
existance were forged, and when the forging stopped, and output
became mere recapitulations (permuted and recombined) of input.
And one could argue, afterall, that with respect to individual
development, lawful descripive trends and relationships are
evidence that change does occur.
Finite state input/output processing models commonly used
in psychology. To say that there are also causes or
determinants for the input leads to an infinite regress that at
some point must be terminated by some sort of event generator
(something like a random number generator (or table) except such
a generator generates qualitatively different events). Such an
event generator may be random or may be orderly, but in either
case it cannot have prior causes without again re-encountering
the problem of the infinite regress. Why the postulation of
such an event generator is resisted is because of our difficulty
in knowing at what point in time such an event generator is
placed. Is the point of observing Brownian movement in physics
the point at which we say that there is NO prior "cause"? The
point at which one places an event generator in our causal
regress governing any human behavior becomes the problem of
where we can truly say that we have backed up to the point at
which we have discovered its source. In summary, then, in order
to account for as well as the variability and novel, creative,
permanent change in output that we observe throughout history,
we need to assume a notion of INDEPENDENT RANDOM OR NON-RANDOM
EVENT GENERATION in the input of the "processing device".
My central thesis in this paper is this: WE COULD NEVER
HAVE A SENSE OF PROGRESSIVE HISTORY--OF TRUE CHANGE--WITHOUT THE
POSTULATION OF A RANDOM (OR NON-RANDOM) EVENT GENERATOR.
(To have all of that "in principle" obtainable knowledge,
would result in our having a perfectably predictable universe.
Ahhhhh--when we only attain that state. The computers would be
generating a continuous role of hard copy just ahead of all of
our actions so that we would know what our actions and the
actions of others are going to be. And what do you think that
we would do with that kind of knowledge? Mmmmmm. The computers
could literally never catch up with our endless capacity to
politic.)
Psychologists interested in operant conditioning have the
same problem. The difficulty is not in explaining how
conditioning occurs, but how the first response comes about.
The "randomly" emitted response. From where does such a
response come?
HOMUNCULI. Where does this unpredictable stuff come from?
Neisser says from where the person "decides" to look. One could
also say from the unpredictableness of actions of others. But
you could say: If we know enough about those other persons to
predict what they would do at any given instant, then we could,
in principle, predict what e.g., John will do.
(Neisser, then, is doing philosophy too (which he denies)
when he invokes theoretical mechanisms such as Piaget's or any
other model which he suggests. It is actually speculative
philosophy rather than analytic philosophy or linguistic
philosophy.)
You can postulate a homunculous for the selection of
information from an infinite array, however then you have the
same difficulty as U. Neisser does in his book, COGNITION AND
REALITY, in explaining the governing mechanism of the
homunculus. How can you explain its activity? Either way, one
is dealing with a variable of non-predictability--in one case
the locus of non-predictibility is in the nature of nature; in
the other case, the locus resides in the ultimate freedom of
humankind. Again, a finite state model ONLY will allow for
output which is the permutation or combination (recombination)
of input (which is of course all that the computer or language
user can do).
It seems to me that one either has to posit a homunculous
or else a God in order to account for infinite variability of
historical change.
REGULARITIES IN CULTURE. Of course, lawfulness itself
could be argued to be an illusion in that what is observed as
orderly and lawlike could in fact be due to certain constants in
the culture which work against true uniqueness. (U. Neisser's
experiments in human capabilties with practice indicate that
some pretty powerful NEW performances can be brought about WHICH
HAVE NEVER BEEN DOCUMENTED BEFORE.) And we DO afterall, have a
"sense" of historical change and human inventiveness in the
qualitative aspects of things.
Psychological theory building is really doing philosophy
that changes our consciousness, yes, but does not state "truth".
It is often said that it "connects facts together," but rather
it arranges facts around a particular ideological
position/statement which divides the "facts" into certain
conglomerates for ease of discourse. In itself though, theory
does NOT reflect facts, nor is it "proven" by facts.
So the question ultimately becomes, do scientific regu-
larities reflect regularity of the universe or regularity in the
culture? OR REGULARILY IN OUR STATES OF MIND (Foucault), IN OUR
CONCEPTS VIS A VIS THE CULTURE (Benedict), OR ULTIMATELY IN OUR
LANGUAGE (Wittgenstein)?
The world has a continuous existance in time and space, as
do our experiences. One never crosses the same river twice. On
the other hand our concepts and language have discrete binary
existances. Either x exists or it doesn't. Either I have a
pencil in my hand or I do not. And the pencil is the "same"
pencil from one moment to the next.
And our language, continually evolves over time through the
homuculated selection (discovery) of new random or non-random
events. Thus, newness and change in human discourse about
things is due not so much to the nature of a language processing
apparatus capable of infinite output as it is due to the
continual event generator in the universe.
RANDOM OR NON-RANDOM EVENT GENERATION. Are psychological
"laws" due to a non-random event generator? That is, is there
truly "order", in the universe? And from where does such order
come? Or is there really random event generation, and our sense
of having psychological "laws" is a result of the orderliness of
our shared cultural concepts (language) about the nature of
things?
And if there is indeed either random or non-random event
generation, are the "choices" of what our homunculus "picks up"
influenced by those events, or does the homunculus remain
inpervious to such effects? And how can we know? If influenced,
from where do these influences come? It seems that it comes
back to a deterministic model. If not, we are assuming a
homunculus that is relatively impervious to such determinants.
And how, then, and under what conditions do we observe some
"lawful" influences over our relatively impervious homunculus?
I think that the solution to these problems is in
postulating either random or non-random event generation in the
universe, and assuming also, a model whereby human beings create
shared linguistic concepts about the nature of nature. Such a
collective human endeavor must then, in fact, be influenced "in
the beginning" by random or non-random event generation, but
whether apparent "non-randomness" (or "lawfulness") is due to
the nature of nature, or the nature of our language as governed
by the random or non-random nature of nature, or the nature of
an, at least, partially deterministically autonomous homunculus,
remains difficult, if not impossible, to answer.
GODS AND EXISTENTIALISTS. Theologians are those who
typically argue for the existance of an orderly event generator.
Existentialists (and anarchists) are those who typically argue
for a random event generator. Whatever we decide is acceptable
for psychology, we must no longer turn away from the problem of
the infinite regress which is implicitly assumed if such a
generator is not assumed to exist, and the impossibility of
accounting for new creative and novel change in the human being
without such a generator to supplement our finite state
input/output processing models.