|
Introduction
It seems clear to me that my subjective mental states are my own.
Furthermore, this act of introspection shows that I have some ability to
think about my mental states. I would say that I am self-aware.
Furthermore, my experience indicates that most humans have such mental states
of their own, including self-awareness. The question thus arises where do
these subjective mental states, including self-awareness, come from? What
causes or explains their presence and content? Has this human capability
evolved?
Continuity between human and nonhuman animals
Studies of a number of nonhuman animals (especially chimpanzees, other apes,
parrots, and dolphins) have indicated that these organisms can respond to
complex stimulation, such as encoded queries and requests, in ways that
resemble our own use of language. These studies are unusual among biological
and psychological experiments in two ways: their small samples of subjects
and their intensive involvement of humans. Nevertheless, there has been an
accumulation of similar results with similar protocols, so that it is
becoming more difficult to exercise broad skepticism about the kinds of
responses evoked.
Interpretations of these responses often devolve into a polarity between
attributing the observed responses to no more than thorough (rote) learning
as opposed to the spontaneity or creativity that human language seems to
show. There are twin problems here -- it is difficult to distinguish
complicated from random patterns of responses and it is also hard to
distinguish repeated from rote responses.
The issue of thorough learning, as opposed to volitional thought, is a
pervasive problem in comparative studies of consciousness. Consider
abilities to respond to oneself in a mirror and to attribute mental states to
others (and by extension therefore to oneself). Even in humans these
abilities clearly require learning. Humans nowadays master some
superordinate associations involving mirrors that most nonhumans have not,
but humans have not mastered all the possible associations. On the other
hand, humans often master well-practiced tasks to the extent that they are
performed unconsciously.
Neurophysiology of consciousness
Suggestive is an experiment that seems to reveal a half-second or so delay
between the initiation of a spontaneous action, on one hand, and awareness of
it, on the other. Such an experiment does not necessarily reveal that action
precedes volition. Instead each of these two operations requires different
neural events lasting finite, and evidently not exactly equal, amounts of
time. Furthermore, the memory is encoded in language, which becomes the sole
means of obtaining the datum actually recorded by the experimenter. Again we
are back to questioning how we can know what another organism feels or
thinks, unless that organism tells us in some way.
The inability so far to find a locus in the human brain specialized for
consciousness has led to proposals that awareness results from distributed
networks of neural interactions. Although some computer programs include
"neural networks", it is still not clear how closely they resemble operations
in a brain. Only at a superficially general level can we suppose that
distributed operations in the brain share the features of computational
"neural networks".
Continuity between brains and other machines
The relationship between consciousness and language also arises in proposals
to distinguish humans from other machines -- or by extension to determine
whether or not any machine is conscious. The issue is whether a
| |
human (conscious) contestant can be distinguished reliably from a nonhuman
(unconscious) one. Searle contends that Turing's test would not distinguish
between a human who understood a language and one who just followed rules by
rote. It thus could not distinguish a conscious human from an unconscious
machine. By extension, it would also not distinguish between a conscious and
an unconscious machine. A fundamental question here is whether or not
conscious behavior, such as language, is strictly rule-following or not.
Noise as a determinant of consciousness
The evolution of consciousness thus depends in a fundamental way on the
evolution of communication. It is thus remarkable that noise influences the
evolution of communication in a way that provides a straightforward
explanation for the evolution of subjective experience.
At the moment of perception a perceiver has no way to determine whether or
not the perception corresponds to a particular external situation or to an
erroneous illusion. All the perceiver knows at the moment is its perception.
Nevertheless, memory of repeated perceptions, especially in combination with
communication with other individuals, could reveal these discrepancies. In
this way such an organism, capable of thought and language, could develop a
sense that its own perceptions differed, in some respects and on some
occasions, from those of others.
Because of noise in perception or communication, a perceiver or receiver must
make a decision every time it acts on any sensation. It must decide whether
the sensation warrants a response (and also which response). In other
words, it must decide whether a sensation is a signal (with some relevance
for the perceiver) or noise (with no, or misleading, relevance). Evolution
by natural selection provides a mechanism that can optimize, within limits,
neural capabilities to make decisions that promote survival and reproduction
for the organism. In a fundamental way, nervous systems are decision-making
organs.
Every organism must confront its subjectivity with some decisions,
no matter how crude the mechanism. Awareness of subjectivity in perception,
however, requires a nervous system to form higher-order associations. The
evolution of this capability seems likely to have emerged gradually by
successively more complex mental associations.
Conclusions
The mathematical analysis of optimal behavior in noisy situations thus
indicates that (1) noise is an inescapable component of communication, (2)
subjective awareness of self is a higher-order association of perceptions and
responses, (3) decision-making is fundamental component of all communication
and perception, and (4) both processes are as unpredictable as the
unavoidable noise. An advantage of this analysis of the evolution of
communication in noise is the framework it provides for addressing the
questions posed at the start of this essay.
To account for the source and the content of self-awareness, previous
discussion has always relied either on supernatural intervention or on vague
neural operations on purified sensations. Supernatural intervention of
course obviates any mechanistic explanation, including evolution. Response
to pure sensations, on the other hand, leaves each organism encased in its
own perceptions, without a way to distinguish between subjective and
objective events.
The evolution of noisy communication, in contrast, shows
that self-awareness (consciousness) results directly from the operations of
nervous systems exposed to noisy sensations. The resulting explanation for
self-awareness will require no unnatural or unspecified components.
|
