From Aaron Sun Dec 22 10:18:59 GMT 1996

Newsgroups: comp.ai.philosophy
Distribution: world
Keywords: consciousness evolution
References: <59htsd$hks@nntp.hut.fi>
Subject: Re: Consciousness and evolution?
Summary: There are discontinuities

phoyer@alpha.hut.fi (Patrik Ole Hoyer) writes:

> Date: 22 Dec 1996 00:03:25 GMT
> Organization: Helsinki University of Technology
>
> ....if one believes in the theory of evolution,
> one would almost have to consider consciousness as a property which is not
> on/off but rather continous-valued, at least according to my reasoning:
>
> If we agree that the earliest life-forms (single-cell organisms etc) where not
> conscious, and consciousness really is an on/off thing, then there would have
> to be some point in time when suddenly, the first conscious animal/being
> appeared, due to some small miraculous mutation... This doesn't sound very
> likely to me, so I would have to give up the on/off theory.
>
> Are there any serious errors in my reasoning?

> ....

I believe many philosophers and scientists have reasoned like this.

There is an error, namely the assumption that there are only TWO
options:

    (a) There's global dichotomy: everything either has
        consciousness or doesn't

    (b) There's a continuum of cases: consciousness is a matter of
        degree and there are no discontinuities.

There is a third option, which I believe is closer to the truth,
namely

    (c) Between the simplest organisms (or designs for behaving
        systems) and the more sophisticated ones there are MANY
        interesting and important discontinuities.

Case (c) can arise if the sorts of organisms that we normally agree
have consciousness (e.g. humans) have a complex COLLECTION of
capabilities, such that different subsets of those capabilities are
capabile of existing.

My claim is that what we normally refer to as consciousness is a
very large, and somewhat ill defined collection of capabilities,
some of which are independent of others, so that different subsets
can occur.

If there are N such capabilities, then in theory there are 2**N (2
to the power N) possible distinct subsets. However, some of the
capabilities may be linked in such a way that not all of the
mathematically possible subsets can occur: e.g. the ability to learn
to read verbal texts and the ability to learn to sightread music
cannot occur without certain other visual abilities. So not all 2**N
subsets can occur.

Nevertheless, between the simplest forms of awareness of some aspect
of the environment, or the simplest forms of self monitoring in
organisms, and the most sophisticated collections of perceptual and
self-monitoring capabilities, very many steps are possible.

I conclude that there are probably large numbers of DIScontinuities
in design and in functionality that are worth studying, in
psychology, brain science, ethology, evolutionary theory and, of
course, AI.

It's not always noticed that Darwinian theory implies that
evolutionary development MUST be discontinuous in the mathematical
sense, because there can only be a finite number of generations
between any two time points, whereas continuous evolution would
require infinitely many intermediate states.

There may be continous change in an individual (e.g. physical
growth) and if Lamarckian evolution (inheritance of acquired
characteristics) were possible, then evolution plus individual
development could form a continuous process.

But not as things are.

The discontinuity also follows from the fact that DNA is made up of
discrete components, hence changes in DNA must be discontinuous.

Often when people say evolution is continuous, what they mean is
that there are lots of small changes. Explicitly acknowledging the
discontinuities is important because that identifies an important
research problem: to identify and explain them. Maybe some of the
changes are much bigger than others.

My own expectation is that when we understand these things fully
we'll find many discontinuities even between the "consciousnes" of a
new-born infant and the "consciousness" of a normal adult, as well
as discontinuities in evolutionary development of mental
capabilities.

The various kinds of abnormality resulting from brain damage, brain
disease, genetic abnormality, environmental abuse, effects of
ageing, etc. will also reveal many interesting discontinuities
between human individuals.

For more on this see the following:

    http://www.cs.bham.ac.uk/~axs/misc/consciousness.rsa.text
        Summary of a talk on consciousness presented at the RSA
        London Feb, 1996.

    http://www.cs.bham.ac.uk/~axs/misc/consciousness.lecture.ps
        Postscript file containing slides prepared for the
        lecture (needs A4 paper for printing)

There are other relevant things in this directory
    http://www.cs.bham.ac.uk/~axs/misc/

[Please note that I am NOT one of those who regard it as important
to maintain a discontinuity between humans and other animals for
religious reasons nor because of any desire to maintain that humans
are better, more important, more advanced, etc. For all I know, or
care, evolution or robotic development may be able to produce
animals or robots that are far superiour to human beings, morally,
intellectually, etc.]

Aaron



(perhaps not all 2**N