On Sun, 5 Mar 2000, Cliffe, Owen wrote:
> does evolution have a mind?
>
> it has the ability to respond to stimuli, create, adapt, interact...
> and it is derived from an underlying symbol system
Good question.
You already know the answer intuitively: Call it the "Pinch Test": If
you pinch "evolution," can you hurt it? If not, it's not the kind of
thing that can have a mind.
Besides, what is "evolution"? If you look at the planet, you don't see
"evolutions" walking around, you see organisms that have evolved. They're
the ones with minds.
You say "evolution" has the ability to respond to stimuli. But you
don't literally mean that. What you mean is that when the environment
changes (that's the "stimulus"), and that happens to make some
organisms survive and reproduce better, then that causes the
symbolic genetic code that built the organism do get successfully passed
to the next generation [that's what we MEAN by evolution].
The genetic code itself is revised and updated in this way (whatever
chance genetic change happens to result in survival/reproductive
advantages gets passed on, by definition: that's what "selective
advantage" MEANS).
So what has the ability to respond to stimuli is the organism's genome,
its inherited code, which codes for all of its basic structures and
functions. But even there, it is not the genome that has the mind, but
the organism. You can't pinch and hurt a genome any more than you can
pinch and hurt evolution. Our genetic code is in every one of the cells
in our bodies (and those cells probably don't have minds either, though
they might have, because they too are organisms).
And "evolution" (which is really just a process) can't pass the Turing
Test any more than it can pass the Pinch Test either; only the
organisms shaped by evolution can.
But there is one way in which the question of having a mind does come up
in evolution very much the same way it does in AI: How could "evolution"
ever tell whether an organism had a mind? Or, to put it another way, if
there were a genetic code for having a mind, how could it get selected
by evolution? How could it confer survival/reproduction advantages to
have a mind?
For evolution itself is just a Turing Test too! Passing the
survival/reproduction test depends on FUNCTION, which means it depends
on exactly what a TT can test: It can test what you can DO, and select
an outcome based on your success or failure. So evolution can select
traits that have helped you survive and reproduce better. But it can't
read your mind, and select on the basis of whether or not you have
one. So evolution can't tell whether or not you have a mind, any more
than an AI designer can tell whether his model has a mind; all he can
do is Turing-Test it -- and maybe a mind will somehow "piggy-back" on
the TT success!
With the hierarchy from genes to genomes to one-celled organisms, to
many-celled organisms to families to societies to species to the entire
biosphere, some biologists have asked themselves whether the entire
biosphere (all living organisms, plus their environment) might be one
big living creature ("Gaia") roughly corresponding to the Earth. For
just as organisms come and go in evolution, so do groups, and species,
and, in principle, the whole biosphere.
Back to the Pinch Test: You can pinch and hurt an organism, but not a
family, society, species, or planet. So these are probably all entities
with a few similar functional properties rather than entities that all
have or can have minds, like us.
If you are interested in this question, here is something we won't be
reading in this course (but you are free to read and use it in the exam
if you like):
Harnad, S. (1994a) Levels of Functional Equivalence in Reverse
Bioengineering: The Darwinian Turing Test for Artificial Life.
Artificial Life 1(3): 293-301. Reprinted in: C.G. Langton (Ed.).
Artificial Life: An Overview. MIT Press 1995.
http://www.cogsci.soton.ac.uk/~harnad/Papers/Harnad/harnad94.artlife2.html
Harnad, S. (2000) Turing Indistinguishability and the Blind
Watchmaker. In: Mulhauser, G. (ed.) "Evolving Consciousness"
Amsterdam: John Benjamins (in press)
http://www.cogsci.soton.ac.uk/~harnad/Papers/Harnad/harnad98.turing.evol.html
Stevan
This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:36:27 GMT