[From Bill Powers(2002.09.25.0229 MDT)]
Dan Palmer (2002.09.25.1157 Melbourne Time)--
Some recent posts to CSG-net have made partial reference to Gregory Bateson's
criteria of mind or what he also calls mental process. Just for the record, I
thought I'd list them more completely here:
1. A mind is an aggregate of interacting parts or components.
2. The interaction between parts of mind is triggered by difference, and
difference is a nonsubstantial phenomena not located in space or time;
difference is related to negentropy and entropy rather than to energy.
3. Mental process requires collateral energy.
4. Mental process requires circular (or more complex) chains of determination.
5. In mental process, the effects of difference are to be regarded as
transforms
(i.e., coded versions) of events which precede them. The rules of such
transformation must be comparatively stable (i.e., more stable than the
content)
but are themselves subject to transformation.
6. The description and classification of these processes of transformation
disclose a hierarchy of logical types immanent in the phenomena
I've seen these statements before, but it always seemed to me that in
seeking for the ultimate degree of generality this approach to knowledge
loses substance. Consider the first one:
A mind is an aggregate of interacting parts or components.
Is there anything in nature of which this is _not_ true? Quarks, perhaps,
assuming they don't get broken down into still smaller particles. If
everything is an aggregate of interacting parts or components, then this
ceases to be a "difference that makes a difference" for the purpose of
distinguishing mind from non-mind. It's like saying that paintings have
weight. This admittedly true statement doesn't distinguish one painting
from another, or from things that are not paintings. So it's not helpful
for defining paintings.
My problem with the second statement is of a different kind: it asserts a
theory without proof.
2. The interaction between parts of mind is triggered by difference ...
It's not "difference" that caught my attention, but "triggered." That is a
very specific mechanism: a trigger stays in one stable state until a great
enough force is applied, whereupon it switches all the way to a second
stable state. Furthermore, a process that is triggered is restrained or
inactive until the moment of triggering, and then it begins to proceed
according to its own internal rules, no longer connected to the triggering
event -- that is, it can't be un-triggered or be further affected by the
triggering mechanism.
So this is very clearly a digital view of mind, at least in great part
(parts of the triggered process could be analog in nature). And I think
that Bateson's conception of "difference" is, in the final analysis, also a
digital concept: there is either a difference that makes a difference, or
there isn't. This is far from the analog world where big differences make
big differences, and little ones make little differences. In the world of
continua, "zero" is an extremely rare condition.
Statements 3, 4, and 5 have, in my opinion, the same deficiency as the
first statement. They apply to almost everything and thus don't distinguish
mind from non-mind. This is a big problem with this sort of generalization.
You can make generalizations that appear to be completely and
self-evidently true of the thing you're trying to define, but in the flush
of this positive success it's easy to overlook the great number of _other_
applications in which the very same generalizations would also be true,
thus rending them trivial. Paintings have weight. By golly, that's true of
every painting I can imagine, so it must be a great Truth about paintings,
right? Wrong, because it's also true of everything else on earth made of
matter.
As to 6., I have to agree, of course. Look at the date on Bateson's list,
though. Note, however, that hierarchical organizations are not confined to
_conscious_ minds. I'm pretty sure that Bateson equated mind and
consciousness. I don't.
I have had much the same reaction to many cybernetic concepts -- for
example, Ashby's Law of Requisite Variety. This Law is certainly true --
the "variety" or number of distinguishable states in the output of a
control system must match the variety of the environment in which control
is to take place, and if you look at any successful control system, you
will see that this condition holds true. However, given only a control
problem and the Law of Requisite Variety, you will find the law of no use
whatsoever. It can be applied only after the fact. What happens is that as
a control engineer identifies the variables to be controlled, the "variety"
of the environment becomes defined, and as he provides sensors and
actuators to deal with each of the variables, the "variety" of the control
system comes to match that of the environment. But he never uses the
concept of variety in doing these things.
If you simply omitted the step of evaluating the varieties, you would end
up with the same design. You don't need to know the values of the varieties
to do a successful design; variety is merely a derived quantity that has no
functional role in a control process. Variety-matching occurs if control is
successful, but if control is not successful, you can't make it so by
"matching varieties."
Oh, well -- I'm still involved in arguments that went away 20 years ago
when we stopped going to meetings of the ASC. I didn't win them then, either.
Best,
Bill P.