[From Bill Powers (960922.0630 MDT)]
Rick Marken (960921.1500) --
Thanks for the commentary on Kleiber -- you saved me the trouble. Klieber's
is the sort of discussion that keeps turning up among those who don't
understand how goals and purposes work. This argument against purpose -- the
future can't affect the present -- is based on the syllogism that says (a)
all events are caused by antecedent events, (2) the future is determined by
the present, and (3) therefore future events cannot affect the events that
lead to them. This argument reveals that the person making it can't conceive
of a way in which something that has not yet happened can be determined by a
present-time intention and a continuing process of control.
That argument against purpose or intention has been around for at least 75
years. The other main argument confuses intentions with predictions. "Sure,
you intend very strongly that you will arrive at the grocery store today,"
goes the second argument," but how do you know something won't happen to
interfere with your plan?" This shows that "intention" is being understood
to mean some magical force of will, so that the more strongly the intention
is held, the more likely it is that the predicted future event will occur.
What is missing, obviously, is the understanding that when interference does
occur, threatening the outcome, the actions will simply change and the
outcome will occur anyway -- if physically possible. Not only the outcome
but the path to the outcome is controlled.
Suppose a scientist made this prediction: "The temperature of this vat of
chemical reactants will increase linearly to 80 degrees c and then remain at
that temperature." The reagents are dumped into the water bath, and we then
observe that the scientist is slowly manipulating a knob that affects the
current in a heater under the water bath. What will we say when the
scientist triumphantly points to a record of the temperature, showing that
it followed exactly the course he "predicted?" We will say that he cheated.
He didn't make a real prediction because he was controlling the water bath
temperature himself all through the "experiment." We all recognize
intuitively that a prediction isn't really a prediction if the effect being
predicted is under continuous control by someone. Anyone would say "If you
can add or subtract heat to the reaction whenever you want, you can make any
reaction seem to follow any temperature curve you want." In other words, we
all recognize that human intentions can change apparent causal relationships
in the non-living world.
But such intuitive arguments made by a person who doesn't understand PCT
can't stand up against the cool logical reasoning of a scientist who doesn't
understand PCT either.
The biologists and others who argue "functionally" are in the same fix: they
don't understand how certain outcomes can become all but inevitable because
they are part of a control process. And they can't distinguish between
outcomes that are just side-effects and those that are controlled relative
to a specific reference value. To them, the statements "a bird flies back to
the nest in order to feed its young" and "evolution produced wings in order
to allow birds to fly" seem perfectly equivalent. They don't distinguish the
purposive process from the non-purposive outcome for the simple reason that
their underlying beliefs say that NO behavior is REALLY purposive -- we just
talk that way for convenience.
···
------------------------------------
Bruce Abbott (960921.1145 EST) --
Evolution as described by
Darwin is undirected and apurposeful; "nature" intends nothing: what is,
simply is. The turtle's shell is the product of a "walk" during which
current properties of the organism at any given time, the situations in
which it found itself at those times, blind variation, and selective
retention combined in such as way as to result in the current properties of
turtles. There was no "intention" to produce a shell and, had circumstances
been slightly different during the course of the animal's evolution, a
different arrangement doubtless would have emerged. The shell is not a
"design" in the sense of a planned arrangement, and it has no "purpose" in
the sense of an intended function.
On the other hand, it can be awkward to talk about biological systems while
assiduously avoiding such language.
But why should one "assiduously avoid" such language in talking about
biological systems just because it is incorrect in talking about evolution?
Render unto Caesar that which is Caesar's. If you're talking about a
purposive system, use the language of purpose; if not, don't.
Teleological argument: The animal aims to survive. Brain is more >
essential for survival than fat tissue. Therefore the animal maintains >
brain weight at the cost of fat.
Biological argument: Animals which are so organized that in times of
starvation their fat deposits are maintained at the cost of the brain have
less chance of survival than those of animals in which brain is maintained
at the cost of fat deposits, and which are therefore favored as the >
fittest by natural selection.
Neither way of putting it is correct, because both attribute purposes which
are unlikely to be represented as reference signals in any given animal. And
the two statements are not about the same thing.
First statement: Animals do not have the aim of surviving. This aim does not
involve knowledge (in a mouse, for example) that brain is more essential
than fat for survival. And the animal does not maintain brain weight "at the
expense of" fat "in order to" survive. There are control systems in the
animal that vary fat content, but none that vary brain weight. Period.
Second statement: Animals survive or they do not survive. A given animal
does not have a "chance of surviving." Chance implies multiple instances;
one animal gets exactly one opportunity to live with the equipment it has.
The first statement is about each animal; the second is about a species and
not any individual. It is possible that animals which have survived contain
control systems for varying fat content but not brain weight. But animals do
not acquire control systems for varying fat content, and avoid acquiring
control systems for varying brain weight, in order that the species should
survive. Neither does evolution vary the kinds of control systems that a
species has in order for the species to survive or in order that it be more
"fit."
We should assiduously avoid purposive language in speaking of systems and
processes that are not purposive, and we should assiduously avoid
non-purposive language in speaking of systems and processes that are
purposive. But in order to do that we must be able to tell the difference.
Best,
Bill P.
-------------------------------------------------------------------------