(The listserver will see to it that this is my last post of the day!)
On Mon, 11 Dec 2006, Paul Brandon went:
[Skinner] never said that thoughts and feelings were irrelevant to
behavior (that was Watson -- a difference that Chomsky seems to have
missed), simply that they were behaviors to be explained, not
explanations of behavior.
So Skinner is rejecting model A in favor of model B:
(A) external events --> thoughts and feelings --> observable behaviors
(B) external events --> internal behaviors (thoughts and feelings)
\-> observable behaviors
In model B, thoughts and feelings *are* irrelevant to any discussion
of *causes* of behavior. They are, as Skinner calls them,
epiphenomema.
(Personally, I think that both models are probably correct in
different instances, and I'm not sure that either one can be falsified
with available technology. But let's go back to Skinner.)
I believe that his point was that it was not necessary to invoke
inner agents in order to understand and predict behavior. His point
in regard to data was that it supported his statement that he could
adequately account for (control and predict) behavior simply through
knowledge of contingencies past and present, within the bounds set
by genetics.
Even if that were true, it would not logically imply that thoughts and
feelings can't _mediate_ the relationships between contingencies and
behavior, as in model A above. Skinner repeatedly asserted that such
mediation does not occur. Based on what? He doesn't say.
Actually, Beyond Freedom and Dignity (1971) should be read as a
package with About Behaviorism (1974),
OK, I'll take a look at that.
On Mon, 11 Dec 2006, Bryan Midgley went:
To say that thoughts/feelings are irrelevant is too extreme, as has
already been pointed out by Paul. Irrelevant to what? To
understanding the human condition? To controlling behavior? They
are links in the chain (Skinner, 1953) and therefore relevant.
When you say thoughts/feelings "are links in the chain," do you mean
they *do* mediate relationships between contingencies and behavior?
If that's what Skinner said in 1953, he turned around and explicitly
disavowed it in _Beyond Freedom and Dignity_.
But to control behavior, we go back to the environment, as I _think_
every psychologist must.
But if cognition is a "link in the chain"--a mediator--doesn't it
follow that behavior should also be changeable through, say, cognitive
restructuring? Any practicing clinician would probably want to know
that such an option is available.
for every successful environmental intervention, for every behavior
that is successfully controlled via contingencies, without
consideration of the inner agents, that's evidence that they need
not be the focus (as in "autonomous man").
There's a tremendous difference between "they need not be the focus"
(a reasonable and empirically supported position) and "they should not
be the focus because they have no effect on behavior" (an
extraordinary claim).
On Mon, 11 Dec 2006, Marc Carter went:
One was an argument from parsimony. If you can explain the behavior
with genetics, situation and history, then you have no need to
resort to mentalisms to explain the behavior, and thus probably
should not.
Yes, I got the impression that Skinner thought parsimony was on his
side. But parsimony is only a dictum, not a type of evidence. I'll
see his dictum and raise him two other dicta: first, extraordinary
claims require extraordinary evidence; second, things should be made
as simple as possible, but no simpler. We could play this game all
day, but all we're doing is tossing the burden of proof back and
forth. :)
Skinner might have been justified in concluding: "At present, we can't
know whether conscious internal states are the proximal causes of
behavior. Considerations of parsimony lead me to believe that they're
not." He was not justified in concluding: "They're not." He says the
latter with utter authority.
The second was an inductive argument rather than a directly
empirical one. He showed that many behaviors could thus be
explained, and with the exception of language (to my thinking, not
his), there are no qualitative differences between keypecks and
complex human behavior; ergo, complex human behavior could be
explained in the same way.
Actually, even Skinner (to his credit) acknowledges throughout _Beyond
Freedom and Dignity_ that language adds a dimension to the control of
behavior that is otherwise not present--for example, the ability to
learn a new contingency without having directly experienced it. He
stops short of admitting that this is a qualitative difference, but it
seems to stretch the boundaries of a merely quantitative one. Which
makes me wonder whether Skinner was being deliberately obscure in
_Beyond Freedom and Dignity_ about the extent to which his assertions
were based on data from nonhuman animals.
With that, I'm going to re-re-reiterate my main point one more time
and then go away: I'd have no quarrel with _Beyond Freedom and
Dignity_ if Skinner had made any distinction between what can be
directly inferred from available data and what can only be argued at a
philosophical level. I'll see whether he cleared things up in _About
Behaviorism_.
--David Epstein
[EMAIL PROTECTED]
---
To make changes to your subscription go to:
http://acsun.frostburg.edu/cgi-bin/lyris.pl?enter=tips&text_mode=0&lang=english
