[From Bruce Abbott (950624.1030 EST)]
Bill Powers (950623.1850 MDT) --
Bruce Abbott (950423.1535 EST)
Now that I've recovered from paroxysms of laughter, I believe I'm ready to
respond to your post concerning "a big strong rat." So, down to business...
I can't help horning in and seizing the opportunity.
I'm not surprised--but let's not get confused about what we're doing here.
What I was trying to do was to indicate how reinforcement theory would
handle the effect described. I know how PCT does it. This is not an
argument about which view is correct, but rather, it is an attempt on my
part to impart some knowledge about reinforcement theory. I believe a
proper understanding of the theory is essential if it to be successfully
attacked.
If it's a higher-level control system, then the variable involved
needs to be included in the reinforcement model, doesn't it? I
would have to know what is reinforcing the behavior of changing the
target position and thus violating the instructions given for the
task (i.e., keep the cursor on the visible target as much as
possible).That's the nice thing about doing away with superstitions like
reinforcement. NOTHING is reinforcing the behavior of changing the
target position and violating the instructions. The subject did that
because that is what the subject wanted to do, to accomplish some
purpose not yet announced. The instructions have no more power over the
subject than the target does. It is not necessary for there to be a
reinforcer to account for any behavior, or even a stimulus when it comes
to that (another superstitious idea). We don't need these intervening
variables to account for behavior using PCT. We simply accept the goals
we find to be in effect, and the control systems that accomplish them by
the means available.
I understand that this is your position, which is derived from PCT. But it
is also your position that reinforcement theory cannot apply here, and this
is incorrect. Let's examine your statements carefully:
That's the nice thing about doing away with superstitions like
reinforcement. NOTHING is reinforcing the behavior of changing the
target position and violating the instructions. The subject did that
because that is what the subject wanted to do, to accomplish some
purpose not yet announced.
From the point of view of reinforcement theory, SOMETHING must be producing
this behavior; the question is what. The answer, for the radical
behaviorist, is to be found in the current environment and in the lasting
effects of the organism's history of experience with the environment and its
contingencies. If the participant "wants" to keep the cursor on the visible
target, it is because this history and the contingencies currently in effect
(as perceived by the participant as a result if his or her experience)
indicate that such behavior will be reinforced under the present
circumstances. So, in theory at least, If I could become privy to the
appropriate details of your personal history and the sources of
reinforcement in the present environment, I could fully account for your
failure to keep the cursor on target as instructed. In practice, I am not
in a position to know all the relevant details, although I can provide an
educated guess about some of them based on the limited evidence available.
The instructions have no more power over the
subject than the target does. It is not necessary for there to be a
reinforcer to account for any behavior, or even a stimulus when it comes
to that (another superstitious idea). We don't need these intervening
variables to account for behavior using PCT. We simply accept the goals
we find to be in effect, and the control systems that accomplish them by
the means available.
Reinforcement theory, also, would "simply accept the goals we find to be in
effect" given our ignorance as to the conditions that brought them about.
This is not to say that we would be satisfied with guessing about them, not
if our aim is to provide a scientific explanation as opposed to merely
providing a possible account that fits the data. We would devise
experiments to identify the source or sources of reinforcement, the
reinforcement equivalent of performing the Test in PCT. However, my purpose
here is not to prove that your behavior is controlled by some particular
source of reinforcement, but to show that reinforcement theory can offer a
reasonable, internally consistent explanation for that behavior.
In searching out these "sources of reinforcement," reinforcement theorists
are at least taking the position that there are causes of the behavior that
can be identified. In your analysis based on PCT, you seem to be saying
that the organism does what it does because it "wants to," without
suggesting why.
Changing the (actual) target location is a response, too, and has
its own reinforcer, not the one involved in keeping the cursor
close to target.No, it's not a response and it has no reinforcer. Changing the target
position is an _action_, not a response. Actions occur when reference
levels change -- i.e., when intentions change. Nothing outside the
organism causes that. There is no stimulus, so calling the action a
response is simply a mistake. A response to what?
You are giving me the PCT view again, rather than attempting to understand
the reinforcement view.
Reinforcement theorists recognized long ago that certain processes going on
in the brain could not be observed from the outside, e.g., making a
decision. One solution is to ignore these processes and refer instead
strictly to the organism's history of reinforcement, on the assumption that
these unobservable intermediaries must be the products of such a history and
therefore that the observable relationships can be accounted for just as
well by staying with the external observables (the Skinnerian approach).
Another solution is to assume that these internal processes work just like
the external, observable ones do. Why invoke new processes if you can
assume that the same processes are at work internally as externally? This
solution has the advantage of parsimony, although testing for such internal
processes is necessarily indirect.
No, it's not a response and it has no reinforcer.
I can certainly observe that you're keeping the cursor somewhere other than
the place where you were instructed to keep it. That's a response. As to
the reinforcer, what about this one:
I, the participant, will do what the whisperer suggests just to show
that the reinforcement-devotee is wrong (I don't believe he'd pay me
that much, anyway, unless he wanted something else from me).
So, because of your personal reinforcement history, you (a) find showing
reinforcement-devotees wrong to be reinforcing and (b) do not perceive a
contingency between staying on the experimenter's selected target and
earning the cash. It is entirely understandable, from a reinforcement
theory point of view, why your cursor does not stay on the designated target.
Actions occur when reference
levels change -- i.e., when intentions change. Nothing outside the
organism causes that.
And reference levels change because.....? I intend to mail a package at the
post office this afternoon. On arrival, I discover that the post office is
closed. So I give up my intention, and intend to go home instead. Nothing
outside me has caused that? Hmmmm....
There is no stimulus, so calling the action a
response is simply a mistake. A response to what?
Operants are simply emitted; they require no stimulus, not since 1938. Call
them operants rather than responses if you like, it's all the same to me.
In the remainder of this post, you offer a series of possible reasons why
you would not conform to the experimenter's wishes, even when given a
relatively strong inducement to do so ($1.00 per minute). You seem to think
that this behavior shows that reinforcement theory is wrong. It's a common
criticism of reinforcement theory in applied settings that reinforcement
"doesn't work" because Johnny still misbehaves in class even though the
teacher offers Johnny a bribe for his compliance. The problem, from the
point of view of reinforcement theory, is not that reinforcement doesn't
work, but that there are other, more powerful sources of reinforcement at
work which are controlling Johnny's behavior, such as the attention he gets
from misbehaving. You've named several of those sources for your misbehavior:
Actually I would pay no attention to your offer. I
would go on tracking the target because I like Rick and wish to see him
calm and happy, and I find your suggestions improper. What are you going
to want for your $1.00 per second? My mother told me about people like
you.
Actually, Rick offered me $10 per second, unbeknownst to you. I didn't
believe he's pay me anything, either, but I did what he asked just
because that's what I decided to do. It's fun just to watch you and him
fight. When the experiment is over, I'm going to take both of your
wallets anyway, and use the credit cards.
So, as a behavior analyst, my problem is that my paltry $1.00 per minute is
no match, in terms of reinforcing power, for the reinforcement you receive
from Rick's approval, from frustrating people whose suggestions you deem
improper (because of a reinforcement history involving your mother and her
stories about "people like me"), from the money Rick offered you not to
conform to my instructions, and from the fun you get when Rick and I fight
about why you didn't behave as expected. To get you to do as I wish, I must
attempt to (a) eliminate these competing sources of reinforcement, and (b)
identify a source of reinforcement that will work in your case.
Fortunately for me, you have been kind enough to list these competing
sources, which makes my job easier. In the process, you have inadvertently
provided all the evidence any reinforcement theorist could want that your
nonconforming behavior is indeed under the control of a number of potent
reinforcers. Thanks!
How do you like dealing with a BIG STRONG rat?
Little rats, big rats, tigers, elephants, blue whales, size makes no
difference, not when you've got a BIG STRONG theorist who knows how to
handle them. Pretty soon, they're all eating out of his hand . . . (;->
Regards,
Bruce