Autoshaping; Herrnstein's girlfriend

[From Bruce Abbott (951129.1930 EST)]

Bill Powers (951129.0600 MST) --

    Bruce Abbott (951128.1750 EST)

    People are reluctant to give up a theory that seems to make sense
    of a great number of observations in favor of a new view whose
    ability to account for at least some of these phenomena has yet to
    be demonstrated.

If you allow people to know the results of an experiment before they are
asked to explain them, most theories can be made to work.

Post hoc analyses are always less satisfactory for this reason. Still,
there are potential findings that can't be made to fit a reasonably "tight"
theory no matter how hard you try. When that happens you know the theory is
in trouble.

But you seem to be assuming that reinforcement theory can only account for
data in this post hoc fashion. On what evidence?

    An example: How about autoshaping?

At what point was reinforcement theory used to predict this result? Was
it before or after the phenomenon had been observed? From what you say,
it was afterward:

In this case, it was: the finding was not expected because it had been
assumed up until then the pigeon's pecking was a pure discriminated operant,
as opposed to a respondent, or stimulus-elicited response. However,
experimental work undertaken to discover what was going on in autoshaping
disclosed that the initial keypecks were in fact conditioned respondents,
which obey the laws of classical, not operant, conditioning. Thus the
pigeon's peck turned out to resemble the eyeblink, which a person can
voluntarily perform but which also occurs as an unlearned reflex. Once this
was understood, autoshaping was no longer either a mystery or an anomaly.

This explanation works because new assumptions are introduced to make it
work. An illuminated key looks as if it might be edible. Classical
conditioning is the cause of the initial key-pecking. The role of the
illuminated key changed from CS to discriminative stimulus.

Incorrect. The explanation emerged as a result of experimental testing to
determine what was actually going on in this situation. It had nothing to
do with guesses about what the illuminated key looked like to the pigeon.
Classical conditioning was demonstrated, not conjectured. The key did not
change roles. Once pecking to the key is reinforced when it is illuminated,
one can demonstrate that presence/absence of key illumination controls (in
the EAB sense) the peck, because key pecking occurs during illumination and
not otherwise. This discriminative stimulus function coexists with the
illuminated key's function as a classical CS. One can demonstrate every
component of this account empirically.

None of these assumptions is itself the outcome of a test conducted
during the same experiment.

Right. Further experiments had to be conducted to discover what was going on.

Rather, they are brought in in order to make
the theory explain what has already happened, because the theory, by
itself, offers no explanation and could not have predicted this effect.

Wrong. The proposed explanation was not offered to explain what had already
happened, but as a conjecture to be tested in subsequent experiments. Those
studies supported the analysis, which is why I offered it.

If you can pick and choose your imagined facts and which other theories
will be invoked when needed (and then be dropped again when no longer
needed), you don't have an explanation at all; you have a plausible
story and nothing more. You can "predict" only after you know what
actually happened, which is no prediction.

You bet! But this isn't what was done.

Suppose I were to ask that reinforcement theory explain behavior in a
new situation strictly on the basis of assumptions which are stated
beforehand, including any adjustable parameters you wish to provide. How
many phenomena would you then claim that reinforcement theory has
explained and can explain in the future? I think that the number would
shrink dramatically. It might even turn out to be zero.

That's an interesting conjecture. And that's all -- a conjecture. I'd like
to know the answer to that one myself. While we're asking, how many new
phenomena will PCT explain in the future?

But you DO have to add new rules, when and as needed. What determines
beforehand whether you will use classical conditioning as part of an
explanation? What determines when you will assume that a lighted key is
treated as something edible? On what basis will you decide that a CS has
turned into a discriminative stimulus? You have so many wild cards up
your sleeve that you can hardly go wrong -- if you know in advance what
results you will have to account for.

Assumes facts not in evidence. Not the procedure actually followed.

predictions of PCT are true predictions; the explanations given after
the fact are the same as those given before the fact, with no new
assumptions except, of course, in the cases where we simply used the
wrong model and had to start over. But those cases are important too,
because all predictions, if they are really predictions, can go wrong.

Brown and Jenkins made predictions too, but the predictions went wrong
because the model was wrong. The model was wrong because at that time
researchers in this area did not know that the pigeon's peck was subject to
respondent conditioning.

An explanation given after the fact, with new assumptions allowed that
were not originally part of the explanation, is only a fabrication, with
no scientific value. It has very little danger of failing, because you
can always come up with a new assumption that makes the theory right
after all. The problem is that you never know WHICH assumptions to make
before the experiment is done. You have to wait until you see the
result, and THEN choose the assumptions that make the explanation fit.
That is truly cargo-cult science: the appearance without the substance.

True, but then I don't know anyone in EAB who would disagree. It's not the
way they work.


Re: Herrnstein's girlfriend [Bruce Abbott (951129.1040 EST)]:

With Herrnstein's experiment, I've offered the exception that tests the rule.
Unless you can demonstrate that the exception does not apply, the rule is
hereby disproven.

I'm still waiting for your reply, may I expect it soon? (:->

Rick's thermostat argument is irrelevant, because the only systematic
difference between the slides was that they either did or did not contain an
exemplar of the concept being tested for, and the exemplars did not all
share the same features in common.