Starting Over?

[Note from Gary Cziko: For some reason Rick Marken is unable to post
directly to CSGnet and as you might expect, it's driving him crazy.
Everything he posts now has to go through me first. At least this way I
get to clean up some Rick's language and spelling!]

[From Rick Marken (930304.1400)]

There are two kinds of psychologists who embrace Perceptual Control
Theory (PCT). One kind (including myself) believes that we should
start psychology over from scratch based on the principles of PCT.
The other kind (probably the majority, unfortunately) believes that
PCT should build on the psychology that has already been done. I
think it is clear that many non-PCT psychologists are turned-off by
the revolutionary implications of PCT; so PCTers of the first kind
are not likely to win many recruits from the ranks of conventional
psychology. On the other hand, it is also clear that many non-
revolutionary PCTers (like Carver and Scheier), by not abandoning
convention, have had to abandon some of the fundemental tenets of
PCT; so PCTers of the second kind are likely to recruit conventional
psychologists to a model with the right name (PCT) and the wrong

PCTers of the first kind argue that MOST conventional psychological
research can be ignored because research carried out in the framework
of the wrong model is not likely to provide the kind of information
that is needed by a PCT modeller. PCTers of the second kind argue
that data is data; if the data can be handled by a conventional model
then PCT should be able to handle it too (if PCT is really worth its
salt); moreover, by ignoring conventional data PCTers of the first
kind give the impression of being evasive.

So here is a chance for the two kinds of PCTer to slug it out again.
The most recent issue of American Scientist (Jan/Feb 1993, v.81)
contains an article by Harold Pashler (a psychologist at UC San
Diego) on "Doing two things at the same time". It is straight
cognitive psychology; no mixer, no ice. It is also apparently
"state-of-the-art" because American Scientist (I think) tries
to present the latest, the best and the most interesting science
in various fields to a general scientific (or intelligent lay)
audience. Thus, I submit that it might be to discuss this article
from the points of view of the two kinds of PCTer.

The experiments described are VERY simple (from a conventional
perspective); the subject is presented with a higher or low
tone (S1) followed, after a variable delay, by one of two possible
letters (S2). The subject is to say, as quickly as possible, whether
the tone was high or low and press a button to indicate whether the
letter was an A or B. So the subject must do two things -- respond
to S1 (tone) and S2 (letter). As the interval between S1 and S2
increases, the time to respond to S2 decreases. This result
(and the results of other experiments) is consistent with the
idea that there is a "bottleneck" in the "processing" of the
two stimuli; the "processor" must finish processing S1 (deciding
which response to make, say the auther) before it can START
processing S2. The model of these results treats the stimuli as
"clients" that must be delt with one at a time by the processor.

Now, how would PCT explain these results?