IV/DV

CSGnet Archive CSG1998
21

[From Bruce Abbott (980206.0955 EST)]

Rick Marken (980205.1100) --

Bruce Abbott (980205.0945 EST)

I think, Sarsi, that you do know what I am getting at

I think you don't even know who I am. It's Rick here, boobala.

Whichever you prefer. Your call. But you sure _sound_ like Sarsi.

If the perceptual function produces no (zero) signal regardless
of the value of i (assuming i = s + o) then there would be no
systematic relationship between s and o. I don't understand
what you mean by the question "how would the relationship
between s and o _change_?". If a person perceives nothing
regardless of the value of i then the person doesn't perceive i.
This means that there was _never_ a systematic relationship
between s and i for this person so there can be no _change_
in the relationship between s and o.

Wrong. During the course of the psychophysical experiment, the center
intensity of the tone was systematically varied across trials. On some of
those trials, the tone was always clearly perceptible to the participant,
and this was evidenced by the fact that she was able to keep the tone at a
constant intensity despite small disturbances to that intensity. In this
case a specific systematic relationship between s and o appeared. On other
trials, the tone was never perceptible to the participant, and this was
evidenced by the fact that she was unable to keep the tone's intensity
against disturbances. In this case the systematic relationship between s
and o was absent. It would appear that something has changed between these
two conditions. It isn't h() and it isn't g(), and it isn't the ratio
-h()/g(). It isn't a property of the environment that this experiment
reveals. It is a property of the participant's perceptual input function.
At some point along the tone intensity continuum, that function begins to
return a constant p for every i, and the s-o relationship observed at higher
tone intensities breaks down.

OK. So I have answered your question. But I have a problem;
I still think that the relationship between s and o in a
conventional psychophysical task tells you nothing about the
subject's perceptual function. Maybe what I'm supposed to get
is that psychophysical experiments can tell you whether or not
subject's can perceive a variable at all. Is that it? If so, I don't
even agree with that; remember my friends at the draft physical?

If this is your objection, then you'd better be prepared to apply it to
ordinary tracking studies, too. If a given participant fails to track a
target, is it because the participant cannot control this particular
variable, or just doesn't care to? There are ways to find out whether a
participant is being uncooperative. And what about the case where you
yourself serve as participant? The objection does not apply.

Now let's return to a traditional psychophysical study of tone thresholds,
one in which the person has control over tone intensity and is asked to keep
the tone crossing just above and below threshold by pressing or releasing a
button. When the button is pressed, tone intensity increases at a constant
rate. When the button is released, tone intensity decreases at a constant
rate. By pressing until the tone is heard, releasing until it is lost, then
pressing again, and so on, the participant can keep the tone intensity
oscillating closely around the threshold boundary. We have asked the
participant to use the threshold as a reference level, and can infer what
that reference level is from the data. We assess this level at a variety of
tone frequencies, and thereby discover that threshold varies in a smooth way
with frequency, being higher at both low frequencies (e.g., 30 Hz) and high
frequencies (e.g., 16,000 Hz) than at intermediate frequencies (e.g., 2,000
Hz). I observe similar functions in most participants I test. I conclude
that for humans with normal hearing, the tone intensity threshold follows an
inverted U-shape as a function of tone frequency. Apparently, however, you
would claim that what I have learned is merely a behavioral illusion, a
characteristic not of my participants but of the environmental feedback
function.

Regards,

Bruce

22

[From Bruce Abbott (980209.0950 EST)]

Rick Marken (980207.1000) --

Bruce Abbott asked:

How would the relationship between s and o change if for every
value of i, the PIF returned zero?

I replied:

If the perceptual function produces no (zero) signal regardless
of the value of i (assuming i = s + o) then there would be no
systematic relationship between s and o.

Now Bruce Abbott (980206.0955 EST) says:

Wrong.

Let's reexamine your _entire_ paragraph, not just the part you chose to
reproduce:

If the perceptual function produces no (zero) signal regardless
of the value of i (assuming i = s + o) then there would be no
systematic relationship between s and o. I don't understand
what you mean by the question "how would the relationship
between s and o _change_?". If a person perceives nothing
regardless of the value of i then the person doesn't perceive i.
This means that there was _never_ a systematic relationship
between s and i for this person so there can be no _change_
in the relationship between s and o.

The experiment I had described had two phases. In the first phase, the tone
was at a higher intensity, the participant was able to control it, and a
specific relationship appeared between s and o. In the second phase, the
tone was at a lower intensity, because of this the participant was unable to
control it, and the relationship between s and o dissolved. Your claim that
"there was _never_ a systematic relationship . . . so there can be no
_change_ in the relationship . . ." is what drew my reply of "wrong," not
the sentence you choose to quote.

This shows that the person was able to control tone intensity.
It doesn't seem to have much to do with your question or my
answer to it. You asked a theoretical question about the
behavior of a control system that could not perceive variations
in i. I told you how that system would behave. What does a person's
ability to control auditory intensity have to do with how that
person would behave if they could not perceive that variable? Why
does the person's ability to control intensity make what I said
about the behavior of a "deaf" control system "wrong"?

What we were discussing is whether phychophysical experiments could tell you
anything about the perceptual input function. I provided an example to
illustrate how the loss of control (and loss of the behavioral illusion) in
such an experiment can tell you at what point the perceptual input function
goes flat, and you claim not to see any relevance of that fact to the issue
at hand!

It would appear that something has changed between these two
conditions. It isn't h() and it isn't g(), and it isn't the ratio
-h()/g(). It isn't a property of the environment that this
experiment reveals. It is a property of the participant's
perceptual input function.

Or of the participant's reference signal. But you are correct; the
results of this experiment do tell you something about the organism.
That's because this study used the method of adjustment, a
psychophysical procedure that is equivalent to The Test. By applying
disturbances to a variable (tone intensity) that the subject could
also influence you can determine that this variable was under
control; you tested for a controlled variable.

Now you change your mind (apparently) and agree that I am correct that such
a study would tell you something about the organism, but then claim (without
supporting evidence) that this is because the experiment used the method of
adjustment, which gives the participant control over tone intensity.

In reply, I will assert (also without proof) that you will get the
essentially the same result using other psychophysical methods (e.g., the
method of constant stimuli). As you know, in the method of constant
stimuli, the participant would be given a series of tones varying in
intensity, and would reply after each trial "present" or "absent," depending
on whether whe did or did not hear the tone, respectively. The intensity
region within which the response goes from "present" to "absent" or vice
versa will be the same region within which the participant using the method
of adjustment would change from pressing the button to releasing it or vice
versa. I will leave it up to you to demonstrate either emiprically or
through a proper logical analysis that my assertion is incorrect.

If a given participant fails to track a target, is it because the
participant cannot control this particular variable, or just
doesn't care to?

You can't tell. When people are _not_ controlling a variable you
can't learn much about their ability to control that variable.

True. You have to trust that your participant is really trying, but this is
the same for any application of the Test. But if the relationship between s
and o deteriorates (and those famously high correlations become low
correlations) when the tone drops below a certain intensity, it is almost
certainly the case that control is failing because the tone can no longer be
perceived (p = 0, although i is > 0).

Now, here is a puzzle for you. As f(i) appears nowhere in the equation
relating o to s, how is it that this experiment tells us something about f(i)?

. . . This
is why I say that the results of conventional psychological research
can be a good _strting point_ for doing PCT aware research; but,
as they sit, the results of conventional research tell us little
or nothing about the nature of the organisms under study.

This conclusion stands in stark opposition to the facts I've introduced. I
think you need to think this out again!

Regards,

Bruce

23

[From Rick Marken (980209.0815)]

Bruce Abbott (980209.0950 EST) --

You have to trust that your participant is really trying, but
this is the same for any application of the Test.

Oded also made this claim and it is simply not true. There is
no trust involved in the Test. If a person is controlling a
variable then the Test will reveal that the variable is under
control. The only way to "fake" the Test (conceal the fact that
you are controlling for some variable) is to give up control
of that variable -- at least as long as the Test is in progress --
but in that case the Test will correctly reveal that the variable
is _not_ (currently) under control.

Me:

This is why I say that the results of conventional psychological
research can be a good _strting point_ for doing PCT aware
research; but, as they sit, the results of conventional research
tell us little or nothing about the nature of the organisms under
study.

Bruce Abbott (980209.0950 EST) --

This conclusion stands in stark opposition to the facts I've
introduced. I think you need to think this out again!

We are obviously getting nowhere. You apparently believe (despite
our demonstrations of the existence of a "behavioral illusion"
when you use conventional methods to study control systems) that
the results of conventional psychological research tell us something
about the nature of the organisms under study; I believe that the
results of such research tell us nothing about the organisms under
study -- in particular these results tell us nothing about what
the organisms are controlling (their CVs), though it may suggest
possible CVs, the existence of which would then have to be proved
(or disproved) using the Test.

Let me ask you this: Do you think that the behavioral illusion
creates a problem for _any_ conventional psychological research?
That is, should any of the findings of conventional psychological
research (and you mentioned quite a few of these findings around
Christmas time, I think) be considered suspect and reevaluated
given Powers' discovery of the behavioral illusion?

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

24

[From Bruce Abbott (980209.1145 EST)]

Rick Marken (980209.0815) --

Bruce Abbott (980209.0950 EST)

You have to trust that your participant is really trying, but
this is the same for any application of the Test.

Oded also made this claim and it is simply not true. There is
no trust involved in the Test. If a person is controlling a
variable then the Test will reveal that the variable is under
control. The only way to "fake" the Test (conceal the fact that
you are controlling for some variable) is to give up control
of that variable -- at least as long as the Test is in progress --
but in that case the Test will correctly reveal that the variable
is _not_ (currently) under control.

I must say that you are having a terrible time understanding me. The claim
I am making is that when a person fails to control a variable during the
Test, this is no proof that the person cannot control that variable, using
the means available during the Test. Before you can conclude on the basis
of the Test that the person cannot control the variable, you must be assured
that the person was actually trying to do so, and failed. All you can say
based on the Test alone is that the person was not controlling the variable,
not that she can't.

We are obviously getting nowhere. You apparently believe (despite
our demonstrations of the existence of a "behavioral illusion"
when you use conventional methods to study control systems) that
the results of conventional psychological research tell us something
about the nature of the organisms under study; I believe that the
results of such research tell us nothing about the organisms under
study -- in particular these results tell us nothing about what
the organisms are controlling (their CVs), though it may suggest
possible CVs, the existence of which would then have to be proved
(or disproved) using the Test.

We are getting nowhere because you refuse to reason through this with me.
Where is your reply to my thought experiment's results? Where do you show
that my reasoning -- or the data which I assert the experiment would produce
-- are incorrect? Rick Marken, you are stonewalling me, and you are doing
it because you have no other choice but to do so in order to avoid
capitulating to my position. You _can_ learn quite a bit of value about the
organisms under study through the use of psychophysical methods, and to
conclude otherwise, as I have demonstrated, is wrong, wrong, wrong. Fess up.

Regards,

Bruce

25

[From Oded Maler (970209)]

Rick Marken (980209.0815)]

Bruce Abbott (980209.0950 EST) --

> You have to trust that your participant is really trying, but
> this is the same for any application of the Test.

Oded also made this claim and it is simply not true. There is
no trust involved in the Test. If a person is controlling a
variable then the Test will reveal that the variable is under
control. The only way to "fake" the Test (conceal the fact that
you are controlling for some variable) is to give up control
of that variable -- at least as long as the Test is in progress --
but in that case the Test will correctly reveal that the variable
is _not_ (currently) under control.

Me:

> This is why I say that the results of conventional psychological
> research can be a good _strting point_ for doing PCT aware
> research; but, as they sit, the results of conventional research
> tell us little or nothing about the nature of the organisms under

This is a very strong (and apparently naive) claim which I think fails
even for the uni-deimensional case. Or else it should be prefixed by
some constraining conditions. Are you claiming that *whatever* is
the perceptual variable (i.e. a function of the cursor position)
a system (person or program) is controlling for, your test
will be able to detect it? Do you really mean it? Or do you restrict
yourself to certain form of functions?

--Oded

26

[From Rick Marken (980209.1030)]

Bruce Abbott (980209.1145 EST) --

The claim I am making is that when a person fails to control a
variable during the Test, this is no proof that the person cannot
control that variable, using the means available during the Test.

I agree. All the Test shows is that a person is or is not
controlling a variable. It doesn't tell you whether or not
the person _could_ control the variable if the person is not
currently conrolling it.

Rick Marken, you are stonewalling me, and you are doing it
because you have no other choice but to do so in order to avoid
capitulating to my position. You _can_ learn quite a bit of value
about the organisms under study through the use of psychophysical
methods, and to conclude otherwise, as I have demonstrated, is
wrong, wrong, wrong. Fess up.

I meant my last post to be a capitulation. I fess up. I capitulate
to your position. You win. You are right.

Now, could you answer the question I asked you in the last
post:

Do you think that the behavioral illusion creates a problem
for _any_ conventional psychological research? That is, should
any of the findings of conventional psychological research (and
you mentioned quite a few of these findings around Christmas
time, I think) be considered suspect and reevaluated given
Powers' discovery of the behavioral illusion?

I'm just curious to see if you think these little demos of
ours have any relevance to psychology.

Oded Maler (970209) ==

Are you claiming that *whatever* is the perceptual variable
(i.e. a function of the cursor position) a system (person or
program) is controlling for, your test will be able to detect it?

Sure. I'm not saying that I (personally) would be clever
enough to be able to use the Test to detect any arbitrary
controlled variable. But I think it's possible in principle.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

27

[From Oded Maler (970209)-2]

Rick Marken (980209.1030)]

Oded Maler (970209) ==

> Are you claiming that *whatever* is the perceptual variable
> (i.e. a function of the cursor position) a system (person or
> program) is controlling for, your test will be able to detect it?

Sure. I'm not saying that I (personally) would be clever
enough to be able to use the Test to detect any arbitrary
controlled variable. But I think it's possible in principle.

How about someone controlling for the fact that the sequence of the
integer parts of the distances from the cursor (as sampled at integer
time points) form the decimal expansion of Pi? I think *even* cleverer
people than yourself will find it hard to discover. One can define a
lot of crazy perceptual functions, even for one basic "objective"
perceptual variable such as distance from a cursor.

--Oded

28

[From Bruce Abbott (980209.1505 EST)]

Rick Marken (980209.1030)

Bruce Abbott (980209.1145 EST)

Rick Marken, you are stonewalling me, and you are doing it
because you have no other choice but to do so in order to avoid
capitulating to my position. You _can_ learn quite a bit of value
about the organisms under study through the use of psychophysical
methods, and to conclude otherwise, as I have demonstrated, is
wrong, wrong, wrong. Fess up.

I meant my last post to be a capitulation. I fess up. I capitulate
to your position. You win. You are right.

I'd feel better about this if you had said exactly what it is that you think
I am right _about_. Right about what? What are you agreeing with now, that
you were not agreeing with before? Are you agreeing that one _can_ learn
something about the organism's perceptual input function based on
traditional psychophysical methods? Or what?

Now, could you answer the question I asked you in the last
post:

Do you think that the behavioral illusion creates a problem
for _any_ conventional psychological research? That is, should
any of the findings of conventional psychological research (and
you mentioned quite a few of these findings around Christmas
time, I think) be considered suspect and reevaluated given
Powers' discovery of the behavioral illusion?

The answer is yes. The behavioral illusion is the illusion that certain
relationships which appear to describe the organism actually describe only
the environment. These relationships emerge when the organism is
controlling some variable against some disturbance s via output o. Many
conventional research finding were obtained by manipulating some independent
variable (IV) and observing how some aspect of the organism's behavior (the
dependent variable or DV) varies as a function of this manipulation. _If_
the independent variable were playing the role of a disturbance to some
controlled variable, and _if_ the behavior being measured consisted of the
output being used by the organism to counter this disturbance, and _if_ it
were asserted that the resulting relationship between the IV and DV was in
fact the organism function relating input to output, _then_ one would have
to consider the finding suspect.
However, I think that very many of the findings of conventional research
fail to meet these criteria. I don't see how Ebbinghaus'es serial position
curve (to take one example) or behavioral contrast (to take another) would
fit into this paradigm, and I can think of plenty more examples that provide
the same difficulty for a "behavioral illusion" interpretation.

Regards,

Bruce

29

[From Rick Marken (980209.1430)]

Bruce Abbott (980209.1505 EST)--

Are you agreeing that one _can_ learn something about the
organism's perceptual input function based on traditional
psychophysical methods?

No. I was agreeing so that I could hear your answer to my question
about the behavioral illusion. Your answer helps me understand
why it's really useless for us to continue our discussion of the
merits of conventional research (psychophysical or otherwise);
you simply will not allow yourself to see that what is done in any
conventional experiment is exactly equivalent to what is being done
in the behavioral illusion situation, such as the one at:

http://home.earthlink.net/~rmarken/ControlDemo/Illusion.html

You even come right out and say it:

I don't see how Ebbinghaus'es serial position curve (to take one
example) or behavioral contrast (to take another) would fit into
this [behavioral illusion] paradigm...

But it is very easy for me to see how these experiments fit
into this paradigm. For exmaple, in the Ebbinghaus situation
the location of the nonsense syllable in the list is the IV;
the number of times a syllable is recalled from each position
is the DV. The "inverted U" serial position curve is the
typically observed relationship between IV and DV in this
experiment. But you would surely see quite different serial
position curves (IV-DV relationships) if the subject controlled
a different variable.

The variable that is probably controlled by most subjects in the
Ebbinghaus task is "repeat to yourself the syllables in order
of presentation"; this gives an advantage, in recall, to the
first and last syllables (the last because these syallables
can be quickly written down or said at recall time).If, instead,
the subject controls for "be sure to repeat to yourself the
syllables in the middle of the list, even if you forget some of
the first syllables", the shape of the serial position curve will
be quite different than usual (I have done experiments where
subjects seem to have used this strategy occasionally; remember,
the "classic" inverted U serial position curve results from
averaging over trials).

Anyway, the message of the behavioral illusion to conventional
psychology is "you forget about the controlled variable". You
seem to think this message doesn't apply to most conventional
research because (as you say) you believe that most of the
findings of conventional research fail to meet the criteria
for the behavioral illusion. Fine. Let's just leave it at that.
Just go off and do your research anyy way you like; maybe it will
turn out to be excellent PCT research; Maybe not. If what you
publish isn't good PCT research, then we'll let you (and your
readers) know;-)

Oded Maler (970209)-2

How about someone controlling for the fact that the sequence of
the integer parts of the distances from the cursor (as sampled at
integer time points) form the decimal expansion of Pi?

This would be tough, sure. But certainly not impossible. We had
people cotnrolling for very complex perceptions when we did a
demonstration of the "coin game" at a CSG meeting. Clever people
like Ed Ford were quite good at detecting these controlled
variables.

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

30

[From Bruce Abbott (980209.1945 EST)]

Rick Marken (980209.1430) --

Bruce Abbott (980209.1505 EST)

Are you agreeing that one _can_ learn something about the
organism's perceptual input function based on traditional
psychophysical methods?

No. I was agreeing so that I could hear your answer to my question
about the behavioral illusion.

Or to put it more starkly, you were lying. Why am I not surprised?

Your answer helps me understand
why it's really useless for us to continue our discussion of the
merits of conventional research (psychophysical or otherwise);
you simply will not allow yourself to see that what is done in any
conventional experiment is exactly equivalent to what is being done
in the behavioral illusion situation, such as the one at:

Or in other words, you are unable to mount a defense against my line of
reasoning and have opted instead for a diversionary tactic. Ah, my dear
Sarsi, the lengths to which you will go to avoid confronting the awful truth
that that which you have proclaimed so loudly and for so long is absolutely
and demonstrably wrong. As our sharp-witted readers will doubtless recall,
we were debating whether conventional psychophysical research can tell us
anything about the organism. Your position is that it cannot, mine that it
can. I offered a thought experiment whose result demonstrates that such
experiments do indeed tell us something about the organism, behavioral
illusion or no behavioral illusion. I am still waiting for a cogent
refutation of my argument, and I expect I shall have to continue waiting.
If you had such a reply, you would by now have offered it. Rather than
providing such an argument, you have attempted instead to shift attention to
another issue entirely and put me on the defensive. CSGnet readers are far
too intelligent to fall for this pitiful attempt to distract them. They
(and I) want to hear you defend your assertion that "psychophysical research
tells us little or nothing about the organism" against the argument I have
offered.

I am still waiting (oh so tactfully, my dear Sarsi!) for your defense. And
so is everyone else.

Regards,

Bruce

31

[From Rick Marken (980209.2040)]

Bruce Abbott (980209.1945 EST)

you are unable to mount a defense against my line of reasoning

I don't feel the need to defend myself against your line of reasoning
becuase it doesn't feel like a threat to me. You might want to think
about defending _yourself_ against your line of reasoning since it
is preventing you from seeing what Bill Powers pointed out to Baum
and Reese over twenty years ago (Science, 1973); that conventional
psychology "has been in the grip of a powerful illusion [the
"behavioral illusion"] since its conceptual bases were laid".

Ah, my dear Sarsi

Who's this Sarsi person?

I am still waiting for a cogent refutation of my argument

I'm sorry. I don't remmember what you argument was. Feel free
to try it again.

They (and I) want to hear you defend your assertion that
"psychophysical research tells us little or nothing about
the organism" against the argument I have offered.

Again, I don't remember the argument you offered. My argument
is simply that in all experiemnts where the subject has no
influence over the value of the "independent variable" (making
the value of that variable _independent_ of the subject's
actions, ergo, an independent variable, IV) the IV is, at best,
a disturbance to some variable being controlled by the subject;
the dependent variable (DV) is, at best, an output that also
influences the state of the controlled variable (CV). If this
is the case, then the relationship between IV and DV is
proportional to the inverse of the feedback function (see
Powers, LCS, pp 156-158). If IV and DV have effects on different
controlled variables then any relationship between them tells us
nothing coherent about _anything_ -- the organism, the environment,
nothing.

But come on, Bruce. Who cares? If you don't want to believe
that the results of conventional research gives us the illusion
that we are learning about the nature fo the organism then don't
believe it. I can't make you believe it. Just do your work. It
may be that you will end up doing terrific research that tells
us all about some of the interesting variables that organisms
control. I'll just wait and see how you do.

Best

Rick

···

--

Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken/

32

[From Bruce Abbott (980210.1310 EST)]

Rick Marken (980210.0815) --

Now that Bill Powers (980209.2010 MST) is back it might be
a tad more difficult for Martin Taylor, Bruce Abbott, Dan
Miller, et al to blame what they don't like about PCT on me.
I realize that many of the implications of PCT, such as the
behavioral illusion, are very unpleasant for those who have
a stake in conventional social science research and pure
"observation". But don't blame me for this; I'm just the
messenger. If you don't like the message of PCT, please
blame PCT, not me.

What a bizarre statement! First, there is the assertion that Martin, Dan,
and I are blaming Rick Marken for "what they don't like about PCT."
Speaking for myself, the only things I "blame" Rick Marken for are making
unsupported allegations, misrepresenting facts, drawing incorrect
conclusions, flat-out lying, and the like. Second, there is the assertion
that we find "many of the implications of PCT" "very unpleasant" (and the
behavioral illusion is singled out as ans example of such a supposedly
unpleasant fact). Again speaking for myself, what I find "very unpleasant"
are certain assertions as to what PCT supposedly implies, which on closer
examination turn out to be unsupportable nonsense. [Note carefully: I am
NOT asserting that the behavioral illusion itself is unsupportable nonsense;
I fully understand its origins and its implications, and accept those
implications.] It is a _fact_ that psychophysical experiments yeild
information about the organism, even though Marken has been asserting
otherwise. He has been unable to defend his assertion against cogent
argument, and so he now attempts to pursuade readers that his position is in
fact correct and that my unwillingness to accept it is due to my inability
to tolerate the supposed unpleasantness of its implications for
"conventional research."

And so, Rick, I do not blame you for believing that "psychophysical
experiments tell us little or nothing about the organism," but I do hold you
personally responsible for ignoring facts presented in evidence against that
position, failing to address the argument, pretending to accept my
conclusion when in fact you did not, and launching a campaign of propaganda
to deflect attention from the fact that you are unable to come up with an
adequate defense of your claim. Moreover, I do not find ad hominum arguments
particularly convincing, and I would hope that the same is true of other
CSGnet subscribers. Your claim is NOT an implication of PCT, however much
you want others to believe so, and that is a good thing for PCT, because the
claim is wrong, wrong, wrong.

So there.

By the way, perhaps Bruce Gregory can tell you who Sarsi was. Bruce?

Regards,

Bruce

33

[From Rick Marken (980210.1100)]

Bruce Abbott (980210.1310 EST)--

the only things I "blame" Rick Marken for are making unsupported
allegations, misrepresenting facts, drawing incorrect conclusions,
flat-out lying, and the like.

OK. That's what you call what I'm doing. I call it presenting PCT.
(You're not related to Kenneth Starr, are you? Are you planning to
send me a subpoena? :wink:

what I find "very unpleasant" are certain assertions as to what
PCT supposedly implies, which on closer examination turn out to
be unsupportable nonsense.

That's what you find on closer examination. But those of us who
don't find the implications of PCT very unpleasant find, upon
closer examination, that these implications turn out to supportable
sense.

It is a _fact_ that psychophysical experiments yeild information
about the organism, even though Marken has been asserting
otherwise.

Bill Powers (980209.2010 MST), another PCT messenger, asserts
otherwise as well (read Bill's comments to Martin in particular).

Richard Kennaway (980219.1717 GMT) --

Brilliant post, Richard.

Rick, is "PCT is a useful strand of experimental psychology" on
your list of standard reasons for rejecting PCT?

I'll add it right away! And while I'm at it, I think I'll also
add "Sociologists don't make the stupid statistical and
methodological errors that psychologists make so they don't
really need PCT" and "Sociologists are ready to deal with the
really important issues; they are way past PCT". :wink:

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

34

[From Bruce Abbott (980210.1540 EST)]

Rick Marken (980210.1100) --

It is a _fact_ that psychophysical experiments yeild information
about the organism, even though Marken has been asserting
otherwise.

Bill Powers (980209.2010 MST), another PCT messenger, asserts
otherwise as well (read Bill's comments to Martin in particular).

Bill and Martin are not talking about the case I've been discussing, and
Bill's comments to Martin are not relevant to it. Also, reliance on the
Method of Authority went out with the Middle Ages. If it happens that Bill
does agree with you on this issue, then that would make both of you wrong.

For those who may have joined this discussion late, my claim is that one can
learn from psychophysical studies the intensity of a CV below which control
will be lost. This is the point along the intensity dimension below which
the perceptual input function relating physical stimulus intensity to its
perceptual correlate returns p = 0 for any i. This, obviously, is
information about the organism's input function with respect to the
stimulation in question (e.g., a particular tone). Rick's position is to
keep insisting (without evidence) that such psychophysical studies "tell us
little or nothing about the organism." Thus far the only arguments Rick has
presented against my position are (a) the ad hominum and (b) a dubious
appeal to authority. This is all very strange, because my argument is
rather simple and, I think, its conclusion obvious.

Bill Powers (980209.2010 MST) --

Martin Taylor 980201 10:40

Anyway, the upshot is that I think that (as a minimum) experiments that
put a bound on the ability of people to perceive changes in a perceptual
variable are useful to PCT.

Yes, that would be useful -- provided you can show what perception is
influenced by changes in a physical variable. My point is that
psychophysical experiments do not do this. You still have to ask the
subject what is being perceived, and the subject can only try to tell you
in words. At least, using the Test, you don't have to use words.

Incorrect. You don't have to use words in psychophysical experiments,
either. A simple button press or release will do, and there are other
nonverbal methods available as well. In addition, as you previously noted,
one psychophysical method (the method of adjustment) essentially _is_ the Test.

But to satisfy my curiosity, in your view, Bill, how would the mere use of
words invalidate the results of psychophysical studies? I can say "yes" and
"no" as easily as I can press or release a button, and thereby report the
same states of a particular perception. An in a tracking study, I can just
as easily report whether the cursor is above or below target as I can push a
key to make the cursor, now above the target, move down (thereby signifying
that I perceive the cursor to be above the target) or release the key to
make the cursor, now below the target, move up (thereby signifying that I
perceive the cursor to be below the target).

Regards,

Bruce

35

[From Tim Carey (980211.0705)]

[From Bruce Abbott (980210.1540 EST)]

For those who may have joined this discussion late, my claim is that one

can

learn from psychophysical studies the intensity of a CV below which

control

will be lost.

But Bruce this is where I get confused right here. In the above statement
aren't you making an assumption that you know what the CV already is? At
the very least don't you have to assume that the participant is controlling
for complying with your instructions?

How can you study the intensity of a CV unless you first test to find out
what the CV actually _is_?

Regards,

Tim

This is the point along the intensity dimension below which

the perceptual input function relating physical stimulus intensity to its
perceptual correlate returns p = 0 for any i. This, obviously, is
information about the organism's input function with respect to the
stimulation in question (e.g., a particular tone). Rick's position is to
keep insisting (without evidence) that such psychophysical studies "tell

us

little or nothing about the organism." Thus far the only arguments Rick

has

presented against my position are (a) the ad hominum and (b) a dubious
appeal to authority. This is all very strange, because my argument is
rather simple and, I think, its conclusion obvious.

>Bill Powers (980209.2010 MST) --

>>Martin Taylor 980201 10:40

>>Anyway, the upshot is that I think that (as a minimum) experiments that
>>put a bound on the ability of people to perceive changes in a

perceptual

>>variable are useful to PCT.

>Yes, that would be useful -- provided you can show what perception is
>influenced by changes in a physical variable. My point is that
>psychophysical experiments do not do this. You still have to ask the
>subject what is being perceived, and the subject can only try to tell

you

>in words. At least, using the Test, you don't have to use words.

Incorrect. You don't have to use words in psychophysical experiments,
either. A simple button press or release will do, and there are other
nonverbal methods available as well. In addition, as you previously

noted,

one psychophysical method (the method of adjustment) essentially _is_ the

Test.

But to satisfy my curiosity, in your view, Bill, how would the mere use

of

words invalidate the results of psychophysical studies? I can say "yes"

and

"no" as easily as I can press or release a button, and thereby report the
same states of a particular perception. An in a tracking study, I can

just

as easily report whether the cursor is above or below target as I can

push a

key to make the cursor, now above the target, move down (thereby

signifying

···

that I perceive the cursor to be above the target) or release the key to
make the cursor, now below the target, move up (thereby signifying that I
perceive the cursor to be below the target).

Regards,

Bruce

36

[From Rick Marken (980210.1350)]

Bruce Abbott (980210.1540 EST) --

We seem to be hitting pretty close to home here, eh?. I haven't
seen you hyperventilate this much since the reinforcement
discussion;-)

Burce:

It is a _fact_ that psychophysical experiments yeild information
about the organism, even though Marken has been asserting
otherwise.

Marken:

Bill Powers (980209.2010 MST), another PCT messenger, asserts
otherwise as well (read Bill's comments to Martin in particular).

Bruce Abbott (980210.1540 EST) --

reliance on the Method of Authority went out with the Middle Ages.

I did not mention Bill as an appeal to Authority. I just wanted
to point out that I was not the only one "asserting otherwise"

my claim is that one can learn from psychophysical studies the
intensity of a CV below which control will be lost... Rick's
position is to keep insisting (without evidence) that such
psychophysical studies "tell us little or nothing about the
organism."

In most psychophysical studies, the stimulus whose intensity
threshold you are trying to meansure is _not_ a CV; your claim
that "one can learn from psychophysical studies the intensity
of a CV below which control will be lost" is true only in
psychophysical studies where the stimulus whose threshold is
being measured _is_ a CV.

The stimulus is (possibly) a CV in psychophysical studies using
the method of adjustment (like the one you described in what
you must think of as your "decisive reasoning" post), where the
subject does have control over the stimulus. This type of
psychophysical experiment uses a version of the Test to determine
whether the stimulus is a CV; if it is, then you _can_ use this
method to learn the intensity of the CV below which control will
be lost. So I agree with your claim that "one can learn from
psychophysical studies the intensity of a CV below which control
will be lost". You can, if the psychophysical method you use
incorporates a Test for the CV.

My argument with you results from the fact that the Test for the
CV is simply not part of the typical psychophysical experiment
(just as it is not part of the typical operant conditioning,
reaction time, two person game or any other type of behavioral
experiment). So the experimenter doesn't know what perception
he's measureing the "threshold" (or whatever) of. When I say that
"psychophysical studies tell us little or nothing about the
organism" I a referring to the _typical_ psychophysical study
where the stimulus is a disturbance (IV) and there is no Test to
determine what perception is under control.

Tim Carey (980211.0705) to Bruce Abbott --

How can you study the intensity of a CV unless you first test
to find out what the CV actually _is_?

Tim managed to say in one sentence what it took me several
paragraphs above to say. Will you pay attention, Bruce, to
wisdom when it comes from the mouths of PCT babes?

Best

Rick

···

--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken

37

[From Bruce Abbott (980210.2140 EST)]

Tim Carey (980211.0705) --

Bruce Abbott (980210.1540 EST)

For those who may have joined this discussion late, my claim is that one can
learn from psychophysical studies the intensity of a CV below which control
will be lost.

But Bruce this is where I get confused right here. In the above statement
aren't you making an assumption that you know what the CV already is? At
the very least don't you have to assume that the participant is controlling
for complying with your instructions?

Of course. As I noted quite a while ago when this same question was asked,
there are ways of checking. (Martin Taylor said this as well.) If you are
really paranoid about whether your participant is lying to you (e.g., the
participant is Rick Marken), you can always use yourself as participant. (I
assume that you don't make a habit of lying to yourself!)

How can you study the intensity of a CV unless you first test to find out
what the CV actually _is_?

Let us say that I conduct a study in which the participant is given control
over tone intensity and asked to maintain that intensity at a given value.
Are you saying that I don't know what the CV actually is? How about in a
tracking study? The participant is asked to keep the cursor over the
target, and can control the position of the cursor by using the computer's
mouse. I don't know what the CV actually is? So much for the Test, huh?

Assume that my psychophysical study reveals that the participant cannot hear
a tone below a certain intensity level. (In the study, the participant has
no control over the tone's intensity, but can only signal whether she does
or does not hear the tone.) This finding leads to the prediction that the
person's ability to control the intensity to a particular reference level
will fail when the reference level is set well below the participant's
threshold for tone intensity. Unless my participant is trying to dupe me
(and ve have vays of finding out), it would seem that the psychophysical
study has indeed discovered something important about the participant's
perceptual input function for tone intensity, important enough to determine
whether or not the participant can or cannot establish control over the CV.

Rick Marken (980210.1350) --

We seem to be hitting pretty close to home here, eh?. I haven't
seen you hyperventilate this much since the reinforcement
discussion;-)

I don't like being lied to, or about. Personal quirk, I guess.

Speaking of hyperventillating, when will you recover your breath well enough
to respond to my puzzle? In case you've already forgotten it, I'll repeat
it here:

But if the relationship between s
and o deteriorates (and those famously high correlations become low
correlations) when the tone drops below a certain intensity, it is almost
certainly the case that control is failing because the tone can no longer be
perceived (p = 0, although i is > 0).

Now, here is a puzzle for you. As f(i) appears nowhere in the equation
relating o to s, how is it that this experiment tells us something about f(i)?

If I'm lucky, Bill won't jump in and ruin it for me by answering for you.

Regards,

Bruce

38

[From Tim Carey (980211.1250)]

[From Bruce Abbott (980210.2140 EST)]

Of course. As I noted quite a while ago when this same question was

asked,

there are ways of checking. (Martin Taylor said this as well.) If you

are

really paranoid about whether your participant is lying to you (e.g., the
participant is Rick Marken), you can always use yourself as participant.

(I

assume that you don't make a habit of lying to yourself!)

I didn't think we were talking about what you could or couldn't do, I
thought we were talking about what _currently_ happens in conventional
psych research. Do most researchers of psychophysics use themselves as
subjects at your university? It doesn't happen at mine?

Let us say that I conduct a study in which the participant is given

control

over tone intensity and asked to maintain that intensity at a given

value.

Yep, where this happens it seems like you're getting pretty darn close to
the Test. Assuming of course the subject always responds when the tone gets
to threshold. Are there ways to detect a person just memorising a
particular tone and indicating whenever the intensity drops below that even
though the can still detect it (I'm assuming there are by the way, I'm just
curious)

Again, though, the issue for me isn't the validity of this _particular_
type of experiment where the subject has some control over the tone
intensity, I'm more interested in how regularly this particular technique
is used?

If it's not used, can you draw _any_ conclusions (that are relevant to PCT)
about the organism?

Assume that my psychophysical study reveals that the participant cannot

hear

a tone below a certain intensity level. (In the study, the participant

has

no control over the tone's intensity, but can only signal whether she

does

or does not hear the tone.)

OK, this is what I was talking about above but probably didn't express very
well. Are there measures in place to detect that the level at which they
_report_ they can't hear the tone is indeed that level? What is to stop
someone remembering a particular level and just indicating each time the
intensity falls below that level?

Cheers,

Tim

39

[From Rick Marken (980211.2310)]

Bruce Abbott (980210.2140 EST)

Speaking of hyperventillating, when will you recover your
breath well enough to respond to my puzzle? In case you've
already forgotten it, I'll repeat it here:

But if the relationship between s and o deteriorates (and those
fammously high correlations become low correlations) when the
tone drops below a certain intensity, it is almost certainly
the case that control is failing because the tone can no longer
be perceived (p = 0, although i is > 0).

Now, here is a puzzle for you. As f(i) appears nowhere in the
equation relating o to s, how is it that this experiment tells
us something about f(i)?

I don't understand why you think this is a puzzle. In this
experiment you have apparently determined (using the Test) that
i is a controlled variable. Since what is actually controlled
(according to PCT) is p, which is a function of i [f(i)], we
learn about f(i) by building a model containing an estimated
perceptual function [f'(i)] that controls like the subject.

We don't learn about f(i) by looking at the relationship between
o and s; we learn about f(i) by modeling the subject's ability
to control i;. If the model controls i (and loses control of i)
as the subject does, then we can assume that we have a pretty
good estimate of f(i).

Your puzzle pertains to a type of psychophysical experiment
(one that involves a Test for the controlled variable) that
is not seen very often (to say the least) in psychophysics.
But you are presenting this puzzle as though this is the way
psychophysicists go about their business all the time. I think
you would be hard pressed to find many studies like the one you
describe in the pages of _Perception and Psychophysics_ or
_JEP:HPP_. Nevertheless, to the extent that studies like this
have been done, we can model the results to estimate f(i).

In most psychophysical experiments, stimuli (like tone bursts)
are presented and the subject responds to indicate something
about the stimulus (is it there?, is it louder than another?, etc).
In such an experiment, we don't know what perception is being
controlled. We don't know the perception that is influenced by
the stimuli (disturbances); this was Bill's point to Martin.
All we know is that there is some relationship between our
stimuli and the subject's responses. This relationship, I argue,
tells you nothing about what the subject is perceiving (if the
subject is a control system). You have to Test to determine what
perception is being protected from variations in what is typically
though of as the stimulus.

Best

Rick

···

--

Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken/

40

[From Bill Powers (980211.1528 MST)]

Bruce Abbott (980210.1540 EST)--

But to satisfy my curiosity, in your view, Bill, how would the mere use of
words invalidate the results of psychophysical studies?

No.

I can say "yes" and
"no" as easily as I can press or release a button, and thereby report the
same states of a particular perception.

Hard to establish _what_ perception, since you can't observe anyone else's
perceptual signals.

An in a tracking study, I can just
as easily report whether the cursor is above or below target as I can push a
key to make the cursor, now above the target, move down (thereby signifying
that I perceive the cursor to be above the target) or release the key to
make the cursor, now below the target, move up (thereby signifying that I
perceive the cursor to be below the target).

All true. However, all irrelevant to the point that we can't observe anyone
else's perceptual signals. We can see the cursor going up and down (since
that is our own perception), but we can't see the perceptions of the other
person. Your guesses about what the other person's movements "signify" are
only that: guesses.

Best,

Bill P.