Shannon's message

[From Bruce Abbott (951020.1055 EST)]

Shannon Williams (951019) --

Bruce Abbott (951019.1540 EST)

Why would you look for what is being controlled unless you were puzzled
about the purposes of the termite's mound?

Let me re-phrase this question:

  "Why would you look for what is being controlled unless you were
   puzzled about overt signs of control?"

Ah, a more general restatement.

My answer is: All living things control. I am interested to know what
it is they control. And I am just as interested in organisms that do not
appear to control as I am in those which obviously control.

No, no, no, this misses the point. I was NOT asking why you would want to
study what living things control, IN GENERAL. I was asking how you would
know what controlled perceptions to look for in the termite unless you had
some information about how termites live their lives.

Wouldn't a detailed description of the life of the termite (what it does,
where, and under what circumstances) guide PCT research by revealing
patterns of behavior that occur reliably under specific circumstances and
which appear to the external observer as though they might serve specific
functions in the life of the species?

If you gave a detailed description of what I do, where, and under what
circumstances, you still would not know my motives. To know my motives
you would have to systematically guess my motive, and then test to see
if you could be right. Without this "guess/test/test/guess/test/test..."
procedure, you mostly have "guess/guess/guess...".

I contend that I would have a very firm foundation on which to build my (now
highly educated) guesses. Without that I would only be reasoning from
analogy from my own experience, which may or may not be relevant. In the
case of the termite, it would be even less relevant. So why flounder in the
dark?

Don't forget the context in which I have framed my argument. I am arguing
that there are many behavioral phenomena discovered in the course of
ordinary scientific investigation that provide excellent starting points for
a PCT-style investigation of controlled variables; many offer strong hints
about the underlying organization of the systems involved in the behavior,
and some have even identified a few of the controlled variables. I am
arguing that findings from neuroscience likewise can guide the search for
the physical structures in the brain thought which this control is carried
out. Your response indicates that you think (as apparently Bill P. and Rick
M. do) that I am advocating a different position, that pure observation can
replace experimental testing.

In fact, as I keep repeating, this is not my position at all. What you
advocate (guess/test/test) is in fact quite ordinary procedure and one that
I heartily endorse. Taken broadly, it is nothing less than the experimental
method.

So, when Bill Powers (951020.0545) responds to this statement with
"yes/yes/yes" it would appear that he agrees that you have made a telling
point against my argument; what you say is correct (hence the
"yes/yes/yes"), but in fact you have not addressed my argument at all.

Until we determine that the child crys when its mother leaves the room, why
would we even ask whether crying is being used as a means to control its
distance from its mother?

We would ask because before we can test a guess, we have to have a guess.

Even "smiling" is a means to control. The question is "*what* is being
controlled?" So we make a guess, then test it. Then test it again. Then
make more guesses. Then test them. Etc.

But Shannon, we have about 30 years of quite excellent research on this
control system (or perhaps more accurately, the collection of systems of
which this is a part). Bill and Rick are claiming that it must all be
thrown away in order to start anew from the PCT framework. My position is
that it need not and should not be. I am sticking to that position.

In so doing, I am not advocating simply adopting currently accepted
interpretations of the phenomenon. Go ahead, make a guess and
test/test/test. This is what science is about. But let's start from an
informed position, not from abject ignorance of the phenomenon of interest,
as if no prior research on the subject exists. Starting from the known is
also part and parcel of the scientific method.

Regards,

Bruce

[From Shannon Williams (951020)]

Bruce Abbott (951020.1055 EST) --

I was asking how you would
know what controlled perceptions to look for in the termite unless you had
some information about how termites live their lives.

In other words you are asking:

  How might I guess that termites control for temperature unless I know
  that termites build mounds?

I do not believe that the information about how a termite lives, is what
leads me to guess that they control for temperature. I think I need
information like: what do termites do when you heat their mounds, or air
condition their mounds, or something. In other words, I need information
on how to make termites *change* how they live.

If you gave a detailed description of what I do, where, and under what
circumstances, you still would not know my motives.

I contend that I would have a very firm foundation on which to build my
(now highly educated) guesses.

With what facts do you build this foundation? Facts interpreted through
your own comprehension of reality? What if I seem to act like you do, but
I perceive the world differently than you?

Without that I would only be reasoning from
analogy from my own experience, which may or may not be relevant.

You are saying:
   Without "a description of what [Shannon] does" I would only be
   reasoning from analogy from my own experience.

If I give you "a description of what [Shannon] does", then what would you
be reasoning from?

I am arguing that there are many behavioral phenomena discovered in
the course of ordinary scientific investigation that provide excellent
starting points for a PCT-style investigation of controlled variables;

When you say "behavioral phenomena" above, to me you are saying
"not understood behavior". And you are saying that PCT should be applied
to these phenomena so that we can understand them. But what about all of
the behavior that is "understood"? Our understanding of behavior is
grounded in how we explain these things that we understand. If we do not
recognize the single law of PC behind simple behavior, how do we recognize
it behind complex behavior?

Should a PCT-style comprehension start with the behaviors that we realise
that we do not understand, or must we first comprehend all behaviors in
terms of PC?

Shannon

[From Bruce Abbott (951021.1120 EST)]

Shannon Williams (951020) --

Shannon, you ask some excellent questions. Let's see if I can address them
to your satisfaction.

Bruce Abbott (951020.1055 EST)

I was asking how you would
know what controlled perceptions to look for in the termite unless you had
some information about how termites live their lives.

In other words you are asking:

How might I guess that termites control for temperature unless I know
that termites build mounds?

No, I was asking the question I asked. The observation that termites build
mounds may be what gets you thinking about termites and how they live their
lives. So you begin an investigation. You learn how they go about
constructing their mounds, what the division of labor is among the termites
in the colony, and what activities they pursue both within and outside the
mound. You make certain measurements, such as the temperature of the mound,
as compared to the temperature of the air and the ground around the nest,
the inside and outside relative humidity, and so forth. You become
intimately acquainted with the lives of termites and from these observations
you identify what you believe to be the functions of the behaviors you
observe: defense, foraging, reproduction, nursery management, ventillation
of the nest, housekeeping. You begin to do systematic studies to pin down
the sensory/phermonal inputs involved in specific behaviors. Not being
schooled in PCT, you begin to develop an explanatory structure in which
these sensory inputs, when combined with the proper hormonal or other
preexisting conditions, serve to "trigger" certain of these patterns (e.g.,
a termite from a rival mound enters the mound and by its "wrong" smell
elicits attack and is destroyed or driven out). You note that certain other
odors are left by the termites as they forage outside the mound, and that
other termites from the mound use these odors to follow the same path. (You
conclude that odors are the relevant input from careful experiment). You
publish your results, others follow up on your work, and soon there is a
large body of research on these termites.

But none of this research was done with PCT in mind, and no attempt was made
to formulate a working control-system model that would behave as these
termites behave, which could be tested against appropriate moment-by-moment
observations of termite behavior as certain environmental variables were
disturbed. Therefore, none of your work is of any value for the PCT
theorist (so it is claimed), and in a post-PCT world it will be viewed in
the way that the work of alchemists is viewed today. No, we must start
over, with an entirely clean slate. Your methods, PCTers will tell you, are
worthless, and your conclusions suspect. How could you possibly have
learned anything of value about termite behavior if you did not start from
the realization that all behavior is control of perception?

Now Bill Powers has stated that he will happily accept the observations so
long as they are observations and not really interpretations pretending to
be observations. I am willing to go a bit further because in my judgment
many of the conclusions that have been reached are reasonable, consistent
with impressive evidence, and consistent, so far as I can tell, with the
view that all behavior is control of perception. Furthermore, I believe
that I can provide a better account for many of those observations than is
currently available, by applying PCT. I can agree with ethologists on the
basic observations to be explained (which include experimental findings
identifying which specific perceptual inputs are the effective ones within a
given system) but differ with them on their theoretical interpretation.
With this as a starting-point, I can then perform my own studies, PCT-style,
to nail down the specific control structures mediating a particular set of
organized behaviors. In this way, my studies build upon previous work,
advance our detailed understanding of the control-system structures of the
termite, and provide a significant test of PCT.

I do not believe that the information about how a termite lives, is what
leads me to guess that they control for temperature. I think I need
information like: what do termites do when you heat their mounds, or air
condition their mounds, or something. In other words, I need information
on how to make termites *change* how they live.

Yes, you need this information. You can certainly begin your investigation
ignorant of all that has been learned about mound-building termites. You
might just happen to see a termite mound and wonder whether one of its
functions is to keep the termites inside within a specific range of
temperatures. You might then launch a careful investigation guided by PCT
and using the approved methods. I'm not arguing that you couldn't, nor am I
arguing that previously developed information would _necessarily_ turn up
anything useful for your investigation. The question is whether it is
better to build on an established foundation of knowledge and previous
theoretical analysis or to start out each investigation blissfully ignorant
of what is already known, what hypotheses have been tested, and what
interpretations have already been ruled out by the evidence.

If you gave a detailed description of what I do, where, and under what
circumstances, you still would not know my motives.

I contend that I would have a very firm foundation on which to build my
(now highly educated) guesses.

With what facts do you build this foundation? Facts interpreted through
your own comprehension of reality? What if I seem to act like you do, but
I perceive the world differently than you?

With facts derived from objective observation. If I observe that you spend
much of your time writing computer programs, I know something about you that
can guide further research. I may discern patterns to your behavior,
particular ways in which you characteristically behave under certain
conditions. With enough observation, I'll bet I can infer many of your
higher-level controlled perceptions, although such inferences would remain
to be tested. These "educated" guesses would certainly be an improvement
over uninformed ones.

But this example is misleading, because it does not really apply. Science
is concerned with establishing general principles (within which specific
cases can be understood, to be sure). The fact that you in particular might
be controlling for the color blue in your office (you ask for your office to
be painted blue and would attempt to prevent its being repainted some other
color) might be of interest to someone, but I'm afraid it would contribute
little to a science of behavior. There is no established research
literature on you that can provide relevant information to guide research
into the organization of the human control hierarchy. My prior observations
of you would not be equivalent to a long history of naturalistic observation
and experimental research on a specific phenomenon of human behavior.

Without that I would only be reasoning from
analogy from my own experience, which may or may not be relevant.

You are saying:
  Without "a description of what [Shannon] does" I would only be
  reasoning from analogy from my own experience.

If I give you "a description of what [Shannon] does", then what would you
be reasoning from?

The "description of what Shannon does" would include many details about the
conditions under which these patterns of behavior tend to repeat, including
prior conditions, present ones, and the objective consequences of Shannon's
actions. If I am reasoning only by analogy, I am guessing what I would do
under similar circumstances. If I am reasoning from direct observation, I
am reasoning from a knowledge of what you actually do, under what
circumstances, and what the effects of these actions are, at at least
insofar as they can be observed from the outside. But again, this argument
uses an example that does not properly apply to my thesis. In this example,
I would agree that the most direct approach would be pick a specific set of
observations, formulate hypotheses as to the possible controlled
perceptions, and start testing.

I am arguing that there are many behavioral phenomena discovered in
the course of ordinary scientific investigation that provide excellent
starting points for a PCT-style investigation of controlled variables;

When you say "behavioral phenomena" above, to me you are saying
"not understood behavior". And you are saying that PCT should be applied
to these phenomena so that we can understand them.

I am saying that PCT research into a given area often can begin on an
already established foundation of scientific research. In many cases other,
non-PCT explanations have already been offered. These areas of research are
already established and being actively pursued. If PCT is the better
mousetrap, why not demonstrate its superior explanatory power in those
arenas? One has to start somewhere. Why not there? And why not make good
use of the information that is already available?

But what about all of
the behavior that is "understood"? Our understanding of behavior is
grounded in how we explain these things that we understand. If we do not
recognize the single law of PC behind simple behavior, how do we recognize
it behind complex behavior?

You've lost me. Are you arguing for walking before we can run? Are you
arguing that behavioral phenomena such as those I have described are too
complex for a beginning analysis? The questions seem more like assertions
to me, but I'm not quite sure what those assertions are.

Bear in mind that nothing I have said precludes one from doing basic
research into control, starting from scratch, as in the tracking studies.
My argument is not that we should necessarily plunge into these established
areas (perhaps we do need to do some basic spadework before applying the
understanding gained in such research to more complex areas). My argument
is that an investigation into the control structures of, for example,
mound-building termites, should one be undertaken, should build on
established observations if such exist. This argument has been offered in
opposition to the assertion that all or most of such previous work is
worthless to the PCT researcher and my be safely ignored.

Regards,

Bruce

[From Shannon Williams (951021)]

Bruce Abbott (951021.1120 EST) --

none of your work is of any value for the PCT theorist (so it is claimed),
and in a post-PCT world it will be viewed in the way that the work of
alchemists is viewed today.

What is the primary difference between the concept of alchemy and
chemistry? (Do chemists deal with different substances than alchemists? Or
do they mainly understand these substances differently?)

If I tell you that Billy is five years old and both parents smoke, we know:
1) the percentage chance that Billy will not smoke. But we do not know why.
We do not know what Billy's parents do to influence him not to smoke.
(Short of not smoking themselves). 2) if Billy's parents have PhD's,
we can tell the percentage chance that Billy will not go to college. But
we cannot tell why. 3) we can tell the percentage chance that he will not
wreck his car when he is 16. But we cannot tell why. We cannot tell
'why', because we do not know 'what' influences people to do the
behaviors that they do.

Current psychology has described different goups that people fit into.
The description also includes notes like: Behavior A is reinforced by
behavior B which exists because of a combination of C and D... These
are the elements of alchemy. The alchemist can describe the
characteristics of his elements, but he cannot explain why these
characteristics exist. Nor does he know his elements' potential for change.

The question is whether it is
better to build on an established foundation of knowledge and previous
theoretical analysis or to start out each investigation blissfully ignorant
of what is already known, what hypotheses have been tested, and what
interpretations have already been ruled out by the evidence.

Ok. Before there were chemists, there had to be alchemists. Chemists are
not spontaneously generated.

With enough observation, I'll bet I can infer many of your higher-level
controlled perceptions, although such inferences would remain to be
tested.

It depends upon what you are going to guess about. If you are going to
guess that I would get irate if the park cancelled a volleyball league, you
would be correct. And you would not make this guess without knowing
something about me. But does this tell you why I want to play volleyball?
Does this tell you what drives me? Does it hint to you if I could replace
volleyball with something else? What variable could change in my life so
that I am not driven to play?

If PCT is the better mousetrap, why not demonstrate its superior
explanatory power in those arenas? One has to start somewhere. Why not
there? And why not make good use of the information that is already
available?

Are you arguing that behavioral phenomena such as those I have described
are too complex for a beginning analysis?

I am arguing that if you believe in an earth-centered universe, you will
not visualize Kepler's laws. And even when the laws predict correctly,
you will not be able to apply them to your comprehensions.

If you believe that you can intuitively understand another being's
behavior just by observing it, you will not be able to visualize PC.
Your objective data could be useful to a PC experimenter, but he could not
build upon your theory or analysis.

Sincerely,
Shannon