THE model for sociology

From Greg Williams (920923 - 2)


      >CLOSED LOOP GW 920919

      >My vote is for an issue on "Influence and Control" (seconding
      >Hugh's vote actually)...

Perhaps for the January 1993 CLOSED LOOP. I'm already well into the next
(October) issue, which will be on "Standards" and closely related topics. I
like your suggestions (not re-posted here) for structuring an
"Influence/Control" issue (and perhaps a follow-up "comments by others" issue,
to which I would want to see added "final remarks" by Bill and me. But I
wouldn't feel particularly comfortable with editing either issue, since I am a
principal in the argument. Might someone else (or more than one person) agree
to be guest editor(s)? If nobody will volunteer, then I might be willing to go
through with it, providing that Bill and everybody involved in the "comments"
issue have an opportunity to pass on the final version. So, let's discuss the
possibility further -- we have a bit of time to make decisions (work would
start in late November/early December).




      >I have reached the tenative judgement that there is no way for
      >one human being to influence (in Greg's terms) another human
      >being and still maintain the principles and processes outlined
      >in the HPCT or PCT model as proposed by Bill Powers.

I'd like to hear more about the basis for your judgement. Do you mean to say
that no one actually will sign his/her name in rubber-banding if presented
with certain kinds of disturbances by another? This is an instance of what I
call "purposive influencing." It is a simple instance, to be sure, because the
influencee's (relevant) controlled perception (keeping the knot over the mark)
is "obvious" to the influencer (who simply asked for the influencee's
agreement that that is the influencee's controlled perception), and because
there is no deception involved. But that is the sort of interaction which I
call "purposive influencing": the influencer makes a model of what the
influencee wants, then manipulates the influencee's environment so that the
influencee gets what he/she wants in such a manner (different from the manner
which would be employed by the influencee in the absence of the influencer)
that the influencer also gets what he/she wants. Or did you think I meant
something else by "purposive influence"? If so, sorry; I didn't.

      >Greg keeps insisting that THE TEST will demonstrate that person
      >A has influenced person B to control for X.

No. What I have been insisting is that a person who wants to purposively
influence someone else would be well advised to use some form of The Test to
make a model of what the prospective influencee wants, rather than just
guessing about what the influencee wants. The Test DOES NOT demonstrate that
person A has influenced person B to control for X. The Test FACILITATES making
a model of X, which A DOESN'T WANT TO ALTER (that would require FORCE or
influencing the outcome of reorganization by B)! A USES his/her model of B's
(and solely B's) want, X (which B is controlling for THROUGHOUT the entire
episode of purposive influence), to INFLUENCE B to successfully control for X
(assuming A is correctly modeling it) by ACTING so as that A will get what
he/she wants.

      >But that is not so,
      >THE TEST only shows (if done repeatedly and systematically with
      >careful records) what B is controlling for BUT IT CAN'T SHOW
      >THAT B TOOK X FROM A. A can only say that the X that B is
      >controlling for seems to be exactly like the X I use and asked
      >(told, demanded, etc.) B to use but since B is now using X it
      >belongs to the "control system" of B and no longer to A even
      >though A might use it also.

What is this "taking"??? I never said that A takes ANYTHING from B. A needs to
make an accurate guess ("model") of what B is controlling for if A is to
succeed in purposively influencing B. A DOESN'T have to TAKE ANYTHING from B!
A's model of B's X certainly "belongs to" A, not B. But, so what?

      >Now this seems like a very "picky"
      >point to make but it is crucial to PCT: my perceptions are mine
      >even though they may look exactly like the perceptions that you
      >say are yours.

I have no problems with this. That's why I said that A makes (his/her OWN)
MODEL of B's want X; A doesn't "TAKE" X "FROM" B. Using A's model of B's X, A
tries to arrange B's environment so B performs actions to achieve X so that A
gets what he wants. A could be wrong about his/her model being correct, in
which case A would try to make a better model -- and The Test is the key to
making models of other peoples' wants (sez PCT-science).

      >The PCT model has to insist on this autonomy or it will very
      >quickly lose its differences from all other theories (as Rick
      >indicates in his "elephant" paper). Thus, anything that looks
      >like a "collective social act" is actually two or more persons
      >controlling their own individual conduct in what appears to each
      >and every party to action as influencing the others.

I agree wholeheartedly. This consideration in no way obviates my notion of
[THE] ACTION AS INFLUENCING THE OTHERS." What is crucial to recognize is that
the purposive influencer is arranging things SO THAT he/she gets something
OTHERWISE OCCUR. The non-purposive influencer does NOT so arrange things, but
influences the other(s) in ways which he/she DOES NOT CONTROL FOR ("accidental

      >careful analysis would show that each party to the "collective
      >act" is acting on their own.

Each is WANTING on their own, but each is not ACTING on their own. Given the
wants, the actions of each depend on what influences (both purposive and
accidental) the other has. The influences might be symmetric or asymmetric.

      >If my tenative judgment (i.e., that influence [A affects or
      >determines what B controls for] is not possible between human
      >beings if they are negative feedback control systems) is turned
      >into a "warranted assertion" with testing and evidence the
      >question arises: Is the PCT (or HPCT) model useful for
      >collective social action? That is an important question to
      >answer for those of us who have a concern for human group life,
      >social life, collective behavior or even Society. Maybe a new
      >model has to be constructed !?!

I AGREE with your judgement/warranted assertion. That is what PCT science says
(barring threats/force/guided reorganization). I also think that "purposive
influence," as based on PCT science, is all we need to explain "collective
action." The key to "collective action" in general is that one person can make
a model of what another person wants, and use that model to purposively
influence the other person. No new model is needed for sociology; the PCT
model is The Model!


      >Thanks Greg for the statements from Skinner which indicate to me
      >that he (and perhaps other behaviorists) did not consider the
      >"innards". I guess I just find it strange that someone would
      >not wonder more than he did about the contribution of the
      >"innards" to behavior...

He believed he could solve the problems he wanted to solve sufficiently well
without modeling "innards." It's as simple as that: he thought he was
controlling well (in PCT terms), and so he thought he didn't need to act

      >but people I pay attention to (especially
      >myself) firmly believe that the CNS is important to say nothing
      >of the liver and bladder.

I, too, think that Skinner was deluding -- handicapping -- himself. Yet his
argument that modeling "innards" is PREMATURE was more reasonable in the 1950s
and earlier, when much less was understood about the CNS than now.

Best wishes,