# paper, IV-DV

**URL:** <http://discourse.iapct.org/t/paper-iv-dv/14533>\
**Category:** CSG1993\
**Created:** [April 27, 1993, 7:36pm UTC](http://discourse.iapct.org/t/paper-iv-dv/14533 "1993-04-27T19:36:17Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Marken\_Richard\_S1](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/m/977dab/32.png) [@Marken\_Richard\_S1](http://discourse.iapct.org/u/Marken_Richard_S1)\
**Post date:** [April 27, 1993, 7:36pm UTC](http://discourse.iapct.org/t/paper-iv-dv/14533/1 "1993-04-27T19:36:17Z")

</div>

[From Rick Marken (930427.0900)]

Chuck, Avery and Isaac -- thanks for asking for a copy of the  
rejected paper. I will send an e-mail version of the paper to  
each of you. The only thing missing from this version of the  
paper is the figures, which are really not essential; you have  
probably seen them in other contexts anyway. Besides, if I'm  
lucky I'll be able to get the thing published in Closed Loop.

Chuck Tucker (930426) --

> Since, everyone is interested  
> in control (in the ordinary sense) then maybe they would publish your  
> paper as a better way to do it. [Your problem will be hiding this from  
> Bill]

As I mentioned in the post, I think Greg Williams has already  
proposed this as a good way to attract the attention of behavioral  
scientists who might otherwise see PCT as little more that a  
masturbatory exercise in applied solipcism. I don't think there is  
anything wrong with this approach -- I wish some people who do  
operant research, for example, would set up some PCT experiments.  
I think we could show them how to control animal outputs far more  
precisely than they can do this now. Indeed, looking for increased  
precision of IV-DV relationships is one way to do the test for  
controlled variables. An example of this is given in the graph  
on p. 55 of "Mind Readings" (I'm going to sell another one of those  
books if it kills me). The x axis of the graph is the DV (subject's  
output) and the y axis is the IV (disturbance). The curvilinear  
relationship between IV and DV can be predicted perfectly once you  
know that the subject is controlling the area of the object on the  
screen.

By the way, I didn't mean to give the impression that there is anything  
wrong with the fact that behavioral scientists want to control. We ALL  
want to control -- it's our nature to do this (if you buy PCT). In most  
cases, the controlling done by behavioral scientists is quite benign;  
these scientists are not always trying to bend people to their will.  
In many aeras of behavioral science the desire to control shows up as  
nothing more sinister than the desire to see a relationship between  
variables. Look at cognitive psychologists for example; none of them  
would claim to have any interest in controlling behavior -- and they  
don't. But they do want to control their perception of relationships  
like those between target types and search times or between character-  
istics of a problem space and the number of trials to solution. PCT  
researchers are also trying to control their perceptions, of course.  
We are trying to control our perceptions of the correspondance between  
model and subject behavior; between our description of the controlled  
variable and what the subject is actually controlling. PCT is just  
suggesting that behavioral science researchers should try controlling  
different variables -- IF their goal is tounderstand HOW the objects  
of their study actually work. PCT provides the ability to control at  
a higher level -- it is not the ability to control people better; it  
is the ability to control our understanding of how people are able  
to control at all.

Marcos Rodrigues (930427.1155 BST)--

> I'm afraid that the impregnated S-R model  
> in the life sciences will last for quite a while.

I'm afraid you're right. There is probably enough "entertainment"  
value in the IV-DV approach to keep it going for quite some time.  
I think that all we PCTers can do is keep trying to show the  
benefits of a "subject centered" approach to understanding living  
systems. I think that robotics is one area where the practical  
benefits of PCT modelling will eventually become evident. After  
all, you can argue about theory all you want but if the theory  
actually produces a better mousetrap, who's going to argue (other  
than the mice)? I think robotics would take off like crazy if some  
of the talented people working in that area started architecting  
their beasts as hierarchical perceptual control systems.

Best

Rick