# Re: The (Clinical) Point

Lois Shawver (rathbone@crl.com)
Sat, 9 Mar 1996 12:21:26 -0800 (PST)

Bill,

Thanks for explaining things in such adequate detail. It helps me
picture where you're coming from a bit better, and it, as always it
seems, raises more questions.

From your account, I see that the technique of the rep grid can be
applied widely, using either individual events or generalized type of
events as variables, and using either individual subjects to give us
infomation on a wide variety of events or many subjects to give us
information on perhaps a fewer number of events.

I am confused, however, by your bringing in the notion of causality.
How do you make the conceptual leap from prediction from one correlated
variable to another to causation? I understand and agree with your point
that for the technique of regression does not imply that we treat any
particular variables as dependent or independent. Our decision to treat
some variables as dependent reflects, most often, our implicit
understanding of the situation prior to our applying the math. Thus, it
is natural that you would want to treat "cutting behavior" as the
dependent variable and "constructs" as independent, because this is a
natural way to think about things. The model beneath this is that
certain thoughts cause us to do certain things. Right?

But couldn't we imagine that cutting herself the patient affects the
patient's feeling "empty" or "anger with a boyfriend talking after
school"? That the variables work both ways? That the simple single
direction of "causation" is really something in our constructs that we
bring to this observation?

You say:
With CR we should be able to say that
across a number of events (including
arguing with boy friend, seeing parents fight,
etc.) the client's feeling empty inside is a
component of later cutting. We might also
conclude that although her getting angry
does correlate with both feeling empty and
cutting, anger does not make up one of the
components of cutting. Thus the anger is a
red herring, in a therapeutic sense. The
emptiness would be the thing to get at.

When you say "the client's feeling empty inside is a component of later
cutting" might this not be part of YOUR construct? This seems to be more
likely if you are thinking of general events, not particular ones. So
you find that someone who "cuts herself" also "feels empty", which
produces which? I can imagine that someone who "cuts herself" feels cut
off from the people who find her perplexing and frightening, and thus
finds herself in shallow conversation with people who do not "trust her"
to be reasonable. Then "cutting herself" would be the IV and "emptiness"
would be the dependent variable. No? I can't see a way around this.
Can you? You might, of course, argue that it seems a reasonable way to
approach the problem, or study the case from both perspectives.

Also, if you do decide that the empiness is the cause and the cutting is
the effect, then how do you reason that explaining this to the patient
reverses this effect? Isn't the explanation itself taking place in a
context in which it is a presumed cause? Wouldn't you want to do a study
to see if explaining "the cause" produced a cessation of the problem?

Here I am reminded by the Barnum studies a few years ago. Do you
remember them? Subjects were given phoney diagnostic statements about
themselves that were ostensibly based on psychological tests. Most
subjects simply accepted the validity of these findings. So, even if
you're "wrong" about the causal quality of "emptiness" it seems it is
possible that it could have a positive effect. And, also, if you're
right about about the causal quality of variables like empiness it could
have a negative effect, or no effect at all. All stuff to be looked at,
no? Not just presumed?

..Lois Shawver

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%