[From Rick Marken (2010.09.13.1410)]
By popular request (well, one person asked me off line if I would do
this) I am replying to Gavin's post on honesty.
Gavin Ritz (2010.09.12.11.29NZT)--
GR:I would like to hear what other list members have to say about Honesty
as a �Controlled Variable� an important one in human affairs,
"Honesty" is a word that, like all words, is a perception that points
to other perceptions. It's the perceptions pointed to by the word
"honesty" that are controlled. The best way to show what honesty is,
then, is to "look at" examples of what is being "pointed at" when
people say "honest" (or "dishonest" or "moderately honest", etc). It's
the perception that's important, not the word, or the definition
thereof in terms of other words.
The fact is that people can perceive, in their own behavior as well as
that of others, different degrees of honesty. When they see a drama
like Hamlet they can see that Claudius is not being perfectly honest
with Hamlet about having murdered his Dad. And they can see that
Hamlet is not being perfectly honest with Claudius about why he wants
to put on a play. What we perceive as "honesty" is a very complex
perception of various consequences of our actions, such as what we say
or don't say to other people.
Because we can perceive what we call "honesty" we can control for it.
That is, we can perceive ourselves producing a particular degree of
honesty and we can protect the perception of honesty that we intend to
produce from disturbances. Hamlet controlled for being less than
perfectly honest about why he was putting on a play; he could have
controlled for being perfectly honest and told his Uncle that he was
trying to expose him as a murderer. But he controlled for being
dishonest about it instead, for higher level reasons (just as we
sometimes control for being less than perfectly honest when our wives
or lovers ask "How do I look in this dress?"). Hamlet also protected
that level of honesty from disturbances, such as his Uncle's attempts
to figure out what Hamlet was up to by asking him questions.
So that's what control of honesty is; it's control of a very complex
perception that we point to with the word "honesty". In order to be
able to control "honesty" we have to be able to perceive it. In PCT
that means that neural networks in the brain (the perceptual
functions) must convert the sensory results of the physical events
going on in the environment outside the system into a neural signal
that represents, by firing rate, a degree of honesty of that
spatio-temporal sensory input. Of course, we currently don't how such
a perceptual function might derive a measure honesty from sensory
variables. But in order to control "honesty" something like this
computation -- the derivation of this perception of honesty -- must be
done (at least according to PCT).
According to PCT the perceptual functions that compute perceptions
like "honesty" are at a high level in the hierarchy of perceptual
control systems, taking inputs from lower level perceptual functions,
like those that compute perceptions of events, relationships and all
the other perceptual components of the perception of honesty. This
gets to why I asked Gavin how his proposal for a new and improved PCT
model controls for honesty. His improved model seems to have
eliminated the systems that perceive (and can thus control) honesty,
as well as the lower order systems that perceive (and control) the
most important components of honesty; events, relationship, programs,
etc.
As far as the importance of honesty of a controlled variable, I think
it's importance depends on the goals of other control systems that use
the honesty control system as a means of achieving their goals. The
level of honesty one controls for seems to depend on goals such as the
goal of not hurting another person (as when answering your wife's
question about how she looks in a dress; it's a problem I don't have,
by the way, because Linda looks great in anything) or more nefarious
goals, like Claudius's goal of not getting caught for murder.
Best
Rick
···
--
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com