[From: Bruce Nevin (Thu 920414 07:44:16)]
(Greg Williams (920513, 920513-2) ) --
I didn't just mean to tweak your beak about the seat of volition, Greg,
I'm asking quite seriously for some elaboration on your remark that we
can't adjust our own reference signals, just as no external agent can
adjust them. What follows spins off my thoughts. I still would like to
know yours.
It sure appears as though there is a driver in the driver's seat,
doesn't it? Awareness identifies with some set of purposes, and among
them is the presumption of having initiated those purposes. Is the
driver a volunteer or a draftee? You argue draftee (determinism). I
would leave that question open still.
I agree with you that Control Theory does not entail a liberal ideology.
To foresee a great amelioration of the human condition with the
awakening out of S-R slumber into the New Age of Control Theory is good
motivation, but naive. Mr. Nobel thought his invention of dynamite
would do the trick. However, I come to that view on a different basis.
Failure of control theory to guarantee autonomy is due to the propensity
of living control systems to bring free variability in their behavior
outputs under control for conformity to one another for social purposes.
People (mammals) bind their freedom to social purposes.
Let me try a quasi-postulational approach.
Manner: A particular conventionalized choice among free or available
variants.
Variants: A range of variation in behavioral outputs.
Free: Variants are free if the differences between them make no
difference (disburbance) for other ECSs.
Available: Variants are available if (a) gain on the control of a
manner exceeds gain on some other ECS to which it makes a
difference, and (b) said other ECS can control the net disturbance
by changing reference signals to other ECSs with lower gain.
Living control systems control free variation for conformity to the
manner of co-members of a social group. By doing so, they assert
membership in that group.
Control for conformity to the manner of co-participants in real time may
conflict with control for conformity to the remembered or imagined
manner of (intended) co-members of another group.
Disparity of manner that goes unnoticed is socially unmarked. Disparity
of manner that occasions conflict (often in both parties) is socially
marked.
A shibboleth is a manner that control systems take as a sign of
membership in a social group, such that disparity with it is always
socially marked.
Children who are taught shibboleths and other socially marked manners
for their social advantage, rather than learning them as co-members of a
social group that controls for them, may come to associate them with
emotional valuations of good (praiseworthy) and bad (punishable).
An assymetric social transaction is one in which one-down participants
control for conformity to the expectations of one-up participants.
People who experience a one-down role in an assymetric social
transaction typically seek ways to play a one-up role in the same kind
of transaction with other players.
Children experience themselves as one-down a lot. Children play one-up
roles with other children a lot. Children who learn shibboleths and
socially marked manners by having them enforced in a one-down way turn
about and enforce them on other children in a one-up way.
And so it goes.
Control theory does provide a framework in which people may understand
and cultivate non-hierarchical social arrangements. It does not
guarantee that they will do so. That hoped-for evolution into a world
of fraternal/sororal amicability depends upon resolution of emotional
craving for redress, in a vast number of individual living control
systems, out of a long heritage of assymetrical social arrangments.
The abused child who grows to an adult who abuses children marks only
the tip, emerging over the horizon, of an iceberg whose distant shores
have yet to be touched by any boatload of social benefactors.
Control theory has very little to say, yet, about the basis of
attachment (craving, addictive demand) or emotion.
Even if we do all agree on non-hierarchical social arrangements, or when
we admit only leadership based on personal attributes needed for a
shared purpose presently at hand (the anarchist program), people will
still control free variation for conformity to the manner of co-members
of a social group. They will still bring free variability in their
behavioral outputs under control for conformity to one another for
social purposes. Like mammals in general, and perhaps lower organisms
too, people will still bind their freedom to social purposes.
Withal, it is important I think to distinguish autonomy from
independence. Independence is freedom from external constraint.
Autonomy is control within one's own domain. But one's domain is
defined by participation in social groups and social arrangements. They
incur a cost in independence. If the cost of particular social
arrangements is not outweighed by a gain in autonomy, one seeks ways to
change them. One may exchange one set for another. Attempts to abandon
them entirely don't appear to work. Perhaps that is a biased view from
within my social arrangements, but if anyone knows better from
experience they (by definition) aren't telling us. Conversely, if we
exclude from Control Theory an account of control for social conformity,
we are not talking about people.
Bruce
bn@bbn.com