Beach model, decisions

[From Rick Marken (940616.0800)]

Jeff Vancouver (940615) --

Beach (1990) has a PCT-like model that describes when DM
[Decision Making?] is likely to be conscious and when it is not.
His model needs more data, Rick,

Could you give a quick description of the model? Is it based
on ANY data? What is the data? How well does the model account
for the data? Is it implemented as a working model?

but basically the idea is that processing occurs unconsciously

>(and presumably in parallel) unless the alternative results

in too much error.

You've really got to get that spreadsheet model of a control
hierarchy fired up. You will see that the term "processing"
can have implications that are misleading regarding the operation
of the HPCT model. "Processing" can imply a sequence of operations
(such as those that happen in a meat processing plant) that are
performed on input (like cattle or perceptions) and result in a
product (like steaks or decisions). This is not the way "processing"
works in the HPCT model. All variables in the HPCT model are changing
simultaneously and all inputs are at the same time products of them-
selves. The model always operates as a parallel (analog) processor; it
makes no sense to talk about an "alternative" to parallel processing;
the processsing is always parallel, though systems can be switched
in a out of the hierarchy by various means; in an Excel version of the
spreadsheet, you (acting as consciousness) can do this by switching
any control system in the hierarchy into imagination mode. Consciousness
is not part of the spreadsheet model. It could be, however. All you
need is a "super" control systems that perceives the state of
perceptual variables in the model (awareness) and injects inputs into
the output paths of the model (volition).

This seem very compatible with Mary's notion of "unless
there is conflict."

Perhaps. But I think Mary was referring to the fact that, under
ordinary circumstances, controlling does not involve "decision
making"; you just make your perceptions match your reference
signals. But, when you are in conflict, there is error in (at least)
two control systems and any effort to decrease the error in one
system increases the error in the other. Consciousness (in HPCT)
presumably moves to control systems with chronic error so we
become conscious of these "damned if I do, damned if I don't"
situations. Our consciousness of this vascillation is what we call
"decision making " -- but there is no "decision making" process
per se in the HPCT model (though I suppose the branch points in a
program perception could be called a decision; but these are perceptual
decisions that are not necessarily conscious; in order to control
a program you have to be able to perceive that there is a choice of
options; that there is a criterion for selection of an option and
that one of the options was selected -- and that choice was either
consistent with the program or it was not --See my post on program
control).

Best

Rick