Association and Degrees of Freedom

[From Bruce Abbott (950113.1245 EST)]

(Bruce Nevin Fri 930113 10:17:51 EST) --

I don't understand how habits and skills work. They appear to be
familiar pathways through the field of available degrees of freedom,
rather than control of the manner of controlling a higher-level
perception (this path through the field of dfs, rather than the other
possible paths). I don't know enough to think clearly about it--my
speculations are too wobbly for words. Probably something obvious.

Just after reading Bruce's post, I went into the room next door to retrieve
a cup of coffee I was reheating in the Departmental microwave. I had about
half a cup left and was going to supplement the reheated stuff with fresh
coffee from the pot, which was about one-quarter full. My mind was
apparently elsewhere (typical for me), because the next I knew, I was
loading the coffee filter to make a new pot.

Now, I had no intention of making a fresh pot when I left my office, and
there was clearly still too much relatively fresh coffee in the pot for that
to be appropriate behavior. Yet there I was, going through all the proper
motions to achieve a goal I had no conscious intention of carrying out. The
traditional explanation for this common phenomenon is that repeated
execution of certain sequences of intentional behavior forms and strengthens
associative connections so as to form links in a chain, with stimuli
signaling successful completion of one link somehow activating behavior
appropriate to the next link.

In PCT such organized sequences would be attributed to the sequence or
program level of control, I would think, depending on the complexity of the
sequences. (In my example the program level would be appropriate.) Bruce
says that he doesn't understand how habits and skills work; I suspect that
none of us do. Which brings me to a question: what role should association
play in PCT? Obviously the traditional notion of skilled execution as a
stimulus-response chain won't wash, but it seems to me that realization of
one goal in a sequence of goals that must be accomplished could be the
"switch" that passes control to the next control system in the sequence.
But what causes the wrong system to be selected next? Why do I find myself
fixing a new pot of coffee after entering a room, when my intended goal was
to retrieve my cup? The concept of association seems to handle this fairly
well by suggesting that the association between the stimuli identifying the
room and making coffee had been repeated often enough to make the "response"
of making coffee highly probable in the absence of any interference from
higher-level (conscious) systems. For that matter, what causes the "right"
sub-goal to be selected next, as is more typically the case?

It does seem that following well-worn pathways to a goal would serve to
reduce the need to deal with "excess" degrees of freedom, as Bruce suggests.
Research project, anyone?

Regards,

Bruce

[Avery Andrews 950107.1402]
(Bruce Abbott (950113.1245 EST)

I certainly have no story about these effects, but it certainly does seem to
me that classical `response chaining' is something that actually happens.
In memorizing long poems (one of my occasional hobbies), it's quite easy
to get stuck in loops, whereby, since a passage A is similar to a later
passage B, you wind up following B with what comes after A. This is
a problem with Homer, because of the extensive use of formulas, repeat
lines, and even repeat paragraphs.

Avery.Andrews@anu.edu.au