[from Mary Powers 9405.30]
Well, the Iron Horse race is over, and a big success as far as we
are concerned, since our daughter Barbara was in it and finished
in a respectable 4:20 (fastest man, 2:10, fastest woman, 2:30).
It's a 47 mile race, beginning at 6500 feet and ending at 9300,
with two passes to cross, one at 10,600 and one at 10,900. It
makes my legs ache to think about it. (Iron Horse refers to
racing the train from Durango to Silverton - in Barbara's case,
the train won, but heck, it follows the river and doesn't have to
go over the passes)
Anyway, to get back to PCT, and Bob Clark's post of 5.27. Was
Barbara's ride a planned action? Hardly. It was a planned
perception: herself and bike in Silverton on the 28th of May. In
order to get there, she did indeed plan: out riding every
possible day, working up to getting up the hill to Fort Lewis
College easily, and later other higher hills, and riding for an
hour, then two, then three. Did she use "stored programs" to do
this? No, I believe the program of training was very general -
get out and train - but the details were constructed as she went
along - had to be, as they depended on constantly changing
circumstances, such as weather and wind, and the availability of
baby sitters. I don't think a "DME" switching switches had
anything to do with this. Simply a higher level control system
setting reference levels for levels below.
"Bringing programs up from storage as needed" seems to me to be
really alien to PCT. Obviously when she was cooking dinner or
changing diapers she was not training for a bike ride, she was in
the midst of other programs satisfying other reference signals.
So in a sense her bike riding program was "stored", but only in
the sense that her reference signals for being a better and
better bike rider were not being acted on at the time. Nor did
the fact that they were not being acted on constitute an error,
since her reference signals for being in training did not require
bike riding to the exclusion of everything else.
This suggests that the higher level reference signals are not
"unchanging", but rather than they change all the time. As one
perception or another begins to deviate from the reference state,
that is the error one attends to - hunger, the fussy baby, the
training schedule. I totally disagree that "the entire system
becomes essentially mechanical in the sense that its performance
is completely determined". This is planned action again. Control
systems don't work that way. For instance, taking care of the
fussy baby by changing his diaper will gradually give way to
taking the little guy to the potty - a totally different group of
actions. And meanwhile changing the diaper can take place on a
bed, the couch, the floor, in a car, whatever. Hardly mechanical
or completely determined.
On the subject of consciousness. My example of discovering that I
was imagining rather than reading only involved consciousness
when I realized what I was doing. The shift from reading to
imagining was not conscious. Bob seems to define unconsciousness
as only being states like sleep and anesthesia, but control
systems go on working whether or not you are paying any attention
to them. Otherwise you would quickly be dead.
There's more in Bob Clark's post that I will not comment on,
mainly because I don't understand what it's about.
Mary P.