From Bob Clark (931105.2130 EST)
Another view of "Feedforward."
The discussion of Feedforward seems to involve several different
Viewpoints as well as different Time Scales. And there seems to be a
general assumption that the systems involved operate, literally,
simultaneously.
It is appropriate to consider that current in-coming perceptions
arrive simultaneously. But the related output actions, at all
levels, require various amounts of time, depending on the length and
complexity of the networks involved. As we found 30 years ago, the
lowest level, muscle systems, involve times of the order of ten
milliseconds. Higher level systems require longer times, and if one
considers interactions among people, the time scales can easily
become not only minutes, but also days or months.
Regarding the skidding incident, we have Hans Blom's report.
This report consists of a fairly detailed and dramatic description of
the events as he remembers them. When writing his report, he seems
to be taking the viewpoint of an observer describing the sequence of
events as they were recorded in his memory. At report writing time,
there is time enough to examine other related memories as aids in the
writing process and, perhaps, other memories of skidding. And, thus,
take an observer's viewpoint.
At the time of the SKID, his viewpoint appears to have been that of a
User. The User is the person attempting to control his environment
as currently perceived. He has barely time enough to consider (very
briefly) the alternatives. He reports keeping the wheel in place,
applying the brakes lightly, etc. Clearly, he has some knowledge of
skids and how to control them. Such knowledge is available, of
course, from his memory via imagination. Doubtless he has several
related memories from which he selected the procedure reported.
Attention can shift among perceptions very quickly. [Is there a way
to measure this speed?] Thus, having selected his strategy
("procedure?"), he continues to monitor ("pay attention to") the
approach of the metal obstacle and his own actions. Upon detecting
the presence of the edge of the pavement, he shifted his procedure to
other familiar procedures, again selected from his memory. And he
continued to pay attention to the movement of the car, its relation
to its surroundings and his actions.
From the viewpoint of an external Observer, who uses a longer
time-scale than the User: Hans is perceived to have entered a skid
and successfully brought the car back under control by means of his
mechanical skills.
It is easy to perceive these events in terms of a hypothetical
"feedforward" if one cares to. But as Bill and others have pointed
out, there are difficulties with a "pure feedforward system,"
especially if it is to be applied in other situations.
To the extent that "feedforward" implies "anticipation of future
perceptions," it is necessary to examine memories that resemble
current situations. Such memories can be extrapolated by established
methods. Of course any such anticipation is limited by the relevance
and accuracy of the available data. But even a limited extrapolation
can be very helpful in making decisions (that do not require the
Reorganizing System).
I have been puzzled for some time by the emphasis on automatic
systems and the exclusion of memory and imagination from the
theoretical analysis. They seem to be accepted concepts -- they are
discussed in BCP. However I do find occasional references to
"awareness" and "attention," at least in the form of "monitoring"
on-going events.
The slippers-by-the-bed situation can be treated similarly in terms
of a remembered route and the act of walking combined with continued
"paying attention" to the process as it occurs. The "wife's
slippers" call for a conscious, and very brief, shift to a "modified"
walking action. From the User's viewpoint this doesn't look much
like an "automatic system."
If you don't like these suggested applications of memory and
awareness, what are your alternatives?
Regards, Bob Clark