Million references

[Dag Forssell (950305 2300)]

[Bill Leach 950303.22:40]

Is it not possible that an individual might have several million
references ABOVE the intensity level? All as active control loops
with 'nothing going on' (ie: the reference condition is
satisfied)?

Let me answer your question with some of my speculation.

We recognize that by way of imagination, we can think about a lot
of things. As suggested in B:CP chapter 15, we do this by
specifying desired perceptions and delivering them from memory
(which has to be addressed or searched somehow) instead of from
"real" inputs which have to be built from the intensity level.
When you think or imagine about anything at all, you do so by
specifying reference signals from memory and playing them back up
as perceptions. You probably don't have millions of reference
signals, but billions. Suppose you are watching a sunrise with
your girlfriend. Are you just watching with passive observation up
a ladder of perceptual functions? Or are you actively imagining
with the help of your control systems and memory as you watch?
Suppose the sun does not rise as you expect. Would you
experience an error signal?

In my talk in Durango last July, I suggested that perhaps the
perceptual and memory switches shown in B:CP Fig 15.3 are not
switching at all, but are "on" all the time in all directions.
I propose that some difference between signals or clusters of
signals prevents confusion. This difference could be a difference
in complexity, richness or completeness of an associated cluster of
perceptual signals, much like dot matrix letters compared with
fully formed type.

Let us assume you are just passively watching the sunrise or
whatever. Assuming that all connections are active, a current
perception would be continuously stored in memory over or along
with previous sunrises, updating "sunriseness." Memory at one
level is associated with addresses to comparators and comparators
with output functions which in turn address other memories. The
entire control hierarchy is active and controlling, but without low
level intensity output to muscles or physiology (or perhaps with
some).

I observe that a few fragments of low level intensity signals can
give rise to a complex high level perception. When I write
"sailboat," you fill in the details -- if you have any memories at
all associated with that word. We autocorrect words so we hear
them clearly even though they were spoken with a slur or
incorrectly. We follow the progress of speech and anticipate what
the speaker will say next: Things go better with... (Coke). We
jump to conclusions. It seems to me that a high level perception
can be based on 1% input and 99% stored memory or vice-versa.
Either makes sense to me if we allow that somehow the entire
hierarchy is active when we passively observe, not just perceptual
functions, and that we play back associated stored memories (as
reference signals) along present perceptions, supplementing them.

I agree with you. Our memories are addressed and serve as
reference signals. We have lots of them. They combine to
constitute our personal understanding or beliefs about the world
and what we want it to be.

Best, Dag

<[Bill Leach 950306.02:34 EST(EDT)]

[Dag Forssell (950305 2300)]

I will need to give some more thought to what you posted but I feel like
you were sort of talking about something that is a bit different that
what I was thinking.

When I said 'several million references ABOVE the intensity level?', I
was trying to specify 'normal' control loops having perceptions
(primarily sensory), reference values (usually satisfied), and a
(usually) zero output feeding into maybe the 'program level' for the
ultimate purpose of causing physical action to restore perception if it
exceeds the limits set by the references.

-bill