[From John E. Anderson (950507.1715 EDT)]
Rick Marken (950505.0900)
NSD is a biological model of synaptic activity in the brain.
and Rick commented:
This would lead me to believe that NSD is an attempt to explain measured
properties of synaptic activity. I would expect to find no conflict between
NSD and PCT at this level. But then you started talking about how NSD
accounts for behavioral (not synaptic) phenomena.
and later, Rick said:
Perhaps NSD really is just an attempt to model neural activity in the brain.
If so, I could see this better if you told me what data NSD is designed to
explain and what experiments would be done to test the model.
You are right that NSD is not an attempt to explain synaptic activity in
the brain. I'm sorry what I said was confusing; I didn't intend for it to
be misleading. NSD is an attempt to use hypothetical properties of
synaptic function (that active synapses generate neuroactive envelopes)
and brain organization (that spatially-adjacent neurons have very similar
functions) to explain the evolution of functional localization in the
brain. A definitive test of these ideas requires a method of functional
imaging that could resolve brain activity on a spatial scale of microns
and a temporal scale of milliseconds in the same experiment. This might
be possible now using optical microscopy with voltage-sensitive
fluorescent dyes, but there is no guarantee that this would work. I hope
this gives you a clearer idea of what NSD is about. But I still don't see
where NSD is incompatible with PCT.
...NSD has a neurally implemented "predictive mechanism". This property
of the NSD model is not designed to account for synaptic activity; it is
designed to account for "quicker reactions" to perceptual changes. At this
point we have entered cause-effect land. In a loop "quicker reactions" to
perceptual changes are 1) ambiguous (are they lags or integration constants?)
What do you mean?
2) not necessarily an improvment (reactions that are too quick can cause
In NSD, neuroactive envelopes amplify any input that occurs adjacent
to very recent activity, namely that representing the previous state
of whatever perceptual quantity the activity represents. This leads
to a quickening of the response to that input: the neuron fires an
action potential in response to the input sooner than it would without
the envelope. According to NSD, the wiring of the perceptual,
reference, and error signal pathways of the control systems of
existing organisms places neurons with similar functions next to each
other -- ie exhibits a fine-grained functional localization -- because
then the envelopes' amplifying effect made their ancestors "mentally
quicker" and thus better able to survive and procreate. Evolution
will have weeded out those patterns of wiring that were maladaptively
unstable due to the quicker neural responses or any other cause.
and 3) are not really "reactions" because they occur in a loop;
if you build a model of an apparent cause-effect relationship (like that
between perception and neural "reactions") that actually occurs in a loop,
your model is probably on the wrong track.
But activity in the control loops of a real brain is carried by neurons,
is it not? And neurons are essentially input-output devices, albeit with
tremendous convergence of inputs and divergence of outputs. So I think it
is ok to talk about their "reactions" to input. It seems to me that if
PCT is based in biology, it has to live with the properties of biological
neurons, including their inherent input-output nature.
In a previous posting [Rick Marken (950502.0830)], you said:
part of a control loop so they are not caused by inputs. Rather, responses
are varied, as necessary, to produce the controlled result while
compensating for unpredictable (and usually undetectable) disturbances.
Without sensing changes in the controlled quantity caused by the
disturbance, how are the control systems to vary responses to control the
controlled quantity? And how can the animal compensate for disturbances
without detecting them?
Bill Powers (950502.0910 MDT)
occur at much lower bandwidths than suggested by the 9 millisecond
transport lag in a spinal control loop.
"Bandwidth" is often mentioned on CSG-L; what does it mean?
Bill P. to Martin again:
how much difference do the perceptual
lags we know about make in behavior? We have a rough estimate from
modeling tracking behavior. An integral-output model with no delay
accounts for 95% of the behavior; adding the best delay factor to the
model brings us to 97% or so. The behavior of the system with delay is
almost identical to the behavior of the system without delay. The only
reasonable conclusion is that while neural delays do have a detectable
effect, the effect is at the lower limits of our ability to measure it,
and plays no major role in behavior.
I think a source of confusion might be that the models and the real
organisms already have their "delays" optimized, the real systems through
evolution, and the model systems through parameter adjustment to fit data
from real systems. For example, you say:
Once we have the right data, we can propose a control model with
appropriate parameters and reference signal, and fit it to the data. We
can try various models and pick the one that reproduces the behavior the
best, and that best predicts new behavior under changed conditions. Then
we can be comfortable with claiming that we have a control-system model
that explains the observations.
It might be instructive to construct a control model using simulated
biological neurons, with biologically-realistic delays, instead of the
formal approach now used. I am not in a position to do so now, but will
when I am.
I am thinking about writing a grant, ...
That, of course, gives me the shivers because you might find out that
there is no evidence for PCT in the nervous system. But I've had those
shivers many times in my career, and have always gone ahead to take the
chance. If the theory isn't exposed to potentially lethal challenges, we
will never know if it's any good. Have at it!
I hope, however, that in making the judgments you will prepare yourself
by learning as much as you can about the PCT model as it applies to
hierarchical control processes. Feel free to consult CSG-L as much as
you like to get obscure points explained. Also, feel free to read the
literature described in the monthly Intro to PCT.
I intend to.
One specific thing you might be able to help me with now is to
explain to me the difference between PCT and other control system-
based models of brain function, for instance those of Scott Kelso,
whom my neurologist friend has mentioned.
All right, that's a very good way to start. I suggest that those who
have more ready access to the literature than I do start supplying some
quotes from Kelso (and others in the same line) which show what he and
they really say, so we can formulate specific statements about
differences from PCT. The models proposed by Kelso et. al. are very
definitely different from PCT, although in a few regards not hopelessly
so. Let's start a RE: Kelso thread, meaning Kelso's theories and others
That would be very helpful. When I told my neurologist collaborator,
Vinod Deshmukh, who has had an interest in control systems for some time,
about PCT, and how PCT considered the nervous system to be a hierarchy of
simple control systems, one of the first things he said was, "But this is
accepted!". He went on to mention Kelso and someone named Yingling at
UCSF as examples of theorists who have used control systems in their
theories. I am in the process of getting the book by McFarland and
Bosser, INTELLIGENT BEHAVIOR IN ANIMALS AND ROBOTS, mentioned by Rick the
other day, from the library. When writing the grant, which I am thinking
of submitting to NIMH or NSF (any other suggestions?), it will be good to
be able to contrast PCT with these other theories, in order to show why it
is worthwhile to support the research I would propose.