PCT & evolutionary epistemology

[T_B_Harms1]

From Tracy Harms 1997;01,10.12:00 UCT

Bill Powers (970109.2110 MST)

Tracy Harms 1997;01,09.19:00 MST --

... we may indeed say that a claim
communicates an objective fact of reality. In saying this we might be
wrong, but we might perhaps be right.

Ah, that's the basic problem of epistemology, isn't it? The problem is not
whether our statements about reality are right or wrong, but that we DON'T
KNOW WHICH. If they're right, but we can't prove it, we're no better off
than if they were wrong, knowledge-wise (survival-wise might be a different
matter).

The problem is that our certainty (confidence) in articles of knowledge has
*something* to do with correctness of knowledge, but the standard approach
of thinking of knowledge as a subset of certainty is bankrupt. Aware of
that failing, you assert "that we DON'T KNOW WHICH." Here "know" is
synonymous with inerring certainty. Unfortuanately that very
conceptualization of knowledge is a flaw which produces and entrenches most
of the problem you indicate. Especially indicative of this is your
reference to proof. The desire for something like proof as a certification
of knowledge is the chief obstacle to improved understanding of knowledge.

While I have become convinced that abandoning pursuit of such certification
is vital to a sound rationalism, doing that is not in itself a solution to
key questions of epistemology. Even after we decide to do without warrant,
rational integrity suggests that we should obtain improved explanation.

It was in light of a most impressive explanation of knowledge that I came
to pay attention to perceptual control theory. Knowledge, in this view, is
a class of situational relationships. Your model of perceptual control
systems is the most simple and straightforward example I have found to date
of a knowledge structure. It allows many facets of the problem-solution to
be discussed as concretes. For example, it is handy for clearly involving
subjective experience (perception), which most people unfortunately (but
understandably) conflate with knowledge. This offers a neat example for
attempts to communicate one aspect of this evolutionary epistemology: the
*subjective* qualities of a perceptual control system are the signals and
signal-processing within the system; the *objective* aspect, which is much
more properly called 'knowledge', is the complete system itself.

In addition to the ideas of science and philosophy there is also biography.
I don't know what the relationship between you and Donald T. Campbell was,
but it seems that you were colleagues and friends. I know Gary Cziko
worked with him also. I bring this up to touch on the fact that our ideas
are in a sense larger than we are, and that authorship and access to
authors does not privilege work with ideas, excellent though such presence
is. It is a bit awkward to assert, as I am now asserting, that there is a
convergence to the work of Donald T. Campbell and William Powers which you
two men did not yourselves identify despite your longstanding personal
communication. Whether or not the two of you were more or less aware of
this, there is great value in this confluence of your legacies. From the
way you write about knowledge it appears that you never really grasped the
epistemological reformation which Campbell identified. Given your
opportunities for learning that in the past, I'm not taking up the hope of
tutoring you in it now. But I am stating that your personal opinions
regarding the nature of knowledge are not wedded with your major theory,
and that PCT is more fruitful in regard to epistemology a la Campbell than
it is with your own views on the topic.

Thanks for you contribution, Bill.

Tracy Bruce Harms
Boulder, Colorado

[Bill_Powers1]

[From Bill Powers (970110.0745 MST)]

Jeez, the 10th of January already. Where did 1997 go?

Tracy Harms 1997;01,10.12:00 UCT --

The problem is that our certainty (confidence) in articles of knowledge
has *something* to do with correctness of knowledge, but the standard
approach of thinking of knowledge as a subset of certainty is bankrupt.
Aware of that failing, you assert "that we DON'T KNOW WHICH." Here "know"
is synonymous with inerring certainty. Unfortuanately that very
conceptualization of knowledge is a flaw which produces and entrenches
most of the problem you indicate. Especially indicative of this is your
reference to proof. The desire for something like proof as a
certification of knowledge is the chief obstacle to improved understanding
of knowledge.

When I speak of "proof" I don't mean the establishment of objective truth. I
mean only applying criteria of proof to see if they are met by observation.
We make up the criteria just as we make the observations.

To me, proof is simply what happens when we run out of alternative
explanations and find that only one remains that we can't find anything
wrong with. The big question in my mind is what relationship that kind of
knowledge has to the extra-sensory world, the existence of which all
evidence seems to support, but the nature of which we can't verify by any means.

I think you may have miscontrued my statement about "not knowing which".
Every time we construct a model of what we can't directly observe, we test
it by applying experimental manipulations to that invisible world and seeing
what changes in our perceptions result. Our best models are so good that
everything, EVERYTHING, we can think of to do to that other world results in
changes in our perceptions that are EXACTLY CONSISTENT with what the model
predicted, as closely as we can measure. Of course such ultra-reliable
models tend to deal only with simple phenomena, but that's beside the point.

One can't help getting the feeling that when models work that well, they
must be approximating something that's really out there. And this is where
my statement comes in. It's possible that these models do in fact reflect
structures or organizations in the actually-existing world. It's also
possible that they do not. But we can never know which is the case, because
we can't open the black box and look inside. Even when we seem to do so, as
when we finally managed to get a picture of the double helix, we find that
the black box contains other black boxes.

You may have noticed, on the Web page and in some CSG books, the CSG logo.
The environmental feedback function is shown in the logo as a black box.
This was Greg Williams' idea, and it was Ernst von Glasersfeld who inspired
it by saying, "The organism isn't the black box; the environment is." Note
that "the disturbance" is hidden inside the black box, too.

I don't think it's possible to arrive at final truth by philosophical means
any more than it is possible to do so by experimental physical means.
Evolutionary epistemology is, like all other ideas, a human idea that
somebody made up. It's a _system of ideas_, and as such it is a model of
reality. When you or I say that we can't know directly what is going on out
there, that, too, is an idea that might be right or wrong -- that is, it
might explain our experiences perfectly or incorrectly. But how can we ever
find out if it's the _right_ idea about the relation of human experience to
reality? It, like all statements about reality, is a proposition, and like
all propositions about what we can't observe directly it is unprovable. It
may be right, or it may be wrong. But we can't tell which.

Hey, when are you going to go to California and quit bugging me?

RE: Don Campbell

My relationship to Don Campbell was rather strange. I met him in 1960 at a
meeting on Self-Organizing Systems (I might have known OF him in my
undergraduate days before 1950), and showed him a draft of the first
in-press paper on what we now call PCT. He was immediately supportive. In
fact, he was supportive of my ideas all through our long sporadic
interaction. It took me a long time to realize that he didn't actually
understand control theory in any technical sense, even though, after 1973,
he assured me he kept my book on the table by his bedside. The thing about
Don Campbell was that he was a true scientist in every good sense of the
word. He knew I had a new idea that had some validity to it. And his
unvarying credo was that every competently expressed new idea deserves full
respect whether one understands or agrees with it or not, and especially
without regard to the status or reputation of its author. I think he lent
his support so unstintingly in part BECAUSE I had no Ph.D., BECAUSE I was a
lowly technician in a medical physics department, BECAUSE I had neither
prestige nor authority to demand serious attention from other scientists.
Don would defend to the death the right of any scientist to be heard whether
he supported the ideas or not. He didn't just support this principle: he
lived it. If it weren't for Don, B:CP would probably never have been
published. Fortunately, most fortunately, I was able to tell him how much
his support had meant to me, before he died. I guess I'm still mourning his
absence.

Best,

Bill P.