Tom & Rick (inadequately) on hf-control

From Greg Williams (931020)

From Tom Bourbon [931019.0829]

I think I understand what you (as a PCTer) mean by "controlling their
own perceptions." Do you think you understand what the nonPCTers meant
by "control [of] people"?

Had you asked me that question, I would answer, "Yes, I do. They mean
exactly what they (people in the HF community) say: information, stimuli,
cues, error and the like in the environment control our actions."

You might indeed understand what they mean, but what you say certainly does
not convince me of that. Good thing you don't edit dictionaries; it isn't
very helpful to specify the meaning of a word in terms of that word. What I
was looking for from Rick (and now from you) is evidence regarding how
various HF folks use the word "control." I am still looking.

The
HFers even play along with the test, as Rick indicated. All you need to do
is say, or suggest, that people control their own perceptions and you
encounter the corrective actions -- raised eyebrows, restatements of what
you said that "correct" your mistake, indifference, changes of subject, and
so on.

Don't you think it at least possible that if they were using the term
"control" differently than PCTers, then it might not make as much sense as
it does to you to think of perception as "controlled"? When you "say, or
suggest" that to them, do you explain the idiosyncratic meaning of
"control" which you have in mind? Or do you let them struggle with fitting
their own conceptions of "control" to "control of perception" and thereby
"play along with the test"? Have you made sure that you are testing what
you think you are testing?

Bridges can't be built from the beach into the ocean. There must be another
side.

Certainly. A preliminary question is whether anyone -- who? -- is hopelessly
at sea because they misunderstand the meanings of terms being used by those
on the other side.

And when people build bridges, isn't it a common practice for the
construction to come from both sides toward the middle, rather thanm from
just one side?

Yes, but it is more difficult to meet at the middle when the opposing
builders are using different standards of measurement. Of course, one side
can impose their standards upon the other, and likely end up building a
bridge which will be little used by the resentful other side.

Might their "control" translate to your "influence"?

It is not for us to tell people their words mean something other that what
they insist they mean.

I am not asking Rick or you to do that. I am asking what you think they
mean by "control," without using the word to do it, and with some supporting
evidence. I myself think that they might be talking about what PCTers call
influence of perceptual signals on actions -- especially when what some of
them term "intentions" (internal to organisms) of relevant sorts are kept
constant. For example, given that I keep intending to track a cursor, then
my actions might be said to be "hf-controlled" by the cursor's movement --
that is, my actions would be influenced by the cursor's movement -- in ways
which can be accurately predicted by some hf models. But I have not met
face-to-face with so many hfers as Rick and you, and I thus haven't had the
opportunities which you have had to ask them whether this is what they mean
by "control." Please add to my knowledge on this, if you can.

Many of them are certain that our phrase, "behavior
controls perception," translates to their "behavior is a response to, or is
controlled by, perception."

That seems to be more evidence that their meaning of "control" has more to
do with influence than PCT-control. When approaching these folks, do you make
it a point to define your meaning of "control" for them, and to ask their
meaning of "control," to avoid confusion on both sides?

It is the idea behind the words that counts:
lineal causal systems rather than perceptual control systems.

The idea behind at least some hfers' ideas is the same as that promulgated
by PCTers, at least at the model/mathematics level, as I have documented
with references. At least some hfers show reference levels inside people
with negative feedback systems which PCT-control perceptions. These models
are LOCALLY causal systems, the same as PCT models; they are CIRCULAR
causal systems, the same as PCT models. In fact, they are generically
identical to PCT models, although some are more complicated in their transfer
functions because the hfers are trying to model a broader range of control
difficulty than are most PCTers.

Do you think what
they mean by "control" is completely incompatible with what you (as a
PCTer) mean by "control"?

Yes. When "they" (the majority of HF people I know of) speak of the
causes of behavior.

I am suggesting that their meaning is different than that of PCT-control,
but not incompatible. Unless they are contradicting their own models. Do
you have evidence that that is in fact the case? Is there interpretation
generally inconsistent with their own models? Please give some references.

To quote you (more or less) on this topic:
what do the their models say, as opposed to what do they say about the
models?

For many of them, the models "say" things like, "environmental information
guides/directs/causes actions," or "error (in the environment)
guides/controls/directs behavior," or "people (as control systems) control
their actions/outputs/behavior." From what I saw in Rick's posts, those
must have been the kinds of things the models said to him.

The mainstream hf models I've seen do NOT "say" that people PCT-control
their actions. One more time: they are the SAME as PCT models, PCT-
controlling their perceptions. Nor do they "say" that error (in the
environment) PCT-controls behavior. As for what the models "say" about
hf-control, hf-guidance, and hf-direction, I am still hopeful that you
and Rick can inform me what those terms mean.

Greg, have you talked to any "real" human factors people recently who
understood and accepted the idea that people control many of their own
perceptions?

That is what the mainstream models say (here, I am taking you to mean
PCT-control). I have not talked with ANY hfers recently, but I believe
that at least those who have PCT-like models would readily understand and
accept the idea -- PROVIDED that the notion of PCT-control is explained
to them. I don't expect ANY of them to accept the notion that people
hf-control their perceptions, but that is a guess, since I still don't
have good data on what hf-control means.

Have you read many recent articles and chapters where HF people
state that idea clearly and unambiguously, so the reader does not need to
read the meaning in, or wonder if that might be what the writer intended to
say? If so, pass them along. I haven't come across them during our
discussions with the Feds, and it looks like Rick didn't find them at the
professional meeting he attended.

Several hfer sources (some of which I've posted to the net) state that idea
as clearly and unambiguously as possible: with mathematical models. If you
want to quibble over the clarity and ambiguity (to whom?) of the appended
text, I think (once again) you must first try to understand how some of
the terms used by hfers have possibly different meanings for hfers than
they do to you.

The point of all of this is that there IS another side, which has already
built a fair portion (perhaps more than halfway! :slight_smile: ) of the bridge between
PCT and HF. They have built it on the same rock-solid foundation of PCT-
style mathematical models as you could build your side on. It would be a
shame for the two sides to miss by a mile because each didn't know that
the other was using a different-sized meter stick.

From Rick Marken (931019.0900)

Well, I see Tom Bourbon already got to this. He said it all much
better than I could -- and with a lot more cool.

Does that mean you aren't going to try to answer my questions?

All I can say
is: Didn't you read the the posts from Wickens and Lintern!?!?

Yes, I did. Were the answers to my questions in them? Where?

Did
you see ANY evidence of an interest in bridge building on their part?
Come on.

What I think I saw was confusion between HF-control and PCT-control. Also
confusion about other terms, such as (dynamical systems) "attractors."
That might help to account for the attitudes on both sides, despite the fact
that some of their models are the same. So I guess there was some evidence
there for my suggestion that some PCTers and some hfers talk "past" each
other. Do you think I am wrong about this? Why or why not?

Also, bridges aren't built until they are built; people who want
to build them keep working on them. Haven't you noticed what happens
to those to whom you would have us build bridges -- they DISAPPEAR.

People disappear for many reasons. Some after attempting to build bridges
for a long time, to no avail. Some have less patience. And some have
previous commitments.

Where is Wickins' answer to Powers' explanation of why error is NOT
the cause of action?

You'd have to ask him. Perhaps he agrees sufficiently to not need to
argue. Perhaps differences in meaning caused communication difficulties.
Perhaps he is attending to unrelated reference signals which he judges
to be more important. Without asking him, I would hesitate to brand him
a bridge-building-quitter.

Where are the descriptions of examples of the
test for controlled variables -- a test that is presumably well
understood by manual control theorists?

We've been through this before. There are many examples of hf model-
building which have required experimental tests to determine controlled
variables (see Milsum's book and back issues of HUMAN FACTORS and
BIOLOGICAL CYBERNETICS).

I imagine that you would say that these people duck out of the bridge
building activities because those of us on the PCT side are so rigid
and uncompromising; or because we are just plain assholes.

My hypothesis is that it is at least sometimes due to misunderstanding of
terms used by those on each side. I certainly would never think of YOU as a
"just plain asshole." :slight_smile: And you have a good excuse to be rigid: (I
think) the PCT model is the best one around. I'm just saying that at
least some hfers subscribe to the same model, and I'm conjecturing that
there is more a definitional/meaning problem than a genuine disagreement
between those hfers and you. You can easily prove me wrong by showing
that hfers deny PCT-control-of-perception and/or hold to PCT-control-of-
output. It will not do to show that they deny control-of-perception
and/or hold to control-of-output if you aren't sure they mean PCT-control.

Believe me, the people like those at the HF meeting don't want to build
bridges to PCT because they see it as too simple minded, ignoring all
the important new "discoveries" (like chaotic attractors), and way
out of date.

Give me some evidence and I will believe you. Even if some are simply
unimpressed by PCTers' accomplishments, meaning misunderstandings could
also play a part in their eagerness (or lack thereof) to build bridges.

They know this becuase we use simple little models to
make simple little points.

Some of them use the same "simple little models." Some use elaborations
of them to make more complex points, like understanding how control
deteriorates when certain tasks are made more and more difficult. I
think some PCTers could benefit from looking at those models, just as I
think some hfers could benefit from learning about PCT. Of course, I
also recognize that the goals of PCTers are generally different from
the goals of hf engineers.

Of course, if they bothered to try to
understand these simple little points they would realize that all
their complex theories and data are useless (as in "Useless Bay").

Maybe some are -- it's hard for me to judge solely on the basis of an
assertion -- but I've seen some that aren't -- the ones which are
elaborations of PCT-type models.

Rick Marken (931019.0915)

Rick Marken thinks that most HF engineers and other conventional
social scientists are INTENTIONALLY clueless; they don't know
anything about the nature of control (purposeful behavior), how it
works (control of perception) or how to study it (test for controlled
variables) -- and they treat attempts to teach them about these things
as threats (disturbances).

Abundant evidence that you are not a just plain asshole! To the contrary,
you make my day. I love that "most." So tell me -- who are among the
fortunate few? I promise I won't tell those who didn't make the cut.

As ever,

Greg