Tower of Babel

(Still getting the formatting right)
[From Bill Powers (960825.0745 MDT)]

Bruce Abbott and Martin Taylor seem to agree that a "model of the
environment" is any feature of a system that is somehow related to
properties of the environment. This leaves me in somewhat the position I
was in with respect to usages of "control." I have to try to find
another word to designate what I have always meant by the term "model,"
because it seems that "model" will now be understood to mean a lot of
things I don't mean.

I'm in somewhat the predicament of a physicist who is told that "mass"
doesn't mean just the ratio of force to acceleration, but also means
"size" and "weight" and "crowd of people" and "religious observance" and
"the process of brewing tea." The physicist, of course, would complain
that all these other meanings have nothing to do with each other, and
certainly nothing to do with the ratio of force to acceleration, and
that using them interchangeably will hopelessly confuse any discussion.
His opponents might argue that these are just symbols and their usages
are completely arbitrary, but the physicist would argue that the
different meanings should not be confused with each other, and that we
need some formal system for assigning symbols to keep them separate. The
physicist might even mutter audibly about the "Tower of Babel."

It seems to me that scientific discourse ought to be organized much in
the manner of a mathematical analysis. It doesn't matter what symbols
are used to designate variables, parameters, and functions, but once the
assignments have been made they must remain the same throughout the
argument. If it is agreed that "mass" is to mean "a religious observance
in the Catholic tradition," then it simply wouldn't do to begin speaking
of how difficult it is for social forces to cause changes in the rate of
change of religious traditions, and label this phenomenon "mass." That
sort of free association is the province of poetry, not science.

Perhaps a term I could use is "formal analog." A perceptual signal is a
formal analog of some external variable, in the sense that a physical
measure of the signal (its frequency) stands for and covaries with a
physical measure of that which is being perceived. A control system
diagram is a formal analog of a real control system, in that each
variable shown in the diagram (and used in the mathematical
representation) is in direct correspondence with a single variable in
the represented system. Furthermore, the boxes in the diagram indicating
dependencies are formally analogous to the properties of the real system
that make one system variable depend on other system variables in a
particular way; in the formal analog, those properties are represented
as mathematical functions. In a formal analog, we expect each variable
to correspond to one and only one variable in the real system, and each
function to correspond to one and only one property of the real system.
Furthermore, the connections among variables in the formal analog
correspond to connections in the real system; the formal analog has the
same structure as the real system.

The idea of a formal analog is a third-party concept: it exists in the
mind of someone who can see both the analog and the system being
formally analogized. What is interesting and logically challenging is
the proposal that a human being actually experiences only the formal
analog and not that which is analogized, with the formal analog being
the only thing experiencable as a "real world." For any person,
experience is therefore only a proposal about how the external world
MIGHT be organized. When the formal analog fails to behave consistently
with itself, it has to be reorganized until consistency is again
achieved -- so that when the analog acts, all perceptions affected by
the action behave in a non-contradictory way.

For me, the term "formal analog" in the above discussion is exactly
interchangeable with "model." But this is evidently not true for others,
so I will try to remember not to say "model" when I mean "formal
analog."

ยทยทยท

------------------------------------------------------------------------
--------------
Best,

Bill P.
Bill Powers
73 Ridge Place, CR510
Durango, CO 81301-8132
powers_w@frontier.net

[From Rick Marken (960825.0945)]

Bill Powers (960825.0745 MDT) --

It seems to me that scientific discourse ought to be organized much in
the manner of a mathematical analysis. It doesn't matter what symbols
are used to designate variables, parameters, and functions, but once the
assignments have been made they must remain the same throughout the
argument.

Yes! But if the other "players" in the discourse don't want to play by
these ground rules, then there you jolly well are, aren't you?

Perhaps a term I could use [instead of "model"] is "formal analog."

But the "Babylonian" will point out that any feature of a system that
is somehow related to properties of the environment is still a "formal
analog" of the environment because an analog (according to _the
dictionary_) is "a mechanism in which data is represented by a continuously
variable physical quantity".

I think it is impossible to solve the "Tower of Babel" problem by
inventing new words. It didn't work with "control" ("the X phenomenon"
never really caught on) and it won't work with "model" (though I
think "formal analog" has a nice, if somewhat starchy, ring to it).

The "Tower of Babel" can only be solved (as you imply) if people
voluntarily agree to a particular assignment of symbols to perceptions
(variables, parameters, phenomena) and stick with that assignment
throughout the discussion. I think the chances for such agreement are
greater for individuals who are seeking understanding than for
those who are trying to prove a preselected point.

I think your proposal that we use new symbols to refer to what we have
been calling "control" and "model" is a good one, however, because
it gives people who want to have a coherent discussion about purposeful
behavior a chance to form a "new covenent" regarding the arbitrary
symbols that will be used to discuss this phenomenon.

Your post made me realize that the "Tower of Babel" is not a curse imposed
by an external agent on a renegade population; it is a curse unintentionally
imposed by un-cooperative living control systems upon themselves.

Best

Rick