Viatractor?

[From Rick Marken (961018.1300)]

Martin Taylor (961018 12:00) --

[Description of reorganization model of the development of social
conventions].

Is this a new mechanism?

No. It's the good ol' PCT reorganization model.

Does it describe how social conventions are attractors of the dynamics of
interacting reorganizing systems?

No. It describes how social conventions (convergence by a group to the
control of certain perceptual variables) can result from reorganization. The
convergence can be _described_ as an attractor; so it would be more
appropriate to say that reorganization theory _explains_ the appearance of
social conventions as "attractors of the dynamics of interacting reorganizing
systems".

If one starts by postulating a mechanism that may lead to an attractor,
then one can assert that the attractor will be of one of the three kinds.

I don't see how it could _not_ be possible to predict, given the mechanism,
that the resulting behavior will be like one of three kinds of attractor. I
can see how the discovery that social conventions can be described as
"attractors" could lead to the invention of mechanisms (like reorganization)
to explain why that appearance is observed. But I really can't see how the
description of a phenomenon (as an attractor) can lead to the prediction of
further phenomena. For me, the theoretical success has already occurred when
we find that the reorgnaization mechanism can explain the development of
social conventions. Seeing that the development of social conventions looks
like an "attractor" is interesting and trendy but (once we understand
reorganization theory) pretty much irrelevant.

Best

Rick

[Martin Taylor 961018 16:05]

Rick Marken (961018.1300)

Martin Taylor (961018 12:00) --

[Description of reorganization model of the development of social
conventions].

Is this a new mechanism?

No. It's the good ol' PCT reorganization model.

Aaaah! (==sigh of relief) So glad you agree.

Does it describe how social conventions are attractors of the dynamics of
interacting reorganizing systems?

No.

How do you reconcile that "No" with what immediately follows?

The convergence can be _described_ as an attractor;

That seems to demand a "Yes".

so it would be more
appropriate to say that reorganization theory _explains_ the appearance of
social conventions as "attractors of the dynamics of interacting reorganizing
systems".

That's what I've been trying to tell you, yes:-) Hurrah!

If one starts by postulating a mechanism that may lead to an attractor,
then one can assert that the attractor will be of one of the three kinds.

I don't see how it could _not_ be possible to predict, given the mechanism,
that the resulting behavior will be like one of three kinds of attractor.

Actually, given the mechanism, one can go further, and assert _which_ of
the three kinds of attractor it will be. I didn't, because the statement
is that no matter what the mechanism, there are only the three kinds of
attractor, and if the mechanism leads to an attractor, it will be of one
of those three kinds.

Since the specific mechanism is reorganization, and reorganization rate
depends on some criterion (we say it is a function of error, but that's
not essential) and is a scalar, the dynamics of the social interactions can
be described in terms of a potential surface, with the criterion being the
"potential." The lower the reorganization rate, the lower on the surface
is the social structure, rather like a marble running into a rather
weirdly shaped bowl. (Of course, reorganization being in part random, the
path of the "marble" won't be straight downslope, but will jitter, tending
on average to be downslope.)

The end result is that the attractor for the social dynamics of an undisturbed
set of interacting control systems must be a point attractor, not a cyclic
or strange attractor (neither of which can occur on such a potential
surface). In other words, communities that do not interact with the outer
world and that retain a stable size will eventually come to a never-changing
social system. Perhaps some small isolated communities did so before
the recent impact of the Western world upon them, but we'll probably
never know.

For me, the theoretical success has already occurred when
we find that the reorgnaization mechanism can explain the development of
social conventions. Seeing that the development of social conventions looks
like an "attractor" is interesting and trendy but (once we understand
reorganization theory) pretty much irrelevant.

Relevance depends on one's viewpoint. It does not disturb me that it is
irrelevant to you, and nor should it disturb you that I see this particular
example of an attractor to fit within a much more general class of behaviours
of dynamical systems.

For the particular case of developing social conventions, the mecahnism
is important. It would be even more important if it were possible to
single out some parameters that determine which conventions are likely
to be universal or nearly so, and which are flukes of particular historical
circumstances.

Personally, I find it relevant, but you may not, to know that in nonlinear
interactions, attractor basins tend to have highly interwoven (actually
fractal) boundaries, because it means that when conventions are beginning to
be established (or re-established following the merging of two groups),
historical fluke can be very important in determining just which of many
possibly stable sets of conventions form the attractor toward which the
structure evolves.

In an even more trivial sense, it could be relevant to know that there almost
certainly (again in the statistician's sense) are many point attractors,
a.k.a. stable social systems, and one should not be surprised that in some
societies, behaviours that we consider immoral should be taken as normal.

But I really can't see how the
description of a phenomenon (as an attractor) can lead to the prediction of
further phenomena.

I guess my examples are pretty weak. However, once one recognizes that a
system is just a specialized exemplar of a wider class of systems, everything
that is true of the wider class becomes useful in looking at the special
system of interest (not necessarily true, because the special system may
involve some kind of exceptional condition, but useful nevertheless).

That's why I so often bring into the discussion facts that are true in
general of systems of which control system collections (hierarchies or
groups) or single control systems are exemplars. And since you tend to be
focused specifically on control systems _as such_, you tend not to see
the relevance of thes facts or ideas. It doesn't present a problem that
you fail to see the relevance. It does generate a disturbance when you say
that they therefore _are_ irrelevant (or worse, wrong).

To use hammers, chisels, bandsaws, sandpaper and screwdrivers in making
a cabinet is not to say one is uninterested in proper cabinet-making,
which _ought_ to be done with a pen-knife alone (or was it teeth:-).

Martin