[From: Jeff Dooley 920326.1500]
(Martin Taylor 920324.1900)
Thanks for reminding me of the Skarda-Freeman article, "How
Brains Make Chaos to Make Sense of the World," in _Behavioral
and Brain Sciences_ (June 1987). Looking back over the
article, I'm struck by how much of it I failed to understand
on first reading. In retro, after conceptually adding control
dynamics to the mechanism that may be said to "achieve" chaos
in neural activity, this "achievement" makes more sense as a
product of control. Reference would be set (in individuals),
on this conjecture, through the lawlike (control?) activities
of the ensemble's generic organizing function (Kauffman's
model) manifesting ontogenetically. But why chaos? Actually,
here it seems to help to have Kauffman's notion of the "edge
of chaos"--a distinction I do not find explicit in Freeman
(perhaps I just missed it?) Anyway, Skarda and Freeman spell
out the value: *adaptability, or the ability to accommodate
and learn.* Quoting from p. 171:
"Without chaotic behavior the neural system cannot add a
new odor [sensation, vector, second level output?] to its
repertoire of learned odors. Chaos provides the system
with a deterministic "I don't know" state within which new
activity patterns can be generated. . .If the odor is novel
and the system does not already have a global activity
pattern corresponding to the odor, then instead of
producing one of its previously learned activity patterns,
the system falls into a high-level chaotic state rather
than into the basin for the background odor. This "chaotic
well" enables the system to avoid all of its previously
learned activity patterns and to produce a new one."
Obviously, a cognitive system whose livelihood depends upon
the inductive establishment and taxonomy of perceptual vectors
(as I think ours does) would benefit from achieving a degree
of structural plasticity like that of chaos and then to
maintain equilibrium around it. Is this possible "evidence"
of control activity on, perhaps, the phylogenetic level of
organization (as Kauffman would allow)? The authors even
suggest that "chaos is controlled noise with precisely defined
properties (p. 165)." But I don't see reference to *how* or
by what mechanisms such control may be explained. They invoke
local feedback (same-level, as you point out?) as a mechanism
by which bulb-wide activation vectors are enabled, but I don't
see further discussion of higher-order feedbacks. These are
all blank spaces that it seems PCT could help fill in.
(Gary Cziko 920325.1720)
Kuhn could only say, "In a sense that I am unable to explicate
further, the proponents of competing paradigms practice their
trades in different worlds" (_Structure of Scientific Rev._ p.
150). Your simple notion of these proponents controlling for
their different (incommensurable) perceptions potentially
rescues the whole concept of "theory-laden" observation from
the intentional depths where it has lurked out of empirical
reach for nearly 50 years.
Theory change. He remarks on the dynamics of theory change
with the question: which programme solves the more significant
question?. It seems that "significant" here has contingent,
pragmatic overtones, and is, in any case, a problematic
predicate. A new theory might also be judged superior for,
according to Kuhn, its 1. accuracy, 2. consistency, 3.
broadness of scope or generality, 4. simplicity ("bringing
order to phenomena"), and 5. fruitfulness, as in explanatory
power over previous anomalies and as in disclosure of novel
relations or predictions of phenomena. (Essential Tension, p.
322). Beyond these lies the subjective component--value-laden
rationality--that plays a crucial role in theory selection, as
I take it, regardless of the five criteria above. This notion
lands us right back in the explication muddle that Kuhn could
only shrug off but which your suggestion seems in a position
to help clarify. I'd like to explore this further but
parsimony forbids.
If it were possible to find someone who had grasped PCT (how to
determine?) and then went back to S-R we could ask "why?" and
gather a wealth of information on these questions of theory
choice, objective criteria, subjective criteria.
jeff dooley
dooley@well.sf.ca.us