Re: opinions vs knowledge - The Cave is Constructed

Just a short post as I am having problems with my computer right now as I
gradually upgrade to Windows XP.

At 04:35 PM 9/9/2002, Ted wrote:
>>It is really have some difficulty to understand how we can think that
>>"logic" is a (narrow) human invention!!!????
>
>Fenzl--
>
>(My second posting this week. Limiting to two seems a good rule.)
>
>We are talking about four domains here:
>
>--1. How things actually work
>
>--2. How things appear to work
>
>--3. How we express and reason about how things work
>
>--4. How the entities involved "express and reason about how things work"
>
>Edwina used the term "logic" in the context of Socratic caves, Aristotle
>and Peirce. She used terms like "rational" and "reasonable." To my mind
>(and many I presume) her use of the term puts her use in domain 3. My
>point was that most workers in logic have many different logics. All are
>invented for different purposes and distort the reality they represent and
>about which they reason.
>
>You used the term "logic" in such a way that means (I think) that there is
>some coherent set of laws in domain 1 that result in reproducable effects.
>That's different.
>
>As common terms of art in computer science, I use "mechanics" and
>"physics" for domain 1, and "logic" variously for the rest. I believe
>there is a large distance of abstraction among these four. Edwina at least
>believes 2 and 1 are the same or close and "logic" covers both. Whatever
>you choose is a religious decision, but there are tens of thousands of
>places where you can recount what people said about the the relationships
>among 1, 2 and 3. I suppose it has pedagogical value, but that Christmas
>tree is pretty well decorated.
>
>I interpret the FIS to agenda to be concerned with domain 4 and
>relationships with the others. There are very few forums for this, so I
>suppose the focus here to be worth preserving.

I think that it is correct that the FIS agenda is primarily in domain 4. In
order to carry out this task in more than an intuitive way, we need to have
a common language to relate domains 1 and 2, and for putting the
limitations of 3 into perspective. It seems to me the only possibility is
logic, which I tend to equate with information theory (at least in terms of
formal syntax and semantics -- the pragmatics and rhetoric are different,
on the surface, anyway).

Thus I agree with Edwina that logic applies to domain 1. If it doesn't then
we are in trouble. This is possible, I would say (my colleague Cliff Hooker
has joked about the physicist who has a revelation and reports "I have seen
the truth and it doesn't make sense"). So we have to make a defeasible
presumption that there is a discoverable order in the world that is
isomorphic in some non-trivial way to some logical order. The limits of
this methodology are empirical, but I think they are the limits of
knowledge as opposed to opinion in the form of hocus pocus and good luck.
Anything that orders the domain of 1 would, ipso facto, be a logic. If it
only appears to order (applies only to domain 2), it is still a logic, but
perhaps no the right one.

Hertz proposed that a theory tries to get a congruence of its logical order
and the causal order of its object. This congruence makes the causal order
of the object a logic in the semantic/syntax sense. This ignores the
pragmatic and rhetorical aspects of logic, which is in tune with the
treatment of logic in contemporary mathematics. These aspects can severely
restrict our ability to think with a logic, but it does not rule out the
world being logical. See also Rosen. More anon.

Complex systems are more or less by definition not completely analyzable (I
have a four-fold account of complete analyzability, which includes
definability, empirical diagnosability, predictability and reducibility --
the four are separate, not as the Logical empiricists thought different
versions of the same thing). This is also true of arithmetic, so it is
debatable whether the Logicist program was successful -- Gila Sher argues
that this depends on ones view of logic -- if it is just the constructive
part of formal science then no, but if the non-constructive part is
included, then pretty much yes -- I prefer the latter, but it is a matter
of taste (or religion, if you want). On the former (nominalist) view there
can be multiple logics that are not unified; on the latter account
(realist) they are all part of one logic (see, for example Zinove'ev for an
argument that all alternative logics are really a part of one logic). On
the logicist/realist view there can be logics that are not fully analyzable
(basically, mechanical in one traditional sense of the term, equivalent to
Rosen's usage). So the world can be logical, and in a sense explicable,
even if it does not satisfy certain pragmatic and rhetorical desiderata,
such as empirical decidability (testability) and analytic provability
(again, the Positivists were wrong here -- they assumed the wrong logic). I
think that the nominalist tendency is a hangover of the early assumptions
that something like Hilbert's Formalism is true (I include Conceptualism
and various forms of Psycholigism with respect to logic in this class).
This is not now sustainable. Even uninterpreted formal systems like Post
canonical systems can have theorems that are not provable. There is no
purely constructive formal system, just ones we construct, and the limits
of our constructions. To say that these are the limits of order in the
world is a form of hubris, I think. We cannot say with certainty what
formal systems on the complexity of arithmetic and set theory or higher
apply to. We must give up (the hope of) certainty and embrace fallibility.
Peirce was very prescient in recognizing this is the proper way
(pragmatically justified way) to embrace pragmatism.

I also agree with Edwina that the domain of 2 and 1 are not distinct. 2 is
contained in 1. We should not forget that out theories are a part of the
world as well. This also goes for 3. Our logics are thus at best theories
of an object, confusingly also called 'logic'. Our theories of logic do not
exhaust logic (and cannot fully).

So I think that domain 4 is:
         - the discovery of logic (Peirce again would agree -- for him
semeiotics and logic were one).
         - the development of a suitable pragmatics for logic (I think this
is semiotics -- analytic philosophers have dispensed with this, thinking
they can formalize pragmatics -- e.g., Montague. This has been proven to be
impossible, unless we follow Hilary Putnam and argue that there is no
domain 1, just an idealized version of domain 2 -- Putnam cites Peirce, but
he uses a verificationist criterion of meaning that is foreign to Peirce,
and leads to a substantive notion of truth rather than Peirce's
methodological/regulative principle of truth.)
         - the application of suitable rhetorics to disarm dis-integrative
approaches to information and logic, including the forcing of
methodological fallibilism (not to be confused with scepticism, even of
Hume's weakish academic -- for a modern version of the latter, see David
Lewis's paper "Lost Knowledge" -- I think it is a reductio ad absurdum of
verificationist notions of truth).
         - the integration of the above three in an open ended project (see
next paragraph).

Finally, domain 4 is embedded in domain 3 which is embedded in domain 1,
but our senses of 1, 2, and 3 depend on and presuppose 4 as we have it,
either tacitly or explicitly (usually a mixture). This looks like a
hopeless entanglement. Entangled yes, but hopeless no. I think we must
resist the tendency to reduce one the other domains to one of them, or we
will end up with a circularity. Likewise we should not seal the domains
from each other, or they become self-justifying and dis-integrated. So, we
need to keep our systems open, and recognize that all systems in the world
are also open. Just as we can be under committed and miss the truth, we can
be over committed and miss the truth. There is no final closure to the
project. This is not a reason to think that there can be no progress, which
stems first of all from anomalies among the four domains. Empirical facts
can impinge on mathematical theory, and vice versa (to think otherwise is
to buy into the analytic-synthetic distinction, which Quine effectively
trashed, along with Hanson, Feyerabend and perhaps Kuhn). In order to avoid
getting caught in vicious circles, one has to adjust ones pragmatics to
leave interpretations open ended, and ones rhetoric to dismiss categorical
(absolute, non-context-dependent) statements of dependency and closure.

Just some thoughts. See my doctoral dissertation for details.

John

----------
Dr John Collier john.collier@kla.univie.ac.at
Konrad Lorenz Institute for Evolution and Cognition Research
Adolf Lorenz Gasse 2 +432-242-32390-19
A-3422 Altenberg Austria Fax: 242-32390-4
http://www.kli.ac.at/research.html?personal/collier
Received on Fri Sep 13 19:55:13 2002