"John J. Kineman" wrote:
> On Intelligent Design:
>
> The real issue involves the underlying world view. Science takes place
> within a world view and cannot comment on its validity, except for its
> value in developing useful explanatory and predictive models. A theological
> world view precludes models by summing all processes to God. It is not
> science because it precludes models (by design). The basic concept of
> intelligence and information in nature comes about from the scientific need
> that other approaches cannot meet. It is only its religious/theological use
> that is not science.
>
> Here are two URL's for the Intelligent Design discussion, ...
> Perhaps the religiously motivated folks would be less successful if the
> scientists would flesh out the ideas better rather than putting them off as
> irrelevant when they are obviously not.
OK. The last quoted phrase in the e-mail of John motivated me definitely to
present to the list the following note on logical models for the design
process. I reject the distinction between "intelligent causes" and "natural
causes" which simply mean that "natural causes are stupid" and "intelligent
causes are non-natural". (The obtuseness of such a distinction can be seen, in
my opinion, with the example of the "stupid" natural cause known as
gravitation, which is a key agent for the existence of radiating stars and in
particular our sun - which with olympic indifference goes on supplying
low-entropy energy to evolutionists and creationists... ;-).
The note has as objective to argue that from a logical standpoint, design may
be seen as implicit in the operation of an agent endowed with a logical system.
The agent is indeed teleological, but this happens because of its structure and
processes, not for acting as an "un-natural cause".
A NOTE ON A LOGICAL MODEL FOR DESIGN
Lets take an agent A possessing
- a language L with an associated semantic relation RS which assigns:
- (eventually ambigous) meanings to the object signs of L as objects O
in the world W of A;
- (eventually ambiguous) meanings of the phrases in L as subsets of
relations R in W;
AND their inverses:
- object signs of L for Os in W;
- phrases of L for subsets of Rs in W.
- a logical system S over L, with
- a set of axioms Ax, ie, phrases selected from L
- a deductive process which allows:
- the generation of phrases of L (theorems) from Ax according to a
truth preserving criterion TC and its inverse
- the demonstration of a given phrase P of L as being or not being
a theorem entailed by Ax, under TC. (Note that one is not requiring that the
agent can demonstrate not-P from Ax. One is just requiring that the agent
recognizes the impossibility of deducing P from Ax. Technically, one is
rejecting CWA - the closed world asssumption.)
Now let us assume also that A can formulate in L the properties some inexistent
object O' in W should possess, as a set D of phrases in L, ie, effectively
describing the object by its potential belonging to relations R in W.
Then, if the agent is able to deduce (demonstrate) D from A, then one can say
that A found a design process for the object O' indicated by D, as the
demonstration record can be seen as a linguistic plan for making O existent in
W.
This is the simple approach to model design in logical terms. It can be argued
that because the process is purely deductive, the agent cannot design anything
which is not implicit in Ax. Something really new, ie, something not an
instance of some general D designable in Ax, cannot be designed.
So we go for abduction, and assume that A, besides being able of deductive
reasoning is also able of abductive reasoning.
In abductive reasoning and facing the evidence of being impossible to deduce
some D from Ax, the agent is free to enlarge Ax to Ax+ with other phrases of L,
restricted to the condition that Ax+ should be consistent according to the
truth preserving criterion TC.
If from Ax+ it is possible to deduce D, then an (ingenious!, creative!) process
design for O' has been achieved, with a linguistic plan composed by the
demonstration record of D in Ax+ and the enlarging of Ax to Ax+.
"Natural causes" required for such an agent to exist (an agent able of design)
are not so un-natural as eventually it may appear. They are:
1) - a language L and its associated semantic relation RS
2) - a logical system S over L which requires:
- a selection process for Ax from L
- a theorem proving process whith deductive and abductive modes of
operation
- a truth preserving criterion
3) - autonomy of the agent in putting in operation its S system.
For 1) I would reference Luis Rocha lastly announced paper.
For 2) I would reference Emil Post model of computation and the subsequent work
of Malcev (not the one of Algebra) which shows that logical reasoning amounts
to operations of pattern matching and
I would reference also the feedback principle for the TC
For 3) I would refer that globally speaking the operation of the S system, can
be seen as a relaxation, under inicial excitation D (or any other phrase of L).
So, I would finally argue:
- that per the above, design is not miraculous, and need not even be
"intelligent" (Im beginning to suspect that 'intelligence' is an oxymoron, a
kind of self-vanity invented by humans to self-justify the destruction of other
species... ;-)
- that abductive designs effectively configure an evolutionary,
self-developmental path, constrained (or guided) by the TC, ie, feedabck
processes between agents and their environment
- that such an agent A is very general, as restrictions in the physical
dimensions of the closure defining it are not necessary.
I would thank any feedback on this note.
Greetings
Paulo Garrido
========================================
Posting to pcp-discuss@lanl.gov from Paulo Garrido <Paulo.Garrido@dei.uminho.pt>
This archive was generated by hypermail 2b29 : Fri Jun 09 2000 - 15:32:32 BST