Subject + Object (Reflective Systems?)
Yuriy Guskov
ygheimth@adm.univd.kharkov.ua
Tue, 21 Sep 1999 16:32:20 +0300
A reflective system is impossible without the subject.
Reflection means at least comparison with previously
gained knowledge about something. Moreover, different
relations between the subject and the object should be
present in any approach related to information.
I begin this discussion in previous messages "Highlevel +
Lowlevel" and "Hyperprogramming". And it is not
occasionaly. The background of Uniform Abstract
Language (UA) (which I try to promote) is the very
relation between the subject and the object. Generally
saying, it is very universal principle. Because, it
embraces all. Usually we consider, say, a star or a tree
as the object of researching. But if we want to take into
account either we should consider them as the object.
Or if you want, we can call a machine "black box" then
the object is "that gets in" and the subject is "that
results in". Or, in other words we will deal with a pair
"components-result".
Further, if we want to build really universal system we
should not care about integers, floats, strings, or
something else. We should just assume we have some
pieces of information and we can associate them in
some way. Kinds of association can be represented by
mere known operations like - + * / . But there is a
distinction between these operations in math (in which
we have almost always something in a result) and in
UA (in which we can have something as a result of an
association with the help of one of these operations
or nothing). The next step is defining different contexts
for elements of information. And again we base only
on relations between the subject and the object to
obtain 4 kinds of contexts (quantitative, qualitative,
relational, and modal). These contexts has basic
elements they based on. They are correspondingly
values, names, types, and modes.
Maybe it is too abstract for many ones. But interesting
is that I got the standard model of programming
language basing on these principles. And, moreover,
we could not walk in the dark any more. But we did. I
say it because it is the fact that names, and types are
present in programming NOT originally but appeared
after. Now, using the principle "the subject-the object"
we can even foresee what we will need in the future.
At least I have already tried to do this... :)
http://www.geocities.com/ResearchTriangle/Facility/2320/ua/