[om-list] Re: What is Real?

[Mark Butler]

Luke Call wrote: 

> However, I've been realizing more that we do need some kind of logical statements for a few reasons. First, they represent a sort of query-cache of knowledge, which in turn might be verifiable by querying against the data in the object storage (observations) area. It would clearly take too long to determine on the fly that humans usually have one head and four limbs, or to determine things like "all ___ are ____"...


You have hit upn a key issue, which is what is considered a legitimate object
in the system, i.e. what do you consider to be real?  So far, you appear to be
willing to only consider the measurable attributes of concrete objects to be
real.  This is generally known as philosophical materialism.  The oppositive
end of the spectrum is philosophical idealism, which emphasizes the reality of
ideas above the reality of the physical world.

Now the first problem with philosophical materialism for truth modeling is
that you can make a nearly perfect argument that all classes of objects are
ideals, i.e. abstractions that only exist in minds and have no independent
existence.  Take "chair" for example - everyone knows what a chair is and what
a chair generally looks like, but while there may be objects that we assign to
the class "chair", the actual class has no independent reality.  The Greeks
started arguing about which was more important, the instance or the class,
several millenia ago.

So to even have any classes in the system you have left the domain of
incontravertable physical reality and entered the world of generalities.  And
as is generally the case, "all generalizations are false", and as everyone
knows, "the exception proves the rule".

However, in another sense, all ideals and beliefs, though usually of only
partial absolute validity, can be considered completely REAL in the sense that
someone believes them and uses those principles to guide their actions. 

We have talked much about the field of knowledge representation.  That begs
the question "Whose knowledge are we representing?"  We can put in our own
personal knowledge of physical reality and end up with a very dry database. 
To me, one of the most fascinating parts of my knowledge isn't what I know
about external reality, but rather what I know about what other people know /
believe about the real world, i.e. meta-knowledge rather than knowledge per
se.

Both knowledge and meta-knowledge are imperfect, inexact, and prone to error. 
The reason why it is so much better than an enumeration of physical objects
and their properties is that a rule, even a very rough and imperfect rule,
condenses an enormous amount of useful information into a small number of
bits.  On the other hand, mere mortals, let alone computers, are hardly
capable of remembering innumerable events and making new conclusions from the
source data for every possible situation.

I believe the key is to model all ideas, no matter what the source, as a
contigent part of reality that are believed by someone at worst and and nearly
divine at best.  Then as long as the sources and contexts are carefully
accounted for, one can actually derive results of a reasonable degree of
accuracy suing a fraction of the time and effort required to rederive every
result from scratch.  But more importantly,  one can do better than just
enumerate facts, but rather refine and analyze thoughts, even if only to
discover where and why they are wrong.

Storing data about a set of concrete objects and properties and events is not
nearly sufficient to model the laws of nature or any divine law for that
matter.  The nature of a law is a logical statement not a collection of
observations.  A logical statement is an idea.  For a logical statement like a
law to be REAL, it has to be enforced.  The laws of men are only enforced
haphazardly, but are sufficient to organize societies of millions of
individuals.  Divine laws, like the laws of nature, are much harder to escape.

Even entities as incontravertably REAL as the laws of nature are thoughts not
things.  Does the law of gravity have an independent existence outside the
mind of God?  In my view*, it does not.  I cannot conceive of a complete model
of even physical reality that does not admit certain types of thoughts as
first class citizens.

I believe that is far more important to model thoughts, even complete false
and misleading thoughts in their proper context, than it is to model
"things".  Modelling things and the properties thereof are only important
insofar as they affect thoughts.  Does it really matter what the physical
world is like if there is no one around to think or to feel about it?

So while I admit that a comprehensive database of objects and events may be
very interesting from a scientific point of view, I think that it barely
scratches the surface of the things that are worth knowing, storing, and
recording.  That is why I favor meta-models capable of representing the full
gamut of human experience rather than just the single shade of scientific
reductionism.

No doubt, thoughts are very hard to model.  Even the Cyc people are primarily
concerned with officially distilled & certified principles of common knowledge
rather than comparative analyses of competing bodies of thought.  However, to
me the way to truth is much more likely to involve an inspired form of the
latter than attempting to turn a computer into a inductive pattern recognition
machine.  That is why I would much rather work on a system that focuses on
modeling thoughts rather than things.

 - Mark


* See D&C 88:13.