Posted by Mark Meckler on July 20, 1997 at 16:00:40:

In Reply to: The issue of sharing knowledge posted by Umesh on July 17, 1997 at 05:36:47:

One of the dimensions of Knowledge one of the discrete constructs, is 'context.'
There are other members of the set of course, discovered and undiscovered
constructs of unknown quantities. Context is one of the ones we have
discovered, and my re search tries to figure out how context works.

One of the reasons to believe that context is important is that it crops up
every time we talk about how we don't fully understand something or other.
When we admit that "it depends." What it depends upon, this uncertainty,
is the context.

So we look very carefully at the context, to become more certain of the
dynamics of the issues. How much, I wonder, does certainty and internal
consistency change when context changes. Is there a criti cal mass at which
certainty and consistency become stable across applications? I am busy
finding out.

The use of this is tremendous effectiveness and efficiency of our knowledge management systems.
If this critical mass of knowledge can be kno wn, it can be stored, sorted,
retrieved, transmitted in "packets" as objects, with an minimized average
level of misunderstandings across usages. The only major limitation is
this recourse to "understanding," a phenomenon which is not fully
u nderstood.

Understanding, in knowledge management at least occurs when an idea can be
put, exactly defined, into language. It is both a verb and a noun. The
when we use it in the verb form, we discuss the 'process of understanding.'
When we use it as an adjective, it describes having similar cognitive
state, as in the hoped for shared understanding when one asks "Do you
understand?"

We sometimes also use it a noun. But when we do, I believe we are
mistakenly confusing in with kn owledge. When we say "my understanding" we
could better say, "according to my knowledge."

Understanding only gets in the way when we need to share knowledge. We want
to find out if the other understands, but the same knowledge can lead to
diff erent understandings. People make sense of things with the help of
past individual experiences, attitudes, beliefs, etceteras. Again, context
is the limiting issue.

To find out if knowledge is true we have to appeal to someone else. One
coul d always ask oneself "perhaps I am crazy, perhaps it's just me and not
the world?" Because of this we have to try to establish the degree of
"shared understanding."

Shared understanding implies a like state of mind. This is so far,
impossibl e to test empirically, and that, I guess is the edge of science in
that region of the net. Quine goes so far as to say "we can never"
directly compare the content and 'feel' of two minds. Empirical agreement
in the form of acquiescence, is all we have to go on. Inside is a black
box, more or less, for Quine.

On the other hand, he and we all should agree that there is a certain
amount of shared structure around our understanding, and a good amount of
detail to the sub-structures, so that discovered laws of probability
especially the central limit theorem, begin to take effect.

It is a bifurcation error to eschew that because some laws of logic do not
hold, it should not be used to describe the 'real world.' Use of logic to
create 'rational' arguments has turned out to be the most successful
survival tool within the known world. No area of study or theory, NONE,
has relationships which all hold at all times. There is always an
inconsistency that has to be ignor ed or assumed away. But that does not
mean they are not nearly perfect or very useful.

We implicitly, 'tacitly' if you will, use what is called probability
because we can't help but *impute*. All day as we act, we keep relying
upon induction from past events. We even think of that as a process of
learning and developing habitual actions based upon what worked before.

When a theory has worked before, we will tend to turn to it again. When a
theory work again, we may start believing, t hat the same chain of
circumstances will be seen again if similar other things were input at a
different time or in a different place.

As scientists, Through contemplation and observation, we think we discover
relationships between events or cl asses of events, we put them in
linguistic form, and name these 'discoveries' "theories." We then test
them by trying to see if they work again, or somewhere else. We try enough
times to be able to say how reliable the theory is, i.e. how often it has
worked, and how often it will probably work in other places, at other
times.

We never arrive at truth. We are always to some extent uncertain. We have
all taken the oath to admit possible corrections. But do we never have
knowle dge?

Yes, as long as the word is purged of any idea of absoluteness. That idea
is perhaps well replaced by probability. We now have knowledge, with the
caveat that although it may be true now, it might be false at any future
moment. Tomorrow or in a hundred years, it will it turn out false. But
probability also tells us how likely it is to stay in it's current state
with the empirical world.

that if other things stay more or less the same in the net, it is likely
that any given theory remains in its truth state.

When you spread something out evenly, no matter how hard you try, you can
never get it perfectly even. Even at the extreme, when we cut a diamond,
so smooth, we see uncountable imperfections at the mico level. Moreover at
the sub-mico level, (called subatomic), we find down right random behavior.
Does randomness at the sub-micro level mean we can't predict the future of
macor events by appealing to observations on other levels? Thank to
recent di scoveries and tests in non-linear dynamics, we now know that we
can predict future events.

What was this discovery? We have found that despite micro level
randomness, distinct macro order often becomes apparent after repeated
observations. "Path dependencies," I guess they have been named, form
predictable structures. Perhaps there is some similarity here to the
central limit theorem.

Some, let us call them loosely "Kantians" for short, might argue that the
order we observe my n ot be in the external constructs, but applied, like
import classifications, by the human mind. We cannot help but classify and
relate things, a Kantian might argue, it is part of being human and alive.

David Hume provided W.V. Quine with the ans wer. "OK, but so what?" It is
now *what we believe* is the real world, instead of simply 'the real
world.' It makes no practical difference. We now separate science from
religion, develop a science of probability, likelihood and prediction th at
has allowed us a better situation on earth. Life is simply not as hard,
and survival of the average individual human is more likely than ever.

Can knowledge be false? Yes. Can knowledge be true? Yes. Does it matter?
No. What does matte r is the question 'can the same knowledge be both
truth and false at the same time in a working system?' By digging into the
dynamics of context, that is what I am trying to find out.

Mark Meckler                  | "For the perfection of

Doctoral Candidate            | things is to be judged

Department of Management      | by their nature and power

and International Business    | alone; nor are they more or

Florida Atlantic University   | less perfect because they

777 Glades Road               | delight or offend the human

Boca Raton FL 33431           | senses, or because they are

mmeckler@acc.fau.edu	      | beneficial or prejudicial to

(561) 367-2964                | human nature." Spinoza, Ethics

---

Follow Ups:

---

Click Here to Post Follow Up in New Forums

    Knowledge Management Think Tank (New)

  

 Subject: 

Message: