Re: From WWW to Super-Brain (new PCP node)

Onar Aam (onar@HSR.NO)
Wed, 11 Jan 1995 15:58:56 +0100


>I have some remarks and questions. I feel that the answer to the question
>wether any knowledge system is semantical or not, presuming you're able to
>define the concept of semantical knowledge, mainly depends on what your
>viewpoint to the system you're looking at is. Posing this kind of question
>is actually making the mistake of trying to fit reality into the wrong
>framework or set of distinctions.
>
>Psychology has also been messing with this problem. The discussion is/was
>wether human knowledge storage and retrieval is semantic or associative (or
>anything else). Most research suggests that there is considerable evidence
>for both positions: it just depends on how you investigate the matter, what
>the conceptual framework of your setup is.

Ok, let me put it this way. Knowledge is fractally clustered into many
perspectives. Each of these perspective have a _locally_ consistent structure.
But the range of the perspective is always limited and it is therefore
impossible to construct a perspective with a semantic structure that extends
infinitely. This means that different perspectives may not be bridged
semantically. The only way to bridge them is associatively via structural
similarity.

>Epidemics would probably affect the functioning of any system. Why would a
>semantic structure be more (or less) vulnerable?

It isn't. My point was that if our wish is for a semantic structure to emerge on
the web then we may wait hard and long for it. But often a locally semantic
network will do just fine and for many (most!?) practical purposes knowbots and
similar devices may be used unambiguously.

Onar.