Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: SUO: Re: Focus and Volume




Adam, Pat, et al.,

AP> It would be incorrect to place me at the opposite end of the spectrum
> from Doug.  I'd summarize his position as that an upper ontology is
> essential and relevant but that there are many "good enough" upper
> ontologies that one might construct and all would be equally useful.  The
> point is to pick one (or generate one) and move on, as agonizing over the
> "one true ontology" is pointless.  I agree with Doug whole-heartedly.

There is certainly a wide range of opinions, but I don't think that
a linear spectrum is adequate for representing them.  I'll describe the
positions of a few of the darts on the board:

 1. Doug L. doesn't think the top level is very important, and I agree
    with him:  his top level isn't important.  But that is merely
    because his top-level categories are so weakly specified that
    there is very little to inherit from them.

 2. I have a sample top level in my KR book, which is also weakly
    axiomatized.  However, that is a temporary situation that I hope
    to remedy with future R & D.  I believe that a well designed
    top level should contain enough significant axioms that it would
    reduce the complexity and clarify the relationships among the vast
    numbers of lower-level categories.

 3. Some of the other people who contribute to this list also have
    some strong ideas about what belongs at the top level, but we have
    not yet reached a consensus about those ideas.  Until we do, I
    believe that it is premature to endorse any proposed top level
    as an agreed upon working document.

 4. My concerns about the current merged ontology can be deduced from
    my points #1, #2, and #3 above:  there isn't enough inheritance
    in it to make it any more relevant (and perhaps less so) than
    Lenat's; although I am pleased that Ian Niles has included my
    top level in the merged ontology, I still believe that more work
    is needed to relate it to the lower levels (I do some of that in
    my KR book, and I am working on more, especially in my causality
    boook (forthcoming)); and I do believe that we need more discussion
    about fundamentals, about which there are still major discrepancies.

I disagree that we can just "pick one" as Adam has suggested.  As long
as the top level is weakly integrated with the rest, as it is in Lenat's
ontology, it doesn't matter very much.  But if we had a good one, it
would be much more strongly integrated.

Bottom line:  There may be more than one "good" top level, but there are
surely an infinite number of "very bad" top levels.  If we just pick one,
the odds of picking a good one are against us.

John Sowa