[Top] [All Lists]

Re: [ontolog-forum] A "common basis"

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: bateman@xxxxxxxxxxxxx
Date: Wed, 02 May 2007 21:09:14 +0200
Message-id: <20070502210914.cnj5l7cqo4wgsgog@xxxxxxxxxxxxxxxxxxxxx>
Quoting Adam Pease <adampease@xxxxxxxxxxxxx>:    (01)

> ...  While some folks want to keep claiming a common
> ontology is somehow impossible, some of us having been doing the
> "impossible" for years.    (02)

There is an interesting shift in appraisal when one moves the
"doing the impossible" to "spending one's time attempting
the impossible", which may be a more accurate semantics
for the "doing" here I think. The former is a wondrous feat, the latter is
possibly a waste of time. We can let which it turns out to be, as you
suggest, remain an empirical issue to avoid a lot of ideological keystrokes.    (03)

> If it's really impossible, it should be easy to
> give a concrete, fully axiomatized example of irresolvable
> incompatibility (not English text or hand-waving, but a real formalized
> example).  No one has yet done so.
>    (04)

We are working on this. I don't think it is easy to come up with because
getting clear what a "fully axiomatized example etc." is not in itself
a simple matter. You suggested yourself, Adam, that finding relations
between different ontological categories may be impossibly difficult:
remember the discussion of SUMO:Process, DOLCE:Perdurant, PSL:Process
last year? (or was that another list...?). When the ontologies being
compared have no common starting point, proving incompatibility is
not at all easy (unfortunately, proving compatability is not so easy
either!). But, as I said, we are working on this because our framework
does allow heterogenous and distinct theories to co-exist: we then try  
and prove
(formally, not English text of hand-waving, but with theorem provers)
relationships between those theories. We are considering methods for
proving inconsistency that work for more than toy examples, but there
is still some significant groundwork to get done. All of that groundwork
requires richer notions of structure and modularities in ontologies than
is commonly found in current offerings.    (05)

My own take on this (but it will be an empirical matter :-) is that we
won't *need* 'a common basis' when we get this right.
I spend a lot of my time teaching linguistics
students not to take what they read in dictionaries too seriously, given
that they are inherently out of date (by the time they are printed)  
and have foundational compromises that make them less than revealing  
about many of the more useful words. So I'm afraid Pat C's sign of  
hope doesn't really
compel either.    (06)

I think it may be possible to pick a reasonably broad basis and then  
go about defining everything one can think of in its terms: certainly  
gives one plenty to do. Whether one believes in the value of this as  
an activity is another matter. But the issue will not be settled  
"easily" I fear and to suggest otherwise preempts some rather more  
solid results in the area.    (07)

John B.    (08)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (09)

<Prev in Thread] Current Thread [Next in Thread>