| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
(Documentation connects "The relationship between three things, when one of the three things connects the other two. More formally, (&%connects ?OBJ1 ?OBJ2 ?OBJ3) means that (&%connected ?OBJ1 ?OBJ2) and (&%connected ?OBJ1 ?OBJ3) and not (&%connected ?OBJ2 ?OBJ3).")
Why, if OBJ1 connects OBJ2 and OBJ3, should OBJ2 and OBJ3 not be connected? (since OBJ1 connects them, isn't that contradictory?)
Does the following represent one relation between three things? :
OBJ2 ----- OBJ1 ---- OBJ3
I see two relations.
JM
Jon Awbrey wrote:
¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤One of the biggest threats to the viability of this project
continues to be people who propose to standardize without
recognizing the standard languages and terminologies
that are already in place.There is a way that everybody can have their own favorite usage
and still relate it to the usages of other interpreters, and that
would be to open up a sign-relational context for inter-com, but
so far that possibility is being ignored and so we have to go back
and do it the old way, which is start up a knock-down drag-out fight
until one usage or the other emerges victorious. I am sorry it has
to be that way, but it's not me who picked the method. I am here
to inform you that the battle is already lost over the usage of
words like "class", "collection", "set", and so on, and so you
might as well just get used to it.You basically only get two choices:
1. You can say that you are using the words "class", "collection", "set"
in their everyday, informal, ordinary senses -- in which case these
three words are now and always will be synonyms.2. You say that you are using the words with the technical senses that
they are commonly given in logic and mathematics -- in which case
there is a distinction made between "class" and "set", while the
word "collection" is now and ever will be a floater between them.Jon Awbrey
¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤
West, Matthew R SITI-ITPSIE wrote:
>
> Dear Ian, David, and Pat,
>
> Pat was asking for an example of when a 4D approach
> can be helpful. This might be such a case.
>
> See below.
>
> Matthew West
> Principal Consultant
> Shell Information Technology International Limited
> Shell Centre, London SE1 7NA, United Kingdom
>
> Tel: +44 20 7934 4490 Other Tel: +44 7796 336538
> Email: matthew.r.west@is.shell.com
> Internet: http://www.shell.com
>
> > > I was looking at the SUMO browser at:
> > >
> > > http://128.136.11.33:8080/rsigma/SKB.jsp?req=SA&skb=Merge&id=260
> > >
> > > where the axiom:
> > > Formula
> > > (=>
> > > (instance ?COLL Collection)
> > >
> > > (exists
> > > (?OBJ)
> > > (member ?OBJ ?COLL) ) )
> > >
> > >
> > >
> > > Now, as I read this, if a collection exists, it MUST have a member.
> > > I understand that the SUMO semantics for member and Collection are:
> > >
> > > (documentation member "A specialized common sense notion of
> > > part for uniform
> > > parts of &%Collections. For example, each sheep in a flock of
> > > sheep would
> > > have the relationship of member to the flock.")
>
> MW: So member is supposed to be a relation like part, but relates
> members to classes (see below).
>
> MW: The problem with this is that member is non-transitive, but part
> is transitive. I therefore declare an inconsistency (member is both
> transitive and non-transitive).
>
> MW: Lets see if we can understand how it should be.
> > >
> > > (documentation Collection "Collections have &%members like
> > > &%Classes, but,
> > > unlike &%Classes, they have a position in space-time and
> > > &%members can be
> > > added and subtracted without thereby changing the identity of the
> > > &%Collection. Some examples are toolkits, football teams, and
> > > flocks of
> > > sheep.")
>
> MW: Well this sounds like Collections are spatio-temporal extents,
> since that's what it is to have a position in space time. Spatio-
> temporal extents are individuals (the terms are synonyms for me).
> Individuals are not classes.
> > >
> > > Now, it seems to me that this axiom requires a collection to have an
> > > instance.
> >
> > That's right, and this is as it should be, I think, since
> > collections are
> > supposed to be situated in space/time and this wouldn't be
> > possible for a
> > collection that had no members.
>
> MW: Well this is another indication that collections are individuals,
> and that we are talking abotu mereology not set theory. Mereology
> has many similarities to set theory, but there are a couple of key
> differences:
>
> a) Mereology has no equivalent of membership (whole-part is
> transitive like sub-type/supertype).
> b) Mereology has no zero (as your intuition above tells you).
>
> MW: I want to be clear here, there is a set that has as members the
> sheep that are parts of the flock, but this is an abstract object,
> and does not have a position in space and time (though of course
> each member does).
>
> MW: The secret is to understand which you are talking about.
> Apparently here Ian is talking about spatio-temporal extents, and
> not classes, despite the language he uses.
>
> no further comments.
>
> >
> > >
> > > Maybe I'm being obscure, but I would hope that this doesn't
> > > say that there
> > > are no such things as intentional collections in the SUMO ontology.
> >
> > I think you need to spell out what you mean by an intensional
> > collection.
> > Note that, in the usage of analytic philosophers, the term is
> > spelled with
> > an "s" when it's opposed to "extensional".
> >
> > >
> > > I don't know how many left-handed leprechauns that play
> > > trombone exist, but
> > > I would think that I should be able to express in SUMO the
> > > collection of them
> > > without wandering around looking for a pot of gold to give
> > > the guardian a
> > > survey (to see if he/she is one) first.
> >
> > Well, we can, within the framework of the SUMO, create the *class* of
> > left-handed leprechaus that play trombone, since this is a meaningful
> > concept. Why do you think we should also be able to create a
> > collection
> > corresponding to this class? If you think this should be
> > possible, you need
> > to spell out what your notion of *collection* is and you need
> > to show how it
> > is distinguished from the SUMO concepts of class and collection.
> >
> > >
> > > David Whitten (713) 791-1414 ext 6116¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤
-- Jean-Marc Orliaguet ( jmo@medialab.chalmers.se ) - http://www.medialab.chalmers.se/people/jmo/ - Tel: +46 31 772 8581