Re: Roles, again

Gerard Ellis <>
From: Gerard Ellis <>
Message-id: <>
Subject: Re: Roles, again
Date: Wed, 13 Sep 1995 15:13:01 +1000 (EST)
In-reply-to: <> from "Don Dwiggins" at Sep 12, 95 10:52:50 pm
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 2136      
Precedence: bulk
Don Dwiggins writes:

> Actually, OO _programming_ languages don't express relations directly at
> all; one has to represent relations in terms of variables containing
> pointers, and suchlike tricks.  Some OOA&D notations do allow expressing
> n-ary relations, much like CGs do.  (Aside: interestingly, in such cases,
> relations wind up looking a lot like compound objects.  Or perhaps
> compound objects are just relations: could an auto be usefully thought
> of as just a relationship among its components?)

We have started prototyping 3D (and nD) conceptual graphs in OpenGL on a
SGI after realising that many of these views are exactly that. You can take
a semantic CG of a relation between states and turn it so that the
post-state is facing the viewer and you will see the OOCG. In
[Ellis95OOCGS] (below), I thought I was doing OO, but I realise now that
it was mainly a state based view, and that OO is really a specific view
of that (or at least OO statebased logical views). These 3D
objects are useful for showing the relationship between OO and relational
views. Recently John Sowa has been making contexts the primitives on
which relations are built, but I see HasA, and IsA etc as the primitive
relations (or -1stArg->, -2ndArg->) as being primitive and that contexts
can be seen as views and foldings of relational graphs in space. The
views are useful, but so is having the relations in hand.

    author = {Gerard Ellis},
    title = {Object Oriented Conceptual Graphs},
    editor = {G. Ellis and R.A. Levinson and W. Rich and J.F.Sowa},
    booktitle = {Proceedings of the 3rd International Conference on
    Conceptual Structures},
    address = {University of California at Santa Cruz},
    publisher = {Springer-Verlag},
    month = {August 14-17},
    series = {Lecture Notes in Artificial Intelligence},
    year = {1995}}

Regards, Gerard.
Gerard Ellis ph:61-3-660-5090 FAX:61-3-662-1617 Rm:10.9.11
WWW:    Computer Science Dept, Royal Melbourne 
Institute of Technology, GPO Box 2476V,  Melbourne, Victoria, 3001, AUSTRALIA