Does every concept have a set of attributes?
Lee
I'm working on finding attributes for a group of concepts. I
personally believe that it is hard, if not impossible, to find the
attributes for abstract concepts, such as ART. But my friend insists
that every concept should have a set of attributes. What do you think?
Thanks,
Lee
[ comp.ai is moderated. To submit, just post and be patient, or if ]
[ that fails mail your article to <com...@moderators.isc.org>, and ]
[ ask your news administrator to fix the problems with your system. ]
Jorn Barger
> I'm working on finding attributes for a group of concepts. I
> personally believe that it is hard, if not impossible, to find the
> attributes for abstract concepts, such as ART. But my friend insists
> that every concept should have a set of attributes. What do you think?
I think it's always possible to find _some_ attributes,
but if you need to fully define a complex concept, using
attributes won't be enough. But it's the complexity
that's the problem, not the abstractness-- you can create
a well-behaved hierarchy of abstractions, so long as
they're simple abstractions.
AI tries to fit everything into hierarchies where children
inherit attributes from parents, because that data-
topology is easy to work with. (Yahoo and eBay both
include 'art' in their navigational hierarchies without
major problems.)
Complex concepts cause problems because they represent
_relationships_ among several simpler concepts, so
they inherit attributes from all of these, and can't
be adequately classified under any single parent.
Art can be analysed as a complex relationship between
an artist with an idea using materials, and a viewer
who experiences the work. Each of these has
characteristic attributes-- artist, idea, materials,
viewer.
Philosophers have always wrestled with the problem of
distiguishing art from craft, or art from science, etc.
If you allow the artist and the viewer to have a basic
_motive_ to experience art, you can avoid having to
solve this philosophical problem, but I'm confident
that someday we'll be able to explain the art-motive
in terms of the nervous sytem's special pleasure in
deep new elegant patterns. (RG Collingwood did a
pretty good philosophical exploration of this back
in the 1930s, I think.)
More on complex data-topologies:
http://www.robotwisdom.com/ai/thicketfaq.html
--
Optimal info-density: Logarithmic timeline of the universe
http://www.robotwisdom.com/science/logarithmic.html
Randolph M. Jones
> Hi, all,
>
> I'm working on finding attributes for a group of concepts. I
> personally believe that it is hard, if not impossible, to find the
> attributes for abstract concepts, such as ART. But my friend insists
> that every concept should have a set of attributes. What do you think?
Consider the point of view that an attribute is really just a relation
to some other concept. Then, your friend's statements reduces to "every
concept has a set of relations to other concepts". That seems true for
art...to me, anyway. However, the less well defined a concept is (by
definition), the more difficult it will be to explicitly identify those
relations.
Jorn Barger
> > I'm working on finding attributes for a group of concepts. I
> > personally believe that it is hard, if not impossible, to find the
> > attributes for abstract concepts, such as ART. But my friend insists
> > that every concept should have a set of attributes. What do you think?
>
> Consider the point of view that an attribute is really just a relation
> to some other concept.
If you generalise your fundamental ontological distinctions
to this degree of abstraction, is anything useful left???
To make it programmable, you have to recognise that relations
between simple objects are different from relations between
relations. And a classic Aristotelian attribute like color
is probably more accurately described as a relation to a
constant (color 'FFFFFF') than to an object 'whiteness'.
> Then, your friend's statements reduces to "every
> concept has a set of relations to other concepts". That seems true for
> art...to me, anyway.
If so, please give examples of these concepts and relations.
I'd propose the relation 'HasUsualStory' to a series of
story-elements linked by 'FollowedBy' relations. Story
elements might include:
person's art-motive gratified by viewing artworks
person studies construction of artworks
person conceives new idea for artwork
person constructs artwork
person critiques and modifies artwork
person offers artwork for sale
etc