"Chris Wood" <standsongrace(a)hotmail.com> wrote in message
news:ch65tl$ose$1@sea.gmane.org...
What I'm arguing is, "is-a" is not going
to work, it's too POV. What we
have
now is sets of related articles, which is all we need (is-a relationships
*can* be represented this way, but are more flexible).
Is-a relations work just fine for noncontroversial categories. Quartz is a
type of mineral. Bill Clinton is (was) a U.S. President. Arnold
Schwartzennegger is a governor of California. Canada is a country in North
America. Then there are some areas that are a little more controversial, due
to judgements about the classification system used: should a particular
author be included in the category Fantasy writers, Science fiction writers,
or Horror writers, or all three? But I think most such questions could be
decided relatively amicably. (I realize that there are problems even with
such simple classes as well--e.g., the discussions about which counties
should be included in the Europe template--the line dividing Europe and Asia
is not self-evident at all points).
Where is gets much more difficult is where there are intense POV values
built into the categorization (especially where the defining criteria depend
on subjective determinations). Such as whether something is a work of
propaganda. Or whether a person is an alcoholic. Or if an incident was an
act of terrorism or of patriotism.
But I do not see how your distinction between related-to and is-a makes any
difference in such situations. I think the difficulties of NPOV
categorization are not that different from that of NPOV language in an
article in general. The language used to describe a person/thing/event
inevitably carries some subjective value judgements about the topic. The
criticism you make of categorization seems to be a variaton on a critique of
language in general. In your earlier post, you
I guess I simply do not understand how related-to is any better than or so
very different from is-a categorization. If you are suggesting that we
should have a very flat category structure with many more categories for
each article, then I very strongly disagree. Hierachichical categorization
is useful, despite the inherent difficulties with any categorization schema.
I do not see any significant attempts to make categorization in Wikipedia is
an endorsement of any single ontological worldview--there are mutliple
competing/complementary hierarchies. I think your assertion (in your earlier
post) that "Sets are much less POV" is simply wrong. A set is just as POV as
a hierarchy. Whether there is a list/set of alcoholics or of propaganda or
of acts of terrorism--inclusion in such a list/set is expressing a POV just
as much as if including within a hierarchical categorization schema.
Bkonrad