[Sigia-l] Question about the use of the term 'ontology'

[Ruth Kaufman]

On Apr 1, 2005, at 1:57 AM, Lars Marius Garshol wrote:
>
> "Reification" was indeed the terminology that was used for the
> topic-subject relationship, but in the new standard (after the TAO was
> written) it's been abandoned in favour of "representation".
>
> So a topic is a symbol that represents a subject.
>
Hmm, I liked that the term "reification" was used. That made sense to 
me, although I'm sure not that many people know what reification is, 
and I am fairly new to the concept, myself. As I understand it, 
reification and representation are not mutually exclusive. Something 
may be reified by being represented by something else. It is reified 
*as* it is represented. I would say that reification is a process that 
the "subject" undergoes, whereas representation is a task of the newly 
formed sign, in this example, the Topic. I'm not sure if this tangent 
will prove useful, so that's where I'll leave it.

[...]
> | In other words, I don't see a topic as a type of subject, but as
> | something that a subject may become.
>
> If in the sentence above you use "subject" in the library science
> sense that's absolutely right. In topic map terminology topics and
> subjects are 100% disjoint.
>
Okay. In this sense, the subjects are only represented by topics and 
not reified as topics :-). If they were truly reified, then there 
wouldn't be 100% disjoint, I don't think. (Off on that tangent again, 
but this is making logical sense now, even if it seems less elegant.)

[...]
> | By the same token, I'd think that subject-based classifications
> | reify ontologies.
>
> I'm not sure I know what you mean by "reify" here.
>
What I mean by 'reify' is the process by which something whose natural 
state is not concrete (e.g., a notion, a cultural generalization, a 
collective memory, a value, a sense of self, a way of being...) 
manifests some aspect of itself, or is somehow denoted, in a concrete 
way (e.g., a term, a picture, a convention...)


> | Maybe what we need is a way to express the way these things become
> | each other -- a sort of state change diagram perhaps?
>
> I tried with a taxonomy, but since that didn't work the next step is
> perhaps a topic map. :-)
>
Perhaps, if "becomes" is an association type. However this implies an 
agent (how does one thing become another?) It appears that topic maps, 
as well as other subject-based classifications, are 1-dimensional with 
respect to processes and sequence. What I think is missing from a lot 
of our representation conventions is a sense of transformation. This 
could be another thread altogether.


Cheers,
Ruth