[sc34wg3] [topicmapmail] the praeteritio explanation
patrick at durusau.net
Wed Jan 13 09:02:28 EST 2010
I think the more important question for topic maps is not where we may
or may not have been but where we want to be going.
That requires answering new questions, such as:
Do we need for topics to reify associations? Or occurrences for that
matter? In light of our *present* experience with topic maps. Which I
assume is different from our experience almost ten years ago.
Why not make associations and occurrences (which are a form of
association) first class representatives of subjects? And so capable of
supporting properties that now require reification with a topic.
Or, why not add to topics (and associations and occurrences as modified)
the ability to have typed properties? Or in other words, allow
"occurrences" to return to the indexing use of the term "occurrences."
I am sure there are a number of other suggestions that merit discussion.
My only request would be that such discussions be mirrored on the
sc34wg3 at isotopicmaps.org mailing list so that members of NBs could be
aware of the ongoing discussions.
Hope you are having a great week!
Steve Pepper wrote:
> Hi Steve,
> Happy New Year...
> * Steve Newcomb
> | With all that in mind, I now ask my question about the design of TMDM in
> | a new way: Is an unreified association in a TMDM topic map an example of
> | praeteritio? I.e., is the purpose of TMDM's design to support the
> | possibility of talking about something while pretending not to be
> | talking about it?
> No, it has nothing to do with any of that.
> | Or maybe there's a better explanation for the necessity of redundant
> | representation of reified associations in TMDM, and for the
> | underprivileged status of association items, which represent things that
> | can't be talked about?
> As Lars Marius pointed out, the "explanation" is to be found in ISO 13250:2000.
> There you will find things called 'topics' that represent subjects, and things
> called 'associations' that represent relationships between subjects but are not
> themselves topics. The same model is found in the XTM specification from 2001.
> That model dates back at least as far as the mid-90s, when it consisted of three
> architectural forms called "semantic assignment" (= topic), "semantic title"
> (= topic name), and "topic relation" (= association).
> The sole "purpose of TMDM's design" was to capture in a formal model what in
> ISO 13250:2000 and XTM 1.0 was expressed less rigorously in syntax. This was
> something that needed to be done in order to make Topic Maps interoperable across
> multiple implementations.
> To uncover the original thinking behind the "underprivileged status of
> association[s]" you have to go back way before the TMDM and ask people like Michel
> Biezunski and Steve Newcomb :)
> Best regards,
> PSI: http://psi.ontopedia.net/Steve_Pepper
> Blog: http://topicmaps.wordpress.com
> topicmapmail mailing list
> topicmapmail at infoloom.com
patrick at durusau.net
Chair, V1 - US TAG to JTC 1/SC 34
Convener, JTC 1/SC 34/WG 3 (Topic Maps)
Editor, OpenDocument Format TC (OASIS), Project Editor ISO/IEC 26300
Co-Editor, ISO/IEC 13250-1, 13250-5 (Topic Maps)
More information about the sc34wg3