Re: Semantic ambiguity [was: Typing was Re: DBpedia as possible project]

Are,

On Fri, 2010-11-05 at 13:23 +0100, Are Gulbrandsen wrote:
> Hi
> 
> This is getting interesting!  :)
> 
> Alexander Johannesen questioning the basis of linked data and subject identification. Andrew S. Townley's thoughts on scalability, questioning if this can work outside a controlled environment.
> 
> Steve Pepper on 4. nov:
> > The categories of human knowledge are better expressed using a prototype model than a criterial-attribute (or "Aristotelian") model. In such a model, roles and types are not sharply differentiated, but rather exist on a continuum.
> 
> 
> Patrick Durusau on Semantic ambiguity:
> "Since we are trying to communicate with other people, there isn’t any escape from semantic ambiguity. Ever."
> 
> We're kind of back to the start. In a messy semi-structured world with information overflow, - what kind of technology can help us find a way?
> 

No surprise but probably no single technology if by that you mean one
way to identify all subjects and in the darkness bind them. ;-)

The point of legends in the TMRM is that different domains will have
different subject identifications practices.

You need only look at any bioinformatics book to realize that they
identify subjects quite differently from people in astroinformatics. 

Neither one is better than the other but they are *equally essential* in
both domains. 

True that it would be rare that you would need a mapping between the two
but you will need mappings from both into the published literature. 

Of course that is only an example that emphasizes the present. 

It is also the case that terms change meaning over time. 

For example, prior to the "decoding" of Egyptian hieroglyphics based on
the Rosetta Stone, there was a lot of activity "reading" those
inscriptions. And learned works quoted each other, etc. 

So if I were to write a topic map about Egyptian hieroglyphics today, it
would be *misleading* to apply our current interpretations to historical
texts that were relying on another interpretation of the same materials.

The same is true for the term "democracy," which has changed meaning
quite a bit over the last 200 or so years. 

Having a single URI identify that term and many others is *semantic
flatland* thinking and while acceptable to some, it isn't to me.

In defense of the *semantic flatland* approach, let me hasten to say
that it may well be adequate for Amazon, ITunes and pornography,
intellectual and otherwise, sites on the WWW. 

My objection is to any *one size fits all* model, which attempts to fit
a complex world into the rather pinched imagination of any designer.

I find the complexity of the world refreshing and try to discover
something new every day. 

Hope you are at the start of a great weekend!

Patrick

> In the meantime, - here's some friday edutainment: "Semantic ambiguity and how human communication fails - except by accident", - revisiting Wiio's laws:
> http://everythingisasubject.blogspot.com/2010/11/semantic-ambiguity-and-how-human.html
> 
> Laugh today, - it's a sunny friday afternoon (and sink into depression on monday when you realize that it's not only sarcastic humour)  :)
> 
> 
> Best Regards,
> Are D. Gulbrandsen
> The XML-group,
> Center for Information Technology Services
> University of Oslo
> _______________________________________________
> topicmapmail mailing list
> [email protected]
> http://www.infoloom.com/mailman/listinfo/topicmapmail
_______________________________________________
topicmapmail mailing list
[email protected]
http://www.infoloom.com/mailman/listinfo/topicmapmail