[sc34wg3] a new name for the Reference Model

Sam Hunting sc34wg3@isotopicmaps.org
Thu, 2 Jan 2003 19:35:09 -0500 (EST)


This post vanished in thread noise, and I want to revive it.

> Anthony B. Coates wrote:
> 
> >** Reply to message from Patrick Durusau <pdurusau@emory.edu> on Tue, 31 Dec
> >2002 07:18:01 -0500
> >
> >>I did not mean to imply that you were deliberately inflating the title 
> >>but I don't see "TM Canonical Metamodel" as indicating (to me at least) 
> >>anything meaningful about the contents. I am with you through "TM" and 
> >>possibly "Canonical" (assuming some meaningful word followed it) but I 
> >>get lost at "Metamodel." If you mean a model of a model, is it still not 
> >>just a model? Have I added anything meaningful by appending "meta" to 
> >>it? That is assuming that we don't already share a common view of some 
> >>set of particular models in a discipline and also a common view of what 
> >>a model that encompassed all those models would consist of, then in a 
> >>particular context "metamodel" might be a meaningful term. But that 
> >>presumes a shared understanding of the term, which I think is lacking in 
> >>this case.
> >>

> >You do have me wondering now whether "metamodel" is really the best term.  That
> >is to say, a "metamodel" is just a model of a model.  You define the family of
> >possible structures using a model, and you describe the family of possible
> >models using a metamodel.  I had been thinking of the RM as something you could
> >use to model the SAM, which is itself a model.  That would make the RM a
> >metamodel.  However, it can also be viewed as a lower-level model of TMs than
> >XTM or the SAM provide.  I'm not sure how it is likely to be viewed/used in
> >future.  So, comments (by e-mail or on the back of a postcard) are welcome as
> >to whether the RM is a metamodel or not.
> >

[patrick durusau]
> Let me use an example completely divorced from topic maps to see if I 
> can clarify what I was attempting to say. This may or may not have any 
> resemblance to your reply.
> 
> As a child I had a "model" of the Volkswagen Beetle that was made in 
> Germany. The doors opened, but in terms of functions, that was the 
> extent to which it "operated" like the car of which it was a model. More 
> complex models of cars are possible and I assume there may even be 
> models that have all the functions of the cars of which they are "models."
> 
> It would be possible to construct a "Model" for developing "models" of 
> cars. The "Model" for "models" would specify things like the size ratio, 
> what things had to operate (or even be present, like the engine), etc. 
> The essential point being that the target (I hesitate to use "subject" 
> for fear of more confusion) of the Model is not the same one as the 
> "models" of cars. One has as its target a general class of things called 
> "models of cars" and the other has as its target a particular car.
> 
> What I don't see (contrary to Lars' repeating "model use to construct 
> other models") is what adding "meta" to "model" obtains in terms of 
> information? A model is always a "model" of something and simply 
> prepending a term borne of the physical arrangement of books some 
> centuries after an author's death does not seem to me to add anything.

I'm with Patrick on this one. (Reminds me of the old jokes about data,
metadata, metametadata ...)

I keep asking people (mathematicians, programmers, analysts) what
(meta)model means, and I can't get answers that are consistent, or that
are simple enough for me to think of using. (One answer I got was
that "model" == "abstraction" which was so, well, abstract that I didn't
find it useful.)

Using Patrick's metaphor: I have a "model" of a VW -- a kit of plastic
parts to assemble according to a set of printed of directions. Then again
there is a "meta-model" which could be the molds used to make the parts
and the directions considered as content. Then again there is a
"meta-meta-model" which is the ... principles used to design the molds and
write the directions....

But assembling a kit is not at all the same as making a mold is not at all
the same as crafting a set of principles. And the "intent" to form (say) a
plastic part is already built in to the word "mold", as the "intent" to
assemble (say) a kit from plastic parts is already built in to the word
"directions" as the "intent" to achieve a certain objective is built in
to the word "principles." 

So in terms of work actually done, and messages communicated, considering
all these activities as modeling, and prepending the the particle "meta"
to some such activities, doesn't seem to communicate intent, or indeed to
add much of value, except for the pleasure of the word-play, of course,
which is high.

Sam Hunting
eTopicality, Inc.

---------------------------------------------------------------------------
"Turn your searching experience into a finding experience."(tm)

Topic map consulting and training: www.etopicality.com
Free open source topic map tools:  www.gooseworks.org

XML Topic Maps: Creating and Using Topic Maps for the Web.
Addison-Wesley, ISBN 0-201-74960-2.
---------------------------------------------------------------------------