Re: [xtm-wg] Re: parallel development of syntax and concept models

["Steven R. Newcomb" <srn@techno.com>]

-------------------------- eGroups Sponsor -------------------------~-~>
GET A NEXTCARD VISA, in 30 seconds!  Get rates
of 2.9% Intro or 9.9% Ongoing APR* and no annual fee!
Apply NOW!
http://click.egroups.com/1/7872/4/_/337252/_/967612957/
---------------------------------------------------------------------_->

[Steve Newcomb:]
> > OK.  I see that we don't agree about this.  I would ask you a
> > question: Is there nothing about the Platonic forms that is not
> > captured in a UML model of them?  It worries me that you don't see any
> > distinction between your preferred expression of the Platonic forms (a
> > set of object classes) and the Platonic forms themselves.

[Luis Martinez:]
> I am completely aware of the distinction. I know that anything that
> represents the Platonic forms is an imperfect copy.

I think maybe I see what is the nature of our communications problem.
I do not believe that a representation of a Platonic form is "an
imperfect copy" of it.  There can never be such a thing as a "copy" of
a Platonic form.  A representation, yes.  A copy, no.  Just as a
person is one kind of thing, and a representation of a person is a
completely different kind of thing, a Platonic form is one kind of
thing, while a representation of a Platonic form is a completely
different thing.  For example, it's not meaningful to say that a
photograph of Luis Martinez is "an imperfect copy" of Luis Martinez.

A statue of Luis Martinez is another kind of representation of Luis
Martinez.  Still, the statue is not an "imperfect copy" of Luis
Martinez.  It's the sculptor's vision of how to portray Luis
Martinez's physical form by means of a statue.  (In fact, a statue
doesn't copy the physical form; instead, it is crafted in a way that
evokes in the viewer some idea of the physical form of Luis Martinez.
Besides, there is much more to Luis Martinez than his physical form.)

There is no such thing as a copy of a notion.  There can be copies of
*representations* of notions, but not of notions.  A notion is the
same, unique notion no matter how many ways it is represented, and no
matter who does or does not understand it.  Notions (thoughts) exist
on a higher plane of existence.  They can be forgotten, but they
cannot be destroyed.

> But, if we do not base our implementations on an imperfect
> representation, what are we basing on?  That is what I am not
> understanding.

We base our implementations on our models, and we base our models on
our best understandings of the Platonic forms.

> This is my point. I do not think that an object model scope and
> field of application is to just implement Application Program
> Intefaces. It is broader than that. Is to help us understand what we
> are trying to design here.

I agree with this statement.

Do you claim that the scope and field of application of a syntax
specification is so much more limited than the scope and field of
application of an object-oriented model that, by creating such a
syntax specification, we are not engaged in a process that helps us
understand what we are trying to design?

> <!ATTLIST
>   name
>    scope          -- List of references to scoping topics that
>                      collectively define the limited contexts within
>                      which the name is applicable to the topic. --
>       CDATA       -- Reference --
>                   -- Reftype: topic+ --
>       #IMPLIED    -- Default: scope is unconstrained. --
> >
> 
> scope CDATA #IMPLIED with a bunch of comments next to it is just not
> good enough for me. Prose is great for lawyer but not for programmers.

OK.  If what you're saying is that the above model doesn't provide as
much useful guidance to software developers as an object-oriented
model might, I agree with you.

Others can't match your skills in the programming arena, but some of
us can do other things, like create standards that act as contracts
between various people about what constitutes the agreed-upon syntax
for the interchange of certain notions.  It's not systems design, nor
systems implementation, but it's honest work, and somebody's got to do
it.  It's not the same kind of work that you do, but your system
design for the support of some set of notions can be based on the
*same* set of notions on the basis of which others may create
syntactic models.  There is no reason why these people need to base
their work on a model that is to be used as the basis for a systems
implementation.  Both models are intended to represent the same
notions, but in different terms and for different purposes.

******************************

I've been discussing this discussion with others, and it has been
pointed out to me that may find my assertions incomprehensible because
they have a different definition of "model" than the one I'm using.

For many people, the word "model" means the same thing that I mean
when I say "information set".  When I say, "it's OK for an information
resource, such as a topic map, to have multiple models", what
people seem to understand from that is that there is no single,
well-understood set of notions that constitutes the topic maps
paradigm.  I'M NOT SAYING ANY SUCH THING.  If that's your definition
of the word "model", let me use that definition now to assert very
clearly and loudly, "THERE IS EXACTLY ONE SINGLE WELL-UNDERSTOOD MODEL
OF THE TOPIC MAPS PARADIGM."

For me, a "model" is a representation of some notion.  A single notion
can have several representations ("models").

Again, a "model" is not a poor copy of a notion or an idea.  It's
just a representation that's intended to direct our inner attention
toward the notion of which it's a representation.  Consider the
tinkertoy-like model of the DNA molecule that Watson and Crick made in
order to communicate the nature of their discovery to others.  That
tinkertoy-like model was not "a flawed copy of a DNA molecule".  There
were very few features of DNA molecules that were duplicated in that
model: in fact, the only features that were "copied" were the bonding
angles of the base pairs.  (Still, this model had incredible impact,
which, fifty years later, is still only beginning to be felt.)

Many computer models of DNA molecules have been made.  These computer
models have different purposes than Crick & Watson's model, which
still has enormous and compelling value for teaching purposes.  It's
more useful to have both Crick & Watson's physical, tangible model of
the bonding angles, *and* various computer models, than to have either
one alone.  But neither kind of model is the last or only word on DNA.
Neither model is based on the other.  Both are based on the reality of
the notion of DNA.  It is not necessary, and it is extremely
undesirable, that there be one single master representation of DNA on
the basis of which all further theorizing and learning about DNA must
be done.  We must allow all kinds of theories about DNA to be
represented, because some of them will turn out to be right, and some
of those will challenge and, ultimately, overthrow or change our
beliefs about DNA.

******************************

> I understand the distinction. If we don't have an explicit,
> imperfect, flawed representation of topicmaps, all we have is an
> implicit model of topicmaps left to interpretation.

I'm *not* urging that we have no "explicit... representation of
topicmaps".  Far, far from it.  I'm simply pointing out that it's
wrong to insist that there must be a single "master representation"
from which all other representations must be wholly and slavishly
derived.  I'm saying that, beyond *any* model, 'way, 'way off in our
inner universe of notions, is the notion (or set of notions) which the
model is attempting to represent.  Good models represent the relevant
notions as well as they can.  It's not essential that all models of
any particular set of notions be completely consistent with each other
in every way.  It *is* essential that all models of certain notions be
true to those same notions, in whatever ways they can and should.
Consistency with those notions is what makes a model useful.
Consistency with other models of those same notions is not necessarily
useful.

> [If we don't have a "master representation" against which to compare
> all other models] ...all we have is an implicit model of topicmaps
> left to interpretation.

No, that's not all we have.  What we have is the possibility of
creating multiple models (representations!), and that's a good thing
to do.  It helps us understand and deal with reality better.

It's completely OK with me if there are people who, like you, believe
that there is "one, true model" of topic maps, and that that model is
an object-oriented model.  

I hope it's OK with you if there are people who disagree with you
about that, and who believe, as I do, that no model can ever possibly
be complete, and therefore that it's very good and very appropriate to
have multiple models, especially where there are different
requirements for different models.

This whole business of deciding which kind of model is the right kind,
and even that there *is* a correct master model is an extremely
religious thing.  In fact, I'll even go so far as to say that choosing
or developing a religion for oneself is exactly the same thing as
choosing or developing a model, and then believing that it is the one,
true "master" model.

If I were you, Luis, at this point I'd ask, "So.  If the syntax model
of topic maps is created independently of the object-oriented model,
rather than based on the object-oriented model, how can we tell
whether the two models are actually good models of the set of abstract
notions that is the topic maps paradigm, and that the information set
is actually the same information set?"  The answer to this question
is:

  We can tell by seeing what abstract notions are implied by the two
  models, and by comparing the two sets of abstract notions.

-Steve

--
Steven R. Newcomb, Consultant
srn@techno.com

voice: +1 972 359 8160
fax:   +1 972 359 0270

405 Flagler Court
Allen, Texas 75013-2821 USA

To Post a message, send it to:   xtm-wg@eGroups.com

To Unsubscribe, send a blank message to: xtm-wg-unsubscribe@eGroups.com