Review of TUNES, arrow and the semantic web in the context of the introspector

[Brian T Rice]

Hello again,

On Sat, 30 Aug 2003, James Michael DuPont wrote:

> Dear Tuners,
>
> After being accused (Rightfully) of not understanding tunes or doing my
> homework, I have started from the beginning. The arrow system is the
> first part of tunes that I think could have be grasped and processed by
> an machine.  I am starting a very detailed review of TUNEs in the
> context of the semantic web and the introspector project.

Thanks for finally admitting it. What follows seems like a step, but you
are still displaying some interaction patterns that are troublesome.

A post-facto note: I think you should have surveyed the mailing list
discussions from the past before summarizing as you did.

You are actually wrong about Arrow. Arrow is a superset of TUNES ideas to
a large extent. But I don't really have time to explain this at the
moment. (I've said it on the mailing list and explained it fairly often.
Look there.)

> I have been reviewing the arrow philosophy file, and for this purpose,
> I converted it to text,
> http://introspector.sourceforge.net/2003/08/ArrowPhilosophy.txt

You should have just asked for an alternative format, like HTML. I'll dig
up my sources and export them soon.

> PhilosophyText CHUMP entry
> http://rdfig.xmlhack.com/2003/08/30/2003-08-30.html#1062208848.959835

Would you mind asking me before re-publishing my work? At least you could
inquire to compare your interpretation with mine.

> I am not done with this process, but as far as I can tell, arrow and
> the semantic web are very related, more than noted in the tune cliki.
> In fact, the definition of the MetaText seems to do a disservice to the
> semantic web, I can only assume that this is because lack of deeper
> research.
> MetaText  CHUMP
> http://rdfig.xmlhack.com/2003/08/30/2003-08-30.html#1062231989.305187
>
> I have extracted the beginnings of an OWL ontology for the classes of
> arrow here :
> http://introspector.sourceforge.net/2003/08/arrowclasses.n3
> CHUMP :
> http://rdfig.xmlhack.com/2003/08/30/2003-08-30.html#1062234793.644775

Why didn't you bother to actually ask me what the relationship between the
code and paper was? The Arrow code was just a tiny prototype that didn't
express even a tiny percentage of the whole idea. And Squeak was a really
poor means to do it in, but it's something average people could at least
hope to understand. Equational semantics, proper closures, and other
things are at the minimum needed to express it. Probably my next version
would be in Maude or some equivalent.

Representing that code is foolish, and as the author, I do not condone it.

> When I am finished reviewing arrow, then I can make the connection
> between the semantic web and the arrow system. This will put the
> introspector in context.

You should have asked me first. The arrow concept occurred to me first in
1994, when I had first studied and grasped category theory, and had
already learned Lisp and the basics of formal logic and inference calculi
and been programming for about 9 years.  Now, the Semantic Web and Arrow
both make use of the term "ontology", but there is a different heritage
among them...

Arrow's use of the term "ontology" relates to the philosophy of Martin
Heidegger, and is more poetic than technical. The Semantic Web's turn on
it is related to "formal ontologies"  whose research started in the 1980's
on knowledge representation and is basically a really limited, fairly
useless perversion of the idea.

A readily accessible interview about Heidegger that might help those of
you allergic to serious philosophy is here:
http://archive.salon.com/books/feature/2000/11/13/heidegger/print.html

``what do we mean by "is," and how does it come to pass that "is" means
anything to us at all?'' -- This kind of question is what Being and
ontology are about, and encoding this in the nature of a textual format
and placing primacy on the text is the mistake of the knowledge
representation groups and the Semantic Web in terms of philosophy.
Basically, it's an unsound idea.

The fundamental difference between the Arrow system's handling of this
question and the Semantic Web may seem trivial to your perspective, but
really are profound.

By the way, (and this is REALLY IMPORTANT!) the Arrow Philosophy paper was
written when I knew FAR less about CS ideas than I do now, probably by a
large multiple, and was also in very poor conditions to write. If you look
for comments about Arrow over the years on the TUNES mailing list, you
will see a lot of background information for the before and after periods
relating to my writing the paper:

http://www.google.com/search?as_q=Arrow&num=10&hl=en&ie=UTF-8&oe=UTF-8&btnG=Google+Search&as_epq=&as_oq=&as_eq=&lr=&as_ft=i&as_filetype=&as_qdr=all&as_occt=any&as_dt=i&as_sitesearch=lists.tunes.org&safe=images

So, reviewing that paper is not very productive. I need to write a much
better one, more technical and specific, and probably about 200 pages at
least, but I don't have the time/funding to spend on it. But I, as the
author, can say that understanding of that document does not constitute an
understanding of the idea, except within the limitations of my writing
abilities and state of mind at the time.

I am obliged to note, also, that you're still focussing on your own
project's perspective. I've heard recently that you do this with all of
the other projects that you discuss with, and that they resent this
attitude and your mode of interaction. We've discussed this already to a
minimal extent, and I don't want to repeat the details, but you do display
an alarming pattern of trumpetting your own project's perspective in the
context and discussion areas of other projects. What's critical is that
you're not handling the necessary social tasks of "doing as the Romans
do", that is, letting someone who's created their own project and has
their own specific and well-formed goals do what they like in their own
forums without harassment.

> Specifically  : Based on the definition of MetaText,
> my hypothesis  is so far :
> This is very similar to the work being done now by the introspector
> project. The de/re-composable objects are in the introspector handled
> by RDF gateways and de-reassemblers, not by rewriting everything in
> lisp/smalltalk or someother funky language.

Sure, but TUNES MetaText is not an idea you can isolate from the rest of
TUNES. It's the idea of applying the HLL and its philosophy to human
expressions. That means that the reflective means has to be very direct,
succinct, and malleable (and multi-paradigm).

Therefore, if you don't accept HLL ("funky language"?), you're missing the
point of MetaText. There are no exceptions to this inference. You either
accept it, or you wholly disagree or do not understand TUNES' application
to the WWW, depending on your perspective. My bet is that you will judge
the former of yourself, since that is consistent with your previous
behavior.

The Semantic Web is one possible reification, within an inefficient
(less-expressive in a few ways) XML format (or syntactic equivalent), and
is only one standard, even if it is a meta-standard. Yes, it's useful, but
it's not of the same scale or scope.

Why don't you bring up the points of MetaText explanations and explain why
you think each is satisfied by the Semantic Web? Actually, you are a
really poor person to ask, for the self-promotion issues I have alluded to
above.

Here's a source to cite: http://tunes.org/Interfaces/tunesvswww.html

A horrible source, but it's Fare's writing and I only cleaned it up as
much as occurred to me to be possible. I definitely know much more than
this page implies, but I don't have a concrete plan to rewrite it
accordint to, or time to develop one.

Finally, the "funky language" comment is incredibly rude. If you don't
understand why languages like those would be preferred by the TUNES
project, then I suggest you go try to use them and learn from them or just
leave. The TUNES concept is that language (any fundamental means of
expression) should be malleable and exist knowingly within a continuum,
and Lisp or Maude or some (greatly) extended version of Haskell or ML are
at /least/ required in order to get to that level of sophistication.

If you can't understand why language needs to be malleable and flexible,
then you're hypocritical, since you trumpet XML standards and such, which
are a reaction to the limitations of "non-funky languages". In Lisp or
Smalltalk or related languages, there are fewer barriers between the base
language and data-representation schemes or validators or even
meta-standards. Lisp and Smalltalk as they are certainly need a lot of
improvement, but you are hypocritical not to acknowledge the applicability
of our reasoning in this matter.

I would spend more time on this, but I have a lot of things to work on,
and I'm not sure how effective any argumentation is with you, having read
your interactions on other mailing lists.

-- 
Brian T. Rice
LOGOS Research and Development
http://tunes.org/~water/