SUO: Re: Formalizing SUO Purposes
In John Sowa's post of 24 Aug with the above subject-line (in response to Mike
Uschold and posted under James Schoening's name) he wrote :
> Oh, no! I wasn't thinking at all about formalizing the
> SUO purpose statements. I considered them to be at a very
> high metametametalevel where natural language is necessary
> because it is "sufficiently vague" that people can discuss
> plans and projects at a level where no one really knows what
> the likely outcome will be.
In my post of 24 Aug (as also repeated by James Schoening), with
subject-line "SUO: A fabric of many threads", I introduced a SUO-like
project which has already advanced, via what I described as a lengthy
"demand/supply dialectic", considerably further than that "vague" state and
much more clearly indicates the "likely outcome", to the extent that it has
already formalized a "SUO Scope and Purpose" to a considerable degree.
Thus that post rather abstractly described "MACK", the "Mainstream Architecture
for Common Knowledge" (as at present being implemented in the "Metaset" boot or
seed product), and affirmed as follows:
'What's more, [the MACK SUO-equivalents'] full "Scope and Purpose", along
with all "documentation", will be expressed in canonical MACK form (greatly
mutatis mutandis from the present SUO Scope and Purpose, of course!) In that
form it will be the integrated driver of all implementations such as Metaset.
There is in fact some very satisfying epistemology in that design and consequent
reformulation, but for now I'll just add that it fulfils the expectations or
hopes of John Sowa and Mike Uschold as apparent in the thread "Formalizing SUO
Purposes" started by Mike on 18 August.'
Evidently I had read too much into the latter "expectations or hopes", but - the
supply-side pointing the way here - I must nonetheless reaffirm that ability of
MACK as it is at present taking ever firmer shape.
It appears that the relevant technology is quite radically different in many
respects from what this list generally seems to have in mind, but if one
faces the epistemological or knowledge-making role of the market (and
more effective and better-sharable knowledge-making is surely why we are talking
in terms of ontologies!) then the high-level end-user-view is quite simple, and
could - for present purposes - be stated along these lines:
The market is where the implemented creativity of some is validated and
propagated when others put it to their own benefit. There are continual
iterations of that democratic process, which seeks - in what sometimes seems a
Sisyphian way - to do ever better justice to the demand-side, where our given
human complexity is to be sought despite its resistance to formulation. With a
view to better enabling that process especially in respect of systems and
theories, we are seeking to construct an Internet-based marketplace supporting
such full-cycle marketing of "ontology-based" products.
A truly generic technology to help formulate and meet the needs of others should
be able to address its own needs. It would thus provide for the gradual process
of those needs' formalization, refinement and elaboration in reliably-usable
architected terms.
Thus, as mentioned above, the Metaset/MACK project has already advanced
considerably along that path, to the extent that the present formalization
is the integrated driver of Metaset.
That also largely explains why the initial Metaset product is called a Boot
or Seed product. It is designed to be self-leveraging in that way, exploiting
for its own evolution the creativity, energy and eventual directedness, in so
many ways, of the wider market.
Technologically, self-leveraging requires self-referencing, or what is often
called reflectivity, and MACK, as related in the above-mentioned post, has
that quality (along with many other qualities!) in terms of an effectively
axiomatized logical construct that this list would call an ontology.
The Uschold/Sowa dialog continued:
>
> >These are more interesting points and examples. However, they do not
> >shed any light on the question of whether you are arguing for the SUO
> >purpose statements that we are discussiong to be formalized.
>
> What I was talking about was something like the lattice of
> all possible theories, in which each theory would have something
> like the Java beans reflection information so that both people
> and machines could check the assumptions about how the theory
> is intended to be used.
MACK goes further than that in its reflectivity. But possibly in rather a
different way. I know Java is only used as an example there, but as it happens
I have often found it necessary to distinguish MACK's reflectivity strongly from
that of Java Beans or Java's so-called introspection. Reflectivity has to
assume
that it is looking at something relevant, but from the MACK point of view,
Java'a application decomposition and recomposition approach all too often
produces a very poor "equivalent" of reality. More generally, my
long-standing criticism of classical OO's "island class" approach (the
latter being Grady Booch's words) is that it often very poorly provides for the
interconnectedness of reality that semantic webs or conceptual graphs are trying
to capture. The problem is even inherent in conventional OO's encapsulation
imperative, and there are other ways of achieving the always-necessary
"complexity-hiding" that encapsulation aims for. Such criticisms - and
associated predictions - have long been proven very valid, as I related in my
earlier post. Now, in such complex areas as we are treading with the very
concept of a SUO, bold yet confirmed predictions may be regarded as probably
quite significant, and at least partly indicative of a correct high-level view.
I apologize for such smugness (and I admit again that I am not taking John's
reference to Java as he had intended it), but I do think that we must see
that the reflectivity that we are looking for - which of course goes far
beyond an integrated and canonically-expressed Scope and Purpose! - has to
be very much better than some current models of reflectivity seem to allow.
I think I can even go further and assert that MACK's reflectivity is more
pertinently useful than that of conventional Higher-Order Logics.
So I have made the opportunity (albeit rather twistedly...) to try to
emphasize once again that MACK is something very different and could well be
worth this list's more serious attention. And that is despite some aspects
in my proposal that will at first sight seem rather way-out.
So at this stage may I recommend another consideration of my earlier post? And
since it went out when there was a problem with the list-server, it is also
possible that some of you may have missed it. So I repeat: it was on 24 Aug,
under James Schoening's name, with subject-line: A fabric of many threads.
Best regards,
Christopher