Fwd: [Ontology Summit] Ontologies for describing the context of systems projects

[Jack Ring]

John Sowa’s ideas are aways enlightening.

Begin forwarded message:

From: John F Sowa <so...@bestweb.net>

Subject: [Ontology Summit] Ontologies for describing the context of systems projects

Date: October 17, 2017 at 08:41:17 MST

To: ontolog...@googlegroups.com

Cc: ontolog-forum <ontolo...@googlegroups.com>

Reply-To: ontolog...@googlegroups.com

Jack and Janet,

I agree with both of you.  This topic is so fundamental that I
added Ontolog Forum to the list of recipients.

Jack

The focus is not on the technology of ontologies per se but on the
content, structure and synergy of ways that "describe the problem”
so that system engineering can proceed.

Janet

Jack provides an excellent scoping of what any proposed ‘ontology
for systems engineering’ needs to account for.

Note:  Jack Ring's slides (3.1 MB) are large for an attachment.
They should be uploaded somewhere.  See below for the URL of a
64-slide presentation that has considerable overlap with them.

For those who don't have the slides, I'll summarize some significant
points.  By the way, the BFO ontology does not support any of them.

On page 3, the diagram shows an oval that emphasizes three aspects:
value, purpose, and system.  Around the oval are the fundamental
stages of any large engineering project:

Community situation.  Problem discernment.  Problem system understood.
Solution effect envisioned.  Intervention strategy.  Problem control
system (PCS) specified.  PCS designed and architected.  Components
specified, developed, assembled.  PCS tested.  Operational readiness.
PCS activated. Operational results.  The purpose of a system is what
it does (POSWID) regardless of the intent of the sponsor or design.
Context adapted. Effects on problem known. Value of system quantified.

The arrows indicate that the stages are implemented and iterated
in a cycle around the oval.  POSWID implies that the cycle must be
repeated as long as necessary to match what the system does to the
sponsors' and designers' intentions.  Since the intentions evolve
over time, the cycle won't end until the system is abandoned.

(IBM euphemistically calls that endpoint "functionally stabilized".)

Quotation from page 4:

Example intervention systems are the air conditioner in your house
or car, the autopilot on a commercial airliner, an anti-ballistic
missile system (with 30 million software instructions), a driverless
automobile (with upwards of 100 million software instructions
containing 30,000 conditional execution threads), First Responders
in national emergencies, riot and cyber-attack preventions, etc.

Pages 6 and 7 discuss some questions for ontologists:  What's the
problem?  Who cares?  What will help them?  How can we be sure that
our recommendations will or did help them?  Are we too focused on
procedural models?  Do requirements establish ontology?

Note that all those questions involve intentions, social relations,
and value judgments, including ethics.

Page 9 shows a diagram of a "human interaction model" that is taken
from the book _Amity and Enmity: Two Archetypes of Social Existence,
An Interdisciplinary Study_ by Rudolf Starkermann.  See URL below.

After the slides is a copy of the chat during the presentation.
There was some discussion of SysML diagrams, which include a subset
of UML diagrams with extensions for representing systems, their
requirements, and interaction diagrams that model their operation.
See URLs below.

The Basic Formal Ontology (BFO) says nothing about any of those issues.
Its total content can be summarized in one short paragraph:

Everything that exists is either a continuant or an occurrent.
Some continuants (objects) are independent; some (qualities) depend
on objects; and the rest are spatial regions.  Occurrents include
processes, temporal regions, and spatiotemporal regions. A version
of mereology is used as a theory of parts and subparts.

Issues that BFO ignores:  The oval with its focus on value, purpose,
and system; and engineering projects, which begin with problems and
intentions and produce results that solve the problems.

A system is much more than a collection of parts and subparts.
It's a tightly organized structure designed to serve some purpose.
For any system, natural or artificial, purpose (telos) is critical
to the definition, use, and even existence of the system.

In a note to Ontolog Forum (Sept. 8), Barry Smith made the following
replies to some issues that I raised:  (URL below)

The treatment of intentionality might indeed be part of ​a TLO
(though some would argue that it belongs rather to a domain ontology
of, for example, psychology)...
purpose, goals, plans for the future, design issues, and value
judgments, while important and interesting in their own right,
may not be essential to every TLO...
From my own point of view they would belong more properly in a
domain ontology relating to human action and matters associated
therewith...

By claiming that some other ontology (either a different TLO or
some theory at a lower level) might handle these issues, Barry
supports my point that BFO, by itself, does nothing for them.

For the topics that BFO does support, I'll cite a book by the
analytic philosopher Peter Unger, who became disillusioned
with the field he devoted his career to research and teaching:
_Empty Ideas: A Critique of Analytic Philosophy_, Oxford U. Press.
See below for the URLs of an interview with Peter U. and some
reviews of his book (both pro and con).

In his book and interview, Unger explains that when analytic
philosophers make claims about scientific issues, their ideas
add nothing to the facts and theories developed by the experts
-- the scientists themselves.  As Unger said in the interview,

To say new and interesting things about the world... you really
have to engage with a lot of science.  And very few philosophers
do any of that, at least in any relevant way.

In reviewing Unger's book, Joel Dittmer (URL below), said

He doesn't touch ethics, political philosophy, or aesthetics.
Okay, these are all normative.  But he also certainly doesn't
touch epistemology.  And although epistemology is at times
normative, much of it is not (at least not directly)...

These have been central issues of philosophy from Socrates to
the present.  They are also critical issues for engineering,
as Jack presented them in his slides and talk:  requirements
(value judgements), community situation (political and social),
intervention strategy (political and social), problem discernment
(epistemology), problem system understood (epistemology)...

The reason why analytic philosophy (including BFO) ignores those
issues is that they are outside the scope of scientific method.
Science, by itself, does not make value judgments.  The methods
of science *and* the mainstream of analytic philosophy can only
determine what is.  They cannot determine what ought to be.

But engineering depends heavily on intentionality and value
judgments -- not what is done, but what *ought* to be done.
It involves aesthetics (design), ethics (moral and legal issues),
politics (management and funding), and social approval by the
sponsors, employees, users, and the surrounding community.

This implies that intentionality and value judgments must be included
at the top level.  They cannot be pushed down into some lower level
because there is no way to define 'ought' in terms of 'is'.  If the
foundation does not support them, there is no way to construct them.

These are the reasons why I said that BFO should be considered a
lower-level microtheory.  It is not sufficiently general to solve
or even to state the issues that Jack described.

The references below say more about the proposed ISO standard for
ontology, Jack's slides, Unger's book and reviews of it, and the
use of SysML as graphic notations for ontology design.

John
____________________________________________________________________

An Ontolog thread with notes by Matthew West, Barry Smith, and me:
https://groups.google.com/forum/#!topic/ontolog-forum/ZCcEiM2xIi4

Part 1 of the draft ISO standard for ontology (June 2017).  Some
minor revisions since then have done nothing to address the issues:
http://stl.mie.utoronto.ca/upper/21838-1_WD1--20170813.pdf

A two-page review by John Sowa of Part 1 of the draft ISO standard:
http:\\jfsowa.com\ikl\revISO21838p1.pdf

"Signs and Reality" by John Sowa.  Issues about including semiotics
at the top level of an ontology:  http:\\jfsowa.com\pubs\signs.pdf

"Worlds, Models, and Description" by John Sowa.  Modal logic and the
issues of deriving 'ought' from 'is':  http://jfsowa.com/pubs/worlds.pdf

Slides by Jack Ring in August 2009:
https://www.researchgate.net/profile/Jack_Ring/publication/265039235_Evolving_An_Autonomous_Evolving_An_Autonomous_Test_and_Evaluation_Enterprise_Test_and_Evaluation_Enterprise/links/544137bf0cf2e6f0c0f604ea/Evolving-An-Autonomous-Evolving-An-Autonomous-Test-and-Evaluation-Enterprise-Test-and-Evaluation-Enterprise.pdf

A 40-page synopsis of Volume 1 of the book _Amity and Enmity:
Two Archetypes of Social Existence, An Interdisciplinary Study_
by Rudolf Starkermann:
http://www.parshift.com/AgileSysAndEnt/WrkPap/WrkPap683-Sommer071015-StarkermannVol1.pdf

Amazon excerpt and description of the book by Peter Unger (2014):
https://www.amazon.com/Empty-Ideas-Critique-Analytic-Philosophy/dp/0199330816

Interview with Peter U. about his book
http://www.3quarksdaily.com/3quarksdaily/2014/06/philosophy-is-a-bunch-of-empty-ideas-interview-with-peter-unger.html

Review of Unger (2014) by Joel Dittmer
https://journals.uvic.ca/index.php/pir/article/viewFile/15110/6071

Review of Unger (2014) by Katherine Hawley
http://ndpr.nd.edu/news/empty-ideas-a-critique-of-analytic-philosophy/

Review of Unger (2014) by Sven Ove Hansson
https://www.researchgate.net/publication/270598578_Is_Philosophy_Just_a_Set_of_Empty_Ideas

"An Ontology for State Analysis: Formalizing the Mapping to SysML"
by David Wagner et al.  The use of SysML for developing an ontology
at Jet Propulsion Laboratory for the Phoenix Lander on Mars.
http://www.omgsysml.org/State_Analysis_Ontology%20_in_SysML.pdf

See also https://en.wikipedia.org/wiki/Systems_Modeling_Language.

--
All contributions to this forum by its members are made under an open content license, open publication license, open source or free software license. Unless otherwise specified, all Ontology Summit content shall be subject to the Creative Commons CC-BY-SA 4.0 License or its successors.
--- You received this message because you are subscribed to the Google Groups "ontology-summit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ontology-summ...@googlegroups.com.
To post to this group, send email to ontolog...@googlegroups.com.
Visit this group at https://groups.google.com/group/ontology-summit.
To view this discussion on the web visit https://groups.google.com/d/msgid/ontology-summit/ab7537ba-bd6f-257a-bf1b-25391ed5f1ea%40bestweb.net.
For more options, visit https://groups.google.com/d/optout.