|
Jack, many thanks for the wide-ranging and so
rapidly-produced pointers.
Some first reactions from me:
1. I shall definitely look at the Dijkstra,
though maybe for reasons different from expecting an answer to my
question. I have in fact been strongly into self-stabilizing systems since
1972, though doubtless very simplistically then, when compared to ED's and even
my own later work. As I've already written on the web in 1996, the
introductory workshop for a 1980s subset of a precursor of Ontology Chemistry
was entitled "Errors: their prevention, detection and management". (Note the
optimism of the first subject, and the realism of the remaining two.) But
it's true of my own work, and I'd guess of ED's too, that none of it really
answers my main question. Defined structure, and function only
in-the-small, is easy. Big Systems is the problem domain
here.
2. Until corrected, I shall assume the Cox,
Love work never got past toy level.
3. Japaridze or von Wright? Hmmm, no thanks, not now,
at least. Kineman? I'm still to try google.
4. Anything specific from Pizzarello? (Google gave me no quick and easy
pointer.)
5. Bayesian Belief Networks? Maybe, but much later
from my project point of view.
6. Conceptual Blending is very interesting. I have
quite widely perused Fauconnier and Turner; Lackoff and Mark Johnson
too. Like them, I acknowledge a debt to Koestler and his Act of Creation,
having bought and absorbed that book the moment it was published in
1964. I don't doubt it had a lot to do with my emphasis on MI and multiple
contexts. But very relevant though all that is, none of it really
addresses my main question in this thread.
To summarize: I regret to observe you mostly seem
to fail to dislodge my damned preconceptions! However, do you have an
answer re Pizzarello? Or any other insistence re my superficial
assessments?
Many thanks again.
Christopher
----- Original Message -----
Sent: Tuesday, January 31, 2012 4:48
PM
Subject: Re: [ontology-summit]
[BigSystems and SystemsEngineering]Relationship between system function and
system structure
Christopher,
Djikstra described recipes for self-healing systems.
E. W. Dijkstra,
“Self stabilizing systems in spite of distributed control,” CACM, Vol17, 11,
643-644 (1974).
In the 1980's Brad Cox, Tom Love, et al, arranged Smalltalk classes
which, when asked a question, organized 'themselves' to produce a response.
(before the Object-oriented crowd reverted almost everyone back to prescient
programming).
Recently,
G. Japaridze, “Introduction to Computability Logic,” Annals of Pure and
Applied Logic, 2003 pp 1-99.
R. Back and J. von Wright, Refinement Calculus, Springer Verlag,
2008.
and the work of Prof. John Kineman, U. of CO, extending Rosen's R-theory
to a relational algebra
appear promising for composing systems and even for automating
composition.
Meanwhile, my OntoPilot LLC associate, A. Pizzarello, pioneers
system-wide integrity checking.
Also relevant, IMO, are the instances of using Bayesian Belief Networks
to simulate and even emulate systems in order to understand the
function/structure effects in various contexts and formulate purposeful
systems.
and the work of Prof. Ockie Bosch, U. Queensland, Australia.
Even cognitive science is relevant, IMO, c.f., the Conceptual Blending
work at Stanford U. and other places,
Lot's to do. You appear to be the kind of guy who can
do it.
Jack
On Jan 31, 2012, at 6:41 AM, Christopher Spottiswoode wrote:
Joe, thank you for the distinctions you make
here. (And sorry I have only now opened this input from
you.)
You have prompted me to open a new
thread.
Whether or not you would approve of the words I
use, you at least remind me of a question I have often thought but promptly
forgotten to put to formal ontologists:
What work has been done on formal deduction
of system function from system structure, or vice
versa?
As far as I am aware, at most very little of
practical relevance to our Big Systems has been achieved along those lines,
but I would be absolutely delighted to be enlightened, and to study the
techniques and their limits.
Please, some formal ontologist out there, point
me in the right direction?!
And why might I be so interested?
Hint: As one talking so much of "Ontology Chemistry" and composition
from components, I am intrigued to note a predominance of links in the areas
of chemistry and physiology.upon now, for the first time, googling
/structure function deduce/. Adding "ontology" to the search string
brings about an enormous reduction in the number found, from 23M to 3M, and
many of the 3M are in some biochemical field.
Perhaps I should add that I have long already
proceeded on the basis of a largely negative answer to my question, and that
the way Ontology Chemistry deals with that presumed problem is particularly
interesting. Oh, yes: and it would remain as interesting however much
good work I stand to learn of in answer to my question, if anyone can point
me to it. Any of that would be a special bonus for Ontology
Chemistry.
Hopefully,
Christopher
----- Original Message -----
Sent: Sunday, January 29, 2012 8:59
PM
Subject: Re: [ontology-summit] [BigSystems
and SystemsEngineering]Systemofsystems
For example three, now consider:
--- "system of (X)"
--- "part of
(X)"
Where X can be , laws, games, airplanes, cars, plants.. and so
on..
The "system concept" may be viewed as a real world
relationship that is used to order or constrain the
environment.
Using this basic view, two types of definitions for a
system can be constructed as well as two main types of activity for system
concepts.
The two definition types are, function (rule) and
constructive (rule).
The two main activity types are discovery and
design.
The functional rule definition for a system was given
previously and is restated here, "A system is a constraint on variety,
where the constraint identifies and defines the system of
interest."
The construction rule definition for a system is, " A
system is a non-empty set of objects and a non-empty set of
relationships mapped over these objects and their
attributes."
Humans tend to use the concept of a system for two
main activities:
--- Discovering, documenting
and discussing natural systems (systems not constructed by
man).
--- Designing, documenting and discussing
artificial systems (systems constructed by man).
Johannes Kepler's
laws of planetary motion that describe the behavior of solar system under
the defining constraints of natural physical forces is one example of
using the system concept in the discovery mode.
The Wright brothers
are an example of the application of the system concept used in the design
mode.
These modes of application have different approaches, methods
and techniques.
Mixing these modes may generate a high degree of
semantic conflict.
Have fun,
Joe
On Sun, Jan 29, 2012 at 3:36 AM, Christopher
Spottiswoode <cms@xxxxxxxxxxxxx> wrote:
Joe, Anatoly,
You both make very useful points.
Here I highlight just 2 of them:
AL:
> This
ontologizing-in-the-large lead to your need to define not only
>
ontology-as-algorithm but also communication protocol between
ontology
> components that reside in different nodes. I doubt that
mantra about
> "federation" is helpful here. If you have web
programming (that is in
> essence programming-in-the-large) you
speak not about "federating" of
> web-server, load balancer,
database, web-page generation, ad banner
> importing, etc. but
have another engineering approach (while all that
> software
developed by different organizations and reside on different
>
computers).
As I shall be describing in some detail later,
appropriate architecture
leads to good 'Separation of Concerns',
hence reliable and flexible
application modularity while also
enhancing the various other qualities
usually sought. That is
what a properly ontology-based architecture
should of course produce,
and "federation" is a good word to describe
the result at the
in-the-large level.
In contrast to what I shall be describing,
the conventional web
programming you highlight is
complication-inducing rather
than
complexity-respecting
JS:
> I suggest that the "binding force" or "binding
concept" that forms a
> number of items in to one entity is
a key feature.
Yes! That is indeed most strongly the
case in the architecture I shall
be describing (or trying once again
to describe, lessons hopefully
having been learnt...).
All of
which recalls that now very mainstream IS programming precept:
Larry
Constantine's "high module cohesion with loose module coupling".
We
don't have to reinvent that wheel.
> Have fun,
>
>
Joe
Yes thanks, Joe, we sure will!
Christopher
--
Joe
Simpson
Sent From My DROID!!
_________________________________________________________________
Msg
Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community
Files: http://ontolog.cim3.net/file/work/OntologySummit2012/
Community
Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2012
Community Portal: http://ontolog.cim3.net/wiki/
_________________________________________________________________ Msg
Archives: http://ontolog.cim3.net/forum/ontology-summit/ Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/ Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxxCommunity
Files: http://ontolog.cim3.net/file/work/OntologySummit2012/Community
Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2012 Community Portal: http://ontolog.cim3.net/wiki/
_________________________________________________________________
Msg
Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config:
http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community
Files: http://ontolog.cim3.net/file/work/OntologySummit2012/
Community
Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2012
Community Portal: http://ontolog.cim3.net/wiki/
|