[Top] [All Lists]

Re: [ontolog-forum] Model or Reality

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Ed Barkmeyer <edbark@xxxxxxxx>
Date: Thu, 16 Aug 2007 12:02:20 -0400
Message-id: <46C4750C.1080806@xxxxxxxx>
I only want to take up John Sowa's point about building bridges.    (01)

 > Engineers regularly build bridges that don't fall down.    (02)

This is so because we have codified "knowledge" about the mechanics of 
structure that has been validated mathematically and by physical experiment, 
and has been accepted as "generally perceived truth" since about 1860.  We 
also "know", from Einstein's work, that that codified knowledge is not 
entirely accurate, but it is accurate enough for the mechanics of bridges.  It 
is not "reality"; it is a *model* of reality that has been proven to work.  In 
fact, we can now reliably bound the difference between the model and the 
possible realities it represents.    (03)

 > Sometimes engineers build bridges that do fall down.    (04)

This is so because there is more to the stability of a bridge than the 
mechanics of its structure.  There is also the quality of the building 
materials and the nature of the terrain on which the bridge is being built, 
and the behavior of that terrain in times of varying natural phenomena. 
Bridges fall down because engineers don't always know enough about the actual 
materials and terrain and climate.  And the famous Seattle bridge disaster 
occurred because of an acoustical phenomenon -- the wavelength of the 
vibration of the bridge in a high wind -- that had never previously been 
studied.    (05)

So bridges stand because we have a certain amount of useful "knowledge" and 
they fail because we are not omniscient.    (06)

When we build ontologies for public use, we have a responsibility to codify 
the knowledge that has been validated by theory and experiment, and to label 
that knowledge as such.  It may not be "truth", but it represents a level of 
comprehension of our world that human experts accept and use, and we can 
hardly do better than that.  Ontological engineering is not epistemology, and 
it is not metaphysics.  But we do need a means of separating the "good" models 
(that generally produce results that can be validated by experiment) from the 
"bad" ones (that often produce nonsense).    (07)

Finally, the bridge-building example teaches us that no ontology or 
combination of ontologies, no matter how well-founded, can be guaranteed to be 
*sufficient* for any given task.  What you don't know can *always* hurt you.    (08)

-Ed    (09)

P.S. The World Trade Center is another example of the success of accepted 
knowledge and the disaster from the missing information.  The impact of a 
jetliner at 480 knots did not bring down the twin towers -- one of the 
buildings swung 6 degrees off of vertical from the impact, but because it had 
been built to withstand earthquakes and hurricanes, it swung back to upright! 
  What brought the towers down was the fact that the particle wind from the 
disintegrated aircraft stripped the heat insulation off the lateral supports, 
and the heat from the slow-burning office furniture then weakened the supports 
-- a combination of bizarre factors for which we only made the predictive 
model after the fact.    (010)

Edward J. Barkmeyer                        Email: edbark@xxxxxxxx
National Institute of Standards & Technology
Manufacturing Systems Integration Division
100 Bureau Drive, Stop 8263                Tel: +1 301-975-3528
Gaithersburg, MD 20899-8263                FAX: +1 301-975-4694    (011)

"The opinions expressed above do not reflect consensus of NIST,
  and have not been reviewed by any Government authority."    (012)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (013)

<Prev in Thread] Current Thread [Next in Thread>