Rich,
You overstate the relationship between SE worlds and reality. The entire relationship between any SE world and “really exists” is the assumption that the variable
values of the SE world have a counterpart in the actual/real world.
In those cases where the software is connected to a sensor, the variable value mirrors some measurement of some aspect of reality, and thus supports at least
that ontological commitment. Similarly, software connected to an actuator causes the variable value to have some relationship to reality. The firm belief in these ontological commitments allows us to drive partially automated vehicles and fly on commercial
airliners and react to appointment notifications from our cell phones. (And these sensor/actuator devices can fail, causing a disconnect between the value and the reality, as recently occurred in an airline disaster. )
Having spent some part of my life developing simulation models of what could be and what might be true of them, however, I am convinced that many SE worlds
have no relationship to the actual/real world at all. For example, one model of long ago predicted aspects of a hypersonic shock wave, except that for a wing angle of attack just outside a certain value, the mathematical series failed to converge. The real-world
explanation was “wing stall”, and the mathematical model was invalid in that case; there was no relationship between the variable values and the reality.
Many real-world problems are created by the false assumption by someone in a position of responsibility that a particular software data set is faithful to reality.
Data entry errors, software implementation errors, and misapplied models are among the reasons why routinely making such ontological commitments can create havoc. As in logic, the commitment goes beyond what the software itself can express or validate.
-Ed
From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx]
On Behalf Of Rich Cooper
Sent: Monday, October 19, 2015 2:37 PM
To: '[ontolog-forum] '
Subject: Re: [ontolog-forum] Existential Commitment in Logic: a Tabula Rasa
Tom and Pat,
You (both) wrote stuff leading up to Tom's statement:
3) I think Pat offered a distinction between "exists" and "really exists". This is a classic move to avoid accepting the conclusion
of a specific syllogism; break the middle term in two.
In the Software Engineering (SE) ontological OO world, the variables correspond to your quantifiers in logic. Whether any of those variables influence reality
(e.g. through an interface to a sensor/effector), has little to do with the SE ontology EXCEPT as needed to interface the device drivers to said real world.
He makes a felicitous terminological suggestion: associate "ontological commitment" with "really exists", and commit to it by explicitly
using "really exists" as a predicate. (I think this commits Pat to Anselm's Ontological Argument, but I'm not sure.)
That's a very effective solution in the Database world where you could devote a column in the Entities table to store the predicate's value each time the row
is changed and the predicate is reapplied. Low cost, good performance for most uses.
The other term he offers is "referential commitment", which he associates with the "exists" that logicians tie to the "Some" quantifier.
And he explains referential commitment as ontological commitment relativized to a specified universe of discourse.
Likewise, in the SE world, all commitments relate to the calculation of those state variables.
That's a very nice move, and the explanation is persuasive. But it doesn't work for ordinary un-typed FOPL, since the universe of discourse
for that logic (I think) is [Everything that Exists and Nothing that Doesn't].
I.e., that universe of discourse could be calculated as everything with the predicate column value = True, and nothing where the predicate column value = False.
However, perhaps when we talk about dragons, in natural language, Gricean conventions lead us to a presupposition of a universe which
is not the "real" one", but is rather the real one supplemented with some fictitious entities (e.g. Tolkien's universe).
Fantasy and imagination, as well as creative alternatives, are what makes logic useful. Without it, you may as well use a video.
Sincerely,
Rich Cooper,
Rich Cooper,
Chief Technology Officer,
MetaSemantics Corporation
MetaSemantics AT EnglishLogicKernel DOT com
( 9 4 9 ) 5 2 5-5 7 1 2
http://www.EnglishLogicKernel.com
On 10/19/2015 11:50 AM, Thomas Johnston wrote:
I accidentally deleted the entire thread, including your responses to my comment of the 18th. So please re-post them. Nonetheless, this gives me a chance
to jump out of the details, and try to see what the big picture looks like, at this point in the discussion.
1) Quine and most other logicians are find with associating ontological commitment with the existential quantifier. Arthur Fine (an outlier on this issue,
I would think) prefers to associate ontological commitment with the universal quantifier. But those quantifiers, as quantifiers, are associated with "All" and "Some", and so calling the latter the "existential quantifier" begs the question I raised.
2) Sometimes, in natural language, we want to say "All ...." in a context in which we presuppose, and want and believe others to understand that we are
presupposing (cf. Grice), the existence of what is referenced by the NP of the statement. At other times, we presuppose that the reference of the NP does not exist, and want and believe others to understand that we are presupposing that the reference of the
NP does not exist. Same thing for "None ..." statements, same thing for "Some ..." statements, same thing for "Some ... not" statements. It was my belief that we want to be able to say all these things in natural language that led to my original question.
3) I think Pat offered a distinction between "exists" and "really exists". This is a classic move to avoid accepting the conclusion of a specific syllogism;
break the middle term in two. He makes a felicitous terminological suggestion: associate "ontological commitment" with "really exists", and commit to it by explicitly using "really exists" as a predicate. (I think this commits Pat to Anselm's Ontological Argument,
but I'm not sure.) The other term he offers is "referential commitment", which he associates with the "exists" that logicians tie to the "Some" quantifier. And he explains referential commitment as ontological commitment relativized to a specified universe
of discourse. That's a very nice move, and the explanation is persuasive. But it doesn't work for ordinary un-typed FOPL, since the universe of discourse for that logic (I think) is [Everything that Exists and Nothing that Doesn't]. However, perhaps when we
talk about dragons, in natural language, Gricean conventions lead us to a presupposition of a universe which is not the "real" one", but is rather the real one supplemented with some fictitious entities (e.g. Tolkien's universe).
Thomas,
It is best that an ontology is defined for a specific use. There is a vast difference between the database world and the otherwise closed or fully open world. In fact, every existence has some scope associated with it. It is just that it gets ignored or we
assume that it is such and such. The "really exists" is appropriate for an open world application which should also allow for speculative, fictional and fantasy contexts. That is why, in Euro/US culture, we tell children "Once upon a time...". We are defining
the context to make it clear and easier for the child or reader to understand the setting.
If you say "really exists" you are contrasting it with some other context or scope and the other(s) should be defined. And "really exists" is from a perspective and therefore subjective, and should have a frame or facet disclosed as part of that definition.
If the context is not defined for the ontology then we should assume it is...what? Undefined, defined for closed world, from an unknown POV, or for a fully open world? If there is not a specified definition then we should not be surprised if there are miscommunications.
-John Bottoms
Concord, MA USA
4) So I'm ok with this: NPs have referential commitment no matter what universe of discourse their statements are relativized to. But NPs have ontological
commitment relativized only to universes of discourse which contain as referents things we believe really exist. This feels like an elaborate tautology to me, but it's what I make of Pat's distinction between referential and ontological commitment, and I don't
know any non-tautologous way of stating it.
5) Referential commitment is made with referring expressions such as names. It is no more closely tied to the "Some" quantification than it is with the
"All" quantification. This seems obvious to me but, I believe, does not seem obvious to either of you. But intuitions, expressed as "seems obvious" or otherwise, are not arguments. FOL, as John said, ain't broke. OK. But I'd add that FOL isn't a Swiss Army
knife. It has limitations, illustrated by the extensive work done on various forms of intensional logics.
6) Prior to Pat's referential vs. ontological commitment distinction, I said that I thought ontological commitment could be made with specific uses of the
"All" quantifier, and withheld from specific uses of the "Some" quantifier. Now I will use Pat's distinction and say: referential commitment can be made with any NP, regardless of whether "All" of "Some" is the quantifier used with it. But this is to rephrase
my original intuition, so I suppose that the intuition you both share is different.
7) I had things I wanted to say about Pat's reply last night. One I will mention here is that, in my original quotation from the SEP Ontological Commitment
article, I was not trying to pull the wool over anyone's eyes by leaving off the phrase "According to the third objection to sufficiency". I realize how lonely it is, where I stand on this, and never meant to suggest anything else.
8) I am grateful for this extended discussion. I now have a much better understanding of why I asked the original question, and where in the beautiful labyrinth
of formal logic I might end up (at the Minotaur?) if I continued down the path of separating commitment from quantification.
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: 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 join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J