Chris, (01)
CP> In other words, "Someone who holds 4D position will find an
> extensionalist position more congenial". This is a point often
> made in the literature, and one that seems sensible to me. (02)
Fine. I just wanted to make it clear that I hadn't agreed to
something that had included the term 'congenial'. (03)
CP> By "extensional terms" do you mean (a) by extensional definitions
> or (b) by an extensional criteria of identity. There seems to be
> confusion about these terms, so can we agree to try and be clear
> which we are using? (04)
Although I'm familiar with nonconstructive definitions in mathematics,
I have strong a preference for constructive ones. And when it comes
to ontology, my preference is extremely strong: any definition that
cannot be used by a human being to recognize instances is useless
for designing an intelligent system. (05)
Some people complain that I am "confusing" ontology with epistemology.
My response is that I want to (a) understand human psychology and
(b) simulate something like it in a robot. For those purposes, a
definition that does not support a clear-cut method for identifying
instances is useless. (06)
CP> I and lots of others have been happily working on ontologies of
> contracts, money, etc. for decades using an extensional criteria
> of identity without any problems. Lewis, Heller, Sider and other
> philosophers working in this area (also for decades) have not had
> any problems with these kinds of entities. (07)
I have read such things, and I have a very serious problem with them:
They can only be tested by God, who can not only view all instances
spread out for all times and places, but who can peer into the hearts
and minds of the actors to detect their innermost intentions. I do
not consider any such definition to be a serious proposal for any
computable application. (08)
CP> You earlier made an argument that said as money, etc. were
> extrinsic properties they were intensional - which I pointed
> out is incorrect. (09)
No. My objection was to defining intentional concepts by extension.
Many extrinsic properties are easy to define by observable criteria
-- for example, the "upper" berth of a bunk bed. (010)
But what counts as money depends on intentions, which are often
difficult, if not impossible to detect without peering into the
hearts and minds of the actors. A list of instances cannot, in
general, show observable criteria for the intentions of the actors. (011)
JFS>> Physicists have been doing such transformations for the past
>> century. Just look at the light cone, which maps the 4-D view
>> to or from a view that can be expressed in terms of English tenses. (012)
CP> And if it was so easy and presumably reproducible, why has
> no-one formally mapped a 4d to or from a 3d ontology? (013)
My statement was about mapping a single "view". Mapping an entire
ontology (a system for describing all possible views) is a much
bigger problem. People have found it hard to align large ontologies
of any kind. Furthermore, there is no funding agency that is paying
anybody to map ontologies from 4d to or from 3d. But if somebody
were spreading around millions of euros, a lot of people might work
on the problem. (014)
CP>> Take two spatiotemporal points in a four-dimensional manifold,
>> ascribe to both exactly the same mass density, then you have
>> (on my view) two 4D-points, one property, and two property
>> instances. (015)
My solution is to describe the situation in some version of logic.
For example, I'll use ordinary predicate calculus: (016)
(E p1)(E p2)(point(p1) & point(p2) & loc(p1,x1,x2,x3,x4) &
loc(p2,y1,y2,y3,y4) & mass(p1)=mass(p2)). (017)
CP>> For some 4D extensionalists, you only have two 4D-points (also
>> called property instances) and one property. So we have two
>> different (mainstream) views to reconcile. (018)
My view is to prefer the logic as the primary metalanguage, rather
than any talk about properties and instances. The difference is
a quibble about how to map the formula into English: (019)
a) "the mass of p1 is equal to the mass of p2" (020)
b) "p1 and p2 have the same mass" (021)
I prefer choice (a), which agrees with your preference. But since I
consider the logical formula to primary, I would dismiss the argument
as a trivial quarrel about how to map a formula to English. (022)
Note, by the way, this so-called 4D view is isomorphic to a view that
calls x4 and y4 "times". (023)
In any case, I'd like to mention that I actually prefer a 4D ontology,
but I would never consider any single ontology ideal for all purposes. (024)
John (025)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/uom-ontology-std/
Subscribe: mailto:uom-ontology-std-join@xxxxxxxxxxxxxxxx
Config/Unsubscribe: http://ontolog.cim3.net/mailman/listinfo/uom-ontology-std/
Shared Files: http://ontolog.cim3.net/file/work/UoM/
Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?UoM_Ontology_Standard (026)
|