Re: SUO: RE: 4_d vs. 3-D views
Matthew,
Thank you for your example and comments.
Since you don't object to the coexistence of 3-D
and 4-D views, we seem to be in agreement on that
issue.
I can see that using 4-D is one solution to the
question of how to relate "roles" and the objects that
fill those roles. I would treat the same question
differently in 3-D, but there might be an accurate
translation between the views. I do believe that
ontologists should be allowed to use the views that
they find most useful or easiest to work with, provided
that there are no logical inconsistencies. So I do hope
that good translations between 3-D and 4-D views can be
found. I also commented on this in a reply to the
note of Chris Partridge.
I can't resist, however, commenting on one of your
points:
>
> MW: Let me try an analogy on you. It used to be that the only
> sort of drawing you could have for an engineered object would
> be 2D. With the advent of computers, it became practical to have
> a 3D representation. The 3D representations did not really
> hold much information that could not be held in the 2D
> representations, but some things like clash detection (2 objects
> occupying the same space) became very much easier. You will not
> be surprised to hear that 2D drawings are now little used to
> represent complex physical objects.
>
For product designers with access to CAD-CAM programs, that may be
true. Yet, people still find 2-D views in printed publications highly
useful, and I suspect that even those with access to CAD-CAM programs
use 2-D drawings between each other for some non-trivial exchange
of information. Specifically, when 2-D is adequate for a purpose,
using 3-D may just waste time and effort. More to the point, when a
person **looks** at a physical object, it is the 2-D view (or 2-1/2-D)
that is seen, with no interior parts, and people are used to
comprehending objects with such a mental image in mind. Likewise, when
a person **looks** at an object, it is the instantaneous object that is
in view at any one time point, and that may be why some of us feel that
the 3-D view is more "natural". For myself, I have always been
uncomfortable trying to envision a whole 4-D object when I can only see
a temporal part (the future part is invisible, except in possible
worlds). This is quite different from a computerized CAD-CAM image,
which can have the whole 3-D object in view.
So, I hope that we can have both 3-D and 4-D views for those
who find either more congenial, and perhaps also so for those who
find one more congenial for some types of representation and the
other more congenial for different purposes. I have an open mind on
4-D, I just haven't seen enough specifics to know where I would
find it a better representation than 3-D. I may well decide that
each is better for a different purpose, like 2-D and 3-D drawings.
Pat Cassidy
=========================
West, Matthew R SITI-ITPSIE wrote:
> Dear Pat,
>
> See comments below.
>
>
> Matthew West
> Principal Consultant
> Shell Information Technology International Limited
> Shell Centre, London SE1 7NA, United Kingdom
>
> Tel: +44 20 7934 4490 Other Tel: +44 7796 336538
> Email: matthew.r.west@is.shell.com
> Internet: http://www.shell.com
>
>
>
>>-----Original Message-----
>>From: Patrick Cassidy [mailto:pcassidy@bellatlantic.net]
>>Sent: 25 October 2001 19:57
>>To: Adam Pease
>>Cc: West, Matthew R SITI-ITPSIE; Seth Russell; John.Velman@HSC.com;
>>standard-upper-ontology@ieee.org; Frank Farance
>>Subject: 4_d vs. 3-D views
>>
>>
>>Matthew --
>> I would very much like to learn what issues
>>caused you to choose 4-D rather than 3-D for
>>your ontology.
>>
>
> MW: I have explained this a number of times already, still
> I dare say one more time will do no harm.
>
>> >> Now the bad news:
>> >>
>> >> It took us 8 years to get the top bit sorted out
>> >> in metaphysical terms - eventually being driven
>> >> to a 4D ontology as the only one that could resolve
>> >> the issues that were being raised.
>> >
>>
>> Since both are supposed to represent
>>the real world, I would imagine that either one
>>would arrive at the same inferences from the same
>>data, or else one or both would be simply wrong.
>>
>
> MW: "All models are wrong - some are useful". Or in this
> case more useful.
>
>> In one of your earlier posts you stated:
>>
>> > MW: Yes, if you time index everything to do with
>>individuals you can
>> > get to something which can capture most of the history of
>>things. But
>> > as I have pointed out elsewhere, 3D also invites a "current state"
>> > approach, which I consider somewhat inadequate.
>> I agree that explicit time-indexing of 3-D objects
>>is required for assertions about the real world, unless there
>>is an implicit time-indexing which is included in the axioms
>>defining atemporal predicates on 3-D objects. The feeling
>>that '3D also invites a "current state" approach' does not
>>seem to be an argument that it is inadequate, since
>>it would doubtless be simple to carelessly encode nonsense
>>in a 4-D or any other type of approach.
>>
>
> MW: Let me try an analogy on you. It used to be that the only
> sort of drawing you could have for an engineered object would
> be 2D. With the advent of computers, it became practical to have
> a 3D representation. The 3D representations did not really
> hold much information that could not be held in the 2D
> representations, but some things like clash detection (2 objects
> occupying the same space) became very much easier. You will not
> be surprised to hear that 2D drawings are now little used to
> represent complex physical objects.
>
>> From what I have seen so far, it should be possible to
>>relate 4-D and 3-D approaches, and I am trying to determine
>>how predicates on 4-D views can be related to predicates
>>on 3-D views.
>>
>
> MW: Yes, a 3D plus time approach can always be mapped to a 4D
> approach (if it makes sense in the first place) and usually
> vice-versa. If you would care to give some 3D examples, I
> would be happy to show you how the 4D equivalent can be
> constructed. But the real secret is to learn to think in 4D.
>
>
>>One big problem is in finding sufficient
>>examples of predicates on 4-D views and how predicates
>>relating 4-D objects relate to the temporal parts of those
>>4-D objects.
>>
>
> MW: This is rather easy. If a predicate applies to a
> 4D object, then in general it applies to any temporal part
> of the 4D object. The kind of predicate that is an exception
> would be one that says "This is a whole-life 4D object".
>
> MW: Whatever is asserted for a 4D object must be true for the
> whole period of the 4D object. So if a 4D object is asserted
> to be red, then any temporal part of that must also be red.
>
> MW: The ordinary language that is used to describe this is
> that of state and substate (occurences).
>
>
>> If you can provide examples of 4-D views and why they
>>are better than 3-D views, it would be very helpful to
>>this discussion.
>>
>
> MW: One of the main things that convinced me that 4D is
> better than 3D is in the way it allowed a physical explanation
> of objects that I had not been able to see as physical before.
>
> MW: Ordinarily, in a 3D view, only objects that have
> essential material continuity can be seen as physical objects.
> Other objects, such as "the chairman of Shell" would be seen
> in some sense as abstract - certainly we did. This gave us
> some problems with some apparent contradictions, because it
> was quite clear that "the chairman of Shell" did things, and
> this contradicted the idea of being abstract.
>
> MW: A 4D approach allows you to see that "the chairman of
> Shell" consists of temporal parts of the people who hold
> that position - the parts are just co-incident with the
> two objects. This in turn allows you to see "the chairman
> of Shell" as a physical object, but just one that does
> not have material continuity, and which shares its parts
> with other physical objects. This in turn enables you to
> see that "the chairman of Shell" is some sort of physical
> object, that it is perfectly reasonable to talk about
> "doing" things.
>
> MW: I should say that formally I am not arguing for a 4D
> view instead of a 3D one, but as well as - unless no-one
> else wants to include a 3D view.
>
>> Pat Cassidy
>>
>>=============================================
>>Patrick Cassidy
>>
>>MICRA, Inc. || (908) 561-3416
>>735 Belvidere Ave. || (908) 668-5252 (if no answer)
>>Plainfield, NJ 07062-2054 || (908) 668-5904 (fax)
>>
>>internet: cassidy@micra.com
>>=============================================
>>
>>
>
--
=============================================
Patrick Cassidy
MICRA, Inc. || (908) 561-3416
735 Belvidere Ave. || (908) 668-5252 (if no answer)
Plainfield, NJ 07062-2054 || (908) 668-5904 (fax)
internet: cassidy@micra.com
=============================================