Re: SUO: Functional Roles
My ears were burning....
Nicola and I used a notion of "unity" to account for the mass vs.
count distinction.
Nicola, in some previous work that is covered in our AAAI tutorial
(http://www.cs.vassar.edu/faculty/welty/aaai-2000/), proposed "levels
of stratification" for accounting for things like the well known
"castle and the pile of bricks" it is made of. The pile of bricks is
different from the castle, the latter being constituted of the former.
As for this:
At 7:29 PM -0700 9/26/00, Michael Uschold wrote:
>A hunk of granit can be a table.
So, in our account, a table can be constituted of a hunk of granite.
The two things are still different.
Next, it is important to realize that there are many subtly different
ways in which the word table is used. Our techniques can be used to
make more clear which of these you mean.
For example, one view of table is as an artifact that was
specifically created to be a table. Such a table is always and
forever an artifact-table. Even after it is destroyed, it can not
stop being "intended to be a table".
We talked about the notion of "functional unity" which helps to
specify that a table is a "whole" as long as it meets some functional
criteria (ie can still serve as a table).
Being used as a table is certainly a "role", it is a non-rigid
property and many things may be used as a table.
>A table can be a chair (if you sit on it)
So, again, we woudl at minimum have "intended to be a chair" for
chair artifacts, and "used as a chair" for objects filling that role
at a time.
>, a
>table can also be a hammer, if you pick it up and hammer somethign with it.
>I wonder if CHris Welty and Nicola can do some of their ontological analysis
>to see how to model" 'table'. One view is to say it is just a role. Something
>can be a table or not be a table at different times. Is there anythign that is
>alwayas and forever more a table? I'm not sure.
>
>Mike
>
>PS this is no longer on the mass vs count distinction, so I changed
>the subject header
>
>>From owner-standard-upper-ontology@ieee.org Mon Sep 25 23:05:17 2000
>X-Sender: apease@pop.teknowledge.com
>Date: Mon, 25 Sep 2000 21:55:09 -0700
>To: "West, Matthew MR SSI-GREA-UK" <Matthew.R.West@is.shell.com>,
> David Whitten <whitten@lynx.eaze.net>,
> standard-upper-ontology@ieee.org
>From: apease <apease@teknowledge.com>
>Subject: RE: SUO: Mass vs Count distinction
>Mime-Version: 1.0
>X-MDaemon-Deliver-To: standard-upper-ontology@ieee.org
>X-Return-Path: apease@teknowledge.com
>X-Resent-To: Multiple Recipients <standard-upper-ontology@majordomo.ieee.org>
>X-Listname: standard-upper-ontology
>X-Info: [Un]Subscribe requests to majordomo@majordomo.ieee.org
>X-Moderator-Address: standard-upper-ontology-approval@majordomo.ieee.org
>
>
>Folks,
> Maybe we can refine this with a few more boundary cases. A block of
>granite used as a table can be subdivided and still be both granite and
>function as a table. One might take a large steel screw and machine it
>into two smaller steel screws. However, it would be harder to take a
>wooden chair and turn it into two smaller wooden chair.
> I would submit that these examples only cause a problem because we are
>confusing function and identity. What we are really saying is
>
>(instance-of Gensym1 Wood)
>(shape Gensym1 RodShaped)
>
>or
>
>(instance-of Gensym2 Granite)
>(shape Gensym2 RectangularSolid)
>
>or
>
>(instance-of Gensym3 Chair)
>(madeOfStuff Gensym3 Wood)
>
>Of course, this begs the question of when to define a thing as having an
>essential nature as stuff or as object but I can't seem to think of an
>ambiguous case at the moment.
>
>Adam
>
>
>At 12:07 PM 9/22/2000 +0200, West, Matthew MR SSI-GREA-UK wrote:
>
>>Dear David,
>>
>>I'm afraid I have to disagree strongly with your analysis of mass-thing vs
>>countable thing below, as this relates to fusion and assembly. The basic
>>problem is that you have a view point from human scale. Let me try to
>>explain.
>>
>>There are two important types of aggregation (amongst others) that I shall
> >call here fusion, and assembly.
>>
>>Fusion
>>
>>For fusion, an aggregate consists of usually a large number of objects, all
>>of the same type (the type may be broad or narrow).
>>
>>So for example, a glass of water (not counting the glass) is made up of a
>>large number of water molecules. I could count the number of water
>>molecules, but it would be at best tedious, and totally uninteresting.
>>
>>Another example is a bag of screws (not counting the bag). Here it is a
>>screw that is the "atomic element" that the fusion consists of, but again it
>>is unordered.
>>
>>It should be noted that in each case the fusion is made up of things that
>>are themselves "ordered" (e.g. a water molecule, a screw).
>>
>>Assembly
>>
>>In assembly, the parts are arranged in a particular way, that means that the
>>whole is more than the sum of the parts, e.g. a car is something different
>>from a pile of car parts, because of the way they are put together. A water
>>molecule is more that two hydrogen atoms, and an oxygen atom because of the
>>way they are put toegether.
>>
>>Regards
>> Matthew
>>============================================
>>Matthew West
>>Asset Information Management
>>Shell Services International
>>H3229, Shell Centre, London, SE1 7NA, UK.
>>Tel: +44 207 934 4490 Fax: 7929
>>Mobile: +44 7796 336538
>>E-mail: Matthew.R.West@is.shell.com
>>http://www.shellservices.com/
>>============================================
>>
>> > -----Original Message-----
>> > From: David Whitten [mailto:whitten@lynx.eaze.net]
>> > Sent: 21 September 2000 19:29
>> > To: standard-upper-ontology@ieee.org
>> > Subject: Re: SUO: Mass vs Count distinction
>> >
>> >
>> >
>> > John Sowa asked:
>> > > But what about a wooden rod?
>> > >
>> > Michael Uschold had said:
>> > > >Distinction: Mass-Count
>> > > >Proposer: Adam Pease / Mike Uschold
>> > > >Options:
>> > > > X: something that cannot be divided into parts, each of which
>> > > > is the same type/class of thing as the whole was.
>> > > > E.g. A person, a car.
>> > > >
>> > > > Y: something that can be divided into parts, each of which
>> > > > is the same type/class of thing as the whole was.
>> > > > E.g. a quantity of sand, an amount of water
>> > > >
>> > > >Proposed name for Y: Mass-Thing
>> > > >Proposed name for X: Countable-Thing
>> > >
>> > John Sowa continued:
>> > > Suppose I have 5 wooden rods. Then I cut each one in half,
>> > > and I get 10 wooden rods. I can count wooden rods, but when
>> > > I divide them, I get more wooden rods than I started with.
>> >
>> > This is one the reasons I enjoy participating in the SUO mailing
>> > list. We have some very capable people who are willing to ask hard
>> > questions about what other very capable persons have presented.
>> >
>> > Incidentally, Mike's original statement above has been altered
>> > by me to match my understanding (and John Sowa's pg 98 of KR)
>> > of what a Mass-Noun and a Count-Noun refer to. (I swapped X & Y).
>> >
>> > In my opinion, being countable is saying that you can count
>> > the elements in a collection, not that you necessarily want to.
>> > Count-Things can be represented by sets, since you can discern
>> > individual elements. Mass-Things require a collection that
>> > tracks quantity, not identity of the individual elements.
>> > Perhaps we need to expand this format of Mike's to include
>> > this distinction as well as boundary conditions.
>> >
>> > One boundary seems to be on the division into parts operation
>> > mentioned. I can imagine dividing each wooden rod into parts,
>> > each of which would be a wooden coin.
>> > In this case, the operation would be taking something I thought
>> > of as a Mass-Thing (a specific group of wooden rods)
>> > and producing another Mass-Thing (a specific group of wooden
>> > coins) but these two groups are not of the same type/class.
>> >
>> > Another related boundary is interepreting the division into
>> > parts operation as a division of the group of wooden rods into
>> > two or more subgroups. Clearly, one wooden rod does not make a
>> > group. So this division into parts operation, when applied
>> > to the group as as whole, must produce subgroups, each of which
> > > are the same type/class. Thus this looks like the group of
>> > wooden rods is a Mass-Thing which has parts that are Count-Things.
>> >
>> > Another boundary is the fact that any object made of wood,
>> > when separated into two or more parts, will produce objects made
>> > of wood. This seems to require the group of wooden
>> > rods to be composed of parts, each of which are a Mass-Thing.
>> >
>> > Does this match anyone else's analysis ?
>> >
>> > David (whitten@lynx.eaze.net) (713) 791-1414 ext 6116
>> >
>
>-----------------
>Adam Pease
>Teknowledge
>(650) 424-0500 x571
****NOTE NEW AREA CODE:
Christopher A. Welty http://www.cs.vassar.edu/faculty/welty/
Vassar College Computer Science Dept. Voice: (845) 437-5992
Poughkeepsie, NY 12604-0462 Fax: (845) 437-7498