RE: SUO: Exposition
Dear Lee,
> On Wed, 11 Apr 2001, West, Matthew MR SSI-GREA-UK wrote:
>
> > Dear Lee,
> >
> > >
> > > On Thu, 5 Apr 2001, West, Matthew MR SSI-GREA-UK wrote:
> > >
> > > >
> > > > Dear Lee,
> > > >
> > > > > Matthew,
> > > > >
> > > > > Since I do not view your position as philosophically or
> > > > > logically adequate
> > > > > to the full range of human experience,
> > > >
> > > > MW: I note this. What I lack is evidence or proof that
> your view is
> > > > correct. If you can identify where the position is
> inadequate and
> > > > which ranges of human experience it does not apply to
> then we might
> > > > make progress.
> > >
> > > There was a Polish philosopher named Tadeusz Kotarbinski who took
> > > something like your position to great lengths. He variously
> > > called his
> > > position reism, pan-somatism, concretism, semantic reism.
> >
> > MW: I'm not familiar with this person, but I support his approach of
> > taking things to great lengths. This way you see if it breaks, and
> > where, and then you know the limitations.
> >
> > > Tarski, insofar
> > > as he bothered to work out a metaphysical position, acknowledged
> > > Kotarbinski as his master and his early and famous work on
> > > truth in formal
> > > systems is couched to make place for a reist approach.
> > >
> > > I once spent a good bit of effort to show how this position,
> > > which is not
> > > my own, can be stretched to cover just about anything,
> > > including ethics
> > > and aesthetics. But it has one great limitation that may have
> > > some bearing
> > > on an SUO effort. It covers experience after the fact, after the
> > > experience has become manifest in some spatio-temporal
> phenomenon.
> >
> > MW: This is no longer true when you introduce a possible worlds
> > approach which allows you to consider possible futures, either
> > hypothetical or intended, as well as possible pasts, which may be
> > linked to or detatched from the present.
>
> Not quite right, since the notion of a possible world (that is,
> possibilities made manifest in a hypothetical spatio-temporal
> ensemble)
> and (to use Kotarbinski's early contribution to possible
> world semantics)
> of an "empty name" as what computer people might call a 'wild card'
> standing for such a world is something short of the
> philosophical category
> of modality. Modality in its best developed sense is the notion of
> possibility detached from (or better, prescinded from) time
> and space.
MW: We used to have a plethora of abstract individuals in our models.
Our analysis has shown that either these can be represeted by possible
worlds, or by classes. Being detached from time and space suggests that
we would treat them as classes. Some specific examples of what you mean
would help to determine this, or whether you are talking about
something we have not come across.
> The view you espouse is workable but more limited; in it
> possibility is
> postulated to be always conjoined with some space-time instantiation
> (which may be placed hypothetically in the past, present or
> future) rather
> than seen a fundamental category in its own right. Possible
> worlds theory
> is certainly an advance over the attempt to absorb possibility into
> potentiality, but it is a step short of a view of multiple, indeed,
> infinite possibilia for each and every sign.
MW: What I have seen of modality has left me singularly unimpressed,
however, you may easily be talking about a version of modality I am
not familiar with. However, my standard question you will be expecting
by now is please give me a specific example of what you think can be
expressed using modality that cannot be equally well expressed using
possible worlds.
>
>
> >
> > > It
> > > thus sees little point in asserting experience, human or
> > > otherwise, as a
> > > broader category than empiria. Thus in your recent exchange
> > > on 'giving'
> > > with Chris Menzel, you cover giving after an act of
> giving has become
> > > manifest, while Chris would like to assert some ternary
> aspect to the
> > > logic of giving, or in natural language to the dative case,
> > > whether or not
> > > it is made manifest in an activity.
> >
> > MW: You can express as many aspects to giving as you wish.
> Whether you
> > use binary or ternary relations is largely a matter of how
> you want to
> > view things, not what you are viewing.
>
> If your distinction between the 'how' and the 'what' of the
> view of giving
> is designed to highlight the issue of meaning, then I agree.
> If I see two
> people grasping an object and releasing it from one to the
> other, that, I
> take it, is the 'what' you are referring to. Viewing this
> 'what' as an
> act of giving is the 'how' I might view it. Similarly, two people
> emitting sounds at each other and responding mutually in body
> language in
> a kind of dance of the eyes, gestures and facial expressions,
> is a 'what'.
> Viewing this as a "conversation" is a 'how'. If this
> captures what you
> have in mind, then my argument is that a 'how' that captures the
> meaningfulness of the act to the actors requires some tertium
> quid beyond
> the physical dynamic.
MW: From where I stand you are just talking about a couple of activities
that conform to a particular pattern that marks them outr as members of
a class. Is there anything else here?
>
> >
> > >
> > > Now I would not argue that you fail to 'cover' giving,
> since you can
> > > provide a verbal account of each and every instance after it
> > > occurs, and
> > > once an act has occurred to every 'extension' of it in the
> > > sense you use
> > > this term, but I would argue that there is something about
> > > the logic of
> > > the dative case that is missing in your view of it.
> >
> > MW: Our model can deal with an arbitrary number of things involved
> > with giving. We simply choose to represent them in a binary form
> > (from a KIF point of view). It can then from this base of
> information
> > support views that might be ternary or what ever is wished for.
> >
>
> Yes, I agree: your key term here is "support views". What is actually
> happening is that the ternary view is added by the user, who
> is supplying
> the element of interpretation that the binary structure
> leaves out. It is
> the user's "view" that is being "supported."
MW: We work particularly hard to capture these things. Our approach
to context information is to capture it explicitly, and relate it to
the objects concerned. This is aided by our treatment of relations
as first class objects, i.e. objects that can be referred to in
other relations.
>
> > > And I
> > > would also argue
> > > what is missing in the binary view of giving is also missing
> > > in the binary
> > > treatment of other concepts crucial to an account of human
> > > institutions,
> > > including the institution of science.
> >
> > MW: Some examples rather than assertions would be really helpful.
>
> See the addition of conversation above, which may be
> generalized to the
> example of language as such, and on the micro-level to the example of
> representation as such.
MW: So far you have only talked in generalities, I would really like to
get you to the "John said to Mary ..." level. Until we are talking about
specific examples of things that can or cannot be expressed, I'm not
going to be convinced.
>
> > >
> > > With respect to a computational ontology to be used in
> > > conjunction with a
> > > relational database, my argument may make no practical
> > > difference so long
> > > as we are interested in using the ontology to search and
> mine what has
> > > already been achieved and certified in some sense as
> > > information. But to
> > > the extent that we may wish to use an ontological artifact
> > > for sciences of
> > > discovery or for deliberation, which contain a purely
> > > hypothetical or even
> > > imaginative element, and in which we wish to be
> especially alert to
> > > emergent categories and to the repurposing and
> reconceptualizing of
> > > information for unanticipated needs, we may want to provide
> > > architectonically for a position that concedes a degree
> of reality to
> > > concepts without grounding them dogmatically in some 'concrete'
> > > manifestation and its extensions.
> >
> > MW: As stated above all this is supported through the
> possible worlds
> > approach we incorrporate, as stated above.
>
> And as replied above, the possible worlds approach still contains the
> concretist grounding but attempts to correct its obvious
> limitations by
> detaching it from the actual world to hypothetical worlds and
> quasi-worlds. Within a certain range this will 'work'.
MW: Again I am looking for specific examples that do not work of the
"John said to Mary ..." variety.
>
> >
> > MW: Also think about how likely it is that I would be able
> to have a
> > position that was not capable of lookihng forwards as well
> as backwards.
> > Shell is one of the most successful companies in the world,
> and has been
> > continuously for about 100 years. One of the reasons for
> this is its
> > expertise in planning ... What chance would I have internally if I
> > couldn't support that?
> > >
>
> Shell is certainly a great and successful company, which has had to
> combine economic, financial, geopolitical, legal, personnel,
> engineering
> and scientific considerations, among others. It is a complex world in
> itself and anything that works within it shows good prospect
> of working
> internally for other modern institutions constructed along
> similar lines.
>
> But 100 years is not a very impressive span of time in these
> matters. If
> Shell exists as a distinct institution 100 years from now, there is no
> guarantee that it will be animated by a mode of thought based
> on the years
> 1901-2001.
MW: Well since most current modes of thinking are based on 2000 year old
ideas I don't have a huge expectation of change over the next 100 years.
Anyway, how do you expect us to cater for modes of thinking that are
not in practice now?
>
>
> > > I am, of course, able to participate cheerfully in an
> > > ontology constructed
> > > on principles in the general family to which your
> position belongs, so
> > > long as the purpose of the effort is clarified to be limited
> > > to uses for
> > > which this position is appropriate.
> >
> > MW: One reason I am pursuing this discussion is that if there really
> > are things it cannot handle, I would really like to know.
>
> The key is how far one can rely on the shared language and mutual
> recognition of purposes of a collection of users to supply the missing
> interpretative element. To the extent one can rely on a
> *community* of
> users, much can be left tacit, and we need not provide
> architectonically
> for computational elements that community members can supply better on
> their own. But giving some computational form to the tertium
> quid will
> prove increasingly relevant to the extent that we wish to automate
> mutually untranslatble patterns of interpretation and
> orientation and to
> embody them in agent protocols.
MW: I would hope to get to a point where everything can be defined
either by formal axioms, or by pointing, both of which I consider
to be adequately unambiguous to need no further qualification. My
intuition tells me this should be possible, but I await the outcome.
>
> >
> > > But if, as in our current effort,
> > > there is strong resistance to narrowing the purpose,
> >
> > MW: I do not wish to rule out other view points. If you
> read up on the
> > IIDEAS architecture you will see that I think it is important to be
> > able to support a range of different viewpoints for
> different purposes,
> > but that also I want a view that is able to
> support/integrate all of them
> > (if possible - I recognise that it may not be).
> >
>
> I was not referring to any resistance on your part-- to the
> contrary, you
> have been by my lights a very reasonable voice on this
> issue-- but to the
> persistent failure of those in greater control of this
> process to show an
> openness to the repeated requests from many other
> participants for greater
> clarity in purpose and attendant methodology.
MW: Well I could not agree more. I am happy enough about
purpose, but until we have a methodology we have no chance
of engineering a solution. To me it therefore follows that
any ontology development work at present is about
getting the methodology developed.
>
> > > I would
> > > counsel that
> > > what I would regard as a more capacious and open-ended view
> > > not be ruled
> > > out prematurely.
> >
> > MW: And what would that be?
> >
>
> Though this is not really an email topic, let me try a quick, somewhat
> fanciful example.
>
> Suppose we drew a line and cut it at an arbitrary point, and
> at the cut
> there exploded an infinity of possibilities not only for resolving the
> point into a continuum of infinitesimals, but for the 'how' (in your
> sense) of viewing each infinitesimal. Is it useful to construe one as
> opposed to another infinitesimal variation as defining a distinct
> 'possible world'? Or would we want a less inflated vocabulary for
> handling these variations?
MW: Actually, number of objects required is not something I see as a big
issue, compared to precision and consistency. So I have no problem here
with there being a possible world with each possible outcome you're
interested in. Of course I might not present it to a user that way where
some context has been established.
>
Regards
Matthew
===============================================================
Matthew West http://www.matthew-west.org.uk/
Principal Consultant Shell Visiting Professor
Operations & Asset Management The Keyworth Institute
Shell Services International The University of Leeds
http://www.shellservices.com/ http://www.keyworth.leeds.ac.uk/
H3229, Shell Centre, London, SE1 7NA, UK.
Tel: +44 207 934 4490 Fax: 7929 Mobile: +44 7796 336538
===============================================================