SUO: Re: Ontology, Epistemology, Semiotics
> Pierre, Murray, Gary, et al.,
>
> What I'm trying to do in this thread is, in Wittgenstein's
> terms, to present a series of reminders of what everybody
> knows, but not everybody has considered important enough
> to mention, much less to adopt as the foundation for
> ontology and related issues.
So you are proposing to send platitudes to two mailing lists. I doubt
Wittgenstein had this concept.
[...]
> I believe that without
> a solid semiotic foundation (in Peirce's sense) we cannot
> develop a satisfactory SUO.
I don't know much about Peirce and your parenthetic comment makes me doubt.
When I say semiology is irrelevant to upper-level ontology, does that cover
what's relevant to semiology in Peirce then? Just checking.
> PG> This project is a a worthy endeavour in a narrow
> > subdomain of theoretical KR.
>
> I'll keep piling up examples that show it is central to
> any SUO that could ever be practical enough to support
> applications (which is the ultimate goal of the SUO project).
Unfortunately the two examples you piled up so far come from the narrow domains
of (ontology of?) experimental activities and of engineering practices (boring
subdomains of social ontology) where apparently there are intends and
purposes...
I wonder if your claim is that we need a social ontology module in SUO.
[...]
> PG> There are other domains: basic subdomains of biology and
> > medicine (not only human of course), anatomy, physiology;
> > chemistry; molecular biology; physical geography and its
> > subdomains; physics and its subdomains, particularly applied
> > in engineering sciences; mathematics, arguably; and so on and
> > so forth.
>
> Yes. And my point is that every experimental and applied
> subject (i.e., with the exception of pure math or pure
> theoretical physics) cannot be defined without introducing
> the purposes and intentions of the agents who do the
> experiments or the applications.
This may be what you are repeating, all right. There is more than
experimentation to these domains. And there is more than intend and purpose in
an ontology of these domains (you seem to black out completely on physical
causality). Or rather, there actually is no such qualm to have before you
introduce agents and this preliminary introduction is your basic petitio
principi.
> In chemistry, for example, there are terms from physics,
> such as "hydrogen" and "energy". But you cannot go
> any further without introducing words like "solution",
> "precipitate", or "result" -- which cannot be defined
> without considering the intentions of the agent who
> is dissolving, precipitating, or producing results.
Ontology of chemistry is not a conceptualization of laboratory activity. There
are precipitations outside laboratories, everyday in my toilet bowl for
instance. And indeed there is more to chemistry than mixing.
> In medicine, the entire subject is based on the normative
> assumption that being healthy is better than being sick
> or dying.
This is old school stuff I think. Now we have a relativist medicine which has
shown death to be normatively bad only from certain perspectives.
> Every aspect of every drug or medical procedure
> is determined by what it does to promote those goals.
This sounds like a gratuitous claim and it's probably false.
> You
> might be able to describe the physical structure of some
> drug without considering its purpose. But you can't talk
> about the relationship of a drug to a disease or of the
> relative efficacy of different drugs without considering
> purpose and intention.
Drugs do not have intend nor purpose, unless there is an agent to stick them up
their ass. You can talk nearly as much about the relationship of a drug to a
disease as you can of the relationship of a disease to its causes without
worrying about these.
> I'll skip to the end of your note, which gets to the
> heart of the issue:
>
> JS> They are the deciding factors in determining what to
> > represent and how to represent it.
> >
> PG> Even there, from a representation point of view, you
> > don't need a sophisticated exegesis of social reality,
> > suitability can probably be a stand alone relation.
> > It's not always clear that you care for expliciting what
> > suitable for a purpose means.
>
> Notice that you want to replace my words "purpose and
> intention" with another word "suitable", which is just
> a disguised synonym for one aspect of "purpose".
Is there a word then which isn't a disguised synonym for one aspect of
"purpose"? Incidentally, this is a general treat of your approach. Terms are
meaningless and concepts completley useless which apply without restrictions. I
already made this remark discussing a putative predicate SocialEntity which you
would seem to apply to everything and nothing.
Now back to your remark. I had no intend nor purpose of replacing your terms.
In the part you sniped from my email I claimed that the concepts you find so
pervasive do not apply but that that of suitability is suitable... I introduced
suitability in terms of constraint problem solving (in a physical context).
> How many "stand alone" synonyms for purpose do you want?
You are missing the point. I am speaking about ontology, not worried about
language. No need for synonyms, I'm interested in entities (substitute
'concept' if needed).
To the question, how many stand alone entities/concepts? It depends on your
purpose as a knowledge engineer.
> In chemistry, you have "result". In medicine, you have
> "efficacy". In engineering, suitability means that the
> "results" "satisfy" the "specifications" within the limits
> of some "criterion". And you can't define specifications
> or criteria without talking about the people whose
> purposes determine the spec's and the criteria.
See the part you sniped in my previous email. There is more to all of these
than testing and experimenting! That's precisely why people test and
experiement.
> Another example is Matthew's recent 4D ontology, which
> has words like "state" and "system", which he and Jon
> have been trying to define. But those words also have
> disguised purposes built into them:
>
> 1. A "system" is an "integrated" collection of
> interacting parts "designed" for some "purpose".
Recall that Matthew's vocabulary comes from Process Engineering. My urinary
system was not designed and it has no purpose, although it may have a function.
Please, don't tell me purpose and function are synonymous, you'll soon break my
long lasting belief in the richness of your mother tongue.
> 2. A "state" is a time period during which the
> "relevant" aspects of some "system" do not
> change "significantly".
>
> All the quoted words in these definitions and in the
> previous paragraphs cannot be defined independently
> of the purpose or intentions of some agent.
Obviously in this context, since 1. defines systems and mentions 'purpose' and
2. mentions 'system'. Still in this context, I agree with respect to 'designed'
and 'purpose'. 'Relevant' is unclear, sorry I can't bother. 'Significantly' can
be indexed to a scale. I wish you wouldn't reply that scales can't "be defined
independently of the purpose or intentions of some agent".
Pierre
--
Pierre Grenon, IFOMIS Uni Leipzig
http://people.ifomis.uni-leipzig.de/pierre.grenon/