[Top] [All Lists]

Re: [ontolog-forum] Watchout Watson: Here comes Amazon Machine Learning

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: John F Sowa <sowa@xxxxxxxxxxx>
Date: Sat, 02 May 2015 14:37:50 -0400
Message-id: <5545197E.8000009@xxxxxxxxxxx>
Tom and Rich,    (01)

I was giving a brief summary of the article by McCarthy & Hayes (1969),
which analyzed various issues in knowledge representation in AI.
That article was the basis for many years of publications.  I put
some of them in the collection at http://www.jfsowa.com/ikl .    (02)

> The classical definition of a statement, as I understand it, is that
> it is a declarative sentence with all indexicals resolved.    (03)

> Please define the word "indexicality" as you are using it.    (04)

In ordinary language, a statement is any sentence that is stated.
An indexical is a "pointing word" like pronouns and noun phrases
such as 'the president'.  In sign languages, the index finger is
used instead of indexicals.    (05)

> I think you mean the binding between FOL variables and their
> substituted constants    (06)

Yes, but very few formal logics have a notation for indexicals, and
people almost never resolve the indexicals before making a statement.
So the process of resolving indexicals occurs during the mapping.    (07)

> So, as a statement, a fluent is a statement schema    (08)

That definition depends on the syntax, and it gets tangled up
with all the issues of mapping NLs to formal logics.  It's better
to use a notation-independent definition -- such as JMc & PJH's:
"A fluent is a function whose domain is the space Sit of situations."    (09)

> More generally, we can conceive of a statement-family as a statement
> schema in which one or more indexicals are not resolved -- place,
> time, person, perhaps even propositional attitude.    (010)

I would say that defining a fluent as a function is more general
because it's independent of any notation, formal or informal.    (011)

> Are such statement families worth reifying? Or is it enough simply to
> understand that many apparent statements are not statements because
> of indexicality?    (012)

Those are philosophical puzzles for nominalists like Quine and
Goodman.  When it comes to irritating Quine, I'm with Alonzo Church.    (013)

There's a very good reason why natural languages permit arbitrary
nominalizations:  People find them concise and convenient for saying
what they want to say.  (Note your use of 'indexicality'.)    (014)

If nominalists don't like them -- excuse my French -- tant pis.
Besides Church, there are many worthy philosophers who agree:
http://www.jfsowa.com/pubs/signproc.htm#static    (015)

And by the way, very few formal languages have a notation for
indexicals. In my 1984 book, I allowed indexicals in conceptual graphs
and marked them with the prefix '#'.  For an example of the notation
and the methods for resolving indexicals, see pp. 20 to 22 of    (016)

    From existential graphs to conceptual graphs    (017)

John    (018)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (019)

<Prev in Thread] Current Thread [Next in Thread>