Re: Grounding in English (was Re: semantics status of RDF(S))

>pat hayes:
> >Sandro Hawke:
> >It seems to me that the only way two agents (eg you and I, or two
> >computer processes) can communicate (at least electronically) is by
> >exchanging linguistic expressions in a language they (we) both know.
> >
> >We can, of course, define a language (KIF, Prolog, FOPC, RDF, etc might
> >qualify) and then use it.  But we have to define THAT language using a
> >language we already know.
> 
> No, we don't!! This is a KEY POINT, and is the basic reason why we 
> 'logics types' keep going on and on and on about semantics.  Giving a 
> translation into another language is NOT the only way to fix meaning. 
> A much, much better way is to give rules for attaching the language 
> to a world.

Interesting.  I'll need to learn more about that.

But my learning more about that will still involve reading English
text.   So "model theory"/"logical semantics" is still defined in
English, right?

I'm certainly not proposing that English Sentences be used as the
logic language -- of course not! -- just that it gives us an
interesting and extremely solid place from which to identify the
possible logics and anything else we might need to get started.

> What you are calling 'semantics' is just using a very long reference 
> number. The machines doing the matching have no idea what the English 
> is saying, or even that the strings are in English.

Correct.  That is what I'm saying.  The strings are reference numbers
which are easy to assign and which solve some of the persistent
problems of connecting abstract information to human minds.  I imagine
(perhaps in error) that a legal contract in this notation would have a
better chance of standing up in court than one written using terms
which were URIs for web pages which maybe said something about the
denoted objects.  

> couldn't read. You are proposing having blank punched cards with 
> writing on them, and calling it 'semantics'.

I don't think so.  It's more like a filesystem and a set of tools like
cmp, cp, tar, mv, etc, which I use to handle thousands of files in
thousands of directories with billions of bytes (some of which are use
by another subsystem to actually be those tools).  It all works fine,
even though the filesystem and tools have not a clue what any of it
means.  I call that that clean, modular design.

I need to figure out what you're thinking would be a clean, modular
design to address the needs of the RDF Logic community.  (I'm not
asking you, because I think you've probably already said it, just not
in terms I yet understand.)

BTW, "trolling" is being provocative (usually taking a position you
don't really believe in) to start a discussion which you really don't
care about.  Unfortunately, that defining difference of apathy is in
the mind of the writer and can only be guessed at by the reader.  I am
confident that we all have much better things to do with our time than
argue about unimportant issues.  But we also sometimes need to be
provocative to start the productive discussions we need, which might
look like trolling.

   -- sandro

Received on Friday, 6 April 2001 19:09:50 UTC