W3C home > Mailing lists > Public > www-tag@w3.org > March 2008

Re: New version of URI Declarations [Usage scenarios]

From: Pat Hayes <phayes@ihmc.us>
Date: Fri, 14 Mar 2008 23:11:14 -0500
Message-Id: <p06230909c400b16c8c8f@[]>
To: John Cowan <cowan@ccil.org>
Cc: "Booth, David (HP Software - Boston)" <dbooth@hp.com>, "www-tag@w3.org" <www-tag@w3.org>

At 3:33 PM -0400 3/14/08, John Cowan wrote:
>Pat Hayes scripsit:
>>  But this isn't how people in fact use names. They use them to refer,
>>  and if they say wrong things about the referents using the names,
>>  they correct those things rather than invent a new name. I very much
>>  doubt if people will change these deeply held habits to suit some
>>  proclamation from on high about declarations, nor indeed should they.
>Quite right so far as natural proper names embedded in natural languages
>are concerned.  But for formal, artificial names like URIs, meant to be
>embedded into formal, non-natural languages, the case is quite otherwise.

Its strange to be on the other side of this 
debate from where I usually am, but it does seem 
to be that URIs are often treated (and thought 
of) more like NL names than formal identifiers. I 
don't mean to say this is right, only that it is 
a widespread intuition: and do, violating it so 
blatantly is likely to cause confusion.

>>  >> OK, then it follows that YOU are using your own name incorrectly,
>>  >> right?
>>  >
>>  >Sorry, this is unintelligible to me.
>>  I don't believe you. Its perfectly intelligible, even to you. 
>Let's not get carried away with assertions of malice here.  I (quite
>naturally, I think) interpreted your phrase "your own name" in a posting
>addressed to me to mean the name "John Cowan", which left me bewildered.

Ah, sorry, I didn't even THINK of that interpretation.

>  > I'll give you an example. Suppose that you were to declare jc:moon in
>>  the way you described, and then another natural earth satellite is
>>  discovered (perhaps extremely small and dark, which is why it was not
>>  noticed before, but spectral analysis shows that it is a genuine
>>  piece of moon-type rock, formed at the same approximate time and in
>>  the same event as the moon was formed.) Now, your declaration of
>>  jc:moon no longer identifies the big moon. But I bet that, just like
>>  everyone else, you will will think of the URI as still denoting the
>>  big moon, and would prefer to correct your definition rather than
>>  invent a new name for the very same thing.
>If you are referring to me personally, I would much rather invent
>new artificial names jc:moon-1 (the moon anciently known of old) and
>jc:moon-2, and then copy over most or all of the assertions about jc:moon
>to create novel assertions about jc:moon-1.

How can you possibly do that, when your URI has 
been (re)used by who knows how many other 
ontologies to refer to the moon?

>  Remember that "jc:moon"
>in this context is an artificial name too, an abbreviation for some URI,
>not the English-language name "the Moon".

I realize that. Still, I bet you will have used 
the English name in some helpful documentation 

>Let us take a more realistic example.  Until recently, the registrar
>of ISO 639-3 promulgated a URI which we may abbreviate as 639-3:ccy,
>referring to a purported language called "Southern Zhuang" in English.
>In 2006, they concluded that there were in fact five distinct languages
>formerly lumped as "Southern Zhuang".  Rather than worrying about which,
>if any, of these might be the true Southern Zhuang, or if all were,
>they simply deprecated 639-3:ccy and introduced five new URIs.
>>  That is precisely what I want to avoid. Note however that it
>>  certainly can be "Im not sure exactly what this denotes, but it
>>  doesn't matter; because if you agree with what I say about it, we can
>>  communicate on that basis", which does not require any kind of
>>  declaration at all.
>Well, there you are: with which parts am I bound to agree on pain of
>talking about something else altogether


>, and which may I freely
>contradict, when speaking in a formal language?

You may contradict anything you like. Hopefully, 
we can come to some sort of resolution in time 
and agree on the facts. If not, there are several 
ways to resolve this. One, but only one, is for 
us to coin different URIs. But this is only the 
best way to proceed if it seems that we are in 
fact talking about different things. If we simply 
stubbornly disagree about the facts, then we 
should indeed have ontologies which stubbornly 

>>  I agree its not circular, but I also don't think it is is useless. In
>>  fact, there is good reason to suppose that most communication between
>>  human beings has exactly this character. None of us know exactly what
>>  others are referring to, and it almost never matters.
>What do you mean by "exactly"?  When I refer to Gell-Mann, I know who
>he is and that he is distinct from Feynman,

True, but are you referring to a continuant 
Gell-Mann, or an occurrent Gell-Mann? Or are you 
referring to something that is neither of these? 
Are you referring to Gell-Man at a particular 
time, or throughout his life? None of these are 
'common-sense' distinctions, but they are all 
distinctions made, and in some cases insisted 
upon, by various ontologies. So if your reasoner 
is using one of these ontologies, you will only 
be referring to something unique if you are able 
to answer questions like this.

>  even though I cannot give
>necessary and sufficient conditions for him just at present.

I didn't mean to imply that necessary and 
sufficient conditions would be necessary.

>  But if
>I wanted to, I could, at least in principle, and that in all possible

Thats quite a claim. I wouldn't dare say what I 
could or couldn't do in other possible worlds, 


>John Cowan <cowan@ccil.org>             http://www.ccil.org/~cowan
>It's like if you meet an really old, really rich guy covered in liver
>spots and breathing with an oxygen tank, and you say, "I want to be
>rich, too, so I'm going to start walking with a cane and I'm going to
>act crotchety and I'm going to get liver disease. --Wil Shipley

IHMC		(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32502			(850)291 0667    cell
http://www.ihmc.us/users/phayes      phayesAT-SIGNihmc.us
Received on Saturday, 15 March 2008 04:11:56 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:32:55 UTC