Re: another crack at 'information resource'

On Thu, Nov 11, 2010 at 10:43 AM, Pat Hayes <phayes@ihmc.us> wrote:
>
> On Nov 8, 2010, at 4:46 PM, Jonathan Rees wrote:
>
>> On Mon, Nov 8, 2010 at 8:17 PM, Pat Hayes <phayes@ihmc.us> wrote:
>>>
>>> Hmm. Not sure if you mean to cast your net this widely, but there are plenty of counterexamples to this as a general principle. A particular photograph of an artwork might well be credited to (and owned by) the photographer, not the artist of the original work. Prints of engravings vary enormously in price depending on whether they were pulled by the artist or by someone else from the artist's plates: often it takes arcane expertise to be able to tell them apart. Im sure there are similar examples from the worlds of music and other arts.
>>
>> Suppose we are talking about some book, say Pale Fire by Nabokov,
>> about its prose style, wit, length, etc., but have never needed to
>> discuss anything about the particular physical copies each of us has
>> in hand. I may not even know what language your copy is written in.
>> Suddenly you say "and it's cheap - I only paid $2000 for it" -
>> something I can't corroborate. I'm completely baffled by this until,
>> after a long and confusing conversation, we finally figure out that by
>> "Pale Fire" you meant your particular signed copy of it, not the book
>> in general.
>>
>> I think a linguist would tell you that this scenario will not happen
>> with ordinary competent speakers of English. You would have said "*my
>> copy* was cheap". The price is simply not a property of Pale Fire. If
>> you try to formalize this, you'll have three distinct individuals:
>> Pale Fire, my copy of it, and your copy of it.
>>
>
>
> Agree with all that. But that seems to be on a different topic than the one we (I) was referring to. For a book example, suppose that one edition of Pale Fire was a hand-made presentation copy made by a famous (and expensive) bookmaker, and collectable as part of a hand-made-book collection, regardless of its literary merits. Then if someone were to ask about the provenance of this copy, a full answer might well refer to Nabokov *and* the guy who made that particular book.

Ah, I think I get it, let me try echoing it back to you. You could
have an "information resource" (Pale Fire) with dc:creator Nabokov but
NOT dc:creator FamousBookmaker, and one of its "representations" could
have dc:creator Nabokov AND dc:creator FamousBookmaker.

I think that's consistent with what I was saying (or maybe I said it
wrong), if dc:creator isn't functional (and it isn't). dc:creator
Nabokov would transfer from all of the IR's 'representations' to the
IR itself, and back again. But  I would have to stipulate that the
negation "not dc:creator" was not a "content property", so that the
property of not having  creator FamousBookmaker would not transfer
from the generic book (IR) back to the specific book (the
'representation' with creator FamousBookmaker).  I don't know what
principle would justify treating dc:creator and its negation
differently, but it's what I want to be true.

If a book can only have one title, and the FamousBookmaker changed the
title, I'd recommend that either the generic book not have any title,
or that we deny that the modified 'representation' was a
'representation' of the generic book. Or that the IR 'located at' a
particular URI did not refer to Pale Fire at all, but to some oddity.

I didn't want to dwell on dc:creator or provenance, though, as that
was just an example. dc:title or dc:date would have worked as well for
my purposes.

I could enumerate lots of properties ("content properties") that I
think should follow the  universal quantification rule - e.g. most or
all of the DC and BIBO properties - but (in light of the negations of
these properties) I don't know how to enable someone else to
generalize to additional properties, i.e. what the boundaries of the
meta-category of content properties are.

Before spending a lot of time trying to figure that out, though, I
want to convince at least one other person that this idea (universal
quantification to extend representation properties to IR properties)
has promise as a possible way to motivate the httpRange-14 rule.

I should also check to see if the idea agrees with Dan Connolly's
work. He may have assumed existential quantification, which I consider
less useful (though not useless) and more error-prone.

Thanks
Jonathan

> Pat
>
> ------------------------------------------------------------
> IHMC                                     (850)434 8903 or (650)494 3973
> 40 South Alcaniz St.           (850)202 4416   office
> Pensacola                            (850)202 4440   fax
> FL 32502                              (850)291 0667   mobile
> phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
>

Received on Thursday, 11 November 2010 22:23:55 UTC