Re: dates

From: Jonathan Rochkind <rochkind_at_nyob>
Date: Tue, 26 Jul 2011 17:19:13 -0400
To: NGC4LIB_at_LISTSERV.ND.EDU
On 7/26/2011 2:49 PM, Eric Lease Morgan wrote:
>
> Instead, I want an additional date denoting when -- in all likely hood -- the idea expressed by the creator of the work was conceived/embodied. Put another way, I want to sort search results from older to new. Find all of Shakespeare, sort by date, and then I want to read the oldest one first. Find all things regarding New-Platonism, sort by date, and start reading at the end and go back in time.
>

One way to get at that in terms of the data we actually have, in FRBR 
terms, would be:

The date the first Manifestation of the Work was published.

Now, that's not neccesarily the exact same thing as "was conceived", but 
it's probably a close enough proxy to give you the sort of results you 
want. And "date first Manifestation of Work" was published is... sort of 
kind of usually encoded somehow in our bibliographic data, although it 
won't be easy to get it out, largely because work-set relationships are 
encoded kind of haphazardly in legacy data (if FRBR were taken 
seriously, it'd be easier!).  While "date of conception" simply isn't 
recorded at all (because it's not really clear how a cataloger would 
know that, it would generally require more research than cataloging 
entails).

Now, what vocabulary to use to to encode that or how to encode it in RDF 
is another story.... But I think it becomes somewhat more 
straightforward if we stick to FRBR, and realize what we want to encode 
is about the Work a particular Manifestation belongs to, the list of all 
Manifestations in that Work, and then the earliest published 
Manifestation in that list.  This is the value of FRBR, to give us the 
language and conceptual model to talk and think about these things.
Received on Tue Jul 26 2011 - 17:20:56 EDT