Aha! Ok, I think I get it. although I'm not sure what to do with it in 
our project.

Here's my attempt at an explanation for those who haven't taken a 
cataloging class :-) And correct me, since my own cataloging training 
was a very long time ago:

There are some parts of the library record that are a transcription the 
title page of the book as precisely as possible, including errors. This 
part of the record can be seen as a surrogate for the title page. So if 
you have a publisher whose official name is "Giangiacomo Feltrinelli 
Editore" but the title page says simply "Feltrinelli", then it's the 
latter that goes into your publisher name field.

I can see this as the "label" for that publisher for this particular 
publication. Except that you have a label in the record but not the 
thing it is labeling. That is, the publisher as entity isn't represented 
in the record, and the label may or may not be sufficient to connect to 
the entity. Admittedly, it may not be necessary to identify the 
publisher of the item in a precise way, but I can imagine that many 
people would assume that library records do attempt to unambiguously 
identify the publisher. In fact, they don't, although the "labels" of 
place and publisher name from the title page are good clues if someone 
wants to try to make that connection.

In contrast, we do create a precise identity for the named creators, in 
the form of the authority controlled entry, as well as the transcribed 
form. So, if the cover of the book says "T.C. Boyle" and the title page 
says "T. Coraghessan Boyle" then it is the latter that you put in the 
field for the transcribed data. ("Title: The tortilla curtain / T. 
Coraghessan Boyle") The author field, however, gets "Boyle, T. 
Coraghessan". That entry is a precise identifier, and the identifier in 
the bib record links directly to the record in the authority file that 
represents that creator. We do have the problem that we use the display 
form as the identifier, but that's just another adjustment that we need 
to eventually make.

I think we'll need to make a clear distinction between those data 
elements that represent entities (creators, etc.) and those that are 
presenting a surrogate of the item based on some rules (e.g. title page 
transcription). So I would take Mikael's example:

Label:
   Publisher name
   Publisher address
   Font size
   color
   dimensions

Publisher:
   name
   address
   no of employees
   legal contact
   website
   owned by

and make it something like:

Label:
   title transcription
   publisher transcription
     publisher ID [link to publisher "record"][optional]
   series transcription

Publisher:
   name
   address
   no of employees
   legal contact
   website
   owned by

Yes? No? Maybe?

kc

John Attig wrote:
> It is possible that I am using the term "label" in a different sense 
> than you are expecting.  What I am intending to convey is that the name 
> of the publisher is presented as a text string on the resource being 
> described and that this text string is included ("transcribed") as 
> identifying information in the description of that resource.  This is 
> different from the name of the publisher as an identifying attribute of 
> the publisher as an entity (although the same text string may be involved).
> 
> In our RDA discussions we have been using the term "label" to refer to 
> the text strings (title, statement of responsibility, edition, etc.) 
> that are included (printed) as identifying information on a 
> bibliographic resource. The transcription of these "labels" in 
> appropriate data elements is one of the principal tasks involved in 
> creating the description of a resource.
> 
> I'm not clear whether this makes the description monolithic -- or that 
> we ought in fact to be getting away from doing that. Perhaps you could 
> expand a bit on that point.
> 
>         John
> 
> At 10:23 AM 8/4/2007, Karen Coyle wrote:
>> John,
>>
>> I'm having trouble with this publisher-as-label thing because I don't 
>> understand what it's labeling. I could see the publisher name as being 
>> a label for the entity that published the book, but I don't understand 
>> the publisher label in relation to the item being described. Unless, 
>> of course, the entire bibliographic description is being considered a 
>> label for the item being described, but then that makes the 
>> description monolithic, which I think is exactly what we're trying to 
>> get away from.
> 
> 

-- 
-----------------------------------
Karen Coyle / Digital Library Consultant
[log in to unmask] http://www.kcoyle.net
ph.: 510-540-7596   skype: kcoylenet
fx.: 510-848-3913
mo.: 510-435-8234
------------------------------------