Mac, and all,

The project that Roy Tennant is working on around OCLC stats has 
produced a list of 100 $e terms and their frequencies. You can see this at:

There are nearly 11,000 such unique terms in that subfield. The 700 $e 
has not be provided yet - I assume it will be as various. Here are the 
first 11 in that list:

4101390	Auteur.
224450	ed.
163390	zhu.
135210	comp.
92854	zhuan.
84300	photographer.
79274	author.
38507	bian zhu.
35361	artist.
27213	interviewee.
25046	Author.

"ed." is the one I recall seeing most often, and there it is near the 
top of the list.

The instructions in MARC say:

"*$e - Relator term*

Designation of function that describes the relationship between a name 
and a work, e.g., ed., comp., ill., tr., collector, joint author.

700 	1#$aSmith, Elsie,$d1900-1945,$eillustrator.
700 	1#$aHecht, Ben,$d1893-1964,$ewriting,$edirection,$eproduction."

I could compare the list of terms in the LC MARC relator list to Roy's 
list of $e's, but first need to remove the ending periods, and that 
complicates things because many of the terms are in the list both with 
and without ending periods, so we don't really know how often the *term* 
appears... I'll also need to normalize the case. But if I get a chance 
I'll do some analysis.

To me the key thing about Roy's list is that it shows that there will be 
*many* lists of valid role terms, not just one. And in some library 
communities, who knows?, the role may  not be from a controlled list but 
might be uncontrolled or transcribed from the title page. I still think 
BIBFRAME needs a way to deal with this that doesn't mess with the name 


On 5/29/13 9:15 PM, J. McRee Elrod wrote:
> Karen said:
>> The list that I was working from was not coded in MARC. The 'names' were
>> not the abbreviated names that would be in $e, but were the 'terms'
>> listed in the LC relator list beside each of the codes.
> The terms which go in $e are not abbreviated.  Some are one word, and
> some are phrases.  The codes go in $4.
> The code definitions (which I assume you mean) are in some cases the
> same as, and in some cases differ from, RDA relationship terms.
> We plan to export those code deffinitions in $e for non Canadian
> cleints; we have them in both French and English, and don't want to do
> the work of pairing up the codes with RDA terms.  It would be nice if
> our various standard setting bodies would work together on one set of
> wheels,
> I assume ILS which can interpret codes would display the code
> deffinitions, as opposed to the RDA terms.
> If interested, here are the French code deffinitions we will use:
>     __       __   J. McRee (Mac) Elrod ([log in to unmask])
>    {__  |   /     Special Libraries Cataloguing   HTTP://
>    ___} |__ \__________________________________________________________

Karen Coyle
[log in to unmask]
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet