I can think of two cases where LC/NACO/SACO, playing by the rules, 
alters the meaning of an LCCN-identified authority:

1. The already discussed undifferentiated personal name authorities 
(UPNAs). Many of these start their careers as unique name headings, 
and presumably valid candidates for being linked. Then they become 
undifferentiated for a time, hosting a changing cast of identities; 
then, if all but one identity  are migrated off to their own 
authorities, the UNPA can become unique again, possibly for a 
different person from the one originally established on the 
authority. Then the cycle starts over.

As was noted in the discussions of RDA at Midwinter, the need for 
UNPAs could evaporate if there was an allowance for using some 
exceptional device (e.g., a neutral sequence number--"001," "002," 
etc.) to differentiate identities when none of the standard 
differentiating data (date, fuller form, etc.) are available. It is 
actually very rare that the different identities referenced by an 
undifferentiated personal name authority cannot be distinguished. 
What's lacking is an allowable addition to the heading to distinguish 
the headings for the differentiable identities from one another.

2. Subject headings for open-ended chronological periods.  SCM:SH 
recommends closing the date on an existing open-date span when a new 
open-date span is proposed, which can occur several years after the 
proposed closing date. For those years, the scope of the heading 
includes the years after the proposed closing date; when the closing 
date is added to the existing subject authority, that scope is 
altered, and some of the bib headings linked to the authority may no 
longer be appropriate for the resource they're describing when the 
closing date is added.

SCM:SH does call for deleting an existing record and using all new 
authority records to handle subject heading "splits." Topical date 
closings might better be handled as heading splits. That's how 
they're usually treated when doing heading maintenance.

Most of the cases of "reuse" of LCCNs have already been addressed 
(e.g., the old practice of adding 410s for valid AACR2 headings to 
corporate name authorities rather than establishing them, but 
allowing that they could be established in the future). If we could 
deal with the few that are left, and do a better job of tracking 
heading histories and marking them up for machine manipulation, we 
could put automated heading maintenance on a sounder footing.

Lastly, I agree with Mary Charles Lasater's note on this 
topic--authorized once doesn't mean authorized indefinitely. We need 
a system to maintain headings on PCC records in OCLC, and to remove 
the PCC label when headings are not maintained, if the PCC label is 
going to retain its value. Most of this work should be done 
automatically, and the bits that can't be (fixing links to the wrong 
authority, dealing with heading splits) should earn some kind of 
acknowledgment.

Stephen

At 07:30 PM 5/12/2008, you wrote:
>Sorry, I thought it went without saying that one should make sure 
>one has used the *correct* heading before controlling it :-) Thank 
>you for the reminder, though. I've had to fix headings that have 
>been linked to the wrong authority record too.
>
>When I said it was a convenient way to ensure that all needed 
>authority work had been done I didn't mean we should just let the 
>headings automatically control without checking. I meant that once 
>I'm done with the record, if I've controlled the headings it's 
>perfectly plain (since the controlled headings are a different 
>color) if I've forgotten to do authority work on any heading.
>
>And I understand the problem with the undifferentiated headings and 
>agree that it probably isn't a good idea to control them.
>
>Bob
>
>Robert L. Maxwell
>Head, Special Collections and Metadata Catalog Dept.
>6728 Harold B. Lee Library
>Brigham Young University
>Provo, UT 84602
>(801)422-5568
>
>
>-----Original Message-----
>From: Program for Cooperative Cataloging [mailto:[log in to unmask]] On 
>Behalf Of john g marr
>Sent: Monday, May 12, 2008 6:18 PM
>To: [log in to unmask]
>Subject: Re: Horrible record [going off on the tangent of Enc lvl]
>
>On Mon, 12 May 2008, Robert Maxwell wrote:
>
> > ... I always control the headings and think it is good practice.
> > Controlling the headings is an excellent way to make sure the BIBCO
> > records are "kept in synch with the authority records." It's also a
> > convenient way to ensure that all the needed authority work has (or has
> > not) been done.
>
>Hi Robert:
>
>    You may not have meant to dredge up a conundrum with your comment, but
>one does have to look at the authority records before controlling to
>them-- not rarely (albeit infrequently), I have had to change
>already-controlled headings to other forms because the persons associated
>with the bib. records were not the persons represented by the controlling
>authority records.
>
>                                              John G. Marr
>                                              Cataloger
>                                              RMBA, UNMGL
>                                              Univ. of New Mexico
>                                              Albuquerque, NM 87131
>                                              [log in to unmask]
>                                              [log in to unmask]
>
>          **"I really like to know the reasons for what I do!"**
>                                              Martha Watson
>
>Opinions belong exclusively to the individuals expressing them, but
>sharing is permitted.

****************************************************
Stephen Hearn
Authority Control Coord./Database Mgmt. Section Head
Technical Services Dept.
University of Minnesota
160 Wilson Library               Voice: 612-625-2328
309 19th Avenue South              Fax: 612-625-3428
Minneapolis, MN 55455      E-mail: [log in to unmask]