Since Series two is a multipart monograph like Series one and Series three, then the fixed field Series (008/12, type of series code) should be "b" rather than "a."   I'll wait a few days and then change that, unless someone has managed to convince me otherwise.  (The Project History page of The Frederick Douglass Papers Edition website supports the idea that Series two is complete in 3 vols.)
A multipart monograph can get series treatment (it can be analyzed, traced and have an SAR) but it can't actually be a monographic series, the two types of resources being mutually exclusive.  There are monographic series, and then there are several other types of things that, for practical reasons, may get series-style treatment to some degree:  multipart monographs, series-like phrases, analyzable periodicals & other serials, and integrating resources.  I think it is especially important to note this distinction as we implement RDA, because under the new instructions the attribute (Series) could be applied to access points where it does not belong and where it will be very unhelpful, very confusing.   (Series) = Form of the work (, i.e. the work or expression is a monographic series.  When added to the access point or given as an attribute for a multipart monograph, (Series) doesn't accurately indicate Form of the work*; instead, it refers to a library tech services procedure, with local variations.  It says something like:  "This resource could be analyzed, and some catalogers have chosen to do so."   
*This is also the case when (Series) is added to the authorized access point or given as an attribute for a periodical (or annual, or biennial).  Adding the qualifier (Series) to series-like phrases (when the whole point of the series-like phrase SAR is to say This is not a series and it doesn't get series-style treatment) also troubles me, but that's another conversation.
Mary Jane Cuneo
Serials cataloging and NACO
Information and Technical Services
Harvard Library
-----Original Message-----
From: Program for Cooperative Cataloging [mailto:[log in to unmask]] On Behalf Of Davis, Carroll N
Sent: Monday, July 15, 2013 11:17 AM
To: [log in to unmask]
Subject: Re: series vs. title proper?
FYI: The Library of Congress does not direct the series treatment practices that PCC (or other) institutions follow and I was not directly involved in cataloging The Frederick Douglass papers.  However, perhaps I can clarify a little what seems to be going on in the different records. 
"The Frederick Douglass papers. Series one, Speeches, debates, and interviews", "The Frederick Douglass papers. Series two, Autobiographical writings", and "The Frederick Douglass papers. Series three, Correspondence" are all multipart monographs (= multipart items).  Of the three, however, only "The Frederick Douglass papers. Series two, Autobiographical writings" is a "series" in the first sense that RDA and AACR2 use:
"A group of separate resources related to one another by the fact that each resource bears, in addition to its own title proper, a collective title applying to the group as a whole." 
In Series one and Series three, the individual parts are identified by spans of years, not individual titles proper.  No analytic titles, so no series.  Therefore, LC created only collective descriptions of Series one (LCCN 78016687) and Series three (LCCN 2009023125).  
In the case of Series two, it appears LC applied the same collective-only treatment at first (in LCCN 98026125), when the first volume was called only "Narrative", likely because the cataloger did not consider that to be an analytic title proper.  (That volume is still controlled on collective record LCCN 98026125 in LC.)  Series two at that point, like Series one, was considered a multipart monograph, but not a series.  The nature of Series two as a monographic series only became apparent with the second volume, which has analytic title "My bondage and my freedom".  At that point, an LC cataloger created a separate analytic description for the analytic and the series authority record for Series two as a series; and the cataloger noted in the SAR "Item is more than a publication intended to reproduce the works of Douglass," perhaps to help as a "bridge" between the treatments of the first and second volumes.
A more accurate description of LC's series treatment might have been "partially analyzed", since the first volume remains under control on the collective record and the second volume is covered by its analytic record.  That is moot now, because LC's policy since June 2006 for cataloging analytics in monographic series is no longer affected by or reflected in series treatment information in series authority records: the LC policy is simply "analyze fully, classify separately" (which matches LC's practice with Series 2 since the second volume), and LC does not consult, create, or modify SARs.  That policy does not apply in the cases of Series one and Series three because they are not monographic series.
PCC and other institutions should make their own choices about cataloging of Series one, Series two, and Series three: do only collective cataloging for each of the three; analyze Series two to the extent the institution chooses and classify the analytics separately or as a collection, as the institution prefers; or any combination of collective and analytic descriptions that is considered practical and useful.
Carroll N. Davis, ABA, U.S. Arts, Sciences & Humanities Division, Library of Congress (on behalf of Policy and Standards Division) [log in to unmask], 202-707-2156
-----Original Message-----
From: Program for Cooperative Cataloging [mailto:[log in to unmask]] On Behalf Of Amy Turner
Sent: Monday, July 15, 2013 7:36 AM
To: [log in to unmask]
Subject: Re: [PCCLIST] series vs. title proper?
LC created that authority record, and also cataloged series 1 (78016687) and series 3 (2009023125) with the 245 $n and $p that you noticed on most titles in OCLC.   Series 2 differs from the other two in that the individual volumes have distinctive titles, thus justifying treatment as a analyzed multi-part item. 
(This is a little murkier than it might be because LC created a record for series 2 as well (98026125).  Their policy has evolved; just before they gave up series authority work they would do either an authority record  or a bibliographic record for a multi-part item, in the past they sometimes did both, and now that they don't do series authority work there is probably less consistency in when they see something as a 4XX and when they see it as part of the title of the individual volume.)
Hope this helps,
Amy Turner
Monographic Cataloger and Authority Control Coordinator Duke University Libraries
[log in to unmask]
-----Original Message-----
From: Program for Cooperative Cataloging [mailto:[log in to unmask]] On Behalf Of John Reese
Sent: Sunday, July 14, 2013 11:25 AM
To: [log in to unmask]
Subject: Re: [PCCLIST] series vs. title proper?
"Fox, Chris" <[log in to unmask]> wrote:
I've been cataloging The Frederick Douglass papers, published by Yale, which consists of 3 "series" each containing several separate books.  I have chosen to use OCLC bib record # 39290009 for the second series within the papers.  I noticed while searching OCLC that most titles put The Frederick Douglass papers in the 245 field, with subfields n and p providing more specificity.  However, there is one book in the second series called My Bondage and My Freedom (OCLC # 51900353), which has only that as the title proper, while The Frederick Douglass papers is in a 490/830.  There is a series authority record (ARN # 5998264) for it, but no corresponding auth records for series 1 and 3.  I'm just wondering about the appropriateness of this authority record.  Maybe there's something I'm just not understanding here, and if anyone can explain it to me, I would feel very enlightened.
Chris Fox
Catalog Librarian
McKay Library
Brigham Young Univ.-Idaho
[log in to unmask]<mailto:[log in to unmask]>