[rak-list] (Fwd) FRBR

[Bernhard Eversberg]

Hier eine weitergeleitete Message aus AUTOCAT mit einigen interessanten 
Ausfuehrungen ueber FRBR, insbes. ueber "work" und "expression":

philip davis wrote: >>>
In an article entitled "A new direction for bibliographic records" which
appears in Catalogue & Index, No. 149, Autumn 2003 and is a continuation
of the article which appeared in the previous issue, Ann Chapman and Alan
Danskin mention a number of studies which have been carried out to test
the FRBR (Functional Requirements for Bibliographic Records) model. One
conclusion, from OCLC, is that identification of the Expression level is
virtually impossible. The work that OCLC was doing on this has ceased. <<<

There's an article in LRTS (v. 46 no. 4, Oct. 2002) by Edward T.
O'Neill, about this work.

Speaking from memory, it seemed to me that what this work showed was that
the data necessary to recognize and coordinate the expression level as
postulated by FRBR either wasn't recorded, or wasn't recorded sufficiently
systematically, in the data in WorldCat for the test set of records.  The
set taken was the records for _Humphrey Clinker_ by Tobias Smollett, a
work which I believe has been studied before in the OCLC database.

The FRBR concept "expression" is, like the concept "work" itself (a notion
rather more familiar to cataloguers), an abstraction.  It's a particular
state of a work, as prepared by an editor or enhanced by illustrations;
it's what's common between different editions of a text prepared by the
same editor, irrespective of the different physical characteristics of
each published instance (publisher, typesetting, also features not
reckoned to be significant, like a foreword by one person instead of a
foreword by another).  Just what differences constitute a difference
sufficient to be regarded as a different expression is a matter of
cataloguing tradition -- in our situation, the Anglo-American tradition.
Indeed, the distinction between one work and another is also a matter of
tradition: while a change of medium (book to film, or musical notation to
recorded sound) is taken to constitute a new work, change of language
(translation) does not; is a paraphrase a new expression or a new
(related) work?

What FRBR has given us is not so much concepts to be used to reconstruct
AACR2 (or any code), as a framework to help us recognize what elements
have to be accounted for in a cataloguing code, in order that different
levels of abstraction (different works and different manifestations of the
same work) can be coordinated: what information has to be present in the
record.  It's then up to the code to specify *how* that information is
recorded, and to the data framework (MARC, XML, DC, whatever) to furnish
the mechanisms to specify it.

To take a different example: there are editions of Tolkien's _The Hobbit_
illustrated by various hands, including the author's own.  The *work* is
known as "The Hobbit".  The versions with illustrations by Tolkien, and
those by Pauline Baynes, are different *expression*.  When published in an
edition, those constitute different *manifestations* (and the physical
copies *items*).  The author's manuscript with his original illustrations
(if it exists) is also a *manifestation* (a unique one, in its nature, so
the set of items it encompasses has but one member).  If the information
about illustrations has not been recorded by cataloguers, it can't be used
to collocate the manifestations of that expression; and even if it has, it
may not be possible to collocate them if it hasn't been recorded uniformly
(fullness, tags and subfields used, etc.).

By my reading of O'Neill, that's why the concept *expression* can't be
usefully applied to _Humphrey Clinker_ in WorldCat.  I don't think that
proves that the concept is invalid, nor that it has no practical
application.

As an aside, while uniform titles (solo or in combination with author's
name) are usually regarded as denoting works, some (especially for sacred
texts) are better understood as denoting expressions: e.g. Bible.
English. Jerusalem Bible.

Hal Cain
Joint Theological Library
Parkville, Victoria, Australia
hecain at ormond.unimelb.edu.au
------- End of forwarded message -------

Bernhard Eversberg
Universitaetsbibliothek, Postf. 3329, 
D-38023 Braunschweig, Germany
Tel.  +49 531 391-5026 , -5011 , FAX  -5836
e-mail  B.Eversberg at tu-bs.de