The book, the bug and the bangle: a parallel and a paradox Washington, International CRM Symposium...

25
The book, the bug and the bangle: a parallel and a paradox Washington, International CRM Symposium “Sharing the knowledge” March 26-27, 2003 Patrick Le Bœuf, Bibliothèque nationale de France, member of CRM-SIG &

Transcript of The book, the bug and the bangle: a parallel and a paradox Washington, International CRM Symposium...

The book, the bug and the bangle:a parallel and a paradox

Washington, International CRM Symposium “Sharing the knowledge”

March 26-27, 2003Patrick Le Bœuf, Bibliothèque nationale de France,

member of CRM-SIG & ISO/TC46/SC4/WG9, chair of the IFLA FRBR WG

Introduction

• Mappings = a tool for testing CRM’s robustness

• 2000-2001: mappings from AMICO, Dublin Core, & EAD DTD to CRM

• 2001-2002: I mapped FRBR [Functional Requirements for Bibliographic Records] to CRM– It induced some changes in CRM– It revealed some inconsistencies in FRBR

(For those who are not familiar with FRBR:)

FRBR’s core = Group 1 of Entities:

Work

Expression

Manifestation

ItemI purchased a book; it is on my shelf (and it occasionally serves to wedge a table leg)

Do you have this book, identified by ISBN 0-1234-5678-X?

I wrote a book; it is in English

My book is also available in Japanese (though I don’t know a single word of

Japanese)

The word book has at least four distinct meanings…

Mapping issues

• 1st problem: “vagueness”: too frequent use of:P2 has type (is type of) E55 Type

• For instance:• FRBR Expression attribute “Expected Frequency

of Issue (serial)” maps to:• E73 Information Object

P2 has type E55 Type (Expression)P2 has type E55 Type (Serial)P2 has type E55 Type (Weekly or Monthly

or Quarterly etc. (Expected Frequency of Issue))

Mapping issues• 2nd problem: difference in scope:• CRM covers the whole “universe of discourse”

of museums’ cultural information• FRBR is incomplete and only covers

“bibliographic records” (i.e., it excludes info about authors, subjects, etc.)

• We’re waiting for FRANAR [Functional Requirements And Numbering of Authority Records] for a model that holds for the whole “bibliographic universe”

Mapping issues

• 3rd problem: logical flaws in FRBR:– FRBR does not distinguish between “things”

and their “names”

– FRBR (and catalographic tradition, beyond FRBR) does not distinguish between “title-as-appellation” and “title-as-inscription”, “name-as-appellation” and “name-as-inscription”

Mapping issues• Above all, working on this FRBRCRM

mapping was an opportunity for me to focus on:– A parallel between library catalogues as

reflected in FRBR and museum documentation as reflected in CRM:books are closer to bugs than to bangles

– A paradox underlying the very activity of cataloguing in current library practice:librarians focus on things that may not “exist”

A parallel

• CRM’s focus for Fine Arts Museums = on individual objects (or collective objects)

• CRM’s focus for Natural History Museums = on individual objects and the classes (“taxons”) they belong to

• FRBR’s (and Catalographic Tradition’s) focus = on classes (“publications”)

A parallel

• What does “Sharing the knowledge” mean?• CRM: sharing knowledge about contexts for the existence

of individual objects, not the description of these (unique) objects

• FRBR (& ISBDs):– exchanging actual descriptions of classes (i. e., “publications”);

– the context for their coming to existence does not matter

– info about items is not deemed relevant for interchange and therefore not coped with

A (paradoxical) parallel

• Focusing on classes rather than instances makes cataloguers’ job closer to CRM-as-for-Natural-History-Museums than to CRM-as-for-Art/Archaeology-Museums:

A (paradoxical) parallel

National Bibliographic Agencies

Natural History Museums

Book Bug

Describe the

“item” “prototype” or “specimen”

Assume common features for all other instances of the

“manifestation” “taxon”other libraries

check if their item exemplifies the

same manifestation

other museums check if their

specimen belongs to the same taxon

Physical item Physical item Physical item

Type

Taxon Publication

Content Content

materialises

has type

belongs to belongs to

embodies

Art/Archaeology museums:the bangle

Natural history museums:

the bug

Libraries:the book

To sum it up:

Physical item

Type

Content

P128 is carrier of (is carried by)

P2 has type (is type of)

Art/Archaeology museums:the bangle

(or painting, sculpture, etc.)CRM

E20 Biological Ob-ject (bug) or E84 Information Car-rier (bangle, book)

FRBR

E55 Type

E73 Information Object

Item

Manifestation

WorkExpression

Physical item

Publication

Content

belongs to(exemplifies)

embodies

Libraries:the book

(or CD, movie, map, etc.)

Physical item

Taxon

belongs to(P2 has type (is type of))

Natural history museums:

the bug(or plant, etc.)

To put it in CRM & FRBR terms:

A paradox

• In “traditional” library catalogues, the 2 FRBR entities that prove to be crucial =

MANIFESTATIONWORK (though at a lesser degree)

• In CRM’s “philosophy”, these 2 notions do not even seem to exist as entities…

• What is it then that librarians catalog?

A paradox• An example:

FRBR and CRM would not model these “things” the same way:

Seikilos Song (original notation)

Seikilos Song (modern transcription)

Seikilos Song (recorded performance)

CRM E73 Information Object

CRM E73 Information Object

CRM E73 Information Object

CRM P130 shows features ofP130.1 kind of similarity: E55 Type “Transcription”

(CRM E73 Information Object)

FRBR Expression

FRBR Expression

FRBR Expression

FRBR Work

FRBR is realized through

CRM P130 shows features ofP130.1 kind of similarity: E55 Type “Performance”

(CRM P130 shows features of)

WORK• = analogous to signs in linguistics, &

= signifying, concrete set of ideational conceptions realized through semantic or symbolic expressions (Smiraglia)

• = commonality of content between and among the various expressions of the work (FRBR’s recursive definition)

• = a conceptual similarity among instances of Information Object (E73) (cf. P73) (CRM)

• = on the whole, rather a relationship than something that “exists” (“entity”)

MANIFESTATION• = physical embodiment of an expression of a work; all

copies produced that form part of the same set are considered to be copies of the same manifestation (FRBR’s contradictory definition: a set and a physical embodiment at the same time)

• = a subclass of Type (E55) to which instances of Information Carrier (E73) are linked through P2 has type property, and= a subclass of Information Object (E23) and of Type (E55), and a metaclass (CRM)

• = on the whole, rather an assumption than a physical evidence

A paradox

• Librarians mainly catalog assumptions and, at a lesser degree, arbitrary conceptual similarities, i.e.:

• They mainly focus on Manifestations and, at a lesser degree, on Works…

• … whereas in “real life” we can only deal with Items and Expressions, i.e. instances of Information Carrier (E84) and of Information Object (E73)

Conclusion (1/2)

• This is not to blame librarians (I’m one of them…): info about Publications (Manifestations) and Works is actually vital to us

• This is just to highlight that Sharing the Knowledge does not mean quite the same for libraries and museums. And yet…

• And yet, we have to find some common ground…

Conclusion (2/2)

• … That common ground could perhaps be found in some FRBR attributes for Work/Expression that are not consistently recorded in current library practice, but that fit in well with CRM:

4.2.7. Context for the Work4.3.10. Context for the Expression4.3.11. Critical Response to the Expression

and in a consistent documentation of relationships defined by FRBR between and among Works and Expressions

• This way, we might really Share the Knowledge between libraries and museums

FRBR attribute « Dates of Person » for a Person, or « Medium of Performance » for a (musical) Work:

1830-1886

100 10 |a Mozart, Wolfgang Amadeus, |d 1756-1791. |t Adagio und Allegro, |m musical clock, |n K. 594, |r F minor

Musical clock

actually no data elements per se, but parts of larger, conventional appellations in the context of library catalogues:

100 1_ |a Dickinson, Emily, |d 1830-1886

“Title proper” (attribute of a Manifestation):

What hapenned [sic] on March, 72 [i.e., 27]=

Inscription interpreted as Title: What hapenned on March, 72

+

“Ideal” title for filing & retrieving: What happened on March, 27

“Place of publication” (attribute of a Manifestation):

Caerdydd [i. e., Cardiff]=

Place Name as Inscription: Caerdydd

+

Alternative Place Name, deemed more “comprehensible”: Cardiff