Re: Itemprop for person

Hi, Karen - 

Bowker is one of the 2 active ISNI registration agencies now, and we've been sending information to OCLC for ISNI assignment.  Other information used to seed the ISNI database comes from Musicbrainz, the British Library, and ProQuest databases.

The pseudonyms in ISNI (and hence in some of the VIAF database - I know for a fact that Agatha Christie's entry was done manually by ISNI staff) are determined by the quality control team, which consists of the British Library and Bibliotheque Nationale. 

Which is to say that yes, in the case of Agatha Christie, the assignment of a pseudonym starts with information in the library authority files, but is added to and corroborated by information from other files as well.

Important to remember that while there may be ISNIs in VIAF, ISNI=!=VIAF.

On Nov 13, 2012, at 11:33 AM, Karen Coyle <kcoyle@kcoyle.net> wrote:

> 
> 
> On 11/12/12 8:45 AM, Graham Bell wrote:
>> Karen, Richard
>> 
>> But increasingly, with adoption of VIAF and ISNI identifiers for names,
>> I guess the library world _will_ be able to differentiate between 'real'
>> names and pseudonyms.
>> 
>> For example, Agatha Christie (ISNI 0000 0001 2102 2127) and Mary
>> Westmacott (ISNI 0000 0003 6613 0900) are different public identities
>> (or personal names) for the same person. The two identifiers are linked
>> within the ISNI registry, and the latter is clearly marked as a
>> pseudonym. Note also that the ISNI 0000 0001 2102 2127 includes 'Agatha
>> Christie', 'Agatha Mallowan', 'Агата Кристи', 'Agatha Miller', 'アガサ・
>> クリスティ' and a bunch of other variations, since these are all names
>> for the same public identity -- whereas 'Mary Westmacott' is a separate
>> public identity. And although Agatha Christie and Mary Westmacott are
>> linked, this isn't always the case (see for example ISNIs 0000 0000 7320
>> 7425 and 0000 0000 4340 7282).
> 
> It would be nice to know where they got the pseudonym information from. Most of the information in the record comes from VIAF (all of those language variants) which means that it comes from library authority files. If they used the data in those files to arrive at the pseudonym it would be interesting to find out how they did it. There are fields in the LC name authority records for "related names" (as opposed to "other name forms" which are coded differently) -- did they assume those to be pseudonyms? Here's what ISNI has:
> 
> [on the Agatha Christie record] Related identities:
>  
> Westmacott, Mary (other identity, same person)
> Westmacott, Mary (pseud; other identity, same person)
> Westmacott, Mary (pseudonym; other identity, same person)
> 
> [on the Mary Wesmacott record] Related identities:
>  
> Christie, Agatha (other identity, same person)
> 
> I wonder if some library authority files indicate which is the pseudonym? LCNA does not, AFAIK. Anyone know?
> 
>> 
>> So you can say that there are three levels to distinguish:
>> -- the person
>> -- the public identity (or 'persona')
>> -- the exact detail of the name
>> Book metadata generally identifies only the public identity of the
>> author (via one or more variations of author name listed in an authority
>> file), and there isn't really a formal link to the real world object --
>> the person -- at all (except perhaps a private one in the contracts or
>> the royalties system of the publisher). Indeed, sometimes, the lack of
>> this link between the real-world person and the public identity (or
>> between two public identities) is critical.
> 
> Exactly. Amanda Cross (really Carolyn Heilbrun) kept her real identity secret because she thought it would hurt her career. She outed herself only after she had gotten tenure in her institution.
> 
> One thing I want to remind people of about VIAF: each entry is a cluster of name authority records. The VIAF identifier identifies the cluster, but all of the information about the name is to be found in the contributed library records. This means that within a cluster you can find a variety of forms of a name, and sometimes even conflicting information. However, it is also possible that one of the participating libraries provides you with the particular bit of information you are looking for, such as:
> 
> Person
> ISNI:
>  
> 0000 0001 0939 7604 [searched on Amanda Cross]
> Name:
>  
> Cross, Amanda (pseud. de Carolyn G. Heilbrun)
> Gold Heilbrun, Carolyn
> Heilbrun, Carolyn
> Heilbrun, Carolyn G
> Heilbrun, Carolyn G.
> Heilbrun, Carolyn Gold
> 
> This is what appears under "Amanda Cross." The "pseud" information in parentheses comes from what looks like a Spanish-based authority file.
> 
> Note also that her name is given both as Amanda Cross AND Carolyn Heilbrun (and various forms of Carolyn Heilbrun). This is because VIAF has pulled these together (and I don't know what the algorithm is, nor why in this case VIAF treats the two names as equivalents. It looks like VIAF *is* making a decision about pseudonyms and RWOs, but it isn't clear to me how it does so or why it does so. (Richard, do you know if VIAF is attempting to bring together personas? If so, that would be good information to have since that is different from what some or most library authorities do, and I haven't heard that discussed in library-land. Is there documentation on this?) VIAF also provides a lot of "variety" of name forms, in part because it is a cluster of records not a single authority "declaration." There is no "preferred VIAF name form," just a cluster of preferred forms from contributing libraries.
> 
> If you want a tightly coherent authority file for your name display you need to choose one and only one of the participant files to pull from.
> 
> kc
> 
>> 
>> If there are links between person and public identity, they are at least
>> optionally one way or private.
>> 
>> This three-level model also helps clarify the 'two authors writing under
>> a single name'. Multiple real people have (often private) relationships
>> with a public identity like 'Franklin W. Dixon'.
>> 
>> It may seem 'mostly good enough' to assume an author is either a person
>> or an organization, but it might be hiding some important issues.
>> 
>> Graham
>> 
>> 
>> 
>> 
>> Graham Bell
>> EDItEUR
>> 
>> Tel: +44 20 7503 6418
>> Mob: +44 7887 754958
>> 
>> EDItEUR Limited is a company limited by guarantee, registered in England
>> no 2994705. Registered Office: United House, North Road, London N7 9DP,
>> UK. Website: http://www.editeur.org
>> 
>> 
>> 
>> On 12 Nov 2012, at 15:27, Richard Wallis wrote:
>> 
>>> The thing we need to take into account when using something like
>>> Schema.org <http://Schema.org/>
>>> is that they use classes to describe real world Things, such as people -
>>> their names, and possibly pseudonyms, being just properties of that
>>> Person.
>> 
>> 
>> 
>> On 12 Nov 2012, at 15:07, Karen Coyle wrote:
>> 
>>> 
>>> 
>>> On 11/12/12 8:31 AM, Richard Wallis wrote:
>>> 
>>>> 
>>>> Two authors writing under a single name has me stumped at the moment,
>>>> and I think we would have difficulty in convincing the Schema.org
>>>> <http://Schema.org> folks
>>>> to make changes to cope with such an edge case.  Perhaps we should
>>>> default to describing them as a Person with an explanatory note as a
>>>> description (not liking that I have just said that.)
>>> 
>>> Richard, I don't think that the library data that we have today
>>> distinguishes between "real" names and pseudonyms. I see nothing in the
>>> authority record that encodes this either. So what we have is "personal
>>> names" (and it's not "person" it's "personal name"), corporate names,
>>> family names, and each of these can either be an Agent (1XX, 7XX, 8XX)
>>> or a subject (6XX). That's it, at least in the MARC world. It would be
>>> interesting to hear from a wider international group if there are
>>> library standards that include more information about the relationship
>>> between the name and a Real World Object.
>>> 
>>> kc
>>> 
>> 
> 
> -- 
> Karen Coyle
> kcoyle@kcoyle.net http://kcoyle.net
> ph: 1-510-540-7596
> m: 1-510-435-8234
> skype: kcoylenet
>

Received on Wednesday, 14 November 2012 13:14:14 UTC