Identifiers
The Loon thought that the announcement of the ISO ISNI standard was thoroughly charming. (Yes, she is a strange bird.) ISNI CEO Olav Stokkmo was quoted, “The fundamental goal of ISNI is to provide a persistent identifier that can be shared and used by any organization that holds information on people and characters, reducing duplication of effort and improving accuracy.”
Characters, eh? Under certain lights, the Loon is one of those. Could she get an ISNI? According to ISNI’s home page, yes: “Those Parties can be natural persons (a human being like an book author), legal entities (like a Record Label) or even fictional characters (like Peter Pan).”
The Loon loves ISNIs already!
No, but let’s step back a moment and think about this. Identifiers are tricky beasts, and that’s before one arrives at lies programmers believe about names, names of people or names of pseudonyms.
Since OCLC is already part of the ISNI constellation, ISNI has presumably already run into problems of pseudonyms and fictional characters, since library authority files contain both. So ISNI presumably knows whether the Loon and her BAE receive different identifiers, and if so, whether ISNI will establish a link between one and the other. (Please don’t, ISNI.)
ISNI presumably also has metadata distinguishing natural persons from pseudonyms from fictional characters. Lovely. Which distinguisher does Violet Blue receive? Lady Gaga? The chronicler of the Quixote? (Trust the Loon, Cervantes was playing all sorts of mask games with that one; he learned to from the picaresque genre the immortal Quixote draws heavily on.)
Is the Hamlet of Hamlet the same Hamlet as the Hamlet of Rosencrantz and Guildenstern Are Dead? Is Defoe’s Robinson Crusoe the same character as Coetzee’s (who, if the Loon recalls correctly, is not even named within the book’s pages)? Does the historical emperor Claudius receive the same ISNI as Robert Graves’s character? Suetonius’s, in several ways equally fictionalized?
ISNI would no doubt prefer to punt on such questions. They just mint identifiers, right? It’s up to the rest of us to decide the boundaries of their application. Unfortunately, ISNI can’t punt, because ISNI must decide whether one or several ISNIs are minted around these edge cases. Perhaps ISNI builds equivalence and referral mechanisms itself, or perhaps it leaves such dilemmas to the evil Modeller and his henchlings, but the question of what merits an ISNI can’t be dodged.
The easiest solution may well be “mint ISNIs for everything and let the Web sort it out.” That solution will not, the Loon fears, endear itself to librarians, or much of anyone else, really.
The Loon could take a stab at some of these questions, but she knows that her background in literature study would govern her answers, and she isn’t sure that’s the most productive perspective from which to regard them. She looks forward, perhaps with more glee than she ought, to watching ISNI work out the answers.
You wrote: “… ISNI must decide whether one or several ISNIs are minted …”
But that doesn’t seem to be the case:
“Individuals, their agents, or in the special case of fictional characters, their creators can apply for a unique number through a registered ISNI agency. ”
(The exceptions seem to be the ISNI numbers assigned to existing VIAF URIs and entities from databases created by “rights management organizations, professional societies, government grant organizations and the supply trade.”)
So, as the creator of the Loon you could apply for an ISNI for that character. You could apply for a second one for yourself, too. Whether or not Violet Blue or Lady Gaga were considered to be a fictional characters would depend on whether the applicant applied as an individual or a creator. Stoppard and Coetzee could apply for ISNIs for their characters; Shakespeare and Defoe, sadly, cannot.
Claudius (0000 0001 0792 3260) and Suetonius (0000 0001 2102 9695) have ISNI numbers courtesy of VIAF. Graves and other fictionalizers presumably could apply for one for their creations.
Questions of identity can’t be treated as ontological questions, though it’s fun to try. The answers are grounded in particular social contexts, and the context that seems to govern ISNI is contemporary intellectual property law (this seems to be a way for creators and famous people to assert ownership over their names).
In any case, they don’t seem to be making any identity decisions of their own. They’re either giving them to individuals upon request, or assigning them to entities in existing databases.
You are correct; the Loon’s use of “ISNI” as a shortcut for “the entire ISNI ecosystem” was slipshod.
Of course these matters are contextual… but to the Loon’s mind they take on greater and greater salience as formerly-siloed data contexts start to rub elbows on the Web. What else has all the owl:sameAs controversy been about, really?
I’m not involved in ISNI, and can’t speak for it, but given the groups behind the intiative and some of the applications I’ve heard discussed for it, I suspect they will be taking a rightsholder-based approach. That is, a rightsholder can create and manage an ISNI for any public identity they control, which can include their “wallet name”, their pen names, and their characters. So the BAE of the Loon could register distinct ISNIs for the BAE and for the Loon, if desired.
While “Library Loon” strikes me more as a distinct public identity for an actual person than as a fictional character, it would be up to the BAE to decide whether to register it as an authorial persona or as a character (assuming those are typed differently in ISNI), since she’s the Loon’s “rightsholder”. (She could also register “Gavia Libraria” as an alternate-language label under the same ISNI as “Library Loon”, if desired.)
Similarly, it would be up to her as the rightsholder to decide whether or not to link the Loon and the BAE persona in ISNI. (In addition, if she doesn’t want to be traceable via common rightsholder data for both identities, the ISNI registration agency she used would have to agree to keep registrant information private. Some DNS registries already do that, and I’d imagine ISNI agencies could too.)
It’s easy enough, then, to see how fictional characters that still have rightsholders would be managed. I believe Barrie’s estate still has some copyright and trademark rights over Peter Pan in some jurisdictions (there’s still a US copyright on the published play script, for instance), so they would control Peter’s ISNI.
This doesn’t give any guidance on fictional characters like Shakespeare’s Hamlet that are completely in the public domain, but I’d imaging those could be managed by the libraries involved in the project, just as those libraries now manage such characters in authority schemes like LCSH. The decision on whether to have a separate ISNI for Hamlet as portrayed in Tom Stoppard’s more recent play “Rosencrantz and Guildenstern are Dead” in this case would be up to Stoppard, or whoever he passes the rights on to.
This isn’t necessarily the only or the best way to run a global naming registry. I suspect a registry run by or for literature scholars would work somewhat differently. But this seems to me the most likely way that ISNI will be run, given its background; and it seems workable enough on its own terms.
It seems perfectly dreadful to the Loon, who would prefer to keep rightsholders’ sticky (and often ignorant) fingers out of authority control.
But we may well be stuck with it, more’s the pity.