I concur with Richard's analysis[1]. Each identifier type serves a 
different community. In particular, ORCID identifiers will tend to 
identify faculty and researchers whose sole output is journal articles 
-- thus who would not normally appear in a library authority file. The 
ISNI is sometimes seen as an interloper from the publishing community, 
but most likely is integrated into the publisher workflow (e.g. writing 
checks to authors).

Like Richard, I don't see anything to worry about. You can use one, 
some, or all of the identifiers based on your needs. So a faculty 
digital repository may need to used ORCIDs because there are authors who 
are only identified by those. (Repositories are beginning to require 
ORCIDs for deposit.) The same repository can also use LCNA for some 
authors -- you are in no way limited to one identifier per person or 
system. If you are hoping to pull in author data in your library catalog 
from wiki/DB/pedia, then you might favor the VIAF identifier, since this 
is being linked to the "pedia" world.

This seems to me to be quite similar to other data and metadata choices 
that we make: define your use case, then choose the data that meets that 

[1] One possible difference is that I would consider ORCID a viable URI 
for linked data purposes, although at the moment ORCID does not export 
its data in RDF. All of the identifiers listed below are HTTP URIs.

On 6/20/14, 7:56 AM, Richard Wallis wrote:
> Hi Eric,
> What distinguishes one from another?
> The communities behind them, the [often overlapping] communities they
> are intended to serve, and the technical implementation.
> As a librarian, why should I care?
> I would, as a non-librarian, suggest that once you are happy with
> the ‘authority’ of them, you shouldn’t have to care. Ideally, we are not
> there yet, systems should be flexible and accommodating enough to link to
> any appropriate authority.
> I will probably get flamed for over generalisation here but - VIAF is
> an aggregation of National Libraries Authority files.  - ISNI is a more
> publisher focused but similar effort.  - OCID comes from and and tries to
> serve individual academic institutions, their researchers and falsity
> authors.
>            authority control |simple identifier |Linked Data capability
>           +-----------------+------------------+--------------+
>    VIAF   |        X        |    X             |      X       |
>           +-----------------+------------------+--------------+
>    ORCID  |                 |     X            |              |
>           +-----------------+------------------+--------------+
>     ISNI  |        X        |     X            |    X         |
>           +-----------------+------------------+--------------+
> ~Richard
> On 20 June 2014 15:42, Eric Lease Morgan <[log in to unmask]> wrote:
>> On Jun 20, 2014, at 10:31 AM, Richard Wallis <
>> [log in to unmask]> wrote:
>>> ISNI has a suite of programs that detects pseudonyms coded as name
>> variants
>>> and changes them into related name and generates related identity
>> records.
>>> It is a while since it was run and will be re-run in the next few weeks.
>>> This should change Currer Bell into a related name of Charlotte Brontë .
>> Please humor me as I ask this question again. What is the difference
>> between ISNI and other identifiers systems (like ORCID, etc.)? What
>> distinguishes one from another? As a librarian, why should I care? Was as a
>> faculty member/scholar, why should I care? Under what context is one
>> identifier expected to be used instead of another? Maybe a picture/graph is
>> in order:
>>            authority control simple pointer
>>           +-----------------+--------------+
>>    VIAF   |        X        |              |
>>           +-----------------+--------------+
>>    ORCID  |                 |     X        |
>>           +-----------------+--------------+
>>     ISNI  |                 |              |
>>           +-----------------+--------------+
>> —
>> Eric Lease Morgan

Karen Coyle
[log in to unmask]
m: 1-510-435-8234
skype: kcoylenet