It's not at all clear to me that a class like (I'll give it a plausible name) bf:IdentifierScheme would be associated with a specific vocabulary. Certainly, (say) LC could publish a handy suite of popular URIs for use in that slot. As is convenient, they could be formed into a vocabulary by some formality. Without conflict with that activity, I could publish at

<> a bf:IdentifierScheme ;
other triples…

and at

<>   a   bf:Instance ;
 			bf:identifier [
					a   bf:IdentifierDescriptor ;
					bf:identifierScheme   < > ;
  					bf:identifierValue   "1qaz2wsx" ; ] .

or the like. One reason for which I suggest the existence of the type is so that there will be some means by which to organize extension work. It's not dissimilar to some of the thinking that turned up in the conversation about classifier schemes.

A. Soroka
The University of Virginia Library

On Aug 5, 2014, at 2:18 PM, "Denenberg, Ray" <[log in to unmask]> wrote:

> A.      Soroka asked:
> Is there an intention to provide a class for the range of this predicate? (bf:identifierScheme)
> It’s a good question, and merits further study.  If we do, then we would have to restrict it to a specific vocabulary, wouldn’t we?