There are many questions about names and how they might be handled in a repository.
There are two major reasons for using name authorities:
- consistency in data entry for search, display, aggregation within and between repositories and harvest services, and for use with supplying dynamic data for author pages. Let's call that DATA.
- providing access to nominated users to specific kinds of data within a repository. Let's call that ACCESS.
Work is being done to provide name authorities on a national level for those purposes. It is clear that there is not going to be one global name authority, but that there will be many, including those provided by commercial services.
DATA authority is being addressed by People Australia. Since 1981 the National Library of Australia has maintained a name authority. A PID is assigned to each person. A Group Record is created, and all other PIDs from other sources (such as the Dictionary of National Biography, or the Dictionary of Australian Artists Online) are brought under that record.
An ACCESS authority is being built by the Australian Access Federation. The aim is to build a unified list of people information designed to be used in authentication and authorisation.
There are other registers, including MAM's People Picker, Thomson's author registry, and others.
MACAR has discussed properties associated with creator. Some of these properties are directly related to the resource in the repository, such as creator name, creator affiliation. Other properties may be needed for statistical or administrative purposes - they are properties of the creator rather than the resource per se.
MACAR has recommended the following properties for creators:
creator display names
Number of authors
personal creator firstnames
personal creator lastname
If external name authorities are available, with namespaces, then there is an argument for including only two pieces of information about any resource creator - the PID and the namespace for that PID. Any other information required for search or display could be linked from that external source. I can see how display would work, but not really the search aspect. There is still a need for indexing.