On 10/27/2010 2:41 PM, Stephen Hearn wrote:
> If all we could get was a list of names or names with meaningless
> differentiating numbers, I wouldn't see the point of this either. But
> computers can do better than that. Lots of list display entries are
> actually concatenations of data from more than one source. For
> example, our system can show title and main entry data as additions to
> its call number lists, which is helpful for shelflisting, even though
> the title and main entry don't appear in the call number field.
> Similarly, bits of data from elsewhere in an authority record, or from
> elsewhere in a data system if the necessary linking data is present,
> could be combined with the name heading in a list display, as occurs
> in the IMDb Names index. The added information doesn't have to be part
> of the heading to become part of the list display.
>
> But--adding these kinds of details is much more useful if the name
> entries are differentiated for those persons catalogers have judged to
> be different. Hence my argument that we need to abolish
> undifferentiated personal name authorities by uncoupling
> differentiation from the form of the heading.
I agree. In the context of the web, what differentiates any entity is
its unique resource identifier. We should stop creating
undifferentiated name records and start establishing separate records
for each person or identity. We can then clearly record all the factual
information about each person in a clear and unambiguous way. We don't
even have to worry about whether the text string that we use as an
access point is unique; the identifier will keep the identities straight
-- and if we need something textual to differentiate in a display, we
have all the data elements associated with the person to choose from
(and by "we" I don't necessarily mean that a cataloger needs to make
that decision; intelligent machines are good at this sort of thing).
I've been arguing for this change of approach ever since I understood
that the FRBR entity PERSON is not defined solely by the name of the
person and that our authority records should be less about controlling
the form of name and more about recording information about the person
(name usage being only one such piece of information).
John Attig
Authority Control Librarian
Penn State University
[log in to unmask]
|