Okay, this is a great example to talk about what the difference is
between 'a web page' and an authority record.
I agree that we need more fine-grained relationships in our authority
records. "See also" is too broadly ham handed. But we need that
information encoded in a structured way that our software can make use
of. Just putting it as text on a web page doesn't do this anymore than
putting it as text in a MARC record. (Because MARC authority records can
and do already have such clarifying 'notes'---the problem is, that they
are narrative notes, there's no way for software to be sure _which_ "see
also" note goes with _which_ actual "see also" reference).
You can easily create web pages from the MARC records that are there
now, that's not a problem. It might be a good idea, but it doesn't give
you any more data than we currently have. If the data we currently have
isn't being displayed usefully by our current software, that's our
software's fault.
But if the data we currently have doesn't _include_ what we _want_, or
doesn't include it in a _way_ that software can algorithmically access,
that's the data's fault.
We need structured authority and bibliographic data that has
semantically explicit components that are what we need. The FRBR/FRAD
model is an attempt to explicitly lay out what those components are.
For this particular example, let's imagine some XML data. It doens't
need to be XML, but XML is just one easy way to lay out structured data
that is also human-readable. [And this example is just a broad
over-simplication to illustrate a particular thing, I know my 'code'
below has lots of problems]
<person>
<name>Brutus, Marcus Junius</name>
<relationship>
<destination_name>Pseudo-Brutus</destination_name>
<type>erroneus_attribution</type>
<note>For the Greek letters erroneously attributed to this
person, see </note>
</relationship>
</person>
If you have structured data like this, you can use it to _generate_ a
web page, sure. You can also use to effect your search algorithms. If a
person enters "Brutus", you can ask them if they'd like to include
Psuedo-Brutus in their search too, or not. You can ask them in general
if they are interested in 'erroneous attribution' type relationships or
not. You can use it for all sorts of things.
Of course, we have most of this data in MARC Authority already, that's
of course the point of our MARC records. XML isn't magically different
than MARC in that sense, although it is a bit more flexible in some
ways. But it's more about our data than it is about in what format we
store our data--so long as we format it in a way that software can get
it out again without losing information! One piece of data in that
example that we don't have in our current authority records is that
"type" above, all we have is "see also" relationships, our data gives
the computer no good way to distinguish a bibliographic identity from a
seperate related person from a parent organization, etc. And we probably
would like to be able to record even _more_ kinds of relationships than
we do now---so our software can act upon it, in ways we can think of
now, and in ways that people will only invent in the future.
But we have much of this in MARC Authorities now. If our software isn't
taking full advantage of it, that's our software's fault. If we're
missing data that we need, that's our data's fault. But "replacing MARC
with web pages" is not the answer, that would be a step backwards. Web
pages are a _presentation_ of authority data, the actual authority data
needs to live in a structured way such that software can extract all the
meaning out of it that the cataloger's put in.
Hope this make some sense, Thomas. The distinction between presentation
and underlying structured data is an important one.
Jonathan
Brenndorfer, Thomas wrote:
> Further to the issue of authority data and presentation...
>
> There has always been one key example in AACR2 that has bothered me
> greatly about the way the automation of catalogue records has
> progressed, and that is rule 26.2D "Explanatory references".
>
> Raw SEE and SEE ALSO references provide great functionality, but what's
> missing is the ability to display clarifying data such as:
>
> Brutus, Marcus Junius
> For the Greek letters erroneously attributed to this person, see
> Pseudo-Brutus
>
> Catalogue cards had some free-form potential, and I think web pages, as
> the presentation mechanism for authority information, can recover what
> has been lost. The bonus in thinking of authorities displayed as web
> pages is that we can add additional structure to authority records to
> draw in more data to help users identify what they are after. Where the
> user encounters that data is important as well. Presenting users with a
> browseable list of headings or a set of bibliographic records with only
> headings visible limits what data can be presented. Presenting users
> with a web page with more options to integrate explanatory data strikes
> me as a very good starting point in discussing ways of improving
> catalogues.
>
> As with explanatory references easily accomplished with catalogue cards,
> I do believe the form of presentation influences the choices of what
> data is included and what underlying structures are used.
>
> Thomas Brenndorfer, B.A, M.L.I.S.
> Guelph Public Library
> 100 Norfolk St.
> Guelph, ON
> N1H 4J6
> (519) 824-6220 ext. 276
> tbrenndorfer_at_library.guelph.on.ca
>
>
--
Jonathan Rochkind
Sr. Programmer/Analyst
The Sheridan Libraries
Johns Hopkins University
410.516.8886
rochkind (at) jhu.edu
Received on Tue May 15 2007 - 09:37:21 EDT