[uf-discuss] Citation format straw proposal on the wiki

[C. Hudley]

On 3/28/06, Michael McCracken <michael.mccracken at gmail.com> wrote:
> Hi, so since we've got a large number of examples on the examples
> page, I think it's time to float an initial version of a citation
> microformat.

I think this is a good start.  I'd vote for volume and issue number
being separately demarcated.  Also, I've asked before about spinning
off a single-class microformat for a URI; this is another case where
having that might save some work here.

In general, though, there's a disconnection in discussion here, on the
cite-examples page, and on the cite-formats page that should be
cleared up to really make progress.  I think the cite-brainstorming
document makes a usefully accurate and crucial distinction that we
need to be more rigorous about (under "Citation vs. media-info"):

  "What distinguishes a cite from say media-info (e.g.
media-info-examples) is that a cite is a reference to something
explicitly external to the current piece of content or document,
whereas media-info describes information about content embedded or
inline in the current document."

The collected examples and notes don't exactly follow that.  For
example, on cite-examples there are snippets from "product pages" and
publisher "detail pages" (see the first several listed examples).  By
the statement above, these are not citations regarding external
content.

As a librarian, I'd thump on the distinction between "bibliographic
records" and "citations/references" here; the non-citation examples
are "bibliographic" in that they are descriptions of content which
provide people with enough information to make a choice about whether
the content at hand is the content they want.  Citations, as
references to external content, provide just enough information to
people interested in the content at hand can follow a line of thinking
through some other external content.

A citation, once somebody decides they need to see what it references
for themselves, should provide just enough information so that that
person can then go find that item in whatever way makes sense,
immediately or thirty years later... ultimately, if they can find it,
they will have to make the bibliographic decision -- the "is this
content what I'm looking for" bit -- about the item that was just the
citation at first.

Sure, that probably sounds perfectly academic, and the lines between
these activities a lot more blurry than this, since the use cases
overlap all the time, especially online.

But, the helpful part of the distinction is that when you're examining
a reference external to the content at hand, you don't need the full
bibliographic details.  This immediately lets you eliminate stuff like
subjects/keywords, pricing, copyright, licensing, owner, format,
coverage, and audience, all of which are mapped in the
citation-formats.

You wouldn't put pricing information in a resume or CV.  You wouldn't
put a copyright statement in a reference at the end of a journal
article.  You wouldn't list "audience" in a reading list for a
seminar.  Etc.

Y'all have your hands full enough dealing with media-info. :)  To make
progress on citations, I'd go for the simplest possible subset(**),
and your strawman looks close to that.  I'd also suggest that we clean
up the cite-* pages to reflect this distinction more aggressively so
that folks new to the process might be directed more clearly.  If this
seemed agreeable to people I'd be happy to help with that, but I won't
touch anything unless somebody says to.


(** to be precise, I'd just reuse the profiles defined in OpenURL for
journals and books to start, since those are explicitly barebones
subsets for following references, but you've heard that from me
before...)


--
C. Hudley
We Know The Truth, Inc.