Jonathan Rochkind wrote:
Hal Cain wrote:
But then I've been proceeding from headings in sequential lists to
works and back again, as a means of exploring resources, for more
years than I care to recite. Is the contemporary frame of reference
really so different?
I believe/suspect it is, yeah. (and you were really proceeding from
headings in a sequential list to MANIFESTATIONS, weren't you? :) ).
Well, not necessarily. Often enough I was pursuing WORKS, and often
works in particular MANIFESTATIONS (e.g. speeches of Demosthenes with
Sandys's notes -- setting aside, for the moment, whether Sandys's notes
constitute an additional, contained work; I simply wanted text +
Sandys's commentary) rather than any particular publication in any
particular publisher's edition. FRBR provides a framework for
developing ways of working with different levels of data; these are
abstractions and may complicate the process of identification, selection
and obtaining the best match to the searcher's inquiry.
You'll note that there I'm speaking as a user, rather than as a
cataloguer. The cataloguing is important but the user's search is the
determining characteristic for assessing usefulness. And if 79% of
works in WorldCat exist in only a single manifestation and expression
(as I seem to recollect) it may be we're constructing problems where
none exist.
If you think I'm doubtful of the value to users of a lot of what's being
discussed, you may be right.
Chris Beer commented: "Any good system (not just OPACS, legacy or not)
will provide multiple means of use and access - usability is about
addressing user needs, and as long as some users feel the need to
browse, it should be provided as a function.)"
That's a most useful observation. There is no one single path for a
search. Maybe Jonathan dislikes having to decide at the outset whether
to start a search for title, for author/editor/other creator, or by
subject (topical, geographical, chronoogical, other)? I think following
Google down the undifferentiated-search path is a disservice to searchers.
I understand that there are things you can do with our data in a
legacy-style browse list that you can't do in a legacy-style keyword
search. And that's all our legacy-style OPACs provided. But I do not
think that contemporary users are happy with the "browse search". It's
an odd sort of display for people who actually are used to software
search interfaces, at least for the contexts we expect them to use it in.
I absolutely fail to understand what's so different about those contexts
-- beyond looking for information through a search engine as compared
with using an encyclopedia (or a whole library reference collection) --
for sure, the search engines find things the encyclopedias probably
won't, without a whole lot of digging and cross-checking; but then it
often happens that one needs to cross-check web sources too, and
establish their authority and reliability.
Beyond launching a whole lot of fragmented pieces of bibliographic data
on the internet (mostly bereft of context which is what enables us to
assess their significance at a glance), what will RDA make possible that
can't be done by disintegrating (maybe repurposing) the
finely-categorized elements of data now contained in MARC records?
I think we're going to start seeing
more better interfaces that try different ways, and that in general, I
suspect (evidence would be good) our users are NOT happy with a browse
search, and would be happier by alternative interfaces filling the same
purpose.
I am at a loss to understand what you are suggesting. Examples, please?
To get back to the list topic, the point of RDA's "vocabulary
explication and formalization" aspect (what I actually think is the most
important part of the RDA effort, although the part that's had the LEAST
resources dedicated to it) -- ought to be modelling our data in a
robust enough way that is _interface independent_.
I'm already on record as saying I think that library system vendors and
their offerings are the weak link in the chain of bibliographic control.
I'll also admit I don't bother any more with using what I think is the
all-time greated browse toll for information retrieval: the white pages
telephone directory. The online search is far more convenient!
However, occasionally it do need the alphabetical listing, typically
when I don't have the complete, accurate information (is it Smith, J P
or Smith, P J?).
Part of the way forward must be to construct interfaces that encourage
users to use what they already know in searching. Of course, the
process of entering chunks of data -- maybe not precise data -- in half
a dozen boxes in a search screen is itself forbidding and encourages
indiscriminate entry of generalized terms; and personally I hate long
drop-down lists of terms (MARC 21 geographic codes; language codes;
chronological codes -- a complete turn-off!) even if I understand the
instructions on the screen! This is why systematic lists of possible
terms remain useful, at least for some of us. They are efficient ways
of making choices.
Hal Cain
Dalton McCaughey Library
Parkville, Victoria, Australia
hec...@dml.vic.edu.au