On Sun, 20 Jan 2013 19:42:22 -0800, Tennant,Roy <[log in to unmask]> wrote:

>Yes, a field must be used before anyone will pay any attention to it.
>A cautionary tale: Nine years ago, almost to the day, MARBI voted 7-0 to
>pass the proposal for a new field[1]: 258 Philatelic Issue Data.  Now in
>2013 I find exactly 8 records that have a 258 in the nearly 300 million
>record WorldCat aggregation. Eight.
>Now how was MARBI to know that nine years later no one would care about the
>258? They couldnąt, really, they just had to judge the proposal in front of
>them. But if we are going to chart a way forward it must be with full
>knowledge of where we may have gone astray in the past.

With reference to the cautionary tale, I wonder how many records have been 
created, in those nine years, in which 258 would have been appropriate? (In 
45 years involvement in cataloguing and with cataloguers, I have never 
created such a record).

On the other hand, current experience of mine, reviewing possible WorldCat 
records against book in hand, shows that the level of duplication among 
records for pre-AACR2 materials (and sometimes for later ones) is high, let 
alone the results of wholesale loading of non-AACR2 foreign records from 
around the globe, many of which have to stand because of the language-of-
cataloguing rule.  Perhaps the base number calls for discounting?  If a 
good, unique record requires an uncommon field in order to create a record 
which fully reflects a resource's attributes, perhaps the true conclusion 
may be that WorldCat was judged not to be the correct repository for that 
metadata to reveal the resource to its likely seekers?  

MARC systems have, after all, been used time and again for 
nonbibliographical materials -- even institutional furniture and equipment 
records -- because a library has a local database with good control and 
indexing features; I have myself entered a set of patron locker keys so that 
they could be controlled through the system checkout/checkin functions (I 
admit I cheated by bypassing many fields altogether).  We have probably 
passed that stage now, seeing as pretty well everyone has access to a 
spreadsheet program, but if a mistake had been made in selecting records for 
export, those key records might have ended up in Libraries Australia and 
thence even been copied into WorldCat.  And then turned up in some kind of 
exception table.

By all means look where we have been.  Uncommon data categories might not be 
inappropriate.  Uncommonness is a reason to prompt reflection and inquiry, 
not necessarily to exclude.  I wonder about some of the provisions and 
distinctions in RDA.

Hal Cain
Melbourne, Australia
[log in to unmask]