[DCRM-L] hand coloring and new descriptions

[Noble, Richard]

[Gmail "send" button damnably close to font choice--click and groan. Trying
again]

​​

I agree absolutely with John Lancaster *in principle*, and I think the
principle is applicable to a focused bibliographical database--ESTC, for
instance. Our 990-pound gorilla of a utility, the WorldCat, is another
context altogether, and the one in which most of us have no choice but to
do our work.



Distinctions of issue at the level of colored/uncolored, alternate
dedications, partial re-settings, variations in inserted matter that may or
may not have mattered to the producers, c19 variant binding issues--none of
these can be maintained without well-informed cataloger judgment, which is
rarely available and infrequently applied, above all and most frequently in
copy cataloging, which has a definite bias towards one set of the choices
among "fast/cheap/good". At the *item* level--the basic scholarly research
level--all is simply chaos: God only knows what lies behind a holdings
symbol attached to a record, no matter how specific that record may be,
when the cataloger is constrained to employ the crudest possible matching
protocols.



I personally strive to maintain, at least, the distinction between
"concealed" editions, where MARC at least gives us the 250 field at the
basic matching level (don't know about Bibframe--not their highest
priority, I'd bet). Beyond that, however, there's practically no way to
evade the dreaded merge. Besides, when I discover a concealed edition for
what had been treated as a single manifestation and has attracted a fair
number of holdings (quite possibly in a record which is itself a merger of
multiple, often very sketchy records that may conflate some rather
obviously variant editions), I have *no idea* which of those holdings
represent one edition or the other. (On the other hand, some records
represent a false distinction of issue where there is only variation in
state of one or a few components of the edition--we've all seen it all
after the first 10,000 or so books. See for example OCLC #37043635.)


I cannot, in this situation, legitimately refashion an existing record
absent the ability to assess each copy and reassign the holdings. In such
cases I mercilessly enhance the record, and add a note on the order of
"This record represents two editions, printed from entirely different
settings of type" (the wording is redundant, but few people know what heck
an "edition" is, so I lend the others a clue), and provide a few diagnostic
details and, if I can, a bit of history by way of explanation. A researcher
is then alerted to the fact that individual copies need to be censused.
(See for example OCLC #13455452).

The colored/uncolored issues fit very well into this category, which
requires being realistic about what the facilities allow you to do
effectively. Those who care can then make a local note to indicate which
issue is in hand (I do wish that the WorldCat would allow for annotated
holdings, which function very well indeed in ESTC). *But*--this is a losing
battle, if my own experience is representative. At best, I'm still
considered useful for the ability to do this, or at least tolerated--I'm a
"legacy", a term less and less frequently used in libraries in its good
sense. Bibliography itself is such a legacy.

If we're talking about records at the institutional level, I agree that the
stated DCRM(B) principle is correct; but at the level of rules designed to
govern common practice in large shared databases, I think we need to
accommodate reality and leave stones on the path behind us, not crumbs, on
our way into the woods.

RICHARD NOBLE :: RARE MATERIALS CATALOGUER :: JOHN HAY LIBRARY
BROWN UNIVERSITY  ::  PROVIDENCE, R.I. 02912  ::  401-863-1187
<Richard_Noble at Br <RICHARD_NOBLE at BROWN.EDU>own.edu>


> On Thu, Feb 26, 2015 at 7:37 PM, JOHN LANCASTER <jjlancaster at me.com>
> wrote:
>
>> From DCRM(B) (which was based on a substantial amount of scholarly
>> discussion, not least Tanselle’s seminal paper, “The bibliographical
>> concepts of issue and state” (PBSA 69 (1975), 17-66, and the responses to
>> it over the years):
>>
>>  Issue
>>
>> A group of published copies which constitutes a consciously planned
>> publishing unit, distinguishable from other groups of published copies by
>> one or more differences designed expressly to identify the group as a
>> discrete unit.
>>
>> It seems pretty clear that versions of a printing designed to sell for
>> different prices, with different physical characteristics, constitute
>> different issues, whether those differences are in the illustrations, the
>> quality or size of paper, or the quality of binding, to name a few common
>> ones.  Both bookseller and purchaser would be quite clear which group of
>> copies they were dealing with in any given transaction, and would not
>> likely consider them the same.
>>
>> Appendix E states:
>>
>> As a default approach, the rules contained in DCRM(B) assume that a
>> separate bibliographic record will be created for each bibliographic
>> variant that represents what is referred to as an "edition" in AACR2 and
>> an "issue" in bibliographic scholarship.
>>
>> The fact that it may be difficult to determine for a specific copy
>> whether that copy was issued colored or not, does not invalidate the
>> fundamental distinction between the types of copies as issued.
>>
>> As to confusing researchers, I guess it depends on the researcher - if
>> one is interested in the physical characteristics, publication conditions,
>> and the like, it would be more confusing to have all the copies of both
>> versions lumped together as holdings on a single record, and to have to
>> sort them out by querying individual libraries (even if only by consulting
>> each of their on-line catalogues).
>>
>> John Lancaster
>>
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserver.lib.byu.edu/pipermail/dcrm-l/attachments/20150227/f0c2e7e8/attachment-0001.html>