Professional Documents
Culture Documents
Cooking The Perfect Custard
Cooking The Perfect Custard
27
9 2003 Kluwer Academic Publishers. Printed in the Netherlands.
JEAN DRYDEN
Project manager of the Canadian~US Task Force on Archival Description (CUSTARD);
consultant CopyRight/- Copyright Advisory Services (E-mail: jedryden @interlog.com )
Abstract. A unified Canadian/US descriptive code is the aim of the Canadian/US Task Force
on Archival Description (CUSTARD). Within the framework of the International Council on
Archives' General International Standard Archival Description (ISAD(G)) and the Interna-
tional Standard Archival Authority Record (ISAAR(CPF)), this Task Force, an NEH-funded
project, is reconciling the US cataloging code, Archives, Personal Papers, and Manuscripts
(APPM), with the Canadian Rules for Archival Description (RAD). In this article, the learn-
ings of the process of blending two established descriptive standards are presented in a
preliminary way.
The connection between cooking and descriptive standards may seem at first
to be a tenuous one, arising only from the acronym for the Canada-U.S. Task
Force on Archival Description, known as the CUSTARD project. The Task
Force is one component of a project to produce a data content standard for
the description of archival materials by rationalizing the Canadian and the
American standards (Rules f or Archival Description (RAD) 1 and the second
edition of Archives, Personal Papers and Manuscripts (APPM2) 2 respec-
tively) within the framework of international descriptive standards. While
this seems far removed from the creamy dessert consisting of milk, eggs, and
sugar, the analogy is more apropos than is initially apparent. As the quotation
above suggests, custard is a simple, yet versatile, dish which can take many
forms and flavours. However, the cookbook also suggests that it is not neces-
sarily easy to produce a perfect product every time. A lot can go wrong - the
eggs can curdle, the mixture may separate, the heat may be too high or too
1 Rules for Archival Description (Ottawa: Bureau of Canadian Archivists, 1990).
2 Archives, Personal Papers and Manuscripts, 2nd ed. Comp. Steven Hensen (Chicago:
Society of American Archivists, 1989).
28 JEAN DRYDEN
low. Once prepared, the product must be carefully refrigerated or it will spoil.
A project to blend two established descriptive standards does in fact have
some similarities with cooking custard. What kind and flavour should it be -
a complicated zabaglione or a simple blancmange? Who should the cooks be?
Are the proposed ingredients compatible? When experienced cooks disagree
over the ingredients or method, how should the matter be resolved? The goal
is not to produce the perfect custard but to produce a recipe for custard and
all its variations that will enable archivists to produce many kinds of custard
that will turn out perfectly every time. As this article is being written, the
project has been underway for a year, and the cooking is still going on. The
final product will not be available for another year. However a number of
observations may be made about the process, and this article presents those
learnings in a preliminary way.
Background
The genesis of the CUSTARD project lies in the requirement to revise
APPM2 as part of a revision schedule instituted when it was adopted as a
standard by the Society of American Archivists (SAA). However, a number
of developments in the decade since the publication of the second edition
of APPM in 1989 made a strong case for a broader, more thoroughgoing
revision effort. For one thing, the proliferation of other national standards
such as RAD and the British Manual of Archival Description (MAD3), 3
and the development of international standards such as ISAD(G): General
International Standard Archival Description (ISAD(G)) 4 and ISAAR(CPF):
International Standard Archival Authority Record For Corporate Bodies,
Persons and Families (ISAAR(CPF)), 5 meant that there were more stand-
ards that must be taken into account in any revision process. Furthermore,
the development of the internet and related technological advances made
it possible to communicate information faster and more easily than ever
before. Closely related was the development of Encoded Archival Descrip-
tion (EAD) 6 which provided a data structure standard for communicating
3 Margaret Procter and Michael Cook. Manual of Archival Description, 3rd ed. (Aldershot:
Gower, 2000).
4 ICA Committee on Descriptive Standards, 1SAD(G): General International Standard
Archival Description, 2nd ed. (Ottawa: International Council on Archives, 1999).
5 ICA Ad Hoc Commission on Descriptive Standards, ISAAR(CPF): International
Standard Archival Authority Record For Corporate Bodies, Persons and Families. (Ottawa:
International Council on Archives, 1996).
6 Encoded Archival Description Tag Library, Version 2002. (Chicago: Society of Amer-
ican Archivists, 2002) and Encoded Archival Description Application Guidelines Tag Library,
Version 1.0. (Chicago: Society of American Archivists, 1999).
COOKING THE PERFECT CUSTARD 29
The Ingredients
One of the most important developments over the past two decades has been
the development of national and international standards for the description
of archival material and communication of descriptive information about
archival holdings, not just among archivists and repositories, but also to a
wider audience of potential users. However archivists were developing stand-
ards much earlier (whether they knew it or not) as soon as they tried to get
everyone to describe archival material the same way, be it within a single
repository or when developing rules/guidelines for a union list such as the
Union List of Manuscripts in Canada or the National Union Catalog of Manu-
script Collections in the U.S. Developing descriptive standards for archives
has always been a derivative process, that is, no-one develops standards
completely from scratch. All standards, no matter how narrow their applica-
tion, started from existing standards or from current practice. Those who
developed the formal national and international standards that are the focus of
this paper were no different. For example, the two North American standards
are first and foremost based on an existing standard, i.e., the bibliographic
model of Anglo-American Cataloguing Rules (AACR2R). 1~ However, RAD
also borrowed much from the first edition of APPM, and APPM2 borrowed
much from the first available RAD chapters. The international standards
subsequently developed were largely based on existing archival standards
(APPM2 and RAD), and started to move away from bibliographic models.
10 Anglo-American Cataloguing Rules, 2nd ed., 1988 Revision (Ottawa: Canadian Library
Association, 1988).
COOKINGTHE PERFECTCUSTARD 31
11 Jean Weihs, ed. The Principles and Future of AACR: Proceedings of the International
Conference on the Principles and Future Development of AACR, Toronto, 1997 (Ottawa:
Canadian Library Association, 1998).
32 JEAN DRYDEN
The Method
The project began in July 2001. The project budget provided for only four
meetings over the course of the project, so most of the work was accom-
plished through email. Drafts of elements were prepared by the project
manager and sent out to members of the group for their comments. After some
experience, it was decided that it would be helpful to focus the responses
by including with each draft a list of particular questions or issues to be
commented on, and a summary of changes made from earlier drafts. To
ensure that the debate about an element did not go on endlessly, it was also
decided that each element would undergo three rounds of comments, with
the second draft incorporating the comments on the first draft, and the third
draft incorporating the comments on the second draft. While working through
email is much faster than distributing paper copies of drafts and comments,
it is still not as satisfactory as working face to face. In an ideal world, the
process would permit frequent meetings at which key issues could be thor-
oughly thrashed out and clearly resolved. This was demonstrated at one of the
meetings, when drafts of particular elements were at a stage where a number
of key points needed to be resolved, and the ability to resolve them face to face
at the meeting was very productive. However, with so few meetings, it was
largely up to the project manager to make decisions about what comments to
incorporate, and justify what to exclude. Cost considerations precluded the
luxury of frequent meetings, but to have such an opportunity to debate and
resolve issues would make a stronger standard.
One of the biggest challenges for this project was the need to blend two
established archival traditions. While to the rest of the world Canada and
the U.S. may look like a monolith, the differences in archival practice are
significant. People were certainly aware of specific differences between RAD
and APPM2 (e.g., sources of information for archival description, inclusion
of dates as part of the rifle, use of 'papers' and 'records' as opposed to 'fonds'
in the title, use of subject in the scope and content, etc.), and some of these
issues were the subject of intense discussion. However the differences extend
beyond specific rules.
Totally aside from archives, the national cultures are very different. Amer-
ican pragmatism and the desire to produce something as quickly as possible
COOKINGTHE PERFECTCUSTARD 33
archivists become more familiar with them, they will recognize the need for
rules that cover very specific situations. If such rules are presented with the
element they relate to, rather than in another place in the standard, people will
be aware of them, and as their understanding of the rules and their application
becomes more sophisticated, they will start to see their utility and use them.
Inherent in this debate is a tension between seeing this standard develop-
ment process as an opportunity to raise the bar in order to improve the level
of descriptive practice, and the need not to make the rules either so difficult,
or so different from entrenched practices, that people will not use them. To
continue the culinary metaphor, providing a recipe that will permit moving
beyond simple baked custard to crbme caramel or crSme brul6e requires
adding more ingredients and learning new skills. The group did agree in
theory that this was an opportunity to provide more rigourous rules which
would result in better, more consistent descriptions, but when discussion
centred on actual rules, it was not easy to achieve the delicate balance between
more robust rules and straightforward user appeal. It must be noted, however,
that the success of the product does not depend on the standard alone. The
development and approval of the standard must be accompanied by education
and training, and institutional implementation. The standard itself need not
be diminished because it will require people to be retrained or will cause
institutions to change their procedures. It should be as complex as it needs to
be with training appropriate to the needs.
Underlying this appears to be a reluctance to acknowledge just how
complex description is. For example, the matter of dates may seem uncom-
plicated, but is enormously complex. Earlier standards recognized only dates
of creation. This standard is the first to acknowledge that there are many
types of dates and identify them. Dates of record-keeping activity, dates
of reproduction, dates of publication are all valid and may differ from the
dates of creation. This innovation was not accomplished without extensive
and sometimes heated discussion within the group. The Extent and Medium
element proved to be similarly difficult. While expressing what a fonds or
series consists of and how much there is of it may seem elementary, the
debate revolved around complex and inter-related issues such as the appro-
priate way to express medium, the overlap between physical and intellectual
forms of material, how meaningful extents expressed in number of containers
or linear measurement are to users, and whether the element should include
information about other physical details.
Closely related to the issue of complexity is what was called the "one-
stop shopping" issue, which came up repeatedly. On the one hand were those
who wanted "one-stop shopping", by which they wanted the standard to
include all relevant rules from other standards (e.g., AACR2R, Cartographic
36 JEAN DRYDEN
Materials, etc.) that archivists could ever possibly use so that they would
not have to consult anything else in order to describe their holdings. On the
other hand were those who wanted the standard to include only those rules
that applied to archival material, with numerous referrals to other standards,
particularly to AACR2R. Regarding the first possibility, it became clear that
including all possible rules would not only result in a very large volume, but
would also require regular monitoring of the standards from which they were
derived, and revisions as those standards changed. On the other hand, those
who wanted archival rules only (combined with referrals to other standards)
were reminded that many smaller archives would not necessarily have access
to all the other standards, and archivists who did not come from a library
background in which they had been exposed to AACR2R would appreciate
having rules which would assist in describing bibliographic materials found
in archival holdings. What resulted was an uneasy truce which might be
called "convenience shopping" in which selected bibliographic rules were
included. While the result may not be completely consistent, the validity of
this compromise will be tested once the standard is in use. If necessary, the
ingredients can be adjusted.
Terminology alone is a great problem. Any set of rules for archival
description must include a glossary which clearly defines key terms. This
project provided an opportunity to examine a number of definitions criti-
cally and redefine or clarify them as required. It quickly became clear that
definitions of terms tend to be borrowed repeatedly from one standard to
another, so that some definitions are decades old. In other cases, the standards
writers have made slight changes to definitions for no apparent reason - are
such changes the result of a deeper understanding of archival description, or
are they merely someone's preference? This project has reviewed the defini-
tions of some of the key concepts and attempted to clarify them so that the
terminology is fixed and used consistently throughout the standard.
The concepts of multilevel description are not well understood. The
application of the principles of multilevel description need to be more
thoroughly explored. A matter of great contention in this project was the
determination of what is the highest level of description. Some institutions
in both countries begin description not at the fonds level, but at the series
level. Using the term "at the highest level" seemed to beg the question and
permit description at any level. In order to accommodate the differences in
practice Principle 3 (that archival description is based on the principles of
multilevel description) was reworked numerous times before it was accept-
able to everyone. The group were mindful that the representation of multilevel
description works well in the paper world where the entire description is
presented and levels can easily be represented using typography or layout.
COOKING THE PERFECT CUSTARD 37
The Cooks
cooks spoil the broth, the rationalizing of two established standards clearly
requires a group effort. Obviously the team must include people who are not
only knowledgeable about the principles and practice of archival description,
but also are committed to the value of standards. During the past two decades,
an increasing number of people in the community have been involved in
developing standards. As a result, the CUSTARD team consisted of a number
of people who had been involved in various standards development initiatives
at the national and/or international levels, as well as in their own institutions.
Furthermore, not only must the team members be experts, but they must
also be seen to be experts and have credibility in the wider community. A
standard usually requires at least some change to the way that people work,
and it is normal to ask "Who did this, and what gives them the right to change
our customary practices?" While some in the wider community still might not
be happy with the changes, there is a better chance of acquiescence (however
grudging) if it is acknowledged that the people developing the standard do
know what they are doing. For that reason the team must be carefully chosen
to represent a variety of interests. If there was anything missing from this
group, it was archivists with expertise in non-textual media.
Everyone comes to the process with baggage and biases derived from their
training, their institutional practice, and their own experiences. However, the
team must be prepared to set aside their baggage, in order to produce the best
standards possible. This means being open to new ideas and being able to put
forward alternatives to the way they have always done it. It also means that the
role of each group member must be clear. If they come from large institutions,
are they there to ensure that their institutional practice is enshrined in the new
standard? Or are they there to contribute to developing the best rules possible,
even if the new rules do not match current institutional practice? Unless these
issues are clarified, there is potential for conflict of interest and uneasiness
among other members of the group who may see the process being driven by
protection of institutional interests.
When APPM and RAD were being developed, they replaced a hodge-
podge of idiosyncratic institutional practices. While people were anxious
about the effect that a new standard would have on their work, there were no
concerns about the implications for large databases containing records from
many institutions. Now, however, there are databases containing thousands
of descriptive records based on APPM2 and RAD, and thus concomitant
concerns that any new standard not contain major changes. It was not easy
to find the balance between maintaining continuity with existing structures,
and improving the rules so that better descriptions result. However, because
these rules were moving away from the bibliographic tradition on which their
predecessors were based, it was important that people were able to let go of
COOKING THE PERFECT CUSTARD 39
the bibliographic bias in favour of rules which better meet the requirements
of archival materials and archival practice.
No matter what their level of expertise or respect within the community,
the members of any group that is developing rules must also be willing and
able to devote the time necessary to review drafts and comment on them.
Some might think that blending two established standards would be a simple
cut and paste job, but the task proved to be very labour intensive. It was long
ago recognized that, if progress is to be made in a timely fashion, writing rules
for description cannot depend on volunteers trying to squeeze it in during
their evenings and weekends. This project had the luxury of a "chief cook"
who was able to spend much of her time in the kitchen writing and revising
the recipe - the process would have taken far longer had the CUSTARD
members been responsible not just for reviewing drafts but for writing the
rules as well. However the reviewing is not a trivial task, and this aspect
of the process was a concern. Draft chapters were circulated to the group
by email; their comments were posted to the CUSTARD webpage so that
all could read them. The first drafts of the first elements elicited a lot of
comment and an interesting debate on the purpose of title and the need for
detailed bibliographic rules. People responded not only to the draft rules, but
also to each others' comments. Unfortunately that level of debate was not
sustained as subsequent drafts and revisions were produced. Even though the
debate was interesting, it was ultimately unsatisfying because the discussion
just trailed off, and the issues were not clearly resolved online.
As more draft elements were sent out, the participation rate declined
considerably. Not everyone commented on the drafts themselves, and an
exchange of views on others' comments was rare. In an effort to improve
participation, the drafts were accompanied by a series of questions intended
to focus the debate on particular problem areas, and elicit responses to these
particular questions if there wasn't time to do a full review of the draft. Even
with that, people found it difficult to get through all the drafts, let alone
everyone's comments. While the project officer was clearly responsible for
reading and evaluating everyone's comments for possible incorporation into
the next version, that placed a large burden on one person. It was occasionally
helpful to have informal discussions with individual members in order to
resolve particularly thorny issues. The overall success of the process could
not assume that silence meant assent. Rather, success depended on getting
informed responses from the members of the group at each version of each
element.
40 JEAN DRYDEN
Conclusion
As this is being written, the recipe for custard is not yet perfected. However
the process has already revealed a number of lessons about developing
descriptive standards. The availability of new standards and revisions of
existing standards is evidence that much more attention is being paid to
archival description, and that people are thinking and discovering more about
it as they try to codify practice. However, despite all the work done in this
area over the past decade, this process has brought home how little we know
about arrangement and description. Discussions of arrangement in particular
are very limited in the professional literature, and a better understanding of
arrangement can only benefit our understanding of description. We also need
to know more about how archivists use rules for description. However, the
biggest gap in our knowledge concerns those who use the end product - the
archival descriptions themselves. What are their needs? Do they understand
the products based on our lovingly crafted rules? While work has begun to
learn more about users, it is far from complete. The development of this
standard unfortunately took place without the benefit of that knowledge.
However, if one were to embark on a similar exercise, certain lessons
relating to both process and people have been learned from the CUSTARD
project. Before a similar project begins, it should be possible to identify at
a high level the stages of the project, and settle a number of aspects of the
process, i.e., number of drafts, expected response time, specific questions
to focus responses, etc. The process should also include intentional team-
building, for example, participants should come to the first meeting prepared
to discuss their understanding of the project so that the group can agree on
a common vision of the goals and end product. If at all possible, the budget
should allow for more frequent meetings at key points. It is also very clear
that the budget for such a project must include sufficient funds to pay a full-
time staff person. The staff member cannot work alone and must be supported
by a steering committee or advisory group of some sort. The composition
of such a group must be carefully considered. Not only must the members
of such a group be acknowledged experts in the field, but they must also be
willing to consider new ideas and to work together to achieve a common goal.
The issue of whether one is representing one's institution must be clarified in
advance. One of the key learnings, however, is the need to make it clear to the
members of the advisory group the time commitment involved in reviewing
drafts, meeting deadlines, etc.
Has CUSTARD been a success? The proof of the pudding is in the eating
and as this is being written the recipe is not available. As the initial quotation
suggests, custard can be made in a variety of ways with different ingredients
COOKING THE PERFECT CUSTARD 41
to serve a variety of needs f r o m the simple dish for a child with a sore throat
to an exquisite cr~me brulre for an elegant dinner party. It is hoped that the
standard will enable institutions to describe a range of holdings in a variety
of descriptive products that can be as simple or as sophisticated as required.
As time goes on, the dish will only improve as m o r e people use it and as we
discover more about description. In the meantime, bon apprtit!
Appendix A
II. As a first step in the process of developing a common descriptive framework, the
participants have committed to the following:
a) Canadian archivists will explore the suitability of Encoded Archival
Description as an appropriate data structure for archival descriptions
compliant with the Rules for Archival Description.
b) American archivists will investigate the suitability of using the Rules for
Archival Description as a data content standard for finding aids encoded
using Encoded Archival Description.