This research paper was originally written in April 2002 for the LIS 593 course in Archives Administration at
the University of Alberta School of Library and Information Studies. It was proofread
and converted to a Web document to satisfy the Capping Exercise requirement of LIS 600. Please
address any questions or comments to Albert Sandor (formerly Peter Cherny).
ARCHIVAL ARRANGEMENT RECONSIDERED
Provenance and original order are fundamental principles used in archives for the arrangement of documents. These notions have attained the status that they currently enjoy with the gradual evolution of the archiving profession. Unlike the librarian, the archivist is concerned with describing and arranging records, which he or she accomplishes by adhering to the principle of respect des fonds. The two more basic principles underlying respect des fonds are provenance and original order. Although modern archives would not have developed as they are known today without the guidance of these principles, there is a continuing debate about the usefulness of arranging documents according to provenance and original order that relates to access, retrieval, finding aids, and total archives. The debate is not entirely about the validity of provenance and original order but about their application.
The purpose of this paper is to discuss the manner in which archival principals were interpreted and applied before the development of Rules for Archival Description (RAD). Provenance and original order will first be explored separately, and the relationship of each to finding aids and the dispute over total archives will later be established. Prior to the creation of a standard for archival description, much concern was expressed about how to make archives more user friendly in terms of document retrieval without departing from the idea of respect des fonds. The same concern arose about the desirability of separating records according to their formats or media in total archives and thereby eroding the principle of provenance. RAD has not entirely solved the problem of total archives, or even retrieval, because these are larger issues in the profession that cannot be addressed solely through archival description. However, the debates among archivists with regard to arrangement were especially vigorous in the absence of a widely accepted standard. These debates form the topic of this paper.
As organizations become more complex and interrelated, the principle of provenance has the potential to be an increasingly powerful tool for the management of information. According to Bearman and Lytle, its greatest weakness is not its irrelevance but the fact that it has not been fully exploited:
The key to the archivists' contribution to information management lies in their unique perspective provided by the principle of provenance as it concerns organizational activity, especially how organizations create, use, and discard information. Despite the insights provided by provenance, however, archivists have not exploited its potential for retrieval in traditional archival applications, and have not even attempted its wider application to the management of all information within their organizations. (14)
The authors cover territory in the area of records management, and their primary concern is to bring the application of the archival concept of provenance in line with the requirements of modern organizations. These requirements are the product of the structure and operation of modern organizations and the way in which they create, share, and store information.
Bearman and Lytle contend that archival theory, especially its view of organizations, is outdated and needs to be altered to suit current realities:
Archival theory has been strongly influenced by the nineteenth-century view of organizations. Classical organizational theory assumes that the typical organization is autonomous and sovereign. At the highest levels, the organization's actions and
the structures it produces are assumed to be the result solely of internally formulated policy. Even if this view were valid for a simpler time, it is far too simplistic for modern organizations operating in a world of multi-national corporations, inter-governmental units, regulatory organizations, and federal
programmes administered by state, provincial, local governments. (16)
They point to what is called a mono-hierarchical structure whereby a linear chain of command exists within a largely self-contained organization. Unfortunately, this traditional archival model does not take into account the apparent absence of hierarchy at the higher levels of some organizations or the fact that a hierarchical model may not always be applied across organizations because a level in one organization may have a completely different function from the same level in another (Bearman and Lytle, 17). To complicate matters further, two or more such organizations may be closely linked. Authority within an organization can be diffuse, delegated to task forces and committees or sub-contracted to staff with multiple responsibilities.
According to Bearman and Lytle,
Current archival practice overemphasizes the importance of hierarchy. [. . .] Provenance of archival records is indicated in archival information systems by terms which identify offices of origin and subsequent custodians (including offices responsible for the records); these terms are then linked in archival retrieval systems in a hierarchical schema which serves as a proxy for relationships between the actual offices of origin (19).
However, this distortion affects records-keeping practices because organizations are fluid and constantly changing. Offices that create records often do not control them, and agencies, offices, departments, divisions, and bureaus are created, abolished, merged, and divided (Bearman and Lytle, 19). A strictly hierarchical organizational scheme cannot adequately represent all the possible shifts and changes. Yet, the authors argue that archival theory remains wedded to such a model.
The overemphasis on hierarchy stems from the identification of the record group concept with the principle of provenance. The record group concept is similar to the shelf-order systems of traditional library classification schemes. In the authors' view, the first step toward making provenance as effective as the title of their article suggests it should be is to separate it from the record group concept and thus free archival arrangement from the intellectual constraints that a physical shelf-order classification system imposes on it. The authors summarize the origin of, and the reason for the attachment to, the notion that archives are hierarchical:
The intellectual lineage of this misconception is clearly tied to the implementation of the record group concept in archival institutions, specifically the attempt to arrange records on shelves to reflect the hierarchical structure of organizations (Bearman and Lytle, 21).
Using provenance information for providing access points to archival records is one solution that Bearman and Lytle propose to improve upon the current state of archival arrangement (21). Provenance can serve the same function as chronological, geographical, or subject information for retrieving records. It is further argued that the form and the function of records are elements within provenance information that can themselves be used together as a source of access points:
Because archival records are the consequences of activities defined by organizational functions, such a vocabulary can be a powerful indexing language to point to the content of archival holdings, without need for actual examination of the materials themselves or for detailed subject indexing (Bearman and Lytle, 22).
The concept "form of material" means something different than either the information derived from a detailed reading of the contents of records or the media on which they are written. Rather, it refers to the "commonalities in their structure." Journals, diaries, and day books would be nineteenth-century examples of different forms, and each form conveys something about the kind of information that the documents represented in that form contain. As a result, "archivists can thus know from provenance rather than from subject indexing certain elements of the intellectual contents of records" (Bearman and Lytle, 22-23).
Another solution proposed by the authors is to establish provenance authority records and separate them from archival descriptions of documents. They would provide access points for the name of the creating agency, its mission, its functions, the source of its authority, its dates of activity, and the name of the agency to which it reported or was related. The authors write that while archival practice does include a tradition of creating authority files, authority control issues have received less attention than in other information services such as librarianship (Bearman and Lytle, 23). The situation has changed somewhat since the article was written with the arrival of Rules for Archival Description (RAD). Like the Anglo-American Cataloguing Rules (AACR2r) on which it is based, RAD devotes some space to subject headings and other access points, obviously modified to accommodate the requirements of archival documents. Unfortunately, it is limited by its purpose, which is the creation of archival descriptions, and is not equipped to incorporate into its proscribed authority records the level of detail that the authors suggest would be needed to describe adequately documents produced in a complex, modern organization. In other words, RAD rules for making authority records may be fine for most archives, but are insufficient to expand the role of provenance as much as Bearman and Lytle propose in their article.
The authors go through several other important points about provenance that are beyond the scope of this paper to discuss in further detail. The main theme of the article, that provenance
is an indispensable archival principle that can be an even more powerful tool for arrangement than it has been up to this point, if it is applied with greater vigour, is stated in the following recapitulation:
Immediate steps for archivists to take in order to improve archival information systems have been suggested. In summary, they are as follows: view provenance information as a provider of retrieval access points; emphasize form of material and function in retrieval systems; establish provenance authority records; rigorously separate authorities from description or control of records; and integrate archival processes from records creation through records appraisal to records description. (26)
Like provenance, the concept of original order has been a part of archival practice for a long time. Although it is widely accepted by archivists dealing with governmental or institutional records, those who deal with personal papers have been more hesitant to embrace it (Boles, 26). It has been argued that objections to original order were valid in the past when manuscript repositories contained small collections but do not necessarily apply in the twentieth (and twenty-first) century when massive amounts of personal papers as well as organizational records are kept in one place (Gracy, 8). Nevertheless, archivists such as Frank Boles identify inconsistencies between usability and the principle of original order and question the latter's efficacy. The debate in this case is more fundamental than the debate over provenance because it concerns whether or not original order should always be respected under all circumstances. It is about the principle itself, rather than merely its application.
In "Disrespecting Original Order," Frank Boles quotes previous authors as saying that, despite the importance of original order, it ought not to be the sole basis for organizing archival material. Keeping user needs in mind is equally important, and provenance and chronology should be used together with original order for arranging records: "In fact, they go so far as to suggest that for personal papers original order is not even the best of the competing organizational schemes" (Boles, 27). Muller, Feith, and Fruin concern themselves primarily with arrangement on the record group level in their definition of original order, and, even at that level, their conception of the principle leaves room for rearrangement by the archivist. Below that level, these three representatives of archival orthodoxy "become ever more liberal in allowing the archivist to vary the documents' original order" (28).
Boles gives two reasons for the continued entrenchment of original order as a guiding principle in archives. The first is that it has adequately served its purpose, regardless of how imperfect it may be. In other words, it works. That fact alone gives it practical value. As for the reasons why it works, exceptions notwithstanding, Boles elaborates in detail:
Historically, the ordering of documents by their creator reveals information about the character and organization of the creator independent of the documents' content. Original order has evidential value. It also has archival value in that it
broadens the cardinal principle of provenance. Provenance insists that archivists respect a creator of a body of documents by maintaining that body of records as a distinct unit, neither adding to nor subtracting from the files. Maintaining the documents' original order amplifies this respect. (29)
The other reason, besides the functional, is more mundane but no less significant:
The second practical mainstay of original order is its economy. Faced with perpetually underfunded budgets, the archival community often has not had the capability to implement alternative organizational schemes upon large bodies of documents, whatever the theoretical considerations. Original order makes virtue out of necessity. (Boles, 29)
There is evidential value in the order imposed on documents by their creator, and it is for this reason that the distinction between the historical and the archival value of original order is muddy and rarely addressed in the literature.
Yet Boles points out that original order can diminish the evidential value of documents because filing is a secondary activity that normally involves less care and effort than the actual creation of the documents, and files are often maintained and organized by people other than the creator, a situation which decreases the evidential value of the filing system (30). Therefore,
when a filing scheme imposed on documents by their creator proves unworkable it becomes legitimate for the archivist to destroy the original order insofar as it is necessary to insure that the evidentially superior documents may be successfully used (Boles, 30).
The perceived connection between provenance and original order, however, makes it especially difficult for many archivists to tamper with the latter lest they violate the principle of provenance. To circumvent this problem, Boles makes a strong case for separating the two principles and considering original order independently so that it can be reformulated (Boles, 30). In his view, access and usability should be higher priorities than original order. He considers usability in particular to be a more pragmatic and vastly superior principle for respecting the evidential value of records:
Usability acknowledges the evidential superiority of documents over filing systems by placing primary emphasis on access to documents. Usability, however, is also mindful of the evidential value of filing order and mandates the preservation of the original filing scheme if it is usable, the simplest system available almost invariably being the creator's filing system. The framework of original order has no mechanism through which relative evidential values can be weighed. (31)
Usability has a closer relationship than original order to provenance because it allows the archivist to respect the integrity of the creator's work by moving beyond a simplistic adherence to the order in which the documents were created and embracing the thoughts of the creator as they unfolded.
The debate over total archives is a particularly interesting one within the topic of archival arrangement because it involves other broad topics such as finding aids and has not yet been entirely resolved, even after the creation of RAD. The exchange in Archivaria between Terry Cook and Andrew Birrell is a good place to start exploring what is meant exactly by total archives and why it has been a source of disagreement in the professional community.
In "The Tyranny of the Medium: A Comment on ‘Total Archives,'" Terry Cook, who sees total archives as a threat to the principle of provenance, states emphatically at the beginning that
Schemes to classify records by subject or some other artificial system, whether alphabetical, geographical, or chronological, are considered quite un-archival. They destroy utterly the evidential value represented by the original order of the records and render arrangement and description of large bodies of material virtually impossible (141).
The fact that original order is not the only, or even the best, way to preserve the evidential value of documents has already been discussed and needs no reiteration. Total archives are defined in the article as repositories housing records in various media. The problem for Cook lies not in the inclusion of differing media formats such as film, photographs, sound recordings and maps in one collection, but in the separation of records according to their formats, thereby violating the principle of provenance and destroying their respective fonds. He describes what has happened in the past when documents were arranged according to media:
For example, in the past many maps, plans, and pamphlets have been stripped from government files and removed to the custodial control of the appropriate division. In several cases, no reference was left behind that anything had been removed or the reference was so vague as to be useless (144).
Another problem resulting from fragmentation into archival divisions is the duplication of effort and expertise. This redundancy naturally occurs because each division functions independently from the others, and acquisition, custody, and public service are replicated across media (Cook, 145). The same waste happens in the process of description and the production of finding aids, although the author grudgingly concedes that the arrangement does have its advantages:
In the custodial arena, the media fragmentation admittedly lends itself to easier storage, handling, and circulation, although these aspects do not alone justify the present situation (Cook, 145).
Terry Cook's essentially conservative argument seeks to reassert the primacy of provenance in archival arrangement. Some of his solutions to the "problem" posed by total archives include increased coordination among divisions and using common, cross-media reference aids produced through such descriptive tools as PRECIS (Preserved Content Indexing System) (147). Since the article was written, RAD has come along as another potential solution. It has provisions for creating reference aids and authority records across different media. Its use makes arrangement in total archives more manageable and weakens, though by no means demolishes, Terry Cook's objections to arranging records by media. It is also suggested in the article that the administrative structure of archival institutions could be reorganized in a manner that would curtail the current tendency of archivists to become media specialists.
Andrew Birrell contends that Terry Cook's attack on media separation is based on an attachment to orthodoxy rather than on an assessment of the genuine requirements of archival collections as they exist today:
[. . .] I feel that he is stating little anew, but suggesting rather that we adhere even more rigidly to the principle of provenance. What he has seen as a growing erosion of a sanctified tradition is merely a practical difficulty of operation as we expand our concepts of what constitutes archives (249).
He invokes Schellenberg to support his claim that the application of provenance can be modified to fit changing circumstances. Historically, provenance was devised to serve the needs of textual records found in a chronological sequence and should be applied differently to records that do not fall into this category (Birrell, 249). Paraphrasing Schellenberg, he writes "[. . .] it does not matter if archival materials are physically separated as long as the principle of provenance is observed in the medium" (Birrell, 249). Quoting Schellenberg directly, Birrell goes even further and shows that provenance information is relatively unimportant in the case of pictorial records because photos can stand alone, a position generally regarded as "archival heresy" (249). He summarizes Schellenberg's interpretation of provenance as follows:
Schellenberg thus believes that the ‘universally venerated' principle of provenance in some cases is unimportant and in others need only be loosely applied. He plainly accepts that different media should be maintained, arranged and indexed separately" (Birrell, 250).
Birrell uses the example of paintings to illustrate his point. The works of a famous painter document his or her career or the cultural and intellectual history of his or her nation irrespective of the content of the paintings themselves. That is, the mere choice of certain media in which to create documents often conveys information and contains both aesthetic and evidential value that would be lost if documents sharing the same medium were not kept together. Recognizing the value of documents based on their media and not just provenance alone requires expert knowledge:
It is here that specialization has an advantage over generalization, for the medium specialist acquires an expertise and comprehension that would be denied if we were always to concentrate on becoming a ‘compleat' archivist (Birrell, 250).
Extending his criticism of generalization, the author disagrees with Cook's suggestion that one archivist can handle all media in an administrative unit because that view is suited only for textual public records collections and "Not all media can fit the textual tradition of archival handling" (Birrell, 251).
Neither does Birrell agree that arrangement according to media poses as great a danger to intellectual control as Cook would have his readers believe:
We are told that separation by medium has led to fragmentation of intellectual control. If this is true, it need not be, since non-textual records can quite easily be handled according to the principle of provenance. Intellectual control resides not in one archivist's having an overview of a whole record group, but rather in having proper finding aids and accession data which will provide an effective entrance into
a collection for the researcher. (Birrell, 251)
Once again, we return to the subject of finding aids, and, once again, RAD needs to be mentioned as a partial remedy. In this case, it has the tools to resolve some of the dilemmas posed by total archives. Finally, the supposed intellectual fragmentation caused by separate arrangement is mitigated by the fact that textual records normally attract a different clientele than visual or computerized records. Historians especially have tended to show a preference for textual records until very recently, and, as a rule, there has been little overlap among users of textual and visual or digital media (Birrell, 251).
In "Total Archives Come Apart," Ernest Dick et al. discuss the merits of both sides of the debate but clearly favour the traditionalist call for unified archives. They see problems with media separation in the area of acquisition:
Researchers who wish to trace the growth and development of a particular medium may well be served by current practise but others will find themselves faced, at worst, by peculiar gaps and absences in the provenances of collections and, at best, by odd mini collections of multi-media material within large media archives" (Dick et al., 225).
They acknowledge that separation has advantages in conservation, but even that acknowledgement is qualified by the observation that the technological sophistication brought to archives by separation sometimes came as a result of a misplaced emphasis occurring at the expense of technologically less impressive but more worthy programmes (Dick et al., 225). Another concern mentioned is the lack of coordination among archivists specializing in different media and therefore among different archival divisions. From this reiteration of one of Terry Cook's points, the following conclusion is drawn:
[. . .] finding aids in each medium become more detailed, more unique, and more specialized. What impetus will there be to search for commonality among finding aids of all archival media if it is not in the interest of any of the media to do so? Automated information processing is heralded as the magic tool that will painlessly reassemble the provenance of multi-media archives. Technology can only succeed in this task if there exists a collective will to proceed in this direction and that collective will can only exist if it reflects the necessities and priorities of the administrative structure. Separation of media, it seems to us, in its pursuit of total archives, is headed inexorably in the direction of separate archives. (Dick et al., 226)
This is a valid interpretation of perceived developments in total archives, but the authors did not explain why separate archives must be avoided at all cost.
What sets the archival profession apart from librarianship is its inherent conservatism and reluctance to embrace new technologies. Because it serves a different purpose in its emphasis on the preservation of unique records rather than the organization of published documents, a more cautious assessment of emerging technologies is understandable. Archives embody continuity and tradition to a greater extent than do libraries since they treat the form and media of their records with just as much respect as the content. The profession's cautiousness applies in equal measure to the topic of finding aids, which is in a much more advanced stage of development in librarianship. The questions posed by Nancy Sahli are indicative of the archival attitude:
Can we develop finding aids systems unified not only within an institution but throughout an entire country, that are capable of intermingling various types of media in the same basic descriptive format? Should we even attempt to? What problems will we encounter, and what possible benefits will we derive from making the attempt? (15-16)
RAD has since given us some workable answers to these questions.
Without mentioning total archives specifically, she addresses the issue of finding aids for different media, a crucial consideration for both supporters and opponents of arrangement according to media separation:
Despite specific differences inherent in the nature of the records and the information they contain, certain common descriptors such as title, date span, and volume, apply to all materials. Similarly, certain common index and access terms can apply, regardless of the type of record involved (Sahli, 18).
This quote is also an answer to critics who charge that media separation is disruptive and violates the integrity of records collections, not to mention the profession as a whole.
With the advent of RAD, archival arrangement is likely to become a less contentious topic. As more archives adopt a common procedure for the description of documents, the advantages of a universal standard of bibliographic control will become just as apparent to archivists as they have been to librarians for a long time (Abraham, 375). Of course, the subtleties of the profession and the judgement of individual archivists will remain. Also, the wide diversity of records held in archival collections makes the implementation of a universal standard more difficult and complex than it is in library collections. For these reasons, it would be accurate to say that the archival profession is more of an art than a science.
Abraham, Terry. "Oliver W. Holmes Revisited: Five Levels of Arrangement and Description in Practice." American Archivist 54 (1991): 370-377.
Bearman, David A., and Richard H. Lytle. "The Power of the Principle of Provenance." Archivaria 21 (1985-86): 14-27.
Birrell, Andrew. "The Tyranny of Tradition." Archivaria 10 (1980): 249-52.
Boles, Frank. "Disrespecting Original Order." American Archivist 45 (1982): 26-32.
Cook, Terry. "The Tyranny of the Medium: A Comment on ‘Total Archives.'" Archivaria 9 (1979-80): 141-50.
Dick, Ernest J., et al. "Total Archives Come Apart." Archivaria 11 (1980-81): 224-27.
Gracy, David B. Archives and Manuscripts: Arrangement and Description. Chicago: Society of American Archivists, 1977.
Sahli, Nancy. "Finding Aids: A Multi-media Systems Perspective." American Archivist 44 (1981): 15-20.
This page was last updated on: 5/26/03