The Making and the Keeping of Records : (1) What Are Finding Aids For?

Chris Hurley has been General Manager (National Archives Business) in New Zealand since 1997. He worked at the National Archives of Australia in the 1970s and was Victoria's Keeper of Public Records from 1981 to 1990. He writes on archives legislation and description.

Abstract

Different approaches to descriptive standardisation usually betoken alternative means for achieving a common purpose. The purpose is often said to be facilitating retrieval for research use. This view limits the scope of resulting standards and excludes those holding a wider view of the descriptive function. Agreement on the purpose of description (which is often assumed rather than debated) must logically precede attempts to achieve standardisation. This article asks what other purposes there are besides retrieval and argues that description is just as relevant to the creation and management of records.

The Making and the Keeping of Records : (1) What Are Finding Aids For?

By Chris Hurley

This is a refereed article. The principle governing all operations such as making-up, handling, repairing, etc. should be that the Custodian should endeavour to add nothing to, and take nothing from, his Archives, however insignificant the addition or subtraction may seem.1

British Records Association, 1937

Is the purpose of description to provide a documentary representation or descriptive surrogate for a record or a body of records (fonds) which has passed out of the recordkeeping process across an "archival boundary"? Or, is it to capture such knowledge of creation, management, and use as may be needed for records to exist?

Are finding aids composed, in other words, of data input into a finding aids system or data output from a recordkeeping system?

Should descriptions be encoded entities standing for the object of description in order to facilitate access? Or, should they be recordkeeping tools used to establish relationships with contextual knowledge?

The possibility of encoding the products of a recordkeeping system to aid and assist in retrieval is not the link some of us now aspire to make between archival and recordkeeping systems. We do not want to track the location of records of "enduring value" as they pass from one system to the other. We want to establish archival systems as the source for metadata needed for the recordkeeping task - providing recordkeepers with the kind of contextualising knowledge archivists are used to managing. That cannot happen if description remains enmeshed in collection description - circumscribed by location of records and by their appraised value.

Descriptions are as much a part of recordkeeping as a register and movement book. In a paper registry, incoming papers are put on a file which passes from hand to hand as business processes take place. The registration (date-stamping, classification and filing) is evidence of the processes through which the documents pass. Similarly, movements are recorded - both on the file cover and in a central movement register. The annotations and the register and index entries are the equivalent of the metadata encapsulating a record-object in an object-oriented system, giving the record-object its context 2.

The metadata encapsulated object (MEO) is a record because it has both content and context. The encapsulating metadata binds content to knowledge of circumstance. It is part of the record because without it a record cannot exist. For the same reason, register and movement cards do not merely describe the files they control but, together with the files, make up the record of transactions.

Finding aids support the continued existence of records after their migration from one system into another - out of an environment where it can be assumed that users bring contextual understanding with them. Recordkeepers have hitherto been witnesses whose testimony could (and would) have been called upon to verify the probative value of the records should a court have felt the documents did not speak for themselves. Records are kept "in use" as evidence, not only by practising negative virtues (adding nothing, etc.) but also by actively intervening to preserve knowledge of business and recordkeeping processes, without which they cease to be records at all - merely estrays.

A court, having to decide whether or not to admit a document and believing its context was not self-evident, would hear testimony from witnesses who had that contextual knowledge in their heads - requiring not merely the "testimony" of the document itself but also sworn evidence (knowledge) of recordkeeping activity. In a creating agency, the fonds has not traditionally been documented because all the records in that place belong to the same fonds. The fonds has re-formed inside the archives with each successive transfer. Archival description has simply been a postponement of what could have been documented at creation.

From this perspective, description does not formulate new knowledge, it captures knowledge which was always a component of the record - kept not in written but in living finding aids, the minds of record creators and users. Knowledge of the circumstances of creation and use, not inscribed or encoded onto the record itself, must be documented when the record moves into an environment where it is joined with records emanating from other creating environments, different systems, and other business processes. In this larger environment, contextual knowledge can no longer be assumed to exist in the minds of keepers and users.

Records have stewards who are unfamiliar with the arrangements in which the records were generated and it is necessary to distinguish records belonging to one process from those belonging to another. We must now document what has hitherto not been written down.

This is a perfectly workable statement of the role of recordkeeping in cyberspace and the purpose of archival description. In an electronic environment without physical boundaries the record has to be able to survive outside of the realm of the creator's contextual understanding from the moment of creation. The purpose of description in cyberspace is not to list records as they pop out of a record making machine but to sustain an environment in which records can continue to exist.

Record making thus continues while we are setting down or encoding essential knowledge about the circumstances of its generation. Documenting it is analogous to giving oral testimony in support of an exhibit and it can be distinguished from copying, remaking or rewriting the exhibit.

Some might call this "preserving" a record and I would not quibble with that. My disagreement is with those who want to distinguish making from preserving. That view portrays the record as a self-sustaining object independent of description. I see little difference between the pen-stroke which inscribes the record, the annotation which documents its use, and the description which tells us the context.

If any distinction is to be made, it is that the record maker documents a business process and the archivist documents a documentation process. It's all part of the recordkeeping business, defined as embracing record making and record keeping - creation and preservation. Contrast this with the view that archivists make finding aids after the recordkeeping process has stopped and stand outside the process, being part of a different (preservation) process3.

When manuscripts librarians describes deposits by documenting knowledge about the depositors and their activities, they (the librarians) become participants in the recordkeeping process, co-creators of the fonds, in partnership with depositors4. They add the hitherto undocumented metadata which cocoons the manuscripts and ensures their continuing evidential value. This "intervention" is necessary because the depositors neglected (if you like) to fully document those elements of context necessary for the records to be used by anyone else. That neglect needs to be repaired when records move out of the donors' possession, where such knowledge existed in their minds, and into the manuscripts collection.

It is neither here nor there whether the knowledge is documented at the same time the content is captured or at some other time - before or after that moment - provided, in all cases, that authenticity is guaranteed.

The need to guarantee authenticity may invalidate some attempts to capture contextual knowledge. It could be argued, for example, and without conceding the point, that the best guarantee of authenticity is to ensure that contextual metadata is captured at creation and never subsequently changed. Just as a court will sift testimony brought to establish probative value, we cannot accept just any archival description as validating a record.

Wendy Duff and Kent Haworth5 have described a comprehensive model for archival description and Barbara Reed6 has analysed the requirements for metadata in recordkeeping. Duff and Haworth posit the existence of an emerging consensus which Reed and I are simply not part of.

Is the difference (as Duff and Haworth imply7) simply a matter of "Australians" pursuing a different path? No. For that would suggest commonality of purpose and disagreement over method. This debate is not about the merits of different methods for achieving the same purpose. It is about differences of purpose.

Internationally, the last opportunity to resolve this was in Stockholm in 1993 when the ICA Descriptive Standards Committee decided not to discuss further a draft Statement of Principles which originally supported ISAD(G). Instead, the Committee moved straight into a consideration of the text of the standard. This has left the descriptive standards discourse rudderless (for everyone except those who didn't have problems with the Statement of Principles) because there is no agreed bench-mark (no statement of common purpose) against which to test ISAD(G) or anything else.

Proponents of the ICA standards8 hold them to be "theory neutral". I think this claim confuses more than it helps. Whether or not a standard is theory neutral depends on your point of view. A standard might appear to be "theory neutral" if it supports purposes you think you have in common with everyone else. But it will appear to be full of theory if their purposes aren't the same as yours. A flat-earther may propose a "theory neutral" route to Cathay: "always sail east after rounding the Cape of Good Hope". This will satisfy the Dutch and Portuguese, who never do anything else, but it will seem anything but "theory neutral" to Christopher Columbus (who wants to sail west) or the Wright Brothers (who don't want to sail anywhere).

The ICA standards are "theory neutral" for those working within what they think are the agreed bounds of "archival principle" and what they suppose are commonly accepted perceptions of the "purposes of description" 9. One doesn't even have to disagree with their principles and purposes to find their standards to be theory laden and very confining. A recordkeeping view will not reject those principles and purposes. It will comprehend and go beyond them. A recordkeeping descriptive standard would suit very well those wedded to the ICA principles, since it would serve their purposes and more besides. On the other hand, the ICA standards do not satisfy recordkeepers because their needs are the same only up to a point.

The recordkeeping view seems to be a minority view and losing strength. I believe that this minority view will prove to be what archivists (or those who replace us when we are shown to be unequal to the task) will need to re-engineer archival work into cyberspace. I fear that what is becoming the majority view will prove to be a dead end. If we had agreement on the purpose of archival description (really, and not just in the minds of some), that could be resolved now, intellectually, without further ado. Without that, we must await the verdict of history to find out who is right and who is wrong.

Encoding any data requires knowledge of how it is going to be used. If it relates to airline bookings, you must know how airline booking is done and what it's for. Archivists assume they know how archival description is done and what it's for - either out of reverence for traditions which (it is believed) settled these questions long ago or because the consequences would otherwise be uncomfortable. Some archivists refuse to discuss what we do - just how to do it. But you can't usefully discuss how to do anything if you don't really understand what it is.

System- or theory-neutral standards require a consensus of some kind about how archival descriptions are done and what they're for. Otherwise, the standards are exposed to a wider context in which alternative theories are possible. An "agreed" purpose is the theory upon which (paradoxically) any theory-neutral standard must depend.

Here is one statement of the purpose of archival description:

to assist researchers in locating materials relevant to their research ... [and] ... to identify and request the physical entities of interest to them10

On this view, descriptions are for locating, identifying, and requesting research "materials" (presumably by "researchers" frequenting archival facilities). The alternative view is that :

1. they are not primarily for retrieval, they are needed to make and keep records, and
2. they are primarily for use by the makers and keepers of records - not just for research in an archives facility.

This view deposes "research use" as the primary objective. It follows that description belongs as much outside as within the archives, begins when records are first made, if not before, and is shaped by requirements which are far more complex and diverse than merely satisfying "research" needs (as that term is used within this debate).

When records are received, archivists once thought they had licence to preserve and describe, but not to add or subtract. This was the principle set out in the 1931 statement from the British Records Association (negatively) forbidding addition or subtraction and (positively) requiring maintenance of the record in the form in which it is received - because otherwise it ceases to be the record that was received and becomes something else. Archives were inert objects to be catalogued like books in a library.

One can encode airline data by assuming that airline bookings are sufficiently like appointment diaries for the technology which supports the one to be adapted when dealing with the other. One can encode archival data by assuming that description is sufficiently like bibliographical activity for the technology which supports the one to be adapted when dealing with the other.

For a bibliographer archival descriptions are themselves objects or documents which can be standardised by type and format so that:

a set of rules for defining and expressing the logical structure of an archival finding aid ... allows software products to control searching, retrieval, and structured display of those finding aids. The rules themselves are applied by tags (or mark-up) embedded in the electronic finding aid.11

Here, the finding aid has the logical structure, not the records or the recordkeeping process. What a bibliographer seeks to encode is the product of a descriptive process. The process simply manufactures descriptive surrogates for records as an aid to retrieval.

The importance of structure in encoding archival descriptions is made by Kent Haworth : ... The difference in our point of view (US and Canada) reflects differing archival cultures. There are two archival traditions in the US: an historical manuscript tradition and a public archives tradition. There is one "tradition" in English-Canada: the total archives tradition.

The only reason I am using MARC in my archives is because I am based in a university library and have access to it. Most other archives in Canada dont have that access. Most archivists in Canada have not "graduated" from a library school with a foundation in library cataloguing. Hence we have never "taken" to MARC the way our many of our colleagues, notably manuscript curators, have in the United States.

Now that we have a data structure standard, the EAD, which is specifically designed for archival description, and accomodates one of the most essential features of archival description, the multi-level technique, which is defined in RAD, it is not surprising that many archives and archivists in Canada are assessing its usefulness in their settings and are beginning to apply it. It is interesting as well to note that where before archivists in the UK and elsewhere stayed pretty much clear of MARC-AMC, there is now extraordinary interest in the application of the EAD.

It would seem to me that many archivists in the US from the "Public Archives Tradition" would also do well to assess its benefits as a data structure standard. In the last analysis, I think we are all coming to appreciate that there are different lenses, (MARC, EAD,)that we can use to view archival desriptions and that this will be a positive benefit for our users as much for ourselves.

I am hoping that the EAD might just be the standard that will break down divisions amongst various "archival cultural traditions", both within the US and with descriptive traditions in other countries.12

Retrieving a descriptive surrogate and retrieving a record are seen effectively as being the same thing because once the description has been retrieved, getting your hands or eyes on the record is purely mechanical.

These surrogates are rightly distinguished from descriptions of books by how well structure can be represented. From a recordkeeping point of view, however, it is the recordkeeping system which has the structure - not descriptions of it. Description is part of recordkeeping, not a different process altogether.

No recordkeeper would deny that retrieval needs, including some beyond those of the records creator, may need to be met. Usability, which must include retrievability, is for some a functional requirement for recordkeeping. The same real world system can accomplish both purposes. Those who want to argue a logical separation between the roles of making and keeping records, however, necessitating a conceptual separation of the functional requirements for systems of archival description and recordkeeping, also seem to be pursuing a real-world separation.

Such a separation appears justified so long as archivists focus their attention on -

finding aids created by repositories ... [whose] ... common purpose is to provide detailed description of the content and intellectual organization of collections ... [in order to] ... provide access ... in a platform-independent electronic format ... [which will] ... assist scholars in determining whether collections contain material relevant to their research ...13

Methodologically, this approach represents a middle path between crude word-processing and a full database application. It gives structure to the information and a logical search path created by imposing a ranking on the items in the documents, the production of which is perceived to be the purpose of the process14. There are now plenty of examples available on the Internet of finding aids which follow this path15.

The aim of such systems is to formulate, structure, encode, standardise (whatever) surrogate descriptions of records so that these surrogates can be managed more effectively. They cannot adequately document the process or system which produces collections of records or comprehend how that process really differs from one which produces collections of books because they don't aspire to. If all I wanted to do was deal with collected records, I too would be happy to use systems which were essentially designed for dealing with collected books. It is because I don't want to deal with collected records that I won't borrow from systems which were developed to deal with collected books16.

Approaches to descriptive standardisation have mirrored early attempts at automation. A paper-based work process was "automated" merely by encoding the forms used to carry out that process in a paper environment. They were transformed into electronic versions of their paper counterparts and the same data was simply manipulated faster and in more imaginative ways. The work process which the paper forms represented was not re-engineered in any meaningful or useful way. Instead it was merely duplicated electronically. The possibility of achieving the desired outcome in a different manner (or achieving new outcomes) was not realised until a second or third generation of users began to understand the possibilities for re-engineering the processes themselves, instead of just duplicating them.

It was not to be expected that archivists would be able to reach a quick or easy agreement on issues surrounding the re-engineering of their descriptive methods. Agreement has not come by changing time-honoured processes into something relevant, but by duplicating them electronically.

Recordkeepers posit a completely different idea about archival description at the very centre of the process. For them, finding aids are not aggregations of surrogate records, the documentary products of a system for producing descriptions of collected records. Recordkeepers want to document the business and recordkeeping processes which are being undertaken, firstly, so that records can be made and, (only) secondly, to provide pathways along which records may be found. They seek data input which "allows software products to control searching, retrieval, and structured display" of the records themselves - not "of those finding aids". It is the records we have to manage, not the finding aids.

I once quoted the formulation given at the beginning of this article with approval17. Since then, I have learned better to distinguish between principles and the archival methods used to carry them out.

The British Records Association was upholding (as I was when I quoted it) the principle of originality. Nowadays we speak of authenticity and, while I would still wish to affirm that the BRA's formulation is a valid methodological application of a principle it was championing and I still subscribe to, I would no longer hold that the preservation of originality is the only method of ensuring it.

Preserving originality is an acceptable method for upholding authenticity, but authenticity can be achieved in other ways. This does not invalidate methods built on preserving originality. It simply recognises them as a means to an end rather than an end in themselves. Something very similar has occurred in the field of law. The Australian Evidence Act 1995 contains the following breathtaking provision :

Original document rule abolished
51. The principles and rules of the common law that relate to the means of proving the contents of documents are abolished.

Taken in conjunction with the rest of the Act, this section replaces a set of legal methods developed over many centuries dealing with admissibility in a paper world and substitutes a revised methodology for the virtual world.

I once incautiously said that this returned us to the thirteenth century. I meant to say that modern rules of evidence on admissibility of documents18 represent an accumulation of decisions reached in successive judgements on particular cases19. I thought the new law was taking a sensible approach by returning us back to the time when these rules began to be formulated around the practical problems of determining the reliability of paper (or parchment). Thus the courts could begin again to build up, case by case, new rules in a world of electronic documents where originality no longer guarantees authenticity. The new statute makes clear (s. 48, for example) what the courts should take into account when determining proof of content.

In other words, the law comes to terms with the modern world, recognising that methods based on examination of an original are no longer adequate, though not invalidated. The aim has not changed, viz. to know when documents are good evidence. What has changed is the way of ensuring it.

Archivists face similar issues. Once, we placed a boundary between record making and record keeping, between records and archives. Records evidenced action. Archives preserved evidence. Preservation meant maintaining originality - neither adding to nor subtracting from - because change was held to obliterate the evidence.

Such ideas may seem quaint to some of us now, but although the life-cycle is today increasingly repudiated many are still trapped intellectually within its paradigm.

In the European cultural tradition, recordkeeping developed to meet pragmatic not research needs:

... it would be little of an exaggeration to say that all of the successor states to the Roman empire are marked by their employment of writing in governmental and private transactions, and by their attempts, however circumscribed, to preserve the resulting records, and that for practical rather than antiquarian reasons.20

Reliability and utility depended upon satisfying requirements which can be readily understood today:

Where there is certainly an area of significant differentiation ... is ... between those societies in which scribes were employed by the courts to draw up records of the proceedings and other related texts, copies of which might be presented to the successful party in a dispute, and those in which the recording or otherwise ... was left entirely to the latter's discretion. Although the second way of proceeding would usually involve attestation of the record ... it invariably produced a simplification in the character of the records, a decline in generic sophistication ... [and] ... Ultimately an imbalance in the availability and employment of written records affected the judicial processes themselves.21

One way to ensure authenticity was transmission into an archives - private (e.g. manorial), public (e.g. the gesta municipalia), or semi-private (e.g. monastic)22. They appear to have left few, if any, examples of what we would think of as finding aids. Procedures existed, however, to provide contextualisation needed to ensure the preservation of evidence:

Even in societies and periods in which the written record predominated it was never allowed an exclusive role. Procedures existed to subject documents to testing, not only in terms of the internal soundness of their drafting but also by requiring support from evidence produced orally by witnesses and/or by the invocation of spiritual sanctions through oath-taking and ordeals.23

Even when something very like a finding aid was produced, e.g. cartularies or calendars containing copies or summaries of charters, it has been argued that their purpose was not to preserve or retrieve the originals:

Traditionally, diplomatists have given low priority to the study of cartularies as such, using them primarily to reconstruct texts of lost originals with little regard to the nature, function, and history of this genre. Examination of their contents focuses on the identification of genuine, forged, or interpolated texts which, properly categorized by the techniques of diplomatics, can then be exploited as though they were originals. When editing cartularies, most nineteenth- and twentieth-century editors have ignored the organization of the cartularies themselves ... In other words, most scholarly attention has focused on eliminating the cartulary itself in order to provide transparent windows into the original archives of an institution. This process was considered legitimate because the cartulary was considered a self-evident attempt to preserve the contents of the institution's archives ... Because of such assumptions about the unproblematic nature of these collections, the history of cartularies and similar collections has yet to be written.24

Archival descriptions designed to meet long term antiquarian interest in those documents which happen to survive is a modern idea. Life-cycle archivists gather in materials from the different places in which they were kept. Their antiquarian collections are unlike mediaeval gesta municipalia (the "place" in which documents were lodged as part of the recordkeeping process). Documents are moved, by life-cycle archivists, from the "place" in which they were kept to "another place" after they have ceased to be part of a recordkeeping process - as a method of preserving them. In these "other places", finding aids ease the path of scholars who would otherwise be unaware of their context.

Their descriptions are the handmaiden to preservation and (while respecting this and preserving that) they play no part in records making. Their job is to keep records which somebody else made. They are like photographers taking baby photos. They make representations of the end product but they never participate in the creative act.

This noble, if flawed, mission statement somehow became debased in modern archives parlance into an exhortation to assist researchers to locate materials and identify and request the physical entities which interest them.

A distinction must be drawn between their finding aids (guides produced by archivists) and control records (registers, indexes, and so on produced by record-creators). Transmittal lists, used by records creators to propel their creations across the archival boundary, seem to have been fitted, without any sense of conceptual difficulty, into the category of finding aids when it was inconvenient to redo that work ourselves.

In the dreary world of the life-cycle, records were authentic if original, unchanged, and produced from an unbroken chain of custody from the creator to the archivist - ensuring against falsification.

The common law never accepted such ridiculous notions - never regarded originality per se as a guarantee of authenticity. Originality, under the best evidence rule, was neither necessary nor sufficient for evidential value. Evidence is not a quality like a colour or texture adhering to a document, unchanging and unaffected by circumstance. The law evaluates probative value by taking into account circumstances - and testimony about (knowledge of) a document and its use. A document may be good evidence of one thing or in one circumstance, and not of or in another.

In a legal sense, documentary evidence was always a compound of the original document itself (internal soundness) and knowledge about the document given in testimony by witnesses as to the making, keeping, and uses of the document (contextualisation). Archival description (knowledge of creation, maintenance, and context) is a form of testimony going to the credit of the documents with which we deal.

Contextualisation can be assured by placement. Preservation of contextual knowledge derived from placement has been an important strand in archival thinking. In the virtual world, we are questioning whether placement is the only or the best way.

Recent debates in this journal on the issue of archives as place (continued on the aus-archivists listserv) brushed past some questions concerning the "archival boundary". Debaters stalk round this matter like a mongoose approaching a cobra. Understandably. Mere mention of it can call forth fountains of (not always comprehensible) prose from some of the least retiring members of our profession - including me!

In my contribution to that debate I said the archival boundary was principally about method, not about "place":

... I can bring records into my repository without taking them across an archival boundary (depending on how I choose to treat them) while my neighbour (who follows a custodial path) erects such a boundary and forces records to cross it when he takes them in. It follows that it is equally possible to construct (or choose not to construct) such a boundary when deciding to "leave records with the agency".

In other words, the archival boundary is a creation of our choice of archival methods. I believe it is possible to fulfil the archival mission by using methods which do not result in the creation of such a boundary - indeed that the creation of the boundary is inimical to fulfilment of the archival mission. Others disagree.

The alternative is to believe that the boundary or threshhold is essential to its fulfilment. I have always assumed that my disagreement is not with those who wish to assume physical custody, but with those who believe in the archival threshhold.

This has been quoted back to me without an important qualification (well, I thought it was important) which I made:

At the end of the day, if you believe that the archival boundary or threshhold is necessary (or that it is not co-extensive with the recordkeeping boundary - i.e. if you think that records can exist on either side of it) then you are ultimately committed (I think) to the custodial view.

Advocates of the archival boundary attach at least two meanings to the concept and the qualification was intended to indicate disagreement with only one of them.

One meaning seems to do with ensuring evidence is maintained. The argument seems to be that this can only be done within the archival boundary - where a set of rules and procedures protects the record from threats to its "record-ness", e.g. from tampering.

I have no difficulty with an archival boundary which establishes a set of rules and procedures within which the requirements for recordkeeping are satisfied. I might say recordkeeping or evidential boundary, but most of what is said by those who use the other term I can subscribe to and, if the two boundaries are identical, then I have no quarrel with anyone.

The second meaning seems to be about distinguishing the role of the records creator from that of the archivist or record preserver - logically, if not temporally, along the life-cycle. From this viewpoint, the boundary separates two activities and, by extension, different processes or systems. With that, I cannot agree.

Some of the most impressive finding aids I have seen are the scholarly products of the English County Records Offices of about forty years ago. These are substantial volumes, many of them, handsomely bound and representing a high level of scholarship. They are immensely helpful, I imagine, in "assist[ing] researchers in locating materials relevant to their research".

Their data content is not very different from what one might find in any piece of archival description, though more fully and elegantly presented. It is not organised, however, in any very systematic or standardised way (consistently between one finding aid and another) into the strata or levels analysing structure or relationships (superior/subordinate - controlled/controlling - predecessor/successor).

Each finding aid stands alone - giving a homogeneous description of the archives being described. Where it is necessary to deal with collateral records or recordkeeping processes, references to these are incorporated into the description.

These references do not operate as entities within a descriptive system. Data concerning collateral records or recordkeeping processes is not linked in through cross-references systematically established. It is this missing process of structuralisation, rightly emphasised in Canadian and international work on descriptive standards, that differentiates archival from bibliographical description.

These standards distinguish between data content and the way data is organised, presented, and used (the system). When discussing data content, however, it matters very much what assumptions you make about the kind of descriptive entities it will populate and how they will be used. Data which is identical as to content but is used differently constitutes different kinds of data. Content standards, in other words, cannot be theory-neutral.

It is essential to know whether data is related or associated. Associated data is part of the archival description (a characteristic of the entity being described). Related data depicts a reciprocal connection between two entities. The same idea - who created these records? - may be related data or associated data, depending on how it is used. Record groups associate provenance data because the provenance statement is incorporated into the description of the records. Series descriptions relate provenance data because the provenance statement simply points to a separate descriptive entity showing how related and when related.

Archivists can participate in recordkeeping processes by documenting complex relationships between records and context. Records must be placed in context - in time and place - by fashioning descriptive entities and documenting relationships.

This is how we can locate them into a time-bound, evidential cocoon of meaning. To understand the record and derive evidence, it must be interpreted not by reference to our observation of it in the circumstances obtaining when we access it, but by understanding the circumstances which existed at its creation and the changes since.

Observe how confusion is dispelled when associated data (Verse One) is transformed into related data (Verse Two) :

Relationships must be reciprocal, however. Applying the multi-level rule turns perfectly good related data back into associated data26 by constricting the nature of the relationships which it is possible to show between separated entities. Here is what happens (with apologies to W.S. Gilbert) :

Though I was a tiny prince
Twenty years ago.
She ain't gained upon me since
Twenty years ago.
Now, she's twenty-one, it's true;
But, blow me down, I'm forty-two!
Unless you want this happening to you,
Many to many show.

The two fundamental issues for discussion concerning archival description are, therefore, what the descriptive entities should be and what are the relationships we need to show between them. In the second part to this article (sub-titled "The Tyranny of Listing") I will endeavour to use Frank Upward's continuum matrix27 to explore these issues.

ENDNOTES

1 Sir Hilary Jenkinson et al. quoted by Maurice F. Bond, 'The British Record Association and the Modern Archive Movement', in Essays in Memory of Sir Hilary Jenkinson edited for the Society of Archivists by A.E.J. Hollaender, Chichester, 1962. Bond cites the source as a Statement of Principles (not available in print) which was appended to the Agenda Sheet for the 1931 B.R.A. Conference in B.R.A./A/2.
2 David Bearman, 'Item Level Control and Electronic Recordkeeping' Archives and Museum Informatics, Vol 10 No 3, 1996, pp.207-245.
3 See Frank Upward, 'In Search of the Continuum' in Sue McKemmish and Michael Piggott (eds), The Records Continuum: Ian Maclean and Australian Archives first fifty years, Ancora Press, Clayton, 1994, esp. p.114.
4 Opponents of this view will point out that the hand of the archivist must be visible. This is true. We will want to know the identity and role (function) of every separate actor in the record making process (including the archivist). The fact that the archivist (and her role) should be as visible as every other actor does not disenfranchise her from a role in creating the record. What uniquely distinguishes the "archivist" from all other actors in the recordkeeping drama is her special mandate to think outside the process and build that perspective into her work. In Jenkinson's terms, the archivist has a licence not to be unselfconscious.
5 Wendy M. Duff and Kent M. Haworth, 'Advancing Archival Description : A Model for Rationalising North American Descriptive Standards, Archives and Manuscripts, Vol 25 No. 2, Nov 1997, pp.194-217.
6 Barbara Reed, 'Metadata: Core Record or Core Business?', Archives and Manuscripts, Vol 25 No 2, Nov 1997, pp.218-241.
7 Duff and Haworth, op. cit., p.215.
8 To view the ICA standards go to http://www.archives.ca/ica/cds/isad(g)e.html for ISAD(G) and to http://www.archives.ca/ica/cds/isaar_e.html for ISAAR(CPF
9 Duff and Haworth, op. cit., p.205.
10 Archival Finding Aids downloaded from http://www.loc.gov/ead/eadfind.html on 23 December 1997, at 1138 hrs (New Zealand Summer Time).
11 Downloaded from PRO(UK) website http://www.open.gov.uk/pro/ead.htm at 1205 (New Zealand Summer Time) on Thursday 8 January, 1998.
12 Kent Haworth posting to EAD@loc.gov on 13 January, 1998, at 03:34:18 am re : MARC and EAD (fwd).
13 Downloaded from Berkeley EAD site at http://sunsite.Berkeley.EDU/ead/ on 9 January 1998, at 1130am (New Zealand Summer Time). I have, of course, chosen quotations about descriptive purpose from EAD sites which illustrate my argument, but compare these with the statement of "Purpose and Methods..." given by Duff and Haworth on p.204 of their article. The purpose of description is to "provide access". Authenticity is proved by documenting the "chain of custody", reflecting "arrangement", and documenting provenance and use. This last is close to a recordkeeping goal, but it is damned by mixing it up indiscriminately with archival methods. The same problem occurs in the list of ways of enabling understanding - especially in the exhortation to describe "from" the general "to" the specific (the dreaded multi-level rule).
14 Many archivists, of course, believe that the standardised archival descriptions which have developed so far have structure and ranking. My contention is that the use of the so called multi-level rule deprives them of much of the usefulness this would have for recordkeeping - see below where I distinguish between associated and related archival data.
15 Go to http://sunsite.berkeley.edu:28008/DWEB-COLLECTIONS.
16 Though, as I indicate above, I accept their utility for custodial programmes in dealing with inventories and the like if a full database option is unavailable.
17 Chris Hurley, 'Personal Papers and the Treatment of Archival Principles', Archives and Manuscripts, Vol 6 No 8, Feb 1977, p.364.
18 In common law, documents are (or were) strictly inadmissible. They were excluded under the hearsay rule. The common law of evidence relating to documents is (or was) in fact a mountain of exceptions to the fundamental exclusionary rule. By 1995, these exceptions had become so numerous and the importance of documents as evidence in almost every branch of the law had become so fundamental that the exceptions were, in effect, a body of positive law regarding their admissiblity.
19 This matter came up again recently on the aus-archivists listserv. The difference between the new law on documentary evidence and other parts of the new Australian Act is significant. Elsewhere, the Act codifies and builds on existing law. In the parts dealing with documentary evidence there is a much more radical sweeping aside of the existing body of law, though as Livia Iacovino rightly points out this too builds on recent common law and other statutory developments). In this area, the new Act jumps the groove (as it were) in important ways will require a new body of case law to support it.
20 Wendy Davies & Paul Fouracre (eds), The Settlement of Disputes in Early Medieval Europe, Cambridge, 1992, p.207.
21 Ibid., p.211.
22 Such distinctions, of course, would be incomprehensible to the medieval mind.
23 Wendy Davies & Paul Fouracre, op. cit., p.213.
24 Patrick J Geary, Phantoms of Rememberance: Memory and Oblivion at the End of the First Millenium, Princeton University Press, 1994, p.83. Geary goes on to argue that cartularies (as copies) had no value or function as legal instruments, but were intended by the copyists to serve a completely different purpose unrelated to the evidential requirements of the original charters.
25 W.S. Gilbert, Princess Ida (1884) Act I.
26 Chris Hurley, 'Ambient Functions - Abandoned Children to Zoos', Archivaria No 40, Fall 1995, pp.22-26, illustrates the point.
27 Frank Upward, 'Structuring the Records Continuum, Part One: Post-custodial Principles and Properties, Archives and Manuscripts, Vol 24 No 2, Nov 1996, pp.268-285 and 'Structuring the Records Continuum, Part Two: Structuration Theory and Recordkeeping', Archives and Manuscripts (May, 1997), Vol 25 No 1, May 1997, pp.10-35.

...