Skip to the content | Change text size

Structuring the Records Continuum - Part One: Postcustodial principles and properties

Frank Upward

Frank Upward is a senior lecturer in the Department of Librarianship, Archives and Records at Monash University. He is an historian of the ideas contained in the Australian records continuum approach, and an ex- practitioner within that approach.

The records continuum is becoming a much used term, but has seldom been defined in ways which show it is a time/space model not a life of the records model. . Dictionary definitions of a continuum describe such features as its continuity, the indescernibility of its parts, and the way its elements pass into each other. Precise definitions, accordingly, have to discern the indiscernible, identify points that are not distinct, and do so in ways which accomodate the continuity of change. This article, and a second part to be published in the next volume, will explore the continuum in time/space terms supported by a theoretical mix of archival science, postmodernity and the 'structuration theory' of Anthony Giddens. In this part the main objectives are to give greater conceptual firmness to the continuum; to clear the way for broader considerations of the nature of the continuum by freeing archivists from the need to debate custody; to show how the structural principles for archival practice are capable of different expression without losing contact with something deeper that can outlive the manner of expression.

This article was first published in Archives and Manucscripts, 24 (2) 1996.

This is the first instalment of a two part article exploring the records continuum. Together the articles will build into a theory about the constitution of the virtual archives. In this part I will examine what it can mean to be 'postcustodial', outline some possible structural principles for the virtual archives, and present a logical model for the records continuum.

In the above paragraph, concepts of the 'continuum', 'postcustodial', and the 'virtual archives' are thrown into the one melting pot. Some archivists may try to reduce the issues to one of custody, but the challenges are much broader than that. I will be building upon themes begun in the pages of this journal by Terry Cook in 'Electronic Records, Paper Minds'. The emphasis, however, is more upon 'electronic minds, any kind of records'. I am following a path suggested by Sue McKemmish in The Records Continuum where she wrote:

The loss of physicality that occurs when records are captured electronically is forcing archivists to reassess basic understandings about the nature of the records of social and organisational activity, and their qualities as evidence. Even when they are captured in medium that can be felt and touched, records as conceptual constructs do not coincide with records as physical objects. Physical ordering and placement of such records captures a view of their contextual and documentary relationships, but cannot present multiple views of what is a complex reality.

If the archival profession is to avoid a fracture along the lines of paper and electronic media, it has to be able to develop ways of expressing its ideas in models of relevance to all ages of recordkeeping, but do so in ways which are contemporaneous with our own society. The terminology and manner of expression within the ages will vary, and will be different for electronic recordkeeping. We need more of the type of construct provided by the Pittsburgh Project's functional requirements for evidence which are 'high modern' but can apply to recordkeeping over time.


Given the breadth of the issues facing us it may seem inappropriate to start this article with considerations of the term postcustodial. The virtual archives, according to Terry Cook, will be a dynamic place with multiple realities. So far the debate in the profession has largely centred on the challenge the concept allegedly makes to one reality - the custodial role of the archives.

One of the basic tenets of a postcustodial stance is that the archival profession can no longer afford to be seen primarily as physical caretakers if we are to exercise an appropriate role in relation to electronic recordkeeping. As Greg O'Shea writes in the introduction to Keeping Electronic Records:

This strategy is a departure from the traditional custodial approach taken by archival institutions but recognizes that in the electronic age physical custody is no longer an essential element of preservation strategy. What is essential is for electronic records to be identified, controlled and accessible for as long as they have value to Government and the Community.

As indicated above, Sue McKemmish and others have broadened this challenge. McKemmish argues that the 'traditional custodial role takes on another dimension when it is accepted that the physical record is only partly manifest in what is in the boxes on the repository shelves'.

There is nothing in Australian Archives' broad philosophy which means they will not have to explore custody as an aspect of identification, control and accessibility. It is more that our thinking has to expand. As a way of opening up consideration of this wider issues it is useful to consider John Francois Lyotard's ruminations upon the three meanings of 'post'. Lyotard is his own form of postmodernism. Never-the-less as one of the chief explainers of the 'postmodern condition', and as a target of criticism, he has been through a useful mill.

'Post' as Simple Succession: Lyotard's first meaning of 'post' as a prefix is that it can be interpreted as simple succession. Within this reductionist view the custodial purposes and structures are at an end. This, for Lyotard, is a form of postmodernism that forgets the past. Lyotard supports initial forgetting, but is a deep respecter of the past as a reservoir of experience.

The postcustodial literature, to date, suggests there will be no permanent forgetting of the past, although many of the physical models for archiving not only will be replaced by new models but already have been in the recordkeeping profession's own past. Indeed many postcustodial authors share a common interest in the history of the profession and of the record. A foundational debate in the area more than ten years ago, the one between the then proto-postcustodialists Hugh Taylor and Terry Cook, reveals a complexity and remembrance of the past which can ground postcustodial approaches within continuing traditions. Hugh Taylor argued for the revival of the tradition of the registrar/ archivist, taking us back to an understanding of the importance of the organisational archive. One of his main themes was that historicism had taken archivists on to a shunt away from the records management main line. Terry Cook's response did not deny the need for Taylor's re-emphasis of the registrar/archivist tradition. Cook argued that the historicity of the archivist could run as a parallel line - that historical archivists were needed to explore 'knowledge, communications and societal dynamics', tasks which the organisational registrar/archivist was not in a position to undertake.

'Post' as a sign of the collapse of confidence in progress: Lyotard's second meaning of the 'post' in postmodern refers to a loss of faith in progress, a form of postmodernity of which Lyotard himself has been a leading contributor. As a marxist and political activist, Lyotard has experienced his own disillusionments. In the passage of text under consideration in this article he cites Adorno's painful one word rebuttal of notions that the story of the modern era has been one of progress - Auschwitz.

Electronic recordkeeping can be seen as a destroyer of faith in linear custodial approaches. Within my own writing and thought, for example, there is an assumption that an archival institution which over- emphasizes physical custody will lack credibility, articulateness, and strategic force. Even before electronic recordkeeping became an issue, Australian Archives was led to consider postcustodial approaches when it became obvious that their existing strategies were ineffective when applied to electronic material such as petroleum data.

Failures in relation to data archiving, or more recent considerations of electronic records, may have been catalysts of the loss of faith in custodial approaches, but they are not the lone cause. In Australia we formulated our national approaches to archives and records management in the 1950's and 1960's. They were built up around the notion of continuous custody and grounded very much in a re-interpretation of Jenkinson's slant on the European registrar/archivist tradition. The assumption, which proved to be correct, was that the archives would build up over time if what has come to be known as a continuum approach was followed. That approach connected logical models for continuity with physical custody. It worked well in the 1970's but we were unprepared for the 1980's. Some of the changes experienced in Australia which reduced faith in notions of physical custody included

on the Registrar/Archivist side:

on the Historical Recordkeeping side:

All these issues are complex, and all raise questions about over-dependence upon the signification of archives as a physical space within which we hold society's most important legal, administrative and historical records. Perhaps the practical physicalities were beginning to supress the maintenance of the healthy logical models we had started out with. In this context any new approaches can be seen as a salvation, not a threat.

The postmodern assault on large metatheories such as those which viewed history as a story of the march of progress extends to a deep doubt that any metatheory could be valid. Perhaps archival theory will go through a similar phase. Postmodernity places an emphasis on respecting difference and diversity within a realization that culture over-rides and dictates reason. All thinking is revisable.

Lyotard notes that a feature of modern society is its capacity for complexification. Custody is an issue going through this process of modern complexification. In societies such as ours is it becoming too complex to be a focus for anything other than explorations of meanings and for working out those meanings in particular contexts? We have to face up to the complexification of ownership, possession, guardianship and control within our legal system. Even possession can be broken down into into physical possession and constructed possession. We also have to face the potential within our technology for ownership, possession, custody or control to be exercised jointly by the archives, the organisation creating the records, and auditing agencies. The complexity requires a new look at our way of allocating authorities and responsibilities.

'Post' as a way of thinking: Lyotard's third meaning of the post in postmodern is that it can mean thinking which is analytical, anamorphic, anamnetic and anagogic. It is a philosophical strut to the sometimes blind processes of re-invention. Lyotard is justifying his own subjective, diverse, experience based and interconnected way of thinking. Lyotard published these comments in a book titled 'Postmodernism explained for children', but it is not the sort of book that will be found in the juvenile reading section of a library. He does not explain his four 'anas', so like a child I had to look most of them up in a dictionary, and reflect upon them in the light of experience.

Lyotard seems to be suggesting that 'post' thinking is an everpresent undercurrent pushing against the trend of settled adult wisdom. Within this meaning those who are labelled 'postcustodial' would be likely to be pushing against the tide in any society at any time. One of the major unsettling techniques of the postmoderns was deconstruction of modern 'dualisms'. This has had its largest impact in english speaking countries where the method of 'rational' thought has been built upon divisions between related concepts. Archivists are only at the beginning of deconstruction. Much of the 'modern' archival thinking in the United States and to a lesser extent in Australia, for example, is based on such 'divisions' as those between archives and records, archives and manuscripts, government and non-government records, information and records, or current records and historical records.

Dualisms are pairs of terms treated as opposites. Most dualisms are amenable to examination as dualities within approaches that look for shared qualities and differences. The custodial debate could furnish us with another dualism: custody and non-custody. From the postcustodial viewpoint, why bother debating the issue? The postcustodial position is not non-custodial, although in particular contexts and in its formative stages it may need to be expressed in such terms. Let's work out suitable regimes for the societies we live in. By all means respect the past, and remember it, but none of us can afford to forget the present and its greater complexity.

In summary, postcustodial approaches do not have to mean a rupture with the past, despite their de-emphasis of physical custody. Postcustodial approaches have grown out of a collapse in confidence in the coping ability of linear regimes of physical custody. They involve a new way of thinking about archives and records management and as such provide a new analysis, fresh shapes for us to consider, different associations of ideas, and a paradigm that is felt intuitively (it is 'ana' based thinking and supports re-invention). And postcustodial approaches involve a working through the meaning of continuous custody, and of the responsibilities and authorities involved. They consist of new orientations to archival issues, different means of applying our skills and knowledge and new professional directions. While they have arisen out of consideration of electronic recordkeeping, this need not result in a split in the profession on the basis of media, unless we allow a new dualism to develop.

Theory for the Continuum

Custody, for the postcustodial archivist, is only one of the many issues that must be faced in the 'multiple realities' of the virtual archives. Do we have to meet those multiple realities with diverse responses, or are there ways which can give us an overview. As indicated above, Lyotard is one of the writers who has done much to undermine the notion that there can be 'grand theory'. Such theory, however, is undergoing a recovery, in academia at least. Anthony Giddens, for example, has done much to put metatheory back into ways of analysing social systems. His 'theory of structuration' provides an overview which explains complexity. That theory is one which a few archivists, including myself, believe has relevance to our many realities. In what follows in the remainder of this article (and all of the next) , I will explore the relevance of Giddens' theory to the structuring of the records continuum.

Structural Principles

Anthony Giddens describes the process of identifying structural principles in social systems as one which involves an "analysis of modes of institutional articulation; and of factors involved in the overall institutional argument of a society or type of society". In the archives and records profession this can mean analysing the way our work is institutionalized and the way the profession itself argues for its existence. Such an analysis can be done at any point of time and in different places, and will produce different results - for the very basic reason that the manner of articulation is developed in time and space contexts.

Structural principles in pre-modern archives - in some times and in some spaces - have been discussed by Luciana Duranti using the terms 'archival bond' and 'archival limit'. In doing so Duranti points to structural principles that seem to have a life beyond the manner of their expression and their time. These terms refer respectively to the relationships established between records during recordkeeping and the role of archival institutions in authenticating records that have been transferred across their boundary.

By the time Jenkinson came to write his manual in the early 1920's these traditional principles, in England, were receiving different expression. He writes of moral defence, physical defence, and the chain of custody. The 'bond' becomes a moral defence issue, the chain of custody is the succession of legitimate 'limits', and preservation is added as a 'physical defence' principle.

A major rewrite of these principles for mid-twentieth century requirements was carried out by Australian Archives and explained to the wider archival community by Ian Maclean. In what has come to be known as the continuum approach Maclean argued that archivists should base their profession upon studies of the characteristics of recorded information, recordkeeping systems, and classification (the way the records were ordered withing recordkeeping systems and the way these were ordered through time). The bonding of records and moral defence became a process of ordering within time and space. The archival limit and the chain of custody disappeared from direct statement within the big picture and became part of the physical model ('classification') for implementation.

The above quick encapsulation is meant to support the contention that structural principles for archival practice are flexible in their manner of expression and implementation. My colleague Sue McKemmish and I have recently tried to re-express Jenkinsonian/Maclean principles for a late modern society. The re-expression took place in a fifteen minute discusssion so we would not claim the points made below are structural principles for the continuum but they provide, for us at least, a starting point:

1. A concept of 'records' which is inclusive of records of continuing value (= archives), which stresses their uses for transactional, evidentiary and memory purposes, and which unifies approaches to archiving/recordkeeping whether records are kept for a split second or a millenium.

In the expression of pre-modern structural principles provided by Luciana Duranti the archival institution's role is to guarantee the continuity, authority and reliability of records after they cease to be used in business. Jenkinson allowed for multiple legitimate successors including the archival institution. Both ways of expressing the principles are disconnected in their manner of expression from 'high modernity' which is a risk based society where the reliability of none of the links in the custodial chain can be counted upon, including the 'integrity' of the archival institution. The concern in the above statement of principle is more with what David Bearman has termed the 'independence of the record'. A significant role for today's archival institution is to help to identify and establish functional requirements for recordkeeping that enable a more systematic approach to authentication than that provided by physical custody. It is the sort of task for which they themselves can be held accountable, unlike the physical holding task where they can always plead shortage of resources and hide behind financial accountability. The emphasis is on continuity (the continuum) and the role of records as trace, evidence and memory. The physical models, however, are not ignored, in so far as physical models have to be established to implement the concepts. The physical models are not, however, determined in advance of considerations of what the logical models suggest should be done.

2. A focus on records as logical rather than physical entities, regardless of whether they are in paper or electronic form,

This clarifies the first principle. The strongest critiques of physical approaches in the archival profession have been presented by David Bearman. His writings show how, too often, the way we think about our tasks has been determined by the physicality of the records and of our actions in relation to them. Notions such as the 'archival bond' and 'moral defence' are in fact logical models (concepts), and are capable of being given different verbal and physical expression in different contexts.

3. Institutionalization of the recordkeeping profession's role requires]a particular emphasis on the need to integrate recordkeeping into business and societal processes and purposes.

This should be a fairly self-evident structural principle within archival theory. In an electronic work environment it means, in part, that the objectivity, understandability, availability, and usability of records need to be inherent in the way that the record is captured. In turn the documents need to be captured in the context of the actions of which they are part, and are recursively involved. Archival documents are firstly documents embedded in action, and then are records disembedded from that action.

The above principles require further exploration. The manner of expression is abstract, the principles need to be more fully reviewed against the past, and they also need to be continuously reviewed against a future in which electronic records will be separated from the externalities of place. Accordingly, Sue and I added in a fourth principle, which is that archival science is the foundation for organising knowledge about recordkeeping. We added the qualification that this should be combined with an acceptance that we need to continue to identify the knowledge and skills in other disciplines of relevance to our endeavours. We could have added further qualifications. This is not an archival science of 'universal laws'. That form of science has disappeared from science itself. It is still a science of knowledge. Such knowledge is revisable but can be structured and can be explored in terms of the operation of principles for action in the past, the present and the future.

Figure 1 - Records Continuum Diagram

Figure 1 - Records Continuum Diagram

The Structural Properties of the Records Continuum

Archival Science, of course, is concerned with properties as much as principles. On the page facing this one you should find a presentation of a model for the records continuum indicating its properties. It is built around four axes: identity, evidentiality, transactionality and recordkeeping entity. The axes encapsulate major themes in archival science, and each axis presents four co-ordinates which can be linked dimensionally.

A records continuum is continuous and is a time/space construct not a life model. [If one wishes or needs to do so, one can still talk of the lifespan of records within the model]. No separate parts of a continuum are readily discernible, and its elements pass into each other. In establishing a depiction of the continuum I have attempted, as far as possible, to label the features independently of any time period or professional loading. For example, a word like 'function' can be used in ways which are dimensionally different from its positioning in the model which follows. The model, however, is self- referencing. Within it, 'function' has its location and that is where it is. Within any implementation environment other than a teaching one, the terms will be given specific interpretations and meanings and the way the elements will join with each other will vary. In what follows I will make brief comments on the axes, co-ordinates and dimensions of the model, using it in teaching mode.

The recordkeeping entities axis

The recordkeeping axis: this deals with the vehicles for the storage of recorded information about human activities. Its co-ordinates are those of the document, the record, the archive and the archives.

The document within the model is based in an act and is a pseudo representation of that act. In this form it has content, structure and a context of creation. The document has not yet been communicated, or if it has been then there is no way of establishing the context of that communication, other than by the content, context and structure of the document itself. The record is a memorialized [disembedded] form of the document usually linked with other documents. It should have additional layers of context to those present in a document, and may be a surrogate record of that document. It is this additional information about context which is the key to 'disembedding' the document from its narrower contexts of creation and carrying it through time and space as a record.

The archive is the aggregated record viewed as all the archival documents of an organisation, which broadly corresponds with the first rule in Muller, Feith and Fruin's manual written 100 years ago. The archives is the archive in plural form. It contains the records of a number of organisations, either because of spatial spread or temporal transmission to another organisation or institution.

The evidence axis: This axis consists of the trace of actions, the evidence which records can provide, and their role in corporate and collective memory. Within the Australian continuum approach, as developed in the 1950's, for example, there was an assumption that a well constructed recordkeeping system controlled documents as a trace of an action, aggregated them into an evidential record, and gave organisations a corporate memory. Peter Scott added the final 'building block' to the Australian continuum approach when he developed the 'series system' as an archival information system which could be placed over recordkeeping systems in their current operational environments in ways which helped order subsequent disposition activities.

The transactional axis: this axis presents the act, activities, functions and purposes as co-ordinates. In the model the terms are simply labels. This axis reflects an emphasis upon records as records of activities undertaken in the conduct of affairs, and upon the way these activities create links between documents. It reflects, in a basic manner, functions of organisations, and the way these are broken down according to subcategories of activity, or built up from the acts themselves. Purpose is the function viewed from a broader societal perspective.

The identity Axis: This axis represents the actor, the work unit with which the actor is associated (which may be the actor alone), the organisation with which the unit is associated (which may also be the actor or the unit) and the manner in which the identity of these elements are institutionalized by broader social recognition.This is the axis of structural provenance, and of the authorities and responsibilities within which the archives are made and used. It accounts for two main themes in archival science, the notion that an archive should be linked to a records creator, and that the records reflect the authorities and responsibilities that support an act.


A dimensional analysis can be constructed from the model and explained in a number of ways including a recordkeeping system reading. When the co-ordinates of the continuum model are connected, the different dimensions of a recordkeeping system are revealed. The dimensions are not boundaries, the co-ordinates are not invariably present, and things may happen simultaneously across dimensions, but no matter how a recordkeeping system is set up it can be analysed in terms such as:

As an example, if one analyses Australian archival developments in the 1950's and 1960's in these terms something like the following brief summary is possible:

  • even greater attention was paid to how registration systems should operate. The regime of control for material not included in the registry system - which in the average agency was 90% of its records - was left for organisations and their units to consider on the grounds that this less complex material was amenable to being held in 'sets' in accordance with the way the work units assessed their own needs (second dimension);
  • the memory of the agency was controlled by survey and disposal techniques which assumed there would be a main source of memory in the registry, and that the necessary information about additional records in the separate sets of records throughout the agency could be consolidated within survey and disposal documentation (third dimension);
  • the archival system built upon the other dimensions, particularly once the series system was developed and could be applied to records not in custody. That system created a way of treating the elements encountered on the vertical axes of the model separately from those on the horizontal axis. The two separate strands of documentation were interconnected through inbuilt cross referencing techniques (fourth dimension).

The logical model in this approach was that of a continuum in which the various elements passed into each other. At times practice co-incided with the logic of the model itself. In the high modern recordkeeping environment of the 1990's a continuum has to take into account a different array of recordkeeping tools. These tools, plucking a few out at random but ordering the list dimensionally, include: document management software, Australian records system software, the intranet and the internet. The conceptual model of a continuum will still be relevant to this new array of recordkeeping, but the physical models for implementation will be different from any form of recordkeeping in our past.

In terms of a records continuum which supports an evidence based recordkeeping approach, the second dimension is crucial. This is where the document is disembedded from the immediate contexts of the first dimension. It is this disembedding process that gives the record its value as a 'symbolic token'. A document is embedded in an act, but the document as a record needs to be validatable using external reference points. These points include the operation of the recordkeeping system into which it was received, and information pertaining to the technical, social (including business) and communication processes of which the document was part.

The model presented above has already proved itself of value as a teaching tool and as a way of conceptualizing the nature of recordkeeping. It is the first checklist of continuum elements in the archives and records management literature, and is suggestive about how these elements merge into one another. In one sense it does not need revision. The continuum operates in particular ways within particular operational environments. It will, for example, have its own manifestation in relation to electronic recordkeeping, and the base model can be of use in these further explorations.


In rejecting the symbolism of external physical space there can be dangers of the type pointed to by Luciana Duranti in her critiques of what she chooses to call 'postcustodial' approaches. Physical space is a strong symbol in anyone's life. For its power one only has to look at how conventional office space is surviving despite the emphasis on virtual space in so much popular culture and technological hype. In tossing out the physicalities of existing models one can, of course, throw out the conceptual base of those models.

Janus, however, can be represented by more than a temple building. He can also be represented on a coin in daily usage, which is just as well for archivists because in late modern societies the externalities of place are becoming less significant day by day. In the virtual archives the location of the resources and services will be of no concern to those using them, and records authentication processes will have to implemented in accordance with new strategies which take into account new realities. Postcustodial approaches to archives and records cannot be understood if they are treated as a dualism. They are not the opposite of custody. They are a response to opportunities for asserting the role of an archives - and not just its authentication role - in many re-invigorating ways, a theme which I will explore further in the next edition of Archives and Manuscripts.

© 1996-1998 Frank Upward. All Rights Reserved. Licence: Limited to on-line viewing and the making of one (1) printout for off-line reading purposes only.

Back to top