back to conference papers page

Conference Papers Online

Session 5: The Archival Control of Systems

Recordkeeping systems: wanted dead or alive


David Roberts

Manager, Government Recordkeeping
Archives Authority of New South Wales

I would like to start by expressing my thanks to Gavan McCarthy and the other members of the Australian Science Archives Project for giving me the opportunity to participate in this excellent program. 

The title of this paper is whimsical, but it hints at two of the themes that I wish to pursue: 

  • that understanding recordkeeping systems is crucial to understanding the records generated through them: as archivists, we want information about those recordkeeping systems;
  • that, particularly in the electronic environment, archivists need to be at least as concerned with live systems - documenting, and even designing, them - as with dead systems.

Recordkeeping Systems in Conventional Archival Description

Describing the recordkeeping system as part of the context of records has long been an accepted component archival practice. Conventionally, however, the concept of a recordkeeping system that we have used for this purpose has been limited to a group of related series used together in their original context. The classic example of this concept of a recordkeeping system is a correspondence file series, with a register and one or more indexes. 

Certainly the recordkeeping system, thus understood, has an important place in archival descriptive practices. It is mentioned frequently in the writings of Peter Scott on the Australian series system. It provides an important basis for identifying what is an agency and what is not an agency under the series system, as seen in the former Australian Archives definition: Agency - an administrative unit that is a recognisable entity, generates records and has its own general recordkeeping system. 

(Recently the Australian Archives, now the National Archives of Australia, have adopted a definition placing less emphasis on the recordkeeping system, but the concept is still there.) 

This concept is also used as a primary basis for describing the relationship between record series under the series system. Thus, for example, the series description procedures in the CRS Manual mentions the recordkeeping systems 24 times. (I have used Australian/National Archives examples here, not to single them out, but because their practices are widely known and followed by other Australian archivists and they have been willing to make their procedures and related material accessible on the World Wide Web.) 

Despite the importance of the recordkeeping system in conventional descriptive practice, it has not been adopted as a structural element in archival control and description systems, that is, as an entity. Thus, the CRS Manual describes the following entities: 

It is a similar story with my own Tabularium system and with ASAP's onQ ADS that has been described elsewhere in this conference. 

In his exposition of the series system (reference), Chris Hurley has suggested that there are three options for building recordkeeping systems into descriptive models: 

  • Option 1
  • Option 2
  • Option 3

In the conventional practice that I have described, we have followed option 1. Hurley notes interest in option 2 when dealing with electronic records. 

I know of no archives institution that has incorporated the recordkeeping systems as an entity in its control and description system, that is, following option 3. 

The Recordkeeping System in Modern Recordkeeping Theory and Practice

We see little mention of recordkeeping systems in traditional records management texts, nor in information management texts that purport to cover records management. Yet, at the same time as archivists have been reviewing the role of the recordkeeping system in descriptive practice, recordkeeping systems have earned a prominent place in modern recordkeeping theory and practice.

Why has this occurred? One reason has been a greater need to make recordkeeping more systematic, especially in devolved or decentralised organisations and in the electronic environment. Systematic recordkeeping is widely recognised as a key support for accountability and, especially in the public sector, a lack of systematic recordkeeping is a frequent subject of criticism by royal commissions, Auditors-General and other accountability watchdogs. As the Standard on Full and Accurate Records, developed by the Archives Authority of New South Wales, notes, "Records that will meet business needs, accountability requirements and other organisational needs cannot be made, maintained or managed in the absence of system. Laissez faire recordkeeping is inevitably poor recordkeeping" (Principle 3). In practice, systematic recordkeeping depends on the effective operation of recordkeeping systems. 

Modern recordkeeping theory and practice have also taken a more comprehensive view of what a recordkeeping system is than in the past. As David Bearman has reminded us, recordkeeping systems are information systems "organised to accomplish the specific functions of creating, storing and accessing records for evidential purposes." (David Bearman, "Recordkeeping Systems", Archivaria 36, 1993, p. 17) This understanding is reflected in the definition of a recordkeeping system in the Australian Standard on records management: Recordkeeping Systems - information systems which capture, maintain and provide access to records over time. (AS 43901996: Records Management, Part 1: General, Clause 4.20) 

Importantly this concept is not confined to systems that are designed primarily for keeping records. Information and communication systems that support business processes are also recordkeeping systems, if they capture, maintain and provide access to records as part of their functionality. This is the case, for example, with transaction processing systems and with database systems with audit trails built in. As recordkeeping systems, these kinds of systems are of as vital interest to archivists and records managers as systems designed specifically to keep records. 

Importantly also, the focus in this concept is on the processes of recordkeeping, not the products. This is reflected in the definition of a recordkeeping system in the standard Australian text on the practice of archives management, Keeping Archives: Recordkeeping System - the principles, methods and processes devised for capturing, arranging and maintaining the records of an agency or person.' (Judith Ellis, ed., Keeping Archives, 2nd edn, Thorpe in association with the Australian Society of Archivists, 1993, p. 477) 

The notion of a recordkeeping system as merely a related group of record series is clearly inadequate here. Computer science has reminded us that a system has many elements: 'the interrelation of personnel, procedure, hardware and software, which combine to accomplish a set of e year. 

The following graphic summarises the major stages in the methodology: 

In stage (a) we identify and document the organisation's role, purpose, organisational structure and legal, regulatory, business and political environment. We also seek to identify and document critical factors affecting recordkeeping. 

In stage (b) we identify and document each business function, activity and transaction, producing:

  • a hierarchical business classification scheme; and
  • an analysis of the flow of business processes and the transactions comprising those processes.

In stage (c) we identify and document the requirements for evidence associated with each business function, activity and transaction and how each requirement may be satisfied through recordkeeping. These requirements are called 'recordkeeping requirements' and their satisfaction is the raison d'etre for recordkeeping activities and for the existence of recordkeeping systems. 

In stage (d) we identify and analyse existing recordkeeping and other information systems used in or in conjunction with business processes, measuring their performance against the recordkeeping requirements identified in stage (c). To what extent do existing systems satisfy those recordkeeping requirements? If there is a significant gap, we proceed to subsequent stages. 

In stage (e), we identify and document possible strategies for recordkeeping that might be built into a new recordkeeping system. These strategies include: 

  • using policies, procedures and practices to affect human behaviour;
  • designing new systems that can help or automate recordkeeping processes;
  • implementing systems in ways which satisfy a requirement; and
  • adopting technological standards for record formats that promote the longevity of technology dependent records.

Important factors in the choice of a mix of strategies are: 

  • the degree of risk involved in failure to satisfy a recordkeeping requirement within the relevant business function;
  • the existing systems environment; and
  • the corporate culture in which the strategy should succeed (especially relevant when trying to influence behaviour).

In stage (f), we bring the chosen strategies together into the design of the recordkeeping system in the broad sense described previously and we document that design. We ensure that the design will support business processes, rather than hinder them. If necessary, we redesign existing business processes and systems, if these are a significant part of the problem. 

In some cases, we incorporate documentary products and tools developed from work in the earlier stages into the design of the system. An example is the records disposal schedule: we use those recordkeeping requirements that relate to the retention of records to develop a disposal schedule specifying how long the records that document particular business functions and activities, or types of recurring transactions, need to be kept for evidence of the activity: 

In stage (g), we identify, document and use a suitable mix of implementation strategies to implement the recordkeeping system and integrate its operation with business processes and systems. 

In stage (h), we collect information and assess the performance of the recordkeeping system against the recordkeeping requirements identified in stage (c), documenting the results. We initiate and document any corrective action and set up an ongoing monitoring regime. 

A key feature of each stage is to document it. What do we end up with at the end of this process? 

  • a range of environmental information from preliminary investigationp;
  • a business classification scheme;
  • a business process analysis;
  • a comprehensive set of recordkeeping requirements;
  • a gap analysis of existing systems against those requirements;
  • a documented selection of recordkeeping strategies;
  • a risk assessment of failure to satisfy recordkeeping requirements;
  • a range of forms of documentation of design processes;
  • implementation plans and reports;
  • assessments of performance; and
  • decisions and reports about corrective action.

In short, we end up with a great steaming pile of metadata - documentation that tells us what the recordkeeping system was meant to do, how it was to do it and how well it did it. We end up with a rich source for the archival description of the recordkeeping system. 

Now all we need is standards for using this metadata in archival description . . .

 Copyright David Roberts 1998


back to conference papers page

Published by: Australian Science Archives Project on ASAPWeb
Comments or questions to: ASAPWeb (
Prepared by: Helen Morgan
Graphics by Lisa Cianci
Date modified: 7 October 1999