Which of these is not the purpose of the iso 15489 standard for a records management system?

From arrangement to accountability

If information architecture represents the Buffonian end of the ancient opposition between context and categorisation, records managementrepresents the Linnaean end, and just as the information architect will regard you as an interfering conservative, so the records manager will look on you as a dangerous radical.

Records management is important because it is a body of management practices that, in theory at least, preserves and maintains access to a reliable memory of an organisation’s activities. This is important for a whole host of reasons, among which are:

enabling us to see the historical contexts of decisions;

referring us to plans, designs and specifications that have material or management value for very long periods of time;

finding out what we actually did in contrast to what we think we did;

demonstrating to everybody’s satisfaction what contracts and agreements we have entered into and what accountabilities we face;

proving to the satisfaction of a court of law what we did or did not know and do at any particular point in time, and so on.

Records management is about accountability in the old fashioned sense, i.e. giving us the ability to render an account (together with satisfactory evidence) of our actions that is reliable and trustworthy. To do all that we need to process the special class of documents that are records in such a way that we can always retrieve them, so that we know that they are the ‘original’ unaltered records of the activities and decisions in question, and we need to be able to pull together the records related to any significant action or decision in such a way that we can give an account of the context surrounding that action or decision. We need a taxonomy (among other things), but a taxonomy that is focused on the special features of records and what ‘reliable’ memory means.

The definition of ‘records’ is disturbingly vague. The international standard for records management defines it as: ‘information created, received, and maintained as evidence and information by an organisation or person, in pursuance of legal obligations or in the transaction of business’ (AS ISO 15489.1-2002). It’s hard to take that definition and immediately know which documents or datasets need to be managed as records.

Records management specialist Marita Keenan reports that a commonly accepted understanding in the profession is that ‘in a business context a record is a by-product of the business transaction.’ However, she points out that records may not simply comprise just the direct documentation of the transaction, but also associated documents that provide context to the primary records of that transaction. ‘Staff need to think holistically and legalistically in deciding whether the business transaction has any risks connected with it, and hence requiring associated records to be captured into a recordkeeping system’ (Keenan, 2006). For example, with the passage of time a given policy document may be difficult to understand without reconstructing the reasoning process involved in formulating that policy, accessed through drafts and minutes of discussions.

Another accessible rule of thumb might be that a record is any document or dataset that allows you to reliably reconstruct any significant action or decision you made in the past. This can be for internal management use or for external purposes, whether in dealing with partners or suppliers or demonstrating that you have met your legal responsibilities, and it encompasses the risk management aspects suggested by Marita Keenan. In some organisations you will also be working in areas that have historical significance, and so your records will at some point be destined for archival status.

This means, from your point of view as an enterprise taxonomist, that a significant part of the content covered by your taxonomy will likely be records. Whether you (or they) like it or not, sooner or later your taxonomy will bump up against the very special requirements of the records management function. Let’s look at what that means.

I have used the word ‘reliable’ several times. Actually in records management, there is a cluster of four very precisely defined concepts that assure the trustworthiness of records over time, each of which has implications for specific records management practices and processes. Again in the slightly stilted language of the records management standard (AS ISO 15489.1-2002):

Reliability. A reliable record is one whose contents can be trusted as a full and accurate representation of the transactions, activities or facts to which they attest and can be depended upon in the course of subsequent transactions or activities. Records should be created at the time of the transaction or incident to which they relate, or soon afterwards, by individuals who have direct knowledge of the facts or by instruments routinely used within the business to conduct the transaction.

Integrity. The integrity of a record refers to its being complete and unaltered. It is necessary that a record be protected against unauthorised alteration. Records management policies and procedures should specify what additions or annotations may be made to a record after it is created, under what circumstances additions or annotations may be authorised, and who is authorised to make them. Any authorised annotation, addition or deletion to a record should be explicitly indicated and traceable.

Authenticity. An authentic record is one that can be proven (a) to be what it purports to be, (b) to have been created or sent by the person purported to have created or sent it, and (c) to have been created or sent at the time purported. To ensure the authenticity of records, organisations should implement and document policies and procedures which control the creation, receipt, transmission, maintenance and disposition of records to ensure that records creators are authorised and identified and that records are protected against unauthorised addition, deletion, alteration, use and concealment.

Usability. A usable record is one that can be located, retrieved, presented and interpreted. It should be capable of subsequent presentation as directly connected to the business activity or transaction that produced it. The contextual linkages of records should carry the information needed for an understanding of the transactions that created and used them. It should be possible to identify a record within the context of broader business activities and functions. The links between records that document a sequence of activities should be maintained.

Records management, even more than librarianship, has a reputation for being a somewhat solipsistic profession, not entirely adjusted to the ordinary needs of the here and now. This is surely partly conditioned by its unique requirement to be accountable to the future and not just the here and now. But when combined with a technical set of processes and rules there has evolved an ethos within the profession that says the ‘ordinary user’ cannot be entrusted with the sacred tasks associated with preserving the status of records. Records management processes should be undertaken by records management staff.

As the discipline has developed it has focused very much on providing the overriding ability to produce a record when required, especially when required by law. In the paper-based documentary environment this meant ensuring that records were captured quickly into the appropriate files, that files were organised in a structured way so that records staff could locate and retrieve them easily, and that the file movement was carefully controlled.

Hence, while we would expect records managers to be highly attuned to taxonomy work (to organise their records for effective retrieval), in practice what has happened is the opposite. In any given organisation, the number of people actively using the file classification structure as a taxonomy is very low (for most people the file reference number is just a number and has no navigational value). Any record-keeping taxonomy that is not exposed to the working needs of a large audience is in great danger of reflecting the very narrow needs of records administrators.

And this is what seems to have happened. Stuart Orr, when studying the use of functions-based classification schemes among organisations in the UK, Canada and Australia, found varying and inconsistent understandings of classification, classification principles and the construction and use of thesauri. While records management theory had relatively sophisticated guidance on the principles according to which records should be classified, in practice records offices diverged considerably and were often hostage to historically accreted schemes.

Despite the rising influence of functional or business activity based classification approaches through the 1990s, interpretations of how this should be done varied considerably, and other classification principles such as organisation structure, record type or record topic were just as common. Orr concluded: ‘The understanding of classification in records management has not been as well developed as in library science’ (Orr, 2005: 110).

And while ‘usability’ is one of the four fundamental attributes of a trustworthy record, this does not mean ‘usability’ in the everyday sense. Records offices have routinely sacrificed ease of use now for assuring proper control of the records by properly authorised staff. This worked with varying effectiveness in a paper-based environment, but as a profession, records management has made a poor transition to the high-speed digital environment of multiple electronic copies of every document, decision turnaround times that have accelerated from weeks to minutes, and to transactions by e-mail and instant message that are never captured within records systems. There are notable exceptions to this statement and there are progressive records managers and organisations, but the profession as a whole has done a poor job of adjusting consistently.

Records offices that have existed for years as file processing and file movement logistics departments are poorly equipped to deal with their underlying role – assuring reliable organisational memory of key actions, transactions and decisions, whether expressed on paper or otherwise (it’s now mostly otherwise). The speed and volume of information transactions are now simply too much to manage if the records management staff are the sole records processors, so increasingly, reluctantly, but painfully slowly, record-keeping functions are being devolved to the producers and consumers of records – who, unfortunately, frequently have more of a commitment to the now than to the future (McMullin, 2004).

Specifically, in relation to taxonomy work, the profession’s limited experience with taxonomy work for large and diverse audiences has resulted in two major (from this book’s perspective) mistakes. First, although faceted taxonomy design and the use of metadata are occasionally given polite nods in the literature, it does not seem to be generally understood in practice. Experienced records managers still say, in the words of one recent contributor to a forum, ‘in a business classification scheme there should be only one place for a record to be located.’ A business classification scheme (BCS) is a common term for the master taxonomy that records managers use to generate their file classification structure – called variously a file plan, a file schema, a records classification plan, and so on.

This commitment to hierarchy has resulted in tree structure record-keeping taxonomies that require extraordinary investments in familiarity – rote learning – to navigate. Orr, for example, found in his study that 12 out of 15 of the practitioners in his sample had more than ten categories at the top level of their taxonomy. One respondent had over two hundred. Eight out of 20 respondents had five or more levels in their taxonomy. Two had nine or more levels which by my calculation would give them a capacity for 38 billion records (Orr, 2005: 99–100).

The second major mistake is the frequent adoption of a ‘top-down’ approach to taxonomy development. Where record-keeping taxonomies have been developed based on business functions and activities, most records management experts told Orr that a top-down approach working from the mission and objectives of the organisation downwards was the most effective way to build the scheme. Some said that it should be supplemented by checking with users, and there are documented cases of systems and process mapping approaches such as we have described in this book. But over 80 per cent of Orr’s expert panel agreed that the top-down analytical approach was the most effective (Orr, 2005: 78–80). In fact, two widely influential approaches to record-keeping classification scheme design, the Australian DIRKs methodology and the British BCS methodology, embed top-down analytical approaches as an explicit part of their guidance (NAA, 2003; National Archives, 2003).

This approach directly contradicts the central thesis of this book, which argues that any taxonomy with a hope of being effective must be grounded in – and built from – the working language of its user population. The issue for us is that the records management approach seems to be still organisation-and-record-centric first and user-centric second.

Small wonder then that the functional classification schemes built on this principle throughout the 1990s seem to have met with considerable user resistance (Orr, 2005: 12; Robertson, 2004). Tina Calabria documented an electronic records management project in a city council in Australia and found that one in two users routinely failed to identify the correct top-level category for their document and found great difficulty in distinguishing between functions. The search terms created by the user population were altogether different from the taxonomy terms, and – perhaps most embarrassing of all – the records managers performed little better than ‘normal’ users (Calabria, 2004). An emerging trend in records management systems is to provide for two separate metadata fields, one for record-keeping purposes and one for user retrieval purposes – as if the needs of the two parties are permanently irreconcilable (Keenan, 2006).

The problem with you as a taxonomist is that, to the extent that your taxonomy includes items of content that are records, then to that extent you will sooner or later come up against the accountability issues served by records management. This will have implications for your metadata framework, but also directly for your taxonomy, because you may be compelled to use alien categories associated with the organisation’s records classification scheme – to which access permissions and retention and disposal rules are all attached. The records classification scheme performs a record control function that cannot be ignored or jettisoned.

Your risk is compounded by the fact that in many organisations the records management function is not highly visible or proactive. The realisation that there are factors you should have considered may well ambush you after your implementation has started. In some cases, records management processes may be lax until your organisation encounters business or regulatory risks associated with its record-keeping (e.g. contract disputes, litigation, official investigations and inquiries, enforcement of regulations such as Sarbanes-Oxley, accreditation in standards such as ISO 9001). The risk of a late ambush is high.

You are likely to face several educational and change management challenges in dealing with these risks because, as we have seen, the records management function, despite good intentions, is on the whole not oriented towards user-centred taxonomy development. The records manager is going to worry much more about whether or not any given record can be produced in a (legally) trustworthy fashion together with all its associated documentation, consistently and reliably every time than about whether Jack down the hall can find readily to hand the piece of information he needs right now to pass on to a customer.

And unlike most of the managers and employees you will deal with, the records manager thinks he knows what a taxonomy is. It’s called a file plan or a records classification scheme, and the fact that it may follow no consistent taxonomic principles, is not intuitive and almost impossible to use unless you learn it through constant use is not relevant to the purpose, which is to provide a unique pigeonhole with a file reference number for every record that is created. Period.

This is not to say that your goals are at odds with theirs. There is in principle no reason why a faceted taxonomy, especially if it contains a facet that catalogues all the key business activities, should not serve a valid records management function as well. But this facet needs to be integrated (via metadata) with your records control functions that govern retention, disposal and access. This is why in the planning phase we suggested that the records management function be included on your governance committee.

The real challenge is in bridging the psychological distance that many records management practitioners have from the concepts and ideas expressed by the versatile and adaptive information architect. Your role, as the taxonomist, lies between these two poles, servicing arrangements for ready use and simultaneously assuring reliable access over time together with preservation of the key characteristics of a record.

If this presentation of the challenges posed by records management issues seems somewhat jaundiced, it is only because such sharply drawn lines do exist in many organisations, and because taking a stark view will bring the important issues sharply into the foreground. There are also organisations where these issues are recognised and approached in a manner which combines present usability with future accountability, where records managers work with information architects and where taxonomy work supports both. It would be prudent to belong to that class.

It should by now be obvious to you that the taxonomist’s task is largely a thankless one from the perspective of the general population. Our role in an organisation is deep in the information and knowledge infrastructure where, if it works, we will not be seen and other actors closer to the outcomes of a good taxonomy (such as information architects and records managers) will take the credit that is due.

Especially when serving the needs of structure and organise, building common ground and boundary spanning, the more visible we are, the less we are really doing our job. As taxonomy processes become more important than taxonomy products, from sense-making to discovery, we emerge with tools and activities that help people do what they need to do. But we are still enablers rather than primary actors. It’s wise to remember this, lest our taxonomic pirouettes distract our clients from what it was they set out to do in the first place – to achieve their objectives, meet their goals, avoid business risk, find new opportunities and improve their effectiveness.