Where Have All the Files Gone? Lost in Action Points Every One?

At the Hutton Inquiry into the death of Dr David Kelly, the weapons expert, James Dingemans QC, the senior counsel, quizzed Jonathan Powell, Tony Blair’s chief of staff, about the apparent lack of minute taking in the Cabinet Office:

A. No minutes were kept of this meeting or subsequent meetings we are probably going to discuss, no.

Q Is that the normal course, that they are just discussions and no-one is bothering to write them down because they are free flowing, as it were?

A. Yes, I thought I might be asked that question because it may seem odd to people from outside, so I looked through the diary for the two weeks of the period we are talking about and the usual pattern is about three written records for 17 meetings a day is sort of the average you get to because there is no purpose served by minutes unless they are either recording people visiting from outside, the president of Nigeria, or something like that, or if they are action points that need to be taken forward, something on school funding for example.

Q. The only documents that Lord Hutton effectively is going to look at, and I think you have very kindly supplied some of these over the weekend, are emails that people sent to each other after meetings or before meetings following up on discussions. ¹

The Review of Intelligence on Weapons of Mass Destruction, chaired by Robin Butler, a former Cabinet secretary, criticized ‘the informality and circumscribed character of the Government’s procedures’ in the events leading up to the war on Iraq that was reflected in this cross-examination. ² Talking on BBC’s Newsnight programme on 18 October 2010 about his newly published memoirs, ³ Powell was unrepentant, even when challenged in the interview by Butler, who stated that in his investigation of the evidence for the existence of weapons of mass destruction there were some excellent papers, prepared by civil servants, which were never tabled at meetings of the Cabinet that was only treated to PowerPoint presentations. This exchange was as much about a cultural shift in the way of conducting business in an environment permeated with technology, as it was about the character of cabinet government with its principle of collective responsibility. The report of the Conservative Democracy Task Force – End to Sofa Government – Better Working of Prime Minister and Cabinet, ⁴ published in 2007, was careful to eschew nostalgia, but inevitably looked back to practices and procedures that were still more or less in place when the Conservative party left office in 1997. It laid much of the blame, although not exclusively, at the door of the Labour administration, with only one reference to technology and then in relation to managing ministerial diaries. ⁵ Implicit in the report’s recommendations were improved flows of information to support decision making, but with no explanation as to how this was to be achieved in an electronic environment. Attention was drawn to the ‘General Rules of Business for the Federal Government’ of Germany that stipulates that the Cabinet cannot make decisions in the absence of the necessary paperwork. ⁶

Butler was a member of the task force and gave advice on a non-party basis, as did the economist Sir Christopher Foster, whose political sympathy lies with the Labour party. In looking back on his early days in the civil service in his book British Government in Crisis, Foster praised the record-keeping practices of the civil servants he encountered:

Underpinning their knowledge of their jobs was an exhaustive, but marvellous, filing system. Everything was written down. Almost every meeting discussed a paper, which was revised as it rose through the hierarchy towards the minister or was sent down again by her. (There was no problem identifying who suggested, and who authorised, every change to a paper, as there was to the 2002 Iraq dossier.) Every meeting, every lunch, was minuted. ⁷

He contrasted the character of the evidence presented to the Hutton Inquiry with that presented to the BSE Inquiry at the beginning of the decade, which found the record keeping in the Ministry of Agriculture was ‘as meticulous as ever’. ⁸ Foster’s solution was the creation of a minister of information within the Cabinet Office, which would be distanced from the Prime Minister’s office; but he did not address the cultural shift brought about by the interaction of the widespread introduction of new technology with the changing composition and behaviour of the workforce. Sir Rodric Braithwaite, the British ambassador in Moscow from 1988–92, has no doubt that subtle changes were already taking place at the very time Foster described:

When I returned to the [Foreign] office from abroad in the 1960s, things had already begun to change. Registry staff were no longer content to remain stuck in one job, the supply of highly intelligent women without university educations and willing to work at the bottom of the pile was drying up, and the personal continuity was therefore fraying fast. The arrival of photocopiers meant that documents could be produced in unlimited numbers, and more and more people could be consulted, and insisted on being consulted, about the most trivial matters. Needles to say, that neither speeded up the decisions, nor improved their quality. It also meant that the decisions about where to file things began to become more arbitrary. But one couldn't say the system was beginning to break down. ⁹

Although Andrew (now Lord) Turnbull, who succeeded Robin Butler as Cabinet Secretary, in his evidence to the 30 year rule review in 2008, referred to the ‘casualisation’ of government under Tony Blair, ¹⁰ he touched, albeit in passing, on the consequences of this cultural shift 30 years later when discussing email – ‘A particular problem was with e-mail. With a flurry of e-mails going backwards and forwards it was difficult to authenticate an audit’. ¹¹ The committee was too pre-occupied with discussing special advisers to pursue this critical observation. Lord Hunt of Wirrall, who had been a cabinet minister under Margaret Thatcher, in giving evidence believed: ‘A return to a “proper process” with a paper trail would result in a sudden improvement in behaviour’, but failed to understand just how difficult that is in an electronic environment. ¹² Only the archivists who gave evidence addressed the ‘technical’ challenges the electronic environment posed, but even they did not appear to appreciate the nature of the cultural shift, of which the greater use of special advisers is just one manifestation. None of the evidence referred to the abortive attempts to introduce document management systems that themselves effected record-keeping, tellingly described by Rodric Braithwaite:

In the 1970s and 1980s mechanisation - secure fax, for example - continued to spread, and the FCO made its first tentative experiments with the computerisation of records. I think the first experiment was in the 1970s. We and the registries were all instructed to give documents a very constricting range of file names that the computers could recognise. This got nowhere, because the computer systems didn't work. There was a similar fiasco in the middle to late 1980s, when a great deal of money was spent on a system of electronic document handling and messaging which in the end was not introduced: it was undercapitalsed, and the senior management never gave it the attention or understanding it needed during the development phase. There were endless problems about sharing information across Whitehall, because departments had incompatible systems, and for often grossly exaggerated reasons of security (and departmental jealousy) were unwilling to share information anyway. ¹³

Such initiatives, which disrupt rather than improve work-flow, only serve to undermine effective record-keeping and support the arguments of those, such as Powell, who question the need for it.

In November 2008, Gordon Brown, as a new Prime Minister, endorsed Information Matters, building government’s capability in managing knowledge and information that was published by the Knowledge Council, ‘a strategic body established to lead government in the better use and management of its knowledge and information’ ¹⁴ with a foreword by the then Secretary of the Cabinet, Sir Gus (now Lord) O’Donnell. ¹⁵ This document sets out a proactive strategy with little analysis of what went before or any reference to the trenchant criticisms of either Butler or Hutton, despite the fact that it bears the imprint of The National Archives (TNA). It is couched in terms of service delivery that relates largely to information about individuals or specific projects, known in the Civil Service as ‘particular instance papers’ (PIPs), which is not of concern here. Where files are discussed they are dismissed as a legacy of another age:

Information management, in a paper age, relied upon clerks and the organised storage of paper files. In an electronic age, it relies upon sophisticated technology and strategic leadership. Technology is evolving at a rapid pace, and our approach to strategic leadership in this area must evolve with it. ¹⁶

This is questionable. As long ago as 1977 in her novel A Quartet in Autumn Barbara Pym poignantly illustrated the lack of agreement of what managing information is about when she described the retirement of two of her central characters, Miss Crowe and Miss Ivory, both registry clerks of the kind Rodric Braithwaite encountered early in his career:

The activities of their department seemed to be shrouded in mystery – something to do with records or filing, it was thought, nobody knew for certain, but it was evidently ‘women’s work’ the kind of thing that could easily be replaced by a computer. ¹⁷

Although even then it would have been unlikely that two replacements could be found who would have been prepared to work for their whole careers in a basement, the consequences of the unthinking replacement of registry practice with electronic systems can readily be seen in the documents submitted to the Hutton Inquiry that are available on its website. ¹⁸ Confirmation of the rapid deterioration in government record keeping is provided not only by the BSE Inquiry but by the documents, also available on line, obtained by the Campaign Against the Arms Trade (CAAT) relating to arms contracts with the Saudi Arabian government in the 1980s. These bear all the references from the registry system that are largely absent from those submitted to the Hutton Inquiry. ¹⁹ Moreover the Treasury tags that hold the documents together in the files are clearly visible. The problem is that the Information Technology (IT) community is largely concerned with short-term development of systems and tools, data flows, efficiency and receptiveness of service delivery to the customer, rather than the management of content in the longer term which it is believed, naively, can be addressed by the use of even more sophisticated free-text search engines. As Sir Gus O’Donnell pointed out in his evidence to the Chilcot Inquiry on 28 January 2011:

I would stress, that if you look at – just take the Internet. There is a mass of information there. The really difficult bit is how do you search for it? I mean, it comes up with respect to – one of the issues is record-keeping. I must admit when you put different things into a search machine you get very different answers. ²⁰

Deficiency in record keeping in the electronic era is not confined to the United Kingdom and may be being compounded by concerns about information finding its way into the public domain by way of Freedom of Information (FOI) requests or leaks, something that Butler drew attention to in his evidence to the 30 year rule committee. The evidence is patchy and often anecdotal. Although the United Kingdom Information Commissioner’s Office denies any change in practice, ²¹ in July 2002 the Swedish National Audit Office concluded that ‘the generous principle of public access to official records has led to a situation in which processes are not documented and decision making processes are collected in informal personal networks that are completely inaccessible to the general public’. ²² Such informality, which may have as much to do with a cultural shift than FOI, is enabled and reinforced by technology, particularly, as Andrew Turnbull pointed out, email that can avoid official channels of communication much more easily than the analogue equivalent. In the conduct of the war in Iraq, it is reported that the Coalition Provisional Authority (CPA) so distrusted the State Department that Colin Powell encouraged staff working in Baghdad to communicate with colleagues in Washington using remarkably insecure Hotmail and Yahoo email accounts to avoid detection by either the CPA or the Pentagon and, by so doing, ensured that no trace would ever find its way to the National Records and Archives Administration (NARA). ²³

The historian John Vincent has rightly worried that history itself is threatened by such new forms of communication:

Whether there can be history under late twentieth century electronic conditions, remains to be seen, but the only safe assumption is that it will bear little resemblance to post-Renaissance historical study, above all where individual motive is concerned. A study based on the written word cannot survive the marginalization of paper. We may be on the verge of a new pre-history, with the era of serious intricate, intimate knowledge of the past merely a fortuitous interlude. Electronic communication means no history. ²⁴

Richard Evans, in his much read book In Defence of History, dismisses as ‘foolish’ Vincent’s concern that history itself is threatened by electronic forms of communication. ²⁵ It is not difficult to reach the conclusion that the historical and archival communities are talking past each other, as much as the IT and knowledge management communities. While records managers and archivists can provide the infrastructure to address Vincent’s worries, there is little they can do to mandate it, when the instruction to break the rules comes from the highest authority and the IT community designs and implements systems which disregard long term access and preservation. The publication of Information Matters under the signature of the Cabinet secretary is part of a tradition. In 1909 Sir Gus O’Donnell’s predecessor Sir George Herbery Murray wrote, himself, the memorandum spelling out registry procedures. ²⁶

The introduction of computers and the accompanying cultural shift did not simply presage the end of registries, it also brought to an end a way of doing business in the United Kingdom civil service (and the many Commonwealth countries in the world that followed its practices and procedures) that dated back in an unbroken chain to the Tudor revolution in government in the late fifteenth century, if not earlier. As Lord Panmure, then Secretary of State for War, opined in 1855: ‘The great desiderata for the easy and efficient discharge of the duty of a public office is a simple and efficient system of registration of the papers of the department’. When the United Kingdom Foreign Office Registry was reorganized in 1918 it was unequivocally accepted that: ‘A Registry is not an end in itself. It is a necessary, even if expensive means of facilitating the rapid and smooth working of any large administrative machine’. Not only was the system designed for efficient office management, it also drew a clear distinction between the back and front office, both to prevent fraud and, as importantly, to protect officials; very necessary if they might find themselves threatened with the scaffold. There was inevitably a tension between the efficient management of an office and the careful procedures and processes needed to detect and prevent fraud and protect officials. ²⁷

By the early eighteenth century procedures for the creation, distribution and ‘putting away’ or custody were specified in some detail in office manuals. ²⁸ These were refined during the following century as the bureaucracy expanded. ²⁹ The file held together with Treasury tags, as we know it, was introduced in the Treasury in the 1850s and had become ubiquitous across government by the end of the century. The drafting of letters for typing, checking, referencing in accordance with file plans and finally registration was an integral component of good practice and efficient office management. ³⁰ Such processes may have been slow, at times cumbersome and perhaps militated against efficiency, but they did provide ample opportunity for reflection and guarded against the inappropriate use of language and government distribution channels. Crucially it enshrined fiduciary protection even for minor officials as a third party was always aware of the contents of communications. As Barbara Craig has shown the system was fallible like any other ‘social practice’, but anyone who has had cause to use the papers and files it generated cannot fail, as Christopher Foster was, to be impressed by the quality of the information that was recorded even after use of the telephone became widespread. ³¹

With the rapid introduction of personal computers in the 1970s coupled with pressures to reduce costs and modernize government, ministers and civil servants began to type their own letters and memoranda and later and more significantly to send their own emails. Crucially, receiving and sending emails bypassed the elaborate post room registration of all incoming and outgoing mail. Quite apart from the obvious collapse of the back office into the front office, which was often referred to by management consultants and information technologists as ‘leveraging the back office’; born digital objects, particularly emails, have characteristics that are inimical to the generation of files. ³² Digital objects differ from their analogue equivalents, as they are stored automatically by the operating systems as single independent instantiations and can only be made to relate to one another by embedding references, now termed metadata, at the time of creation. Although, as we have seen, referencing was mandatory in the paper world so that records could be easily retrieved, few in the front office were familiar with the file plans and registration arrangements and some thought them unnecessary. In any event senior managers cannot be expected to act as if they were filing clerks. Because of the way electronic systems work, all born digital objects, unlike their analogue equivalents, automatically leave multiple traces on portable devices, desktops, servers and backup tapes unless deliberate action is taken to destroy them. This characteristic results in the unnecessary retention of a mass of duplicate and inconsequential information, which has to be managed. In the paper world ephemeral material and duplicates were either never filed or destroyed shortly after creation. ³³ Another feature of digital objects is that they can be distributed widely, often with little trace as to whom, and can therefore subsist as multiple instantiations with little evidence as to their context or even provenance, unless steps have been taken to embed it in metadata at the outset. Ease of distribution not only complicates decision making, but also presents a huge security risk, very evident in the recent Wikileaks controversy, ³⁴ that can often divert attention from effectively managing the content. To compound matters, despite the potentially large number of instantiations, there is no guarantee that they will survive or at least can be read over time as software is upgraded and platforms change. ³⁵ This situation was described as the ‘digital dark age’ at the Time and Bits conference held at the Getty Center in Los Angeles in February 1998, leading Stewart Brand, one of the co-sponsors, to observe in The Library Journal as recently as 2003: ‘There is still nothing in the digital world like acid-free paper’. ³⁶ Although a great deal of money globally has been poured into digital preservation in the last decade, Seamus Ross, an international authority, reported gloomily in 2007 that ‘after more than twenty years of research in digital curation and preservation the actual theories, methods and technologies that can either foster or ensure digital longevity remain startlingly limited’. ³⁷

Nevertheless, The National Archives (TNA) in the United Kingdom, in common with those in other countries, is committed to ‘ensuring continued access to digital information in the future’. ³⁸ This is supported by advice to government departments on digital continuity – ‘Digital continuity is the ability to use your information in the way you need, for as long as you need. If you do not actively work to ensure digital continuity, your information can easily become unusable’. ³⁹ Direct engagement in the process of current record keeping marks a departure from the thinking of Sir Hilary Jenkinson, who in 1922 first systematized British government archival practice in his seminal work A Manual of Archive Administration. ⁴⁰ Jenkinson’s fear was that by becoming engaged in the process of record creation the archivist might compromise ‘the sanctity of the evidence’, as he had seen happen in Nazi Germany. ⁴¹ Jenkinson was unequivocal in describing the duty of the archivist: ‘His Creed, the Sanctity of Evidence; his Task, the Conservation of every scrap of Evidence attaching to the Documents committed to his charge; his Aim, to provide, without prejudice or afterthought, for all who wish to know the Means of Knowledge’. ⁴² In 2006 Eric Ketelaar, General State Archivist of the Netherlands from 1989–97, echoed Jenkinson’s concern about the moral defence of the archive: ‘Archives are memory because they are evidence. They are not only evidence of a transaction, but also evidence of some historic fact that is either part of the transaction itself, or that may be traced via the transaction, or that which is otherwise embodied in the record, or in the context of the archiving process’. ⁴³

Jenkinson would not have been troubled by the TNA providing the building blocks for the wider ‘corporate governance and what to keep’ initiative within the Cabinet Office, led by Knowledge and Information Management Unit. As has always been the case with any record-keeping system in bureaucracy, ‘the programme seeks to address the cultural issue of balancing business needs and practices with the corporate need to capture an adequate record’. Apart from providing the technical infrastructure that in some senses is straightforward, the TNA has, however, to come to an understanding of what represents the historical record in the digital age. Jenkinson believed that this was the task of the creators alone, as they were best placed to assess the historical significance of the records they created. Despite the risks of disclosure through FOI and the breakdown in trust between ministers and civil servants, anecdotal evidence suggests that senior civil servants still uphold a Weberian view of their function and wish to leave some testimony of what they contributed to public life. In the digital environment this is not straightforward as documents are held as single instantiations within electronic document record management (EDRM) systems and are linked to the file plan through associated metadata. One of the problems is that such systems are simply bolt-ons to software and hardware used to create documents, which only have limited tools for managing them, if they have them at all.

It is not enough for archivists and historians simply to lament the loss and perversion of the public record that these developments presage; they need to consider what they can do to influence policy and procedure to ensure that history can be written. In Richard Evans’ words, that ‘it really happened, and we really can, if we are very scrupulous and careful and self-critical, find out how it happened and reach some tenable though always less than final conclusions about what it all meant’. ⁴⁴ Despite all the rhetoric of accountability and transparency, many aspects of government actions can only be properly scrutinized long after the event. Crucial areas of policy are exempt in the ‘public interest’ from FOI inquiries. Historians must trust governments to maintain an adequate record of its actions and both support archivists in their endeavours and contribute to the debate about the function of both the ‘archive’ and ‘history’ in a democratic society. ⁴⁵ British archival theory has its foundations in the rule of law and the TNA has a juridical function through the Advisory Council on National Archives and Records, a statutory body chaired ex officio by the Master of the Rolls, whose title derives from his original duty to supervise the safe custody of legal and state papers. ⁴⁶ The authority of the rule of law over the executive has recently been reaffirmed by Lord Justice Moses and Mr Justice Stevens in the judicial review of the case brought by Corner House and the Campaign Against the Arms Trade (CAAT) against the Director of the Serious Fraud Office and BAE Systems PLC in which they stated: ‘No-one, whether within this country or outside is entitled to interfere with the course of our justice’. ⁴⁷ Such authority depends crucially on the creation of records in accordance with processes that leave little room for doubt about their integrity and are held securely and fiduciarily by the archive. In their evidence to the Chilcot Inquiry, both Andrew Turnbull and his predecessor as cabinet secretary, Richard Wilson, compared unfavourably information management practices in the public sector with the very strict rules that govern the creation and release of information in public companies in the private sector. By placing information management at the heart of corporate governance in a sense in juridical context, they were moving well beyond the administrative efficiency advocated by Lord Panmure and breaking new ground in terms of the accountability of the cabinet that the Inquiry may well comment on its findings. ⁴⁸ Richard Wilson observed caustically: ‘I have been thinking to myself there is a deep irony – I used to think this after I left the Civil Service – that the Blair government, which imposed rigorous or pretty rigorous standards of governance on companies in the Higgs Report and other codes, was itself so non-compliant’. ⁴⁹

If we accept that the archive, however constructed, must have evidential value, as Jenkinson would have insisted, for a democratic society to function under the rule of law, the challenge for archivists and historians is to suggest a route out of ‘the digital Dark Age’. They may not be able to contribute much to the debate about the conduct of government except to raise their voices, but they can call attention to past practice in the analogue that forms the bread and butter of much contemporary history. There has been relatively little writing about the diplomatic of modern documents and the way in which such study might inform practice in the digital. A shining exception is Luciana Duranti of the School of Library Archives and Information Studies at the University of British Columbia, who has constantly championed the use of diplomatic in shedding light on the digital. ⁵⁰ She draws inspiration from the legal forms of her native Italy, which she has applied in three sequential projects titled: The International Research on Permanent Authentic Records in Electronic Systems (InterPARES). ⁵¹ What is of interest here is the diplomatic of the files that the historian encounter in the reading rooms of national archives. In the United Kingdom there were strict rules governing the creation and registration or filing of documents, as we have seen, that persisted until the introduction of the personal computer. ⁵² The British government file had a specific form. The cover sheets had a reference number, which linked it to a filing plan mandated by the department’s central registry. ⁵³ The file had a title created by the registry clerks to describe the contents that appears, more often than not, in the TNA catalogues. The cover also bears the initials and dates of the civil servants and ministers who had cause to consult it. The contents are filed in date order on the right and held together with Treasury tags. Ephemeral material, such as notes and memos, not considered worthy of preservation or ‘putting away’ was filed on the left for destruction in weeks or months. In 1909 Sir George Murray, the permanent secretary of the Treasury, instructed that:

The papers in the file should be arranged in chronological order beginning with the letter, minute, or other documents which originated the file, together with its enclosures if any.

Only papers which it is necessary or desirable to keep permanently on the file should be treated in this way. Other memoranda or enquiries should be written on separate pieces of paper, which should be detached immediately when they have served their purpose. ⁵⁴

The file grew organically within British registries out of the docquet system, as the Notes for the Use of Registry Branches in 1919 explained:

The jacket or file cover is historically a survival of the eighteenth century double sheet of foolscap, on which was written a minute, giving the instructions as to the action to be taken or the reply to be sent, and that minute, before the general introduction of the letter copying press served as the only record of the reply. ⁵⁵

The docquet consisted of foolscap sheets at first stitched together and later held together with Treasury tags. It was folded twice horizontally to create a docquet, endorsed with a description and the date by a clerk, registered with a unique number, indexed and cross-referenced, since all documents inevitably have a one to many relationship. They were pigeon-holed in specially designed cabinets that resembled pigeon nests in dovecotes and the location recorded in a separate register. Sometimes docquets relating to a common subject were brought together in so-called ‘long bundles’ that could extend over a number of years. Such assemblies were noted in the location registers. When they ceased to be current, the docquets were ‘put away’, often in sacks that until relatively recently was the way they were produced in the TNA search rooms. In the late nineteenth century the introduction of means of mechanical reproduction allowed outgoing correspondence and memoranda to be added to docquets. For convenience of adding material, instead of being folded they were stored flat, resulting in what we know as the file. What is of interest here is not the effectiveness of the registry; but the docquet system itself.

EDRM systems, where each object is held as a single instantiation and cross-referenced, are the digital equivalent of the docquet registry and accompanying pigeon holes before the file was invented. The one exception is that it is much more difficult to find mis-registered papers in EDRM systems, not just because of the digital format, but because the back office has been collapsed into the front office. In the nineteenth century the clerk in the front office would describe the contents of a document in the endorsement and the registry clerk (back office) would allocate a place in the file plan and place it in the correct pigeon hole. Now the creator has to perform both functions. Research has shown that the creators of documents will only apply metadata to documents if it adds value to their work. ⁵⁶ They cannot be expected to be glorified filing clerks, however much knowledge and records managers would like them to be. Moreover in the analogue civil servants can review well ordered files for preservation or destruction, which is not possible in EDRM systems. This in turn militates against the weeding (selective destruction) of ephemeral material and review of contents for sensitive items and compliance with data protection legislation. This is an enormous challenge for any ‘what to keep’ project. The trick is to develop tools and protocols, as the Effective Records Management project at the University of Glasgow proposed, that make use of the potential of the technology to improve efficiency while at the same time managing the information objects:

A document in the digital order can and should be used for further processing within the transactions of which it forms a part, rather than simply being a step in the creation of a representation of information on paper. In the paper order documents do not have this potential. If advantage is not taken of these possibilities, we are simply using a more expensive set of tools to create the documents that we previously created with typewriters, with few advantages to show for the additional expense. ⁵⁷

With its preoccupation with short-term objectives and data flows, the IT and data preservation communities are re-inventing the paper order in the digital:

Digital preservation not only deals with migration to new carriers and new formats and with maintenance of functionality, it starts with a very basic requirement: namely, that documents can be identified unambiguously and located by those who need them. ⁵⁸

Like so much in the digital order, this statement of the obvious assumes that this problem only arose with the emergence of the electronic computer and no one ever before had persistence in mind when creating documents in the paper world. The conclusions of the report, from which this quotation comes, cover ground that those involved in the management of information have grappled with for a long time:

A system for persistent identification is built just to ease the administration, not to make it obsolete …

Institutions should recognize that the implementation of persistent identifiers always comes with some costs. ⁵⁹

Compare these with statements written in 1918 and 1919:

A Registry system ought to follow to some extent the anatomy of the organization to which it belongs. ⁶⁰

A registry is not an end in itself. It is a necessary, even if expensive, means of facilitating the rapid and smooth working of any large administrative machine. ⁶¹

What the digital preservation community is proposing looks remarkably like the docquet system the file replaced, but with no added functionality. If the file was simply a way of holding single instantiations of documents together, taking advantage of available technologies – hole punches, Treasury tags and critically the letter copying press – the Knowledge Council should not be so quick to dismiss them as the legacy of another age, but rather learn from past experience and explore ways in which they can seamlessly be created in the digital environment. The Knowledge Council may have been formed ‘to address ways in which knowledge and information management can better support good public service delivery, well-informed policy making, and the wider exploitation and reuse of information’; ⁶² but that after all is the very mechanism that Jenkinson argued generated the evidence on which historians rely to develop their hypotheses. It begs many questions and overlooks the critical place of information management in corporate governance, to which Andrew Turnbull and Richard Wilson drew attention in their evidence to the Chilcot Inquiry.

If Evans is correct in dismissing Vincent’s concerns as ‘foolish’ then historians need reassurance that EDRM systems will deliver in the digital environment what the docquet system did for almost 400 years before the file emerged. They certainly do not do so in their current configuration. There are many organizational and technical impediments; but they should not be allowed to get in the way of the need for individual instantiations of documents that have been preserved in a secure fashion and can be linked together unambiguously so that government can be called to account in the courts of law and history. Technical issues aside, much of the discussion in the digital community is bogged down by broken links – the 404 Error – File not found syndrome – and the potential that the new technology proffers to capture everything. Such issues have never troubled either the courts or historians in the analogue world. There have always been broken links and there has never been an expectation that everything will survive – heaven forbid. The historian has always accepted appraisal and loss as a necessary part of the function of the archive that supports the process of research and writing. In such discussions the voice of the historian is noticeably silent, although those of former civil servants, the press and the public increasingly strident.

We are now a long way from where we began and we have traversed a great deal of territory. Although it is tempting simply to blame the loss of the file, with which we, as archivists and historians, are comfortable, on sloppy business practices (action points) or the uncritical use of new technology, the explanation is much more complex as Peter Seeger’s lyric, of which the title of this article is a parody, suggests. The constant refrain of ‘Where have all the flowers gone?’ is ‘When will they ever learn?’ For historians and archivists alike, in confronting the challenge of the digital and what it appears to enable, learning cannot be delayed for ever, otherwise there is a danger that the essential ingredients that allow government to be called to account in the courts of law or history will be lost and Vincent will not be as foolish as Evans supposes. The historical community needs to be reconnected with the archival and the two communities need to speak to rather than past each other in addressing issues fundamental to our democratic freedoms. Jenkinson observed perceptively in 1955:

the mere manufacture of documents is only one element in the creation of Archives: another and much more potent one is their preservation for reference; that is to say their substitution not merely for the spoken word but for the fallible and destructible memory of the people who took part in whatever the transactions may have been that gave rise to them. Recordari still means, as it meant in the twelfth century, to remember. So long as memory is a necessary part of the conduct of affairs so long will it be necessary to put that memory into a material form, and so long as that is necessary so long will you have Archives, whether they take the form of writing on paper or parchment or palm leaves by hand or that of steel tape (shall we say) engraved by mechanical means with microscopic grooves which enable you to reproduce at will the voices of men who forgot or have been themselves forgotten. ⁶³

This is something government and those in positions of power and authority need to be reminded of, whether ‘engraving by mechanical means’ is a groove on a disc or in a bit pattern. As the Canadian archival thinker Heather MacNeil, following in Jenkinson’s footsteps, has emphasized recently: ‘Archival custodianship has always been linked inextricably to the protection and safeguarding of evidence. Physical ownership of the records is merely the means by which, historically, archivists have assured that protection. The advent of information technologies does not change the substance of our custodial responsibility; it simply changes the means by which we exercise it’. ⁶⁴