The Past Speaks

Back in May, Stefan Schwarzkopf  of Copenhagen Business School posted an essay about the relationship between business historians and corporate archives on NEP-HIST.  You can download the entire paper here.

Abstract:  Archival records are a constitutive element of business historical research, and such research, in turn, is fundamental for a holistic understanding of the role of enterprise in modern capitalist societies. Despite an increasing debate within business history circles about the need to theorize the historian as author and creator of narratives, a fuller reflection on the uses and limitations of the archive in business historical research has not yet taken place. This article takes its lead from theories of organisational epistemology, and asks to what extent business historians are trapped by an outdated, realist methodology and epistemology which is in danger of ignoring the multiple roles that archives play in their knowledge production.

Essentially, Schwarzkopf  is asking for business historians to be more critical in their use of this particular set of primary sources. Scholars working in other branches of history have recently become much more conscious of the ways in which the selectiveness of their archives bias their work. For instance, historians of criminal law are aware that police archives give us the perspective of State employees, not those of the people who were deemed to be criminals in a given era. Religious historians are acutely aware that church archives give us the perspectives of the missionaries, not the so-called “pagans” they were attempting to convert. The adoption of critical stances to archives by other groups of historians  has been driven by the emergence of postmodernist and postcolonial perspectives on the sociology of knowledge.  Most business historians, according to Schwarzkopf, are stuck in an outdated and uncritical mindset towards the corporate archives that are the foundation of their research. They are, in his view, naive empiricists.

Stephanie Decker, a business historian at Aston Business School, has followed up Schwarzkopf’s piece with a short reaction essay of her own, Decker’s piece draws on her research into the area of African business history and the development policies, which has involved trips to the World Bank Archive in Washington, DC.

In my view, the most interesting part of the essays by Decker and Schwarzkopf relate to digital technology.  Since the 1960s, government and corporate archivists have been struggling with the issue of how to save data recorded on punch cards, magnetic tapes, and successive generations of electronic storage material.  (You can read about the first generation of digital archivists in the US government here). More recently, there have been efforts to put documents online.  For instance, you can now read the handwritten letters of Abraham Lincoln from the comfort of your own home.

Decker applauds organizations such as the World Bank for digitizing parts of their vast holdings and putting scanned images of certain document categories online.  She points out that since digitization of hard copies is inevitably selective, it may bias future historical research towards topics and perspectives supported by those documents which happened to put online.  Decker writes:

How does digitisation affect how archives are used, and vice versa? Will it determine what the collection stands for, more so than the entire body of files? Perhaps not a new problem for libraries that contain individual high value items that eclipse the totality of their collection, but certainly a phenomenon that will spread with digitisation. Just consider decisions to digitise parts of archival collections that are of greater public interest, such as World Bank’s digitisation of the Robert McNamara’s files. Faced with the impossibility of digitising an archive as vast as theirs, files of greater relevance to present-day audiences are prioritised, negating the need for people to physically enter 1818 H Street, NW, and engage with the overall collection. Is this a manipulation by the archivists, or is this it the pressure of demand shaping organisational responses?

Schwarzkopf  asserts that digital records are easier to manipulate and delete than the hard copies. (For the time being, let’s assume this claim is correct, although I’m sceptical because many electronic documents leave traces that can be recovered by experts.)  Selective editing of archived emails may create problems for future business historians interested in the early internet era (i.e., the present). Observing that much business communications (reports, emails, memos, etc.) are increasingly becoming  digital-only, he suggests that there is little to stop governments and corporations employing twenty-first  century Winston Smiths to deal with their own digital records in the same way?

Schwarzkopf is here alluding to the protagonist of George Orwell’s 1984.  I’m not suggesting that Schwarzkopf’s concerns about the deletion of incriminating documents are invalid.  There have been examples of such documents “disappearing” from archives or the archives simply remaining personally closed to researchers. Enron employees shredded many documents in the last few days of that company’s existence. Deutsche Bank kept records related to the Holocaust secret for years. In 2005,  Hydro One, a Canadian SOE, suddenly closed its archive to all researchers when a non-academic began to use their archive to find material to support a lawsuit against the company. Art has imitated life:  the plot of the film Michael Clayton revolves around the efforts of a fictitious company to hide a document

Personally, however, I am less worried than Schwarzkopf  about the selective editing by people looking to hide incriminating documents than about the simple accidental deletion of documents. The most serious problem is the deliberate deletion of documents related to storage costs.

Today’s business historians depend on documents that others have kindly saved for us.  Since 1934, the Business Archives Council in the UK has been helping companies to save their archives and make them available to outside researchers.  There are equivalent organizations in other countries. These efforts, which were supported by the business historians of earlier decades, make our research possible today. We have an obligation to help preserve today’s corporate records for the future.

This means that we need to think about how we can save the data formats that are being created today. It occurs to me that cloud computing might allow us to do this cheaply.  Many companies now outsource the storage of their email and other data to trusted firms such as Amazon Web Services. Interestingly enough, the cloud computing divisions of Amazon and IBM are now suing each other for the right to store the data created by the CIA and the NSA in the United States.

It occurs to me that the Business Archives Council or some other charitable organization might undertake to save all or part of the data that a cross-section of companies upload to the cloud.  Perhaps it could be a completely separate organization. Let’s call it the Business E-Archives Council or BEAC for short. Under this scenario, the customers of the cloud computing firm would consent to the release of part of their data to BEAC on the understanding that the data would be kept in a secure environment and would only be released to the researchers for a predetermined period.

It seems to me that there are three basic reasons  a company might be reluctant to consent to a heritage organization copying a cross-section of their electronic files for posterity.

1) The first is concern that the data might fall into the hands of enemies of the firm.

2) The second  is the sheer administrative hassle of asking IT people in their company to liaise with with the archivists at the heritage organization to decide which email accounts to copy and how to go about transferring the files over.  Sharing information with others costs real resources, most notably time. That’s true regardless of whether one is setting up data backup account for one’s home computer or arranging cloud computing services for a major bank. It is difficult enough to synchronize data systems within a given organisation (such as a university at the start of term) let alone ask IT staff to allow outsiders to get involved.

3) The third concern relates to Public Relations in the post Edward Snowden era. A company might be reluctant to do business with a company that had announced it was allowing the BEAC to record some documents. After all, they might worry that consumers would be concerned about the protection of their data.

The first and third concerns could be addressed through a variety of legal and social mechanisms. First, you could reinforce the confidentiality agreements between the companies and the BEAC by incorporating the BEAC through a special piece of legislation that removes any doubt about whether the BEAC’s right to protect the data for term prescribed in the contract. In other words, the BEAC’s charter would be a special act of parliament or Congress.  Including representatives of the country’s top companies and business leaders on the board of the BEAC would also bolster the credibility of the firm with the data-generating companies. Including prominent citizens of the country in question on the board of the committee might help to allay consumer fears about the BEAC.

Cloud computing would help to reduce to the costs of participating for the companies. Since company X has already done the difficult work of making its systems work with those of the cloud computing company, allowing the BEAC to record a cross-section of their archived emails would not cost them any person hours. I admit that the connection between the cloud computing companies and BEAC would cost some money to set up, but surely it is easier for the BEAC to deal with just one or two cloud computing firms than with all of the companies served by these firms.

The benefits of recording the electronic data for future generations of business historians would be massive. Exciting things happen when academics meet big data. Consider what Google N-gram allows literary scholars and people in the field of corpus linguistics to do. I would love to do keyword frequency counts of the internal correspondence of the companies I study.