A Question of Authenticity

I’m just on my way back from Wigan, having given a presentation this afternoon on the role of the records manager in digital preservation to a group of, you’ve guessed it, records managers. I was really encouraged to see that the group had decided to dedicate their entire meeting today to tackling the thorny issue of digital preservation. It doesn’t happen very often – usually there’s just the odd session on digital preservation at this type of event – but it’s really great that they recognised from the start that digital preservation can’t be covered in 45 minutes! That’s not to say it can be covered in a day of course, but at least it gave them the opportunity to hear from a number of different speakers, all of whom approached the subject from a different perspective.

One of the things I decided to focus on was the issue of authenticity. It’s a real interest of mine and I’ve been wondering for a while if perhaps we might not be paying as much attention to it as we ought to be, particularly insofar as office records such as text documents, spreadsheets and emails are concerned. The immediate value of this type of record lies, to a great extent, in their evidential value. Their evidential value rests on their authenticity. If the authenticity of the records is compromised then their evidential value is too, and then we run into all sorts of issues like legal accountability and so on.

One of the reasons I think it’s such a relevant issue for records managers is the fact that even small migrations through different applications and application versions have the potential to impact on the authenticity of a record. Such migrations are known to have the ability to alter, for example, automatically generated content such as date and author fields in a text document. In a spreadsheet that derives strings of cell content from embedded formula, a migration of this sort can also impact on the end calculation – particularly if the spreadsheet has been badly created or has errors in it.

The problem is, of course, that migrations on this scale can take place fairly frequently. An organisation may decide to ‘upgrade’ the software they use, or a user may decide to access and develop a shared file with a different application to the one that’s commonly used, with unexpected results. If there’s no attempt to control these circumstances (and I’m not necessarily saying they can be controlled but that we should at least make an attempt to control them) then the risks to a record’s authenticity are increased. And all this before it even reaches an archive.

We talk a lot in our community about Trusted Digital Repositories and preserving digital objects once in the archive. But I’d really love to see some more discussion of how we can ensure records are maintained in an authentic way before they are actually ingested. This is where the records managers – and the records creators – come in. Because otherwise, we run the risk of preserving records whose authenticity could be in question despite their storage/preservation in a TDR. And, because they are stored/preserved in a TDR, their authenticity prior to the point of ingest may never even be questioned and non-authentic records therefore find their way into use.