Already in 2007 Ross Harvey wrote a “provocative position paper” for Digital Preservation Europe under the title “So where is the black hole in our collective memory?” to draw attention to the lack of evidence of loss of digital material. I think his cry for evidence is still valid. Not only to warn new comers for the possible errors they could make, but also to convince funding bodies of the need for digital preservation.
He states that
From the literature it is only possible to conclude that the
evidence of digital data loss is, overall, anecdotal, and that
data can be recovered, albeit at considerable expense.
There is an urgent need in his opinion to have more evidence of data loss, to have an answer to the question to what extent data was lost and to the question whether the problem of digital preservation is as big as was assumed when people were talking about the digital dark ages and the digital black hole.
(Although in my opinion the attention paid to preserve digitial material in an organized way and concepts like OAIS, preservation metadata, Representation Information, etc. have saved a lot of data already.)
And he finishes with
Finding answers to these questions is crucial to securing resources to ensure the future of digital preservation. Inability to answer these questions will lead to scepticism about whether the problem is as great as claimed, and to lack of appropriate and adequate resourcing. This is an excellent reason to put some effort now into quantifying the extent of digital information loss or compromise, or, at the very least,to document more examples to supplement the few specific studies currently available.
Based on the limited evidence you can find under Stories
I intend to make an overview of reasons, time and costs involved in rescuing these data. But from the information I collected until now, one could be wondering that, while on the list of reasons for data loss in general Number One is human errors, the digital preservation community does not report on these kind of errors at all. Even not anonymously.