Some definitions, and distinctions

Very quickly the need has become apparent to take a first stab at some definitions of the terminology already peppering this blog. What, for example, is 'preservation'? This word alone, when used across different disciplines, has the capacity to induce instant confusion.

One example of this from my own experience is the digitisation process. This is a moment where hard-copy meets digital and two theories of preservation collide - one is 'traditional' preservation, the other is digital preservation. While the preservation of born-digital items seems inherently understood from its context (digital items need digital preservation), the problem comes when you apply the word (preservation) to collections of digitised materials themselves - is hard-copy 'preserved' by being digitised?

While the digital image data files of these digitised collections have entered the domain of preservation in the digital sense and become 'digital assets', the original hard-copy has not in fact been preserved as an object in its own right. Take the extreme example of digitised documents being discarded in the name of economy of space, only to find that the digital image data files have become corrupt, or the system required to interpret them has become obsolete. All evidence of the documents are then lost - they were never preserved. It seems important to make this distinction.

Certainly, stories of the ephemeral nature of digital assets are legion, and businesses are still forced to retain hard-copy. The problem of the previous example is as old as sound recording technology, well summarised in a Wired article on Digital Archaeology from 1993, when CD-ROM was becoming firmly established. Nevertheless, my impression at this early stage is that the digital asset manager is primarily concerned with born-digital material since, once a digital data file is created from hard-copy, that is the creation of a new asset (in addition to the original hard-copy) that enters the domain of digital preservation, just like any born-digital asset.

Once a digital data file is created, the term 'preservation' can then be comfortably applied within this understood digital context. For what that might then mean, I've just been introduced to OAIS (Open Archival Information System) designed by the Consultative Committee for Space Data Systems after the data migration problems that befell NASA after the Apollo Program (also mentioned in the Wired article). OAIS identifies two divisions of digital preservation: the preservation of digital data as it is migrated across media and across formats; and the preservation of access services to digital data as technology changes and software is ported (adapted) to new systems, including emulation. A pithy explanation of OAIS itself can be found here.