It's been interesting to see that the data coming from digital library professionals shows that there is an emphasis on access via digital to the physical originals, and preservation of physical holdings by creating digital 'surrogates'. It's interesting because, while there's been general support for the idea of integrating digital and physical collections, it would seem that the current reality in libraries is that digital collections are very much in a supporting role, always leading back to the physical original.
This can be explained in several ways, not least by the resource pressures that most libraries are under. Digital collections are an addition to a library's existing holdings, and demand different models for their construction and upkeep, so using them in such a supportive and economical manner is understandable. The idea of digitisation as preservation is certainly the most overt example of this.
Let's be clear: we are not talking about the mistaken notion that digital versions of physical originals will allow them to be kept in perpetuity within that medium - anyone who knows about the digital medium knows that to be naive. Rather, it's essentially the idea that once you've digitised your originals, they don't need to be consulted to the same extent, if at all, thus 'preserving' them from the most likely cause of damage, which will always be transport and handling. This makes sense when original material is at death's door (photographic and moving image collections come to mind), but unless this is the case across the board, the idea seems worth further consideration.
As an example, say you've got a few thousand ancient manuscripts kicking around - they're frequently accessed and you decide to digitise them. You want more people to be allowed to see them and maybe increase your library's profile a bit, but you also intend to minimise access to the originals once they've been digitised. If you follow the logic of preservation to its conclusion, you might try to restrict access to the originals altogether. That logic is: protect an item for the future by preventing causes of deterioration now, which is inherently at odds with an open access policy.
A couple things can happen here. First of all, much of the anecdotal evidence I've come across suggests that requests for access to originals increases once they have been put online in digital form. You then have more people seeking to view material to which they can't have access, which might upset the punters. Even with more tightly controlled access, if requests do in fact go up due to the library's online content, even requests that fit these new parameters should also be expected to rise, leaving no drop in physical handling in overall terms.
Next, having re-written your access policy to achieve this, you have to wonder whether someone higher up might not query why a library is continuing to use expensive storage (think real estate, climate control, and so on) for material that no one is actually looking at. Of course, it depends on the collection, and perhaps a collection of nineteenth century books might be removed to a salt mine, if you have one, without much ado, but it sits awkwardly alongside the library's mission.
In my experience there seems to be a disjoin between preservation ethics and the major factors influencing library development today, and I'm worried about a self-imposed obsolescence for physical holdings, paradoxically driven forward by those who treasure them the most. Add to that the fact that physical original and digital version have their own distinct attributes and seem unable to replace each other, combined with the fragile digital medium as realistically only a temporary solution, and this leaves the notion of digitisation as preservation perhaps not everything it appears to be.
