I agree 100% with Paul Conway’s approach. In fact, I think most digital preservation strategies that are prepared today have adopted this viewpoint, i.e., (a) there’s no point preserving something if you can’t provide access to it and (b) digital preservation starts at creation (i.e. file format standardization, metadata capture, organizational resource commitment to long-term storage, etc.).

I have enjoyed reading your blog since I heard you speak at SAA last year. Thanks for some thought-provoking postings.

The discussion of the relationship between preservation and access in the digital context is something we are all grappling with. I like your idea that you cannot deliver the “original” digital object but can only rebuild it.

Coming at it from a slightly different angle, you might look at it the way Paul Conway does in “Rationale for Preservation and Digitization” in the NEDCC’s Handbook for Digital Projects. http://nedcc.org/oldnedccsite/digital/dman.pdf

He talks about preservation as the “creation of digital products worth maintaining over time” and says that preservation is inextricably intertwined with access to become what he calls the “preservation of access.” Thus, preservation becomes not only an aspect of access, but an aspect of content creation as well.

Ideally the institution has integrated its researcher registration and retrieval requests across both its physical and digital collections. In practice, though, this is typically not the case.

As well, in the case of digital objects, even though the bitstreams may have been retrieved from digital storage and are now available to the user in a specific network location, additional software and hardware may be required to render and view the archival materials. This is however, where I draw a distinction between access and preservation and use.

The other big difference (if this is what you are asking) is that it is possible to retrieve and deliver the original for physical archival materials. However, due to the engimatic nature of digital objects it is not possible to retrieve and deliver the original digital object. It is only possible to reproduce or ‘rebuild’ the digital information object based on what we know about its original content, context and structure. I discuss this issue in a previous post called The Anatomy of a Digital Information Object (see the section called “If a Digital Information Object Falls in the Forest…”).

Also, see section ’4.1.1.7 Access’ in the Open Archival Information System (OAIS) standard for more details on the request and retrieval process. The OAIS is now an ISO standard. PDF copies of ISO standards are sold on their website but you can still get the free Blue Book version that was submitted to ISO by the Consultative Committee for Space Data Systems (CCSDS).

I agree about access – it’s a movable feast, and you must define your own terms. But I’m glad you agree that it has to mean more than “locating”.

I like your IO definition particularly.

access
the ability to identify, locate and retrieve relevant archival materials

One thing to keep in mind is that these definitions are intended to work within the scope and context of how I am defining archives access systems and may not work universally. The list above is the first draft of a glossary that helps to define very specific boundaries for this particular type of system.

Your question about the definition of ‘access’ is very valid. This is one of the definitions I have not yet worked out fully. Firstly, I am looking at it purely from a functional point of view. There are lots of discussions about access restrictions and accessibility (i.e. the quality or universality of access). However, right now I am purely interest in the mechanics of the process. Essentially I am trying to draw a distinction between being able to identify and locate archival materials for retrieval from the ability to render and use those materials. However, maybe it should read ‘identify, locate and retrieve’ to make that point more clearly.

Lastly, ‘software-intensive system’ is a term used commonly in software engineering (e.g. see the IEEE standard). A typical information system consists of people, computer hardware, computer software, procedures and data (that is processed to make information). Each of these components are important but the focus of the design, development and user-interface work for such a system is typically on the software component. This is why the term ‘software-intensive’ is used.

Coupl’a questions:

Access: If I “identify” and “locate” something, is that really the same as “accessing” it?

Software-intensive: by what criterion are you assessing predominance?

All the best

JB