« March 2008 | Main | May 2008 »
April 15, 2008
Authenticity
So, if I show you the page image of a book scanned from Special Collections here at the Kelvin Smith Library, how do you know that IT IS the page from the book I'm telling you it's from? If I show you a WPA Print, how do you know that I didn't Photoshop something into it? Take away a line here or there, insert a new form, and erase a smudge?
This is the type of problem that is addressed by DigiCULT in its work on the Integrity and Authenticity of Digital Cultural Heritage Objects.
It is unfortunate that this article and a subsequent interview only outline the problem and provide no real solutions or suggestions. What is fortunate is that they outline the problem and that the article was written in 2002, so, much time has passed for answers to emerge. I'll be doing some lit searches soon to see what I can find out and will report back.
The issue I'm looking at is the first thematic issue (there are seven) and it:
concentrates on a question that is critical to all organizations that archive and provide access to cultural heritage objects: How to preserve and prove the integrity and authenticity of digital objects?
In the first case, authenticity, the article states, "When we work with digital objects we want to know they are what they purport to be." As for the second, integrity, we want to know that the objects are "complete and have not been altered or corrupted." In his article, Seamus Ross states that "digital objects are more easily altered and corrupted than say paper documents and records, creators and preservers often find it challenging to demonstrate their authenticity."
While I certainly agree with the latter part of the quote, I take issue with the former: that is, Ross presumes (I suppose) that not many people have uncontrolled access to the print or hard copy documents and records, as they can easily be corrupted (as an object), though the content may still be available; (Or the content can be changed or forged). In the former regard, one need look no further than the unfortunate incident with Samuel Berger Probe. This is to say, just because they are hard copy or print doesn't guarantee anything with regard to Authenticity or Integrity. But certainly the latter point of Ross's quote is valid: how do you demonstrate that the prints in your digital exhibit are what they claim to be?
The first and most obvious answer is strong, but upon close inspection has frightening elements: namely, the institution guarantees it. That is, who would argue with the Cleveland Museum of Art that the print they put up is by Kagy Sheffield or whomever? After all, it is a world renown museum and that is what museums do, right? They display their art and describe it on the tombstones. What point would CMA have in not putting up a "true" print? On the face of it the answer is clearly that they would have no reason and that, in general, it is a proposition in trust that should be accepted. But there is a shadow area of doubt--and for conspiracy theorists everywhere, the gap between what is absolutely proven and what is absolutely suggested is grand-canyonish--a buffet table to be laden with tasty nightmares and a swirling demon punch to wash it all down.
So, Ross states next, "if this is the case what are the requirements of authenticity and integrity functionality and what can be done to ensure that they are present in digital objects or in the systems that maintain them"?
Well, before you get to excited, Ross goes into a discussion of what exactly is meant by "authenticity" after all, a creator of an object might have a different view of what is authentic to him than will a preservationist or a connoisseur. So, underpinning authenticity and integrity and their preservation over time are the concepts of fixity, stabilization, trust, and the requirements of custodians and users. And further that, an authentic digital object is one whose genuineness can be assumed on the basis of one or more of the following: mode, form, state of transmission, and manner of preservation and custody.
Further, the different types of digital objects, including records, online journals, databases, audio-visual materials each appear on the surface to have their own requirements in relation to authenticity.
So, finally, Ross proposes a few areas for further investigation including:
Can general characteristics of authenticity be identified that would apply to all digital objects?
Do different types of digital objects, record keeping procedures and digital object creation practices, alongside the variety of institutional requirements mean that digital object preservation would require a range of mechanisms for enabling user and preservers to ascertain the authenticity of material?
One area that is especially fruitful and aligns with my own experience has to do with the following: More research needs to be centered on the creation of digital objects to establish how they might acquire these properties. That is, when the process of creating digital assets is actually underway, what can be done to assure and ensure that authenticity is proved and that integrity is assured? Is there some certification process that should be undertaken? Should the process be witnessed or observed by credentialed persons? What stamp can be placed on the process? After all, this is where the majority of issues will occur. (Sure, some post-processing can damage an object and so can massive re-distribution of an object.) But it all comes back to the source. There may be 1,000,000 digital copies of the Mona Lisa floating around, but the Louvre should have the certified, by-God original that always can be pointed back to.
This leads Ross to ask the following questions:
How in practice can the creators be influenced to produce and create digital objects in ways that can guarantee and provide evidence of authenticity and integrity?
How can vendors be encouraged to produce products that will allow creators to create records and objects with characteristics that enable authentic digital objects? It is currently impossible to purchase a "preservation solution" off-the-shelf.?
Ross notes that one way is to break the checks into domains: "agreeing to two sets of requirements for authenticity: one for the creator and the other for the preserver."
But an even larger concern to Ross is that "the communities that create and preserve digital objects forming the backbone of this discussion" are public administrations, broadcasters, publishers and libraries, and each seem to think that their way is very specific to them and that no general manner of processing items can be undertaken; and that "few controls existed on the creation, preservation and dissemination of digital objects" in any of these environments. So, he asks, "Can there then be one solution that all types of organizations, regardless of size and institutional culture, could adopt that would support the creation and preservation of authentic digital objects"? A very important question, which, again, I'll do a lit search on and try to figure out.
Later, in an interview with Dr. Ulrich Kampffmeyer, DigiCULT examined the question of Authenticity and Integrity again, with Dr. Kampffmeyer speaking about the "digital flood that threatens to drown us. Unless", he says, "the ICT industry and the public sector are able to co-operate in developing solutions for document life cycle management. A life cycle that not only comprises the generation and use, but also the long-term availability, and the guaranteed authenticity of documents." That "the sensible thing to do is to try and develop standards, predefined structures, metadata and interchangeable formats, through co-operation between the public sector and industry." According to Kampffmeyer, the traditional problems of archiving are the kilometres of paper documents, whose integrity is threatened by their fragility and a fading consistency. "These problems seem small when compared to the archiving problems created by the use of new technology." He continues: "A lot of this information is digital only, meaning it has no physical representation. That makes it much more volatile. An XML-document for instance is created while you view it. So how do you keep it" Other problems exist, though, including one related to what I mentioned above: a "certification" process for files. Even if we do manage to find a way of approving or certifying something as being Authentic or having a defined Integrity, would the "stamp" even last as long as the file: Kampffmeyer notes that "you can create electronic documents and you can keep them in an electronic archive. But, for example, "an electronic signature to authenticate a document is invalid after three years; the migration of data and the accompanying loss is still an unsolved problem, and the history of electronic records can be tampered with quite easily." Ultimately, Kampffmeyer points out that "archives of physical documents have been in existence for over 6000 years, whereas, electronic documents have only been around for the last twenty years, or so. The people who develop the technologies, haven't made the mental transition yet, from creating and using a document to the long term availability of that same document. And this is what will be necessary," he asserts, if we are to realize the goal that "electronic archives are the memory of the information society."
Posted by twh7 at 07:31 PM | Comments (0) | TrackBack