Authenticity of digital objects affects us all regardless of whether we curate, preserve, access, or create them. As an authoritative body, our integrity is on the line. In a world where digital objects are moved from one physical storage media to another because the first media has reached the end of its lifespan, and where an object must be migrated from its original data format to one or more other formats because the original format is no longer accessible, what are the principles that guide our choices and by which the quality of our objects can be measured?
For example, a digital image captured in a Sony Mavica (proprietary) data format on flash memory is migrated to a TIFF data format on a gold-based CD. What will be our rationale for asserting the authenticity of the image once it is a TIFF and on a CD?
So I ask the question: by what standards do we determine a digital object's authenticity?
Subscribe to:
Post Comments (Atom)
5 comments:
Let me draw a parallel here and also touch on the previous thread (objects born digital).
In the astronomical community, as in many observational sciences, almost all observed signals are immediately digitized and recorded as digital data. So, our (raw) objects (images, photon event lists, spectra, radio signals) are born digital and in their (authentic) form as a specific bitstream autheticity is not an issue.
However, quite commonly further processed versions of these digital objects (calibrated, corrected, etc.) will be created and here the question of authenticity does come up. My preference would be to consider such a data product authentic if it is produced by the entity under whose authority the original was created, or an entity to which this authority was conferred. What that means is that for Chandra X-ray Observatory data, the objects containing raw telemetry streams are considered authentic, but also all the derived data products that are generated by the Chandra X-ray Center which happens to be the creator, as well as the publisher, of Chandra observations. On the other hand, if Joe Astronomer pretties up an image or derives another data product, this would not be considered authentic.
In the context of Ricc's example, the question comes down to: "Who made the TIFF file?" If NMXYZ was the curator of the original and it also published the TIFF file, I would consider it authentic; if SAO, on the other hand, created and published the TIFF file, I would not consider it authentic. I suppose what I am really proposing is to link "authentic" to "authoritative".
It may be helpful to point out that all of this deals with the question "What is authentic?". The flipside of this is: "How do we know something is authentic?" But that is a subject for another thread at some other time.
This is such a a great question. There has been much discussion in digital readers about "Does the concept of an original have meaning in the digital environment?" In Charles T. Cullen's essay, Authentication of Digital Objects:Lessons from a Historian's Research, he says "The issues stemming from authenticating digital objects are quite similar, and in some cases identical, to those relating to holographs or printed books. Everyone dealing with important material in any form should approach it with a bit of skepticism, but scholars especially need to question what it is they are using. In other words, they need to authenticate all documentation they use in the processes of learning and of creating new scholarship." This idea of the New Scholarship is something very exciting and a great topic for the Smithsonian to tackle.
Arnold, I think you've captured a very important aspect - that of documented governance.
I would like to extend it further and assert that the same rigor is necessary when the preserved edition, or in your case the derived data products, is in a different data format. An example is a PSD image (PhotoShop) migrated to a TIFF format. I contentd that the authenticity of the TIFF format stems from:
1 - the documented governance, i.e. the TIFF was created by the authorized custodian
2 - the migration used valid and tested methodology in selecting the preservation format
3 - appropriate quality control measures have verified that the newly formatted object exists solely of the original source code's content.
3a - Authentic content subsumes data appearance, layout, behavior, and relationships.
4 - full and complete documentation of all of the above is maintained in the appropriate place - either with the object or in a related database.
This approach provides a means to assert authenticity without sacrificing accessibility. Regardless, one still needs to hang onto the original source as well as the preserved object or derived data product (see Arnold Rots's comment).
I think we're talking about three types of authenticity here:
One is the issue of what Ricc calls governance - basically, that the digital object was generated/created by the authoritative body for that object (whether it was born digital or was created as a digital surrogate of a physical object).
The second issue might be called the experential or use authenticity -- does the digital object in its immediately accessible form provide an authentic representation of the digital or analog original. E.g. is a PDF printout of a website a reasonably authentic experience of visiting that site.
The third issue, which is more of a preservation issue, is how do we measure the integrity of the bits that are permanently storing the object. This is where the use of a checksum mechanism (e.g. SSH or md5sum) -- as mentioned by Arnold in his great metadata article -- can be incredibly useful.
I've posted a short list of articles relating to authenticity here
Post a Comment