Abstract Digital collections of historical objects are increasingly the first interface when setting up new academic enquiries in the Humanities. This influences how research questions are formed. The size of these collections is posited to grow even more as the costs of digitization decrease and access widens. Framing digital collections and their contents correctly is therefore of great importance. However, in many cases the conceptualization of digitized collections remains in practice tied to the mistaken notion that digital copies are surrogates for the analogue originals they represent, with serious implications for how these items are organized. This article builds from earlier seminal contributions to argue that digitized objects are a different entity altogether, neither copy nor surrogate. This difference can be captured more faithfully by the notion of ‘digital translation’: digital objects are a selective description of the physical original; this description is translated into digital language and written in symbols bound by rules. This structured information can be read algorithmically. From this definition, we can derive a taxonomy based on the depth of digitization, that is, how many layers of description exist about the object. These layers often include shape, colour, and recorded metadata. Adopting a by-depth taxonomy has ramifications for existing digital collections: if digital objects are individualized translations of an original, their digital collections are collections of commentaries in digital language reflecting specific ways of seeing the original. If digitization depth differentiates these sources, our archival and dissemination strategies must respond to this reality.
Giovanni Maria Pala (Mon,) studied this question.