Digitisation describes the process of taking an analogue object and creating a reproduction, copy or backup copy1 of it that can be read on a computer, while digitalisation involves using the result as the afterlife, long-term preservation and public accessibility of a digital record and the reuse of digitised materials. But even the basic terms providing conceptual framework and meaning to the digitisation process of artistic heritage, are already problematic.

The main aim of digitising in the art museum is to produce reproductions of pieces in the collections of sufficient resolution and with a colour rendering index which would correlate as closely as possible to the original piece within a scale visible to the human eye, which would allow for the pieces to be reproduced in various sizes and on different materials. Digitising is primarily part of the work of a contemporary museum that supports the general activities of the museum. Museum publications, the opportunity to buy reproductions of works in the museum, museum webpages and databases introducing the collections are the most important outputs of the digitising process and help spread awareness of the museum’s collections.

The relationship between art and the terms copy or backup copy used widely in the field of digitising as the general aim of memory institutions is a little more problematic. These notions describe mainly the activities of libraries and archives in the digital age. A library digitises a page from a book and saves the information it contains in digital format; the same process also takes place with documents in an archive. The appearance or material characteristics of the page of an archival document or book are relatively unimportant. One can also consider the production of a backup copy of a photograph through digitisation. A photograph on paper is already essentially a copy and the production of a digital copy from an analogue copy are not contradictory activities. There are also certainly exceptions among the aforementioned types of cultural records. A book may be copiously illustrated with unique works, the base- material of an archival document may be as important a source of information as the information it holds and the original photographs as negatives may be just singular examples of exceptions.

How should we understand the nature of a backup copy in the case of art? A backup copy is designated a “copy, which has been produced to preserve and make available the information in case the original is destroyed”2 and according to a broader definition “the high quality of a backup copy and its safest possible long-term preservation (digital archiving) allows for a digital copy as close to the original as possible to be produced from the backup copy in the case of the destruction of the original”3. Here the terms ‘destruction of the original’ and ‘as close to the original as possible’ are important.

The digital recordings that adhere to the current digitising standards only provide information concerning the visual appearance of the piece in the case of the destruction of the artwork. A large proportion of the information, which may be important in terms of the meaning of the artwork, would be lost in the case of the destruction of the piece. In the case of books, archival documents or photographs, we can consider the production of backup copies because the information contained within the document can be copied and recorded. In the case of an artwork, the term backup copy is misleading and does not consider the possibilities and limitations of current technology. With artworks, it would be more precise to talk only of the production of digital reproductions. Technically, the digital reproduction still has to adhere to the requirements of archival backup copies—using tiff file formats, a resolution of at least 300 dpi, colour correction charts and scales—but still the digitising process can only record the appearance of the piece. The materiality of the piece is lost in the process. Digitising artworks according to current means and understandings is inevitably restrained by the shallowness of the production of digital reproductions.

There are certainly exceptions among artworks, which allow for the use of current widespread technical tools and practices for the production of reproductions which are very close to the originals, an example of this is printmaking. But here too the results of the digitising process may differ greatly (see illustrations).

The technical manuals that advise memory institutions on achieving the standardised quality for digitising are also coming up against the intricacies of the process of digitising artworks, especially paintings. For example, the Technical Guidelines for Digitizing Cultural Heritage Materials (2016) by the Federal Agencies Digital Guidelines Initiative (FADGI) claims that colour maintenance is problematic with paintings and other two-dimensional artworks because contemporary digital technology does not allow all colours to be correctly recorded and adds the requirement that all paintings should be digitised as three-dimensional objects.4 This is due to the fact that the relief of the surface of the painting may convey information and be an inseparable part of the expression of the piece.

How would it be possible to overcome the immateriality of artworks brought on by the digital age and achieve a backup copy of the digitised piece? Technically? According to a technological-positivist understanding and allowing for a flight of fancy, we can imagine how a piece could be recorded in all its finest details, including unconventional base materials in the structure and all the characteristics of the techniques used by the artist, in addition to the materials used in the creation of the piece and layers which would remain invisible upon normal observation. In the case of a few singular projects, digitisation which considers numerous aspects of a piece has been attempted, but the aims of these projects have been to record information pertinent to conservation using imaging technology,5 not to produce a model of the piece, according to which it could later be restored as precisely as possible.

With enough information it is possible to create a three-dimensional model of the piece, which along with the necessary metadata about the materials and techniques of the piece could be 3D-printed. But this approach would also require a new type of technology. In the case of producing a backup copy of a painting, we could consider the capability of a 3D-printer to reproduce the support material and fill it layer-by-layer with paint of the authentic consistency. With current technology, this is only a faint vision of the future, which poses the question that museums are already coming up against while digitising their collections today even more pointedly: “What is the artwork in the age of digital reproduction?”6

Even the most precise model of an artwork would not be its backup copy. Without attempting to designate what an artwork is, the formal characteristics it should adhere to or how it should come about, I will focus in the following solely on the meaning of originality. An artwork is usually original and authentic. Even if it were possible to create a model of the piece that could be described as a backup copy, the process would lose the originality of the piece and after the destruction of the original, only an intricate reproduction of the artwork could be restored using machines, which could be viewed as a copy of the piece, if the materials and techniques used were close or identical to those of the original artwork. This doesn’t mean that the possibilities offered by technology in the digital age should not be implemented, although the greater aim of digitising artworks in a way that would allow them to be restored based on the data recorded in the case of the original piece being destroyed remains unachievable. Quite the contrary, the means offered by technology for recording artworks should be implemented much more broadly.

Digitisation as a technological process is only the first step in a chain of actions, that have changed the work of memory institutions in the digital age and which will also alter how art history is studied and written in the future.

The files recorded through digitisation require long-term preservation because the need to reproduce pieces inside as well as outside museums occurs repeatedly. The files also need to be easy to find, there is no point spending large amounts of time searching for the necessary piece of information from disorganised piles of data. In the digital age the “symbolic forms”7 for organising data that describe the world are databases. If a database has been implemented by a memory institution, there is a temptation as well as a need to make the information about the collections of the museum more widely accessible over the internet.

Electronic catalogues with images of the pieces are publicly available in most of the world’s memory institutions by now, be they solutions created by the museums themselves or broader developments that combine a region or whole country like the Museums Information System (MuIS) and its public user interface Museums Public Portal. In one sense, internet technology provides art historians a more effective way to study art, actualising on a new level the Mnemosyne8 project started by Aby Warburg about the pictures that affected (Western) civilisation and their use in various eras.

Digitalisation has created the prerequisites for a more thorough study of art, but technologically there are still many limitations keeping us from digital art history. For digital art history we needed to teach computers to see and to look, important advances have been made in this field, but computers still have difficulties understanding pictures and analysing them. Implementing developments in AI can also improve these capabilities and reach a general visual literacy.9

Digitisation is a process that relies on technological developments. Active digitisation started in memory institutions in the 1990s, with the broadening of technological means it started to be conceptualised and standardised. The standard technical requirements for digitisation and the aim of creating digital backup copies of cultural heritage were best met by the types of objects, which could be reproduced in large quantities also without the use of digital technology: books, photographs and documents. Art heritage, foremost with the example of painting, continues to be a contradictory and problematic field in terms of digitisation, which requires the implementation of the most contemporary imaging technology as well as the use of a new kind of technology. This would allow for the creation of visual studies, in which technology supports the study of art and studies of art help technology achieve better results.

1 See Raamatukogusõnastik https://termin. nlib.ee/view/6242 [checked 15.01.2020]

2 Kurmo Konsa. Bitid purki. Teabe säilitamine digiühiskonnas. Tartu. Tartu Art College, 2018, p 159.

3 Mari Siiner, Martin Sermat. Museaalide digiteerimine.—Renovatum Anno 2015, https://renovatum.ee/autor/museaalide- digiteerimine [checked 15.01.2020]

4 Technical Guidelines for Digitizing Cultural Heritage Materials (2016), p 47, http://www.digitizationguidelines.gov/ guidelines/FADGI%20Federal%20%20 Agencies%20Digital%20Guidelines%20 Initiative-2016%20Final_rev1.pdf [checked 20.01.2020]. Vahur Puik has started rewriting the FADGI guidelines in Estonian and conducted symposiums on standardized digitization based on the guidelines and workshops mostly with a focus on digitizing photographic heritage.

5 Andres Uueni, Hembo Pagi, Hilkka Hiiop. Getting in between the Paint Layers by Way of Natural Sciences. The Use of Imaging Methods in Documenting Heritage.—Rode Altarpiece in Close-Up, Art Museum of Estonia, 2016, pp. 69−90.

6 Paraphrase of the title of Walter Benjamin’s cultural-critical essay “The Work of Art in the Age of Mechanical Reproduction”.

7 Lev Manovich. Database as Symbolic Form.—Museum in a Digital Age. Ed.
R. Parry. London and New York: Routledge, 2010, pp. 64–71, here: p. 65

8 Leonardo Impett and Franco Moretti have studied the Mnemosyne project with digital tools. Totentanz. Operationalizing Aby Warburg’s Pathosformeln.—Pamphlets of the Stanford Literary Lab, 2017, https://litlab. stanford.edu/LiteraryLabPamphlet16.pdf [checked 31.01.2020]

9 Maria D. Avgerinou and Rune Pettersson provide an overview of the themes and problems of visual literacy. Toward a Cohesive Theory of Visual Literacy.—Journal of Visual Literacy, 2011, Vol. 30, no 2, pp. 1–19, https://www.researchgate.net/publication/ 267988880 [checked 10.10.2019]