Digital recovery of removed texts
Photographic methods have been applied to recover documents since the birth of photography. Document photography had already become widespread by the late nineteenth century. Some documents created challenges to photographers in their attempts to produce legible results. The development of early photography coincided with the alternative chemical methods that were applied to palimpsests. At the end of the nineteenth century, Ernst Pringsheim and Otto Gradenwitz developed a method which was capabale of separating the upper script from the lower one, often already treated with chemical reagents. The virtual separation of the two scripts required a precise photographic operation. Photographers first processed two negative plates, an overexposed one and an underexposed one, which were perfectly congruent in a geometrical way. Each produced an image of the lower script with a different intensity. Then they used a positive plate made from the underexposed negative plate, using it as a mask, in order to process a positive reproduction without the upper script [cf. Pringsheim – Gradenwitz, Photographische Rekonstruktion von Palimpsesten].
In the context of palimpsests, the use of ultraviolet light is the most widespread method, although it was first used to this purpose only in 1914 by Father Raphael Kögel, OSB, who worked in the Palimpsest-Institut der Erzabtei Beuron (Germany) where the series Spicilegium Palimpsestorum was launched in 1912. With Knight’s expert technical summary, “he [Kögel] explained how ultraviolet radiation from an electric arc or a mercury vapor lamp could be used to excite fluorescence in parchment, but the fluorescence would be blocked (quenched) where the ink had originally been. A photograph of the visible fluorescence could then be taken, using filters to exclude the invisible ultra-violet which would obscure the image” [Kögel, Die Palimpsestphotographie; the citation is from B. Knight, Father Kögel and the ultra-violet examination of manuscripts, 24 March 2014]. This method increased the contrast between the two scripts, making it easier to detect and read each one. In addition, if combined with Pringsheim’s and Gradenwitz’ method, the technique allowed for the making of a more legible positive image of the lower scripts. Even today, scholars use the “Wood lamp”, named after the American physicist and optician, Robert W. Wood (1868–1955), or UV lamps with LED to read palimpsests. These lamps emit long-wave (UV-A) lights which generate a fluorescent response on the traces of iron ink of the removed scripts, ideally in a dark environment.
In the context of photography and palimpsests, the idea of applying document photography to conservation is important. The condition of old, often damaged documents, can change fast. Texts that are legible now can very soon become illegible; their state of conservation can deteriorate to the point of becoming inaccessible for consultation and scholarly use. The frequent use of old documents so important for many scholars is inconvenient for the conservation of these documents. It was considered feasible that photography be able to offer a remedy. An international project was launched in 1895 to encourage the reproduction and international distribution of photographic copies of important manuscripts. Franz Ehrle, the Vatican prefect, was especially sensitive to the conservation of documents and had the idea to apply this principle to palimpsests as well. The first facsimile edition of an entire manuscript made up of recycled parchment sheets was made for a Vatican manuscript originating from Bobbio (Vat. lat. 5750). The photographic facsimile edition was made in Milan at the publishing house of Hoepli (Milan 1908) [cf. K. Krumbacher, Die Photographie im Dienste der Geisteswissenschaften, Leipzig 1906, pp. 7-20].
The first volume of Codices Latini Antiquiores was edited by Elias Avery Lowe in 1934, which collected the Latin scripts prior to the ninth century from the Vatican Library and was dedicated to Cardinal Franz Ehrle. The volume included many details of palimpsests wherein the lower script often had to be presented as visible and legible. Very likely glass plates were used for these images, even though they do not survive in the Photo Archive of the Vatican Library. Since the 1930s, UV light was also applied to palimpsests at the Vatican Library already in the very early stages of its own photo laboratory.
Digital images opened new horizons for palimpsest research. Today, multispectral photography offers the best solution for enhancing the difference between faded inks and their surrounding areas. To do so, it utilizes multiple numbers of images taken from the same page and captured in the same position, while being illuminated by a pre-set succession of fixed wavelengths of light (various waves of natural, ultraviolet, infrared, and other types of lights). These images capture the different reactions of the parchment to different light waves, which can be compared fruitfully. In the process, mathematic algorithms are manipulated in such a way as to bring these differences from the invisible spectrum to the spectrum that is visible to human eyes. Multispectral photography, however, requires the specialized knowledge of image scientists who have to work together with philologists who are able to read these texts. Only in the event of collaboration between the scholars from both fields will the texts finally be able to come to light; otherwise, each sector will not be able to come to the best results [cf. Easton – Kelbe, Statistical Processing of Spectral Imagery].
Multispectral imaging produces raw data which cannot be interpreted directly by the final users, textual scholars or conservators. Various photographic projects for palimpsests that use this principle seek to publish the raw data together with a more-or less legible version that can be interpreted by their own standards. The Vatican Library normally publishes two versions of each palimpsest page, one captured in natural light and another one captured with UV fluorescence. Both can be interpreted on its own without being combined with the other and the image enhancing tools allow scholars to manipulate these images to satisfy individual research tasks. The digital recovery of the removed texts has multiple advantages. First of all, it is a non-intrusive method and does not involve long exposure to UV or infrared lights. Second, the results can be manipulated without the risk of damaging the source file which functions as a point of departure for digital enquiry. Third, it is very easy to distribute and circulate such images and their processed variants among scholars even in multiple copies in exactly identical forms. However, the maintenance of huge repositories of digital images creates challenges for future generations.