© Ségolène M. Tarte, 2010. Creative Commons Attribution-NonCommercial licence
Born digital
Citations from the text of this article should be by paragraph number (found on the ID attribute of the p element).
The second of the series Schriften des Instituts für Dokumentologie und Editorik, this volume, entitled
Website: http://www.i-d-e.de/ ― last checked: September 5, 2010
The volume is also available online: http://www.i-d-e.de/schriften-2/kodikologie-und-palaographie-im-digitalen-zeitalter
The “Codicology” section discusses the constitution,
publication, and (to a lesser extent) maintenance of both online/digital
catalogues and the research environments accessing such catalogues; all
themes that yield questions of standardizations, taxonomies, and ontologies
(Bernardi et al., 2009; Cartelli et al., 2009; Speer, 2009; Stinson, 2009;
Kalning and Zimmermann, 2009; Uhlíř and Knoll, 2009; Deckers et al., 2009;
Wolf, 2009). Each paper presents the authors' own approach to such
questions, and discusses their procedures of digitizing, encoding, and
making available online scholarly material, whether representations of
codices―including images, descriptions and transcripts―or research output
emanating from the study of codices. Also discussed are their choices of
advanced content management systems, software architecture, and data
storage; their strategy to organize metadata; and their methods for
harvesting data, whether collecting it for inclusion in a catalogue or
querying a catalogue itself. Most of the papers in this section have an
active online counterpart that implements the paper's detailed
principles. Save for the
papers that are more prospective papers, and for which at the time
of writing this review (September 5, 2010) I could find no such
publicly accessible active website.
The papers in the “Palaeography” section can be split in three groups: those that concentrate on the didactics of palaeography (Kamp, 2009; Cartelli and Palma, 2009; Muir, 2009); those that present advanced imaging and/or computing techniques to study scripts (Cayless, 2009; Shiel et al., 2009; Fusi, 2009; Tomasi and Tomasi, 2009; Gurrado, 2009); and those that (reflexively) consider palaeographical methodologies and their relationship(s) to digital tools (Ciula, 2009; Stansbury, 2009; Hofmeister et al., 2009; Aussems and Brink, 2009; Stokes, 2009). The didactically oriented papers explore the impact of digital tools on teaching: how new methods of teaching can be tested, especially through the accessibility of material such as digitized collections of letter shapes, and how these methods compare with more traditional ones (lookup tables); how digital tools are likely to impact the creation of knowledge; and how additional teaching material can be produced, e.g. ductus animations, videos of scribal hands in action. The next group of papers present advanced computer technologies and how they can be used to reveal the script in a document. They span imaging and image processing techniques: hyperspectral imaging followed by spectral and component analyses of the main directions of variations is one method; another is to develop geometrical shape descriptions of the script via Scalable Vector Graphics (SVG) that can be formatted and integrated into XML. Intelligent Character Recognition (ICR) is also presented as a kind of intelligent Optical Character Recognition (OCR). Finally, neural networks are used to train systems to perform pattern recognition whilst reducing the search space by injecting in the Support Vector Machine (SVM) some theoretical expert knowledge about letter shapes. In the third group of papers, the authors take a reflexive view on palaeography and, beyond developing digital tools that support palaeography, they try to evaluate how those tools influence palaeographical practice.
One marked trend and aspiration amongst these papers is a search for either objectivity, or at least some kind of intersubjective commonality in palaeographical studies, often via statistical tools. The section (and the volume) are concluded by an intelligent last paper (Stokes, 2009) that expands on those subjects and very appropriately incites (digital?) scholars to make sure that in digital codicology and palaeography, like in the digital humanities, an essential balance be found between the digital and the humanities components; so that the humanities aspects of the research shine through without being overshadowed by the digital aspects, and so that the use of ICT tools for the advances of Humanities research is clearly promoted, advocated, published and disseminated.
I found this volume very enlightening in that it covers a wide range of subjects and yet, thanks to elegant and clever editing work, it skillfully guided me through the varied landscape of digital tools and digital research in codicology and palaeography. Indeed, I often found myself asking a question at the end of a paper only to find it addressed in the very next paper. This occurred in particular when I started formulating to myself questions about the transformative effect of digital tools on actual practices. Straight away, these questions were met by an elaborate and detailed reflexive answer in the context of codicological descriptions (Stinson, 2009). The author identifies three major changes in the constitution and usage of codicological descriptions: an evolution in purpose; a shift from a one-to-one (description to codex) to a one-to-many (description to codex and images of codex) relationship between the description and its subject; and a drifting away from the linear book form of correspondence between two texts (description and codex) as de-linearization and asymmetry are advantageously injected by marking up the descriptions. After which my new question was: “So, in practice, how are these digital tools used, really?” And again, the very next paper (Kalning and Zimmermann, 2009) provided me with an answer, presenting clearly the possible uses of digital catalogues, and explaining how they enable optimization of the work done in the presence of the actual physical object by allowing so much to be done remotely with its digital avatar.
The few minor regrets I might have about the
“Codicology” section have to do with: the rarity of the mention and
availability of links between transcripts and the digital avatar(s) of
codices; As a side remark, I
use here on purpose the expression “digital avatar” rather than
“surrogate” or “virtual book”. While “surrogate” is a perfectly
acceptable term, I’m slightly concerned about the use of “virtual
book”, as the word “book” carries with it a string of implicit
expectations that are not quite moderated by the use of the
“virtual” qualifier, and that might yield to think that “virtual
book” means a digital version of a book on which pages can be turned
―a thought I particularly dislike as it usually involves using a
template “virtual book” that does not attempt to render the
materiality of the codex apart from its page-turning property (which
itself might even be limited!). The advantages of the term “avatar”
is that it implicitly conveys that it is one of the many possible
ways to represent an object; it is moreover a term that has already
been widely used in the digital context (to the point that in
everyday language “avatar” is commonly associated with the virtual
world).
The “Palaeography” section is somewhat less unified than the “Codicology” section, mostly due to the wide-ranging nature of the subjects broached by the papers. All papers raise interesting points and present exciting and promising technology. It is worth noting in this section the presence of two papers that have a more commercial tone and present products (Muir, 2009; Tomasi and Tomasi, 2009), which goes to prove, if needed be, that there is a demand in the Digital Humanities for elaborate technological tools.
One point I would like to reinforce as a side note to the papers that deal specifically with image processing (Cayless, 2009; Gurrado, 2009; Tomasi and Tomasi, 2009; Shiel et al., 2009) is that all the algorithms and methods presented require input images that have undergone one form or another of (pre-)segmentation (variably meaning image enhancement or binarization, depending on the authors and conventions). Segmentation is an arduous task, mostly because it poses one of those typically circular problems, where discriminating text, script, or other features from a background requires some kind of prior knowledge of the properties specific to those features and/or background. Segmentation is thus a non-trivial task, and has no universal solution. It requires taking expert knowledge into account either via machine learning, or directly in the segmentation model that is applied, and it is vastly affected by both the quality of the images and the physical state of the document. It also is a processing stage that conditions the downstream results of further processing. Often, the underlying assumption is that segmentation has been performed on the images before text characterization and script study can be carried out.
Finally, one theme that I was very glad to find present and that shone like a more or less salient watermark throughout the volume is that of the bidirectional influences between practices in codicology and palaeography, and digital tools. The development and adoption of digital tools open up a wide range of possibilities: sharing of resources, virtual reunion of dispersed corpora, remote scholarly collaborations, advanced image processing, elaborate statistical study of scribal hands. But not only do these digital means enable scholars to ask new research questions, to access and process complex and large amounts of data, they also contribute to the evolution of the discipline, in that the practices of codicology and palaeography are changing with their use. It is at this crossroads of the many disciplines involved that digital technology and the fields of palaeography and codicology both enrich and transform each other. Traditional codicological and palaeographical methodologies are both informing the development of tools and being modified by their use. Teaching palaeography becomes a different activity with digital tools; communicating research outcomes now also involves the notion of reproducibility of results; classification of scripts becomes more elaborate due to the large amount of accessible data, thus potentially transforming the taxonomies and descriptive criteria of scripts. This theme is covered by both the introduction and the conclusion, and reveals the importance of the study of what types of expertise are developed in the Arts and Humanities, of how expertise can be supported by digital tools, and of how expertise is naturally also impacted by those tools. As often in such volumes, I thought that the Introduction (Vogeler, 2009) was easier to read after having read the full volume. It is well structured and regroups the papers into different pools than those chosen by the editors, which is always both instructive and interesting. In general, to return to the idea of expertise and the transformative effects of technology on expertise, I would have wished for more contextualizing stories than were actually present. By “contextualizing story”, I mean a narrative or an illustration of how humanities research results are produced with the assistance of a digital tool, and to what extent the research methodology was modified by the tool. Examples of papers that achieved doing this masterfully are those by Speer, Stinson, Stansbury, and Aussems and Brink, 2009.
To conclude, I think the editors of the volume were able
to draw a good picture of the various research directions undertaken in the
fields of digital codicology and palaeography. Despite the fact that the
volume is in four languages, which made it somewhat arduous to read as
French, Italian, German and English have very different ways of structuring
both sentences and arguments, Multilingual reading is an interesting mental gymnastics that is
both inherent to the Humanities and understandable from an editorial
point of view. Yet it is undeniably a complicating factor for the
reader ―and also likely for the editors?