In January 2006, an international conference was organised in Arezzo, Italy, to discuss the principles and purposes of the critical edition produced with the support of humanities computing tools and methods. With , editors Arianna Ciula and Francesco Stella have published a multi-lingual selection of the conference proceedings, bringing together ten papers in Italian, four in English and one in French.

The book comes with a CD-ROM, containing all papers as PDF files, presentation slides of a selection of published papers and, in addition, of unpublished papers by Kevin Kiernan (), Paul Spence and Harold Short (), and Arianna Ciula (). Furthermore, the CD-ROM provides additional material from Ferrarini's and Hagel's contributions. Although quite helpful in illustrating their papers, these add-ons are not accessible via the index file of the CD-ROM and are therefore a bit tricky to locate.

The volume is organised according to the presentation of the papers in the conference. They can also be categorised thematically: contributions with a theoretical approach (Robinson, Maggioni, L. Leonardi, Orlandi, Stella), papers introducing or discussing tools and guidelines for digital work (Hagel, Schreibman, Ferrarini, Fusi, Del Turco) and project descriptions (Poupeau, Feliziani, Del Corso, Boccini and D'Imperio, Cartocci). While most of the papers cover philology in particular medieval texts, the scope of the volume is extended by contributions on epigraphic texts (Fusi), papyrology (Del Corso) and watermarks (Cartocci).

The range of styles of presentation is too diverse to offer an overall assessment of them. While, for example, Robinson addresses the reader using the casual diction of his conference talk (if there is anyone in this room [...] [1], if we press this button here [...] [4]), others like Ferrarini and Del Turco support their statements with substantial bibliographical notes.

Three prefaces representing the co-hosts of the conference (Università di Siena, King's College London, Fondazione Ezio Franceschini Firenze) are followed by Francesco Stella's introduction in which he points out the goals of the conference and of the proceedings. He presents some advantages of digital scholarly editions in comparison to print editions: these points are certainly worth making but are relatively obvious and will not be discussed further here. Regarding the disadvantages of digital publication, suffice it to repeat his statement that the former are capable of providing everything that the latter provides except paper (viii). Stella does not hesitate to summarize the surplus value (viii ff) and ends with a fairly Italian analogy comparing digital and print to the Ferrari and the Fiat (xiii). However, the skepticism that exists towards digital editing was repeatedly expressed during the discussion at the conference as well as in several papers (Maggioni, L. Leonardi). Accordingly, Claudio Leonardi emphasizes in his the demand for completeness of digital editions and judges the question is future philology digital? as still being legitimate (ii).

In his illustrative essay , Peter Robinson discusses the capacities of electronic texts and concludes with a list of prerequisites which he considers the most urgent and most important for further promotion of electronic editions: these are collaborative tools and frameworks on the one hand and unhindered access to high-quality digital images on the other. In order to come to this conclusion, Robinson names six general aims of an edition and argues that five of them are better realised electronically than in printed form. He presents this theory with a case-study of Shaw's CD-ROM edition of Dante's , choosing this as an example of a very complex collation of text witnesses. Robinson manages to employ well chosen and illustrated examples in order to underline his theory, convincing the reader that all things are possible (11).

A disturbing picture of a future academic world is drawn by Giovanni Paolo Maggioni (). In an analogy to the Dying Earth genre classic by H.G.Wells, he utters his fear of scholars being divided into two classes. The first is the Eloi who work superficially on the user interface while in complete ignorance of the second class, the terrible Morlocks who underneath to keep the data and machines running. Maggioni’s concern arises from his own experiences while producing digital multi-textual editions (edizioni multitestuali) in order to present different stages of the genesis and transformation of a particular text, namely the by Jacobus de Varagine. Even though valuable results have been provided in digital formats, maintaining a digital edition and adapting it to ongoing technical progress can only be achieved by the editor himself and an IT elite and, thus, would exceed the normal budget of any academic institution, whereas a mere print edition is supposed to remain a stable and reliable point of reference.

Gautier Poupeau () describes the experiences of work at the École des Chartes on digital scholarly editions which are provided online and are freely accessible (<http://elec.enc.sorbonne.fr/>). According to Olivier Guyotjeannin, director of the edition of the Cartulaire blanc (<http://lemo.irht.cnrs.fr/42/mo42_12.htm>), these digital editions change nothing, that is, they do not have any methodological impact since they follow the same principles of printed scholarly editions. But at the same time, digital editions change everything in usability and availability. This change is modestly labelled as being evolutionary rather than revolutionary and is illustrated with several examples taken from the projects in question. Regarding the encoding of texts and auxiliary information, Poupeau emphasizes the importance of the XML data format and the TEI standard respectively. In order to allow enhanced research using the resources provided, the open source database management system eXist has been tested for creating user-friendly research interfaces.

Ombretta Feliziani () reports on the achievements of the natural born digital edition of the , an autograph manuscript of Boccaccio (<http://rmcisadu.let.uniroma1.it/boccaccio/>). This edition (EDIC – Edizione Diplomatica Interpretativa Codificata) is intended to be methodologically different from a traditional print one inasmuch as the transcription of the manuscript is not considered to be a preliminary stage of the critical text but rather the centre of the philological work itself. The importance of the manuscript legitimates the enormous effort that has been made to maintain the honesty of the text (l’onestà del testo, 40, cf. 37) and to encode in a meticulous way a large amount of detail dealing with linguistic, codicological, iconographic and, above all, palaeographical phenomena. In doing so, the TEI guidelines have been followed as far as possible but were found to be too generic. In consequence, further specifications and modifications have been made. The article is illustrated by a series of tables, among them a complete list of the details that have been recorded (letters, numbers, graphemes, abbreviations, and so on) along with a facsimile example of each.

The concern that there will be more informatics than philology in a future philology is expressed by Lino Leonardi (). To prevent this, Leonardi tries to take the discussion back to the principles of philology expressed in either one of the two main branches in the pre-digital tradition of scholarly editing, namely to the academic tradition of Bédier and the truth of the manuscript on the one hand, and to that of Lachmann and the truth of the author on the other. According to Leonardi, digital editions support the idea of the edition as an archive that virtually gathers all the relevant documents, texts and contextual material; but the digital nature of the edition has not contributed for nothing to the constitution of the critical text (69). Avoiding the editor’s decision between one reading and another, digital editions stop at the threshold where the crucial and distinguishing work of the philologist is supposed to begin. In conclusion, Leonardi claims that digital editions must not merely present material in the most complete and efficient way; rather digital philology must go beyond and consolidate a methodology and theory that benefits from the enormous capacity of digital media in order to justify the editor’s choice when establishing a text that is intended to be read (73).

is a paper by Stefan Hagel, author of a supportive tool widely used for creating critical editions in printed form, the Classical Text Editor (CTE). Despite observing that his customers' demand for digital output has been negligible (78), Hagel outlines how the CTE can be used both to generate a printer's copy (PDF) and simultaneously to prepare an electronic edition (TEI/XML and HTML). He illustrates the latter by some samples that can be accessed on the CD-ROM, although unfortunately it remains quite unclear to what extent the CTE created this output automatically. Hagel defines three prerequisites for optimal editorial software (80) and discusses the balance within them, in particular between designing a tool that is easy to use, on the one hand, and providing a wide range of capabilities that electronic texts could offer, on the other. For Hagel this is a contradiction. He regards his own compromise in response to this as a major sacrifice of the CTE approach (83) and concludes that it is better to have an imperfect electronic version to search in than to have nothing but the book (83), a statement that is understandable in his context. Whether his evaluation down to the most boring orthographic variation, if this is what the editor wants to spend her or his life on (83) really supports his argument, may be better assessed by scholars for whom these variations are an important aspect of their work.

Ontological distinctions between a digital and a print edition are made by Tito Orlandi in his short contribution on theory and practice of a digital edition (). His initial point of analysis is the observation that, until now, the guiding principles that have been developed for scholarly print editions have not changed for editions in a digital format; that editors still apply the same criteria when producing one text that is intended to represent what is meant to be the most genuine reconstruction of an original (86). However, any theory of digital editing must be aware of four essential distinctions between this and a print one:

A digital edition is first of all an electronic resource, not a (re-)presentation of text. Distribution and usage of the text is based on hidden digital processes (producing two dimensional text representations from a one dimensional digital resource). The digital editor’s work is focused on the resource (file) that contains binary elements (bits); the process of giving significant value to a certain group of bits must be explicitly – in completeness and accuracy – operationalized by the editor. The digital editor’s consciousness of the text as a synthesis of various textual aspects and notions (text as system) is fundamental.

is substantial article in which Susan Schreibman introduces the Versioning Machine (VM) in the context of the historical-critical theory of editing: as a toolkit to perform text out of various witnesses. Schreibman briefly describes the principles that lie behind the first two releases of the software (version 3.2 has been released since), and with that she presents the fundamental encoding strategies based on TEI/XML: parallel segmentation and individual encoding of each witness. In the second part of her paper, Schreibman looks deeper into the textual theory that motivated the development of VM: employing deformance (Samuels and McGann 1999) to create an illusion of text in print (97) rather than providing a new display paradigm. VM is a tool that is the stage on which [text encoding] performance is enacted allowing an editor to see a work in a fluid state (99) and therefore serves not only as an interface for readers but also as a scholarly tool for editors. Reflecting Tanselle’s meaning of literal text and his distinct classification of scholarly editing, Schreibman concludes that this model has to be modified to use deformance to enable machine processing and analysis (101) of the textual data to allow several theories of the text to be embedded within the encoding.

The importance of very accurate transcriptions for digital editions has been emphasized by Feliziani (regarding the honesty of a manuscript) and by Orlandi (regarding the need for a meticulous palaeographic analysis as basis for operationalizing any processing work). Both philological approaches are manifest in the instructive and stimulating article by Edoardo Ferrarini () who focuses on the question of how a digital transcription should be made (104 ff) and concludes by giving reasons when (§2, 114 ff) and why (115 ff) it must be done. A scholarly transcription, according to Ferrarini, must be documented, portable among platforms, exhaustive and normalized; transcription starts by using ASCII code and ends by encoding in XML. A series of illustrative and instructive slides as well as a complete TEI encoded transcription of Heu heu vita mundi from MS 325 of the Biblioteca Città di Arrezzo (XIV century) is available on the CD-ROM, as is the DTD schema.

The article by Daniele Fusi on digital epigraphics () is the longest and most comprehensive contribution to the volume. Fusi exhaustively elaborates all aspects of the digitization and publication of epigraphic texts in both Latin and Greek. Almost all-embracing, Fusi’s project is born, above all, in the perspective of a philologist and of a computer scientist, and it is fundamentally characterized by the decisive distinction between content and its presentation, based on the crucial notion of transformation: any content, highly structured from the semantic point of view, can take potentially infinite forms, with regard to the selection and order of the material as well as to the electronic format (122). Accordingly, the structure of his argument (which is clearly laid out and illustrated through presentation slides included in the CD-ROM) is as follows: § I General aims of digital editions: (1) media and public, (2) levels of specialization, (3) sustainability, (4) semanticizing; § II Realization and expansion: (1) characteristics of epigraphic documents, (2) orthography and editorial practice, (3) levels of semanticizing, (4) mark-up, (5) structure and transformation of XML databases, (6) text layering, (7) delivery and publication.

The last paragraphs (§ III) are dedicated to perspectives on digitally editing epigraphic documents, and the author envisages a series of promising applications of databases and digital text corpora. In glaring contrast to the highly instructive article, the bibliography is a major disappointment comprising as it contains only four items: three by Fusi himself, and the fourth dating from 1980. Few annotations are provided and there is no reference to current academic debates or related studies, which leaves the impression that Fusi is the only scholar working on digital editions, at least regarding epigraphics.

In contrast, the Papiri della Società Italiana (PSI) is an online project discussed in Lucio del Corso's article (), where it is explicitly presented as a collaboration between archaeologists, classicists, IT specialists and web designers, as well as between several institutions (listed below). This short article gives first a brief overview of the short history of creating systematic inventories of papyri that began with the in 1901, followed by the introduction of databases from 1982 onwards by the Duke Data Bank of Documentary Papyri (DDBDP, now accessible under <http://www.perseus.tufts.edu/hopper/>), the Leuven Database of Ancient Books (LDAB, <http://ldab.arts.kuleuven.be/>) and the Advanced Papyrological Information System (APIS, <http://www.columbia.edu/cu/lweb/projects/digital/apis/index.html>). Against this background, the PSI online project (<http://psi-online.cea.unicas.it/>) is being developed at the University of Cassino to build up a virtual reconstruction of the papyrus collection of the Italian Society for the research on Greek and Latin papyri in Egypt (henceforth Istituto Papirologico Vitelli); this collection comprises some 1500 items that are now dispersed between institutions in Florence and Cairo and in the Library of Alexandria. This virtual space aims to reproduce the functions of a real and proper ideal library (168) and provides scheduled descriptions as well as digital reproductions of the items. However, Del Corso modestly states that the project does not intend to substitute traditional (i.e. print) inventories but to be a mere complementary tool (172). In order to aid future usability and sustainability, the resource is organised as an XML database and as far as possible the most reliable open source software is used for its implementation. Del Corso’s greatest concern, however, is the durability of databases in the humanities in general: Within ten or twenty years, quale sarà il loro destino? (173)

The SISMEL Special Library project presented by Fabiana Boccini and Francesca Sara D'Imperio () collects in one database all the available information on 8,412 manuscript witnesses of the writings of Gregory the Great. The results will be published as a hybrid edition (print and CD-ROM) under the title . The project is clearly a fundamental and important contribution to research in the writings and tradition of the fathers and Pope Gregory in particular, and it is probable that this vast amount of information will be available to researchers digitally on CD-ROM. Unfortunately, use of this resource by a wider public is restricted by the fact that neither online access nor a connection with similar resources is intended, even though the project itself profits from collecting data from online resources.

In his article on editing Old English texts (), Roberto Rosselli del Turco presents a series of encoding problems and solutions that arose out of the Digital Vercelli Book Project (<http://islp.di.unipi.it/bifrost/vbd/>). The focus of this thorough contribution is on character encoding and metrical markup. A series of examples is given in order to discuss several solutions to characteristic problems. The argument is supported by an instructive English (!) slide presentation on the CD-ROM.

The short article by Cristiana Cartocci () reflects on the importance of watermarks for dating texts from the fifteenth century onwards. The enormous value of digital inventories is evident particularly in their capacity for reproduction and visualization of watermarks as well as in the systematic organization of related information through a database. In order to make such inventories usable for research, Cartocci suggests establishing hierarchical search criteria and cursorily refers to the International Standard for the Registration of Papers with or without Watermarks and to the experiments that have been carried out in the field of Content Based Image Retrieval. The article is followed by a list of relevant digitisation projects dealing with watermarks, inadvertently omitting Piccard online (<http://www.piccard-online.de/>). The WIES project (Watermarks in Incunabula printed in España: <http://www.ksbm.oeaw.ac.at/wies/>) went online in January 2007 after this volume was published.

The final words in this volume are left to Francesco Stella. In , he draws a positive picture of existing resources for digital philology of medieval texts in general but observes that in the field of medieval Latin and Humanistic philology no editions that can aspire to these high standards have yet been produced (225). This statement about the present situation is followed by some of his own thoughts on existing editorial theory such as new philology, and Stella illustrates their implementation in electronic form with a number of examples. He describes four domains of technical innovations and methodological innovations (232): quantity of data, relationability, interoperability and multimediality. Suprisingly, he leaves out one domain which he describes later in his paper and which perhaps best characterises the digital edition: user interaction. Closing this theoretical part of his paper, Stella classifies digital editions in three categories: hypertext editions, codified editions and database editions. The final two sections of his contribution are devoted to the presentation of , an international joint project to facilitate research on early medieval poetry and music. Employing this interesting case-study, Stella not only underlines his editorial theory but also outlines the implementation of the project as an interdisciplinary database edition and presents its capabilities in an illustrative way. This article, however, deserved more editorial care than it received.