Report on the Digital Humanities 2010 conference (DH2010) General Information Digital Humanities 2010, the annual international conference of the Alliance of Digital Humanities Organisations (ADHO), was hosted by the Centre for Computing in the Humanities (CCH) and the Centre for e-Research (CeRch) at King's College London from 7th to 10th July. The conference, whose theme was “Cultural expression, old and new”, featured opening and closing plenary talks given by Charles J. Henry and Melissa Terras respectively, the Busa Award and Lecture given by Joseph Raben, three days of academic programme with four strands each, and four poster sessions, as well as a complementary social programme. The highlights of the latter included a conference dinner at the Great Hall at Lincoln's Inn, and excursions to Hampton Court, to the Tate Britain and Tate Modern museums, and to the Globe Theatre and Exhibition. The conference website is at <http://dh2010.cch.kcl.ac.uk/> from which the conference abstracts can be downloaded. Discussions could also be followed on Twitter at <http://twitter.com/search?q=%23DH2010>. The three-day programme of parallel sessions began on Thursday morning, the following is a summary of the panels and paper sessions attended. Session one The first session attended was a panel session, entitled “Building the Humanities Lab: Scholarly Practices in Virtual Research Environments” and moderated by Charles van den Heuvel and Smiljana Antonijevic of the Royal Netherlands Academy for Arts and Sciences, in which six panelists presented their work on virtual research environments: 1. David Bodenhamer, Director of the Polis Center at Indiana University Purdue University Indianapolis (IUPUI), introduced his centre <http://www.polis.iupui.edu/>, a multi-disciplinary research unit founded in 1989. The centre adopted GIS as its technology of choice in 1992 and has six focus areas: geoinformatics, community informatics, health geographics, spatial informatics, humanities GIS, and community collaborations. The centre is a matrix organisation with 25FTE and employs a central strong project management system and industry-standard practices. The centre is entirely projects-based, it runs between 50-100 projects, and self-funded with a diverse funding base that includes consultancy work. One of its key focus area is the Virtual Center for Spatial Humanities, a Humanities GIS with book and journal publications in preparation. 2. Fotis Jannidis of the University of Würzburg presented on the BMBF-funded Textgrid project <http://www.textgrid.de/en.html>. The project is now entering its second phase, there are already a number of resources and tools at researchers' disposals for retrieval and re-use. The integrated tools in the user environment include data capture, editing, analysis, workflows, annotation, and publication. The front-end is the Textgridlab software, the services on the service layer include among others tokenize, lemmatize, sort, transform, and collate, the backend is provided by the Grid. Textgrid is based on lucene and eXist. Textgrid is now a library-based project, academics serve 1 3. 4. 5. 6. as advisers. Some of the challenges include: providing robust, production-ready tools and sustainability for both services and data; need for structure and an architecture, but openness is a necessity, and development should be collaborative; availability of an integrated user interface, but user-defined processing pipelines are needed. Bethany Nowviskie of the University of Virginia Scholars' Lab presented on the NINES VRE <http://www.nines.org/>. NINES is a collaborative research environment for nineteenth-century scholarship, which models an alternative to traditional peerreviewed publishing. Peer-review is the core function of NINES, currently there are 670,000 digital objects available in an entirely scholar-driven environment. While the social functions in the community have been less successful, work concentrates on resource discovery rather than on an VRE as an online environment that needs to be studied to be used. Essentially, NINES establishes federated peer-review and resource discovery of quality online materials. Geoffrey Rockwell of the University of Alberta, Canada, gave a presentation on the TAPoR Portal <http://portal.tapor.ca/>, which is being put on a Grid-infrastructure. The Portal is a Web-services broker for the TAPoR tools, the second-generation tools known as Voyeur, formerly TAPoRware tools. There is also an accompanying Web of tutorials, documentation, and a wiki. Some challenges remain: building a portal is tricky, the complexity of learning about and using it is prohibitive; tools pose a challenge as well, as humanists think of texts not tools, therefore tools need to be plugged into texts; lastly, community building is challenging, methods and research practices are important aspects. There are however benefits too: Web-based tools are easier to develop and maintain, and adaptable to different environments; it is possible to offer APIs to the community. Funding is always an issue: TAPoR seeks sustainability through collaboration with the HPC infrastructure; libraries are considered to be permanent homes for projects like this. Mike Priddy of King's College London gave a presentation on the EU-funded DARIAH project <http://www.dariah.eu/>, which develops a digital research infrastructure for the Arts and Humanities. DARIAH's roots are in VREs, it was originally planned to widen the reach of e-science. The ultimate aim of the project is a Virtual Research Community (VRC) that is based on the collective intelligence of the community and establishes an architecture of participation. It aims to be a Social VRE, one that is institutionally grounded in everyday research and empowers research that is open and re-usable. The challenges include interoperability (integration of data), collaboration that can lead to collective intelligence, integration of tools and services into everyday research, software engineering vs. humanities practices. Conceptually, VREs are second-generation digital libraries, in which the concept of “research objects” is foregrounded. Joris van Zundert of the Department of Software R&D at the Huygens Institute, Royal Netherlands Academy of Arts and Sciences, presented on AlfaLab, a digital research infrastructure <http://virtualknowledgestudio.nl/current-projects/alfalab/>. The project is a collaboration between computer science and humanities divisions. It takes the “distant reading” paradigm, coined by Franco Moretti, as its underlying concept. AlfaLab conceptualizes VREs as highly specific applications for very specific research questions for specific communities. AlfaLab comprises TextLab, GISLab, LifeLab, underlying them is a common technical infrastructure that facilitates very basic functions, e.g. annotations across corpora and collections. AlfaLab 2.0 is currently 2 planned as a support service for digital researchers enabling them to innovate, providing support, and offering sustainability. Session two In the second session, entitled “TEI”, Piotr Bański of the University of Warsaw and Adam Przepiórkowski of the Institute of Computer Science, Polish Academy of Sciences, presented on “TEI P5 as a Text Encoding Standard for Multilevel Corpus Annotation”. The presenters introduced the National Corpus of Polish (NCP) project <http://nkjp.pl/>, a three-year project that involves four institutions. It builds on several previous corpora projects and employs multiple levels of linguistic annotation. One million words have been annotated manually at a detailed level. The workflow starts with the source text, which is segmented into its morphosyntax, into syntactic words (syntactic groups and named entity versions), and word senses (kept in a sense inventory). Several standards have been investigated to provide the encoding for the corpus: TEI P5, XML Corpus Encoding Standard (XCES), and the ISO TC37/SC4 family of standards, including LAF (Linguistic Annotation Framework). The pros and cons of the standards were reviewed and weighed and a decision for TEI P5 was made primarily due to its good documentation and solid user base. The NCP project has customized the P5 schema for its specific purposes by comparing the solutions offered by TEI to the other standards and then recommend the customization as best practice for the project. The segment element is used for sentences, phrases, words, etc. instead of feature structures, ptr is used to point to immediate constituents as it allows for encoding discontinuities, which are very frequent in the Polish language. TEI has proven to be a reasonable choice as it is sufficiently flexible to implement data models proposed by a variety of standards. John Walsh of Indiana University gave a paper entitled “'It’s Volatile': Standards-Based Research & Research-Based Standards Development”. Digital Humanities scholarship connects humanities scholarship with technological research and development, such as standards development. Unicode is one of the most important standards for the digital humanities. The Chymistry of Isaac Newton project <http://www.chymistry.org/> uses Unicode extensively to digitise and edit, study and analyse the alchemical works of Isaac Newton and to develop various scholarly tools around the collection. The project has made a proposal to the Unicode Consortium to register all the alchemical symbols in Newton's alchemical writings. The proposal process is a thorough, peer-reviewed process, the project's proposal is currently in its final stages and due to be included in the Unicode 6.0 release of the standard. One of the main challenges of the project was that by its nature alchemical language is imprecise, full of allusion and uncertainty, which is in conflict to the stringent requirements of character standardisation. The graphic and pictorial nature of alchemical language makes glyphs very important carriers of meaning, which contradicts the idea in Unicode of abstract characters that can be represented by a number of different glyphs. Some of the solutions employed involved additional layers of meaning used in combination with standard Unicode characters, additional markup, use of images, and representing variants of characters and glyphs. The Unicode proposal process has been a rewarding experience and as it was successful it has benefited the scholarly community and potentially a much wider audience. Deborah Anderson of UC Berkeley presented on “Character Encoding and Digital Humanities in 2010 – An Insider's View”, a paper related to the previous paper in the same session. 3 Unicode <http://unicode.org/> is now the most used encoding standard on the Web, TEI's endorsement of the standard has been phenomenally helpful in establishing it as the encoding standard of choice for the electronic transmission of text. Unicode 5.2 currently defines over 100,000 characters for over 75 scripts, Unicode 6.0 is about to be finalized and published. While Unicode offers a vast array of character choices to the digital humanist, its breadth and depth can also be confusing. This is particularly true of the approval process for new characters by the Unicode Technical Committee and the relevant ISO committee. The presenter explained this process, which involves two standards committees, namely the Unicode Technical Committee (UTC) and the ISO/IEC JTC1/SC2 WG2, and usually takes more than two years. The UTC is primarily concerned with technical questions and computer science applications, discussions are often based on potential issues for computer systems and electronic text. The ISO committee is comprised of national representatives of all member countries and support from a national body is highly advisable and sometimes essential for proposals to be accepted into the international standard, discussions are sometimes more political in nature. Being aware of the interests and backgrounds of each standard group and their members can help explain what appears to be a spotty set of characters in Unicode. Session three The third session, headed “Archives”, began with a paper by Peter Doorn and Dirk Roorda of the Data Archiving and Networked Services, Netherlands, entitled “The ecology of longevity: the relevance of evolutionary theory for digital preservation”. The presenter borrowed notions from evolutionary theory in thinking about digital longevity or “digital survival”. Thus, the concept of “sustainability”, which in biology refers to ecological systems that remain diverse and productive, is analogous to preservation in the digital environment. Digital objects are intended for the long term, but only short-term survival can realistically be accomplished. The preservation of individual objects may be less effective than making them richer and interoperable. Preservation strategies are either emulation, which preserves the environment of the data, or migration, which adapts the data to the environment. Current digital preservation thought states to first preserve and then re-use, however, this has the perverse consequence that preserving too well actually prevents re-use. Moreover, in evolutionary theory systems get rid of unused functions, therefore it is more suitable to first re-use and thus preserve. Evolution is based on mutation and copies, not originals, making copies preserves data. Reproduction in ecological systems can therefore be directly linked to the interoperability of data, mating your data will preserve it. An ecology of sustainable information must be governed by certain rules: a financial incentive to preserve is helpful, optimizations of both value and cost are necessary, data use is instrumental in preservation, and the actors involved must constantly be aware of the chances and risks of survival of the digital data they curate and use. The second paper, entitled “The Specimen Case and the Garden: Preserving Complex Digital Objects, Sustaining Digital Projects”, was given by Melanie Schlosser and H. Lewis Ulman of Ohio State University. The presenters outlined the aspects of preservation of digital humanities projects from a librarian's and a scholar's point of view. Preservation too often means taking a snapshot of living projects and storing them away where they die from lack of use. From a scholarly perspective, digital projects led by scholars are often ambitious and complex, more often than not a labour of love of individuals, often situated on the fringes of the official institutional preservation infrastructure. From a librarian's perspective, in order to support 4 faculty research effectively, standardisation is paramount, there are few resources available for unique projects, however the library is committed to offer advice and to preserve projects. The NEH-funded project "Reliable Witnesses: Integrating Multimedia, Distributed Electronic Textual Editions into Library Collections and Preservation Efforts" <https://digitalhumanities.osu.edu/>, a collaboration between faculty and librarians, can serve as an example. For preservation purposes ('the specimen case'), the library has adopted a lowlevel approach to preservation of projects, namely preserving all pieces of a project as METSfiles, employing a dedicated, registered METS profile. The digital projects are ideally sustained in their production environments ('the garden'), this requires to work closely with and connect all the staff involved in digital projects in order to support staff in creating robust, sustainable projects. This makes it possible to curate a set of completed projects that comply with certain criteria and to sustain them beyond the time of involvement of the original creators. The final paper of the session, entitled “Digital Libraries of Scholarly Editions”, was authored by George Buchanan of the School of Informatics, City University London, and Kirsti Bohata, Centre for Research into the English Literature and Language of Wales (CREW), Swansea University. Digital editions are complex, often bespoke, dynamic objects, which are difficult to preserve. The presenters are working on establishing common denominators of the scholarly electronic editions created in Wales that will allow them to create a central “library” of digital editions. Digital libraries have been a major movement in Computer Science since at least 1995. Digital editions are best organized into collections, but often digital libraries are not designed to hold complex objects, instead they reduce, make static, and atomize content. While digital library systems can provide preservation and collection-level description, little focus is on the internal organisation of scholarly editions. A user study of producers and consumers of digital editions has been undertaken to establish the common features of scholarly editions. The presenters aim to develop extensions to the Greenstone Digital Library system to bridge the gap between digital library software development and the specific requirements of digital editions. In particular, the extension of the Greenstone SW is intended to support both text and commentary, dynamic content, collections of content, and provide interoperability. Session four In a session headed “Literature”, John Walsh of Indiana University gave a paper entitled “'Quivering Web of Living Thought': Mapping the Conceptual Networks of Swinburne's Songs of the Springtides”. The presenter introduced a new section of the Swinburne Archive <http://swinburnearchive.indiana.edu/> with experimental work. The starting point for the experiments is the postmodernist view of texts as nodes in networks that work on both intratextual and intertextual levels. Swinburne himself used the metaphor of networks and nodes in a Web of intertextuality. His poetry is highly allusive and referential, and employs its own collection of symbols, images, and concepts. His “Songs of the Springtide” collection is an excellent example of such a deliberately designed volume of networks: the poems have been organized into three parts with a transitional sonnet and a final ode. The thematic networks employed include: literary figures; poetry, music, and song; the passage of time; celestial bodies: sun, moon, and stars; the sea. This highly structured volume of poetry is almost schematic in its explorations of these topics. The texts have been marked up using the TEI P5 encoding standard, and keywords are used to encode these concepts. A P5 taxonomy is used and keywords are encoded as <seg type=”keyword”>...</seg>. The presenter demonstrated an 5 interface that allowed for the selection of clusters of keywords, which were highlighted in individual poems, but could also be seen in a bird's eye view of the entire volume of poetry thus offering a unique view of the nodes at work in the overall structure of the volume. The second paper of the session was given by Mark Byron of the University of Sydney and was entitled “Digital Mediation of Modernist Literary Texts and their Documents”. The presenter introduced a project that will digitise and transcribe all of the manuscripts of and documents relating to Samuel Beckett's novel Watt (1953). The project is part of the larger international Samuel Beckett Digital Manuscript Project, which aims to have all of Beckett’s literary manuscripts transcribed and represented in digital form. The Watt editorial project has identified the problem of MS transmission as key to the complexities of the novel. The MS, written during the Second World War, suggests a series of major revisions of the text, it is littered with emendations, revisions, and additions. The addenda section at the end of the novel is far more than a metatextual game, it points towards concepts and nodes in the work itself. The opportunity to represent a large corpus of materials in a single online environment, which allows for their manipulation and exploration in new ways, as well as exploring the implications of textual evidence for the authority of “the text” are two of the unique opportunities offered by digital scholarly editions. The innovation of the project will be the coherent representation on a platform of all the documents required for a scholarly edition, showing the MSS in their contexts as well as their relation to the published text. The documentary transcription of the MSS will follow the TEI/XML Guidelines and will seek to capture both text and the many illustrations. A future critical edition of the novel will be able to build upon the findings of the documentary edition produced in this project. Joshua Sternfeld of the NEH gave a paper entitled “Thinking Archivally: Search and Metadata as Building Blocks for a New Digital Historiography”. The paper explored the relationship between data of historical evidence and the interpretative implications of finding and using such data. As data is collected and selected, the archivist serves as a mediator of records, not merely as a curator. Selection criteria themselves are synonymous with interpretation and since no coherent selection policy can be applied across collections, it must be made transparent for any users of the same materials and open for interrogation by historians. In the online world, the interface and search functionalities offered contribute to the ways in which knowledge is produced and is responsible for the preference of certain records over others (e.g. tag clouds). The archive organises information and makes this organisation available, which in itself is an interpretative process that favours certain ways of exploration and certain methodologies. Digital historiography therefore needs to pay more attention to these often implicit decisions. Another concern is the reliability of historical data. Provenance is a key archival principle and it should be extended to digital historiography. Searching in particular can lead to the decontextualisation of records and the obscuring of relations and dependencies of historical data. Historians should therefore always think “archivally” when selecting and evaluating historical data. Session five The fifth session, headed “Annotation/Markup”, saw the presentation by Wendell Piez, Mulberry Technologies Inc., of a paper entitled “Towards Hermeneutic Markup: An architectural outline”. The presenter explained that hermeneutic markup is essentially process- 6 oriented, not product-oriented, it is devoted to interpretation, not re-usability, it is focussed on a specific enquiry and may contain multiple perspectives on the same text. XML encoding is always hierarchical, thus multiple structures need to be merged, e.g. literary and linguistic markup. Any anomalies in texts pose problems for the markup, sometimes they are in direct conflict with the markup employed. Therefore, a descriptive, not a prescriptive, schema mechanism is required to overcome these issues. Markup today serves the end of “production”, it assumes that schemas are complete and “given”. In a new, more flexible architecture, we need to include schema design and query design as components of the process. We need openness, a gentle learning curve, build on the current infrastructure, a data model that supports overlap, many hierarchies, a flexible and extensible schema, ad-hoc schemas, multiple schemas for documents, extensible annotation frameworks, and structured and taggable annotations. Jane Hunter of the University of Queensland gave a paper entitled “The Open Annotation Collaboration: A Data Model to Support Sharing and Interoperability of Scholarly Annotations”. The Open Annotation Collaboration (OAC) <http://openannotation.org/> is an 18-month project funded by the Andrew W. Mellon Foundation. It addresses a number of problems with scholarly annotations: there is a plethora of annotation/tagging tools, but a lack of consistency, interoperability, and standards; the annotations are neither persistent nor addressable, neither sharable nor aggregatable. The OAC data model is intended to enable sharing and interoperability. The guiding principles have been: annotations must be publishable and discoverable; be Web- and resource-centric; have unique URIs; be based on W3C principles (Semantic Web and Linked Data). The basic assumptions (based on use cases) are: the act of annotation is an event (creator/created); there is a body of the annotation and a target of the annotation; an OAC annotation is an ORE (Object Re-use and Exchange) aggregation, which can be described and serialized in a number of ways (RDF, RSS etc.); segments of objects are annotable; the annotation can also refer to a proxy (which can have additional qualifiers for more detailed descriptions) which points to the target. OAC annotations are automatically propagated across various instances of digital objects, e.g. resized or 3D versions of videos. OAC also allows for annotations of aggregations and annotations of annotations. Currently, the AustLit Annotation Service Architecture is being developed based on the OAC specifications. Furthermore, OAC makes the annotation of dynamic resources possible (blogs, twitter postings, etc.), these can be divided into timeless annotations (that never change), uniform time annotations (which have the time stamp of the target), and variable time annotations (which have both the time stamp of the target and of the annotation). Next steps include finishing version 2 of the OAC data model, the publication of the OAC ontology, and a detailed evaluation of the OAC model, as well as a number of workshops. Session six In a session headed “Geography”, Natasha Smith of the University of North Carolina at Chapel Hill gave a paper entitled “Unfolding History with the Help of the GIS Technology: a ScholarLibrarian Quest for Creating Digital Collections”. The presenter reported on the Carolina Digital Library and Archives (CDLA) and Documenting the American South (DocSouth) collaborative digital library laboratory that creates, develops, and maintains online collections on the history of the American South with materials drawn primarily from the archival 7 holdings of the UNC library. User involvement/requests/responses are an important factor for digital collection development in the libraries and maps have been a frequent request. Historical maps in particular are very popular, however, they are difficult to digitize and represent in an online environment. The North Carolina Maps project has nevertheless managed to put over 3,000 maps of the state of N.C. online. The project has developed its own digital workflow and infrastructure for this purpose in parallel with another map collection, the “Going to the Show” history project <http://www.docsouth.unc.edu/gtts/>, which maps cinemas across the state in the first three decades of the 20th century. The project has found the Sanborn Fire Insurance maps extremely valuable for its purposes. The project also makes use of the Google Maps API. As a method, the project applies geo-referencing on stitched together Sanborn maps that cover the entire state, over 1,300 movie theater venues have thus been identified. Another project is the “Driving Through Time” project, which layers historical maps and makes it possible to “travel through time” by using different levels of opacity for the maps. “Main Street Carolina” finally is a collaborative project and toolkit (Web application) for content creation that can be re-used by other projects (JavaScript API) and has already been used by several demonstration projects. The second paper of the session, entitled “GIS, Texts and Images: New approaches to landscape appreciation in the Lake District”, was given by Ian Gregory of Lancaster University. The presenter reported on the work of an interdisciplinary research group at Lancaster University that works in the field of Historical GIS, the application of GIS to historical and humanities data in general. The group has employed GIS in the field of literary studies and has been influenced by Franco Moretti's concept of 'distant reading', which summarizes large bodies of text rather than focussing on a few texts in detail. In a project funded by the British Academy, the group has studied Coleridge's and Gray's tours of the Lakes <http://www.lancs.ac.uk/mappingthelakes/>. The aim is to establish whether we can create a GIS of text that is useful in literary research. For this purpose, the two accounts have been manually converted, analysed, and encoded. Place names have been matched to a gazetteer and turned into GIS form. The project has thus produced a number of different maps, e.g. simple dot maps, density smoothed maps, and maps of emotional response to the landscape. Thus both the physical characteristics of the tours (altitudes, population densities of places visited etc.) and the literary means by which they have been recorded have been taken into consideration. A combination of using a historic map of 1815 of the Lakes and modern aerial imagery provided by Google Earth has been found a valuable approach for an alternative “close reading” approach. The project has also experimented with geo-referencing photos of the Lake District on Flickr to visualize the areas described. It was thus possible to achieve a distancing from the text and a subsequent close reading in the light of the new data gathered for individual passages in the text. The project has concluded that text and GIS can work together very fruitfully by including various sources in the analysis process and offers many potential applications in humanities research. Elton Barker of the Open University gave the third paper of the session, entitled “Mapping the World of an Ancient Greek Historian: The HESTIA Project” <http://www.open.ac.uk/Arts/hestia/>. HESTIA (the Herodotus Encoded Space-Text-Imaging Archive) is an interdisciplinary project, sponsored by the AHRC and involving the collaboration of academics from Classics, Geography and Archaeological Computing from Oxford, Birmingham, and Southampton. Herodotus' Histories about the conflict between Greece and Persia are at the centre of the investigation. The digital text (TEI/XML) of the 8 Histories has been provided by the Perseus Project, they have been enhanced as part of the project and been resubmitted to Perseus. The aims of the project are: to explore geographic concepts in the texts; investigate the textualization of these geographic concepts; and the use of ICT in order to visualize these concepts. The XML-encoded text is encoded for use in the project's GIS system (QGIS) by creating references that specify the location, the section in the text, and the toponym. The location is recorded by its name, category (settlement, territory, physical features), coordinates, and its super-region. For the purpose of visualization, the project has investigated a number of approaches: generated GIS maps (simple point maps); Google Earth has been found to be both easy to use and valuable for the visualization of the data as it can display the text alongside the maps and images; TimeMap, developed by Nick Rabinowitz, is a mashup that investigates narrated time in Herodotus' text by allowing for data plotted on Google Maps to appear and disappear as a timeline is moved. The project has also done a detailed analysis of networks by applying a qualitative analysis in just one chapter of the text. All geographic entities have been defined in the section and all phenomena that occupy a place in the topographic reality have been marked up. Using a system of weighting relations between geographic concepts in the text, new insights about the networks drawn in the text have been won. Session seven The final panel session, headed “Networks of Stories, Structures and Digital Humanities”, saw three presentations on social network analysis, the mapping of relationships as networks, and its application to the humanities, the techniques and tools used, the potential of the approach and some of the problems encountered. Network analysis is a new social sciences methodology. A network is defined as a set of nodes connected by lines. Anything can constitute a node and networks usually contain different types of nodes. The lines represent relations, they can be directed or undirected. Network analysis focusses on the lines to detect meaningful patterns from a set of systematically selected nodes and lines. The characteristics (categorical qualities and quantities) of nodes and lines are called attributes. Furthermore, there are structural properties of the system connecting the nodes. Networks are observed over a period of time, which allows for the analysis of network evolution over time, the diffusion of changes through a network, and the analysis of an (intellectual) genealogy. Patterns of particular interest include those concerning the overall network structure (identifiable subgroups, centrality and brokerage, ranking) and those that show how a single actor responds to his local network structure (e.g. reciprocity, balance, preferential attachment, transitivity, homophily, etc.). Visone <http://visone.info/> is one of the tools available to visualize social networks. 1. Zoe Borovsky of UCLA presented on “Dickens' Double Narrative: Network Analysis of Bleak House”. The presenter demonstrated an analysis of the novel using network content analysis tools, focussing in particular on gender issues in the text. The narrative voice in the novel has been analysed from the perspective of male/female spheres, the hypothesis being that there is an interesting play at work between the first-person female narrator and the third person narrator in the text, as they share a number of unusual collocates. The Wordij software <http://wordij.net/> has been used to analyse the text as a network, calculating the distance between concepts in both narrative sections. Thus concepts like “man” and “woman” can be visualized for each narrator and the analysis shows that the third-person narrator’s change in style (employing 9 collocate patterns used in the first-person narrator's text) reflects and mirrors the character development in the novel. A shift from “one sex” to “two sex”, a theory that was very influential in 19th century sociology, can thus be observed to be played out in the text. 2. Wouter de Nooy of the University of Amsterdam gave a presentation entitled “Network analysis of story structure”. The presenter demonstrated a network analysis of Russian folk tales drawing on Vladímir Propp's Morphology of the Folktale (1928). The folk tale is characterized by a limited set of characters, a simple character-function relation, and the functions served by the characters occur in the same order. Taking the example of the “Swan-geese” story, the relation of characters and actions was analysed as a network. As with all fairy-tale roles, temporal patterns influence the actions and drive forward the morality of the story exemplified by the actors and their actions. Together with the attributes (e.g. positive vs. negative) of the relations of the characters, it is possible to construct a network of the story. Each character fulfills a series of positive or negative actions, his role in the story can be deducted from these actions alone. The assumption is that moral values are attached to patterns of interaction rather than the roles characters fulfill in the story, the result is that children are morally educated through the recognition of the right or wrong doing of the characters in their interactions. Interesting patterns usually occur when authors play with the conventions of the genre in which they write and structural patterns are turned upside down. Shifts between genres and developments of genres over time are possible applications of network analysis. In conclusion, network analysis requires formalization of the story structure; time is an important factor to be formalized; interesting comparisons can be made between stories and across genres. 3. Alkim Salah of the Royal Netherlands Academy of Arts and Sciences gave a paper entitled “Mapping the Flow of 'Digital Humanities'”. The presenter used scientometrics/bibliometrics as the methodology in her research, using information on journal papers collected from abstracting services, to map out activities in digital humanities research. The presenter found that digital humanities research is less widely distributed and visible than might be expected, it is usually confined to journals dedicated to the topic and journals on library and information sciences. There has been very little impact in traditional humanities publications in the various disciplines, despite the interdisciplinary approach of humanities computing. It is however to be expected that publications in online journals will alter these findings. It is also necessary to take a closer look at all the interactions between scholars and digital humanists that happen in virtual research environments to be able to fully appreciate the work that is going on in the digital humanities. These social networks pose interesting challenges and may yield fruitful insights into how the digital humanities have evolved (institutionally and globally) and how they have contributed to humanities research as a whole. It was also announced that next year’s Digital Humanities conference will be hosted by Stanford University, and that the 2012 conference will be hosted by the University of Hamburg. 20/08/2010 Alexander Huber Bodleian Library, University of Oxford 1 0 <http://users.ox.ac.uk/~bodl0153/> 1 1