Search

Nous n’avons guère à rappeler que, depuis son fondement il y a quelque vingt-cinq ans, la Sator cherche à conjuguer la recherche littéraire avec les nouvelles technologies aptes à aider à réaliser le travail à la hauteur des ambitions du projet, c’est-à-dire à mieux comprendre et étudier la récurrence narrative dans les textes français du Moyen Âge à la fin du XVIIIe siècle. Diverses tentatives se sont succédé, que l’on pense à Toposator, SatorBase, TopoScan, PBLit, et d’autres initiatives connexes, chacune laissant entrevoir un nouveau monde de possibilités tout en nous laissant sur notre faim. Qu’est-ce qui distingue alors ce nouveau projet portant le nom Toucher (Textes, Outils, chercheur en réseau) ? Très peu, à certains égards : il s’agit toujours d’un groupe de chercheurs à la fois convaincus et circonspects au sujet du potentiel de l’informatique pour enrichir et multiplier les façons d’étudier les phénomènes littéraires. Cela dit, quelques différences essentielles existent, motivées par l’expérience et une réflexion honnête sur l’appui réel que peut apporter l’informatique aux chercheurs individuels et à l’équipe. Le concept du réseau nous a paru propice pour structurer nos objectifs, tant pour les composantes individuelles (les textes, les outils, les chercheurs), que pour le réseau qui peut se dessiner entre les composantes (un véritable réseau de réseau, ou internet). Dès lors, il ne s’agit plus simplement de créer un outil en isolation qu’on espère saura convaincre un utilisateur éventuel, mais plutôt de prendre conscience de l’ensemble des textes disponibles et de réfléchir à quelles sortes d’outils s’insérerait le plus naturellement dans les pratiques actuelles des chercheurs. Nous présenterons dans cet article les premières balbutiements du projet Toucher. En particulier, nous décrirons notre utilisation de Zotero, un outil bibliographique collaboratif, pour compiler un grand nombre de textes déjà numérisés (en différents formats et états). De là, nous présenterons un outil de balisage topique développé pour le projet qui permet de classifier des occurrences de mots-clés ainsi que les termes contribuant à la classification. Le but de cet outil est de permettre à la machine d’apprendre à reconnaître elle-même des occurrences possibles, ce qui permettrait de constituer une ressource extrêmement puissante : un utilisateur pourrait chercher dans un grand corpus des exemples possibles d’occurrences topiques afin d’enrichir la compréhension de phénomènes locaux ou d’alimenter une réflexion sur des phénomènes plus larges. Ainsi, textes, outils et chercheurs fonctionnent en réseau.

In 1971, Paul Fortier created a computer program to save significant time in analyzing French literary theme words connected to semantic fields. The system, aptly called THEME, harnessed the capabilities of computer-generated keyword concordances with frequency and distribution calculations to create research reports for user-defined literary themes. Fortier's system represents a significant achievement for digital humanities, not only due to its impressive capabilities but also for the precedents the system created in conceptualizing the role of the computer in text analysis. This paper discusses efforts to recover the THEME system and create a working approximation of the system in Python. This effort is part of Stéfan Sinclair and Geoffrey Rockwell's Epistemologica project that seeks to recover, valorize, and interpret historical text analysis in the humanities.

Visualization is the new knowing. From big data to complex processes, visualization tools and data walls are deployed for discovering knowledge and representing it back to others. The tools, however, are not accessible to most scholars, especially humanists who are trained to work with text. The Visual Matters project takes a speculative design approach to prototyping alternative visualization environments that challenge ideas about the surveillance of data. We do this by adapting an existing text analysis environment, Voyant Tools (voyant-tools.org), so that it can be used in new ways by textual scholars to learn through the graphical. In this paper we will discuss: ● Significance and Methodology: We will outline the speculative design approach of prototyping that we have taken and why we think it is important. ● Auditing Voyant: We will discuss how we audited how Voyant works (or not) on data walls and touch tables. ● Book Map: We will demonstrate a first working speculative design called the Book Map. ● Prototypes: We will discuss two sketches for future research and development. ● Conclusion: We will end with a few thoughts on visualization, ubiquitous screens and knowledge.

BACKGROUND In 1956, the Jesuit scholar Roberto Busa, S.J. founded the first dedicated center for humanities computing. (This is not the same as undertaking the first act of humanities computing. Nor does this historical "first"--however significant--mark the only intellectual lineage of what we now call the digital humanities.) Busa named the center CAAL, the Centro per L'Automazione dell'Analisi Letteraria--the Center for the Automation of Literary Analysis ("Letteraria" was sometimes replaced with "Linguistica"--or "Linguistic"). The idea was to use IBM punched-card machines to process language instead of numbers, to treat language as data. After several years in various temporary locations, in 1961 Busa's operation moved into a former textile factory in Gallarate, Italy, outside Milan, where IBM punched-card data processing machines were installed. Teams of student operators, most of them young women, in conjunction with scholars based at Father Busa's college in Gallarate, the Aloisianum, worked there for six years (1961-1967) on a lemmatized concordance to the works of St. Thomas Aquinas, the Index Thomisticus , as well as other projects in punched-card data processing. In the 1970s, the Index Thomisticus was finally printed in 56 bound volumes. In other configurations and locations, in Boulder Colorado, for example, and back in Italy, CAAL continued to operate for years. But that six-year period in the former factory stands as a significant moment in the history of humanities computing, though its details have become obscured with time. The machinery was returned to IBM in 1967 and the main building was demolished sometime between January 2010 and March 2012. As evidence of the site, how it was actually configured and equipped, we are left with (1) Busa’s own published accounts of the story of CAAL; (2) accounts by some of the former punched card operators; (3) the documentary record in the Busa Archive, including about 80 photographs, commissioned and staged by Busa himself, often as illustrations to press reports; (4) a few outside contemporary reports, for example in the typed manuscript still in the Busa Archive for a Dutch book on early humanities computing by Félicien de Tollenaere, Nieuwe wegen in de lexicologie (1963); and (5) some documents in the IBM Archives and other collections. PROJECT GOALS This project is part of an ongoing widely-collaborative scholarly effort to “reconstruct” (by researching and modeling) the 1961-1967 instantiation of CAAL in its multiple dimensions, not in any final way or to settle every question, but in order to raise new questions and to reveal what we don't know about the first humanities computing center and the work that was done there. Our goal is to explore in detail the infrastructure, workflow, and the historical and institutional contexts for this significant site (literally and figuratively) in the history of humanities computing. Busa’s work was a historic milestone for humanities computing, one strand of which fed into what we now call the digital humanities, an interdisciplinary field of increasing significance in the academy and the wider public arena. Busa’s work is often treated as a myth of the origin of humanities computing. We want to complicate the myth with history. One of our premises is that technology includes material and social infrastructure, and we’ve tried to illuminate issues important to the humanities broadly considered, such as the gendered labor involved in this kind of work, the specific roles of Cold-War funding and corporate sponsorship for humanities research, and the emergence of new forms of interdisciplinary scholarship across science and the humanities.

What graduate students want […] is simply answered at the present time: they want a job. —John Guillory One can think through a digital humanities curriculum in three ways. One can ask what should be the intellectual content of a program and parse it up into courses; one can imagine the skills taught in a program and ensure that they are covered; or one can ensure that the acculturation and professionalization that takes place in the learning community is relevant to the students. This chapte...

Using a computer to analyze a text intimidates many humanities students, but the reality is that text analysis is becoming a fundamental and naturalized part of how we operate in a digital society. Text analysis is what enables Google to compile and index tens of billions of web pages so that our search terms produce results; it is fundamental to building IBM's Watson, a computer-system that was able to beat two of the top human Jeopardy! players of all time; it allows smartphone developers t...

Academic institutions are starting to recognize the growing public interest in digital humanities research, and there is an increasing demand from students for formal training in its methods. Despite the pressure on practitioners to develop innovative courses, scholarship in this area has tended to focus on research methods, theories and results rather than critical pedagogy and the actual practice of teaching. The essays in this collection offer a timely intervention in digital humanities scholarship, bringing together established and emerging scholars from a variety of humanities disciplines across the world. The first section offers views on the practical realities of teaching digital humanities at undergraduate and graduate levels, presenting case studies and snapshots of the authors' experiences alongside models for future courses and reflections on pedagogical successes and failures. The next section proposes strategies for teaching foundational digital humanities methods across a variety of scholarly disciplines, and the book concludes with wider debates about the place of digital humanities in the academy, from the field's cultural assumptions and social obligations to its political visions. Digital Humanities pedagogy broadens the ways in which both scholars and practitioners can think about this emerging discipline, ensuring its ongoing development, vitality and long-term sustainability.

Pour avoir une idée précise des possibilités apportées par le numérique à l’édition, il est indispensable de pouvoir assumer le point de vue du lecteur : que peut-on faire avec un texte numérique ? Comment peut-on l’exploiter ? Que peut-on en apprendre en utilisant des outils d’analyse ? Se poser ces questions est indispensable lorsque l’on s’interroge sur les bonnes pratiques de la production des contenus numériques. Seul le fait d’être conscient des potentialités offertes par le texte numér...

L’apparition du numérique a entraîné ces dernières années une transformation profonde des modèles de production et de circulation des livres, qui ont peu changé depuis le XVIIIe siècle. Le web, en particulier, a provoqué une remise en question du sens même du partage des connaissances : d’une économie de la rareté, nous sommes passés à la surabondance. Auparavant, une poignée d’institutions centralisatrices, privées et publiques, étaient garantes du choix, de l’évaluation et de la distribution des contenus ; aujourd’hui, il n’y a plus de systèmes de légitimation, ou alors ils sont déstructurés. Après avoir fait le constat de la crise de ces modèles et de la difficulté d’en proposer de nouveaux, ce livre présente les enjeux et les défis complexes du nouveau monde de l’édition numérique.

In this article, we discuss the first year research plan for the INKE interface design team, which focuses on a prototype for chaining. Interpretable as a subclass of Unsworths scholarly primitive of discovering, chaining is the process of beginning with an exemplary article, then finding the articles that it cites, the articles they cite, and so on until the reader begins to get a feel for the terrain. The chaining strategy is of particular utility for scholars working in new areas, either through doing background work for interdisciplinary interests or else by pursuing a subtopic in a domain that generates a paper storm of publications every year. In our prototype project, we plan to produce a system that accepts a seed article, tunnels through a number of levels of citation, and generates a summary report listing the most frequent authors and articles. One of the innovative features of this prototype is its use of the experimental oil and water interface effect, which uses text animation to provide the user with a sense of the underlying process.

Introduction: Using a computer to analyze a text intimidates many humanities students, but the reality is that text analysis is becoming a fundamental and naturalized part of how we operate in a digital society. Text analysis is what enables Google to compile and index tens of billions of web pages so that our search terms produce results; it is fundamental to building IBM’s Watson, a computer-system that was able to beat two of the top human Jeopardy! players of all time; it allows smartphone developers to build predictive texting capabilities; it also enables a humanist to study the relationship between Agatha Christie’s dementia and the richness of her vocabulary over the course of her writing career. Significant transformations of how we handle the written record are occurring as more and more of it is digitized and made available for computer analysis. Analytics are no longer an exotic preoccupation of digital humanists and computational linguists: humanities students need to understand automated methods if only because we are surrounded by their use—in everything from our email to the news. This chapter will therefore: Briefly describe what text analysis is; Make the case that analytics should be taught; Discuss how it can be integrated into humanities courses; Discuss recipes as a way of introducing students to text analysis; and Introduce the idea of notebooks for advanced students. Our goal is to start by making the case for teaching text analysis, then to provide ideas as to how it might be taught, and to end with reflections on advanced support in the form of notebooks—where the analysis becomes a form of research writing.

Introduction: One can think through a digital humanities curriculum in three ways. One can ask what should be the intellectual content of a program and parse it up into courses; one can imagine the skills taught in a program and ensure that they are covered; or one can ensure that the acculturation and professionalization that takes place in the learning community is relevant to the students. This chapter will focus mainly on the third approach, but use that to touch on issues of content and skills.

The With Criminal Intent project will create an intellectual exemplar for the role of data mining in an important historical discipline–the history of crime–and illustrate how the tools of digital humanities can be used to wrest new knowledge from one of the largest humanities data sets currently available: the Old Bailey Online. It will create a seamlessly connected environment, the Newgate Commons, in which scholars can use data mining techniques to select themed texts from the 120 million words of trial records contained in the Old Bailey, and employ these texts as the basis of a study collection in Zotero where they will in turn be available for analysis using TAPoR tools (including quantitative text analysis and visualization). In the process, this project will showcase the integration of online textual resources with bibliographical and analytical tools emerging from Digital Humanities.