The German Historical Institute in partnership with other research institutions has organized the conference “Digital Humanities Experiments – #dhiha6″ that will take place in Paris on June 12th. I’ve been invited to introduce a framework that we’ve developed for researching on texts.
The title of the conference, “Digital Humanities Experiments (DHE)” is somehow ambiguous and can be understood in at least two different ways:
- DHE as experimental scholarly communication: Using the liberties of the novel discipline Digital Humanities for breaking new ground in communication between researchers.
- DHE as experiment-driven research: Using the capabilities of digital technologies to adopt the paradigm experiment, primary endemic to Natural Sciences for handling issues in the Humanities.
An example of the first viewpoint can be found in the article “Process as Product” (Coble et al. 2014), where “Digital humanities experiments in post-publication filtering, open peer review, middle-state publishing, decentering authority, and multimodal and nonlinear publication platforms are discussed in the context of broader library scholarly communication efforts.” Not least the upcoming conference is arranged as experimental, where unconventional ways of scholarly communication should be proved.
The second viewpoint -still less frequently represented in the discourse- was picked up by Anne Baillot within her contribution to the #dhiha6 Blogcarnival “Two h words and one e word”. Anne describes arising issues due to the friction between hermeneutical oriented humanists and digital/experimental oriented methods. As computational linguist, I still can’t get the point, about which kind of research might work better without computers. Ted Underwood seems to have the same problem: “Claim information technology as something that belongs in the humanities — not an invader from some other field.”
For my part, I think that both perspectives can be intergrated into one holistic view. That’s exactly what I’ve tried to express with the headline of this post and is also the main objective that I’m going to pursue with the presentation of our framework at the conference. We named our framework Tesla, after … Nicola Tesla. As acronym stands for Text Engineering Software Laboratory.
Tesla is an in silico laboratory for performing experiments on textual data. It provides functions to upload text corpora in various formats, to select, configure and – last not least – combine a variety of software tools to use different methods for analyzing data. A very important feature of Tesla is that all performed experiments are documented within an electronic laboratory notebook. Such a notebook is commonplace in natural sciences. Why social and human sciences should document their reserach too, was outlined by Suzanne Dumuochel within her contribution to #dhiha6. Individual experiments can be exported to a XML file containing information about corpus data as well as the configurations of the employed software components. These XML files can be exchanged between various researchers. Such an exchange of complete in silico experiments allows scholars to re-perform experiments of other scientists, which ensures reproducibility, one of the main principles of the scientific method (or at least should be).
When a scholar accesses other researchers’ experiments, she/he can take a look at the data, analyze methods and compare results. Even more, the scholar can start her/his research exactly where others stopped working on their experiments.
On the conference I want to show how to use a workflow-management-tool (WMT) like Tesla to perform experiments on textual data in silico and how to use workflow-sharing-plattforms (WSP) like MyExperiment for sharing them. In my opinion, the combination of both, WMT and WSP, has the potential to significantly improve the exchange between researchers. To use another metaphor (Mareike König already deconstructed the experiment metaphor within her contribution to #dhiha6): WMTs can be seen as packaging machines, whereby one researcher’s work can be packed together and be shipped out via WSPs to all other scholars that are interested in His/her approach.
Anne Baillot, Two h words and one e word (#dhiha6), in: Digital Intellectuals, 24.4.2015 http://digitalintellectuals.hypotheses.org/2610.
Zach Coble, Sarah Potvin, Roxanne Shirazi: “Process as Product: Scholarly Communication Experiments in the Digital Humanities.” Journal of Librarianship and Scholarly Communication eP1137 http://jlsc-pub.org/cgi/viewcontent.cgi?article=1137&context=jlsc
Suzanne Dumouchel, Comment et pourquoi construire un protocole en SHS?, in: Digital Humanities à l’IHA, 29.5.2015, http://dhiha.hypotheses.org/1519.
Ted Underwood: “Seven ways humanists are using computers to understand text”, 4.5.2015 http://tedunderwood.com/2015/06/04/seven-ways-humanists-are-using-computers-to-understand-text/