J. Berenike Herrmann (Basel University, Switzerland), Aletta G. Dorst (Leiden University, Netherlands), Jan Horstmann (Hamburg University, Germany) & Nils Reiter (Stuttgart University, Germany)
From 3 to 20 participants
Contact the organizer here.
When researching real-world contexts, scholars need a valid procedure for identifying and analyzing metaphorically used language. This means that they need a handle for systematically gauging the reliability of annotations, possibly for ensuing machine learning, but also a way of spotting patterns and principles behind creative and otherwise ‘messy’ cases. One such procedure for metaphor identification is MIPVU, the Metaphor Identification Procedure Vrije Universiteit (Steen et al. 2010), which celebrates its tenth anniversary in 2020. Originally developed at VU Amsterdam, it is now being widely used in research projects around the globe, being adapted for different languages (see Nacey et al. [in press] Metaphor identification in multiple languages: MIPVU around the world) and different types of authentic data, including spoken and multimodal discourse (see VISMIP for visual metaphor [Sorm & Steen 2018], and FILMIP for filmic metaphor [Bort-Mir 2019]. However, many scholars are challenged by developing their own computational environment for applying MIPVU and analyzing the annotations, including reliability tests. Some use an XML editor, others MS Excel or MS Word. So far, a tool is missing that facilitates an intuitive annotation and analysis environment.
Our hands-on tutorial introduces metaphor scholars to a computational tool for applying MIPVU called CATMA, which stands for Computer Assisted Text Markup and Analysis, see https://catma.de. CATMA was developed at the Hamburg University and is currently used by over 60 research projects worldwide. It supports collaborative annotation and analysis as well as explorative, non-deterministic practices of text annotation. It integrates text annotation, text analysis, and visualization in a web-based working environment, combining the identification of textual phenomena with their investigation in a seamless, iterative fashion. The workshop is endorsed by the Special Interest Group “Digital Literary Stylistics” (SIG-DLS; https://dls.hypotheses.org/) of The Alliance of Digital Humanities Organizations (ADHO).
Workshop format: The workshop will be hands-on. After a short run-through of the basics of MIPVU, participants will learn how to work with CATMA. This introduction to CATMA includes the core annotation and analysis functionalities, text upload, annotation and specification of annotation categories, as well as text queries of source text and its annotations. Participants will also generate the visual output of query results and learn the export of markup data in XML format. In a last section, we will discuss several methods of annotation validation, including inter-annotator agreement.
Participants will be asked to acquaint themselves with the English version of the annotation manual MIPVU in preparation. Access to language-specific MIPVU protocols from Nacey et al. (in press) may be provided upon request. Participants will, however, need no prior knowledge of computational text annotation and can work with their own laptop computers. CATMA runs on Laptop or PC (Windows, Unix or MacOS) with a current web browser (MS Explorer or Edge; Firefox, Chrome, Safari) with a mouse or touchpad (touchscreen navigation is not yet supported).
About the organizers
J. Berenike Herrmann, Basel University. Berenike is Assistant Professor (‘Oberassistentin’) at the DHLab Basel. She is a literary/linguistics digital humanist with a track record in metaphor studies, computational stylistics, and cognitive stylistics. Among her research topics are modernist and realist literature from a comparative perspective, the epistemology and methodology of computational textual studies, and evaluative discourse on the web 2.0. She is co-developer of the original and the German MIPVU.
Jan Horstmann, Hamburg University. Jan is a digitally working literary scholar and currently coordinates the forTEXT project ( https://fortext.net) for the dissemination of digital routines, resources, and tools for text annotation and analysis. He has taught CATMA in many seminars and workshop in recent years and uses the tool especially for the annotation and analysis of German literary texts. His particular interests lie in narratological categories as well as renunciation and irony in Goethe.
Aletta G. Dorst, Leiden University. Lettie is an Assistant Professor in English Linguistics and Translation Studies at Leiden University Centre for Linguistics. Her research interests and main publications are in the fields of metaphor studies, stylistics, translation studies, genre analysis, contrastive linguistics and corpus linguistics. She was part of the team that developed MIPVU for English and Dutch, and co-author of German MIPVU. She has applied MIPVU to texts from different genres and registers and in different languages.
Nils Reiter, Stuttgart University. Nils has a background in natural language processing and works on the analysis of literary texts. He is principal investigator in several DH-related projects concerning dramatic, narrative and historic texts and has supervised several workshops, tutorials and classes on statistical methods for text analysis, including annotation and metaphor detection. Starting in September 2019, he will be a visiting professor at Cologne University.