Computational Literary Studies: History, Practices, Opportunities, Requirements and Challenges

Prof. Dr. Christof Schöch
Trier University, Germany

Institute for Georgian Literature – Tbilisi State University – Georgia

13 Mar 2025

Introduction / Overview

• Before I begin, please allow me to thank you all for being here today
• Thank you, also, for your interest in Digital Humanities and Computational Literary Studies.
• Many thanks as well to Prof. Ratiani for hosting this event
• And of course a special thank to Dr. Irakli Kvedhelidze for the idea and the coordination of this event.
• It was a priviledge to host Irakli at Trier University, last year, and it is really great to be here now.

Overview

• Context: Digital Humanities
• Computational Literary Studies: History, Methods, Projects
• Computational Literary Studies: Requirements, Benefits, Challenges
• Conclusion

• My goal in this first talk is to provide an introduction to the field we now call “Computational Literary Studies” (but it has had, and still has, many names)
• I will first try to provide a brief outline of its history, the most important methods, and some recent projects
• Against this background, I will then try to outline some of the requirements and challenges that might be relevant when applying the paradigm of CLS to the literary tradition such as the Georgian one.
• But before I start, I will provide a little bit of background on myself and the context I work in, so that you get a sense of where I am speaking from, so to say.

Context: Digital Humanities

My Profile

• French Studies (.e.g La Description double)
• Scholarly Digital Editing (e.g. Essai sur le récit)
• Computational Literary Studies (e.g. CLiGS group)
• Digital Humanities (at Trier University)

Digital Humanities at Trier University

• Kompetenzzentrum – Trier Center for Digital Humanities (since 1998)
• Department for Computational Linguistics and Digital Humanities (since 2012)
• Digital History (several domains, e.g. Medieval Jewish History, Enlightenment)

The Trier Center for Digital Humanities

• Founded in 1998, now a central research unit
• About 35 team members, 30 parallel projects
• Three main research areas, many topics

See: tcdh.uni-trier.de

Digital Humanities in German-speaking areas

• DHd-Verband (scholarly association, since 2012)
• Annual DHd conference (500-600 participants, since 2014)
• NFDI consortia (Culture, History, Objects, Text)
• Professorships, centers, study programs

Computational Literary Studies: History, Methods, Projects

Origins: Stylometric Authorship Attribution

• Mendenhall, “The Characteristic Curves of Composition”, 1887
• Mosteller and Wallace, The Federalist Papers, 1963/1964
• John Burrows, “Delta”, 2002
• Maciej Eder, Jan Rybicki, Mike Kestemont: “stylo” for R (2012–)

Diversification: A Plurality of Methods

• Stylometric Authorship Attribution
• Stylometry beyond authorship (genre, period, gender)
• Topic Modeling (thematic analysis)
• Sentiment Analysis (positive / negative attitude)
• Network Analysis (for drama, correspondences, novels)
• Contrastive Analysis (keyness)
• Character Speech Recognition (in novels)
• Literary Mapping / spatial analysis
• Linked Open Data
• Machine Learning (clustering, classification)
• Large Language Models (BERT and GPTs)
• …

Opening up paths…

(Jockers 2013)

(Underwood 2019)

(Paige 2020)

(van Dalen-Oskam 2023)

• I have been a bit light on references
• I do want to acknowledge, however, a few milestones.
• These books, and the larger work they stand for, really opened up new paths for us.
• In hindsight, beyond all the methodological innovation, or the important findings,
• what these books meant to us, to me at least is:
  
• An enlarged sense of what was possible, what we could dare to do.

Consolidation: Computational Literary Studies

• Distant Reading for European Literary History (COST Action, 2017–2021)
• Priority Programme Computational Literary Studies (DFG, 2020–2026)
• Computational Literary Studies Infrastructure (CLS INFRA, H2020, 2021–2025)
• Journal of Computational Literary Studies (journal and conference, 2022–)

Principles and Opportunities of CLS (and DH)

• Open Science (e.g. Survey of Methods, 2023)
• Decanonization (e.g. Corpus for Beyond Words)
• Multilingualism (e.g. ELTeC: 21 languages)
• Sustainability (e.g. LOD and TEI in Mining and Modeling Text)
• Exploration (e.g. wine label project)

Computational Literary Studies: Requirements and Challenges

Digitization and Corpus Building

• With digital methods, we can scale up analysis, but…
• No analysis is possible without digital corpora
• More texts is good, more metadata is even better
• Challenges:
  • Digitize the canon and the archive
  • OCR models for historical languages
  • Metadata scarcity
  • Diversity paradox
  • Sustainability (formats, accessibility)

Challenge: The Diversity Paradox in ELTeC

English ELTeC corpus

Romanian ELTeC corpus

• ELTeC design goals: enable meaningful cross-language investigations
  • Balance with respect to key text characteristics
  • (text length, author gender, prestige)
    
  • Inclusivity with respect to language-based literary traditions
• Consequence: the ‘diversity paradox’
  • If the criteria are too loose, balance is compromimsed (many, but invalid, corpora)
  • If the criteria are too strict, inclusivity is compromised (valid, but few, corpora)
  • In both cases, meaningful cross-language investigations are impossible

Tools and Methods

• Most relevant tools are readily available, but…
• Challenges:
  • Competent use of tools requires training (e.g. stylo)
  • Some tools require language-specific NLP resources (tokenization, lemmatisation, POS-tagging, BERT models, etc.; cf. Topic Modeling)
  • Tools and methods need to be trust-worthy: evaluation (e.g. Beyond words)

Infrastructure (in a wide sense)

• Infrastructure is much more than just tools
  • Corpora with metadata
  • Tools and research environments (such as FuD)
  • Competency building capacity (e.g. workshops)
  • Community building capacity (see below)
• Challenges
  • Resources
  • Time / patience

Steps in community building

1. “Stammtisch” (regular social meeting)
2. Network (e.g. mailing list, Signal group, etc.)
3. Events (e.g. lectures, training)
4. Center (consolidation, locally)
5. Association (national scope)

Conclusion

Conclusion

• Computational Literary Studies is a well-established area of DH
• Today, it uses a diverse set of methods and tools
• There is a number of opportunities and challenges when establishing CLS for Georgian Literature (to be discussed…)

Thank you! | დიდი მადლობა [didi madɫoba]

References

Dalen-Oskam, Karina van. 2023. The Riddle of Literary Quality: A Computational Approach. Amsterdam: Amsterdam University Press.

Jockers, Matthew L. 2013. Macroanalysis: Digital Methods and Literary History. University of Illinois Press.

Paige, Nicholas D. 2020. Technologies of the Novel: Quantitative Data and the Evolution of Literary Systems. New York: Cambridge University Press.

Underwood, Ted. 2019. Distant Horizons: Digital Evidence and Literary Change. Chicago: The University of Chicago Press.