With DH2020 Ottawa converted into a Virtual Conference, the ADHO-Special Interest Group “Digital Literary Stylistics” (SIG-DLS) has decided to convert also its new Tool criticism workshop into an online, hybrid format.
These contributions will provide food for thought for an asynchronous discussion on Humanities Commons. The discussion thread will be open for the entire week before the official opening of the Conference.
Second phase (synchronous)
On Monday 20 July, at 3pm CEST, an online live event will take place on Zoom (at a link provided on the “DH2020” Humanities Commons Group), where invited respondents will discuss the contributions and open the discussion among participants. Here is a provisional schedule:
After the success of the second edition in Utrecht, the ADHO-Special Interest Group “Digital Literary Stylistics” (SIG-DLS) organizes a new version of its Tool criticism workshop for DH2020 Ottawa. This year, it will be open to contributions of all interested researchers and it will also set a focus on historical tool criticism. The main goal of the workshop is to provide more critical awareness on the use of tools in DLS.This type of tool criticism can only be stimulated in two ways: (a) by a direct confrontation with user feedback and reports,which usually remain “in the background” of research papers; (b) via reconstructions of the historical evolution of tools and methodologies (which are never a given, but are built on top of complex –and sometimes unpredictable– research paths). The workshop is a natural extension of the DLS Tool Inventory (DLS-TI), which gathers information on the practices of the various traditions present in DLS. The workshop is scheduled for a pre-conference slot (the date pending confirmation). It will be structured as a mini-conference, where participants will be asked to present short critical reports (20 minutes each) on their use of tools in DLS. We encourage submissions on the following topics:
user interface criticism;
criticism of implied/hidden epistemologies (anatomy of tools);
theoretical tool criticism (e.g. issues in modeling);
historical tool criticism;
biases in DLS studies;
Proposals should be sent by 15 April 2020 to email@example.com. They should contain a title and a brief abstract (max 300 words). Also feel free to enter contributions to the DLS Tool Inventory (DLS-TI), if relevant. Acceptances will be notified by 15 May, 2020.
Organizers: Simone Rebora (University of Verona & University of Basel), J. Berenike Herrmann (University of Basel), Francesca Frontini (UPVM & Praxiling), Geoffrey Rockwell (University of Alberta), and Thierry Poibeau (CNRS & LATTICE)
Scientific committee: Anne-Sophie Bories (University of Basel), Julia Flanders (Northeastern University), Francesca Frontini (UPVM & Praxiling), J. Berenike Herrmann (University of Basel), Thomas C. Messerli (University of Basel), Thierry Poibeau (CNRS & LATTICE) , Simone Rebora (University of Verona & University of Basel), Geoffrey Rockwell (University of Alberta), Jan Rybicki (Jagiellonian University), and Peter Verhaar (Leiden University).
Please note that all participants and attendees will need to be registered for DH 2020. Full-day workshop registration for participants costs $70CAD and includes lunch. Conference registration information can be found here: https://dh2020.adho.org/registration/
Most of us are back from the post-conference summer break: Time to revisit! Seven weeks have passed since the DH2019 conference in Utrecht: presented as the biggest DH conference so far (with about 1,100 participants and more than 400 presented papers), it also distinguished itself for the number and quality of contributions in digital literary stylistics. With this post, through highlights and summaries, we try to provide an overview of the most recent outcomes of stylistics research in DH. Well aware that (as our colleague José Calvo Tello already highlighted when reporting on DHd2018) these notes cannot be considered but a preliminary, incomplete, and subjective report on a much more complex and stratified event. As usual, more information can be found in the conference programme and book of abstracts, both available on the conference website. To extend the variety of documentation, then, the #dh2019 and #sig_dls hashtags were remarkably active on Twitter, while participants like Geoffrey Rockwell even shared their conference notes online.
But well, if you would like to know the specific point of view of the SIG-DLS (and, more specifically, of the member who’s writing this report), here are a few notes.
The first step, inevitably, is the pre-conference workshop organized by the SIG-DLS steering committee. Held on the morning of 9 July, DLS Tool Criticism. An Anatomy of Use Caseswas conceived as a moment of self-reflection (and of self-criticism) on the tools and methodologies most frequently used in digital literary stylistics. During the workshop, Clémence Jacquot presented her use of the TXM software in the study of Apollinaire’s poetry (stimulating a discussion on limits and potentials of pre-compiled, user-friendly tools for the study of style); Geoffrey Rockwell presented a case study and a theoretical reflection on the concept of replication (thus rekindling the still lively discussion sparked by Nan Da’s criticism, but going beyond, querying ‘replication’ as a more flexible concept); and Steffen Pielström resisted the temptation to advertise his efficient topic modeling tool, inciting on the contrary an acute disputation on the pros and cons (and on the aleas) of such a technique. One of the major achievements in the final discussion was the shared impression that the quantitative/qualitative dichotomy is simply misleading, and digital literary stylistics will have to move beyond it, looking for a more organic integration of approaches.
When reaching the bulk of the conference, the impressions we can record here become inevitably more random and scattered. A general trend was evident, however, with digital stylistics contributions divided roughly in two main categories: methodological analyses, providing very advanced and stimulating discussions of (more or less) shared approaches; and applications on complex and engaging case studies. This can be considered as a reflection of the state of things in stylometry, where methodological research is needed to ground more and more steadily the scientific rigor of the approaches (with the risk of an excessive technicism), and new applications even more so, if our research field is to gain recognition beyond the boundaries of DH.
Among the methodological papers, some titles call for a specific mention. In “Comparing Assonance and Consonance for Authorship Attribution”, Lubomir Ivanov proposed prosodic elements as possible new features in stylometric classification, considering the extremely relevant role of prosody in 18th century prose (and beyond). His experiments showed how the combination of assonance/consonance with traditional stylometric methods generates significant improvements in the classification. Further research on the subject is definitely advised! In “Feature Selection in Authorship Attribution: Ordering the Wordlist”, Maciej Eder and Joanna Byszuk worked on the effects of re-ordering the (traditional) list of features. Main discovery: a combination of term frequency (the widely-used MFW) and a coefficient of variation slightly improves the results. Finally, in “Identifying Similarities in Text Analysis: Hierarchical Clustering (Linkage) versus Network Clustering (Community Detection)”, a team of four researchers led by Jeremi Ochab tried to provide an answer to a still under-studied–but ever-present–question in stylometric classification: how do I cluster the texts to visualize similarity/dissimilarity? Results showed that network clustering tends to underestimate the actual number of clusters, while hierarchical clustering overestimates it (and of course, which one is better will depend on the research question!). Traditional algorithms (Ward, Louvain) were confirmed as the most efficient.
Placing itself between method and application, David L. Hoover proposed his “Invisible Translator Revisited”. Given the (supposed) invisibility of the translator in stylometry, Hoover proposed a method for making her visible again, by filtering out the words preferred by the author. The approach showed surprisingly effective, but will it work for all pairs of authors-translators? As it happens, the idea found an indirect confirmation in “Attributions Of Early German Shakespeare Translations”, where a team of three researchers led by David Lassner was equally successful in revealing the translator (here, by combining advanced classification methods based on sub-word-level features). Shifting the focus towards themes and genres, Fotis Jannidis and his colleagues (in “Thematic complexity”) used two approaches, LDA and Zeta, to model genre distribution and to measure thematic complexity (via the “Gini index”) for dime novels and high-brow genres. Their results support their hypothesis of high literature comprising a broader spectrum of subject areas than pulp fiction genres. On a similar line of enquiry, a team of seven researchers led by J.D. Porter used stylometric approaches to identify passages in novels that adopt stylistic traits from other genres (in the paper “Microgenres”).
Moving then to the applications, it is worth mentioning “Challenging Stylometry: The Authorship of the Baroque Play La Segunda Celestina”, by Laura Hernández Lorenzo and Joanna Byszuk (who wins our award of the most prolific stylometrician, with three co-authored papers, and who on top of that also has contributed a SIG-DLS blog post!). That of La Segunda Celestina is a real “challenge” for authorship attribution: in fact, the results of the analysis did not limit themselves to “solving” the problem, but suggested possible new lines of enquiry. In “Stylometric Analyses of Character Speeches in French Plays”, then, Ioana Galleron provided an effective example of “scalable reading”, combining multiple approaches to distinguish the style of male and female characters. Finally, Massimo Salgaro and Simone Rebora (i.e. myself, in the panel “Digital Humanities for the Study of Social Reading”) showed how stylometric approaches can also be used to distinguish professional and non-professional book reviews.
A conclusive note should be dedicated to “Stylometry for Noisy Medieval Data: Evaluating Paul Meyer’s Hagiographic Hypothesis” by Ariane Pinche, Jean-Baptiste Camps, and Thibault Clérice. The work stands out as a perfect realization of a processing pipeline (from handwritten text recognition to stylometric classification, with advanced techniques to reduce noise and manage spelling variants) that instantiates a fruitful dialogue with traditional literary research, by confirming (and in part disputing) Paul Meyer’s hypotheses about the composition of the French saint’s Lives. The paper was awarded the Paul Fortier Prize for the best young scholar paper of the conference. A further success that confirms the increasingly central position of digital literary stylistics in DH research.
The SIG-DLS steering committee reiterates their congratulations to Ariane, Jean-Baptiste, and Thibault, but also to the Programme Committee chairs Elena and Fabio, and to the local organizers Joris and Franciska, for making such a spectacular conference possible. Looking forward to meeting you all again in Ottawa (hopefully, with even more stylometry to discuss)!
About the workshop on the Recreation of Harry Potter, endorsed by the SIG-DLS at DH2018 on 25 June 2018, led by Mike Kestemont and Enrique Manjavacas; and developed in cooperation with Greta Franzini and Marco Büchler, part of the 2018 Digital Humanities conference in México City
By Corina Koolen
Harry Potter novels, not one of the first topics one might think about when performing stylometry. But, as Mike Kestemont argues, popular literature is one of the unrightfully overlooked areas of digital literary studies, which still often focuses on the classics. With the workshop run by him and Enrique Manjavacas (many thanks for all your help!) during the international Digital Humanities Conference 2018 in México City, they show us exactly why this is undeserved. I am impressed by the setup chosen, which combines computational analyses with thoughtful reflection on a number of humanistic issues and interests.
One of them is the legal aspect of researching contemporary novels. After a discussion on the differences per country, where the basic conclusion is that there it is very hard to determine what exactly is legal in which country, Kestemont remarks: “We always talk about the author’s rights. I believe I have rights, too, as a researcher.” I could not agree more, this is going to be ammunition for legal discussions in the future.
Then the stylometry. Generally, stylometrists are known for performing authorship recognition. J.K. Rowling herself found out about the discipline when she was unmasked by Patrick Juola as the real author behind Robert Galbraith. There are other cool things to do, however, through stylometry. And this is where it becomes interesting for me as a researcher of popular literature: we are going to look at stylistic similarity between the novels and HP fan fiction. There are two cases that we will test, both of which follow the original structure of the novels. The first case is Aidan Chase. This fan fiction author, of whom little biographical information is available, reframed the originals to create a story world where Harry’s parents never died, but stayed true to the main story line. The second is Norman G. Lippert. He created new novels, based on the original characters, because his children were so disappointed that the series had ended. And indeed, the tools we apply show that Lippert deviates more from the originals than Chase. By using text-matcher by Jonathan Reeve, a not-yet-completely-polished but impressive tool nonetheless, in combination with the visualisation tool Bokeh, it is possible to visualize the overlap (see image); including the opportunity to make a line-by-line comparison of the sentences that have similar word usage. Chase is proven to copy large parts of the originals literally; showing how we can also apply stylometric methods to examine intertextuality, including reuse of materials.
Bokeh visualizes the overlap between two texts – size and location in the text. Credit: Christof Schöch
It gets even more interesting when we start to examine a larger body of Harry Potter fan fiction. Kestemont uses the database Archive of our Own to mine metadata and texts of fan fiction novels. This gives the researcher information that is never as easy accessible as it is here: who are the main characters, who have relationships, what are the fandoms that the author chooses to include — fan fiction authors provide this meta information because potential readers can use it to more easily to select fiction on the relationships and fandoms they prefer. The creativity is astounding: Lord of the Rings combined with Harry Potter is not as rare as I had expected it to be; and there are TV shows being crossed with Harry Potter that no one in the room has ever heard of and appear to not even be decently Google-able.
When we dive into the contents of these fan fiction stories, of course, there is sex. Lots of sex. (I for one never thought I would hear the term ‘elf porn’ in an academic context.) A computational topic model of fan fiction versus the original novels shows that topics specific to fan fiction are pornography, transportation and modern technology — that last one, interestingly, is also a topic The Riddle project team found as more typical (pdf) of ‘popular fiction’ as opposed to ‘literary’ fiction. But apart from the giggles the porn topic generates, it also shows something about how readers engage with characters. ‘Slash’ is, as fanfic researchers know, an important genre within fan fiction. Central characters, especially male ones, are paired in a romantic and often pornographic relationship. One of the first pairings was Kirk/Spock; the ‘/’ sign gives slash its name. Kestemont focuses on how attention to certain characters deviates from the original novels. Draco Malfoy and Severus Snape for instance, are much more present in fan fiction than in the Harry Potter novels, whereas Ron Weasley has the opposite effect. As Kestemont states: it gives us the possibility to research reception in a different way, to see how readers/writers engage with the original materials and characters.
That, I would stress, is an important academic outcome of this workshop, but I would like to end on another: fun. What always strikes me about Mike Kestemont’s work, is the joy he appears to get from his materials, the humour he brings to it. As he stresses, this is enhanced by working with bright and motivated colleagues such as co-tutor Enrique Manjavacas. But it is also partially explained by the type of material. As popular fiction is as much part of our cultural heritage as Great Literature is, this serves a dual purpose. First, we get a better view of fiction in general. Second, with that fun and humour comes creativity — something we could use a little bit more of every now and then. Because that, I think, is where the magic happens. Bombarda Maxima!
A few months ago, our COST Action Distant Reading for European Literary History (CA16204) was approved and started running. Even more recently, our Working Group ‘Scholarly Resources’, aka its corpus backbone, met in Prague, for the first time! We are very excited about getting started and would like to briefly report on our work in this blog.
The COST Action Distant Reading has the goal to “develop the resources and methods necessary to change the way European literary history is written” (Memorandum of Understanding). To approach this goal, the Action’s Working Group ‘Scholarly Resources’ aims to create a big open source benchmark corpus of novels from 1850-1920, called ELTeC (European Literary Text Collection). During the course of the Action, we will examine ELTeC with different computational distant reading methods such as topic modelling, authorship attribution, network analysis, stylistic analysis, and different types of character analysis.
The Action’s Working Group ‘Scholarly Resources’ will coordinate the task of creating the ELTeC; It consists of members from all over Europe, from literary studies, computational and corpus lingustics, and from information science. The COST Action is a great opportunity to collaboratively work with researchers and bring together expertise for different languages as well as computational methods.
A challenging task has been defining the corpus selection criteria, a common annotation model and potential workflows for corpus creation for the ELTeC which can be applied to European novels from different languages as well as various publication contexts.
During our first Working Group meeting, we developed corpus selection and balancing criteria which follow a simple but consistent corpus design approach and represent the variety of novels in this period. Among other things, we do not want to solely rely on normative canon-based selection criteria and set a bias to our understanding of novels. The focus of the ELTeC encoding scheme is to uniformly represent historical novels from different languages with a basic TEI mark-up. The standard markup is necessary for applying different types of distant reading methods rather than representing the original text structure. ELTeC and the encoding scheme will be freely available at our GitHub Organization.
In this way, we would like to contribute a big chunk to the creation of a digital basis for cross-national and cross-language analysis of European literature!
A month ago (26th February – 2nd March) the Conference of Digital Humanities in the German-speaking countries (DHd) took place, this year in Cologne. In this post I report on the topics of some of the talks and events that were most relevant for our Special Interest Group “Digital Literary Studies”. Surely, the post is not completely comprehensive in this way, as it reflects my subjective perspective and I am just one conference attendee; there were many other interesting talks! For further inquiries you can check the conference program here or download the Book of Abstracts.
Cologne and Berlin are on the top, followed by Würzburg, Vienna and Stuttgart. It surprises me how few researchers came from Switzerland (only 10, in comparison to 60 from Austria and 444 from Germany); the non-German speaking place best represented was Moscow (12 researchers). Another very nice visualization they made was a network of authors (node) and co-authorships (edges) with the colour marking the working place. The fun part was to try to recognize which colourful bubble you are (I found myself!):
For more information, visit the blog post in the DHd blog (in German).
Even before the conference started, several pre-conference workshops took place that were interesting for Digital Literary Stylistics: CorpusExplorer v2.0, Machine Learning with CRETA, Annotation with ATHEN and a workshop about Embedded Humanities. I attended the latter workshop, organized by Fotis Jannidis and Mike Kestemont (sponsored by DARIAH and Digital Humanities Flanders). You can check the program here, with contributions by David Bamman, Fotis Jannidis, Mike Kestemont, Marc Kupietz, Fabio Ciotti, Christof Schöch and Pim Huijnen. With so many international presenters and everyone talking in English, I almost forgot that it was still part of the German-speaking conference. However I was reminded at the end of each talk, when the audience knocked repeatedly on the tables: the German academic version of an applause! The speakers presented different approaches and use-cases to word embeddings, mainly Topic Modeling and word2vec. The workshop ended with a hands-on tutorial about the new DARIAH tool for topic modelling: Easy Topic Modeling in Python.
The morning of Wednesday had two consecutive sessions in a text mining track, with two talks about word embeddings (one with word2vec with historical texts, the later about religious posts been analyzed with topic modelin), two about stylometry (one about a complicated case of authorship attribution analyzed with Delta, one about new versions of Zeta), as well as on networks of characters in social media and an analysis of the style of the prose of German studies using n-grams.
After lunch, a panel about resources and reviews was held, which among others included the people from the Review journal for digital editions and resources (RIDE). By the way, in the last months, this journal has published two special numbers exclusively about digital collections, the kind of data that we work with, so go ahead and enjoy the RIDE!
In the afternoon two good things happened. First, Christof Schöch, a current Steering Committee member of this SIG, was elected chair (Vorsitzender) of DHd! And second, a new format was presented: Fight Club! At the stage of a bar in downtown Cologne (or was it uptown?), four researchers (three of them professors) each had three minutes to present their take on different topics of the DH.
Who says that the Germans don’t like having fun? This was the less conference-like format that I have seen in a conference anywhere. One of the winners, Professor Henning Lobin, defined the DH as the “Humanities with NSA methods”, which summarizes in only five words pretty well a part of our job.
Thursday, among other things, had a section about Digital Literary Stylistics with talks about digital close reading, direct speech recognition with Deep Learning and recognition of subgenre learning from prototypes.
During the afternoon the poster session took place. Some posters were especially interesting for DLS, like the ongoing work on Delta by Thomas Proisl and Stefan Evert from the University of Erlangen, the corpora and results of the DLINA group (with its sub projects that included a deck of cards of networks based on German literature and a way of play with them), or the brand new DISCO corpus (>4000 sonnets in Spanish, TEI+LOD).
Friday offered a sentiment analysis section across different periods, genres and methodologies. Later on, several talks tackled the annotation of corpora, including the presentation of a shared task on the Systematic Analysis of Narrative Levels Through Annotation (SANTA), and an analysis of the historical critique in Germany that addressed the Verdi Requiem (texts in TEI). The issue of automatic corpus annotation had been addressed on Wednesday by another SIG-DLS member, Berenike Herrmann, who reported on an evaluation of the POS-Tagging of the German Text Archive (DTA).
And the conference was closed with an inspirational keynote hold by C.M. Sperberg-McQueen, who besides being a XML expert speaks German perfectly (Medieval German Literature was part of his PhD). You can still watch the keynotes on the website of the conference.
I hope this was a useful overview of selected topics handled at this year’s Dhd, and that some of you will join us the next year in the double venue Frankfurt and Mainz. Remember: the deadline for abstract submission will be around August and the conference will take place between 25th and 29th of March 2019.
In fact I have no idea why I’m blogging about this, since the whole event was covered by “La Repubblica” (so nice to see that there are still countries in this world where literature is taken seriously by mainstream meadia) and the local press. They all published relatively sensible accounts of the event, also with diagrams; and a photo of some of the participants (I don’t want to brag, but…) made it into Greek press as well! Better than that: you can still (I think) watch the proceedings on Livestream!
Still… The speakers were either already members of our SIG, or will be soon, or are our close cousins from Quantitative Linguistics. Arjuna Tuzzi and Michele Cortelazzo (University of Padova), Jacques Savoy (University of Neuchâtel), Jan Rybicki (Jagiellonian University), Maciej Eder (Polish Academy of Sciences), Patrick Juola (Duquesne University), George K. Mikros (National and Kapodistrian University), Pierre Ratinaud (University of Toulouse II) all agreed that stylometric evidence overwhelmingly points to Italian writer Domenico Starnone (he has always been one of the main suspects anyway) rather than his wife, translator Anita Raja; yet her participation of some sort cannot be ruled out. This is very uncharacteristic agreement among people who use different methods and come from different backgrounds; and suggests this is a result that should be reckoned with.
I think this event was noteworthy for two things: first of all, DLS is becoming more and more visible in the media, but there is still a long way to go. Second, it ended in a spontaneous discussion on the ethics of the whole thing. Do anonymous writers have a right to privacy, especially when they’re (probably) alive? Starnone’s and Raja’s privacy had already been trampled upon by journalists who claim to have traced Ferrante’s royalties. Should they not be left alone?
My own answer is the following: I don’t really care who wrote which book, any book, and who took the money, as long as I learn something interesting about the process of literary creation. I think I did this time. While it is quite plausible that both Raja and her husband are in it together, the stylometric fingerprint is that of the latter. If it’s a collaboration, then this is a very valuable insight that might lead us further on: what is, if any, the “silent partner’s” contribution? In what way might Raja’s work as translator of Christa Wolf contribute to the Ferrante phenomenon? If they are, or one of them is, Ferrante, are they becoming Ferrante themselves?