Navigation auf uzh.ch
Students who produce outstanding scholarly work are nominated for a semester award. In this series, we look at a few examples to show what makes an excellent thesis, how it can benefit teaching, what motivates students to produce outstanding work, and how they are supported and guided by teaching staff.
Switzerland’s Federal Supreme Court issues between 10,000 and 15,000 rulings each year. Law student Leander Etter compared around 100,000 of these rulings and the national language in which they were written. In doing this, he highlighted aspects of legal practice at the federal level that were not previously known with such clarity. He focused on the question of which language the judgments and the literature cited in the Federal Supreme Court rulings are written in. Citations are an important part of the reasoning for any judgment and serve to support the legal arguments, take account of legal precedents and place the judgment in the context of the existing case law and literature. His thesis supervisor was Tilmann Altwicker, who holds the Chair of Legal Data Science and Public Law at the Faculty of Law. “Leander Etter worked meticulously like a detective. With analytical precision, he shed light on correlations that hadn’t previously been documented in this way,” says Altwicker.
In Switzerland, rulings by the Federal Supreme Court are generally written in one of the four national languages: French, German, Italian or Romansh. The choice of language depends on which region of Switzerland the case is tried in, or in which language the parties made their submissions. Roughly two-thirds of all rulings from the Swiss Federal Supreme Court are written in German because the proceedings originate from the German-speaking part of Switzerland. Rulings in French and Italian originate from the corresponding language regions. Rulings written in Romansh are rare and are only issued in very specific cases.
Leander Etter grew up in an area in which Romansh is spoken and speaks the language himself. He says that this heightened his awareness of linguistic issues. “After I graduated from high school, I was initially unsure what degree I should study; it was a choice between law and informatics and ultimately I chose law because I was also interested in language.” In the end, his Master’s thesis in legal data science combined all his different interests – in law, informatics and language. The impact of a specific national language on case law had rarely been comprehensively analyzed before now. “I found this topic particularly fascinating and was delighted that Professor Altwicker supported it,” recounts Etter. He suggested the topic and question to explore in his thesis himself. “Leander Etter came up with an original research question, but also displayed a fine aptitude in realizing what you can actually do in the time you have to complete a Master’s thesis,” says Altwicker.
Legal data science research aims to use quantitative methods to find structures or patterns in the legal data and identify new correlations. “My students are developing automated methods for classifying or comparing court rulings or even making predictions on the likelihood of judgments,” explains Altwicker. Etter devised the data analysis methods he needed himself. “This is one of the reasons why this thesis is worthy of an award,” says the law professor.
Federal Supreme Court rulings are between 10 and 40 pages long, and they’ve only been digitized since 2000. When Etter conducted his analysis, he was able to make use of a dataset compiled by the former doctoral candidate Florian Geering comprising roughly 100,000 Federal Supreme Court rulings.
Etter classified the rulings of the Federal Supreme Court by their language and analyzed the citations using data science and statistical methods. “The outcome is controversial,” says Altwicker. “The thesis sheds a critical light on the practice adopted by the Federal Supreme Court because it shows that language barriers are rarely crossed in providing the reasons for a judgment. By sticking to individual language areas in the reasons for the judgment it reaches, the Federal Supreme Court is diminishing the basis for its arguments. It would be better for the quality of the legal practice for the Federal Supreme Court to ensure greater penetrability between the linguistic regions.”
Even before Leander Etter wrote his Master’s thesis, he worked as a student assistant at Altwicker’s chair. He was part of a team of lawyers, computer scientists and statisticians and received a great deal of support from everyone involved. “It’s this cooperation between disciplines that makes legal data science so thrilling,” he says. Brief meetings with Tilmann Altwicker to discuss the progress of his work also gave him the assurance he needed that he was following the right approach.
“When the first results emerged, I was motivated by the prospect of being able to present my thesis and make a genuine contribution to the legal discourse,” says Etter. He has also learned how to combine legal questions with modern data analysis techniques. This will be a good foundation for the PhD thesis that he’s planning to write after his bar examination. In the last few years, digitization and the increasing availability of large volumes of data has had a significant impact on the legal sciences, according to Etter. “This is opening up lots of new research questions and I’d very much like to play my part in helping to clarify these questions in the future.”