What arguments were used for Stuttgart 21?
An interdisciplinary project at the University of Konstanz funded by the Federal Ministry of Education and Research examines persuasion strategies in the political debate
In the project "VisArgue. Why and when do arguments win? – An analysis and visualization of political negotiations" at the University of Konstanz, an automated text analytical tool is being developed using methods from political science, linguistics and computer science that will provide new insights into the workings of deliberative political discourse. Under the heading "eHumanities", the project is being funded until 2015 by the Federal Ministry of Education and Research (BMBF) with a total of 876,000 euros. The idea behind funding "eHumanities" is to support research in the humanities and social sciences with the help of yet to be developed new methods of modern information science.
The overall objective of the project is to contribute to a more effective communication between citizens and decision-makers. The implementation of large-scale public projects often leads to conflicts between state and citizens, such as in particular the controversy over the expansion of the Stuttgart train station ("Stuttgart 21") has shown. The theory of deliberative democracy claims that such conflicts can be resolved through mediation, discourse or civil dialogue. Such methods are based on the principle of deliberative communication: the rational exchange of arguments to convince the opponent.
In "VisArgue", large amounts of text, including the transcription of the mediation process "Stuttgart 21" are examined with the help of computational linguistic methods and so-called text mining for argument structures. The Konstanz political scientist Prof. Dr. Katharina Holzinger will use the elaborate analytical tool to look for patterns in this immense body of text that provide information on how argumentation is carried out in politics and when a deliberative process is being used.
She receives support from the computer scientist Prof. Dr. Daniel Keim and linguist Prof. Dr Miriam Butt. Miriam Butt is the project spokeswoman and responsible for the computational linguistic method with which the entire text is annotated, that is, divided linguistically into syntactic and semantic structures. For this purpose, their workgroup uses automatic grammars which they themselves have developed in part and are not only used in this context but are also further developed according to the task. "This is a new research topic for us as well. The issue here is not what the subject is and what is the object is, but rather: Is this clause a counter argument,” said the linguist.
She sees her work as a link between the political scientist and the computer scientist Daniel Keim, who uses the analysis method of text mining to first search for basic patterns of a text such as quick verbal exchanges, emotional discussion passages, etc., providing basic data about the text. Above all, however, it is the task of the computer scientist to make visible and comprehensible relevant features in the large amounts of text as well as complex linguistic structures. This analysis software - like the attained annotated data and the developed annotation scheme – is to be made available to a broad research audience on the CLARIN infrastructure platform that bundles infrastructures created in eHumanities projects.
The cooperation project traces back to a research initiative, an institution at the University of Konstanz that is temporally limited to the preparation of collaborative research projects and is funded within the Konstanz institutional strategy by money from the Excellence Initiative. In this particular case, it was the research initiative "Computational Analysis of Linguistic Development" (CALD). "With this cooperation, which has only come about due to the promotion of the University, we are also well-placed internationally. We are often cited as pioneers,” said the project spokeswoman Miriam Butt.
