Other Projects

In this project, José Alfonso Lomelì Hernández and I seek to understand how the gender of the speaker influences the way their arguments are perceived in terms of persuasiveness. Gender bias is an issue that has been discussed in a number of contexts such as psychology, communication studies, as well as linguistics. However, in-depth studies on the way the gender of the speaker affects their persuasiveness in an argumentative context are still lacking. To investigate the role that speaker gender plays in the persuasiveness of arguments, we adopt an experimental approach. We designed a first experiment with 20 non-problematic, i.e., more factually oriented statements, that took either the form of a pragmatic argument or an argument from expert opinion. These statements were then recorded by a male and a female speaker. Participants are exposed either to the male or the female version of each statement and have to respond to two assessment questions targeting the statements persuasiveness and accuracy.

When a person talks, the hearer intuitively makes themself an image of the speaker. Consequently, the perceived image, the speaker's ethos, can be influenced, for the better or the worse, by what has been said. In this project, Thierry Herman (University of Neuchâtel) and I investigate the perception of different types of speaker ethos. Although ethos has been an object of inquiry for quite some time, its effect on the hearer has never been assessed. For this reason, we adopt an experimental approach and tackle the question of whether the orientation (to the addressee or a third party) and polarity (positive, negative or neutral) of an ethotic statement have an impact on the way people perceive the speaker's ethos. To do so, we designed a series of experimental items, containing a contextual statement followed by a statement targeting ethos types. With these statements, we measure to what degree it impacted the image of the speaker on one hand, and the three qualities Aristotle associated with ethos, namely competence, benevolence, and integrity on the other hand.

This project is a collaboration between researchers specialised in computational linguistics and natural language processing from the University of Passau, Annette Hautli-Janisz and Zlata Kikteva, and researchers specialised in experimental and cognitive pragmatics from the University of Fribourg, Steve Oswald and myself. The researchers on the computational side of this project are interested in building a model for the persuasiveness of responses to questions in a political debate setting. To achieve that goal, they annotate and label a large number of responses from a debate corpus, identifying argumentative features. The experimental and cognitive-pragmatic side then uses the corpus data to design specifically tailored experiments assessing the individual effectiveness of the identified features (e.g., the use of inference or rephrase, causal connectives or epistemic markers) in statements containing argument-conclusion pairs. The results of the experimental studies are then fed back to the computational model.

The straw man fallacy has been the main focus of interest in my PhD project. There, I have focused on 3rd personal straw men (people evaluating straw men attacks that were performed on a third person). In the current project, I collaborate with two researchers, Scott F. Aikin (Vanderbilt University) and John P. Casey (Northeastern Illinois University) to assess the way straw men are perceived by people who are attacked by another person with the fallacy. To do that, we are working on an experiment that includes the participants as much as possible as part of the argumentative exchange so that they have the impression of being part of the dialogue. We then confront them with a straw man attacking the views that they are supposed to defend. The participants' impressions will be measured with three assessment questions asking whether 1) they are persuaded by the straw man argument, 2) they would oppose the straw man argument, and 3) they would use the straw man argument.