The Pragmatics of Straw Man Fallacies

An Experimental Approach

This PhD thesis provides the first experimental pragmatic investigation of the straw man, uncovering a variety of factors that increase or decrease the acceptability of this fallacy. The thesis starts from an overview of the mostly theoretical frameworks on argumentation/fallacies on the one hand, and the straw man fallacy on the other hand. The aim is to provide a synthetic review of the research conducted in argumentation and the study of fallacies with a strong focus on modern approaches, followed by a detailed discussion about the way the straw man has been approached in the past decades. It emerges that the straw man is a more recent addition to the list of fallacies, that it has mostly been approached from a theoretical angle, and that the linguistic formulation hasn't been analyzed in detail yet. This observation lays the foundation for the experimental work presented in the thesis. The core is constituted by an in-depth analysis of different linguistic factors and their impact on the acceptability of the straw man. To do so, the book presents original experiments analysing factors such as the locus of misrepresentation, the type of reformulation, the information structure, the use of connectives to introduce the fallacious argument in French, and the use of connectives to introduce the fallacious argument across languages. The analyses of these different factors provide a deeper insight into the misrepresentation dimension of the straw man. In order to gain a complete picture of the straw man, it was necessary to investigate the perception of the underlying disagreement in straw men such as used in the experiment as well. In addition, the chapter also focuses on the role of question wording and puisque as a maker reinforcing the disagreement. Taken together, this thesis provides a more fine-grained understanding of the way specific linguistic formulations contribute to the success of a fallacy such as the straw man.

The thesis is structured as follows: Chapter 1 provides the introduction to the thesis. The reader will be familiarised with the topic and introduced to the straw man as a form of fallacious argument. This will lay the foundation for the research questions which will be addressed further in the following chapters.

Chapter 2 provides an overview of argumentation and fallacy theory. Although the discussion starts with Aristotle, the focus lies on more modern approaches to argumentation and fallacies. The aim of this chapter is to show how different theories of argumentation approach the topic in general and fallacies in particular. The chapter also addresses the language component in these approaches, as the main research question of the book is concerned with the linguistic factors that influence the acceptability of straw men.

Chapter 3 discusses various approaches to the study of the straw man fallacy. It shows that the fallacy has received a lot of attention in the past years, but that empirical research in regard to this fallacy is still lacking. This chapter lays the foundation for the experimental work presented in chapters 4 to 8.

Chapter 4 presents a series of exploratory studies, discussing three different factors, namely the locus of misrepresentation (misrepresentation of the standpoint vs. misrepresentation of the argument), the type of reformulation (explicit reformulation using similar wording vs. implicit reformulation using different wording), as well as the way of introducing the fallacious segment (with the French connective puisque vs. juxtaposition) that affect the acceptability of the straw man fallacy. Overall, this chapter shows that linguistic choices matter in the formulation of the fallacy, as the straw man was better accepted when the misrepresentation focused on the argument, was reformulated in an explicit way, and was juxtaposed to the previous segment without connective.

Chapter 5 addresses the questions of information structure for the communication of straw man fallacies. The experiment presented in this chapter assessed if there is a difference in acceptability when the fallacious argument is introduced after or before the standpoint. The results of the experiment showed that the straw man is more likely to be accepted when it follows a more traditional given-before-new pattern, i.e., when the standpoint is presented before the fallacious argument.

Chapter 6 presents a more detailed analysis of the role of causal connectives used to introduce an argument. The experiments focus on four causal connectives with an attributive function in French (puisque, étant donné que, vu que, and comme). The analysis showed that the connectives each have a different profile and diverge in their underlying functions. This reflects on their impact in the context of straw man fallacies. The results showed that different connectives lead to different effects on the acceptability of straw men.

Chapter 7 presents a cross-linguistic examination of the role of causal connectives with attributive functions. The results from Chapter 6 are thus compared to a series of experiments on English (since, given that, and as), Spanish (ya que, puesto que, and como), and German (da) connectives. The results show that the connectives are different within and across languages regarding their effects on the straw man fallacy.

Chapter 8 focuses on the refutational dimension of the straw man fallacy. The previous chapters have mainly investigated factors related to the misrepresentational dimension of the straw man. To gain a more complete picture of the straw man fallacy, Chapter 8 assesses whether people are sensitive to the disagreement underlying the straw man. Two additional factors are also analysed, namely the role of question-wording (positive or negative formulation of the experimental question, i.e., perceived agreement vs. perceived disagreement) and the role of the connective puisque as a linguistic marker of disagreement. The results show that people can spot the underlying disagreement, but that the wording of the experimental questions and the presence of the connective puisque did not affect the perception of the disagreement.

Chapter 9 provides a summary of the main findings and opens toward future lines of investigation.