Invited speakers

Ruth Amossy

Les schèmes argumentatifs et leur mise en mots : l’argument par la cause

Le repérage des arguments, leur catégorisation et, éventuellement, leur évaluation, font partie des tâches que se donne le spécialiste d’argumentation lorsqu’il se penche sur des corpus choisis. Ce travail consiste principalement à dégager de la gangue des mots des schèmes abstraits qui correspondent à des types d’arguments répertoriés dans les théories de l’argumentation : syllogismes, analogies, arguments par la cause ou la conséquence, règle de justice, argument ad hominem, etc. Pour pleinement comprendre l’entreprise de persuasion, il ne suffit pas de voir comment divers modèles servent à structurer des contenus particuliers de manière à faire triompher une thèse. Il faut aussi replonger l’argument dans son contexte, c’est-à-dire dans un interdiscours où il prend sens – l’argumentation en situation est nécessairement socio-historique. Mais il faut également voir (par un mouvement inverse et complémentaire du premier) comment ces arguments sont mis en mots. Du schéma abstrait, on revient à la matérialité du langage – au cadre d’énonciation et à la dynamique de l’échange, aux choix lexicaux et aux connecteurs, aux couches d’implicite, à la disposition de l’énoncé ou du texte. Ce processus qui lie l’argumentation et l’analyse du discours comme étude des fonctionnements discursifs à la croisée du linguistique et du social – ce que j’ai appelé l’argumentation dans le discours (2012 [2000]) sera examiné dans des exemples choisis d’arguments par la cause.

Marianne Doury

Le marquage langagier des types d’argument

Cette présentation s’inscrit dans le cadre d’une approche de l’argumentation qui cherche à prendre au sérieux sa dimension verbale. On considère en effet que l’argumentation n’est pas « accidentellement » langagière : elle n’est pas avant tout une activité de pensée, ou une structure logique, dont la forme langagière ne serait qu’un habillage imparfait mais nécessaire pour lui permettre de « circuler ». Il ne s’agit donc pas pour l’analyste de l’argumentation d’atteindre « l’essence » de l’argumentation en la dépouillant de ses oripeaux langagiers, mais de reconnaître « l’importance de la matérialité langagière pour la construction du sens », et de revendiquer « au plan méthodologique, la connaissance et l’utilisation de savoirs élaborés dans les différents domaines de la linguistique » (Micheli 2010 : 21). Plus spécifiquement, je chercherai ici à adopter une démarche que Plantin (2010) qualifie d’onomasiologique, et qui consiste à identifier d’abord les catégories cruciales pour l’analyse argumentative, puis à en repérer les manifestations langagières sur la base de l’examen d’un grand nombre de données argumentatives diversifiées (relevant de genres variés). Je m’arrêterai en particulier à la question du marquage langagier des types d’arguments (et plus spécifiquement, au marqueurs de réfutation par analogie logique, marqueurs d’argumentation par le précédent, marqueurs d’argumentation par l’absurde, d’argument pragmatique, de réfutation ad hominem).

Hugo Mercier

Argumentation and reasoning

Historically, the relation between reasoning, understood as a private mental act, and argumentation, understood as a public act, has oscillated widely. One old tradition—dating back to ancient Greece—sees them as identical in many respects. Another tradition, developed in early modern Europe, in particular by Descartes and Kant, draws a clear line between the two, often to favor reasoning over argumentation. Using data from experimental psychology, I offer new arguments in support of a strong link between argumentation and reasoning. In particular, I defend the idea that argumentation is the function of human reasoning: that reasoning evolved so that we could produce arguments in order to convince others, and evaluate others’ arguments in order to change our mind when the arguments are strong enough.

This new perspective on argumentation and reasoning has implications for several questions of interest to argumentation scholars, such as: What is the difference between an argument and other types of inferences drawn from communication, such as implicatures? How is the difference between understanding and accepting an argument cognitively fleshed out? Are there genuine fallacies of argumentation?

Francisca Snoeck Henkemans

Argumentative patterns in over-the-counter medicines advertisements:
The presentation and support of health claims

Due to institutional requirements, according to the pragma-dialectical theory of argumentation, specific context-dependent patterns of argumentative moves can be observed in particular communicative practices in various communication domains. Such argumentative patterns consist of a specific type of standpoint, a specific type of main argument and further supporting arguments in which the arguer attempts to deal with critical questions that might be raised against the main argument.

In this paper, a number of such characteristic argumentative patterns in health product advertisements will be investigated. Just as in prescription drug advertisements, in health product advertisements pragmatic argumentation generally plays a central role. By making therapeutic claims that the product will have a beneficial effect on a particular health condition, advertisers try to convince consumers that buying the product will bring about desirable consequences for them. Since claims concerning the efficacy of the health product play such a central role in the attempt to convince the consumer that the product is worth buying, there are many advertising guidelines that aim to regulate the presentation and support of such claims, with the aim of protecting the health and safety of consumers and allowing them to make an appropriate and informed choice. These regulations require advertisers to answer particular critical questions and to do so in accordance with specific requirements.

In the paper, it shall first be shown how the argumentative patterns that are prototypical for health advertisements can be explained as a way of dealing with the institutional preconditions for strategic maneuvering in the communicative activity of health product advertising. Secondly, the presentation of health claims will be examined to make clear how advertisers attempt to present their health claims in such a way – for instance by using particular implicating maneuvers – that they are able to make the strongest claims possible without openly violating the regulations for this type of advertisement.

Douglas Walton

Automated Assistants for Finding Arguments in a Debate

The expression ‘finding arguments’ in the title of this abstract is ambiguous. It can refer to two recently developed computational argumentation tools, both of which depend on extensive use of linguistic markers currently studied in a growing body of research in computational linguistics. One of these tools is argument mining, the task of identifying (finding) arguments in a natural language text. Leading work uses argumentation schemes in conjunction with linguistic markers to identify arguments used in the corpus. The other is argument construction, the task of helping a user to find arguments to support or contest a claim made in a debate. This task has traditionally been called argument invention, or the art of finding arguments. Two computational tools for argument invention have now been implemented, the IBM Debater System and the Carneades Argumentation System (CAS). Both depend on the use of linguistic markers but in different ways. Debater uses Wikipedia as its natural language knowledge base, and can assist an arguer to find pro and con arguments on a topic of debate such as ‘Should violent videogames be prohibited to minors?’ The argument assistant of CAS requires that the user input an argument diagram representing an argument extracted from a natural language text before proceeding to the step of using the automated ‘find arguments’ assistant. The prior task of building an argument diagram depends on interpreting a natural text supposedly containing an argument, or chain of argumentation, a task assisted by the use of linguistic markers.

This presentation explains how the leading systems of argument invention and argument mining currently work, and surveys how current research initiatives in computational linguistics are finding linguistic markers that can aid both tools for finding arguments to improve. It is argued that those working in the areas between linguistics and argumentation can better focus their efforts in a practical direction by seeing how these computational tools work, and by understanding how they relate to the use of linguistic markers for argumentation.