The present research focuses on politeness in candidates' discourses in U.S. political campaign debates of the 2000 elections from a socio-cognitive approach to social interaction. This approach entails an eclectic perspective on communication that intends to account for its cognitive, linguistic, relational and socio-cultural aspects in a determinate communicative encounter. This eclectic perspective is based on Brown and Levinson's (1987) Politeness Theory on the one hand, and Sperber and Wilson's Relevance Theory (1986/1995) on the other hand, with the latter constituting a cognitive complement to the former on theoretical grounds. From this eclectic approach, politeness has been conceived as the context-sensitive cognitive-based linguistic instantiation of social bonds. Therefore, politeness constitutes the linguistic enactment of social relationships in a specific communicative situation, and the internal knowledge on what is appropriate or inappropriate therein underlying such enactment. Politeness may thus consist of a) 'mitigating' behaviour, whereby the speaker (S) attends to his/her own and/or the hearer's (H) face or social image one wants for him/herself in a specific society (Brown & Levinson, 1987), or b) 'aggravating' behaviour, that is, damage of one's own and/or H's face. In view of this, the following research questions were posited in this study: 1) what are the main features of politicians' face mitigating and aggravating sequences in terms of: type of politeness prevailing in these (if any), recurrent linguistic elements (if any), and typical location of these sequences in the whole discourse debates themselves constitute (if any)?; 2) what are the specific forms face mitigating and aggravating sequences adopt (if any), and which are their features?. In order to provide an answer to these questions, a total of 89 North-American electoral debates were collected together with other secondary data (e.g. newspaper articles, television programmes, etc.). These debates were organised into Corpus of Analysis (Corpus A) and Corpus of Reference (Corpus B), out of which the former consists of 16 debates corresponding to a total of 20 hours of on-going talk, and the latter contains the rest of the debates collected. Corpus A was transcribed in its entirety and analysed according to the units of analysis of the 'pragmatic sequence' and the 'micro strategy'. Overall, face mitigation, which is commonly directed towards the audience, appeared to be the predominant shape of politeness in candidates' discourses in debates, more specifically, mitigation of the non-pure type oriented towards H's positive face or his/her desire to be approved of (ibid.). This positive face attention characteristic of mitigating sequences in these events was found to be principally based on the strategies 'presuppose/raise/assert common ground', 'assert/presuppose S's knowledge of and concern for H's needs and wants', and 'offer and promise'. These results back one of the main claims of this study, namely, that political campaign debates are essentially persuasive discourses besides antagonistic exchanges as the debate literature has commonly shown. Face aggravation, which is typically targeted at the opponent, was observed to primarily consist of aggravation of the pure sort oriented towards H's negative face or his/her desire to be unimpeded upon (ibid.). Negative face aggravation was found to usually lie in the strategies 'increase imposition weight', 'refer to rights, duties and rules not respected, fulfilled or complied with respectively', and 'challenge'. A possible explanation for the predominance of this variety of aggravation over others is that a) pure aggravation leaves no doubt as for a politician's intent to discredit the adversary, and b) negative face aggravation is not as hostile as positive face attack in debates and political discourse (cf., e.g. Lakoff, 2001), and enables the speaker to attack the rival without a potential boomerang effect on S's own image.

El presente trabajo explora el fenómeno de la cortesía lingüística en los discursos de los candidatos en debates electorales de las elecciones Norte-Americanas del año 2000. Para ello se ha partido de un enfoque ecléctico al estudio de la interacción social basado en la Teoría de la Cortesía de Brown y Levinson (1987) de carácter social, y la Teoría de la Relevancia de Sperber y Wilson (1986/1995) de carácter cognitivo, constituyendo esta última un complemento cognitivo a la primera a nivel teórico. En base a este enfoque, hemos concebido la cortesía en esta investigación como la manifestación lingüística de las relaciones sociales en contexto fundamentada en la cognición de los individuos. La cortesía se compone, por tanto, de comportamientos o actitudes lingüísticas de atención a la imagen social de un interlocutor (mitigadoras), y comportamientos o actitudes lingüísticas dañinas hacia dicha imagen (agravadoras). Con el objetivo de encontrar posibles tendencias o patrones definitorios de la cortesía en debates políticos Norte-Americanos de campaña, se procedió a la recogida de un corpus de datos consistente en 89 debates, de los cuales 16 (i.e. 20 horas de habla) fueron transcritos al completo y analizados según las unidades de análisis de la 'secuencia pragmática' y la 'micro estrategia'. En general, la mitigación, la cual se dirige comúnmente hacia la audiencia, resultó ser la forma predominante de la cortesía en estos eventos, en concreto, la mitigación no pura con orientación hacia la imagen positiva del oyente o su deseo de ser aprobado por los demás (Brown y Levinson, 1987). Por otro lado, la agravación, la cual se dirige normalmente al oponente, no resultó ser tan frecuente como la mitigación y aparecía sobre todo en forma de agravación hacia la imagen negativa del oyente o su deseo de tener libertad de acción. Estos resultados apoyan en general uno de los argumentos principales de esta investigación, a saber, que los debates electorales son fundamentalmente discursos persuasivos más que encuentros antagónicos como la bibliografía sobre debates se ha centrado en mostrar.