Ethos, the character of the speaker, is a powerful tool used to influence others through communication. Together with logos (argumentation) and pathos (emotions of the audience), it constitutes a key element of Aristotle’s Rhetoric. This chapter introduces Computational Ethos, a model of ethotic structures derived from, and verified on, real-life large-scale argument data. The model is applied to several domains with the ultimate goal of implementing technologies such as ethos mining and ethos analytics. For example, the analysis of parliamentary debates reveled that politicians frequently use structures of ethos supports, such as in the Mr. Moore’s utterance “I bow to my hon. Friend’s [Miss Widdecombe’s] distinguished past and detailed knowledge of these matters”, and ethos attacks, such as in the Mr. Forsyth’s utterance “When the hon. Gentleman [Mr Canavan] was the Member for part of my constituency, he fled the field because he was scared that he would lose”. These structures were annotated resulting in the largest publicly available corpus of appeals to ethos. Statistics derived from this corpus showed, e.g., that attacks are significantly more frequent in ethotic structures than in the structures of logos featuring 74% of ethotic attacks vs. 20% of conflicts and 26% ethotic supports vs. 80% inferences in two of our corpora.

Appeals to ethos, the character of the speaker, play the key role in dialogues where speakers emphasise, establish, weaken or undermine their own or others’ credibility and trustworthiness. Despite some valuable attempts at studying discursive features of ethos associated with speakers’ identity, the dynamics of appeals to speakers’ character is an area of pragmatics that has not yet been explored in a systematic way. The aim of the paper is to provide a way of grasping the dynamics of ethos interactions. Thus we focus rather on analysing linguistically difficult cases of appeals to ethos than proposing an \emph{a priori} model of those appeals. For that purpose we propose a method of linguistic analysis of ethos elements. As a resource we take Hansard, the UK parliamentary debate record as a genre that has a relatively high density of appeals to ethos. We treat Aristotle’s account of ethos types (Practical Wisdom, Moral Virtue and Goodwill, further referred to as WVG-categories) as a source of inspiration for building the corpus of ethos structures annotated with OVA+: Online Visualisation of Arguments software. As the theory of rhetoric turned out to be not directly applicable to the natural language data, we proposed an iterative process of annotating ethos discourse units that consists of three iterations of ethos annotation. This method allows for reformulating rhetorical theory to get closer to real data. This linguistic account of ethos elements encompasses the `agile corpus creation’ methods. In contrast to traditional corpus studies, agile methodology emphasises the need for designing an iterative corpus creation procedure that employs conclusions from the previous iteration of data annotation for improving the annotation scheme.

Cultural heritage considers the legacy from our past that should be preserved in the present for future generations. The key place in this domain is occupied by controversial cultural objects, such as Palace of Culture and Science in Warsaw, associated with \textit{historical figures}, in this case Joseph Stalin, which provoke heated public debates and dividing the society. This paper aims to introduce the new concept specific for this domain of historical ethos building upon the classic notion of the character of the speaker introduced by Aristotle. We show that the debates about whether to demolish a cultural object are in fact the debates about whether a historical figure associated with this object should still be revered by it.

The theory of conventional implicatures is a relatively new way of understanding one class of discourse material that is implicit in language use. Conventional implicatures (in contrast to conversational implicatures) are triggered by lexical items, form commitments which are made by the speaker of the utterance and are context-independent. Despite their ubiquity in language and the critical role they play in argumentation, they have heretofore been almost entirely absent from theories of argument and the linguistic expression of reasoning. By extending existing accounts that support the close analysis of argument structure with conventional implicature it becomes possible to unpack the wide variety of ways in which arguments are triggered by, composed of, and demolished by such implicit discourse material. The complex interplay between conventional implicature and argumentation that can then be modelled in this way in turn facilitates new explanation of a wide range of argumentative phenomena.