Here’s an interesting section from a Deirdre Wilson paper. It touches on some of the ideas in Mercier & Sperber, 2011, which describes reasoning as a tool for argumentation. It seems potentially relevant to what I’m doing because I’ve been interested in persuasion and argumentation in mathematics talks and this brings that right to the level of logical terms such as if, and, or and unless.
Procedural information and epistemic vigilance
As noted above, the speaker’s goal in producing an utterance is not only to be understood, but to be believed. I have given some evidence that hearers are equipped with epistemic vigilance mechanisms which protect them from being accidentally or intentionally misinformed. What resources do speakers have for getting past the hearer’s epistemic vigilance mechanisms and convincing him after all?
Suppose I want you to believe a certain proposition, but I realise that it conflicts with some background assumption you have in mind. One way to get past your vigilance mechanisms would be to produce an argument showing that this proposition follows logically from, or is strongly supported by, other background information you have available that you would be reluctant to give up. Producing an argument of this type would involve the use of logical or discourse connectives to display the intended logical or evidential relations. As Sperber (2001) puts it, Displaying [logical/evidential relations] requires an argumentative form, the use of logical terms such as if, and, or and unless, and of words indicating inferential relationships such as therefore, since, but, and nevertheless. It is generally taken for granted that the logical and inferential vocabulary is – and presumably emerged as – a tool for reflection and reasoning.
From an evolutionary point of view, this is not particularly plausible. The hypothesis that such terms emerged as tools for persuasion may be easier to defend. (Sperber, 2001: 410) This opens up a possible alternative to the standard relevance-theoretic account, on which the main function of discourse connectives is to guide the hearer’s path in inferential comprehension. On this new account, the main function of discourse connectives would be not so much to guide the comprehension process as to trigger argumentative procedures which yield intuitions about evidential relations, and form part of the capacity for epistemic vigilance directed at the content of communicated information. In the light of recent work on the argumentative theory of reasoning (e.g. Mercier & Sperber, 2011), this possibility seems well worth exploring further.
Returning to linguistic indicators of epistemic modality and evidentiality such as the particles yo, kana and tte in Japanese, I want to suggest that they may also be linked to epistemic vigilance mechanisms, this time geared to assessing the reliability, honesty and trustworthiness of the speaker. As noted in Sperber et al. (2010), it is in the interest of speakers to appear generally reliable, honest and trustworthy. If we regularly interact with the same people, giving them false or inaccurate information, even if it is to our own immediate advantage, may damage our reputation and end up being costly in the long run.
Conversely, doing our best to be systematically trustworthy may cost us some extra effort in the short term, but may be beneficial in the long run. The trade-off between the short term costs and long term benefits of a policy of trustworthiness may vary from person to person, so that different speakers may end up following different policies. However, speakers who opt for a policy of systematic trustworthiness would stand to benefit from a reputation for being highly trustworthy, which would be fed by common knowledge of their past actions, and might be advertised by their everyday public behaviour and demeanour.
Suppose, now, that I want you to believe some proposition, but I am not sure you will take my word for it in the absence of information about the type of evidence I have available or my reliability on that topic. An obvious way to get past your epistemic vigilance mechanisms would be to display openly the type of evidence I have, or my degree of confidence in the truth of my assertion, by using linguistic indicators of epistemic modality or evidentiality. On this account, the function of evidentials and epistemic modals would not be to guide the comprehension process (the proposition expressed by the utterance would have been understood just as well without them), but to display the communicator’s competence, benevolence and trustworthiness to the hearer.