Mercier and Sperber (M&S) argue that the function of reasoning is argumentative: “It is to devise and evaluate arguments intended to persuade” (see their abstract). This contrasts with a more familiar deliberative view of reasoning, which holds that the function of reasoning is to draw new conclusions and form new beliefs. Reasoning within that more familiar view is then seen as a special kind of inference, perhaps one with a distinctive relationship to consciousness and the rational faculties of the whole agent. Such views also tend to be individualistic; they hold that the psychology of reasoning has no special relation to social life.

M&S do allow that sometimes reasoning leads to new conclusions on practical and theoretical matters being drawn by the reasoner, conclusions that can be put to use in guiding action. But this is an incidental by-product of reasoning's main function, where “function” is understood in evolutionary terms.

There is also a third option, however, one drawing on the views of the Russian psychologist Lev Vygotsky (Reference Vygotsky and Kozulin1986). On this view, reasoning is deliberative in function but dialogic in structure (Yegnashankaran Reference Yegnashankaran2010). Reasoning is an internalized analogue of interpersonal discourse. Interpersonal discourse itself might be typically a complicated mix of attempts to persuade, attempts to think things through and form new conclusions, and other activities, but what results in our psychology is a tool whose function is primarily deliberative. We do not think that this view is clearly superior to M&S's, but we do think it is an important option to have on the table when considering the evolution of reasoning and the opposition between deliberative and argumentative views.

Once we have the contrast between M&S's view and the Vygotskian version of the deliberative view in mind, the message of the empirical evidence is less clear. M&S say that, on their view, “reasoning should produce its best results when used in argumentative contexts, most notably in group discussions” (sect. 1.2, para. 11). This, they say, is what we actually find. But if the aim of reasoning is to help in persuasion, one would think that a context of dialogue would promote more and more agile deployment of justifications for whatever each agent antecedently believes, not a willingness to respond to others' arguments by changing one's mind. M&S see people as poor individual reasoners but “skilled arguers,” where skilled arguers “are not after the truth but after arguments supporting their views” (see their abstract). But that picture is at tension with the fact that people interacting in groups are, as M&S accept, quite good at finding the truth by exchanging ideas, and not merely at buttressing their own positions. And on the M&S view as we understand it, any similarity between changes of mind induced by the social exchange of ideas and changes of mind induced by private reflection is incidental.

On the other side, some forms of confirmation bias do fit better with M&S's view. On a Vygotskian deliberative view, an agent has no good reason to prefer a search for confirmation of a hypothesis they are inclined to believe, to a search for disconfirmation of the hypothesis. On M&S's view, this tendency does make sense.

Finally, we suggest that M&S may underestimate the adaptive value of the directions agents may be in led by conscious reasoning. For example, they discuss an experiment where individuals are asked to choose between a small heart-shaped chocolate and a larger chocolate shaped like a roach. Most individuals chose the roach-shaped one, because making the other choice would be harder to rationally justify. M&S say that “in the light of the results from the psychology of disgust …, we can tell that their choice was certainly the wrong one” (sect. 5.3.4, para. 2). But if an analogue of this chocolate choice was faced in an evolutionary setting, a reasoner would win out.