Should Mercier and Sperber Change the Way We Teach and Study Reasoning?
Groarke, Leo, Argumentation and Advocacy
In this short comment I ask whether Mercier and Sperber's seminal account of reasoning should change the way we teach and study reasoning--not in cognitive psychology, but in argumentation theory (understood as a cross-disciplinary amalgam of informal logic, rhetoric, dialectics, AI and related disciplines that study arguments in natural language). In attempting to answer this question, I focus on two aspects of their approach: first, their evolutionary account of reasoning; second, their attempt to use it to explain the persistence of cognitive errors that tend to characterize ordinary reasoning (most notably, in instances of confirmation bias and motivated reasoning).
In making my own arguments, I put aside some concerns about Mercier and Sperber's project that are rooted in the thought that we can reason (i.e., offer reasons for some belief or action) in radically different ways: by constructing geometrical demonstrations, by declaiming, by drawing political cartoons, by thinking to ourselves, by engaging in parliamentary debate, by performing in a play, and so on and so forth (e.g., Birdsell & Groarke, 2007; Carozza, 2007; Gilbert, 1997). This makes reasoning an umbrella concept which is, arguably, ill suited to a general account ("a grand theory") of reasoning built on one its many variants.
But that is a concern I leave for elsewhere. Assuming a general approach, why do humans reason? Mercier and Sperber (2011) answer that reasoning has evolved as an aid to communication: "Reasoning enables people to exchange arguments that, on the whole, make communication more reliable and hence more advantageous" (p. 60). It does so by acting as a mechanism of epistemic vigilance which allows one to evaluate the content of messages we receive (p. 96). "Reasoning contributes to the effectiveness and reliability of communication by allowing communicators to argue for their claim and by allowing addressees to assess these arguments. It thus increases both in quantity and in epistemic quality the information humans are able to share" (p. 60). Mercier and Sperber dub this account the "argumentative" theory of reasoning.
The claim that reasoning is a mechanism for persuading others is in keeping with well-established rhetorical accounts of argument. Mercier and Sperber's unique perspective is their claim that the argumentative theory explains a body of psychological research, which is usually taken to show that humans are poor reasoners. In particular, they cite research on confirmation bias that shows that humans tend to reason in a manner that rationalizes and justifies their prior perspectives, prejudices and interests, ignoring and dismissing countervailing evidence, even when it is logically compelling. In the face of this finding, Mercier and Sperber argue that we should understand confirmation bias as a phenomenon that demonstrates that humans are good reasoners in the argumentative sense--i.e., in the sense that they are good at finding ways to support and defend the perspectives they present and defend in argument. From this point of view, confirmation bias is not a problem, but a trait of strong reasoners consistent with the argumentative theory.
One might object that this rehabilitates confirmation bias only at the expense of reasoning's supposed evolutionary value, for the value of reasoning appears to depend on its ability to expose our mistakes and open our minds to new and better ways of understanding. Mercier and Sperber propose a dialectical solution to this apparent inconsistency, suggesting that reasoning in a group context corrects the problems with confirmation bias:
When one is alone or with people who hold similar views, one's arguments will not be critically evaluated. This is when the confirmation bias is most likely to lead to poor outcomes. However, when reasoning is used in a more felicitous context--that is, in arguments among people who disagree but have a common interest in the truth--the confirmation bias contributes to an efficient form of division of cognitive labor. …