Mercier and Sperber (M&S) make the provocative case that, in evolutionary terms, reasoning is better seen as an adjunct to communication than as a guide to decision making. However, since there are also evolutionary advantages to effective consequential choice, broadly interpreted, what might this ability be based on? We argue that emotional responses, specifically those associated with regret aversion and justification, may serve such a role, linking argument making of the sort described by M&S to consequential decision making.

In a continuing program of research, we have shown that regret aversion can help in overcoming decision errors. Much of this research draws on decision justification theory (Connolly & Zeelenberg Reference Connolly and Zeelenberg2002; Connolly et al. Reference Connolly, Ordóñez and Coughlan1997), which distinguishes regret associated with a (comparatively) poor outcome (outcome regret) from that associated with the judgment that the focal decision was wrong or poorly made – that is, was “unjustified” (self-blame or process regret). Efforts to avoid regret of this latter sort facilitates improved decision processes (Reb & Connolly Reference Reb and Connolly2010), information search (Reb Reference Reb2008) and task learning (Reb & Connolly Reference Reb and Connolly2009).

It also appears to reduce or eliminate reason-based decision errors, such as those discussed in M&S sections 5.2 and 5.3. For example, Connolly et al. (Reference Connolly, Reb and Kausel2010) compared the effects of external accountability and regret priming on the attraction (or decoy) effect, in which an option is seen as more desirable when it dominates an irrelevant decoy option. Replicating earlier studies (Simonson & Nye Reference Simonson and Nye1992; Slaughter et al. Reference Slaughter, Bagger and Li2006), we showed that accountability (a demand to justify one's choice to others) exacerbated the attraction effect, consistent with M&S's argument. Regret priming, in contrast, with its demand to justify one's decision to oneself, eliminated the effect. It seems that making regret salient may have led to a more balanced use of reasoning whose goal was less to convince others and more to arrive at a choice that satisfies one's own values and standards.

Reb (Reference Reb2005) showed that regret priming also reduced or eliminated other “reason-based” effects such as the compromise effect (Simonson Reference Simonson1989), in which an option is more desirable when presented as a compromise, and the accept/reject effect (Shafir & Tversky Reference Shafir and Tversky1992), in which the same option tends to be both rejected and selected. In all these reason-based choice effects, the justifying arguments do not withstand close scrutiny. They are simply “shallow but nice-sounding rationales” (Simonson Reference Simonson1989, p. 170) that might serve to convince an uncritical external audience but not one's thoughtful self. In contrast, regret priming did not reduce the most important attribute effect (Slovic Reference Slovic1975) where the justifying argument can reasonably be construed to both self and others as a legitimate tiebreaker between equally valued options (Reb Reference Reb2005).

Regret priming appears to involve both motivational and attention-directing effects, which are sometimes quite subtle. For example, Reb and Connolly (Reference Reb and Connolly2009) used unobtrusive priming of either outcome or self-blame regret in a repeated decision task where feedback on outcomes of unchosen options was offered. Subjects primed for outcome regret rejected such feedback more often, learned more slowly, and ultimately performed less well than those primed for self-blame regret (thus falling victim to the myopic regret aversion trap: avoiding short-term regret led them to experience greater long-term regret). Both groups were motivated to avoid regret, but one did so by avoiding painful comparisons, the other by following a justifiable decision process.

In summary we find persuasive M&S's case that reasoning is primarily for persuasive argumentation rather than for effective consequential choice. Given the evolutionary advantages of the latter, however, it is plausible that other systems may have developed to support such choice processes. A growing body of evidence suggests that mechanisms of regret, regret avoidance, and justification can serve such a decision-improving role. Specifically, aversion of process regret may complement the fluent argument maker and tweak it to pay more balanced attention to and weighing of the pros and cons associated with a decision problem. Because of the anticipatory nature of regret, attention may further be directed to future consequences that are predicted to impact experienced regret. Mechanisms of regret and justification thus suggest important linkages between the argument-rich communicator sketched by M&S and the purposive consequentialist demanded by rational choice models of human decisions. We see such evidence as dovetailing neatly with, and modestly extending, the findings compiled in the target article. Perhaps ironically, as the central role of reasoning in assuring good choices has come increasingly into doubt in recent decision research, emotions, earlier seen as an obstacle to effective decision making, are increasingly being found to perform crucial functions in facilitating such decisions.