A number of claims are made in the target article that are difficult to dispute. People do appear to be skilled at everyday argumentation while struggling to solve reasoning problems presented in the psychological laboratory. There is a great deal of evidence in the psychology of reasoning and decision making–as the authors demonstrate with admirable scholarship–that people frequently use reasoning to justify intuitively cued responses. And of course it is much easier to argue that people evolved skills of argumentation than the ability to solve reasoning problems. However, as is commonly observed by evolutionary theorists, a facility which evolves for one function may become “exapted” for another. Hence, my only concern is that the authors may have used their own formidable skills of argumentation to overstate the case against explicit reasoning as a tool for rational thinking and decision making.
As Mercier and Sperber (M&S) note, their analysis has implications for the now ubiquitous dual-process theories of higher cognition (Evans Reference Evans2008). It is interesting to note that the very first such theory in the psychology of reasoning was consistent with the authors' proposals. Wason and Evans (Reference Wason and Evans1975; Evans & Wason Reference Evans and Wason1976) distinguished between type 1 (intuitive) processes, which unconsciously cued the relevance of certain cards on the Wason selection task, and type 2 (explicit reasoning) processes, which served to confabulate justifications for these choices. Although it seems clear now that people do reason on the selection task, there is strong evidence that this is focused on finding justifications for unconscious cued cards. However, where such a justification cannot be found, recent evidence indicates that participants may withhold an intuitively prompted response (Evans & Ball Reference Evans and Ball2010). Hence, even in this case, reasoning is performing some beneficial role in decision making.
M&S are also correct to say that in more recent years dual-process theorists have emphasised the efficacy of type 2 processing in problem solving and decision making, especially for novel problems (e.g., Evans Reference Evans2010; Evans & Over Reference Evans and Over1996; Kahneman & Frederick Reference Kahneman, Frederick, Gilovich, Griffin and Kahneman2002; Stanovich Reference Stanovich1999; Reference Stanovich2010). In particular, theorists have emphasized the ability of some individuals, especially those of higher cognitive ability, to engage in hypothetical thinking and mental simulation, decoupling their actual beliefs in order to support suppositional reasoning (Evans Reference Evans2007; Stanovich Reference Stanovich2010). Such theorists attribute the unique cognitive and cultural achievements of the human species to the development of a second system of cognition or a new mind, largely unavailable to other animals. Moreover, it is proposed that while both new and old minds have instrumental rationality, they pursue different goals by different cognitive mechanisms (Evans Reference Evans2010; Stanovich Reference Stanovich2004). In particular, the old mind (broadly shared with higher animals) is driven by the past, replicating by evolution or individual learning past successes, whereas the new mind (distinctively human) can conduct mental simulations and reason consequentially, anticipating the future.
M&S deal with this issue rather dismissively, stating that “the possibility to deal with novelty and to anticipate the future is less a characterization of reasoning than it is of learning” (sect. 1.2, para. 3), going on to argue that the occasional successes of reasoning in helping us to solve problems would be insufficient to explain its evolution. It is possible to agree with the latter claim while strongly disputing the former. First, the learning mechanisms of the old mind do only enable us to repeat what has worked in the past. For novel problems which require simulation of future events, such a mechanism is largely useless. And while we are admittedly fairly poor at consequential decision making (Baron Reference Baron1994) we can do it. Since 1945, politicians have so far resisted their traditional practice of applying their most powerful weapons to their enemies when such weapons are nuclear, presumably because of the anticipated catastrophic consequences. And while it is debatable whether the human race will in fact avoid the disastrous effects of climate change, it could hardly be denied that the world is currently engaged in detailed hypothetical thinking (aided by computer models) about the future of the planet, and that much current policy is being driven by this. No other animal can even remotely think and act in this way.
M&S, as other evolutionary psychologists before them, assert the domain-specific nature of human reasoning and apparently overlook the importance of heritable general intelligence (for detailed critique of such arguments, see Stanovich Reference Stanovich2004; Stanovich & West Reference Stanovich, West and Over2003). Research on dual-process theory has provided a mass of evidence that the ability to solve novel reasoning and decision problems is related to individual differences in general intelligence, working-memory capacity, and other (highly correlated) measures of cognitive ability. Solving novel problems is also related to rational thinking dispositions, which help to determine (Stanovich Reference Stanovich2009; Reference Stanovich2010) the tendency to engage reflective reasoning rather than rely on intuition (see also Frederick Reference Frederick2005). Such findings fly in the face of fashionable claims (e.g., Dijksterhuis et al. Reference Dijksterhuis, Bos, Nordgren and van Baaren2006b; Gigerenzer Reference Gigerenzer2007; Gladwell Reference Gladwell2005) that we are better off relying on intuition than engaging our powers of reasoning.
The difficulty which such advocates of intuition have faced is to explain why humans evolved a capacity for reasoning which is best not trusted. M&S attempt to fill that gap in the target article, thus supporting what I believe to be a dangerously flawed line of reasoning about human intelligence. It is not necessary to follow them down this path in order to respect the integrity of their arguments about the evolution of reasoning via argumentation. Unique human abilities for reflective thought have required the evolution of a number of facilities, including language, metarepresentation, and large forebrains, none of which could plausibly have been driven by some Darwinian need for a new mind. If there were such a driver, surely other animals would have evolved human-like intelligence. It is more plausible to argue that the new mind was an evolutionary accident, in which case an exapted ability for reasoning derived from argumentation may well be part of that story.
A number of claims are made in the target article that are difficult to dispute. People do appear to be skilled at everyday argumentation while struggling to solve reasoning problems presented in the psychological laboratory. There is a great deal of evidence in the psychology of reasoning and decision making–as the authors demonstrate with admirable scholarship–that people frequently use reasoning to justify intuitively cued responses. And of course it is much easier to argue that people evolved skills of argumentation than the ability to solve reasoning problems. However, as is commonly observed by evolutionary theorists, a facility which evolves for one function may become “exapted” for another. Hence, my only concern is that the authors may have used their own formidable skills of argumentation to overstate the case against explicit reasoning as a tool for rational thinking and decision making.
As Mercier and Sperber (M&S) note, their analysis has implications for the now ubiquitous dual-process theories of higher cognition (Evans Reference Evans2008). It is interesting to note that the very first such theory in the psychology of reasoning was consistent with the authors' proposals. Wason and Evans (Reference Wason and Evans1975; Evans & Wason Reference Evans and Wason1976) distinguished between type 1 (intuitive) processes, which unconsciously cued the relevance of certain cards on the Wason selection task, and type 2 (explicit reasoning) processes, which served to confabulate justifications for these choices. Although it seems clear now that people do reason on the selection task, there is strong evidence that this is focused on finding justifications for unconscious cued cards. However, where such a justification cannot be found, recent evidence indicates that participants may withhold an intuitively prompted response (Evans & Ball Reference Evans and Ball2010). Hence, even in this case, reasoning is performing some beneficial role in decision making.
M&S are also correct to say that in more recent years dual-process theorists have emphasised the efficacy of type 2 processing in problem solving and decision making, especially for novel problems (e.g., Evans Reference Evans2010; Evans & Over Reference Evans and Over1996; Kahneman & Frederick Reference Kahneman, Frederick, Gilovich, Griffin and Kahneman2002; Stanovich Reference Stanovich1999; Reference Stanovich2010). In particular, theorists have emphasized the ability of some individuals, especially those of higher cognitive ability, to engage in hypothetical thinking and mental simulation, decoupling their actual beliefs in order to support suppositional reasoning (Evans Reference Evans2007; Stanovich Reference Stanovich2010). Such theorists attribute the unique cognitive and cultural achievements of the human species to the development of a second system of cognition or a new mind, largely unavailable to other animals. Moreover, it is proposed that while both new and old minds have instrumental rationality, they pursue different goals by different cognitive mechanisms (Evans Reference Evans2010; Stanovich Reference Stanovich2004). In particular, the old mind (broadly shared with higher animals) is driven by the past, replicating by evolution or individual learning past successes, whereas the new mind (distinctively human) can conduct mental simulations and reason consequentially, anticipating the future.
M&S deal with this issue rather dismissively, stating that “the possibility to deal with novelty and to anticipate the future is less a characterization of reasoning than it is of learning” (sect. 1.2, para. 3), going on to argue that the occasional successes of reasoning in helping us to solve problems would be insufficient to explain its evolution. It is possible to agree with the latter claim while strongly disputing the former. First, the learning mechanisms of the old mind do only enable us to repeat what has worked in the past. For novel problems which require simulation of future events, such a mechanism is largely useless. And while we are admittedly fairly poor at consequential decision making (Baron Reference Baron1994) we can do it. Since 1945, politicians have so far resisted their traditional practice of applying their most powerful weapons to their enemies when such weapons are nuclear, presumably because of the anticipated catastrophic consequences. And while it is debatable whether the human race will in fact avoid the disastrous effects of climate change, it could hardly be denied that the world is currently engaged in detailed hypothetical thinking (aided by computer models) about the future of the planet, and that much current policy is being driven by this. No other animal can even remotely think and act in this way.
M&S, as other evolutionary psychologists before them, assert the domain-specific nature of human reasoning and apparently overlook the importance of heritable general intelligence (for detailed critique of such arguments, see Stanovich Reference Stanovich2004; Stanovich & West Reference Stanovich, West and Over2003). Research on dual-process theory has provided a mass of evidence that the ability to solve novel reasoning and decision problems is related to individual differences in general intelligence, working-memory capacity, and other (highly correlated) measures of cognitive ability. Solving novel problems is also related to rational thinking dispositions, which help to determine (Stanovich Reference Stanovich2009; Reference Stanovich2010) the tendency to engage reflective reasoning rather than rely on intuition (see also Frederick Reference Frederick2005). Such findings fly in the face of fashionable claims (e.g., Dijksterhuis et al. Reference Dijksterhuis, Bos, Nordgren and van Baaren2006b; Gigerenzer Reference Gigerenzer2007; Gladwell Reference Gladwell2005) that we are better off relying on intuition than engaging our powers of reasoning.
The difficulty which such advocates of intuition have faced is to explain why humans evolved a capacity for reasoning which is best not trusted. M&S attempt to fill that gap in the target article, thus supporting what I believe to be a dangerously flawed line of reasoning about human intelligence. It is not necessary to follow them down this path in order to respect the integrity of their arguments about the evolution of reasoning via argumentation. Unique human abilities for reflective thought have required the evolution of a number of facilities, including language, metarepresentation, and large forebrains, none of which could plausibly have been driven by some Darwinian need for a new mind. If there were such a driver, surely other animals would have evolved human-like intelligence. It is more plausible to argue that the new mind was an evolutionary accident, in which case an exapted ability for reasoning derived from argumentation may well be part of that story.