In Regard for Reason in the Moral Mind, May (Reference May2018) acknowledges that moral judgments and behaviors are frequently produced by automatic intuitions. May argues that intuitive cognitive processing is best categorized as “reasoning” because it is not heavily dependent upon emotional responses. Thus, May aligns these intuitions with a rationalist (rather than sentimentalist) framework and suggests that these intuitions are not substantively threatened by debunking arguments. However, to successfully vindicate moral cognition on the grounds that it is rooted in reason, it is crucial to determine that intuitive moral cognition truly arises from inferential processes – ideally, those that move from well-justified premises to logically warranted conclusions. Otherwise, moral intuitions can more easily be dismissed, because debunking arguments rely primarily on the irrationality or unreliability of everyday moral judgments rather than on their emotionality (e.g., Sinnott-Armstrong Reference Sinnott-Armstrong2011). Therefore, regardless of whether emotions are the primary fuel for moral judgments and actions, it is crucial to determine the extent to which these judgments and actions are aligned with reason to prevent them from being discredited.
Moral cognition, like all cognition, involves information processing. However, the complexity of this processing can vary widely. Some moral evaluations result from careful consideration of clearly represented concepts, whereas others involve no internal representations and are therefore considerably more inflexible and error-prone (e.g., Crockett Reference Crockett2013; Cushman Reference Cushman2013). Therefore, even if moral competence can be described as operating in accordance with certain principles (e.g., intentionally causing harmful outcomes is morally worse than inadvertently allowing harm to occur), this is consistent with a range of psychological mechanisms ranging from reasoned inference to unreasoned instinct. Although the latter is not necessarily aligned with sentimentalism (as it may not be driven by emotional responding), it is also not clearly aligned with rationalism. Instead, many automatic intuitions defy this binary opposition and instead exist in a liminal space between these philosophical strongholds. A crucial empirical question is therefore raised: Are most moral intuitions produced by processes of inductive or deductive reasoning, or are they formed by less rational means?
Developmental psychology provides a crucial tool for assessing the rationality of moral intuitions, as it can uncover the sources of intuitive responding. Some cognitive developmental processes are clearly aligned with rationalism, whereas others can reveal moral intuitions to be independent of reasoning (see Shweder et al. Reference Shweder, Turiel, Much, Flavell and Ross1981). As May (Reference May2018) proposes, intuitive automaticity could eventually result from an extended rehearsal of conscious reasoning, just like a chess expert is able to spontaneously make adept moves after internalizing the careful thinking that she exerted across many previous games (see also Pizarro & Bloom Reference Pizarro and Bloom2003; Saltzstein & Kasachkoff Reference Saltzstein and Kasachkoff2004). However, this seems unlikely, as people are often unable to consciously recover the principles that underlie their moral judgments (e.g., Cushman et al. Reference Cushman, Young and Hauser2006; Rottman et al. Reference Rottman, Kelemen and Young2014), suggesting that these intuitions may have never been consciously produced to begin with. Alternatively, some developmental psychologists have argued that young children can acquire intuitive frameworks for moral reasoning as a result of rational inference (e.g., Rhodes & Wellman Reference Rhodes and Wellman2017). However, just because intuitions can result from rule-governed inferences does not mean that they typically do, and recent research on moral development has indicated that babies and children possess a wide array of adaptive moral predispositions that do not appear to be the result of rational inference (see Bloom Reference Bloom2013; Rottman & Young Reference Rottman, Young, Decety and Wheatley2015). Therefore, I suspect that children's (and plausibly adults’) moral competence can most accurately be described as occupying a middle ground between rationalism and sentimentalism.
From an evolutionary standpoint, it would be maladaptive to rely on one's logical reasoning abilities to reach moral conclusions, as reason would not necessarily converge upon beliefs that successfully promote social status and coordination (see Krebs Reference Krebs2008). Instead, it is likely that moral competence is primarily composed of innately prepared intuitions and learning mechanisms that are modulated by relevant environmental inputs during childhood (Rottman & Young Reference Rottman, Young, Decety and Wheatley2015). Recent evidence from research with infants and young children suggests that many morally relevant intuitions are in fact the nascent products of evolutionary adaptation. These intuitions exhibit signatures of evolved psychological traits, for example, being spontaneously acquired in ways that do not rely heavily on protracted learning (see Dunham et al. Reference Dunham, Baron and Banaji2008) and emerging so early in life that it is unlikely that they result from rational inference or relevant experiences (see Hamlin Reference Hamlin2013). Young children also think about morality in domain-specific ways, and these features of moral cognition that transcend domain-general reasoning tendencies appear suited to resolve adaptive problems related to sociality (Cummins Reference Cummins1996).
Other moral intuitions rely heavily upon individual learning and enculturation, but it is similarly unlikely that this acquisition process typically involves reasoning. Rather, children are prone to blind conformity in the moral domain and are predisposed to promiscuously moralize a wide range of actions upon brief exposure to normative behaviors (see Chudek & Henrich Reference Chudek and Henrich2011; Rakoczy & Schmidt Reference Rakoczy and Schmidt2013; Tomasello Reference Tomasello2016). A recent set of studies has indicated that learning new moral beliefs is not always a rational endeavor (Rottman et al. Reference Rottman, Young and Kelemen2017). This research failed to support a strong sentimentalist view, as incidentally elicited disgust was insufficient for producing moralization. However, children acquired novel moral beliefs in irrational and undiscerning ways. Participants were equally persuaded by “well-fitting” and “poor-fitting” explanations, suggesting that children do not attend to the rationality of the testimony they are provided during the process of forming new moral beliefs, and they often lacked the ability to reconstruct the processes leading to their formation of moral beliefs when learning from emotion-laden testimony. Of course, this research is not conclusive by itself, particularly as it does not align with theoretical perspectives that children should only learn from testimony that they discern to be appropriate and relevant (e.g., Grusec & Goodnow Reference Grusec and Goodnow1994; Nucci Reference Nucci1984; also see Sobel & Kushnir Reference Sobel and Kushnir2013), and considerably more research is needed to more fully understand typical processes of moral acquisition.
Turning from moral thought to moral behavior, children sometimes appear to be motivated by virtue; they are spontaneously prosocial in certain affiliative situations when they can help others at a small cost to themselves (e.g., Warneken & Tomasello Reference Warneken and Tomasello2006). However, this prosociality is selective and strategic (see Martin & Olson Reference Martin and Olson2015). In particular, when children stand to achieve a relative advantage, their behaviors are typically motivated by selfish gains. Even when they clearly understand how they should act in moral situations, they often choose to act in self-interested ways instead (see Blake et al. Reference Blake, McAuliffe and Warneken2014). Children are strongly motivated by appearing moral rather than by actually being moral (e.g., Engelmann et al. Reference Engelmann, Over, Herrmann and Tomasello2013; Leimgruber et al. Reference Leimgruber, Shaw, Santos and Olson2012; Shaw et al. Reference Shaw, Montinari, Piovesan, Olson, Gino and Norton2014), and it takes many years for them to begin to overcome these egocentric tendencies (to the extent that they succeed at all).
On the whole, a review of recent developmental research uncovers sparse evidence that rationalism successfully accounts for moral cognition in infants and children. Instead, there is reason to conclude that moral intuitions are often irrational. Children's moral intuitions are constrained by innate representational biases that are diversified through sociocultural learning, rather than actively formed through reasoned inferences about social interactions. Of course, biased intuitions that are similarly irrational, motivated, and inaccessible to introspection have also been argued to characterize much of adult moral cognition (e.g., Greene Reference Greene2013; Haidt Reference Haidt2001), but May (Reference May2018) argues that the evidence for these biases either falls short or is limited in scope. Developmental evidence has the potential to bolster the pessimists’ claims even further, however. First, studying development can rule out some alternative interpretations of automaticity (e.g., that it results from initial judicious deliberation). The processes leading to moral belief formation may be more generally defective than is evident from studies of adult moral cognition, thus surmounting the “Debunker's Dilemma.” Second, early development may be a time when motivations are particularly egocentric and situational, and thus poor motivations are sometimes sufficiently pervasive for surmounting the “Defeater's Dilemma.”
Overall, although I disagree with many of May's (Reference May2018) conclusions, I applaud the many redeeming qualities of this impressive treatise. Throughout its thorough consideration of a wide swath of evidence, this book provides an important counterweight to oppose the strong force of the sometimes overblown claims that morality is wholly driven by emotions and egoism rather than by reason and virtue. The sentimentalism that has largely taken hold in social psychological approaches to moral psychology (e.g., Haidt Reference Haidt2001) has sometimes obscured the cases in which reason can play at least a limited role in moral cognition (e.g., Holyoak & Powell Reference Holyoak and Powell2016; Paxton & Greene Reference Paxton and Greene2010; Pizarro & Bloom Reference Pizarro and Bloom2003). This emphasis on vapid emotional responding is reflective of a more general tendency to focus on the irrational, motivated, and biased nature of human thought that has prevailed in the field of social psychology as a whole (see Alter Reference Alter2013; Bargh Reference Bargh2017; Nisbett & Wilson Reference Nisbett and Wilson1977). On the contrary, many developmental psychologists have sought and often found evidence for the rational, scientific, and objective nature of children's thought (see Gopnik Reference Gopnik2012; Schulz Reference Schulz2012; Xu & Kushnir Reference Xu and Kushnir2013). In recent years, dozens of elegant studies have demonstrated that children can use scientist-like reasoning to form and revise beliefs (e.g., Schulz et al. Reference Schulz, Bonawitz and Griffiths2007; Sobel & Kirkham Reference Sobel and Kirkham2006), indicating that children rely heavily on reason in certain contexts. This characterization has also reigned in classical theories of moral development, which posit that moral judgments are produced by careful reflection (e.g., Kohlberg Reference Kohlberg and Mischel1971; Nucci & Turiel Reference Nucci and Turiel1978; Piaget Reference Piaget and Gabain1932; Smetana Reference Smetana, Killen and Smetana2006). However, just as adults are not as asinine as social psychologists often characterize them, children are not as astute as developmental psychologists often characterize them. This may be especially true in the moral domain, for which affiliative motivations tend to reign over truth-seeking and it is difficult (if not impossible) to construct knowledge through individually acting on the world.
Descriptively, there are myriad possibilities for characterizing the nature of moral cognition. As reviewed here, research in moral development has indicated that emotional forces do not ubiquitously drive moral evaluations and behaviors, but neither does careful inductive reasoning. There is an intermediate space between sentimentalism and rationalism that may most accurately characterize everyday moral psychology. Therefore, regardless of whether emotions are shown to be unnecessary or insufficient for moral development to occur, despite some arguments to the contrary (e.g., Eisenberg Reference Eisenberg2000; Hoffman Reference Hoffman1975; Kagan Reference Kagan, Kagan and Lamb1987), the veracity of rationalism would not necessarily hinge upon the success of these demonstrations. Even if sentimentalism is found to be empirically false, the unreasoned and heuristic nature of many moral intuitions prompts a cautious pessimism regarding the nature of moral cognition. While this stance is certainly less pleasant than optimism, it may be beneficial for avoiding complacency. A healthy dose of pessimism can serve as motivation for fostering a more humane world, perhaps by investigating ways to encourage future generations to overcome natural moral inclinations. By resisting the tendency to consider moral “truths” to be self-evident and by vigilantly entreating children to apply careful reasoning to crucial moral issues, it may be possible to nurture moral cognition in the direction of rationalism.
In Regard for Reason in the Moral Mind, May (Reference May2018) acknowledges that moral judgments and behaviors are frequently produced by automatic intuitions. May argues that intuitive cognitive processing is best categorized as “reasoning” because it is not heavily dependent upon emotional responses. Thus, May aligns these intuitions with a rationalist (rather than sentimentalist) framework and suggests that these intuitions are not substantively threatened by debunking arguments. However, to successfully vindicate moral cognition on the grounds that it is rooted in reason, it is crucial to determine that intuitive moral cognition truly arises from inferential processes – ideally, those that move from well-justified premises to logically warranted conclusions. Otherwise, moral intuitions can more easily be dismissed, because debunking arguments rely primarily on the irrationality or unreliability of everyday moral judgments rather than on their emotionality (e.g., Sinnott-Armstrong Reference Sinnott-Armstrong2011). Therefore, regardless of whether emotions are the primary fuel for moral judgments and actions, it is crucial to determine the extent to which these judgments and actions are aligned with reason to prevent them from being discredited.
Moral cognition, like all cognition, involves information processing. However, the complexity of this processing can vary widely. Some moral evaluations result from careful consideration of clearly represented concepts, whereas others involve no internal representations and are therefore considerably more inflexible and error-prone (e.g., Crockett Reference Crockett2013; Cushman Reference Cushman2013). Therefore, even if moral competence can be described as operating in accordance with certain principles (e.g., intentionally causing harmful outcomes is morally worse than inadvertently allowing harm to occur), this is consistent with a range of psychological mechanisms ranging from reasoned inference to unreasoned instinct. Although the latter is not necessarily aligned with sentimentalism (as it may not be driven by emotional responding), it is also not clearly aligned with rationalism. Instead, many automatic intuitions defy this binary opposition and instead exist in a liminal space between these philosophical strongholds. A crucial empirical question is therefore raised: Are most moral intuitions produced by processes of inductive or deductive reasoning, or are they formed by less rational means?
Developmental psychology provides a crucial tool for assessing the rationality of moral intuitions, as it can uncover the sources of intuitive responding. Some cognitive developmental processes are clearly aligned with rationalism, whereas others can reveal moral intuitions to be independent of reasoning (see Shweder et al. Reference Shweder, Turiel, Much, Flavell and Ross1981). As May (Reference May2018) proposes, intuitive automaticity could eventually result from an extended rehearsal of conscious reasoning, just like a chess expert is able to spontaneously make adept moves after internalizing the careful thinking that she exerted across many previous games (see also Pizarro & Bloom Reference Pizarro and Bloom2003; Saltzstein & Kasachkoff Reference Saltzstein and Kasachkoff2004). However, this seems unlikely, as people are often unable to consciously recover the principles that underlie their moral judgments (e.g., Cushman et al. Reference Cushman, Young and Hauser2006; Rottman et al. Reference Rottman, Kelemen and Young2014), suggesting that these intuitions may have never been consciously produced to begin with. Alternatively, some developmental psychologists have argued that young children can acquire intuitive frameworks for moral reasoning as a result of rational inference (e.g., Rhodes & Wellman Reference Rhodes and Wellman2017). However, just because intuitions can result from rule-governed inferences does not mean that they typically do, and recent research on moral development has indicated that babies and children possess a wide array of adaptive moral predispositions that do not appear to be the result of rational inference (see Bloom Reference Bloom2013; Rottman & Young Reference Rottman, Young, Decety and Wheatley2015). Therefore, I suspect that children's (and plausibly adults’) moral competence can most accurately be described as occupying a middle ground between rationalism and sentimentalism.
From an evolutionary standpoint, it would be maladaptive to rely on one's logical reasoning abilities to reach moral conclusions, as reason would not necessarily converge upon beliefs that successfully promote social status and coordination (see Krebs Reference Krebs2008). Instead, it is likely that moral competence is primarily composed of innately prepared intuitions and learning mechanisms that are modulated by relevant environmental inputs during childhood (Rottman & Young Reference Rottman, Young, Decety and Wheatley2015). Recent evidence from research with infants and young children suggests that many morally relevant intuitions are in fact the nascent products of evolutionary adaptation. These intuitions exhibit signatures of evolved psychological traits, for example, being spontaneously acquired in ways that do not rely heavily on protracted learning (see Dunham et al. Reference Dunham, Baron and Banaji2008) and emerging so early in life that it is unlikely that they result from rational inference or relevant experiences (see Hamlin Reference Hamlin2013). Young children also think about morality in domain-specific ways, and these features of moral cognition that transcend domain-general reasoning tendencies appear suited to resolve adaptive problems related to sociality (Cummins Reference Cummins1996).
Other moral intuitions rely heavily upon individual learning and enculturation, but it is similarly unlikely that this acquisition process typically involves reasoning. Rather, children are prone to blind conformity in the moral domain and are predisposed to promiscuously moralize a wide range of actions upon brief exposure to normative behaviors (see Chudek & Henrich Reference Chudek and Henrich2011; Rakoczy & Schmidt Reference Rakoczy and Schmidt2013; Tomasello Reference Tomasello2016). A recent set of studies has indicated that learning new moral beliefs is not always a rational endeavor (Rottman et al. Reference Rottman, Young and Kelemen2017). This research failed to support a strong sentimentalist view, as incidentally elicited disgust was insufficient for producing moralization. However, children acquired novel moral beliefs in irrational and undiscerning ways. Participants were equally persuaded by “well-fitting” and “poor-fitting” explanations, suggesting that children do not attend to the rationality of the testimony they are provided during the process of forming new moral beliefs, and they often lacked the ability to reconstruct the processes leading to their formation of moral beliefs when learning from emotion-laden testimony. Of course, this research is not conclusive by itself, particularly as it does not align with theoretical perspectives that children should only learn from testimony that they discern to be appropriate and relevant (e.g., Grusec & Goodnow Reference Grusec and Goodnow1994; Nucci Reference Nucci1984; also see Sobel & Kushnir Reference Sobel and Kushnir2013), and considerably more research is needed to more fully understand typical processes of moral acquisition.
Turning from moral thought to moral behavior, children sometimes appear to be motivated by virtue; they are spontaneously prosocial in certain affiliative situations when they can help others at a small cost to themselves (e.g., Warneken & Tomasello Reference Warneken and Tomasello2006). However, this prosociality is selective and strategic (see Martin & Olson Reference Martin and Olson2015). In particular, when children stand to achieve a relative advantage, their behaviors are typically motivated by selfish gains. Even when they clearly understand how they should act in moral situations, they often choose to act in self-interested ways instead (see Blake et al. Reference Blake, McAuliffe and Warneken2014). Children are strongly motivated by appearing moral rather than by actually being moral (e.g., Engelmann et al. Reference Engelmann, Over, Herrmann and Tomasello2013; Leimgruber et al. Reference Leimgruber, Shaw, Santos and Olson2012; Shaw et al. Reference Shaw, Montinari, Piovesan, Olson, Gino and Norton2014), and it takes many years for them to begin to overcome these egocentric tendencies (to the extent that they succeed at all).
On the whole, a review of recent developmental research uncovers sparse evidence that rationalism successfully accounts for moral cognition in infants and children. Instead, there is reason to conclude that moral intuitions are often irrational. Children's moral intuitions are constrained by innate representational biases that are diversified through sociocultural learning, rather than actively formed through reasoned inferences about social interactions. Of course, biased intuitions that are similarly irrational, motivated, and inaccessible to introspection have also been argued to characterize much of adult moral cognition (e.g., Greene Reference Greene2013; Haidt Reference Haidt2001), but May (Reference May2018) argues that the evidence for these biases either falls short or is limited in scope. Developmental evidence has the potential to bolster the pessimists’ claims even further, however. First, studying development can rule out some alternative interpretations of automaticity (e.g., that it results from initial judicious deliberation). The processes leading to moral belief formation may be more generally defective than is evident from studies of adult moral cognition, thus surmounting the “Debunker's Dilemma.” Second, early development may be a time when motivations are particularly egocentric and situational, and thus poor motivations are sometimes sufficiently pervasive for surmounting the “Defeater's Dilemma.”
Overall, although I disagree with many of May's (Reference May2018) conclusions, I applaud the many redeeming qualities of this impressive treatise. Throughout its thorough consideration of a wide swath of evidence, this book provides an important counterweight to oppose the strong force of the sometimes overblown claims that morality is wholly driven by emotions and egoism rather than by reason and virtue. The sentimentalism that has largely taken hold in social psychological approaches to moral psychology (e.g., Haidt Reference Haidt2001) has sometimes obscured the cases in which reason can play at least a limited role in moral cognition (e.g., Holyoak & Powell Reference Holyoak and Powell2016; Paxton & Greene Reference Paxton and Greene2010; Pizarro & Bloom Reference Pizarro and Bloom2003). This emphasis on vapid emotional responding is reflective of a more general tendency to focus on the irrational, motivated, and biased nature of human thought that has prevailed in the field of social psychology as a whole (see Alter Reference Alter2013; Bargh Reference Bargh2017; Nisbett & Wilson Reference Nisbett and Wilson1977). On the contrary, many developmental psychologists have sought and often found evidence for the rational, scientific, and objective nature of children's thought (see Gopnik Reference Gopnik2012; Schulz Reference Schulz2012; Xu & Kushnir Reference Xu and Kushnir2013). In recent years, dozens of elegant studies have demonstrated that children can use scientist-like reasoning to form and revise beliefs (e.g., Schulz et al. Reference Schulz, Bonawitz and Griffiths2007; Sobel & Kirkham Reference Sobel and Kirkham2006), indicating that children rely heavily on reason in certain contexts. This characterization has also reigned in classical theories of moral development, which posit that moral judgments are produced by careful reflection (e.g., Kohlberg Reference Kohlberg and Mischel1971; Nucci & Turiel Reference Nucci and Turiel1978; Piaget Reference Piaget and Gabain1932; Smetana Reference Smetana, Killen and Smetana2006). However, just as adults are not as asinine as social psychologists often characterize them, children are not as astute as developmental psychologists often characterize them. This may be especially true in the moral domain, for which affiliative motivations tend to reign over truth-seeking and it is difficult (if not impossible) to construct knowledge through individually acting on the world.
Descriptively, there are myriad possibilities for characterizing the nature of moral cognition. As reviewed here, research in moral development has indicated that emotional forces do not ubiquitously drive moral evaluations and behaviors, but neither does careful inductive reasoning. There is an intermediate space between sentimentalism and rationalism that may most accurately characterize everyday moral psychology. Therefore, regardless of whether emotions are shown to be unnecessary or insufficient for moral development to occur, despite some arguments to the contrary (e.g., Eisenberg Reference Eisenberg2000; Hoffman Reference Hoffman1975; Kagan Reference Kagan, Kagan and Lamb1987), the veracity of rationalism would not necessarily hinge upon the success of these demonstrations. Even if sentimentalism is found to be empirically false, the unreasoned and heuristic nature of many moral intuitions prompts a cautious pessimism regarding the nature of moral cognition. While this stance is certainly less pleasant than optimism, it may be beneficial for avoiding complacency. A healthy dose of pessimism can serve as motivation for fostering a more humane world, perhaps by investigating ways to encourage future generations to overcome natural moral inclinations. By resisting the tendency to consider moral “truths” to be self-evident and by vigilantly entreating children to apply careful reasoning to crucial moral issues, it may be possible to nurture moral cognition in the direction of rationalism.