A major claim underpinning the approach taken by Baumard and colleagues is that partner choice (in which social agents choose mutualistic partners and advertise their cooperativeness) plays a critical role in the emergence of human cooperation and fairness. In particular, partner choice may be more important than partner control (in which agents decide whether to cooperate or defect in a dyadic situation). These claims mainly derive from data with human adults. However, given that this moral sense is supposed to be unique to our species, it is important to understand the evolutionary changes and ontogenetic origins of these behaviors, as well.
In fact, evidence from both human children and other species suggests that partner choice is a fundamental mechanism shaping cooperation both across ontogeny and in other species. In nonhuman primates, observational and experimental studies provide abundant evidence that individuals engage in long-term reciprocal relationships that result from seeking out other cooperators (Schino & Aureli Reference Schino and Aureli2010). For example, chimpanzees selectively choose skillful over unskillful cooperators for a mutualistic task (Melis et al. Reference Melis, Hare and Tomasello2006) and choose a partner who had chosen them previously over one who ignored them (Melis et al. Reference Melis, Hare and Tomasello2008). In contrast, evidence for reciprocal exchanges in which individuals temporally modulate their cooperation within a dyad contingent upon the partner's prior behavior (such as tit-for-tat) is weak to nonexistent (Hammerstein Reference Hammerstein and Hammerstein2003). In human infants, a similar pattern has emerged. In the first few years of life, children begin to differentiate between cooperators and defectors (Kuhlmeier et al. Reference Kuhlmeier, Wynn and Bloom2003), show a preference for cooperators over defectors (Hamlin et al. Reference Hamlin, Wynn and Bloom2007), and tend to cooperate with cooperators over defectors (Dunfield & Kuhlmeier Reference Dunfield and Kuhlmeier2010). However, temporally contingent reciprocity in a dyadic relationship seems to emerge much later: Children do not begin to selectively decrease or increase their giving in response to what they received from a partner until 3.5 years of age (Warneken & Tomasello Reference Warneken and Tomasello2009a).
Together, these data suggest that both nonhuman and human primates might be better equipped for partner choice than for partner control. On the one hand, this seems to support the claim by Baumard et al. that partner choice is an important mechanism supporting cooperative activities more generally. However, this also raises a major challenge to this model's explanatory power in illuminating human cooperation and morality more specifically. If nonhuman primates also engage in mutually beneficial interactions and seek out other good cooperators, why does this not scale up to a “full-fledged moral sense” (sect. 4, para. 2) characterizing humans? Thus, while morality may be a “consequence” of mutualistic cooperation that includes social selection (see target article, sect. 2.1.1, para. 4), this seems unlikely to be the full story, given these comparative and developmental findings. In general, this suggests that some other factors are necessary to explain human-like morality beyond mutualism and partner choice.
What might account for the emergence of the moral systems that we see in humans? I suggest that one relevant feature is the coupling of fairness norms with concerns for other people's welfare. As Baumard et al. suggest, merit-based principles (based on assessments of work contributions) might emerge from the dynamics of selecting partners and divvying up the resulting benefit of mutualistic interactions. However, this does not appear to account for distributive justice more broadly construed. That is, moral considerations in the domain of resource sharing are not restricted to merit alone, but also can be used to improve the situation of disadvantaged individuals. This distinction is already important in the domain of mutually beneficial cooperative interactions that are the focus here. In addition, they become crucial in situations in which individuals must decide whether to share resources with unrelated individuals who are prevented from engaging in such mutualistic interactions in the first place. Prescriptive theories of justice try to account for this situation. For example, Rawls' difference principle suggests that not everything should be left to talent and effort: inequalities are permissible if they accrue benefit to the disadvantaged (Rawls Reference Rawls1971). Moreover, descriptive models of adult behavior suggest that people's reasoning and behavior do not only concern equitable distributions, but also involve adjustments based upon others need (Deutsch Reference Deutsch1975). Such processes where fair distributions account for other's needs, moreover, are often fueled by empathy and sympathy with the welfare of others (Hoffman Reference Hoffman2000). Along these lines, developmental studies indicate that children progress through a developmental sequence reflecting the integration of these different principles: Younger children focus on strict equality and individual work contributions, but older children make need-based adjustments (Damon Reference Damon1977). In conclusion, it seems that the essence of genuinely moral behavior in humans is to partly override mutualistic strategism, which poses a challenge for the current model to integrate this characteristic of human behavior.
A major claim underpinning the approach taken by Baumard and colleagues is that partner choice (in which social agents choose mutualistic partners and advertise their cooperativeness) plays a critical role in the emergence of human cooperation and fairness. In particular, partner choice may be more important than partner control (in which agents decide whether to cooperate or defect in a dyadic situation). These claims mainly derive from data with human adults. However, given that this moral sense is supposed to be unique to our species, it is important to understand the evolutionary changes and ontogenetic origins of these behaviors, as well.
In fact, evidence from both human children and other species suggests that partner choice is a fundamental mechanism shaping cooperation both across ontogeny and in other species. In nonhuman primates, observational and experimental studies provide abundant evidence that individuals engage in long-term reciprocal relationships that result from seeking out other cooperators (Schino & Aureli Reference Schino and Aureli2010). For example, chimpanzees selectively choose skillful over unskillful cooperators for a mutualistic task (Melis et al. Reference Melis, Hare and Tomasello2006) and choose a partner who had chosen them previously over one who ignored them (Melis et al. Reference Melis, Hare and Tomasello2008). In contrast, evidence for reciprocal exchanges in which individuals temporally modulate their cooperation within a dyad contingent upon the partner's prior behavior (such as tit-for-tat) is weak to nonexistent (Hammerstein Reference Hammerstein and Hammerstein2003). In human infants, a similar pattern has emerged. In the first few years of life, children begin to differentiate between cooperators and defectors (Kuhlmeier et al. Reference Kuhlmeier, Wynn and Bloom2003), show a preference for cooperators over defectors (Hamlin et al. Reference Hamlin, Wynn and Bloom2007), and tend to cooperate with cooperators over defectors (Dunfield & Kuhlmeier Reference Dunfield and Kuhlmeier2010). However, temporally contingent reciprocity in a dyadic relationship seems to emerge much later: Children do not begin to selectively decrease or increase their giving in response to what they received from a partner until 3.5 years of age (Warneken & Tomasello Reference Warneken and Tomasello2009a).
Together, these data suggest that both nonhuman and human primates might be better equipped for partner choice than for partner control. On the one hand, this seems to support the claim by Baumard et al. that partner choice is an important mechanism supporting cooperative activities more generally. However, this also raises a major challenge to this model's explanatory power in illuminating human cooperation and morality more specifically. If nonhuman primates also engage in mutually beneficial interactions and seek out other good cooperators, why does this not scale up to a “full-fledged moral sense” (sect. 4, para. 2) characterizing humans? Thus, while morality may be a “consequence” of mutualistic cooperation that includes social selection (see target article, sect. 2.1.1, para. 4), this seems unlikely to be the full story, given these comparative and developmental findings. In general, this suggests that some other factors are necessary to explain human-like morality beyond mutualism and partner choice.
What might account for the emergence of the moral systems that we see in humans? I suggest that one relevant feature is the coupling of fairness norms with concerns for other people's welfare. As Baumard et al. suggest, merit-based principles (based on assessments of work contributions) might emerge from the dynamics of selecting partners and divvying up the resulting benefit of mutualistic interactions. However, this does not appear to account for distributive justice more broadly construed. That is, moral considerations in the domain of resource sharing are not restricted to merit alone, but also can be used to improve the situation of disadvantaged individuals. This distinction is already important in the domain of mutually beneficial cooperative interactions that are the focus here. In addition, they become crucial in situations in which individuals must decide whether to share resources with unrelated individuals who are prevented from engaging in such mutualistic interactions in the first place. Prescriptive theories of justice try to account for this situation. For example, Rawls' difference principle suggests that not everything should be left to talent and effort: inequalities are permissible if they accrue benefit to the disadvantaged (Rawls Reference Rawls1971). Moreover, descriptive models of adult behavior suggest that people's reasoning and behavior do not only concern equitable distributions, but also involve adjustments based upon others need (Deutsch Reference Deutsch1975). Such processes where fair distributions account for other's needs, moreover, are often fueled by empathy and sympathy with the welfare of others (Hoffman Reference Hoffman2000). Along these lines, developmental studies indicate that children progress through a developmental sequence reflecting the integration of these different principles: Younger children focus on strict equality and individual work contributions, but older children make need-based adjustments (Damon Reference Damon1977). In conclusion, it seems that the essence of genuinely moral behavior in humans is to partly override mutualistic strategism, which poses a challenge for the current model to integrate this characteristic of human behavior.
ACKNOWLEDGMENTS
I thank Alexandra Rosati for helpful comments.