Baumard et al. argue that partner choice leads to fairness and mutualism, and that these form the basis for morality. While I very much agree with the focus on partner choice as a key driving force, I believe that mutualism takes us only so far, and that moral behaviour is better viewed in a signalling context. Through competitive signalling, evolution may lead to individuals displaying traits that are more generous than others, not just acting in a way that is the equal of others. Strategic investment in behaviours that benefit others so that actors can be seen as desirable partners may at least partly explain what we see as moral behaviour.
The basis for my argument lies in the theory of “competitive altruism” (Roberts Reference Roberts1998) or reputation-based partner choice. Baumard et al. do refer to this work, and to the evidence for competitive altruism. Indeed, the hypothesis that “individuals may compete for the most altruistic partners and non-altruists may become ostracized” (Roberts Reference Roberts1998, p.427) is very close to the arguments they develop (with altruism here implying a short-term cost, in common with usage in the term “reciprocal altruism” and with usage in the fields of psychology and behavioural economics). However, I believe they miss a crucial distinction. A defining feature of competitive altruism is that it explicitly envisages a two-stage process, whereby individuals first build up cooperative reputations and then choose partners for further interactions. The rationale is that if individuals benefit by attracting a cooperative partner for the second stage, then it may pay to display generosity in the first stage. This two-stage structure is important because it means that generosity can go beyond mutualism. This contrasts with Baumard et al.'s assumed structure in which there are envisaged to be two-way exchanges with fair outcomes.
For any non-kin cooperation, there must be a correlation such that cooperators receive more cooperation in return. In reciprocal altruism, this relationship is supplied by discrimination such that we only give to those who give to us (direct reciprocity) or to others (indirect reciprocity). It is these types of matching relationships on which Baumard et al. focus, but competitive altruism goes beyond reciprocity. In contrast, giving in competitive altruism can be unconditional.
Competitive altruism is based on three assumptions: (1) that individuals differ in quality as potential partners; (2) that public behaviour provides a reliable guide to another individual's future behaviour; and (3) that individuals pair up through mutual partner choice. From these assumptions, the theory infers that assortative partner choice will drive competition to be seen as a desirable partner. The framework combines theoretical work on the correlation between generosity and choosiness (Sherratt & Roberts Reference Sherratt and Roberts1998) with models of mutual partner choice (Johnstone Reference Johnstone1997). Honest signalling (Maynard Smith & Harper Reference Maynard Smith and Harper2003) and market effects (Noë & Hammerstein Reference Noë and Hammerstein1994) may also play a role. Evidence now shows that a strategy of investing in a cooperative reputation can indeed reap rewards, in that the highest contributors to a public goods game obtain the most profitable subsequent partnerships (Sylwester & Roberts Reference Sylwester and Roberts2010).
While competitive altruism is a theory of costly, cooperative behaviour, it has been developed in relation to morality (Van Vugt et al. Reference Van Vugt, Roberts, Hardy, Dunbar and Barrett2007). As Baumard et al. say, humans don't just cooperate, they have a sense that this is the right thing to do. This is what lies behind our use of the term “moral altruism” (Van Vugt et al. Reference Van Vugt, Roberts, Hardy, Dunbar and Barrett2007): we don't just cooperate but reward altruists and punish non-altruists. “Moralistic” punishment of defectors is itself a behaviour which contributes to reputation (Barclay Reference Barclay2006; Kurzban et al. Reference Kurzban, DeScioli and O'Brien2007). Like altruism itself, moral altruism is costly, and as with other costly displays, sexual selection may well play a role in driving the evolution of morality through partner choice for traits that honestly indicate underlying quality (Miller Reference Miller2007).
Indirect reciprocity has also been developed in the context of morality, and indeed Alexander's thesis on this (Alexander Reference Alexander1987) is notable by its absence from Baumard et al.'s discussion. The “moral rules” (Sigmund Reference Sigmund2012) which emerge from models essentially specify how we should treat others. They specify who is deserving of our cooperation, as opposed to how we ourselves benefit from cooperating. In consequence, questions remain about when such rules will be evolutionarily stable, given the tension between the need for discrimination rules (otherwise defectors receive as much as cooperators) and what rules best increase an actor's own reputation (which in the case of image scoring may be indiscriminate giving; Leimar & Hammerstein Reference Leimar and Hammerstein2001; Milinski et al. Reference Milinski, Semmann, Bakker and Krambeck2001; Nowak & Sigmund Reference Nowak and Sigmund2005). Furthermore, there are questions about how well a theory based on scenarios in which donors never meet recipients again can be applied to human societies based on group living (Roberts Reference Roberts2008).
An evolutionary explanation for morality must explain why moral traits are favoured by selection. To some extent a rule of doing unto others as one would do for oneself could be partly explained by one's own welfare being linked to that of others. In other words, there may be an element of direct self-interest due to the stake we have in others' welfare (Roberts Reference Roberts2005). However, conflicts of interest are widespread, so this can take us only so far. Mutualism, as championed by Baumard et al., takes us further, and could form the basis for a sense of fairness. However, being moral can go beyond mutualism and fairness and may be more profitably viewed as a display. As such, those who behave in a more moral way may be reaping rewards through being selected as desirable partners. It doesn't have to be fair, provided they are making a strategic investment in future benefits. Consider charitable donations. These may be moral, but they have more to do with signalling (Bereczkei et al. Reference Bereczkei, Birkas and Kerekes2007; Lyle et al. Reference Lyle, Smith and Sullivan2009) than fairness. As long as moral individuals are in demand, moral conduct can pay – which is surely a hopeful message.
You have
Access
Baumard et al. argue that partner choice leads to fairness and mutualism, and that these form the basis for morality. While I very much agree with the focus on partner choice as a key driving force, I believe that mutualism takes us only so far, and that moral behaviour is better viewed in a signalling context. Through competitive signalling, evolution may lead to individuals displaying traits that are more generous than others, not just acting in a way that is the equal of others. Strategic investment in behaviours that benefit others so that actors can be seen as desirable partners may at least partly explain what we see as moral behaviour.
The basis for my argument lies in the theory of “competitive altruism” (Roberts Reference Roberts1998) or reputation-based partner choice. Baumard et al. do refer to this work, and to the evidence for competitive altruism. Indeed, the hypothesis that “individuals may compete for the most altruistic partners and non-altruists may become ostracized” (Roberts Reference Roberts1998, p.427) is very close to the arguments they develop (with altruism here implying a short-term cost, in common with usage in the term “reciprocal altruism” and with usage in the fields of psychology and behavioural economics). However, I believe they miss a crucial distinction. A defining feature of competitive altruism is that it explicitly envisages a two-stage process, whereby individuals first build up cooperative reputations and then choose partners for further interactions. The rationale is that if individuals benefit by attracting a cooperative partner for the second stage, then it may pay to display generosity in the first stage. This two-stage structure is important because it means that generosity can go beyond mutualism. This contrasts with Baumard et al.'s assumed structure in which there are envisaged to be two-way exchanges with fair outcomes.
For any non-kin cooperation, there must be a correlation such that cooperators receive more cooperation in return. In reciprocal altruism, this relationship is supplied by discrimination such that we only give to those who give to us (direct reciprocity) or to others (indirect reciprocity). It is these types of matching relationships on which Baumard et al. focus, but competitive altruism goes beyond reciprocity. In contrast, giving in competitive altruism can be unconditional.
Competitive altruism is based on three assumptions: (1) that individuals differ in quality as potential partners; (2) that public behaviour provides a reliable guide to another individual's future behaviour; and (3) that individuals pair up through mutual partner choice. From these assumptions, the theory infers that assortative partner choice will drive competition to be seen as a desirable partner. The framework combines theoretical work on the correlation between generosity and choosiness (Sherratt & Roberts Reference Sherratt and Roberts1998) with models of mutual partner choice (Johnstone Reference Johnstone1997). Honest signalling (Maynard Smith & Harper Reference Maynard Smith and Harper2003) and market effects (Noë & Hammerstein Reference Noë and Hammerstein1994) may also play a role. Evidence now shows that a strategy of investing in a cooperative reputation can indeed reap rewards, in that the highest contributors to a public goods game obtain the most profitable subsequent partnerships (Sylwester & Roberts Reference Sylwester and Roberts2010).
While competitive altruism is a theory of costly, cooperative behaviour, it has been developed in relation to morality (Van Vugt et al. Reference Van Vugt, Roberts, Hardy, Dunbar and Barrett2007). As Baumard et al. say, humans don't just cooperate, they have a sense that this is the right thing to do. This is what lies behind our use of the term “moral altruism” (Van Vugt et al. Reference Van Vugt, Roberts, Hardy, Dunbar and Barrett2007): we don't just cooperate but reward altruists and punish non-altruists. “Moralistic” punishment of defectors is itself a behaviour which contributes to reputation (Barclay Reference Barclay2006; Kurzban et al. Reference Kurzban, DeScioli and O'Brien2007). Like altruism itself, moral altruism is costly, and as with other costly displays, sexual selection may well play a role in driving the evolution of morality through partner choice for traits that honestly indicate underlying quality (Miller Reference Miller2007).
Indirect reciprocity has also been developed in the context of morality, and indeed Alexander's thesis on this (Alexander Reference Alexander1987) is notable by its absence from Baumard et al.'s discussion. The “moral rules” (Sigmund Reference Sigmund2012) which emerge from models essentially specify how we should treat others. They specify who is deserving of our cooperation, as opposed to how we ourselves benefit from cooperating. In consequence, questions remain about when such rules will be evolutionarily stable, given the tension between the need for discrimination rules (otherwise defectors receive as much as cooperators) and what rules best increase an actor's own reputation (which in the case of image scoring may be indiscriminate giving; Leimar & Hammerstein Reference Leimar and Hammerstein2001; Milinski et al. Reference Milinski, Semmann, Bakker and Krambeck2001; Nowak & Sigmund Reference Nowak and Sigmund2005). Furthermore, there are questions about how well a theory based on scenarios in which donors never meet recipients again can be applied to human societies based on group living (Roberts Reference Roberts2008).
An evolutionary explanation for morality must explain why moral traits are favoured by selection. To some extent a rule of doing unto others as one would do for oneself could be partly explained by one's own welfare being linked to that of others. In other words, there may be an element of direct self-interest due to the stake we have in others' welfare (Roberts Reference Roberts2005). However, conflicts of interest are widespread, so this can take us only so far. Mutualism, as championed by Baumard et al., takes us further, and could form the basis for a sense of fairness. However, being moral can go beyond mutualism and fairness and may be more profitably viewed as a display. As such, those who behave in a more moral way may be reaping rewards through being selected as desirable partners. It doesn't have to be fair, provided they are making a strategic investment in future benefits. Consider charitable donations. These may be moral, but they have more to do with signalling (Bereczkei et al. Reference Bereczkei, Birkas and Kerekes2007; Lyle et al. Reference Lyle, Smith and Sullivan2009) than fairness. As long as moral individuals are in demand, moral conduct can pay – which is surely a hopeful message.