The mystery of human intergroup aggression has been studied in various fields of social sciences, from anthropology to political sciences. Previous psychological research mainly focused on attacking behaviour, leaving the psychology of defence considerably neglected. Although experimental economic games can be used to separately measure attack- and defence-based interactions between multiple individuals, research has so far not focused on the asymmetry that results when attack and defence co-occur, as they most often do. In this sense, the Attacker-Defender Game (AD-G) and IAD-C offer significant methodological and theoretical developments. In this commentary, we mainly focus on how these games can be used to uncover the mystery of the evolution of out-group aggression. We discuss their future directions, as well as their limitations.
Modern investigations of intergroup behaviour rely on integrating theories from evolutionary sciences with economic games. Particular focus is given to the human tendency to cooperate with in-group members and display aggression towards out-group members. Here, we will focus on the co-evolution model (CM; Choi & Bowles Reference Choi and Bowles2007). The CM suggests that frequent intergroup conflict in human evolutionary and cultural past caused parochial cooperation, which psychologically co-occurs with out-group aggression. In other words, at its core, the CM predicts that humans should be unconditionally motivated to not only increase their own group's standing, but also decrease the standing of other groups. Many subsequent studies attempted to test this prediction using various experimental paradigms, from third-party punishment (Bernhard et al. Reference Bernhard, Fischbacher and Fehr2006), intergroup prisoner's dilemma – maximizing difference (Halevy et al. Reference Halevy, Bornstein and Sagiv2008), and the pre-emptive strike game (Simunovic et al. Reference Simunovic, Mifune and Yamagishi2013). While some researchers found support for the predictions of the CM, others did not (e.g., Yamagishi & Mifune Reference Yamagishi and Mifune2016).
De Dreu and Gross (D&G)’s AD-G and IAD-G protocols offer new experimental paradigms to test the validity of the CM, as they may possibly and uniquely be able to appropriately reproduce the complex intergroup environment assumed by the CM in the laboratory. We discuss each game separately to illustrate what it tells us and what it can tell us about the co-evolution of out-group aggression and in-group cooperation.
In the case of IAD-C, the results reported by D&G do not provide unique support for the CM. Participants invested more in defence than in attack. Defence was more common and intuitive a choice than attack, which goes against the idea that humans are equally motivated for in-group cooperation and out-group aggression. There is some ambiguity here, however, as a non-negligible number of attacks occurred. In addition, as the authors suggest, we may suppose only defensive aggression was the mechanism that supported in-group-directed cooperation, thus making defensive, but not offensive, out-group aggression “the midwife of altruism and xenophobia” (sect. 5.1, para. 3).
AD-G, by contrast, does not at present inform new insights into the CM, but it has great potential value to do so. By introducing minimal group manipulation into the AD-G, we would be able to compare individual attack rates towards in-group and out-group members. We expect out-group members to be attacked and defended against more often than in-group members. The CM predicts no change to these attack and defence rates if the ties of interdependence between in-group members are broken (e.g., Balliet et al. Reference Balliet, Wu and De Dreu2014; Yamagishi & Mifune Reference Yamagishi and Mifune2016). However, it is possible that we observe a reduction of offensive but not defensive actions in such circumstances, thus providing support for the idea that defence was the driver of the psychological adaptations responsible for in-group-bounded cooperation.
D&G also report that defence was a more intuitive decision for participants to make than was attack, suggesting that it is a default strategy in conflict situations. This is evident in both the AD-G and the IAD-C, but once again we contemplate an intergroup AD-G. Previous research may have shown that defence is not the automatic response to the presence of out-group members. Defensive aggression, measured by the pre-emptive strike game, has not emerged as either the dominant individual choice (Mifune et al. Reference Mifune, Simunovic and Yamagishi2017) or a 3-on-3 paradigm (Mifune et al. Reference Mifune, Hizen, Kamijo and Okano2016). This could be because defensive behaviour in intergroup situations occurs only when people perceive distinct attack intentions in out-group members, thus negating its unconditional nature as predicted by the CM. This should be investigated using the AD-G and with special reference to the prediction of the out-group partner's behavioural preferences.
Attacking behaviour can, at least in part, be deconstructed into preference and belief, as can be done for cooperation (Yamagishi et al. Reference Yamagishi, Mifune, Li, Shinada, Hashimoto, Horita, Miura, Inukai, Tanida, Kiyonari, Takagishi and Simunovic2013). Preference, such as loss or inequity aversion, refers to one's outcome goals. Beliefs about or expectations of others’ behaviours, partly driven by the perception of the ongoing situation, refer to the individual's outcome expectations (Pruitt & Kimmel Reference Pruitt and Kimmel1977). D&G's results open the possibility that expectation is a more important factor in attack/defence relations than is preference. This has already been demonstrated for in-group cooperation (Balliet et al. Reference Balliet, Wu and De Dreu2014; Yamagishi & Mifune Reference Yamagishi and Mifune2008). The two novel games have the potential to become the key to deciphering that puzzle while simultaneously providing special ecological validity inherent in the fact that participants would be interacting in the same paradigm, rather than in separate ones, to measure attack and defence.
In conclusion, D&G suggest methodological and theoretical improvements to the field of intergroup relations, which we find important and fruitful. One of the avenues their work opens has to do with the in-group cooperation and out-group aggression CM. By demonstrating the dichotomy and asymmetry inherent in a large proportion of intergroup conflicts, their results suggest possible refinements for the CM. In particular, the AD-G and IAD-C show that defensive rather than offensive out-group aggression could be the theoretical underpinning that would allow for the development of an in-group-bounded cooperative psychology in humans. In addition, these games provide opportunities to separate attack from defence while retaining ecological validity and to test the relative importance of outcome preference and outcome expectations. Nevertheless, more research is needed to claim anything more conclusive.
You have
Access
The mystery of human intergroup aggression has been studied in various fields of social sciences, from anthropology to political sciences. Previous psychological research mainly focused on attacking behaviour, leaving the psychology of defence considerably neglected. Although experimental economic games can be used to separately measure attack- and defence-based interactions between multiple individuals, research has so far not focused on the asymmetry that results when attack and defence co-occur, as they most often do. In this sense, the Attacker-Defender Game (AD-G) and IAD-C offer significant methodological and theoretical developments. In this commentary, we mainly focus on how these games can be used to uncover the mystery of the evolution of out-group aggression. We discuss their future directions, as well as their limitations.
Modern investigations of intergroup behaviour rely on integrating theories from evolutionary sciences with economic games. Particular focus is given to the human tendency to cooperate with in-group members and display aggression towards out-group members. Here, we will focus on the co-evolution model (CM; Choi & Bowles Reference Choi and Bowles2007). The CM suggests that frequent intergroup conflict in human evolutionary and cultural past caused parochial cooperation, which psychologically co-occurs with out-group aggression. In other words, at its core, the CM predicts that humans should be unconditionally motivated to not only increase their own group's standing, but also decrease the standing of other groups. Many subsequent studies attempted to test this prediction using various experimental paradigms, from third-party punishment (Bernhard et al. Reference Bernhard, Fischbacher and Fehr2006), intergroup prisoner's dilemma – maximizing difference (Halevy et al. Reference Halevy, Bornstein and Sagiv2008), and the pre-emptive strike game (Simunovic et al. Reference Simunovic, Mifune and Yamagishi2013). While some researchers found support for the predictions of the CM, others did not (e.g., Yamagishi & Mifune Reference Yamagishi and Mifune2016).
De Dreu and Gross (D&G)’s AD-G and IAD-G protocols offer new experimental paradigms to test the validity of the CM, as they may possibly and uniquely be able to appropriately reproduce the complex intergroup environment assumed by the CM in the laboratory. We discuss each game separately to illustrate what it tells us and what it can tell us about the co-evolution of out-group aggression and in-group cooperation.
In the case of IAD-C, the results reported by D&G do not provide unique support for the CM. Participants invested more in defence than in attack. Defence was more common and intuitive a choice than attack, which goes against the idea that humans are equally motivated for in-group cooperation and out-group aggression. There is some ambiguity here, however, as a non-negligible number of attacks occurred. In addition, as the authors suggest, we may suppose only defensive aggression was the mechanism that supported in-group-directed cooperation, thus making defensive, but not offensive, out-group aggression “the midwife of altruism and xenophobia” (sect. 5.1, para. 3).
AD-G, by contrast, does not at present inform new insights into the CM, but it has great potential value to do so. By introducing minimal group manipulation into the AD-G, we would be able to compare individual attack rates towards in-group and out-group members. We expect out-group members to be attacked and defended against more often than in-group members. The CM predicts no change to these attack and defence rates if the ties of interdependence between in-group members are broken (e.g., Balliet et al. Reference Balliet, Wu and De Dreu2014; Yamagishi & Mifune Reference Yamagishi and Mifune2016). However, it is possible that we observe a reduction of offensive but not defensive actions in such circumstances, thus providing support for the idea that defence was the driver of the psychological adaptations responsible for in-group-bounded cooperation.
D&G also report that defence was a more intuitive decision for participants to make than was attack, suggesting that it is a default strategy in conflict situations. This is evident in both the AD-G and the IAD-C, but once again we contemplate an intergroup AD-G. Previous research may have shown that defence is not the automatic response to the presence of out-group members. Defensive aggression, measured by the pre-emptive strike game, has not emerged as either the dominant individual choice (Mifune et al. Reference Mifune, Simunovic and Yamagishi2017) or a 3-on-3 paradigm (Mifune et al. Reference Mifune, Hizen, Kamijo and Okano2016). This could be because defensive behaviour in intergroup situations occurs only when people perceive distinct attack intentions in out-group members, thus negating its unconditional nature as predicted by the CM. This should be investigated using the AD-G and with special reference to the prediction of the out-group partner's behavioural preferences.
Attacking behaviour can, at least in part, be deconstructed into preference and belief, as can be done for cooperation (Yamagishi et al. Reference Yamagishi, Mifune, Li, Shinada, Hashimoto, Horita, Miura, Inukai, Tanida, Kiyonari, Takagishi and Simunovic2013). Preference, such as loss or inequity aversion, refers to one's outcome goals. Beliefs about or expectations of others’ behaviours, partly driven by the perception of the ongoing situation, refer to the individual's outcome expectations (Pruitt & Kimmel Reference Pruitt and Kimmel1977). D&G's results open the possibility that expectation is a more important factor in attack/defence relations than is preference. This has already been demonstrated for in-group cooperation (Balliet et al. Reference Balliet, Wu and De Dreu2014; Yamagishi & Mifune Reference Yamagishi and Mifune2008). The two novel games have the potential to become the key to deciphering that puzzle while simultaneously providing special ecological validity inherent in the fact that participants would be interacting in the same paradigm, rather than in separate ones, to measure attack and defence.
In conclusion, D&G suggest methodological and theoretical improvements to the field of intergroup relations, which we find important and fruitful. One of the avenues their work opens has to do with the in-group cooperation and out-group aggression CM. By demonstrating the dichotomy and asymmetry inherent in a large proportion of intergroup conflicts, their results suggest possible refinements for the CM. In particular, the AD-G and IAD-C show that defensive rather than offensive out-group aggression could be the theoretical underpinning that would allow for the development of an in-group-bounded cooperative psychology in humans. In addition, these games provide opportunities to separate attack from defence while retaining ecological validity and to test the relative importance of outcome preference and outcome expectations. Nevertheless, more research is needed to claim anything more conclusive.