Jones hypothesizes that the grammar faculty is an adaptation for playing coordination games (sect. 5.3, para. 6), allowing an individual to discover constraints, match his or her own constraint rankings with those of other speakers, and generate mutually intelligible classifications of kin. Jones does not fully develop an evolutionary account that explains why human propensities for coordination games should be applied to the domain of kinship, however. In other words: Why do we need to coordinate when it comes to classifying kin?
Evolutionary explanations of human kinship often begin with theories of kin selection (Hamilton Reference Hamilton1964). As cultural anthropologists often remind us (Sahlins Reference Sahlins1976), however, kin terminologies rarely classify kin in ways that correspond with genetic relatedness. As a result, many cultural anthropologists see kin selection – and evolutionary theory more broadly – as irrelevant to our understanding of human kinship (McKinnon Reference McKinnon, McKinnon and Silverman2005). Focusing on the role of coordination games in the structure and function of kin terminologies may provide a solution to the apparent disparity between the ways that kin terminologies define relatedness and the evolutionary advantages of nepotistic behavior.
Several researchers, including Jones himself (Reference Jones2000), have emphasized how kinship enables individuals to identify common interests and coordinate their actions accordingly (Chagnon Reference Chagnon, Cronk, Chagnon and Irons2000; Cronk & Gerkey Reference Cronk, Gerkey, Dunbar and Barrett2007; Fox Reference Fox, Chagnon and Irons1979; Irons Reference Irons, Alexander and Tinkle1981; Van den Berghe Reference Van den Berghe1979). We suggest that there are actually two coordination games when it comes to understanding kin terminology. The first game determines whether two individuals can arrive at a mutually intelligible and agreed upon term for different kinds of kin. This is simply a specific instance of the broader coordination game presented by language in general (Hume Reference Hume1740; Lewis Reference Lewis1969; Sugden Reference Sugden2004). The second game builds on the shared meaning of kin terms by combining them with cultural norms and values that inform how two individuals should act toward kin. Jones focuses primarily on the first coordination game and alludes only briefly to the second. Exploring the relationship between these two coordination games may provide a more comprehensive understanding of the structure and function of kin terminologies.
Efferson et al. (Reference Efferson, Lalive and Fehr2008) demonstrated that individuals can spontaneously use symbolic markers to solve coordination problems. Although these markers were arbitrary at the start of the experiment, they acquired meaning and became reliable guides for solving the coordination problem when two conditions were met: (1) individuals differed from one another in an important but unobservable way, and (2) individuals were allowed to choose markers freely and flexibly (p. 1848). Efferson et al. note that the conditions enabling symbolic markers to serve as guides for solving coordination games should apply “whenever people have a shared interest in distinguishing among themselves in terms of their unobservable information” (p. 1848). Identifying kin and interacting with them often requires a significant amount of coordination, and kin terminologies may allow related individuals to distinguish among themselves in the way that Efferson et al. describe.
Although there is evidence suggesting that kin can to some extent recognize one another without kin terms or other symbolic markers (Lieberman et al. Reference Lieberman, Tooby and Cosmides2007), there are many important ways in which related individuals differ from one another that are difficult or impossible to observe. Genetic relatedness is one such trait, but, as Jones rightly emphasizes, there are others. In human social groups, where kinship is often inextricable from economic, political, religious, and reproductive affairs, an individual's sex, age, rank, descent group, and alliances may be as important as genetic relatedness, if not more so. This is because the kinds of coordination problems that humans need to solve involve complex calculations of costs and benefits in multiple currencies that eventually have consequences for reproductive success.
The next step is to investigate how kin terminologies help individuals solve coordination problems beyond the domain of classification. Alvard and Nolin's (Reference Alvard and Nolin2002) research on cooperative whale hunting in Lamalera, Indonesia, shows how kinship can help people solve coordination games. In Lamalera, descent groups coordinate whaling by providing the equipment, skill, and labor that lets individuals earn greater returns for their effort than solitary productive activities. Whaling crews are composed of related individuals, but Alvard (Reference Alvard2003) has shown that descent group membership better predicts the composition of whaling crews than genetic relatedness. Unlike genetic relatedness, which varies from individual to individual in a whaling crew, descent group membership can be the same for all members. Interestingly, research by Nolin (Reference Nolin2010; in press) shows that the subsequent distribution of whale meat in Lamalera follows genetic relatedness between households. Lamaleran whalers use descent groups defined by kin terminologies to solve the coordination game of whaling, and then they use the logic of kin selection to spread the spoils.
If humans possess an adaptation for solving coordination games, we should expect this adaptation also to apply in contexts that do not necessarily involve kin. Efferson et al.'s experiments with symbolic markers support this idea, and there is evidence from other approaches, as well. Cronk (Reference Cronk2007) conducted experiments in Kenya with trust games that were framed with a reference to osotua, a need-based gift-giving relationship among Maasai. Maasai participants responded to the framed games in ways consistent with the central principles of osotua: They were attuned to signs of need, transferring more money when the other player appeared to need help.
Given that Maasai are familiar with the osotua concept, it may not be surprising that the osotua framing influenced how they played the game. However, when the experiment was conducted with Americans who were learning about osotua for the first time, the results were nearly identical (Cronk & Wasielewski Reference Cronk and Wasielewski2008). This quick adoption of osotua norms and values may stem from a broader human susceptibility to being influenced by cultural norms that facilitate coordination. If solving coordination problems has been important throughout our evolutionary history, then we may have developed an alertness for and ability to quickly adopt such norms.
Jones hypothesizes that the grammar faculty is an adaptation for playing coordination games (sect. 5.3, para. 6), allowing an individual to discover constraints, match his or her own constraint rankings with those of other speakers, and generate mutually intelligible classifications of kin. Jones does not fully develop an evolutionary account that explains why human propensities for coordination games should be applied to the domain of kinship, however. In other words: Why do we need to coordinate when it comes to classifying kin?
Evolutionary explanations of human kinship often begin with theories of kin selection (Hamilton Reference Hamilton1964). As cultural anthropologists often remind us (Sahlins Reference Sahlins1976), however, kin terminologies rarely classify kin in ways that correspond with genetic relatedness. As a result, many cultural anthropologists see kin selection – and evolutionary theory more broadly – as irrelevant to our understanding of human kinship (McKinnon Reference McKinnon, McKinnon and Silverman2005). Focusing on the role of coordination games in the structure and function of kin terminologies may provide a solution to the apparent disparity between the ways that kin terminologies define relatedness and the evolutionary advantages of nepotistic behavior.
Several researchers, including Jones himself (Reference Jones2000), have emphasized how kinship enables individuals to identify common interests and coordinate their actions accordingly (Chagnon Reference Chagnon, Cronk, Chagnon and Irons2000; Cronk & Gerkey Reference Cronk, Gerkey, Dunbar and Barrett2007; Fox Reference Fox, Chagnon and Irons1979; Irons Reference Irons, Alexander and Tinkle1981; Van den Berghe Reference Van den Berghe1979). We suggest that there are actually two coordination games when it comes to understanding kin terminology. The first game determines whether two individuals can arrive at a mutually intelligible and agreed upon term for different kinds of kin. This is simply a specific instance of the broader coordination game presented by language in general (Hume Reference Hume1740; Lewis Reference Lewis1969; Sugden Reference Sugden2004). The second game builds on the shared meaning of kin terms by combining them with cultural norms and values that inform how two individuals should act toward kin. Jones focuses primarily on the first coordination game and alludes only briefly to the second. Exploring the relationship between these two coordination games may provide a more comprehensive understanding of the structure and function of kin terminologies.
Efferson et al. (Reference Efferson, Lalive and Fehr2008) demonstrated that individuals can spontaneously use symbolic markers to solve coordination problems. Although these markers were arbitrary at the start of the experiment, they acquired meaning and became reliable guides for solving the coordination problem when two conditions were met: (1) individuals differed from one another in an important but unobservable way, and (2) individuals were allowed to choose markers freely and flexibly (p. 1848). Efferson et al. note that the conditions enabling symbolic markers to serve as guides for solving coordination games should apply “whenever people have a shared interest in distinguishing among themselves in terms of their unobservable information” (p. 1848). Identifying kin and interacting with them often requires a significant amount of coordination, and kin terminologies may allow related individuals to distinguish among themselves in the way that Efferson et al. describe.
Although there is evidence suggesting that kin can to some extent recognize one another without kin terms or other symbolic markers (Lieberman et al. Reference Lieberman, Tooby and Cosmides2007), there are many important ways in which related individuals differ from one another that are difficult or impossible to observe. Genetic relatedness is one such trait, but, as Jones rightly emphasizes, there are others. In human social groups, where kinship is often inextricable from economic, political, religious, and reproductive affairs, an individual's sex, age, rank, descent group, and alliances may be as important as genetic relatedness, if not more so. This is because the kinds of coordination problems that humans need to solve involve complex calculations of costs and benefits in multiple currencies that eventually have consequences for reproductive success.
The next step is to investigate how kin terminologies help individuals solve coordination problems beyond the domain of classification. Alvard and Nolin's (Reference Alvard and Nolin2002) research on cooperative whale hunting in Lamalera, Indonesia, shows how kinship can help people solve coordination games. In Lamalera, descent groups coordinate whaling by providing the equipment, skill, and labor that lets individuals earn greater returns for their effort than solitary productive activities. Whaling crews are composed of related individuals, but Alvard (Reference Alvard2003) has shown that descent group membership better predicts the composition of whaling crews than genetic relatedness. Unlike genetic relatedness, which varies from individual to individual in a whaling crew, descent group membership can be the same for all members. Interestingly, research by Nolin (Reference Nolin2010; in press) shows that the subsequent distribution of whale meat in Lamalera follows genetic relatedness between households. Lamaleran whalers use descent groups defined by kin terminologies to solve the coordination game of whaling, and then they use the logic of kin selection to spread the spoils.
If humans possess an adaptation for solving coordination games, we should expect this adaptation also to apply in contexts that do not necessarily involve kin. Efferson et al.'s experiments with symbolic markers support this idea, and there is evidence from other approaches, as well. Cronk (Reference Cronk2007) conducted experiments in Kenya with trust games that were framed with a reference to osotua, a need-based gift-giving relationship among Maasai. Maasai participants responded to the framed games in ways consistent with the central principles of osotua: They were attuned to signs of need, transferring more money when the other player appeared to need help.
Given that Maasai are familiar with the osotua concept, it may not be surprising that the osotua framing influenced how they played the game. However, when the experiment was conducted with Americans who were learning about osotua for the first time, the results were nearly identical (Cronk & Wasielewski Reference Cronk and Wasielewski2008). This quick adoption of osotua norms and values may stem from a broader human susceptibility to being influenced by cultural norms that facilitate coordination. If solving coordination problems has been important throughout our evolutionary history, then we may have developed an alertness for and ability to quickly adopt such norms.