Guala rightly draws attention to the fact that human decision-making under experimental game conditions cannot be extrapolated directly to decisions made under the real conditions. For example, the Ju/‘hoansi (!Kung san) make decisions in real life that contradict their behavior in experimental game conditions (Wiessner Reference Wiessner2009). The disconnect relates to decisions being predicated on both a biological and cultural heritage (Read Reference Read2010a); hence, the behaviors observed in a game context are a complex mixture of background predispositions and the conditions specified in the game context, and need not mirror decisions made during daily life.
Guala does not go far enough, though, in his discussion of the disjunction between experimental and real conditions. In an endnote he observes that current theories of reciprocity based on game theory have not drawn upon the concept of reciprocity previously developed in anthropology (e.g., Sahlins Reference Sahlins1972/1974) to account for the informal exchange of goods and services that is part of social life in human societies (see Note 2 in the target article). However, he does not follow up on his observation and instead limits his argument to a discourse on weak versus strong reciprocity, as if the only matter at issue is whether we account for cooperative behavior in human societies by one or the other of these two competing theories.
Running deeper than the surface issue of whether experimental evidence for strong reciprocity can be extrapolated to behaviors in natural conditions, is whether our perception of cooperative behavior in human societies has been framed correctly in the first place. Guala, like most researchers in this area, accepts uncritically the notion that small-scale human societies such as hunter-gatherers can be characterized as “acephalous social orders based on spontaneous cooperation” (sect. 8, para. 2, emphasis added). From the assumption of “spontaneous cooperation,” it follows that relevant evolutionary questions are: How and under what conditions will there be selection at the individual level for cooperation as a trait? And, how and under what conditions will a population composed of individuals with a cooperation trait be stable against invasion by a “free-rider” trait?
The problem with formulating cooperation in this manner, along with its attendant questions, lies in the lack of evidence that individuals in small-scale societies are spontaneous cooperators. Consider how food resources are shared. In a hunter-gatherer society such as the Ju/‘hoansi, food resources “in the wild” are collectively owned, in the sense of collective rights of access, by a residence group of families closely related through culturally constituted kinship relations (see Read Reference Read, Feinberg and Ottenheimer2001; Reference Read2007, for the difference between biological and cultural kinship). Rights to those food resources depend on membership (temporary or permanent) in a residence group. Collective ownership of resources changes into individual ownership by the mode of procurement and characteristics of the resources. Resources that come in small units are accessible to any able-bodied adult on a regular basis, and have low risk of failure on each procurement episode (such as, but not limited to, vegetal foods); these are transformed from collective into individual ownership through foraging. Resources that come in relatively large units are differentially accessible by adults according to individual skills, and have high risk of failure on a given procurement episode (such as game animals); these resources are considered to be collectively owned through hunting. For the latter, ownership changes from collective to individual according to culturally specified rules of sharing that remove decisions about sharing from the individual hunter to the group as a collectivity. Among the Ju/‘hoansi, for example, the cultural rule is that the owner of the arrow that killed the animal (who need not have been present during the hunt) distributes the meat from the animal (Marshall Reference Marshall1976). Among the Netsilik Inuit, seals killed in winter hunting through their breathing holes in the pack ice were distributed in accordance with a culturally constituted system of “sealing partners” (Balikci Reference Balikci1970). Cultural rules like this make meat-sharing a group-level, not an individual-level, trait (Read Reference Read2012).
In general, resources that are individually owned are not subject to cultural rules of sharing. Individually owned resources are shared within a family (with a culture specific definition of what constitutes a family) and without cultural rules. Sharing within a family corresponds to “spontaneous cooperation.” However, we need neither weak nor strong reciprocity to account for sharing and cooperation within a family.
Individually owned resources allow for individual decisions about whether they should be kept or given to others, but the act of gift giving is a social one (Mauss Reference Mauss and Halls1924/1990). Gift-giving is subject to cultural rules such as generalized reciprocity (Sahlins Reference Sahlins1972/1974) in which A gives to B, with the (usually unstated) understanding that B will reciprocate at some indefinite time in the future and by an unspecified amount, as in hxaro gift giving among the Ju/’hoansi (Wiessner Reference Wiessner1977; Reference Wiessner, Leacock and Lee1982). Generalized reciprocity is dependent upon trust by the parties concerned (Sahlins Reference Sahlins1972/1974), and trust depends on close kin relations (kin being understood in a cultural and not a biological sense). This is the context where “punishment” comes into play, but punishment, as Guala discusses, is not of the kind invoked in the theory of strong reciprocity. Rather, it is social punishment in which the transgressor is made to understand by various means that her or his behavior is unacceptable as a kinsman. This, as pointed out by Guala, is punishment by the collective against the individual, by one's kin against oneself. Among close kinsmen, social punishment is effective because of one's dependency on kin for surviving in hunter-gatherer and other small-scale societies, not because of the magnitude of the punishment in a material sense.
Cooperative behavior characterizes sharing of collectively owned resources, but it is not “spontaneous cooperation” and instead is determined through cultural rules. The specificity of the cultural rules relates to the degree of risk that failure to succeed in a resource procurement episode has for the survival of members of the group. For the Netsilik Inuit, disputes over sharing of seals hunted through the pack ice in winter could lead to the breakdown of a winter sealing camp, thereby exposing camp members to the risk of death through starvation (Balikci Reference Balikci1970). Correspondingly, the Netsilik had extensive and highly specific rules for the sharing of seals that transformed the individual hunter into an agent for the group and removed decisions about sharing from the hunter. In effect, the cost of a dispute over sharing of seals was too high and too immediate to allow for the sharing to be subject to individual decision-making even if there were costly punishment as hypothesized under strong reciprocity. In contrast, the Tiwi living on Melville and Bathurst Islands off the northern coast of Australia had low risk and relatively simple rules for sharing of hunter meat (Goodale Reference Goodale1971). (Risk can be measured indirectly by the complexity of implements used in resource procurement [Read Reference Read2008; Torrence Reference Torrence and Torrence1989]. The Netsilik used complex implements, and the Tiwi simple ones.)
The notion of cooperative behavior used in weak or strong reciprocity theories is at odds with the facts of meat-sharing in hunter-gatherer societies. These theories do not take into account the major transformation that took place in the basis for social organization during the evolution of human societies (Read Reference Read, Dunbar, Gamble and Gowlett2010b; Reference Read2012). That transformation is from societies in which patterns of social organization and structure emerge from face-to-face interaction of group members to relation-based societies (Read Reference Read2012) predicated on behaviors formed in accordance with systems of organization for the society as a whole (Leaf & Read, in press), such as culturally constructed systems of kinship relations that define the boundary for, and internal organization of, the small-scale human societies from which are derived all larger-scale human societies.
Guala rightly draws attention to the fact that human decision-making under experimental game conditions cannot be extrapolated directly to decisions made under the real conditions. For example, the Ju/‘hoansi (!Kung san) make decisions in real life that contradict their behavior in experimental game conditions (Wiessner Reference Wiessner2009). The disconnect relates to decisions being predicated on both a biological and cultural heritage (Read Reference Read2010a); hence, the behaviors observed in a game context are a complex mixture of background predispositions and the conditions specified in the game context, and need not mirror decisions made during daily life.
Guala does not go far enough, though, in his discussion of the disjunction between experimental and real conditions. In an endnote he observes that current theories of reciprocity based on game theory have not drawn upon the concept of reciprocity previously developed in anthropology (e.g., Sahlins Reference Sahlins1972/1974) to account for the informal exchange of goods and services that is part of social life in human societies (see Note 2 in the target article). However, he does not follow up on his observation and instead limits his argument to a discourse on weak versus strong reciprocity, as if the only matter at issue is whether we account for cooperative behavior in human societies by one or the other of these two competing theories.
Running deeper than the surface issue of whether experimental evidence for strong reciprocity can be extrapolated to behaviors in natural conditions, is whether our perception of cooperative behavior in human societies has been framed correctly in the first place. Guala, like most researchers in this area, accepts uncritically the notion that small-scale human societies such as hunter-gatherers can be characterized as “acephalous social orders based on spontaneous cooperation” (sect. 8, para. 2, emphasis added). From the assumption of “spontaneous cooperation,” it follows that relevant evolutionary questions are: How and under what conditions will there be selection at the individual level for cooperation as a trait? And, how and under what conditions will a population composed of individuals with a cooperation trait be stable against invasion by a “free-rider” trait?
The problem with formulating cooperation in this manner, along with its attendant questions, lies in the lack of evidence that individuals in small-scale societies are spontaneous cooperators. Consider how food resources are shared. In a hunter-gatherer society such as the Ju/‘hoansi, food resources “in the wild” are collectively owned, in the sense of collective rights of access, by a residence group of families closely related through culturally constituted kinship relations (see Read Reference Read, Feinberg and Ottenheimer2001; Reference Read2007, for the difference between biological and cultural kinship). Rights to those food resources depend on membership (temporary or permanent) in a residence group. Collective ownership of resources changes into individual ownership by the mode of procurement and characteristics of the resources. Resources that come in small units are accessible to any able-bodied adult on a regular basis, and have low risk of failure on each procurement episode (such as, but not limited to, vegetal foods); these are transformed from collective into individual ownership through foraging. Resources that come in relatively large units are differentially accessible by adults according to individual skills, and have high risk of failure on a given procurement episode (such as game animals); these resources are considered to be collectively owned through hunting. For the latter, ownership changes from collective to individual according to culturally specified rules of sharing that remove decisions about sharing from the individual hunter to the group as a collectivity. Among the Ju/‘hoansi, for example, the cultural rule is that the owner of the arrow that killed the animal (who need not have been present during the hunt) distributes the meat from the animal (Marshall Reference Marshall1976). Among the Netsilik Inuit, seals killed in winter hunting through their breathing holes in the pack ice were distributed in accordance with a culturally constituted system of “sealing partners” (Balikci Reference Balikci1970). Cultural rules like this make meat-sharing a group-level, not an individual-level, trait (Read Reference Read2012).
In general, resources that are individually owned are not subject to cultural rules of sharing. Individually owned resources are shared within a family (with a culture specific definition of what constitutes a family) and without cultural rules. Sharing within a family corresponds to “spontaneous cooperation.” However, we need neither weak nor strong reciprocity to account for sharing and cooperation within a family.
Individually owned resources allow for individual decisions about whether they should be kept or given to others, but the act of gift giving is a social one (Mauss Reference Mauss and Halls1924/1990). Gift-giving is subject to cultural rules such as generalized reciprocity (Sahlins Reference Sahlins1972/1974) in which A gives to B, with the (usually unstated) understanding that B will reciprocate at some indefinite time in the future and by an unspecified amount, as in hxaro gift giving among the Ju/’hoansi (Wiessner Reference Wiessner1977; Reference Wiessner, Leacock and Lee1982). Generalized reciprocity is dependent upon trust by the parties concerned (Sahlins Reference Sahlins1972/1974), and trust depends on close kin relations (kin being understood in a cultural and not a biological sense). This is the context where “punishment” comes into play, but punishment, as Guala discusses, is not of the kind invoked in the theory of strong reciprocity. Rather, it is social punishment in which the transgressor is made to understand by various means that her or his behavior is unacceptable as a kinsman. This, as pointed out by Guala, is punishment by the collective against the individual, by one's kin against oneself. Among close kinsmen, social punishment is effective because of one's dependency on kin for surviving in hunter-gatherer and other small-scale societies, not because of the magnitude of the punishment in a material sense.
Cooperative behavior characterizes sharing of collectively owned resources, but it is not “spontaneous cooperation” and instead is determined through cultural rules. The specificity of the cultural rules relates to the degree of risk that failure to succeed in a resource procurement episode has for the survival of members of the group. For the Netsilik Inuit, disputes over sharing of seals hunted through the pack ice in winter could lead to the breakdown of a winter sealing camp, thereby exposing camp members to the risk of death through starvation (Balikci Reference Balikci1970). Correspondingly, the Netsilik had extensive and highly specific rules for the sharing of seals that transformed the individual hunter into an agent for the group and removed decisions about sharing from the hunter. In effect, the cost of a dispute over sharing of seals was too high and too immediate to allow for the sharing to be subject to individual decision-making even if there were costly punishment as hypothesized under strong reciprocity. In contrast, the Tiwi living on Melville and Bathurst Islands off the northern coast of Australia had low risk and relatively simple rules for sharing of hunter meat (Goodale Reference Goodale1971). (Risk can be measured indirectly by the complexity of implements used in resource procurement [Read Reference Read2008; Torrence Reference Torrence and Torrence1989]. The Netsilik used complex implements, and the Tiwi simple ones.)
The notion of cooperative behavior used in weak or strong reciprocity theories is at odds with the facts of meat-sharing in hunter-gatherer societies. These theories do not take into account the major transformation that took place in the basis for social organization during the evolution of human societies (Read Reference Read, Dunbar, Gamble and Gowlett2010b; Reference Read2012). That transformation is from societies in which patterns of social organization and structure emerge from face-to-face interaction of group members to relation-based societies (Read Reference Read2012) predicated on behaviors formed in accordance with systems of organization for the society as a whole (Leaf & Read, in press), such as culturally constructed systems of kinship relations that define the boundary for, and internal organization of, the small-scale human societies from which are derived all larger-scale human societies.