Wood (2006) describes two alternative hypotheses that attempt to account for the fact that meat transfer in ethnographic hunter-gatherers doesn't accord simply with a family provisioning strategy:
The showoff hypothesis, developed by Hawkes (1990, 1991) using data gathered among the Ache as well as the Hadza, proposes that men's hunting is best understood as camp provisioning that is repaid with social attention including increased mating opportunities. In the showoff account, men target large game because it can be shared widely and exchanged for fitness benefits including mating opportunities. Hadza men are said to pursue the showoff strategy at the expense of more dependable forms of family provisioning such as gathering or the hunting of small game (Hawkes, O'Connell, and Blurton Jones 1991).
The costly-signaling hypothesis, presented by Smith and Bliege Bird (2000), Bliege Bird, Smith, and Bird (2001), and Smith, Bliege Bird, and Bird (2003) using data gathered among Meriam Islanders, draws from signaling theory developed by Zahavi (1975, 1977). This framework proposes that men may be motivated to hunt as a means to broadcast information that is of interest to others. Establishing status as a good hunter is seen as a way for a man to advertise genotypic or phenotypic qualities and promote mutually beneficial arrangements. This hypothesis proposes that higher-status hunters display inherent qualities that make them sought as mates, preferred as allies, and avoided as competitors (Wood 2006:383).
These hypotheses clearly share much in common -- both propose that the value of pursuing costly hunting strategies (for example, big impressive kills as opposed to small but more reliable kills) is toward male competition in sexual selection. In the "showoff" model as described by Wood, that competition is mediated by exchange (excess meat that can be give away); in the "costly-signaling" model the competition is solely informational. But as Wood points out, Hawkes, O'Connell and Blurton Jones in recent publications (2001) have also focused on the informational aspect of hunting large game -- after all, the whole idea of "showing off" is that men are signaling others about their quality (bravery, prowess, etc.).
If there is a way to separate the informational content of these models, I would guess it is that the "showoff" model as proposed here includes information about generosity or magnanimity as well as hunting skill or bravery, and the "costly-signaling" hypothesis need not include such information. Of course, to that extent the "costly-signaling" hypothesis need not predict that hunters will actually share their food widely -- a jealous and hungry woman can just as easily get the information as a happy woman with a full belly.
The difference between these surely is by degrees of information about different aspects of hunting and social exchange. Being difficult to seperate, Wood lumps them both as "prestige" hypotheses for hunting.
So Wood showed a bunch of Hadza men two pictures: one showed men apparently returning to a group bearing lots of dead animals; the other showed men returning with only one little dik-dik. He asked the men which of the two camps they would choose to join. Wood predicted that men who wanted to increase their prestige would choose to join the one-dik-dik camp, because they would have a chance to show they were superior hunters to the resident men. It's sort of like the "big fish in a small pond" story. In contrast, men following the the "provisioning" strategy should choose the group with lots of meat, because they would be sure to have plenty to eat, even if they were not the best hunters.
Wood found that a large majority of the Hadza men would choose to join the camp with plenty to eat. He interprets this as support for the "provisioning" model. He does inject another interesting data point -- although very few of the Hadza men would choose the one-dik-dik camp, that small group included the only two polygynously married men, who were apparently confident that they could teach their new group to hunt better.
The study does not, of course, propose that prestige is unimportant for understanding men's behaviour among the Hadza. Rather, it should be taken as a decision situation in which most men value the material benefits of living with good hunters over the costs suffered to their own hunting reputation by comparison with such company. In this particular instance, men and women's responses are mostly in harmony and produce outcomes beneficial for household provisioning. Such mutually advantageous situations promote the persistence of pair bonds (Wood 2006:386).
It really seems strange to me that someone might predict the opposite result -- the result where more men choose to join the one-dik-dik camp. I mean, that's like asking people if they would rather work for Google or Enron, and they choose Enron! Sure, maybe you could be the smartest Enron employee and quickly move up the management ranks, but you'd better watch your 401k.
Of course, that very comparison suggests that people should really perceive the circumstances under which they benefit from cooperation, and that hunting success is one of those circumstances. The reactions of the Hadza men are certainly consistent with the idea that they want to make sure they and their families have more reliable meat. The result is also compatible with the idea that Hadza men understand their hunting success (or failure) to be partly influenced by the success or failure of other men. This might include a broad range of perceived influences -- for instance, the men might suspect that a camp with no meat is just in a bad place for hunting; or that the men work together poorly.
I would also suggest another scenario -- perhaps a snap assessment of hunting success (based on a picture) carries other connotative information about a group. If the men are bad hunters, and most or all of the women in the group are related to the men, then one might tend to doubt the mating quality of the women. For example, if women have not had a reliable nutritional history, then they might face many fitness risks. If the men cannot support a strong hunting tradition, then other essential traditions might be lacking among the women. Or supernatural influences may have affected the camp.
Notably, several of the Hadza men who choose the one-dik-dik camp explicitly reference a possible exchange of information as a reason for their choice -- as in, "I can teach them how to hunt", or "The camp right now needs lots of help" (Wood 2006:386). None of the men say "The men over there are clueless, and I will have an easy time getting another wife" (all these men already had at least one wife). Now, there is no necessary reason for the men to be honest about (or even conscious of) a reproductive motive for their choice. But it is interesting that the choice is couched in terms of exchange, which contains an implicit promise of cooperation. As Richard Alexander pointed out, human competition occurs through the forum of cooperation, and intragroup and extragroup dynamics need not be the same.
There are so many options, some of which complicate the scenario and others which may simplify it a bit. So it probably should be expected that men would tend to err on the side of having enough to eat!
Finally, Wood suggests another way that prestige that accrues from hunting may affect mating behavior -- through its effect on group size:
In terms of signaling theory, these results support the idea that hunting ability and especially its material consequences are strong audience attractors (Smith and Bliege Bird 2000; Hawkes and Bliege Bird 2002). Most respondents wanted to join the camp of good hunters and, when asked why, said something similar to "I want to go there for lots of meat." Consequently, we can reasonably assume that good hunters find themselves in large camps more often than poor hunters (Wood 2006:386).
Wood points out that large groups have disadvantages, as perceived by the Hadza, and that such problems may manifest as costs for good hunters. But large groups with many good hunters will clearly spread risk more effectively than small groups with many poor hunters. Choosing to join a group of strong hunters may therefore have a synergistic effect as group size also increases -- at least to the point where the carrying capacities of group-typical territories are reached.
In this context, the choice to join a group of good or poor hunters is clearly context-dependent. The opportunity to join a group of good hunters is more valuable when food is scarce. The costs of joining a group of poor hunters are less severe when food is abundant. The fitness consequences of getting this choice right may outweigh the importance of within-group competition when food is plentiful.
Wood BM. Prestige or provisioning? A test of foraging goals among the Hadza. Curr Anthropol 47:383-387.