Success or Prestige? Hunters’ cultural biases
Rob Boyd and Pete Richerson have identified two biases, one based on success, the other on prestige, that might influence which individual is most imitated. If you were living in a foraging society, would you rather imitate prestigious hunters or successful ones? Successful ones, you say? It may not be so easy or so argue Kim Hill and Keith Kintigh in "Can Anthropologists Distinguish Good and Poor Hunters? Implications for Hunting Hypotheses, Sharing Conventions, and Cultural Transmission" (in Current Anthropology, Volume 50, Number 3, June 2009).
Here is the Abstract:
Numerous articles examine the relationship between men's hunting skill and other important biological and social traits. We analyzed more than 14,000 hunter days during 27 years of monitoring the Ache of Paraguay by using resampling methods to demonstrate that large sample sizes are generally required in order to distinguish individual men by hunting skill. A small published study on !Kung hunters shows that large‐game hunters are even more difficult to distinguish by individual skill level. This is a serious problem because regressions using noisy hunting data as the independent variable systematically underestimate the association of hunting ability with other biosocial traits. The analysis suggests that some coresidents in many small‐scale societies will be unable to accurately distinguish hunters by skill level, possibly favoring groupwide meat‐sharing conventions and biased cultural transmission that emphasizes prestige rather than perceived hunting skill.
Olivier Morin 30 April 2009 (12:15)
"The analysis suggests that some coresidents in many small‐scale societies will be unable to accurately distinguish hunters by skill level, possibly favoring (...) biased cultural transmission that emphasizes prestige rather than perceived hunting skill." Well, there is another possibility : that people just don't have to select one individual to learn from, as these skills are often widely known and shared in the population, and one benefits from the redundancy brought about by exposure to many different models.
Christophe Heintz 30 April 2009 (14:23)
Olivier rightly notes that "one benefits from the redundancy brought about by exposure to many different models". Yet, this does not provide a means to choose among the models. It seems that a consequence of Kim Hill and Keith Kintigh's analysis is that people will not be able to pick the best model on the basis of its success [b]as it is revealed by statistical data[/b] (with the idea that each hunter implements one type of hunting model). Another possibility, however, is that people pick the best hunting models or strategies not on the basis of statistical evidence, but because they come to understand why some strategies are more likely to be successful and efficient than others. This would be a type of [i]content based biased transmission[/i] rather than an [i]observed-success based biased transmission[/i] or [i]prestige based biased transmission.[/i]
Olivier Morin 30 April 2009 (15:11)
"Yet, this does not provide a means to choose among the models." But, precisely, my point is that you don't have to chose one model. Learning from one model only is a requirement of Boyd and Richerson's mathematical models of imitation. It is not a requirement in real life. One may simply learn from everyone around, then adopt, as you say, the skills that work best for oneself (which may come from a variety of models or a recombination of these). I think that the debate around success- or prestige- biased imitation is a false debate. Social learning does not have to be biased to begin with, and it need not consist in learning from one person only, let alone in doing everything like her and neglect all other sources of expertise.
Dan Sperber 30 April 2009 (17:37)
Even without a specific bias, and on grounds of general rationality, one would expect people to imitate successful individuals, at least in the type of action in which they are being successful. What this article suggest is that success is not always easy to evaluate. This may be true not just for hunters but also for many types of cultural competence. How good are we at identifying a successful parent, a successful doctor, or a successful philosopher? Typically our ground for assuming that So-and-so is successful is her reputation to be sucessful. But reputation can be wholly dissociated from the facts of the matter, as illustrated by the fact that some diviners have the reputation of being more reliably successful than others. Now, of course, reputation and prestige are at least overlapping properties. So we have three interesting questions here: 1) What are the respective role of success and prestige in providing cultural models? 2)To what extent are the two really independent? 3)To what extent do they have to exist as independent biases as opposed to regularities following from rationality considerations (a 'rationality bias' if you want). I myseld don't have answers to these questions. PS. In response to Olivier's last comment: For the Boyd-Richerson-Henrich theory to work, it is not necessary, I believe, that individuals imitate just a single model, or only one type (e.g. prestigious, or successful) of model. It is the statistical effect of these biases at the level of the population that matter.
Olivier Morin 4 May 2009 (12:08)
Hill and Kintigh conclude their paper by supposing that, since actual success is not such a good measure of skill, observers should rely on skill instead, in order to choose a model. But that, too, might be more difficult than the theory assumes. A recent paper in Evolution and Human Behavior finds that, among the Tsimane, the correlation between pharmacopeic knowledge and prestige is extremely weak and not robust at all. The paper by Reyes-Garcia et al. contains many thoughtful observations on what prestige is and is not. PS In response to Dan : Granted. But this part of dual inheritance theory is extremely influenced, beyond Boyd and Richerson, by the important work of Henrich, Gil-White, etc. In many of these models, it is crucial that people choose one and only one person to imitate (I am thinking in particular of Henrich's explanation of technological decline among the Tasmanians). URL of the paper: http://linkinghub.elsevier.com/retrieve/pii/S1090513808000275