A recent paper in PLOS ONE by Anderson et al. concluded that the frequency of hunter-gatherer societies in which women hunted was much higher than people thought (79%), which was taken to discard the “myth” that in such societies there was a sex-based division of labor, with men hunting and women foraging. Further, Anderson et al. concluded that in 37% of the societies women hunted large game or game of all sizes, so that the hunting was not just things like birds or rabbits. That was taken to explode another myth—that of women who didn’t hunt the big game.
Here’s what I wrote about the paper on June 30:
On June 30 I reported on a PLOS One paper whose data showed that, in existing hunter-gatherer societies, women participated in hunting a lot more than most people (well, at least I) thought. (The original paper is here.) Here’s what I wrote, giving the authors’ conclusions (the authors’ words are doubly indented, my short bit singly indented).
The authors’ conclusions:
Here we investigated whether noted trends of non-gendered hunting labor known from the archaeological record continued into more recent, ethnographic periods. The descriptive sample described here is sufficient to warrant the conclusion that women in foraging societies across the world participate in hunting during more recent time periods, a finding that makes sense given women’s general morphology and physiology . The prevalence of data on women hunting directly opposes the common belief that women exclusively gather while men exclusively hunt, and further, that the implicit sexual division of labor of ‘hunter/gatherer’ is misapplied. Given that this bimodal paradigm has influenced the interpretation of archeological evidence, which includes the reluctance to distinguish projectile tools found within female burials as intended for hunting (or fighting) , this paper joins others in urging the necessity to reevaluate archeological evidence, to reassess ethnographic evidence, to question the dichotomous use of ‘hunting and gathering,’ and to deconstruct the general “man the hunter” narrative.
Of the 63 foraging societies with clear descriptions of hunting strategies, 79% of them demonstrated female hunting. The widespread presence of female hunting suggests that females play an instrumental role in hunting, further adding to the data that women contribute disproportionately to the total caloric intake of many foraging groups. Additionally, over 70% of hunting done by females is interpreted as intentional, meaning that females play an active and important role in hunting—and the teaching of hunting—even if they use different tools and employ different acquisition strategies. For example, among the Aka, women’s participation in net-hunting was required, whereas men’s participation was not.
It’s clear from these data that hunter-gatherer societies do not show a strict division of labor, though I’d like to see data on the frequency of hunts in which women participate, not just the frequency of societies in which women hunt. Men still do most of the hunting and most of the big-game hunting, but this shows only moderate rather than extensive division of labor.
Since then, critiques of the data and conclusions of Anderson et al. have appeared first in a series of tweets by Vivek V. Venkataraman in the original post and then in a paper in Aporia Magazine by the pseudonymous “Alexander”.
Now Venkataraman and one of his students have provided a longer critique of Anderson et al. that isn’t yet peer-reviewed, but is on Venkataraman’s blog. I can’t remember where I found this quote from him, as I copied it out a while back, but it links to the critique that is online:
A group of who know the literature have been discussing this paper on Twitter. One of us along with his students has developed a critique of many of the data problems and data censoring. For example, the paper is on hunter-gatherers but at least five groups in Table 1 are not hunter-gatherers. Here is the link: https://www.vivekvenkataraman.com/blog/2023/7/5/debunking-a-debunking
Venkataraman is an assistant professor in the Department of Anthropology and Archaeology of the University of Calgary. His university webpage describes his interests:
Dr. Venkataraman is an evolutionary anthropologist who is broadly interested in the evolution of the human diet and food systems, and their relation to life history and behavior. He is assistant director of the Guassa Gelada Research Project ,and also the co-founder and co-PI of the Orang Asli Health and Lifeways Project (OAHeLP)
As I said I’d keep people posted as analysis of the Anderson et al. paper proceeds, here’s the latest (click to read):
The general tenor of all the criticism so far is that Anderson et al. selected a group of hunter-gatherer societies that wasn’t random, and was weighted towards those in which women hunted. Further, the critics have argued that Anderson et al.’s literature analysis was slipshod, exaggerating the propensity of women who did hunt to take bigger game. This paper (and remember that it is neither submitted nor peer-reviewed) makes the same criticisms. I’ll give a few excerpts, all indented. Where bolding is mine, I say so.
We present a critique of the methodology employed by Anderson et al (2023) in their study of women’s hunting in foraging societies. When this new article was published, my students (undergraduate Jordie Hoffman and master’s student Kyle Farquharson) and I were putting the final touches on our own cross-cultural study of women’s hunting. This paper will be posted as a pre-print shortly. Given our recent immersion in this literature, we feel well-positioned to comment on this new study. We have identified a number of issues with Anderson et al’s (2023) methodology. While we applaud their investigation of this important issue, and we thank the authors for drawing our attention to sources that we ourselves had not discovered, we believe the main conclusions of their paper are invalidated by these issues.
We argue two main points about this paper:
-The sampling methodology is likely biased toward reports of female hunters, producing an inflated estimate of the frequency of women’s hunting (80%). More realistic estimates are offered.
-The claim that women commonly participate in the hunting of large game does not hold up under closer scrutiny.
This critique is by no means exhaustive. It does not deal with other significant problems about this paper, including the misrepresentation of consensus views on divisions of labor in foragers and of the intellectual history of Man the Hunter (Venkataraman 2021). Other critiques may be read here and here.
Here is their own range of estimates of hunter-gatherer societies in which women hunted:
Anderson et al (2023) argue that 50/63 (80%) forager societies show evidence of women’s hunting. For the estimate of 80% to hold up across all foragers, their sample of 63 societies would need to be representative of the broader literature. However, there are a number of societies with detailed information on subsistence behavior, including hunting, that were not included in their study. These might be found in sources with large compilations of foraging behavior such as Kaplan et al (2000) and Kraft et al (2021), among others. Dozens of such societies could appear in such a list. Because there are so many populations with ‘explicit’ – if not in-depth – descriptions of foraging returns, we are not confident that Anderson et al’s (2023) sampling strategy was not biased toward reports of women hunters. The lack of description in their methods does little to allay this concern.
ESTIMATING THE FREQUENCY OF WOMEN’S HUNTING [JAC: the bolding below is mine]
Based on our forthcoming paragraph-level survey of women’s hunting (Hoffman et al. in prep) using the Human Relations Area Files (HRAF), we found that reports of women’s hunting were genuinely rare. Our analysis revealed only 232 paragraphs across 53 societies in this large database. If we consider data from the Standard Cross-Cultural Sample (SCCS), which consists of 186 societies, we found that 29 societies had evidence of women’s hunting. This translates to a frequency of 15.5%( 29/186). We also conducted a search on the D-PLACE database, focusing on the “Sex Difference: Hunting” variable. Out of 965 societies, only 14 (1.4%) were coded for the presence of women’s hunting. Of those 14 groups, 12 were hunter-gatherers, and the other two were agriculturalists. If we treat 965 as a reasonable denominator for the number of societies and consider the 53 societies we found, this would raise the frequency of women’s hunting to 5.5% (53/965). This estimate covers all subsistence types. If we consider only the forager cases (n=391), the frequency would be calculated as 13.5% (53/391). Finally, Koster et al (2020) synthesized data on hunting across 40 small-scale societies. There was little data on hunting by women. Though the authors caution against generalizing about sex differences based on low female sample sizes, it is important to recognize that much of the data in that study came from well-studied populations (e.g. Ache, Batek, Pume). This would suggest the behavior is indeed uncommon.
Our search of reports on women’s hunting using the SCCS in the HRAF revealed nearly an 80% match with the ethnographies found by Anderson et al (2023). This convergence suggests our study and that of Anderson et al (2023) have synthesized the vast majority of ethnographic reports on women’s hunting, and do not support the idea that there is a large unsampled literature on the topic. Therefore, assuming that 80% of the 328 societies unsampled by Anderson et al (2023) will also show evidence of women’s hunting is not warranted.
In summary, the frequency of women’s hunting depends on the type of calculation performed. We argue that any reasonable calculation puts this number well below 20%. Our findings confirm previous cross-cultural work on the topic and contrast starkly with those of Anderson et al’s (2023) estimate of 80%, which, we suggest, likely results from biased sampling.
Venkataraman and Farquharson (henceforth, V&F) then show several coding errors or “problematic cases” in the data of Anderson et al. that seem to show that their original paper recorded “women hunting” when the original data didn’t show it. But V&F also note one paper in which women apparently hunted big game but it wasn’t recorded as such. In general, though, their list of eight “miscodes” tend to exaggerate the frequency of societies in which women hunt. Here are just three cases they point out:
In Table 1, the authors reference Hewlett (1996), which has only one reference to the Ganij in a paragraph that is about breastfeeding. We were unable to locate any mention of women’s hunting.
Memmott et al (2008) focused on fishing techniques and the trapping of fish, dugong, and turtle. The latter two prey types would, in fact, be quite large and perhaps better classified as large game. However, it was not explicitly noted that women were participating in the hunting.
This Bantu ethnic group from Kenya are agriculturalists, not foragers.
If this is true, it’s embarrassing for the original authors.
Finally, the rest of the preprint is a society-by-society analysis of whether women hunted large game. The consensus of anthropologists before Anderson et al.’s paper is that this was quite rare, and V&F support that conclusion, by criticizing Anderson et al.’s analysis. I’ll end with V&F’s conclusion (bolding is theirs):
We have critiqued a number of aspects of the Anderson et al (2023) paper. We find, on the whole, that their data do not support their conclusions. We have argued that their sampling methodology suffers from significant problems related to sampling bias. Moreover, a re-examination of their data contradict the conclusion that one-third of their sample of foraging societies show evidence of female participation in large-game hunting. With the exception of a few cases (such as the Agta and some Arctic cases, totaling roughly six), the data do not present compelling ethnographic examples of women participating habitually in large-game hunting; and in virtually every case, hunting is communal. The findings of Anderson et al (2023) do not pose a significant challenge to current consensus views on divisions of labor among foragers.
Here we see a paradigm of how science is done: the use of replication, in this case of literature analysis. First, data are presented, then several authors go through the original data with a fine-tooth comb and conclude that it doesn’t support the original conclusions. No doubt Anderson et al. will respond to this spate of criticism.
One thing that should not be taken from this back-and-forth is that any authors have a substantial investment in finding the results they did. We have no idea of whether there was confirmation bias in any of these papers, though time will tell if there was cherry-picking of data.
Along with that goes a caveat that we shouldn’t say that “men are better” if they hunt more often, or hunt big game. The roles of both men and women even if separated by task, are both absolutely essential for these societies to function. If it turns out that men hunt mor, and take bigger game than Anderson et al concluded, that doesn’t mean that men had a more “important” role.