The headlines are all over the media: The Paleo Diet May Need a Rewrite, Ancient Humans Feasted on a Wide Variety of Plants writes the Smithsonian and defines the Paleo diet as a baseless “fad”. Amanda Henry, a prominent researcher, specializing in the analysis of ancient tooth plaque, was quick to state: ““We need plant-derived nutrients to survive – vitamin C and fibre, for example, Hominins were probably predominantly vegetarians”. The lead archaeologist of the GBY site in Israel, Naama Goren-Inbar, cited the ethnographic record of the Hadza to justify a similar conclusion.
The reason for this burst of media attention in the Paleolithic diet was a paper by Melamed et al., including Goren-Inbar, published at PNAS. The paper describes 55 food plant taxa from the Acheulian site of Gesher Benot Ya’acov in Israel (henceforth GBY – “the bridge of Jacob’s daughters” in Hebrew). We are talking 780,000 years ago and the reason the plant remains were preserved is that the layers were quickly covered by sediment and stayed permanently under water, a very unique situation for an archaeological site.
The results were presented as a big surprise but the drama is, as is the norm these days, highly exaggerated in my opinion. I have some 4000 scientific papers on my hard drive, most of which deal with aspects of the Paleolithic diet. I can’t recall any of them claiming that humans did not eat plants at any time during their evolution. So where is the surprise? Quite made up if you ask me or in another name – a “straw man”.
So does the paleo diet need a rewrite to a largely vegetarian version? My short answer is no. The long answer is listed below.
Malemed et al. identified 9,148 remains from 55 taxa as “plant food”, however what does it say about the relative role of plant food in the diet of the Homo erectus, the species that most probably occupied GBY?
The simple answer is NOTHING. At the same site 1778 bones were identified as belonging to some 41 species of animals from insects through turtles to a nice quantity of ungulates and elephant bones (Rabinovich and Biton, 2011) including an elephant scull in the process of brain extraction (Goren-Inbar et al., 1994). Straight Tusked Elephants, the extinct species that was consumed at GBY, packed some 4 million calories each. This quantity is equivalent to the calories in 50 thousands apples. Elephant was present in each of the 18 layers at the site so we can say that there were at least 18 elephants at the site and probably more. In other words if we assume that each of the 9148 remains of “plant food” represent an apple the total calories that the “plant food” assemblage represent don’t amount to the calories in one elephant and there were at least 18 of those, not to mention tens of other large animals. So can we say anything from these data on the relative consumption of meat in the erectus diet? Unfortunately NOTHING. GBY is only one location where food was consumed by groups of people that wondered constantly about the landscape. We have no way of telling if GBY is representative of their average diet and we can’t even tell that the examined remains represent the relative plant-meat/fat ratio at GBY.
This inability to determine the relative plant-meat/fat ratio from a simple analysis of the archaeological record is quite frustrating to archaeologists. The classical go-to scientific discipline, when it comes to estimating the relative plant-meat/fat ratio, is Ethnography as demonstrated by Goren-Inbar’s comment. Ethnography is very appealing because presumably we have reliable data, gathered recently by scientists, pertaining to groups with a similar lifestyle to Paleolithic hunter gatherers.
Unfortunately nothing can be further from the truth. I have just written a paragraph for my PHD thesis about the falsehood of using the ethnographic record to estimate plant-meat/fat ratios in Paleolithic diets, so I am attaching it to the end of the post for those who wish to read a more detailed analysis. For now I will just state that doubts about the relevancy of the ethnographic record to the Paleolithic lifeways are not new. In 1978 Wobst published a paper titled “The archaeo-ethnology of hunter-gatherers or the tyranny of the ethnographic record in archaeology”. In short, unreliable, anecdotal reporting at the initial contact periods, use of iron and clay tools, use of dogs, intensive long term contact with neighbor herders and farmers, devastation or complete change of adjacent environments and megafauna extinction, all make even the estimate of the historic diets of the present groups completely unreliable, let alone their relation to Paleolithic diets.
The “true” Paleolithic diet will never be “known”. There will always be hypotheses that demand careful judgement of the data and interpretation of the data supporting them. I believe that collecting data from as many scientific fields as possible is the best way to arrive at a good hypothesis. At the same time that Melamed et al. published their paper on GBY, Inchley et al. (2016) published their finding that an increase in the number of gene copies involved in increased consumption of carbohydrates, evolved only after the split of Homo sapiens and Neandertal, meaning that the Homo erectus of GBY probably didn’t have the ability to consume large quantities of starchy plant foods. These findings confirm previous identical findings published last year by Perry et al. ( 2015) that did not get much media attention. I have presented some additional evidence in this direction in 2013 in Atlanta (AHS13). Later, I hope to publish a more comprehensive set of evidence in the form of a scientific paper. At present it seems that the evidence for human evolution toward carnivory is very compelling.
But this is not the main reason we shouldn’t rewrite the Paleo diet as it is practiced today. The best reason not to change it is the unbelievable success of the diet, as it is presently practiced by millions, at eliminating and preventing many of the ills that have mushroomed since the Dietary guidelines were established in 1977. This success may also serve as another evidence that the reconstruction upon which it is based is not so far from the “true” Paleolithic diet.
Bias in ethnographic estimates and Paleolithic predictions of carbohydrates consumption
The debate concerning the relevance of the ethnographical record to historical HG and Paleolithic era reconstruction is old and wide (eg. Wobst, 1978; Wilmsen, 1983; Lee, 1992; Marlowe, 2010). Ethnographic based reconstruction of the Paleolithic diet is very appealing at first glance since it presumably relies on hard evidence, witnessed directly by researchers. However this perception may also provide a false feeling of security in the estimates. Methodological dietetic data gathering at time of contact with hunter-gatherer (HG) groups is completely lacking. For example, Australia is in theory an ideal source of ethnographic data as its local hunter-gatherers had never met pastoralists and farmers prior to contact with Europeans and they, almost uniquely among HG worldwide, were still practicing a stone tools economy at time of contact. Amanda Lee (1996) who performed a comprehensive review of the literature on Australian aboriginal diets, states that little reliable information is available from the fertile temperate seaboard areas which were settled rapidly by Europeans starting 1788 and that reports from desert and other, more remote areas only cover short periods of time. The very limited quantitative data she could find was from studies of groups who reverted to traditionally-oriented lifestyle for experimental purposes for a small number of days. Amanda Lee (1996) came to the conclusion that, contrary to previous assessments, the aboriginal diets were meat oriented with vegetable foods providing an important supplement rather than an alternative to animal foods. Of the 50 or so Australian groups in Marlowe (2005)’s analysis, 8 groups appear in the Ethnographic Atlas (Murdock, 1967) which served as the basis for many of Marlowe (2005)’s data on other groups. However, in stark contrast to Amanda Lee (1996), none of the 8 listed Australian groups is reported as having the majority of its subsistence from hunting. It is clear that whatever data was used for Australia in the Ethnographic Atlas, its reliability is questionable and may be indicative of the questionable reliability of a significant portion of the data in ethnographic databases. However even if the data were accurate its applicability to Paleolithic circumstances is in great doubt. Researchers recognize that there are substantial technological differences between recent hunter-gatherers and Paleolithic groups and Marlowe (2010:Table 4) even lists them in detail and speculates on their effects on subsistence. It is doubtful however that the complex and comprehensive change that the transformation from lithic to metal culture had on all aspects of the Homo HG subsistence can be predicted with any degree of certainty. The Hadza may provide a good case in point since, arguably, it is one of the most researched hunters-gatherers group and consequently has been used extensively to elucidate and support Paleolithic human evolutionary paradigms (eg. Woodburn, 1982; Hawkes et al., 1991; Marlowe, 2010; Hawkes, 2016). The Hadza do not make or use stone tools, a fact that is rarely mentioned, let alone dealt with, in the literature when Paleolithic behavior is deduced from their record. They most probably have relied on traded iron from neighboring farmers for at least 500 years, since Bantu speaking farmers arrived in the area, (Marlowe, 2010:18). At time of contact by European researchers, many of them were speaking their neighbors’ language as a second language (Marlowe, 2010:16). The effects of the absence of stone culture, and reliance on iron instead, on the many aspects of HG behavior, and indeed economy, is a subject for deeper analysis than is possible here, so just two examples of the many questions which are relevant to diet will be mentioned. Has the time saving achieved by using iron arrows led the Hadza to a higher efficiency of hunting and therefore higher quantities of meat compared to the Paleolithic as Marlowe speculates? Couldn’t one also speculate with equal degree of certainty (or uncertainty) that the time saved by using metal would have allowed more plant gathering which is much less time efficient than hunting (Stiner and Kuhn, 2009). Also, wouldn’t plant consumption benefit significantly from metal knifes and metal (and clay) containers for cooking (Marlowe, 2010:Table 4.2)? In ethnographic settings bow and arrows are used to hunt mainly smaller, faster prey (Churchill, 1993). Dogs, which many H-G groups (including the Hadza) possess, may also enable hunting of smaller faster animals rather than increasing the total caloric contribution of hunting, compared to a situation where large animals are more prevalent. In the Levant, for example, the Upper Paleolithic and Epipaleolithic, when presumably bows and arrows were first used and dogs domesticated, is actually associated with evidence of a reduction in the average size of hunted fauna (Bar-Yosef, 2014:263) and increased consumption of plant foods (Bar-Yosef, 2014:260). Marlowe (2010:35,36) states that the availability of prey animals and specifically large animals in the Hadza territory diminished and ventures that they probably ate more meat in historic times. Similar evidence exist for other HG groups (eg. Hill et al., 2003). In another example, Lee (1979), who also documented a high plant diet for the San, quote a 70 year old /Xai/Xai San man so: “When I was young the elephants, buffalo, and rhino were thick at /Xai/Xai. Before I was born white hunters would visit /Xai/Xai and shoot the elephants with guns”. Another witness describes the /Xai/Xai area as a place where he went to hunt elephants, and hippopotamuses and waterfowl abounded. Today the area is a “dustbowl” in the words of Lee. On the effect of the arrival of pastoral herders to the San area Malherbe (1983:36) writes: “After a while the San noticed that livestock scared the game away, damaged the veldkos (plant food) and made the water dirty. Their own food supply decreased…”. The evidence may amount to a complete replacement of a faunal intensive ecology by a faunal depleted ecology for the San. A similar process must have begun for the Hadza at least 200-300 years ago with the arrival of Nilotic-speaking cattle herders from Sudan, gradually taking over the best foraging spots (Marlowe, 2010:17,18). Elephants were hunted for their tusks by foreign hunters for more than 100 years in the area (Wilmsen, 1983 with regard to the San; Marlowe, 2010:19,30 with regard to the Hadza). One of the most significant consequences of the diminishing prey and especially diminishing larger prey was that the three largest animals that were hunted in the past by the Hadza; rhino, hippo and elephants, are no longer hunted today. One hippo may be equal to 32 impalas in caloric terms with elephants two to three times that much, while the time taken to hunt one of these large animals is most probably not much longer than that of hunting smaller animals. Just one of these animals can satisfy a group’s needs for many days in which they could and probably did eat nothing but meat and fat, with a potential to completely change the recent hunting/gathering ratio. Similarly, large game extinction took place in most of the world outside Africa starting some 40-50 thousands years ago and peaking some 10 thousands years ago (Koch and Barnosky, 2006; Gillespie, 2008; Sandom et al., 2014). These extinctions must have affected the risk-reward ratio of hunting globally in a manner that significantly reduced the relative hunting contribution in many HG societies. The importance of large animals to hunting yield can be demonstrated by the fact that although the number of animals hunted by the Hadza was approximately equal between large and small animals (Marlowe, 2010: figuer 8.7), the contribution of large animals to the total weight of hunted meat was close to 90%. Another intensively researched group, the Ache, which reportedly consume a high meat diet (Hawkes et al., 1982), live permanently around a Christian mission which supply them milk, sugar, rice etc. to supplement what they grow agriculturally and obtain during hunting trips in an area to which they have been transferred by the government. In summary, using Ethnographic evidence to infer Paleolithic human behavior trends may sometimes yield cautiously acceptable results but gauging a typical plant/animal ratio in their pre-contact diets can be utterly speculative and more probably misleading by providing false certainty as it doubtfully represent their own historical diets and in addition contain no “Paleo” parallel in the ecological sense of prevalence of large fauna, nor does it contains a “lithic” parallel in terms of basic technology.
The ethnographic record may provide however useful clues as to general trends and preferences, especially if a trend crosses groups from different ecological and historical background (Kuhn and Stiner, 2001). Kuhn and Stiner identified a common ranking of food items with large game animals on top and plants at the bottom. They also identified an association of the plant-animal ratio of food sources with latitude showing a decline in dependence on plant food away from the Equator. Tellingly, a high plant food dependency is associated with reduced availability of large animals and increased population density rather than an increase in plant food availability (Kuhn and Stiner, 2001). This phenomenon means that high plant diets are consumed as a result of ecological stress rather than choice and support our notion that they are a result of the large animals’ extinction during the UP. A similar trend with regard to diets was identified by Cordain et al. (2000) stating that “whenever and wherever it was ecologically possible hunter-gatherers consumed high amounts (45-65%) of animal foods”. Another interesting trend that Cordain et al. (2000) identified was that although the simplest solution to the physiological ceiling on protein consumption would have been gathering more plants, data from the Ethnographic record clearly indicate that the preferred solution was acquiring more fat. This conclusion is supported by a study (Ben-Dor, 2015) of the cultural attitudes of many traditional, geographically widespread, societies, including HG groups, toward fat. The study found that all the studied groups attribute extreme importance to animal fat in many life dimensions including, for example, fertility and the life source itself. In a more general manner Kelly (2013) states in his seminal book on HG that they prefer fat as a food source and that it may be animal fat rather than protein that drive the desire for meat in HG societies. Since fat and plants serve as alternative protein releasing foods (Speth and Spielmann, 1983) this preference for fat signals a strong preference for animal food.
Comparing the ethnographic record with the late UP record and the MP record in terms of material and behavioral trends that are associated with food acquisition and consumption, Kuhn and Stiner (2001) found that the ethnographic record is a continuation of trends that were first practiced in the LUP but that it doesn’t seem to have anything in common with the MP, especially in regard to plant consumption. Since the late UP represents a period in which humans existed for 1.8 million years, it can be safely argued that relying on the ethnographic record as a representative of the conditions during human evolution is not bore by the evidence.
Because of the centrality of nutrition in humans’ (and any animal’s) life, a wide variety of hypotheses and conclusions can potentially lose their validity if they base themselves on the poorly documented and poorly representative ethnographic record. The importance of a sound identification of a diet pattern and magnitude of the potential inaccuracy of the ethnographic prediction in the anthropologic literature can be illustrated by the fact that Cordain et al. (2000)’s paper in which they estimated HG diet based on Murdoch Atlas have been cited over 600 times. Examples of just two of the papers that cited Cordain et al. 2000 can demonstrate typical hypotheses and conclusion that will probably require reconsideration. One is the “Grandmother” hypothesis (Hawkes and Coxworth, 2013) of the reason for humans’ extended longevity which assumes high gathered component in Early Paleolithic diets in East Africa and another is a paper (Rodríguez-Gómez et al., 2013) concluding a lack of competitions between humans and carnivores in Atapuerca, Spain, 1.2 Mya, based on an estimate of animal sourced food in some groups in Cordain et al. (2000) paper. Another example is attempts to predict plant/animal ratio based on comparison of dental microwear and macrowear with HG groups of different diets. Thus, a more substantiated estimate of the plant/animal ratio of Paleolithic humans has the potential to lead to a significant review of the scientific literature and open the doors for new hypotheses and conclusions in a wide spectrum of anthropological research.
Bar-Yosef O. 2014. Upper Paleolithic Hunter-Gatherers in Western Asia. In: Cummings V, Jordan, P., Zvelebil, M., editor. The Oxford Handbook of the Archaeology and Anthropology of Hunter-Gatherers. Oxford: Oxford University Press.
Ben-Dor M. 2015. Use of Animal Fat as a Symbol of Health in Traditional societies Suggests Humans may be Well Adapted to its Consumption. Journal of Evolution and Health 1(1):10.
Churchill SE. 1993. Weapon technology, prey size selection, and hunting methods in modern hunter‐gatherers: implications for hunting in the Palaeolithic and Mesolithic. Archeological Papers of the American Anthropological Association 4(1):11-24.
Cordain L, Miller JB, Eaton SB, Mann N, Holt SHA, and Speth JD. 2000. Plant-animal subsistence ratios and macronutrient energy estimations in worldwide hunter-gatherer diets 1 , 2. The American journal of clinical nutrition 71(3):682-692.
Gillespie R. 2008. Updating Martin’s global extinction model. Quat Sci Rev 27(27):2522-2529.
Goren-Inbar N, Lister A, Werker E, and Chech M. 1994. A butchered elephant skull and associated artifacts from the Acheulian site of Gesher Benot Ya’aqov, Israel. Paleorient:99-112.
Hawkes K. 2016. Ethnoarchaeology and Plio-Pleistocene sites: Some lessons from the Hadza. JAnthArch.
Hawkes K, and Coxworth JE. 2013. Grandmothers and the evolution of human longevity: a review of findings and future directions. Evolutionary Anthropology: Issues, News, and Reviews 22(6):294-302.
Hawkes K, Hill K, and O’CONNELL JF. 1982. Why hunters gather: optimal foraging and the Ache of eastern Paraguay. Amer Ethnol 9(2):379-398.
Hawkes K, O’connell JF, Jones NB, Oftedal O, and Blumenschine R. 1991. Hunting income patterns among the Hadza: big game, common goods, foraging goals and the evolution of the human diet [and discussion]. Philosophical Transactions of the Royal Society B: Biological Sciences 334(1270):243-251.
Hill K, McMillan G, and Fariña R. 2003. Hunting‐Related Changes in Game Encounter Rates from 1994 to 2001 in the Mbaracayu Reserve, Paraguay. Conserv Biol 17(5):1312-1323.
Inchley CE, Larbey CD, Shwan NA, Pagani L, Saag L, Antão T, Jacobs G, Hudjashov G, Metspalu E, and Mitt M. 2016. Selective sweep on human amylase genes postdates the split with Neanderthals. Sci Rep 6.
Kelly RL. 2013. The lifeways of hunter-gatherers: The foraging spectrum: Cambridge University Press.
Koch PL, and Barnosky AD. 2006. Late Quaternary Extinctions : State of the Debate.215-252.
Kuhn SL, and Stiner MC. 2001. The antiquity of hunter-gatherers. In: Panter-Brick C, Layton R, and Rowley-Conwy P, editors. Hunter–Gatherers: Interdisciplinary Perspectives. Cambridge: Cambridge University Press. p 99-142.
Lee A. 1996. The transition of Australian Aboriginal diet and nutritional health. Metabolic Consequences of Changing Dietary Patterns: Karger Publishers. p 1-52.
Lee RB. 1979. The! Kung San: men, women, and work in a foraging society: Cambridge University Press Cambridge.
Lee RB. 1992. Art, Science, or Politics? The Crisis in Hunter‐Gatherer Studies. Amer Anthrop 94(1):31-54.
Malherbe C. 1983. These Small People: Shuter & Shooter.
Marlowe F. 2010. The Hadza: Hunter-gatherers of Tanzania: University of California Press. 325 p.
Marlowe FW. 2005. Hunter‐gatherers and human evolution. Evolutionary Anthropology: Issues, News, and Reviews 14(2):54-67.
Murdock GP. 1967. Ethnographic atlas: a summary. Ethnology 6(2):109-236.
Perry GH, Kistler L, Kelaita MA, and Sams AJ. 2015. Insights into hominin phenotypic and dietary evolution from ancient DNA sequence data. J Hum Evol 79:55-63.
Rabinovich R, and Biton R. 2011. The Early–Middle Pleistocene faunal assemblages of Gesher Benot Ya ‘aqov: Inter-site variability. J Hum Evol 60(4):357-374.
Rodríguez-Gómez G, Rodríguez J, Martín-González JÁ, Goikoetxea I, and Mateos A. 2013. Modeling trophic resource availability for the first human settlers of Europe: The case of Atapuerca TD6. J Hum Evol 64(6):645-657.
Sandom C, Faurby S, Sandel B, and Svenning J-C. 2014. Global late Quaternary megafauna extinctions linked to humans, not climate change. Proc R Soc B: The Royal Society. p 20133254.
Speth JD, and Spielmann KA. 1983. Energy source, protein metabolism, and hunter-gatherer subsistence strategies. JAnthArch 2:1-31.
Stiner MC, and Kuhn SL. 2009. Paleolithic diet and the division of labor in Mediterranean Eurasia. The Evolution of Hominin Diets: Springer. p 157-169.
Wilmsen EN. 1983. The ecology of illusion: anthropological foraging in the Kalahari. Reviews in Anthropology 10(1):9-20.
Wobst HM. 1978. The archaeo-ethnology of hunter-gatherers or the tyranny of the ethnographic record in archaeology. AmerAnt:303-309.
Woodburn J. 1982. Egalitarian societies. Man:431-451.