Do we know what paleolithic humans ate?


Warning message

Welcome to the site! This content is old and may not reflect my current opinions. I keep it up mainly for reference and because I hope at least some of it is still good, but I encourage you to check out more recent posts as well as my Start Here page

 Hands down the best health book I read this year was The Definitive H.P. Lovecraft: 67 Tales of Horror in One Volume. Despite being about fictional creatures of terror from unholy abysses, I learned quite a bit from Lovecraft's depiction of the universe. The humans in Lovecraft's stories are baptized into the knowledge that the universe is older and more incomprehensible than they could have ever imagined. While the monstrosities and sublime ancient temples are quite terrifying, what is even more terrifying to the humans in the stories is their realization of how little they can ever really know. Those that get a taste of the mysteries often only do so at a very high price. 

They called up some image from deep cells and tissues whose retentive functions are wholly primal and awesomely ancestral

I'm not sure I have any sort of particular cause in terms of diet anymore. It's gotten to the point where I'm just interested in the Paleolithic and not really very concerned with arguing about whether or not a potato is safe to eat or not. 

Wouldn't it be nice if our nice little narratives worked out? The ones in which Homo sapiens sapiens is the protagonist and you can trace his illustrious evolution neatly through the ages. And he fits rather nicely in your romantic stories about hunters and mammoths so you can tell people that this is their heritage. 

But in reality you don't get your nice story. Instead, you get ages and ages of dust and bones, in which every little shred of a skeleton is a prized, but dim, glimpse into ages long past. 

In my anthropology class last year, one of the skull casts that caught my attention was the Kabwe skull, which is estimated to be between 125,000 and 300,000 years old. Not quite Homo Sapiens, the skull has some features of modern humans and some of Neanderthals. Homo rhodesiensis? Homo heidelbergensis? Homo sapiens rhodesiensis? Anthropologists could argue about it all day. Either way, this person died a miserable death. The first known incidence of dental infection in a hominid as far as I know, and the infection was bad enough to put some ugly holes in the bone and eventually kill the individual. 

There is only so much you can tell from bones, which leaves lots of room for people to make stuff up. Stable isotope analysis seems quite promising, as they can potentially tell you the source of protein in the diet, but they can only tell that and nothing else, and the isotopes are subject to interpretation. For example, Lierre Keith in her error-ridden Vegetarian Myth claims that stable isotope analysis showed Australopithecus africanus ate meat, but in reality the data only said that the protein was from carbon-13 enriched foods, which could include grasses and sedges as well. Later investigations revealed that the carbon-13 probably was more likely from grasses and sedges, but the data is up for interpretation. Before you tear up your lawn to make dinner, it might be worth remembering that Australopithecus africanus is only thought to be a possible human ancestor and was quite a bit different from a modern human. 

That said, stable isotope analysis puts to bed the idea that early Homo sapiens were getting their protein from the Paleolithic equivalent of tofu or the idea that Neanderthals definitely only ate meat (turns out that some ate fish too...maybe). 

"Maybe", "later investigations revealed", "thought to be"- these are things that should give you pause whenever you encounter stable isotopes being used to argue about ancient diets. Have I confused you? Good, now you are less vulnerable to the abuse of bones in the name of various causes one way or another. 

It can be used to estimate the trophic level and origin of the protein, but it cannot tell you whether the person ate a teeny tiny auroch steak and then 17 potatoes or whether they only ate mammoth. It cannot tell you the percentage of protein in the diet. It cannot tell you how much protein in grams. That information was lost when the person died. 

Then there is the use (and mainly misuse) of animal bones and modern data from wild game species to argue various things about ancient diets. I read this latest paper, Man The Fat Hunter, with absolute glee because it uses many of the same questionable methods and comes to an opposite conclusion of many past papers, which overemphasize protein. The questionable method is taking bones of animals possibly consumed by ancient humans and plugging them into an equation with the modern wild game data and then saying this or that about the amount of fat or protein in an ancient diet. In this paper we have elephants featured, which is great, since elephants are very fatty, but unfortunately their presence or absence in bone assemblages is not a food diary. There is no way to know how often elephants were eaten, so there is no way to make an even sort-of accurate conclusion about %elephant and therefore %elephant fat in the diet. Whether or not the hominids in question were able to cook is also a point of contention.

One good thing about the paper is that it does try to address one issue, which is ceilings. In this case, the paper mentions possible ceilings for protein consumption and fiber consumption that could be used to build diet-estimating equations. Unfortunately, there are quite hard to determine, as they are affected by human genetic variation, culture, and environment. For example, there is possible a ceiling on the consumption of raw plant materials based on gut morphology (though if you have only skeletons you can only speculate on this) and toxins, but that ceiling can be raised with access to cooking and processing. To complicate matters further, their food sources may have been things you haven't even thought about eating. You can try to figure it out based on local paleobotany and starch microfossils, which can be hard to read. Once you've established that a microfossil on an ancient tooth is possibly Bromus secalinu, you might be able to figure out a little about how it was processed based on microfossil shape and local conditions and if you have a rich lab you might be able to collect it and do a full nutritional analysis, but you still have no idea how much of the diet it made up. 

And what is the protein ceiling? It depends on the rest of the diet, an individual's health, and possibly genetics. Modern genetics adds some depths to the picture. For example, the fact that genetic adaptations for a starch-based diet seem to be part of fairly recent selective sweeps may give us a clue that Paleolithic human ancestors probably weren't eating mainly starch, but statistical genetics is in its infancy.

But genetic variation can add more confusion if we are talking about what to eat now. Many "paleo" dieters have learned the hard way that they carry alleles for hemochromatosis, which means they can over-accumulate iron, which has some pretty nasty effects. It would be interesting to know where this came from, as it clearly would be a liability if an ancient human ate meat-based diet, but ultimately whether or not Paleolithic hominids carried such alleles in high frequency is irrelevant to the millions of men (and some women) who are at risk. This represents a ceiling for them, though it can be modified through modern medical treatment.

Normal is of limited use if you are on the end of the bell curve- this is where personalized medicine and self-experimentation is important

So while it's not completely true we have no idea what Paleolithic hominids ate. We do have some good clues, but reconstructing the diet is pretty hard. That doesn't stop people from trying, but their results are on some pretty shaky ground. 

My own method, which is about as accurate as some of these equations, is to observe the fact that a medallion of relatively lean wild boar goes absolutely perfectly with a seared hazelnut crust and dollop of mashed celeriac or potatoes cooked in broth. Maybe there is a reason that dishes containing a protein on a bed of delicious carbs AND fat (but not overpowered by them) is so appealing to so many? Who knows.