Posts Tagged ‘diet’

A Recipe for Sentience: The Energetics of Intelligence

“No man can be wise on an empty stomach.”

- Mary Anne Evans, under the pseudonym George Eliot


We humans have been suffering from a bit of a self-image problem for the last half century.

First we were Man the Tool-Maker, with our ability to reshape natural objects to serve a purpose acting to  separate us from the brute beasts.  This image was rudely shattered by Jane Goodall’s discovery in the 1960s that chimpanzees also craft and use tools, such as stripping leaves from a twig to fish termites out of their nest to eat, or using the spine of an oil palm frond as a pestle to pulverize the nutritious tree pulp.

Then we were Man the Hunter.  We’d lost our tool-making uniqueness but we still had our ability to kill, dismember, and eat much larger animals with even simple tools, and it was thought that this ability unlocked enough energy in our diet to fuel the growth of larger body size and larger brains1.  This idea rather famously bled into popular culture and science fiction of the time, such as the opening to the movie 2001: A Space Odyssey.  However, we would later find out that although it is not a large component of the diet, chimpanzees eat enough meat to act as significant predators on other primates in their forest homes.  We would also find out that the bone piles we had once attributed to our ancestors belonged to ancient savannah predators, and that the whole reason hominid bones showed up in the assemblage at all is because we were occasionally lunch.

So meat eating by itself doesn’t seem to make us as distinct from our closest living relatives as we had previously thought, and the argument of what makes us special has since moved on to language.  That does leave a standing question, though: if it wasn’t meat-eating that allowed us to get bigger and more intelligent, what was it?

While there is evidence in the fossil record that eating raw meat allowed humans to gain more size and intelligence, it is both unlikely that we were the hunters and that this behavioral change was enough to unlock a significant jump in brain size.  Instead, there is another hypothesis and human identity that has been gaining more traction as of late: the concept of Man the Cooking Animal, the only animal on Earth that can no longer survive on a diet of raw food because of the energy demands of its enormous brain2.

Napoleon is famously said to have declared that an army marches on its stomach (at least, after what may be a loose translation).  That is, the power of an army is limited by the amount of food that a society can divert to it.  What we have come to realize more recently is that this same limitation exists inside the body, be it human, animal, or speculative alien species.  No matter what the diet, a creature will only have a fixed amount of energy available to divert to activities such as maintaining a warm-blooded body temperature (homeothermy), digestion, reproduction, and the growth and maintenance of tissues.  We can track some of these changes in the human line in the fossil record, but others must at best be more speculative due to the difficulty of preserving evidence of behavioral changes (which of course, do not fossilize) as well as limited research on modern examples.  We’ll start by looking at the evolutionary pathway of humans to see what information is currently available.


 The Woodland Ape and the Handy Man


Size comparison of Australopithecus afarensis and Homo sapiens (by Carl Buell)

Some of the oldest human ancestors that we can unequivocally identify as part of our line lie in the genus Australopithecus.  These have been identified by some authors as woodland apes, to distinguish these more dryland inhabitants from the forest apes that survive today in Africa’s jungles (chimpanzees, bonobos, and gorillas).  They are much smaller than a modern human, only as tall as a child, but they have already evolved to walk upright.  They still show adaptations for climbing that were lost in later species, suggesting they probably escaped into the trees at night to avoid ground predators, as modern chimps do.  Their brains were not much larger than a modern chimpanzee’s, and their teeth are very heavy, even pig-like, as an adaptation to a tough diet of fibrous plant material – probably roots, tubers, and corms, perhaps dug from plants growing at the water’s edge2,3.

The hominids thought to have first started eating meat are Homo habilis, the “handy man”, and the distinction between them and the older Australopithecus group from which they descended are not very large.  The two are close enough that Homo habilis has been suggested it might be more properly renamed to Australopithecus habilis, while the interspecies variation suggests to some researchers that what we now call habilis may represent more than one species4Whatever its proper taxonomic designation, H. habilis shows a modest increase in brain size and evidence that it was using simple stone tools to butcher large mammals, probably those left behind by the many carnivorous mammals that lived on the savannahs and woodlands alongside it.

The transition between H. habilis and H. erectus is far more distinctive, with a reduction in tooth size, jaw size, and gut size, and an increase in brain volume.  They are also believed to have been larger, but the small number of available hominid fossils makes this difficult to verify.  H. erectus is also the first human to have been found outside of Africa.  While the habilis-erectus split has been attributed to the eating of significant amounts of meat in the Man-the-Hunter scenario (recall that habilis, despite its tool-using ability for deconstructing large animals, does not appear to have hunted them), the anthropologist Richard Wrangham has suggested that the turnover instead indicates the first place at which humans began to cook2,3.  Because the oldest solid evidence of cooking is far younger than the oldest known fossils of erectus, what follows is largely based on linking scraps of evidence from modern humans and ancient fossils using what is known as the Expensive-Tissue Hypothesis.


 Brains versus Guts: The Expensive-Tissue Hypothesis

 The Expensive-Tissue Hypothesis was first proposed by Leslie Aiello and Peter Wheeler in 19955, and it goes something like this.  Large brains evolve in creatures that live in groups because intelligence is important to creating and maintaining the social groups.  This is known as the social brain hypothesis, and it helps to explain why animals that live socially have larger brains than their more solitary relatives.  However, not all social primates, or even social animals, have particularly large brains.  Horses, for example, are social animals not known for their excessively large brain capacity, and much the same can be said for lemurs.  Meanwhile, apes have larger brains than most monkeys.  This can’t be accounted for purely by the social brain hypothesis, since by itself it would suggest that all social primates and perhaps all social animals should have very big brains, rather than the variation we see between species and groups.  What does account for the difference is the size of the gut and, by extension, the quality of the diet.

Both brains and guts fit the bill for expensive body tissues.  In humans, the brain uses about 20% of the energy we expend while resting (the basal metabolic rate, or BMR) to feed an organ that only makes up 2.5% of our body weight2.  This number goes down in species with smaller brains, but it is still disproportionately high in social, big-brained animals.  Aiello and Wheeler note that one way to get around this lockstep rule is to increase the metabolic requirements of the species5 (i.e., throw more calories at the problem), but humans don’t do this, and neither do other great apes.  Our metabolic rates are exactly what one would expect for primates of our size.  The only other route is to decrease the energy flow to other tissues, and among the social primates only the gut tissue shows substantial variation in its proportion of body weight.  In fact, the correlation between smaller guts and larger brains lined up quite well in the data then available for monkeys, gibbons, and humans5.  Monkeys and other animals that feed on low-quality diets containing significant amount of indigestible fibers or dangerous plant toxins have very large guts to handle the problem and must expend a significant amount of their BMR on digestion, and have less extra energy to shunt to operate a large brain.  Fruit-eating primates such as chimpanzees and spider monkeys have smaller guts to handle their more easily-digested food, and so have larger brains.  Humans spend the least amount of time eating of any living primate, with equally short digestion times as food speeds through a relatively small gut.  And ours, of course, are the largest brains of all2.

These tradeoffs are not hard-linked to intestinal or brain size, and have been demonstrated in other species.  For example, there is a South American fish species with a tiny gut that uses most of its energy intake to power a surprisingly large brain, while birds with smaller guts often use the energy savings not to build larger brains, but larger, stronger wing muscles2.  Similarly, muscle mass could be shed instead of gut mass to grow a larger brain or to cut overall energy costs.  The latter strategy is the one taken up by tree-dwelling sloths to survive on a very poor diet of tough, phytotoxin-rich leaves, and although it makes them move like rusty wind-up toys it also allows them to live on lower-quality food than most leaf-eating mammals.

Modern humans have, to a degree, taken this approach as well.  When compared to one of our last surviving relatives, H. neanderthalensis, humans have a skeletal structure that paleontologists describe as “gracile:” light bones for our body size, anchoring smaller muscles than our shorter, heavier relatives.  Lower muscle and bone mass in H. sapiens gives us an average energy cost on the order of 1720 calories a day for males and 1400 calories a day for females in modern cold-adapted populations, which are thought to have similar metabolic adaptations for cold weather as the as extinct Neanderthals.  By contrast, H. neanderthalis has been estimated to need 4000-7000 calories a day for males and 3000-5000 calories for females, with the higher costs reflecting the colder winter months6.


Cooked versus Raw


Tribe of Homo erectus cooking with fire (from

At the point where human brain size first increases dramatically (H. erectus, as you might recall), both guts and teeth reduce significantly while the brain increases.  The expensive tissue hypothesis explains the tradeoff between guts and brains, but cooking provides a possible explanation for how both the teeth and the guts could reduce so significantly while still feeding a big brain.

Data on the energetics of cooked food are currently limited, but the experiments that have been performed so far indicate that the softer and more processed the food the more net calories are extracted, since less calories need to be spent on digestion.  A Japanese experiment with rats showed that they gained more weight on laboratory blocks that had been puffed up like a breakfast cereal versus rats on normal blocks, even though the total calories in the food were the same and the rats spent the same amount of energy on exercise2.  Similarly, experiments with pythons show that they expend about 12% more energy breaking down whole meat than either meat that has been cooked or meat that has been finely ground.  The two treatments reduce energy cost independently of each other, meaning that snakes fed ground, cooked meat used almost 24% less energy than pythons fed whole raw meat or rats2.

There is even less data on how humans utilize cooked food versus raw food.  Because it only recently occurred to us that we might not be able to eat raw food diets like other animals, only a few studies exist.  So far the most extensive is the Giessen Raw Food study performed in Germany, which used questionnaires to collect data from 513 raw foodists in Germany who eat anywhere from a 75% to 100% raw food diet.  The data are startling.  Modern humans appear to do extremely poorly on diets that our close relatives, the forest apes, would get sleek and fat on.  Body weights fall dramatically when we eat a significant amount of raw food, to the point where almost a third of those eating nothing but raw had body weights suggesting chronic energy deficiency.  About half of the women on total raw food diets had so little energy to spare that they had completely ceased to menstruate, and 10% had such irregular cycles that they were likely to be completely unable to conceive at their current energy levels2.   Mind you, these are  modern first-world people with the advantage of high-tech processing equipment to reduce the energy cost of eating whole foods, far less energy expenditure required to gather that food, and a cornucopia of modern domestic plants that have been selectively bred to produce larger fruits and vegetables with and lower fiber and toxin contents than their wild counterparts.  The outcome looks more dismal for a theoretical raw-food-eating human ancestor living  before the dawn of civilization and supermarkets.


Fantastic Implications

What this all ultimately suggests is that there are tradeoffs in the bodies of intelligent creatures that we may not have given much consideration: namely, that to build a bigger brain you either need a much higher level of caloric intake and burn (high BMR) or the size and energy costs in something in the body have to give.  Certain organs do not appear to have much wiggle room for size reduction, as Aiello and Wheeler discovered; hearts for warm-blooded organisms need to be a certain size to provide enough blood throughout the body, and similarly lungs must be a particular size to provide enough surface area for oxygen to diffuse into the blood.  However, gut size can fluctuate dramatically depending on the requirements of the diet, and musculature can also reduce to cut energy costs.

Humans seem to have done an end-run around some of the energy constraints of digestion by letting the cultural behaviors of cooking and processing do the work for them, freeing up energy for increased brain size following social brain hypothesis patterns.  This is pretty classic human adaptive behavior, the same thing that lets us live in environments ranging from arctic to deep desert, and should therefore not come as a great surprise.  It does, however, give us something to think about when building intelligent races from whole cloth: what energy constraints would they run up against, and assuming they didn’t take the human path of supplanting biological evolution with culture, how would they then get around them?

You're going to need to cook that first. (From

Fantasy monsters and evil humanoids in stories tend to be described as larger and stronger than humans (sometimes quite significantly so) and as raw meat eaters, particularly of humanoid meat.  There’s a good psychological reason for doing so – both of these characteristics tap into ancient fears, one of the time period not so long ago when humans could end up as prey for large mammalian predators, and the other a deep-seated terror of cannibalism without a heavy dose of ritualism to keep it in check.  However, both the Neanderthal example and the Expensive Tissue Hypothesis suggest that such a species would be very difficult to produce; there’s a very good reason why large mammalian predators, whatever their intelligence level, are rare.  It wouldn’t be a large shift, however, to take a monstrous race and model them after a hybrid of Neanderthal and grizzly bear, making them omnivores that can supplement their favored meat diet with plant foods and use cooking to reduce the energy costs of digestion.  Or perhaps their high caloric needs and obligate carnivory could become a plot point, driving them to be highly expansionistic simply in order to keep their people fed, and to view anything not of their own race as a potential meal.

On the science fiction front, it presents limitations that should be kept in mind for any sapient alien.  To build a large brain, either body mass has to give somewhere (muscle, bone, guts) or the caloric intake needs to increase to keep pace with the higher energy costs.  Perhaps an alien race more intelligent than humans would be able to do so by becoming even more gracile, with fragile bones and muscles that may work on a slightly smaller, lower-gravity planet.  Or perhaps they reduce their energy needs by being an aquatic race, since animals that swim generally use a lower energy budget for locomotion than animals that fly or run7.

From such a core idea, whole worlds can be spun: low-gravity planets that demand less energy for terrestrial locomotion; great undersea empires in either a fantastic or an alien setting, where water buoys the body and reduces energy costs enough for sapience; or creatures driven by hunger and a decidedly human propensity for expansion that spread, locust-like, across continents, much as we did long ago when we first left our African cradle.

Food for thought, indeed.



1.  Stanford, C.B.,  2001.  The Hunting Apes: Meat Eating and the Origins of Human Behavior.   Princeton, NJ: Princeton University Press.

2. Wrangham, R., 2009.  Catching Fire: How Cooking Made us Human.  New York, NY: Basic Books.

3. —-, 2001.  “Out of the Pan, into the fire:  from ape to human. ”  Tree of Origin: What Primate Behavior Can Tell us About Human Social Evolution.  Ed.  F.B.M. de Waal.   Cambridge, MA:  Harvard University Press.   119-143.

4. Miller, J.A., 1991.  “Does brain size variability provide evidence of multiple species in Homo habilis?”  American Journal of Physical Anthropology 84(4): 385-398.

5. Aiello, L.C. and P. Wheeler, 1995.  “The Expensive-Tissue Hypothesis: The Brain and the Digestive System in Human and Primate Evolution.”  Current Anthropology 36(2): 199-221.

6. Snodgrass, J.J., and W.R. Leonard, 2009.  “Neanderthal Energetics Revisited: Insights into Population Dynamics and Life History Evolution.”  PaleoAnthropology 2009: 220-237.

7. Schmidt-Nielsen, K., 1972.  “Locomotion: Energy cost of swimming, flying, and running.”  Science 177: 222-228.


Alien Psychology by Diet

Editor’s note: This post was scheduled for last Friday, but something mysterious happened in the technology that we didn’t become aware of right away. Please forgive the delay!

When we expand into outer space, we may contact other sapient species, and it can be guaranteed that they will think differently than us. But we may be able to predict their behavior by their diet. Something that Larry Niven touched on with the carnivorous Kzinti and the herbivorous Puppeteers.

Carnivores: The first thing you should remember is that predators are opportunists, they always take the route that is least expensive. If they get to colonizing other planets I would expect them to have the capability to adapt considering their need for complex ecosystems that would be a hassle to terraform from scratch. Due to their opportunistic natures I doubt that they might go to the trouble of eating other sapient species like us humans, though if there is a massive technological difference they might enslave or domesticate the less advanced species. Relations with humanity would probably be neutral or even allies, we might colonize and terraform dead worlds while they adapt to living ones. However they would likely be very territorial. If we found them as primitives they would probably try eating the explorers, but once that failed would leave them alone unless they managed to communicate, in which case they might end up domesticated (which might make uplifting easier).

Herbivores: To an herbivore, any other animal is a potential enemy, a predator or a competitor. I expect that their planets would be ecological disasters devoid of any other animal species except in the most extreme regions they never got around to colonizing. Since their supporting ecosystems would be comparatively simple they would probably terraform their colonies. Because of that I expect that they would use a lot of weapons of mass destruction in inter-species wars. In addition the fact that extreme paranoia would have been a survival trait in their early history (more than humans anyway) would make diplomacy with them very difficult. If contacted before spaceflight they would inevitably hide or attack, depending on how advanced they were we might have to conquer them or at least quarantine their homeworld, if they haven’t exterminated all their predators yet we might convince them we’re on their side by helping them.

Omnivores: Would probably be closest to humans psychologically, as we are omnivores ourselves. Kind of a wild card, they might terraform, they might adapt, they might exterminate, they might enslave. They might even join forces with us and form the galactic federation, or not. Fortunately (or unfortunately) trends on earth make it seem like most sapient species will be omnivorous.

Plants: I don’t see any reason for plants to become sapient, but maybe a machine civilization would be similar. In short, they wouldn’t care about consumers unless we got in their way. Will definitely be capable of space travel.

Reprinted with permission.