Chapter 2: Design Flaws

There is a prevalent and arrogant notion that humans are outside of nature: that our bodies tend toward dysfunction, steadily succumbing to entropy; that our instincts are primitive and barbaric, and left to them we would suffer. According to this reasoning, we must leverage our clever, medical knowledge to hack our bodies into wellness by correcting imbalances with drugs and supplements, and codifying behaviour with rules we have devised. The truth is, the functioning of our bodies has been honed by millions of years of evolution, including the ability to resist damage (heal) from common assaults. Insofar as we are now afflicted with chronic diseases, this may reflect a mismatch between our modern environment and our evolved environment. One environmental factor with particularly high potential to affect our physical health is food. As such, many of our current medical issues may be largely the result of radical divergence from the food we evolved eating and relentless exposure to inappropriate, or even toxic food input.

Naturally unnatural

A car is an amazing feat of engineering. If you have one, you know that even a new car requires frequent maintenance to keep it running smoothly. Parts need replacing, as they wear down with use. Lubricators must be added, residues removed. Some of us take proactive steps aimed at extending the car’s “life”, but a car isn’t literally alive, of course, even if you’ve given it a pet name and coaxed it with loving incantations.

If a car were alive then we would expect it to have an innate ability to maintain itself, provided it had access to required materials and the opportunity to sense and act on its own need for more stress (we do need some stress to function and grow!) or more recuperation time. Self-maintenance, that is, healing, is a fundamental process of living things. That is not to say that all damage is reversible. Obviously some wounds cannot heal, and the process of aging is sometimes understood as an accumulation of damage that ultimately we cannot recover from. Nonetheless, living things routinely repair themselves. Those that don’t do this well are evolutionarily outcompeted by those that do.

A fundamental mistake many make is to treat the human body more like a car than a living organism. Maybe this comes from our general confusion about the concept of “natural”. The meaning of the word “natural” is elusive. Humans are animals, and therefore everything we do is, in a certain sense, natural, by definition. Yet, the things we most often consider unnatural are products of the human mind. In many cases, the distinction we are trying to make when we use the word natural is to contrast it with “on purpose” or “designed”. This distinction between design and nature figures prominently in our understanding of evolutionary adaptation, and not just in the sense of compatibility or incompatibility with creationism!

An organism’s form and function is normally thought of as dependent on its natural habitat, but humans change their habitats in extraordinary, unprecedented ways; our environments are largely products of our own designs. This puts us in a strange and unique position in an evolutionary sense.

What does selection select?

Evolution proceeds in large part by natural selection. This process tends to shape a species to be better and better equipped to survive and reproduce. It’s just a consequence of organisms that happen to be better suited, or “fitter”, being more likely to survive and pass on their genes. In some cases, being fitter means enhancing the survival of their offspring who also carry their genes [1]. Given a relatively stable and uniform aspect of an environment, like climate for example, we might expect a species to fit that aspect more and more specifically over time. That is, that aspect of the environment can be said to reward specialisation. In diet, koalas are well-known specialists, because they can eat only specific types of leaves, mainly eucalyptus, that are high in toxins, which koalas have adapted to break down efficiently.

Given an aspect of the environment with higher variability, for example because of seasonality, or because a species migrates frequently among different regions, we expect a species to fit more loosely. Variable environments reward flexibility and generalisation strategies. Racoons are canonical dietary generalists, thriving across two continents in many different ecosystems, from wooded northern climes, to urban garbage scapes, or southwestern deserts.

Sometimes, environments undergo significant changes over short time periods. This can result in an animal that was optimised for the previous environment suddenly no longer being so fit. If these changes are gradual, or if a change is abrupt but then becomes newly stable, species have a fighting chance of adapting. The more specialised the animal was to the original environment, the more fragile it is in the face of abrupt and ongoing changes.

In order to understand what kinds of things can break (or adapt) when an environment changes, it can be useful to think of living things as having these general functions. First there must be some kind of ability to detect a boundary between self and non-self. As the biological philosopher Daniel Dennett puts it, “if you are setting out to preserve yourself, you don’t want to squander effort trying to preserve the whole world: you draw the line.” [2]

Second, we sense and move. At a basic level this means that we are able to distinguish categories of things in the environment in order to approach or withdraw from them. For example a cold-blooded reptile will move into the warmth of the sun. Our senses can also be used to inform contextual processes. For example, in humans, fading light at dusk induces the production of melatonin, a hormone that helps to orchestrate sleep repair processes during the night.

Third, we eat and excrete. That is, we take stuff from the outside, transform it for our own use, and return what’s unused and unwanted. This, of course, relies on sensing and moving. Each species is tuned to distinguish what it needs or can use. Some plants open their leaves and orient them following the sun. Many animals have exquisite senses of smell leading them to food. Some of us have taste receptors on the tongue that guide us toward what to consume heartily and what to spit out.

Finally, we grow or maintain ourselves, and while we might not individually reproduce, we each come from an unbroken line of those who did. Growth is determined by a genetic script and a cascade of signals, but it is also dependent on eating. The specifics of the impetus to grow affect what we seek as food, and that may change at different life stages or through seasons. But the quality of the food we find can also influence our physiological “decision” of whether or not it’s a good time to grow. For example, some animals, such as the nematode worm C. elegans, can put themselves into a form of hibernation called dauer arrest, if they sense too little food in the environment. In this state, they age much more slowly. If, on the other hand they sense abundance, they will seek to reproduce. This ability to change state depending on sensory input allows them to live out famines. We, too, have different metabolic modes that are activated depending on what our bodies sense from our food.

These processes aren’t due to conscious choices, but are simply a series of biological pathways triggered by the environment. That is, they are physiological responses our bodies do by “nature”. In order for a body to do something by nature, it means that those responses were selected for, or at least not selected against. So they were likely adaptive in our evolutionary environment [3]. Therefore, it is precisely these responses that can become maladaptive if the environment changes. So, for example, the C. elegans dauer adaptation works great when the environment shifts from abundance to scarcity and back with regularity. However, one could imagine a new environment that is permanently much less abundant. Now the worm that readily hibernates would be outcompeted by one that is “willing” to reproduce under worse conditions, while the first one waits for a feast that never comes. Similarly, if we find new forms of food that imperfectly match our ability to sense food properties, and they send signals and elicit responses based on biological assumptions we “learned” as a species previously, then ingesting them could have unexpected effects on our entire physiology.

Biological assumptions

These “assumptions” about what the environment is like make results from unexpected input unpredictable. It would be a bit like using an old family recipe but substituting some of the ingredients. It might work, but it might not. For example, some people use ground flax seeds, applesauce, or yogurt to substitute for egg whites in baking, with varying success. However, this would be disastrous for a meringue recipe. Those ingredients may provide some kinds of properties useful in a cake, but they don’t have the chemical structure needed for meringues.

Perhaps an even more apt analogy comes from computer programming. When you enter information into a computer form, on the Internet for example, the program is supposed to take the data you entered, and do something with it, like make a calculation, or store it in a database. Only the right kind of data will work. If you put your name into the “interest rate” field of a mortgage calculator, the calculation can’t proceed in any sensible way. Usually, the computer will just stop and tell you there’s an error in the input. That’s because the programmer who made the calculator thought about what form the data should be in, and made the program verify it was in the right form before continuing. But there are some kinds of data that are “almost” right, and don’t get noticed. For example, maybe the programmer forgot to check for negative numbers, assuming that that would never happen.

In bad cases, computer input assumptions can lead to a type of security vulnerability called a “code injection”. It works more or less like this: Imagine Alice signed a generic contract giving someone the ability to deposit money into her bank account. All that’s needed is for the would-be-donor to fill in his name. The contract she signed says “I hereby permit _______ to deposit cash into my account. Signed, Alice” If you were malicious, you could put in the blank the words “Bob to withdraw $10,000 from and”, so the whole contract now reads “I hereby permit Bob to withdraw $10,000 from and to deposit cash into my account.” Alice was expecting you to put in only a name, but you put a name along with extra instructions!

When a computer has a “code injection” vulnerability, hackers are able to compromise the system by sending unexpected input to a program that causes it to unwittingly interpret it as instructions, not just data! If the input data is in the form the programmer expected, then everything works fine, but if it is in the wrong form, it allows the hacker to change what the computer does, with potentially catastrophic results. Because the vulnerability isn’t normally triggered, it can exist for a long time before it ever causes an issue.

Modern programmers are usually careful to design their programs to avoid this kind of vulnerability, by checking all incoming data and only accepting it if it conforms to certain parameters. But evolution isn’t design! Organisms have no reason or even mechanism to build adaptations to inputs they never see. Physiology doesn’t think ahead to anticipate environmental changes. It doesn’t think at all. Organisms normally only respond to inputs they actually encounter, and survive better or worse for it. Natural selection provides no way to optimise for a hypothetical future environment. If the “input” an animal receives is in some important way different from what it “expects”, the results will be unpredictable.

The medicalised model of health

In light of these evolutionary principles, it seems reasonable to assume as a starting point that for a given bodily response that’s normal for a species now, that response was likely to have been adaptive, or at least not detrimental in the environment it developed under. Yet much of conventional wisdom and modern medicine seem to revolve around the assumption that the human body is haphazard and unregulated.

For example, you may have heard assertions such as “You should drink water before you get thirsty, because once you feel thirsty, you’re already dehydrated.” But think about it. How could a situation like that have possibly developed? Imagine a creature that didn’t become thirsty until the lack of water was detrimental to its functioning. It seems that an inability to accurately detect something as basic as water needs would be selected against. To suggest that waiting until you are thirsty before drinking is bad for your health should at a minimum be accompanied by an explanation of how such a trait could have survived. Likewise, any theory of obesity that suggests our hunger can’t be trusted to tell us when to stop eating seems at least suspect.

Interestingly, there is an evolutionary mismatch argument that contends that our appetites are not properly regulated, and that this dysregulation was actually selected for. The original hypothesis was from James Neel in 1962 [4]. Many others have reiterated versions of it since, but the basic argument, which you’ve probably heard, remains something like this: in our evolutionary past, we were often subject to famines. Therefore, it was advantageous to fatten easily, because those who were fatter were more likely to survive through those famines. So our appetites have been manipulated to tell us to eat more than we need. However, the environment that used to give us regular famines is gone. So our current obesity epidemic comes from unrelenting easy access to food.

Many problems have been identified with this hypothesis, though [5]. For example, the majority of those who die during famines are either beyond reproductive age, so they can’t pass on their genes, or are children. And until very recently, there were not obese children to select for! Moreover, what people die of in famines is typically disease, not starvation, so being fatter might not even help. Fertility rates, which are more directly related to natural selection, are likely to be just as adversely affected in the obese as in the lean, since fertility signals are based on levels of incoming energy, not just body fat levels.

Perhaps more importantly, in humans, as in other animals, the decrease in births seen when food is scarce is compensated for by baby booms when food returns, and there is no reason to believe the booms favour fat people. As we typically see in other animals, when there is more food, rather than individuals eating more, they reproduce more.

The most damaging evidence against the hypothesis might be that even just a few hundred years ago there was not remotely the rate of obesity we see now, even though there were long periods between famines, with ample time for people with the presumed fattening genes to get fat. It’s much more plausible to me that there is something about the quality of modern food that disrupts our hunger regulation, than that our appetite systems never worked properly in the first place and this happens to be the first time it’s been put to test.

At further extremes, we have prevalent theories about chronic diseases that essentially state that our regulatory mechanisms are broken by random genetic accident. The surge in depression and mood disorders is because people are being born with genetic inability to keep their neurotransmitters in balance. Our immune systems are randomly overactive and thus turn against us creating a new prevalence of autoimmune disorders. Cholesterol, a ubiquitous and completely essential substance is suddenly “clogging” our arteries simply because there is too much of it. None of these ideas make much sense in a model of the human body that assumes it evolved to self-regulate. All of them conceive of us as a machine out of tune to be tinkered with. In other words, we are assumed to have design flaws.

We often underappreciate that eating is an interaction with the living things in our environment. We pass food right through our bodies day after day, multiple times. Putting biological substances into our mouths and swallowing them invites intimate chemical interactions. Eating something is unprotected gastrointestinal intercourse, and promiscuity can have undesirable consequences.

Much of the food that we eat now differs from the food we are adapted to eating in many ways. Some of these ways are likely to be unimportant, but others may be extremely important. To overlook or minimise these differences when considering systemic health problems seems myopic.

Many ideas in this book are influenced by the notion that modern diseases come in large part from some form of “evolutionary mismatch”. That is, in most cases we are not so much malfunctioning, but ill-adapted. I hope we can use this paradigm to help reverse some common problems.

Effective treatment does not imply its lack is the cause!

Although I believe that evolutionary mismatch is a leading factor in modern disease, I do not believe that because a diet (or other treatment) resolves a condition it should therefore be held as our “true” adapted diet and that any deviation from it will cause illness, or limit our health potential. Although removing plants from the diet appears to be therapeutic in some conditions, this is not proof that humans are poorly adapted to plant eating, any more than syphilis reveals an evolutionary requirement for penicillin. Indeed, we must be careful with evolutionary arguments, particularly when they are untestable, lest they become “Just So Stories” [6] distorting our judgment with a false sense of confidence. Not only are plausible evolutionary stories far too easy to invent, but not all evolution even proceeds by natural selection.

If the thrust of this book were to attempt to persuade you not to eat plants based primarily on a story about the relative lack of prominence of plants as food in human evolution, it would not only be a flimsy argument, but a questionable motive. Rather, what I would like to convey is that fashionable ideas about the importance of plants in the diet are greatly exaggerated, and that many of the most common varieties and forms of plant foods we eat now are not part of our heritage. They may be contributing to disease epidemics. Even when they are not disease causing on their own, their elimination may help allow the body to heal.

There are many good reasons to expect that a diet high in meat and low in plants can be very healthy for humans, including evolutionary reasons which I will describe in detail. However, humans are much more like the flexible, generalist racoon than the koala. As I will cover in subsequent chapters, there is an important sense in which humans should be viewed as omnivorous. Nonetheless, our ability to thrive without plants — the definition of “facultative carnivore” — makes it a viable treatment option without important downsides characteristic of the current alternatives. Downsides often critically include lack of efficacy. The true efficacy of plant avoidance remains to be seen.

Many non-infectious diseases appear to be on the rise, including those related to insulin resistance such as diabetes and heart disease, autoimmune conditions, and mood disorders. What if the source of these illnesses is not random genetic drift causing random errors, but damage that’s accruing from prolonged attempts to cope with food our systems perceive as foreign. Such damage could also make us vulnerable to smaller insults from foods that would present no challenge to the healthy. In other words, we could become intolerant of foods that we ought to tolerate. Maybe part of the answer to why a plant-free diet has been helpful for so many people, is that it relieves the stress of unexpected input, allowing healing processes to complete, and chronically active components of the immune system, such as inflammation, to retreat.


[1] I make this distinction, because some people insist that evolution doesn’t care about the health or fitness of an organism after it reproduces. This seems misguided. For one thing, it’s not necessarily known by your body whether or not you’ve reproduced. It seems to me that in men, there would be advantage to staying fit and continuing to reproduce as long as possible. In women, the very fact that we stop being able to reproduce at an age far before death suggests that there might be an advantage in allocating resources to the offspring already alive. This is a generalisation of the “grandmother theory”. The grandmother theory comes from the observation that if you put resources into aiding the survival of your offspring’s offspring, this confers a powerful selective advantage to your genes.

[2] Dennett, D C. 1991 Consciousness Explained. Boston: Little, Brown.

[3] In the interest of simplicity, I’m ignoring the possibility of a genetic mutation that affects multiple traits, some of which are adaptive, and others which are neutral, or at least less detrimental than the adaptive trait. Moreover, some traits are simply a result of genetic drift, or constraints independent of the environment. This is important to keep in mind.

[4] Neel JV (1962) Diabetes mellitus: A “thrifty” genotype rendered detrimental by “progress”? Am J Hum Genet 14:353-2

[5] Speakman, J R. “Thrifty Genes for Obesity, an Attractive but Flawed Idea, and an Alternative Perspective: The ‘Drifty Gene’ Hypothesis.” International Journal of Obesity 32, no. 11 (November 2008): 1611–17.

[6] Gould, S J, and R C Lewontin. “The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme.” Proceedings of the Royal Society of London. Series B. Biological Sciences 205, no. 1161 (September 21, 1979): 581–98.


Chapter 3: If It Ain’t Broke