Wednesday, April 4, 2012

Evolution of cooking

Or 'Food for thought'

What follows is a speculative idea about human evolution that occurred to me about two years ago. I’m not a biologist, but it’s a very simple idea, so its one I feel confident enough about to express it here. What is surprising, though, is that from the limited sources I have seen on the subject, none of the experts working in the relevant field has discussed this idea. I’d be delighted to hear from anybody who actually knows about biology whether the case I make here has anything going for it.

On the 2nd of March, 2010 I watched a BBC ‘Horizon’ documentary outlining a controversial theory detailing the possible impact of the invention of cooking upon human evolution, in particular, suggesting that the cooking of food was originated by earlier hominids than usually accepted, and that this was instrumental in allowing Homo sapiens to evolve. (I made some quick notes the next days, which is why I know the date, but unfortunately I did not record the title of the program.)

According to wikipedia, ‘most anthropologists believe that cooking fires began only about 250,000 years ago.Here, however, you can read a brief summary of some evidence that cooking may have begun about 1.9 million years ago (substantially before H. sapiens appeared).

The main benefits of cooking food mentioned in the Horizon program were:

§     greatly increases the amount of energy gained from the food: > 20 % energetic profit for some foods studied
§     reduces the amount of energy needed to digest the food

These advantages are significant (and can be added to by the enormous saving in time required to chew a large meal of raw meat, compared to its cooked equivalent). One might well postulate that such advantages could have significant evolutionary effects. In particular, a reduction in size of the gut, and concomitant increase in size of the brain (not a necessary consequence, but perhaps enabled by it).

One of the major objections to the theory seems to be (according to the program) that the practise of cooking may not have been around for long enough to affect evolutionary change.

It seems to me, however, that a major piece of evidence on this front was overlooked by the program and by the experts who contributed to it. One of the professors advocating the theory even alluded to this evidence, but in a totally backwards way. In casually speculating on how the practice of cooking may have got started, he said something like “a man accidentally drops his meat into a fire, takes it out and tries it, only to find that it is delicious.” My question to that expert is this: why on Earth would that man have found the cooked food delicious?

Now a reasonable hypothesis space concerning the question of why cooking makes food taste good consists of 4 options:

  1. The molecules in cooked food are intrinsically delicious
  2. The good taste of cooked food comes from the same molecules as are present in raw food – they are just released more readily following the heating process
  3. The experience of enjoyment occurring when we eat a cooked meal is something we learn over the course of our lifetimes, in response to habitually eating such things
  4. The experience of enjoyment occurring when we eat a cooked meal is an evolved response, brought about by natural selection resulting from certain advantages to eating food in this form

Number 1 would seem to be manifestly preposterous.

About number 2, as a non-chemist, I’m going further out on a limb here, but I expect that the molecules contributing to the delicious taste and smell of cooked food are mostly totally different from those that are emitted by and contained in raw food. The Maillard reaction, for example between denatured proteins and sugars requires a temperature of over 150ÂșC, and is responsible for much of what tastes good when we eat cooked meat and some other cooked foods (unless it has been boiled, I suppose).

We could well imagine that an organism regularly eating raw meat would have receptors and neural circuits adapted to provide a pleasurable experience from the taste and smell of raw meat – it would be to its evolutionary advantage. But the smell and taste of cooked meat comes, I expect, not just from more of the same scent molecules, liberated more copiously by the cooking process, but mainly from totally different molecules, produced by the exposure to heat. The first individual to taste cooked meat would, therefore, have had no reason to experience enhanced pleasure, as there would have previously been no selective pressure to evolve any special response to those cooked scent molecules, to which the individual’s ancestors were never exposed.

Option number 3 is harder to quickly refute. It is clear that different nations and cultures have different ideas about what constitutes yummy. We can quickly confirm, however, that at least some reactions to tastes are innate, rather than conditioned. Many natural poisons taste very bad to us, even though we do not receive much childhood exposure to them. Conversely, some plants seem to use noxious tasting chemicals that are not poisonous as a defence against would-be grazers – they seem to make use of an animals innate expectation that something tasting like this is sure to be bad for you. Thus, we can see that a substantial part of an animal’s response to certain tastes is likely not due to learning acquired during the animal’s own experience. As I briefly discuss below, one can imagine experiments that might shed light on the extent of the influence of hypothesis number 3 for the current debate.

It seems to me then, that there is a reasonable argument in favour of possibility number 4, the theory that cooking has been in use for long enough to affect our evolution. Specifically, that it has led to the evolution of a rewarding experience of pleasure when we eat foods that are cooked. This feeling of ‘wow, that piece of scorched flesh was damn delicious,’ could easily be interpreted as a mechanism to overcome that urge: ‘god, I just spent the whole day chasing this bloody rabbit, can I really be bothered starting up a fire?’ The evolutionary advantage of overcoming this lazy urge is that the energy saved by not getting a fire going and preparing a cooked meal is much less than that gained by consuming the food in a partially pre-digested, cooked state.

If, as I suppose, the first people to eat cooked meat got no special pleasure out it (this holds for both hypotheses 3 and 4), one might well ask why they persisted in doing it. One must speculate here, but it is quite easy to hypothesize a good reason. We don’t even need to assume the ability to control fire, though it is quite possible that fire was controlled during the important period. It is known that humans have used fire as a tool for hunting – either for driving game, or as a means of killing game directly. In this way, early humans may have got into the habit of eating scorched foods, simply as a matter of convenience, thanks to this particularly efficient method of hunting. Alternatively, it may have been sufficient for our ancestors to make similar use of fires that started naturally. That there was a huge energetic bonus, on top of the efficiency of the hunting method, may have contributed to the evolution of mechanisms leading to enhanced satisfaction from eating food in this form. Maybe it was enough for some humans to recognise that cooked food is, in many cases, much easier to eat and takes less time to digest.

One aspect of the argument that might lend itself to experimental investigation is the question of whether an animal evolved for eating raw meat would prefer to eat cooked meat, given the opportunity. One might imagine experiments with such an animal placed an equal distance from two sources of hidden food, one cooked and one raw. One then simply needs to observe which food type the animal goes to first. A refinement of the experiment might be to replace the foods with standardized sources of scent molecules corresponding to the two different food types – standardized so that the concentrations in the air are equal for each. We can control for the influence of hypothesis number 3, above, by dividing the test subjects into two groups – one that has previously been fed exclusively on raw food, the other having been fed on cooked food. We could even imagine doing similar experiments on human babies, young enough that they have never tasted any meat: incorporate the flavours of raw and cooked meat into two bowls of rice porridge, and see which they prefer.


  1. I'd say there must be something in a crunchy texture of burnt meat :)

    1. I'm fairly sure Homer Simpson has something to say about that.

    2. ... or H. simpson as he should be known, according to convention