In current MRI studies, the correlation of brain volume and IQ is about 0.4 – and a good thing, too. If there was no significant correlation, we would have to conclude that natural selection didn’t work. Here’s why.
Brain tissue is expensive: altogether it burns up something like 20% of the food we eat. A big brain is expensive in more ways than that: for example, it makes human birth much more difficult and risky than in other mammals. If we could think just as well with a smaller brain, individuals with smaller brains would obviously have higher fitness. They’d do better when food was short, and for them having babies would as easy as spitting out a watermelon seed. Since humans brains are expensive and troublesome, there must be a positive relationship between brain size and brain performance – otherwise natural selection would already left us with shrunken heads.
That doesn’t mean that size is the only factor affecting IQ: obviously there can be differences in neural organization, in the performance of individual neurons, etc. So it is not surprising that the brain size/IQ correlation is significantly above zero while at the same time being well below 1.0. That is just what you would expect.
Of course, it is possible in principle that the brain contributes to fitness in some way other than thinking. Aristotle taught that its primary purpose is to cool the blood, but as Will Cuppy pointed out, this is true only of certain persons.
We know that IQ increases with brain size, on average – but the costs increase as well. It seems likely that costs grow more and more rapidly as brain size increases. If nothing else, the difficulty of giving birth to a big-headed baby must grow rapidly as head size increases. It could also be the case that the payoff of extra brain volume gradually decreases. For one thing, as the brain expands, more and more of its volume has to be used for connections between neurons (white matter), leaving less room for neurons.
This line of analysis suggests that there is an optimum brain size – smaller brains are cheaper but performance is lower still (resulting in lower net fitness), while larger brains have higher performance, but not enough to pay for their higher costs. Of course optimum means optimum for past environments. Changes in the environment – which includes the aspects of the environment created by humans – could move that optimum up and down. Since brain size has decreased noticeably during the Holocene (by about 10%), we know that evolutionary response to changes in selection pressures affecting the brain is certainly possible in 10,000 years or less. Since major human populations spent tens of thousands of years with almost zero gene flow, average brain size could differ between populations, and of course it does.
One of the interesting implications is that brains that work less well – that produce less fitness per cubic centimeter – will also be smaller. If the costs are the same, lower brain performance means that the optimum shifts downward. That lower fitness payoff per cubic centimeter could stem from physiological differences that caused lower performance, or, in principle, from environmental factors that were not well addressed by human intelligence: in some sense incomprehensible, at least for humans in prehistory.
Consider a graph that shows fitness as a function of brain size. It has an optimum, but since we see quite a bit of variation in brain size within populations, we suspect that the curvature is not that high near the optimum. Modest differences in IQ do not (or, to be more exact, did not) cause huge changes in fitness. Mutational pressure will also cause deviations from the optimum: adaptation happens, but there is always some noise in the system.
By the way, there is a lot of talk about how nobody can really define or measure intelligence, IQ is not related to brain size, brain size doesn’t vary between human populations, and brain volume is hard to measure. That’s all nonsense.