Are Big Brains Not For BEING Smart, But For STAYING Smart?

Another Edge essay on scientific ideas that should be retired:
Nicholas Humphrey 
Emeritus School Professor, The London School of Economics; Author, Soul Dust 
The bigger an animal's brain, the greater its intelligence. You may think the connection is obvious. Just look at the evolutionary lineage of human beings: humans have bigger brains—and are cleverer—than chimpanzees, and chimpanzees have bigger brains—and are cleverer—than monkeys. Or, as an analogy, look at the history of computing machines in the 20th century. The bigger the machines, the greater their number-crunching powers. In the 1970's the new computer at my university department took up a whole room. 
From the phrenology of the 19th century, to the brain-scan sciences of the 21st, it has indeed been widely assumed that brain volume determines cognitive capacity. In particular, you'll find the idea repeated in every modern textbook that the brain size of different primate species is causally related to their social intelligence. I admit I'm partly responsible for this, having championed the idea back in the 1970's. Yet, for a good many years now, I've had a hunch that the idea is wrong. 
There are too many awkward facts that don't fit in. For a start, we know that modern humans can be born with only two thirds the normal volume of brain tissue, and show next to no cognitive deficit as adults.

"Next to no" <> "no"

The brain scan volume to IQ correlations found in recent years have been in the 0.3 to 0.4 range, which is approaching a "moderate" correlation for the social sciences. In other words, the glass is part-full as well as part-empty, but very few people know that. Far more assume Stephen Jay Gould had the last word on the subject.

We know that, during normal human brain development, the brain actually shrinks as cognitive performance improves (a notable example being changes in the "social brain" during adolescence, where the cortical grey matter decreases in volume by about 15% between age 10 and 20). And most surprising of all, we know that there are nonhuman animals, such as honey bees or parrots, that can emulate many feats of human intelligence with brains that are only a millionth (bee) or a thousandth (parrot) the size of a human's.  
The key, of course, is programming: What really matters to cognitive performance is not so much the brain's hardware as its onboard software. And smarter software certainly does not require a bigger hardware base (in fact, as the shrinkage of the cortex during adolescence shows, it may actually require a smaller—tidier—one).

But the history of computers suggests that while there are occasionally advances in software that allow the same work to be done by less powerful hardware, the more typical pattern is for software to elaborate over time so that the minimum processor size keeps going up: look at the "Requirements" small print on an old Word Perfect box, say, and marvel over the tiny amount of hardware capacity (as measured in transistors) once required to provide highly adequate word processing capability.

What has happened of course (Moore's Law) is that the number of transistors you can place on a square centimeter has doubled every year or two, but that seems to be somewhat different from what we see in the brains of animals.

It's true that programs to deliver superior performance may require a lot of designing, either by natural selection or learning. But the fact is that, once they've been invented, they will likely make less demands on hardware than the older versions. To take the special case of social intelligence, I'd say it's quite possible that the algorithm for solving "theory of mind" problems could be written on the back of a postcard and could be implemented on an iPhone. In which case, the widely touted suggestion that the human brain had to double in size for humans to be capable of "second-order mind-reading", makes little sense.

Okay, but our relatively large-brained primate relatives don't seem as well-programmed as honey bees to benefit from cooperation. I've spent maybe 20 hours in the Lincoln Park zoo Great Ape House, and the chimps struck me as jerks. So, considering our relatives, maybe primates have to cogitate their way to cooperating more than social insects have to?

Then why did the human brain double in size? Why is it much bigger than you might think it needs to be, to underpin our level of intelligence? There's no question that big brains are costly to build and maintain. So, if we are to retire the "obvious theory", what can we put in its place? The answer I'd suggest lies in the advantage of having a large amount of cognitive reserve. Big brains have spare capacity that can be called on if and when working-parts get damaged or wear out. From adulthood onwards humans—like other mammals—begin to lose a significant amount of brain tissue to accidents, haemorrhages and degeneration. But because humans can draw on this extra reserve, the loss doesn't have to show. This means humans can retain their mental powers into relative old age, long after their smaller brained ancestors would have become incapacitated. (And as a matter of fact the unfortunate individual born with an unusually small brain is much more likely to succumb to senile dementia in his forties).

Not wholly implausible. Has anybody studied it?

True, many of us die for other reasons with unused brain power to spare. But some of us live considerably longer than we might have done if our brains were half the size. So, what evolutionary advantage does longevity bring, even the post-reproductive longevity typical of humans? The answer surely is that humans can benefit—as no other species could do—from the presence of mentally-sound grandparents and great-grandparents, whose role in caretaking and teaching has been key to the success of human culture.

Or, say, that having a crafty old mother or grandmother to act as matchmaker and social arbiter increases your chance of marrying well? Some of the predicament of the five unmarried Bennett daughters in Pride and Prejudice is not only that their father hasn't provided them with dowries, but also that their mother isn't all that bright at manipulating the social whirl for them.

The basic brain size equals intelligence in old age theory seems quite testable. For example, Ian Deary continues to give IQ tests to Scots born in 1921 who were first given IQ tests in 1932. What are their hat sizes today? (Granted, hat sizes are pretty crude measures.)