04 JANUARY 2007
Human brains are capable of amazing things, including, perhaps, figuring out why human brains are so capable.
It is a bit of a mystery. Human intelligence emerged rather suddenly
in evolutionary time. Life evolved for 3 billion years without producing
anything smarter than an amoeba, and then hundreds of millions of years
more passed before a brain came along that was smart enough to know how
to use opposable thumbs. 
But once human-class brains appeared, it took only a few dozen millennia for their owners to conquer the planet. What's more, humans can use those brains not only for fighting and farming, but for creating art and music, programming computers and discovering the vastness of the universe. Brains with these abilities existed long before there was any need for such modern mental talent.
So where human intelligence came from, and why it became so powerful, have been mysteries that challenge intelligence itself.
"The two burning questions are what factors drove the evolution of brain size and why our ancestors 50,000 years ago needed the brains they had," write biomathematician Sergey Gavrilets and computer scientist Aaron Vose of the University of Tennessee.
One possible answer links the rapid increase in individual intelligence
to the rise of human social interaction. Once people began living in groups,
certain social skills became helpful for males 
desiring
to spread their genes to future generations (by acquiring the most mates).
Those social skills included deception, manipulation, and beating up the
other males -- strategies that men used to outcompete their rivals for
access to females.
Presumably, a bigger brain is better able to learn such strategies. So at some point evolution encouraged rapid growth in the brain's size and complexity. This idea, known as the Machiavellian hypothesis, suggests that human sociality and culture evolved hand in hand with the expansion of brain power.
Gavrilets and Vose have tested that idea by devising equations capturing the essential features of the process. Those equations quantify things like how easy it is for the brain to learn a strategy (the brain's learning ability) and how many strategies the brain can remember (its cerebral capacity). Each strategy is quantified by its complexity (how hard it is to learn) and by its fitness (its effectiveness at winning the competition for mates).
Being able to learn strategies faster and remember more of them should provide an evolutionary payoff. Genes that improve learning ability and increase cerebral capacity will be passed on to more offspring, and subsequent generations will be more and more likely to be smarter and smarter -- if the Machiavellian idea is correct.
Sure enough, if you throw all those equations in the computer, it appears that the Machiavellian approach could very well explain the rapid growth in human brain power.
Starting with groups of 50 to 150 individuals (with zero learning ability and zero cerebral capacity at the outset), the computer simulations show that mutations or new combinations of genes in offspring can eventually produce a runaway increase in brain ability. But not at first. For several thousand generations, learning ability and capacity stay close to zero. Then a "cognitive explosion" occurs, with a dramatic increase in intelligence. The explosion stops when the benefits of a large brain no longer compensate for the cost of having it.
Brains require a lot of energy, hogging about 20 percent of a body's
metabolic resources while making up only 2 percent of the body's weight.
That's why it takes so long for advanced intelligence to appear in the
first place. A dimwit male unable to learn doesn't waste resources on
a brain. A 
bigger
brain is no help if there's nothing to learn with it, and a clever new
strategy won't help intelligence evolve if nobody can learn it.
"Learning ability and cerebral capacity are advantageous only if the individual has both of them simultaneously," Gavrilets and Vose point out in a paper appearing recently in the Proceedings of the National Academy of Sciences.
Basically, evolution has to wait for random mutations to produce both learning ability and cognitive capacity at the same time -- and that can take up to 25,000 generations, the simulations show. But once the explosion is triggered, learning ability and capacity rise quickly, reaching high values within 5,000 to 10,000 generations.
And once evolution provides a big brain for use in winning the war for mates, that same brain power can then be turned to other pursuits -- kind of the way powerful computers invented to crunch numbers can be used for word processing or photo editing.
This idea also explains why human brain power seems not to have advanced much since cave dwelling days, The simulations show that at some point, the high costs of a bigger brain exceed the benefits, and brain power reaches a plateau. Or even drops. In modern society, the scientists note, brain power may be less necessary to reproduce, so cognitive capacity may tend to diminish. Hmmm.
Gavrilets and Vose warn, though, that their results do not prove that this story describes what really happened with humans. It's merely a case of demonstrating the mathematical plausibility of the Machiavellian scenario. Their model has several limitations.
For one thing, it ignores any strategies that females might be learning and using to manipulate the males. Which almost certainly must be a factor in the evolution of male intelligence, or lack thereof.
E-mail: tsiegfried@nasw.org
