Is Our Brain as Good as It Gets?

Evolution is typically presented as the ultimate Horatio Alger story: up from the primordial muck we came, with each generation and successive species fitter--better, faster, stronger, smarter--than the one before. Tales of brain evolution offer few exceptions to this script: mammals are smarter than reptiles, primates smarter than other mammals, and we're smarter than other primates (electoral politics notwithstanding). Popular science (and science fiction) abounds with speculation about what heights we will scale next.

But an article in this month's issue of Scientific American by Douglas Fox (http://www.scientificamerican.com/article.cfm?id=the-limits-of-intelligence) showers a bit of acid rain on this ticker-tape parade. If you take a look at the facts--the coldest, hardest facts of physics--there doesn't seem to be much room left for improvement. This may be is a good as it gets.

Consider: the human brain already jams 100 billion neurons into our skulls, giving us the biggest brain:body ratio of any species on the planet. To achieve this inspired feat of packing that should be the envy of vacation goers everywhere, evolution had to make a series of trade-offs that may have boxed our brains, and us, into a corner--what theorists call a local maximum. Being at local maximum is like scaling a lesser peak. There are taller mountains, somewhere, but for you there's no place to go but downhill. In our case the local maximum looks like this:

We could try to improve our brains by making our neurons faster--able to carry and process signals with greater speed and efficiency. The only known way to do this is to increase the diameter our axons, the impulse-transmitting parts of neurons. But thicker axons need more energy and more space, and to fit the neurons (and any nutrient-carrying supports) into our skull we would require either that we use fewer neurons (defeating the purpose), or grow bigger heads. Putting aside the question of whether we could give birth to babies with even bigger heads (ask your mother), a bigger brain forces neurons further apart, and so requires longer connections, which in turn requires more energy and--you guessed it--makes signaling slower, once again defeating the purpose.

How about going with more connections between the neurons we already have? Well, more wires likewise take up more space (see above) and longer wires take longer to transmit signals (ditto).

Finally, what if we try to implement one of these options by making neurons smaller, so we can fit more, or more connections, or both, into the skulls we already have? Thinner neurons slow down signal transmission (as we noted). But more importantly, they also tend to fire more randomly--and noise is signal processing's worst enemy.

You see the evolutionary bind we are in? Whatever hope there is for increasing human mental potential, the fix won't come from biology. And so that leaves culture to pick up the slack. Perhaps the most important moral of this article is this: it is up to us to design the social systems such as schools and work teams, and technologies like encyclopedias and smart phones that will enhance human intellectual capacities, rather than deaden them. Are we up to this challenge? I certainly hope so. I'd hate to think we will collectively stay exactly this dumb.

Of course, fully embracing this challenge will require a much better understanding of how our brains actually do their job--and although I commend Mr. Fox for producing this fine piece of reflective science journalism, he did miss some opportunities for education in this area. In particular, when he is discussing the structural compromises that allow us to fit so many neurons into so little space, he writes that our brain is organized into localized, functionally specialized regions dedicated to things like speech comprehension or face recognition. But it is not so.

It is certainly true that our brain is organized into "modules" consisting of regions of dense local interconnections, with sparser connections to other more distant regions. This organization--called a "small world" architecture--allows for fast communication across the whole brain without the need for wires connecting everything to everywhere, which would take up too much room. But it is not the case that these network modules are functionally specialized. In fact, as noted in my previous posts, they are put to many different uses under different circumstances.

I don't know whether Dr. Changizi (who is quoted in this context) actually said we have specialized modules, or whether Mr. Fox filtered what Dr. Changizi said through the lens offered by the popular understanding of a brain consisting of segregated localized functions, but either way it is a mistake. The brain doesn't work that way.

Intelligence Essential Reads

How Do You Know What You Know?

How Gifted Children Change in the Teenage Years

The irony is that Mr. Fox makes a very deep fact about the brain quite clear: intelligence is less about local processing than about cooperative connectivity. Function is about establishing and coordinating neural partnerships across the brain. Thus the reversion to an outdated localizationist view is in fact out of sync with the main thrust of the article.

So, to Mr. Fox (and good science journalists everywhere): thank you for a very interesting, thoughtful piece about the organization of the brain. I look forward to even better ones.