Two of the smartest people you'll ever meet are the guys who used to operate the M. Coy bookshop on Pine Street in Seattle. Business pressures recently forced them to shutter their shop, but for 20 years, they sold their books, and from the moment you walked into their store, they had you figured out. They noticed where your gaze would go; they noticed where you paused. They noticed what books you picked up and how long you lingered over them. They recalled earlier customers who had bought the same titles and remembered other books those shoppers bought. They flashed through their entire 20,000-book inventory and then approached you with the single most important thing they had to offer: a recommendation.
Across town, in the Art Deco headquarters of Amazon.com the booksellers are good at making recommendations too. Log on to their site, and you've walked into their store. There, Amazon computers also keep an eye on you. They see where you click; they see where you pause. They recall every book you've ever bought and what other customers like you have bought. They shovel through data about millions of buyers and tens of millions of sales and then, like the shopkeepers, come up with a suggestion. However, the computers don't do all this in a 1,400-g (3 lb.), walnut-wrinkled mass of brain tissue but in a vast network of computers. It's easy to say that one approach is more complex than the other. It's a lot harder to say which one.
Of all the things that confuse human beings, perhaps nothing trips us up so much as what it means for something to be simple or complex. A houseplant, with its microhydraulics, fine-tuned metabolism and dense schematic of nucleic acids, may be more complex than a manufacturing plant. A modern army, with its thicket of bureaucracy and static encampments, may be simpler than a nimble guerrilla group. A guppy, with its symphony of biological systems and subsystems, is vastly more complicated than a star.
Human beings are not wired to look at things this way. We're suckers for size, for flash, for speed, for scale; we mistake immensity for complexity and subtlety for simplicity. That has very often been our undoing. Shock and awe should win a war, until an insurgency beats it back. An election should be sealed by storming Super Tuesday, until the campaign dies of a thousand little losses. The 2003 Yankees, with their $180 million payroll, should win the World Series, until the $63 million Marlins send them packing.
These may be lessons most of us must repeat again and again, but science increasingly is learning something from them. A generation ago, the paradigm-shifting understanding of chaos theory revealed the power of disorder in meteorology, marketing, plate tectonics and more. Similarly, investigators across the social and scientific spectrums are today studying how systems that seem simple or complex may be just the opposite--and how that fact can expand our understanding of our world. "Ask me why I forgot my keys today, and the answer may be that something was on my mind," says neuroscientist Chris Wood of the Santa Fe Institute (SFI) in New Mexico, a multidisciplinary think tank devoted to complexity theory. "Ask me about the calcium channels in my brain that drive remembering, and you're asking a much harder question."
There are a lot of ways the push-pull between simplicity and complexity is being explored and explained. Consider how babies learn to speak--a job so complicated that by some measures they shouldn't be able to do it at all. By the time babies are 18 months old, they have a core vocabulary of 50 words they can pronounce and 100 more they understand. By their sixth birthday, children have a working vocabulary of 6,000 words--meaning they've learned, on average, three new words every day since birth. Mastering conversational English requires about 50,000 words. What's more, since babies can't know where they'll be born, they must start life able to learn any of the world's nearly 7,000 tongues. It's processing speed that makes all this possible.
At Rutgers University in Newark, N.J., neuroscientist April Benasich fits prelingual babies with caps that read electrical activity in the brain. Benasich then plays one-syllable word bits to them--da and ta sounds, for example--and watches as their brains process the difference. At first, the sounds are separated by 300 milliseconds, very fast but well within the brain's ability. She then speeds things up so that the gap shrinks to 200 milliseconds, then 100, then 35--the point at which the length of the space is less than the length of the syllable itself. Even then the babies keep pace, getting all the way down to 10 milliseconds before the sounds run together.
Not all kids, however, have the same gifts. Benasich has found that some children fall out of the word-break race at about 70 milliseconds. Find the kids who later develop reading or speech disabilities, and they may also turn out to be the ones who had trouble keeping up with the sounds. "If you can't make a precise phonological map of a word," Benasich says, "you can't recognize it or reproduce it." If therapists could spot kids with such processing problems early, they could provide programs better targeted to their needs. No matter how the children's disability is corrected, it's a mark of the simple things on which speech stands or falls that the need for such retraining may turn on a few milliseconds of hearing either way.
Mundane matters like traffic move through simplicity choke points too. On any given day, about a million cars stream into and out of Manhattan. At any given moment, however, only about 8,000 of them are in operation in the heavily traveled midtown area. Keep those cars moving, and traffic flows smoothly all over the island. Jam them up, and gridlock can spread like ice freezing. "In fact," says urban-planning consultant Sam Schwartz, a former New York traffic commissioner who helped the city prepare for the 1980 transit strike, "in the case of true gridlock, the streets are actually 60% empty. All of the crowding is at the intersections, with nothing getting to midblock."