The Best American Science and Nature Writing Part 2

Notes.

1. This article is based on a talk given on September 25, 2009, at the Harry Ransom Center for Humanistic Studies of the University of Texas at Austin, to commemorate its exhibition "Other Worlds: Rare Astronomical Works," on view September 8, 2009-January 3, 2010.

2. Of course the stars are not visible during the day, but some of them can be seen just after sunset, when the sun's position in the sky is still known.

3. A gnomon is different from a sundial because the pole that casts a shadow on a sundial is not vertical but set at an angle chosen so that the pole's shadow follows about the same path during each day of the year. This makes the sundial more useful as a clock but less useful as a calendar.

4. It may be wondered why Calypso did not tell Odysseus to keep the North Star on his left. The reason is that in Homer's time the star Polaris, which is now the North Star, was not at the North Pole of the sky. This is not because of any motion of Polaris itself but because of a phenomenon known as the precession of the equinoxes, discovered by Hipparchus. In modern terms, the axis of Earth's rotation does not keep a fixed direction in the sky, but precesses like the axis of a spinning top, making a full circle every 25,727 years. It is a measure of the accuracy of Greek astronomy that the data of Hipparchus indicated a period of 28,000 years.

5. I say "direct" applications because experimental and theoretical work in particle physics that pushes technology and mathematics to their current limits occasionally spins off new technology or mathematics of great practical importance. One celebrated example is the World Wide Web. This can provide a valid argument for government support, but it is not why we do the research.

6. I have written about this at greater length in "The Wrong Stuff," New York Review of Books, April 8, 2004.

7. This opinion was most recently expressed by Giovanni Bignami, the head of the European s.p.a.ce Agency's Science Advisory Committee, in "Why We Need s.p.a.ce Travel," Nature, July 16, 2009.

TIMOTHY FERRIS Cosmic Vision.

FROM National Geographic.

WHEN YOU START STARGAZING with a telescope, two experiences typically ensue. First, you are astonished by the viewa"Saturn's golden rings, star cl.u.s.ters glittering like jewelry on black velvet, galaxies aglow with gentle starlight older than the human speciesa"and by the realization that we and our world are part of this gigantic system. Second, you soon want a bigger telescope.

Galileo, who first trained a telescope on the night sky four hundred years ago this fall, pioneered this two-step program. First, he marveled at what he could see. Galileo's telescope revealed so many previously invisible stars that when he tried to map all of those in just one constellation, Orion, he gave up, confessing that he was "overwhelmed by the vast quant.i.ty of stars." He saw mountains on the moona"in contradiction of the prevailing orthodoxy, which declared that all celestial objects were made of an unearthly "ether." He charted four bright satellites as they bustled around Jupiter like planets in a miniature solar system, something that critics of the Copernican sun-centered cosmology had dismissed as physically impossible. Evidently Earth was a small part of a big universe, not a big part of a small one.

And soon, sure enough, Galileo went to work making bigger and better telescopes. Large light-gathering lenses were not yet available, so he concentrated on making longer telescopes, which produced higher magnifying powers and reduced the halos of spurious colors that afflicted gla.s.s lenses in those days. Subsequent observers took the design of gla.s.s-lensed, refracting telescopes to great lengths, sometimes literally so. In Danzig, Johannes Hevelius deployed a telescope 150 feet long; hung by ropes from a pole, it undulated in the slightest breeze. In the Netherlands, the Huygens brothers unveiled lanky telescopes that had no tubes at all: the objective lens was perched on a high platform in a field, while an observer up to 200 feet away aligned a magnifying eyepiece and peered through it. Such instruments proffered fleeting glimpses of planets and stars, which, like the dance of the seven veils, only aroused a burning desire to see more.

The reflecting telescope, pioneered by Isaac Newton, made it practical to gratify such desires: mirrors required that only one surface be ground to gather and reflect starlight to a focal point, and since the mirror was supported from behind, it could be quite large without sagging under its own weight, as large lenses tended to do. William Herschel discovered the planet Ura.n.u.s with a handmade reflecting telescopea"he cast his metal mirrors in his garden and bas.e.m.e.nt and once had to flee from a coursing river of molten metal after the horse-dung mold fractured. Spiral-armed galaxies were first glimpsed through a ma.s.sive reflecting telescope with a six-foot-diameter primary mirror that Lord Rosse constructed on his estate in Ireland.

Today's largest telescopes have mirrors up to some 10 meters (33 feet) in diameter, with quadruple the light-gathering power of the legendary 5-meter Hale Telescope at Palomar Observatory in southern California. Looming as large as office buildings, some of these giants are so highly automated that they can dust off their optics at sundown, open the dome, sequence and carry out observations throughout the night, and shut down come threatening weather, all with little or no human intervention. Yet humans, being human, still intervene a lot, if only to make sure nothing goes awry: to lose just one night's work at a big telescope these days is to squander as much as $100,000 in operating costs.

Three of today's largest telescopesa"Gemini North, Subaru, and Kecka"stand within hailing distance of one another atop the nearly 14,000-foot peak of Hawaii's Mauna Kea, an inactive volcano. The alt.i.tude puts them above 40 percent of Earth's atmospherea"and most of its water vapor, which is opaque to the infrared wavelengths the astronomers like to studya"but also makes it difficult for the astronomers and engineers who work there to breathe and think. Many wear clear plastic oxygen tubes in their nostrils as routinely as we might wear eyegla.s.ses. Others rely on the body's ability to adapt but worry about making what they call a CLM, or "career-limiting mistake." "At alt.i.tude, we don't improvise; that would be a disaster," says the Gemini astronomer Scott Fisher. "We're kind of trained monkeys up here. The real thinking goes on at sea level."

These big Mauna Kea observatories are similarly smart and costly, yet each exudes a distinct personality. The 8.1-meter Gemini telescope is housed in an onion-shaped silver dome ringed by a set of shutters that, when closed during the day, make the observatory look as ungainly as a fat man in an inner tube. But the shutters open at dusk to create an enormous set of windows, three stories tall and stretching nearly three-quarters of the way around the observatory, that let in the night air and happen to afford a panorama of the blue Pacific all the way to Maui and beyond. Gemini's four main digital detectorsa"cameras and spectrometers, as heavy as cars and costing around $5 million eacha"are attached to a carousel surrounding the telescope's focal point, where they can be rotated into place in minutes. Computers run the telescope by night, shuffling requested observations to make the most of every minute. "We're all about nighttime efficiency," says Fisher.

The Subaru telescope's instruments are housed in alcoves like jeroboams of champagne in a heavenly wine cellar. (The comparison is not entirely fanciful; one leading j.a.panese astronomer propitiates the G.o.ds at the start of each Subaru observing run by pouring vintage sake on the ground outside the dome at the four points of the compa.s.s.) When a particular instrument is required, a robotic yellow trolley makes its way to the alcove, picks up the detector, ferries it to the bottom of the ma.s.sive telescope, and locks it in place, attaching the data cables and the plumbing for the detector's refrigeration system. Subaru happens to be one of the few giant telescopes that anybody has ever actually looked through. For its inauguration in 1999, an eyepiece was attached so that Princess Sayako of j.a.pan could have a look through the scope, and for several nights thereafter eager Subaru staffers did the same. "Everything you can see in the Hubble s.p.a.ce Telescope photosa"the colors, the knots in the cloudsa"I could see with my own eyes, in stunning Technicolor," one recalled.

Keck consists of two identical telescopes. Both have 10-meter mirrors made of thirty-six segments; with its support structure, each segment weighs close to a thousand pounds, costs close to a million dollars, and would suffice to create a fine, university-grade telescope on its own. The telescopes' "tubes" are spindly steel skeletons that look as delicate as spiders' webs but are more precisely configured than a racing sloop's rigging. "We use the telescope's mission to motivate ourselves," one Keck astronomer told me. "If a little wire or something is found intruding into the optical path, we think, If the light has been traveling through s.p.a.ce for 90 percent of the history of the universe, and it got this close to the telescope, we'd better make sure it gets the rest of the way."

Few of the astronomers awarded time on the big telescopes actually go there to observe anymore. Most submit their requests electronicallya"on a recent night at Gemini, the scheduled projects ranged from "Primordial Solar System Ma.s.ses" to "Magnetic Activity in Ultracool Dwarfs"a"and the results are sent back to them. Geoff Marcy, a modern-day Prince Henry the Navigator, whose team has discovered more than 150 planets...o...b..ting stars other than our sun, gets more observing time than most at Keck but has not been there for years. Instead, his extrasolar-planet team observes from a remote operating facility at UC Berkeley. During observing runs, Marcy reports, "we settle into a routine of working all night. We have all our books and other resources here at hand, plus enough normal life so our spouses don't forget us."

In addition to their unprecedented light-gathering power, today's big telescopes benefit from their adaptive optics (AO) systems, which compensate for atmospheric turbulence. The turbulence is what makes stars glitter; telescopes magnify every twinkle. A typical AO system fires a laser beam into a thin layer of sodium atoms 56 miles high in the atmosphere, causing them to glow. By monitoring this artificial star, the system determines how the air is churning and adjusts the telescope's optics more than a thousand times each second to compensate. Gemini pays a pair of students ten dollars an hour to sit outside the dome all night, walkie-talkies in hand, ready to warn the astronomers to turn off the laser should an airplane approach. "It's incredible to see in practice," says Scott Fisher. "When the AO system is off, you see a nice, pretty star that looks a little fuzzy. Turn the AO on, and the star just goes phonk! and collapses to a tiny point."

Objects in the night sky are measured in degrees, the full moon spanning about one half of a degree. Without AO, a powerful telescope on a fine night can perceive objects separated from each other by as little as one 3,600th of a degree, or one arc second. Thanks to Keck's AO system, UCLA astronomer Andrea Ghez was able to make a motion picture of seven bright stars whirling around the invisible black hole at the center of our galaxy over a period of fourteen years: the entire movie takes place inside a box measuring only one arc second on a side. Based on the frenzy of the stars in the grip of the black hole, Ghez calculated that it has the ma.s.s of 4 million suns, generating enough gravitational force to slingshot some stars that pa.s.s too close right out of our galaxy. Several such hypervelocity stars have been located, speeding off toward the depths of intergalactic s.p.a.ce like party crashers ejected from an exclusive nightclub.

What's next? Even bigger telescopes, of course, with the capability to shoot cosmic pictures faster, wider, and in even greater detail. Among the behemoths due to come on line within a decade are the Giant Magellan Telescope, the Thirty Meter Telescope, and the 42-meter European Extremely Large Telescopea"a scaled-down version of the 100-meter Overwhelmingly Large Telescope, which was tabled at the planning stage when its projected budget turned out to be overwhelming too.

Particularly innovative is the Large Synoptic Survey Telescope, or LSST, whose 8.4-meter primary mirror was cast last August in a spinning furnace under the stands of the University of Arizona Wildcats' football stadium in Tucson. (The rotation technique produces a mirror blank that is already concave, reducing the amount of gla.s.s that must be ground away to bring the mirror to a proper figure.) Conventional telescopes have narrow fields of view, typically spanning no more than half a degree on a sidea"much too narrow to take in the enormous patterns that grew out of the big bang. The LSST will have a field of view covering 10 square degrees, the area of fifty full moons. From its site in the Chilean Andes, it will be able to image galaxies far across the universe in exposures of just 15 seconds each, capturing fleeting events to distances of over 10 billion light-years, 70 percent of the way across the observable universe. "Since we'll have a big field of view, we can take a whole lot of short exposures anda" bang, bang, bang, banga"cover the entire visible sky every several nights, and then repeat," says LSST Director Tony Tyson. "If you keep doing that for ten years, you have a moviea"the first movie of the universe."

The LSST's fast, wide-angle imaging could help answer two of the biggest questions confronting astronomers today: the nature of dark matter and the nature of dark energy. Dark matter makes its presence known by its gravitational attractiona"it explains the rotation speed of galaxiesa"but it emits no light, and its const.i.tution is unknown. Dark energy is the name given to the mysterious phenomenon that for the past 5 billion years has been accelerating the rate at which the universe expands. "It's a little bit scary," says Tyson, "as if you were flying an airplane and suddenly something unknown took over the controls."

The LSST could help solve these immense riddles thanks in part, oddly enough, to the science of acoustics. The big bang was noisy. Although sound cannot propagate through the vacuum of today's s.p.a.cea"as pedants are fond of reminding the directors of science-fiction filmsa"the early universe was a thick plasma and as alive with sound as a drummers' convention. Certain tones resonated in the primordial plasma like the tones of struck winegla.s.ses, and these harmonies, etched into sheets of galaxies that today shamble across billions of light-years, contain precise information about the nature of dark matter and dark energy. If astronomers can map these large-scale structures accurately, they should be able to identify the signatures of dark matter and dark energy in the big bang's harmonics. The Sloan Digital Sky Survey, a pioneering wide-angle study, captured some of this information when it mapped the sky from 1999 through 2008. The LSST is designed to go much deeper into cosmic s.p.a.ce. It may not resolve the mysteries, but, predicts Tyson, "it will go a long way toward showing what dark energy and dark matter aren't."

The LSST's photographic "speed" will also give astronomers a better look at events too short-lived to be readily studied today. Most astronomers, even amateurs using backyard telescopes and off-the-shelf digital cameras, regularly record fleeting events of unknown origin. You take a series of digital exposures, and in one of them a spot of light appears where none was before or after. It may have been a cosmic ray hitting the light-detection chip, a high-velocity asteroid hurtling through the field of view, or a blue flare on the surface of a dim red star. You just don't know, so you shrug and move on. Because the LSST will take so many repeat exposures of the entire sky, it could resolve many such riddles.

Tomorrow's enormous telescopes will do as much in one night as today's do in a year, but that will not necessarily render the older telescopes obsolete. When the giants come on line, says Scott Fisher, "the Geminis of today will become the telescopes that get to go out and do the surveys," finding interesting phenomena for the largest scopes to investigate in detail. "It's like a pyramid, and it feeds both ways: when a really big telescope finds something exciting that we can't spend every night observing, the astronomers can apply for time on a smaller telescope to, say, check it out every clear night for a year and see how it changes over time."

Orbiting s.p.a.ce telescopes are opening up another dimension. NASA's Kepler satellite, which launched in March 2009, is methodically imaging the constellation Cygnus, looking for the slight dimming of light caused when planetsa"some perhaps Earthlikea"transit in front of their stars; Geoff Marcy's team will then use Keck to scrutinize stars flagged by Kepler to confirm that they have planets. In the future, pairs of mirrors deployed in orbit and linked by laser-ranging systems could attain the resolving power of telescopes measuring thousands of meters across. One day, observatories sitting in craters on the far side of the moon may probe the universe from surroundings ideally quiet, dark, and cold. The coming combination of smart satellites talking to big, increasingly automated ground telescopes, themselves linked together by fiber-optic networks and employing artificial intelligence systems to search out patterns in the torrents of data, suggests a process as much biological as mechanical, akin to the evolution of global eyes, optic nerves, and brains.

Film directors like to say that each movie is really two moviesa"the one you make and the one you say you're going to make while raising the money. The point is that n.o.body can accurately predict the outcome of any genuinely creative venture. The same is true of scientific discovery: scientists can explain what they expect to accomplish with bigger and better telescopes, but such predictions are mostly just extrapolations from the past. "If you're going to Washington to seek funding for a new telescope and you make a list of what you'll see through this new window on the universe, you know that the most interesting thing it will discover is probably not on your list," says Tyson. "It's likely to be something totally new, some out-of-the-box physics that's going to blow our minds."

The marvelous model of the big-bang universe pieced together in the twentieth century arose largely from just such unantic.i.p.ated discoveries. Edwin Hubble discovered the expansion of the universe accidentally, at the telescope: cosmic expansion had been implied by Einstein's general theory of relativity, but Hubble knew nothing of the prediction, and not even Einstein had taken it seriously. Dark matter was discovered accidentally; so was dark energy. A telescope doesn't just show you what's out there; it impresses upon you how little you know, opening your imagination to wonders as big as all outdoors. "The spygla.s.s is very truthful," said Galileo.

TIMOTHY FERRIS Seeking New Earths.

FROM National Geographic.

IT TOOK HUMANS thousands of years to explore our own planet and centuries to comprehend our neighboring planets, but nowadays new worlds are being discovered every week. To date, astronomers have identified more than 370 "exoplanets," worlds...o...b..ting stars other than the sun. Many are so strange as to confirm the biologist J. B. S. Haldane's famous remark that "the universe is not only queerer than we suppose, but queerer than we can suppose." There's an Icarus-like "hot Saturn" 260 light-years from Earth, whirling around its parent star so rapidly that a year there lasts less than three days. Circling another star 150 light-years out is a scorched "hot Jupiter," whose upper atmosphere is being blasted off to form a gigantic, cometlike tail. Three benighted planets have been found orbiting a pulsara"the remains of a once mighty star shrunk to a spinning atomic nucleus the size of a citya"while untold numbers of worlds have evidently fallen into their suns or been flung out of their systems to become "floaters" that wander in eternal darkness.

Amid such exotica, scientists are eager for a hint of the familiar: planets resembling Earth, orbiting their stars at just the right distancea"neither too hot nor too colda"to support life as we know it. No planets quite like our own have yet been found, presumably because they're inconspicuous. To see a planet as small and dim as ours amid the glare of its star is like trying to see a firefly in a fireworks display; to detect its gravitational influence on the star is like listening for a cricket in a tornado. Yet by pushing technology to the limits, astronomers are rapidly approaching the day when they can find another Earth and interrogate it for signs of life.

Only eleven exoplanets, all of them big and bright and conveniently far away from their stars, have as yet had their pictures taken. Most of the others have been detected by using the spectroscopic Doppler technique, in which starlight is a.n.a.lyzed for evidence that the star is being tugged ever so slightly back and forth by the gravitational pull of its planets. In recent years astronomers have refined the Doppler technique so exquisitely that they can now tell when a star is pulled from its appointed rounds by only one meter a seconda"about human walking speed. That's sufficient to detect a giant planet in a big orbit or a small one if it's very close to its star, but not an Earth at anything like our Earth's 93-million-mile distance from its star. Earth tugs the sun around at only one-tenth walking speed, or about the rate that an infant can crawl; astronomers cannot yet prize out so tiny a signal from the light of a distant star.

Another approach is to watch a star for the slight periodic dip in its brightness that will occur should an orbiting planet circle in front of it and block a fraction of its light. At most a tenth of all planetary systems are likely to be oriented so that these mini-eclipses, called transits, are visible from Earth, which means that astronomers may have to monitor many stars patiently to capture just a few transits. The French COROT satellite, now in the third and final year of its prime mission, has discovered seven transiting exoplanets, one of which is only 70 percent larger than Earth.

The United States' Kepler satellite is COROT's more ambitious successor. Launched from Cape Canaveral last March, Kepler is essentially just a big digital camera with a .95-meter aperture and a 95-megapixel detector. It makes wide-field pictures every thirty minutes, capturing the light of more than 100,000 stars in a single patch of sky between the bright stars Deneb and Vega. Computers on Earth monitor the brightness of all those stars over time, alerting humans when they detect the slight dimming that could signal the transit of a planet.

Because that dimming can be mimicked by other phenomena, such as the pulsations of a variable star or a large sunspot moving across a star's surface, the Kepler scientists won't announce the presence of a planet until they have seen it transit at least three timesa"a wait that may be only a few days or weeks for a planet rapidly circling close to its star but years for a terrestrial twin. By combining Kepler results with Doppler observations, astronomers expect to determine the diameters and ma.s.ses of transiting planets. If they manage to discover a rocky planet roughly the size of Earth orbiting in the habitable zonea"not so close to the star that the planet's water has been baked away nor so far out that it has frozen into icea"they will have found what biologists believe could be a promising abode for life.

The best hunting grounds may be dwarf stars, smaller than the sun. Such stars are plentiful (seven of the ten stars nearest to Earth are M dwarfs), and they enjoy long, stable careers, providing a steady supply of sunlight to any life-bearing planets that might occupy their habitable zones. Most important for planet hunters, the dimmer the star, the closer in its habitable zone it liesa"dim dwarf stars are like small campfires, where campers must sit close to be comfortablea"so transit observations will pay off more quickly. A close-in planet also exerts a stronger pull on its star, making its presence easier to confirm with the Doppler method. Indeed, the most promising planet yet founda"the "super Earth" Gliese 581 d, seven times Earth's ma.s.sa"orbits in the habitable zone of a red dwarf star only a third the ma.s.s of the sun.

Should Earthlike planets be found within the habitable zones of other stars, a dedicated s.p.a.ce telescope designed to look for signs of life there might one day take a spectrum of the light coming from each planet and examine it for possible biosignatures such as atmospheric methane, ozone, and oxygen, or for the "red edge" produced when chlorophyll-containing photosynthetic plants reflect red light. Directly detecting and a.n.a.lyzing the planet's own light, which might be one ten-billionth as bright as the star's, would be a tall order. But when a planet transits, starlight shining through the atmosphere could reveal clues to its composition that a s.p.a.ce telescope might be able to detect.

While grappling with the daunting technological challenge of performing a chemical a.n.a.lysis of planets they cannot even see, scientists searching for extraterrestrial life must keep in mind that it may be very different from life here at home. The lack of the red edge, for instance, might not mean a terrestrial exoplanet is lifeless: life thrived on Earth for billions of years before land plants appeared and populated the continents. Biological evolution is so inherently unpredictable that even if life originated on a planet identical to Earth at the same time it did here, life on that planet today would almost certainly be very different from terrestrial life.

As the biologist Jacques Monod once put it, life evolves not only through necessitya"the universal workings of natural lawa"but also through chance, the unpredictable intervention of countless accidents. Chance has reared its head many times in our planet's history, dramatically so in the many ma.s.s extinctions that wiped out millions of species and, in doing so, created room for new life forms to evolve. Some of these baleful accidents appear to have been caused by comets or asteroids colliding with Eartha"most recently the impact, 65 million years ago, that killed off the dinosaurs and opened up opportunities for the distant ancestors of human beings. Therefore scientists look not just for exoplanets identical to the modern Earth but for planets resembling Earth as it used to be or might have been. "The modern Earth may be the worst template we could use in searching for life elsewhere," notes Caleb Scharf, head of Columbia University's Astrobiology Center.

It was not easy for earlier explorers to plumb the depths of the oceans, map the far side of the moon, or discern evidence of oceans beneath the frozen surfaces of Jovian moons, and it will not be easy to find life on the planets of other stars. But we now have reason to believe that billions of such planets must exist and that they hold the promise of expanding not only the scope of human knowledge but also the richness of the human imagination.

For thousands of years we humans knew so little about the universe that we were apt to celebrate our imaginations and denigrate reality. (As the Spanish philosopher Miguel de Unamuno wrote, the mysticism of the religious visionaries of old arose from an "intolerable disparity between the hugeness of their desire and the smallness of reality.") Now, with advances in science, it has become gallingly evident that nature's creativity outstrips our own. The curtain is going up on countless new worlds with stories to tell.

PART TWO.

Neurology Displacing Molecular Biology.

JONAH LEHRER Don't!

FROM The New Yorker.

IN THE LATE NINETEEN SIXTIES, Carolyn Weisz, a four-year-old with long brown hair, was invited into a "game room" at the Bing Nursery School, on the campus of Stanford University. The room was little more than a large closet containing a desk and a chair. Carolyn was asked to sit down in the chair and pick a treat from a tray of marshmallows, cookies, and pretzel sticks. Carolyn chose the marshmallow. (Although she's now forty-four, Carolyn still has a weakness for those air-puffed b.a.l.l.s of corn syrup and gelatin. "I know I shouldn't like them," she says. "But they're just so delicious!") A researcher then made Carolyn an offer: she could either eat one marshmallow right away or, if she was willing to wait while he stepped out for a few minutes, she could have two marshmallows when he returned. He said that if she rang a bell on the desk while he was away, he would come running back, and she could eat one marshmallow but would forfeit the second. Then he left the room.

Although Carolyn has no direct memory of the experiment, and the scientists would not release any information about the subjects, she strongly suspects that she was able to delay gratification. "I've always been really good at waiting," Carolyn told me. "If you give me a challenge or a task, then I'm going to find a way to do it, even if it means not eating my favorite food." Her mother, Karen Sortino, is still more certain: "Even as a young kid, Carolyn was very patient. I'm sure she would have waited." But her brother Craig, who also took part in the experiment, displayed less fort.i.tude. Craig, a year older than Carolyn, still remembers the torment of trying to wait. "At a certain point, it must have occurred to me that I was all by myself," he recalls. "And so I just started taking all the candy." According to Craig, he was also tested with little plastic toysa"he could have a second one if he held outa"and he broke into the desk, where he figured there would be additional toys. "I took everything I could," he says. "I cleaned them out. After that, I noticed the teachers encouraged me to not go into the experiment room anymore."

Footage of these experiments, which were conducted over several years, is poignant, as the kids struggle to delay gratification for just a little bit longer. Some cover their eyes with their hands or turn around so that they can't see the tray. Others start kicking the desk, or tug on their pigtails, or stroke the marshmallow as if it were a tiny stuffed animal. One child, a boy with neatly parted hair, looks carefully around the room to make sure that n.o.body can see him. Then he picks up an Oreo, delicately twists it apart, and licks off the white cream filling before returning the cookie to the tray, a satisfied look on his face.

Most of the children were like Craig. They struggled to resist the treat and held out for an average of less than three minutes. "A few kids ate the marshmallow right away," Walter Mischel, the Stanford professor of psychology in charge of the experiment, remembers. "They didn't even bother ringing the bell. Other kids would stare directly at the marshmallow and then ring the bell thirty seconds later." About 30 percent of the children, however, were like Carolyn. They successfully delayed gratification until the researcher returned, some fifteen minutes later. These kids wrestled with temptation but found a way to resist.

The initial goal of the experiment was to identify the mental processes that allowed some people to delay gratification while others simply surrendered. After publishing a few papers on the Bing studies in the early seventies, Mischel moved on to other areas of personality research. "There are only so many things you can do with kids trying not to eat marshmallows."

But occasionally Mischel would ask his three daughters, all of whom attended the Bing, about their friends from nursery school. "It was really just idle dinnertime conversation," he says. "I'd ask them, 'How's Jane? How's Eric? How are they doing in school?'" Mischel began to notice a link between the children's academic performance as teenagers and their ability to wait for the second marshmallow. He asked his daughters to a.s.sess their friends academically on a scale of zero to five. Comparing these ratings with the original data set, he saw a correlation. "That's when I realized I had to do this seriously," he says. Starting in 1981, Mischel sent out a questionnaire to all the reachable parents, teachers, and academic advisers of the 653 subjects who had partic.i.p.ated in the marshmallow task, who were by then in high school. He asked about every trait he could think of, from their capacity to plan and think ahead to their ability to "cope well with problems" and get along with their peers. He also requested their SAT scores.

Once Mischel began a.n.a.lyzing the results, he noticed that low delayers, the children who rang the bell quickly, seemed more likely to have behavioral problems, both in school and at home. They got lower SAT scores. They struggled in stressful situations, often had trouble paying attention, and found it difficult to maintain friendships. The child who could wait fifteen minutes had an SAT score that was, on average, 210 points higher than that of the kid who could wait only thirty seconds.

Carolyn Weisz is a textbook example of a high delayer. She attended Stanford as an undergraduate, and got her Ph.D. in social psychology at Princeton. She's now an a.s.sociate psychology professor at the University of Puget Sound. Craig, meanwhile, moved to Los Angeles and has spent his career doing "all kinds of things" in the entertainment industry, mostly in production. He's currently helping to write and produce a film. "Sure, I wish I had been a more patient person," Craig says. "Looking back, there are definitely moments when it would have helped me make better career choices and stuff."

Mischel and his colleagues continued to track the subjects into their late thirties. Ozlem Ayduk, an a.s.sistant professor of psychology at the University of California at Berkeley, found that low-delaying adults have a significantly higher body-ma.s.s index and are more likely to have had problems with drugsa"but it was frustrating to have to rely on self-reports. "There's often a gap between what people are willing to tell you and how they behave in the real world," he explains. And so last year Mischel, who is now a professor at Columbia, and a team of collaborators began asking the original Bing subjects to travel to Stanford for a few days of experi ments in an fMRI machine. Carolyn says she will be partic.i.p.ating in the scanning experiments later this summer; Craig completed a survey several years ago but has yet to be invited to Palo Alto. The scientists are hoping to identify the particular brain regions that allow some people to delay gratification and control their temper. They're also conducting a variety of genetic tests, as they search for the hereditary characteristics that influence the ability to wait for a second marshmallow.

If Mischel and his team succeed, they will have outlined the neural circuitry of self-control. For decades, psychologists have focused on raw intelligence as the most important variable when it comes to predicting success in life. Mischel argues that intelligence is largely at the mercy of self-control: even the smartest kids still need to do their homework. "What we're really measuring with the marshmallows isn't will power or self-control," Mischel says. "It's much more important than that. This task forces kids to find a way to make the situation work for them. They want the second marshmallow, but how can they get it? We can't control the world, but we can control how we think about it."

Walter Mischel is a slight, elegant man with a shaved head and a face of deep creases. He talks with a Brooklyn bl.u.s.ter, and he tends to act out his sentences, so that when he describes the marshmallow task he takes on the body language of an impatient four-year-old. "If you want to know why some kids can wait and others can't, then you've got to think like they think," Mischel says.

Mischel was born in Vienna in 1930. His father was a modestly successful businessman with a fondness for caf society and Esperanto, while his mother spent many of her days lying on the couch with an ice pack on her forehead, trying to soothe her frail nerves. The family considered itself fully a.s.similated, but after the n.a.z.i annexation of Austria in 1938, Mischel remembers being taunted in school by the Hitler Youth and watching as his father, hobbled by childhood polio, was forced to limp through the streets in his pajamas. A few weeks after the takeover, while the family was burning evidence of their Jewish ancestry in the fireplace, Walter found a long-forgotten certificate of U.S. citizenship issued to his maternal grandfather decades earlier, thus saving his family.

The family settled in Brooklyn, where Mischel's parents opened up a five-and-dime. Mischel attended New York University, study ing poetry under Delmore Schwartz and Allen Tate, and taking studio-art cla.s.ses with Philip Guston. He also became fascinated by psychoa.n.a.lysis and new measures of personality, such as the Rorschach test. "At the time, it seemed like a mental X-ray machine," he says. "You could solve a person by showing them a picture." Although he was pressured to join his uncle's umbrella business, he ended up pursuing a Ph.D. in clinical psychology at Ohio State.

But Mischel noticed that academic theories had limited application, and he was struck by the futility of most personality science. He still flinches at the naivete of graduate students who based their diagnoses on a battery of meaningless tests. In 1955 Mischel was offered an opportunity to study the "spirit possession" ceremonies of the Orisha faith in Trinidad, and he leapt at the chance. Although his research was supposed to involve the use of Rorschach tests to explore the connections between the unconscious and the behavior of people when possessed, Mischel soon grew interested in a different project. He lived in a part of the island that was evenly split between people of East Indian and of African descent; he noticed that each group defined the other in broad stereotypes. "The East Indians would describe the Africans as impulsive hedonists who were always living for the moment and never thought about the future," he says. "The Africans, meanwhile, would say that the East Indians didn't know how to live and would stuff money in their mattress and never enjoy themselves."

Mischel took young children from both ethnic groups and offered them a simple choice: they could have a miniature chocolate bar right away or, if they waited a few days, they could get a much bigger chocolate bar. Mischel's results failed to justify the stereotypesa"other variables, such as whether or not the children lived with their father, turned out to be much more importanta"but they did get him interested in the question of delayed gratification. Why did some children wait and not others? What made waiting possible? Unlike the broad traits supposedly a.s.sessed by personality tests, self-control struck Mischel as potentially measurable.

In 1958 Mischel became an a.s.sistant professor in the Department of Social Relations at Harvard. One of his first tasks was to develop a survey course on "personality a.s.sessment," but Mischel quickly concluded that while prevailing theories held personality traits to be broadly consistent, the available data didn't back up this a.s.sumption. Personality, at least as it was then conceived, couldn't be reliably a.s.sessed at all. A few years later, he was hired as a consultant on a personality a.s.sessment initiated by the Peace Corps. Early Peace Corps volunteers had sparked several embarra.s.sing international incidentsa"one mailed a postcard on which she expressed disgust at the sanitary habits of her host countrya"so the Kennedy administration wanted a screening process to eliminate people unsuited for foreign a.s.signments. Volunteers were tested for standard personality traits, and Mischel compared the results with ratings of how well the volunteers performed in the field. He found no correlation; the time-consuming tests predicted nothing. At this point, Mischel realized that the problem wasn't the testsa"it was their premise. Psychologists had spent decades searching for traits that exist independently of circ.u.mstance, but what if personality can't be separated from context? "It went against the way we'd been thinking about personality since the four humors and the ancient Greeks," he says.

While Mischel was beginning to dismantle the methods of his field, the Harvard psychology department was in tumult. In 1960 the personality psychologist Timothy Leary helped start the Harvard Psilocybin Project, which consisted mostly of self-experimentation. Mischel remembers graduate students' desks giving way to mattresses, and large packages from Ciba chemicals, in Switzerland, arriving in the mail. Mischel had nothing against hippies, but he wanted modern psychology to be rigorous and empirical. And so, in 1962, Walter Mischel moved to Palo Alto and went to work at Stanford.

There is something deeply contradictory about Walter Mischela"a psychologist who spent decades critiquing the validity of personality testsa"inventing the marshmallow task, a simple test with impressive predictive power. Mischel, however, insists there is no contradiction. "I've always believed there are consistencies in a person that can be looked at," he says. "We just have to look in the right way." One of Mischel's cla.s.sic studies doc.u.mented the aggressive behavior of children in a variety of situations at a summer camp in New Hampshire. Most psychologists a.s.sumed that aggression was a stable trait, but Mischel found that children's responses depended on the details of the interaction. The same child might consistently lash out when teased by a peer but readily submit to adult punish ment. Another might react badly to a warning from a counselor but play well with his bunkmates. Aggression was best a.s.sessed in terms of what Mischel called "if-then patterns." If a certain child was teased by a peer, then he would be aggressive.

One of Mischel's favorite metaphors for this model of personality, known as interactionism, concerns a car making a screeching noise. How does a mechanic solve the problem? He begins by trying to identify the specific conditions that trigger the noise. Is there a screech when the car is accelerating, or when it's shifting gears, or turning at slow speeds? Unless the mechanic can give the screech a context, he'll never find the broken part. Mischel wanted psychologists to think like mechanics and look at people's responses under particular conditions. The challenge was devising a test that accurately simulated something relevant to the behavior being predicted. The search for a meaningful test of personality led Mischel to revisit, in 1968, the protocol he'd used on young children in Trinidad nearly a decade earlier. The experiment seemed especially relevant now that he had three young daughters of his own. "Young kids are pure id," Mischel says. "They start off unable to wait for anythinga"whatever they want they need. But then, as I watched my own kids, I marveled at how they gradually learned how to delay and how that made so many other things possible."

A few years earlier, in 1966, the Stanford psychology department had established the Bing Nursery School. The cla.s.srooms were designed as working laboratories, with large one-way mirrors that allowed researchers to observe the children. This past February, Jennifer Winters, the a.s.sistant director of the school, showed me around the building. While the Bing is still an active center of researcha"the children quickly learn to ignore the students scribbling in notebooksa"Winters isn't sure that Mischel's marshmallow task could be replicated today. "We recently tried to do a version of it, and the kids were very excited about having food in the game room," she says. "There are so many allergies and peculiar diets today that we don't do many things with food."

Mischel perfected his protocol by testing his daughters at the kitchen table. "When you're investigating will power in a four-year-old, little things make a big difference," he says. "How big should the marshmallows be? What kind of cookies work best?" After several months of patient tinkering, Mischel came up with an experi mental design that closely simulated the difficulty of delayed gratification. In the spring of 1968, he conducted the first trials of his experiment at the Bing. "I knew we'd designed it well when a few kids wanted to quit as soon as we explained the conditions to them," he says. "They knew this was going to be very difficult."

At the time, psychologists a.s.sumed that children's ability to wait depended on how badly they wanted the marshmallow. But it soon became obvious that every child craved the extra treat. What, then, determined self-control? Mischel's conclusion, based on hundreds of hours of observation, was that the crucial skill was the "strategic allocation of attention." Instead of getting obsessed with the marshmallowa"the "hot stimulus"a"the patient children distracted themselves by covering their eyes, pretending to play hide-and-seek underneath the desk, or singing songs from Sesame Street. Their desire wasn't defeateda"it was merely forgotten. "If you're thinking about the marshmallow and how delicious it is, then you're going to eat it," Mischel says. "The key is to avoid thinking about it in the first place."

In adults, this skill is often referred to as metacognition, or thinking about thinking, and it's what allows people to outsmart their shortcomings. (When Odysseus had himself tied to the ship's mast, he was using some of the skills of metacognition: knowing he wouldn't be able to resist the Sirens' song, he made it impossible to give in.) Mischel's large data set from various studies allowed him to see that children with a more accurate understanding of the workings of self-control were better able to delay gratification. "What's interesting about four-year-olds is that they're just figuring out the rules of thinking," Mischel says. "The kids who couldn't delay would often have the rules backward. They would think that the best way to resist the marshmallow is to stare right at it, to keep a close eye on the goal. But that's a terrible idea. If you do that, you're going to ring the bell before I leave the room."

According to Mischel, this view of will power also helps explain why the marshmallow task is such a powerfully predictive test. "If you can deal with hot emotions, then you can study for the SAT instead of watching television," Mischel says. "And you can save more money for retirement. It's not just about marshmallows."

Subsequent work by Mischel and his colleagues found that these differences were observable in subjects as young as nineteen months. Looking at how toddlers responded when briefly separated from their mothers, they found that some immediately burst into tears or clung to the door, but others were able to overcome their anxiety by distracting themselves, often by playing with toys. When the scientists set the same children the marshmallow task at the age of five, they found that the kids who had cried also struggled to resist the tempting treat.

The early appearance of the ability to delay suggests that it has a genetic origin, an example of personality at its most predetermined. Mischel resists such an easy conclusion. "In general, trying to separate nature and nurture makes about as much sense as trying to separate personality and situation," he says. "The two influences are completely interrelated." For instance, when Mischel gave delay-of-gratifi cation tasks to children from low-income families in the Bronx, he noticed that their ability to delay was below average, at least compared with that of children in Palo Alto. "When you grow up poor, you might not practice delay as much," he says. "And if you don't practice, then you'll never figure out how to distract yourself. You won't develop the best delay strategies, and those strategies won't become second nature." In other words, people learn how to use their mind just as they learn how to use a computer: through trial and error.

But Mischel has found a shortcut. When he and his colleagues taught children a simple set of mental tricksa"such as pretending that the candy is only a picture, surrounded by an imaginary framea"he dramatically improved their self-control. The kids who hadn't been able to wait sixty seconds could now wait fifteen minutes. "All I've done is given them some tips from their mental user manual," Mischel says. "Once you realize that will power is just a matter of learning how to control your attention and thoughts, you can really begin to increase it."

Marc Berman, a lanky graduate student with an easy grin, speaks about his research with the infectious enthusiasm of a freshman taking his first philosophy cla.s.s. Berman works in the lab of John Jonides, a psychologist and neuroscientist at the University of Michigan, who is in charge of the brain-scanning experiments on the original Bing subjects. He knows that testing forty-year-olds for self-control isn't a straightforward proposition. "We can't give these people marshmallows," Berman says. "They know they're part of a long-term study that looks at delay of gratification, so if you give them an obvious delay task they'll do their best to resist. You'll get a bunch of people who refuse to touch their marshmallow."

This meant that Jonides and his team had to find a way to measure will power indirectly. Operating on the premise that the ability to delay eating the marshmallow had depended on a child's ability to banish thoughts of it, they decided on a series of tasks that measure the ability of subjects to control the contents of working memorya"the relatively limited amount of information we're able to consciously consider at any given moment. According to Jonides, this is how self-control "cashes out" in the real world: as an ability to direct the spotlight of attention so that our decisions aren't determined by the wrong thoughts.

Last summer the scientists chose fifty-five subjects, equally split between high delayers and low delayers, and sent each one a laptop computer loaded with working-memory experiments. Two of the experiments were of particular interest. The first is a straightforward exercise known as the "suppression task." Subjects are given four random words, two printed in blue and two in red. After reading the words, they're told to forget the blue words and remember the red words. Then the scientists provide a stream of "probe words" and ask the subjects whether the probes are the words they were asked to remember. Though the task doesn't seem to involve delayed gratifi cation, it tests the same basic mechanism. Interestingly, the scientists found that high delayers were significantly better at the suppression task: they were less likely to think that a word they'd been asked to forget was something they should remember.

In the second, known as the Go/No Go task, subjects are flashed a set of faces with various expressions. At first, they are told to press the s.p.a.ce bar whenever they see a smile. This takes little effort, since smiling faces automatically trigger what's known as "approach behavior." After a few minutes, however, subjects are told to press the s.p.a.ce bar when they see frowning faces. They are now being forced to act against an impulse. Results show that high delayers are more successful at not pressing the b.u.t.ton in response to a smiling face.

When I first started talking to the scientists about these tasks last summer, they were clearly worried that they wouldn't find any behavioral differences between high and low delayers. It wasn't until early January that they had enough data to begin their a.n.a.lysis (not surprisingly, it took much longer to get the laptops back from the low delayers), but it soon became obvious that there were provocative differences between the two groups. A graph of the data shows that as the delay time of the four-year-olds decreases, the number of mistakes made by the adults sharply rises.

The big remaining question for the scientists is whether these behavioral differences are detectable in an fMRI machine. Although the scanning has just beguna"Jonides and his team are still working out the kinksa"the scientists sound confident. "These tasks have been studied so many times that we pretty much know where to look and what we're going to find," Jonides says. He rattles off a short list of relevant brain regions, which his lab has already identified as being responsible for working-memory exercises. For the most part, the regions are in the frontal cortexa"the overhang of brain behind the eyesa"and include the dorsolateral prefrontal cortex, the anterior prefrontal cortex, the anterior cingulate, and the right and left inferior frontal gyri. While these cortical folds have long been a.s.sociated with self-control, they're also essential for working memory and directed attention. According to the scientists, that's not an accident. "These are powerful instincts telling us to reach for the marshmallow or press the s.p.a.ce bar," Jonides says. "The only way to defeat them is to avoid them, and that means paying attention to something else. We call that will power, but it's got nothing to do with the will."

The behavioral and genetic aspects of the project are overseen by Yuichi Shoda, a professor of psychology at the University of Washington, who was one of Mischel's graduate students. He's been following these "marshmallow subjects" for more than thirty years: he knows everything about them, from their academic records and their social graces to their ability to deal with frustration and stress. The prognosis for the genetic research remains uncertain. Although many studies have searched for the underpinnings of personality since the completion of the Human Genome Project in 2003, many of the relevant genes remain in question. "We're incredibly complicated creatures," Shoda says. "Even the simplest aspects of personality are driven by dozens and dozens of different genes." The scientists have decided to focus on genes in the dopamine pathways, since those neurotransmitters are believed to regulate both motivation and attention. However, even if minor coding differences influence delay abilitya"and that's a likely possibilitya"Shoda doesn't expect to discover these differences: the sample size is simply too small.

In recent years, researchers have begun making house visits to many of the original subjects, including Carolyn Weisz, as they try to better understand the familial contexts that shape self-control. "They turned my kitchen into a lab," Carolyn told me. "They set up a little tent where they tested my oldest daughter on the delay task with some cookies. I remember thinking, I really hope she can wait."