Professional Documents
Culture Documents
The Eureka Hunt: Why Do Good Ideas Come To Us When They Do?
The Eureka Hunt: Why Do Good Ideas Come To Us When They Do?
The Eureka Hunt: Why Do Good Ideas Come To Us When They Do?
By Jonah Lehrer
guage is so complex that the brain has to subtle associations coming from the right puzzle known as “the candle problem,” for
process it in two different ways at the side of the brain. “That’s when I realized instance, subjects are given a cardboard
same time,” he said. “It needs to see the
forest and the trees. The right hemi- Brain-imaging techniques are revealing how our minds produce insight.
box containing a few thumbtacks, a book that can be combined with all three—in sumed to be gradual. Kounios, a man
of matches, and a candle. They are told to this case, “apple” (“pineapple,” “crab with a shock of unruly wavy hair and an
attach the candle to a piece of corkboard apple,” “apple sauce”). The subjects have affinity for rumpled button-up vests, had
so that it can burn properly. Nearly ninety up to thirty seconds to solve the puzzle. been working with electroencephalogra-
per cent of people pursue the same two If they come up with an answer, they phy, or EEG, which measures the waves
strategies. They try to tack the candle di- press the space bar on the keyboard and of electricity produced by the brain by
rectly to the board, which causes the can- say whether the answer arrived via in- means of a nylon hat filled with greased
dle wax to shatter. Or they try melting the sight or analysis. When I participated in electrodes. (The device looks like a bulky
candle with the matches, so that it sticks the experiment in Jung-Beeman’s lab, I shower cap.) Because there is no time
to the board; but the wax doesn’t hold and found that it was surprisingly easy to delay with EEG, Kounios thought it
the candle falls. Only four per cent of peo- differentiate between the two cognitive could be useful for investigating the
ple manage to come up with the solution, paths. When I solved puzzles with anal- fleeting process of insight. Unfortu-
which involves attaching the candle to the ysis, I tended to sound out each possible nately, the waves of electricity can’t be
cardboard box and tacking the cardboard word combination, cycling through all traced back to their precise source, but
box to the corkboard. the words that went with “pine” and then Kounios and Jung-Beeman saw that
To isolate the brain activity that de- seeing if they also worked with “crab” or combining EEG with fMRI might
fined the insight process, Jung-Beeman “sauce.” If I worked toward a solution, I allow them to construct a precise map,
needed to develop a set of puzzles that always double-checked it before pressing both in time and space, of the insight
could be solved either by insight or by the space bar. An insight, on the other process.
analysis. Doing so was a puzzle in itself. hand, felt instantaneous: the answer ar- The resulting studies, published in
“It can get pretty frustrating trying to find rived like a revelation. 2004 and 2006, found that people who
an experimentally valid brainteaser,” Jung-Beeman initially asked his sub- solved puzzles with insight activated a
Jung-Beeman said. “The puzzles can’t be jects to solve the puzzles while inside an specific subset of cortical areas. Although
too hard or too easy, and you need to be fMRI machine, a brain scanner that the answer seemed to appear out of no-
able to generate lots of them.” He even- monitors neural activity by tracking where, the mind was carefully preparing
tually settled on a series of verbal puzzles, changes in blood flow. But fMRI has a itself for the breakthrough. The first
based on ones used by a psychologist three-to-five-second delay, as the blood areas activated during the problem-
in the early nineteen-sixties, which he diffuses across the cortex. “Insights hap- solving process were those involved with
named the Compound Remote Associ- pen too fast for fMRI,” Jung-Beeman executive control, like the prefrontal cor-
ate Problems, or CRAP. (The joke is be- said. “The data was just too messy.” tex and the anterior cingulate cortex. The
ginning to get old, and in his scientific pa- Around this time, he teamed up with scientists refer to this as the “preparatory
pers Jung-Beeman decorously leaves off John Kounios, a cognitive neuroscientist phase,” since the brain is devoting its
the final “P.”) at Drexel University, who was interested considerable computational power to the
In a C.R.A. word puzzle, a subject is in insight largely because it seemed to problem. The various sensory areas, like
given three words, such as “pine,” “crab,” contradict the classic model of learning, the visual cortex, go silent as the brain
and “sauce,” and asked to think of a word in which the learning process was as- suppresses possible distractions. “The
cortex does this for the same reason we
close our eyes when we’re trying to think,”
Jung-Beeman said. “Focus is all about
blocking stuff out.”
What happens next is the “search
phase,” as the brain starts looking for an-
swers in all the relevant places. Because
Jung-Beeman and Kounios were giving
people word puzzles, they saw addi-
tional activity in areas related to speech
and language. The search can quickly
get frustrating, and it takes only a few
seconds before people say that they’ve
reached an impasse, that they can’t think
of the right word. “Almost all of the
possibilities your brain comes up with
are going to be wrong,” Jung-Beeman
said. “And it’s up to the executive-
control areas to keep on searching or, if
necessary, change strategies and start
searching somewhere else.”
But sometimes, just when the brain is
“Lawn care has been good to us, Ed.” about to give up, an insight appears.
“You’ll see people bolt up in their chair nected.” When the brain is searching for actually prevent the insight. While it’s
and their eyes go all wide,” Ezra Weg- an insight, these are the cells that are commonly assumed that the best way to
breit, a graduate student in the Jung-Bee- most likely to produce it. solve a difficult problem is to focus, mini-
man lab who often administers the C.R.A. mize distractions, and pay attention only
test, said. “Sometimes they even say ‘Aha!’
before they blurt out the answer.” The
suddenness of the insight comes with a
T he insight process, as sketched by
Jung-Beeman and Kounios, is a del-
icate mental balancing act. At first, the
to the relevant details, this clenched state
of mind may inhibit the sort of creative
connections that lead to sudden break-
burst of brain activity. Three hundred brain lavishes the scarce resource of atten- throughs. We suppress the very type of
milliseconds before a participant commu- tion on a single problem. But, once the brain activity that we should be encourag-
nicates the answer, the EEG registers a brain is sufficiently focussed, the cortex ing. Jonathan Schooler has recently dem-
spike of gamma rhythm, which is the needs to relax in order to seek out the more onstrated that making people focus on the
highest electrical frequency generated remote association in the right hemi- details of a visual scene, as opposed to the
by the brain. Gamma rhythm is thought sphere, which will provide the insight. big picture, can significantly disrupt the
to come from the “binding” of neurons, “The relaxation phase is crucial,” Jung- insight process. “It doesn’t take much to
as cells distributed across the cortex Beeman said. “That’s why so many in- shift the brain into left-hemisphere mode,”
draw themselves together into a new net- sights happen during warm showers.” An- he said. “That’s when you stop paying at-
work, which is then able to enter con- other ideal moment for insights, according tention to those more holistic associations
sciousness. It’s as if the insight had gone to the scientists, is the early morning, right coming in from the right hemisphere.”
incandescent. after we wake up. The drowsy brain is Meanwhile, in a study published last year,
Jung-Beeman and Kounios went unwound and disorganized, open to all German researchers found that people
back and analyzed the information from sorts of unconventional ideas. The right with schizotypy—a mental condition that
the fMRI experiment to see what was hemisphere is also unusually active. Jung- resembles schizophrenia, albeit with far
happening inside the brain in the sec- Beeman said, “The problem with the less severe symptoms—were significantly
onds before the gamma burst. “My big- morning, though, is that we’re always so better at solving insight problems than a
gest worry was that we would find noth- rushed. We’ve got to get the kids ready for control group. Schizotypal subjects have
ing,” Kounios said. “I thought there was school, so we leap out of bed and never give enhanced right-hemisphere function and
a good possibility that whatever we ourselves a chance to think.” He recom- tend to score above average on measures of
found on the EEG wouldn’t show up on mends that, if we’re stuck on a difficult creativity and associative thinking.
the brain imaging.” When the scientists problem, it’s better to set the alarm clock a Schooler’s research has also led him to
looked at the data, however, they saw few minutes early so that we have time to reconsider the bad reputation of letting
that a small fold of tissue on the surface lie in bed and ruminate. We do some of our one’s mind wander. Although we often
of the right hemisphere, the anterior su- best thinking when we’re still half asleep. complain that the brain is too easily dis-
perior temporal gyrus (aSTG), became As Jung-Beeman and Kounios see it, tracted, Schooler believes that letting the
unusually active in the second before the the insight process is an act of cognitive mind wander is essential. “Just look at the
insight. The activation was sudden and deliberation—the brain must be focussed history of science,” he said. “The big ideas
intense, a surge of electricity leading to a on the task at hand—transformed by ac- seem to always come when people are
rush of blood. Although the function of cidental, serendipitous connections. We sidetracked, when they’re doing some-
the aSTG remains mostly a mystery— must concentrate, but we must concen- thing that has nothing to do with their re-
the brain is stuffed with obscurities— trate on letting the mind wander. The search.” He cites the example of Henri
Jung-Beeman wasn’t surprised to see it patterns of brain activity that define this Poincaré, the nineteenth-century mathe-
involved with the insight process. A few particular style of thought have recently matician, whose seminal insight into non-
previous studies had linked the area to been studied by Joy Bhattacharya, a psy- Euclidean geometry arrived while he was
aspects of language comprehension, chologist at Goldsmiths, University of boarding a bus. “At the moment when I
such as the detection of literary themes London. Using EEG, he has found that put my foot on the step,” Poincaré wrote,
and the interpretation of metaphors. (A he can tell which subjects will solve insight “the idea came to me, without anything
related area was implicated in the pro- puzzles up to eight seconds before the in- in my former thoughts seeming to have
cessing of jokes.) Jung-Beeman argues sight actually arrives. One of the key pre- paved the way for it. . . . I did not verify
that these linguistic skills, like insight, dictive signals is a steady rhythm of alpha the idea; I should not have had the time,
require the brain to make a set of dis- waves emanating from the right hemi- as, upon taking my seat in the omnibus, I
tant and unprecedented connections. He sphere. Alpha waves typically correlate went on with the conversation already
cites studies showing that cells in the with a state of relaxation, and Bhattacharya commenced, but I felt a perfect certainty.”
right hemisphere are more “broadly believes that such activity makes the brain Poincaré credited his sudden mathemati-
tuned” than cells in the left hemisphere, more receptive to new and unusual ideas. cal insight to “unconscious work,” an abil-
with longer branches and more den- He has also found that unless subjects ity to mull over the mathematics while he
dritic spines. “What this means is that have sufficient alpha-wave activity they was preoccupied with unrelated activities,
neurons in the right hemisphere are col- won’t be able to make use of hints the re- like talking to a friend on the bus. In his
lecting information from a larger area searchers give them. 1908 essay “Mathematical Creation,”
of cortical space,” Jung-Beeman said. One of the surprising lessons of this re- Poincaré insisted that the best way to
“They are less precise but better con- search is that trying to force an insight can think about complex problems is to im-
THE NEW YORKER, JULY 28, 2008 43
merse yourself in the problem until you For now, though, the science of pro- which lights up whenever people are
hit an impasse. Then, when it seems that moting insight remains rooted in anecdote, shown the right answer—even if they
“nothing good is accomplished,” you in stories of people, like Poincaré, who haven’t come up with the answer them-
should find a way to distract yourself, were able to consistently induce the neces- selves. Pressed tight against the bones of
preferably by going on a “walk or a jour- sary state of mind. Kounios tells a story the forehead, the prefrontal cortex has
ney.” The answer will arrive when you about an expert Zen meditator who took undergone a dramatic expansion during
least expect it. Richard Feynman, the part in one of the C.R.A. insight experi- human evolution, so that it now represents
Nobel Prize-winning physicist, preferred ments. At first, the meditator couldn’t solve nearly a third of the brain. While this area
the relaxed atmosphere of a topless bar, any of the insight problems. “This Zen guy is often associated with the most special-
where he would sip 7 UP, “watch the en- went through thirty or so of the verbal puz- ized aspects of human cognition, such as
tertainment,” and, if inspiration struck, zles and just drew a blank,” Kounios said. abstract reasoning, it also plays a critical
scribble equations on cocktail napkins. “He was used to being very focussed, but role in the insight process. Hallucinogenic
you can’t solve these problems if you’re too drugs are thought to work largely by mod-