IF YOU THINK mind reading is outside the realms of science, think again. In the past decade a revolution in brain-imaging technology has made it possible to see your private mental world in real time. These days no self-respecting university or teaching hospital is without a PET or fMRl scanner and, as the technology comes of age, its reach has extended. Where once researchers were most interested in sweeping generalisations about how humans think, now their studies are becoming ever more personal, uncovering details about our individual motivations, desires and prejudices.
One potentially fruitful area for brain imaging is in understanding why we make the choices we do. Economists and strategists have a vested interest in making these sorts of predictions. They have devised a whole field of mathematics - game theory - to try to uncover the rules of decision making, but it turns out that we lesser mortals just won't comply with their mathematical formulae. Now, by combining game theory with brain imaging, scientists are beginning to see why our decision making isn't necessarily rational, and how factors such as emotions and social context influence the choices we make. Already they can predict the choices monkeys will make by the pattern of activity in a single neuron. Could we be next? Will imaging one day allow others to consider your options even before you have? ("Why you do what you do", page 32).
From knowing what you want, it is a short leap to manipulating what you want. Advertising agents and marketers have been trying to do this for decades, but their methods are not exactly scientific. Now, with the dawn of neuromarketing, selling could go high-tech. This particular incarnation of brain imaging has had some high-profile publicity and, despite the old adage, it is not all good.
Consumer groups are enraged by the thought of multinationals trying to sneak inside our heads in search of the "buy button". But researchers say that what they are finding is much more subtle and could actually lead to more ethical ways of selling. Is this simply the latest gimmick in the corporate world's attempt to drum up business or the beginning of a new era for consumerism? ("They know what you want", page 36).
If all this seems like an infringement on your free will and right to privacy, you ain't seen nothing yet. Brain imaging could also be used to reveal your inner secrets. Studies have already put the spotlight on such things as racial prejudice, deception, sexual fantasy and personality. Though researchers urge caution there is little doubt that it is only a matter of time before the technology moves out of the lab and into the courtrooms, our workplaces and our everyday lives. It is an ethical minefield: will people be forced to undergo scanning, will the results raise insurance premiums, affect their job prospects, or determine whether or not they go to jail? Whatever the future of brain scanning, we need to think about these things now. ("Private thoughts, public property", page 38).
Why we do what we do
Wouldn't it be useful if you could predict the choices people
were going to make? Laura Spinney finds out how close we are to a
science of decision making.
BACK in the 1950s at the height of the cold war, President
Truman got an alarming piece of advice from the scientists at his military
think-tank, the RAND corporation. The Soviets have nuclear capability they
said, we have nuclear capability better nuke them before they nuke us. This,
according to game theory, their latest piece of mathematical wizardry, was
the only rational course of action. While an agreement not to push the
button would benefit both parties, cooperation is also
the riskiest
strategy because you stand to lose everything if the other guy gets in
there first. Better, then, to act - and hang the consequences.
History records that Truman made a different choice, and that the Soviets
also kept their nuclear arsenal in check. How could the RAND mathematicians
have got it so wrong? Common sense might argue that their mistake was to
assume that science has anything to say about the highly subjective process
of decision making. If you accept that
humans have
free will, what hope is there of predicting someone else's choices? But
neuroscientist Read Montague from Baylor College of Medicine in Houston,
Texas, disagrees. He believes that a science of decision making is possible;
where Truman's advisers went wrong was in assuming that the best decisions
are totally rational. By looking inside the brain, Montague and other researchers
are starting to see how the choices we make are swayed by a complex range
of factors such as
emotions,
social context and uncertainty and how we weigh up the potential costs and
benefits of alternative options before we make up our minds.
Gaining scientific insight into decision making is no trivial matter Our
whole lives are defined by the choices we make. Every day we make countless
decisions, from whether to read this book or that magazine, to
which person
to date, whether to
buy a certain
stock and when to ask for a pay rise. Economists and strategists would
dearly like to know what goes on inside our brains, but it could also help
the rest of us understand why we want what we want. Recent findings are even
suggesting new approaches to treating psychological conditions where decision
making is impaired, such as depression, dementia and
addiction.
On a more sinister note, this line of research could one day allow others
to uncover your secret motivations and predict, or even manipulate, the decisions
you make.
Game theory was the first real attempt to understand decision making in
mathematical terms. Its architect, the mathematician John von Neumann,
was influential at the RAND corporation in the 1950s, and an ardent supporter
of the pre-emptive strike, based on his belief that humans act as "rational
agents". Since then, however, numerous studies have shown that people simply
don't behave in real life as his mathematical models predict. In the game
called the prisoner's dilemma - which is analogous
to the situation that Truman faced -the maths says that two rational players
should not cooperate for a reward if they stand to gain a bigger reward by
"defecting". That's because even though the game is rigged so that if both
players defect the payout is minimal, it is in their interests to do so because
they will receive nothing at all if their partner defects and they do not.
Yet, countless repetitions of this game in psychology and economics laboratories
worldwide reveal that if the same two players play repeatedly, their strategies
evolve into a complex mix of cooperation and defection. Von Neumann got it
wrong - people do not behave rationally. Half a century later, we are now
starting to understand why, and the key lies in brain imaging.
Two years ago, a group led by Gregory Berns from Emory University, Atlanta,
used functional magnetic resonance imaging (fMRI) to scan the brains of 36
women playing repeated rounds of the prisoner's dilemma against the same
opponent. When his subjects cooperated, a network of brain regions including
the orbitofrontal cortex and striatum was activated; these areas are associated
with reward processing and learning by reinforcement. What's more, when
interviewed afterwards, the women typically said that even though defecting
could be more profitable, they found it less satisfying -either because they
were profiting at their opponent's expense, or because they realised they
were destabilising the game, prompting their opponent to defect and leading
to lower cumulative earnings (Neuron, vol 35, p395).
It is clear that when we weigh up the costs and benefits of various courses
of action, we do not just consider the material gains but also social and
emotional ones. "Yes, it is riskier financially to cooperate, because you
could get nothing," says Berns. "But defecting carries a social risk, being
labelled as a back-stabber." Each of us puts a different value on the various
components we must consider to reach our final decision. "Individually, it
comes down to the personal motivations of each person:
more money or be a nice person?" says Berns. This might suggest that you
can never reduce decision making down to a mathematical formula, but the
neuroscientists are not daunted. On the surface, the problem looks like comparing
apples and pears, but the brain must use a common currency to encode all
the different elements of cost and benefit so that they can be weighed up.
If you can find the part of the brain where this common currency is computed,
perhaps then you'll have the key to predicting the choices people make.
| Imaging the activity in particular parts of the brain is starting to reveal some very personal details about our lives such as how we make up our minds and what persuades us to buy certain products | |
| 1.Orbitofrontal Cortex Part of the frontal lobe with connections to the amygdala which suggest that it plays a role in integrating the cognitive and and emotional aspects of decision making |
 |
| 2.Medial prefrontal Cortex Seems to light up in fMRI studies when people identify with something or someone |
|
| 3.Striatum Links with orbitofrontal cortex,amygdala and neurons in the midbrain that assess reward make this a likely hub for decision making |
|
| 4.Putamen Area of the striatum associated with reward that lights up when people taste or see something they like |
|
| 5.Limbic System Emotion processing system |
|
| 6.Amygdala Activity relating to emotions,especially fear |
|
Berns's fMRI scans suggested a good starting point. They revealed that the
greater the activation in the orbitofrontal-striatal circuits, the greater
the probability that the women would cooperate. These circuits looked like
being an important hub for the decision-making process, given their connections
with emotional processing areas such as the amygdala, and with
dopamine-signalling neurons starting out in the midbrain, which play a role
in assessing reward (see Diagram, above). Montague and Berns suspected that
the orbitofrontal-striatal circuits might be where the brain integrates
information about the anticipated benefits of a particular choice, using
input from the midbrain neurons to encode a measure of actual reward and
the probability of achieving it, and input from the emotional centres to
encode the more subjective value of various options.
In 2002, Montague and Berns set about trying to work out systematically how
the orbitofrontal-striatal circuit responds in various situations. To do
this, they drew together all the existing findings about decision making
from brain imaging and behavioural studies and began looking for patterns.
Based on these findings and some specially designed fMRI experiments of their
own, the pair constructed a model that would predict how the
orbitoftontal-striatal network will respond to different options. When they
applied this predictor-valuation model (PVM) to situations in which monkeys
were faced with a choice of actions that would earn them different food rewards,
they found that the PVM accurately predicted the pattern of neural activity
in the orbitofrontal-striatal circuits just before the monkey made its choice,
though it fell short of indicating what choice it would actually make. The
challenge now, says Berns, is to discover the relationship between the neuronal
response and a monkey's behaviour, so that its choices become predictable.
"When we weigh up the costs and benefits of various courses of action,
we do not just consider the material gains but also the social and emotional
ones"
An inkling that this may indeed be possible comes from research done by Paul
Glimcher from the Center for Neural Science at New York University. He has
developed a similar model to the PVM to see what's happening at the level
of the individual neurons that encode the value judgements monkeys make when
they are weighing up their various options. Working with Michael Dorris,
Glimcher found that the choices monkeys made could not always be predicted
by equations describing rational behaviour in terms of game theory. Their
choices departed from the rational in the same way that human decisions do
-but could be predicted by changes in the firing rates of these individual
neurons (Games and Economic Behavior, in press).
Compared with monkeys, we probably weigh up a wider array of possible rewards
and costs when making most of our choices, but Berns still believes that
the PVM could shed light on what motivates us to make the decisions we do.
"Wouldn't it be great," he says, "to understand why people often forgo long-term
benefits for short-term gains - addictions for example - or why we pursue
wealth, when we know that it is a game with diminishing returns?" But Montague
points out that to get a handle on these kinds of questions you need first
to appreciate that many of the reasons why people do seemingly irrational
things are social. "I don't think [von Neumann's] 'rational agent' is dead,"
he says, "but I think it's pretty clear that it doesn't live in an individual
human's head." Instead, Montague argues,
it lives in the interactions between people -which explains why human decisions
are so powerfully shaped by social context. By taking the effects of this
context into account, he and Hems hope to create a PVM that more closely
resembles human behaviour.
"If decision-making behaviour
can be altered with just a single shot of propranolol or amino acids, could
we be on the verge of a brave new world"
Enter the hyperscan
A couple of years ago, they realised they would need to scan several people's
brains simultaneously to properly study brain mechanisms underlying the social
aspect of decision making. They came up with a technique called hyperscanning,
in which INIRI machines scanning people's brains are linked via the internet
so their responses can be compared in real time. Now they have a whole raft
of studies under way using the technique, looking at trusting behaviour and
trading behaviour, among other things.
Montague is now looking at how the value we place on something depends on
how other people value it. To do this, he staged the equivalent of an auction
or commodities market, with all the bidders in scanners, but able to see
what their rivals are bidding via computer screens. First he scanned one
person's brain while he or she made a bid, then two, and so on. The work
is preliminary, but he is finding that the more people who enter the bidding,
the greater the activity in the orbitofrontal-striatal network. Montague
describes this as a "sort of instinctual response to overvalue something"
when we see that other people want it. He points out that it makes sense
for us to be swayed by others in this way because much of our success as
a species depends on our ability to learn what is good and what is not from
those around us. This copycat mentality probably also explains why stock
markets are apt to boom and crash as people buy and sell when others do.
"The propensity of investors to create bubbles reflects some deep evolutionary
feature of our mental machinery," says Glimchen.
As well as accounting for social context, an accurate model of decision making
would also have to consider the effect of human emotions. Emotional context
can have a very powerful influence on the choices we make, as was shown
brilliantly two decades ago by American psychologists Daniel Kahneman of
Princeton University and Amos Tversky of Stanford University in California.
Imagine, they said to their subjects, that the US is bracing itself for the
outbreak of an unusual Asian disease which is expected to kill 600 people.
You must chose one of two options to combat the disease: programme Awill
save 200 lives; programme B offers a 1-in-3 chance that all 600 will be saved,
and a 2-in-3 chance that no one will be saved. Most people played safe and
chose programme A. The researchers then presented a second group with the
same problem. This time they asked, would you choose programme A, which will
guarantee 400 people die, or programme B, which carries a '4 probability
that no one will die, and a '4 probability that 600 will die? Nothing had
changed mathematically, yet the majority now went for programme B.
The only difference between the two trials was that in the first the
psychologists emphasised the benefits of programme A, while in the second
they emphasised the costs.
The inescapable conclusion is that the way we see our options can colour
the decisions we make. Spin doctors, pollsters and advertisers will not be
surprised. Its influence can also be seen in the different decision-making
styles of people who tend to see the glass as "half full" and those who see
it as "half empty". "Maximising your reward or minimising your loss feel
different emotionally," says Robert Rogers of the University of Oxford, who
is interested in the neurochemical pathways involved. His studies, published
earlier this year, suggest that the brain actually processes benefits and
costs in different ways, at least under certain circumstances.
In the first experiment, Rogers and his team gave healthy volunteers 80
milligrammes of a common beta-blocker called propranolol -roughly equivalent
to the daily dose prescribed for a person with very mild hypertension. This
drug blocks the action of the neurotransmitter noradrenaline, which is known
to be involved in coping with stress. The researchers then put their guinea
pigs to work on a computerised decision-making game. The results were striking.
Propranolol altered people's choices -but in a very selective way. They became
less able to discriminate between large and small losses when their chances
of losing were high.
Meanwhile, in a separate study volunteers took an amino acid drink designed
to deplete their brains of tryptophan - a precursor for the neurotransmitter
serotonin, which is important in controlling mood and is thought to be involved
in depression. When this group played the same decision games as the first
group, the results were very different. They were uniformly less able to
discriminate between small and large rewards, regardless of whether the stakes
were high or low. Rogers concludes that noradrenaline plays a role in processing
the costs associated with an option - which may be related to its known role
in coping with stress - while serotonin is important for assessing benefits
(Neuropsychopharmacology, vol 28, p153).
Distorted decisions
"This suggests that maybe, just possible when you make risky choices these
neurochemical systems might be doing complementary things," says Rogers.
His interest in the work is purely clinical, A novel way to think about
depressipn, he suggests, is that lowered serotonin levels in the brain lead
to distorted decision making in social situations, which may trigger feelings
of alienation. He is also intrigued by the finding that tryptophandepleted
volunteers behave just like chronic cocaine abusers when tested on the same
decision-making task. Abnormal decision making is a feature of a range of
common neuropsychiatric conditions, including substance abuse, manic depression,
dementia and personality disorder, says Rogers, who hopes his findings will
offer new treatments.
Rogers is reluctant to consider the more sinister implications of his findings
but, no doubt, they will not go unnoticed. If the decision-making behaviour
of healthy individuals can be altered with just a single shot of propranolol
or amino acids, could we be on the brink of a brave new world where cautious
CEOs might choose to pop a pill before taking a particularly risky decision,
or a shy teenager might drink a "neuro-pop" to give him the courage to approach
a girl? Perhaps this prospect is not as terrifying as it first appears. As
Glimcher points out, we already have some pretty good candidates for
mind-bending, decision-altering drugs: cocaine is the most obvious, alcohol
the most popular.
But if scientists do ever come up with a reliable neurobiological model of
human decision making, the implications will be far-reaching. Imagine, for
example, if your employers knew what you wanted, and why you wanted it. With
a simple MRI scan they could predict the minimum salary you would accept.
Glimcher thinks it would take just 20 to 40 quick questions in the scanner
to do this. "My guess is that if a dedicated group tried to be able to make
this prediction, it would take them about six months:' Scary! But equally,
your employer might also want to assess your particular motivations and fit
you to the job you would find most rewarding.
Economists and politicians would undoubtedly like more power to predict what
choices people are likely to make. It is naive to suggest that analysts would
ever be able to second-guess the future. But a greater understanding of what
sways people could make politicians more persuasive. It could also lead to
less volatile stock markets. The RAND corporation might even be persuaded
to give better advice to US presidents than it gave to Truman. Whether they
would choose to take that advice is another matter.