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
Links with orbitofrontal cortex,amygdala and neurons in the midbrain that assess reward make this a likely hub for decision making
Area of the striatum associated with reward that lights up when people taste or see something they like
5.Limbic System
Emotion processing system
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.


Laura Spinney is a freelance writer in London





Chaos Quantum Logic Cosmos Conscious Belief Elect. Art Chem. Maths

NewScientist July 2004  File Info: Created 20/5/2004 Updated --/--/-- Page Address: http://members.fortunecity.com/templarseries/mask.html