by David Eagleman


Should I eat the ice cream or not? Do I answer this email now or later? Which shoes? Our days are assembled from thousands of small decisions: what to do, which way to go, how to respond, whether to partake. Early theories of decision making assumed that humans are rational actors, tallying the pros and cons of our options to come to an optimal decision. But scientific observations of human decision making don't bear that out. Brains are composed of multiple, competing networks, each of which has its own goals and desires. When deciding whether or not to gobble down the ice cream, some networks in your brain want the sugar; other networks vote against it based on long-term considerations of vanity; other networks suggest that perhaps you could eat the ice cream if you promise yourself you'll go to the gym tomorrow. Your brain is like a neural parliament, composed of rival political parties which fight it out to steer the ship of state. Sometimes you decide selfishly, sometimes generously, sometimes impulsively, and sometimes with the long-view in mind. We are complex creatures because we are composed of many drives, all of which want to be in control.

The sound of a decision

On the operating table, a patient named Jim is undergoing brain surgery to stop tremors of his hand. Long, thin wires called electrodes have been lowered into Jim's brain by the neurosurgeon. By applying a small electric current through the wires, the patterns of activity in Jim's neurons can be adjusted to reduce his tremors.

The electrodes create a special opportunity to eavesdrop on the activity of single neurons. Neurons talk with one another via electrical spikes called action potentials, but these signals are invisibly tiny, so surgeons and researchers often pass the tiny electrical signals through an audio speaker. That way, a miniscule change in voltage (a tenth of a volt that lasts a thousandth of a second) is turned into an audible pop!

As the electrode is lowered through different regions of the brain, the activity patterns of those regions can be recognized by the trained ear. Some locations are characterized by pop!pop!pop! while others sound quite different: pop!....poppop!...pop! It's like suddenly dropping in on the conversation of a few people somewhere randomly on the globe: because the people you land upon will have specific jobs in diverse cultures, they'll all have very different conversations going on.

The monitor showed these tiny spikes of electrical current known as action potentials. Every idea Jim generates, every memory he recollects, every choice he contemplates is written in these tiny, mysterious hieroglyphics.

I'm in the operating room as a researcher: while my colleague performs the surgery, my goal is to better understand how the brain makes decisions. To that end, I ask Jim to perform different tasks - like speaking, reading, looking, deciding - to determine what's correlated with the activity of his neurons. Because the brain has no pain receptors, a patient can be awake during a surgery. I ask Jim to look at a simple picture while we're recording.

What happens in your brain when you see the old woman? What changes when you see the young lady?


In the figure, you may see a young lady with a bonnet looking away. Now try to find another way of interpreting the same image: an old woman looking down and to the left. This picture can be seen in one of two ways (this is known as perceptual bi-stability): the lines on the page are consistent with two very different interpretations. When you stare at the figure, you'll see one version, and then eventually the other, and then the first again, and so on. Here's the important part: nothing on the physical page changes - so whenever Jim reports that the image has flipped, it has to be because of something that changed inside his brain.

The moment he sees the young lady, or the old woman, his brain has made a decision. A decision doesn't have to be conscious; in this case, it's a perceptual decision by Jim's visual system, and the mechanics of the switchover are hidden completely under the hood. In theory, a brain should be able to see both the young lady and the old lady at the same time - but in reality a brain doesn't do that. Reflexively, it takes something ambiguous and makes a choice. Eventually, it remakes the choice, and it might switch back and forth over and over. But our brains are always crushing ambiguity into choices.

So when Jim's brain lands on an interpretation of the young lady - or the old woman - we can listen to the responses from a small number of neurons. Some leap into a higher rate of activity (poppop!.pop!..pop!), while other neurons slow down (pop!....pop!., pop!....pop!). It's not always about speeding up and slowing down: sometimes neurons change their pattern of activity in more subtle ways, becoming synchronized or desynchronized with other neurons even while maintaining their original pace.

The neurons we happen to be spying on are not, by themselves, responsible for the perceptual change - instead, they operate in concert with billions of other neurons, so the changes we can witness are just the reflection of a changing pattern taking hold across large sweeps of brain territory. When one pattern wins out over the other in Jim's brain, a decision has been landed upon.

Your brain makes thousands of decisions every day of your life, dictating your experience of the world. From the decision of what to wear, whom to call, how to interpret an offhand comment, whether to reply to an email, when to leave - decisions underlie our every action and thought. Who you are emerges from the brain-wide battles for dominance that rage in your skull every moment of your life.

Listening to Jim's neural activity - pop!pop!pop! - it's impossible not to be awed. After all, this is what every decision in the history of our species sounded like. Every marriage proposal, every declaration of war, every leap of the imagination, every mission launched into the unknown, every act of kindness, every lie, every euphoric breakthrough, every decisive moment. It all happened right here, in the darkness of the skull, emerging from patterns of activity in networks of biological cells.

The brain is a machine buit from conflict

Let's take a closer look at what's happening behind the scenes during a decision.

Imagine you're making a simple choice, standing in the frozen-yogurt store, trying to decide between two flavors you like equally. Say these are mint and lemon. From the outside, it doesn't look like you're doing much: you're simply stuck there, looking back and forth between the two options. But inside your brain, a simple choice like this unleashes a hurricane of activity.

By itself, a single neuron has no meaningful influence. But each neuron is connected to thousands of others, and they in turn connect to thousands of others, and so on in a massive, loopy, intertwining network. They're all releasing chemicals that excite or depress each other.

Neural populations compete against each other, like political parties struggling for dominance.

Within this web, a particular constellation of neurons represents mint. This pattern is formed from neurons that mutually excite each other. They're not necessarily next to one another; rather, they might span distant brain regions involved in smell, taste, vision, and your unique history of memories involving mint. Each of these neurons, by itself, has little to do with mint - in fact, each neuron plays many roles, at different times, in ever-shifting coalitions. But when these neurons all become active collectively, in this particular arrangement ... that's mint to your brain. As you're standing in front of the yogurt selection, this federation of neurons eagerly communicates with one another like dispersed individuals linking online.

These neurons aren't acting alone in their electioneering. At the same time, the competing possibility - lemon - is represented by its own neural party. Each coalition - mint and lemon - tries to gain the upper hand by intensifying its own activity and suppressing the other's. They fight it out until one triumphs in the winner-take-all competition. The winning network defines what you do next.

Unlike computers, the brain runs on conflict between different possibilities, all of which try to out-compete the others. And there are always multiple options. Even after you've selected mint or lemon, you find yourself in a new conflict: should you eat the whole thing? Part of you wants the delicious energy source, and at the same time part of you knows it's sugary, and perhaps you should be jogging instead. Whether you polish off the whole container is simply a matter of the way the infighting goes.

As a result of ongoing conflicts in the brain, we can argue with ourselves, curse at ourselves, cajole ourselves. But who exactly is talking with whom? It's all you - but it's different parts of you.

Simple tasks can make internal conflicts even more obvious. Name the color of the ink in which each of these words is printed:




Under special circumstances it becomes particularly easy to witness internal conflict between the different parts of the brain. As a treatment for certain forms of epilepsy, some patients undergo "split-brain" surgery, in which the brain's two hemispheres are disconnected from each other. Normally the two hemispheres are connected by a super-highway of nerves called the corpus callosum, and this allows the right and left halves to coordinate and work in concert. If you're feeling chilly, both of your hands cooperate: one holds your jacket hem while the other tugs up the zipper.

But when the corpus callosum is severed, a remarkable and haunting clinical condition can emerge: alien hand syndrome.The two hands can act with totally different intentions: the patient begins to zip up a jacket with one hand, and the other hand (the "alien" hand) suddenly grabs the zipper and pulls it back down. Or the patient might reach for a biscuit with one hand, and their other hand leaps into action to slap the first hand into failure. The normal conflict running in the brain is revealed as the two hemispheres act independently of each other.

Alien hand syndrome normally fades in the weeks after surgery, as the two halves of the brain take advantage of remaining connections to begin coordinating again. But it serves as a clear demonstration that even when we think we're being single-minded, our actions are the product of immense battles that continually rise and fall in the darkness of the cranium.

Information from the left half of the visual field goes to the right hemisphere, and vice versa. As a result, if a flashed word straddles the midline, each independent hemisphere of a split-brain patient will only see half the word.


Difficult, right? Why should this simple task pose any difficulty at all, especially when the instructions are so simple? It's because one network in your brain takes on the task of identifying the color of the ink and putting a name to it. At the same time, competing networks in your brain are responsible for reading words - and these are so proficient that word reading has become a deeply ingrained, automatic process. You can feel the struggle as these systems contend with each other, and to get to the right answer you have to actively suppress the strong impulse to read the word, in deference to concentrating on the ink color. You can directly experience the conflict.


To tease apart some of the major competing systems in the brain, consider a thought experiment known as the trolley dilemma. A trolley is barreling down a train track, out of control. Four workers are making repairs farther down the track, and you, a bystander, quickly realize that they will all be killed by the runaway trolley. Then you notice that there's a lever nearby that can divert the trolley onto another track. But hang on! You see that there's one worker on that track. So if you pull the lever, one worker will be killed; if you don't, four will be killed. Do you pull the lever?

When people are asked what they would do in this scenario, almost everyone pulls the lever. After all, it's far better that only one person is killed rather than four, right?

Now consider a slightly different, second scenario. The situation begins with the same premise: a trolley is barreling down the tracks, out of control, and four workers are going to be killed. But this time you're standing on the deck of a water tower overlooking the tracks, and you notice there's a large man standing up there with you, gazing out into the distance. You realize that if you push him off, he'll land right on the track - and his body weight will be sufficient to stop the trolley and save the four workers. Do you push him off?

The trolley dilemma, scenario 2. 

In this situation, almost no one is willing to push the man. Why not? When asked, they give answers like "that would be murder" and "that would just be wrong." But wait. Aren't you being asked to consider the same equation in both cases? Trading one life for four? Why do the results come out so differently in the second scenario? Ethicists have addressed this problem from many angles, but neuroimaging has been able to provide a fairly straightforward answer. To the brain, the first scenario is just a math problem. The dilemma activates regions involved in solving logical problems.

In the second scenario, you have to physically interact with the man and push him to his death. That recruits additional networks into the decision: brain regions involved in emotion.

When considering pushing an innocent man to his death, networks involved in emotions become more involved in the decision making - and that can flip the outcome.

In the second scenario, we're caught in a conflict between two systems that have different opinions. Our rational networks tell us that one death is better than four, but our emotional networks trigger a gut feeling that murdering the bystander is wrong. You're caught between competing drives, with the result that your decision is likely to change entirely from the first scenario.

The trolley dilemma sheds light on real-world situations. Consider modern warfare, which has become more like pulling the lever than pushing the man off the tower. When a person hits the button to launch a long-range missile, it involves only the networks involved in solving logical problems. Operating a drone can become like a video game; cyber attacks wreak consequences at a distance. The rational networks are at work here, but not necessarily the emotional networks. The detached nature of distance warfare reduces internal conflict, making it easier to wage.

One pundit suggested that the button to launch nuclear missiles should be implanted in the chest of the President's best friend. That way, if he chose to launch nukes, he'd have to inflict physical violence on his friend, tearing him open. That consideration would recruit emotional networks into the decision. When making life-and-death decisions, unchecked reason can be dangerous; our emotions are a powerful and often insightful constituency, and we'd be remiss to exclude them from the parliamentary voting. The world would not be better if we all behaved like robots.

Although the neuroscience is new, this intuition has a long history. The ancient Greeks suggested that we should think of our lives like chariots. We are charioteers trying to hold two horses: the white horse of reason and the black horse of passion. Each horse pulls off-center, in opposite directions. Your job is to keep control of both horses, navigating down the middle of the road.

Indeed, in typical neuroscientific fashion, we can unmask the importance of emotions by seeing what happens when someone loses the capacity to include them in decision making.

States of the body help you decide

Emotions do more than add richness to our lives - they're also the secret behind how we navigate what to do next at every moment. This is illustrated by looking at the situation of Tammy Myers, a former engineer who got into a motorcycle accident. The consequence was damage to her orbitofrontal cortex, the region just above the sockets of the eyes. This brain region is critical for integrating signals streaming in from her body - signals that tell the rest of the brain what state her body is in: hungry, nervous, excited, embarrassed, thirsty, joyful.

Tammy doesn't look like someone who has suffered a traumatic brain injury. But if you were to spend even five minutes with her, you would detect that there's a problem with her ability to handle life's daily decisions. Although she can describe all the pros and cons of a choice in front of her, even the simplest situations leave her mired in indecision. Because she can no longer read her body's emotional summaries, decisions become incredibly difficult for her. Now, no choice is tangibly different from another. Without decision making, little gets done; Tammy reports she often spends all day on the sofa.

Tammy's brain injury tells us something crucial about decision making. It's easy to think about the brain commanding the body from on high - but in fact the brain is in constant feedback with the body. The physical signals from the body give a quick summary of what's going on and what to do about it. To land on a choice, the body and the brain have to be in close communication.

Consider this situation: you want to pass a misdelivered package over to your next-door neighbors. But as you approach the gate to their yard, their dog growls and bares its teeth. Do you open the gate and press on to the front door? Your knowledge of the statistics of dog attacks isn't the deciding factor here - instead, the dog's threatening posture triggers a set of physiological responses in your body: an increased heart rate, a tightening in the gut, a tensing of the muscles, pupil dilation, changes in blood hormones, opening of sweat glands, and so on. These responses are automatic and unconscious.

In this moment, standing there with your hand on the gate latch, there are many external details you could assess (for example, the color of the dog's collar) - but what your brain really needs to know right now is whether you should face the dog or deliver the package another way. The state of your body helps you in this task: it serves as a summary of the situation. Your physiological signature can be thought of as a low-resolution headline: "this is bad" or "this is no problem." And that helps your brain decide what to do next.

Blood pressure, heart and respiratory rate increase

Increased tension and blood flow m large muscles

Clotting factors and sugars made more available in blood

Most situations involve too many details to reach a decision purely through logic. To guide the process, we need abridged summaries: "I'm safe here" or "I'm in danger here." The physiological state of the body maintains a constant two-way dialog with the brain.

Physiological response to fear

Pupils dilate, tear and salivary glands dry up.

Every day we read the states of our bodies like this. In most situations, physiological signals are more subtle, and so we tend to be unaware of them. However, those signals are crucial to steering the decisions we have to make. Consider being in a supermarket: this is the kind of place which leaves Tammy paralyzed with indecision. Which apples? Which bread? Which ice cream? Thousands of choices bear down upon shoppers, with the end result that we spend hundreds of hours of our lives standing in the aisles, trying to make our neural networks commit to one decision over another. Although we don't commonly realize it, our body helps us to navigate this boggling complexity.

Take the choice of which kind of soup to buy. There's too much data here for you to grapple with: calories, price, salt content, taste, packaging, and so on. If you were a robot, you'd be stuck here all day trying to make a decision, with no obvious way to trade off which details matter more. To land on a choice, you need a summary of some sort. And that's what the feedback from your body is able to give you. Thinking about your budget might make your palms sweat, or you might salivate thinking about the last time you consumed the chicken noodle soup, or noting the excessive creaminess of the other soup might put a cramp in your intestines. You simulate your experience with one soup, and then the other. Your bodily experience helps your brain to quickly place a value on soup A, and another on soup B, allowing you to tip the balance in one direction or the other. You don't just extract the data from the soup cans, you feel the data. These emotional signatures are more subtle than the ones related to facing down a barking dog, but the idea is the same: each choice is marked by a bodily signature. And that helps you to decide.

Earlier, when you were deciding between the mint and lemon yogurt, there was a battle between networks. The physiological states from your body were the key things that helped tip that battle, that allowed one network to win over another. Because of her brain damage, Tammy lacks the ability to integrate her bodily signals into her decision making. So she has no way to rapidly compare the overall value between options, no way to prioritize the dozens of details that she can articulate. That's why Tammy stays on the sofa much of the time: none of the choices in front of her carry any particular emotional value. There's no way to tip one network's campaign over any other. The debates in her neural parliament continue along in deadlock.

Because the conscious mind has low bandwidth, you don't typically have full access to the bodily signals that tip your decisions; most of the action in your body lives far below your awareness. Nonetheless, the signals can have far-reaching consequences on the type of person you believe you are. As one example, neuroscientist Read Montague has found a link between a person's politics and the character of their emotional responses. He puts participants in a brain scanner and measures their response to a series of images chosen to evoke a disgust response, from images of feces to dead bodies to insect-covered food. When they emerge from the scanner, they are asked if they would like to take part in another experiment; if they say "yes" they take ten minutes to answer a political ideology survey. They are asked questions about their feelings on gun control, abortion, premarital sex, and so on. Montague finds that the more disgusted a participant is by the images, the more politically conservative they are likely to be. The less disgusted, the more liberal. The correlation is so strong that a persons neural response to a single disgusting image predicts their score on the political ideology test with 95% accuracy. Political persuasion emerges at the intersection of the mental and the corporal.

Traveling to the future

Each decision involves our past experiences (stored in the states of our body) as well as the present situation (Do I have enough money to buy X instead of Y? Is option Z available?). But there's one more part to the story of decisions: predictions about the future.

Across the animal kingdom, every creature is wired to seek reward. What is a reward? At its essence, it's something that will move the body closer to its ideal set points. Water is a reward when your body is getting dehydrated; food is a reward when your energy stores are running down. Water and food are called primary rewards, which directly address biological needs. More generally, however, human behavior is steered by secondary rewards, which are things that predict primary rewards. For example, the sight of a metal rectangle wouldn't by itself do much for your brain, but because you've learned to recognize it as a water fountain, then the sight of it comes to be rewarding when you are thirsty. In the case of humans, we can find even very abstract concepts rewarding, such as the feeling that we are valued by our local community. And unlike animals, we can often put these rewards ahead of biological needs. As Read Montague points out, "sharks don't go on hunger strikes": the rest of the animal kingdom only chases its basic needs, while only humans regularly override those needs in deference to abstract ideals. So when we're faced with an array of possibilities, we integrate internal and external data to try to maximize reward, however it's defined to us as individuals.

The challenge with any reward, whether basic or abstract, is that choices typically don't yield their fruits right away. We almost always have to make decisions in which a chosen course of action returns reward at a later time. People go to school for years because they value the future concept of having a degree, they slave through employment they don't enjoy with the future hope of a promotion, and they push themselves through painful exercise with the goal of being fit.

To compare different options means assigning a value to each one in a common currency - that of anticipated reward - and then choosing the one with the highest value. Consider this scenario: I have a bit of free time and I'm trying to decide what to do. I need to get groceries, but I also know I need to get to a coffee shop and work on a grant for my lab, because a deadline is coming up. I also want to spend time with my son at the park. How I do arbitrate this menu of options?

It would be easy, of course, if I could directly compare these experiences by living each one, and then rewinding time, and finally choosing my path based on which outcome was the best. Alas, I cannot travel in time.

Or can I?

As in the movie Back to the Future, humans time travel daily.

Time travel is something the human brain does relentlessly. When faced with a decision, our brains simulate different outcomes to generate a mockup of what our future might be. Mentally, we can disconnect from the present moment and voyage to a world that doesn't yet exist.

Now, simulating a scenario in my mind is just the first step. To decide between the imagined scenarios, I try to estimate what the reward will be in each of those potential futures. When I simulate filling my pantry with the groceries, I feel a sense of relief at being organized and avoiding uncertainty. The grant carries different sorts of rewards: not only money for the laboratory, but more generally the kudos from my department chairman and a rewarding sense of accomplishment in my career. Imagining myself at the park with my son inspires joy, and a sense of reward in terms of family closeness. My final decision will be navigated by how each future stacks up against the others in the common currency of my reward systems. The choice isn't easy, because all these valuations are nuanced: the simulation of the grocery shopping is accompanied by feelings of tedium; the grant writing is attended by a sense of frustration; the park with guilt about not getting work done. Typically under the radar of awareness, my brain simulates all the options, one at a time, and does a gut check on each. That's how I decide.

How do I accurately simulate these futures? How can I possibly predict what it will really be like to go down these paths? The answer is that I can't: there's no way to know that my predictions will be accurate. All my simulations are only based on my past experiences and my current models of how the world works. Like all animals in the animal kingdom, we can't just wander around hoping to randomly discover what results in future reward and what doesn't. Instead, the key business of brains is to predict. And to do this reasonably well, we need to continually learn about the world from our every experience. So in this case, I place a value on each of these options based on my past experiences. Using the Hollywood studios in our minds, we travel in time to our imagined futures to see how much value they'll have. And that's how I make my choices, comparing possible futures against one another. That's how I convert competing options into a common currency of future reward.

Think of my predicted reward value for each option like an internal appraisal that stores how good something will be. Because grocery shopping will supply me with food, let's say it's worth ten reward units. Grant writing is difficult but necessary to my career, so it weighs in at twenty-five reward units. I love spending time with my son, so going to the park is worth fifty reward units.

But there's an interesting twist here: the world is complicated, and so our internal appraisals are never written in permanent ink. Your valuation of everything around you is changeable, because quite often our predictions don't match what actually happens. The key to effective learning lies in tracking this prediction error, the difference between the expected outcome of a choice and the outcome that actually occurred.

In today's case, my brain has a prediction about how rewarding the park is going to be. If we run into friends there and it turns out even better than I thought, that raises the appraisal the next time I'm making such a decision. On the other hand, if the swings are broken and it rains, that lowers my appraisal the next time around.

How does this work? There's a tiny, ancient system in the brain whose mission is to keep updating your assessments of the world. This system is made of tiny groups of cells in your midbrain that speak in the language of a neurotransmitter called dopamine.

When there's a mismatch between your expectation and your reality, this midbrain dopamine system broadcasts a signal that re-evaluates the price point. This signal tells the rest of the system whether things turned out better than expected (an increased burst of dopamine) or worse (a decrease in dopamine). That prediction error signal allows the rest of the brain to adjust its expectations to try to be closer to reality next time. The dopamine acts as an error corrector: a chemical appraiser that always works to make your appraisals as updated as they can be. That way, you can prioritize your decisions based on your optimized guesses about the future.

Dopamine-releasing neurons involved in decision making are concentrated into tiny regions of the brain called the ventral tegmental area and the substantia nigra. Despite their small sizes, they have a wide reach, broadcasting updates when the predicted value of a choice turns out to be too high or too low.

Fundamentally, the brain is tuned to detect unexpected outcomes - and this sensitivity is at the heart of animals' ability to adapt and learn. It's no surprise, then, that the brain architecture involved in learning from experience is consistent across species, from honeybees to humans. This suggests that brains discovered the basic principles of learning from reward long ago.

The power of now

So we've seen how values get attached to different options. But there's a twist that often gets in the way of good decision making: options right in front of us tend to be valued higher than those we merely simulate. The thing that trips up good decision making about the future is the present.

In 2008, the US economy took a sharp downturn. At the heart of the trouble was the simple fact that many homeowners had over-borrowed. They had taken out loans that offered wonderfully low interest rates for a period of a few years. The problem occurred at the end of the trial period, when the rates went up. At the higher rates, many homeowners found themselves unable to make the payments. Close to a million homes went into foreclosure, sending Shockwaves through the economy of the planet.

What did this disaster have to do with competing networks in the brain? These subprime loans allowed people to obtain a nice house now, with the high rates deferred until later. As such, the offer perfectly appealed to the neural networks that desire instant gratification - that is, those networks that want things now. Because the seduction of the immediate satisfaction pulls so strongly on our decision making, the housing bubble can be understood not simply as an economic phenomenon, but also as a neural one.

The pull of the now wasn't just about the people borrowing, of course, but also the lenders who were getting rich, right now, by offering loans that weren't going to get paid. They rebundled the loans and sold them off. Such practices are unethical, but the temptation proved too enticing to many thousands.

This now-versus-the-future battle doesn't just apply to housing bubbles, it cuts across every aspect of our lives. It's why car dealers want you to get in and test-drive the cars, why clothing stores want you to try on the clothes, why merchants want you to touch the merchandise. Your mental simulations can't live up to the experience of something right here, right now.

To the brain, the future can only ever be a pale shadow of the now. The power of now explains why people make decisions that feel good in the moment but have lousy consequences in the future: people who take a drink or a drug hit even though they know they shouldn't; athletes who take anabolic steroids even though it may shave years off their lives; married partners who give in to an available affair.

Can we do anything about the seduction of the now? Thanks to competing systems in the brain, we can. Consider this: we all know that it's difficult to do certain things, like go regularly to the gym. We want to be in shape, but when it comes down to it, there are usually things right in front of us that seem more enjoyable. The pull of what we're doing is stronger than the abstract notion of future fitness. So here's the solution: to make certain you get to the gym, you can take inspiration from a man who lived 3,000 years ago.

Overcoming the power of now: the Ulysses contract

This man was in a more extreme version of the gym scenario. He had something he wanted to do, but knew he wouldn't be able to resist temptation when the time came. For him it wasn't about getting a better physique; it was about saving his life from a group of mesmerizing maidens.

This was the legendary hero Ulysses, on his way back from triumph in the Trojan War. At some point on his long journey home, he realized that his ship would soon be passing an island where the beautiful Sirens lived. The Sirens were famous for singing songs so melodious that sailors were rapt and enchanted. The problem was that the sailors found the women irresistible, and would crash their ships into the rocks trying to get to them.

Ulysses desperately wanted to hear the legendary songs, but he didn't want to kill himself and his crew. So he hatched a plan. He knew that when he heard the music, he would be unable to resist steering toward the islands rocks. The problem wasn't the present rational Ulysses, but instead the future, illogical Ulysses - the person he'd become when the Sirens came within earshot. So Ulysses ordered his men to lash him securely to the mast of the ship. They filled their ears with beeswax so as not to hear the Sirens, and they rowed under strict orders to ignore any of his pleas and cries and writhing.

Ulysses knew that his future self would be in no position to make good decisions. So the Ulysses of sound mind arranged things so that he couldn't do the wrong thing. This sort of deal between your present and future self is known as a Ulysses contract.

In the case of going to the gym, my simple Ulysses contract is to arrange in advance for a friend to meet me there: the pressure to uphold the social contract lashes me to the mast. When you start looking for them, you'll see that Ulysses contracts are all around you. Take college students who swap Facebook passwords during the week of their final exams; each student changes the password of the other so that neither can log on until finals are over. The first step for alcoholics in rehabilitation programs is to clear all the alcohol from their home, so the temptation is not in front of them when they're feeling weak. People with weight problems sometimes get surgery to reduce their stomach volume so they physically cannot overeat. In a different twist on a Ulysses contract, some people arrange things so that a violation of their promise will trigger a financial donation to an "anti-charity." For example, a woman who fought for equal rights her whole life wrote out a large check to the Ku Klux Klan, with strict orders to her friend to mail the check if she smoked another cigarette.

In all these cases, people structure things in the present so that their future selves can't misbehave. By lashing ourselves to the mast we can get around the seduction of the now. It's the trick that lets us behave in better alignment with the kind of person we would like to be. The key to the Ulysses contract is recognizing that we are different people in different contexts. To make better decisions, it's important not only to know yourself but all of your selves.

The invisible mechanisms of decision making

Knowing yourself is only part of the battle - you also have to know that the outcome of your battles will not be the same every time. Even in the absence of a Ulysses contract, sometimes you'll feel more enthusiastic about going to the gym, and sometimes less so. Sometimes you're more capable of good decision making, and other times your neural parliament will come out with a vote you later regret. Why? It's because the outcome depends on many changing factors about the state of your body, states which can change hour to hour. For example: two men serving a prison sentence are scheduled to appear before a parole board. One prisoner comes before the board at 11:27 am. His crime is fraud and he's serving thirty months. Another prisoner appears at 1:15 pm. He has committed the same crime, for which he had been given the same sentence.

The first prisoner is denied parole; the second is granted parole. Why? What influenced the decision? Race? Looks? Age?

A study in 2011 analyzed a thousand rulings from judges, and found it likely wasn't about any of those factors. It was mostly about hunger. Just after the parole board had enjoyed a food break, a prisoner's chance of parole rose to its highest point of 65%. But a prisoner seen towards the end of a session had the lowest chances: just a 20% likelihood of a favorable outcome.

In other words, decisions get reprioritized as other needs rise in importance. Valuations change as circumstances change. A prisoner's fate is irrevocably intertwined with the judge's neural networks, which operate according to biological needs.

Some psychologists describe this effect as "ego-depletion," meaning that higher-level cognitive areas involved in executive function and planning (for example, the prefrontal cortex) get fatigued. Willpower is a limited resource; we run low on it, just like a tank of fuel. In the case of the judges, the more cases they had to make decisions about (up to thirty-five in one sitting) the more energy-depleted their brains became. But after eating something like a sandwich and a piece of fruit, their energy stores were refueled and different drives had more power in steering decisions.

Traditionally, we assume that humans are rational decision makers: they absorb information, process it, and come up with




We spend plenty of energy cajoling ourselves into making decisions we feel we ought to. To stay on the straight and narrow, we often fall back on willpower: that inner strength which allows you to pass on the cookie (or at least the second cookie), or which allows you to hit a deadline when you really want to be out in the sunshine. We all know what it feels like when our willpower feels run down: after a long, hard day at work, people often find themselves making poorer choices - for example, eating a larger meal than they intended to, or watching television instead of hitting their next deadline.

So psychologist Roy Baumeister and colleagues put it to a closer test. People were invited to watch a sad movie. Half were told to react as they normally would, while the other half were instructed to suppress their emotions. After the movie, they were all given a hand exerciser and asked to squeeze it for as long as they could. Those who had suppressed their emotions gave up sooner. Why? Because self-control requires energy, which means we have less energy available for the next thing we need to do. And that's why resisting temptation, making hard decisions, or taking initiative all seem to draw from the same well of energy. So willpower isn't something that we just exercise - it's something we deplete.

The dorsolateral prefrontal cortex becomes active when dieters choose the healthier food options in front of them, or when people choose to forego a small reward now for a better outcome later.


an optimal answer or solution. But real humans don't operate this way. Even judges, striving for freedom from bias, are imprisoned in their biology.

Our decisions are equally influenced when it comes to how we act with our romantic partners. Consider the choice of monogamy - bonding and staying with a single partner. This would seem like a decision that involves your culture, values, and morals. All that is true, but there's a deeper force acting on your decision making as well: your hormones. One in particular, called oxytocin, is a key ingredient in the magic of bonding. In one recent study, men who were in love with their female partners were given a small dose of extra oxytocin. They were then asked to rate the attractiveness of different women. With the extra oxytocin, the men found their partners more attractive - but not other women. In fact, the men kept a bit more physical distance from an attractive female research associate in the study. Oxytocin increased bonding to their partner.

Why do we have chemicals like oxytocin steering us toward bonding? After all, from an evolutionary perspective, we might expect that a male shouldn't want monogamy if his biological mandate is to spread his genes as widely as possible. But for the survival of the children, having two parents around is better than one. This simple fact is so important that the brain possesses hidden ways to influence your decision making on this front.

Decisions and society

A better understanding of decision making opens the door to better social policy. For example, each of us, in our own way, struggles with impulse control. At the extreme, we can end up as slaves to the immediate cravings of our impulses. From this perspective, we can gain a more nuanced understanding of social endeavors such as the War on Drugs.

Drug addiction is an old problem for society, leading to crime, diminished productivity, mental illness, disease transmission - and, more recently, to a burgeoning prison population. Nearly seven out often prisoners meet the criteria for substance abuse or dependence. In one study, 35.6% of convicted inmates were under the influence of drugs at the time of their criminal offense. Drug abuse translates into many tens of billions of dollars, mostly in terms of drug-related crime.

Most countries deal with the problem of drug addiction by criminalizing it. A few decades ago, 38,000 Americans were in prison for drug-related offenses. Today, it's half a million. On the face of it, that might sound like success in the War on Drugs - but this mass incarceration hasn't slowed the drug trade. This is because, for the most part, the people behind bars aren't the cartel bosses, or the mafia dons, or the big-time dealers - instead, the prisoners have been locked up for possession of a small amount of drugs, usually less than two grams. They're the users. The addicts. Going to prison doesn't solve their problem - it generally worsens it.

The US has more people in prison for drug-related crimes than the European Union has prisoners. The problem is that incarceration triggers an expensive and vicious cycle of relapse and re-imprisonment. It breaks people's existing social circles and employment opportunities, and gives them new social circles and new employment opportunities - ones that typically fuel their addiction.

Every year the US spends $20 billion on the War on Drugs; globally, the total is over $100 billion. But the investment hasn't worked. Since the war began, drug use has expanded. Why hasn't the expenditure succeeded? The difficulty with drug supply is that it's like a water balloon: if you push it down in one place, it comes up somewhere else. Instead of attacking supply, the better strategy is to address demand. And drug demand is in the brain of the addict.

Some people argue that drug addiction is about poverty and peer pressure. Those do play a role, but at the core of the issue is the biology of the brain. In laboratory experiments, rats will self-administer drugs, continually hitting the delivery lever at the expense of food and drink. The rats aren't doing that because of finances or social coercion. They're doing it because the drugs tap into fundamental reward circuitry in their brains. The drugs effectively tell the brain that this decision is better than all the other things it could be doing. Other brain networks may join the battle, representing all the reasons to resist the drug. But in an addict, the craving network wins. The majority of drug addicts want to quit but find themselves unable. They end up becoming slaves to their impulses. Because the problem with drug addiction lies in the brain, it's plausible that the solutions lie there too. One approach is to tip the balance of impulse control. This can be achieved by ramping up the certainty and swiftness of punishment - for instance, by requiring drug offenders to undergo twice-weekly drug testing, with automatic, immediate jail time for failure - thereby not relying on distant abstraction alone. Similarly, some economists propose that the drop in American crime since the early 1990s has been due, in part, to the increased presence of police on the streets. In the language of the brain, the police visibility stimulates the networks that weigh long-term consequences.

In my laboratory, we're working on another potentially effective approach. We are giving real-time feedback during brain imaging, allowing cocaine addicts to view their own brain activity and learn how to regulate it.

Meet one of our participants, Karen. She is bubbly and intelligent, and at fifty years old she retains a youthful energy. She's been addicted to crack cocaine for over two decades, and she describes the drug as having ruined her life. If she sees the drug right in front of her, she feels no choice but to take it. In ongoing experiments in my lab, we put Karen into the brain scanner (functional magnetic resonance imaging, or fMRI). We show her pictures of crack cocaine, and ask her to crave. That's easy for her to do, and it activates particular regions of her brain that we summarize as the craving network. Then we ask her to suppress her craving. We ask her to think about the cost crack cocaine has had to her - in terms of finances, in terms of relationships, in terms of employment. That activates a different set of brain areas, which we summarize as the suppression network. The craving and suppression networks are always battling it out for supremacy, and whichever wins at any moment determines what Karen does when offered crack.

Using fast computational techniques in the scanner, we can measure which network is winning: the short-term thinking of the craving network, or the long-term thinking of the impulse control or suppression network. We give Karen real-time visual feedback in the form of a speedometer so she can see how that battle is going. When her craving is winning, the needle is in the red zone; as she successfully suppresses, the needle moves to the blue zone. She can then use different approaches to discover what works to tip the balance of these networks.

By practicing over and over, Karen gets better at understanding what she needs to do to move the needle. She may or may not be consciously aware of how she's doing it, but by repeated practice she can strengthen the neural circuitry that allows her to suppress. This technique is still in its infancy, but the hope is that when she's next offered crack she'll have the cognitive tools to overcome her immediate cravings if she wants to. This training does not force Karen to behave


Some networks in the brain are involved in craving (red); others in suppressing the temptation (blue). Using real-time feedback in neuroimaging, we measure the activity in the two networks and give a participant visual feedback about how well they're fighting the battle.


in any particular way; it simply gives her the cognitive skills to have more control over her choice, rather than to be a slave to her impulses.

Drug addiction is a problem for millions of people. But prisons aren't the place to solve the problem. Equipped with an understanding of how human brains actually make decisions, we can develop new approaches beyond punishment. As we come to better appreciate the operations inside our brains, we can better align our behavior with our best intentions.

More generally, a familiarity with decision making can improve aspects of our criminal justice system well beyond addiction, ushering in policies which are more humane and cost-effective. What might that look like? It would begin with an emphasis on rehabilitation over mass incarceration. This may sound illusory, but in fact there are places already pioneering such an approach with great success. One such place is Mendota Juvenile Treatment Center in Madison, Wisconsin.

Many of the twelve to seventeen-year-olds at Mendota have committed crimes that might otherwise qualify them for life in prison. Here, it qualifies them for admission. For many of the children, this is their last chance. The program started in the early 1990s to provide a new approach to working with youths the system had given up on. The program pays particular attention to their young, developing brains. As we saw in Chapter 1, without a fully developed prefrontal cortex, decisions are often made impulsively, without meaningful consideration of future consequences. At Mendota, this viewpoint illuminates an approach to rehabilitation. To help the children improve their self-control, the program provides a system of mentoring, counseling, and rewards. An important technique is to train them to pause and consider the future outcome of any choice they might make - encouraging them to run simulations of what might happen - thereby strengthening neural connections that can override the immediate gratification of impulses.

Poor impulse control is a hallmark characteristic of the majority of criminals in the prison system. Many people on the wrong side of the law generally know the difference between right and wrong actions, and they understand the threat of the punishment - but they are hamstrung by poor impulse control. They see an older woman with an expensive purse, and they don't pause to consider other options besides taking advantage of the opportunity. The temptation in the now overrides any consideration of the future.

While our current style of punishment rests on a bedrock of personal volition and blame, Mendota is an experiment in alternatives. Although societies possess deeply ingrained impulses for punishment, a different kind of criminal justice system - one with a closer relationship to the neuroscience of decisions - can be imagined. Such a legal system wouldn't let anyone off the hook, but it would be more concerned with how to deal with law breakers with an eye toward their future rather than writing them off because of their past. Those who break the social contracts need to be off the streets for the safety of society - but what happens in prison does not have to be based only on bloodlust, but also on evidence-based, meaningful rehabilitation.

Decision making lies at the heart of everything: who we are, what we do, how we perceive the world around us. Without the ability to weigh alternatives, we would be hostages to our most basic drives. We wouldn't be able to wisely navigate the now, or plan our future lives. Although you have a single identity, you're not of a single mind: instead, you are a collection of many competing drives. By understanding how choices battle it out in the brain, we can learn to make better decisions for ourselves, and for our society.








Keith Hunt