On Rumors : How Falsehoods Spread, Why We Believe Them, What Can Be Done

by Sunstein, Cass R.

• ISBN13: 9780809094738
• ISBN10: 0809094738
• Format: Hardcover
• Copyright: 2009-09-29
• Publisher: Farrar, Straus and Giroux
• More Book Details
• Purchase Benefits

Summary

You're being misled. From whom President Obama pals around with to the causes of the current economic crisis, rumor propagators know precisely how to start a rumor and see it entrenched. They are attempting to derail political candidates, ruin companies and reputations, even damage democratic governance. And in the era of the Internet, they know more about manipulating the mechanics of false rumorssocial cascades, group polarization, and biased assimilationthan you do. They also know that the presumed correctivespublishing balanced information, issuing corrections, and trusting to the marketplace of ideasdo not work.On Rumorsexplains what the rumor propagators know, describes why the marketplace of ideas is broken, and offers a solution that challenges a legal tradition tracing back to Oliver Wendell Holmes: reset the "chill" on free speech to dampen false and damaging rumors. A pioneer in the effort "to design regulation around the ways people behave" (The Wall Street Journal), Cass R. Sunstein uses examples from the real world and from behavioral studies to explain why certain rumors spread like wildfire, what we can do to avoid being misled, and what the government could do to discourage those who are out to mislead us.

Author Biography

Cass R. Sunstein is the Felix Frankfurter Professor of Law at Harvard Law School (on leave).

Table of Contents

The Problem
Rumors are nearly as old as human history, but with the rise of the Internet, they have become ubiquitous. In fact we ..tions,

and they often resist correction. They can threaten careers, policies, public officials, and sometimes even democracy itself.
Many of the most pervasive rumors involve famous .panies, large and small. Still others involve people who are not at all in the

public eye. All of us are potential victims of rumors, including false and vicious ones.
In the 2008 election, many Americans believed that Barack Obama was a Muslim, that he was not born in the United States,

and that he “pals around with terrorists.” Rumors are pervasive about the allegedly terrible acts, beliefs, and motivations of

public officials and about the allegedly scandalous private lives not only of those officials, but also of many other people with a

high public profile. Rumors can harm the economy as well. If it is rumored that a company is about to fail, stockholders might

well be frightened, and they might sell. Because of the rumor, the company might be seriously harmed. Rumors can and do

affect the stock market itself, even if they are baseless. It should not be entirely surprising that the Securities and Exchange

Commission has taken a keen interest in the pernicious effects of false rumors, and that New York has made it a crime to

circulate false rumors about the financial status of banks.
In the era of the Internet, it has become easy to spread false or misleading rumors about almost anyone. A high school

student, a salesperson, a professor, a banker, an employer, an insurance broker, a real estate agent—each of .ful, damaging,

or even devastating effect. If an allegation of misconduct appears on the Internet, those who Google the .tion will help to define

the person. (It might even end up on Wikipedia, at least for a time.) The rumor can involve .gence Agency, General Motors,

Bank of America, the Boy Scouts, the Catholic Church. Material on the Internet has considerable longevity. For all practical

purposes, it may even be permanent. For this reason, a false rumor can have an enduring effect.
This small book has two goals. The first is to answer these questions: Why do ordinary human beings accept rumors, even

false, destructive, and bizarre ones? Why do some groups, and even nations, accept rumors that other groups and nations

deem preposterous? The second is to answer this question: What can we do to protect ourselves against the harmful effects

of false rumors? As we shall see, part of the answer lies in recognizing that a “chilling effect” on those who would spread

destructive falsehoods can be an excellent idea.
We will also see that when people believe rumors, the believers are often perfectly rational, in the sense that their belief is

quite sensible in light of their existing knowledge. We lack direct or personal knowledge about the facts that underlie most of

our judgments. How do you know that the .ter is made of atoms? That the Holocaust actually occurred? That Lee Harvey

Oswald assassinated President Kennedy? .ple, other nations, other cultures, other religions. We rarely know for sure whether

a particular company is in terrible trouble, or whether a particular public official has taken a bribe, or whether an influential

person has a terrible secret agenda or a shameful incident in her past. Lacking personal knowledge, we tend to think that

where there is smoke, there is fire—or that a rumor would not have spread unless it was at least partly true. Perhaps the truth

is even worse than the rumor. Certainly we should be cautious before entrusting our nation or our company to the hands of

someone who is rumored to have said or done bad things. Our willingness to think in this way causes special problems when

we rely on the Internet for our information, simply because false rumors are so pervasive there.
There is no settled definition of rumors, and I will not attempt to offer one here. To get the discussion off the ground, let us

acknowledge the crudeness of any definition, put semantic debates to one side, and take the term to refer roughly to claims of

fact—about people, groups, events, and institutions—that have not been shown to be true, but that .ity not because direct

evidence is known to support them, .stood, rumors often arise and gain traction because they fit with, and support, the prior

convictions of those who accept them. Some people and some groups are predisposed to .ible with their self-interest, or with

what they think they know to be true. In 2008, many Americans were prepared to believe that Governor Sarah Palin thought

that Africa .ulous confusion fit with what they already thought about Governor Palin. Other people were predisposed to reject

the same rumor as probably baseless. Exposure to the same information spurred radically different beliefs.
Many of us accept false rumors because of either our fears or our hopes. Because we fear al-Qaeda, we are inclined to believe

that its members are plotting an attack .pany will prosper, we might believe a rumor that its new product cannot fail and that

its prospects are about to soar. In the context of war, one group’s fears are unmistakably another group’s hopes—and

whenever groups compete, the fears of some are the hopes of others. Because rumors fuel some fears and alleviate others,

radically different reactions to the same rumor are inevitable. The citizens of Iraq may accept a rumor that has no traction in

Canada or France. Those in Utah may accept a rumor that seems preposterous .crats ridicule. And to the extent that the

Internet enables people to live in information cocoons, or echo chambers of their own design, different rumors will become

entrenched in different communities.
Many rumors spread conspiracy theories.1 Consider the rumor that the Central Intelligence Agency was responsible for the

assassination of President John F. Kennedy; that doctors deliberately manufactured the AIDS virus; that the ..erate fraud; that

the Trilateral Commission is responsible for important movements of the international economy; that Martin Luther King, Jr.,

was killed by federal agents; that the plane crash that killed the Democratic senator Paul Wellstone was engineered by

Republican politicians; that the moon landing was staged; that the Rothschilds and .dents and for economic distress in Asian

nations; and that the Great Depression was a result of a plot by wealthy people to reduce the wages of workers.2 Or consider

the work of the French author Thierry Meyssan, whose book 9/11: The Big Lie became a bestseller and a sensation for its

claims that the Pentagon explosion on 9/11 was caused by a missile, fired as the opening salvo of a coup d’état by the

military-industrial complex, rather than by American Airlines Flight 77.3
Rumors spread through two different but overlapping processes: social cascades and group polarization. Cascades occur

because each of us tends to rely on what other people think and do. If most of the people we know believe a rumor, we tend to

believe it too. Lacking information of our own, we accept the views of others. When the rumor involves a topic on which we

know nothing, we are especially likely to believe it. If the National Rifle Association spreads a rumor that a political candidate

wants to “confiscate guns,” or if an environmental organization spreads a rumor that someone believes that climate change is

“a hoax,” many people will be affected, because they tend to believe the National Rifle Association or the environmental

organization.
A cascade occurs when a group of early movers, sometimes called bellwethers, say or do something and other people fol­low

their signal. In the economy, rumors can fuel speculative bubbles, greatly inflating prices, and indeed speculative bub­bles help

to account for the financial crisis of 2008. Rumors are also responsible for many panics, as fear spreads rapidly from one

person to another, creating self-fulfilling prophe­cies. And if the relevant rumors trigger strong emotions, such as fear and

disgust, they are far more likely to spread.
Group polarization refers to the fact that when like-minded people get together, they often end up thinking a more extreme

version of what they thought before they started to talk to one another.4 Suppose that members of a certain group are inclined

to accept a rumor about, say, the malevolent intentions of a certain nation. In all likeli­hood, they will become more committed

to that rumor after they have spoken among themselves. Indeed, they may have moved from being tentative believers to being

absolutely certain that the rumor is true, even though all they know is what other group members think. Consider the role of

the Internet here: any one of us might receive numerous com­munications from many of us, and when we receive those

communications, we might think that whatever is being said must be true. 
What can be done to reduce the risk that cascades and polarization will lead people to accept false rumors? The most

obvious answer, and the standard one, involves the system of free expression: people should be exposed to bal­anced

information and to corrections from those who know the truth. Freedom usually works, but in some contexts, it is an

incomplete corrective. Emotions can get in the way of truth-seeking. People do not process information in a neutral way. Their

preconceptions affect their reactions. Biased assimilation refers to the fact that people assimilate new information in a biased

fashion; those who have accepted false rumors do not easily give up their beliefs, especially when they have a strong

emotional commitment to those beliefs. It can be exceedingly hard to dislodge what people think, even by presenting them

with the facts.
Many people firmly believe in the “marketplace of ideas.” They think that the marketplace is the best way to ensure that

people arrive at the truth. In one of the greatest opin­ions in all of American law, Justice Oliver Wendell Holmes argued that

“the ultimate good desired is better reached by free trade in ideas—that the best test of truth is the power of the thought to get

itself accepted in the competition of the market.”5 This powerful claim has exerted an enduring and salutary influence on the

law of free speech, not merely in the United States, but throughout the world.
For some rumors, however, the marketplace does not work so well. Consider, for example, the potential conse­quences of a

rumor of criminal behavior by a neighbor of yours, someone with no access to the media and without credibility on the Internet.

Or suppose that an emotion­ally gripping rumor is starting to spread about the leader of a local company. Far from being the

best test of truth, the marketplace can ensure that many people accept false­hoods, or that they take mere fragments of lives,

or small events, as representative of some alarming whole. The problem is serious and pervasive, and—with the growing

influence of the Internet and new kinds of surveillance—it seems to be increasing. On occasion, it results in serious harm to

people’s lives, damages the prospects of businesses, hurts investors, and undermines democracy itself.
We should underline the last point in particular. Free speech is meant, in part, to promote self-government; a well-functioning

democracy cannot exist unless people are able to say what they think, even if what they think is false. But if people spread

false rumors—most obviously about public officials and institutions—democracy itself will suf­fer. For no good reason, people

might lose faith in particular leaders and policies, and even in their government itself. At the same time, false rumors impede

our ability to think well, as citizens, about what to do about a crisis, whether large or small.
These points should not be taken as a plea for any kind of censorship. It is true and important that any effort to reg­ulate

speech will create a chilling effect. Punish people for spreading falsehoods, and you will find yourself “chilling” truth. Suppose

that the law will hold people accountable if they circulate a false rumor about a bank. To be sure, it is good if people are not

injured as a result of that false rumor. But that very law might discourage someone else from disclosing, on the basis of

credible evidence, the fact that a bank is in real trouble. Pointing to the risk of a chill­ing effect on free speech and hence on

the transmission of truth, reasonable people often suggest that the government should allow a great deal of breathing space for

falsehoods, even damaging ones. They suggest that the less regulation of the marketplace, the better.
Under certain assumptions, they are probably right. But there is a countervailing consideration. Sometimes a chilling effect

can be an excellent safeguard. Without such an effect, the marketplace of ideas will lead many people to spread and to

accept damaging falsehoods about both individuals and institutions. If false rumors create serious problems, we must be

careful to ensure that the fear of a chilling effect does not itself have a chilling effect on public discussion or on our practices.

These falsehoods can hurt or even ruin individual lives. They can also have serious economic consequences. This risk is

precisely what led New York to enact a law making it a crime to spread false rumors about banks. As we have seen, false

rumors can undermine democracy itself. For all these reasons, it is sensible to hope that social norms and even law will

impose a certain chill on them. We need, in short, to find ways to discourage the harmful effects of false rumors.
One of my major goals here is to sketch the mecha­nisms that lie behind false rumors—their propagation, their transmission,

and their entrenchment. Many of those who seek to spread rumors have an intuitive awareness of those mechanisms;

sometimes their understanding is highly sophisticated. Many propagators know exactly what they are doing. It follows that

those who would protect themselves or others from false rumors must understand the underlying mechanisms as well. We

shall see that while old-style cen­sorship is out of the question, it is legitimate for courts to use libel law to protect

people—whether or not in public life—from falsehoods. But part of my goal has nothing at all to do with law. It is to suggest

the possibility of what social scientists call debiasing—in this case, through an improved understanding of how information

spreads. That understanding might lead us to be more cautious in accept­ing false rumors, and in the process help to create a

kind of culture that avoids injury or even destruction to personal lives and valuable institutions, both large and small.

Propagators
Why do rumors start? Why do some rumors obtain large audiences while other rumors fall from their own (lack of) weight? Let

us begin by making some distinctions.
Rumors are often initiated by self-conscious propa­gators, who may or may not believe the rumors that they spread. Rumor

propagators have diverse motivations. To understand the current situation, we need to identify them.
Some propagators are narrowly self-interested.

 

Excerpted from On Rumors by Cass R. Sunstein.
Copyright © 2009 by Cass R. Sunstein.
Published in 2009 by Farrar, Straus and Giroux.
All rights reserved. This work is protected under copyright laws and
reproduction is strictly prohibited. Permission to reproduce the material in
any manner or medium must be secured from the Publisher.

Excerpts

The Problem
Rumors are nearly as old as human history, but with the rise of the Internet, they have become ubiquitous. In fact we ..tions,

and they often resist correction. They can threaten careers, policies, public officials, and sometimes even democracy itself.
Many of the most pervasive rumors involve famous .panies, large and small. Still others involve people who are not at all in the

public eye. All of us are potential victims of rumors, including false and vicious ones.
In the 2008 election, many Americans believed that Barack Obama was a Muslim, that he was not born in the United States,

and that he “pals around with terrorists.” Rumors are pervasive about the allegedly terrible acts, beliefs, and motivations of

public officials and about the allegedly scandalous private lives not only of those officials, but also of many other people with a

high public profile. Rumors can harm the economy as well. If it is rumored that a company is about to fail, stockholders might

well be frightened, and they might sell. Because of the rumor, the company might be seriously harmed. Rumors can and do

affect the stock market itself, even if they are baseless. It should not be entirely surprising that the Securities and Exchange

Commission has taken a keen interest in the pernicious effects of false rumors, and that New York has made it a crime to

circulate false rumors about the financial status of banks.
In the era of the Internet, it has become easy to spread false or misleading rumors about almost anyone. A high school

student, a salesperson, a professor, a banker, an employer, an insurance broker, a real estate agent—each of .ful, damaging,

or even devastating effect. If an allegation of misconduct appears on the Internet, those who Google the .tion will help to define

the person. (It might even end up on Wikipedia, at least for a time.) The rumor can involve .gence Agency, General Motors,

Bank of America, the Boy Scouts, the Catholic Church. Material on the Internet has considerable longevity. For all practical

purposes, it may even be permanent. For this reason, a false rumor can have an enduring effect.
This small book has two goals. The first is to answer these questions: Why do ordinary human beings accept rumors, even

false, destructive, and bizarre ones? Why do some groups, and even nations, accept rumors that other groups and nations

deem preposterous? The second is to answer this question: What can we do to protect ourselves against the harmful effects

of false rumors? As we shall see, part of the answer lies in recognizing that a “chilling effect” on those who would spread

destructive falsehoods can be an excellent idea.
We will also see that when people believe rumors, the believers are often perfectly rational, in the sense that their belief is

quite sensible in light of their existing knowledge. We lack direct or personal knowledge about the facts that underlie most of

our judgments. How do you know that the .ter is made of atoms? That the Holocaust actually occurred? That Lee Harvey

Oswald assassinated President Kennedy? .ple, other nations, other cultures, other religions. We rarely know for sure whether

a particular company is in terrible trouble, or whether a particular public official has taken a bribe, or whether an influential

person has a terrible secret agenda or a shameful incident in her past. Lacking personal knowledge, we tend to think that

where there is smoke, there is fire—or that a rumor would not have spread unless it was at least partly true. Perhaps the truth

is even worse than the rumor. Certainly we should be cautious before entrusting our nation or our company to the hands of

someone who is rumored to have said or done bad things. Our willingness to think in this way causes special problems when

we rely on the Internet for our information, simply because false rumors are so pervasive there.
There is no settled definition of rumors, and I will not attempt to offer one here. To get the discussion off the ground, let us

acknowledge the crudeness of any definition, put semantic debates to one side, and take the term to refer roughly to claims of

fact—about people, groups, events, and institutions—that have not been shown to be true, but that .ity not because direct

evidence is known to support them, .stood, rumors often arise and gain traction because they fit with, and support, the prior

convictions of those who accept them. Some people and some groups are predisposed to .ible with their self-interest, or with

what they think they know to be true. In 2008, many Americans were prepared to believe that Governor Sarah Palin thought

that Africa .ulous confusion fit with what they already thought about Governor Palin. Other people were predisposed to reject

the same rumor as probably baseless. Exposure to the same information spurred radically different beliefs.
Many of us accept false rumors because of either our fears or our hopes. Because we fear al-Qaeda, we are inclined to believe

that its members are plotting an attack .pany will prosper, we might believe a rumor that its new product cannot fail and that

its prospects are about to soar. In the context of war, one group’s fears are unmistakably another group’s hopes—and

whenever groups compete, the fears of some are the hopes of others. Because rumors fuel some fears and alleviate others,

radically different reactions to the same rumor are inevitable. The citizens of Iraq may accept a rumor that has no traction in

Canada or France. Those in Utah may accept a rumor that seems preposterous .crats ridicule. And to the extent that the

Internet enables people to live in information cocoons, or echo chambers of their own design, different rumors will become

entrenched in different communities.
Many rumors spread conspiracy theories.1 Consider the rumor that the Central Intelligence Agency was responsible for the

assassination of President John F. Kennedy; that doctors deliberately manufactured the AIDS virus; that the ..erate fraud; that

the Trilateral Commission is responsible for important movements of the international economy; that Martin Luther King, Jr.,

was killed by federal agents; that the plane crash that killed the Democratic senator Paul Wellstone was engineered by

Republican politicians; that the moon landing was staged; that the Rothschilds and .dents and for economic distress in Asian

nations; and that the Great Depression was a result of a plot by wealthy people to reduce the wages of workers.2 Or consider

the work of the French author Thierry Meyssan, whose book 9/11: The Big Lie became a bestseller and a sensation for its

claims that the Pentagon explosion on 9/11 was caused by a missile, fired as the opening salvo of a coup d’état by the

military-industrial complex, rather than by American Airlines Flight 77.3
Rumors spread through two different but overlapping processes: social cascades and group polarization. Cascades occur

because each of us tends to rely on what other people think and do. If most of the people we know believe a rumor, we tend to

believe it too. Lacking information of our own, we accept the views of others. When the rumor involves a topic on which we

know nothing, we are especially likely to believe it. If the National Rifle Association spreads a rumor that a political candidate

wants to “confiscate guns,” or if an environmental organization spreads a rumor that someone believes that climate change is

“a hoax,” many people will be affected, because they tend to believe the National Rifle Association or the environmental

organization.
A cascade occurs when a group of early movers, sometimes called bellwethers, say or do something and other people fol­low

their signal. In the economy, rumors can fuel speculative bubbles, greatly inflating prices, and indeed speculative bub­bles help

to account for the financial crisis of 2008. Rumors are also responsible for many panics, as fear spreads rapidly from one

person to another, creating self-fulfilling prophe­cies. And if the relevant rumors trigger strong emotions, such as fear and

disgust, they are far more likely to spread.
Group polarization refers to the fact that when like-minded people get together, they often end up thinking a more extreme

version of what they thought before they started to talk to one another.4 Suppose that members of a certain group are inclined

to accept a rumor about, say, the malevolent intentions of a certain nation. In all likeli­hood, they will become more committed

to that rumor after they have spoken among themselves. Indeed, they may have moved from being tentative believers to being

absolutely certain that the rumor is true, even though all they know is what other group members think. Consider the role of

the Internet here: any one of us might receive numerous com­munications from many of us, and when we receive those

communications, we might think that whatever is being said must be true. 
What can be done to reduce the risk that cascades and polarization will lead people to accept false rumors? The most

obvious answer, and the standard one, involves the system of free expression: people should be exposed to bal­anced

information and to corrections from those who know the truth. Freedom usually works, but in some contexts, it is an

incomplete corrective. Emotions can get in the way of truth-seeking. People do not process information in a neutral way. Their

preconceptions affect their reactions. Biased assimilation refers to the fact that people assimilate new information in a biased

fashion; those who have accepted false rumors do not easily give up their beliefs, especially when they have a strong

emotional commitment to those beliefs. It can be exceedingly hard to dislodge what people think, even by presenting them

with the facts.
Many people firmly believe in the “marketplace of ideas.” They think that the marketplace is the best way to ensure that

people arrive at the truth. In one of the greatest opin­ions in all of American law, Justice Oliver Wendell Holmes argued that

“the ultimate good desired is better reached by free trade in ideas—that the best test of truth is the power of the thought to get

itself accepted in the competition of the market.”5 This powerful claim has exerted an enduring and salutary influence on the

law of free speech, not merely in the United States, but throughout the world.
For some rumors, however, the marketplace does not work so well. Consider, for example, the potential conse­quences of a

rumor of criminal behavior by a neighbor of yours, someone with no access to the media and without credibility on the Internet.

Or suppose that an emotion­ally gripping rumor is starting to spread about the leader of a local company. Far from being the

best test of truth, the marketplace can ensure that many people accept false­hoods, or that they take mere fragments of lives,

or small events, as representative of some alarming whole. The problem is serious and pervasive, and—with the growing

influence of the Internet and new kinds of surveillance—it seems to be increasing. On occasion, it results in serious harm to

people’s lives, damages the prospects of businesses, hurts investors, and undermines democracy itself.
We should underline the last point in particular. Free speech is meant, in part, to promote self-government; a well-functioning

democracy cannot exist unless people are able to say what they think, even if what they think is false. But if people spread

false rumors—most obviously about public officials and institutions—democracy itself will suf­fer. For no good reason, people

might lose faith in particular leaders and policies, and even in their government itself. At the same time, false rumors impede

our ability to think well, as citizens, about what to do about a crisis, whether large or small.
These points should not be taken as a plea for any kind of censorship. It is true and important that any effort to reg­ulate

speech will create a chilling effect. Punish people for spreading falsehoods, and you will find yourself “chilling” truth. Suppose

that the law will hold people accountable if they circulate a false rumor about a bank. To be sure, it is good if people are not

injured as a result of that false rumor. But that very law might discourage someone else from disclosing, on the basis of

credible evidence, the fact that a bank is in real trouble. Pointing to the risk of a chill­ing effect on free speech and hence on

the transmission of truth, reasonable people often suggest that the government should allow a great deal of breathing space for

falsehoods, even damaging ones. They suggest that the less regulation of the marketplace, the better.
Under certain assumptions, they are probably right. But there is a countervailing consideration. Sometimes a chilling effect

can be an excellent safeguard. Without such an effect, the marketplace of ideas will lead many people to spread and to

accept damaging falsehoods about both individuals and institutions. If false rumors create serious problems, we must be

careful to ensure that the fear of a chilling effect does not itself have a chilling effect on public discussion or on our practices.

These falsehoods can hurt or even ruin individual lives. They can also have serious economic consequences. This risk is

precisely what led New York to enact a law making it a crime to spread false rumors about banks. As we have seen, false

rumors can undermine democracy itself. For all these reasons, it is sensible to hope that social norms and even law will

impose a certain chill on them. We need, in short, to find ways to discourage the harmful effects of false rumors.
One of my major goals here is to sketch the mecha­nisms that lie behind false rumors—their propagation, their transmission,

and their entrenchment. Many of those who seek to spread rumors have an intuitive awareness of those mechanisms;

sometimes their understanding is highly sophisticated. Many propagators know exactly what they are doing. It follows that

those who would protect themselves or others from false rumors must understand the underlying mechanisms as well. We

shall see that while old-style cen­sorship is out of the question, it is legitimate for courts to use libel law to protect

people—whether or not in public life—from falsehoods. But part of my goal has nothing at all to do with law. It is to suggest

the possibility of what social scientists call debiasing—in this case, through an improved understanding of how information

spreads. That understanding might lead us to be more cautious in accept­ing false rumors, and in the process help to create a

kind of culture that avoids injury or even destruction to personal lives and valuable institutions, both large and small.

Propagators
Why do rumors start? Why do some rumors obtain large audiences while other rumors fall from their own (lack of) weight? Let

us begin by making some distinctions.
Rumors are often initiated by self-conscious propa­gators, who may or may not believe the rumors that they spread. Rumor

propagators have diverse motivations. To understand the current situation, we need to identify them.
Some propagators are narrowly self-interested.

 

Excerpted from On Rumors by Cass R. Sunstein.
Copyright © 2009 by Cass R. Sunstein.
Published in 2009 by Farrar, Straus and Giroux.
All rights reserved. This work is protected under copyright laws and
reproduction is strictly prohibited. Permission to reproduce the material in
any manner or medium must be secured from the Publisher.