Liars and Outliers (36 page)

It's a tough balancing act, but I think we're up to it. Maybe not in the near term, but in the long term. History teaches how often we get it right. As Martin Luther King, Jr., said: “The arc of history is long, but it bends toward justice.”
5

Acknowledgments

I consider writing a book to be a social as well as a solitary activity. As such, I have a great number of people to thank.

First, thank you to all the participants of the four Security and Human Behavior Workshops, for sparking many of the ideas that went into this book. Thank you to my primary researcher, Kathleen Seidel, for her ability to find just what I need whether I know it exists or not, and also to Deborah Pursch for additional research. Thank you to the Hennepin County Library, which already had many of the books I needed, and got me absolutely everything else I asked for via interlibrary loan…even if I could only keep any one thing for three weeks.

Thank you to all the people who read this book in varying stages of completion. Steve Bass, Charles Faulkner, Greg Guerin, Victoria Gaile Laidler, Stephen Leigh, and David Prentiss read an early draft of the book, and were especially helpful in organizing my initial ideas. Dorothy Denning and Jay Walker read a later draft, and both suggested a reframing—on the same day, coincidentally—that turned out to be particularly useful. Rebecca Kessler gave the book a badly needed final edit. Andrew Acquisti, Andrew Adams, Michael Albaugh, Ross Anderson, Marcia Ballinger, Jason Becker, David Brown, Steve Brust, Tyler Burns, Jon Callas, David Campbell, Raphael Carter, Cody Charette, Dave Clark, Ron Clarke, Chris Cocking, Karen Cooper, David Cowan, Tammy Coxen, Cory Doctorow, John Douceur, Kevin Drum, Nicole Emery, Oisin Feeley, Eric Forste, Amy Forsyth, Peter Fraser-Mackenzie, J. Carl Ganter, Edward Goldstick, Sarah Green, Rachael Greenstadt, Jim Harper, Bill Herdle, Cormac Herley, Chris Hoofnagle, Leif Huhn, Owen Imholte, David Kahn, Jerry Kang, Arlene Katz, John Kelsey, Lori Kingerly, David Leach, David Mandel, Chris Manning, Petréa Mitchell, David Modic, Josh More, Doug Morgenstern, John Mueller, Peter Neumann, Andrew Odlyzko, Evan Oslick, Gerrit Padgham, Cirsten Paine, Ross Patty, David Perry, Daniele Raffo, Coe Roberts, Peter Robinson, Dave Romm, David Ropeik, Marc Rotenberg, Stuart Schechter, Jeff Schmidt, Martin Schneier, David Schroth, Eric Seppanen, Susan Shapiro, Adam Shostack, Daniel Solove, Thomas Sprinkmeier, Nisheeth Srivastava, Frank Stajano, Mark Stewart, Steven L. Victor, Alan Wagner-Krankel, Paul Wallich, Chris Walsh, Rick Wash, Skyler White, and Jeff Yan also commented on all or part of the book.

Thank you to all the readers of my blog, who read my intermittent posts on the book's progress and regularly left insightful comments. Thank you to Carol Kennedy, who suggested the title in one of those comments. Thank you to Luke Fretwell, who designed a cover because he didn't like any of the candidates I posted. His idea ended up being the inspiration for the final cover.

Thank you to Josh More, for creating all the diagrams and putting up with my endless changes and tweaks. Thank you to Kee Nethery for proofreading the galleys. Thank you to Beth Friedman, for copyediting the book before it even got to my publisher, for dealing with all the MS-Word style issues, and for inserting herself between me and the copyeditor.

Thank you to Carol Long, Ashley Zurcher, Ellen Gerstein, and Tom Dinse at Wiley for helping turn the manuscript into an actual book.

Finally, thank you to the Suiboku Museum of Toyama, Japan, for the ink painting of a bulbul that adorned the manuscript cover during the writing process.

Notes

Chapter 1

Numbers preceding the notes refer to endnote numbers.

(1)
In his book,
The Speed of Trust
, Stephen Covey talks about
five levels of trust
, which he calls “waves”: self-trust, relationship trust, organizational trust, market trust, and societal trust.

(2)
Piero Ferrucci wrote:

To trust is to bet
. Each time we trust, we put ourselves on the line. If we confide in a friend, we can be betrayed. If we put faith in a partner, we can be abandoned. If we trust in the world, we can be crushed. Far too often it ends that way. But the alternative is worse still, because if we do not put ourselves on the line, nothing will happen.

(3)
Diego Gambetta: “
When we say we trust
someone or that someone is trustworthy, we implicitly mean that the probability that he will perform an action that is beneficial or at least not detrimental to us is high enough for us to consider engaging in some form of cooperation with him.”

(4)
David Messick and Roderick Kramer: “
We will define trust
in these situations as making the decision
as if
the other person or persons will abide by ordinary ethical rules that are involved in the situation.”

(5)
Sociologist Anthony Giddens proposed a similar
three-level progression
of trust:

Trust in persons…is built upon mutuality of response and involvement: faith in the integrity of another is a prime source of a feeling of integrity and authenticity of the self. Trust in abstract systems provides for the security of day-to-day reliability, but by its very nature cannot supply either the mutuality or intimacy which personal trust relations offer….

In pre-modern settings, basic trust is slotted into personalised trust relations in the community, kinship ties, and friendships. Although any of these social connections can involve emotional intimacy, this is not a condition of the maintaining of personal trust. Institutionalised personal ties and informal or informalised codes of sincerity and honour provide (potential, by no means always actual) frameworks of trust….

With the development of abstract systems, trust in impersonal principles, as well as in anonymous others, becomes indispensable to social existence.

(6)
Piotr Cofta covered
similar ground in his book
Trust, Complexity, and Control
.

(7)
Not coincidentally, I, along with colleagues Ross Anderson and Alessandro Acquisti, founded the annual Interdisciplinary Workshop on Security and Human Behavior in 2008.

(8)
Coming from mathematical security—cryptography—where research results are facts, it can be unsettling to research fields where there are theories, competing theories, overturned theories, and long-standing debates about theories. It sometimes seems that nothing is ever settled in the social sciences, and that for every explanation, there's a counter-explanation. Even worse, a reasonable case can be made that most research
findings are false
and there is
sloppy methodology
in the social sciences, primarily because of the pressure to produce newsworthy results. Also, that many results are based on experiments on a
narrow and unrepresentative
slice of humanity. The only way I can see to navigate this is to look at both the individual research results and the broader directions and meta-results.

(9)
Adam Smith
wrote:

If there is any society among robbers and murderers, they must at least, according to the trite observation, abstain from robbing and murdering one another. Beneficence, therefore, is less essential to the existence of society than justice. Society may subsist, tho’ not in the most comfortable state, without beneficence; but the prevalence of injustice must utterly destroy it.

Chapter 2

(1)
Chimpanzees have been observed using
sticks as weapons
, and wrasses have been observed
using rocks
to open up shells.

(2)
Some of this can be pretty complex; a single
Brants's whistling rat
builds a burrow with dozens or hundreds of entrances, so there's always one close by to retreat to. There's even an
African rat
that applies a tree poison to its fur to make itself deadly.

(3)
Just recently, an
entirely separate
, probably older, immune system was discovered in bacteria and archaea, called
Clustered Regularly Interspaced Short Palindromic Repeats
or CRISPRs.

(4)
In an
earlier book
, I mistakenly called this the “establishing reflex.”

(5)
In one experiment, children were faster at picking out a
picture of a snake
than pictures of more benign objects.

(6)
Stephen Jay Gould used to call these “Just So Stories” because they rarely have any proof other than plausibility (and the fact that they make a good story). So while these seem like possible evolutionary explanations, there is still controversy in evolutionary biology over the levels of selection at work in any given instance. Certainly not all evolutionary biologists would accept these necessarily simple descriptions, although they would concur with the general outline that there was some evolutionary advantage to the possession of certain genes manifesting certain phenotypes in certain populations.

(7)
Among other things, human intelligence is unique in the complexity of its expression, and its ability to comprehend the passage of time. More related to security, humans are vastly ahead of even chimpanzees in their ability to understand
cause and effect
in the physical world.

(8)
No other creature on the planet does this. To use the words of philosopher Alfred Korzybski, humans are the only
time binding species
: we are the only species that can pass information and knowledge between generations at an accelerating rate. Other animals can pass knowledge between generations, but we're the only animal that does it at observable rates.

(9)
All 5,600 or so species of mammals are at least minimally social, if only in mating and child-rearing.

(10)
To use the words of philosopher Daniel Dennett, we need to adopt an
intentional stance
in order to understand each other. That is, instead of looking at people as physical objects or even biological systems, we have to look at them in terms of beliefs, intents, and thoughts.

(11)
There's evidence from rodents that
social group size
is directly correlated with individuality.

(12)
There's even a theory that reasoning evolved not because we needed to make better decisions, but because we needed to
win arguments
and convince other humans.

(13)
Of course, this does not necessarily mean that the sole purpose of the neocortex is to deceive.

(14)
It's actually a range between 100 and 230; 150 is the most common value. Dunbar has often said “150, plus or minus 50.”
Others posit
the number is 200-ish. Groups that are more focused on survival tend to be larger, because “there's safety in numbers.”

(15)
Larger group sizes aren't as stable because their members don't know each other well enough. We interact with people outside this circle more as categories or roles: the mailman, the emergency room nurse, that guy in the accounting department. We might recognize them as individuals, but we tend not to know a lot about them. A modern human might have a virtual network of 2,000 Facebook friends, but it's unlikely that he'll have more than a casual acquaintance with even a tenth of them.

(16)
Modern data from
primitive peoples
validates this number. In the primitive tribes of the New Guinea highlands, who lived apart from the rest of the world until the 1930s, about 25% of men and 5% of women died in warfare. The
Yanomamö
live in the upper reaches of the Orinoco River in Venezuela and Brazil. While they once had only sporadic contact with other cultures, they still lived apart in their traditional manner. They lost 24% of men and 7% of women to warfare.

(17)
Big-game hunting
is inefficient because: 1) big game's low density means fewer encounters, 2) it's harder to catch, 3) it can hurt you when you hunt it, 4) it requires a lot of people to catch, 5) it takes a lot of work to butcher and preserve, and 6) it's perishable, and must be eaten quickly or preserved before it spoils.

(18)
Chimpanzees' aggression rates
are two to three orders of magnitude higher than humans', although their lethal aggression rates are about the same as those of human subsistence societies.

Chapter 3

(1)
There is evidence that
increased specialization
is a function of group size. To be fair, there are researchers who maintain that division of labor is
not what makes
leafcutter ants so successful.

(2)
To some extent, this is also true of other social insects that don't have polymorphism. Bees, for instance, tend to change specializations as they age, but they can change early if some task is going undone. Leafcutter ants can't do this; they're physiologically distinct according to role.

(3)
This
startling statistic
comes from the fact that there are a lot of other organisms in our digestive tract: “The adult human organism is said to be composed of approximately 1013 eukaryotic animal cells. That statement is only an expression of a particular point of view. The various body surfaces and the gastrointestinal canals of humans may be colonized by as many as 1014 indigenous prokaryotic and eukaryotic microbial cells.” Note that the percentage is by number, not by volume or weight. All those digestive organisms are much, much smaller than our own cells.

(4)
The initial paper is actually more complicated. In addition to hawks and doves, there are bullies who only pick on doves, retaliators who respond as hawks against hawks and as doves against doves, and so on. And many other game theorists wrote papers analyzing this or that variant of the Hawk-Dove game, looking at other strategies or more complications in the simulation. But this simple representation is sufficient for our needs. Adding some sort of “fighting skill” parameter is a complexity that doesn't add to our understanding, either.

(5)
Researchers have also conducted Hawk-Dove games with more fluid strategies. Instead of being 100% hawk or 100% dove, individuals could be a combination of both. That is, one individual might behave as a dove 80% of the time and as a hawk 20% of the time. What that individual does in any given situation might be random, or depend on circumstance. This complication better mirrors the behavior of real people.

Another way to make strategies more fluid is to allow individuals to use
some mixture
of hawk and dove strategies in a
single
encounter. So instead of either being all hawk or all dove, an individual might be 20% hawk/80% dove. That is, she might cooperate a lot but not fully or exclusively. This is definitely a more realistic model; we cooperate to different degrees with different people at different times.

In this more complicated model, it's much harder for cooperative behavior to appear. If everyone is constantly switching from the dove camp to the hawk camp and vice versa—as happens in most species—a genetic mutation that enables a small amount of cooperation doesn't confer enough benefit to take hold in the broader population before it gets stamped out by the defectors taking advantage of it.

(6)
Of course this is simplistic. The effects of laws on crime isn't nearly as direct and linear as this example. We'll talk about this more in Chapter 9. But the basic idea is correct.

(7)
The costs and benefits of being a hawk also depend on
population density
. In simulations, dense populations have more doves, and sparse populations more hawks.

(8)
South African meerkats
raise their young communally; even distantly related non-breeders will pitch in to protect newborn pups in their burrows, deliver them beetles, scorpions, and lizards to eat, and even pass along new foods mouth-to-mouth to help them become accustomed to unfamiliar flavors.
Red ruffed lemurs
engage in extensive alloparenting.

(9)
It's
much less common
in the wild. It's
also slow
; there is evidence that mutualism appears to evolve more slowly than other traits.

(10)
Between species, mutualism is more commonly known as symbiosis.
Wrasse cleaner fish
are the canonical example; they eat parasites and dead skin off larger fish. This feeds the wrasses and provides a health benefit to the larger fish. Similarly, clownfish tend to stay within the tentacles of Ritteri sea anemones; each protects the other from predators. Pollination, too: the bees get food, and the plants get pollinated. It is easy for mutualistic relationships to evolve, which is why they are common throughout the natural world.

(11)
Sometimes, the benefit of fighting and winning is so great that most individuals will be hawks. Male
elephant seals
are an example; the winner gets to mate with all the females on the beach. Sometimes, the risk of injury is so low that most individuals will be hawks:
some bullfrogs
function that way, because they can't really injure each other. Sometimes, the risk of injury is so high that almost
everyone is a dove
: oryx and other hoofed animals with nasty horns, rattlesnakes, and so on. Most often, though,
there is a mixture
of hawks and doves within a population. Sometimes it's the more aggressive individuals that are hawks. Sometimes an animal is a hawk within its own territory and a dove outside it.

(12)
Economist
Kaushik Basu
described the problem in the introduction to his book
The Less Developed Economy
. Paraphrasing: Imagine that you are in a strange city, and you've hired a taxi to take you from the airport to your hotel. You and the taxi driver have never previously met, and you'll never meet again. Why do you pay him at the end? If you were just calculating, you might not bother. After all, the taxi driver has already driven you to your destination. Still, you might realize that if you didn't pay, the driver would make a huge fuss, embarrass you in public, perhaps resort to violence, and perhaps call the police. It's just not worth the risk for such a small amount of money. But here's the problem: even if you do pay, the taxi driver could still do all of that. If the taxi driver is just as calculating as you are, why doesn't he accuse you of nonpayment regardless? Double money for him, and he'll never see you again. So if he were going to do that, you might as well not pay. You can take the analysis even further. Maybe you both calculate that if the police got involved, the courts would figure out who wronged the other—maybe there was a camera in the taxi that recorded the whole thing—so it makes sense to be honest. But that doesn't help, either. If the police and the judges are just as calculating as you and the taxi driver, why should they attempt to resolve the dispute fairly, rather than in favor of the side that gave them the biggest bribe? They might fear they would get caught and punished, but that fear assumes those doing the catching and punishing aren't calculating and will attempt to be fair and honest.