by Hugo Mercier
fril s instead of tasty treats.22 In fact, building bowers doesn’t ap-
v i g il a n c e in c o mmuni c at i o n 27
pear to be particularly costly: males are not more likely to die
during the bower- building season, in spite of their construction
efforts.23 So what makes the bower a reliable signal?
The mechanism was discovered, somewhat inadvertently, by
ornithologist Joah Madden when he attempted to trick female
bowerbirds by putting extra berries on some of the bowers.24
Typically, the females prefer to mate with the males that build
the most berry- rich bowers. But the berries added by Madden
had no such effect. It wasn’t that he had bad taste in berries, from
a bowerbird perspective. Instead, rival bowerbirds were sabotag-
ing the bowers that had received Madden’s gift of berries. The
other bowerbirds took these extra berries to mean that the bow-
ers’ owners were pretending to be of higher status than they
actually were, and they vandalized the bowers to put the owners
back in their place.
What keeps the system stable isn’t the intrinsic cost of build-
ing a fancy bower (which is low in any case). Instead, it is the
vigilance of the males, who keep tabs on each other’s bowers and
inflict a cost on those who build exaggerated bowers. As a result,
as long as no male tries to build a better bower than they can af-
ford to defend, the bowers send reliable signals of male quality
without any significant cost being paid. This is costly signaling
for free (or nearly for free, as there are indirect costs in monitor-
ing other males’ bowers).
As we will see, this logic proves critical to understanding the
mechanisms that allow human communication to remain stable.
No intrinsic cost is involved in speaking: unlike buying the lat-
est iPhone, making a promise is not intrinsically costly. Human
verbal communication is the quin tes sen tial “cheap talk” and thus
seems very far from qualifying as a costly signal. This is wrong.
What matters isn’t the cost borne by those who would keep their
28 ch ap t er 2
promises but the cost borne by those who do not keep them. As
long as there is a mechanism to exert a sufficient cost on those
who send unreliable messages—if only by trusting them less in
the future— we’re dealing with costly signaling, and communi-
cation can be kept stable. Undoubtedly, the fact that humans
have developed ways of sending reliable signals without having
to pay a cost every time they do so has greatly contributed to
their success.
The Need for Vigilance
Communication is a tricky business. We find successes and fail-
ures of communication in the most surprising places: prey can
convince predators to give up chasing them, a fetus can’t per-
suade its mother to give it more resources. The logic of evolu-
tion is crucial to understanding these successes and failures. It
tells us when individuals have common incentives: cells in a
body, bees in a beehive. But as the conflicts occurring during
pregnancy illustrate, having some common incentives isn’t
enough. Unless the reproductive fates of two entities are perfectly
intertwined, incentives to send unreliable signals are pretty much
guaranteed to exist. This is when natu ral se lection gets creative,
having stumbled on a variety of ways to keep signals reliable.
Some of these solutions— such as the gazel es’ stots— are fas-
cinating but hardly applicable to human communication. In-
stead, I will argue that human communication is kept (mostly)
reliable by a whole suite of cognitive processes— mechanisms
of open vigilance— that minimize our exposure to unreliable sig-
nals and, by keeping track of who said what, inflict costs on
unreliable senders.
How these mechanisms work, how they help us decide what
to believe and who to trust, is the topic of the next five chapters.
v i g il a n c e in c o mmuni c at i o n 29
What should be clear in any case is that we cannot afford to be
gullible. If we were, nothing would stop people from abusing
their influence, to the point where we would be better off not
paying any attention at all to what others say, leading to the
prompt col apse of human communication and cooperation.
3
EVOLVING OPEN- MINDEDNESS
for humans, the ability to communicate is of enormous
significance. Without communication, we would have a hard
time figuring out what we can safely eat, how to avoid danger,
who to trust, and so forth. Although effective communication
is arguably more impor tant than ever, it was also critical for our
ancestors, who needed to communicate with each other in order
to hunt and gather, to raise their children, to form alliances, and
to pass on technical knowledge.1 Our complex vocal and audi-
tory apparatuses, which clearly serve sophisticated verbal com-
munication, are at least as old as anatomically modern
humans— three hundred thousand years. That our cousins the
Neanderthals, from whom our ancestors split more than six
hundred thousand years ago, appear to have had the same ana-
tomical equipment, suggests complex verbal communication is
considerably older.2
If, from a very early stage in their (pre)history, humans stood
to gain enormous benefits by communicating with each other,
they have also been at risk from the abuse of communication.
More than any other primate species, we are in danger of being
misled and manipulated by communication. The existence of an
evolutionarily relevant prob lem creates se lection pressures that
favor the development of cognitive mechanisms dedicated to
30
e v o l v in g o p e n - mind e d ne s s 31
solving this prob lem. The same is true of communication, with
its promises and its perils.
Indeed, the stakes are so high that it would be puzzling if we
hadn’t evolved specialized cognitive mechanisms to deal not only
with the potential but also with the danger of communication.
In an article from 2010, cognitive scientist Dan Sperber and some
colleagues (including yours truly) called these mechanisms epis-
temic vigilance, but I will call them here open vigilance, to stress
that these mechanisms are at least as much about being open to
communicated information as being vigilant toward it.3 How-
ever, even if we agree such mechanisms must exist, there are dif-
fer ent manners in which they could function.
One way of thinking about the evolution of communication,
and thus our open vigilance mechanisms, is to use the arms race
analogy. An arms race is a competition between two entities in
which each, in response to the other’s move, progressively ups
the ante. The analogy emerged during the Cold War, as Rus sia
and the United States built more nuclear weapons as a reaction
to the other power building more nuclear weapons, which was
a reaction to the other power building more nuclear weapons,
and so on.<
br />
In the case of communication, an arms race could take place
between senders, which evolve increasingly sophisticated means
of manipulating receivers, and receivers, which evolve increas-
ingly sophisticated means of rejecting unreliable messages. This
is what we get, for instance, with computer viruses and security
software. In the human case, this model leads to an associa-
tion between lack of mental acuity and gullibility. Many com-
mentators throughout history have suggested that some
humans— from women to slaves— have stringent intellectual
limitations, limitations that would make these populations gull-
ible (in my terms, by precluding the use of more sophisticated
32 ch ap t er 3
mechanisms of open vigilance). Even assuming we are all en-
dowed with the same cognitive equipment, we are not always
able to rely on it. Thus the arms race model predicts that when
receivers, because they are exhausted or distracted, cannot use
properly their most refined cognitive mechanisms, they would
be defenseless against the senders’ more advanced cognitive de-
vices, in the same way that a security software system that has
not been updated leaves a computer vulnerable to attacks.
Brainwashers and Hidden Persuaders
For Amer ica in the 1950s, fear of manipulation was in the zeit-
geist. With Joseph Stalin still at the helm of the Soviet Union, the
perceived communist threat was at its height, and the United
States had reached peak McCarthyism. The “Reds” were thought
to have infiltrated every thing: the government, academia, de-
fense programs. Even more insidiously, they were supposed to
have wormed their way into the minds of that most dedicated,
most patriotic American: the soldier. During the Korean War,
thousands of U.S. soldiers were captured by the Koreans and the
Chinese. Those who managed to escape brought back tales of
horrible mistreatment and torture, from sleep deprivation to wa-
terboarding. When the war ended, and the prisoners of war
(POWs) were repatriated, these mistreatments acquired an even
darker meaning. Not simply an example of the wanton cruelty
of the enemy, they were seen as an attempt to brainwash U.S.
soldiers into accepting communist doctrine. Twenty- three
American POWs chose to follow their captors to China instead
of going back to their homeland, which was surely, as the New
York Times stated at the time, “living proof that Communist
brainwashing does work on some persons.”4
e v o l v in g o p e n - mind e d ne s s 33
Brainwashing was supposed to function by shattering people’s
ability for higher reflection, as it involved “conditioning,” “debili-
tation,” and “dissociation- hypnosis- suggestibility.”5 For U.S.
rear admiral Daniel Gallery, it made of men “borderline case[s]
between a human being and a rat struggling to stay alive.”6 The
techniques used by the Koreans and Chinese were thought
to be derived from those developed earlier by the Rus sians,
turning the POWs into “prisoners of Pavlov,”7 referring to the
psychologist famous for making dogs salivate at the sound of a
bell. Ironically, Americans in their “war on terror” would go on to
use many of the same techniques— waterboarding being a case
in point— when attempting to extract information from suspected
terrorists.
In 1950s Amer ica, the idea that people are more easily influ-
enced when they cannot think also showed up in a very dif-
fer ent context. The targets weren’t POWs suffering the hell of
Korean prison camps, but moviegoers comfortably watching the
latest Hollywood blockbusters. In the midst of the movie, mes-
sages such as “drink Coke” were presented so quickly that they
could not be consciously perceived.8 These messages would soon
be called subliminal, meaning “below the threshold,” here the
threshold of awareness. Subliminal messages created a scare that
would last for de cades. As late as 2000, a scandal erupted when
a Republican-funded advertisement that attacked policy pro-
posals made by Al Gore— the Demo cratic candidate for the
U.S. presidency— was found to have presented the word rats sub-
liminally to its viewers.9 The power of subliminal messages was
also harnessed for nobler causes. Companies started producing
therapeutic tapes—to enhance self- esteem, say— that people
could listen to in their sleep. Because people don’t tend to
exercise much conscious control when they are asleep, the
34 ch ap t er 3
recordings were aimed directly at their subconscious and were
thus believed to be particularly effective.
The scares surrounding brainwashing and subliminal influ-
ence rely on a pervasive association between inferior cognitive
ability and gullibility: the less we think, the worse we think, and
the more we will be influenced by harmful messages. This as-
sociation between lack of intellectual sophistication and gull-
ibility is historically pervasive. Already when Heraclitus, in 500
BCE, talked about “the people” who “let themselves be led by
speechmakers, in crowds, without considering how many fools
and thieves they are among [them],” he was talking about the
masses, the common people, not the aristocrats.
Twenty- five centuries later, the same trope pervaded the dis-
course of crowd psychologists. These Eu ro pean scholars, work-
ing in the latter half of the nineteenth century, grappled with the
growing impact of crowds in politics, from revolutionary mobs
to striking miners. These scholars developed a view of crowds
as both violent and gullible that would prove im mensely popu-
lar, inspiring Benito Mussolini and Adolf Hitler and being, to
this day, common among those who have to deal with crowds,
such as members of law enforcement agencies.10 The best known
of these crowd psychologists, Gustave Le Bon, suggested that
crowds shared the “absence . . . of critical thought . . . observed
in beings belonging to inferior forms of evolution, such as
women, savages, and children.”11 In a beautiful illustration of mo-
tivated reasoning, Le Bon’s colleague Gabriel Tarde claimed
that because of its “docility, its credulity . . . the crowd is femi-
nine” even when, as he admitted, “it is composed, as is usually
the case, of males.”12 Another of the crowd psychologists, Hip-
polyte Taine, added that in crowds, people were reduced to the
state of nature, like “servile monkeys each imitating the other.”13
At about the same time, on the other side of the Atlantic, Mark
e v o l v in g o p e n - mind e d ne s s 35
Twain was depicting Jim as a “happy, gullible, rather childlike
slave.”14
In the twenty- first century, we still find echoes of these unsa-
vory associations. Writers for the Washington Post and Foreign
Policy claim Donald Trump was elected thanks to the “gullibil-
ity” of “ignorant” voters.15 A common view of Brexit— the v
ote
for Britain to leave the Eu ro pean Union—is to see the Brexiters
as “uneducated plebs” while those who voted remain are “sophis-
ticated, cultured and cosmopolitan.”16
In con temporary academic lit er a ture, the link between unso-
phistication and credulity mostly takes two forms. The first is
in children, whose lack of cognitive maturity is often associated
with gullibility. A recent psy chol ogy textbook asserts that as
children master more complex cognitive skills, they become
“less gullible.”17 Another states, more sweepingly, that “ children,
it seems, are an advertiser’s dream: gul ible, vulnerable, and an
easy sel .”18
The second way in which lack of cognitive sophistication and
credulity are linked is through a popu lar division of thought pro-
cesses into two main types, so- called System 1 and System 2.
According to this view— long established in psy chol ogy and re-
cently pop u lar ized by psychologist Daniel Kahneman’s Think-
ing, Fast and Slow— some cognitive pro cesses are fast, effortless,
largely unconscious, and they belong to System 1. Reading a
simple text, forming a first impression of someone, navigating
well- known streets all belong to System 1. The intuitions that
form System 1 are, on the whole, effective, yet they are also
susceptible to systematic biases. For instance, we seem to make
judgments of people’s competence or trustworthiness on the
basis of facial traits. These judgments may have some limited
reliability, but they should be easily superseded by stronger
cues— such as how the person actually behaves.19 This is when
36 ch ap t er 3
System 2 is supposed to kick in. Relying on slow, effortful, reflec-
tive pro cesses, System 2 takes over when System 1 fails, correct-
ing our mistaken intuitions with its more objective pro cesses
and more rational rules. This is the common dual- process
narrative.20
Maybe the best- known task exemplifying the function of the
two systems is the Bat and Ball:
A bat and a ball cost $1.10 in total. The bat costs $1.00 more
than the ball. How much does the ball cost?21
If you haven’t encountered it before, give it a shot before read-
ing further.
This prob lem has fascinated psychologists because, in spite of
its seeming simplicity, most people provide the wrong answer—