by Hugo Mercier
pro cess of calling nature what you exclude from the mere fact
that you are interested in something, this something that
distinguishes itself from being named, nature doesn’t risk
anything but to affirm being a potpourri of non- nature.3
Degrees of Counterintuitiveness
What do these very diverse ideas, from the Trinity to plate tec-
tonics to Lacan’s musings, have in common? First, they have
proved at least somewhat influential, at most widely culturally
successful. Across the world, around 2.4 bil ion people share the
Christian faith. Belief in the god of the Bible is accepted by
sh a l l o w g ur us 219
56 percent of Americans (as of 2018).4 Most people in rich
countries trust science to a significant extent and accept the
majority of the theories scientists agree on (with the odd but
worrying exception).5 Obviously, Lacan could never claim the
same reach, but his authority ran deep, and he boasted many
distinguished intellectuals as his groupies. Twenty years after
his death, Lacan’s teachings were still influential, in France at
least— I should know, having had to suffer through them when
I started my BA in psy chol ogy. More generally, postmodern
thinkers held center stage in the Western intel ectual world for
a good chunk of the twentieth century and exert their influ-
ence to this day. Bruno Latour, who used to be one of them,
now bemoans that “entire Ph.D. programs are still running to
make sure that good American kids are learning the hard way
that facts are made up, that there is no such thing as natu ral,
unmediated, unbiased access to truth; that we are always pris-
oners of language.”6
Besides their popularity, these ideas share another trait: they
do not fit with our intuitions. They either challenge them or pass
them by altogether.
Concepts can be more or less intuitive.7 Take the concept of
“ human.” Once we categorize an agent as human, we can make
a wide variety of inferences: that this agent perceives things,
forms beliefs, has desires and overcomes obstacles to fulfill them,
likes some people more than others, needs to eat and drink, has
a material body, has ancestors who were also humans, eventu-
ally dies, and so forth. Because these inferences come naturally,
the concept of “ human” is intuitive.
Some ideas fail to tap into any of our intuitive concepts: they
are essentially incomprehensible. “Something that distinguishes
itself from being named” doesn’t trigger any concept I’ve mas-
tered, and it fails to ring any inferential bel s.
220 ch ap t er 14
Other concepts yet go against the grain of our intuitions.8 For
instance, because we don’t have well- worked- out concepts of
super natural entities, we have to rely on our concept of human,
even though super natural beings by definition violate a number
of our intuitions. Ghosts are a kind of human that can walk
through walls. Zeus is a kind of human who is immortal and
shoots lightning bolts. The Christian god is a kind of human who
is also an omnipresent, omnipotent, omniscient, all- loving being.
All of these concepts are, in some ways, counterintuitive.
Religious concepts are often counterintuitive, but not all to
the same degree. Pascal Boyer has argued that the vast majority
of concepts of super natural agents found across the world are
only minimally counterintuitive.9 For example, Zeus violates
some of our assumptions about human agents—he is immortal,
for one. But he still re spects most of our preconceptions: he per-
ceives things through his senses, forms beliefs, has desires and
overcomes obstacles to fulfill them, likes some people (or gods)
more than others. Likewise, ghosts are immaterial, but they stil
perceive things through their senses, and so forth.
By contrast, the Christian god, in his ful theological garb, vio-
lates just about every assumption we have about humanlike
agents. Not only is he immortal and immaterial, but he doesn’t
perceive things through his senses or form beliefs (he already
knows every thing), he doesn’t need to overcome obstacles (he
can do every thing he wants), and he doesn’t prefer some people
to others (he loves every one).
Much like the theologically correct Christian god, many sci-
entific concepts are full-on counterintuitive. Our concept of what
moving entails— the feeling that we’re moving, movements of
air, and so forth—is violated by the idea that we’re barreling
through space at a tremendous speed. Our naive sense of biology
tel s us that like begets like, and that microorganisms definitively
sh a l l o w g ur us 221
don’t beget humans. Our naive sense of physics suggests that
humongous rocks don’t drift about with no apparent cause.
To be accepted, ideas that don’t tap into our intuitive con-
cepts, or that go against them, face severe obstacles from open
vigilance mechanisms. We have no reasons to accept ideas we
don’t understand, and we have reasons to reject counterintuitive
ideas. When we engage in plausibility checking, we don’t tend
to reject only ideas that directly clash with our previous views
but also ideas that don’t fit with our intuitions more generally.
For instance, you’ve prob ably never thought about whether there
are penguins on Jupiter. Yet if I told you that some had been re-
cently discovered, you would be skeptical: you have an intuition
that no animals, and especially no terrestrial animals, would be
found there.
Open vigilance also contains mechanisms to overcome plau-
sibility checking and accept beliefs that clash with our previous
views or intuitions: argumentation and trust.
Argumentation is unlikely to play a significant role in the wide
distribution of incomprehensible ideas or counterintuitive con-
cepts. Argumentation works because we find some arguments
intuitively compel ing. This means that premises and conclusions
must be linked by some intuitive inferential pro cess, as when
someone says, “Joe has been very rude to many of us, so he’s a jerk.”
Every one can understand how being repeatedly rude entails
being a jerk. But if a proposition is incomprehensible, then it
can’t properly be argued for. That’s prob ably why Lacan asserts,
rather than argues, that “nature’s specificity is to not be one.”10
Argumentation plays a crucial role in the spread of counter-
intuitive religious and scientific concepts, but only in the small
community of theologians and scientists who can make enough
sense of the arguments to use and construct them. Beyond that,
few people are competent and motivated enough to properly
222 ch ap t er 14
evaluate the technical defense of the Christian god’s omnipo-
tence, or of relativity theory. For example, most U.S. university
students who accept evolution by natu ral se lection don’t under-
stand its princi ples properly.11
Precio
us Shallowness
If argumentation can’t explain the widespread ac cep tance of
incomprehensible or counterintuitive beliefs, then it must be
trust. Trust takes two main forms: trust that someone knows
better (chapter 5), and trust that they have our best interests
at heart (chapter 6). To really change our minds about some-
thing, the former kind of trust is critical: we must believe that
someone knows better than we do and defer to their superior
knowledge.
The preceding examples suggest that people are often so
deferential toward individuals (Lacan), books (the Bible), or
specialized groups (priests, scientists) that they accept incom-
prehensible or counterintuitive ideas. From the point of view
of open vigilance, the latter is particularly problematic. Accept-
ing counterintuitive concepts, concepts that could wreak havoc
with our cognitive systems, seems eminently dangerous, as it
would involve letting other people play around with our way
of thinking. For example, believing that an agent can have the
properties of the Christian god could jeopardize our ability to
draw inferences about humans more generally— after all, our
assumptions about humans are quite sound, and it would be a
shame if something happened to them.
Experiments have shown that, in fact, counterintuitive con-
cepts do not have much of an influence on our intuitive way of
thinking. In the religious domain, psychologist Justin Barrett has
shown that many Christians abide by a form of “theological cor-
sh a l l o w g ur us 223
rectness,” but that their theologically correct beliefs have little
impact on how they actually think about god.12 The Christians
Barrett interviewed were able to describe god’s canonical
features—he knows every thing, is everywhere, and so forth.13
However, when praying, they saw god “like an old man, you
know, white hair,” even though they “know that’s not true.”14
Moreover, when asked to retell a story about god intervening to
save a drowning child, many described god’s actions as sequen-
tial: first, he finishes answering one prayer, then he turns his
attention and powers to the child.15 Omniscient and omnipotent
beings aren’t supposed to get busy or distracted.16
This doesn’t mean Christians can’t draw inferences from their
theologically correct views. If asked whether god is omnipres-
ent, and then whether god is in both this room and the next, they
would answer “yes.” Stil , Barrett’s observations suggest that the
ac cep tance of counterintuitive ideas remains shallow: we can as-
sent to them, even draw inferences from them when pushed,
but they do not affect the way we think intuitively. On the con-
trary, our intuitive way of thinking tends to seep into how we treat
counterintuitive concepts, as when Barrett’s participants implic-
itly thought that god had a limited attention span.
The same logic applies to scientific concepts. Psychologist
Michael McCloskey and his colleagues were among the first to
systematically investigate students’ intuitive physics: how they
answer simple physics prob lems intuitively, without having re-
course to the explicit knowledge of physics acquired in the class-
room.17 One of the experiments involved students at an elite
U.S. university, most of whom had taken some physics classes.
McCloskey and his colleagues confronted the students with a
series of prob lems, such as the one illustrated in figure 4.
Fewer than half of the students were able to provide the cor-
rect answer, namely, that the ball goes on in a straight line. Most
224 ch ap t er 14
Figure 4. What path does a ball launched at the arrow follow when it exits the tube?
Source: Redrawn from McCloskey, Caramazza, & Green, 1980, p. 1139.
said it would keep going in a curve. This means that fewer than
half of the students were able to apply the understanding of iner-
tia they had acquired in school, according to which, in the ab-
sence of any force exerted on them, objects move in a straight line
at constant speed. This notion of inertia is counterintuitive: for
example, our experience tel s us that objects stop moving seem-
ingly of their own accord, absent the application of any force (a
bal eventual y stops rol ing even if it doesn’t hit a wal ). The
counterintuitiveness of the correct notion of inertia means that
it is easily overridden by the students’ intuitions about object
movement. If that seems unfortunate, we should on the whole be
thankful for the limited cognitive influence of counterintuitive
scientific concepts. If our brains could truly pro cess the idea that
we’re darting across space at tremendous speeds following com-
plex curves, we would constantly suffer from motion sickness.18
sh a l l o w g ur us 225
These observations show that counterintuitive ideas, even
when they are held confidently, have no or very limited impact
on the functioning of the intuitive systems with which they are
at loggerheads. To some extent, counterintuitive ideas are pro-
cessed like incomprehensible ideas: even though, in theory, they
should constantly clash with our intuitions, in practice they sim-
ply pass them by. Like many of the misconceptions we have
explored in the last chapters, they remain reflective, detached
from the rest of our cognition.
Charismatic Authority?
The cognitive shal owness of counterintuitive ideas mitigates the
challenge they raise for open vigilance, as accepting such ideas
is much less risky than it would be if they had affected our intui-
tive cognitive mechanisms. But this shallowness doesn’t explain
why people would accept a bunch of bizarre beliefs, some of
which clash with their intuitions: it still seems that people are often
unduly deferential, seeing some authorities as more knowledge-
able than they really are (except for scientists, whose knowledge,
if anything, is likely underestimated).
A common explanation for this undue deference is that some
people are charismatic: their attitude, their voice, their nonver-
bal language make them uniquely enthralling and even credible.
Anthropologist Claude Lévi- Strauss wasn’t a Lacan groupie, and
yet he described “the power, the hold over the audience that ema-
nated both from Lacan’s physical person and from his diction,
his movements.”19 Lacan’s sycophantic French Wikipedia page
even claims that his “style of discourse” “irrevocably affected” the
French language.20
When it comes to widespread religious or scientific beliefs,
charisma cannot be the main explanation. None of our Christian
226 ch ap t er 14
contemporaries have met Jesus, and I’ve managed to accept the
concept of inertia without meeting Galileo. I don’t think that per-
sonal charisma explains at all why some people are deemed
more credible than others. Instead, I outline three mechanisms
that lead some individuals to be perce
ived as more knowledge-
able than they are, making their audience unduly deferential. I
believe that the spread of incomprehensible and counterintui-
tive beliefs largely stems from a mix of these three mechanisms.
Reputation Credit
To understand why we sometimes end up thinking some people
more knowledgeable than they really are, we must go back to
the cues we use to deem individuals more or less knowledge-
able. One of the main cues we rely on is past per for mance.
Someone who is able to consistently fix computers is deemed
competent in this area, and we are more likely to believe them
when they advise us on how to fix our stalled PC. Past per for-
mance doesn’t comprise only actions but also words. People
who give us valuable information are deemed more competent,
which leads to the question of how we decide what information
is valuable.
In many cases, we can judge whether a piece of information
is valuable after the fact: Did our friend’s advice help us fix our
computer? In other cases, we deem a piece of information po-
tentially valuable, and think its source competent, before being
sure that the information is really valuable. We give a kind of
reputation credit. For information to be deemed valuable, it must
be both plausible and useful.21 For example, information about
threats has the potential to be very useful, as it can help us avoid
significant costs. In a series of experiments, Pascal Boyer and psy-
chologist Nora Parren showed that people who transmit infor-
sh a l l o w g ur us 227
mation about threats, by contrast with other types of informa-
tion, are seen as more competent.22
Attributing competence on credit, before we’re sure whether
a message is actually useful or not, works well on the whole, but
there are some loopholes. For instance, we might tend to over-
estimate the usefulness of threat information, deeming it rele-
vant even when we have few chances of ever being exposed to
the actual threat. In Boyer and Parren’s experiments, one of the
stories given to participants mentioned the risk of encountering
leaches when trekking in the Amazon, a situation few partici-
pants would ever face. This meant not only that the information
was not all that useful but also that the participants would never
find out whether or not it was accurate. Indeed, this is a general
