14
will be one where all the individual beliefs are mutually coherent and rein-
15
forcing.
16
Some planets are not stable. People go through life with a very large
17
number of beliefs, some of which may not be compatible with others, even
18
if they don’t recognize it. We should think of planets of belief as undergo-
19
ing gradual but constant churning, bringing different beliefs into contact
20
with one another, just as real planets experience convection in the mantle
21
and plate tectonics near the surface. When two dramatically incompatible
22
beliefs come into direct contact, it can be like highly reactive chemicals be-
23
ing mixed together, leading to an impressive explosion— possibly even
24
blowing the entire planet apart, until a new one can be reassembled from
25
different parts.
26
Ideally, we should be constantly testing and probing our planets of belief
27
for inconsistencies and structural deficiencies. Precisely because they are
28
floating freely through space, rather than remaining anchored on solid and
29
immovable ground, we should always be willing to improve on our planets’
30
composition and architecture, even to the point of completely jettisoning
31
old beliefs and replacing them with better ones. The new information we
32
receive through our observations is like the rain of meteors and comets that
33
is perpetually falling on real planets, to be incorporated into our view of the
34
world. Occasionally, there may even be an asteroid impact of such magni-
S35
tude that the entire planet is destroyed. These instabilities, either from
N36
117
Big Picture - UK final proofs.indd 117
20/07/2016 10:02:42
T H E B IG PIC T U R E
01
internal inconsistency or from an external shock, are more likely to happen
02
to relatively young planets, ones that have not completely settled down, but
03
we’re all vulnerable.
04
The real problem is that we can imagine more than one stable planet—
05
there can be multiple sets of beliefs that are consistent within the sets, but
06
not among them. One person’s planet might include the scientific method,
07
as well as the belief that the universe is billions of years old; another’s might
08
include a belief in biblical literalism, as well as the belief that the world was
09
created a few thousand years ago. If each planet consists of beliefs that are
10
consistent with each other, how do we ever know which is right?
11
This is a real worry. People do hold beliefs that clash violently with the
12
beliefs of others, even though they may seem consistent with other beliefs
13
of their own. But there is reason to hope that the problem isn’t insur-
14
mountable.
15
As a matter of empirical fact, there are a number of important, common
16
beliefs that almost everyone shares. Most people believe that reason and
17
logic play an important role in finding truth. They might disagree over
18
whether those are uniquely powerful techniques, but very few people reject
19
them outright. We also tend to share the goal of coming up with models of
20
the world that provide accurate representations of what we actually ob-
21
serve. If you confront a young- Earth creationist who thinks that the world
22
came into being 6,000 years ago with scientific evidence for a very old Earth
23
and universe, their typical response is not “Oh, I don’t believe in evidence
24
and logic.” Rather, they will attempt to account for that evidence within
25
their belief system, for example, by explaining why God would have created
26
the universe that way.
27
That’s the way it’s supposed to work, anyway. But mere “coherence”
28
might seem like precious little on which to base a theory of truth. Aban-
29
doning the quest for a secure foundation in favor of a planet of belief is like
30
moving from firm ground to a boat on choppy seas or a spinning teacup
31
ride. It can make you dizzy, if not seasick. We are spinning through space,
32
nothing to hold on to.
33
What rescues our beliefs from being completely arbitrary is that one of
34
the beliefs in a typical planet is something like “true statements correspond
35S
to actual elements of the real world.” If we believe that, and have some reli-
36N
able data, and are sufficiently honest with ourselves, we can hope to
118
Big Picture - UK final proofs.indd 118
20/07/2016 10:02:42
Pl A n E t S O F b E l I E F
construct belief systems that not only are coherent but also agree with those
01
of other people and with external reality. At the very least, we can hold that
02
up as a goal.
03
There is a crucial difference, in other words, between stable planets of
04
belief, ones where all the different pieces attract one another in a consistent
05
and coherent way, and habitable planets, ones where we could actually live.
06
A habitable planet of belief necessarily includes some shared convictions
07
about evidence and rationality, as well as the actual information we have
08
gathered about the world. We can hope that people working in good faith
09
will, after trying hard to understand reality the best they can, end up con-
10
structing planets of belief that are somewhat compatible with one another.
11
12
•
13
We shouldn’t overestimate people’s rationality or willingness to look at new
14
evidence as objectively as possible. For better or for worse, planets eventu-
15
ally develop highly sophisticated defense mechanisms. When you realize
16
that you are holding two beliefs that are in conflict with each other, psy-
17
chologists refer to the resulting discomfort as cognitive dissonance. It’s a sign 18
that there is something not completely structurally sound about your planet
19
of belief. Unfortunately, human beings are extremely good at maintaining
20
the basic makeup of their planets, even under very extreme circumstances.
21
Leon Festinger, an
American social psychologist who was the founder
22
of cognitive- dissonance theory, and his collaborators once studied an apoc-
23
alyptic cult led by a woman named Dorothy Martin (known to generations
24
of psychology students by the pseudonym Marian Keech). Following Mar-
25
tin’s lead, members of her group became convinced that the Earth was go-
26
ing to be destroyed on December 21, 1954, but that the true believers would
27
be rescued by aliens the night before. The cult members were extremely
28
serious; they quit their jobs, left their families, and huddled together to
29
await the big day. Festinger was curious as to how they would react when—
30
as his own planet of belief led him to surmise— nothing special happened
31
on the appointed day. Would they, confronted with the unassailable fact
32
that their leader’s prophecy had been incorrect, change their minds about
33
her mystical powers?
34
The day came and went— and afterward, the believers were more con-
S35
vinced of Martin’s prophetic ability than ever. On the morning of the
N36
119
Big Picture - UK final proofs.indd 119
20/07/2016 10:02:42
T H E B IG PIC T U R E
01
twenty- first, as it happened, Martin had conveyed a new vision: it was pre-
02
cisely the unflagging belief of their small group that had been able to pre-
03
vent the Earth’s destruction. Overjoyed, and very ready to believe, her
04
followers doubled down on their commitment, and proceeded to try to
05
spread the word of their insights as widely as possible.
06
Human beings are not nearly as coolly rational as we like to think we
07
are. Having set up comfortable planets of belief, we become resistant to al-
08
tering them, and develop cognitive biases that prevent us from seeing the
09
world with perfect clarity. We aspire to be perfect Bayesian abductors, im-
10
partially reasoning to the best explanation— but most often we take new
11
data and squeeze it to fit with our preconceptions.
12
It’s worth highlighting two important cognitive biases that we can look
13
to avoid as we put together our own planets. One is our tendency to give
14
higher credences to propositions that we want to be true. This can show up
15
at a very personal level, as what’s known as self- serving bias: when something 16
good happens, we think it’s because we are talented and deserving, while
17
bad things are attributed to unfortunate luck or uncontrollable external
18
circumstances. At a broader level, we naturally gravitate toward theories of
19
the world that somehow flatter ourselves, make us feel important, or pro-
20
vide us with comfort.
21
The other bias is our preference for preserving our planet of belief, rather
22
than changing it around. This can also show up in many ways. Confirma-
23
tion bias is our tendency to latch on to and highlight any information that
24
confirms beliefs we already have, while disregarding evidence that may
25
throw our beliefs into question. This tendency is so strong that it leads to
26
the backfire effect— show someone evidence that contradicts what they be-
27
lieve, and studies show that they will usually come away holding their ini-
28
tial belief even more strongly. We cherish our beliefs, and work hard to
29
protect them against outside threats.
30
Our need to justify our own beliefs can end up having a dramatic influ-
31
ence on what those beliefs actually are. Social psychologists Carol Tavris
32
and Elliot Aronson talk about the “Pyramid of Choice.” Imagine two peo-
33
ple with nearly identical beliefs, each confronted with a decision to make.
34
One chooses one way, and the other goes in the other direction, though
35S
initially it was a close call either way. Afterward, inevitably, they work to
36N
12 0
Big Picture - UK final proofs.indd 120
20/07/2016 10:02:42
Pl A n E t S O F b E l I E F
convince themselves that the choice they made was the right one. They each
01
justify what they did, and begin to think there wasn’t much of a choice at
02
all. By the end of the process, these two people who started out almost the
03
same have ended up on opposite ends of a particular spectrum of belief—
04
and often defending their position with exceptionally fervent devotion. “It’s
05
the people who almost decide to live in glass houses who throw the first
06
stones,” as Tavris and Aronson put it.
07
08
•
09
We’re faced with the problem that the beliefs we choose to adopt are shaped
10
as much, if not more, by the beliefs we already have than by correspondence
11
with external reality.
12
How can we guard ourselves against self- reinforcing irrationality? There
13
is no perfect remedy, but there is a strategy. Knowing that cognitive biases
14
exist, we can take that fact into account when doing our Bayesian inference.
15
Do you want something to be true? That should count against it in your
16
assignment of credences, not for it. Does new, credible evidence seem in-
17
compatible with your worldview? We should give it extra consideration, not
18
toss it aside.
19
A Utopia of rationality might not be achievable by flawed human be-
20
ings, but it’s something to which we can aspire. Robert Aumann, an Israeli
21
American mathematician who shared the Nobel Prize in Economics in
22
2005, was able to prove a wonderful mathematical theorem: two people,
23
both acting rationally, who start with the same Bayesian prior credences for
24
their beliefs, and who have access to the same information, including know-
25
ing what the other knows, cannot disagree about the updated credences for
26
those beliefs. You might think that people can start with common priors
27
but disagree about the likelihoods for observations being obtained, but Au-
28
mann’s theorem shows that this can’t happen if both share “common
29
knowledge”— that is, when everyone knows what everyone el
se knows (and
30
they all know that they all know it).
31
Aumann’s “agreement theorem” sounds too good to be true, in part be-
32
cause it doesn’t comport very well with actual human behavior. In the real
33
world, people are not completely rational, they don’t have common knowl-
34
edge, they misinterpret one another, and they certainly don’t start with the
S35
N36
121
Big Picture - UK final proofs.indd 121
20/07/2016 10:02:42
T H E B IG PIC T U R E
01
same priors. But it gives us hope that we could come to common agreement,
02
even on very contentious issues, if we worked hard enough at it. Even wildly
03
different priors will eventually be swamped by the process of updating if we
04
collect enough evidence. If we try to be as honest as possible with others
05
and with ourselves, we can hope to bring our planets of belief into closer
06
alignment.
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35S
36N
122
Big Picture - UK final proofs.indd 122
20/07/2016 10:02:42
01
02
15
03
04
Accepting Uncertainty
05
06
07
08
09
10
11
12
l
13
et’s say you want to take a scientist down a peg, make her a bit flus-
14
tered. Here’s an easy way to do it: whenever she says that something
15
is true, in her considered opinion as a scientist, just ask, “Can you
16
really prove that?” If your adversary is a good scientist, but not trained in
17
public relations, chances are very high that she will hem and haw, finding it
18
difficult to give a straight answer. Science never proves anything.
19
A lot depends on our definition of “proof.” Scientists will often have in
20
their minds the kind of proof we have access to in mathematics or logic: a
21
rigorous demonstration of the truth of a proposition, starting with some
22
explicitly stated axioms. This differs in important ways from how we might
The Big Picture Page 21