24
25
Syllogisms are not the only kind of logical argument— they’re just a par-
26
ticularly simple form that will suffice to make our point.
27
The first two statements in a syllogism are the premises of the argument,
28
while the third statement is the conclusion. An argument is said to be valid
29
if the conclusion follows logically from the premises. In contrast, an argu-
30
ment is said to be sound if the conclusion follows from the premises and the
31
premises themselves are true— a much higher standard to achieve.
32
Consider: “Pineapples are reptiles. All reptiles eat cheese. Therefore,
33
pineapples eat cheese.” Any logician will explain to you that this is a com-
34
pletely valid argument. But it’s not very sound. An argument can be valid,
35S
and even interesting, without telling us much that is true about the real
36N
world.
396
Big Picture - UK final proofs.indd 396
20/07/2016 10:02:56
W h At I S A n d W h A t Ou g h t t O b E
If we were to try to put a derivation of ought from is into the form of a
01
syllogism, it might look something like this:
02
03
1. I would like to eat the last slice of pizza.
04
2. If I don’t move quickly, someone else will eat the last slice of
05
pizza.
06
3. Therefore, I ought to move quickly.
07
08
At a casual glance this seems like a good argument, but it’s not a logically
09
valid syllogism. The two premises are both “is” statements— my desire to
10
eat the last slice, and the likelihood that I will miss that chance if I don’t
11
move quickly, are both factual claims about the world, whether or not they
12
are actually true. And the conclusion is undeniably an “ought” statement.
13
But if you look past the everyday meaning of the sentences to their underly-
14
ing logical content, something is missing. Premises 1 and 2 don’t actually
15
imply the conclusion 3; what they imply is “Therefore, if I don’t move
16
quickly, I will not get what I like.”
17
To make the conclusion follow validly, we would need to add another
18
premise, along the lines of:
19
20
2a. I ought to act in such a way as to bring about what I
21
would like.
22
23
With this addition, the argument becomes valid. It’s also no longer a can-
24
didate for deriving ought from is— an “ought” statement appears right
25
there in the new premise. All we’ve done is to derive an ought from an
26
ought plus a few is’s, which isn’t nearly as impressive.
27
That’s the problem with attempting to derive ought from is: it’s logically
28
impossible. If someone tells you they have derived ought from is, it’s like
29
someone telling you that they’ve added together two even numbers and
30
obtained an odd number. You don’t have to check their math to know that
31
they’ve made a mistake.
32
33
•
34
And yet, it happens all the time. Over and over again, before and after the
S35
appearance of Hume’s famous passage, many people have triumphantly
N36
397
Big Picture - UK final proofs.indd 397
20/07/2016 10:02:56
T H E B IG PIC T U R E
01
declared they have finally cracked the code and shown how to derive ought
02
from is. Smart, knowledgeable people, with interesting things to say. But
03
somehow they have all gone wrong.
04
Physicist Richard Feynman liked to tell the story of meeting a painter
05
and asking him about his craft. The painter boasted that he could mix red
06
and white paint together and get yellow. Feynman knew enough about how
07
color works to be skeptical, so the painter fetched some paint and com-
08
menced with mixing. After a bit of effort and nothing but pink paint to
09
show for it, the painter mumbled that he should probably add a touch of
10
yellow to the mix, to “sharpen it up a bit.” At that point Feynman under-
11
stood the trick— to get yellow out, you put a bit of yellow in.
12
The painter’s gambit is the same basic move that has been used to per-
13
form the logically impossible, deriving ought from is, many times over the
14
centuries. One presents a set of incontrovertible “is” statements, then sneaks
15
in an implied “ought” statement that seems so extremely reasonable that
16
nobody could possibly deny it. Sadly, all statements about what ought to
17
happen can (and will) be denied by somebody, and even if not, that doesn’t
18
prevent them from being ought statements.
19
A classic example was offered up by John Searle, of Chinese Room fame.
20
Here is Searle’s version of the kind of deductive argument we examined
21
above:
22
23
1. Jones uttered the words “I hereby promise to pay you, Smith,
24
five dollars.”
25
2. Jones promised to pay Smith five dollars.
26
3. Jones placed himself under (undertook) an obligation to pay
27
Smith five dollars.
28
4. Jones is under an obligation to pay Smith five dollars.
29
5. Jones ought to pay Smith five dollars.
30
31
You see the magical appearance of “ought” in the last line, even
32
though all of the other lines were about “is.” Where did the sleight-of-hand
33
occur?
34
It’s not that hard to find. Just as we had to imagine a new premise 2a
35S
above, Searle is relying on a hidden premise between 4 and 5:
36N
398
Big Picture - UK final proofs.indd 398
20/07/2016 10:02:56
W h At I S A n d W h A t Ou g h t t O b E
4a. All else being equal, one ought to do what one is under an
01
obligation to do.
02
03
Searle actually admits the need for a premise like this, right in the text of
04
his paper. But he thinks it doesn’t count as a premise, since it’s a “tautology”—
05
> something that is automatically true by the definitions of the terms involved.
06
Searle is claiming that what it means to say, “Jones made a promise to do
07
something,” is simply “Jones ought to do something” (all else being equal).
08
That’s not true. Hopefully the equivocation is clear. Up in premises 1– 3,
09
the idea of “placing himself under an obligation” referred to a certain fact
10
about the world, a sentence that Jones uttered. But now in 4– 5, Searle wants
11
us to treat an “obligation” as a moral command, a statement about what
12
ought to happen. He’s using the same word in two different senses, to trick
13
us into thinking that factual statements about what happens can somehow
14
lead to evaluative conclusions about right and wrong.
15
This example is worth belaboring because it stands in for an impressive
16
number of attempts to derive ought from is over the years. Inevitably, the
17
argument introduces just a tiny bit of prescription into their list of descrip-
18
tions: the painter sharpens things up with a touch of yellow.
19
20
•
21
This inherent flaw in deriving ought from is has been pointed out many
22
times. The list of thinkers who claim to have successfully pulled off the
23
trick is long and distinguished; they aren’t simply making elementary
24
mistakes. Lurking in the back of their minds is usually some kind of justi-
25
fication along the lines of “Okay, there is some hidden premise that intro-
26
duces an ought into my list of is’s, but surely we agree that this particular
27
hidden premise isn’t so bad, right?”
28
It wouldn’t be so bad if it weren’t for the fact that, when brought out
29
into the clear light of day, the hidden evaluative premises don’t seem to be
30
universally true. Quite the contrary; they tend to be conspicuously conten-
31
tious. The reason why deriving ought from is should be thought of as a
32
philosophical felony, rather than a simple misdemeanor, is because these
33
hidden premises deserve our closest scrutiny. They are, more often than not,
34
where most of the action is.
S35
N36
399
Big Picture - UK final proofs.indd 399
20/07/2016 10:02:56
T H E B IG PIC T U R E
01
You might be tempted to think that Searle’s hidden premise 4a seems
02
pretty unobjectionable, but let’s examine it more closely. Surely there are
03
some kinds of obligations that one ought not to carry out— when they were
04
made under duress, or when they would grossly violate some other moral
05
precept. Searle would say that such examples don’t count, because of the “all
06
else being equal” clause. So what exactly does that clause mean? He tells us:
07
08
The force of the expression “other things being equal” in the
09
present instance is roughly this. Unless we have some reason for
10
supposing the obligation is void (step 4) or the agent ought not
11
to keep the promise (step 5), then the obligation holds and he
12
ought to keep the promise.
13
14
So you ought to do what you are under an obligation to do— unless
15
there is some reason you ought not to do it. This doesn’t seem like a useful
16
foundation for moral reasoning.
17
We shouldn’t hide or downplay the assumptions we make in order to get
18
moral reasoning off the ground. Our attempts to be better people are best
19
served if those assumptions are brought out into the open, interrogated,
20
and evaluated as carefully as we can manage.
21
•
22
23
A modern twist on the ought- from-is campaign is to claim that morality
24
can be reduced to, or absorbed by, the practice of science. The idea is some-
25
thing like this:
26
27
1. Condition X would make the world a better place.
28
2. Science can tell us how to achieve condition X.
29
3. Therefore, we ought to do what science tells us to do.
30
31
In this case, the hidden assumption would appear to be:
32
33
2a. We ought to make the world a better place.
34
35S
This might seem like a tautology, depending on your definition of the
36N
word “better.” But whether we put the hidden assumption into a statement
400
Big Picture - UK final proofs.indd 400
20/07/2016 10:02:56
W h At I S A n d W h A t Ou g h t t O b E
such as this one, or bury it in the definition of “better,” we are still making
01
some positive claim that something ought to be done. Such claims cannot
02
be grounded on factual statements alone. Who decides what is “better”?
03
Proponents of this technique will sometimes argue that all we’re doing
04
is making some reasonable assumptions, and science makes reasonable as-
05
sumptions all the time, so what we’re doing really isn’t any different. That’s
06
missing an important aspect of what science is. Consider the following
07
statements:
08
09
• The universe is expanding.
10
• Humans and chimpanzees share a common ancestor.
11
• We should work to allow people to lead happier and
12
longer lives.
13
14
All of these statements are, by some lights, true. But only the first two
15
are “scientific.” The reason is that each of them could have been false. They
16
are not true by definition or assumption. We can imagine possible worlds
17
in which the universe was contracting, or in which there were species like
18
humans and chimpanzees that had not evolved from a common ancestor.
19
We decide whether such statements are true or not by empiricism, abduc-
20
tion, and Bayesian reasoning— we go out and observe the world, and update
21
our credences appropriately.
22
We don’t imagine carrying out experiments to decide whether we should
23
work to allow people to lead happier and longer lives. W
e assume that it’s
24
so, or we try to derive it from a related set of assumptions. That crucial extra
25
ingredient separates how science works from how we think about right and
26
wrong. Science does require assumptions; there are certain epistemological
27
precepts, like our trust in our basic sensory inputs, that play an important
28
role in constructing stable planets of belief for working scientists. But the
29
assumptions that suffice to get science off the ground don’t do the same
30
trick for morality.
31
32
•
33
None of this is to say that we can’t address “ought” issues using the tools of
34
reason and rationality. There is an entire form of logical thought called in-
S35
strumental rationality, devoted to answering questions of the form “Given
N36
401
Big Picture - UK final proofs.indd 401
20/07/2016 10:02:56
T H E B IG PIC T U R E
01
that we want to attain a certain goal, how do we go about doing it?” The
02
trick is deciding what we want our goal to be.
03
One attractive suggestion was put forward by Bill Preston and Ted Lo-
04
gan, as played by Alex Winter and Keanu Reeves in the movie Bill & Ted’s
05
Excellent Adventure. They proposed the timeless moral axiom, “Be excellent
06
to each other.”
07
As foundational precepts for moral theorizing go, you could do worse.
08
It’s tempting to brush aside concerns about the foundation of morality on
09
the grounds that we know moral goodness when we see it, and what’s really
10
important is how we go about achieving it.
11
But there are important reasons why we have to do a little bit better than
12
Bill- and- Ted-level philosophizing. The truth is that we don’t ultimately all
13
agree on what constitutes happiness, or pleasure, or justice, or other forms
14
of being excellent to each other. Morality and meaning are areas where
15
foundational disagreement doesn’t arise just by someone making a mistake;
16
it’s real and inevitable, and we need to figure out how to deal with it.
17
It’s tempting to say, “Everyone agrees that killing puppies is wrong.” Ex-
18
cept that there are people who do kill puppies. So maybe we mean “Every
19
reasonable person agrees . . .” Then we need to define “reasonable,” and real-
20
ize we haven’t really made much progress at all.
21
The lack of an ultimate objective scientific grounding for morality can
22
The Big Picture Page 67