In Treatise of Human Nature, Hume presents the following argument: “Morals excite
passions, and produce or prevent actions. Reason of itself is utterly impotent in this particular.
The rules of morality, therefore, are not conclusions of our reason” (Hume 457). For Hume,
are simple feelings, and REASON is rational thought1. In this paper, I will discuss
Hume’s argument. Then I will argue against Hume’s view.
The first premise of Hume’s argument states that morals give rise to passions, and
produce or prevent actions. It seems that Hume understands MORALS to be a special class of
judgments which oblige people to behave in certain ways. Hume seems to think the “rules of
morality” (457) are judgments. This characterization of morals is prima facie adequate for
explaining Hume’s argument that morals do not derive from reason by itself.
The first premise states that morals give rise to passions, and produce or prevent actions.
In order to see why this is so, consider how morals influence our passions. If a robber were to
steal from a candy shop, the owner of the shop would likely be angry. It seems this is because the
owner of the shop has the moral judgment that stealing is wrong, which gives rise to the passion
of anger. In this way, the shop owner’s moral judgment gives rise to her passion – the shop
owner has the passion because she judges that stealing is wrong. Now, morals produce or
prevent actions in a similar way. If one judges that it is good to help others, then ipso facto, when
she sees a person bleeding profusely on the street, she will probably help that person. In this
case, she helps the person in need because she thinks it is right to help others. The action is done
My characterization of reason is somewhat imprecise – but Hume’s argument focuses on the effects of reason,
not what reason itself is. Accordingly, it seems my characterization of reason is adequate for the purposes of this
paper. The same is true for my characterization of passions.
in virtue of the moral judgment. Conversely, a person who judges that it is wrong to murder will
ipso facto avoid committing murder. Here, the action of murder is avoided because of the moral
It seems that morals are not sufficient for producing or preventing actions. A person who
judges that it is good to help others might not end up helping others. She might even avoid
helping a person bleeding on the street if, for example, there were a gunman between her and the
person in need. While morals are not sufficient for action, it is still true the case that some
actions are enacted (or avoided) because of moral judgments. This is the sense in which morals
produce (or prevent) actions. Now Hume argues that the negation of this premise is absurd.
Morality obliges people to behave in certain ways – but if morals did not produce or prevent
actions, then instilling morals in people would be pointless, and morals themselves would have
no function. But that does not seem right, as morals seem to have a large role in our everyday
affairs. Hume says it is “confirm’d by common experience” that morals produce or prevent
actions, as we can think of examples in which actions are enacted (or avoided) because of
morals. So it seems clear that the first premise is true.
The argument’s second premise is the following: reason by itself cannot give rise to
passions, nor can it produce or prevent actions. Hume thinks there are only two types of
reasoning (413): (1) demonstrative reasoning and (2) probabilistic reasoning. According to
Hume, demonstrative reasoning is concerned with ideas and their relations, but probabilistic
reasoning is concerned with matters of fact. In other words, demonstrative reasoning is
concerned with anything that can be known a priori (namely logical relations, and mathematical
relations), whereas probabilistic reasoning is concerned with anything that can be known a
posteriori (namely the relations of objects, and causal relations). We know through
demonstrative reasoning that all bachelors are unmarried, for example, since the meaning of
‘bachelor’ includes the property of being unmarried. It is known independently of experience
that all bachelors are unmarried. On the other hand, we know through probabilistic reasoning that
heat causes water to evaporate – and this is known only from experience.
Hume thinks that it is impossible for reason by itself to give rise to passions, and produce
or prevent actions. Before I expound his argument, I will explain some central concepts which
the argument draws upon. An IMPRESSION, for Hume, is a mental state acquired directly through
experience (1). This is different from an IDEA, which is a less forceful or immediately
experienced copy of an impression (1). The difference between impressions and ideas is one of
degree, not of kind (1), since both are PERCEPTIONS, which are mental states (456). Now THE
the impression we are conscious of when we purposefully act (399).
Hume argues that since demonstrative reasoning is concerned with ideas, but the will is
concerned with objects in reality, it is impossible for demonstrative reasoning by itself to
produce or prevent actions. Demonstrative reasoning either have to be concerned with objects in
reality in order to produce action, or be able to produce the will, which produces action. Since
neither are possible, Hume thinks, demonstrative reasoning and the will are distinct, and
demonstrative reasoning by itself cannot produce actions. Still, he acknowledges that
demonstrative reasoning can be useful in practical affairs. A cashier may use arithmetic in giving
a customer her change, but demonstrative reasoning does not by itself produce the action here.
Rather, the cashier uses arithmetic as a means to bring about an effect. Demonstrative reasoning
by itself cannot bring about such effects.
Probabilistic reasoning by itself cannot produce actions either, Hume thinks. Here is why:
we have emotions toward objects which can cause us either pleasure or pain, and these emotions
make us “avoid or embrace” (414) what will give us either pain or pleasure. When we are faced
with the possibility of pleasure or pain, our emotions make us try to comprehend relevant causal
relations. Consider a case where a traveler is walking along train tracks, and sees a train
approaching. At this point, she will feel an emotion which will strive to make her act. We might
expect her to have a reaction of horror, and in order to escape death, the emotion will push her to
try to avoid the oncoming train. In order to do this, the traveler engages in probabilistic reasoning
and discovers that if she does not move out of the way, she will be killed. Consequently, she
moves off the train tracks. In this scenario, Hume would say that probabilistic reasoning just
guided the traveler to act, as the impulse to act occurred independently from probabilistic
reasoning. Since the impulse to act did not come from probabilistic reasoning, probabilistic
reasoning by itself cannot produce actions, nor can it give rise to the will. Since neither types of
reasoning can produce actions, or give rise to the will, reason by itself cannot produce actions, or
give rise to the will. Thus reason by itself can neither produce not prevent actions. Ostensibly,
Hume thinks that passions are necessary for actions. If this is so, then reason by itself cannot
give rise to passions either.
Since moral judgments cannot derive from something which is unable to produce or
prevent actions, and unable to give rise to passions, Hume concludes that morals are not derived
from reason by itself. To be more specific, judgments like ‘it is wrong to steal’ cannot derive
from judgments of demonstrative reason, which is concerned with ideas, nor can they derive
from probabilistic reason, which is concerned with matters of fact. This means that when we are
making moral judgments, we are not making judgments of reason. Judgments of reason might
play a role in moral judgments, but since reason cannot produce or prevent actions, moral
judgments cannot be reduced to judgments of reason.
I disagree with Hume. It seems that probabilistic reasoning can be said to produce actions
in the same sense that morals produce actions. Remember that moral judgments produce actions
in this way: a person does an action because of a moral judgment, such that if that person did not
have that moral judgment, she would not have enacted that action. Now recall the example of the
traveler walking on the train tracks. We can say her action – moving off the train tracks, out of
the way of the oncoming train – is done because of a judgment by reason concerning matters of
fact. It seems that her action would not have been done if she had not made the judgment she did.
Probabilistic reasoning can also prevent actions in this way. Thus, the sense in which
probabilistic reasoning produces or prevents actions seems to be analogous with the sense in
which morals produce actions. If this is so, then it does not matter that reason by itself cannot
produce action, since that just means that reason by itself is not sufficient for action. Because
moral judgments are not sufficient for action, this is not problematic.
It seems, therefore, that the second premise is false. Reason by itself can produce actions
by providing reasons to act. In the case of the traveler, probabilistic reasoning supplied a reason
for her to move out of the way. Because reason by itself can produce actions in the same sense
that morals produce actions, morals might be derivable from reason by itself. This undermines
Hume’s conclusion. In order to avoid this worry, Hume might want to instead hold that moral
judgments are sufficient for action. Since judgments concerning matters of fact cannot be
sufficient for action, the second premise of his argument would then be true. But this does not
seem defensible, since as I argued earlier, one might hold a moral judgment, but be unable to act
accordingly. Since moral judgments cannot be sufficient for action, it seems that moral
judgments can only produce actions in a weaker sense – but this is clearly not problematic. It is
not clear how Hume can save his argument from these objections.
Hume, David. A Treatise of Human Nature. 2nd ed., Oxford University Press, 1978.
Purchase answer to see full