4.

Ethical Anti-rationalism
Hume claims that moral distinctions are not derived from reason but rather from
sentiment. His rejection of ethical rationalism is at least two-fold. Moral rationalists tend
to say, first, that moral properties are discovered by reason, and also that what is morally
good is in accord with reason (even that goodness consists in reasonableness) and what is
morally evil is unreasonable. Hume rejects both theses. Some of his arguments are
directed to one and some to the other thesis, but ambiguities in the text make it unclear
which he means to attack in certain places.

In the Treatise he argues against the epistemic thesis (that we discover good and evil by
reasoning) by showing that neither demonstrative nor probable/causal reasoning has vice
and virtue as its proper objects. Demonstrative reasoning discovers relations of ideas, and
vice and virtue are not identical with any of the four philosophical relations (resemblance,
contrariety, degrees in quality, or proportions in quantity and number) whose presence
can be demonstrated. Nor could they be identical with any other abstract relation; for
such relations can also obtain between items such as trees that are incapable of moral
good or evil. Furthermore, were moral vice and virtue discerned by demonstrative
reasoning, such reasoning would have to reveal their inherent power to produce motives
in all who discern them; but no causal connections can be discovered a priori. Causal
reasoning, by contrast, does infer matters of fact pertaining to actions, in particular their
causes and effects; but the vice of an action (its wickedness) is not found in its causes or
effects, but is only apparent when we consult the sentiments of the observer. Therefore
moral good and evil are not discovered by reason alone.

Hume also attempts in the Treatise to establish the other anti-rationalist thesis, that virtue
is not the same as reasonableness and vice is not contrary to reason. He gives two
arguments to this end. The first he says follows directly from the Representation
Argument, whose conclusion was that passions, volitions, and actions can be neither
reasonable nor unreasonable. This direct argument is quite short. Actions, he observes,
can be laudable or blamable. Since actions cannot be reasonable or against reason, it
follows that “[l]audable and blameable are not the same with reasonable or unreasonable”
(T 458). The properties are not identical.

The second and more famous argument makes use of the conclusion defended earlier that
reason alone cannot move us to act. As we have seen, reason alone “can never
immediately prevent or produce any action by contradicting or approving of it” (T 458).
Morality — this argument goes on — influences our passions and actions: we are often
impelled to or deterred from action by our opinions of obligation or injustice. Therefore
morals cannot be derived from reason alone. This argument is first introduced as showing
it impossible “from reason alone... to distinguish betwixt moral good and evil” (T 457) —
that is, it is billed as establishing the epistemic thesis. But Hume also says that, like the
little direct argument above, it proves that “actions do not derive their merit from a
conformity to reason, nor their blame from a contrariety to it” (T458): it is not the
reasonableness of an action that makes it good, or its unreasonableness that makes it evil.
This argument about motives concludes that moral judgments or evaluations are not the
products of reason alone. From this many draw the sweeping conclusion that for Hume
moral evaluations are not beliefs or opinions of any kind, but lack all cognitive content.
That is, they take the argument to show that Hume holds a non-propositional view of
moral evaluations — and indeed, given his sentimentalism, that he is an emotivist: one
who holds that moral judgments are meaningless ventings of emotion that can be neither
true nor false. Such a reading should be met with caution, however. For Hume, to say that
something is not a product of reason alone is not equivalent to saying it is not a truth-
evaluable judgment or belief. Hume does not consider all our (propositional) beliefs and
opinions to be products of reason; some arise directly from sense perception, for
example, and some from sympathy. Also, perhaps there are (propositional) beliefs we
acquire via probable reasoning but not by such reasoning alone. One possible example is
the belief that some object is a cause of pleasure, a belief that depends upon prior
impressions as well as probable reasoning.

Another concern about the famous argument about motives is how it could be sound. In
order for it to yield its conclusion, it seems that its premise that morality (or a moral
judgment) influences the will must be construed to say that moral evaluations alone move
us to action, without the help of some (further) passion. This is a controversial claim and
not one of which Hume offers any defense. The premise that reason alone cannot
influence action is also difficult to interpret. It would seem, given his prior arguments for
this claim (e.g. that the mere discovery of a causal relation does not produce an impulse
to act), that Hume means by it not only that the faculty of reason or the activity of
reasoning alone cannot move us, but also that the conclusions of such activity alone (such
as recognition of a relation of ideas or belief in a causal connection) cannot produce a
motive. Yet it is hard to see how Hume, given his theory of causation, can argue that no
mental item of a certain type (such as a causal belief) can possibly cause motivating
passion or action. Such a claim could not be supported a priori. And in Treatise 1.3.10,
“Of the influence of belief,” he seems to assert very plainly that some causal beliefs do
cause motivating passions, specifically beliefs about pleasure and pain in prospect. It is
possible that Hume only means to say, in the premise that reason alone cannot influence
action, that reasoning processes cannot generate actions as their logical conclusions; but
that would introduce an equivocation, since he surely does not mean to say, in the other
premise, that moral evaluations generate actions as their logical conclusions. The
transition from premises to conclusion also seems to rely on a principle of transitivity (If
A alone cannot produce X and B produces X, then A alone cannot produce B), which is
doubtful but receives no defense.

Commentators have proposed various interpretations to avoid these difficulties. One

approach is to construe ‘reason’ as the name of a process or activity, the comparing of
ideas (reasoning), and to construe ‘morals’ as Hume uses it in this argument to mean the
activity of moral discrimination (making a moral distinction). If we understand the terms
this way, the argument can be read not as showing that the faculty of reason (or the
beliefs it generates) cannot cause us to make moral judgments, but rather as showing that
the reasoning process (comparing ideas) is distinct from the process of moral
discrimination. This interpretation does not rely on an assumption about the transitivity of
causation and is consistent with Hume's theory of causation.