49

Hume and the Problem

of Induction

Editors? note: We have included

ing induction in order tohighlig

Hume's Problem o f Induction

James E, Taylor e e

_ Hume, David. An Enquiry concerning Hunan

1.

Hackers, 1993, NB.

based on either the present
)
(e.g., ? A l l bachelors are
and existence,? w h i c h are not certain
ue (e.g., ?The sun w i l l rise tomorrow?),
tect;

about matters o f fact and existence are

OUr Memories of what
»

Or reasoning about relations

?nses and memory beliefs, For
nt place on the grounds
that i t was sent by our
er a currently unobserved cause from 4
cause, also infer more generally from
cause-and-effect relationships at
we have WeObserved in the past that similar
effects in the future, Thus, according to Hume,
tters o f fact and existence is
of inferences from experience
 H u m e and the Problem o f Induction 175

are not based on any further reasoning. The way this claim is typicall
today is that there is no rational justification for inductive inferences i this
claim is true, then no one can be rationally justified in believing anythin
that goes beyond what one is currently observing, and if that is the case,
then there is no rational justification for any theory of empirical science.
Since Hume was the first philosopher to make this claim and argue fori t ,
the problem facing philosophers who deny it is called ?Hume?s Problem of
Induction.? Though many attempts have been made to solve this problem,
none of these attempts is widely believed to be successful. Consequently,
Hume?s problem of induction continues to be a central topic of philosophi-
cal conversation.

A l l reasonings may be divided into two kinds, namely, demonstrative

reasoning, or that concerning relations of ideas, and moral reasoning, or that
concerning matter of fact and existence. That there are no demonstrative
arguments in the case seems evident; since it implies no contradiction that the
course of nature may change, and that an object, seemingly like those which
we have experienced, may be attended with different or contrary effects. May
I not clearly and distinctly conceive that a body, falling from the clouds, and
which, in all other respects, resembles snow, has yet the taste of salt or feeling
of fire? Is there any more intelligible proposition than to affirm, that all the
trees w i l l flourish in December and January, and decay in May and June?
N o w whatever is intelligible, and can be distinctly conceived, implies no
contradiction, and can never be proved false by any demonstrative argument
or abstract reasoning @ priori.
If we be, therefore, engaged by arguments to put trust in past experience,
and make it the standard of our future judgement, these arguments must be
probable only, or such as regard matter of fact and real existence, according
to the division above mentioned. But that there is no argument of this kind,
must appear, if our explication of that species of reasoning be admitted as
solid and satisfactory. We have said that all arguments concerning existence
are founded on the relation of cause and effect; that our knowledge o f that
relation is derived entirely from experience; and that all our experimental
conclusions proceed upon the supposition that thefuture w i l l be conformable
to the past. To endeavour, therefore, the proof o f this last supposition NY
probable arguments, or arguments regarding existence, m u s t b e eviently
going i n a circle, and taking that for granted, is
which
very point i n q
the

tion. (IV.ii)

an inductive inference employed by Hume (which can

An example of
represent all inductive inferences) is the inference from (a) ?All the bread
I have eaten has nourished me? to (b) ?The bread I am about t o eat
ill ?sh me.? I w i l l refer to this example i n my reconstruction o f
Hume'sargument for his claim that no such inferences have a rational
foundation.
176 James E. Taylor and Stefanie Rocknak

P1. If the (inductive) inference from (a) to (b) has a rationalf o u n d a t i o n

then it must be based on intuition, reasoning that is based
o n INtUition
(?demonstrative? or deductive reasoning) or reasoning t h a t
direct observation (?experimental? orinductive reasoning).
P2. The (inductive) inference from (a) to (b) is not based on intui t i o n , req.
soning that is based on intuition, or reasoning that is based ON d i r e c
observation.
C1. The (inductive) inference from (a) to (b) does not have a rational
foundation (modus tollens, P1, P2).

Argument for P2:

P3. The connection between (a) and (b) o f the example inference jg not
intuitive (i.e., it isn?t self-evident that i f (a) is true, then (b) is true),
P4. The inference from (a) to (b) is not based on demonstrative reaso
ning
(since demonstrative? reasoning can only establish claims that are not
possibly false and the claim that i f (a) is t r u e , t h e n (b) is t r u e is possibly
false).
PS. The inference from (a) to (b)
is not based on experimental reasoning
(because all experimental reaso
ning presupposes that similar causes have
similar effects and the inference in question is an instance of this very
Presupposition, so an experimental, (inductive) argument for that infer
ence would be circular).
oo

C2. P2 is true: the (inductive) inference f r o m (a) to (b) is not based on

intuition, reasoning that ig based on
di b :
intuition, or reasoning that is
based on direct observation (conjunction, P3, P4, P5; De Morgan's)

Hume?s Negative Argument concerning Induction

Stefanie Rocknak :
fe p
omens,
O TwokRea ye pMEM
a y ete e u
TF

Hume, D a v i d . A Treatise o f Human Nat Meee ee ? oo

Oxford: Oxford University Py ss, 2003edited by D . Band M. J. Norton.

Arnold, N. Scott. ?Hume?s Skepticism 4

of the History of Philosophy 21,1 ( o g gltductve Inferences.? Journal
Baier, Annette. A Progress of Sentiments. 31~$5 +

Press, 1991. |
Cambridge, MA: Harvard UniversitY
Beauchamp, Tom, and Alexander oo
Benhero

py)
Causation. Oxford: Oxford Univers: 8. Hur,6 3 nd the Problem of
Broughton, J. ?Hume's Skepticism, a b e 1991, e Pro
Philosophical Quarterly 64 (1983), 319 Causal Inferences » pacific
 H u m e and the Problem o f Induction 177
W y e wMA Si e E C O E Ft e
ee
ee woe

a cgay meena B e ea

? casrett, Don. Cognition and Commitment is » Phi

oxford University Press, 1997, . ume? Philosophy. Oxford: .

? amerwahr, John. ?The Failure of Hume? a ) :

+ BN A977): 57-71. © Treatise.? Hume Studies 3, 2

» speb, L. E. Stability and Justification in Hume's Treat | |

University Press, 2002. Tume?s Treatise. Oxford: Oxford

- Owen, David. Hume?s Reason.

; Oxford:: Oxford Universi Press, 1999
smore, John. Hume's Intentions. Cambrid : bride Unive

* Pas
Press, 1952/1968. ? - & V R Cambridge University
-
Schmitt, F E. Knowledge and Belief.
. London: Routled ge, 1992.
Smith, Norman K. The Philosophy o f David Hume: A Critical Study o f its
: Origin?a nProbability
Stove, D. d Cenral
C. Doctrines.
and Hume?sNInductive
e w York: Skeptici
Macmillan, 1941 d:
+ v e n a t y Press, 1973. i e Skepticism.O x f o r d : Oxford
.

Strawson, P. E, Introduction to Logical Theory. London: Methuen, 1952.

Where does the necessity that seems to accompany causal inferences come
from? ?Why [do] we conclude that [...] particular causes must necessarily
have such particular effects?? (Hume, 1.3.2.15) In 1.3.6 of the Treatise,
Hume entertains the possibility that this necessity is a function of reason.
However, he eventually dismisses this possibility, where this dismissal con-
sistsof Hume?s ?negative? argument concerning induction. This argument
has received, and continues to receive, a tremendous amount of attention.
How could causal inferences be justified if they are not justified by reason?
If we believe that p causes q, isn?t it reason that allows us to conclude with
some assurance ? that is, w i t h some necessity ? that g whenever we see p?
The responses to these questions are many, but they may be parsed into
four groups. (1) Some argue that Hume?s negative argument shows that he
thought that inductive inferences are worthless. Hume was actually a closet
?deductivist,? where he meant to show that any method that does not rely
on a priori principles is useless (€.g-, Stove). (2) Others have alleged that
Hume?s negative argument only meant to show that we cannot use demon-
Strative reason to justify inductive inferences, but we can, apparently, justify
them with prob able reason (e.B+s Beauchamp and Rosenberg; Arnold,
Broughton, and Baier). (3) Still others argue that Hume?s notiono f justifica-
tion (in regard to beliefs in general,including beliefs in causal inferences)
should be understood in two stages inBook I of the Treatise. In the first,
Hume does lay out 2 theory of justification. In the second (particularly in
Immerwahr, Schmitt, and Loeb). (4)
1.4.
4.7), he retracts irit ( e w hPassmore,
(E-Bes
o claim that no ?justification? is needed for
Finally, there are t h ° . laced demand for
Causali n f e r e n c es. I n fact, a s k i n g f o r i t amounts t o a misp
178 James E. Taylor and Stefanie Rocknak

epistemic explanation; to some degree, this is what the Negative ar me

shows us. What we must do instead is give a descriptive PSYchologie
explanation where this explanation consists o f Hume S "POsitive? accouy
of induction; see, for instance, Treatise 1.3.14, ?of the idea ofnecese, t
connexion? (e.g., Strawson, Garrett, and Owen). ty

In the arguments that rule out demonstrative and Probable teasonin

Hume assumes that the principle o f uniformity is justified by, respective,
demonstrative reason and probable reason, and then he Tespectively show,
why these assumptions are incorrect. In the concluding argument, he
shows that this means that the principle o f uniformity is not justified by
reason, nor is the necessity that obtains o f our causal inferences a function
of reason. To do so, he draws on the premises established in his introduc.
tion and the conclusions established in the arguments that rule out demon-
strative and probable reason.

[We must now] discover the nature of that necessary connexion, which
makes so essential a part of [the relation of cause and effect] [. . .]. Since it
appears, that the transition from an impression present to the memory or
senses to the idea of an object, we call cause and effect, is founded on past
experience, and our resemblance of their constant conjunction, the next ques-
tion is, whether experience produces the idea by means o f the understanding
or of the imagination; whether .we are determin?d by reason to make the
transition, or by a certain association of perceptions. I f reason determin?d us,
it wou?d proceed upon that Principle, that instances, of which we have had
no experience must resemble those, of which we have had experience, and
that the course of nature continues always uniformly the same. In order
therefore to clear up this matter, let us consider all the arguments, upon which
such a proposition may suppos?d to be founded? and as these must be deriv?d
cither rom imowledge o r Probability, let us cast our eye on each of these
cereesOF
evidence, and see whether they afford o f this
ion

nature, (Hume, 1.3.6.3, 1.3.6.4) ¥ attord

any just conclusion0

P1. When the mind makes what appears to be a necessary transition from
a present impression, or a m e m o r y o fa n j ? .
. an impress ea, we
call that transition ?cause and effect.? T h Pression, t o a given idea,
i iti € question is, on what is this
seemingly necessary transition founded? O f what is it a function: undet-
standing (i.e., reason) or theimagination?
P c a t e n i n Goes Sctermine us to make these causal transitions, then this
associations of any two v e the Principle that instances (e.g., particular
1y Objects) that occurred in the past will continue
to occur as such in the future (the Principle of uniform; . "
P3. If causal necessity is a function of feason, where that based
reason
js

on the principle ofuniformity, then t reason is

inci co
ba

in some fashion or other, be F he principle of uniformity musts

 Hume and the Problem o f Induction 179

reason. In symbolic form, this reads (N > P) > J, where ?N? stan ds fo
causal necessity is a function of reason, ?P? stands for a principle of
uniformity, and ?J? stands for ?the principle of uniformity is justified by
reason.
p4, There are only t w o kinds of reason that may justify a principle, includ-
ing the principle of uniformity: (a) ?knowledge? (demonstrative reason-
ing) or (b) ?probable? reasoning.
ps, Assume that the principle of uniformity is justified by demonstrative
reasoning.
p6. If the principle of uniformity is justified by demonstrative reasoning ~ in
other words, it is an instance of demonstrative reasoning ? then the
principle of uniformity cannot be imagined otherwise.
P7, We can imagine that nature will not continue uniformly in the future,
while simultaneously imagining that nature has always continued the
same in the past, without contradicting ourselves.
C1, The principle o f u n i f o r m i t y is not proved; that is, justified by demon-
strative reasoning (modus tollens, P6, P7).
P8. The principle of uniformity is justified by probable reasoning (assump-
tion for reductio).
P9, Probable reasoning is actually causal reasoning, since both are cases
where we are automatically led to think of an idea in virtue of experienc-
ing an impression or remembering an impression.
P10. If the reasoning at hand is an instance of causal reasoning, then such
reasoning is justified by the principle of uniformity.
C2. Probable reasoning is justified by the principle of uniformity (modus
ponens, P9, P10).
is justified by probable reasoning (i.e.,
C3. The principle of uniformity
causal reasoning) and justifies probable reasoning (i.e., causal reason-
ing) (conjunction, P7, C2).
s not justified by probable reason
C4. The principle of uniformity }

o s a u t o n r e e , uniformity is notjustified by either demonstrative

ing (conjunctio ? ?
.

PIL. or p r o b a b l e a r a n f o r m i y is n o t justified by either demonsa

or probable reasoning, then we must reject the
claim that the
princip
i ?ty

justified by reason.
is
_ . a
C o w em u s t r e j e c t the c l a i m t h a t the principle o f u n i f o r m i t y 1s justified
onens, C5; Pit). . woe
P12 py reason ( m o a t ne claim that the principle o f uniformity is justified
br we m n w e must reject the claim that the necessity that seems to
oy reason, a n t relations is a functiono f reason.
ccompany caus t the claim that the necessity that seems to accompany
C7. Wei o n ons is a function o f reason (modus ponens, P12, C6).
causal rela