DebateLabKIT/deepa2-corpus · Datasets at Hugging Face

text stringlengths 0 1.71k
A miracle is a violation of the laws of nature [DEF - m]; and as a fi rm and
unalterable experience has established these laws [DEF - l], the proof against
a miracle, from the very nature of the fact, is as entire as any argument from
experience can possibly be imagined. (Hume, 76)
It is scarcely controversial, although it has not remained unchallenged
(see Swinburne), that Hume ’ s argument is to be read as addressing the
question about whether we could be justifi ed in believing in a miracle on
the basis of empirical evidence; in fact, he leaves to a different part of the
Inquiry a discussion of whether we could have knowledge of God (and so,
on its basis, indirectly of miracles) other than by revelation. It is a bit more
controversial, though, whether Hume is trying to establish the conclusion
that we could be justifi ed in believing in a miracle on the basis of no empirical
evidence whatsoever – and so, in particular, not even if we had an
experience of a miracle – or the considerably weaker conclusion that we
could not be justifi ed in believing in a miracle on the basis of the testimony
of others. In what follows, the argument will be presented as it is more
customarily discussed, that is, as aimed to establish the weaker conclusion.
It is also unclear whether Hume ’ s argument is meant to be a priori or a
posteriori . This difference is worth taking into consideration for the following
reason: while many philosophers think that a priori justifi cation is
indefeasible by experience, it is quite uncontroversial that a posteriori justifi
cation can be so defeated; since the conclusion of an argument cannot
be justifi ed more fi rmly than its premises, this implies that Hume ’ s argument,
depending on whether we read it as a priori or a posteriori , is aimed
to establish a conclusion that cannot – respectively can – be overturned by
further experiences. A possible irenic solution that will be adopted here is
to divide Hume ’ s argument in two parts (corresponding to the parts into
which Chapter X is divided), 1 and to read the fi rst of them as aimed to
assess the question whether testimony (possibly) justifi es belief in a miracle
(at least in part) in light of a priori considerations, and to read the second
1 Buckle explains that both parts refl ect to some degree a division of probability arguments
which was common in Hume ’ s time: that among “ internal ” and “ external ” evidences. The
internal part of any such argument examines the internal credibility of a claim, the external
part examines this claim in light of the evidence available.
46 Tommaso Piazza
as aimed to assess the very same question in light of additional considerations
of a clearly a posteriori nature. As we will see, the (sub)conclusion of
the fi rst part is logically weaker than the conclusion of the second. By the
same token, the conclusion of the fi rst argument has appeared more resistible
than the conclusion of the second. (For this reason, Fogelin contends
that Hume just wanted to offer the second argument and not the fi rst.) The
fi rst argument purports to show that no testimony could ever provide evidence
that is strong enough to undermine our expectation that events that
have always been experienced in conjunction with one another will not (and
did not) appear in conjunction with one another; the second argument aims
just to show that no testimony has provided such evidence. Finally, it is
worth mentioning that some controversy surrounds the scope of the conclusion
that (the fi rst part of) Hume ’ s argument, if successful, would actually
establish. In particular, it has been suggested that the argument quickly (and
undesirably) generalizes to the conclusion that it is always irrational to
accept testimony of something which is very unlikely in light of past experience
(such as water ’ s becoming ice, if you live in a very warm region).
Hume ’ s insistence on the distinction among extraordinary (yet natural)
events and miracles is often invoked to resist this objection (Levine): given
its supernatural origin, a miracle cannot be judged to be analogous to any
other event in experience; on the contrary, an extraordinary event can bear
the relevant similarity to past experience and then become acceptable by
analogy. Yet, it is controversial whether this strategy – since it implies that
Hume ’ s argument should feature as a premise the claim that we could not
have had an experience of a miraculous event – creates more problems for
Hume ’ s argument than it helps to solve.
P1. That A is the case provides evidence E for B ’ s being the case if and only
if the number of times in which we have observed A, and then have
observed B, is greater than the number of times in which we have
observed that A was not followed by B.
P2. The strength of E is proportional to the ratio among the number of
times in which we have observed A, and then have observed B, and the
number of times in which we have observed that A was not followed by
B.
P3. The existence of testimony of type K (henceforth K - testimony) to the
effect that B was not the case provides evidence ET for the proposition
that B was not the case only if the number of times in which we have
received K - testimony for a proposition, and then have observed that the
proposition was true, is greater than the number of times in which we
have received K - testimony for a proposition and then have observed that
the proposition was not true (instantiation, P1).
Hume’s Argument against Miracles 47
P4. The strength of ET is proportional to the ratio among the number of
times in which we have received K - testimony for a proposition and then
have observed that the proposition was true and the number of times in
which we have received K - testimony for a proposition and then have
observed that the proposition was not true (instantiation, P2).
P5. Whenever one ’ s global evidence is constituted by E1 and E2, and E1 is
evidence in favor of a given p, and E2 is evidence for non - p, it is rational
to believe that p only if E1 is stronger than E2, it is rational to disbelieve
that p only if E2 is stronger than E1, and it is rational to withhold belief
as to whether p only if the strength of E1 is the same as the strength of
E2.
P6. There is K - testimony to the occurrence of a miracle M.
P7. There is a law of nature L – say that every A is followed by B – and
the K - testimony is to the effect that A was not followed by B
(DEF – m).
P8. It is rational to accept that M occurred (assumption for reductio ).
P9. That A was the case provides evidence E * for the proposition that B
was the case, which is weaker than the evidence ET, provided by the
K - testimony, for the proposition that B was not the case.
P10. If it is rational to accept that M occurred, then that A was the case
provides evidence E * for the proposition that B was the case, which is
weaker than the evidence ET, provided by the K - testimony, for the
proposition that B was not the case (instantiation, P8, P9).

A miracle is a violation of the laws of nature [DEF - m]; and as a fi rm and

unalterable experience has established these laws [DEF - l], the proof against

a miracle, from the very nature of the fact, is as entire as any argument from

experience can possibly be imagined. (Hume, 76)

It is scarcely controversial, although it has not remained unchallenged

(see Swinburne), that Hume ’ s argument is to be read as addressing the

question about whether we could be justifi ed in believing in a miracle on

the basis of empirical evidence; in fact, he leaves to a different part of the

Inquiry a discussion of whether we could have knowledge of God (and so,

on its basis, indirectly of miracles) other than by revelation. It is a bit more

controversial, though, whether Hume is trying to establish the conclusion

that we could be justifi ed in believing in a miracle on the basis of no empirical

evidence whatsoever – and so, in particular, not even if we had an

experience of a miracle – or the considerably weaker conclusion that we

could not be justifi ed in believing in a miracle on the basis of the testimony

of others. In what follows, the argument will be presented as it is more

customarily discussed, that is, as aimed to establish the weaker conclusion.

It is also unclear whether Hume ’ s argument is meant to be a priori or a

posteriori . This difference is worth taking into consideration for the following

reason: while many philosophers think that a priori justifi cation is

indefeasible by experience, it is quite uncontroversial that a posteriori justifi

cation can be so defeated; since the conclusion of an argument cannot

be justifi ed more fi rmly than its premises, this implies that Hume ’ s argument,

depending on whether we read it as a priori or a posteriori , is aimed

to establish a conclusion that cannot – respectively can – be overturned by

further experiences. A possible irenic solution that will be adopted here is

to divide Hume ’ s argument in two parts (corresponding to the parts into

which Chapter X is divided), 1 and to read the fi rst of them as aimed to

assess the question whether testimony (possibly) justifi es belief in a miracle

(at least in part) in light of a priori considerations, and to read the second

1 Buckle explains that both parts refl ect to some degree a division of probability arguments

which was common in Hume ’ s time: that among “ internal ” and “ external ” evidences. The

internal part of any such argument examines the internal credibility of a claim, the external

part examines this claim in light of the evidence available.

46 Tommaso Piazza

as aimed to assess the very same question in light of additional considerations

of a clearly a posteriori nature. As we will see, the (sub)conclusion of

the fi rst part is logically weaker than the conclusion of the second. By the

same token, the conclusion of the fi rst argument has appeared more resistible

than the conclusion of the second. (For this reason, Fogelin contends

that Hume just wanted to offer the second argument and not the fi rst.) The

fi rst argument purports to show that no testimony could ever provide evidence

that is strong enough to undermine our expectation that events that

have always been experienced in conjunction with one another will not (and

did not) appear in conjunction with one another; the second argument aims

just to show that no testimony has provided such evidence. Finally, it is

worth mentioning that some controversy surrounds the scope of the conclusion

that (the fi rst part of) Hume ’ s argument, if successful, would actually

establish. In particular, it has been suggested that the argument quickly (and

undesirably) generalizes to the conclusion that it is always irrational to

accept testimony of something which is very unlikely in light of past experience

(such as water ’ s becoming ice, if you live in a very warm region).

Hume ’ s insistence on the distinction among extraordinary (yet natural)

events and miracles is often invoked to resist this objection (Levine): given

its supernatural origin, a miracle cannot be judged to be analogous to any

other event in experience; on the contrary, an extraordinary event can bear

the relevant similarity to past experience and then become acceptable by

analogy. Yet, it is controversial whether this strategy – since it implies that

Hume ’ s argument should feature as a premise the claim that we could not

have had an experience of a miraculous event – creates more problems for

Hume ’ s argument than it helps to solve.

P1. That A is the case provides evidence E for B ’ s being the case if and only

if the number of times in which we have observed A, and then have

observed B, is greater than the number of times in which we have

observed that A was not followed by B.

P2. The strength of E is proportional to the ratio among the number of

times in which we have observed A, and then have observed B, and the

number of times in which we have observed that A was not followed by

B.

P3. The existence of testimony of type K (henceforth K - testimony) to the

effect that B was not the case provides evidence ET for the proposition

that B was not the case only if the number of times in which we have

received K - testimony for a proposition, and then have observed that the

proposition was true, is greater than the number of times in which we

have received K - testimony for a proposition and then have observed that

the proposition was not true (instantiation, P1).

Hume’s Argument against Miracles 47

P4. The strength of ET is proportional to the ratio among the number of

times in which we have received K - testimony for a proposition and then

have observed that the proposition was true and the number of times in

which we have received K - testimony for a proposition and then have

observed that the proposition was not true (instantiation, P2).

P5. Whenever one ’ s global evidence is constituted by E1 and E2, and E1 is

evidence in favor of a given p, and E2 is evidence for non - p, it is rational

to believe that p only if E1 is stronger than E2, it is rational to disbelieve

that p only if E2 is stronger than E1, and it is rational to withhold belief

as to whether p only if the strength of E1 is the same as the strength of

E2.

P6. There is K - testimony to the occurrence of a miracle M.

P7. There is a law of nature L – say that every A is followed by B – and

the K - testimony is to the effect that A was not followed by B

(DEF – m).

P8. It is rational to accept that M occurred (assumption for reductio ).

P9. That A was the case provides evidence E * for the proposition that B

was the case, which is weaker than the evidence ET, provided by the

K - testimony, for the proposition that B was not the case.

P10. If it is rational to accept that M occurred, then that A was the case

provides evidence E * for the proposition that B was the case, which is

weaker than the evidence ET, provided by the K - testimony, for the

proposition that B was not the case (instantiation, P8, P9).