In the essay “Of Miracles”, Hume argues that we should not believe in miracles, ‘a violation of

the laws of nature’ (58), on the basis of testimony, accounts of events by someone other than
oneself. Given that we adopt Hume’s standards for wise reasoning — ‘a wise man…
proportions his belief to the evidence’ (56) — Hume demonstrates that no testimony provides
‘sufficient [evidence] to establish a miracle’ (59).

Before we discuss Hume’s attempts to invalidate the validity of testimony, it is important to note
that Hume’s argument is not concerned with the possibility of miracles but only in whether we
are rationally justified in believing in miracles, specifically those we do not experience
ourselves; first hand religious experiences are beyond the scope of his thus this essay. As
such, Hume does not deny the possible existence of miracles or faith-based justifications for
miracles but merely objects (wise) rational belief in them from probability.

For Hume, ‘it is experience that gives authority to human testimony and it is the same
experience that assures us of the laws of nature’ (67). Therefore, when evaluating whether we
should believe in miracles, we should compare the reliability of testimony against the reliability
of the laws of nature. Unless it is more likely that the miracle is true than that the testifier is
mistaken, Hume argues that we should not believe in miracles on the basis of testimony.

Hume offers four reasons in attempts to invalidate the reliability of testimony. First, no miracle
is supported by a sufficient number of trustworthy testifiers to eliminate the possibility of
falsehood (59). Hume questions the education, judgment and intentions of the first-hand
testifiers. Second, ‘mankind’s usual liking for the marvellous’ (61) leads to poor judgement and
creates incentives for testifiers to forge miracles. Third, Hume notes that reports of miracles
tend to stem from ‘ignorant and barbarous nations’ where people lack critical thinking skills to
independently evaluate testimony (61). Fourth, even if we take testimonies at their best,
religions provide testimonies of miracles that aim to invalidate other religions (63). Therefore,
even if we only evaluate testimonies amongst themselves and not against the laws of nature, it
is still unclear why we should trust one religion’s miracle over another if any.

With the four reasons, Hume argues that false testimony is not all that unlikely whereas
miracles, a violation of the laws of nature i.e. an exception to a previously exceptionless
regularity, is much more unlikely. Weighing the probabilities of a miracle and of false testimony,
Hume concludes that we should not believe in miracles on the grounds of testimony, which he
believes is the only evidence we have.

While I agree that testimonies offer rather fragile grounds for belief, I wonder whether serious
entertainment of poorly grounded miracles should be encouraged as a means to advance our
understanding of the laws of nature. If first instances of phenomena are prematurely dismissed
as mistaken, we may be missing out on the very first data points in the potential discovery of a
new law or corrections of an existing one. As Hume notes, ‘nothing is counted as a miracle if it
ever happens in the common course of nature’ (58) and the ‘common course of nature’ is
regarded as reliable precisely because it had proven its regularity over the test of time.
Therefore, it would be a great disservice to the potential discovery/correction to discount its
first instance. For example, if the suspected miracle is in fact regular but occurs only under
extremely specific circumstances that only happen every fifty years, to neglect the first
instance would seriously harm its timeline towards proving to be ‘common course’.

Further, Hume seems to have implicit favouritism towards first hand observations over
testimonies and muddles extraordinary events with miraculous events in the discussion of the
Indian prince. Hume argues two main points. First, the Indian prince who has never seen water
freeze into ice is right to refuse testimonies of ice on grounds that they do not conform to his
experience of water (above the freezing point). Second, ice is an extraordinary event not a
miraculous event because it does not violate what the Indian prince expects when
“circumstances are the same” (58). Testimonies of ice do not challenge the Indian prince’s
observations in his own warm climate. Rather, they offer information that is observed beyond
the scope of his climate.

I believe that ice is a perfect example that illustrates how phenomenon can be “regular” in that
it can be consistently reproduced, but also “irregular” in that it does not follow from direct
extrapolation. In response to Hume’s first point, if we accept “consistent reproduction” as a
sufficient condition of the laws of nature, perhaps it is not so wise for the Indian prince to
extrapolate too far outside the scope of his experiences. Also, even if the Indian prince’s first
hand observations are more reliable than testimonies, only being able to observe from one
position at any given time is a huge drawback: first hand observations are limited to one
perspective, data collection is highly inefficient and harms the sample size, there is little room
for self correction as it is impossible to repeat the same moment in time and “re-observe”, etc.
Most importantly, there is nothing that suggests that being a first hand observer makes one
any wiser than the testifiers Hume discredits.

In response to Hume’s second point, I do not think that a violation of the laws of nature
(miracle) requires comparison between the same circumstances. The kind of observations that
Hume refers to are not exclusively controlled scientific experiments (56), which means that
observations, first hand or through testimony, do not necessarily share the same
circumstances anyway. If all anomalies are explained away by dissimilar circumstances, barely
anything can be a miracle. When the law of nature was established, something along the lines
of “water is [always] a fluid”, I do think that the Indian prince assumed it to be true across all
climates. If so, the first testimony of ice would be contrary and should be regarded as a miracle
by the Indian prince, rather than just an extraordinary property of water as we now know it to
be. If the law of nature was more modestly “water is a fluid according to the experiences of the
Indian prince”, there seems to be little reason to “wisely” resist a testimony of water (ice) that
was completely out of scope to begin with.

Testimonies are frail, but so is human understanding even if less so. Since Hume’s whole
argument rests on comparing the probabilities of the laws of nature with false testimony, I
believe that we must be equally, if not more critical of the law of nature we hold as unwavering
and be open to looser ideas of “regular” as a necessary condition. Testimonies of miracles can
be ignored as faulty anomalies or more constructively, they can be regarded as worthy
challengers that remind us to constantly validate our supposed laws of nature.