Philosophical Disquisitions: The Possibility of Historical Evidence for Miracles (Part 1)

This post is part of my series on miracles. For an index, see here.

Over the next few entries I will be going through the following article:

Luck, M. “Against the Possibility of Historical Evidence for Miracles” (2005) 44 Sophia 7

In this article, Luck assesses Richard Swinburne’s argument for the viability of strong historical evidence for miracles. At the outset, Luck notes that Swinburne’s argument is based on the following three assertions:

Assertion 1: It is logically possible for a violation of a law of nature to occur.
Assertion 2: The existence of particular present events could be indicative of the prior occurrence of particular violations of the laws of nature (thanks to event causality).
Assertion 3: There are conditions under which it is reasonable to assume that a violation of a law of nature was caused by God.

Luck sets out to evaluate the plausibility of each of these assertions. Although he ultimately agrees with assertion 1, he finds reason to doubt each of the other two assertions (particularly the second).

In this part, I will run through Luck’s evaluation of the first assertion. The analysis is hugely significant in the general context of the miracle debate where the appropriate definition of a law of nature, as well as the possibility of its violation, are key concerns.

We begin by considering two arguments to the effect that violations of laws of nature are impossible. These arguments already featured in my posts on Everitt’s discussion of miracles. They are given slightly more precise formalisations here.

1. The Argument from Necessity
The first argument comes from those who have a necessitarian conception of a law of nature. According to this conception, it is impossible to violate a law of nature because such a law prescribes a necessary set of relations between states of affairs. For a necessitarian theorist, a claim like “water boils at 100c at normal pressure” prescribes a relation that cannot be broken.

This gives rise to the following argument:

(1) A law of nature is a necessary relation.
(2) A violation of a law of nature is a violation of a necessary relation.
(3) A necessary relation cannot be violated.
(4) Therefore, a violation of a law of nature cannot occur.

This argument relies on the view that laws of nature are logically necessary. Some people might have a slightly more relaxed view on the type of necessity involved. Specifically, they might think laws of nature prescribe physically or naturally necessary relations. Under such a conception, would it not be possible for there to be non-physical or supernatural violations of laws of nature?

The answer is “no”. Because under that conception a supernaturally caused event would not actually be a violation, but rather a positive instance of a physically necessary law. As follows:

(5) A law of nature is a necessary relation between natural events, iff those events have natural causes.
(6) A violation of a law of nature is a violation of a necessary relation between natural events, iff the natural events have natural causes.
(7) Miracles involve natural events with non-natural causes.
(8) A violation of a law of nature cannot be caused by a miracle.

So it seems that, on either necessitarian view, there cannot be a violation of a law of nature.

2. The Argument from Regularity
The necessitarian conception is not the only game in town. In fact, the more popular conception of a law of nature, at least since the time of Hume, is the regularity one. According to this view, a law of nature is simply a highly generalised description of the regular pattern of events.

And, so the argument goes, there cannot be a violation of a highly generalised description of the regular pattern of events. Why not? Because any putative violation would simply be a new piece of data to be accommodated by the highly generalised description. In other words, it would simply be evidence that the existing description is faulty and needs to be revised.

These sentiments can be given the following formal expression [note: this is a slight revision of the version found in Luck’s article]:

(9) Laws of nature are generalised descriptions of the actual course of events.
(10) Every event that occurs can be accounted for by these generalisations.
(11) A violation of a law of nature would be an event that could not be accounted for by these generalisations.
(12) Therefore, a violation of a law of nature cannot occur.

If this argument, and the necessitarian one, are sound, then it seems like Swinburne faces an uphill battle to defend his first assertion. How does he do it? Unsurprisingly, for those who are familiar with Swinburne’s work, his argument relies on the explanatory virtue of simplicity.

3. Swinburne on the Possibility of a Violation-Miracle
Swinburne begins by seeming to accept a broadly regularitarian conception of a law of nature. In other words, he takes it that a law of nature (L) is simply a highly successful description of events with some predictive power. He then defines a miracle as a “non-repeatable counter instance to a law of nature”.

Let’s assume that L stands for “no human being can levitate”. Now suppose that an event occurs (E) in which a human being levitates into the air. Swinburne points out that the occurrence of such an event could prompt us to formulate a new law of nature (L1) that accounts for the levitation in this particular instance. L1 would effectively add an exception-clause to L.

Swinburne then argues that while we could accept L1 as being the true law of nature, doing so might come at a cost. If L1 only accounted for the exception E, it would be more ad hoc and less simple than L. And Swinburne maintains that if reformulating L to account for E comes at the expense of simplicity, we would be justified in calling E a miracle (i.e. a non-repeatable counter-instance to L).

Luck thinks Swinburne’s argument in this respect is sound. He is persuaded by the fact that one of the most successful modern Humean approaches to the laws of nature -- Lewis’s Systematic theory -- allows for violations of this form.

According to Lewis’s theory, a law of nature is a generalisation that occurs as a theorem (or axiom) in an explanatory deductive system that achieves the best combination of simplicity and strength. As such, Lewis’s view can tolerate some violations of the laws of nature, provided the system as a whole maintains its simplicity and strength. As follows:

(13) Laws of nature are axioms with the best combination of simplicity and strength.
(14) It is possible that these laws may be substantially simpler and insubstantially weaker by not accounting for the occurrence of certain events.
(15) It is possible these unaccounted for events could be in violation of the axioms with the best combination of simplicity and strength.
(16) Therefore, it is possible for a violation of a law of nature to occur.

That concludes Luck’s evaluation of Swinburne’s first assertion. In the next post, we’ll take at look at his evaluation of the second assertion.