A law of nature

Miracles are often defined as events which "violate" the "laws of nature" (or, equivalently, "natural law(s)" or "scientific law(s)"). Whether miracles, as thus defined, ever occur (or even can occur) will thus depend on whether there are such things as "laws of nature", and on whether anything ever "violates" (or even can ever violate) these laws. But before we can know whether there are such laws and whether they ever are or can be violated, we must first determine what is meant by the term "law of nature" (or natural law, or scientific law).

It will be convenient for present purposes to distinguish between a weak sense and a strong sense of the term "law of nature". In the weak sense of this term, a law of nature is simply a universally true empirical generalisation in a sense more or less like that attributed (perhaps unfairly) to Hume. It is universally true simply in the sense that it truthfully applies to whatever it applies to without exception. Its universality, in other words, consists simply in the fact that it is absolutely exceptionless.

It cannot, however, be true by definition, nor can its truth be logically necessary for any other reason. Thus, its truth is a contingent matter of fact, and it is "empirical" in the sense that our knowledge of its truth must be based on observation or experiment. Moreover, since we are here concerned with events in the physical world, which are in principle repeatable ad infinitum, any law of nature which concerns us in this context cannot be conclusively confirmed, but can only be confirmed to a high degree of probability by means of induction.

The general form of any such law will simply be a proposition to the effect that for every event answering such-and-such description, thus-and-so is or will be the case, without exception. Given that anything which counts as a "violation" of a law must, according to any reasonable interpretation of that term, be an exception to that law, it would seem to follow logically that there cannot be any such thing as a "violation" of a law of nature as understood in this weak, somewhat Humean, sense.

For suppose that 'x' is a purported violation, and 'L' is the purported law of nature which 'x' purportedly "violates". If 'x' does in fact occur and is in fact a violation, then 'x' must constitute an exception to 'L', as explained above. But if there are any exceptions to 'L', then, according to our proposed definition of a law of nature, 'L' cannot be such a law, which entails that 'x', contrary to hypothesis, cannot constitute an exception to such a law simply by constituting an exception to 'L'.

Now we have good reason to believe that there are at least some laws of nature as defined above. Consider, for example, the relationship between thunder and lightning, or the fact that water boils at 212 degrees Fahrenheit at sea-level. Although it should not be necessary for our present purposes to craft detailed propositions describing those phenomena which, when appropriately qualified, would satisfy the defining characteristics of a law of nature as delineated above, it seems that it would not be difficult to do so in both of these cases.

Crude approximations of these propositions would be simply that thunder always follows lightning and that water always boils when heated to an appropriate temperature as defined by atmospheric boundary conditions. Both of these propositions, if carefully reformulated and appropriately qualified, would appear to be universally true empirical generalisations satisfying all the conditions specified above for something's being a law of nature.

Let us now suppose, however, that we were to observe clear exceptions to these generalisations. Suppose, in other words, that we were to observe lightning to occur, in a situation in which all the conditions alluded to above were satisfied, but that no thunder clap ensued. Or suppose we were to heat water, under all the appropriate conditions, to the appropriate temperature and it did not boil. Could we then say, in either of these cases, that there had thus been either an exception to or a violation of a law of nature?

The answer, of course, is no, because in both of these cases, the purported law would turn out not to be a law after all, since the generalisation in question would turn out not to be absolutely exceptionless. The purported violation could not then be a violation, since it is not possible to violate a "law" that is not a law. We could, of course, in a sense preserve the purported law in question by reformulating it in such a way as to take this supposed exception into account, thereby creating a new law which incorporates the suacles on a different, more robust view of the laws of nature.

Whatever else we may say about miracles, there is a tendency to think that there must be something especially noteworthy about them. Thus, for example, they often tend to be especially beneficial to particular individuals, and to be of divine origin. Being especially beneficial, however, does not distinguish miracles from simple good fortune. As for divine origins, God is both the Creator and Sustainer of the universe from a theistic point of view, so that all existing things are of divine origin in a fairly straightforward sense of the term.

In particular, everything that exists is thought to exist because and only because He has created it and continues to sustain it. Perhaps, then, miracles are especially noteworthy because they occur (or are thought to occur) despite exto exist, we will never know whether that person has really died, since for all we know, even if all biological function has completely ceased for a prolonged period of time, it may one day resume, in which case it will turn out that this person, all appearances to the contrary, had never really been dead.

The problem here is that defining death as "permanent cessation of biological function" renders (1) tautologous, which has the effect of ruling it out as a possible law of nature in the sense currently under consideration. Resurrection in this case thus becomeudely expressed as the proposition (1) "dead macles on a different, more robust view of the laws of nature. Whatever else we may say about miracles, there is a tendency to think that there must be something especially noteworthy about them.

Thus, for example, they often tend to be especially beneficial to particular individuals, and to be of divine origin. Being especially beneficial, however, does not distinguish miracles from simple good fortune. As for divine origins, God is both the Creator and Sustainer of the universe from a theistic point of view, so that all existing things are of divine origin in a fairly straightforward sense of the term.

In particular, everything that exists is thought to exist because and only because He has created it and continues to sustain it. Perhaps, then, miracles are especially noteworthy because they occur (or are thought to occur) despite exto exist, we will never know whether that person has really died, since for all we know, even if all biological function has completely ceased for a prolonged period of time, it may one day resume, in which case it will turn out that this person, all appearances to the contrary, had never really been dead.

The problem here is that defining death as "permanent cessation of biological function" renders (1) tautologous, which has the effect of ruling it out as a possible law of nature in the sense currently under consideration. Resurrection in this case thus becomes logically impossible, and belief in it unreafficulties concerning the definitions of "death" and "resurrection" and the nature of the "law" that resurrection would supposedly violate, we are left with the same situation as before.

Understanding (1') simply as an absolutely exceptionless empirical generalisation, which it must be in order to qualify as a law of nature in our current Humean sense, renders resurrection as defined above logically inconsistent with the truth of (1'). For the same sorts of reasons given in our previous examples, resurrection cannot both occur and also constitute a violation of (1'), or of any other law of nature, for that matter.

Notice, however, that unless we simply presuppose the truth of (1') as an absolutely exceptionless generalization, there is absolutely nothing here which implies that resurrection cannot or even does not occur. Any accumulation of empirical data which tends to confirm (1') inductively will at most show that resurrection is statistically improbable or unlikely, which believers in resurrection will readily admit, since this simply seems to support the view that resurrection is miraculous.

The only thing that has been shown here is that if resurrection does occur, it cannot constitute a violation of a law of nature, and thus cannot be a miracle as presently understood. There still seems to be something wrong here. We seem to have arrived at a dead end. Perhaps we need to go back to th "complete cessation of heart function," "complete cessation of respiration," or "complete cessation of brain activity," (1) admits of actual or at least potential counter-examples, due to advances in medical technology.

Cardiovascular function can sometimes be artificially restored and maintained almost indefinitely, and although the situation concerning brain activity is more problematic, there is no reason to suppose that such activity could not one day be artificially restored and maintained as well. Moreover, no one believes that any of these actual or potential occurrences are or would be contrary to the laws of nature, whether or not we can now provide satisfactory explanations for them. Thus it would appear that (1), as presently stated, cannot constitute a law of nature in the sense currently under consideration.