But what makes an event a miracle? Hume was very precise about this. It is "a transgression of a law of nature by a particular volition of the deity, or the interposition of some invisible agent." Hume didn't seem to think this was very controversial, since he originally relegated his definition to a mere footnote.
Two hundred and fifty years later, however, this gloss is more contentious. It should not be. Hume was right, and any attempt to make a miracle anything less destroys the phenomenon it strives to name....
Hume is therefore right. Miracles are violations of the laws of nature. The merely extraordinary is not miraculous.
It is nonsense, though, to speak of laws of nature in Hume's sense as being 'obeyed' by everything in the universe in the way Baggini speaks of laws of nature. Hume's laws are uniform regularities adequate for proof. This is not at all Baggini's conception of a law of nature, which is from a Humean perspective spookily normative; there is no possible way to fit such a conception into Hume's version of empiricism. We have no way of knowing that "every single physical event in the universe" obeys a given law; how could we possibly do so based solely on our impressions and the basic rules, derived from experience, that we apply to them? What, on Hume's account makes it a law of nature that the dead do not return to life? Hume tells us that it is because "that has never been observed, in any age or country." That is all. And it is a question of how we are to understand this without begging the question -- because we know what has never been observed only through testimony, and some testimony says that people have occasionally returned from the dead. Ultimately, it appears, Hume's view is based on a sort of parsimonious inference to the best explanation: as he says to Hugh Blair in response to an argument by Campbell, if you run after every silly tale, you eventually end up believing nonsense.
One of the problems with the whole argument is that Hume definitely knew of at least two philosophers (Butler and Malebranche) who held that miracles were interventions by God but denied that they were violations of laws of nature, because they were in fact manifestations of higher-order laws that lower-order formulations could not rule out (and were, in fact, conditional upon). And Hume certainly does not elsewhere assume that we have in hand the most developed laws of nature. When we set aside Baggini's anachronizing, we are, in fact, left with the puzzle of how Hume's conception of a law of nature allows it to function in the particular way he suggests it does. I would suggest that what Hume primarily calls the laws of nature are in fact those features of experience upon which we have become so dependent that we cannot help but assume them in practice: e.g., that gravity will not suddenly and massively change, that what is very hot will burn when you touch it, and so forth. They are simply descriptions of experience supported by induction without exception, and therefore have the strongest sort of psychological force any conclusion from experience can have. Thus if you want to find the closest contemporary analogue to Hume's laws of nature, it will not be 'laws of nature' in the sense Baggini suggests, which have a strong, if never really explained, modal component: physical laws say that things "must" happen, physical events "obey" them, they define "physical possibility". This is all mumbo-jumbo to a genuine Humean. Rather, the closest analogue would be the schema found in cognitive science explanations. Hume's miracles are more or less schema-inconsistent happenings. But this raises a second puzzle: there are extraordinary events that we would tend to allow as "physically possible," however we may mean it, that are schema-inconsistent as well.
If the argument against accepting testimony for miracles is to be accepted, it must be against what believers in miracles count as miraculous, and of the three traditional types of miracle, only one, the counternatural, is plausibly a violation or suspension of a law of nature in Baggini's sense. Another, the supernatural (which Campbell calls the 'preternatural'), evades violation of the laws of nature because any sensible account of laws of nature holds that they say what things will do if no cause external to the system intervenes; any such cause changes the conditions, and while the law still holds, the conditions to which it applies are different, thus resulting in different behavior. It is clear that Baggini would consider this a violation of the laws of nature; but it is also clear that what would ground this conclusion is his very un-Humean conception of what the laws of nature are. Or, in other terms, it is physical closure that is doing the real work, the notion that there can be no causal laws except physical laws governing physical objects. And the third, the preternatural (what Campbell calls the 'natural'), is precisely the sort of thing that Baggini claims is "only a miracle in a figurative sense": "an extremely unlikely and fortuitous sequence of events" that is not outside the power of natural causes. What he overlooks is the point that extremely unlikely and fortuitous sequences of events can nonetheless carry significance because they can be arranged if the laws of nature and initial conditions are themselves effects: events that are extremely unlikely in terms of natural laws and fortuitous in terms of initial conditions can be guaranteed by a cause governing the whole system. (This point was made at great length and sometimes with cleverness against Hume-style arguments by Charles Babbage.) And on Hume's account anything unlikely and fortuitous enough can count as a miracle; this is the very point of the footnote to which Baggini refers. Despite the contentiousness of the 'violation' formulation, Hume's treatment does more justice to actual miracle claims than Baggini's. Again, what is doing the work in Baggini's argument is an entirely non-Humean set of assumptions. And the question has to be asked -- and Baggini does not get around to asking it -- whether these assumptions beg the question.
In any case, my major point here is that Baggini is massively deviating from the spirit of Hume in order to preserve the letter. While Hume's focus is on the miraculous, he clearly does not exempt the extraordinary from his argument: this is explicitly stated in his account of counterpoise. Miracles are for Hume simply the strongest case of the extraordinary; much of his argument depends on the marvelous and the miraculous being on a spectrum. But Baggini is very clearly committed to treating the extraordinary and the miraculous as radically different. A further sign of the deviation is that Baggini interprets Newton's laws as laying out bounds of physical possibility. But Hume's own view of Newton's laws is indifferent; he thinks that they rid us of any such questions by giving us a way of dealing with the phenomena while restoring our skepticism in what underlies the phenomena. Hypotheses non fingo. This is beautifully stated, in a comparison with Boyle's mechanical philosophy, in the History of England. Baggini's argument is inconsistent with Hume's own account of what a violation of a law of nature would be, and appeals to things to which a consistent Humean would not appeal; to make Hume out to be right, Baggini's argument makes Hume wrong on practically every point. The slogan is preserved, the meaning is not.
Baggini has some other interesting articles on Hume and religion at the Comment is Free space. I might possibly comment on them at some point, since the same thing is in evidence there: a handful of Hume's words are preserved as slogans, but Hume's interesting underlying reasoning is stripped away and replaced by something else, and what Baggini replaces it with is not always as clever or insightful as what is replaced, and certainly not consistent with Hume's own expressed views.