I'm going to review Robert Larmer's The Legitimacy of Miracles (Lexington Books, 2014). Larmer is a Canadian Christian philosopher whose area of specialization includes the philosophy of miracles. He's published books and articles on the subject since 1983. The Legitimacy of Miracle is the culmination of 40+ years of research and reflection. This may be his magnum opus on the topic. Incidentally, Larmer has forthcoming book entitled Hume's Muddled Mess. Larmer is also developed a website. Stay tuned!
In chapter 1, Larmer outlines occasionalist, deistic, and supernatural models of divine agency. Larmer compares these models to the pretheoretical concept of miracles in Scripture, concluding that Scripture supports the supernatural model, involving ordinary providence: second causes with natural teleology. That's the context which makes divine intervention meaningful, and miracles detectable.
The raw data of Scripture furnish paradigm examples which in turn provide the basis for a philosophical definition in chapter 2: namely: "a miracle is an unusual and religiously significant event which reveals and furthers God's purposes, is beyond the power of physical nature to produce, and is caused by an agent who transcends nature".
He defines "unusual" in the sense that they are events which unaided physical nature would not otherwise produce. They are "extraordinary" in the sense that they constitute exceptions to what would occur when nature is allowed to run its course.
Quoting Newman, he says miracles are not "unconnected and unmeaning occurrences", but hold a place in the "extensive plan of divine government".
Larmer doesn't consider God to be the only miraculous agent. Creatures like angels can be miraculous agents.
Larmer criticizes the "violation of natural laws" definition in part because there's no agreed upon definition of natural laws. If natural laws are defined as universal generalizations, then miracles don't violate natural laws inasmuch as the definition covers anything that happens, which, if miracles occur, would be included in whatever actually happens. So even if we define a miracle in reference to natural laws, a miracle is consistent with natural laws on nomic necessity and regularity theories.
Causal dispositional theories are ambiguous in reference to miracles. If no event can violate a law of nature, does that mean miracles can't happen, or that miracles don't violate natural laws?
Larmer approvingly quotes J. S. Mill's statement that
in order that any alleged fact should be contradictory to the law of causation, the allegation must be, not simply that the cause existed without being followed by the effect, for that would be no uncommon occurrence; but that this happened in the absence of any adequate counteracting cause. Now in the case of an alleged miracle, the assertion is exactly the opposite of this. It is, that the effect was defeated, not in the absence, but in consequence of a counteracting cause, namely, a direct imposition of act of the will of some being who has power over nature. A miracle is no contradiction to the law of cause and effect; it is a new effect, supposed to be produced by the introduction of a new cause.
Larmer says natural laws are silent on the question of events caused by divine intervention. They don't speak to that issue one way or the other. Natural laws have implicit ceteris paribus clauses–what will transpire all other things being equal.
In the same chapter, Larmer analyses coincidence miracles. Could some miraculous events be the end-result of front-loaded determinism? Larmer rejects that as deistic.
On a related note, Larmer differentiates miracles from special providence.
Larmer draws an analogy between divine intervention and substance dualism–in contrast to physicalism and physical determinism.
Larmer draws an analogy between divine intervention and human agency to alter the course of nature. In that connection, he denies that miracles are intrinsically rare.
Reloading a gun doesn't violate a law of nature, even though it may change the outcome. In vitro fertilization doesn't violate a law of nature, even though it may change the course of nature (my examples).
Larmer addresses the objection that miracles violate the conservation of energy. He says that's only true if we view nature as a closed system. But that begs the question in relation to miracles. Moreover, a Christian will reject the stipulation that God cannot create or destroy energy. Indeed, assuming that miracles "violate" the conservation of energy, this would be evidence that it's not an absolute principle.
It would be viciously circular to adopt the conservation of energy on the grounds that there's no evidence to the contrary, then appeal to the conservation of energy to rule out ostensible evidence to the contrary.
Natural laws, considered in isolation, predict nothing–but only in conjunction with supplementary information regarding initial conditions.
In chapter three, Larmer engages Hume's classic essay on miracles, which is the standard frame of reference in modern philosophy and theology. According to the traditional interpretation, Hume proposed an a priori epistemological argument according to which testimonial evidence is incapable, even in principle of justifying rational belief in miracles. Revisionist interpretations have challenged the traditional interpretation. Larmer defends the traditional interpretation.
Since Larmer rejects the Humean concept of miracles, he considers the a priori argument to be a failure, because Hume misframed the issue.
Larmer says Hume's appeal to the uniformity of nature is at variance with his theory of induction and causation. On his own grounds, Hume can't presume that the future will resemble the past, or even probably resemble the past.
In addition, Hume has no principled basis to accept testimonial evidence for merely unusual events while rejecting testimonial evidence for miraculous events.
Over and above the lack of consistency with Hume's theory of induction and causation, Larmer raises direct objections to Hume's a priori argument:
If natural laws are defined as exceptionless generalizations, can they be revised when exceptions are discovered? By Hume's strictures, natural laws can never be revised, because prior invariable experience automatically discounts subsequent observations to the contrary. And that has the same logical status as a reported miracle.
Larmer says the question of whether an event occurred is logically distinct from the question of what caused it, if it did in fact occur. Criteria for accepting that an event did in fact occur are quite different from criteria for determining whether its cause was natural or supernatural. Testimony must be believed before there's any point in analysing what happened.
Hume attempts to distinguish between warranted belief in unusual events and unwarranted belief in miracles by saying the former are analogous to our general experience whereas the latter are disanalagous to our general experience. Yet that's at variance with his hypothetical case of the Indian prince whom Hume says is justified in disbelieving reports about walking on (frozen) water.
Moreover, Larmer says miraculous events are analogous inasmuch as personal agents, whether divine or human, produce outcomes contrary to the ordinary course of nature.
In addition, Hume's objections are confined to secondhand information about miracles. He never makes allowance for firsthand experience of miracles.
Larmer says Hume's appeal to the uniformity of nature is viciously circular. We only know that uniform experience rules out reported miracles if we know that every such report is false. And we can only know that every such report is false if we already know that miracles never happen. So there's no independent evidence for Hume's appeal.
In chapter 4, Larmer examines the God of the gaps objection, which he construes as the (allegedly) fallacious argument from ignorance. He counters that arguments from silence are often used in historical research lack of knowledge inferences can be reasonable in psychology, natural sciences employ the concept of negative evidence, and philosophers acknowledge the legitimacy of non-see-um inferences.
Lack of evidence is not an argument from ignorance in case there's a reasonable expectation that if a claim were true, we should be able to find supporting evidence.
Larmer quotes Del Ratzch:
Identification of the agency as supernatural depends upon the implicit claim that neither nature alone not finite agent activity is causally or explanatorily adequate for the phenomena in question…if neither nature nor finite agency can produce some phenomenon inarguably before us, then supernatural agency is about the only option left.
Take the evidence that the sabertooth tiger is extinct. Although lack of evidence may not be conclusive, it justifies a provisional assessment regarding the extinct status of that species (my example).
Critics appeal to the stately march of science. Larmer responds by distinguishing between artifacts and natural products.
He notes that even prescientific theologians distinguished between primary and secondary causality, or mediate and immediate agency. They didn't attribute every event to God's direct action. That's an urban legend.
Success in filling one kind of gap by discovering a natural mechanism doesn't imply or predict for success in filling all kinds of gaps by discovering a natural mechanism.
The argument for scientific progress cuts both ways. Superior scientific knowledge can make some reported miracles more naturally inexplicable than ever. Indeed, the difficulty of providing a naturalistic alternative explanation is why skeptics simply deny the occurrence of some reported miracles.
Larmer says appeal to some presently undiscovered natural causes commits the critic to unwarranted skepticism regarding our understanding of how nature works. Ironically, that stands in contrast to a Christian doctrine of ordinary providence. Quoting Lennox, we need to distinguish between how things work and how things came to exist in the first place.
Larmer uses the illustration of a man who puts diamonds in a safe, only to find the diamonds missing. He can either infer that someone else knew the combination (perhaps a safe-cracker) or else there's some unknown natural process by which diamonds dematerialize. Which is more reasonable?
He says biblical miracles are not anomalous surds, but figure in a larger teleological pattern.
To the objection that miracles are "science-stoppers," Larner provides two criteria:
i) the event has religious significance
ii) the event is an exception to established pattern
Likewise, we need to distinguish between events caused by unaided nature and personal agency. Indeed, there are sciences devoted to the role of personal agency (e.g. forensics, archeology, anthropology, cryptography). The fact that some events are caused by agents manipulating nature doesn't automatically foster skepticism about the ordinary course of nature.
Larmer's basic objection to methodological naturalism is that, ironically, it's unscientific. The aim of science is to discover the cause of natural events. Methodological naturalism precludes a scientist from identifying a supernatural cause even if that's the correct explanation. So it stultifies science by prohibiting a scientist from following the evidence wherever it leads. Method mustn't trump evidence.
By the same token, methodological naturalism isn't neutral, but prejudicial. Methodological naturalism is only warranted if metaphysical naturalism is warranted.
Larmer says methodological naturalism cultivates intellectual indolence.
In addition, Larmer says some reported miracles are amendable to scientific confirmation or disconfirmation, such as medically verifiable miracles. Although supernatural causes are not empirical, their effects may be empirical. Larmer draws a parallel with particle physics.
In chapter 5, Larmer examines the claim that miracles are incoherent. The first section overlaps with chapter 2, although it has some distinctive material.
Later on Larmer fields other objections, such as the claim that an incorporeal agent can't exert causal influence on physical objects. As Larmer notes, that parallels objections are raised to interactionist substance dualism. But here and elsewhere, he counterattacks by raising objections to physicalism.
Although he doesn't use this terminology, Larmer counters the objection by saying there comes a point beyond which or below which we must allow for direct causation–otherwise we're stuck in an infinite regress. Every cause-effect relation can't be facilitated by an intervening physical medium without appeal to the infinite divisibility of matter.
Borrowing a page of Hume, Larmer says our understanding of causation is fundamentally descriptive. We simply know from experience that some things cause other things. But why that's the case is ultimately mysterious.
In addition, Larmer says we have immediate knowledge of mental causation. That's more fundamental and unquestionable than how physical objects causally interact. And that's analogous to divine action in the world.
Moreover, Larmer alludes to the hard problem of consciousness. A problem for physicalism, not dualism.
Larmer turns tables on Troeltsch by agreeing with his principle of analogy, but appealing to well-attested modern miracles to demonstrate that the past and the present are comparable in that regard.
Larmer says Nowell-Smith has a flawed definition by making predictability a necessary condition of what constitutes an explanation. Larmer says that fails to distinguish between impersonal agencies and personal agents.
Larmer addresses the question of whether repeatability is necessary to rule out coincidence.
Larmer addresses the objection that miracles depict God as an incompetent engineer who must keep adjusting the machinery. Variations on this objection have been around since Leibniz, Spinoza, and Maimonides. The stock metaphor is the clockwork universe.
Larmer objects to the mechanical metaphor. He says that instead of comparing the world to a machine, what if God designed the world to function like a musical instrument. You don't just make a violin and leave it alone. Rather, you make it to play it. To do something with what you make.
Dropping the metaphor, he appeals to a dynamic rather than static relationship between the Creator and his rational creatures.
In addition, Larmer appeals to the self-imposed limitations of God, according to freewill theism. Such a God may need to adapt to obstacles which recalcitrant free agents pose to his objectives.
Larmer then addresses the objection that the inequitable distribution of miracles is incompatible with divine benevolence. Larmer notes that this isn't unique to miracles, but to the inequitable distribution of certain goods generally, so it goes to larger questions of theodicy. Why not more good and less evil? By the same token, defending this specific objection to miracles can make use of general responses to the problem of evil.
He appeals to the soul-making theodicy. In addition, he seems to indicate that the fact that God only selectively answers some petitionary prayers fosters humility. If God routinely answered prayer, that would foster pride. If that's what he means, the argument is underdeveloped. It may related to a further point he later makes that we never control God.
He says God generally performs miracles through other individuals, which means God will sometimes be frustrated due to lack of human cooperation (e.g. Mk 6:5).
Finally, he addressed the objection that miracles are at odds with divine transcendence, by making God just another agent.
Larmer takes the position that laws of nature are necessary, not in the absolute sense that God couldn't create different laws, but that he couldn't create the same world with different laws.
In chapter six, Larmer begins by reviewing Swinburne's four types of evidence. He takes issue with Swinburne's Humean definition of a miracle.
More generally, Larmer says that to be rational, worldviews must be based on evidence, not dictate what the evidence must be. Worldviews are not an independent source of evidence, but must be responsive to a comprehensive body of evidence.
Larmer discusses the relationship between firsthand observation and testimonial evidence.
In addition, there's a degree of circularity to evidential appeals inasmuch as we must privilege some evidence as the standard of comparison when assessing other evidential appeals.
Larmer says that unless there's conflicting evidence which casts doubt on a reported miracle, the evidence in favor of that reported event should be accepted. In the absence of conflicting evidence, the onus lies on those who wish to dismiss the reported miracle.
Larmer says multiple, independent attestation is strong evidence, but by the same token, that means we must make allowance for minor discrepancies.
Larmer says that while it's possible to personally witness a miracle, most miracles will take place outside the firsthand experience of any particular individual, so testimonial evidence remains pertinent.
Larmer comments on Sagan's famous slogan that extraordinary claims demand extraordinary evidence. Moreover, even some Christian philosophers think warranted belief in miracles must meet a higher evidential threshold.
In that connection, Lamer indicates that facile appeal to Ockham's Razor is viciously circular. You can't rationally justify disbelief in God on the grounds that there's no evidence for his existence and then discount evidence for miracles on the grounds that God's nonexistence has already been established.
Larmer says the demand cuts both ways. He quotes Licona's observation that appeal to group hallucinations (to discredit the post-Resurrection appearances) is an extraordinary claim requiring extraordinary evidence.
Larmer addresses the objection that ostensible evidence for miracles must be balanced against the evidence for God's nonexistence (e.g. problem of evil). Larmer endeavors to turn tables on the critic by saying the existence of evil actually furnishes evidence for God's existence rather than nonexistence inasmuch as moral realism and moral responsibility have no foundation in naturalism. Larmer appeals to moral intuition and libertarian freewill. He says naturalists typically either outright deny the existence of freewill or define it according to compatibilism, but both options negate moral ascriptions. Larmer revisits the same issue in chapter 7.
Larmer says the popular view that miracles require "extraordinary evidence" rests on the mistaken assumption that there are competing bodies of evidence. But since he denies that such a conflict exists in general, he denies the higher evidential threshold that reported miracles must surmount. Larmer agrees with Newman that miracles don't require a type of evidence distinct from what's required for other events. It's impossible to draw a qualitative or quantitative evidential line to prove an earthquake or meteor shower. Testimony can't be more than that of competent and honest men. We must content ourselves with obtaining this kind of evidence rather than some inhumane ideal.
Larmer says even modest evidence for a reported miracle gives good grounds for believing it in the absence of counterevidence.
Larmer concedes that while it's unavoidable to assess claims within a framework of prior beliefs regarding what is possible or probable, that mustn't be allowed to override facts. He cites an amusing vignette about Laplace, who imperiously dismissed reported meteors. He approvingly quotes Newman's statement that experts are at risk of "correcting the evidence for their senses" when confronted with "strange phenomena"; conversely, "the same persons are competent to attest miraculous facts who are suitable witnesses of corresponding ones"; "everyone is apt to interpret facts in his own way; if the superstitious see too many prodigies, men of science may see to few".
Moreover, Larmer says that in some cases, we do have extraordinary evidence for miracles. So even by that artificially high standard, the argument from miracles goes through.
Larmer relates a personal anecdote about respondents who were willing to dismiss reported miracles, not on the basis of evidential considerations relevant to establishing their occurrence, but on the basis that if such events were to occur, they'd be difficult if not impossible to explain naturalistically.
Larmer says Hume's first three a posteriori objections are tendentious assertions rather than arguments. Moreover, they are hasty generalizations. And they cut both ways. For instance, cessationists are predisposed to be unduly skeptical of reported miracles.
The fourth a posteriori argument concerns non-Christian miracles. Larmer indicates verification is not the sole function of miracles, so even if miracles exist in rival religions, that doesn't ipso facto cancel respective claimants.
He says that apart from Christianity, most religions don't even emphasize miraculous attestation. In addition, not all reports are equally well-attested. It's necessary to examine them on a case-by-case basis. Larmer says this is analogous sifting divergent testimony in the court room. Larmer says Hume misrepresents the actual procedure (e.g. process of elimination).
In chapter 7, Larmer says miracles can furnish direct evidence for God's existence. One needn't prove God first. It is the event itself, and not the subsequence classification of the event as miraculous, which functions as evidence for God. The skeptic should consider God's existence as a hypothetical assumption, then ask if that's the best explanation for the event–compared to rival hypotheses.
Although he doesn't say so, this means Larmer is siding with evidential apologetics rather than classical apologetics in that particular regard.
Responding to the objection of J. S. Mill that we can never definitively rule out a naturalistic explanation, Larmer says we need to distinguish between abstract possibilities and realistic probabilities. Just because a naturalistic explanation might be an outside possibility doesn't mean all possible explanations are equally plausible.
Larmer points out that in practice, skeptics typically deny the occurrence of reported miracles rather than attempting to explain them naturalistically.
Larmer classifies the argument from miracles as variation on the teleological argument. He says, however, it has an advantage over the usual examples. What's at issue is not the supernatural pedigree of the event, if it occurred, but whether it occurred. Larmer says the evidence for miracles is often underestimated or simply ignored. In reality, there's a "massive amount of evidence".
Larmer adds that if there's so much evidence of supernatural intervention in human history, then why not natural history (e.g. the origin and development of life, fine-tuning argument)?
Larmer fields the claim that the argument from miracles might be used in support of polytheism, pantheism, or panentheism.
To the objection that our experience of agency is confined to embodied agents, Larder's quotes Alston's contention that the concept of agency is more abstract.
On a related note, Larmer finds total apophatic theology to be incoherent since every negation implies some kind of prior affirmation. If we have no positive knowledge of God prior to what we negate, there's no meaningful way to know what to negate. Larmer says abstract concepts can be univocal, but analogical when predicated of different things.
In chapter 8, he argues that miracles can furnish evidence for a particular religion (i.e. Christianity), but are too coarse-grained to adjudicate intramural Christian disputes. He says post-Christian Judaism doesn't typically appeal to modern miracles. He says Islam doesn't claim to be established by miracles. The best candidate is the ambiguous "splitting of the moon" in surah 54:1-2.
Larmer notes the emphasis on power evangelism, both in the NT and the modern mission field. This includes revelatory dreams, visions, prophecies, and exorcism–as well as miracles. Sometimes this involves the relationship between a supernatural event and a supernatural interpretation of said event. These aren't free-floating miracles, but tied to a particular religion's claims.
He says you can't separate the miraculous incidents in the Gospels from the mundane incidents. Either the Gospels are historically reliable in reporting both kinds of events or neither. Larmer appeals to Lewis's Lord/liar/lunatic trilemma.
Larmer doesn't deny the possible occurrence of non-Christian miracles, but says Christianity supplies the best frame of reference for explaining non-Christian miracles as well as Christian miracles.
Larmer evaluates miracles attributed to Apollonius of Tyana, Hanina ben Dosa, Honi the Circle-Drawer, and Vespasian. While his position allows for non-Christian miracles, these are poorly-attested examples.
He says if miracles are deemed to be maximally improbable compared to other events and explanations, then all reported miracles are equally incredible. But he denies that. Reverting to his critique of methodological naturalism, he says the job of a historian is not to prejudge what can or cannot happen, but to be guided by the evidence.
In the appendix, Larmer records four dramatic, medically verified miraculous healings. In his popular level Dialogues on Miracle, Larmer has an appendix with an additional six cases of miraculous healings.
1. Larmer's monograph is an outstanding contribution to the philosophical defense of Christian miracles. The analysis is sophisticated and detailed. A thoroughgoing, often multiple-point response to stock objections to miracles. In general, this is presently the best work of its kind. The current standard-bearer.
2. Larmer has an impressive bibliography, which includes most of the major modern titles and historical titles. His bibliography could be updated in one or two places. He lists the 1987 edition of Craig Blomberg's The Historical Reliability of the Gospels, but Blomberg published a revised and expanded edition in 2007. In addition, Blomberg issued a more recent edition (2011) of Jesus and the Gospels, although that's only a few pages longer than the original.
There are a few striking omissions in Larmer's bibliography. No mention of Peter van Inwagen's essay, "Of "Of Miracles," in The Possibility of Resurrection and Other Essays in Christian Apologetics (Boulder, Colorado: Westview Press, 1998), chapter 6.
Or Timothy McGrew's SEP entry:
Or McGrew's review of Fogelin's A Defense of Hume on Miracles in Mind, Vol. 114, No. 453 (Jan., 2005), 145-149.
Or Elizabeth Anscombe's essay, "Hume on Miracles," in G.E.M. Anscombe, Faith in a Hard Ground: Essays on Religion, Philosophy and Ethics (Imprint Academic, 2008), chapter 4.
Perhaps, though, it was not his intention to offer a bibliography for reference, but only to include the titles he cites in the body of the text.
3. He quotes Michael Licona favorably on several occasions. In recent debates, Licona has been putting more emphasis on extrabiblical evidence for the supernatural to debunk the presumption against Biblical supernaturalism.
4. We might adapt Chisholm's distinction between methodists and particularists to the question of miracles. Larmer is a particularist. He begins with paradigm-examples. By contrast, "skeptics" are methodists who begin with a priori criteria which they invoke to preempt reported miracles.
5. On the traditional interpretation, Hume presents an a priori argument designed to render belief in miracles rationally inadmissible in principle. Revisionists interpreters claim that Hume's argument was less ambitious and tendentious. I'm not a Hume scholar, so I don't have an independent judgment to offer. What I will say is that the revisionist interpretation poses a dilemma. On the one hand, the revisionist interpretation makes Hume's argument less vulnerable to easy refutation.
But there's a tradeoff. On the traditional interpretation, the value of Hume's argument for "skeptics", if successful, is that it disables the argument from miracles at one stroke. On the traditional interpretation, Hume's argument is a shortcut, by relieving the "skeptic" of any burden to disprove specific evidence for specific miracles. So making Hume's argument more defensible comes at the cost of making his argument less useful to "skeptics," for if his objection was never intended to block the argument from miracles in principle, then the "skeptic" is forced to fall back on a case-by-case evaluation of reported miracles. But wasn't the primary advantage of Hume's contribution to sidestep that daunting task?
6. Larmer cites Berkouwer, whom he who identifies as a theologian in the Calvinist tradition, as a proponent of occasionalism. In the same context, he cites a secondhand quote from Kuyper, via Berkouwer. I'd point out that although Berkouwer began his career as a Reformed theologian, he liberalized his theology over the years so that there came a point where he was no longer a representative of Reformed theology. For instance, his 1955 monograph on election marked a turning point.
Occasionalism is an outlier in Reformed theology. For more representative sample, there is Warfield's essay on "The Question of Miracles," in Selected Shorter Writings of Benjamin B. Warfield, vol. 2, chap 12, John Frame's The Doctrine of God, chap. 13, as well as Paul Helm's discussion of deism and occasionalism in The Providence of God, chap. 3.
In fairness, Larmer isn't attempting to give a historical overview of Reformed theology on this topic, but just citing notable individuals as foils for his own position. Moreover, he later mentions Hodge as an opponent of deism and occasionalism in his definition of miracles.
7. Regarding the burden of proof, a crucial point that I think Larmer neglected to make explicit is that because naturalism disallows supernaturalism in toto, the bar is extremely high for the naturalist and extremely low for the Christian. Naturalism has no give. It can't tolerate a single miracle. Therefore, it only takes a few well-documented miracles to falsify naturalism. An atheist must discredit every single reported miracle whereas a Christian apologist need only establish a few.
8. "Skeptics" accuse Christians of fallacious reliance on the argument from ignorance. Yet "skeptics" deploy the "argument from ignorance" when they justify disbelief in miracles on the grounds that, to their knowledge, there are no well-documented cases of miracles.
9. A stock objection to miracles is that it makes God an incompetent engineer whose rollout is plagued by failure to debug the prototype before he launched. Larmer addresses that objection.
I'd point out, however, that if open theism is true, then God might have to resort to midcourse corrections, due to unforeseen developments. I'm not sure what Larmer's position is on open theism. He's a freewill theist who approvingly quotes Hasker and Pinnock in the course of his monograph.
10. Apropos (9), critics of miracles who treat the clockwork universe as their ideal fail to distinguish between creation and the fall. It's not a design flaw if a watch needs to be repaired because it was damaged after it left the factory or the jewelry store.
11. Larmer is a freewill theist, and he frequently appeals to freewill theism in his defense of miracles. As a rule, if a philosopher can defeat or undercut a position by offering either a more ambitious argument or a less ambitious argument, it's preferable to use the less ambitious argument inasmuch as defending a less ambiguous argument is less intellectually demanding. That would expose less of his flank.
With that in mind, rather than attempting to attack determinism in general, it would be more prudent for Larmer to narrow his objection to the kind of determinism espoused by naturalism. That's typically blind physical determinism.
Although Larmer disagrees with predestinarian theological traditions (e.g. Thomism, Calvinism), objections to blind physical determinism don't equally cut against just any kind of determinism, viz. substance dualistic Calvinism. It isn't necessary for Larmer to engage determinism in general to engage naturalism in particular. Take the observation by freewill theist Richard Swinburne:
It has been argued that any argument for determinism would be self-defeating. For suppose a scientist discovers an apparently cogent argument for determinism. He will conclude that he has been caused to believe that his argument is cogent. But when we discover of people that they are caused to hold beliefs—e.g. as a result of the way they were educated, or of subjection to drugs—we do not regard them as having a rationally justified belief. To be rational in adopting a belief we have to do so freely, i.e. uncaused, the argument goes. So no one can ever be justified in believing determinism to be true. For one who believes determinism to be true must believe his belief to be caused and so unjustified. (There is a statement of this argument, subsequently retracted, by J. B. S. Haldane in his Possible Worlds, Chatto and Windus, London, 1930, p. 209. For references to other statements of it, including one by Epicurus, and discussion thereof, see K. R. Popper and J. C. Eccles, The Self and its Brain, Springer, New York, 1977, pp. 75 ff.) This argument has, I believe, no force at all. The mere fact that our beliefs are caused is no grounds for holding them unjustified. Exactly the reverse. I argued in Chapter 7 ["Beliefs"] that to the extent that we regarded them as uncaused or self-chosen, we could not regard our beliefs as moulded by the facts and so likely to be true. The point is rather that if we see some belief to be caused by a totally irrelevant factor (e.g. a belief that I now am being persecuted being caused by something irrelevant in my upbringing) then we rightly regard it as unjustified. But a belief that determinism is true could be both caused and justified, if caused by relevant factors, e.g. hearing relevant arguments.The Evolution of the Soul (OUP, rev. ed., 1997), 233n2.
12. Regarding the quality of witnesses, here's an interesting test-case. Take the news report of pious Catholics who venerated a bank window because the bank window sometimes had a pattern that resembled the Madonna, which they took to be a Marian apparition.
On the one hand, the witnesses are stereotypically credulous, superstitious religious believers. Wishful thinking combined with the conditioning effect of Catholic iconography induce them to bear witness to a miracle. Yet this is clearly a coincidence. An optical illusion caused by the angle of the light and the angle of the window at a particular season or time of day. There's nothing naturally inexplicable about the phenomenon.
But consider what they witnesses got right as well as what they got wrong. They aren't liars. They aren't hallucinating, whether individually or collectively. On occasion, the bank window does have a reflection that bears an adventitious resemblance to the Madonna.
So here we have an important distinction between what they see and what they think they see. What they see exists outside their minds, but what they perceive only exists in their minds. Anyone in the same physical position could observe the same pattern. That has nothing to do with religious predispositions. They're not mistaken about the evidence, but their interpretation of the evidence.
13. Larmer says the argument from ignorance is sometimes justified. Let's consider a few examples. Classifying some species as extinct is an argument from ignorance. If there's no evidence that the Irish Elk still exists, it's classified as an extinct species. But that's hardly fallacious.
Or take the Loch Ness monster. Lack of adequate evidence is sufficient to doubt its existence. Same thing with Bigfoot.
Ironically, atheists sometimes compare reported miracles to reported sightings of Bigfoot, the Loch Ness monster, &c., but aren't they guilty of resorting to the argument from ignorance? It's either a fallacy for believers and unbelievers alike, or it's not a fallacy, per se.
14. Larmer says inferring incorporeal agency (God, angels, demons) is analogous to inferring elementary particles. Another example would be abstract objects. If they exist, they lie outside of space and time. Incapable of direct empirical confirmation. Yet it's rational to posit abstract objects based on the indispensable explanatory role they fill.
15. Larmer is critical of Cardinal Kasper's claim that miracles are incongruous with divine transcendence. While I agree with Larmer's response, we could make another point. Kasper is a Catholic theologian and prelate. Unlike an outright atheist, Kasper must muster some pious-sounding excuse to rationalize his disbelief in miracles.
16. Larmer conjectures that the inequitable distribution of miracles may be due in part to the fact that God's will is sometimes thwarted by unwilling humans. But even if we grant freewill theism for the sake of argument (which some readers will dispute), that's not a very convincing explanation. For even if God would rather work a miracle through human instrumentality, God may have more than one human vehicle to choose from–and failing that, God retains the fallback option of direct divine agency. I offer a different explanation further down.
17. Larmer is critical of how Swinburne counterbalances the posterior probability of miracles against their prior improbability. That may suggest that Larmer doesn't think Bayesian probability theory is a good framework for the argument from miracles.
To take an example, consider the 9/11 attacks. In theory, we could begin by assessing the mathematical odds of two passengers planes, within minutes of each other, colliding with two adjacent skyscrapers. We could try to calculate the total number of airplanes and total number of skyscrapers over a period of decades, and lay odds. We could then ask why quality and/or quantity of evidence would be needed to overcome the astronomical improbability of that event.
But of course, no one puts these two sets of facts on either side of the scale, then wait to see which one tips the scale. No, we simply go with the evidence that airplanes struck the Twin Towers.
In addition, the abstract odds are irrelevant, because this wasn't an accident.
18. Larmer admits that we must use some evidence to assess other evidence. On the face of it, that appears to be viciously circular. What justifies taking some evidence as the standard of comparison?
Perhaps one justification would be that this isn't absolute. We can try this with different samples. We can treat one sample as the frame of reference and see how that works. We can then treat a different sample as the frame of reference and see how that works. We can compare and contrast different samples.
One might also appeal to certain "truths of reason" for guidance.
19. On the question of non-Christian miracles, I'd like to take one hypothetical example. Abraham's ancestors were heathen. Abraham himself was heathen before Yahweh disclosed himself to Abraham.
It stands to reason that some of Abraham's lineal ancestors faced life-threatening conditions. Mortality was high in the ancient world. It stands to reason that they prayed to their pagan gods for healing. And it stands to reason that Yahweh might perform a life-saving miracle for one of Abraham's linear ancestors, since Abraham wouldn't ever exist if one of is forebears died too soon. That miracle would not be for the benefit of the immediate recipient, but for Abraham, maybe generations down the line.
20. Larmer discusses coincidence miracles, but I don't think he succeeds in getting to the nub of the issue. Here's one definition:
It is important to emphasize that in spite of the widespread belief to the contrary, an event may be the source of marvel and elicit genuine religious response, not only without violating any natural law, but even if all its details may be explained by known laws. As long as an event is genuinely startling and its timing constitutes a mind-boggling coincidence, in that it occurs precisely when there is a distinct call for it to promote some obvious divine objective, then that event amounts to a miracle. The promotion of a divine objective may take many forms: it could be a spectacular act of deliverance of the faithful from the evil forces ranged against them, it might come as a highly unusual meteorological event through which the priests of Baal are discredited, or it might appear as a swift, clear, and loud answer to the prayers of the truly pious. However, whatever form the wondrous event takes, it should have a religious impact on its witnesses. George Schlesinger, “Miracles,” Quinn & Taliaferro (eds.), A Companion to Philosophy of Religion (Blackwell, 1999), 398-99.
What makes this miraculous is not that it circumvents natural processes, but that the outcome is too discriminating, too opportune, to be the result of natural processes alone. Natural processes are uniform. They aren't directed to benefit anyone in particular.
21. Larmer defines a miracle in terms of supernatural "intervention". I'm not sure what he means by that. He uses that framework in opposition to deism and occasionalism, but it raises other question. Does he think God intervenes in the sense that God is a temporal agent in history?
To state my own position, I use divine "intervention" to express the counterfactual truth that some things won't happen absent prayer, some things won't happen if nature ran its course. Miracles and prayers make a difference in that sense. It doesn't mean God is rewriting the plot.
Take a film in which, at one level, the director causes everything. He doesn't "step in" to change the plot in midstream, because he wrote the plot in advance. He's scripted every scene.
However, a film involves an interplay between personal agents and their physical environment. Things happen as a result of human interaction that would not occur in crystal formation.
Likewise, the director can write a "coincidence" into the plot. Timely, opportune meetings between one person and another, or a character and something he needs at that very moment. This doesn't require the director to introduce "breaks" into the continuity of the plot. Rather, they reflect the coordination of otherwise independent chains of events to achieve an intended goal. Something beyond the ability or ken of characters inside the story.
22. A stock objection to miracles is that a miracle is just a coincidence which believers misidentify due to sample selection bias. Larmer attempts to field this objection. I think his response is rather weak.
i) The first thing that needs to be said is that this by no means unique to Christianity. To the contrary, the need to distinguish a coincidence from what is not coincidental is crucial in many different fields and walks of life. Most people don't have sophisticated criteria.
ii) One attempt to provide a rigorous criterion is Dembski's specified complexity. Given Larmer's sympathy for intelligent design theory, or at least criticisms of methodological naturalism by intelligent design theorists, it's odd that Larmer doesn't appeal this principle to help differentiate a miracle from a coincidence.
iii) Finally, here's an older work making the point that repetition is not a necessary criterion to distinguish a coincidence from what is not coincidental:
The order of the phenomena is not a phenomenon. That order is only grasped by the mind; it is an intelligible relation between the phenomena, of which, however, we seek the explanation quite as much as the phenomena themselves. Take the fall of a stone, it is explained by the law of gravitation; le there be a second fall, it is explained by these same law. But let ether be a hundred falls…yet these hundred falls will not longer admit of being explained by the repetition a hundred times over of one and the same cause; and a mind which should not be capable of remarking this agreement of phenomena,and which should continue to explain them indefinitely by the same cause, would on that very account appear to us struck with imbecility. [It would be like] that that man of whom Gassendi speaks, who, half-asleep, and hearing four o'clock strike, say, This clock is mad; lo, four times in succession it has struck one o'clock. Paul Janet, Final Causes (T & T Clark, 1878), 27.But yet one more: what is there here more than in a hundred separable falls? Nothing but their convergence or simultaneity.Repetition…would be insignificant if it merely had reference to the number of facts (since we are always equally remote from the infinite)…A single experiment [may] suffice for proof, because it is such a coincidence as could scarcely occur even once, had it not its own reason in the laws of nature. this is what causes great scientists rarely to mistake the worth of a significant fact, though occurring only once. The Abbé Haüy lets fall a piece of quartz, and merely by observing the fracture, he at once concludes that he has discovered a law of nature; for what is the likelihood that a mineral should break by chance according to the laws of geometry? So in a thousand cases. The knot [of the inductive problem], then, is not in the repetition itself, but in the fact of the coincidence. Ibid. 460-61.
23. Another objection to miracles is that the inequitable distribution of miracles is incompatible with divine benevolence. Larmer responds to this objection, but I'd like to make a few additional observations:
i) A world in which miracles are evenly distributed will have a different world history. Moreover, the world histories increasingly diverge the earlier you change a variable. Different people will be born into a world in which everyone is saved. Mating and procreation depends on timing. Who you meet. That depends on when and where you were. Same thing with procreation. It takes very little to throw that off.
At that point some critics say people who never exist in the first place have nothing to lose. That, however, embroils them in an Epicurean dilemma. I think most philosophers wish to reject the Epicurean symmetry between prenatal nonexistence and posthumous nonexistence.
Ironically, a Christian could accept the symmetry, but reverse the assessment. Rather than taking that to mean if nonexistence is no misfortune at one end (prenatal nonexistence), then it's no misfortunate at the other end (posthumous nonexistence), we can logically take it to mean nonexistence is misfortune at both ends.
For instance: "There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse (William Rowe)."
But that's ambiguous. Does Rowe mean God could preserve the same goods without the attendant evils, or equivalent goods? What if preventing some evils prevents attendant goods?
Let's take a comparison. Many atheists think death is bad. Murder is bad. And they think premature death is worse than dying at a ripe old age.
By contrast, Epicurus and Lucretius posited a symmetry between prenatal nonexistence and posthumous nonexistence. As Mark Twain put it, "I do not fear death. I had been dead for billions and billions of years before I was born, and had not suffered the slightest inconvenience from it."
Consider how secular philosophers struggle with this issue. Cf. J. Fischer, ed. The Metaphysics of Death (Stanford University Press, 1993). One way to cash out the intuition, according to secular philosopher Thomas Nagel, is appeal to the principle of deprivation as well as counterfactual goods. The argument goes like this: death is bad because death is an experimental blank. And that's bad because it robs the decedent of the goods of life. Had he died later rather than earlier, he'd enjoy more of life's goods.
By that logic, both prenatal and posthumous nonexistence are experiential blanks that deprive one of life's goods, including missed opportunities. And if it's worse to die young, worse to cease existing at an earlier age, then it's even worse not to exist in the first place.
By the same token, we don't simply regret actual goods we lose, but lost opportunities for desirable goods.
The problem that poses for Rowe's argument is that preventing certain evils prevents certain lives. And that's a loss for them. Indeed, total loss. They never had a chance to enjoy life's goods.
Even if that's offset by countervailing goods, the people who don't exist in that alternate history aren't compensated for their loss. Rather, other people who take their place are the beneficiaries.
ii) Take a different example: suppose you have two neighborhood boys of the same age. One is disabled. The other boy is a high school athlete. He's hoping for a football scholarship to pay for college. Suppose, if God heals the disabled boy, he becomes a competitive athlete who gets the scholarship instead. The miracle is beneficial to the recipient, but harmful to the other boy.
iii) On a related note, one miracle can impact more than one person. So the apparently inequitable distribution of miracles may be superficial in many cases, because we're looking at the situation as if these are discrete, self-contained events, in a one-to-one relationship between the miracle and the recipient, whereas they may often have a one-to-many relationship down the line.