Home » Library » Modern Library » Nicholas Tattersall Miracles

Nicholas Tattersall Miracles

A Critique of Miracles by C. S. Lewis (2000)

Nicholas Tattersall



More than thirty years after his death, C. S. Lewis is still world-famous as a literary critic, as an author of children’s fiction and as a Christian apologist. His works have been translated into many languages and they are widely read and cited. According to the blurb on the back of one publication of Lewis’ book Miracles, “Lewis challenges the rationalists, agnostics and deists on their own grounds and makes out an impressive case for the irrationality of their assumptions.” The aim of this essay is to provide an assessment of that claim.

Summary And Critique

Lewis begins his book by noting that people approach the question, “Has a miracle ever occurred?” with a spectrum of different presuppositions. At one end of that spectrum, naturalists will be highly sceptical of any claims that miracles have occurred. At the other end of the spectrum, resolute believers in the supernatural will be very much more inclined to accept a supernatural event has occurred based on relatively little evidence. According to Lewis, only when we have decided which position on the spectrum is the rational one can we decide whether the Christian claims of miracles (e.g. Christ’s resurrection) are at all likely. As Lewis puts it,


What we learn from experience depends on the kind of philosophy we bring to experience. It is therefore useless to appeal to experience before we have settled the philosophical question.[1]

Lewis tackles “the philosophical question” in the remainder of his book by arguing against naturalism and for supernaturalism. He concludes that naturalism suffers from “cardinal difficulties” and that supernaturalism is therefore the only reasonable position.

However, it is untrue that we have to decide between naturalism and supernaturalism before we can determine whether an alleged miracle has a high or low prior probability of being a genuine miracle (i.e. a genuine act of some supernatural being). In fact, it is clear that the prior probability of some alleged miracle being a genuine one is very low indeed. This is because there have been countless alleged miracles throughout history. Many of these alleged miracles have been shown beyond all dispute to be due to natural causes, while few or none of them have been shown beyond all dispute to be due to supernatural causes. So even if supernatural beings exist, they evidently only rarely produce miracles. From this, we can conclude that the prior probability of an unexplained event being a miracle is very low, even we suppose supernaturalism is true. Given all this, the rest of Lewis’ book loses some of its relevance. Even if Lewis were successful in his arguments for supernaturalism, he would not have established that Christian miracles are at all likely.

Before we can understand Lewis’ arguments against naturalism, we will need to know precisely what Lewis take the term “naturalism” to mean. He says,


What the Naturalist believes is the ultimate Fact, the thing you can’t go behind, is a vast process in space and time which is going on of its own accord. Inside that total system every particular event (such as your sitting reading this book) happens because some other event has happened. All things and events are so completely interlocked that no one of them can claim can claim the slightest independence from ‘the whole show'[2]

It seems to be the case that Lewis defines “naturalism” such that it entails determinism, the view that every event (except the first one, if there was a first event) has antecedent causes sufficient for its occurrence. Whether Lewis’ naturalism is compatible with the view that there exist abstract objects (things which are not in cause and effect relationships such as numbers, propositions or perhaps even moral rules) is never clarified. Exactly what it means to “go behind” the process in space-time is totally ambiguous. The only indisputable thing we can say about Lewis’ naturalism is that it entails determinism and the view that the universe exists uncreated.

Lewis’ definition of “naturalism” is bound to lead to confusion, since contemporary philosophers who defend naturalism do not take it to entail determinism. One of the big assumptions in Miracles is that either naturalism is true or else supernaturalism is true. Without that assumption, Lewis’ alleged refutation of naturalism would not demonstrate the truth of supernaturalism. But as is seen from Lewis’ definition, those two positions not constitute an exhaustive list. A person could consistently claim that Lewis’ supernaturalism and naturalism are both false and that some alternate view is correct. For example, it is perfectly meaningful to say, “Determinism is false and there are no supernatural beings.” This is precisely what many modern philosophers who call themselves “naturalists” do say.

Lewis’ terminological slip-up is a major mistake, since he has ignored the position held by many of his critics. Although Lewis never grasps this fact, there are places in the book where he comes close to noticing that something has gone seriously wrong. With regard the fact that certain interpretations of quantum theory predict that determinism is false, Lewis says,


Those who like myself have had a philosophical rather than scientific education find it almost impossible to believe that the scientists really mean what they seem to be saying.[3]

Scientific (non-supernatural) indeterminism cannot be fit into the classification system Lewis requires for his arguments, so he dismisses it out of hand!

Lewis goes on to present three arguments against naturalism. His first argument makes use of his claim that naturalism entails determinism. If determinism is true then all the events in the brains of determinists are caused by antecedent events. So when a determinist comes to believe “determinism is true” that was made inevitable by a chain of causes and effects which stretch back through time. According to Lewis, there can never be a reason to think that there is any connection between what comes out at the end of this cause-effect chain and what is true. So when a determinist says, “determinism is true” that only tells us something about the nature of the cause-effect chain, the laws of physics and the initial conditions of the universe. It doesn’t tell us about what is likely to be true. It follows, Lewis claims, that those who profess determinism can have no rational grounds for doing so. Determinism is a self-defeating belief. Lewis says,


For his [the naturalist’s] history is an account in Cause and Effect terms, of how people came to think the way they do. And this of course leaves in the air the quite different question of how they could possibly be justified in so thinking. This imposes on him the very embarrassing task of trying to show how the evolutionary product which he has described could also be a power of ‘seeing’ truths.[4]

Even though this argument is irrelevant to the naturalism/supernaturalism issue (since naturalism does not entail determinism), we might still wonder if it is sound. Lewis’ argument can be formulated as follows:


(1) If determinism were true, then every human’s thoughts would be completely determined by antecedent events.

(2) If every human’s thoughts were completely determined by antecedent events then no human would be able to make rational inferences.

(3) Therefore, if determinism were true, no human would be able to make rational inferences [from (1) & (2)]

(4) Humans are able to make rational inferences.

(5) Therefore, determinism is false [from (3) & (4)].

This argument is deductively valid, but its premise (2) is false or at least not established. It is unclear why Lewis believed it. Presumably that conditional is to be understood in terms of logical implication, but it is hard to see how both affirming the antecedent and denying the consequent involves a logical contradiction. There is no inconsistency in human reasoning being deterministic but rational. What is relevant is not whether human reasoning is caused, but whether it is reliable. Human thought-processes might be deterministic, indeterministic, reliable or unreliable. There is no obvious reason why human thinking being deterministic entails that it is unreliable.

For example, if we are presented with a computer program and asked whether it can reliably be used to process information and output true statements, the answer to neither of the following questions will decide the issue; (i) Will the program be run on deterministic hardware? (ii) Is the interaction of entities (e.g. atoms and electrons) within the hardware purely materialistic?[5]

According to most naturalists, the processes that go on in human brains (and in the brains of some non-human animals) have been shaped by millions of years of natural selection. According to one theory, we know how the capacity to make rational inferences was conducive to survival and reproduction in our ancestors, hence we know this ability is something that natural selection preserved. That is one way of explaining why humans are able to make rational inferences.

Lewis makes a half-hearted attempt to show such a theory would not work by advancing the peculiar idea that if naturalism were true, we would have evolved the ability to act as if we were rational without actually being rational,


A conditioning which secured that we never felt delight except in the useful nor aversion save from the dangerous might serve us as well as reason or in some circumstances better.[6]

It is unclear as to how organisms are to feel delight about what is useful or gain aversions to what is dangerous when in complex situations without ever making any rational inferences. For his suggestion to have any force, Lewis needed to provide a detailed account of how this mechanism might work, how it might evolve and why it would have been preferred by natural selection. But he gave no such account and so his suggestion has no force.

Another way that naturalists might refute Lewis’ argument is to make the more cautious claim that there exists some future scientific theory that explains why the ability to make rational inferences was preserved by natural selection. The onus is on Lewis to show that no such theory exists and he has not done that. In response, Lewis says,


This [i.e. the impossibility of a naturalistic account of reason] is best seen if we consider the humblest and most despairing form in which this could be made. The Naturalist might say, ‘Well perhaps we cannot exactly see – not yet – how natural selection would turn sub-rational mental behaviour into inferences that reach truth. But we are certain this has in fact happened. For natural selection is bound to preserve and increase useful behaviour’ But notice what we are doing. Inference is itself on trial: that is, the Naturalist has given an account of what we thought to be our inferences that suggests they are not real insights at all. If the value of our reasoning is in doubt, you cannot try to establish by reasoning. There can be no question either of attacking or defending it.[7]

In the paragraph above Lewis claims that “inference itself is on trial” and also that “There can be no question either of attacking or defending it.” If we cannot attack or defend the belief that rational inferences are possible, how can that belief be on trial or in doubt? There is little more than confusion and contradiction in what Lewis says here.

Worse still, in the paragraph above Lewis makes use of the objectionable principle that might be summarised, “If a person does not have an explanation for why some proposition is true, then he cannot reasonably believe that proposition to be true.” Specifically, Lewis assumes that if a naturalist cannot give an explanation for why humans can reason, then that naturalist cannot reasonably believe that humans can reason. But Lewis’ principle is obviously false, as can be demonstrated by examples. Consider the following; (i) grass is green; (ii) the electron charge is approximately 1.6*10-19 C; (iii) water expands when it freezes. I estimate that most people who believe (i)-(iii) know of no explanation for why those propositions are true rather than false. Does it follow that those people are unjustified in their beliefs? Of course not. A person who knows nothing about the scientific explanation for why grass is green can obviously have a justified belief that grass is green.

So to summarise my criticisms of Lewis’ first argument; (i) he erroneously supposes that naturalists are committed to determinism; (ii) he offers defective arguments against determinism.

Lewis offers a second argument against naturalism. This argument makes use of his claim that if naturalism were true then there would be no objective values. But according to Lewis, we know that there are objective moral values for two reasons; (i) there is widespread agreement that certain acts are morally wrong; (ii) ethical debate is possible. Of naturalists’ rejection of ethical objectivism, Lewis says,


But they must stick to it; and fortunately (although inconsistently) most real Naturalists do not. A moment after they have admitted that good and evil are illusions, you will find them exhorting us to work for posterity, to educate, revolutionise, liquidate, live and die for the good of the human race.[8]

This argument is irrelevant to the naturalism/supernaturalism issue since naturalism does not entail the view that there are no objective values. It is a gross misunderstanding on Lewis’ part that if a person denies that any supernatural entities exist then that person must deny that there are any objective values. Many naturalists can and do believe that there are objective moral rules.

However, we might still wonder if Lewis’ argument is sound. Lewis’ second argument can be formulated as follows:


(1*) There is widespread agreement that certain acts are morally wrong.

(2*) Ethical debate is possible.

(3*) (1*) & (2*) could only be true if ethical subjectivism is false.

(4*) Therefore, ethical subjectivism is false [from (1*) & (2*) & (3*)].

The problem here is Lewis’ premise (3*). Premise (1*) is compatible with ethical subjectivism since there can be widespread agreement in subjective opinions. For example, there is near unanimous agreement among humans that soap has an disagreeable taste, and yet that is an entirely subjective matter. Premise (2*) is compatible with ethical subjectivism too. For example, perhaps ethical judgements express our attitudes rather than describe them. In that case, we argue for our moral views in an attempt to cause other people to come to hold views and attitudes like our own. We care about the ethical attitudes of others because we care about how others behave. So even if there were no objective values, morality would not have to be an “illusion” as Lewis mistakenly claims. Nor would moral discussions be reduced to describing the speakers’ attitudes with no genuine disagreement between participants.

So to summarise my criticisms of Lewis’ second argument; (i) he erroneously supposes that naturalists are committed to ethical subjectivism; (ii) he offers defective arguments against ethical subjectivism.

Lewis’ third argument against naturalism is based on David Hume’s arguments for inductive scepticism.

In his Enquiries Concerning Human Understanding, Hume noted that in everyday life we suppose that events in the past tell us something about events in the future. For example, we assume that because all emeralds that we have so far discovered are green, this gives at least some support for the hypothesis that the future emeralds we discover will be green. But according to inductive sceptics, we have no justification for this mode of reasoning. Why is it that the colour of emeralds at time t0 should have any relationship (probabilistic or otherwise) to the colour of emeralds at some later time t1?

Of course, we could justify inductive reasoning if we had reason to think nature will always be uniform. But according to inductive sceptics, the only reason we have for thinking that nature will remain uniform is based on inductive reasoning (nature has been uniform in the past so will probably be uniform in the future). So it seems that any justification for induction we might give would be circular.

According to Lewis, naturalists have no justification for their reliance on inductive reasoning, while supernaturalists do have justification.


If Naturalism is true we have no reason to trust our conviction that Nature is uniform. It can only be trusted if quite a different Metaphysic is true. If the deepest thing in reality, the Fact which is the source of all other facthood, is a thing in some degree like ourselves – if it is a Rational Spirit and we derived our rational spirituality from It – then indeed our conviction can be trusted.[9]

This kind of argument has successfully been dealt with elsewhere[10] and I summarise the main criticisms as follows; (i) it has not clear that the arguments for inductive scepticism are decisive or even coherent; (ii) it is not true that rational predictions are only possible if inductive reasoning is possible; (iii) Christian theism, and certainly supernaturalism, are both compatible with inductive scepticism.

Since supernaturalism has now supposedly been established by three independent arguments, Lewis argues for a specific kind of supernaturalism. He argues that since the supernatural being is required to account for human rationality and objective morality (as shown by the first two arguments), it follows that it must itself be rational and moral. Lewis argues that pantheism (the belief that the natural world is identical with God) should be rejected because it is a very old belief.


Pantheism is congenial to our minds not because it is the final stage in a slow process of enlightenment, but because it is almost as old as we are. So far from being the final religious refinement, Pantheism is in fact the permanent natural bent of the human mind; the permanent ordinary level below which man sometimes sinks, under the influence of priestcraft and superstition, but above which his own unaided efforts can never raise him for long.[11]

However, it is not the case that a supernatural agent who creates rational and moral humans must itself be rational and moral. The principle needed to warrant this is, “If x creates y and y has property p then x has p.” But that is clearly false. If Dr Frankenstein created his monster from the body parts of different deceased humans, it does not follow that Dr Frankenstein himself was made from the body parts of different deceased humans. If God creates humans who are mortal, weak, irrational, and immoral, with dandruff and bad breath, it does not follow that God has any of those properties.

Furthermore, Lewis’ arguments against pantheism are unsound. It is absurd to suggest that a discussion of the age or history of pantheism can be of any assistance in determining its truth-value. There are also other views that come under the heading of “supernatural” that Lewis does not consider. For example, he gives no argument against polytheism.

So my conclusion is that even if we grant that Lewis’ earlier faulty arguments are sound, his later arguments for the attributes of the supernatural being(s) are no good.


The Church Times declares Miracles to be, “Dr Lewis’s most substantial and persuasive essay in Christian apologetics.” If that is true, then the arguments presented in Lewis’ other apologetic works need not be examined to conclude they are ineffectual.

In Miracles, Lewis never sorts out a clear and tenable distinction between naturalism and supernaturalism. Without argument, he supposes naturalists are committed to views they are not, yet his arguments against those irrelevant views are unsuccessful. Lewis wrongly supposes that Christian miracles are antecedently probable on supernaturalism and offers deficient reasoning in an attempt to show that any extant supernatural being must have the properties of the God of theism.


See also

A Response to Nicholas Tattersall’s ‘A Critique of Miracles by C. S. Lewis’” (2001) by Darek Barefoot.


[1] Lewis,1998, “Miracles,” Fount, p.1.

[2] Ibid., p. 6.

[3] Ibid., p. 13.

[4] Ibid., p. 21.

[5] On p. 14 Lewis’ argues against materialism (the view that everything that exists can be reduced to matter and energy) as follows,


Thus a strict materialism refutes itself for the reason given long ago by Professor Haldane: ‘If my mental processes are determined wholly by the motions of atoms in my brain, I have no reason to suppose that my beliefs are true and hence I have no reason for supposing my brain to be composed of atoms.’ (Possible Worlds, p. 209).

Even if this argument were sound, it would only show that it is impossible to rationally believe materialism to be true. It would not show materialism to be false. In any case, Haldane’s argument is hopeless for the reasons given above. There is nothing conceptually impossible about the interaction of matter in Haldane’s brain causing him to think rationally. If it is possible for the interaction of matter in his brain to cause him to think rationally, then it does not follow that his belief (i.e. his belief that all his beliefs are caused) is untrustworthy.

[6] Lewis, 1998, “Miracles,” Fount, p. 19

[7] Ibid., p. 21.

[8] Ibid., p. 38.

[9] Ibid., p. 110.

[10] Michael Martin. 1997. “Does Induction Presume The Existence Of The Christian God?” Skeptic Vol. 5, #2, pp. 71-75, also available at <URL:https://infidels.org/library/modern/michael_martin/induction.html>.

[11] Lewis, 1998, “Miracles,” Fount pp. 85-86.