C. S. Lewis’ Miraculous Error

Posted on 2011/07/13


Picture 9

If I put six pennies into a drawer on Monday and six more on Tuesday, the laws decree that – other things being equal – I shall find twelve pennies there on Wednesday. But if the drawer is has been robbed I may in fact find only two. Something will have been broken (the lock of the drawer or the laws of England) but the laws of arithmetic will not have been broken.

Lewis, C. S. (1947) ‘Miracles’ ‘Chapter 8: Miracles and the Laws of Nature’ p 60


After C. S. Lewis’ conversion to Christianity in 1931 (when he was 33) his writings on the subject have achieved canonical status for Christian Apologists.

Contemporary writers and debaters such as Francis Collins, John Lennox, William Lane CraigAlister McGrath and John Polkinhorn (shall we call them ‘New Christians’?) frequently reference the arguments of Lewis in their debates and books. In some of these, Francis Collins’ ‘The Language of God’ for example, you can only find Lewis’ arguments and none of their own.

As a critic of Christianity, it would be too easy to suggest that the reason for Lewis’ dominant influence in their work is the same that lead them to Christianity in the first place. A worship of authority and a distaste for critical investigation. Christians are prone to believe the validity of Lewis’ arguments as readily as they accepted the doctrines these arguments are supposed to prove. It is often the case that in a public debate on religion, the Christians and atheists both adjourn thinking their representative was the victor. I think a contributing factor to this is that while atheists see Lewis’ arguments as transparently false, the Christians see them as opaquely true.

As I said, that’s all too easy. But, the more serious problem with Lewis’ proofs is that they are, mostly, invalid. And in order to maintain this claim, this is the introduction to a series of pieces I shall write in response to the main themes in Lewis’ Christian writings. With efficiency in mind, these are simultaneously responses to most of the writings of the above mentioned authors.

To start, I’ll refute Lewis’ arguments on the possibility of miraculous events, mostly contained within his 1947 book called ‘Miracles’. I will focus my attention on ‘Chapter 8: Miracles and the Laws of Nature’ as this is where Lewis makes most of his scientific claims. It is a chapter that demonstrates Lewis’ breathtaking misunderstanding of physics and the philosophy of science.

As an example of how contemporary apologists can be prone to quoting Lewis’ work in full, here is John Lennox who restates most of this chapter in a video clip from his website:

And here the relevant section of Lewis’ writing for you to compare:

This perhaps helps to make a little clearer what the laws of Nature really are. We are in the habit of talking as if they caused events to happen; but they have never caused any event at all. The laws of motion do not set billiard balls moving: they analyse the motion after something else (say, a man with a cue, or a lurch of the liner, or, perhaps, supernatural power) has provided it. They produce no events: they state the pattern to which every event – if only it can be induced to happen – must conform, just as the rules of arithmetic state the pattern to which all transactions with money must conform – if only you can get hold of any money. Thus in one sense the laws of Nature cover the whole field of space and time; in another, what they leave out is precisely the whole real universe – the incessant torrent of actual events which makes up true history. That must come from somewhere else. To think the laws can produce it is like thinking that you can create real money by simply doing sums. For every law, in the last resort, says ‘If you have A, then you will get B’. But first catch your A: the laws won’t do it for you.

Lewis, C. S. (1947) ‘Miracles’ ‘Chapter 8: Miracles and the Laws of Nature’ p 61

Notice that Lennox only mentions Lewis at 3:35 (when he launches into the quote at the beginning of this article) after lifting the entire passage above as it were his own. Indeed, Richard Dawkins has shown that Lennox is prone to being imaginative with quotations, catching him, in what Dawkins calls, ‘Mining the Eddington Concession’.

A Summary of Lewis’ Argument For The Plausibility of Miracles

We are to assume that a supernatural agent exists and then ask if it is possible for miracles to occur. Then, Lewis’ main thread is as follows:

  • The Laws of Nature are comparable with Mathematical Rules
If the laws of Nature are necessary truths, no miracle can break them: but then no miracle needs to break them. It is with them as with the laws of arithmetic. [p 60]

[The laws of motion] produce no events: they state the pattern to which every event (…) must conform, just as the rues of arithmetic state the pattern to which all transactions with money must conform (…). [p 61]

  • The Laws of Nature have nothing to say about particular events

(…) what [the laws of nature] leave out is precisely the whole real universe – the incessant torrent of actual events which makes up true history. [p 61]

  • God can ‘feed in’ new events into Nature without breaking its Laws

If God annihilates or creates or deflects a unit of matter He has created a new situation at that point. Immediately all Nature domiciles this new situation (…), adapts all other events to it. It finds itself conforming to all the laws. [p 62]

The divine art of miracle of miracle is not an art of suspending the pattern to which events conform but of feeding new events into that pattern. [p 62]

My Response

Lewis works under several faulty assumptions and false distinctions, some of which are only implied by the text. I have grouped the misunderstandings into the themes below.

The Distinction Between The Laws of Nature & The Theories of Physics

Lewis does not seem to have grasped that physics, as an active inquiry, is necessary because we do not know the laws of nature.

More technically, a Law of Nature is to understood as a regularity (a pattern) of the universe. So, part of the assumption of these law is that they describe everywhere and anywhere. These would obviously be useful to know – they would allow us to predict and control our world.

Here’s the vital question: how can we find out what the laws of nature are?

Suppose we have a physical theory such as ‘light travels in straight lines’. This is to say, we think light travels in straight lines anywhere and anywhen in the history of the universe. [These are often, for this reason, called ‘universal hypothesis’.] We might try to test this theory by cooking up cleaver experiments with lasers, mirrors and cameras. All with the plan of catching out the light when it doesn’t travel in a straight line.

How long would this process have to continue before we are certain light does travel in straight lines? A million experiments? A billion? Actually, it could never be proven. Nor could any other physical theory (universal hypothesis).

[In fact, light doesn’t travel in straight lines. Firstly, it was discovered by Einstein & Eddington that space and time are curved, so light, moving through curved space, seems to bend. Secondly, Quantum Electrodynamics (QED) works under the assumption that it is a meaningless question to ask which single path light travels when moving from source to detector.]

A finite, limited number of events that the theory adequately describes is not proof that the theory describes all events. This problem was discovered by David Hume in 1739 and is called ‘The Problem of Induction’.

And it remains a problem until you accept it as part of the human condition. This is our relationship with the universe. Then, it is possible to see scientific inquiry as a process of guessing theories and testing them to see if they are false. The solution, called Critical Rationalism, was found by Karl Popper in 1934.

Richard Feynman humorously put it thus:

Feynman, R. P. ‘Messenger Lectures’ Part 6

I will call these guesses ‘Physical Theories’ for the purposes of distinction (however, it is common for physicists to use ‘law’, ‘theory’ and ‘hypothesis’ interchangeably). Lewis makes no such distinction, and so talks about the Laws of Nature throughout his work.

This is a problem because there is a big difference between the possibility of a miracle occurring and the possibility of knowing a miracle has occurred. In fact, John Lennox states this himself in the previous video, although it would seem he does not understand its implications.

If you are going to recognise a miracle (and after all, a miracle is, by definition, something very much out of the ordinary), you’ve got to know what the corresponding regularities are, otherwise you couldn’t tell if the thing was a miracle or not.

John Lennox
http://johnlennox.org/index.php/en/resource/science_and_miracles/ 6:09

If we can never know the Laws of Nature, how do we know if they have been broken? Feynman put it well in his lecture when he said if ‘[a physical theory] disagrees with experiment: it’s wrong.’ To permit that a miracle has occurred is to do so arbitrarily. How do you know if a miracle occurred, or your physical theory was incorrect?

Lewis’ answer to this seems to be that the laws of nature aren’t in fact broken by a miracle. In doing so, he radically warps the usual understanding of a ‘Law of Nature’ as mentioned above.

Mathematical Rules and Physical Theories

Hopefully it is now clear that Physical Theories are things that cannot be proved from observation. Mathematical Rules, in contrast, are statements of equality that are true by nature of the meaning of the terms. Technically, they are called analytic propositions (or necessary truths), since what you end up with, on inspection, is exactly what you started with, said in a slightly different way. If this weren’t the case, then the rule would be invalid, and would be prone to faulty results.

So, the analogy between mathematical rules and physical theories is incorrect. Lewis does however suggest that Laws of Nature (as distinct from our Physical Theories) are necessary truths. Leaving aside how he knows this, he supposes that a Laws of Nature can be interpreted as saying ‘two sides of the account balance.’ He then goes on to say that ‘[w]hen we understand this we see that of course they must balance.’ [p 58]

On his first point, I am in agreement. The theories we have called conservation laws say exactly that – before and after, the numbers balance. So, he is also correct in saying that they are descriptions, not causal explanations. However, suppose that these theories are the true laws (a likely incorrect assumption as several conservation laws have been proven incorrect), it does not seem obvious that they then must be that way. Indeed, many people have speculated on what alternate realities would be like, most famously in George Gamow’s Mr Tompkins books.

The Content of Physical Theories

When Lewis suggests that the Laws of Nature on their own cannot tell us about actual history, this is to misunderstand that Physical Theories describe patterns but also provide initial conditions. If you know how something started, and the rules for getting from one moment to the next, you know everything about its history.

Lewis views the Laws of Nature as a sort of computer code to the the game that is the Universe. In it, all objects obey the rules laid down by the code. However, at any moment, the programming God can add a new object, which the code accommodates. Think of it like playing with a physics engine such as Algodoo:

What is wrong with this picture? Well, it makes an assumption that the laws of nature are separate from the events they describe. In fact, I find it very difficult to picture ‘Laws of Nature’ as floating independent statements. I see them as the totality of all events. It is us who construct Physical Theories in equations and sentences to summarize these events. I don’t know what it means for there to be Laws of Nature, independent of the objects they describe, or the people describing them. To think of them as independent in this way seems to assume some sort of ‘Mind of God’ where these statements dwell. I speculate this is what Einstein meant with his famous poetic phrase:

I want to know how God created this world. I am not interested in this or that phenomenon, in the spectrum of this or that element. I want to know His thoughts; the rest are details.

Albert Einstein
Clarck, Ronald, W. The life and Times of Einstein. New York: The World Publishing Company, 1971, p. 18-19.

More vital to the argument, our Physical Theories are guessed from certain aspects of physical events. They are relationships between measurable variables. Patterns of properties. How can they be separate from the objects they describe? Lewis suffers from a tremendous confusion of language, physics and philosophy on this topic.

‘Feeding In’ New Events

Feynman once again clarifies the physicist’s position on Physical Theories (and their distinction with Laws of Nature) in this beautiful analogy:

To help show exactly where Lewis has gone wrong, it is useful to extend Feynman’s analogy. We have our two players, who follow the rules of the game. In Lewis’s conception of a miracle, someone sneaks up on them and messes with the board. They could move one of the pieces from one square to any other square, breaking a rule. They could sneak a new piece onto the board , or remove a piece (as long as it isn’t a king!). All the time, the players carry on and make their moves in accordance with the rules of chess.

Now, here’s the problem. How do you tell if a piece has been sneaked on the board, or this is an example of some new rule you’ve never seen before? How can you tell our trickster has grabbed a piece a placed it in an odd spot, or that this is part of an undiscovered pattern?


What if God were to, say, flick Hitler’s head off (to borrow a phrase from Eddie Izzard)? It would falsify the law of conservation of linear momentum. If he did it with a newly created projectile, it would falsify the conservation of mass/energy. How would we tell if he had done this, or that our Physical Theories were in error and it had all occurred within the Laws of Nature? Without presupposing that which Lewis was aiming to prove, I think there is no answer to questions like these.

And if there is an answer, it certainly isn’t to be provided by C. S. Lewis.