The limits of science

If Richard Dawkins has a fault (and who among us hasn't after all), it is not that he is often wrong, but that he is intolerably right. It strikes anyone who is capable of thinking and is not religious very quickly that believers are willing to take on faith a lot of stuff that is not even slightly credible, but we are mostly too polite to say so. Dawkins is a bit like the boy who couldn't stop himself yelling that the emperor was nude though, and yell he does. It enrages the believers, who proceed to say things that are so silly they make Dawkins look all the more rational and sane. The review of a recent book by Alister McGrath in the New Scientist (at this URL but thanks to the old-school model of the NS, you cannot actually read the review for free online; I have not read and will not be reading the book, or for that matter, the Dawkins book that was the stimulus for it) was in equal measure pompous and misguided. I'll be looking at just one paragraph, because this is representative of the Christian challenged to Dawkins.

The argument [in brief, that progress in science poses the question why the universe is explicable] also puts pressure on Dawkins, if not to believe in God, then at least to consider the possibility of the faith-based nature of his own convictions. As Dawkins acknowledges and physicists have shown, the existence of conscious, rational beings is a wildly improbable outcome. To insist that we are simply the products of the workings of, ultimately, physical laws is to avoid the question of the nature and origin of those laws. To say that there is no evidence for God is merely, therefore, an interpretation justified in one context but quite meaningless in another. Everywhere we look, there is evidence of something, but it is by no means clear that that something is, in fact, nothing. Rather, it seems something of a startling intelligibility.

On the face of it, this is a decent argument, but it is a sneaky version of Paley's watchmaker.

Let's look at what our doughty reviewer actually says:

The argument [in brief, that progress in science poses the question why the universe is explicable] also puts pressure on Dawkins, if not to believe in God, then at least to consider the possibility of the faith-based nature of his own convictions.

Bluntly, how can Dawkins look at a universe that is ordered and not believe that something ordered it? He must be relying on faith to be convinced that it isn't. This is a very common argumentative move from creationists, who suggest that Darwinian evolution is believed by scientists as an article of faith. However, scientists believe in Darwinian evolution for much the same reason they generally believe in anything they believe in: it fits the facts well. It's quite simple. We believe random, undirected processes can create order because we can observe them doing so. The processes we consider can create our universe without the need of an ordering intelligence. That's not to say there isn't one; rather, it says we do not need one. One is reminded of Laplace, on presenting his masterwork to Napoleon, and having Napoleon ask him, sir, I see no mention of God in your work, and Laplace's replying, sir, I have no need of that hypothesis. Dawkins is resolute in his scepticism. He will believe what there is evidence for, and there is none for God. God is an extra, an additional layer. Science does not care much for additional layers. Is Dawkins' conviction that there is no God faith based? Yes, I think you can say it is. Dawkins has faith in science and its methods. He believes that they can in time describe the world and everything in it entirely. If you were to say to Dawkins that there is more to the world than science can explain, he would ask you, well, what then? After you are done handwaving and talking about spirituality and that blather, he will point out that none of the stuff you believe is there can be measured or replicated. Now does that mean it doesn't exist or does it mean that it is just personal to you? Arguably, these are both the same thing. For a thing to exist, it must exist separately from you. If it does not, how can it be said to exist and not simply be something you have imagined? (I can imagine a god, no problem, but that does not make it exist, does it? In any case, if it does, this is too broad a meaning of "exist" for Dawkins' argument. He is not saying that God does not exist for you. He is saying that it does not exist in the sense that observables exist: being something that you can share, point to, replicate the experience of: in other words, existence in a scientific sense, a factual sense if you will.) In any case, these two faiths are quite different: on the one hand, a Christian has faith that this thing that exists for them has the other kind of existence -- that it really is a thing that can be shared, pointed to, measured; on the other, Dawkins' faith is that science will work, which is more like having faith that your car will start because you have known it to start in many conditions (whereas the Christian believes his car is a flying machine, or can travel through time in the right circumstances -- in other words, that it is capable of something that it has not been observed to do or has not suggested it can do).

In this connection, here's something to think about when considering "spirituality", or other intangibles that one believes are beyond science. When I go to the coast back home in Cornwall, and I'm blown away by the cliffside scenery, is that not "spiritual" in a broad sense?

But what created the scenery? Not God! The sea. The sea ate away at the coastline for millennia, grain by grain, to make the beauty I feast on. (I do not say that God cannot be responsible for the sea, directing it, but I will claim that I can apply Occam's Razor and put it down to sea alone.) The sea is effectively working at random. It does not choose its path, where to apply its power. The crenellations in the coastline might each have their origin in just one weak rock, crashing down and creating a weak point in the line, at which the relentless sea could pick. And which rock will break at which time is effectively random (determinable if you knew exactly where everything started from).


As Dawkins acknowledges and physicists have shown, the existence of conscious, rational beings is a wildly improbable outcome.

Clearly, the reviewer is not keen on "Chinese odds". These are a fun way of looking at things. What was the chance that I would be born? A hundred per cent. Because if I wasn't, I wouldn't be here asking about it! More usually, Chinese odds say that there's a 50 per cent chance of everything: it either happens or doesn't.

The point is, however improbable the outcome, we're here talking about it. Probability is only really a useful tool to understand what might happen, not what has happened.

This is not to fall foul of the anthropic principle. I'm prepared to accept that the universe is not just so just so I exist, but of course it's true that if it were not just so I would not be here to discuss it. (The anthropic principle is slightly more powerful than this because there were many ways the universe could be, but it is rather finely tuned to produce us.) But it is like an antifallacy to think that it's meaningful that a long shot has come in. (The corresponding fallacy is that outcomes were inevitable. In hindsight bias, it seems obvious that a process should have had the outcome it did. We just knew that horse would win; we knew that we would lose the hand if we called that all-in.) The antifallacy says, not that it was obvious now we look back that it should have been that way (in other words, that it is not meaningful that it was a long shot) but that it is not obvious (in other words, that it is meaningful that it was a long shot).

It should be fairly clear that both the fallacy and the antifallacy are equally possible. One can argue that it's obvious that the universe should be fine tuned because here we are and that it's astonishing that it's fine tuned because here we are. Neither is convincing! It is not obvious that we should have evolved just so; many other outcomes were possible. But it is not particularly striking. The universe is enormous and permits many possibilities; there may even be many universes, not all suitable for our kind of life (some may not have had an inflationary period in their early life and might only be inches across; others might have certain constants at a slightly different level and have collapsed, or never have seen the creation of heavier elements, so that without carbon, there could be no carbon-based life).

Besides, the reviewer's argument has the unspoken premise that we are special, that thinking beings are special. But we don't know that. We seem special to ourselves (particularly because our foremost mythology has it that we are special creations of our god) but from the outside, we could seem to be meaningless whirlings of atoms, throwing up emergent patterns that we overinterpret.


To insist that we are simply the products of the workings of, ultimately, physical laws is to avoid the question of the nature and origin of those laws.

As someone who does insist that we are the products of physical laws, I feel up to answering this charge. The problem the reviewer has here is strictly conceptual, an outcome of the limitations of language. If I make laws for my children, I set rules that I wish them to obey. In the same way, societies set laws for their members to abide by. Laws are stronger than "rules" in the sense that they are considered to be motivated by an underlying sense of order, or one could say that they are thought to emerge from a deeper basis in principle or concept and are not arbitrary (that's not to say that laws cannot be arbitrary, only that they are conceptualised as not being). In any case, both the laws I make for my children and the laws a society makes for its members are made by humans for humans. They are codified behaviours (or antibehaviours).

Physical laws are different. They are not created by humans; they are merely discerned by them. They too are not arbitrary -- they cannot be because they are not invented or created at all. But the difference is deeper than that: bodies do not obey the law of gravity because someone has made a law and they must follow it; they obey the law of gravity because it is simply a description of what bodies do. Both types of law describe what should happen, but the senses of "should" are different. In the one case -- the human law -- it denotes an obligation, a demand; in the other -- the physical law -- it denotes an expectation. (Compare these two sentences "Trains should be clean and tidy" and "The train should arrive in the next few minutes".)

We do not need to question the nature and origin of physical laws because we are aware that they do not have any origin outside us, and their nature is simply to be the things we have noticed that things have in common. They are "laws" because the commonality is so broad. But the broadness is simply an outcome of the obviousness of the commonality: for instance, a body falls towards another because of gravity, and all bodies will do so because bodies do not differ from one another in kind in this sense: they are bodies, made ultimately of the same stuff. That is all that laws in physics say: things are made ultimately of the same stuff so will ultimately be bound to do the same things.

If the reviewer was able to grasp that laws in a broader sense do not need to be imposed or policed, he would see that his question is meaningless. Physical laws do not need an origin: they are descriptions, not prescriptions.


To say that there is no evidence for God is merely, therefore, an interpretation justified in one context but quite meaningless in another.

"Therefore" is often a leap. Sometimes it's a step so short that it is barely needed; but sometimes it is a huge spring into the dark. Here it is impossible. How the reviewer went from Dawkins' not considering the origin of physical laws to saying that there being no evidence for God is purely contextual is beyond me. (Although of course I do understand what he means: he is appealing to Gould's notion of nonoverlapping magisteria -- he wishes to suggest that science can only catalogue nature's laws but cannot inquire into their origin. As I've noted, one has to believe that they have an origin outside ourselves to think that more than one "context" even exists here.) I know that what the reviewer wishes me to understand is that Dawkins believes there is no evidence for God because he will not consider where all these laws came from. But Dawkins would be nonplussed by this line of argument. The laws of physics just are. If you throw a ball into the air, it falls to earth because things fall to earth. There is no deeper foundation for it. It is just what things do. They do it because bodies attract one another. They just do. (One can discuss the mechanism but there is no purpose for it, it just is what things do.)

There is, in this reviewer's muddle, an interesting point, and a problem with Dawkins' position that even I, broadly sympathetic, find in need of an answer. Start from Gould's magisteria. We have the realm of science, all the things it can talk about, all the areas it can be meaningful in. And then we have the realm of religion, those things beyond science. In Gould's schema, each could not speak about the other's realm (they had nothing to say about it). (So for instance, science might describe how a foetus comes into being, but only religion can say whether or when it is alive.) Now one can argue that science's realm is broad enough that it includes all of life and there is no need for the magisterium of religion.

But the problem is, one cannot say that the other realm is empty. One can believe that science includes everything but you cannot know it. Science cannot measure the magisterium of religion, and as a consequence cannot declare it empty. So the reviewer is saying, even if you are right that science seems to cover all there is, you cannot say that there is nothing at all in religion's realm. So your universe might not need God but that is not sufficient to say he does not exist. He might be there in the realm that you cannot measure.

Which is another way of saying, you cannot prove the negative.

Everywhere we look, there is evidence of something, but it is by no means clear that that something is, in fact, nothing. Rather, it seems something of a startling intelligibility.

Tis the clarion call of intelligent design. Ooooh, there's what looks like a watch. Must be a watchmaker.

But I think we see the watch because we are products of the same mad, random universe, and the way we are made makes us pick out patterns, samenesses, structure from even the entirely incoherent.

There are laws because we notice what we notice. Things do just what things do. They are not directed or aware: a law is not something they obey but something that they have just done.

Still, if we have created the watch out of what there is, in a sense we will create the watchmaker out of the same. Our god matches what there is -- and just as what there is cannot live outside the laws of science, neither can our god. He must be something we can imagine. For me, there is the remaining possibility that the realm of religion is not entirely empty, simply because science has no way to know whether its own bounds are broad enough to exclude any other explanations, any additional (in the strict sense of adding something) analyses, religion. Although science has squeezed God into what seems to us to be a vanishingly small margin, science cannot know how broad the margin truly is.

posted by Dr Zen @ 7:58 pm