© Michael Lacewing
Swinburne’s argument from design
Swinburne’s argument
Swinburne distinguishes between two types of order or regularity in nature (‘The Argument
from Design’, pp. 200–2). In regularities of ‘spatial order’, different things, e.g. parts of an
eye, exist at the same time in an ordered way, e.g. being organised to serve a purpose. But
there are also regularities of ‘temporal order’ – an orderliness in the way one thing follows
another. These temporal regularities are laws of nature. The design evident in nature, then,
is in the laws of nature themselves. Swinburne argues that the activity of a designer is the
best explanation of the operation of the laws of nature (p. 203).
The argument from spatial order to a designer is weak for two reasons. There is lots of
disorder in the universe; and science can explain the emergence of order from disorder, e.g.
through evolution. By contrast, the laws of nature are (nearly?) universal and, Swinburne
argues, there is no scientific explanation for why the laws of nature are as they are. Science
can’t tell us why the laws of nature are the way they are. It can only tell us that this is how it
is. For example, science explains why water boils when you heat it in terms of the effect of
heat on the properties of molecules. It explains these effects and these properties in terms of
other laws and properties, atomic and subatomic ones. Some further explanation of these may
be possible, but again, it will presuppose other laws and properties. Laws can be explained in
terms of more general laws, but that’s all. So the most fundamental laws, whatever they are,
can’t be explained by science.
Put another way, science must assume the fundamental laws of nature in order to provide any
explanations at all. It can’t say where they come from or why they are the way they are,
because all scientific explanations presuppose laws. Therefore, scientific laws have no
explanation unless we can find some other kind of explanation for them.
We use another type of explanation all the time, namely ‘personal explanation’. We explain
the products of human activity – this book, these sentences – in terms of a person, a rational,
free agent. I’m writing things I intend to write. This sort of explanation explains an object or
an event in terms of a person and their purposes. So we know of regularities in succession –
things coming about because someone intentionally brings them about – that are caused by the
activity of a person. The hypothesis that a designer exists and created the universe to include
the laws of nature provides a personal explanation for the laws of nature, and so for the order
of the universe.
So Swinburne argues (p. 205):
1
2
3
4
5
6
7
8
There are some temporal regularities, e.g. related to human actions, that are explained in
terms of persons.
There are other temporal regularities, e.g. related to the laws of nature, that are similar
to those explained in terms of persons.
We can, by analogy, explain the regularities relating to the laws of nature in terms of
persons.
There is no scientific explanation of the laws of nature.
(As far as we know, there are only two types of explanation – scientific and personal.)
Therefore, there is no better explanation of the regularities relating to the laws of nature
than the explanation in terms of persons.
Therefore, the regularities relating to the laws of nature are produced by a person (a
designer).
Therefore, a designer exists.
Even if the explanation in terms of a designer is technically the ‘best’ – because there is no
other – we still might not accept it if it isn’t a good explanation. One important criterion for
this is whether it is simple. Ockham’s razor says ‘Do not multiply entities beyond necessity’.
Swinburne’s explanation introduces a new entity – the designer. But if a designer is necessary
to explain the laws of nature, then the explanation respects Ockham’s razor. Swinburne
concludes that whether the argument succeeds depends on premises (2) and (3) – how strong
is the analogy between human action and the designer’s producing the laws of nature?
Swinburne’s response to Hume
Hume famously makes a series of objections to the argument from design (these are
presented more fully in the handout on ‘The Argument from Design: Paley v. Hume’).
1
2
3
4
5
6
The analogy is not very strong. The products of human design, such as a house or a watch,
are not much like nature or the universe as a whole. We cannot, therefore, reasonably
infer that the cause of nature is anything like a human mind.
Even if we could infer from part to whole, there is no good reason to choose design by an
intelligent mind as the explanation of the whole universe. Thought is a tiny, weak and
limited cause that only moves the bodies of animals.
We have experience of the products of human design being caused by a designer, but we
don’t have any such experience in the case of nature. So we can’t know the cause of order
in nature.
Hume’s fourth objection assumes his theory of causation. We cannot tell, from a single
instance of some object or event, what its cause is. (Think of one billiard ball hitting
another and the second moving away. The second ball’s movement could follow many,
many events – your breathing, someone walking about the room, a light going on… How do
you know which is the cause?) But the origin of the universe is unique. To make any
inference about the cause of the universe, we would need experience of the origins of
many worlds.
In order to infer that there is a designer of nature, we have to rule out other possible
explanations of the organisation of parts for a purpose. For instance, couldn’t finite
matter over infinite time by chance come to be arranged as if it were designed?
The inference of a designer is ‘useless’. When we infer from a cause to an effect, we
should only attribute properties to the cause that we need in order to explain the effect.
Anything else is mere speculation. With the designer of nature, all we have to go on is
what we already know – nature. We can’t find out about other designers or other worlds to
draw any useful conclusions about nature or the designer. So the hypothesis of a designer
adds nothing to our knowledge.
Swinburne argues that Hume’s objections are unpersuasive in the context of inference
to the best explanation.
So, why choose human reason as a model for explaining the laws of nature (objection (2))
rather than, say, chance (objection (5))? Because other causes of order in the universe rely on
the laws of nature which they don’t themselves explain (p. 209). Even Hume’s account of how
chance brings about spatial order assumes regular changes over time (p. 210). Could the laws
of nature themselves change randomly over time, sometimes producing regularity, sometimes
producing chaos? Perhaps, but given the amount of order in the universe, and the universal
applicability of laws of nature, this is a poor explanation of our experience. Explaining
temporal regularities in personal terms remains the best explanation.
Swinburne also disagrees with Hume’s account of inferring causes (objection (4)). Hume
objects that we can't reach conclusions about the causes of a unique object, such as the
universe. Swinburne replies (p. 209) that cosmologists have drawn all sorts of conclusions
about the universe as a whole. Uniqueness is relative to how something is described. For
example, many of the processes involved in the universe as a whole, e.g. its expansion, can be
identified repeatedly in other contexts. And Swinburne argues that Hume’s principle in
objection (6) – that we attribute to a cause no more than is needed for the effect – is
mistaken. He agrees with Hume that if this is all we could do, that would be fairly useless.
The cause would only be known as ‘whatever it is that produces this effect’. But clearly we do
more than this in science all the time, e.g. in supposing some cause is another subatomic
particle rather than a goblin.
However, Swinburne doesn’t say how we do this or why it is justified. The answer is that we
draw on what else we already know that is relevant, e.g. that there are subatomic particles
but no goblins. An inference to the best explanation is supported by how the explanation (the
cause) fits with what else we know. Hume’s objection is that in the case of the designer,
there is nothing else we know about the designer apart from nature. So we cannot draw any
informative conclusions.
A better response to Hume’s objection (6) is this: just being able to infer that the universe
has a designer – even if this knowledge is otherwise ‘useless’ – is itself important. If we can,
in addition, show that the designer has the traditional properties of God, then this is a
conclusion that changes our understanding of the whole world.
Is the existence of a designer a good explanation?
On p. 208, Swinburne refers to an objection of Hume’s in Dialogues, Pt 4. Hume, as Philo,
objects that for the designer to be a satisfactory explanation of design in nature, we would in
turn need to explain the designer. A mind is as complex and as ordered as nature, so if the
order in nature requires an explanation, so does the order of the designer’s mind. If we can’t
explain the designer, then it would be better to stop our attempts at explanation at the level
of nature.
Swinburne replies (just as Cleanthes does in the Dialogues) that this misrepresents
explanation. Science is full of examples of explanation that don’t explain what is assumed in
the explanation. Science will introduce an entity – like a subatomic particle – in order to
explain something, e.g. explosions in a nuclear accelerator. However, these new entities now
need explaining, and scientists don’t yet know how to explain them. This is absolutely
normal, and has happened repeatedly throughout the history of science.
If we will always have something we can’t explain, why invoke a designer? Why not just say
we can’t explain scientific laws? Because invoking a designer explains one more thing, namely
scientific laws, and we should explain as much as we can. This is a principle of science and
philosophy. If you give up on this, you give up on pursuing these forms of thought. So we can
still say that the designer is a good explanation for scientific laws even if we can’t explain the
designer.
Multiverse theory There is a recent popular development of Hume’s objection (5). Suppose
that instead of just this universe, there are or have been millions of universes. Each had
randomly different scientific laws, and in most cases, the laws didn’t allow the universe to
continue to exist – as soon as it began, it ended. Others existed for a time, but the
development of order was limited, e.g. there was no life. Given all the possible universes, and
all the possible variations in scientific laws, a universe such as ours would come to exist
through chance eventually. It doesn’t need any special explanation.
This explanation can be compared to winning the lottery. It is very unlikely that you will
win, but it is highly likely that someone will win. For whoever wins, that they won is a huge
coincidence; but we don’t need any special explanation for it (such as ‘someone intended
them to win, and rigged the lottery’). So, likewise, there is nothing special about this
universe, except that it has, by chance, the right laws for order to exist.
We can object that just as the lottery explanation assumes that millions of people are
playing, multiverse theory assumes the existence of huge numbers of other universes. These
other universes are completely inaccessible to us, and we have little evidence for their
existence (multiverse theory is also invoked to explain some phenomena in quantum
mechanics, but this is theory, not evidence). Is the hypothesis that there are millions of
universes better than the hypothesis that there is a designer? Two considerations:
1
We might think that the existence of a designer is a simpler explanation: just one designer,
not millions of universes. On the other hand, the designer is a new kind of thing, whereas
we already know that universes can exist.
2
If the designer is God, then God’s existence is also supported by other evidence, e.g.
religious experience. On the other hand, we also have evidence against the existence of
God, namely the problem of evil. At least we don’t have evidence against the existence of
other universes.
For discussion of whether we can infer that the designer is God, see the handout ‘The
Argument from Design: God’.