Philosiphy

chapter 4

God Earl Conee

‘Religion is entirely a matter of opinion, of course, and you are as

entitled to your religious opinions as I am to mine.’ We’ve all

heard that. We may have said it ourselves. It seems to be a safe

and sensible judgment, until we stop trying to be so agreeable

and take it seriously. Then it starts to look like a premature

judgment, maybe even a dogmatic one.

When a disputed topic is entirely a matter of opinion, there is

no better reason to take one side than another. So if religion is

entirely a matter of opinion, then either the reasons for and

against any religious view balance out evenly, or there are no

reasons at all. But that’s not credible. Religious thinkers and their

opponents have offered lots of reasons and we have no good basis

just to assume that they always balance out.

Religion and metaphysics overlap on the question of God’s

existence. It is a metaphysical matter because part of metaphys-

ics, ontology, concerns the most basic kinds of beings. God is

basic. For one thing, God is the creator of the universe, if God

exists. It would be of tremendous metaphysical interest to learn

that at least all of physical reality depends for its existence on the

creative choice of one being.

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

Several lines of reasoning are aimed at establishing that God

exists. We’ll investigate three sorts of arguments where many of

the surrounding issues are metaphysical.

Getting It All Started

Effects

Our first version of an argument for God’s existence relies on one

fact about the world. The fact is that some things are caused to

happen. Many things that are taking place now are clearly effects

of various causes. This includes things that are happening to you

right now. You see these words as an effect of light that is beamed

to your eyes and you understand these words as an effect of your

learning English and applying your knowledge of it.

Okay, so there are effects. What is the connection to God’s

existence?

We next observe that the causes of effects are themselves

caused. Those causes in turn have their causes, and so on.

The reasoning from this point that gets us to God’s existence is

not supposed to rely on anything that we find out by observing

the world around us. We are supposed to see its force by thinking

about the relation of cause and effect. First we note that a

sequence of cause and effect might go back indefinitely. But

could it go back forever? The argument asserts that each causal

sequence must have gotten started. There must have been a first

cause that was not caused, to get each sequence into existence.

Thomas Aquinas was the major medieval proponent of this sort

of argument. Aquinas observed that, if you take away the cause,

then you take away the effects. We see effects. He inferred that

there must be some first cause of the existing effects. The

argument concludes that the first cause of all is the creator of

the universe, God.

God � 63

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

The argument proceeds in two phases. The premises of the

argument are the assumptions that it relies on. The conclusion

of each phase is the claim that is supposed to be proven by that

phase.

First Cause Argument

Phase 1

Premise 1: There are effects.

Premise 2: Any effect derives eventually from a first cause.

Conclusion 1: There is a first cause.

The second phase builds on the first. It starts with the conclusion

that Phase 1 is supposed to prove. It adds another assumption and

draws the First Cause argument’s final conclusion.

Phase 2

Conclusion 1: There is a first cause.

Premise 3: If there is a first cause, then it is God.

Conclusion 2: God exists.

We have to select ways to understand ‘God’ and ‘exists’. People

mean various things by ‘God’. Sometimes someone’s ‘God’ is

whoever the person idolizes. It might be the person’s favorite

musician. This meaning is no good for present purposes. It is not

a major metaphysical matter whether or not any given musician

exists (however major a musical matter it is). In contrast, it is

metaphysically huge to find out about the existence of a being

like this: a creator of the universe who is all-knowing, all-power-

ful, and morally perfect. We’ll understand ‘God’ so that God is a

being who is all-knowing, all-powerful, morally perfect, and the

creator, if ‘God’ applies at all. By using the word only for a being

with these extreme attributes, we make the issue of God’s exist-

ence a topic of metaphysical significance and we use a meaning

that is recognizable to those in Western religious traditions.

64 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

We’ll understand ‘exists’ in a broad way. ‘Exists’ applies to

anything that is in reality at all, whether past, present, or future,

whether in space or not. What ‘exists’ does not apply to are

merely apparent realities, the merely mythical, the illusory, the

fictional.

An argument relies on its premises. They must be entirely

reasonable to believe if the argument is to establish its conclu-

sion. If there is some serious unresolved doubt about a premise,

then the argument does not prove its conclusion.

Let’s consider Premise 2 (‘P2’ for short). The claim made by

P2—that any sequence of cause and effect must have gotten

started—holds a powerful grip on many people. It can seem

just obvious that a series of things must have a first one. This

grip loosens, though, when we try to spell out anything that

would justify this claim. Exactly why can’t each cause in a series

have its own cause, with no beginning?

‘No beginning’ must be rightly understood. It just means that

nothing is first in the series. There are familiar precedents for

this. The series of numbers known as the integers has no first

one. The integers include �1, which is preceded by �2, which is preceded by �3, and so on. The integers go back infinitely. This infinity is not mind-boggling. We don’t have to think of

all of the integers separately. We understand the infinity ad-

equately if we get the idea that each integer has a new integer

as its predecessor. This arrangement is an understandable way

for a sequence to exist while having ‘no beginning’.

We see how the negative integers are arranged. Why couldn’t

causes and effects be arranged that way too? Why couldn’t there

be causes preceding effects backward in time infinitely into the

past with no beginning? We cannot picture a whole infinite series

like that. But we cannot do this picturing just because we have no

way to picture the series’ ‘far end’, since it has none. We still do

understand the structure of the series, without a picture. So

again, what reason do we have to deny the possibility of an

God � 65

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

infinite series of causes and effects that is structured in the same

way? Nothing comes to mind.

This possibility undercuts the credibility of P2. P2 is supposed

to be worth assuming because we are supposed to see the need

for a first cause in order to have a causal sequence at all. P2 is

doubtful if we don’t see the need. And now we don’t.

Trouble for the First Cause argument does not stop there.

Phase 2 has a weakness as well, namely, P3. Suppose that a causal

series has some first cause. P3 asserts that the first cause is God in

particular. Why so?

Here is an answer: only almighty God is great enough for self-

creation. So God can exist without having something else as a

cause. Anything other than God has to have help in order to

exist.

This answer assumes that each thing has to have a cause. It

assumes that either the cause is something other than the effect,

or the cause and the effect are one and the same. The answer

claims that only God is fit to be a self-cause.

Why must each thing have any cause at all, though? It seems

possible that something just happens without being caused at all.

This possibility does not imply that anything is so powerful, or

otherwise magnificent, that it causes itself. For all we can tell by

thinking about causes and effects, it is possible that something

just does happen in nature, without a cause, and it starts a causal

series. Whether or not this ever actually happens, we don’t seem

to have any way to exclude it as impossible. So thinking about

causes and effects does not give us any good basis to accept the

claim made by P3 that any first cause is God.

There is a different defense of the claim that God is a special

sort of cause. The new idea is that God is so great that God does

not need to get caused into existence. In contrast, all lesser

beings require help in order to exist.

But what does greatness have to do with getting caused? Why

couldn’t some tiny insignificant particle just pop into existence

66 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

without anything making it exist, and then cause other things?

An uncaused first cause of that metaphysically minor sort ap-

pears to be possible. This appearance casts doubt on P3.

Maybe there is a need for God as the first cause that stems

from a need for explanations. We could not correctly explain why a

first cause just pops up, because there would be no explanation.

Is that an objection to the possibility? Yes, if we have some

assurance that everything has some correct explanation. The

claim that there is an explanation for everything is known as

the Principle of Sufficient Reason.

The Principle of Sufficient Reason requires an explanation for

the existence of any first cause. The principle also raises questions

about infinite causal series that do not have a first cause. Maybe

each item in an infinite series is explained as an effect of prior

causes. But according to the Principle of Sufficient Reason, that is

not all that needs explaining. The whole series is something too.

The principle requires an answer to the question of what explains

why the whole series exists. Thus, first causes and infinite series of

causes both require explanation, according to the principle.

A first response to the question that the principle raises about

infinite causal series is that the whole series may have a derivative

explanation. Perhaps when each event in the series has been

explained, the combination of all of those explanations explains

the whole thing.

That first response may seem fishy. Maybe each element in the

series causes the next one. But do those causal facts entirely

explain why that particular contingent series exists at all?

Suppose not. The Principle of Sufficient Reason requires that

there be some explanation. But what assures us that this principle

is true? When we think about how things might possibly have

gone, it seems possible that some things just do exist with no

explanation. Why not? The situation would be intellectually

disappointing. But what guarantee do we have that intellectual

satisfaction is always available? The Principle of Sufficient Reason

God � 67

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

declares that explanations always exist. Again, why believe it?

The lofty title, ‘Principle of Sufficient Reason’, doesn’t make the

principle true. And loftily labeled principles are plentiful. The

Principle of Insufficient Reason says that some things have no

explanation. The two principles conflict. Thinking about possi-

bilities seems to tell us that each of the principles might have

been true. Thinking about how things might be gives us no

reason to believe that the Principle of Sufficient Reason in

particular is the one that is actually true.

If nothing assures us that the Principle of Sufficient Reason is

true, then the principle does not help the argument. It does not

justify our denying apparent possibilities that go against P3. For

instance, it seems possible that everything started with the Big

Bang, rather than God, and the Big Bang has no explanation.

Until we have a sound basis for denying that any such possibility

obtains, P3 is in doubt.

Dependents

Here is an interestingly different version of the argument. The

new version is about a non-causal sort of dependence. Onto-

logical dependence consists in one thing needing another sim-

ultaneously, in order to support its existence. The idea eludes

precise definition, but it has one clear sort of illustration. Consider

a tuna salad sandwich. At any given time, the sandwich derives its

existence from the existence of the bread, the tuna salad, and any

other ingredients that compose it. Without them, it would be

nothing. The sandwich’s ingredients do not cause it to exist.

Rather, they give it existence directly. The sandwich ‘ontologically

depends’ on its ingredients. Anything that does not depend in this

way on any other entity is ontologically independent.

Using this idea of ontological dependence, the new version of

the argument otherwise goes just like the previous one.

68 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

Ontological Dependence Argument

Phase 1

Premise 1: There are ontologically dependent things.

Premise 2: Anything ontologically dependent derives its ex-

istence eventually from something ontologically independ-

ent.

Conclusion 1: Something ontologically independent exists.

Phase 2

Conclusion 1: Something ontologically independent exists.

Premise 3: If something ontologically independent exists,

then God exists.

Conclusion 2: God exists.

The claim made by P1 about the existence of ontological de-

pendence is fully credible. Many things, such as a tuna sandwich,

illustrate its truth. P2 is supposed to be true because an endless

sequence of ontological dependence is supposed to be blatantly

impossible. P3 is supposed to be true because only God is

powerful and knowledgeable enough to be able to exist inde-

pendently of all other entities.

We can be efficient here. The doubts about the Ontological

Dependence argument parallel the doubts about the First Cause

argument.

First, concerning P2, exactly why couldn’t there be an endless

sequence of ontological dependents? For instance, why not an

endless sequence of bigger parts depending for their existence on

ever-smaller parts? The sheer infinity of the sequence does not

make it inconceivable. We saw that we can conceive of a begin-

ningless series by considering the negative integers. If it is other-

wise impossible, why is that? Until we see a good reason, P2

stands in doubt.

And why does only God qualify as ontologically independent?

Suppose that there are point-sized physical particles that have no

God � 69

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

parts. Why think that they would have to depend on anything

just to exist?

Until we have a good answer to this question, we have

grounds to doubt that God is uniquely qualified for ontologically

independence, as P3 claims.

Designing the World

When we stand back from the previous arguments and consider

what they try to do, they seem amazingly ambitious. The only

facts about the world around us that the arguments use are the

facts that there are effects and that there are ontologically depen-

dent things. Simple, abstract, neutral facts like those seem far

removed from the existence of an all-knowing, all-powerful,

morally perfect creator. It is no wonder that arguments on that

meager basis turn out to fall short of proving God’s existence.

The actual facts of the world are much more wonderful than

just any old effects and dependencies that might have existed.

Maybe some awe-inspiring facts about how things actually are

can serve to establish God’s existence.

Suppose that the whole universe was unplanned and purely

accidental. What would it be like? We can apply to this question

what we’ve observed about accidents. Accidents make messes.

Car crashes, bridge collapses, and accidents generally, result in

disarray. Yes, once in a while there is a fortunate accident where

some structure happens to develop. Some inadvertently spilled

paint occasionally forms some neat shape. But that is highly

exceptional. And if the accidents keep coming, any structure in

the situation eventually dissolves. Further accidental paint spills

obliterate a pretty pattern. So, if the universe was entirely acci-

dental, then our observations lead us to expect that it would

display disorderly disarray, with the occasional pattern emerging

briefly.

70 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

That is not what we find. Instead, we find an abundance of

examples of organized structures resembling complex machines.

The most impressive machine-like structures that we know of

involve life. The examples range from the intricate interrelation-

ships of components within single cells to the tremendous com-

plexities of whole organisms and eco-systems. The composing

material at sub-cellular levels is also highly organized, from the

structures of molecules to the structures of atoms and sub-

atomic particles. On larger scales we find planetary systems,

galaxies, and groups of galaxies.

We have observed how order gets introduced. What we ob-

serve is that machine-like order is imposed by minds. We see

such order arise by design in everything from simple tools to

amazingly intricate systems like computers and ocean liners. We

do observe mindless robotic devices at work on assembly lines,

arranging materials into planes, trains, and automobiles. But it

always turns out that minds designed the system.

What does this comparison tell us about the origin of the

universe? Proponents of a design argument for God’s existence

contend that it makes a strong case for a divine mind behind the

whole thing. They contend that the universe has machine-like

structure throughout. They add that the only mind up to the task

of planning all this is the mind of the divine creator, God.

First Version

Here is our first version of this reasoning, in two phases.

Demonstration by Design

Phase 1

Premise 1: The universe exhibits intricate machine-like struc-

ture on every scale of space and time.

God � 71

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

Premise 2: The only possible way for the universe to exhibit

such structure is for it to have been intelligently designed.

Conclusion 1: The universe was intelligently designed.

Phase 2

Conclusion 1: The universe was intelligently designed.

Premise 3: If the universe was intelligently designed, then it

was designed by God.

Conclusion 2: God exists.

P2 links the claim made by P1 about order in the universe to the

conclusion of Phase 1 so that the conclusion follows inescapably.

By doing this, though, P2 runs afoul of the possibility of the

improbable. Consider the most orderly arrangement imaginable

of the largest universe imaginable. Call it a MOHU, for ‘Max-

imally Orderly Huge Universe’. If we somehow knew that we

were in a MOHU, it would be ridiculous to assume that our

MOHU happened to exist for no reason. That is so unlikely as to

be virtually impossible. The problem for Phase 1 is that the

accidental existence of a MOHU is only virtually impossible,

not quite just plain impossible. No matter how much structure

the MOHU has, its materials might possibly have happened to

arrange themselves that way in a fluke random occurrence. If we

doubt this, our doubts can be worn away. We must acknowledge

that some minimal structure could arise by chance, say, a simple

shape arising from random fluctuations. How about just a little

more structure? No doubt that is less likely, but still, it is a

possibility. How about a little more, and more, and more? We

find ourselves acknowledging the possibility of a structure

exactly like a Rolls Royce arising at random. And we can’t stop

there. Only the likelihood decreases; we never reach any impos-

sibility. Finally we have to admit that random typing by monkeys

might possibly type out Hamlet. No defensible stopping place

exists and we end up acknowledging the possibility of a chance

MOHU. P2 denies this possibility, and that is bad for P2.

72 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

Second Version

There is an alternative version of the reasoning. Some arguments

render their conclusions highly reasonable, though they offer

something short of proof. If considerations of design could do

that for the conclusion that God exists, it would be an important

result. We said that a chance MOHU was possible. It would be

highly unreasonable, though, on finding ourselves in a MOHU,

to think that it was a chance MOHU. We would be seeing

maximally orderly arrangements everywhere. They would al-

most certainly exist for some reason, either in the nature of the

MOHU’s laws or in the mind of a creator. It would be way more

sensible to deny that our MOHU existed by a fluke of chance. If

we could be shown that affirming God’s existence is as reason-

able as denying that a MOHU happened by chance, then the

claim that God exists would be very strongly supported. Even

somewhat weaker support would be plenty interesting.

Let’s return to what our observations show us about the

origin of organized structure. Our observations make it grossly

implausible that much machine-like order arose by accident. The

claim that this order exists by chance seems a very poor explan-

ation of it. In contrast, the claim that the order implements a

planned design renders its existence understandable to us. Pro-

ponents of a design argument can offer God’s design as the best

explanation of the structure that we find in the universe. To

capture this idea, we can replace P2 of the Demonstration by

Design with a claim about explanation.

Best Explanation by Design

Phase 1

Premise 1: The universe exhibits machine-like structure on

every scale of space and time.

God � 73

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

Premise 2e: The best explanation of the universe exhibiting

such structure is that the universe was intelligently designed.

So probably:

Conclusion 1: The universe was intelligently designed.

Phase 1 assumes that the best explanation of something is prob-

ably true. Phase 1 offers no proof that its conclusion C1 is true. But

if it succeeds, then it gives very good reason to believe C1.

Phase 2

Conclusion 1: The universe was intelligently designed.

Premise 3: If the universe was intelligently designed, then it

was designed by God.

Conclusion 2: God exists.

Not questioning P1 for now, how credible is P2e? Initially, it seems

quite plausible. What could explain the high level of order that

we observe as well as the explanation claiming that the order

resulted from an intelligent plan?

Here is a rival hypothesis: unplanned physical laws exist—laws

of physics, chemistry, biology, and the other sciences—and these

laws, operating on the physical materials in the universe, produce

the high level of order. This natural sort of explanation does

work. It gives an explanation of the machine-like organization

that we observe in things like molecules, marsupials, and

marshes. We can understand how some laws, operating on

some materials that were in a position to develop into orderly

arrangements by conforming to the laws, would yield the highly

orderly systems that we find in the universe. It is a long story that

science has yet to complete in detail. The point is that we see that

this is one way to explain the development of the order.

An explanation saying that the order implements a creator’s

plan also works. We understand that machine-like order could

have come about by implementing an intelligent design. P2e says

74 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

that the latter explanation is best. But so far, the two explanations

seem equally capable of explaining the phenomenon in question.

So why think that the design explanation is better?

It is sometimes complained that the purely physical explan-

ation just takes for granted that the physical laws and materials

that exist produce the observed order. ‘Yes,’ it is conceded, ‘we

can understand the presence of order, given the presence of

physical laws and materials that just happen to go together to

produce it. But this only pushes back the phenomenon requiring

explanation: why is there this remarkable combination of phys-

ical materials and laws that mesh together so as to produce the

observed high level of organization?’

Notice what this reply concedes. It acknowledges that the

physical account explains the existence of the order. The com-

plaint is that the physical account relies on something else, the

combination of laws and the arrangement of materials, and they

call for explanation.

This reliance does not show that the physical explanation is

worse than the explanation by design. The designing creator

explanation relies on things too. It relies on the existence of an

intelligent designer, the designer’s plan for the universe to be as it

is, and the designer’s capacity to implement that plan in the

universe. The existence of these things could use explaining

too. It would be arbitrary simply to rest content with no explan-

ation of them.

It is far from clear which explanation is in better shape here.

A powerful intelligent being who planned and created the whole

universe would be the most amazing thing in the world. Such a

being can seem much more remarkable than the existence of

natural laws and materials that happen to work together to

generate the observed order. After all, we acknowledged some

possibility of natural things just happening to produce a high

degree of order, however unlikely it was. Is a designing creator

even that likely? If the existence of the right physical ingredients

God � 75

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

calls for explanation, then the existence of a designing creator

cries out for explanation.

Some have claimed that God’s existence is self-explanatory;

others have denied that it requires explanation. These claims are

seriously obscure and doubtful. The first one seems to say that

God exists because God exists. That’s just bewildering. It never

explains anything just to repeat what needs explaining. An all-

powerful God would have what it takes to sustain a continued

existence, if God exists. But God’s existence in the first place is

what we are concerned about now. Similarly, it is baffling to be

told that God’s existence requires no explanation. Why not? If

there is some good reason to take for granted God’s existence,

why doesn’t that reason also apply to the laws and materials of

the physical account? Why do they still need explaining?

If there is anything finally better about the design explanation, it

remains to be seen. Until it is definitely seen, P2e stands in doubt.

What about P1? Is the universe really so well organized all over

the place? That is doubtful. On the largest spatial scale that we

currently observe, the galaxies are not randomly distributed.

They tend to cluster. But that’s it. They are not arrayed in

some pinwheel pattern or any other fancy structure. Being

somewhat clumped together is not an impressive type of organ-

ization. Similarly, on the smallest spatial scale that we currently

have information about, the scale of particles composed of

quarks, we have trios of quarks bound closely together and

jiggling about. That is not much like a complicated machine.

When we look far back and far forward in time, the leading

current cosmological views find considerably less intricate or-

ganization than is present today. Going far back toward the Big

Bang, the theories say that things become ever less machine-like

in structure. Going far forward toward the Big Chill, the theories

say the same thing. So P1 is open to serious doubt.

We could replace P1 with a premise about the more localized

order that is more clearly present in the world. But the smaller

76 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

the portion of reality that displays machine-like order, the more

probable it is that the order is accidental. Recall that our obser-

vations of accidents allow occasional patterns to be purely acci-

dental effects of natural laws in operation. Does the extent of

machine-like order in the whole world, throughout all of space

and time, rise above that level? This is a question of detailed fact

with no obvious answer.

There are other kinds of order that are sometimes cited in

design arguments. One kind is the order that consists in the

unbroken regularity of the operation of natural laws. This order

is present throughout the known universe, including regions

where machine-like structure is absent. If we replace P1 with a

claim about the existence of this lawful order, does that make a

better case for a designing creator?

The second premise will have to be adjusted too. It will have to

claim that intelligent design best explains this lawful order. This

new premise is open to doubt. When it comes to machine-like

order, we are familiar with how minds introduce it. We have

observed minds producing machines. But when it comes to

something as perfectly uniform as the operation of a natural

law, we have not observed minds implementing any such thing.

Natural laws are like rules. Minds do invent rules. But intelligent

minds in our experience do not enforce the same rules with no

variation, ever, no matter what. Attributing such order to an

intelligent design does not enable us to understand why the order

exists, at least until we see a good enough reason for the absolute

constancy.

This is an initial ground for doubt. Some views about God

offer candidate reasons for God to institute unvarying laws. Also,

according to some religious views natural laws are not perfectly

constant, since they have been miraculously violated. These

views in turn are disputed.

Another sort of order that some people point to as evidence of

design is a kind of fine-tuning among physical magnitudes.

God � 77

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

According to current theories, if certain basic physical magni-

tudes had not been almost exactly the quantities that they are,

they would have disallowed the development of complex atoms,

much less human life. Does this argue that the universe was

designed for us to exist in it?

Again, there is some ground for doubt. Suppose that human

life depends on some exactly appropriate basic magnitudes in

nature. Still, human life occupies an extremely small fraction of

the known universe and it is extremely recent by cosmological

standards. If the universe was designed for us by a mind intelli-

gent and powerful enough to adjust physical magnitudes so that

we would eventually get here, why didn’t the mind produce us

more efficiently?

Again, the initial doubt may be answerable. Perhaps the huge

lifeless portion of space and time serves other intelligent pur-

poses. Such purposes have been proposed, and disputed.

This issue will not be resolved here. There is no brief way to

decide the merits of replacing our first premise with one about

these other sorts of order. However the best version of the first

premise finally works out, Phase 2 of the Design arguments has a

problem that deserves our attention.

The doubts that were just raised about explanations by design

are similar in spirit to some of David Hume’s ideas in his

wonderful work, Dialogues Concerning Natural Religion. P3 in

Phase 2 is subject to some of Hume’s other powerful points.

For one thing, Hume suggests that we make use of more specific

details in our observations of the origins of order. For instance,

any large building project in our experience has multiple design-

ers who have limited knowledge and ability. The universe was the

largest building project of them all, if it was created by design. So

our experience would lead us to expect a huge team of limited

designers for such a project, rather than one all-knowing all-

powerful God. Do we know anything else that overrides this

lesson of experience? If not, then P3 is highly questionable.

78 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

Conceptually Guaranteeing God

A concept is a way of classifying something in our thinking. All

of us have approximately a zillion concepts. We have the concept

of a mammal, the concept of molasses, the concept of a toy, the

concept of friendship, the concept of gravity, the concept of

eyesight, the concept of danger, the concept of a boringly long

list, and so on. A singular concept is a classification that brings to

mind a single thing, if the concept applies at all. Singular con-

cepts are familiar. Examples from ordinary life abound. When

Donna’s dachshund Dobson is in Donna’s house alone, he is fond

of luxuriating on the sofa, occupying his chosen pillow in regal

comfort. While Dobson is doing this, we can bring him to mind

in many ways—for example, by conceiving of him as the pooch

on the couch, as the dachshund on the pillow, and as the dog in

the house. These are singular concepts that apply to Dobson.

One important line of thinking has it that God is the greatest

being that anyone could bring to mind. If so, then one singular

concept of God is the concept of the greatest conceivable being.

We’ll need the phrase ‘greatest conceivable being’ a lot. Let’s

abbreviate it with its initials: GCB.

Almost a thousand years ago the medieval philosopher An-

selm argued that the GCB concept has to apply to an existing

entity who is God, because of facts that we can discover by

appreciating the nature of the concept itself. The reasoning is

called ‘Anselm’s ontological argument’. 1 In one version or an-

other, ontological arguments are particularly appealing to many

philosophers. This appeal has something to do with the remark-

able fact that we are supposed to be able to find out, just by

1 The aim of this chapter in considering Anselm’s argument is to think about

whether it shows that God actually exists. In the chapter ‘Why Not Nothing?’ two other ontological arguments are discussed. The aim there is to determine whether they can show that a necessary being exists, whether or not the being qualifies as God.

God � 79

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

thinking correctly, all that we need to know to see them prove

their point. They are pure philosophy with a powerful payoff—if

they work. The ontological argument that we’ll consider is a

reconstruction of Anselm’s highly influential reasoning.

It’ll be helpful to have a label for what a singular concept

singles out. In other words, we want a term for the entity that

meets the specifications of the concept, if anything does. The

concept of Donna’s dog, for instance, calls for a dog that is the

one owned by Donna. The concept applies to such a dog, or it

does not apply. Let’s label the entity that is singled out by a

singular concept the target of the concept.

Typical singular concepts need not have a target. Consider the

concept of the spoon on the moon. If a single spoon happens to

be on the moon—maybe an astronaut left one there—then this

concept has that spoon as its target. Otherwise the concept of the

spoon on the moon has no target. Either way, the concept of

the spoon on the moon is one of our concepts. The same goes

for the singular concept of the pooch on the couch, the singular

concept of the farthest star from the Earth, and so forth.

Key question: Could our GCB concept lack a target?

No, according to Anselm. He asks us to suppose that the GCB

concept has no target. In other words, suppose that the GCB

does not exist. Anselm argues that if this were so, then we could

form another concept that would be a concept of something

greater than the GCB. Starting with our GCB concept, we can

add the idea of existing. This gives us the concept of the existing

GCB (the EGCB for short). Anselm holds that under circumstan-

ces where no GCB existed, our EGCB concept would be the

concept of something greater than the GCB. The reason is that

existing is a better status than not existing and we would be

explicitly requiring existence in our EGCB concept.

But wait! Anselm points out that there is no possible way for

us to form a concept of any being that is greater than the greatest

80 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

conceivable one. The GCB is the greatest being that we can

conceive of—it says so right in the concept itself. Therefore we

cannot conceive of a greater being. Yet in the situation just

described, we are supposed to be conceiving of a greater being.

Since this is impossible, as we just saw, we must have assumed

something untrue in setting up the situation. Anselm holds that

the only questionable assumption in the setup is the initial one,

the assumption that the GCB concept does not have a target. If

that assumption is the mistake, then the GCB concept does apply

to something. So the target of the GCB concept, the GCB, exists.

The GCB is God. So God exists. This reasoning can be summar-

ized as follows.

Anselm’s Ontological Argument

Phase 1

Temporary Assumption (TA): The GCB concept has no target.

Now add this premise:

Premise 1: If the GCB concept has no target, then the EGCB

concept is a concept of something greater than the GCB

concept.

From TA and P1, infer:

Temporary Conclusion (TC): The EGCB concept is a concept

of something greater than the GCB concept.

Add another premise:

Premise 2: No concept is a concept of something greater

than the GCB concept.

Premise 2 says that TC is untrue, so the temporary assumption

TA that got us TC must be false. In other words, infer:

Conclusion 1: The GCB concept does have a target.

God � 81

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

Phase 2

Conclusion 1: The GCB concept does have a target.

Premise 3: If the GCB concept does have a target, then the

GCB exists.

Conclusion 2: The GCB exists.

Phase 3

Conclusion 2: The GCB exists.

Premise 4: The GCB is God.

Conclusion 3: God exists.

Let’s start our critical consideration of this argument on a posi-

tive note by contemplating P3. It is entirely okay. If a singular

concept has a target, then the concept does apply to some existing

thing. For example, since the singular concept of Donna’s dog

has the real dog Dobson as a target, Donna’s dog exists.

Now let’s consider the final assumption, P4. It seems pretty

credible at first that God is the GCB. But maybe we can conceive

of something greater than God. Such as? Well, consider someone

with limited abilitieswho overcomes adversityand acts heroically. In

away, such a person seems to be better than any being of unlimited

power and knowledge who is morally flawless. That sort of being is

tooknowledgeableandpowerfultobeheroic.Maybeheroismisone

feature of a conceivable being who would be overall greater than a

being who has the power and knowledge of the traditional God.

This is debatable. God could still turn out to be the greatest.

For instance, the greatness of God might consist in God’s having

all of the important positive properties, like knowledge, ability,

and moral goodness, to a maximum extent. That sounds like an

unbeatable combination.

This idea that God has the maximum degree of greatness is a

risky one, though. The important positive properties may not

all have a possible maximum. For example, part of being morally

good is doing good. Yet no matter how much good someone does,

it seems possible to have done more good. So moral goodness

82 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

may not have a maximum. If not, then we don’t get the GCB by

conceiving of a being who is maximally morally good, because we

get an impossible being. Any being that does exist and is good

surely outdoes the greatness of any impossible being. Thus, the

maximum idea of God is a problematic way to try to establish

God as the GCB.

Much more thinking is needed to draw a justified conclusion

about the truth of P4. But regardless of how well Phase 3 with P4

works out, successful reasoning through Phase 2 would be

nothing to sneeze at. A proof of Phase 2’s conclusion, C2,

would be mighty metaphysically interesting. Establishing the

actual existence of the greatest conceivable being would show

us something wonderful about reality.

P1 and P2 are taken for granted in Phase 1. If either one of them

is untrue, then C1 is not proven in Phase 1. Without success in

Phase 1, the whole argument collapses. Let’s think more about P1.

P1 says that if the GCB concept has no target, then the EGCB

concept is ‘of’ something greater. The interpretation of the small

word ‘of’ turns out to be crucial to assessing the argument. Two

interpretations should be distinguished. First, for a concept to be

‘of’ a greater being, on one interpretation, is for a greater being

to be the concept’s target. This interpretation gives us:

P1.1: If the GCB concept has no target, then the target of the

EGCB concept is a greater thing than the target of the GCB

concept.

If the GCB concept has no target, then it is easy for some other

concept to have a greater target. The other concept would just

have to apply to something that is greater than nothing. Again,

anything good is greater than nothing. So a concept of a good

thing that exists would qualify as having a greater target than the

GCB concept. But would the EGCB concept in particular have a

greater target, as P1.1 says?

God � 83

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

Suppose that the GCB concept has no target. Recall that this

means that the GCB concept does not apply to anything. If

nothing is the greatest conceivable being, then nothing is the

existing greatest conceivable being either. Thus, if the one con-

cept applies to nothing, then so does the other. Since they both

lack targets, the greatness of their targets is the greatness of

nothing—worthless! Therefore, if the GCB concept has no tar-

get, then the GCB concept and the EGCB concept would be tied

at zero for the greatness of their targets. This denies the P1.1 claim

that the EGCB concept would have a greater target. So if we have

interpreted P1 correctly as P1.1, then it is untrue.

There is another interpretation of P1. The new idea is that if

the GCB concept has no target, then the EGCB concept demands

more greatness than does the GCB concept. In other words, if no

GCB exists, then in the competition for being our way of conceiv-

ing of the greatest being that we can possibly conceive of, the

EGCB concept would beat out the GCB concept. Both concepts

clearly require extreme greatness to apply. But according to P1 as

we are now interpreting it, in the absence of a real GCB, the EGCB

concept would require the greater greatness. This gives us:

P1.2: If the GCB concept has no target, then the greatness

needed for the EGCB concept to apply is more than the

greatness needed for the GCB concept to apply.

P1.2 does not stand scrutiny. The GCB concept goes all out in its

demand for greatness—it demands ‘the greatest’. It demands

maximal greatness, whether or not its demand is met. For

example, existing appears to be part of what it takes to be the

greatest thing that we can conceive of. Any ‘things’ that could

have existed, but don’t exist, at most could have been great.

‘They’ aren’t great. ‘They’ aren’t anything, much less anything

great. If this appearance that existence is needed for greatness is

correct, then the GCB concept demands existence just as much

as the EGCB concept. If this appearance is incorrect, then the

84 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

EGCB concept does not demand more greatness by explicitly

demanding existence.

There is just no way for the GCB concept to be beaten in this

competition. The GCB concept requires ‘the greatest’, and that’s

that! Yet P1.2 alleges that under one particular condition—the

non-existence of the GCB—the EGCB concept demands more

greatness. That must be a mistake. The existence or non-exist-

ence of a GCB does not alter what any concept demands for its

application. A concept’s demands for its application are what

make it the concept that it is. For example, the concept of

chocolate is the concept of chocolate, rather than the concept

of vanilla, or the concept of strawberry ice cream, or any other

concept, because the concept of chocolate is the one that de-

mands for its application precisely chocolate, nothing more or

less. A concept’s demands are just built into it. The non-existence

of the GCB doesn’t affect what the EGCB concept demands,

including whether the EGCB concept demands something

greater than the GCB concept demands. And we’ve just seen

that the EGCB concept does not demand anything greater. So on

this other interpretation P1 is also untrue and does not help

Anselm’s ontological argument. Phase 1 of the argument relies

on the truth of some interpretation of P1. Since the argument

needs Phase 1 to work in order to get anywhere, the argument

goes nowhere if our criticism is correct.

Putting it All Together

We have found problems in each of the arguments for God’s

existence that we have considered. Let’s not leap to any conclu-

sions. Even if we had found problems in all arguments that are

ever made for God’s existence, it would not follow that God does

not exist. Entities whose existence cannot be proven by us might

exist. They might exist without being in any revealing sort of

God � 85

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

relation to us. God could be like that. Or God could be revealed

by an argument that we have not considered.

Let’s not leap away from any conclusions either, though. The

arguments that we have seen for God’s existence do not work.

Sometimes each clue to a crime on its own does not mean

much, while together they argue powerfully for a certain culprit.

Likewise, the thoughts from several arguments might work

better in combination. The most reasonable belief that we can

have about something is one based on all relevant available

evidence. So before we draw any conclusive conclusions about

God’s existence, we would do best to look at the combined

strength of our arguments.

There are initial indications of an improved case. For instance,

it seems to become more reasonable to think that the universe

had God as first cause when we add in the observations from our

discussion of the design argument that support the idea that the

universe displays various sorts of order. On the other hand, the

doubts raised about whether the universe really is organized as

though by an intelligent designer carry over as doubts that God

was its first cause.

Assessing the strength of a combined case for God’s existence

would require assessing together everything in the First Cause

and Design arguments, and the Ontological argument as well.

Having done this, we would still not be in a position to draw the

most rational conclusion. More evidence exists. There are other

arguments for God’s existence. There are arguments against

God’s existence too. The most prominent one—the Problem of

Evil—contends that an all-powerful, all-knowing, morally per-

fect being would never allow all of the bad things that exist in this

world, and so no such being exists. Several versions of this

argument have been developed. They have in turn received

intense critical scrutiny. All of that is more of the evidence

available on the topic of God’s existence. And then there’s the

86 � God

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi

challenge of assembling and weighing the totality of the evi-

dence . . . We never said that metaphysics was quick and easy!

There’s no need to get discouraged, either. We have looked

into the merits of the major metaphysical arguments for God’s

existence. So we have a serious investigation well under way.

further reading

Philosophy on the topic of God’s existence comprises a huge literature.

Here are two significant recent books. The first one is favorably

disposed toward arguments against the existence of God; the second

one is favorably disposed toward arguments for the existence of God.

Jordan Howard Sobel, Logic and Theism (Cambridge University Press,

2004).

Richard Swinburne, The Existence of God, 2nd edn. (Oxford University

Press, 2004).

God � 87

OUP CORRECTED PROOF – FINAL, 20/8/2014, SPi