Soon I will be writing a series of posts on the arguments of natural theology. I’ll focus primarily (but not exclusively) on those defended by William Lane Craig, mainly because he is the most well-known Christian apologist of our day. The arguments I have in mind are:
- the Kalam cosmological argument,
- the fine tuning argument,
- the moral argument,
- the ontological argument,
- the argument from the historicity of Jesus’ resurrection, and
- the argument from personal experience.
I expect to cover more than just these, and I’d be happy to take requests if you’d like a particular argument covered.
But before I get started with that, I think it would be sensible to make a few general points about philosophical argumentation in general. Since these are important things to keep in mind when discussing all of the above arguments, it makes sense to say them once now and refer back here when needed.
Before going further, I should mention that Graham Oppy covers a lot more ground in his excellent book, Arguing about gods, from which I have borrowed the title of this post. As well as extensive discussions of the arguments I intend to discuss, Oppy devotes an entire chapter of nearly 50 pages to general considerations of theistic and atheistic argumentation, including the finer details of what it means to deem an argument “successful”.
All of the arguments I consider will be in the form of logical deductions. There will be a sequence of premises, and a conclusion. There are two aspects to the “soundness” of an argument of this form:
- the conclusion must follow logically from the premises (this is to say that the argument is “valid”), and
- the premises must be true.
If both these conditions are met, then any rational person is required to accept the conclusion. If either or both of the conditions are not met, then the argument is not sound. This is not to say that the conclusion is not true, but only that the argument does not demonstrate that the conclusion is true. For example, consider the following argument:
- The sun is in the Milky Way.
- Anything in the Milky Way is a star.
- Therefore, the sun is a star.
This argument satisfies the first condition (the argument is valid), but not the second (the second premise is false). Therefore, the argument is not sound, even though the conclusion is true. Here is a second example:
- Stars are bright.
- The sun is bright.
- Therefore, the sun is a star.
In this case, the second condition is satisfied (both premises are true), but the first condition is not (the argument is not valid). Even without examining the logical structure of the argument in detail, we should already be suspicious. Indeed, if the argument was deemed sound, then we could mimic it to “prove” things that were obviously not true. For example:
- Stars are bright.
- My torch is bright.
- Therefore, my torch is a star.
But such parodies are unnecessary to defeat the original argument. The unsound nature of the logic of both arguments may be demonstrated by noting that they take the form:
- Anything that has Property P also has Property Q.
- X has Property Q.
- Therefore, X has Property P.
Here, P is the property of being a star and Q is the property of being bright. This is an example of the logical fallacy known as affirming the consequent. Wikipedia’s list of logical fallacies is a useful resource, but since the arguments I consider won’t have such obvious problems, I don’t need to say much more about logical fallacies.
2. Word play
But there is one other kind of fallacy that can sometimes plague an argument that might otherwise appear to be valid. Consider the following well-known example:
- Socrates is Greek.
- Greek is a language.
- Therefore, Socrates is a language.
This is an example of what is known as equivocation. Here, a word or phrase is used in two different senses to connect two otherwise disconnected ideas. In the above example, the word that is equivocated on is “Greek”. In the first sense, it is used to describe a person from the country Greece, and in the second it is used to describe the language such people usually speak. Although both premises are true, the conclusion is clearly not; despite appearances (they share a common word), the premises have nothing to do with each other.
But sometimes equivocation is more subtle than that. Rather than a word clearly being used in two different senses, sometimes it is the scope of the word that changes. Here is another well-known example, taken from Critical Thinking by Lavery et al:
- Some say it is wrong to discriminate.
- But people must often discriminate (choosing a marriage partner, for example).
- So it is not wrong to discriminate.
Here, in both cases, discrimination means making decisions based on the qualities of the options. But Premise 2 is speaking of discrimination based on relevant qualities, whereas Premise 1 refers to discrimination based on irrelevant qualities. The validity of the first kind of discrimination lends no support to the validity of the second kind.
3. Burden of proof
Perhaps the most important concept to grasp in relation to theistic arguments is that of the burden of proof. Suppose Bob presents an argument like the following:
- If X is the case, then unicorns exist.
- X is the case.
- Therefore, unicorns exist.
For the sake of our discussion, let’s suppose also that Premise 1 is true. In this case, since it has a logically valid form, the argument will be successful if and only if Premise 2 is true. So Bob must be prepared to demonstrate that X is the case.
To make the example more concrete, suppose X refers to the existence of a unicorn skeleton. Perhaps I am suspicious that the unicorn skeleton Bob claims to own might be a fake. If I told Bob about this suspicion, he might ask me to prove that it is a fake, and claim that his argument will be valid unless I can succesfully do so. But this would be to try to shift his burden of proof. If Bob wishes to prove that a unicorn skeleton exists, it is up to him to prove that his unicorn skeleton is not a fake. It is not up to me to prove it is a fake. Naturally, if I could show his unicorn skeleton was a fake, then I would have undermined his argument. But it is not necessary for me to do so. Since Bob is making the claim, it is up to him to demonstrate it.
Let’s consider a second example. Suppose Sally claimed that your husband was cheating on you. As evidence for this claim, Sally said that she saw your husband eating lunch with another woman. You might object, and say “Maybe he was eating lunch with a work colleague”, to which Sally might respond “Can you prove she was a work colleague?”. Of course it is not up to you to demonstrate that your husband’s lunch partner was a work colleague. Since Sally is making the claim, she must demonstrate that your husband really is cheating on you. And since the possibility of his lunch partner being a work colleague sheds doubt on the strength of her evidence, it is up to Sally to prove that your husband was indeed having an affair with this woman.
Both these examples show that if you have reason to be suspicious of a premise used in an argument, it is up to the defender of the argument to show that your suspicion can be adequately dealt with. It is not enough for them to ask you to prove that your suspected problem is an actual problem.
For further treatment of this important topic, I highly recommend QualiaSoup’s excellent youtube video on The burden of proof.
When I come to examine the theistic arguments alluded to above, it is not my goal to prove that there is no god. In fact, I doubt that such a goal could be achieved. Rather, my goal will be to simply prove that the theistic arguments fail for one reason or another. To do so, I do not have to prove that any of the premises are false (although this is sometimes possible). Rather, I have to demonstrate that one or more premises are not adequately supported. To do this, it is usually only necessary to shed some reasonable doubt on a premise. Often the defender of the argument will claim that this doubt is not sufficient to undermine their argument, and say that I must disprove the premise. But this is an attempt to shift the burden of proof, and is not a valid criticism.