- 0. Introduction. Matt Slick; evangelical Calvinist, radio presenter, apologist. He has made something of a name for himself by promoting a version of the ‘transcendental argument for the existence of God’. His version is one of the easiest to refute that I have come across. However, in all the debates and online discussions I’ve seen Slick engage in, and to be sure he engages in a lot, I have never seen anyone offer what I consider to be the correct refutation. So I will present it here.
His argument was given on his radio-show/podcast, on 17th December, 2015, in an episode entitled ‘A Proof of God’. In fact only the last 14 mins of the show are dedicated to this topic, when Slick is prompted by a caller – ‘Hollywood dude’. I will use that version as a foil. Here is the link it on his official ‘CARM’ podcast site: http://carmpodcasting.blogspot.co.uk/2015/12/carm-podcast-1214.html
Admittedly, the argument was given in a rather off-the-cuff manner by Slick in that show, and he could be forgiven for not being clear and careful with his words. On the other hand, his presentation on the show was very similar to many other times he has given the argument in the past, in situations where he had the opportunity to prepare and refer to notes as he spoke, such as:
- https://www.youtube.com/watch?v=oPi073maMRw – In debate with Dan Barker in April 2014. Slick makes his case at about 4:40 into the video.
- https://www.youtube.com/watch?v=Dpsz4v_Z8JY – In debate with Matt Dillahunty on the Atheist Experience in 2009. Slick makes the argument at 43:03 in the video.
The argument is also given in written form on his website, here: https://carm.org/transcendental-argument. The version of the argument I am looking at here is found at the end of the written version (section 9).
- Disjunctive syllogism and true dichotomy
At 44:15 into our show, Slick explains his argument. He says that he will use the argument form known as ‘disjunctive syllogism’, which is the following inference rule:
Either p or q
It says that if either p or q is true, and if it is also true that one of them is not the case (say, p), then the remaining one (q) is true. Disjunctive syllogism is valid in propositional logic, and its validity will not be challenged by me here.
Slick also uses the notion of a ‘true dichotomy’, by which he means a strong type of ‘or’-statement. In propositional logic, ‘or’ is a connective that takes two propositions, e.g. p or q. It’s behavior is entirely logical. ‘p or q’ is true when p is true and q isn’t, when q is true and p isn’t, and when they are both true. It is false when they are both false. That is a disjunction.
Slick’s ‘true dichotomies’ are a strong version of a disjunction; true dichotomies are always true, as by definition one of the options is true in exclusion of the other. The way this is achieved is purely logical; the propositional form of ‘true dichotomies’ is a disjunction between a proposition and its direct negation; ‘p or not-p’.
So here is a normal disjunction:
Either Sam or Alex will come to the party.
If it is true, then one of them will be at the party; but it might be false because perhaps neither Sam nor Alex will come to the party. Consider, in contrast, the following:
Either Sam will come to the party, or she won’t.
In this case it has to be true, because there are no other possible options than Sam being at the party, or her not being at the party. A ‘true dichotomy’ for Slick is like this; it has to be true because it covers all possible options.
- Slick’s argument
At 44:15, Slick gives the following monologue:
“If you only have two possibilities to account for something … if one of them is negated the other is necessarily validated as being true … So we have ‘God and not-God’, so that’s called a true dichotomy, God either exists, or it is not the case that God exists, we have the thing and the negation of the thing. So now we have a true disjunctive syllogism … We have, for example, the transcendental laws of logic … Can the no-God position account for the transcendental laws of logic? And the ultimate answer is no it cannot. So therefore because it cannot, the other position is automatically necessarily validated as being true. Because, you cannot negate both options out of the only two possibilities; that’s logically impossible.”
The argument structure being used is as follows:
1) Either God, or not-God.
2) Not-God cannot account for the laws of logic.
3) Therefore God can account for the laws of logic.
He then proceeds to examine objections to premise 2, such as some of the main ways an atheist (a representative of the not-God camp?) might try to account for the transcendental laws of logic. Are they discovered, measurable features of empirical reality? Slicks says they cannot be. Are they ‘linguistic constructs’? Again, no. Do we vote on them? (Sigh) No. Could they be constructs of human minds? No, no, no. No.
At the end of it, Slick summaries how he speaks to his imaginary interlocutor, the poor atheist, who has had his every attempt at accounting for logic rebutted (this is at 48:22):
“When we go through this with them, I’ll say: ‘See, you can’t account for it. Therefore, the other position is valid’. And then I say: ‘Next!’”
So, what is my refutation of this argument? Well, it does not involve giving a better account for the transcendental laws of logic than our poor imaginary atheist. Nor does it require pinning Slick down on precisely what it means to have an account of something. Neither does it involve pointing out to Slick that the premise ‘God or not-God’ is not an instance of a true dichotomy because, strictly speaking, it is not a properly formed sentence at all. Anyway, nothing as fancy as the metaphysics of logic is needed here. And we can forgive a badly formed sentence here and there. We can afford to be so magnanimous because there is a logical problem with the argument, and it is very simple. It is a slight of hand, which can go un-noticed, but is easy to spot when spelled out. It is an instance of the fallacy of ‘false dichotomy’.
A true dichotomy, such as:
- a) ‘Either God exists, or it is not the case that God exists’,
is substituted for the false dichotomy of:
- b) ‘Either God accounts for the transcendental laws of logic or not-God accounts for the transcendental laws of logic’.
The second is not a genuine dichotomy, because it is quite possible that neither God nor his negation has anything to do with the laws of logic. Here is an example, meant as a reductio of Slick’s argument:
1) Either toast, or not-toast.
2) The absence of toast cannot account for the laws of logic.
3) Therefore, toast can account for laws of logic.
Obviously, the absence of toast cannot ‘account’ for anything, especially the notoriously murky metaphysics of logic. Does this mean though that toast itself can? It seems equally obvious that it cannot. Taking one out of the running is not all that is needed to show that the other is the winner by default. Neither toast nor ‘non-toast’ can account for the laws of logic. The unsoundness of the argument is painfully obvious when ‘toast’ is used in place of ‘God’.
To make Slick’s fallacy apparent, let’s spell out the argument a bit more clearly:
- Reconstruction 1:
1) Either God can account for the laws of logic, or not-God can account for the laws of logic.
2) Not-God cannot account for the laws of logic.
3) Therefore, God can.
As we have seen, the problem with this is that the first premise isn’t a true dichotomy. Slick’s premise says:
Either [x can do y], or [not-x can do y]
This leaves the logical space available, where neither x nor not-x can do y, which stops the argument being sound. Maybe it is the case that nothing can play the role of x; i.e. maybe nothing can account for logic. If this were the case, then we could not prove one of these two options by eliminating the other (which is the whole point of using disjunctive syllogism). So if the first premise is as I have indicated, then we can rule out disjunctive syllogism as a useful argument form; that is, unless some independent reason can be produced for thinking that this form of the premise is true.
The point about the first premise, when spelled out like this, is that it is in need of justification. Slick dangles the true dichotomy of ‘God or not-God’ in order to gain assent (as nobody can deny a tautology), but then switches focus to the false dichotomy above without conceding that he now needs to justify the new premise. This is the heart of the Matt Slick Fallacy; it is a bait and switch from a true dichotomy to a false one.
It is clear that that [not-x can do y] is not the direct negation of [x can do y]. The direct negation of [x can do y] is:
not-[x can do y].
This would make the actual true dichotomy:
Either [x can do y] or not-[x can do y]
To get a feel of the distinction, consider the following:
Either God can account for logic, or not-God can account for logic
Either God can account for logic, or it is not the case that God can account for logic.
It is a subtle enough point, but makes all the difference. It is a scope distinction about whether the negation should be thought of as ranging over the entire proposition (as in the true dichotomy), or just one element of the proposition (as in Slick’s false dichotomy). Slick’s mistake is rather like supposing that either the present king of France is bald, or the present king of France has hair. In reality, neither is true.
- Reconstruction 2:
We could get around this problem by making the first premise a true dichotomy:
1) Either God can account for the laws of logic, or it is not the case that God can account for the laws of logic.
2) It is not the case that (it is not the case that God can account for the laws of logic).
3) Therefore, God can account for the laws of logic.
Now the first premise is a true dichotomy (and so definitely true). Also, the form of the argument is definitely that of disjunctive syllogism, so therefore definitely valid.
This is where the good features of this argument end though. All disjunctive syllogisms with true dichotomies as the first premise are doomed to triviality, as is easy to show. This problem is due to the second premise of disjunctive syllogism. In this premise, either of the two options in the first premise (either p or not-p) is negated (it doesn’t matter which one is used). In the example above, it second premise uses not-p rather than p. So it is the negation of not-p, i.e. not-not-p. But this just means we already have our conclusion in our second premise. p is equivalent to not-not-p; the two ‘nots’ cancel each other out. This makes it a case of ‘begging the question’, where the conclusion of the argument is smuggled in as one of the premises.
To make it crystal clear, here is the form of disjunctive syllogism with a true dichotomy as first premise:
p or not-p
If we substitute ‘p’ for ‘not-not-p’ in the second premise (as they mean the same thing), the argument becomes:
p or not-p
The first premise is now clearly redundant. We could drop it and the argument would simply be:
Thus, the argument just boils down to the derivation of p from p. If the argument is formed this way, it becomes entirely trivial. We are left with no reason to think that p is true, other than the simple assertion that p is true in the first place.
In conclusion then, Slick has presented an argument which commits the fallacy of false dichotomy, and if repaired so as to avoid that ends up committing the fallacy of begging the question instead. Thus, the argument is either unsound or trivial.
 The sentence has no verb in it. Also, it is dubious that the negation of a noun, such as ‘not-God’, has any meaning whatsoever. In logic, it is propositions that get negated, not names.