I Expose Penn Jillette’s Logic Trick

May 14, 2013


He is quite right to point out that agnosticism and atheism address completely separate issues. He is also correct when he says that agnosticism answers the question “Is there a God?” with the response, “I don’t know”.

He is also correct that atheism answers the question, “Do you believe in God?” in the negative.

Here’s his mistake. He claims ‘if you don’t know God exists, then you don’t believe God exists.’ This is clearly false. It states that knowledge is a necessary condition for belief. By contraposition (“if not p then not q” is the same as “if q then p”), we get the claim, “if you believe God exists, then you know God exists.” What self respecting atheist would endorse that claim? None! None I tell you! That would be bullshit!

 Penn Jillette in his Wizard attire

[Penn and Teller magic powers happen…]

Wait a minute. Maybe this does make sense.

Let’s back up and go through it slowly.

We’ve got two questions.

1. Does God exist?
Answer: I don’t know = Agnosticism. Check.
2. Do I believe God exists?
Answer: No = Atheism. Check.

But what if I give a different answer to question 1?

1. Does God exist?
Answer: Yes = Theism.

Sure, but is that all? In responding so positively, and so confidently, it seems like you’re really saying “yes God exists, and I know it be true.” So, in answering ‘yes’, it seems like the position of gnostic theism falls out. You can address both issues at once, depending on your answer.

What about this response:
1. Does God exist?
Answer: probably not.

This is the most common response you’ll get from atheists, because most atheists don’t want to answer “no” with such confidence, such certainty, because it feels like they’re saying “I know God exists,” and they don’t want to say that. So they say “probably not,” indicating that they don’t believe God exists without claiming knowledge of such.

So really, I think you can answer that first question in a way that reveals both your position on gnostic vs. agnostic and theist vs. atheist, and those various responses indicate that there is indeed a difference between the two issues. So can we get an answer indicating agnostic theism?

Can anyone ever reasonably answer the first question like this:
1. Does God exist?
Answer: probably… but I don’t know for sure.

This answer represents an agnostic theist. Penn Jillette doesn’t think you can be an agnostic theist, though. I think you can be an agnostic theist. What do you think Teller? 

Teller murdering Penn.

Clearly Teller agrees. How to explain this to Penn? What we need is a Good Analogy. Penn knows a lot about cards, so let’s use a card analogy.

You go to MIT and you are part of the MIT Blackjack Team. You are freaking amazing at counting cards. You’re a regular Rain Man. So you’re sitting there counting, and that count is really high, like +13, which means some face cards are about to start dropping. Pause.

1. Will one of the next few cards be a face card?
Answer: I don’t know = face card agnostic.

So, that means you don’t believe one of the next few cards will be a face card, right? Because face card agnosticism implies afacecardism. If so, you wouldn’t last long on the MIT Blackjack Team. They’d kick you off. For being a loser. When the count is +13, you signal for one of your teammates to come and start betting. The probability is high that one of the next few cards is a face card.

MIT blackjack team member, played by Kate Bosworth.

Not a loser.

But consider:

1. Will one of the next few cards be a face card?
Answer: Very probably, but I don’t know for sure.

Well, that seems like a perfectly fine answer.

What about:

2. Do you believe one of the next few cards will be a face card?
Answer: Yes.

Spacey, coach of the blackjack team.

Doesn’t play with losers.

So, there’s nothing about answering the first question with “I don’t know” that forces you to answer the second with “no”. And there is a perfectly logical reason for that.

Epistemology is the philosophical study of knowledge, and it is also one of the coolest of all philosophy subjects. The consensus is, and pretty much has been since before Socrates, that knowledge entails belief, but that belief does not entail knowledge.

This means that in order to know something, you’ve gotta at least believe it, but that believing is not enough to count as knowledge. If believing were enough to count as knowledge, then you’d have ridiculous things like, “If you believe God exists, then you know God exists.” Ridiculous, right?

Some of you are probably gearing up for the ACT or SAT, or maybe oven some other test involving a combination of capital letters. On those tests, you’ll have to do analogies. So, let’s do another analogy.

Belief is to knowledge as __________ is to ____________ .

a. shape … triangle
b. mode of transportation …. car
c. death … murder
d. injury … assault

Of course, all of those are correct answers. If someone knows something, then they believe it. If someone has a car, then they have a mode of transportation. If someone is murdered, then they have died. If someone has been assaulted, then they have been injured.

These are all false, with (counterexamples in parentheses):
If someone doesn’t know something, then they don’t believe it (It might actually be false, but they wrongly believe it, in which case they don’t know it, but they believe it). If someone doesn’t have a car, then they don’t have a mode of transportation (bicycle). If someone has not been murdered, then they have not died (old age). If someone has not been assaulted, then they have not been injured (tripped on a tree root and broke an arm).

So, in conclusion, Penn Jillette’s logic trick rests on a claim that no self-respecting atheist would ever assert (if you believe X, then you know X), and it fails to recognize the proper relationship between knowledge and belief, namely that knowledge is merely a type of belief, just like murder is merely a type of death.

American Psycho demotivational.

Still, it’s a special kind of death, so make it special.

If you take a belief, add truth, justification, and some special sauce that avoids the Gettier problem, then you get knowledge. If you take a (human’s) death, add unlawfulness and another person as the cause, then you get murder. It’s that simple. Penn Jillette’s claim is formally identical to the claim, “If he died, then he was murdered.” This would make prosecutors’ jobs a lot easier, but thankfully everyone recognizes that it is false. It’s bullshit.

Penn and Teller: Bullshit!