Knowledge, evidence, and terrorism

Science isn’t just good for teaching scientific facts, it’s good for teaching people how to think. This principle was illustrated vividly today on the Rush Limbaugh radio show.

Embassy Riot.Rush Limbaugh spent an hour on his radio show today weaving a tale that can basically be summarized like this: 1) There were riots and deaths in the middle east on September 11th; 2) Some people knew that these riots were pre-planned terrorist attacks that had nothing to do with any video or anything else; 3) The government ignored the fact that these people knew that the acts were terrorism, and tried to cover up the terrorism by claiming that they were investigating or didn’t have enough information; 4) Therefore, Obama is anti-American and refuses to admit that there was a terrorist attack on us.

There are two things I would like to untangle about this argument. Both problems appear because Rush Limbaugh is unable (or unwilling) to reason scientifically about knowledge.

1) What is knowledge?

What does Rush Limbaugh mean when he says that there were some people who “knew immediately” that the riots were preplanned terrorist attacks?

Generally believing, we say that someone knows something (X) when three things hold:

1. X is true
2. The person believes X is true
3. The person has a good reason to believe that X is true

So, for example, if someone shows you a jar of marbles and you pick a number at random and say, “I think there are 153 marbles in there!” and you happen to be right, nobody would normally say that you knew the number of marbles in the jar: they would say you guessed and it was a coincidence that you were right. That’s not the same as “knowledge.”

Similarly, if the person said, “I think there are 153 marbles in there because 153 is my favorite number and the number 153 won the lottery yesterday,” most people would similarly not call that “knowledge.”  Yes, you believe something that is true, but you don’t know that it’s true… you just coincidentally believed something that turned out to be true by accident.

That’s how most people think of the term “knowledge.” So when Rush Limbaugh says that there were people who “knew immediately” that the riots were preplanned terrorist attacks, this presents a problem. Sure, there might have been people who believed that they were terrorist attacks. And after all of the research and investigation is done, it might turn out that the riots actually were preplanned terrorist attacks.  But even if there were people who believed it, and even if those people turn out to be right, to claim that they “knew” that the riots were terrorist attacks is to imply a whole other assumption: namely, that they had some kind of evidence or good reason to believe it.

Did they?

This is where we–meaning people who are not actually members of the government and official security organizations who monitor and deal with these things–simply don’t have enough information to know for sure. President Obama would like us to believe that there wasn’t enough evidence initially to conclude that it was an organized attack, and that therefore nobody could possibly know “right away” that it was an attack, because the data was still being gathered. In other words, it was impossible for anyone to “know” either way, because regardless of whether you believed it or not and regardless of whether it was true or not, there was simply not enough information to have a good reason to believe it. You might have the belief, but it would not be justified.

Rush Limbaugh, on the other hand, wants you to believe that the people in power had enough information to be justified in believing that it was a preplanned terrorist attack… they are simply covering it up.

Is that belief justified?

2) What is your null hypothesis?

Of course, it’s possible that Rush Limbaugh doesn’t think that officials on the ground had access to some kind of secret privileged information that made belief in a terrorist attack justified.  It’s possible that, instead, Rush Limbaugh is treating “terrorism” as the null hypothesis for violence in the middle east.

A “null hypothesis”, in any situation, is the thing that you believe by default. You do not need evidence to believe the null hypothesis, because the null hypothesis is by definition that thing that you believe when there isn’t any evidence to contradict it.  You need evidence to reject a null hypothesis, but without evidence the null hypothesis is what you believe.

In order for a null hypothesis to be valid, though, it really does have to be a “nothing hypothesis”: a hypothesis that there is no special thing going on or no special thing to explain.

A null hypothesis is something like “unicorns don’t exist.” In science, you don’t need evidence that unicorns don’t exist. In the absence of evidence that unicorns do exist, the hypothesis that “unicorns don’t exist” is the thing that you believe by default.

This is why stupid and snarky claims in religious debates like “You can’t prove that God doesn’t exist!” are so enraging for atheists. It is a nonsense argument. The idea that you would believe that something exists just because you don’t have evidence that it doesn’t goes against the very core of the scientific “null hypothesis” idea. You could use the same logic to say that you believe that the universe was created by a 3-million foot tall Italian Pizza Man… because after all, you have no specific evidence that is was not.

So for something to be a rational and sensible “default” hypothesis, it has to be a “nothing to see here” type of explanation. In this scenario, the rational “null hypothesis” would be that there was no plan, there was no organization, there was no forethought… the riots in the middle east were just something that happened.

That’s a proper null hypothesis.

Of course, there is some (circumstantial) evidence against that. The timing, for example: the  fact that it happened on September 11th.  That may be a little bit of evidence to suggest that it was more than a coincidence. But is it enough? Do we have enough evidence to “reject the null hypothesis”  (i.e. no plan) and accept the alternative (i.e. a planned terrorist attack)?  Pretty clearly not.

Or at least: no such evidence existed “immediately” when the events first happened. That is why it doesn’t make sense to say that anyone “knew” that it was a terrorist attack “immediately” … there simply wasn’t enough evidence, at that time, to justify rejecting the null hypothesis.

But the reason why Rush’s argument sounds so convincing to conservatives is because they are perfectly happy to have “It was terrorism!” be the null hypothesis — the default thing that they believe — any time there is violence in the middle east.  Conservatives, and specifically the audience to Rush’s show, have no problem with the idea that if there is violence in the middle east, they can assume that it is terrorism unless they get specific evidence to the contrary.

In short: they don’t understand the idea of a null hypothesis, and they don’t understand what it means for “knowledge” to be justified true belief.