On Evidence

Simen has a few wonderful posts about atheism, including the best put-down of Dawkins’ nascent radicalism I’ve seen. He also has a very good post about the difference between atheists, agnostics, and theists.

In the latter post’s comment thread, a presumably theistic comment asks why atheists believe books that say the Earth spins but not a book that says God exists. It’s an argument I saw a few years ago, back when I was a regular on Hofesh: “Why do you believe Napoleon existed?” Since the unreflective answer is “Because history books say so,” it superficially exposes an analogy between the Bible and mainstream science.

In fact, Napoleon’s existence is more than just something written in a book. There’s a multitude of independent accounts of people who knew him, an even greater multitude of independent accounts of people who fought against or under him. There are maps he used when devising strategy, and contemporary newspapers referencing him. That’s not extraordinary evidence, but the claim that a short-statured Corsican came to be Emperor of France isn’t extraordinary either.

Now, compare Napoleon with an apocryphal historical figure, say Jesus. There’s only one independent source saying Jesus existed, Mark; the other Gospels are derivatives of him. Paul predated Mark, but he refers to Jesus in mythical terms, saying nothing about his life. Mark was not a contemporary of Jesus: he wrote his Gospel in the 70s, while Jesus supposedly died in 30. Mark’s account of Jesus is full of impossible religious miracles, while real historical accounts that leave Jesus out entirely, like Josephus’s, are not.

There is a gray area between the two extremes, of course. There are four independent pieces of evidence Socrates existed: a historical record showing Athens executed someone by that name in 399 BC, and the writings of Plato, Xenophon, and Aristophanes. The writings by and large contradict one another, and although it’s probable that someone by the name existed, it’s entirely plausible Plato invented Socrates’ philosophy. But I digress.

The actual form of the theistic argument fielded in the comment to Simen’s post is not about history, though, but about science: “How do you know the Earth spins?” The best answer to that takes the form of Popper’s scientific principles, which, while not perfect, are by and large true. The objections to Popper that aren’t batshit crazy tend to apply only to established theories, but not so much to statements of scientific fact.

Scientific hypotheses, like “the Earth spins,” make an enormous number of falsifiable predictions that can be independently verified. In case of the statement about the Earth’s spinning, its converse also makes an equally large number of predictions that are falsified: that the Sun will be found to revolve around the Earth, that there will be no centrifugal force, that there will be no Coriolis effect, and so on.

Usually, independent verification takes the form of an appeal to better-established scientific principles. That’s why the falsification of a theory is a big deal: it tends to falsify a lot more than just a single statement. In the case of the Earth’s spinning, the easiest appeal is to the parallax effect, which shows it revolves around the Sun. That’s not why scientists accept heliocentrism, but the last geocentric theory to remain standing, Tycho’s, was based on the falsifiable prediction that no parallax could be observed. A more theoretical appeal will be to Newton, whose theory had a ton of empirical backing in its own right.

It’s this stack of scientific theories, each presuming a more basal one’s soundness, that causes science and religion to superficially appear comparable. The average person barely knows science, so to him, it really is just something he read in a textbook or heard from a teacher. Most people who are asked how they know the Earth spins can’t give any satisfactory answer.

The Enlightenment didn’t make the common people more rational; it made the scientific and philosophical establishment more rational. A century of developments in public education hasn’t been able to extend knowledge of scientific evidence to more than a few percent of any country’s population. It so happens that most people in the modern world believe in true things, but it’s sheer luck. Three hundred years ago, they’d believe in witches.

The increasing complexity of scientific theories makes things even worse. In 1750, an educated person could have a grasp of the entirety of human knowledge available in his locale. In 1900, a scientist could know everything in his field. Right now a scientist is restricted to one subfield; when he knows two, it’s usually as part of a fusion of fields, like biostatistics.

I know the precise evidence for the Earth’s spinning, but I know the evidence for relativity only in general and without the math that distinguishes it from woo. The only way I can know PZ Myers isn’t lying in his science posts is that there are other evo devo-minded commenters who’d check him, who I in turn trust because I have no reason to believe that scientists are only honest when I know enough to check their work.

In contrast, I do have reason to believe scientists are materially different from theologians. Theologians may be individually honest, but their entire enterprise is predicated on premises I know to be shaky. And while scientists concentrate on accumulating evidence for their theories, theologians content themselves with hermeneutics. Therefore, I have grounds to dismiss religion as false, while possessing no similar grounds to dismiss fields I know little about except that they’re scientific, like neuroscience.