The Scientific Method

by Luke Muehlhauser on November 21, 2009 in Reviews,Science

I’m blogging my way through Sense and Goodness Without God, Richard Carrier’s handy worldview-in-a-box for atheists. (See the post index for all sections.) Last time, we finished the section on What There Is. In the naturalist worldview, there is the physical universe, voluntary free will, abstract objects, consciousness, meaning, reason, emotion, and spirituality.

So you don’t have to give up much to be a naturalist. But there are some things that don’t exist according to naturalism. We begin our look at What There Isn’t with Carrier’s section on the paranormal vs. science.

The problem with the paranormal is that every time it is investigated in any kind of rigorous way, it turns out to be the product of fraud, self-deceit, sensory illusion, or myth- making. Rather than rebut every paranormal claim individually, Carrier suggests some resources. I will recommend you start with Skeptoid,, and James Randi’s An Encyclopedia of Claims, Frauds, and Hoaxes of the Occult and Supernatural.

But let us consider supernatural claims more generally. Carrier writes:

There are generally two approaches to defending belief in the supernatural. The first is to claim there are facts that support it… But this exposes such claims to scientific and historical study, and when scientists and historians come knocking, either of two things always seem to happen: those making the claim retreat or refuse to submit their evidence to scrutiny, or their evidence turns up bogus or groundless.

…But there is a second approach, bypassing science altogether… claiming to get at the truth by sitting in a chair and just reasoning to it…

You may have noticed that of the six methods [for truth-finding], I place metaphysics dead last. Though in my scheme reason does come first, what i mean by reason is simply the analysis of our own ideas about things, which can only teach us truths about those ideas, not about whether those ideas apply anywhere in the world outside our imagination. When it comes to that question, I place science in first place…

Why? What’s so great about the scientific method for discovering truth about the external world?

The “Scientific Method”

Science is a three-step process: adduction, deduction, induction. We identify a problem, gather relevant data, form a hypothesis, figure out what predictions are entailed by that hypothesis, then test those predictions.

In other words, we “adduce” a hypothesis from our the data we’ve gathered, then we “deduce” what would have to be the case if the hypothesis were true, and they we use empirical (“inductive”) methods to test if those predictions hold up.

The best reason to use this method is that it works so damn well. Nothing has so thoroughly transformed the world like the working knowledge that comes in vast quantities every day from using this method in every field of study about our universe. That’s why scientists usually don’t care if philosophers can find a logical justification for what they do. When the philosopher explains to the scientist why falsification can’t demarcate science from non-science, the scientist retorts:

Oh yeah? When was the last time you cured smallpox or sent a man to the fucking moon? Now either you can let me do it the way that works and I’ll cure blindness in another ten years, or we can do it your way and run in circles for another thousand years.

Still, let’s think about why science works so well. This method of adduction, deduction, and induction has three major virtues, says Carrier:

First, it accommodates creativity. Looking at nature and coming up with guesses about how things work… is driven from the start by curiosity and imagination, two things humans have in spades. Second, it doesn’t dare let you get down to the business of proving anything without first analyzing the meaning of our guesses, applying rigorous logic to our ideas, forcing us to fully confront just what our theories would entail if they were true or false… it unites creativity with our most powerful tool of all: reason. Third, it doesn’t let you get away with claiming anything without proof – even better, it especially likes an unusual sort of proof: not mere evidence, but evidence no one expected.

Until Einstein’s Theory of Relativity, no one would have even thought that matter could be transformed into energy or that gravity bends light, and so it was a surprise to find that they did, just as Einstein’s theories predicted. Our ability to predict things we could never have otherwise guessed, from when volcanoes will erupt, to the outcome of mixing two new chemicals, is a daily vindication of the correctness of our scientific knowledge….

When a scientific theory predicts a fact no one has ever seen before… then it’s an increasingly good bet it’s right. You can’t get better proof than that. No one can claim scientists are just seeing what they want to see… if their explanations of existing facts turn up, and explain, totally new facts. When you see that, repeated again and again, only the universe itself could be at work.

But the devil’s in the details. There are many ways to apply the standard method of adduction, deduction, and induction. And the particular methods used in one field may not be relevant to another field. But scientists continued to toss out those methods that give misleading results, and they keep the methods that give them reliable results. And this is the real strength of science: it not only checks facts for truth but checks its own methods for accuracy.

And that’s the problem with supernatural speculation. It offers no way to check its methods or facts for accuracy.

Science is also highly skeptical. Science depends on the encouragement of skepticism. Science requires that you submit your results for peer-review and give others ample chance to disprove your theory – all of which they are eager to do, for science gives its highest honors to those who disprove reigning doctrines.

Previous post:

Next post:

{ 12 comments… read them below or add one }

Chuck November 21, 2009 at 6:50 am

This is sort of OT, but does anyone know if String Theory has produced any testable predictions?


Reginald Selkirk November 21, 2009 at 7:20 am

NOVA interviews Edward Witten, a proponent of String Theory and generally acknowledged to be a Really Smart Guy:

NOVA: It seems like the standard criticism of string theory is that it isn’t testable. How do you respond to that criticism?

Witten: One very important aspect of string theory is definitely testable. That was the prediction of supersymmetry, which emerged from string theory in the early ’70s. Experimentalists are still trying to test it. It hasn’t been proved that supersymmetry is right. But there is a very precise relationship among the interaction rates of different kinds of particles which follows from supersymmetry and which has been tested successfully. Because of that and a variety of other clues, many physicists do suspect that our present decade is the decade when supersymmetry will be discovered. Supersymmetry is a very big prediction; it would be interesting to delve into history and try to see any theory that ever made as big a prediction as that.

NOVA: What are some of the other ways that string theory could be confirmed experimentally?

Witten: There are a lot of conceivable ways we could get experimental information that would help with string theory. Explorations of cosmology, studying the cosmic background microwave radiation and hopefully finding gravitational waves left over from the big bang and studying their properties are very plausible avenues for eventually testing string theory, although there isn’t yet to my thinking a satisfactory theoretical understanding of what to expect.

But it’s conceivable that the big bang could have produced a string so large that it would be present in today’s universe and visible in telescopes, perhaps discoverable by the satellites that are now mapping out the microwave sky. If that were discovered, it would be a dramatic confirmation of the existence of strings. Still, that’s a story that will develop over the next decade or two as the experiments progress and conceivably as the theory progresses.

(Web page copyrighted 2003)
Note that whether String Theory has produced testable predictions, and whether those predictions have actually been tested are two different questions.

On the other side: A Great Unraveling
NYTimes review of books by Lee Smolin and Peter Woit criticising String Theory on the very point of testability.


Reginald Selkirk November 21, 2009 at 7:23 am

Wikipedia says

Supersymmetry is also a feature of most versions of string theory, though it can exist in nature even if string theory is incorrect.

I.e. String Theory may support Supersymmetry, but Supersymmetry could exist without a valid String Theory, so it is not clear that successful testing of Supersymmetry would be a validation of String Theory.


lukeprog November 21, 2009 at 7:26 am

Once again, Reginald has the links!

Thanks, Reginald.


Reginald Selkirk November 21, 2009 at 7:34 am

Your welcome. The hard part is getting links from reliable sources with accurate information, but that are accessible to the relevant audience. So for example, I would try not to link directly to an article in Physics Review, but instead find intelligent and accessible commentary which describes that article.


Bradm November 21, 2009 at 11:43 am

“When the philosopher explains to the scientist why falsification can’t demarcate science from non-science, the scientist retorts …”

The answer your scientist gave is fine in and of itself. But all it really says is that science “works” (whatever that means … religion “works” for a lot of things too), not that it leads to anything that is true. The question that some philosophers seem to be asking is not whether science works to help bring people to the moon, but whether science works to help us find truth. The answer you gave has little to do with the question philosophers are concerned with.


Bradm November 21, 2009 at 12:02 pm

And, for the record, there are plenty of scientists say that the hypothetico-deductive method that you describe isn’t an accurate description of how all science is done. For one example, see this paper by the Nobel prize winning physicist Frank Wilczek:

From the paper:

“We put aside the “not wrong” complicated models with spontaneous supersymmetry breaking, and wrote a short paper that, taken literally (with unbroken super-symmetry), was wrong. But it presented a result that was so straightforward and successful that it made the idea of putting gauge symmetry and supersymmetry unification together seem (maybe) right.

This little episode, it seems to me, is 179 degrees or so out of phase from Popper’s idea that we
make progress by falsifying theories. Rather in many cases, including some of the most important,
we suddenly decide our theories might be true, by realizing that we should strategically ignore
glaring problems. ”

In other words, instead of identifying a problem, gathering data, forming a hypothesis, etc. they identified a problem, ignored the data, made a hypothesis that was in conflict with the data because it was simpler, published, and then waited until they could test their hypothesis.


Bill Maher November 21, 2009 at 2:44 pm
lukeprog November 21, 2009 at 6:44 pm

Michio Kaku is the Japanese Neil deGrasse Tyson.


Bill Maher November 21, 2009 at 8:16 pm

I would say he is the Japanese-American Stephen Hawking and Tyson is the new Carl Sagan. :-)

lukeprog: Michio Kaku is the Japanese Neil deGrasse Tyson.  


Ben November 25, 2009 at 8:04 am

I’m sure theistic philosophy will manage to produce something worthwhile one of these days.


Roman December 8, 2009 at 10:36 am

Cool, I like this post. The scientific method is just so darn clever!
I wonder how much philosophers can use this method. Some interesting questions that I thought of when reading this:

1)Do philosophical theories ever make predictions which are testable empirically?

2)If not, how are they testable? Are they testable only by their logical consistency? Or is agreement or disagreement with our pre-philosophical ideas a worthy test? Or is there some other (third) way of testing them?


Leave a Comment