[Previous] Physical and Moral Truth | Home | [Next] Steve Yegge on Epistemology


Anti-Popperians (APs) lack a relentlessly critical and imaginative attitude. Here are four examples followed by explanations:

APs will observe water on mars and say "the theory that there is water on mars is now confirmed."

APs think correlation implies causation.

APs come up with mathematical or logical or deductive "proofs" of things and can't think of any ways they might have made a mistake.

APs think they induce things and then can't see how their set of observations is consistent with alternative explanations.

When you tell them that water-on-mars is not confirmed, in the sense that it may be false and their observation doesn't tell us one way or another, they balk. They say that the only theory compatible with the observation is that there is water on mars. They have a failure of imagination. You have to point out other theories compatible with their observation -- they won't brainstorm them -- and then after you do they will still complain. They will usually say those alternatives are not plausible, and base this on lengthy and controversial arguments. That means if water-on-mars was really confirmed, it was confirmed by the combination of the observation and the lengthy arguments. Even if they are right the observation played only a small role. Their only way out is to declare all of their arguments "obvious" and anyone who disagrees "stupid", so that the arguments only have a small role and the observation is the main event. And, of course, they do that: they say the arguments are trivial and I am an idiot.

APs won't explicitly admit that they think correlation implies causation, but they still do it all the time. They have a lack of imagination when it comes to alternative explanations for correlations. They are the people who find a correlation between wearing heavy coats while driving and car accidents and want to ban coats. They say coats are dangerous, look at the evidence. They can't think of any plausible ways in which it might turn out that coats are not dangerous. They don't notice that this "evidence" equally well is evidence of explanations which assert coats are not dangerous. They are missing the fact that rain causes both coat wearing and elevated accident rates. They are also the people who read a few studies and think genes cause everything, without noticing that the observations made in the studies are consistent, in a variety of ways, with genes not controlling those traits (and also they don't notice that the studies only study differences in traits between people, not whole traits, and various other rather crucial things).

With a mathematical proof, APs can't think of anything that could have gone wrong. This is despite the fact that learning math takes extensive study, and the majority of students understand very little of it and make mistake after mistake. They can't come up with the possibility that one of their teachers wasn't really all that great at math and taught them a falsehood. They can't come up with the possibility that people are different and have different understanding of language and of math and that this can lead to confusion and misunderstanding. They can't come up with the possibility that they spaced out for 15 minutes while doing the proof and didn't properly double check one section and it contains an error. They can't come up with the possibility that they have a disease and their memory is deteriorating. Or that they had a muscle spasm and wrote an incorrect number on the page at some point. Or that they are living in a simulation and the programmer is messing with them. Or that they were distracted by their recent divorce and made a mistake. Or just plain that people make mistakes, all the time, without realizing it. And they conveniently forget that sometimes mathematical and logical proofs of the past, which people were very sure of, have turned out to be flawed.

APs would take a bunch of temperature readings, which go 70 72 70 68 70 71, and induce that the next temperature reading will probably be between 67 and 73. When you tell them it might not be they will say that's very unlikely. They can't think of any plausible way that would happen. So you say maybe those readings were all taken at 8pm and the next one will be taken at 1pm in a different location. And they say you're cheating. But you aren't. The data about the temperatures did not include any mention of when or where they were taken. It's their own fault for not thinking to ask; they should learn to brainstorm; they have no imagination. Next they will say a competent scientist won't make a mistake like that; he won't do anything stupid. So now we can see that induction really has two parts. It has inducing and it has not being stupid. Both are required or it doesn't work. And what does "not being stupid" consist of? Ultimately it must consist of the entirety of CR. Nothing else would work. And what does that leave as the role of inducing? Just guessing. You induce (guess) what the observations mean, then subject it to criticism and full CR approach, and then if it turns out it was stupid you have to induce (guess) again, and repeat until you get a guess that stands up to criticism and argument. So it turns out the entire process is either CR or stupid.

Elliot Temple on October 14, 2008

Comments (7)


Your argument agaisnt the so-called third claim of APs is trivial.

Anonymous at 7:37 AM on November 1, 2008 | #1622

Water on Mars a theory?

How do you get that speculating that there may be water on mars counts as a scientific theory.

Someone may have a theory on the formation of mars from water-bearing comets, and there may be another that it formed from dry asteroids, and both theories may predict a certain water content. However just saying I have a 'theory' that there is water on mars comes across as though the 'theorizer' is actually deeply ignorant as to what is and isn't 'scientific'.

John Murphy

Anonymous at 4:50 PM on December 21, 2008 | #1714
I think "There is water on Mars" is no Theory at all. It is just a singular Statement true or false.
Therefore if they find at least a bit of water on Mars, this observational sentence "There is water on Mars" is confirmed and true.

Many a Mars theory would contain next to many other to-be-fact sentences, that there is water on Mars. And the confirmation would make all waterbased Theories of Mars more likely.

sry for bad english. As german i have sometimes problems to find the correct english words.

Anonymous at 1:06 PM on January 22, 2009 | #1727
'Therefore if they find at least a bit of water on Mars, this observational sentence "There is water on Mars" is confirmed and true.'

When you *think* you find water on Mars, you may be mistaken. Maybe it was a different liquid that looks like water. Maybe you were hallucinating. Maybe your colleague was playing a prank on you and put some water from Earth where you would look. Maybe the device for checking the chemical compositions of liquids was malfunctioning or badly designed.

This is not such a big deal. It's OK that we don't have certain knowledge. We don't need it. We can still say we discovered water on Mars, and this can be the best theory available that we think is true, and there may be no competing theories anyone finds plausible.

But it does mean that confirmations are fallible, not "true" in the sense of justified, certainly-infallibly-true belief. One reason this is important is we should always be open to criticism and revision of our ideas if someone disputes them. We have to address the reasons he gives, and never say "you can't question that fact b/c it's an established, confirmed truth".

- Elliot

Anonymous at 1:12 PM on January 22, 2009 | #1728
The observational aperatus may be defective or the theories making them possible wrong.
But if our waterfinding device works correctly, the "There is water on Mars" sentence is true.

Theories are nets of such sentences. Our observation itself is theory dependent. And our observations or conclusions may be wrong.

Confirmation is not proove. We cant proove that there is water on Mars, and therefore have no assured knowledge of this, even if we find large amounts of water many times over and over.

sry again for bad english

Tobias Mueller-Kortkamp

Anonymous at 1:26 PM on January 22, 2009 | #1731
You say that confirmation is not a proof, so it's possible we actually agree.

When Popper says confirmation or verification is not possible, I think he means two main things.

The first is that if a theory is confirmed or verified by an observation, it still might be mistaken. He's insisting on fallibility. No matter how much confirmation you have, it could still be false, so we must always keep an open mind.

The second thing he means is that induction is false, and the general idea of data pointing us to theories is false.

The right way is to make conjectures and explanations, and then use observation data in the role of criticism.

- Elliot

Anonymous at 1:39 PM on January 22, 2009 | #1733
I agree we are in no real controversy.
I looked up a word in dictionary which I can't properly translate.
It states that the word i look is "prove"....don't know if theres a differnce between prove and proove.

A Theory of Mars containing that Mars helds water would prove? the theory.
This means it would have shown its worth for explaining things we observed.

Anonymous at 2:17 PM on January 22, 2009 | #1735

What do you think?

(This is a free speech zone!)