From Robyn Dawes's Rational Choice in an Uncertain World:
Post-hoc fitting of evidence to hypothesis was involved in a most grievous chapter in United States history: the internment of Japanese-Americans at the beginning of the Second World War. When California governor Earl Warren testified before a congressional hearing in San Francisco on February 21, 1942, a questioner pointed out that there had been no sabotage or any other type of espionage by the Japanese-Americans up to that time. Warren responded, "I take the view that this lack [of subversive activity] is the most ominous sign in our whole situation. It convinces me more than perhaps any other factor that the sabotage we are to get, the Fifth Column activities are to get, are timed just like Pearl Harbor was timed... I believe we are just being lulled into a false sense of security."
Consider Warren's argument from a Bayesian perspective. When we see evidence, hypotheses that assigned a higher likelihood to that evidence, gain probability at the expense of hypotheses that assigned a lower likelihood to the evidence. This is a phenomenon of relative likelihoods and relative probabilities. You can assign a high likelihood to the evidence and still lose probability mass to some other hypothesis, if that other hypothesis assigns a likelihood that is even higher.
Warren seems to be arguing that, given that we see no sabotage, this confirms that a Fifth Column exists. You could argue that a Fifth Column might delay its sabotage. But the likelihood is still higher that the absence of a Fifth Column would perform an absence of sabotage.
Let E stand for the observation of sabotage, H1 for the hypothesis of a Japanese-American Fifth Column, and H2 for the hypothesis that no Fifth Column exists. Whatever the likelihood that a Fifth Column would do no sabotage, the probability P(E|H1), it cannot be as large as the likelihood that no Fifth Column does no sabotage, the probability P(E|H2). So observing a lack of sabotage increases the probability that no Fifth Column exists.
A lack of sabotage doesn't prove that no Fifth Column exists. Absence of proof is not proof of absence. In logic, A->B, "A implies B", is not equivalent to ~A->~B, "not-A implies not-B".
But in probability theory, absence of evidence is always evidence of absence. If E is a binary event and P(H|E) > P(H), "seeing E increases the probability of H"; then P(H|~E) < P(H), "failure to observe E decreases the probability of H". P(H) is a weighted mix of P(H|E) and P(H|~E), and necessarily lies between the two. If any of this sounds at all confusing, see An Intuitive Explanation of Bayesian Reasoning.
Under the vast majority of real-life circumstances, a cause may not reliably produce signs of itself, but the absence of the cause is even less likely to produce the signs. The absence of an observation may be strong evidence of absence or very weak evidence of absence, depending on how likely the cause is to produce the observation. The absence of an observation that is only weakly permitted (even if the alternative hypothesis does not allow it at all), is very weak evidence of absence (though it is evidence nonetheless). This is the fallacy of "gaps in the fossil record"—fossils form only rarely; it is futile to trumpet the absence of a weakly permitted observation when many strong positive observations have already been recorded. But if there are no positive observations at all, it is time to worry; hence the Fermi Paradox.
Your strength as a rationalist is your ability to be more confused by fiction than by reality; if you are equally good at explaining any outcome you have zero knowledge. The strength of a model is not what it can explain, but what it can't, for only prohibitions constrain anticipation. If you don't notice when your model makes the evidence unlikely, you might as well have no model, and also you might as well have no evidence; no brain and no eyes.
Okay, replace my earlier definition of E with
Do you agree that, under some priors, you could have p(Q|E) > p(Q) and p(R|E) > p(R), even though Q implies not-R?
Set aside the question of whether these are reasonable priors. My point was only this; Warren didn't make the simple mistake with the probability calculus that Eliezer thought Warren made. He wasn't simultaneously asserting p(H|E) > p(H) and p(H|~E) > p(H). That would be wrong under any prior, no matter how bizarre. But it's not what Warren was doing.
What Warren said is consistent with coherent Bayesian updating, even if he was updating on a bizarre prior. It might have been wrong to put a high prior probability on subversive activity, but the probability calculus doesn't tell you how to pick your prior. All I am saying is that the Warren quote, in and of itself, does not constitute a violation of the rules of the probability calculus.
Maybe Warren committed such a violation earlier on. Maybe that's how he arrived at such a high prior for the existence of subversive activity. But those earlier steps in his reasoning aren't laid out before us here, so we can't point to any specific misapplication of Bayes's rule, as Eliezer tried to do.
I don't like the way you describe that. It is confusing. The evidence is subversive activity. You cannot go out and look for no subversive activity, that makes no sense. You have to look for subversive activity. I'm not sure why you're fighting so hard for this point, since not finding something suggests just as much as finding something does. The only reason I suggest a change is for clarity. I don't want to think about no subversive activity and not no subversive activity, I want to think abo... (read more)