1

Hitchens's razor states, "What can be asserted without evidence can also be dismissed without evidence," while Carl Sagan argues, "Absence of evidence is not evidence of absence." What are your thoughts on this?

6
  • 1
    Whose perspective is "correct" depends on one's perspective. "What are your thoughts" questions are off-topic here. And one-liners out of context are not particularly cogent, these two can be interpreted to both be correct, for example, see When is absence of evidence not evidence of absence?.
    – Conifold
    Apr 13 at 7:38
  • 3
    Absence of evidence frequently is evidence of absense (not always though). Especially in cases where you're explicitly looking for evidence that you should expect to find. People misuse this phrase a lot and, speaking in a bayesian sense, it's just often not right.
    – TKoL
    Apr 13 at 7:39
  • 1
    Hitchen's razor brings up ambiguity and depending upon the context it is being used might fall flat, Carl's on the other hand is a logical observation regardless of the scenario. Also personally, I think Carl Sagan was more of a scientific and philosophical intellectual than Hitchens was, Hitchens was the face of media more focused with politics and journalism.
    – How why e
    Apr 13 at 8:08
  • Does this answer your question? When is absence of evidence not evidence of absence?
    – g s
    Apr 13 at 15:24
  • 1
    (See Quote Investigator for the origins of Sagan's quote.)
    – gidds
    Apr 13 at 17:56

2 Answers 2

6

They are two aphorisms intended to make a point, not to be considered the last word. You will find similar superficial contrasts in lots of everyday maxims, such as 'penny wise, pound foolish' which might be taken as conflicting with 'look after the pennies and the pounds will look after themselves'.

Hitchens' phrase is related to the principle of the burden of proof, pointing out that unsubstantiated claims are just that, and there is no obligation on anyone to prove that they are incorrect. If you feel entitled to claim X without any justification, I should feel entitled to suppose not X with an equal lack of justification.

Sagan's point is different. He is saying that a lack of evidence for X doesn't prove X must be wrong.

Given that, there isn't really a conflict between the two statements, as they are making different points. For there to have been a conflict, Hitchens would have to have said something like 'what can be claimed without evidence can be proved wrong without evidence', which of course would be a ludicrous claim.

1

"What can be asserted without evidence can also be dismissed without evidence"

This is a recipe for confirmation bias, as it is likely to be used to dismiss ideas that you don't like without giving them any attempt at consideration/understanding. Good polemic/rhetoric "debate", but bad for good-faith truth-seeking discussion. Note that there will be some things where evidence simply isn't available, but that doesn't mean we can't use reason to form an opinion about them, e.g. eternal inflation-style multiverses, whether there are stars beyond the observable universe etc.

"What can be asserted without evidence is unlikely to convince those who don't already agree"

would be better, but not as good as a soundbite.

"Absence of evidence is not evidence of absence"

This is factually incorrect stated as a general rule. For example, if you have a theory that says purple swans are ubiquitous in Norfolk, then if you never witnessed a purple swan in Norfolk, then that would be evidence against the theory (and evidence for it's negation). It would be evidence of absence under that hypothesis. If the theory was "there is A purple swan in Norfolk", then not seeing one would be unsurprising.

If you do a Bayesian analysis, then in the first case if you didn't see a purple swan, then the Bayes factor would be heavily in favour of the theory being false (because you would be very unlikely to fail to spot a purple swan if the theory were correct, and highly likely to fail to spot one if it was false). For the second theory, the Bayes factor would be close to one as you would be more or less equally unlikely to see a purple swan whether the theory was true or not.

So absence of evidence is evidence of absence, but it may be negligible evidence of absence.

However, as Marco (+1) says "They are two aphorisms intended to make a point, not to be considered the last word." They are both rhetoric, I like Sagan's better because it seems more self-skeptical?

BTW there is a branch of statistics called "survival analysis", which is used to estimate the time-to-event (often death). It often involves "censored" data, where the subject was monitored for a period but the event didn't happen. The fact that you didn't observe the event still provides useful evidence about the problem, it tells you the most probably time-to-event was longer than you observed the subject.

Not the answer you're looking for? Browse other questions tagged .