3 ms·

If you want to understand the "huh?" reactions you're getting, you might want to wrap your mind around the idea that it's effectively impossible to prove a nega

by crusso 14y ago

If you want to understand the "huh?" reactions you're getting, you might want to wrap your mind around the idea that it's effectively impossible to prove a negative. Take a look at Russell's Teapot: http://en.wikipedia.org/wiki/Russells_teapot http://en.wikipedia.org/wiki/Russells_teapot Follow that with careful study of Occam's Razor: http://en.wikipedia.org/wiki/Occams_razor http://en.wikipedia.org/wiki/Occams_razor

coldtea 14y ago
>it's effectively impossible to prove a negative. Really? "There is no cat in this box". Opens box. No cat inside. Proved.
- crusso 14y ago
  You're totally correct. I should also consider the fact that you can prove that negative voltages exist. The opportunities for pedantically ignoring the context of the discussion at hand are endless.
  - coldtea 14y ago
    You didn't stop at the "discussion at hand". You invoked the general principle of "proving a negative" and Russel's teapot (which is rather naive reasoning in itself). Well, proving a negative is certainly feasible: >In some circumstances it can be safely assumed that if a certain event had occurred, evidence of it could be discovered by qualified investigators. In such circumstances it is perfectly reasonable to take the absence of proof of its occurrence as positive proof of its non-occurrence. http://en.wikipedia.org/wiki/Evidence_of_absence http://en.wikipedia.org/wiki/Evidence_of_absence You might want to check the whole Wikipedia lemma on "Proving a negative" before retelling old wives tales: http://en.wikipedia.org/wiki/Proving_a_negative http://en.wikipedia.org/wiki/Proving_a_negative
mistercow 14y ago
I'm sort of annoyed by the notion that "you can't prove a negative". It's true when we're talking about empirical evidence, but it's only a specific case of "you can't prove anything". There's nothing at all special about negatives. It's just that at no point can we ever say "Well, that's it, our confidence in this belief is now 100%" (even though we can get damned close). Of course, sometimes people say "proved" to mean "supported by evidence to the point that all reasonable people should believe it". In that case, you can prove negatives just fine. I think that the "can't prove a negative" thing must originate from the subtler "it's hard to confirm the null hypothesis", but even that one isn't a hard rule, so long as you have lots of statistical power.

by crusso 14y ago

coldtea 14y ago

>it's effectively impossible to prove a negative. Really? "There is no cat in this box". Opens box. No cat inside. Proved.

crusso 14y ago
You're totally correct. I should also consider the fact that you can prove that negative voltages exist. The opportunities for pedantically ignoring the context of the discussion at hand are endless.
- coldtea 14y ago
  You didn't stop at the "discussion at hand". You invoked the general principle of "proving a negative" and Russel's teapot (which is rather naive reasoning in itself). Well, proving a negative is certainly feasible: >In some circumstances it can be safely assumed that if a certain event had occurred, evidence of it could be discovered by qualified investigators. In such circumstances it is perfectly reasonable to take the absence of proof of its occurrence as positive proof of its non-occurrence. http://en.wikipedia.org/wiki/Evidence_of_absence http://en.wikipedia.org/wiki/Evidence_of_absence You might want to check the whole Wikipedia lemma on "Proving a negative" before retelling old wives tales: http://en.wikipedia.org/wiki/Proving_a_negative http://en.wikipedia.org/wiki/Proving_a_negative

mistercow 14y ago

I'm sort of annoyed by the notion that "you can't prove a negative". It's true when we're talking about empirical evidence, but it's only a specific case of "you can't prove anything". There's nothing at all special about negatives. It's just that at no point can we ever say "Well, that's it, our confidence in this belief is now 100%" (even though we can get damned close). Of course, sometimes people say "proved" to mean "supported by evidence to the point that all reasonable people should believe it". In that case, you can prove negatives just fine. I think that the "can't prove a negative" thing must originate from the subtler "it's hard to confirm the null hypothesis", but even that one isn't a hard rule, so long as you have lots of statistical power.