On the impossibility of proving a negative

[catsittingstill]

So I was having an argument the other day, and a phrase popped up:

"It is impossible to prove a negative."

At the time I just let it go by.  This phrase pops up a lot in discussions of this type (anything touching on atheism) and it's generally accepted as true and it didn't occur to me to think more deeply about it.  But I was puttering around this morning sweeping the floors and I started thinking about it.

It is impossible to prove a negative.

Let's move it out of knee-jerk territory (at least for many people atheism is knee-jerk territory) and think about some other negatives.

"My plates are not trying to murder me."

"There is not a green unicorn sitting in my lap as I type this."

"The floor will not open up and swallow me when I get up off this chair."

Is it seriously not possible to prove any of these things?  Does the fact that my plates have never tried to murder me before; that I have never heard of plates murdering anyone, that plates have never, in fact, been observed to move of their own accord before as far as I know, much less to move with intent--do those things mean nothing about my plates' current plans with regard to murder?

Well, hmm.  Sometimes something that never happened before happens.  Someone writes a song that never existed before, someone invents something that never existed before...  I guess one can make a case that just because there is no evidence whatsoever that plates have ever murdered anyone or even moved on their own doesn't mean it absolutely can't happen.

So, one could be "aplatistic" on the subject of plates: "I am certain my plates are not trying to murder me."  One could be "platistc" on the subject of plates "I am certain my plates are trying to murder me."  Or one could be "ambiplatistic" on the subject of plates "My plates may be trying to murder me; there is no way for me to be certain."

I suppose one could, technically, make the argument that the ambiplatitisic stance is the only truly correct one.  I mean, if you can't prove that your plates aren't trying to murder you, and you have no evidence that your plates are trying to murder you, then the stance that your plates may be trying to murder you covers both possiblities.

But which one produces more reasonable behavior in real life?  The aplatistic stance would produce the behavior most people (I think) exhibit around plates--we behave as if it never crossed our mind that our plates might try to murder us; we take no precautions whatsoever; we walk through the kitchen without a care, and turn our backs on the plates without a second thought, and handle them with causual, indeed contemptuous, lack of concern.  The platistic stance?  I'm guessing it would involve smashing plates, or possibly banning them from the house.  The ambiplatistic stance--I guess one would tiptoe through the kitchen only when necessary, keeping a wary eye on the plates the whole time.  Or possibly simply ban them from the house, to be on the safe side--though that might be seen as unjust, since there is no evidence that the plates are planning to murder you.

So is it unreasonable to think that a person who smashes her plates to keep them from trying to kill her is...well, psychologically unusual?  Arguably to the point where her ability to live a normal life is compromised?  Is it unreasonable to say that being ambiplatistic--tiptoeing through the kitchen and never turning one's back on one's plates--is silly? 

And if not, does that mean that sometimes it doesn't matter whether or not one can prove a negative?

Comments

[tigertoy.livejournal.com]

Absolute logical proof is only possible in the abstract mathematical realm where we precisely state and agree on the rules ahead of time. In math-land, I can absolutely prove that 2+2=4. I can absolutely prove that there are no even prime numbers larger than 2. It has also been rigorously proven that no map can be drawn that can't be colored with only 4 colors, or that no integers a, b, c, and n, a, b, and c > 0 and n > 2, such that a^n + b^n = c^n (though those proofs are way beyond me). In the realm where it's truly possible to prove anything, it is possible to prove a negative, though it may be more difficult than disproving a negative by providing a counterexample.

In the real world, there are no axioms. There are only working hypotheses and heuristics that we generally agree on, and we often find that when we try to state them exactly we don't agree on them. And there is no truly absolute evidence; all observational data is subject to fallible perceptions and fallible memory. If I interview your plates and they speak and tell me that yes, they are in fact trying to kill you, that doesn't actually prove the platist proposition -- they could be lying, or I could be nuts.

Much more vexing, though, is the fact that human beings don't evaluate and maintain their beliefs through pure logic. We filter evidence that is offered to us based on how it fits with what we already believe and how much we want it to be valid. The more emotionally invested we are in a position, the less likely we are to honestly consider something that appears to contradict our position. And there are few questions that we're more emotionally invested in than the existence and nature of a deity.

[catsittingstill.livejournal.com]

Good points, especially the last paragraph.