Is absence of evidence evidence of absence?

Great Debates
61

I’ve taken philosophical logic and theory of knowledge at MIT, and also am excessively familiar with mathematical and electronic logic. So your assertion is false on the face of it.

If you decide to return - and everyone disagreeing with you is not necessarily evidence that everyone is blind - could you distinguish seeing no elephant from not seeing an elephant? You seem to be asserting that what is going on in the room is the observation of a non-elephant, which is evidence of the existence of the non-elephant. Since there are infinite non-things in the room, this seems an absurd statement.

Another example from science. The neutrino was proposed mathematically, which gave an impetus to look for it, but was not evidence of its existence. The characteristics of the neutrino were such that we would not expect to see any in our normal lives. This absence of evidence was most definitely not evidence of absence. People then built neutrino detectors at great expense. If there were no neutrinos found in them, then that would be evidence that the neutrinos defined by the theory did not exist. They were found, however. It would have been incorrect to consider that evidence foir neutrinos existing or not existing would have been unchanged if the experiment did not produce results. Only a badly designed experiment would result in this.

We can tell that people are cranks by their refusal to accept the absence of expected evidence of their pet theory as evidence against it. Any number of non-religious examples come to mind.

62

It seems the fundamental disagreement is what the phrase means. Of course, it is tautologically true that no evidence is no evidence.

But I don’t think that’s the way the phrase is commonly used or intended. It seems it’s usually supposed to mean something more like “Lack of evidence for something is not evidence against it”, which is clearly false, in the most general case.

63

I just don’t see the distinction here. If you can grant that absence of evidence when evidence is expected is evidence of absence, how can you not grant the same for scenarios where evidence is not expected, so long as one is not forbidden to reassess?

The absence of observable mice/ mice activity in a room is evidence of its absence whether the search is exhaustive or not. In case evidence is found of the presence of mice, one can always change one’s assessment. In other words, ones tentative assessment that there are no mice in the room is always warranted in the absence of evidence of mice/ mice activity.

64

With apologies to Superfluous Parentheses, there’s so many superfluous negatives in this thread it makes my head spin. I think that I agree with the essence of what you’re saying, within certain limits, but those limits are very important.

If you’ve made some prediction, like that you will find dog poop in the yard where a dog has been kept, and you search for and fail to find dog poop, I don’t think you have an absence of evidence. You have set out to gather evidence and found that the total poop is less than the expected poop. That’s an important finding, and it’s evidence. What you found is an absence of dog poop, not an absence of evidence. This is an essential distinction, and I believe it’s what much of the disagreement in this thread hinges upon. So to the extent that it’s possible with different definitions, I agree with **Strinka **about this.

65

Typically, we consider evidence only in relation to a hypothesis that we’ve got in advance, and we weigh evidence against the probability that we would have also seen it if our null hypothesis were true. The goal when considering evidence is to look for something that would be unlikely if the null hypothesis were false. If we find it, we can disprove the null hypothesis and as a consequence our hypothesis is stronger. With the mice problem, we can arrange it like so:

The hypothesis is that we have mice in the room.
The null hypothesis is that there are no mice in the room.
Evidence that would cause us to disregard the null hypothesis include sightings of mice or mice droppings, hearing squeaks or scratching in the walls, seeing holes chewed in our box of cereal, etc. These are good kinds of evidence because they are less likely to occur if the null hypothesis is true, so finding any of them would quickly eliminate the null hypothesis with fairly high confidence. Once the null hypothesis is eliminated, our hypothesis stands somewhat stronger. We would then want to pick a new null hypo that conforms to our evidence and find a new test that would dismiss the new null hypo. Doing so would make our hypo stronger still. The new null hypo might be that there’s a rat in the room, and the standard for evidence that it’s a mouse rather than a rat which is causing the evidence we had already identified would be higher.

In the context that I’ve outlined, not finding evidence that can eliminate the null hypothesis means that we have no good scientific basis for eliminating it. In this sense then, absence of evidence is not evidence of absence. This is the origin of the phrase as the OP presented it, and the reason why it’s true logically and used scientifically as the standard. It means only that we’ve done a poor job coming up with a hypo and we should work harder.

In life outside of the scientific method we rarely have a hypothesis so clearly in mind, and the definition of the word evidence seems to be shifting to something less rigorous. When we’re standing in our kitchen we don’t have the hypothesis that there’s a mouse in the room and we’re not trying to disprove the null hypothesis that there isn’t. Instead we’re trying to pleasantly prepare our breakfast. I’m tripping up while trying to construct this sentence in a way that the terms fit both definitions, because it’s awkward to write out this way… but in this non-scientific context, the absence of evidence that there is a mouse in the room is cause enough to not bother to hypothesize that there may be one. It’s not evidence of absence even under this situation - it just isn’t. But it may be enough to make us comfortable eating our breakfast not worrying about mice, and isn’t that enough?

66

The evidence of absence holds only at a given time. Any conclusions you draw from it are provisional. My position is that you can’t draw any conclusion at all in the case where no evidence is expected (like the aliens) but your confidence in the absence of the thing grows with the amount of expected evidence that is absent. And only one item falsifies the conclusion of absence, of course.

In terms of black swans, if you’ve never seen any swans you can’t tell anything. Seeing the absence of black swans in one pond is less evidence than not seeing black swans in 1,000 ponds. But it all gets changed when you see your first. So, you do reassess constantly.

ETA: Or as twhitt said, a bit more formally.

67

I don’t know why people get confused about this…

Absence of evidence is evidence of absence, but we don’t know how strong that evidence is without knowing the sample size, whether it was a representative sample, whether methods used to try to gather that evidence were valid or not, and so forth.

Lack of evidence for a particular viewpoint is very strong evidence against it when we have ample evidence contradicting that viewpoint and every logical reason to believe that if supporting evidence existed we would know about it.

Lack of evidence for a belief for which there’s no reliable way to gather supporting evidence and no contrary evidence either isn’t evidence one way or another.

68

I just posted a view contrary to Frank apisa’s in another thread, but I am very much with him on this one.

Let’s leave out microscopic elephants, since elephants and mice were obviously specifically chosen to be objects with a large and small volume respectively. (silly people.) And let’s also assume that it’s a visible object. (why do I have to say that?)

Your senses provide evidence. When you look in the room, your senses tell you that there is no elephant or mouse in any area you can see. The difference is in where the respective objects can hide. Only when you have looked everywhere it can hide can you say it isn’t there.

When I don’t see an elephant, (or a mouse,) that is evidence of absense… everywhere I can see, and only where I can see. Since I can see everywhere an elephant could be, I can conclude there isn’t one in this room. But, that is with evidence. I have the same evidence that the mouse isn’t anywhere I’m currently looking. But I haven’t looked behind the boxes or in the cupboard. The evidence that it isn’t here isn’t evidence that it isn’t there.

An absense of evidence is when you stand in the room for an hour every day and never see a mouse come out of one of those hiding places. It’s waiting for the evidence to come to you. You haven’t looked to rule out those hiding places, but you haven’t seen it come to you. If you start looking in the cupboard and in and behind the boxes, you are gathering evidence.

If you stare at a box, and nothing comes out, that is an absense of evidence for what is in the box. You can’t say what is or isn’t in the box based on what hasn’t come out of it.* If something comes out, that is evidence. If you look in the box and see nothing, that is evidence of absense.

Staring at the box and seeing nothing come out isn’t the same as looking in the box and finding nothing. Absense of evidence is not evidence of absense. And that is the sense in which that makes sense.

*(generally. our knowledge of the size of the box is evidence that nothing larger is inside… although, only if the rules of physics stay the same inside. [so many things we just assume in normal life, that you have to state in these discussions. And I’m sure I’ve missed something someone will nitpick about. Something obvious, like: elephants and mice were chosen to display distinct size characteristics.])

69

I took a “vacation” from this thread for a few. I was having troulbe seeing just what the point of it was.

I have not seen a rhinoceros on my desk all morning. I didn’t see one last week, or the week before. I have never seen a rhinoceros on my desk. So, I guess, it’s safe to say there never was a rhinoceros on my desk.

Am I wrong? Must I believe that there may have been one there at some other time that I was not present to observe it? Was it Schroedinger’s Rhino? Do I have to believe in The Great Invisible Rhino Who Magically Doesn’t Crush The Desk?

Sometimes, just sometimes, there is no evidence of something, simply because it wasn’t ever there.

I think the OP was saying :
If you can’t disprove it, you have to believe in it.
Wrong Answer.

Now where’s my cookie?

70

That’s not his problem at all.

The problem is that people don’t know when they’re gathering evidence. And I felt his pain while reading this thread.

71

And???

We observe (or not observe) things all the time. We don’t have to be consciously or deliberately “searching for evidence or the lack thereof” in order to prove or disprove everything. I guess I missed the point, because I thought he was making some sort of vague religious point, and trying to prove that we are all too stupid to get it (whatever “it” is).

72

I can’t believe this nonsense is still being argued. appleciders gets it perfectly right in just a few words. I really can’t see how anyone with even a basic understanding of logic cannot grasp this. Unbelievable.

Wait, maybe this will help:

Absence of evidence is evidence of absence. It’s not, however, proof of absence. - appleciders

73

Absence of correlation implies an absence of causation. Example:

Claim: The death penalty deters murder. However, in the absence of any correlated drop in the murder rate in death penalty states, there is no causation between the death penalty and deterrance. Therefore the claim is false by virtue of the absence of correlation. No correlation, no causation.

74

Why are we debating this again? Oh yeah, the god thing. With the god thing, atheists complain about the absence of evidence because the theists keep saying that we’re up to our eyeballs in evidence (anecdotes, mostly) and present that as proof of God; atheists then disprove the evidence or the argumentive value thereof, and state that the theist argument has had its knees kicked from under it, therefore we’re back to the default case of atheism. And then somebody else walks in an slaps the atheists for arguing from a lack of evidence.

Atheists seem to have a hard time articulating a response to this, possibly because the rebuttal is a completely different line of argument from the ‘argument from anecdotes’ they often ding themselves refuting, despite seeming quite similar at first glance. So, the atheists use the responses for the prior argument and are unable to get solid traction (though that often doesn’t slow them down much).

Against the “absence of evidence isn’t evidence of absence” that gets whipped out by theistic agnostics, the only real response is “what specifically are you proposing?” If the person is really doing some kind of intellectualist spin cycle with no goal, then one can shrug and admit, “Yeah, there could be an undetectable extraphysical spirit spying on us when we shower - and we could be in the Matrix, or solipsim could be correct, or we could all be characters in a book. What do you expect us to do about it - believe in everything that’s not disprovable, from Sauron to Star Wars? Act on these assumptions? 'Cause if I’m a solipsist there’s no problem if I go around killing people, right?”

If the person has some kind of agenda, which would be instantly obvious if they do not give solipsism equal time and respect in their arguments, then they can be reasonably dismissed as shills, in my opinion.

75

Dismissed as shills??!?

Are you suggesting that people are being paid to come to this forum and post things they’d don’t necessarily believe?

76

Yes.

That, or I grabbed at a word that vaguely resonated with me as conveying that the person was asserting to have one agenda (pure agnosticism) while actually trying underhandedly to pave the way for another agenda (theism/theistic agnosticism). I could have been doing this instead, I’m not sure.

77

But staring at a box and not seeing anything come out is more evidence for what is and isn’t in the box. You say so yourself in the * part of your post. That is gathering evidence. Every time you look for something and don’t find it you have gathered evidence.

If you don’t look at the box (or smell, touch, listen, weigh, etc.) then clearly you’ve gathered no evidence which is not evidence of absence. There could be an elephant in the box for all you know if you never look. By looking you’ve eliminated all kinds of things and gathered evidence.

As someone else pointed out, this really comes down to how you define “absence of evidence”. If you define that to be looking and not finding anything, then absence of evidence is clearly evidence of absence. If, on the other hand, you define “absence of evidence” to mean you’ve never looked, then it does follow that absence of evidence is not evidence of absence.

78

looking at it is evidence alright. It’s evidence of what’s around the box. It’s evidence that the box is a certain size. Those pieces of evidence tell us something. (although we can’t be sure the box isn’t extradimensional, we’re just assuming that based on previous experience.)

It’s also evidence that nothing came out of the box. However, are you claiming that nothing coming out of the box is evidence for what is or isn’t inside the box?

So tell me. what more do we know about whats in the box based on what didn’t come out???*

[quote=“cmosdes, post:77, topic:529720”]

As someone else pointed out, this really comes down to how you define “absence of evidence”. If you define that to be looking and not finding anything, then absence of evidence is clearly evidence of absence.QUOTE]

But looking and not finding anything is clearly gathering evidence. It’s absolutely finding evidence; evidence of absense of what you were looking for… Here and now. So, I can’t see how anyone can say that it’s an absense of evidence. A negative result is still a result. When its evidence it can’t be an absense of evidence. But, looking here and not finding it doesn’t mean it isn’t over there where you aren’t able to look.

We have a box. For some reason we can’t look inside it or weigh it. We have one piece of evidence for what is in the box… it’s size. We have an absense of any other evidence. And since you think that absense of evidence is evidence of absense… There must be nothing in the box.

*oooh, I know. I didn’t see radiation come out of the box, so there can’t be a radiation source inside, right? No, the box might be lead shielded. The absense of radiation coming out doesn’t tell us there is an absense of radiation inside. Or, the absense of evidence for radiation isn’t evidence of the absense of radiation.

79

To quote myself (a pleasurable endeavor):

“Absence of evidence means you ain’t got nuthin’”

Assuming evidence will one day materialize to support your contentions may be comforting, but it carries no weight (I’m specifically referencing medicine and other scientific matters here).

When come back, bring evidence.

80

The question, though, is whether that “nuthin’” can reasonably used to reach any tentative conclusions. I think it can, if it’s a situation where you would expect there to be positive evidence for the thing in question.