Is evidence proof?

This seems to come up a lot, where person A asks person B for evidence for a certain proposition: Can we just stop fucking around with this? It's disingenuous at best to draw a distinction between "evidence" and "proof". You can't have evidence for propositions that are in error, just as you can't prove the truthfulness of something that isn't true. You can appear to do those things (i.e. "prove" that F=MA) but only because some part of your methodology has failed; you have bad data or bad conclusions.If your position is wrong, then you can't have evidence for it. If you can't prove it, you don't have evidence for it. There's no difference between evidence and proof; otherwise I can have "evidence" for literally any statement whatsoever.[This message has been edited by crashfrog, 01-11-2004]

So what if the available facts supported one position over another, while leaving both positions a reasonable possibility (just that one is more likely than the other based on what we have).Now wouldn't it be reasonable to say that we have evidence for one position over another ? Even though we are not in a situation where either could be said to be proved ?

Hi crash,I think Peter is having a game of semantics with us.When you say this: Technically speaking this is incorrect. When your position is wrong you still have 'evidence' for it, it is just that the evidence is incorrect. If I go to court and tell complete pack of lies I have still given evidence, however the evidence is false. It wouldbe the same if someone stand up in court and say that his God Jesus told him to kill his wife and kids, this is evidence that Jesus is God but it is crappy evidence.Even if we look at the design argument of William Paley's, it is evidence for the existence of God, however it is very very poor evidence.I think this is what Peter is getting at with the dollar bill example. It is evidence of God, the fact that it is very poor evidence is neither here nor there, it is still evidence. I may be off target with my guess about what Peter is saying, but this is what I think he means.Brian.

Would it be fair to say that when we say "evidence", what we mean is "evidence for the proposition at hand?" (This was a thought I had on the way to go see "Big Fish".)I mean, I guess all things could be evidence. For instance, if I say "What evidence is there that it is cold outside?" And you say "well, the thermometer is made of injection-molded plastic," that is evidence - that the thermometer is cheap. It isn't evidence that it is cold out. Therefore, in the context of a discussion of the outside temperature, we can say that your statement is not in fact evidence.Also I think there's a difference between conclusions that are valid inferences from the evidence; and conclusions that are in fact accurate. "It's cold out" is a valid inference from a low value on the thermometer. Whether or not the conclusion is accurate depends on the validity of your data.Any more thoughts, people? If there's one thing I hate, it's word-games. It'd be nice if everyone could just agree on what evidence means.

I too think the distinction is invalid, and in everyday use 'proof' is somehow supposed to carry more weight than 'evidence.' The admins will remind us that no scientific theory is ever proven, and rightly so.However, I'm not an evidence purist or anything. I'm willing to accept that there is data supporting some crackpot theories, and I guess strictly speaking it constitutes 'evidence.' The reason we're obliged to consider them crackpot theories is that they only explain a fraction of the available data. The more facts the theory ignores, the less legitimacy it can claim.When a theory is supported by the amount of data that substantiates Darwin's theory (and excludes as little data as the same theory), we consider the mountain of evidence 'proof' only because it requires a superhuman capacity for self-delusion to resist the conclusion.

---

The dark nursery of evolution is very dark indeed.
Brad McFall

The thing is, we know that they're crackpot theories because they:1)are invalid conclusions from the presented evidence; and/or
2)fail to explain relevant, contradictory evidence.Not that any of this contradicts what you were saying; it's just an extra wrinkle, and part of the problem - proponents of these theories often claim that they have no responsibility to address other relevant evidence or follow the rules that lead to valid conclusions from the evidence they've presented, because "you didn't ask for proof."

You cannot have 'evidence' on it's own, you can only
have 'evidence for' something. i.e. you need context.Evidence and proof are different, and the debate of evidence
centres around it's quality wrt the 'something' under discussion.Evidence can support a concept, but doesn't necessarily proove
the concept.A good case is built upon diverse relevant evidence, whose
interpretations match the proposition.

Well, now, what do you mean by "proof"? In the scientific context, it usually means "a sufficient weight of evidence" so in that context, they aren't different.In a mathematical/logical context, the difference could be the difference between induction and deduction.So it sort of depends what context, don't you think?

Peter,I don't think Crash is claiming that evidence has meaning out of its context. Data presented as support for a claim shouldn't depend on the context for its validity, though.I heard howls of derisive laughter when Martin J. Koszegi claimed that evolution was not as well demonstrated as the orbit of the Earth around the Sun. I'm fairly sure his point was that the Earth's revolution around the Sun is demonstrable fact, while evolution is not. As a good Darwinist, I saw a better use for his analogy, so I adapted it to point out that we believe the Earth revolves around the Sun because of the cumulative effect of inductive inference, not because that's what we see.I don't find it surprising that ancient civilizations believed the Sun revolved around the Earth, since we still refer to 'sunrise' and 'sunset' in non-ironic speech. The core hypotheses of geocentrism were well supported by observations of lunar and solar motion. The constellations appeared in the sky with predictable regularity throughout the year. These data certainly do constitute 'evidence' of geocentrism. The problem was the data that didn't. Since the heliocentric model explained planetary motion in addition to all the data that the geocentric model claimed, it superseded geocentrism in a manner that would (and did) make Thomas Kuhn proud. The heliocentric model is no more 'demonstrable' than the geocentric one. It simply explains more of the data, and the data are independent of the theory it's being used to support.I'm interested in the 'evidence' issue because it seems creationists are confused about the notion of inference, as well as overly selective about the validity of empirical evidential inquiry.

---

The dark nursery of evolution is very dark indeed.
Brad McFall

You've just said how they are different :-Proof requires a sufficient weight of evidence.If they are the same, one cannot require the other.Evidence is what proof is built from.Even with a sufficient weight of evidence to favour one explanation
over another, new evidence can refute the best-bet-so-far.

By 'context' I just meant that data with no context is
not evidence. Data is only 'evidence for' something if it
is of relevance to the proposition.Just as an aside, I think at least one ancient culture
knew that the solar system was heliocentric -- like 4000 BC or
something -- I remember seeing carvings of a ten-planet
heliocentric solar system claimed to be several thousand years
old.

I'm sorry Crash but this is simply not true; you can have evidence for a false statement. For example, suppose I claim all odd numbers are prime and present the following evidence:3 is prime, 5 is prime, 7 is prime.

Well, Mr. J, in that case it would depend on your epistemology:Mathematician: 1 is prime, 3 is prime, 5 is prime, 7 is prime, 9 is not prime; therefore the hypothesis is falsified.Biologist: 1 is prime, 3 is prime, 5 is prime, 7 is prime, 9 is not prime, 11 is prime, 13 is prime. 9 is attributable to experimental error, therefore the hypothesis is provisionally trueSystematist: 1 is prime, 3 is prime, 5 is prime, 7 is prime, 9 is not prime, 11 is prime; 9 is an outgroup, therefore the hypothesis is proven statisticallyPhysicist: 1 is prime, 3 is prime, 5 is prime, 7 is prime, 9 is prime for sufficiently large values of 9, 11 is prime...Engineer: 1 is prime, 3 is prime, 5 is prime, 7 is prime, 9 is prime, 11 is prime; therefore the hypothesis is true[This message has been edited by Quetzal, 01-12-2004]

German Engineer: 1 is prime, 3 is prime, coffee break, go on strike for shorter working hours, company closes, unemployed, what was the hypothesis again?

Speaking as a pedantic maths graduate I have to point out that 1 is not prime. But I suppose that only the mathematician example needs to be changed because nobody else seems to have noticed it...