A search of the internets yielded quite a few cases of hypotheses that were considered incorrect to due complete lack of evidence, yet shown to be correct later on.

For example:

• http://www.sciencedaily.com/releases/2008/08/080806140128.htm — “theorists had believed since 1926 that a measurement of a quantum particle inevitably forced a collapse”; in 2006 scientists proved otherwise.
• It’s common knowledge that light can’t be bent to create an invisibility cloak, right? Not quite. Turns out you can do fun things with light: http://www.timesonline.co.uk/tol/news/science/article4499914.ece
• And suddenly it becomes possible to create blue roses. Who would have thought? http://www.phenomenica.com/2008/11/blue-rose.html

And, of course, before Newton invented gravity, nobody would ever have believed that things fall. It’s called innovation. Sometimes you only get there if you keep trying. Mathematicians tried for more than 300 years to prove Fermat’s Last Theorem, first proposed without any proof in 1637. They failed, and failed, and failed again. Still, in 1995, British mathematician Andrew Wiles succeeded. Was it reasonable to keep trying? Perhaps not. But he did, and it paid off.

Which things? Science is about refining our model of how things happen. It appears reasonable for science to concentrate on things that can be shown to happen.

Kev

Scientifically speaking, you know nothing at all about the probability that telepathy is possible. You don’t even know anything about the likelihood of its existence… going from the statistical meaning of the two terms, of course. Talking about probabilities does not make a lot of sense without some kind of concrete underlying statistical model (of the probability space and/or the probability distribution). If you have one, I’ll be glad to look at it.

Without any sort of underlying model, you can use neither frequentist probability theory (we could be dealing with an ill-defined non-atomic sample space) nor Bayesian probability theory (there is no plausible way to pick a prior probability).

In fact I’m tempted to claim that statistics can’t be meaningfully applied at all to the “probability of discoveries”, but naturally it’s impossible to prove that (or even say anything about how probable it is… hint hint). At any rate, saying that “the lack of any evidence [...] reduces the probability” is, statistically speaking, utter nonsense. Probability doesn’t work that way and it never has. So if you want to talk about probabilities at all here, it will have to be in the layman’s sense, i.e. as some ill-defined concept of (un-)certainty that is based on nothing but intuitions about the nature of probability… and it’s no big secret that humans are not, by default, very good at intuitively judging probabilities, particularly if things like Bayes’ theorem get involved.

That’s just the statistical side, though. The whole “reasonable demonstrations” thing you mentioned is equally interesting. “Reasonable” is not very well defined. In the end it always comes down to the belief that the human mind is capable of seeing things objectively enough to judge what is reasonable and what isn’t. Anyone who has studied stuff like cognitive psychology knows that humans are chock full of cognitive biases.

You are right that it’s not rational to be convinced of the existence of telepathy if you have no evidence of it. Believing in the possibility is not irrational by any rational (see what I did there?) standard, though. Neither is researching telepathy. I’m not saying we’ll be bound to see scientific evidence of telepathy any time soon just because a bunch of people start actively researching it, or that everyone should be encouraged to research telepathy… but I’m saying that if all people only assumed that things which don’t have any formal evidence speaking for them are impossible, many things we take for granted today would never have been discovered.

Indeed we “know” nothing about the world, if knowledge is understood as involving certainty or objectivity. We can only make educated guesses. That makes it all the more important to remember that we don’t have any level of certainty… clinging to an educated guess that’s not strictly correct might end up preventing a whole new area of discoveries.

Empirical science isn’t about certainty. It’s about having a structured way of dealing with uncertainty. Which brings me back to the main point of the article: science is useful but scientists are often overconfident in their conclusions.

To believe telepathy is likely, or even worthy of investigative time and effort, without any reasonable demonstrations of the phenomena (and certainly without any evidence to support it), is a much more religious and dogmatic view than the ‘scientific’ view that the probability of its existence is tiny.

Don’t get caught only thinking in possibilities. Thinking in probabilities (assuming you are happy to change your calculated probabilities as new evidence comes to light) appears to be much more useful. Possibility approach: “we know nothing about the world”. Probability approach: “we’re not sure, but the world looks like this.”

Kev