But seriously, by my accounting, every standard of evidence I know includes an element of faith. The only differences between them are (a) what is taken on faith and (b) how credulous that faith is.
Namely, I find faith elements in believing:
first/second/third-hand reports, even by trained, neutral observers
expert consensus
single expert opinion
by contradiction
the evidence standards of American civil and criminal trials (note the plural, “standardS”. The standards are different between them.)
induction, including (perhaps) mathematical induction
“engineering quality” proof
“mathematician quality” proof
My professional practice as an aircraft mishap investigator is to identify and apply the highest feasible standard from the above list, based on what information is available. “Best practice” in this industry dictates that selection and application of a standard of evidence is a matter of prudential judgement, based on the consequences and probabilities of being wrong, the resources available (evidence, time and $), while being scrupulously open about ones methods.
On historical questions about events of 20 centuries ago, the quality of evidence is not very good. About pretty much everything. What we are left with is Bayesian stuff. Anybody who goes to 0% or 100% draws the raised eyebrow from me. :-)
This is, in practice, a form of equivocation between epsilon uncertainty and sufficient uncertainty to take seriously as an argument.
There is technically no such thing as certainty.
Therefore, the uncertainty in [argument I don’t like] is non-negligible.
Step 2 is the tricky one. I suggest reviewing But There’s Still A Chance, Right? Humans are, in general, really bad at feeling the difference between epsilon uncertainty and sufficient uncertainty to be worth taking notice of.
I reviewed your link—thanks, that was interesting.
Maybe we’re in agreement. Let me try a more audacious assertion...
All I was saying was that practical demonstration or persuasion takes place within an unquestioned frame of reference. For purposes of the topic at hand, I would say, for example, that using the available evidence, I could convince 9 of 12 jurors under American rules of evidence and jury instructions applicable to civil trials, that Jesus of Nazareth was a flesh and blood historical figure. I think I could do this every week for a year and win 90% of the time. If we change the rules to “establish it beyond a reasonable doubt to 12 of 12 jurors”, then my success rate goes way down, obviously.
(Of course, I am assuming that I make no intentional misrepresentations and call only expert witnesses with recognized technical qualifications and unimpeachable character. I.e., no “funny stuff”.)
In other words, I am claiming to not be anywhere in the neighborhood of epsilon.
I’m almost entirely unconvinced (somewhat greater than epsilon) that’s a useful measure—humans can be convinced of just about anything. But, an upvote for suggesting a measure. Still thinking of what I’d accept as a measure.
Winky-face noted and appreciated!
But seriously, by my accounting, every standard of evidence I know includes an element of faith. The only differences between them are (a) what is taken on faith and (b) how credulous that faith is.
Namely, I find faith elements in believing:
first/second/third-hand reports, even by trained, neutral observers
expert consensus
single expert opinion
by contradiction
the evidence standards of American civil and criminal trials (note the plural, “standardS”. The standards are different between them.)
induction, including (perhaps) mathematical induction
“engineering quality” proof
“mathematician quality” proof
My professional practice as an aircraft mishap investigator is to identify and apply the highest feasible standard from the above list, based on what information is available. “Best practice” in this industry dictates that selection and application of a standard of evidence is a matter of prudential judgement, based on the consequences and probabilities of being wrong, the resources available (evidence, time and $), while being scrupulously open about ones methods.
On historical questions about events of 20 centuries ago, the quality of evidence is not very good. About pretty much everything. What we are left with is Bayesian stuff. Anybody who goes to 0% or 100% draws the raised eyebrow from me. :-)
This is, in practice, a form of equivocation between epsilon uncertainty and sufficient uncertainty to take seriously as an argument.
There is technically no such thing as certainty.
Therefore, the uncertainty in [argument I don’t like] is non-negligible.
Step 2 is the tricky one. I suggest reviewing But There’s Still A Chance, Right? Humans are, in general, really bad at feeling the difference between epsilon uncertainty and sufficient uncertainty to be worth taking notice of.
I reviewed your link—thanks, that was interesting.
Maybe we’re in agreement. Let me try a more audacious assertion...
All I was saying was that practical demonstration or persuasion takes place within an unquestioned frame of reference. For purposes of the topic at hand, I would say, for example, that using the available evidence, I could convince 9 of 12 jurors under American rules of evidence and jury instructions applicable to civil trials, that Jesus of Nazareth was a flesh and blood historical figure. I think I could do this every week for a year and win 90% of the time. If we change the rules to “establish it beyond a reasonable doubt to 12 of 12 jurors”, then my success rate goes way down, obviously.
(Of course, I am assuming that I make no intentional misrepresentations and call only expert witnesses with recognized technical qualifications and unimpeachable character. I.e., no “funny stuff”.)
In other words, I am claiming to not be anywhere in the neighborhood of epsilon.
I’m almost entirely unconvinced (somewhat greater than epsilon) that’s a useful measure—humans can be convinced of just about anything. But, an upvote for suggesting a measure. Still thinking of what I’d accept as a measure.
Indeed. They pretty much fail at thinking.