The question is "what comprises evidence?" is an interesting conundrum in the non-legal sense as well (Hempel's paradox).
If we consider the statement:
All swans are black
How could we prove, or at the very least provide
evidence for this? One way would be to go round observing swans, and every black swan we saw would be
additional evidence towards the proposition, whereas if we saw a blue, or a red swan, this would disprove the proposition immediately, despite all the "evidence".
But I can't be bothered to do this. Instead I take the logically equivalent proposition:
All things that are not black are not swans
And set about confirming this by looking around me. There's a blue pen here, so that is (weak) evidence for my proposition - in fact, I can find thousands, if not millions, of non-black objects here, none of them swans, so, given the above, do my observations then constitute strong "evidence" of my initial claim?
It obviously doesn't since each "non-black" observation only collapses the total possibilities by an infinitesimal amount (although if you lived on a hypothetical distant planet where everything around you was, in fact, a black swan, would this mean that the appearance of a blue pen was then strong evidence for the proposition? No idea
)
The paradox comes in by the fact that the blue pen can be used to confirm the proposition that "all swans are black" and "all swans are white" simultaneously.
In summary then, my first approximation is that evidence for a proposition is "sufficient"
information (in the Shannon sense) which excludes alternative possibilities.
{Or alternatively, what evidence is NOT is NOT "sufficient"
information (in the Shannon sense) which excludes alternative possibilities.
}
PE