Evolving probabilities

By Douglas, January 31, 2016

In the past, when we knew little about orbits, the idea that the sun would rise every day was said to be very likely based on the pattern observed. There were limited stories (likely myths) of days where it did not appear to rise at all. Without knowing how it worked, we could not prove that it would rise every day but we would like to work under the assumption it would for practical purposes.

If ancient civilizations had WL they could not say ‘It is likely the sun will rise’ as WL does not deal in likelihood. They could however say there is a X% chance of it rising and in their argument, record the number of days it seems to have risen vs those few myths where it was said not to rise. This argument would be based on some other work such as a calculation of how many days in recorded history that there are etc.

The end result is that through a slightly more convoluted route, you can use the sun rising tomorrow as a ‘fact’. Of course it isn’t but for the purposes of any other argument which asks whether ti will rise tomorrow or not, you need to assume it will unless you get further evidence. However, by acknowledging the tiny chance it means that in special circumstances where a normally negligible chance is important, it will be able to be recognized. An obvious example of this would be to say the sun will always rise. This strong claim cannot use the high probability of the sun rising as proof as the wording is too concrete.

In modern day where we understand planetary rotations we can actually predict some of the reasons what would cause the sun rising effect to fail, such as if the planets all stopped moving relative to the sun. We can add these extra details to the argument improving its accuracy. If we witness the sudden stop in rotation of the earth we can go from having a near 99.999..% chance of the sun rising to 0.000…% chance f it rising.