Generally, there are many things which are unproven or not tested, a smaller (but still large) number of things which difficult to test or difficult to prove, a smaller number of things which are testable or provable relatively easily, and finally a small number of things which are tested or proven.
One can expect some people to consider only truths which cluster at these ends. Academia I think tends in that direction. At one level this makes sense, one can expect a lot of useful work to be done on things which are relatively easy to prove, while things that are very difficult to prove one can expect a much lower density of utility in the work and discussion on them.
However, one cannot expect interesting and useful truths to cluster at the “provable” or “proven” end of these distributions. Indeed, given the high amount of work done at the provable end, one might expect the most useful provable truths to be pretty well described already, and the supply of provable truths yet to be proven to be more and more abstract and less and less useful. We pick the fruit that is low-hanging, and well we should.
So one would expect the more interesting truths still open to question to be concentrated along the spectrum of harder or very hard to prove.
Indeed. But reframing and carving testables from untestables and provables from unprovables should be an explicit goal.
OK. And so should a theoretical exploration of the space of hypotheses about the not-yet provables with the intent of getting the most truth for the buck when these expensive experiments are finally done.
Generally, there are many things which are unproven or not tested, a smaller (but still large) number of things which difficult to test or difficult to prove, a smaller number of things which are testable or provable relatively easily, and finally a small number of things which are tested or proven.
One can expect some people to consider only truths which cluster at these ends. Academia I think tends in that direction. At one level this makes sense, one can expect a lot of useful work to be done on things which are relatively easy to prove, while things that are very difficult to prove one can expect a much lower density of utility in the work and discussion on them.
However, one cannot expect interesting and useful truths to cluster at the “provable” or “proven” end of these distributions. Indeed, given the high amount of work done at the provable end, one might expect the most useful provable truths to be pretty well described already, and the supply of provable truths yet to be proven to be more and more abstract and less and less useful. We pick the fruit that is low-hanging, and well we should.
So one would expect the more interesting truths still open to question to be concentrated along the spectrum of harder or very hard to prove.
Indeed. But reframing and carving testables from untestables and provables from unprovables should be an explicit goal.
OK. And so should a theoretical exploration of the space of hypotheses about the not-yet provables with the intent of getting the most truth for the buck when these expensive experiments are finally done.