(Curt Doolittle December 19 at 11:24am)
I classify falsifiability under ‘scope consistency’: limits, parsimony, and falsifiability. Technically they are all properties of scope. But to test scope we must test all dimensions of scope.
Also, like internal consistency, I use external correspondence rather than ‘predictability’ since ‘prediction’ generally invites the ludic fallacy (probability). We cannot predict much in the economy, because any observation and measurement we make effects it. the physical sciences progress quickly because they are the most simple, because they are the least variant. social sciences advance more slowly because we adapt where the physical world can’t.
So science requires that we ‘match the data’ recorded in retrospect, not that we predict.
Instead, prediction is a reductio test of simple systems. Ergo, the explanation horizon depends reflects the rate of adaptation. so we must choose more prediction in some cases (physical science) and more explanatory power in other cases (social science) simply because the horizons vary so much between reaction (the physical world) and action (the social world).