|
The reason I asked is because strictly speaking, an unpredictable thing cannot be studied empirically. We can sacrifice some precision or certainty in science, but we can't get rid of reproducibility and still call it "science" just by prefixing it with the term "soft." For example, sound economic models are generally observable in the aggregate despite being imprecise, and high energy physics has many competing theories which are demonstrable but incomplete. The game theoretic principles of market analysis are reliable, as are the principles of gravity. Markets are generally efficient, and the model's conclusions have clear utility that matches real world conditions, even though small pockets of inefficiency also exist. It's fuzzy, but not unscientific. Forests are green, but some trees don't have green leaves. In the abstract, we can tolerate some fuzziness or imprecision as a margin of error, but only if it's compartmentalized to some incomplete theories, and only as long as it's grounded and consistent. We cannot tolerate something being true one day and false the next. Green forests cannot inexplicably and inconsistently become orange without threatening our claim that forests are green. I don't really have an opinion on psychology in particular, though it's pretty clear there's a reproducibility crisis. But as a direct response to your thesis: arguing that a "different mindset" is required to scientifically study subjects which are unpredictable is an untenable position. If humans actually are fundamentally unpredictable - whether due to intrinsic non-determinism or a present lack of sufficient data - they cannot be empirically studied. At that point we're no longer compartmentalizing incompleteness or fuzziness in otherwise sound models. Instead, we're compartmentalizing otherwise sound observations in a sea of chaos. Faced with this sort of reality (and I take no position on whether it is the reality), any scientist, in a "hard" or "soft" discipline, would have to examine if they can reasonably acquire enough information related to the thing in question to make any determination in good faith. An unpredictable thing is an unknowable thing; you may as well try to resolve the three body problem. |