Even though we do have a few more formal knowledge-justification methods that are occasionally useful, the default fallback method remains trial and error.
Furthermore, we can safely assume that quite a bit of trial and error went into developing traditional practices:
Especially in the realm of human behavior, we know that nothing is provable (math) and very, very little is testable (science). Therefore, everything we truly know about it, is the result of endless trial and error.G. K. Chesterton wrote: Don't ever take a fence down until you know the reason it was put up.
In fact, modern western society is quite presumptuous and even outright ridiculous to believe that it can add value, on the fly, in knowledge domains that are necessarily governed by trial and error.
Even worse, the often imbecile beliefs of modern western society even liberally carry over to taking down fences in science. It has now even become a society that does not need to test its vaccines, because hey, vaccines do not need to be tested!