Let's not forget one of the greatest differentiators of what we can think of as "primitive thought" , "theological thought," "philosophical thought", and "scientific thought."

In primitive thought, testing was almost always to destruction. Either you survived to breed then did so successfully -- or not. Costs for experimentation at a broad level were extremely high. Punctuated equilibrium was the only mechanism of accumulating potential energy, as it were, to make the big experimental jumps -- that still often failed.

Jump to post-sentience with theological thought. Thought itself is the big hop that shifted to reducing costs for failure. Religion provided an encoding system that allowed human groups motivation to remember beyond the immediate tactical environment. Failures of religious experimentation were only sometimes to destruction, as one group consumed or eradicated another, but the chances for absorption of partial successes were a huge step forward. Experimentation could happen much more frequently, speeding the rate of non-lethal failures and thus, successes.

Now comes the fork: philosophical versus scientific thinking. Both manage to push experimentation failure costs way, way down. A failure in either one is minimized, in the case of philosophy because the terms are so abstract that it's difficult to even distinguish failure, while science provides a framework for projecting the experiment and reasoning about how it might fail before adjusting accordingly.

They make another trade off, too. Philosophy has almost no grounds for failure but, additionally, has no grounds for success! They have traded off effect on the world for the pleasure of just going through the motions of reasoning. The bits that accidently turn out to reflect processes which has real-world verifiability, they reject from the school as fast as they can manage (and they end up migrating over to the sciences). This is not an entirely wasted approach to reason because it does occasionally generate new forms of thought which are verifiable and falsifiable -- but its a slow, grinding, undirected groping in the dark for them to simply reject them as "failures" because they Succeed!

And here, science: Hypothesis, theory, experiment, restatement, repeat. The most efficient and effective means of functioning within a bounded set of rules the lot of which you don't immediately know. The costs of failure are sometimes lethal, as astronauts can attest, but the successes are surprisingly frequent.

Given the choice, I know which I prefer by far.