Studying process, I see it like this:
- Do it first. You’d do a lot of mistakes, but that’s the point.
- Learn theory then. How you should do what you did.
- Pratice a lot, calibrating what you do against the theory you’ve learned.
- Continue practicing, without calibrating. Just do, do, do.
- Calibrate against theory periodically. Over time, you’d find a better theory than you found at the step 2.
Perhaps that’s too obvious, but I thought I’d note it down.