[newton10471]

Reading to learn is not borne out by research as a good way to build and retain skills, and can just as likely create illusions of competence http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect. Research shows it's far better to just start doing the activity. The reading should just be used to get you just to the point where you can start doing.

[WorldWideWayne]

Understanding fundamentals before doing some activity is empirically more efficient though. Both thinking and doing are necessary for understanding the fundamentals of a given skill. Reading or communicating about things is a subset of thinking about them. It's all necessary at some point, whether you do it before, during or after your activity.

I read to understand fundamentals and to get useful information such as which libraries, techniques, etc are popular in the arena that I'm entering or things to avoid. Many, many times reading has saved me from entering a given arena because I was able to conclude that whatever new technology I was looking into wasn't really going to work the way that I wanted.