-Dudley Young, Origins of the Sacred: The Ecstasies of Love And War
... but not every culture gets the practice it deserves.
How much practice is enough?
Sometimes I watch or help my own students struggle with new concepts (e.g., artificial neural networks and error backpropagation in my AI foundations course), and I think, "well, this is just the first 3-4 repetitions of the 17 or so recommended by educational theorists".1
As an anecdote: a friend of mine is very diligent about revising notes. She's often transcribing or updating them from textbooks. IMO this is a good way for a lot of us to get the practice. Mind you, as an undergraduate, I did not do this - and it's not because I don't need practice, but because I am an infamously prolific note-taker. Anyone who's ever heard "the way Bill takes notes" and has seen me at a conference, writing feverishly (and more recently hunting and pecking on my iPAQ or typing on one of my ThinkPads), knows what the phrase means.
Another anecdote: Some of my CS friends and (more rarely) students like to retake courses to gain a better understanding. Some audit the first time. I always wished I had the time to take courses twice. I felt that I didn't quite grasp automata theory when I took it (600.191) from Mike Goodrich at Hopkins in the fall of 1990. By the time I took it (CS375) from Lenny Pitt at Illinois in the fall of 1993, though, I found I could ace it.
Edit, 03:15 CDT Mon 09 Apr 2004 - As a coda of sorts, I'm sitting here thinking about practice and a friend who is being encouraged to finish a degree in CS ASAP. It's led me back to how the IT field2 is a bit schizophrenic. We want hardworking, capable, and focused software engineers and other programmers; yet we do things that "rush the factory-made engineers" out the door.
you4 yao4 ma3 er2 hao3
you4 yao4 ma3 er2 bu1 chi1 chao33
("To want a horse of quality,
but to wish that it will not need to eat grass")
The above is a Chinese stratagema that is roughly equivalent to to have one's cake and eat it too, and the implied futility of the mindset described is perhaps endemic of the job market in IT (especially of late).
Bjarne Stroustrup once remarked that Denmark has the right idea about scientists and engineers - to wit, that it typically takes five years and not four to make one. K-Staters and other students who have been very, er, Danish, and then some, can take some comfort in this.
1 "17?!", you might well say. Well, it's an application of the Power Law of Practice, much vaunted in many CS/AI areas, especially those inolving learning. Whether you believe the number 17, the point of diminishing returns is definitely pretty high out by most established theories.
2 I use field insofar as this amalgam of applied CS, computer engineering, decision sciences, MIS, etc. can be called one field.
3 Thanks to Haipeng Guo for the Pinyin spelling correction.