When I was a kid my family went to church regularly and I was in the choir for a while. One day I'm sitting in a pew, maybe as part of the choir but probably not, because I'm just watching people stream in. As was common, I'd borrowed a pen or pencil from one place or another and was doodling and making notes on my service program. For whatever reason there's a big dog coming down the center aisle, a mutt most likely. I think there was a pet service once a year, maybe that was it? Anyhow, I think to myself, "I'm going to try to draw that dog!" It's intimidating, ordinarily I didn't draw things from life, but I really wanted to be able to draw a dog. The face was complex, so I simplified it just to get the resemblance of a dog. The details would've overwhelmed me. Then the dog gets into a pew or something, I try to draw the body and fail. But I got the face!

This kind of thing. Bottom I first had it, but over time it was stretched and warp into the top. I was pretty happy with it back then.

For the next eight years or so, that was how I drew a dog face. I remember how it felt, figuring out the relationship between the ears and snout and head and eye and drawing it. Like a flash of inspiration, a sublime math problem, seeing and simplifying and balancing like that (although I never returned to that kind of drawing after the fact, just doodling). I think that's one way to draw, not fully grounded in reality.

The other way is to copy. See a line, draw a line, see a relationship between the size of lines or angle of lines or whatever and then draw the lines accordingly. In a way it's a form of invention and simplification, too. In your mind you figure out the spatial relationships in what you see, and that allows you to accurately draw what you see. One big difference, though, is that copying does not facilitate the invention of actual drawings very well. See the relationship between sizes and angles all you want, but good luck drawing faces with only sizes and angles.

No! Copying is (admittedly) critical for your hand and eye, but much less help to the faculty of invention. Learn to break down what you see more meaningfully: that's what we have to do too. Because in animation you need to draw from all kinds of angles, very accurately and consistently, with little reference for the details. Plus, you need to be able to pick up new character designs and draw them, too.

So, here's what I intend to do. Draw a character's head from a bunch of different angles and try to figure out how it works, then use that understanding to draw it from memory and at new angles. In short, try to learn the head. We'll look at the human skull, the underlying logic of the face, but also eyeball measurements and look at lines to help understand the less strictly logical choices made. I reckon drawing anime isn't copying nor free invention but has restraints and freedom just like English does.

It's one thing to talk and another to do, eh?

Artists, actual ones, have all kinds of quick methods for blocking out and constructing the head: spheres, ovals, t-shapes, boxes, masks, etc, and the variety of methods is even greater for the body. More important than the method is that you use and make those shorthands with knowledge of the real thing. Function over form, so to speak. But truth be told, I don't think you can apply those techniques without that knowledge either. Can't run an engine without fuel.