Wayward paths can help us better understand the correct ones.
Very true, IMO. It's one of the reasons I love my algorithm textbook more than any other school textbook I've ever had: it not only explains why you choose a certain way of proving an algorithm, or how to construct an efficient one, it gives you seemingly-logical, incorrect alternatives that people naturally fall into and explains why they're wrong. It actively helps you to recognize potential mistakes later, rather than having you stumble onto them without understanding why you made the mistake.
I recently picked up Betty Edwards' "Drawing on the Right Side of the Brain" book. Supposedly, it's been updated recently, to contain her most recent knowledge on mark-making and 'seeing' the drawing in life.
I used to be quite a decent pencilist, but after 18 years of not having drawn a lick, and perhaps relating to the drugs that were consumed during my teenage years, I've simply forgotten how to do anything, how to even see the marks that need making, to complete even rudimentary drawings.
I'm only a few pages in, but at least on the premise, the referenced book in this works on similar principles.
I think Edwards actually has a completely different approach. The book referenced in the article uses the idea of making an object out of basic shapes: put together some squares and circles and you have an elephant. The only problem is that that approach only works for making a kind of stencil image of an elephant and totally fails for realistic drawing. Edwards' approach is totally focused on convincing the brain to drop its preconceptions (like the squares and circles) and transfer the reality of the actual lines and shadows that are in front of the eye to the paper. You actually have to stop the mind from trying to substitute it's template of an idealized "elephant" (or the idea that I should draw a circle because someone told me to draw a circle) and let the pencil just follow the eye.
I think Edwards's approach is more compatible with Ames's construction approach than your description implies. Sure, Edwards is trying to get you to see what is in front of you rather than rely on preconceptions. But what is in front of you can often be described as a collection of generalized shapes, as Ames does. And, indeed, this is a good way to get the initial proportions of a drawing correct; I believe Edwards discusses some helpful construction rules in the portrait section of her book.
In other words, you should draw a circle not because someone told you to draw a circle, but because the thing in front of you genuinely has an underlying circular form at that size and in that position -- as Ames must have done when he came up with his drawing books in the first place. And the more you do this, the faster you learn the underlying structure of things, which makes it easier to draw from memory, draw from imagination, or simply draw from life without needing to stop as frequently to measure things.
Generalized shapes might help if you're drawing extremely regular objects (books, balls, apartment buildings) or perhaps if you need to draw a generic "horse" from memory but it's not a big help with life drawing. Edwards and other life drawing teachers often suggest drawing the negative spaces as a way of breaking the habit of using the brain to figure out what the object is "like" (e.g. it's like a square sitting on a circle) and focus on what it is.
I think of them both as techniques for the same purpose - to get you to draw what you see as there rather than what you know is there. (Or even drawing the symbol for what you know is there.)
It's been a while since I've done any drawing, but from what I remember you can easily use both in the same image - blocking out the constituent shapes with approximations with around the right proportions, and then making sure that the negative space looks about right in the gaps in between.
Yes, negative space is useful for focusing a student's attention on the shape itself rather than the subject. But it also has the tendency to encourage a local rather than global focus, so that locally all the shapes look right, but the global proportions are wrong. Or where you work your way around a bicycle wheel drawing all the negative shapes, and the last shape you draw is too wide or too thin.
Which isn't much of a problem for things like still lifes, but it becomes a bigger problem in, for example, portraiture, where the human eye is particularly sensitive to errors in proportion -- which is why Edwards includes some construction rules-of-thumb in her portraiture chapter.
You may be focusing too strongly on "circles and squares" as idealized shapes... If you look at the picture of the elephant in the article, none of the shapes is quite a circle. Rather, they outline masses, like the blobs you might see if you defocus your eyes. Being able to capture the proportions and locations of these masses is actually very important when drawing from life. (see, for example, Nicolaides's "The Natural Way to Draw")
