A short theory of learning how to draw

4 minute read


It’s generally assumed that people either have a talent for drawing or they don’t, and when they do they learn in an organic way over time. There’s a sort of mystique around the idea of drawing. More magic than science.

In the excellent book ‘Drawing on the Right side of the Brain’, Betty Edwards writes in the introduction:

Many people regard drawing as mysterious and even somewhat magical.

My perspective is that we feel it’s magic because it’s a kind of skill that is more holistic than others. We always do the full thing, all the time, and it always gets the merciless internal comparison with what we had in mind. While there are different components that we can learn to draw (color, shapes, materials, etc) the line itself either goes where we want it to go, or it doesn’t. It’s either there, or not there.

My level of skill has peaks and valleys. I can produce high quality work with certain subjects and especially in digital, but I’m not as good with the human figure and with certain techniques.

That’s why a while ago I started studying again in a more structured way, borrowing from my other studies on how our mind learns. I also raised my constraints: black and white only, ink only to start.

I did that for many reasons. Focus for sure, but also because I already know the technique of refining a line until I get the right shape. Instead, I wanted to learn how to do a line right the first time. Which is a bit of an absurd, I know it doesn’t really work that way, but it’s both a challenge and a pleasure for me.

After a while thanks to these minimalist limits I started noticing more clearly how I was learning too. I noticed that every time I copied something, very slowly tracing the lines on paper trying to get the right curve, then I improved my ability in doing that specific shape again from memory.

I started doing that on purpose, oscillating between straight copying for a while, and then trying to repeat by decomposing and recomposing in different ways the things I was copying.

Recalling from memory was worse, like 80% worse of the original copy I did, but still an improvement to the same shape before the copying exercise I did.

That meant that slowly I was acquiring knowledge thanks to the copying. Straight lines that tried to be legs now acquired a slight curvature – still far from perfect, but closer to the real thing. And even the straight lines techniques for a more stylized leg were better.

I was acquiring patterns too. When I had to do a specific body position, in practice my mind wasn’t doing anything else than trying to recall from memory the other time I did it, and shaping it in the same way.

And then dawned of me the thing my art teacher was repeating us over and over, more than a decade ago:

Do not stereotype.

He was obsessive about that – to the point that we joked about it – but that’s the essence behind this process. He was teaching us how to draw, so the bit of copying from memory – stereotyping – was pointless. We had to copy from reality, from another drawing, from anything. But not from memory.

A technique can be acquired and repeated – which means in practice repeated from memory — but when we learn, we have to copy from reality.

As far as drawing techniques go, this could also mean to copy a specific technique, not just the subject. But.

It’s the copying where learning happens.

It’s not seeing the picture.

It’s not recalling the picture from memory.

The key is the moment when the picture is being copied.

The interesting bit of this learning technique, copying and then reassembling from memory, is a general learning technique. Shouldn’t seem surprising now that this is also the genius behind LEGO: we get a box with a specific thing to build but then we can disassemble and reassemble it in many other ways, bringing in also other pieces we had acquired previously to build even better and complex structures.

In short, this theory of learning can be summarized as a continuous repetition of three steps:

  1. Copy — copy as-is, as precise as you can be.
  2. Combine — hide the source, recall from memory and try to reproduce it, disassembling and reassembling the subject in different ways. Also composing it with other parts you already know how to do to see how they can come together.
  3. Check — go back to the source and check how close you were in recalling from memory, highlighting the mismatches, which are where improvement can be made.

So, if you like drawing, why don’t you pick an artist you like, and do the copy, combine, check steps above once every day?