In many of my paintings, I’ll often start with just a couple of colours and perhaps one brush.

Yet if you look at the materials list I ended up using, you’ll see a much larger range.

And that process of how I got there can be slightly misleading.

It can look as though I sat down before starting the painting, carefully selected the perfect palette, laid all the colours out and then simply followed the plan.

That usually isn’t what happened.

The materials list is a record of the colours I discovered I needed during the painting journey.

If I repainted the same subject again from scratch, I may take a slightly different route, based on what I learnt. I might find I don’t need one of those colours at all. I might discover that a mixture I struggled with could have been achieved much more simply.

I often finish a painting on the first attempt and think:

Why on earth did I start with that colour?

But that is part of the process.

Building without predicting

It reminds me of an article Derek Sivers wrote about when designing an off-grid house in New Zealand.

His idea was essentially building without predicting. (You can read the article here: Building Without Predicting)

Rather than designing a house around everything he imagined he and his son might possibly need, they started by living very simply in a bare cabin.

Then they added things when they discovered they actually needed them.

I love that idea.

We took a similar approach when renovating our house and studio. We lived in it for a year before making some of the bigger decisions.

That meant we could follow the seasons. We could see where the morning light came through in Winter, where we naturally sat in Summer, which rooms we actually used and where we tended to end up during the day.

Instead of trying to predict how we would live in the house, we let living in the house tell us what it needed.

A comment at the end of Sivers’ article particularly struck me. He described a neighbour visiting and being horrified by some of the choices they had made.

A neighbor came by for the first time and really hated it, getting all upset and yelling, saying I was crazy for not having his preferred comforts. I love that this home feels wrong for others, and just right for us – Derek Sivers

His conclusion was essentially that he loved the fact the house felt wrong for someone else, but exactly right for them.

And I think there’s something really useful in that when developing a colour palette.

Your palette doesn’t have to look right to somebody else.

Developing a personal palette

There’s a temptation when painting to think there must be a correct palette.

You see another artist using Burnt Umber, Ultramarine Blue, Cadmium Red and Titanium White, so you feel as though you should have those colours too.

Or you look at the reference and think:

There’s a bright red there, so I need a bright red.

There’s a pale colour there, so I need white.

There’s green in the landscape, so I need a green.

But you might not.

A palette should develop from the problems the painting actually presents.

Not from the problems you imagine it might present. And that leads to a really useful painting exercise.

Start with one colour

Find a subject you’d like to paint.

Then choose one colour.

Not five colours.

Not your normal palette.

One.

Start painting and see how long you can survive with it. Really push it, thin it, use it thickly, let the colour of the ground show through, change the pressure of the brush, scrub it in.

Use the paint opaquely in one area and almost like a stain somewhere else.

Your aim isn’t to produce a perfect finished painting with one colour; your aim is to discover the exact moment when the painting genuinely asks you for something else.

A palette developed by questions

Before adding another colour, ask another question.

Suppose you reach an area that needs to be lighter.

Your immediate reaction might be: I need white.

But do you?

Could you thin the paint and allow the lighter ground underneath to create the light?

Could you leave that area unpainted?

Could you make the surrounding shapes darker instead?

A colour doesn’t exist independently; it alters enormously depending on what surrounds it.

So sometimes the answer isn’t another tube of paint.

The answer is changing the relationship between the colours you already have.

Limitations make you notice more

This is where a limited palette becomes much more than a colour-mixing exercise.

If you have twenty colours sitting next to you, every problem can appear to have the same solution:

Reach for another tube.

Take those options away, and you have to become more inventive. You start paying much closer attention to value and notice how dark one shape really is compared with another. You become more sensitive to warm and cool relationships, and to how much paint thickness can affect colour.

You also become more aware of the surface underneath. Your brush handling improves because you’re trying to make the same paint behave in several different ways, rather than immediately reaching for another tube.

Perhaps most importantly, you start looking at the painting as a whole. Rather than asking, What colour is that object? You start asking, what does this area need in relation to everything around it?

I think that’s a much more useful question.

The finished palette is evidence, not instructions.

So when you look at one of my course materials lists and see six, eight or ten colours, try not to think of them as the recipe for the painting. Think of them as a record of the decisions I made while working through several versions of the piece.

I will have made a few small colour studies and swapped brands to get the right hue. I might have started with a couple of colours, then encountered a problem that made me introduce another. A mixture could have almost worked, but needed something slightly cooler, so another colour found its way onto the palette.

By the end, there’s a finished palette sitting there.

But the palette came out of the painting.

The painting didn’t come out of the palette.