The Split Primary Palette Is The Thief of Joy

Today I would like to suggest that the split primary palette–the palette most commonly recommended to beginners–is, actually, far from ideal for people testing out watercolour.

The split primary palette may be frequently recommended, but it is not universally beloved. Its many issues have been pointed out by watercolour blogging greats such as Handprint (who railed against it spicily) and Logan (who politely pointed out several six-paint alternatives).

So, now, readers might well be wondering if this post is necessary at all. To which I say, of course! The fight against Big Split must go on! But also, while I find this palette meh in general, I consider it particularly unappealing as one’s first and only set of paints. And I would like to explain why using Handprint’s “four dimensions of palette value“, i.e. four things to consider when evaluating any palette. These are, broadly:

• Variation in value, the ability to get both darks and lights. In watercolour, where water can be used to make lights, this boils down to “How easy is it to get varied darks?”
• Chroma. Look at that, Logan just defined chroma! And value, for that mattter. Anyway, what Handprint means here is the ability to get vibrant colour all the way around the colour wheel.
• Mixing convenience: surely no explanation is needed for this one, but I will just point out that a palette that requires me to mix three paints to obtain my most-used hues is NOT convenient.
• “Pigment Attributes”: the interesting thing about watercolour (vs, say, gouache, acrylics, or oils) is just how varied the pigments are. We get different levels of opacity, and staining, and granulation/separation. To me, this is what makes watercolour such a magical medium.

Incidentally, I feel a bit ambivalent about having to cite some dude to explain how to judge palettes, since these four concerns seem pretty universal, and are certainly something I consider.. But all serious works require citations! Especially ones that support their thesis.

Anyway, let’s apply these four dimensions to the split primary palette.

Value

Many beginners seem to avoid dark values when they start out, producing soft, genteel paintings without any too-thrilling shadows. Of course, part of that might be the use of student paints, which are often weak and hard to work up to a dark value. So… does the artist-grade split primary help avoid this problem by making darker values easily available?

The answer is… sort of? The easiest way to get darks is by combining contrasting paints that are capable of dark values themselves. The palette has several promising darkish colours on the cool side of the spectrum, but the warms are disappointing: yellows do not reach dark values in themselves. So, our only real option is to mix something with the yellow-leaning red, probably the yellow-leaning blue.

Of course, mixing other dark shades, like a dark green, will require a three-colour mix.

Chroma/Colour Range

The whole marketing spiel about the split primary is that doubling up the “primaries” makes it easier to get vibrant oranges, greens and purples.

When I first read that, I immediately thought “wait, if we want vibrant oranges, greens, and purples, why not just buy them in the first place?” And, indeed, it can be proven (e.g. by Handprint) that a secondary palette (one made up of primaries and secondaries) yields a wider range of bright colours.

Mixing Convenience

Hmm. Perhaps we should skip this topic, since it is completely irrelevant to beginners. After all, all novice painters are supposed to spend the first year or two Learning How To Mix. Giving them access to bright secondaries, darks, or useful landscape shades will delay this learning process. It is important to eat the vegetables of Practice before the dessert of Art.

(Sideline: I like vegetables. I also find learning to mix–or “exploring mixing”, as I prefer to think about it–to be fun. But I enjoy it more if I do it by choice, not by necessity, by allowing myself a large set of paints from which I choose various limited palettes. Triads, for example, are an interesting exercise. Logan has posted about several.)

Coming back to the split primary palette: I have already mentioned that it provides only limited options for mixing darks. The secondary palette is obviously better at this, since orange tends to mix interestingly with blues, and the green and the magenta will yield something cool, too. But even the secondary palette requires one to mix many of the colours a painter might need, like browns, or natural-looking greens, or skin tones. And yes, it is good to practice mixing these hues until the process is intuitive and effortless, but all that mixing practice can feel like a chore, one tedious enough to kill inspiration. Meanwhile, a wide array of ready, luscious colours is fun and inspiring.

Pigment Attributes

I had intended this to be my main point! One big reason to choose watercolour over other media is the variety of special effects made possible by the variety of pigment attributes that are showcased by watercolour. Some paints, like the cobalts or many earths, granulate when used alone, and separate in mixes; some paints, like cadmiums and some earths, are weighty opaque in masstone but dilute to glowing transparent tints; some paints, like the phthalos and many other manufactured organics, stay put after they dry and allow for endless layering; others can be lifted for highlights or texture.

So, what interesting attributes does the usual split primary palette bring to the table? Well… most versions do consist of smooth, staining, (semi-)transparent paints, with Ultramarine as a possible “gateway drug” to granulation.

In a way, this makes sense: there are many painters who dislike paints that are granulating, or opaque, or too-liftable, but I have yet to hear of anyone who hates smoothness or transparency. So the split primaries are a safe selection. At the same time, they are a pretty sad introduction to the magic of the watercolour medium; they do not give beginners the chance to discover properties they might love.

Conclusions

In conclusion, the split primary palette is clearly objectively awful. So, what should beginners start with?

My default answer is “whatever artist-grade paints they can get cheaply. So, perhaps a starter palette from whatever their local brand is?”

So, given that most starter palettes are split-primary-based… Yeah, enjoy your split primary palettes, beginners! At least you will be able to follow the many beginner tutorials that refer to it!

Although…

…it is not a given that a split primary palette is a beginner’s best cheap option. Many companies offer more interesting selections, ones picked out by specific artists, or meant for a specific use-case, or intended to showcase something (e.g., granulation). And second-hand paint is always an option: I got my first artist-grade paints by buying other people’s used palettes on e-bay-type sites. This let me try out many weirdo pigments early on.

Another option is picking out individual colours can be just as economical as buying a set (since, even if a set is cheaper, some of the paints in it might be completely useless). I know picking out a whole palette can be overwhelming–but many paint bloggers are happy to offer their own opinions! I have already mentioned Logan’s post on alternatives to a split primary. (I agree with what he says there 99%.) One can also pick up a palette as described above, and then just grab a few individual colours. Such as the ones in my experimental palette above.

In any case, I think the most important thing is to end up with a set of colours that sparks joy. That joy will lead to painting, and painting will lead to painting skill, and also to developing one’s personal opinions on palettes. (Which are of course, much more valuable than mine.)