No Colour Is Itself

In the 1820s Michel Eugène Chevreul was appointed director of dyeing at the Gobelins tapestry works, where he had a problem to solve. Weavers were complaining that certain blacks came out weak, and the obvious suspect was the dye.

The dye was fine. What Chevreul found, and published in 1839, was that the appearance of a yarn depended on the yarn beside it. The blacks looked weak next to blues. The same colour placed against different neighbours was a different colour to the eye.

He called it simultaneous contrast, and it is the fact that organises everything else about colour in painting. A grey reads warm against blue and cold against orange. A patch surrounded by a light field looks darker than the identical patch on a dark field. Two colours that match on the palette will not match on the canvas if what is around them differs.

Josef Albers built an entire teaching method on it at the Bauhaus and later at Yale, and the first line of his book is the warning: in order to use colour effectively it is necessary to recognise that colour deceives continually. His exercises are all the same shape: take two colours that look different and make them look the same, take one colour and make it look like two.

The working consequence is that a painter never matches a colour to the object being depicted. They set it in relation to its neighbours until the relationship is right, which is why paintings that look nothing like their subject under studio conditions are convincing on the wall, and why a colour lifted from a photograph with a dropper tool so often looks wrong when placed.

The second working rule is that value comes before hue.

Value is lightness and darkness. It carries the structure of a picture, and it survives reproduction: a painting whose values are right reads perfectly well in a black-and-white photograph, while one that is only correct in hue collapses into mush. This is why illustrators and film directors work from grey thumbnails before anything is coloured, and why squinting is a genuine technique rather than an affectation. It discards hue and leaves you looking at the thing you are actually deciding.

Then there is the part of colour theory that is really economics.

Ultramarine was made by grinding lapis lazuli, mined in what is now Afghanistan and carried across a continent. At times it cost more per ounce than gold. Commission contracts specify it by price, 'the blue shall be ultramarine of the value of four florins the ounce', because the cheaper azurite looks similar at first and degrades differently, and a patron paying for the most expensive substance in the workshop wanted it on the most important figure in the picture.

So the Virgin's robe is blue for symbolic reasons and also because that is where the money was visible. Vermeer used ultramarine lavishly and died in debt, leaving a widow with eleven children and a bakery bill. The palette is a budget.

A synthetic ultramarine was developed in 1826 after a prize competition, and the price collapsed by a factor of thousands. Chrome yellow, cobalt blue, cadmium red, viridian and zinc white all arrive in the nineteenth century from industrial chemistry. The bright, high-keyed painting of the 1870s is not only a change in taste; it is painters spending colour they could suddenly afford, out of tubes that had only existed since 1841.

Finally, the hardest thing to hold in mind in a gallery: much of what you are looking at has moved.

Smalt, a cobalt-blue glass, fades to grey. Copper greens brown over time. Red lakes are fugitive and vanish, taking the pinks with them. Varnish yellows across a century and shifts every colour beneath it towards warmth. A landscape whose foliage has gone strangely blue is a landscape whose yellow glaze has gone.

Conservators are careful to distinguish the current state of the object from the painter's decision, and the two are different things. Some of the most admired colour effects in the history of painting are chemistry rather than intention, and in a few cases they are damage that time has made beautiful.