When you stand before a Van Gogh painting in a museum, you assume you are seeing what he painted. It is a reasonable assumption and, in part, false. Much of what you observe is his: the composition, the thick impasto, the direction of each brushstroke, the texture. However, some of the colours are no longer those he applied. This is not because they have been touched, but because chemistry did not cease the day he laid down his brush. The painting has continued to change, alone, for over a century.

The best-documented case is that of red lakes. Van Gogh utilised red pigments derived from organic dyes: cochineal and eosin. They possess a vulnerability: light decomposes them. Over the decades, that red dissipates, and in doing so, it disrupts all the mixtures into which it was incorporated. Where he mixed red with blue to achieve a violet, today only the blue remains. Where he mixed red with white for a pink, today only a soiled white remains. In The Bedroom, the painting of his room in Arles, the walls we see as bluish today were painted by him in a pale violet, and the doors in lilac. We do not deduce this: he wrote it. In a letter to his brother Theo, in October 1888, he described the room colour by colour: 'the walls pale violet, the doors lilac'. The letter and the painting no longer coincide.

The unsettling aspect is that he knew this. In another letter, written the same year, he noted that all the colours popularised by Impressionism were unstable, and that for precisely this reason they must be used raw, without fear: time would soften them regardless. Van Gogh painted with the awareness that his canvases would fade. He accepted the creation of something he knew to be ephemeral. He is not a victim of deterioration: he accounted for it, though he likely did not imagine to what extent.

And it is not only the reds. Some of his yellows, the chrome yellows, darken due to a reaction between the lead and light, shifting towards brown. Not all do so equally, as he utilised different variants: some resist better than others, yet in certain areas the brilliant yellow he sought has become a brownish shadow. Combined with the lost reds, the result is that the colour balance of certain paintings—which in Van Gogh is almost everything—is no longer what left his studio.

Here, science attempts to restore what has been lost. A major research project, convened around the Van Gogh Museum and concluded in 2017, analysed several works using X-rays, hyperspectral imaging, microscopic samples, and even physical recreations of paintings following nineteenth-century recipes. With all this, they produced digital reconstructions of how the original colours may have appeared. They are an extraordinary achievement. However, the term their authors employ is to simulate, not to recover. And that distinction is everything. In many areas, the data were insufficient to provide a single answer, and the team had to make informed decisions, that is, reasoned hypotheses. The reconstruction is not the painting Van Gogh beheld: it is the best available conjecture regarding that painting, executed with rigour and honesty concerning its own limitations.

That honesty is tremendously valuable. It would be simple to project the reconstruction onto a screen and declare 'this is how he painted it', and be satisfied. Yet that would transform an estimation into a false original, exactly what rigorous science refuses to do. What we possess are two distinct and equally true things: the physical painting that hangs today, with its aged colour, and a scientific hypothesis regarding its initial state. Neither is a deception. The former is the work as time has brought it to us; the latter, a rigorous attempt to imagine its starting point.

From this, I derive an idea that extends beyond Van Gogh. We tend to conceive of a painting as a fixed object completed on the day the painter signs it. It is not. A painting is a slow process that continues to occur long afterwards, in the darkness of pigments reacting, in the light that wears them down, in the varnish that yellows. The work we observe is not a frozen moment: it is the sum of that initial gesture by the painter and everything that has transpired for the material since. When we view a Van Gogh from over a century ago, we do not see an instant arrested in 1888. We see 1888 plus everything that followed, inscribed upon the paint itself by a hand that belongs to no one.

On the open conversation

This text proceeds from a verified fact—the colour shift in several of Van Gogh’s works due to pigment degradation—to consider the artwork as a material object possessing its own history beyond the artist. I distinguish the fact, which pigments altered and why, from the scientific reconstruction, which is a reasoned estimation rather than an exact recovery, and I avoid the generalisation that 'we see no Van Gogh as he painted it', which would be excessive. Should anyone wish to contribute from the perspectives of conservation, art chemistry, or the history of painting, this notebook remains open.

Sources

Van Gogh Museum, REVIGO project (REassessing VIncent van GOgh), concluded in 2017: digital simulation of original colours via XRF, hyperspectral imaging, microsamples, and physical recreations; the authors speak of 'simulating', not 'recovering'.

Studies on the degradation of red lakes and chrome yellow in Van Gogh’s works (analyses published in Heritage Science and other scientific journals; European Synchrotron Radiation Facility regarding chrome yellow).

Van Gogh’s correspondence: letter 705 to Theo (Arles, 16 October 1888) concerning the colours of The Bedroom; letter from 1888 regarding the instability of Impressionist colours (vangoghletters.org).

Works cited: The Bedroom (1888), Field with Irises near Arles (1888), Irises and Roses series (1890).