Your description of the Manet study, The dead toreador,2 was well analyzed. And the whole letter proves to me the same as your croquis of Paris made me think at the time, that if you put your mind to it you can paint something in words.3

It’s certain that by studying the laws of colours one can move from an instinctive belief in the great masters to being able to account for why one likes what one likes, and that’s very necessary nowadays when one considers how terribly arbitrarily and superficially people judge.

You must just let me maintain my pessimism about today’s trade, because it definitely does not imply despondency. This is how I reason to myself. Suppose that I’m right when I increasingly see something like tulip mania in the curious haggling about the price of paintings.4 Suppose, I say, that like tulip mania at the end of the previous century,5 the art trade, with other branches of speculation, were to disappear at the end of this as it came, that’s to say relatively quickly.

Tulip mania may have perished, BULB-GROWING REMAINS. And for my part I’m content, for better or worse, to be a little gardener who loves his nursery.

Presently my palette is thawing, and the bleakness of the earliest beginnings has gone.6

I still often run up against a blank wall when undertaking something, but all the same, the colours follow one another as if of their own accord, and taking a colour as the starting-point I see clearly in my mind’s eye what derives from it, and how one can get life into it.  1v:2

Jules Dupré is like Delacroix in landscape, for what enormous diversity of mood he expressed in symphonies of colour.7

Now a seascape, with the most delicate blue-greens and broken blue, and all sorts of pearly tones.

Then an autumn landscape with foliage from deep wine red to vivid green, from bright orange to dark havana, with yet more colours in the sky in greys, lilacs, blue, white, forming another contrast to the yellow leaves.

Then again a sunset in black, in violet, in fiery red.

Then again more capricious, like the corner of a garden by him that I saw and have never forgotten; black in the shadow, white in the sun, bright green, a fiery red, and then again a dark blue, a bituminous greenish brown and a light brownish yellow. Truly colours that can have quite a lot to say to one another.

I’ve always idolized Jules Dupré, and he’ll become even more recognized than he is now. For he’s a real colourist — always interesting, and with something so powerful and dramatic. Yes, he is indeed a brother to Delacroix.

As I said, I think your letter about black very good, and what you say about not doing it in the local colour is also correct.

Still, it doesn’t satisfy me. To my mind there’s far more behind not doing it in the local colour. True painters are the ones who don’t do it in the local colour — that was what Blanc and Delacroix discussed once.8

May I not simply understand by it that a painter does well if he starts from the colours on his palette instead of starting from the colours in nature?  1v:3

I mean, when one wants to paint a head, say, and one looks closely at the nature one has before one, then one might think, this head is a harmony of reddish brown, violet, yellow, all broken — I’ll put a violet and a yellow and a reddish brown on my palette, and break them into each other.

I retain from nature a certain sequence and a certain correctness of placement of the tones, I study nature so as not to do anything silly, to remain reasonable — but — I don’t really care whether my colours are precisely the same, so long as they look good on my canvas, just as they look good in life. Truer by far is a portrait by Courbet, manly, free — painted in all sorts of beautiful, deep tones of reddish brown, of goldish, of colder violet in the shadow, with black as a foil, with a little piece of tinted white linen as a rest for the eye9 — finer than a portrait by anyone you will — who has imitated the colour of the face with hideous precision.10

A man’s head or a woman’s head, looked at very composedly, is divinely beautiful, isn’t it? Well then — with painfully literal imitation one loses that general effect of looking beautiful against one another that tones have in nature; one preserves it by re-creating it in a colour spectrum PARALLEL to, but not necessarily exactly, or far from the same as the subject.

Always and intelligently making use of the beautiful tones that the paints form of their own accord when one breaks them on the palette, again — starting from one’s palette — from one’s knowledge of the beautiful effect of colours, isn’t the same as copying nature mechanically and slavishly.  1r:4

Now here’s another example. Suppose I have to paint an autumn landscape, trees with yellow leaves.11 Very well — if I conceive it as — a symphony in yellow, what does it matter whether or not my basic yellow colour is the same as that of the leaves — it makes little difference. Much, everything comes down to my sense of the infinite variety of tones in the same family.

If you think this a dangerous tendency towards romanticism, a betrayal of ‘realism’ — painting from the imagination12 — having a greater love for the colourist’s palette than for nature, well then, so be it.

Delacroix, Millet, Corot, Dupré, Daubigny, Breton, 30 more names, do they not form the heart of this century where art is concerned, and all of them, do they not have their roots in romanticism, even if they surpassed romanticism? Romance and romanticism are our era, and one must have imagination, sentiment in painting. HAPPILY, realism and naturalism are not free of them. Zola creates, but doesn’t hold a mirror up to things, creates them amazingly, but creates, poetizes. That’s why it’s so good. So much for naturalism and realism, which are NONETHELESS related to romanticism. And I still say that I’m touched when I see a painting from the days of 30-48,13 a Paul Huet, an old Israëls like the Zandvoort fisherman,14 a Cabat, an Isabey. But I find that saying, don’t paint the local tone, so very true — that I would far rather see a painting with lower values than nature than one that’s exactly the same as nature.  2r:5

Rather a watercolour that’s somewhat vague and unfinished, on the other hand, than one that has been worked up to capture reality.

That saying has a broad meaning — don’t paint the local tone — and leaves the painter free to seek colours that form a whole and are related to one another, which comes out all the more through contrast to another series.

What do I care that the portrait of a worthy citizen tells me precisely what the milk and water, pinkish, purplish, nondescript colour of the pious man’s face — which I’d never seen — is? But the fellow citizens of the little town where the individual in question made himself so estimable that he felt it incumbent upon him to keep posterity familiar with his physiognomy — are very edified by the speaking likeness.

COLOUR EXPRESSES SOMETHING IN ITSELF. One can’t do without it; one must make use of it. What looks beautiful, really beautiful — is also right. When Veronese had painted the portraits of his beau monde in the The marriage feast at Cana,15 when he had devoted to it all the richness of his palette in sombre violets, in magnificent golden tones —  2v:6 then — there was still a faint azure and a pearly white he thought of — which doesn’t appear in the foreground. He flung it on at the back — and it was right, changed of its own accord into the surroundings of marble palaces and sky that singularly complete the array of figures.

That background is so magnificent in that it came about of its own accord, spontaneously, from a colour calculation.

Am I wrong about this?

Isn’t it painted differently from how someone would do it who had thought about the palace and about the figures at the same time? As a single whole?

All that architecture and sky is conventional and subordinate to the figures, calculated to make the figures show up well.

That is truly painting — and the result is more beautiful than precise imitation of the things themselves. Thinking about one thing and letting the surroundings belong to it, derive from it.

Making studies from nature, wrestling with reality — I don’t want to argue it away. I’ve tackled it that way myself for years and years, almost fruitlessly  2v:7 and with all sorts of sad results. I wouldn’t want to have missed that — error.

Always carrying on in the same way would be folly and stupid, that’s what I mean — but not that all my effort has been utterly in vain.

One begins by killing, one ends by healing is a doctors’ saying.

One begins by fruitlessly working oneself to death to follow nature, and everything is contrary.

One ends by quietly creating from one’s palette, and nature is in accord with it, follows from it. But these two opposites don’t exist without each other. Swotting, even if it’s apparently in vain, gives a familiarity with nature, a sounder knowledge of things. And a fine saying of Doré’s, who’s sometimes so clever! is I remember.16

Although I believe that the finest paintings are made relatively freely from the imagination, I can’t break with the idea that one can’t study nature, swot even, too much.  2r:8

The greatest, most powerful imaginations have also made things directly from reality that leave one dumbfounded.

In reply to your description of the study by Manet, I’m sending you a still life of an open, hence an off-white Bible, bound in leather, against a black background with a yellow-brown foreground, with an additional note of lemon yellow.17

I painted this in one go, in a single day. This to show you that when I say that perhaps I haven’t swotted entirely for nothing I mean it, because these days it really comes quite readily to me to paint a given object, whatever the shape or colour may be, without hesitation.

Lately I made several studies outdoors, of the autumn landscape. I’ll send you the still life and one of these autumn studies soon.18 I’ll write again in the next few days anyway, and send this letter in haste to say that I was very pleased with what you say about black. Regards.