On inaccuracy in representation

“Thus the picture or the drawing ought to be not only a study of a figure for the sake of the figure, and the incomparably harmonious form of the human body—but at the same time ‘a gathering of mangels in the snow’! Have I made myself clear? I hope so, for, as I once said to Seurat, a nude by Cabanel, a lady by Jacques, and a peasant woman, not by Bastien-Lepage himself, but by a Parisian painter who has learnt drawing at the academy, will always have her limbs and body expressed in the same way—often quite charmingly, and, as far as proportions and anatomy are concerned, quite correctly. When, however, Israels, Daumier or Lhermitte, for instance, draw a figure, one is much more conscious of the form of the body, although—and that is why I include Daumier in the number—the proportions will tend to be almost arbitrary. The anatomy and structure of the body will not always seem quite correct in the eyes of the academician. But it will have life, particularly if it come from the brush of Delacroix.

“I have not expressed myself quite satisfactorily yet: tell Seurat that I should despair if my figures were correct; tell him that if you take a photograph of a man digging, in my opinion, he is sure to look as if he were not digging; tell him that I think Michelangelo’s figures magnificent, even though the legs are certainly too long and the hips and the pelvis bones a little too broad; tell him that in my opinion Millet and Lhermitte are the true painters of the day, because they do not paint things as they are, dryly analysing them and observing them objectively, but render them as they feel them; tell him it is my most fervent desire to know how one can achieve such deviations from reality, such inaccuracies and such transfigurations, that come about by chance. Well yes, if you like, they are lies; but they are more valuable than the real values.”

—Vincent Van Gogh (in a letter translated by Anthony M. Ludovici via Margarete Mauthner and found online here)