Leaving the Window

Alberti's picture was a window. You look through the rectangle at a scene arranged beyond it, and every technical achievement of Western painting for four hundred years is in service of making the glass invisible.

Abstraction closes the window. What is left is paint on a flat bounded surface, and the picture is about what that can do.

The move happened around 1910 to 1915, arrived at by several people who did not agree with each other, and the tidy formalist explanation everyone is taught came afterwards, from critics.

Here is that explanation at its strongest, because it deserves to be stated properly. Each art, on this account, becomes serious by investigating what is unique to its own medium and discarding what it has borrowed from others. Painting's defining condition is the flat surface, bounded by edges. Depicting a scene means using painting to do sculpture's or literature's work. So the history of modern painting is the progressive acknowledgement of flatness: from Manet through Cézanne and Cubism to the point where the surface is the subject.

Clement Greenberg made this the criterion of quality in mid-century criticism, and it did real work: it tells you what to look at in a painting that depicts nothing, and it explains why the best abstract painting is not decorative.

Now the actual motivations of the first abstract painters, which the account edits out.

Kandinsky was a theosophist writing about spiritual vibration, colours that sound like instruments, and an epoch of the spirit he believed was arriving. Malevich's black square was mystical, an object he hung in the corner of the room where an Orthodox household hangs an icon. Mondrian was deep in theosophy and building a visual system for a universal order. Hilma af Klint was making large abstract canvases from 1906 as records of spiritualist sittings, and instructed that they not be shown for twenty years after her death.

None of this is a formal investigation of the medium. The pioneers were not painting about paint; several of them thought they were painting the invisible structure of reality.

The objection that abstraction had to answer is the obvious one and it is not stupid: if the picture does not depict anything, why is it not wallpaper?

The answers are about scale, touch, seriousness and the evidence of decision. A Rothko is sized so that at the distance he specified it fills your visual field and stops being an object across the room. The edges of his rectangles are worked and re-worked until they breathe rather than sit. A Pollock records the entire physical event of its making, and reading it is following a sequence of decisions at speed. A Mondrian's proportions were adjusted over months, with tape, until the thing locked.

Plenty of abstract painting fails that test, which is exactly why the decorative objection retains its force. The category does not confer quality. It never did for figurative painting either.

Formalism's own weakness is a different matter. A story in which everything since Manet leads inevitably to a particular kind of American abstraction is an enormous claim to derive from a preference, and it required discarding a great deal of good painting that was heading elsewhere: surrealism, most figurative modernism, anything narrative. When art moved in the sixties to objects, performance, photography and documentation, the criteria had nothing to grip, and the position fell apart within a decade.

The lesson is not that formalism was foolish. It is that criticism which explains why the work you already like is historically necessary has told you about the critic.