Necker Cube

Twelve lines, two cubes

The Necker cube is a wireframe: twelve lines of equal weight, with no shading to say which face is nearer. Look at it for a few seconds and one square comes forward as the front of the cube. Keep looking and, usually without you deciding anything, the other square takes over as the front and the whole cube turns inside out.

The drawing doesn't change. The flip happens entirely in how your visual system turns flat lines into a solid object.

Why there's no right answer

A real cube seen at an angle projects a pattern of lines onto the retina. So does this drawing, and so would a second cube oriented the other way, viewed from the opposite side. The drawing fits both objects exactly, and with no shading, overlap, motion or stereo depth to break the tie, the visual system has nothing to choose between them.

It still refuses to stay undecided. The brain commits to one cube, holds it for a few seconds, then switches to the other. This is called bistable perception: two interpretations, each stable on its own, taking turns. A common explanation is that the neurons representing the current interpretation gradually adapt, letting the competing interpretation win for a while, until the same thing happens in reverse (Sterzer, Kleinschmidt & Rees, 2009). It's the Gestalt point in its plainest form: perception imposes structure on what it receives, and "a cube" is a much more useful hypothesis than "some lines".

You have some control. Fixing your gaze on one corner, or imagining a particular face as nearer, can make that reading come up more often and last longer, but switches still happen when you don't want them to.

Which face do people see first?

The cube above starts with no face highlighted, so the first reading you get is your own. In this drawing, seeing the lower-left square as the front means you're looking down on the cube from above, and seeing the upper-right square as the front means you're looking up at it from below.

That matters because the visual system has a known preference for seeing objects from above. Nikolaus Troje and M. McAdam showed it with the spinning dancer: when a silhouette is ambiguous, people favour the interpretation that puts them above it (Troje & McAdam, 2010). If the same preference applies to a static wireframe, more visitors should report the lower-left square than the upper-right one. The poll under the demo keeps a running count.

A crystallographer's discovery

The Swiss crystallographer Louis Albert Necker described the effect in 1832. He wasn't studying perception: he noticed that drawings of rhomboid crystals he was examining would spontaneously reverse in depth, and wrote to a scientific journal about it. That makes it one of the oldest illusions in the scientific literature, nearly a century older than Gestalt psychology. It's still used to study how the brain decides between competing interpretations. Rubin's vase and the duck-rabbit also flip while the image stays the same, though between objects rather than depths.