If quantum superpositions are real, why do we never see a chair in two places at once? For decades the standard answer was a shrug: large things are "classical." Since the 1970s physicists have had a much better answer, called decoherence.

Where the idea came from

The German physicist H. Dieter Zeh first described decoherence in 1970. In the 1980s the Polish American physicist Wojciech Zurek developed it in detail and brought it to wide attention.

How it works

No real object is perfectly isolated. A dust grain floating in a room is constantly struck by air molecules and bathed in light. Each collision carries away a little information about where the grain is. Suppose the grain is in a superposition of two places. The air molecules that bounce off it will bounce differently in each case, and soon the environment holds many scattered records of the grain's position.

Here is the key point. To see interference, the signature of a superposition, the two alternatives must be brought back together so that they can overlap. Once information has leaked out, that is no longer enough: the air molecules must be brought back together too. So in decoherence, information about the system leaks into its surroundings, and interference between the alternatives becomes practically impossible to observe. Restoring the interference would mean gathering up every scattered trace of that information from the environment, every air molecule and photon, and that cannot be done.

The process is astonishingly fast. The larger an object is and the more it interacts with its surroundings, the faster it decoheres. For a speck of dust in air, estimates give times far shorter than any instrument could measure. For a single atom in a vacuum chamber, superpositions can last long enough to study. This explains why quantum effects are easy to see in tiny, isolated systems and practically invisible in chairs and cats.

Pointer states

Decoherence does not treat all states equally. Zurek showed that the environment favors certain stable states, such as definite positions for large objects, which survive constant monitoring. He called them pointer states, after the pointer on a measuring dial. The environment keeps "asking" where an object is, so position becomes the property that stays definite.

What decoherence does not solve

It is tempting to conclude that decoherence solves the measurement problem. Most physicists and philosophers who study the question agree that it does not, at least not alone. Decoherence explains why we never see interference between outcomes, but not why only one outcome occurs.

The reason is simple once seen. Decoherence is a consequence of the smooth Schrödinger equation applied to a system and its environment. The smooth equation never collapses anything. So after decoherence, the full state of system plus environment is still a superposition, just one whose parts can no longer interfere. Looked at from the system alone, the result resembles a list of outcomes with probabilities, like a shuffled deck of cards. But in a shuffled deck one card really is on top and we merely do not know which. In the decohered state, nothing in the mathematics picks one outcome as the real one. John Bell put the point sharply: decoherence gives answers good enough for all practical purposes, which is not the same as an answer in principle.

How each interpretation uses it

Decoherence is not a rival interpretation but a tool every interpretation uses. In many worlds, decoherence defines the branches and keeps them from interfering with one another. In pilot wave theory it explains why the "empty" parts of the wave stop affecting the particle. For Copenhagen it explains why the cut can be placed anywhere beyond the point where decoherence has happened. For collapse theories it is a nuisance in experiments, since ordinary decoherence can mimic the effect being tested, and must be carefully shielded out.

It also matters for technology. Decoherence destroys the fragile quantum states that quantum computers depend on, which is why their hardware is cooled to near absolute zero and shielded with great care. QUAN 203 takes up that story.

Worked example: hot molecules

In 2004 a team in Vienna led by Anton Zeilinger and Markus Arndt sent large carbon molecules, each made of seventy atoms, through a grating and recorded the interference pattern on the far side. Then they heated the molecules with lasers before they entered the grating. A hot molecule glows faintly, giving off light in the form of thermal photons.

As the temperature rose, the interference pattern faded. Nothing had pushed the molecules off course. Instead, each emitted photon carried away a little information about where the molecule was, and with enough photons the environment held a record of the molecule's path. The superposition was still there in the full state of molecule plus light, but it could no longer be seen.

The experiment shows decoherence happening step by step, in a way that can be dialed up and down. It also shows exactly what decoherence does and does not do. It explained why the pattern vanished. It did not say which path each molecule "really" took, and that remaining question is where the interpretations of this course take over.