consistent histories

The consistent histories approach, developed by Robert Griffiths, Roland Omnes, and (as decoherent histories) Murray Gell-Mann and James Hartle, reframes quantum theory around whole sequences of events in time rather than single measurements. A history is a chain of properties a system might have at successive moments — here at one time, there at the next — and the theory's job is to assign sensible probabilities to such histories.

The catch is that you cannot do this for every imaginable set of histories. Quantum interference means that, for many sets, the probabilities would not add up properly. A set is called consistent, or decoherent, only when the interference between its alternative histories washes out, so that ordinary probability rules apply and you may reason about the histories as if one of them simply happened. Decoherence with an environment is what naturally picks out such well-behaved sets for everyday objects.

The aim is to talk about events in a closed system, even the whole universe, without ever invoking an external observer or a separate collapse — an attractive feature for quantum cosmology. The honest subtlety is that many different consistent sets exist, and they can describe the same world using incompatible families of questions. Critics worry this undercuts a single objective account; defenders reply that you simply must not mix incompatible sets, and that within any one set the story is perfectly clear.