The Multiverse

Definition

The single universe of everyday experience is only a thin slice of physical reality. Quantum interference phenomena — the fact that light or any other particle behaves differently depending on paths it does not appear to take — can only be explained if there are other, causally connected but mostly imperceptible universes, similar in kind to our own, running in parallel. Deutsch calls the whole of physical reality the multiverse, and argues that accepting its existence is not an "interpretation" of quantum theory but simply what quantum theory, taken as an explanation rather than a prediction machine, says.

In the Book

Deutsch builds the case with a variant of the classic double-slit experiment, worked through with an unusual amount of care (chapter 2, "Shadows"). Widen a barrier from two slits to four and a point on the screen that was previously lit up goes dark — something coming through the newly opened slits is deflecting light that used to arrive there. Repeating the experiment one photon at a time changes nothing: even single photons "know" whether four slits are open. Since a lone photon cannot be interacting with itself, and cannot be splitting and recombining (detectors at each slit never fire in pairs), something else — undetectable directly, but behaving exactly like light — must be travelling through the other slits and deflecting it. Deutsch names these undetectable entities "shadow photons." Scaling the argument up (a barrier stops shadow photons too, which means there must be shadow barriers made of shadow atoms, and so on) leads to the conclusion that particles are grouped into whole parallel copies of the universe, each internally consistent, each only weakly coupled to the others through interference. He traces this to Hugh Everett's 1957 discovery that quantum theory, read literally, describes exactly this structure, and dismisses the rival "Copenhagen interpretation" as a way of refusing to ask what happens between observations rather than an explanation.

Why It Matters

The multiverse is Deutsch's proof-of-concept that a phenomenon too small to feel — a spot of light landing here rather than there — can force a wholesale revision of what exists, once you insist on explanation rather than mere prediction. It reframes "unobservable" as a technological, not a metaphysical, limitation: entities are admitted into your ontology because the best available explanation requires them, not because you can point to them. That move — inferring the necessary furniture of reality from what a theory must explain, rather than from what instruments register — is exportable to any domain where indirect effects (a market price, a symptom, a bug) are best explained by a rich structure you cannot directly see.