Everett and the Measurement Problem

← Back to Philosophy Essays

The measurement problem comes from a basic tension in quantum mechanics. The theory says that systems evolve into superpositions of different outcomes. But when we actually measure something, we only ever see one result. As Tim Maudlin explains, this tension can be made precise: the wavefunction is supposed to be complete, it evolves linearly, and measurements have definite outcomes. These three claims cannot all be true at once. So any interpretation of quantum mechanics has to give up at least one of them.

The Everett interpretation tries to solve this by getting rid of collapse entirely. Instead of saying one outcome happens, it says that all outcomes happen, each in its own branch of the universe. In doing so, it keeps the wavefunction complete and preserves linear evolution, but it gives up the idea that there is a single definite outcome. While this keeps the math clean, I argue that it does not fully solve the measurement problem. In particular, it does not explain why our experience is definite, and it does not give a clear, non-circular account of probability. Rather than solving the problem, Everett changes what the problem looks like.

To see this, we need to be clear about what would count as a solution. At minimum, an interpretation should explain why measurements lead to definite outcomes, and why those outcomes follow the Born rule. If it cannot do this, then it has not actually solved the measurement problem, since the problem itself arises from the gap between superposition and definite experience.

Everett's idea is simple. The Schrödinger equation is always true, and there is no collapse. When a measurement happens, the system, device, and observer become entangled. The result is a superposition where each term corresponds to a different outcome. Each outcome exists in a different branch, and each version of the observer sees one definite result.

This has real advantages. It avoids adding extra rules to the theory, and it treats the wavefunction as a complete description of reality. Decoherence also helps explain why branches do not interfere with each other. Because of interactions with the environment, certain states become stable and behave like classical worlds. This gives a picture of many "worlds" that evolve independently.

But this does not yet explain definiteness. Everett tells us that each version of the observer sees a definite outcome, but it does not explain why experience has the form it does - why there is a single perspective rather than many. After a measurement, there are many equally real versions of "me," each seeing a different result. The theory describes all of them, but my experience is just one of them. The question is not just why observers see definite outcomes, but why experience picks out one perspective at all. If an interpretation cannot explain why experience is definite, then it has not solved the measurement problem. Everett avoids the original inconsistency by denying that there is a single outcome, but it does not explain the feature of experience that made the problem important in the first place.

This problem connects to the preferred basis problem. For the idea of branching to make sense, we need a way to divide the wavefunction into distinct worlds. Decoherence is supposed to do this by picking out stable states. But this selection is not exact. It depends on how we draw the line between system and environment, and it does not give a single, precise set of worlds. Wallace argues that this is acceptable because worlds are emergent rather than fundamental. But this response weakens the explanatory power of the theory. If the worlds are only vaguely defined, then it is unclear how they can explain something as precise as a measurement outcome. The account starts to look less like a physical explanation and more like a rough description.

The deeper problem, though, is probability. In standard quantum mechanics, the Born rule tells us the probability of each outcome. But in Everett, all outcomes occur. As Hilary Greaves points out, this creates a basic problem: it is not just hard to derive the right probabilities, but hard to make sense of probability at all. If every possible outcome happens, then probability cannot mean which outcome will occur, since they all do. This is sometimes called the incoherence problem.

Everettian approaches try to respond by appealing to decision theory. Deutsch and Wallace argue that a rational agent should act as if each branch is weighted by its amplitude squared, which gives the Born rule. At first, this looks like a solution. But it depends on how we think about rationality. The argument assumes that agents should care about future branches in proportion to their amplitudes. But this is exactly what the Born rule says. If the rule must already be assumed in order to justify it, then it has not really been explained. Calling it a principle of rational decision-making does not solve the problem - it just shifts it.

The problem becomes even more serious when we think about evidence. Scientific theories are confirmed by comparing predicted probabilities to observed frequencies. But in Everett, all outcomes with nonzero amplitude occur. As Greaves and Myrvold argue, this makes it unclear how experimental results can confirm the theory at all, since every possible result happens somewhere. Emily Adlam pushes this further by arguing that observing a measurement outcome in an Everettian universe only tells us which branch we are in, not which theory is true. If that is right, then no possible observation can count as evidence for Everett over its rivals. This is a serious issue, because it suggests that Everett cannot even explain how we come to know that quantum mechanics is correct.

An Everettian might reply that their view is still better than the alternatives. It keeps the theory simple and avoids collapse. That is a real advantage. But the question is not which interpretation is cleaner - it is whether it explains what needs to be explained. Everett preserves the mathematical structure of quantum mechanics, but it does so by abandoning the requirement that measurements have single outcomes and by leaving probability and confirmation unclear.

In conclusion, the Everett interpretation is a powerful and appealing idea, but it does not provide a complete solution to the measurement problem. It keeps the formalism intact, but introduces new problems about experience, probability, and evidence. For this reason, it does not resolve the measurement problem, but instead shifts it into a different form.

References

Maudlin, Tim. Three Measurement Problems. Topoi 14 (1995): 7–15.

Greaves, Hilary. "Probability in the Everett Interpretation." Philosophy Compass 2, no. 1 (2007): 109–128.

Greaves, Hilary, and Wayne Myrvold. "Everett and Evidence." (2008).

Adlam, Emily. "The Problem of Confirmation in the Everett Interpretation." Studies in History and Philosophy of Modern Physics 47 (2014): 21–32.

Wallace, David. The Emergent Multiverse: Quantum Theory According to the Everett Interpretation. Oxford University Press, 2012.