Wigner’s Diamond Reveals Quantum Fragmentation of Space-time

A new theoretical analysis has delivered an unsettling verdict on some of the most popular modern interpretations of quantum mechanics: if you insist that quantum measurement outcomes are observer-dependent, you may have to give up the idea that all observers share a single, unified space-time. The work, published in Foundations of Physics by Jacques L. Pienaar of the QBism Group at the University of Massachusetts Boston and the Federal University of Rio de Janeiro, builds on a family of recent “extended Wigner’s friend” no-go theorems and concludes that rejecting a key assumption about the absoluteness of observed events appears incompatible with the “block universe” picture in which all events exist once and for all in a single four-dimensional arena.

The starting point is a thought experiment that has haunted quantum foundations since Eugene Wigner first posed it in the 1960s. Wigner’s friend sits inside a sealed laboratory and measures a quantum system, obtaining a definite outcome. Wigner, standing outside, treats the entire laboratory — friend and system included — as a single quantum system in an entangled superposition. The friend says the outcome is definite; Wigner’s quantum formalism says no outcome yet exists. The recent no-go theorems, developed by researchers including Časlav Brukner, Daniela Frauchiger and Renato Renner, and a team led by Kok-Wei Bong, sharpen this paradox into rigorous impossibility results. They target what philosophers call the Absoluteness of Observed Events, or AOE: the claim that any measurement outcome has a unique, definite value that does not depend on which observer measured it. The theorems show that AOE cannot be reconciled with quantum theory plus other reasonable-sounding assumptions about locality and free choice.

Pienaar’s contribution begins with a careful dissection of what AOE actually asserts. It bundles together two distinct claims: that an outcome is unique, and that it is absolute — the same for everyone. One can reject the package either way. Many-worlds interpretations reject uniqueness: every outcome occurs, each in its own branch, and there is arguably no single classical space-time threading them together, only something like a “block multiverse.” But a second family of interpretations, which Pienaar calls perspectival, retains uniqueness while denying absoluteness. Each outcome really happens, with a single definite value — but only relative to the observer who performs the measurement. Prominent members of this camp include QBism, the quantum-Bayesian approach developed by Christopher Fuchs and colleagues, and Carlo Rovelli’s Relational Quantum Mechanics.

For these perspectival views, the question Pienaar tackles is brutally simple to state: can observers who reject absoluteness still agree that they live inside one shared space-time? The concern was voiced most crisply by Eric Cavalcanti in a 2021 paper on QBism’s response to Wigner’s friend. If we reject AOE, Cavalcanti argued, then the classical notion of an event must be challenged too: outcomes that are definite for the friend but not for Wigner cannot be located in “Wigner’s space-time” at all. They occur, in effect, inside a “Wigner bubble” — a region of reality that exists for one observer but not for another. Fuchs himself has long embraced a related sentiment, describing space-time in QBism as an “abstract diagram” each agent uses to organize their own expectations, and writing in 2011 that the lesson of Wigner’s friend is that the world is more truly a “pluriverse” than a universe.

Yet Cavalcanti’s argument, Pienaar noticed, contains a hidden weakness. It works only if the space-time point where the friend’s outcome occurs genuinely cannot be matched to any point in Wigner’s manifold. In a seemingly devastating counterexample — which Pienaar names “Wigner’s diamond” — that matching looks easy. Inside a sealed, quantum-controlled chamber, the friend opens a box containing a diamond whose nitrogen-vacancy center encodes a qubit. She measures it by shining a green laser, at a wavelength of roughly 637 nanometers, onto the NV-center and watching for fluorescence: light on means outcome “1,” no light means “0.” Wigner watches through a window. A tiny screen blocks his view of the diamond, so no information about the fluorescence escapes, and Wigner’s entangled superposition survives intact. But the chamber is fitted with rigid measuring rods and embedded clocks — a physical coordinate grid — so Wigner can see exactly where the diamond sits and exactly when the laser fires. Both observers agree on the where and the when. The only disagreement is whether an outcome with a definite value actually happened there. For the friend it did; for Wigner the light field is entangled with the friend and there is simply no event to speak of. Does this not defeat the bubble conclusion?

The answer, Pienaar argues, hinges on deep and contested questions in the philosophy of space-time — specifically on Einstein’s famous “hole argument” from general relativity. General relativity’s equations are invariant under smooth transformations of the coordinate system called diffeomorphisms, which means that radically different-looking distributions of matter and metric fields over the space-time manifold can be observationally identical. Physics alone, it seems, cannot tell you which mathematical manifold point “really is” the point where a given material event happens. Einstein’s own solution, appealing to Leibniz’s identity of indiscernibles, was a relational view: space-time points just are the material events and field configurations located there, with no further “container” behind them. On this relational view, Wigner’s diamond is no counterexample at all. The friend’s measurement outcome is not merely located at a space-time point; it helps define that point. Since the outcome does not exist as a material event for Wigner, the point it defines cannot exist for Wigner either — and Cavalcanti’s bubble returns.

Two rival metaphysics, however, open an escape hatch. Space-time substantivalism holds that manifold points exist independently of any matter or fields, so the same bare point can be shared by Wigner and his friend even if the events occurring there differ. More intriguingly, Pienaar discusses a recent “reference frame” view, developed in work on quantum reference frames and the so-called quantum hole argument, according to which space-time points are identified with the material events of a specially chosen subset of matter — the reference frame. Once the rods-and-clocks grid inside the chamber is chosen as the reference frame, Wigner and his friend can operationally agree that the laser strike happens at one and the same juncture of the grid. This yields the almost self-contradictory-sounding sentence that there is a unique space-time point shared by both observers where the outcome occurred for the friend but not for Wigner. On either non-relational view, Cavalcanti’s original argument fails to force the bubble conclusion.

The heart of the paper is a formal proof that this escape route is illusory in general. Pienaar sets up a precise framework: each observer A has a local space-time manifold and a “localization map” that assigns each possible measurement outcome value to the space-time region where it occurs for that observer. The “classical background” condition says all observers’ manifolds can be embedded as submanifolds of one single shared manifold, with every outcome occurring in exactly one region of it. He then defines an “absolute localization map,” a global assignment of outcomes to regions in the shared manifold that is consistent with every observer’s individual localization map, and proves that such a map is a necessary condition for the shared space-time to exist. The key input is a deliberately weak assumption: that it is possible to design a quantum measurement whose different outcome values occur in disjoint space-time regions — as in a photon-polarization measurement where a beamsplitter routes the photon to one of two spatially separated detectors. Rejecting even this assumption, Pienaar notes, would imply that the locality loophole in Bell tests could not be closed in principle, since Alice’s and Bob’s outcomes could never be confined to space-like separated regions — a radically unpalatable position.

With that in place, the main theorem follows. Suppose a set of assumptions in some extended Wigner’s friend scenario forces you to reject AOE. Now redesign the experiment so that every observer’s distinct outcomes land in distinct space-time regions — always possible by the weak assumption. If a shared classical background space-time existed, the corresponding absolute localization map would, in every run, single out a consistent global assignment of one region to each outcome. But the consistency requirement means the tuple of regions uniquely determines a complete set of outcome values — a “cross-section” of all possible results. Assumptions that prohibit AOE prohibit exactly such simultaneously definite value assignments. Contradiction: no absolute localization map can exist, and the classical background condition must fail. In other words, any no-go theorem that topples AOE can be engineered into a version that also shatters the single space-time block. Rejecting absoluteness of observed events commits you, whether you like it or not, to a fragmented space-time in which some points that exist for one observer cannot belong to another’s manifold.

The implications ripple outward. Pienaar shows that perspectival interpretations cannot be “local” in any traditional sense, because notions like space-like separation and light cones presuppose the shared background that rejection of AOE undermines. QBism and RQM have both claimed some form of locality; critics have disputed those claims precisely because locality becomes ambiguous without absoluteness. Pienaar suggests the way forward is to define locality observer by observer — “relative locality,” a phrase that connects his result to an active research program in quantum gravity pioneered by Giovanni Amelino-Camelia, Laurent Freidel, Jerzy Kowalski-Glikman and Lee Smolin. Perhaps, he speculates, the fragmentation demanded by quantum foundations is not a bug but a clue to the quantum structure of space-time itself.

The result also draws a sharp line between two kinds of perspectivism. Special relativity is perspectival but benign: there exists a perspective-free description — Minkowski space-time — from which any observer’s view can be derived. Radical perspectivism, as embodied by QBism and RQM, denies that such a view-from-nowhere is even possible in principle. Philosophers like Dean Rickles had hoped that participatory-realist views could be accommodated within a block universe, much as free will can be lived from within a deterministic block. Pienaar’s theorem closes that door. Whether the resulting picture — observers each inhabiting their own space-time, overlapping only where their events are mutually defined — is a reductio ad absurdum of perspectival quantum theory, or the first sketch of a genuinely new geometry for quantum gravity, is now an open and newly urgent question.