Spacetime and Causal Structure

Operational causal order and spacetime causal structure are related but distinct. A process can lack a decomposition into definite laboratory orders without demonstrating that a gravitational metric has been placed in a superposition. The research question is which process-level predictions remain distinguishable after the laboratories, clocks, and allowed interventions are embedded in a specified physical model. Literature checked 11 October 2026: the selected evidence below runs from the 2012 process framework through a 2023 extended certification scenario; it is not a complete current experimental survey.

Helpful background. Indefinite causal order explains the operational formalism. Relativistic causality distinguishes signaling constraints.

Causal separability is a precise test

In a bipartite setting, a causally separable process can be written

$$W=qW^{A\prec B}+(1-q)W^{B\prec A}, \qquad 0\le q\le1.$$

This is a convex-mixture statement relative to a specified laboratory decomposition. Failure of such a decomposition is causal nonseparability. Violating a causal inequality imposes a different operational criterion; the two concepts must not be equated.

Established mathematical results: Oreshkov, Costa, and Brukner introduced processes without a presupposed global order and demonstrated correlations outside definite-order bounds. Barrett, Lorenz, and Oreshkov studied a cyclic causal-model description. These results concern defined operational frameworks; their gravitational interpretation requires additional assumptions.

The separation between witness types matters in an actual design. Araújo and colleagues showed that the quantum switch is causally nonseparable although it does not violate causal inequalities in the scenario they analyze. Van der Lugt, Barrett, and Chiribella subsequently showed certification in an extended scenario, adding a spacelike-separated observer and assumptions excluding superluminal and retrocausal influences. This is not a contradiction: changing the operational setting changes the admissible classical models and inequalities. “Device independent” does not mean independent of causal and intervention assumptions.

A causal game gives a numerical control

In the Oreshkov–Costa–Brukner game, Alice receives a random bit $a$ and outputs $x$; Bob receives independent random bits $b,b'$ and outputs $y$. They aim for $x=b$ when $b'=0$, and $y=a$ when $b'=1$:

$$p_{\mathrm{succ}}=\frac12\left[ P(x=b\mid b'=0)+P(y=a\mid b'=1)\right].$$

For a fixed order, the earlier laboratory cannot infer the later laboratory’s independent bit better than chance. The other task can succeed perfectly, giving $p_{\mathrm{succ}}\le3/4$; mixtures preserve the bound. Closed laboratories and independent input choices are essential to this reasoning.

The paper’s process and instruments attain $(2+\sqrt2)/4$. If that fixed experiment is mixed with a no-information process giving success $1/2$, linearity gives

$$p_{\mathrm{succ}}(v)=\frac12+\frac{v}{2\sqrt2}, \qquad p_{\mathrm{succ}}>\frac34\ \Longleftrightarrow\ v>\frac1{\sqrt2}.$$

This is a threshold for this noise family, game, and instrument choice, not a universal robustness bound or an implementation of the process by gravity. Failure to violate this inequality below the threshold does not certify causal separability.

From clock shifts to a gravitational interpretation

For stationary clocks in a weak field, neglecting velocity corrections, a potential difference gives $\delta\tau\simeq T\Delta\Phi/c^2$. A classical uncertain potential can therefore produce uncertain clock readings. Such uncertainty alone does not distinguish an incoherent distribution of orders from coherent control of order.

Zych and colleagues propose a gravitational thought experiment in which a spatially superposed source controls temporal order, together with a Bell-type test under stated locality, preparation, and intervention assumptions. Castro-Ruiz and colleagues analyze event localization relative to quantum clocks in gravitationally interacting models. These are concrete theoretical connections, not observations of fluctuating quantum geometry. Active: construct a complete intervention model with clocks of finite resolution and determine which joint statistics exclude its classical alternatives. Conjectural beyond a specified model: identifying an arbitrary nonseparable process with a superposition of spacetime metrics.

Research questions at the interface

Embedding. Which abstract processes can be realized by controlled quantum circuits, spacetime-local interactions, or a specified gravitational model? Positivity and normalization alone need not establish physical implementability in every proposed setting.

Laboratory decomposition. An operation delocalized over time may admit descriptions with different subsystem boundaries. Which witness remains invariant under physically legitimate redescriptions?

Control resources. Coherent control of operation order is an additional resource. A claimed communication or computation advantage must count it and compare with an appropriate definite-order baseline.

Causal dynamics. A fluctuating metric changes proper times and light cones. Connecting those quantities to an operational process requires a consistent matter-plus-clock model, not just a circuit diagram drawn on two geometries.

An entry project: classify a noisy process

Baseline. First reproduce the game’s $3/4$ bound by enumerating deterministic strategies for each order. Then reproduce the cited quantum process, its normalized probabilities, and the analytic noise threshold. This separates a probability-convention error from a semidefinite-program error.

Resources. A laptop, Choi matrices, convex optimization, and basic quantum channels suffice for qubit input/output spaces. State the Choi transpose convention. Use a known definite-order process and its mixtures as negative controls.

Deliverable. Implement the linear process constraints and causal-separability program from the witness paper. Report primal and dual residuals, witness normalization, and noise thresholds with numerical tolerances. Check normalization on a spanning family of local trace-preserving maps, then test additional randomly generated instruments. Compare the optimized witness with the fixed game’s violation; do not call a finite set of tested instruments a proof of universal validity unless its span establishes the constraints.

Failure criterion. Stop interpretation if positivity or normalization fails, the violation is smaller than the solver residual, or a definite-order control violates the supposed bound. If an implementation restricts the allowed local instruments, rederive the witness under those restrictions. A witness optimized and assessed on the same finite noisy data also needs an independent validation sample or a valid statistical correction.

Research extension. Impose one concrete circuit or clock-resolution constraint and characterize the reachable subset or the resulting certification gap. Comparing two feasible sets is a graduate-scale problem; classifying every physically realizable process is not. Check existing witness and implementation literature before claiming a new threshold, since the basic white-noise calculation is a reproduction exercise.

What would count as a stronger result?

A convincing advance links three layers without skipping one: a mathematically valid process, an operationally realizable intervention scenario, and a justified spacetime model. Evidence at the first two layers should not be advertised as an experimental discovery of quantum geometry.

Students with quantum-information training can begin at the process layer. Those extending to spacetime need additional expertise in clocks, localization, and relativistic dynamics. Choose the boundary between layers as the research question rather than assuming the dictionary exists.

References

• M. Araújo, C. Branciard, F. Costa, A. Feix, C. Giarmatzi, and Č. Brukner, “Witnessing causal nonseparability,” New Journal of Physics 17, 102001 (2015). DOI — causal witnesses, semidefinite programs, and the switch distinction.
• J. Barrett, R. Lorenz, and O. Oreshkov, “Cyclic quantum causal models,” Nature Communications 12, 885 (2021). Primary article.
• E. Castro-Ruiz, F. Giacomini, A. Belenchia, and Č. Brukner, “Quantum clocks and the temporal localisability of events in the presence of gravitating quantum systems,” Nature Communications 11, 2672 (2020). DOI — a relational clock model.
• O. Oreshkov, F. Costa, and Č. Brukner, “Quantum correlations with no causal order,” Nature Communications 3, 1092 (2012). arXiv:1105.4464.
• T. van der Lugt, J. Barrett, and G. Chiribella, “Device-independent certification of indefinite causal order in the quantum switch,” Nature Communications 14, 5811 (2023). DOI — an extended scenario with a spacelike-separated observer.
• M. Zych, F. Costa, I. Pikovski, and Č. Brukner, “Bell’s theorem for temporal order,” Nature Communications 10, 3772 (2019). DOI — gravitational thought experiment and explicit causal assumptions.

Entry points

Start with the linear normalization constraints and a definite-order example. Learn what a witness certifies before attempting a gravitational realization.