Bell's theorem can be understood as establishing the incompatibility of quantum correlations with any classical causal explanation. Causal inference, however, relies not only on passive observations but also on interventions. Here we ask when the observational and interventional data collected on a given causal structure can all be reproduced by a single classical causal model, which is a particular instance of the data fusion problem in causal inference [Bareinboim and Pearl, PNAS 113, 7345 (2016)]. We show that quantum systems give rise to a new form of nonclassicality in this setting, namely: when the data tables associated with each (non)intervention paradigm admit classical explanations, but no single classical model explains the different intervention paradigms when considered jointly. We call this nonclassicality in the synthesis of the data fusion. We assess the prevalence of such fusion by partitioning all marginalized causal structures over three observed variables into exactly two categories: those which can and those which cannot exhibit such nonclassicality. For all those that can, we present explicit quantum violations of the associated causal inequalities.
Bell's theorem can be understood as establishing the incompatibility of quantum correlations with any classical causal explanation. Causal inference, however, relies not only on passive observations but also on interventions. Here we ask when the observational and interventional data collected on a given causal structure can all be reproduced by a single classical causal model, which is a particular instance of the data fusion problem in causal inference [Bareinboim and Pearl, PNAS 113, 7345 (2016)]. We show that quantum systems give rise to a new form of nonclassicality in this setting, namely: when the data tables associated with each (non)intervention paradigm admit classical explanations, but no single classical model explains the different intervention paradigms when considered jointly. We call this nonclassicality in the synthesis of the data fusion. We assess the prevalence of such fusion by partitioning all marginalized causal structures over three observed variables into exactly two categories: those which can and those which cannot exhibit such nonclassicality. For all those that can, we present explicit quantum violations of the associated causal inequalities.
From the modern perspective of causal inference, Bell's theorem -- a fundamental signature of quantum theory -- is a particular case where quantum correlations are incompatible with the classical theory of causality, and the generalization of Bell's theorem to quantum networks has led to several breakthrough results and novel applications. Here, we consider the problem of causal data fusion, where we piece together multiple datasets collected under heterogeneous conditions. In particular, we show quantum experiments can generate observational and interventional data with a non-classical signature when pieced together that cannot be reproduced classically. We prove this quantum non-classicality emerges from the fusion of the datasets and is present in a plethora of scenarios, even where standard Bell non-classicality is impossible. Furthermore, we show that non-classicality genuine to the fusion of multiple data tables is achievable with quantum resources. Our work shows incorporating interventions -- a central tool in causal inference -- can be a powerful tool to detect non-classicality beyond the violation of a standard Bell inequality. In a companion article "Quantum Non-classicality from Causal Data Fusion", we extend our investigation considering all latent exogenous causal structures with 3 observable variables.
Since Bell's theorem, it is known that local realism fails to explain quantum phenomena. Bell inequality violations manifestly show the incompatibility of quantum theory with classical notions of cause and effect. As recently found, however, the instrumental scenario-a pivotal tool in causal inference-allows for nonclassicality signatures going beyond this paradigm. If we are not limited to observational data and can intervene in our setup, then we can witness quantum violations of classical bounds on the causal influence among the involved variables even when no Bell-like violation is possible. That is, through interventions, the quantum behavior of a system that would seem classical can be demonstrated. Using a photonic setup-faithfully implementing the instrumental causal structure and switching between observation and intervention run by run-we experimentally witness such a nonclassicality. We also test quantum bounds for the causal influence, showing that they provide a reliable tool for quantum causal modeling.
Bell's theorem is often considered the most stringent notion of nonclassicality. The generalization of Bell's theorem to causal networks offers interesting new perspectives on the phenomenon of quantum nonclassicality and prompts us with a fundamental inquiry: what is the simplest scenario leading to the incompatibility between quantum correlations and the classical theory of causality? Here, we demonstrate that quantum nonclassicality is possible in an entanglement swapping network consisting of only three dichotomic variables, without the need for the locality assumption or external measurement choices. We also show that interventions, a central tool in the field of causal inference, significantly improve the noise robustness of this new kind of nonclassical behavior, making it feasible for experimental tests with current technology.
Explaining observations in terms of causes and effects is central to empirical science. However, correlations between entangled quantum particles seem to defy such an explanation. This implies that some of the fundamental assumptions of causal explanations have to give way. We consider a relaxation of one of these assumptions, Bell's local causality, by allowing outcome dependence: a direct causal influence between the outcomes of measurements of remote parties. We use interventional data from a photonic experiment to bound the strength of this causal influence in a two-party Bell scenario, and observational data from a Bell-type inequality test for the considered models. Our results demonstrate the incompatibility of quantum mechanics with a broad class of nonlocal causal models, which includes Bell-local models as a special case. Recovering a classical causal picture of quantum correlations thus requires an even more radical modification of our classical notion of cause and effect.
In his seminal paper, Bell [Physics Physique Fizika 1, 195 (1964)] considers the correlations that result from space-like separated measurements on a pair of entangled particles. He uses relativity theory to motivate the Bell causal structure, then shows the existence of quantum correlations that cannot be explained classically within this causal structure. Classical explanations of such quantum correlations are possible in alternative causal structures, for instance, those that allow superluminal causal influences, but, as shown in [New Journal of Physics 17 033002 (2015)], all such alternative explanations require fine tuning (causation without correlation). Here we discuss the existence of spurious quantum correlations --- correlations that look quantum in one causal structure, but have a natural classical explanation in another. More precisely, there are causal structures that admit non-classical quantum correlations, but for which the same correlations have a classical explanation in another causal structure without fine tuning. The realisation in the other causal structure can also be achieved without breaking any natural constraints on the causal structure that follow from relativity theory. However, similarly to non-classical quantum correlations in the Bell causal structure, we find other causal structures with non-classical quantum correlations that do not have a classical causal explanation in any alternative causal structure without fine tuning.
Bell scenarios are multipartite scenarios that exclude any signalling between parties. This leads to a strict hierarchy of classical, quantum, and non-signalling correlations in such scenarios. Here we consider a minimal relaxation of non-signalling: each party is allowed to receive a system once, implement any local intervention on it, and send out the resulting system once. Crucially, unlike Bell, we make no global assumption about causal relations between parties, e.g., they could be embedded in some exotic spacetime with indefinite causal order. We do make a causal assumption local to each party, i.e., the input received by it causally precedes the output it sends out. We then ask: Can we device-independently certify the nonclassicality of multipartite correlations in such scenarios, just as Bell inequality violations do so in Bell scenarios? A priori, this is not clear: without some assumptions on the underlying physics (e.g., non-signalling), parties can realize arbitrary correlations. We therefore make a minimal assumption of logical consistency on the underlying physics, i.e., it must be free of time-travel antinomies without imposing any restrictions on local interventions of the parties. We then define antinomicity as a device-independent notion of nonclassicality and prove a strict hierarchy between correlation sets based on their antinomicity. An antinomic correlation cannot be explained by a classical physical theory compatible with free local interventions on pain of logical contradictions in the theory. On the other hand, parties exchanging quantum systems can witness antinomicity while respecting logical consistency. Antinomicity reduces to Bell nonlocality for non-signalling parties. It also resolves a conceptual puzzle, namely, the failure of causal inequalities as witnesses of nonclassicality: antinomicity implies causal inequality violations, but not conversely.