Fig. 1
From: The relativistic causality versus no-signaling paradigm for multi-party correlations

The causal structure of the three-party Bell experiment. The outputs A, B, C are correlated via a common Λ. The inputs X, Y, Z are chosen freely according to the notion of free choice. The input X cannot signal to change the distribution of a remote party’s output B or C, analogously for inputs Y and Z. On the other hand, depending upon the space–time configuration of the measurement events, the input may influence the correlations between remote parties’ outputs without violating causality. In particular, in the measurement configuration shown in Fig. 2, the input Y may influence the correlations \({\cal{C}}_{{\mathrm{AC}}}\)