Delayed-Choice Experiments and the Metaphysics of Entanglement

Entangled states of quantum systems have played an important role in debates on the metaphysics of contemporary science. Teller (1986), French (1989), Maudlin (1998) and Esfeld (2004) all base metaphysical claims on the quantum mechanical description of entangled two-particle states. Glossing over the substantial differences between these accounts, their general thrust is that quantum mechanics pushes us away from a metaphysics of individual things towards a metaphysics of relations, relations which do not supervene on the intrinsic properties of their relata. Of course, such arguments presuppose that what appears as the description of an entangled state in the quantum formalism actually represents a real relation in the world. This presupposition derives support from the fact that measurements on entangled states exhibit significant correlations, as demonstrated, for example, in experimental tests of Bell’s inequality. Furthermore, the flourishing of quantum information theory in the last two decades has resulted in a rather dramatic shift of attitude towards entanglement: While it was once considered a bizarre consequence of the quantum formalism with little importance outside of philosophical debates, entanglement has now come to be recognized as a physical resource which can be experimentally manipulated in various ways.1 In the present paper, I will investigate whether this manipulability really supports a realist stance on entanglement relations. More specifically, I will focus on one particular kind of such manipulations, namely entanglement swapping. In this process, two pairs (A,B) and (C,D) of entangled particles are created by two independent sources. If one then performs the right kind of joint measurement on particles B and C, the pair (A,D) enters into an entangled state even though A and D have never interacted with each other. Interestingly, this phenomenon has given rise to contradicting ontological conclusions. On the one hand, Clifton (2002, S163) takes it to support a realistic view of entanglement: “It appears that there is sufficient substantiality to entanglement that it can be swapped from one pair of particles to another”. On the other hand, Healey (forthcoming, sec. 4) studies an entanglement-swapping experiment in which the measurement