“…If c satisfies a twist condition, meaning x ∈ X −→ c(x, y 1 ) − c(x, y 2 ) has no critical points for y 1 = y 2 ∈ Y , then we shall see that not only is the minimizing γ unique, but its mass concentrates entirely on the graph of a single map f 1 : X −→ Y (a numbered limb system with one limb), thus solving a form of the transportation problem posed earlier by Monge [61,81]. This was proved in comparable generality by Gangbo [52] and Levin [66] (see also Ma, Trudinger and Wang [72]), building on the more specific examples of strictly convex cost functions c(x, y) = h(x − y) in X = Y = R n analyzed by Caffarelli [19], Gangbo and McCann [53,54], Rüschendorf [89,90] and in case h(x) = |x| 2 by Abdellaoui and Heinich [1], Brenier [17,18], Cuesta-Albertos, Matrán, and Tuero-Díaz [31,32], Cullen and Purser [34,35,87], Knott and Smith [64,93], and Rüschendorf and Rachev [91]. Adding further restrictions beyond this twist hypothesis allowed Ma, Trudinger, Wang [72,100], and later Loeper [68], to develop a regularity theory for the map f 1 : X −→ Y , embracing Delanoë [36], Caffarelli [20,21] and Urbas' [102] results for the quadratic cost, Gangbo and McCann's for its restriction to convex surfaces [55], and Wang's for reflector antenna design [106], which involves the restriction of c(x, y) = − log |x − y| to the sphere [57,107].…”