“…However, this raises a well‐known paradox of CC formation that mantle‐derived melts are generally basaltic and differ from CC, which has an overall andesitic‐dacitic composition (Figure 1a; Rudnick & Gao, 2014). Resolving this discrepancy has proven to be a challenge, and three models are generally invoked: (a) delamination of mafic/ultramafic rock of lower crust into the mantle (Figures 1b−1c; Herzberg et al., 1983; Jagoutz & Kelemen, 2015; Lee & Anderson, 2015; Ringwood & Green, 1966); (b) relamination of subducted felsic rock to the base of the arc (Figures 1d−1e; Hacker et al., 2011, 2015; Hess, 1962; Kelemen & Behn, 2016); and (c) partial melting of basaltic oceanic crust (Gazel et al., 2015, 2019; Green & Ringwood, 1968) or eroded arc crust (Gómez‐Tuena, Mori, & Straub, 2018; Straub et al., 2015) in the mantle. All three processes may occur, but whether any mechanism dominates CC formation is unclear.…”