Couldn't it be possible for ear protrusion to be uncorrelated with the other variables but still have a genetic basis - presumably by depending on a set of genes with little overlap over the ones that govern size?