A mostly informal note on how it is possible to recast head movement as external merge in contemporary set-theoretic Minimalism (and some thoughts on why it may be a good idea to shift to graph theory, where the problems that have been identified do not arise).