³⁷This is called hierarchical control (Poole and Mackworth, 2010), hierarchical planning or hierarchical task networks (Georgievski and Aiello, 2015; Nau et al., 2003) or hierarchical reinforcement learning (Sutton et al., 1999; Dietterich, 2000; Botvinick, 2012). A related paradigm is given by what is called “options” in reinforcement learning, see (Sutton and Barto, 2018, p. 461).