Explanatory models in neuroscience: Part 2 - constraint-based intelligibility