> <i>A diatonic scale has only seven pitches, and every melody ever created must work within that seven-note constraint. So the universe of possible melodies forms a grid.</i><p>Well, the diatonic scale is only one of 66 possible scales with seven notes (without rotation symmetric variants); in western music a lot of these scales next to the diatonic one are in regular use (altered, acoustic, hungarian minor, etc.). And you can also add many other scales with more or less than seven notes, which are in regular use in western music. If you also consider non-western music (which is also subject to copyright) a lot more scales become relevant. If you want to check yourself, here is a tool to do so: <a href="https://github.com/rochus-keller/MusicTools/tree/master/ScaleAnalyzer">https://github.com/rochus-keller/MusicTools/tree/master/Scal...</a><p>And these are only one-octave scales. There are also scales which change over octaves (e.g. in Jazz), and melodies usually extend over more than one octave. And this is only the pitch domain. If you also consider rhythm and phrasing, the number of useful possibilities is virtually infinite (countably infinite, but still infinite).<p>The problem from my humble point of view is not the (mathematically) objective difference between melodies, but the (subjective) perception of similarity by the judge and his/her experts. It would therefore not be possible to create legal certainty even if it were really possible to publish all the possibilities in advance, as the human arbitrariness of our legal system still exists. But the whole thing is undoubtedly an excellent marketing idea for a lawyer.