According to this theory, we ignore the question of whether the words of a song are true. But this seems to ignore the long tradition of didactic poetry, which

According to this theory, we ignore the question of whether the words of a song are true. But this seems to ignore the long tradition of didactic poetry, which was typically sung rather than simply read aloud. How would this theory account for Lucretius' De Rerum Natura, for instance, or Hesiod's Works and Days? These are musical works, intended for singing, but they are also supposed to convey fairly detailed factual information to the listener. We still use song for similar purposes today (e.g. in religious hymns) so this shouldn't be too surprising.