I also don't parse singing as language intuitively. It might be a result of context-based parsing mixups in human brains, so it may not be so easy to just try.