Given my present belief in the much greater range of variability as to both order and kind of complexity in the world's musics versus the world's languages, I can hardly imagine how a model developed really satisfactorily for the detailed structureal explanation of one musical language is so easiy modified to another, and all the more so if the original model be evolved from linguistics rather than from the musical disciplines.

Harold Powers, "Language Models and Musical Analysis", p.48.