“I refuse to admit there is no meaning here. I will find meaning. I will force it.”
Lloyd, Phonogram: The Singles Club #6
For over a half century, the use of music and magic has stayed firmly rooted in the model Tolkien created: a bardic, Bombadillian tradition of sorcerous chanting. As a result, we've had wizards extemporising poetry and forcing rhymes. It is way of creating a functional metaphor for the otherwise inexpressible: a way of objectively measuring and comparing the immeasurable and incomparable. Music, like, presumably, magic, relies on wit, vocabulary, memorisation, and artistic prowess: therefore magic-as-music become a recurring metaphor. It was a way for the reader to understand the un-understandable. Music as an expression of an individual's magical skill was, and still is, the dominant paradigm - it can still be found happily chanting away in many a high fantasy series.
In the 1990s, the technology and production of music changed (well, it did in the 1980s, but it took a while for fantasy fiction to catch up). Music was no longer, if you’ll pardon the pun, a one-man band – the bard, furiously racking his memory for demon-repelling doggerel. Music itself began to be made in layers: songs and studios, productions and distortion, sampling and remixing, a task that could involve dozens or hundreds of people… or even one vocally-challenged youth with an Amiga, a kit-bashed mixing board and an army of samples. Stage presence was no longer required – nor, for that matter, a singing voice. Pop stars may swan around like kings and queens, but, as Faithless pointed out, “God is a DJ.”