Sherrylyn Branchaw points me to this quotation from “On Fairy-Stories”:

Therein lies the inherent weakness of the analytic (or ‘scientific’) method: it finds out much about things that occur in stories, but little or nothing about their effect in any given story.

Sounds like, in the opinion of Prof. Tolkien, my project is doomed.  It’s a good thing scientists don’t accept arguments from authority.