If someone were to publish word frequency lists derived from copyrighted films or books on their website would this constitute fair use, or be a breach of copyright?
The word frequency lists would simply consist of a list of individual, unique words found in the film/book and how many times each word occurs, ordered in terms of frequency.
The frequency of words appearing in a given work is a fact. It is not protected by any copyright on the work. Nor would it be protected by any copyright on an article, essay or other publication reporting such a frequency list. Facts, including this one, are not protected by copyright at all (See 17 USC 102(b) for US law on this. Other countries have similar laws). This is no more protected than a statement such as:
J Random Author published a book last moth entitled Adventures of Authors.
Facts about a work are in no way protected by the copyright on the work. The requirement of originality in a work for copyright to apply in US law was strongly stated in Feist Publications, Inc., v. Rural Telephone Service Co., 499 U.S. 340 (1991) (the Feist decision). However the rule that facts are never protected by copyright goes back at least to the 1909 US copyright act, and I think to the Statute of Anne (1710), the first somewhat modern copyright statute.
