Correcting two common mistakes about how much fair use is too much

[Note: This is part 3 of a 4-part series on fair use.]

Plagiarism makes the news all the time, and rightly so. When an author borrows from another author or a musician from another musician, this feels wrong if it happens without permission and it's a ripoff, but okay if it leads to something fresh and new. How can a creator know how much is too much to take?

I've previously addressed the first two parts of the fair use test: looking at whether the use is transformational and looking at the nature of the original work. The third element of the test requires you to consider the quantity and value of the portions of the original material that you're using in your new work.

The best way to explore what this means is to correct two commonly-made mistakes about this part of the test:


Mistake #1: The "10% Rule". I've often seen people say that they think a certain use is legal because they've counted lines of a book or bars of music and it's less than a certain percentage or number of words. And while counting can be helpful, it's never going to get you all the way to safety. After all, if you quote only 17 syllables of a haiku, you've taken the whole thing.

Harper & Row v. Nation Enterprises shows this point clearly. Harper & Row was the publisher of US President Gerald Ford's memoirs and had licensed first serial rights to Time Magazine. But somehow The Nation magazine got its hands on a copy of the memoirs before they were published and before the Time article ran. In the hope of scooping Time, the Nation quickly prepared a 2,250-word article focusing on Ford's pardon of President Richard Nixon and quoting 300 words directly from the unpublished book.

It worked. The Nation beat Time to the punch. Time cancelled its article and terminated its contract with Harper & Row. So Harper & Row sued Nation.

Unfortunately for Nation, although it used only 300 words from the book, they were 300 really important words. In fact the original court hearing the lawsuit held these were "essentially the heart of the book". Put another way: it's not good enough to say "I only copied a little bit" if the parts you copied were the most important ones.


Mistake #2: All counting is irrelevant. Again, not true. Looking at the Nation case again, the court did do some counting. But what it counted was the words in the Nation article. 300 words out of 2,250 = 13% of the article was taken from the unpublished manuscript. This shows that although it matters what proportion of the old work you took, the more important question is how much of your new work comes from someone else.

The Acuff-Rose case, on the 2 Live Crew song "Pretty Woman", gives some indication as to why the counting matters in this context. Where the new work is comprised largely of elements from a prior work or works, it will be even more important that the new work be transformational. The 2 Live Crew song could use a large amount of the Roy Orbison song which it was parodying because otherwise a parody wouldn't work. But if the transformation isn't there, then the amount of copying is going to be very important.

This will be even more important in a non-fiction situation. At least fiction writers can also protect things like characters and plot elements. Non-fiction writers have to share their underlying facts with the world. So all they have is their own unique way of describing them. And the courts will therefore look to this test more in that circumstance.

That's not to say it's only important in non-fiction. In Warner Bros. v. RDR Books, a case about a Harry Potter encyclopedia, the court looked at this element of the fair use test also to determine that the encyclopedia wasn't just a set of "facts" from the Harry Potter universe but rather an overly-intrusive copying of JK Rowling's own words. It required the author to make changes to remove her words while still allowing the book to keep the entries.

This makes sense by the way, at least to a lawyer. Copyright exists to protect and encourage creative work. If you're not taking much of the original work then you're not diminishing the original creation. But if your new work is mostly borrowed then it's less likely (although not impossible) that you've made something truly new.

So here's your takeaways from this part of the test:

1. There is no safe minimum, whether number of words or percentage of the item copied. Even a minimal copying can be illegal.
2. If your own work is mostly made up of pieces from someone else's work, you should be extra-careful.
3. If you're writing non-fiction, be extra-careful not to plagiarize your sources.

I've saved the best for last: effect on the potential market. If you're not costing the other person any money, how much does this matter? Come back and find out.


LINKS:
Legal Minimum post on fan fiction:
http://legalminimum.blogspot.com/2012/08/want-to-make-million-dollars-from-fan.html
Campbell v. Acuff-Rose Music:
http://scholar.google.com/scholar_case?q=acuff-rose&hl=en&as_sdt=2,48&case=16686162998040575773&scilh=0
Harper & Row v. Nation Enterprises:
http://scholar.google.com/scholar_case?case=12801604581154452950&q=acuff-rose&hl=en&as_sdt=2,48&scilh=0
Warner Bros v. RDR Books:
http://scholar.google.com/scholar_case?q=warner+RDR+books&hl=en&as_sdt=2,48&case=13852164224811081270&scilh=0