Crosswords and copyright


A plagiarism scandal has erupted in the strange and esoteric community of crossword puzzle constructors, and—as plagiarism scandals usually do—it sheds some interesting light on what we mean by "plagiarism" and how that relates to what we mean by "copyright infringement."

Matt Gaffney's article at Slate tells the story, building on an earlier analysis, involving comparisons among several thousand published puzzles, by Saul Pwanson and picked up by Oliver Roeder at Basically, dozens of puzzles that were edited by Timothy Parker (crossword editor for USA Today and Universal Syndicate) bear rather striking similarities to earlier puzzles published in the New York Times.

The similarities that have people the most riled up concern similarities in "theme." As those of you who are regular crossword puzzle solvers know, a crossword's theme is some clever trick, or gimmick, that links various answers together in an interesting way.

For instance, a puzzle might have the same one-word clue—exasperate"—for three long answers ("DRIVEUPTHEWALL," "GETONONESNERVES" and "RUBTHEWRONGWAY"). Or it might have the clues "breakfast combo," "lunch combo" and "dinner combo" for the answers "COFFEEANDDANISH," "SOUPANDSANDWICH" and "MEATANDPOTATOES."

It's what gives crosswords their identity, and what makes them, possibly, memorable for the solvers—figuring out what the theme is, and then working it out, is one of the great pleasures of the task of puzzle-solving.

Copying someone else's puzzle theme seems to be considered an unpardonable sin among puzzle-writers. Gaffney, himself a well-known and widely published puzzle editor, writes that "in the crossword world, copying someone else's themes is verboten. A theme is the soul of a crossword, what makes it unique. … The evidence suggests that Parker did in fact swipe themes, and did it repeatedly, over a long period of time. If USA Today and Universal come to the same conclusion, then they should treat this as plagiarism. No other word fits better in this particular puzzle."

To Roeder, "replicating another crossword's theme is like puzzle identity theft." And Will Shortz, the Times' crossword editor, when presented with the evidence of Parker's theme replications by Roeder, said: "I have never heard of something like this happening before. … To me, it's an obvious case of plagiarism. It's unethical, and I would never publish a person who plagiarizes another person's work."

Not surprisingly, Parker disagrees. "For themes to be the same is not an unusual thing in crosswords," and he chalked it up to a kind of statistical inevitability in a population of more than 15,000 puzzles he has edited over the years.

What's interesting, to me, in all this, aside from the light it sheds on puzzle construction, is that it illustrates how "plagiarism," though it is often conflated with copyright infringement, actually covers very different territory and involves very different interests. A crossword's "theme" is probably one element of the puzzle-creator's work that is not protected by copyright; copyright law doesn't protect "ideas," only the expression of ideas, and a puzzle's theme is, in my opinion, just such an unprotectable "idea," free for the taking (as far as copyright law is concerned). But it's precisely this kind of taking—theme theft—that gets the angriest response from those in the puzzle-writing business.

This has a direct parallel in academic writing. There, too, the plagiarism norms focus on a kind of borrowing that the law of copyright deems permissible: taking another's ideas or expression without attribution. Nobody in the academic world will complain if you use their ideas or quote their work—in fact, that's very much the whole point of the enterprise. But to do so without citation—that will get you into the hottest of hot water. [Just ask Doris Kearns Goodwin, or Stephen Ambrose or Joseph Ellis]. Yet copyright law gives an author no enforceable right to have his/her work properly attributed to him/her—a fact that surprised the hell out of many of my law prof colleagues when they first learned of it (insofar as proper attribution was really the only thing they cared about).

I suspect this is an example of a more general phenomenon that either illustrates (a) how norm-creation wisely supplements legal rules to enforce those interests that are important to the community, or (b) how badly copyright law tracks the things that people think are important.

[Thanks to Josh Kershenbaum for the pointer.]