Copyright starts with the written word as its model, then tries to fit everything else into the literary mode. It oscillates between two positions on non-textual creative works such as images - they are either transparent, or they are opaque. When courts treat images as transparent, they deny that interpretation is necessary, claiming that the meaning of an image is so obvious as to admit of no serious debate and that the image is a mere representation of reality. When they treat images as opaque, they deny that interpretation is possible, pretending that images are so far from being able to be discussed and analyzed using words that there is no point in trying. This oscillation between opacity and transparency has been the source of much bad law.Tushnet notes that -
This article explores the ungovernability of images in copyright, beginning with an overview of the power of images in the law more generally. The article then turns to persistent difficulties assessing copyrightability and infringement for visual works. In assessing copyrightability, courts draw lines between artistic choice and mere reproduction of reality, but also treat the artist as a person with a special connection to reality, a way of seeing that ordinary mortals lack. Infringement analysis repeats this doubling, using the representation/reality divide to separate protected elements of a specific work from unprotected ones while simultaneously insisting that works are indivisible gestalts. Current doctrine makes impossible, self-contradictory demands on fact-finders. It should be replaced with a true “reproduction” right against exact or near-exact copying.
Despite this radical proposal, much of my argument is critical and diagnostic. I therefore turn to more specific problems in authorship questions for multimedia works and fair use that highlight the instabilities in current approaches to non-textual works. Greater epistemic humility, recognizing that images make multiple meanings in multiple ways, could combat the judicial tendency to presume that images are nothing more than what they seem.
Larry Lessig has eloquently written about how freedom to quote is the foundation of textual fair use; quotation is the foundation of scholarship, news reporting, and many other important endeavors. While courts have begun to recognize that copying an entire picture may be necessary to critique or analyze it, freedom to quote has historically been far less likely to be recognized for music or video, creating a significant gap between good educational and scholarly practice and the law. Literary analysis often proceeds quite well with quotations, but art history and criticism routinely require whole pictures, not fragments, to make their points. As Lessig points out, it is bizarre that freedom to quote a Hemingway novel is the baseline, but not freedom to quote the filmed version.
Standard fair use analysis, with its prototype of the text, favors partial and limited quotations. The amount of the work used is even an enumerated factor in the statutory definition of fair use In order to protect critical, news-reporting, and similarly fair uses of images. Yet with images, paraphrase is often insufficient to achieve a legitimate objective. In a case from the 1960s, Bernard Geis, the fear of liability led a commentator analyzing the Kennedy assassination to redraw frames of the Zapruder film rather than reproducing them mechanically—and he still got sued, because he‘d reproduced the features of the film that made it valuable. Mechanical reproduction would have been better because it would have been more persuasive.
Later copiers recognized this need for accurate reproduction in their fair uses. The influential Second Circuit has started to accept the need for the special veridical power of images, both in works of fiction and nonfiction, as has the Ninth Circuit. A district court, attempting to show those aspects of Superman that had been established as of Action Comics #1 (which was important because one creator‘s heirs had been able to recapture the copyright in that issue, but not in aspects of the Superman character subsequently developed), even reproduced the entire issue of the comic book as an appendix to its opinion, apparently confident that this was legitimate because it was the best way to show what aspects of Superman the respective parties owned.