Information theory, born shortly after World War II, has long been a fruitful source of interdisciplinary study, producing insights into communication so apparently universal and so intriguingly counterintuitive that fields as diverse as experimental psychology, particle physics, philosophy, biology, economics, and aesthetics have felt its influence.1 Among recent advances in mathematics, perhaps only game theory has inspired theoreticians in so many diverse subjects.2 One field that so far has not been subjected to information theory analysis is copyright law. Perhaps this demonstrates the pragmatic good sense of copyright scholars-law is not, after all, subject to mathematical proofs. Nevertheless, this Article argues that information theory can contribute useful insights into copyright doctrine, in part by suggesting a more objective and more inclusive alternative to the traditional model of "authorship."
The central character in the copyright drama is the "author," whose distinguishing characteristic is "originality." Rights to copyrightable works belong, at least initially, to the author3 and extend only as far as the author's original expression.4 In order to be copyrighted, a work must exhibit at least a small measure of originality.5 Most works surpass this threshold so easily that little inquiry into what we mean by "author" or "originality" is required. Generally, even the poorest sketch and the most hackneyed novel arc unmistakably the products of the mind and abilities of a particular individual. Some works, however, cio not so clearly exhibit an "author's" influence-highly factual works, such as telephone directories, with unremarkable characteristics of selection and organization; works that nearly reproduce existing works by other authors; or works created by mechanical processes with little human intervention. In such cases, the concept of authorship demands closer scrutiny.
The traditional model of authorship is frequently described as the "romantic" model.6 The paragon of romantic authorship is an individual who possesses "privileged access to the numinous."7 The romantic author employs his or her gifts in the service of mankind, creating works to entertain and enlighten-works that are unmistakably and uniquely the product of the author's singular vision. This conception of authorship has by no means disappeared; many authors and artists are still revered for their genius, and new provisions of the Copyright Act protecting artists' "moral rights,"8 particularly in certain works of "recognized stature,"9 suggest the continuing vitality of the romantic model.
Recently, however, both literary critics and legal scholars have questioned the romantic notion of authorship as an expression of individual will. Critics focus on the work, or "text," at the expense of the author. The "text" is a product of the author's cultural influences as much as his or her unique persona.10 In the collaborative endeavor of text consumption, the audience is at least as important as the author;11 consequently, all readings of a text are valid, even if they are not what the author intended.12 Perhaps Roland Barthes exaggerated when he announced "The Death of the Author,"13 but certainly the prevailing analytical perspective leaves the author greatly diminished.
On the copyright front, scholars have similarly questioned whether authorship is correctly envisioned as creation ex nihilo. It is often more accurate, they point out, to imagine authors assembling their works from the scraps of their cultural environment, transforming and adapting rather than making anew.14 Naive acceptance of authorship as a predominantly individual and creative act may foster authorial rights that are too broad or too powerful for the good of society.
Perhaps literary theorists can choose to ignore the author, but copyright law cannot. The principal object of copyright is to encourage authors, by the grant of exclusive economic rights, to create writings for the ultimate benefit of the public. …