Prof. Jane C. Ginsburg, Columbia University School of Law*
Aug. 14, 2009

The recent controversy over the publication of a new, and altered, edition of Ernest Hemingway’s memoir of his early days in Paris, A Moveable Feast (see, e.g., the New York Times account), inspires this exploration of the legal consequences of an heir’s reworking of a work of authorship.  In an op-ed contribution to the New York Times, A.E. Hotchner, Hemingway’s friend and biographer, proclaimed: “Someone who inherits an author’s copyright is not entitled to amend his work.”

Whatever the artistic or ethical merits of Hotchner’s lament, the legal basis of his objection is at best uncertain.  I propose to consider how copyright and other intellectual property norms would apply when the heir “corrects” a work’s content to suit his or her historical or artistic vision, or for whatever other reason alters the work.

The copyright analysis is, initially, straightforward: If the heir holds the rights to reproduce the work and to make new versions of it (the “derivative works right”), she is entitled to exercise those rights, including by authorizing a new edition that departs in significant ways from its predecessor.  Moreover, she may prohibit the reproduction and distribution of new copies of the prior edition, though she may not prevent the continued circulation of copies already made and sold.1

But what if there is more than one heir?  Accounts of the Moveable Feast controversy identify the grandson of Hemingway’s second wife (whose unflattering portrayal in the original edition the grandson revised in the second) as the heir; suppose, however, that other Hemingway descendants were co-owners of the relevant rights – would the outcome change?  Assume a deceased author has four children, by four different spouses.  Unless the author has provided otherwise by will (or previously transferred the relevant copyright interests), the children are co-owners of the copyright, and, unless they have contracted otherwise, each may exercise or license non-exclusive rights in the work, without the others’ permission, and subject only to a duty to account to the other co-owners for profits.

This means that each child could, at least in theory, rewrite the memoir from other fragments left by the author (assuming the fragments are also jointly owned), or indeed, from the child’s own inspiration, to improve the presentation of that child’s parent, and/or to denigrate the various stepparents.  Moreover, if one child liked her parent’s portrayal in the original, she could reissue the original version, without the other heirs’ accord (subject to a duty to account).  Complicating things further, suppose a third party infringed the original edition.  Any heir could enforce the copyright in the original, whether or not the others participated in the action.2

Another variation on the ownership theme: Suppose the deceased author of a work first published before 1978 (like A Moveable Feast) transferred all relevant reproduction and derivative work rights to the publisher.  Under the “termination” provisions of the U.S. copyright law, 56 years following first publication, the heir(s) may terminate the grant to the publisher and reclaim rights (for the United States) to publish and to make new derivative works.3  (For rights transferred after 1977, the author or heirs may terminate 35 years after entering into the contract.5)  Thus, the disgruntled descendant could, having properly terminated the author’s grant of rights, prohibit the former publisher from publishing further, and could issue her own version of the work.

If there is more than one heir, then a majority of the heirs (calculated per stirpes) must effect the termination.  All the heirs jointly obtain the reverted rights, even if a minority did not agree to terminate.  In a departure from the usual rules that joint owners may separately exploit the work on a non-exclusive basis, and that a grant of non-exclusive rights need not be in writing, the termination provisions of the Copyright Act specify that a further grant (apparently either exclusive or non-exclusive) of the reverted rights requires a signature of a majority of the holders of the reverted rights (the minority that do not sign the grant will nonetheless be bound by the majority action).5

As a result, if the deceased author has three children by three different spouses, and the second and third children dislike their parents’ portrayals, they could agree to terminate the rights in the first edition, despite the first child’s unwillingness to withdraw or change the original.  On the other hand, no two of the three children may be able to agree on how a revised version should present their respective parents, and therefore there will be neither an original nor a new edition.

The moral of this story: Even if the original publisher holds the rights at the author’s death, the heirs may subsequently re-enter the picture.  While the termination right was conceived, and is probably most often employed, primarily to improve the financial return to the author or her heirs, it is at least conceivable that the reversion to the heirs of control over the work’s exploitation could result in a withdrawal of and/or alterations to the work.

The role of the heir as the villain of the piece (or at least the bowdlerizer) complicates the recourse that “moral rights” might otherwise allow … in other countries, notably France (where one might expect Hemingway’s works to enjoy a market).  Moral rights entitle the author, or her heirs, to protect the integrity of the work, even after the author has granted rights of economic exploitation.6  The United States (despite accepting an – unenforceable – international obligation to protect the author’s moral rights7) has by and large not implemented these rights, either directly, or through other provisions in the copyright or other laws.

In a moral rights-respecting jurisdiction, if the publisher’s new edition substantially changed the content of a work without the approval of the author or her heirs, they would most likely have a claim for a violation of the work’s integrity, notwithstanding the author’s publishing contract.  But if the heir has allegedly violated the work’s integrity, there may be no claim because the only person who can bring the claim is the person who has allegedly mangled the work.  Where there is more than one heir, however, any heir can bring the claim,8 including, perhaps, against another heir.

Returning to the Moveable Feast controversy, suppose the original version has gone out of print, and that the altered version is the only one offered for sale.  (Currently, the first edition, while available new from at least some booksellers, e.g., Amazon.com, is no longer on the publisher’s website.) The original will remain available in libraries and from second-hand booksellers, but over time, one might imagine that access to the original will become increasingly difficult.

May a scholar, or for that matter, any member of the public, set the literary record straight without incurring copyright liability?  The answer largely depends on how the setting straight is done.  Were the scholar (or the user) to publish a comparative table, identifying the altered chapters in the new edition, and summarizing the information deleted from the text (without direct quotations or close paraphrasing), no copyright interests would be implicated.  Copyright protects expression (Hemingway’s way of telling the story), not the information the story conveys (the fact that in the original version Hemingway depicts wife # 2 as insinuating her way into Hemingway’s susceptible affections).

What if the record-straightener also includes comparative quotations from the original and altered versions?  The fair use defense traditionally immunizes exactly this kind of copying: to support the point for which the critic is arguing.  But the critic should not go overboard: Justifiable quotation to prove an argument does not in turn entitle the user to republish the entire chapter.  Courts typically are vigilant to ensure that the scholar or other user does not gild the lily, bootstrapping more than the critical or scholarly enterprise requires.9

Arguably, this caution places judges in the inapt position of second-guessing the user’s literary judgment, but the alternative is probably worse, abandoning examination any time a defendant waves the flag of criticism or commentary.  Indeed, the fair use provision of the Copyright Act identifies criticism and commentary as uses that may be ruled “fair,” but nonetheless directs courts to apply a variety of factors, including one addressing the amount and substantiality of the copying in relation to the copied work.10

Finally, is there any recourse outside the copyright law if an altered work is purveyed without signaling its departures from the original?  The new edition of A Moveable Feast is subtitled “the restored edition,” which at least suggests some differences with the first edition.  Moreover, the state of the manuscript at Hemingway’s death prompts considerable discussion about the “authenticity” of any version.

Suppose, however, a “definitive” original is substantially altered by an heir, without disclosure.  Arguably, selling the work by its original title, when its contents have significantly changed, misleads the public as to the nature or quality of the goods.  So phrased, the conduct sounds like a kind of false advertising, which might be redressed through a claim for violation of the Federal Trademarks Act’s prohibition of “in commercial advertising or promotion, misrepresent[ing] the nature, characteristics, [or] qualities of [the purveyor’s] or another person’s goods….”11

Unfortunately, in a 2003 decision, Dastar Corporation v. Twentieth Century Fox Film Corporation,12 the Supreme Court held that a related provision, addressing the false designation of the “origin” of the “goods” covered by the Trademarks Act, did not include intellectual goods, but only physical goods.  Thus, apparently, to attribute the provenance of physical books to HarperCollins when they were in fact published by Scribner would violate the Act, but to attribute to Hemingway the authorship of a book written by someone else would not because intellectual provenance does not count.

By the same token, it may appear, giving an altered book the same title as the original, even if it misleads as to the contents of the book, would not count because the work of authorship is not a covered “good.”  While it is not clear the Supreme Court intended to immunize this kind of misrepresentation, lower courts have broadly interpreted Dastar to remove copyrighted works from the ambit of the false representation provisions of the Trademarks Act.13

This survey of the legal landscape suggests that A.E. Hotchner and those who share his dismay had best try their claims in the court of public opinion.