In a recent discussion about Sci-Hub, I took some heat for bringing up the history of copyright and the shifting attitudes over time about what it is good for. So I should probably know better than to appeal to history again as I begin some thoughts about the unique place of translations in copyright law. Nevertheless, I think it is important to note that translations were not explicitly protected by U.S. law until 1870, and that 17 years earlier, an author of no less stature than Harriett Beecher Stowe lost a lawsuit in which she tried to prevent the publication of an unauthorized translation into German of Uncle Tom’s Cabin. As this article about the case tells us, the judge asserted that once it was published, Stowe’s expression was as available to translators — as much “public property” — as were the works of Homer or Cervantes.
Today the peculiar place of translations is still a subject of some debate in copyright discussions. For one thing, the ability to control translations is often asserted as something that academic authors should not give up, as they seem to do when they are required, usually by funding agencies, to license their works under a CC-BY license. Since these requirements usually apply to (mostly) scientific articles, it is legitimate to ask how serious this concern really is. But, to my mind at least, this concern raises a much bigger issue, the place of translations in the attempt to draw a line between derivative works, over which the author holds an exclusive right, and transformative fair uses. Can we make the case (as my colleague Will Cross recently did in conversation) that translations are, at least some of the time, more like transformative uses that should be permitted by fair use, instead of derivative works that are entirely under the control of the original author?
It certainly seems true that translation is a highly creative act and results, by definition, in new expression. So it is possible to imagine that the idea/expression dichotomy in copyright could be used to argue against the appropriateness of giving the original author exclusive rights over translations. Consider a highly creative work like a poem or play; once it is translated into a new language, all of the expression has changed, the aesthetic is very likely quite different, and what remains seems very much like the underlying idea of the original. Should Arthur Rimbaud’s estate (hypothetically, since there is a good chance the original poem is in the public domain) be able to control Samuel Beckett’s translation into English of Le Bateau ivre, which critics acknowledge deserves to be considered a great poem in its own right? Or could that control undermine the purpose of copyright by stifling new creativity? Even if idea/expression does not create space for translations, the central notion of transformation in the fair use analysis points toward recognizing that what Beckett created as The Drunken Boat could well be justified, even if Rimbaud executors objected.
Academic works, of course, are different than poetry in most cases. On the one hand, there is likely to be more room for fair use when a work is critical or factual than there is for a highly creative poem. On the other hand, there may be a stronger argument for giving the academic author some level of control over translations of her work, precisely because it is intended to inform and educate, purposes which could be undermined by a poor translation.
These reflections point up to me the need for more flexibility in copyright law, even beyond what fair use already provides. Ideally, the law should recognize the diverse motivations and incentives that different authors have, and adjust to that diversity. Poets and physicists write and publish for very different reasons. Fair use might be able to account for this diversity with a more sensitive analysis of the second factor, but we have never seen that analysis develop. Instead, I want to suggest two approaches — both hypothetical — that might help us with the problem of translation outlined here.
The first place to look for a better approach to translations would be the moral rights tradition, which is strong in the copyright laws of much of the world, but quite weak in the U.S. The moral right of attribution gives a creator the right to have her name associated with her work, and also, importantly, to have her name removed from any alteration of the work that might impair her reputation. It is reputation, of course, that academic authors care most about, since all of the economic rewards they receive, indirectly, for their published work come through acknowledgement, citation and other reputational mechanisms. So if the U.S. had strong moral rights, an academic author would be able to demand that her name be removed from any incompetent translation of her work. Indeed, the strong appeal of the CC license for academic authors is that it creates a contractual obligation to cite an original creator and to remove that creator’s name from a derivative work if she demands it. Recognizing the moral right of attribution more broadly in the U.S. would help deal with problems of plagiarism and poor, unauthorized translations more directly, rather than via licensing.
My second suggestion is even more speculative, involving the idea of liability rules, which I first wrote about over five years ago. In contrast to an exclusive right, which is what copyright now grants to creators, a liability rule would not prevent someone other than the creator from making a specified use of the work, but would require compensation in certain conditions (such as harm to the creator’s reputation). Under a liability rule regime, then, an author could not prevent a translation by asserting an exclusive right over that form of derivative work, as she can today, but would have to be compensated if, for example, the translation was detrimental to her reputation, or even if, perhaps, it enjoyed commercial success. So in my example above, Samuel Beckett would not need to ask permission from Rimbaud or his estate in order to translate the poem, but might be obligated to share profits if successful or pay damages if harmful. This kind of liability rule could introduce into copyright the kind of flexibility to deal with the different motivations for creativity and the uncertainty around translations.
Both of these ideas are more thought-experiments than they are practical suggestions. I do not believe either one is going to become part of our copyright jurisprudence anytime soon. But it is instructive to think creatively about the relationship between rights holders and users, in order to get past the either/or dichotomy between exclusive rights and fair use. Copyright can be described as a negotiation between the interests of creators and users, and in negotiation it is always helpful to get behind the status quo and ask where the real interests lie for all sides. I hope these speculative attempts to push into that realm can provide some useful perspective on a difficult and unclear area of copyright — the nature, status, and right to control translations.