Is this really a copyright problem?
posted by Lea Shaver
Say we agree there’s very little being translated into languages like Zulu and we agree that’s a problem. Is the problem that copyright makes it too expensive? Or is the real problem that translation or publication is inherently too expensive, given the limited size of these audiences and low ability to pay?
A theoretical model and some examples
It’s clear that the potential profits from Zulu translations aren’t sufficient to cover the costs. But those costs fall into lots of different categories: copyright licensing fees, the translator’s labor, distribution costs… etc. Would merely eliminating the copyright barrier be enough to tip the equation?
Let’s build a theoretical model first and then my research can look for data points…
P = potential income from book sales
C = costs of copyright compliance (including fees to the copyright owner, transaction costs of negotiating the license, and lawyer’s fees to handle copyright compliance and disputes)
T = costs of performing the translation (translator’s labor)
D = distribution costs (printing costs, shipping costs, overhead, leaving a profit margin for retailers, etc.)
For translated works to be produced, we would need to see that potential sales outweigh the copyright, translation, and distribution costs.
P > C + T + D
We can think about this equation as applying generally to works in a particular language market. Or more accurately, we can think about it applying to any specific work. If potential sales appear to be greater than costs, then the market should produce the work. So, let’s look at some specific works that the market is producing in Zulu right now…
The Bible. Broad appeal and high willingness to pay = high potential income. No copyright fees or negotiation costs because the Bible (and the mid-19th century Zulu translations done by missionaries) are in the public domain. T is nil because the work was translated long ago. So it’s really all about the distribution costs.
Newspapers. Low willingness to pay but broad appeal, and advertising revenues as well. No copyright fees because facts and news items are not subject to copyright protection as long as you express it in a different way. It’s all about the translation/authorship cost and the distribution costs… which are cheap b/c it’s newsprint. Sold by sidewalk vendors earning low income.
Textbooks. Copyright fees are there in full, but the government is paying for it as well as the translation/authorship component. (A lot of the content is original, not translated.) Delivery to schools offers a more efficient distribution structure than operating retail stores to sell the books.
I hesitate to draw too many generalizations on the basis of just three examples. But these examples do suggest that for some works where the copyright costs are low (or subsidized by a party with deep pockets) written works are being made available. Decreasing the copyright costs should increase the universe of works for which this equation works. It will more likely be books with mass appeal, which lowers the marginal costs of translation.
The hidden costs of copyright
One of the challenges is that the cost of copyright compliance is not JUST what you contract to pay as the licensing fee. Even if the copyright holder were willing to waive the fee entirely, C might still be significant. It includes the author/publisher’s negotiating costs (probably at New York or London salaries), the South African publisher’s negotiating costs, and a South African lawyer’s fees to advise on copyright compliance and handle any disputes that arise. These transaction costs are a significant part of the cost of copyright.
Still, it would be nice to be able to more clearly quantify how much of the problem is copyright. Is copyright compliance 10% of the cost or 80% of the cost? The answer matters greatly for our confidence that eliminating copyright barriers would make a dramatic difference. I hope to be able to produce a reliable estimate as I get into the empirical side of the research. Any ideas/suggestions from readers about how to go about this, or related literature to look at?
One complication is that I think that copyright protection also significantly pushes up the costs of T (translation) and D (distribution), in ways that are not immediately obvious. For example, the cost of translation could be brought down dramatically by utilizing good automated translation software and crowd-sourcing mechanisms. But copyright issues have inhibited the use and development of these technologies.
Distribution costs are also strongly influenced by copyright law. Can anybody with a printer publish and sell copies or do you have to have personally contracted with a copyright holder? Will multiple translations compete? Can people transfer digital copies costlessly by cell phone or does this expose phone company to secondary liability? Can street vendors sell books on the sidewalk or will police confiscate what they believe to be pirated works?
So copyright law increases the costs even of the portions of the equation that seem initially to be not about copyright law. And these problems can’t be solved by getting permission from the copyright holder as to a single work. There are systemic market issues that require a critical mass of works being translated in order to bring down translation and distribution costs on a system-wide level.
Another systemic issue is readership. Once there are more books to read, more people will want to read books. Literacy rates should improve and the market for future works should expand, causing P to rise. There’s an ecosystem here where one book’s chances of success is tied to the successes of other books.
What about public domain works?
A related question I’ve gotten from Matt and others is this: If the real problem is copyright, shouldn’t we see plenty of translations of public domain works? Does the absence of Zulu translations of public domain works prove that copyright is not the real problem? Maybe, maybe not.
I need to do more research, but my prediction is that I will find more translations of public domain works. There are 10,000 Wikipedia articles in Yoruba and Gaelic. (Sorry, I have no figures for Zulu!) The Bible has been translated into more languages than any other work.
On the other hand, public domain works aren’t representative of works as a whole in terms of their appeal. Zulu readers will probably never rush to snap up translations of Charles Dickens. Like readers in America, they’d probably much rather read Harry Potter, the Hunger Games, and Fifty Shades of Gray.
Harry Potter, for instance, has been translated into Latin. Apparently even for a dead language, there is a sufficient effective demand among Latin students in high-income countries to support the costs of translation and copyright licensing.
I’m not holding my breath that anyone is going to translate my most recent book into Latin anytime soon. Even though it’s under Creative Commons license and they wouldn’t have to pay me a dime.
Thanks for the great questions and comments… keep them coming!