Interview with James Grimmelmann
For podcast release Monday, September 14, 2015
KENNEALLY: What is it that robots cannot do? In 2015, robots not only build our automobiles, but they also can drive them. Robots vacuum floors and work alongside human beings in warehouses, and sleepless, tireless robots read everything they find on the web.
Welcome to Copyright Clearance Center’s podcast series. I’m Christopher Kenneally for Beyond the Book. In factories, human beings must act carefully when around robots, because, of course, the machines are mindless and unconscious, which makes them potentially very, very dangerous. In the view of at least one legal scholar, text-reading machines are also oblivious to and immune from copyright law.
James Grimmelmann is a professor of law at the University of Maryland, where he studies how laws regulating software affect freedom, wealth, and power. In a forthcoming article for the Iowa Law Review, he asserts that humanity and copyright face a kind of marginalization in a world of literate robots. Professor Grimmelmann joins me now from his College Park campus office. And welcome to Beyond the Book, James.
GRIMMELMANN: Thanks for having me, Chris.
KENNEALLY: Well, we’re delighted you can join us, because this piece is very thought-provoking and really relevant to our time. And yet, in order to understand the argument you make, we have to go back to another time and to this notion, as you put it in the piece, of romantic authorship. So for those in the audience who are wondering what we’re talking about here, we have to set the stage by saying why copyright is really intended for human beings first, and therefore that will help us understand why robots may be immune from it.
GRIMMELMANN: I’d say that out of all the fields of intellectual property, copyright is the most explicitly humane. It’s designed to promote human creativity, and that means humans doing artistic things that capture the imaginations of other humans – writing, music, movies. These are the things that fire the imagination. And copyright gets at this idea that they capture something of an author’s essential humanity and readers encounter that humanity and recognize it in the copyrighted work.
KENNEALLY: And what’s interesting about that notion of humanity is that copyright was born sometime in the 18th century, developed over time in many interesting ways, but it was always about a kind of a line between reading and copying. Ordinary acts of reading, as you put it, didn’t involve copying. There was this separate activity of reading directly separate from copying.
GRIMMELMANN: If you go back to the birth of modern copyright, it’s about control of printing presses. And the idea is that the printer contracts with the author and becomes the only persons authorized to make more copies of a book. So somebody else who acts without permission and creates their own printing press edition – they’re a pirate and an infringer.
But the reader who buys a copy of the book and opens it and enjoys it – they’re outside the system of copyright. Copyright never has claimed to prevent reading.
KENNEALLY: Right. And I guess we’ll have to underscore that point, that the notion of copying and infringing were almost an equation at that point. If you were copying something, unless you had the right to do it, by definition you were infringing, and yet today, in our technology-driven environment, copying is really an essential element of communication. We recently spoke with Professor Chris Sprigman from NYU Law School. He was talking about this very point that it may be the essence of 21st century communication that so much of it is about copying.
And in the piece, you make a point about the way that robots, meaning any kind of online ingestion of text, really have changed the activity of reading because it changed it in both the quality of the reading, the kind of reading that’s being done and the quantity of it, and this has implications for copyright.
GRIMMELMANN: There are very familiar human kinds of reading taking place online. Every time you download a book for your Kindle, it comes to you through the magic of the Internet. But the experience of a reader and a text is very similar to what it’s always been.
But there’s another kind of reading now increasingly taking place alongside it in which computers copy thousands, millions of works and read them in bulk. So a search engine spider goes out and reads as much of the Web as it can in order to figure out what’s out there. Or the Google Books project uses computers to scan and copy the words in millions of books. And that scale means it’s a new kind of reading, very different from what humans have done.
KENNEALLY: And that kind of non-human reading is, to use the term that you have created, non-expressive reading. Tell us a bit more about what you mean by non-expressive reading.
GRIMMELMANN: I borrowed the term from the scholar Matthew Sag, who was trying to find a way to explain a trend he saw in case law about fair use and computers and technology. And he noticed that there are some cases in which the defendant uses the work and argues, I wasn’t really making use of it in the same way that a human reader was. My reading was incidental.
So the search engine aren’t human. It’s like, well, we’re not really reading all these web pages to enjoy them or appreciate their content. We’re just indexing them. We’re making a map of the web, and a map is not a territory. We are not competing with the authors or the publishers of these works. We’re just, in fact, directing potential real readers to them.
KENNEALLY: Right. And Professor James Grimmelmann, we’re speaking with you today about a piece for the Iowa Law Review that you’ve written looking at literate robots and whether or not they could ever be infringing as they go about their activities online.
I wonder if you would tell us just a little bit more about some of the case law that has really supported this argument. For example, the one case that many people will be familiar with is of course the Google Books case, and there it was determined by Judge Denny Chin that the activity Google was engaged in was transformative. And I suppose again for listeners who are without a legal background such as yours, we should explain what we mean by transformative, and then you can say just how this applies in the Google Books case.
GRIMMELMANN: Right. The concept of transformative use comes out of case law on authors who build on other authors’ works. So just as West Side Story reworks the story of Romeo and Juliet, authors have always built on each other’s works, and transformative fair use was a way of saying that when a new use makes use of an old work and adds to it new understandings, new meanings, new context, new purposes, then it will be held fair and non-infringing.
And what’s happened is that that concept has taken flight. It’s gone beyond the cases of new works, of parodies that it was traditionally associated with, and it’s now used to catch, I’ll call it new contexts, for works, so that in cases involving search engines, the courts have said showing a small thumbnail of a picture helps search users realize what’s on the page they’re about to go to, that could be a fair use.
And when we get to the Google Books litigation, Judge Denny Chin borrowed that idea from the search engine cases to say that scanning these millions of books and having computers read them to make a giant index is a transformative use of these books. And part of his opinion that really struck me is, in explaining why it’s transformative, he says that Google Books does not supersede or supplant books because it is not a tool to be used to read books, which means that whatever those robots are doing when they copy the books is not reading.
KENNEALLY: Right, or at least not that humane reading that you were speaking of earlier. This has implications for all manner of published material online and for even most of the activities that we engage in online, if all of these robots and spiders are out there reading things, reading not only published material, but in the case of Gmail, for example, reading our e-mails in order to offer us various kinds of advertisements or other commercial incentives.
It is Google’s argument that we shouldn’t be concerned about this because, after all, it’s not human beings who are doing the reading of the e-mails, it’s only the robots. So even they really zero in on the significant difference between the reading activities.
GRIMMELMANN: Yeah, and you see these arguments in a bunch of different places, within copyright and across other bodies of law. There’s the sense that familiar human activities, reading, observing the world, moving around, driving, somehow they don’t count if computers do them. It might have the same results, but because the computer is not itself human, there’s this intuitive sense that the law ought to leave them alone or not treat it as being the same in kind as when a person does it.
KENNEALLY: And what’s fascinating about this, Professor Grimmelmann, is, not only is it perhaps true, at least in the legal sense, now, increasingly, the courts are less inclined to find this to be an infringing activity so computers cannot be accused of infringing, but they also can never really hold copyright themselves because copyright is about incentives to create, and it’s been pointed out that computers simply don’t need those kinds of incentives.
GRIMMELMANN: I think that that’s right for the computers we have now, but as robots become more sophisticated as readers and potentially authors, I think this idea that they’re categorically different and don’t count becomes questionable, because what if you did have a computer that says, I will write you a novel if you keep me plugged in and show me pictures of cats? And it could potentially spit out a copyrighted work that humans might enjoy, and copyright could be a tool for giving that computer the incentive that would cause it to write the copyrighted novel. Is it conscious of this? We don’t know. But given copyright’s goals, this commitment to pure romantic authorship becomes harder and harder to maintain as computers get more and more sophisticated.
KENNEALLY: Last year, Professor Grimmelmann, we had a question regarding whether or not animals could hold copyright, and so this summer you’re asking us whether or not robots or computers can, and you’re suggesting that perhaps not today, but sometime in the future, it’s possible that they could. So IBM’s Watson could have a copyright.
GRIMMELMANN: Right. I don’t see what the path from here to there looks like, but I think it’s possible in principle, and the fact that it’s possible in principle leads us to ask questions about what we have copyright for, and who is it meant to serve?
KENNEALLY: Well, we appreciate you raising the questions and answering some of them for us today. Professor James Grimmelmann is at the University of Maryland, where he studies how laws regulate software and affect freedom, wealth and power. He discussed a forthcoming article for the Iowa Law Review on copyright for literate robots. Professor Grimmelmann, thanks so much for joining us on Beyond the Book.
GRIMMELMANN: Thank you, Chris. It’s been a real pleasure.
KENNEALLY: Beyond the Book is produced by Copyright Clearance Center, a global rights broker for the world’s most sought-after materials, including millions of books and e-books, journals, newspapers, magazines and blogs, as well as images, movies and television shows. You can follow Beyond the Book on Twitter, find us on Facebook, and subscribe to the free podcast series on iTunes or at our website, BeyondtheBook.com. Our engineer and co-producer is Jeremy Brieske of Burst Marketing. My name is Christopher Kenneally. From all of us at Copyright Clearance Center, thanks for listening to Beyond the Book.