u_topn picture
Atlantic Unbound Sidebar

Life, Liberty, and ... the Pursuit of Copyright?
Round Three: Concluding Remarks

Lawrence Lessig

John Perry Barlow fears my resort to law as a way to permit creators to reap a harvest from the information they make. I haven't made myself clear: My resort to law is not to protect creators (or more likely, publishers). Creators will have all the protection that they need -- if not from the copyright code that Congress passes, then from the software code that Mark Stefik and others will write. My use of law is to limit the protection that creators have. The argument is that law should limit the power that code like Mark's gives creators. For if the world goes as it may well go, the protections of this code will be far greater than the protections of present copyright law. And my point was that they shouldn't. Copyright law has always been about striking a balance between protecting the rights of creators and protecting the access of readers. If technology changes in favor of creators, then law should intervene to tilt it back.

From Post & Riposte:

"Most people are breakers of copyright laws on a regular basis. What right have you to go into a bookshop, open a book and read the ideas of an author whose works you have not purchased?... The point is there is a well-recognized fact that if you want people to buy your intellectual efforts you have to give them time (and a place) to evaluate them. The same rule should apply on the Internet. You should have time to browse through an electronic product before buying it, without prejudice. Technology control that restricts the copying or printing of data is one thing, but technology control that restricts access to data is another. Such controls mean that the rich will get richer while the poor get poorer."
--Martin Bryan, 9/21/98

"Missing from Mann's otherwise thorough examination of copyright was the role of copyright collectives in copyright administration. Copyright collectives in fact play the moderating or intermediary role Mann seeks, falling somewhere between no copyright and impenetrable copyright boxes. Copyright collectives manage to collect royalties, individually miniscule to the individual consumer, that add up to significant revenues for publisher and author. At the same time, bulk or blanket licensing does not interfere with the open flow of information between author and reader. Is it perhaps telling that copyright collectives are so poorly developed in the U.S. that they go without mention in Mann's essay?"
--Robert Labossiere, 9/22/98

Via E-mail:

"The discussion on copyright brings to mind the question of enforcement, in regards to information on the Internet. If there are to be any new laws or policies, who is going to enforce them? Should there be an international coalition set up, since the medium involved is not confined to national boundaries? It may come down to copyright, as it is now, being applicable in only those countries that recognize these laws. Countries, like the present-day China, which do not enforce copyright laws will be looked at as 'lost causes.'"
--Craig Murray, 9/24/98

What Do You Think?
Join the debate in the "Life, Liberty . . . Copyright?" forum of Post & Riposte.

There is nothing especially new in this. Creators have always pointed to changes in technology to argue for increasing the protections of law. They pointed, for example, to the Internet itself as reason for Congress to strengthen the law's protection. But if Mark is correct, then technology is about to change once again -- this time, to the advantage of creators, and against fair use. If it does, and if the effective scope of fair use is thereby constrained, then law should do something about it. It should not simply stand by as the balance that the Constitution calls for gets shifted by changes in technology. Here, as in many contexts -- privacy, free speech, access, anonymity -- public values should not be replaced by code.

Mark Stefik doesn't think that the emergence of code means the disappearance of law. Governance, he assures us, will not be turned over to the 'code.' Instead, trusted systems simply represent a partial shift -- from copyright law to contract law -- in the legal basis governing creative works. Shrinkwrap and clickwrap licenses, along with digital (machine-enforced) contracts, will become the basis of the enforceable obligations that constitute cyberspace. And these, Mark admits, can only survive if we develop social and legal institutions as part of the cyberspace infrastructure.

But this misses something important. Shrinkwrap or clickwrap or digital contracts may well be, as Mark says, like a contract (because they enable agreements between creators and users, and provide a mechanism through which those agreements get enforced, one might liken them to the sorts of things we lawyers call contracts). But it is a large step from saying it is like a contract to saying it is a contract -- or, more directly, a large step from saying that since this is like a contract we should consider it to be as benign as if it were actually a contract. A tiger is like a kitten, but one wouldn't bring a tiger home to play with the kids.

The difference is in the mechanism of enforcement, and the conditions under which agreement is recognized. Contracts are those things that envision ultimate enforcement by a court, when the agreement is engaged consistent with substantive rules of formation. They are voluntary agreements, but they get their power only after review by the state. We tend to forget this essential role that the state plays in contract law; we tend to speak as if a contract were just between two parties. But that is simply a mistake. We have never had a law of contracts that simply enforced whatever the parties desired; we have never had a law of contract that let parties structure their agreement however they wished. The law has always (more at some times then at others) supervised.

But supervision is what's hard to see in this emerging architecture of machine-enforced agreements. Not only will the machine permit an increasingly articulated set of norms to be built into the code of the agreement, but the universality of the medium will make it increasingly difficult to effect any review of these norms. A state may have a set of rules about how contracts are to be entered into; those rules could well be displaced by the machine-enforced structures of a clickwrap agreement; yet it will be increasingly difficult to check these machine-enforced structures against the norms imposed through law. My sense is that they will displace law, not reflect it, and that the values that are displaced are values that are significant.

This is not to say that all agreements or exchanges or deals in real space are under the supervision of the state. When I drop 75 cents into a soda machine, the machine enforces the exchange without any appeal to the state. And likewise, when someone gives a bond as collateral for a deal, if the deal falls through, then holding the collateral serves in a sense as an automatic structure for enforcement. But even these agreements have more potential for review than the agreements implied in Mark's code. For these agreements announce themselves as agreements; conditions imposed by the code need not.

My first concern, then, is with the displacement of important contract values that digital contracts will effect. Digital contracts become a kind of privatized contract; they become a privatized system for enforcing agreements. Some might not think that's so bad. Some might prefer it to the state-sanctioned model. But we should at least be clear about the differences.

A second concern connects with a question Charles Mann presses. Charles asks whether contract law can supersede copyright law since copyright law is federal, and contract law is not, and federal law trumps all inconsistent law that is not federal.

A simple answer, though misleading, is no: contract law cannot supersede copyright law. But the problem is in determining what types of contracts would supersede copyright law. As Judge Easterbrook explained in ProCD v. Zeidenberg, copyright law defines the rights of the copyright holder vis-à-vis the world; contract law defines the rights of a copyright holder vis-à-vis a particular person. Copyright law has always permitted individuals to strike deals (over at least some copyright rights) that are different from the deals the law strikes with respect to every other person. For example, the first-sale doctrine limits the rights of the copyright holder over material that has been sold. Once I buy a book, for example, I am free to read it once or a hundred times; I am free to burn it or seal it in wax. The copyright holder cannot simply assert that, as a matter of his copyright rights, I am not free to burn his book.

But the copyright holder could enter into a contract with me, whereby I promise in exchange for his selling me his book not to burn it. That would be an example where the rights of the copyright holder were extended by contract law, but not an example where contract law would supersede copyright law.

As a formal matter, then, there's no inconsistency between the rights a copyright holder has under the Copyright Act and other rights she may be able to secure through contract.

But here, as in many places, formalisms mislead. For if one believes that copyright law has always been about striking a balance between access and control -- assuring, that is, a sufficient degree of control to the creator yet adequate access through fair use to the public -- then one must worry about worlds where contracts become so plentiful as to effectively erode the balance that copyright law seeks. It might be that in a world where it is relatively difficult to enforce contracts with book buyers, and hence relatively unlikely that many book sellers would insist upon them, permitting contract to extend the rights of copyright is untroubling. But if the cost of contracting falls dramatically, and the cost of enforcement falls as well, then the net increase in contracting may well effectively displace values implicit in the law.

My claim is not that this will happen necessarily -- there is little way in theory to argue that an empirical balance will become skewed. My claim is just that we should reject the simple formalism that says since digital contracts are like contracts, and since contracts have always been permitted within at least part of the copyright domain, then digital contracts enforced through software mechanisms can't be thought to be inconsistent with copyright law. It might turn out that that is true. But it is certainly not true necessarily. Here, as in many other areas of cyberspace law, we must watch, and wait. And remain vigilant against those who would sell tigers as kittens.

What do you think?

Join the debate in the "Life, Liberty . . . Copyright?" forum of Post & Riposte.

Roundtable Overview

Introduction by Charles C. Mann

Round One: Opening Remarks -- posted on September 10, 1998

Round Two: Responses -- posted on September 17, 1998

Round Three: Concluding Remarks -- posted on September 29, 1998

Lawrence Lessig is the Berkman Professor of Law at Harvard University. He teaches and writes in the areas of constitutional law, contracts, and the law of cyberspace, and is presently completing a book, Code, and Other Laws of Cyberspace.

Copyright © 1998 by The Atlantic Monthly Company. All rights reserved.
Cover Atlantic Unbound The Atlantic Monthly Post & Riposte Atlantic Store Search