In my last posting, I posed four questions brought to my mind by the Aaron Swartz case. Here, I propose what I think are reasonable answers to those questions. The result is kind of a long post, but hey, it’s the weekend. Tell your spouse that the yard work will have to wait; you’re busy helping to solve the fundamental structural problems of the scholarly information marketplace.

Are the legal penalties provided for copyright infractions proportional to the actual damage caused by such infractions?

In general, I don’t know the answer to this question; I don’t know how many varieties of copyright infraction are identified in the law and I don’t know what all the penalties are. In this particular case, however, I do find it alarming that a U.S. attorney — presumably someone who is well versed in the philosophy and principles underlying American jurisprudence — would be willing publicly to assert that “stealing is stealing, whether you use a computer command or a crowbar, and whether you take documents, data or dollars.” This is hogwash. For one thing, as I pointed out in my previous post, “stealing” electronic documents is not at all the same thing as stealing dollars or cars or any other physical objects; it’s a matter of creating unauthorized copies, not of taking the documents away from someone else. What is taken away in an online piracy case is the exclusivity of control that comes with “ownership” of a document. Such theft can be serious and I believe it should be taken seriously, but it is simply not the same thing as stealing a car. Online piracy and maritime piracy don’t magically become morally equivalent just because we use the word “piracy” to describe both, and the same is true of the word “stealing.”

But there’s another significant difference between illegal online copying and actual theft, and that difference is violence. Most of us have probably noticed that a white-collar criminal who is convicted of embezzling $10,000 will sometimes get a lighter sentence than a working-class criminal who is caught robbing someone of $500 at knife-point. One contributing factor to this pattern is probably the various kinds of structural prejudice that persist in our society. But another is the line we draw, as a society, between violent crime and nonviolent crime. I think that line is reasonable, I think it matters, and I think it’s worth preserving. I think it is blurred dangerously when prosecutors deal publicly in the kind of overheated rhetoric that was employed by the U.S. attorney in announcing the Swartz indictment, and I worry about the precedent such language sets. If Swartz is convicted, it will be very interesting to see whether the prosecution seeks, and succeeds in getting, the kind of punishment that would normally be reserved for those who actually use crowbars to steal dollars.

That said, Swartz is actually accused of doing more than just illegally copying mass numbers of documents, many of them copyrighted. He also stands accused of the online equivalent of breaking and entering. This leads to the next question:

To what degree does it make sense to give network security itself the force of legal protection?

Swartz is alleged to have committed the equivalent of breaking and entering in two ways: first, by actually, physically breaking into a locked network closet at MIT; second, by virtually breaking into a secured network by hacking past its security protocols; the indictment refers to this as “unlawfully obtaining information from a protected computer.” This offense is separate from the offense of pirating copyrighted material; if the allegations are true, then the crime of copyright infringement came after the crime of gaining illegal access to the network. (And in fact, although copyright infringement is part of what Swartz is alleged to have done, it’s not even included in the charges brought against him.)

I think the answer to this question is yes: it makes a great deal of sense for network security to be given the force of legal protection, for much the same reason that it makes sense to give locks on the entrances to privately-owned spaces such as homes and offices legal protection. Suppose that I’ve filled my home with copies of documents that are in the public domain—government reports, phonebooks, 19^th-century monographs, etc. Now suppose that Aaron Swartz wants to have access to those documents; in fact, suppose that he feels passionately that everyone in the world ought to have free access to those documents. He is arguably right. Would the fact that those documents are in the public domain and should be freely available to all give Swartz the right to pick the lock on my front door, enter my home without permission, and make copies of them? Clearly not. The problem with doing so isn’t that he proposes to make illegal use of the documents; his proposed use is perfectly legal. The problem is that I have a legal right to secure access to my home, even if my home contains copies of documents that are in the public domain. The law recognizes a societal value in allowing people to secure their private spaces. Even if my private space contained objects that I actually had no legal right to have, the law makes provision for their removal by legal authorities following due process—it does not allow another citizen simply to break into my home and take those things away.

A data network is not the same thing as a private home. But my analogy doesn’t depend on a network and a home being the same thing; it depends on the logic between the two scenarios being parallel. To me, the fundamental conundrum is this: can a coherent legal system permit me to lock my door (or secure my network) while simultaneously allowing someone else to pick the lock (or hack into the network)? I don’t see how it can. I’m not sure that the legal remedies for network hacking ought to be exactly the same as those for breaking and entering, but it seems clear to me that the law should protect network security (where it has been legally and legitimately applied) and punish those who hack past it.

Does the current system of scholarly publishing currently exist to provide access to content, or to restrict it?

The formulation of this question paraphrases a comment by a Swartz supporter. I should probably have rephrased it in a more responsible way instead of simply repeating it in its original, rather tendentious form; my intent was not to endorse the sentiment but to examine it. Commenter (and fellow Chef) David Wojick caught the absurdity of the question immediately, and responded in what I think is the only realistic way: by pointing out that a system of scholarly communication will inevitably both provide access and restrict access. Since the services provided by a scholarly communication system are costly, those services will only be provided at some cost, and costs always put a limit on access. The question isn’t whether the costs will have an impact on access, but who will absorb the costs and whose access will be affected. Another question is: to what degree should the price of those services be set by a free-market system (and therefore provide for profit-taking)? This brings us to the final, and most fraught question that I believe the Swartz case should prompt us to face:

Should the very concept of intellectual property be rethought—at least in the realm of academic scholarship?

I argued in my previous post that the copyright regime on most university campuses is strange: the university hires people for the specific purpose of creating intellectual content, but then asserts no ownership over the content it has hired them to create. The fact that public funding is used to support some or all of the research that results in these intellectual products makes the arrangement stranger still. I’m not saying the arrangement is bad; on the contrary, I think it has generally worked out quite well for all concerned, except for one enormous problem: leaving copyright in the hands of academic authors has made possible the emergence of a massive commercial marketplace in scholarly documents which (the free market being what it is) have turned out to be objectively so valuable that the authors’ colleagues can’t afford access to them.

I realize it sounds as if I’m suggesting that the free market in scholarly resources should be dismantled and replaced by a different system, one that opens up access by excluding the profit motive and therefore keeping prices lower. But actually, that’s not what I’m saying at all. On the contrary: I believe the answer to this final question is no: we should not rethink the concept (or the application) of intellectual property in the realm of academic scholarship. I think the current system, whereby authors are generally allowed to retain copyright in what are arguably works for hire, and to trade that copyright to commercial publishers is, though flawed, the least bad system available to us.

Let’s consider the alternatives. Or, rather, the alternative, because there’s really only one: some kind of command economy in scholarly documents. Currently, scholarly documents are traded by their authors in what amounts to a free market of copyright and prestige (not readership; readership is incidental to prestige). The system generally works this way: an author has content to offer, and looks for the publisher that will confer the most prestige in return for rights to that content. Other factors may enter into the equation as well (number of readers, rights terms, etc.) but the primary exchange is one of rights for prestige: given the choice between ceding all rights and publishing in Nature on the one hand, and retaining copyright and publishing in Joe’s Biomedical Letters on the other, most tenure-seeking scholars will probably choose the former, because retaining copyright is much less important to their careers than being published in Nature.

Now suppose a university retained all copyright in the documents created by its faculty members, and therefore reserved the right to decide where and how those documents will be published. Those faculty members no longer have the right to decide with whom they will trade content for prestige. If the university deems Nature‘s copyright-transfer terms to be unacceptable, the faculty member can be forbidden from publishing there and can be forced to trade her content for less (or theoretically more, but let’s be honest here: it would be less) prestige from a less-demanding publisher.

Such a system might well have certain beneficial impacts on the system as a whole. The current market-based system has beneficial effects as well. Like the current system, the command system would also have deleterious effects. One such effect: the system would have to be national, even international, and monolithic. Unless all universities agreed to impose the same restriction on their faculty authors (and without exception), the varying amount of restriction from institution to institution would become a competitive tool used to attract high-profile faculty: “You don’t really want to stay at Podunk College, do you?”, the recruiters would coo. “Come to Glamorous University, where we’ll not only pay you more, but we’ll also let you retain copyright in that book you’ve been working on for the last five years.”

In fact, I’ll go further: I think the only way such a system could work would be if it were mandated and administered by the federal government. There would have to be sanctions imposed on colleges that tried to sidestep the requirement that they retain institutional control of faculty members’ copyrights. I realize that by invoking the specter of federal involvement I may be starting to sound like the academic equivalent of a Tea Party activist. But I’m not invoking the government as some kind of boogieman; I’m simply pointing out that universities are unlikely to police themselves effectively when there are potentially great benefits to be realized by policing themselves poorly. Some entity outside of the higher education establishment would almost certainly have to act as referee. And when the establishment we’re talking about consists of thousands of institutions, public and private, distributed across all 50 states, it’s hard to see what that entity could be if not the federal government. (In fact, since the scholarly economy is international, maybe it would have to be the United Nations. Now you can accuse me of invoking a boogieman: cue G. Gordon Liddy screaming on the radio about jackbooted thugs in blue helmets editing the People’s International Journal of Molecular Biology.) One response to this argument might be: come on, college athletics rules are enforced by the NCAA, and that’s not the federal government; why couldn’t there be an academic equivalent of the NCAA? My response would be: read that sentence again and tell me what there is about it that sounds better than the current system.

I’ve thought about this a lot, and in my view, there really is no reasonable option available to us that isn’t built on a foundation of free-market dealings in scholarly documents the copyrights to which are retained by authors and traded to publishers as those authors see fit. I’m not saying that this system doesn’t result in serious problems; on the contrary, there really is a pricing crisis in the serials marketplace, and that crisis is only possible because copyright holders have real monopolies and can therefore set very high prices for documents that are in high demand. Stopgap measures such as the Big Deal have masked the crisis temporarily, but have done nothing to solve it, and the fiscal disasters of the last few years have brought the crisis to a head. At my institution, we’re planning a $300,000 cut to our journal and database subscription portfolio in the coming year, with every expectation that we’ll have to do so again the following year. Why? Because our materials budget is flat, price increases for science journals and databases run about 9% per year, and $300,000 represents about 9% of our materials budget. Res ipse loquitur; the math is very simple. We can argue all we want about whether or not those journals and databases are worth what they cost, but those arguments ultimately mean nothing. Value is subjective; budgets are objective.

But what I’m describing is a pricing problem, not a structural problem. If scholarly journals all cost $15 per year and the price increased at a rate of .05% annually, no one would be saying that the system is broken. The fact is that our current system, though far from perfect, may well be better than any of the alternatives. And if that’s the case, then the Aaron Swartzes of the world are not only too extreme in their actions—they’re fundamentally wrong in their analysis of the problem, and their proposed solutions are fundamentally wrongheaded.