Yesterday the three plaintiff publishers in the copyright infringement case against Georgia State filed their proposed injunction, as the Judge required that they do, and a memorandum of law in support of that proposal.  So now we have a chance to examine their first legal response (as opposed to press releases) to the ruling Judge Evans handed down three weeks ago.

First a couple of comments about the memorandum of law.  There is a statement in it which asserts that the May 11 order found enough infringement to justify imposing the original injunction the publishers suggested back before the trial.  This is wishful thinking indeed, considering how many points the Judge has found in favor of GSU since that sweeping proposed injunction was offered.  And remember, the Court found only 5 instances of infringement out of the 75 excerpts examined or 5 out of 95 if we count the ones consider at trial and subsequently withdrawn by the plaintiffs.   But this hardly seems like a serious hope on the part of the three publishers, and the rest of the memorandum is spent explaining and defending the alternate proposal they offer, which is, they say, more narrowly tailor to track the Court’s fair use analysis.

One other comment in the memorandum caught my eye — the assertion that the GSU policy “did nothing” to limit GSU’s copying for electronic reserves to “decidedly small” excerpts.  This is technically a true statement but seems to misrepresent the Judge’s exact words and their implication (in fairness, the plaintiffs do quote those exact words later on).  What the Judge did say was the the GSU policy did cause the infringement “in that it did not limit copying in those instances [the five infringements] to decidedly small excerpts as required by this order.” She does not say that the policy did nothing to limit the copying, only that it did not conform to an arbitrary standard which was not available to GSU when the policy was implemented.  Hardly a culpable lapse, and the Judge does go on to say that the policy was a good faith attempt to comply with the law.

The phrase “in those instances” may be very important as this case proceeds, because much will turn on whether or not the Judge thinks the instances of infringement that she found rise to the level of “ongoing and continuous” violation of federal law.  The publishers assert that it does, and use those words over and over.  They need to make this argument in order to get an injunction that broadly proscribes future behavior by GSU. The exception to sovereign immunity on which this case is founded allows a court to issue an injunction only to prevent such “ongoing and continuous” violation.  So it is possible that the Judge could decline to issue an injunction at all by ruling that the few infringements she found amongst the many excerpts she examined do not suggest a pattern of ongoing and continuous violation.  Or she could simply order small tweaks to the GSU policy.

In their (alternate) proposed injunction, the publishers ask for much more.  They want GSU to be enjoined from ever using any excerpts from Sage, Oxford or Cambridge works without permission unless a strict set of conditions is met.  Those conditions include meeting all of the fair use factors (it would no longer be a balancing test), as they have been defined by the Judge, although the injunction combines factors three and four into a single set of requirements.  The proposal says that, to be used by GSU without payment, any excerpt from these three publishers must be “strictly without charge for nonprofit educational purposes” AND be “narrowly tailored” to “fulfill a legitimate purpose in the course curriculum” AND not be the heart of the work AND not be more than 10% or a single chapter, whichever is less.  Alternatively, if a digital license is not available from either the CCC or directly from the publisher (they want GSU to pursue both options), the excerpt must still be sufficiently small “as not to cause actual or potential market harm to the work.”

These requirements are taken very closely from the analysis Judge Evans used, but they are shaped into a less flexible test than the Judge actually employed.  For example, in the one case where the use of the “heart” of a work led to a finding of infringement by Judge Evans, that excerpt also exceeded her standard in terms of length (it was two chapters and 12.5%).  Because the third factor did not favor fair use, the factors were tied, and the Judge looked at the issue of the centrality of the excerpt to break the tie.  Thus she never actually asserted, as the proposed injunction would have her do (it connects length and centrality with a conjunction, not a disjunction), that an excerpted could be infringing if it was the heart of the work even if it was sufficiently short to meet her basic third-prong test.

The proposed injunction goes on to impose additional requirements on GSU, which look very like what was suggested in the publishers’ first proposal.  GSU would have to rewrite its policy to conform with the publishers’ reading of Judge Evans’ ruling, and state explicitly that the fair use checklist had been superseded.  They would have to document extensively, including evidence of an inquiry about a license made to both the CCC and the publisher for every excerpt.  Faculty would be required to see the entire proposed order every time they tried to upload anything, and to be threatened with sanctions if they disobey it in any particular.  The Provost would have to certify compliance to the Court each semester for the next three years.  And perhaps most objectionably, GSU would be required to provide the publishers with access to its course management system so that they could verify compliance.

The proposed order is clearly intended to humiliate GSU and to make fair use as difficult as possible for them.  It reads to me like a party who actually won very little at the trial still trying to spike the ball in the other parties’ face.  I hope the Judge will see it as yet another attempt to overreach the evidence on the part of the publishers.

One thing I can say with confidence.  If the publishers are given the right to poke around in GSU’s course management system, faculty will be outraged, at GSU and elsewhere.  That outrage might lead some to look for “underground” methods for sharing academic readings, which would be unfortunate.  It might also lead to difficulties in finding faculty authors who will be willing to continue to provide free content and free labor to these three publishers.

GSU now has 15 days to respond to this proposal, and then the Judge will decide what she will order.  It is likely her order will not be exactly what either side wants, and that the case will then move to the Eleventh Circuit Court of Appeals.

Whenever Jonathan Band writes a “Friend of the Court” brief on behalf of the Library Copyright Alliance, it is sure to be worth careful reading.  One not only learns a lot about the particular case from Jonathan’s filings, but also a good deal about the legal and social place of libraries the U.S.

Jonathan’s most recent brief amici curia was filed ten days ago in the case brought by the Authors Guild against the Hathi Trust and five of its university partners. In some ways it is an unlikely case in which to seek any enlightenment, since the posture and the legal theories advanced by the plaintiffs are odd, to say the least.  While it is hard to see this complaint going very far, the consequences if it did, and especially if the recent motion for partial summary judgment filed by the Authors Guild garnered any credence from a court, would be catastrophic for libraries.  Fortunately, Jonathan’s brief in response to that motion is smart and, I think, devastating.  And, as usual, it tells libraries some important things about themselves.

The motion, which I discussed here several weeks ago, argues that the only exception that libraries can rely on in the Copyright Act is section 108, the specific exception that authorizes some preservation and interlibrary loan activities.  It explicitly claims that fair use is unavailable to libraries, whose rights, it asserts, are entirely circumscribed by section 108.

In short, the AG would transform a safe harbor included in the copyright law promote certain library services into an anchor that would restrain libraries from performing many of their day-to-day activities.  Or, as Jonathan puts it, “They [the Authors Guild] seek to transform an exception intended to benefit libraries into a regulation that restricts libraries.”

Jon goes on to list many of the library activities that the public depends upon that would be of doubtful legality if the Authors Guild’s argument was taken seriously, ranging from ordinary, daily lending of materials to digital exhibits.  One of his most effective arguments is based on the many portions of the Library of Congress’ American Memory project that explicitly rely on fair use.  As the brief says, under the plaintiffs’ theory, the Library of Congress, in which the Copyright Office itself resides, would be “a serial copyright infringer.”

The absurd results of this radical and insupportable theory advanced by the Authors Guild are balanced by arguments both that the plain language of the law (and its legislative history) support the obvious proposition that fair use is available for libraries, and that, in fact, section 108 would permit the orphan works project that Hathi proposed and the Authors Guild seeks to prevent.

So the brief effectively counters the bizarre theory advanced by the AG.  But it also implicitly tells us two things about where libraries stand today that are worth noting.

First, it reminds us that libraries are always adapting to the changing needs of their patrons, many of which today are driven by rapid advancements in technology.  The ways people encounter culture shift with alarming regularity and libraries must stay abreast of these shifts.  Fair use, which has existed in U.S. law for over 170 years, has always been a key part of libraries’ ability to respond to patron needs, and Congress recognized the continuing need for libraries to be able to rely on fair use when it drafted the 1976 Copyright Act.  Section 108 is important for libraries, and it still has a role to play in library services.  But fair use is, perhaps, more important in an era of rapid change.  As Jon writes (quoting a 1990 case):

While the specific exceptions provide courts with no discretion, fair use is “’an equitable
rule of reason’ which permits courts to avoid rigid application of the copyright statute when, on occasion, it would stifle the very creativity which the law is designed to foster.”

The second, more troubling, reminder from this case and Jon’s filing is that the Authors Guild has shown itself willing to launch an extremely broad and devastating attack on libraries in order to protect some strange fantasy about how they can make more money.  Libraries have always been, and should remain, the best ally of authors who seek to find readers. It is foolish and short-sighted of the Author’s Guild to turn on libraries, and to advance a theory that would cripple them, without apparently realizing how much harm those actions could do to authors.

In a previous debate about this case, a commentator wrote “I don’t care about readers, I want buyers.”  It seems this attitude is assumed by the leaders of the Authors Guild, but it is disturbing for three reasons.

First, this attitude neglects the fact that library readers often become buyers.  Marketing their books is something authors expect from publishers, but if libraries are taken out of that equation, it will grow more difficult.  Especially in the world of online purchasing, the ability to discover and browse a book at a library is one of the best routes to placing an order for that book.

Second, not all authors (or even most, I daresay) share this attitude.  The majority of authors, including all scholarly authors, write, I hope, to be read.  Many are simply not motivated by the lure of profit, since profit is unavailable.  When profit is possible, it can be very important, but it does suppliant, in most authors I believe, the fundamental desire to communicate, to be read.

Finally, when the desire to make a profit does overcome the desire to express oneself and to be read, the result is inevitable an uninteresting book.  Those who do not think first about readers do not deserve any.

If the diverse members of the Authors Guild do not all share this assumption that readers and buyers are distinguishable, and only the latter are desirable, than they are seriously misrepresented by this filling, and should be grateful, along with the library community, to Jon Band for his work on behalf of readers everywhere.

On February 28, the Authors Guild filed a memorandum in support of its “motion for partial judgment on the pleadings” in its lawsuit against the Hathi Trust and five of its partner libraries, asking the judge to rule that the activities the AG has complained about – the mass digitization of books and the proposed orphan works project — are not protected by the specific library exceptions found in section 108 of the copyright law and that Hathi Trust cannot even assert, much less successfully rely upon, fair use, which is section 107 of that law.  There is a news report on the motion from Publishers Weekly here.

The memorandum strikes me as a masterpiece of misdirection, trying to make plausible arguments that do not quite fit the actual case in front of the judge.  The problem is that if the judge accepts these arguments, it could be devastating for libraries.  At its heart, the motion argues that libraries do not have any fair use rights, since their entire set of privileges under the copyright act are encompassed by section 108.  I think there are lots of reasons to reject this logic, which runs counter to the express language that Congress used in section 108 itself, which says (in subsection (f)(4)) that “Nothing in this section… in any way affects the right of fair use.

One way to see the flaw in the AG’s argument is to look at the odd results that arise if it is accepted.  For one thing, libraries would thereby become disadvantaged actors under the copyright act.  Other institutions and persons would still have the broad and flexible opportunities under fair use, but libraries would not.  Indeed, in the other lawsuit about mass digitization in which the Authors Guild is a plaintiff, against Google itself, Google will be able to argue fair use to justify its mass digitization, if the case gets that far.  But the plaintiffs argue that libraries cannot assert the same defense in regard to the same activity, simply because they are libraries, and thus disadvantaged by the existence of an exception that was supposed to benefit them.

By the way, this logic runs counter to the legislative history of the codification of fair use.  Congress explicitly stated that it did not intend to change or harden the judicial application of fair use; they wanted judges to continue to be free to apply the factors to specific circumstances to make equitable decisions.  But if placing section 107 in the same law as all the specific exceptions limits its application to situations not covered by those exceptions, that judicial freedom is undermined and the clear intent of Congress frustrated.

The are many other specific exceptions in the copyright law, and viewing them as limits on fair use again shows the absurdity of the argument.  There is an exception that allows photographers to take pictures of publicly visible architectural works, even when those works are protected by copyright.  If that exception is taken as the entire expression of the rights of photographers, then they could not argue fair use when taking photographs of other publicly visible copyrighted works, like a piece of public sculpture. That result is absurd, of course, and was implicitly rejected by cases allowing such photography.  There is also an exception that allows public performance of music in the context of religious worship, but its existence does not mean that someone who sings a song in public outside of a worship service would not be able to argue fair use.

It is interesting to note that a consultation on copyright in Ireland, which issued this consultation paper earlier in the year, suggests the value of a fair use provision for the Irish copyright law and makes the point that each of the specific exceptions can actually be seen as examples of fair use (see page 120).  If each specific exception is read as an instantiation of the fair use analysis for a particular situation, the logic of ensuring that fair use always remains an option (as 108(f)(4) tries to do) is particularly clear. (Hat tip to David Hansen for bringing this paper to my attention)

One more potential absurdity – If libraries have no fair use rights, would it automatically be infringement for a library to capture and print a single still image from a film for a student to include in a paper?  Section 108 excludes film from all but its preservation sections, so making a copy for a patron from a film would not be permitted under the 108 subsections on copying for users.  Yet this activity would seem like an obvious fair use if anyone else did it.  Why, we should ask, would libraries (and their users) be penalized simply for being libraries?

One of the difficulties here is that fair use is sometimes equated to “fair dealing” or “fair practice” in international copyright law.  Those terms tend to be blanket concepts that incorporate but do not expand upon the specific exceptions within each national law.  Fair use is different.  It is a separate and free standing exception within the US scheme. We need to remember that the US law was not written within the context of international copyright harmonization and does not conform, in any number of ways, to the usual pattern of copyright in other countries. That fair use is a separate exception and not simply a blanket term or gap filler is proved, I think, by the specific reference inside section 108 to fair use as an alternative option for libraries.  It is also a fact recognized and pointed out by the Irish paper on copyright linked above, which suggests that fair use can be seen as “a doctrine that defines the ambit of copyrightability and thus not an infringement at all.”  As a boundary definition on the exclusive rights and the basic analysis underlying all of the specific copyright exceptions, fair use would, again, always be an appropriate defense that courts should never rule out as a potential argument.

The most creative part of the memorandum supporting the plaintiff’s motion is its attempt to convince the judge to ignore section 108(f)(4), the fair use “savings clause” quoted above.  That language seems pretty categorical, and the phrase “right of fair use” certainly suggests that fair use is a positive right that does not simply fall away when more specific exceptions to copyright are enumerated.  The plaintiffs ask the judge to ignore this phrase with two arguments.

First, they refer to the general principle that “the specific governs the general.”  Because fair use is, allegedly, general, and section 108 is specific, it is asserted that 108 preempts the application of fair use in libraries.  But as we have just seen, fair use is a positive right that Congress acknowledged by inserting section 107 into the law but did not intend to limit.  So it is a different sort of thing; it is not merely a general principle that can be set aside by specific rules, but a distinct exception — or even a boundary definition — intended to do its own work within the framework of the law.

Second, the plaintiffs rely on a different case, Corley v. Universal Studios, in which a judge dismissed a similar fair use “savings clause” in the Digital Millennium Copyright Act.  The problem is that that dismissal, which has not been followed by other courts even when interpreting the DMCA, is based on an entirely different reason than the one being asserted by the Authors Guild.  The judge in Corley held that the activity in question, circumventing technological measures intended to prevent copying, was simply not the sort of thing that fair use was intended to apply to.  He did not reject the “savings clause,” he merely found that the activity before him was outside its scope.  In the Hathi case, of course, the activities in question are directly within the ambit of fair use, and the judge should respect the clear intention that Congress expressed as powerfully as it could in 108(f)(4).  Whatever the ultimate decision about fair use may be, the defendants must be allowed to argue it if the structure of US copyright law is not to be grossly undermined.