Now that I am back from vacation and have read the GSU ruling and some of the commentary more thoroughly, I wanted to add a few additional comments. In many cases these may be repetitive of things others have already said; I have not read all the commentary, of course, and some of this writing is for my own sake as much as yours. If it is redundant, I apologize.
I do hope most folks either have read or will read Brandon Butler’s “Issue Brief” for the Association of Research Libraries, which does a great job of analyzing the ruling.
The main thing that struck me as I read the fair use discussions for the individual excerpts that were in question in the case is that Judge Evans really did understand that fair use is a balancing test. Although some parts of her reasoning are more mechanical than I would like, she does not take a bright-line view of any of the factors. It seems worth a look, to me, at the way she relates the four factors in her discussion.
First, she finds that the first two factors — the purpose and character of the use and the nature of the original — always favor fair use in regard to the specific use and the specific works before her. This is true even though she declines to hold that the provision of short course readings is transformative, which is the key determinant in most fair use rulings over the past few decades. Instead, she sees this activity as at the heart of what fair use is intended to be, according to its own express terms. On the nature of the original, she holds that the works in question were all published and factual in nature.
This point about the second factor is likely to be the focus of some disagreement from the publishers. They wanted the court to hold that factor two did not favor fair use, apparently because of the effort involved in writing and publishing academic works. No one is more respectful of the labor of scholarship than I am, but hard work by itself does not get one a copyright or insulate one’s copyright from fair use. The Judge observed that educational works are exactly the works that fair use is focused on, with its favored purposes of research, scholarship and teaching. She also noted that permission fees make up no part of the incentive for academic authors to write their works (p. 81ff). This, to my mind, is the key reason for finding that factor two favors fair use in this case (although the Judge called it an “additional consideration”).
On the other side of the analysis, Judge Evans presumed, at least as an initial matter, that factor four, the impact of the use on the market for or value of the original, always favored the plaintiffs, which is to say it counted against fair use.
Which brings us to the third factor, the amount used. This was the “swing vote” in most of the individual analyses, and the Judge applied a clear but quite narrow rule. Ten percent of a work was acceptable as fair use if the book had fewer than 10 chapters, and a single chapter was considered fair use for books that contained 10 or more chapters. This rule, which was applied quite strictly to decide if a “decidedly small” portion of the work had been used (that was the standard Judge Evans applied for this particular type of use), had some odd results; as little as 3% of a work was considered too much for fair use in a situation where the work was long and had many chapters.
But notice something important. When the third factor disfavored fair use, the factor analysis stood at two in favor (one and two) and two against (three and four). In those cases the Judge went on to do additional analysis, asking several further questions. Was the heart of the work used? (she declined to assume that all e-reserve uses involved the heart of a work) Was there a “readily available” license for reproduction of the excerpt in digital form? Were licensing fees an important part of the value of the original work in question? (Judge Evans held that they usually were not, so this was an important question).
The results of this additional analysis were what finally determined whether or not a use was fair if the third factor threshold the Judge had set was exceeded. When the amount used was within her guideline of 10% or one chapter, the availability of a license for digital distribution did not sway the analysis against fair use. And when a license for a digital excerpt was not available, the amount used was less important (although the leeway here is small, for Judge Evans).
If this rule about amount seems unduly restrictive, libraries would do well to remember three other findings from the ruling that impact how we might view this guideline. First, she held that percentages should be calculated based on the total number of pages in the book, which makes for a much easier and quicker determination of amount than the method of counting the plaintiffs sought. Second, she held that individual chapters by different authors in an edited volume should by counted by the same method and not as if each were an individual work, which also simplified the process of decision-making. Third, she rejected the idea that a use that was fair use in one semester must be paid for in a subsequent semester.
All three publisher plaintiffs, along with the two organizations (the Copyright Clearance Center and the Association of American Publishers) that bankrolled them, have now issued initial responses to the verdict. Overall their objections are quite vague. Several of the responses refer to legal or factual “errors” in the ruling, but they do not specify what they are. The AAP does disagree with the Judge’s finding that the loss of permission revenues because of fair use imperils their business; they repeat this absurd claim even after the Judge pointed out, based on figures supplied by the plaintiffs themselves, that these publishers made less than one quarter of one percent of their 2009 revenues from academic book and journal permissions.
All of the publishers assert that there is error in the judge failing to consider what the AAP calls the “full context” of the activities at GSU. Sage and Oxford (who issued identical statements) suggest a “pattern and practice” of infringement , while Cambridge refers to “systematic and industrial-scale unauthorized reproduction.” It is hard to know what to make of these assertions, other than that they arise out of sheer frustration. Since the Judge has just found that only five of the excerpts before her were infringing, “systematic” and “pattern” seem like inappropriate words. The Judge had to decide the case based on the specific allegations and evidence before her, and the plaintiffs were the ones who produced those allegations. So if only five out of 99 (or 75, depending on where you start counting) were infringing, no pattern of systematic infringement has been proved.
The Cambridge statement has the strongest hint of an appeal, saying that they look forward to working with their “partners” to “resolve this issue.” It also has the chutzpah to evoke “our authors” as victims of the ruling. How many times can we repeat what publishers hate to admit in public, that the vast majority of academic authors are not paid for their scholarly writings (by publishers, at least) and do not consider permission fees when deciding whether, what or where to publish?
So where do we go from here? We need to remember that this is only a District Court ruling, which is not binding on any university other than GSU, that no remedy has yet been ordered, and that it is likely to be appealed. Nevertheless, it is a carefully reasoned ruling with lots of specifics for libraries and universities to consider. Some institutions may decide that the Judge has defined a safe harbor and that they want to anchor there; in that case it is worth noting that some of the analysis she applies would be difficult to replicate without detailed financial information from publishers. More likely, schools will look at this ruling and tinker with their own policies a bit. Given both its substance and the still-early stage of the proceedings (really, even after 4 years and 350 pages this is not the final word), this decision is not, in my opinion anyway, a cause for large-sale or precipitous changes at most institutions, a fact that even the publisher statements seem to admit.
Policy on Electronic Course Content
For help deciding whether course content in Blackboard or some other digital form is fair use or requires copyright permission, consult this policy document adopted by the Academic Council in February 2008.
Search the Scholarly Communications Blog
- Authors' Rights
- Copyright in the Classroom
- Copyright Information Notes
- Copyright Issues and Legislation
- Digital Rights Management
- Fair Use
- international IP
- Open Access and Institutional Repositories
- Open Access topics
- Orphan works
- Public Domain
- Scholarly Publishing
- Traditional Knowledge
- User Generated Content
- Dave Fernig on Going all in on GSU
- Gretchen McCord on Going all in on GSU
- In Georgia State University E-Reserves Case, Eleventh Circuit Endorses Flexible Approach to Fair Use | ARL Policy Notes on GSU appeal ruling — the more I read, the better it seems
- Paul Callister on Swimming in muddy waters
- Jim Neal on Free speech, fair use, and affirmative defenses