In an order filed last week, Judge Orinda Evans clarified her recent written opinion in the Georgia State University e-reserves case, holding that the publisher plaintiffs in fact prevailed on just four counts of alleged infringement, rather than the seven originally reported.

In her March 31 remand decision in Cambridge University Press v. Patton (known as the GSU e-reserves case) Evans for a second time declared GSU the prevailing party, writing in her conclusion that 41 of 48 alleged infringements considered at trial (and reconsidered on remand) were protected by fair use. But shortly after the decision, GSU attorneys filed a motion for a clarification, noting that in the body of her decision Evans appeared to have found 44 of 48 uses were fair, rather than 41.

Indeed, a number of observers had wondered about the discrepancy: was there a mistake in Evans's conclusion? Or, a mistake in the body of the opinion?

In her brief, two-page order, issued April 14, Evans acknowledged that she had miscounted in her conclusion. In fact, there were just three infringements from The Sage Handbook of Qualitative Research (rather than the five she had originally tallied); one infringement stemming from the use of Sage’s Utilization-Focused Evaluation; And one excerpt listed as an infringement in the conclusion of the March 31 decision, from Oxford University Press’s edition of C. Wright Mills' 1956 classic The Power Elite, was in fact a fair use.

Notably, Evans did not explain the error in her order last week.

First filed in April, 2008, by three academic publishers (Oxford University Press, Cambridge University Press and Sage Publications, with support from the Copyright Clearance Center and the AAP) the suit alleges that GSU administrators systematically encourage faculty to offer unlicensed digitized copies of course readings to students as a no-cost alternative to traditionally licensed course-packs.

In 2012, Evans first ruled against the publishers. But in October, 2014, the Court of Appeals for the Eleventh Circuit reversed and sent the case back to Evans with instructions for her to re-balance her four-factor fair use analysis.

Now, in yet another twist, despite winning a reversal on appeal, the publishers appear to have actually come out worse. The four infringements Evans found on remand are actually less than the five infringements Evans found in the original decision, in 2012.

In a statement, the Association of American Publishers said they are once again considering their appeal options.