Saturday, October 31, 2009
Star-Brite succeeded in enjoining certain claims Kop-Coat aka Valvtect made about its gasoline additive. Valvtect then issued a press release titled “Court Finds Valvtect’s Comparison Ad Was Not False.” It didn’t mention that the court found the ad misleading and enjoined it, but noted that the court hadn’t ordered the ads recalled and included the statement “nor did the court order prevent Valvtect from placing future ads comparing the performance of Valvtect Ethanol Gasoline Treatment and StarTron.” Star-Brite then issued its own press release, “Starbrite Granted Injunction Against Valvtect.” It variously described the court as finding that the ads were “false or misleading,” “false and misleading,” and “false and deceptive,” which Valvtect contended was itself misleading because of the use of “false,” rather than only “misleading.”
Valvtect moved to increase Star-Brite’s bond based on its allegedly misleading press release. Since that press release was issued in response to Valvtect’s own misleading press release, the court was unimpressed. “While the parties have some freedom to ‘spin’ in the realm of public opinion what court orders mean to the marketplace, neither side is free to mislead the public about the issuance and substance of an injunction.” As for whether Star-Brite’s press release crossed the line, Star-Brite argued that an ad that is not literally false can still be “false” if it’s misleading. At least in the context of this dispute, the court refused to sanction Star-Brite’s characterization. Valvtect’s own press release minimized the actual language of the injunction so significantly as to be misleading; Star-Brite was setting the record straight.
Valvtect also placed an ad referring to some of the same test results the court found misleading but only naming “the Competition.” The court refused to hold Valvtect in contempt because it wasn’t a comparison ad naming Star-Brite by name, citing cases that irreparable harm is presumed when an ad compares two products. I don’t really get this—an ad can be comparative without naming the competition, and there are plenty of cases so holding; otherwise the law would be too easy to evade. In the context of holding a defendant in contempt, perhaps a rule of lenity is more justified. The parties here, however, seem likely to be repeat players before the court.
Friday, October 30, 2009
As entertaining as a Posner opinion always is, I fear this one has more than enough casual language in it to do mischief to the law; it will likely be much cited, whether for heat or for light. Discussion of opinion below, dismissing Schering’s false advertising claims. Schering sells an OTC polyethelyne glycol 3350 laxative, MiraLax. Defendants sell prescription polyethelyne glycol 3350, which has created a problem under FDA rules. I will let Posner explain:
MiraLAX was originally a prescription drug too. After the patent on it expired, the Food and Drug Administration approved defendants' ANDAs (Abbreviated New Drug Applications), which authorized them to sell a generic version of the drug. Later the FDA approved an over-the-counter version of MiraLAX but required that the label contain a warning to "use [for] no more than 7 days." Constipation that lasts longer than that may be symptomatic of a serious medical condition, so a consumer who wants to use a laxative longer should do so under a doctor's supervision.
The Food, Drug, and Cosmetic Act requires that the labeling of the generic drug be the same (with immaterial exceptions) as that of the original drug-- the "pioneer" drug, as it is called, which in this case was the prescription-only version of MiraLAX. And if the generic drug is approved for use as a prescription drug, the label of the generic drug must "bear, at a minimum, the symbol 'Rx only.' " The labels of the defendants' generic versions of MiraLAX do bear that symbol.
Problem: defendants’ labels and package inserts differ from Schering’s in omitting the 7-day warning and in saying “Rx only.” On the one hand, those are required differences; on the other, the labels are required not to differ! Schering thus argues that the labels are false; they say that polyethylene glycol 3350 is sold only by prescription, but in fact there’s an OTC version. (E.g., insert language: “Polyethylene Glycol 3350, NF Powder for Oral Solution is a prescription only laxative which has been prescribed by your physician to treat constipation.”) “And should the patient’s condition change, so that he didn’t need to use a laxative for more than seven days, he might be unaware that he could switch to an over-the-counter version of the laxative that had been prescribed for him.”
The FDA is currently conducting a proceeding to determine whether defendants’ drugs are now misbranded. Because the FDCA doesn’t allow both Rx-only and OTC versions of the same drug to be sold simultaneously, the proceeding encompasses the issue of whether there’s a meaningful difference between the pioneer and the generic. Defendants argue (rather unconvincingly, it seems to me) that the drugs are different because their versions aren’t subject to the 7-day warning.
This is a matter for the FDA. The FDA may determine the generic can’t be sold any more if the drugs are the same, or it may decide that the label needs to change if they’re different. Posner thought that the FDA might or might not agree with Schering’s argument that the defendants’ labels “occlude, in the mind of the consumer, the existence of its over-the-counter version of the drug.” The FDA might just care that the labeling induces doctors, pharmacists, and consumers to use the drug safely. “It is not obvious that the goal of protecting consumers is furthered by making sure that they are aware of the existence of an over-the-counter equivalent, and, if it is not, there would be no conflict between the FD & C Act and the Lanham Act. But we do not know, and see no need to guess while the misbranding proceeding is wending its way through the FDA.”
Schering argued that all that was required was a disclaimer indicating that not all polyethylene glycol 3350 drugs need a prescription. But the court thought this was a potential Lanham Act-FDCA conflict, in which the FDCA would need to prevail. “A disclosure required to protect a competing seller might mislead a consumer, in which event the drug would be mislabeled and could not be sold, so that the seller's concern with unfair competition would be moot; it would have no competitor because there would be no competing product.” This would be inconsistent with the FDA’s permission to sell the drug, in the same way as a copyright lawsuit against a competitor for copying the label of a pioneer drug—as the FDA requires a generic to do—is inconsistent, and thus the non-FDCA law must give way.
But courts do try to give as much effect to both statutes as possible, so Schering argued that the FDCA shouldn’t be interpreted to forbid a misrepresentation-curing disclaimer. But Schering “has been coy” about what the disclaimer should say, and this coyness made the court doubt that the matter could be resolved without FDA involvement. The FDA might worry that a disclaimer might make consumers think that only one product is prescription only (when there are at least two generics that are), or would lead consumers to think the brands might be chemically different. The FDA has more experience with consumer understanding of drug labels and should get a chance to rule.
This case, Posner wrote, could be contrasted to others in which FDA involvement is not required, for example when a defendant says that the FDA has approved a drug for a number of uses that were in fact unapproved. Courts can evaluate such a claim without using the FDA’s insights. Likewise, if the defendants’ label had said that it was “the only polyethylene glycol 3350 that won't make hair grow on the palm of your hand, or that each container contains 727 grams of the drug, when in fact it contains only 527 grams, like its competitors’ containers,” no FDA involvement would be required. This case, however, was “subtler,” because it’s unclear how consumers understand “Rx only” labels, and how a disclaimer could be worded to improve that understanding.
Schering cited FDA regulations allowing a seller to make minor labeling changes without FDA approval, though it must notify the FDA. Moderate label changes require more FDA involvement, and major changes require preapproval. The court held that major changes are the default category, to preserve the strength of the regulatory system, and that Schering’s proposed change did not fall within the definition of “minor” or “moderate.” The FDA would probably want to require pre-approval of a possibly misleading disclaimer. And the issue was further complicated by the requirement that generic labels be identical to those of the pioneer.
As a result of all these concerns, Schering jumped the gun by suing before the FDA ruled on whether the generics are now misbranded.
It’s hard to say whether what comes next is dicta. But the Seventh Circuit also agreed that Schering should have lost its motion for partial summary judgment on falsity regardless of how the FDA rules.
Schering argued literal falsity, because polyethylene glycol 3350 is not Rx only. The court first expressed some uncertainty about, I think, whether the labels and inserts at issue are “advertising,” given that the only people who see them are people who’ve been prescribed the generic. The ad for one of defendant’s drugs on its website had “just the sort of disclaimers” Schering wanted.
Regardless, Posner concluded that Schering had been “misled” by general language in false advertising cases. “[U]ncritical generalization is a path to error. One form of uncritical generalization, ironically in view of Schering’s invocation of the doctrine of ‘literal falsity,’ is reading general language literally.”
Many literally false statements aren’t deceptive. Posner gave the example of the Soviet Union’s slogan “2 2 = 5,” which was designed to spur workers to complete the Five-Year Plan in four years, as well as the example of an ad touting “The Most Fabulous Jewels in the World.” The latter, he wrote, was literally false because the jewels were no more fabulous than the Crown Jewels or the Hope Diamond, but no one was deceived. As he noted, the doctrine used is puffery, “but the principle cuts deeper; if no one is or could be fooled, no one is or could be hurt.”
On the other hand, “a representation may be so obviously misleading that there is no need to gather evidence that anyone was confused.” And misleadingness is often clearer than literal falsity, because literalness is often a semantic question. So, what the cases mean when they say that proof of literal falsity allows the plaintiff to skip proving likely deception is that “the seller who places an indisputably false statement in his advertising or labeling probably did so for a malign purpose, namely to sell his product by lies, and if the statement is false probably at least some people were misled, and since it was a lie why waste time on costly consumer surveys?” “Literal” must therefore be understood in the “common colloquial sense.” “A ‘literal’ falsehood is bald-faced, egregious, undeniable, over the top.” We know this is what literal means because the case law includes the caution that the statements must be assessed in context.
So: a beautifully subtle importation of fault, which is a misdescription of traditional Lanham Act doctrine. Where did that “lie” in the previous paragraph come from? It’s no necessary part of the cost-saving rationale for allowing plaintiffs to dispense with consumer surveys. Perhaps if Posner had recalled the connection between false advertising and trademark law he would have been less seduced into the error of intent. Moreover, many literal falsehoods in the caselaw are far from “bald-faced, egregious, undeniable, [or] over the top.” They are simply, on the weight of the evidence, wrong. Consider the line of cases about “tests prove” claims. If the tests don’t prove the stated claim, because the plaintiff shows the tests to be unreliable or to be unrelated to the claim actually made, the defendant loses. It’s not a matter of obviousness or good faith, as Posner’s list of adjectives suggests. It’s a matter of evidence.
The funny thing is that Posner’s explanation is so much further-reaching and unsupported than his conclusion, which is much more in line with existing caselaw: a literally false statement is one “that means what it says to any linguistically competent person.” That’s a matter of interpretation, not a matter of speaker’s intent, and it leaves open the possibility that a false statement can be false—that is, not true—without any moral flaw in the speaker. Where Posner conflates the process of communicating meaning with the evidence supporting any particular statement, he unfortunately encourages misreading the scope of the Lanham Act.
Here, Posner offered a counterexample of a linguistically simple message: “suppose the labels on the defendants’ products stated: ‘All polyethylene glycol 3350, by whomever made, can be sold only by prescription; there is no over-the-counter version of this drug.’ That would be false and misleading per se; there would be no need to consider context or audience.” But, he noted, that’s not what the labels say. “There is no statement in the ordinary sense, because there is no verb.” (I doubt we’re meant to take this seriously as a reason why claims in general wouldn’t be literally false: consider “55 MPG”; linguistic competence includes the occasional ability to infer some additional words.) Instead, there’s “Rx only,” accompanied by some other information, including the name of the active ingredient. Obviously this product is prescription only, but it’s not obvious that all other products with the same active ingredient are prescription only.
Nonetheless, the court concluded that the district court was right not to dismiss the suit with prejudice. FDA findings might ultimately put the issue of consumer confusion in a different light.
Plaintiff RLI is “the world’s leading provider of computer-based assessment technology for pre-K-12 schools.” Its Accelerated Reader (AR) is a software program first introduced in 1986. Metiri offers educational consulting services. One of its products is Technology Solutions That Work (TSW), which offers a database of research studies on educational software.
RLI alleged that the TSW database contains false statements about AR and that Metiri has made other false statements about AR. The allegedly false statements include: (1) statements touting TSW’s quality and lack of bias; (2) statements that the evidence didn’t support claims that AR is effective, despite its popularity; (3) statements that most of the research on AR was done by RLI and that most was not of sufficient technical quality to be worth reviewing; and (4) statements that AR’s use of rewards is increasingly controversial.
RLI’s first claim was for commercial disparagement, which requires (1) that the person making a statement intend, recognize, or should recognize that the statement is likely to harm another’s pecuniary interest and (2) knowledge or reckless disregard of falsity. The court found genuine issues of material fact sufficient to deny Metiri's motion for summary judgment.
For example, comparing TSW’s evaluations of AR and Little Planet Early Learning Series permitted an inference that Metiri isn’t unbiased. (I have to admit discomfort with treating the term “unbiased” as falsifiable rather than opinion-based, at least when we’re out of the realm of quantitative studies.) The statements “We feel we are being generous with a rating of ‘Inconclusive Evidence’ with respect to Accelerated Reader” and “Despite its predominance in American schools, the research base on Accelerated Reader is weak, idiosyncratic, and contradictory. It seems odd that with tens of thousands of American schools using the program, that most oft-quoted study was conducted on two classrooms in a socio-economically disadvantaged area of Scotland” arguably was evidence of some bias. (Why? The statements are evaluative, but I don’t see why that shows bias. It’s not biased, I would have thought, to come to a conclusion after evaluation. But see next statement.) The court thought that Metiri’s highest recommendation for Read 180 was “generous” in comparison to its analysis of AR, because Read 180’s research was also pretty limited, as was the research behind the Little Planet Early Learning Series, another Metiri-recommended program. Moreover, one of the people Metiri assigned to review AR worked extensively in the place where Little Planet and Read 180 originated while those programs were being developed.
By contrast, in the Department of Education’s What Works Clearinghouse, only 24 of 153 beginning reading programs met the evidence standards; AR was one, receiving a rating of “Potentially positive effects: evidence of a positive effect with no overriding contrary evidence,” so the court could infer that there is scientifically based research to support AR. Moreover, peer reviewed, significant studies of AR were not included in Metiri’s analysis while studies that weren’t peer reviewed or weren’t significant were included, raising an inference that Metiri didn’t include the highest quality research. Worse, Metiri’s practice of updating TSW at least every two years, plus its failure to include a study suggested by RLI representatives (the study cited by What Works Clearinghouse) “raises an inference that TSW is not an up-to-date resource and that Metiri is not always on the look-out for high quality research to update its database.” (Okay, perhaps these claims should have been run through a puffery filter before assessing their truth?)
Finally, there was a factual dispute about what percentage of the “mountains of research” on AR was completed by RLI; a “good bit” are case studies in which data are collected by individual schools and then submitted to and summarized by RLI, allowing an inference that the individual schools rather than RLI “completed” this research. (Really? Just being picky, how could it be recklessly false to say that RLI’s summarization didn’t count as “completion”? And anyway, could anyone reasonably think that the schools that already adopted AR would have interests not aligned with RLI in reporting success?)
In fact, the court didn’t assess the evidence on state of mind—I guess it must have implicitly held that, given these facts in the light most favorable to RLI, it would have been reckless for Metiri to make the statements.
On the false advertising claims, Metiri argued that the allegedly false statements weren’t in ads and that RLI suffered no competitive injury. The court also found material issues of fact. As to ads: many of the statements at issue were in TSW videos or PowerPoints, available on Metiri’s site without a membership, allowing an inference that these statements are commercial ads; other statements were made at conferences where Metiri was promoting its services.
There was a factual dispute over whether Metiri and RLI are business competitors—not in the sale of educational software, but in the areas of policy consulting, professional development, and research and evaluation services. If Metiri discredited one of RLI’s major products, customers might question its other services. Moreover, there was a factual dispute over whether Metiri used ads to discredit AR to spark interest and sales in TSW and divert business that otherwise would have gone to RLI to a TSW subscription which recommends alternative products. (While that first rationale for competition makes sense to me, this second doesn’t—it’s just a restatement of the claim that Metiri cost RLI business, which might be commercial disparagement but is not competition. Compare: “Consumer Reports ran ads for its vacuum cleaner comparisons touting its surprising findings, diverting consumers from vacuum cleaner purchases towards CR, at which point the consumers changed their initial vacuum purchase intentions to buy the CR-recommended product.”)
RLI’s claims for tortious interference also survived. TSW says AR is currently in place in almost half of all American schools, allowing an inference that Metiri knew of RLI’s contracts/business expectancies with schools. (I thought the knowledge about the contract and the interference with same had to be specific to particular clients, though; otherwise any dominant market participant can easily win this tort.) RLI also provided an expert report claiming lost sales from AR when customers subscribed to Metiri’s services. Metiri ads such as “Click here for a special offer and finally settle that argument about Accelerated Reader” allowed an inference that Metiri was motivated to discredit AR to spark interest in TSW.
Anybody looking for evidence that business torts are used to suppress critics could, it seems to me, reasonably cite this case. A little First Amendment defamation medicine would not have been out of place here. It's not that Metiri's motives seem pure from the facts as alleged, but competition in the marketplace might be a better way to sort it out than litigation.
Tuesday, October 27, 2009
Wolfe Tory moved to preliminarily enjoin Sharn from using an allegedly confusingly similar trade dress in medical devices that deliver medicine to the throat and nose (mucosal atomization devices). Sharn used to be a Wolfe Tory distributor, but began to market its own mucosal atomization devices after that arrangement ended. The parties’ devices share many similar features, including a white nasal cone on the nasal device. The court found, however, that Wolfe Tory hadn’t met its burden of showing nonfunctionality; though it identified numerous other designs, it didn’t show that those designs offered the same combination of features, cost and quality it touted as its own.
Likewise, Wolfe Tory failed to meet its burden on its false advertising claims. Wolfe Tory alleged that Sharn’s ads that its devices atomize were false and misleading because at least some of Sharn’s devices don’t atomize. In March 2009, Sharn discovered that a small number of its mucosal atomization devices were defective; it recalled them and replaced them, fully advising the FDA. The court was unpersuaded that this amounted to false advertising: some of Sharn’s devices were defective, and Sharn remedied the defect. A party isn’t liable for false advertising just because some of the products were defective and were recalled. (Why not? Maybe there’s a materiality issue, if the numbers were small enough, but false advertising is generally strict liability.) Anyway, Wolfe Tory couldn’t show irreparable injury; this wasn’t a case of comparative advertising where harm could be presumed, and Wolfe Tory didn’t carry its burden to show irreparable harm in other ways.
We find that it is neither competitively neutral nor otherwise in the public interest to certify a carrier as eligible for universal service funding when that carrier refuses to contribute to legally mandated programs to insure universal service, programs into which the carriers’ competitors pay and which fund the very sort of universal service which is the goal of the ETC program. It also is not in the public interest to grant ETC designation to a carrier that denies the authority of the State agency charged with insuring universal telephone service and protecting the public’s interest, convenience and necessity.It's always interesting to see new market entrants negotiate old regulatory spaces. To what extent should they be able to cast off old regulatory structures aimed at previous architectures? The reference to competition is the key: Tracfone is competing with old market participants who are subject to the full regulations.
Sunday, October 25, 2009
The parties distribute a prescription drug whose sole active ingredient is colchicine. Colchicine has been used to treat gout and Familial Mediterranean Fever (FMF) since before the FDA was established. Until recently, everyone sold colchicine products without FDA approval, and for seven years plaintiffs bought their colchicine from one of the defendants. During this time, the parties’ products were listed in various drug dispensing databases and pricing services known as Price Lists and could be ordered by drug wholesalers or retail pharmacies.
In 2007, plaintiffs applied to the FDA for an orphan drug designation for its colchicine product, and in 2009, the FDA granted plaintiffs a three-year exclusivity period to market their product, Colcrys, for gout flares, and seven years for FMF. The price increased from $9 per bottle to $485 per bottle. (Apparently this generic-to-orphan transition isn't unknown, nor is the resultant price spike.) Nonetheless, defendants allegedly continued to distribute their competing colchicine products at substantially lower prices.
Days after achieving orphan drug status, plaintiffs sued under the Lanham Act and state unfair competition law, alleging that defendants falsely advertised their colchicine products as FDA-approved. Rather than alleging literal falsity, plaintiffs’ claim is that including defendants’ products on Price Lists and drug ordering systems confuses pharmacists into believing that they’re FDA-approved. Plaintiffs also alleged that the labels and product inserts for defendants’ products falsely imply FDA approval and greater safety than Colcrys. Thus, plaintiffs sought to enjoin defendants from listing their products on Price Lists and drug ordering systems and from distributing their products without labels and inserts containing information that the FDA required plaintiffs to put on Colcrys labels and inserts.
Plaintiffs principally relied on an opinion in a very similar action they brought against a different group of defendants who were distributing a different drug for which plaintiffs had also obtained an orphan drug designation and exclusivity period.
See Mutual Pharm. Co. v. Ivax Pharm., Inc., 459 F.Supp.2d 925 (C.D. Cal.2006). In that case, the court rejected defendants’ argument that the matter was one for the primary jurisdiction of the FDA and found likely success on the merits.
Here, however, defendants also attacked the merits of the false advertising claim, including the equity of enjoining defendants from doing exactly what plaintiffs were doing until days before they filed suit. The court found that plaintiffs hadn’t shown likely success on the merits. Ivax read the Lanham Act broadly; by contrast, Mylan Labs., Inc. v. Matkari, 7 F.3d 1130, 1139 (4th Cir. 1993), rejected a false advertising claim that placing drugs on the market falsely implied FDA approval. Plaintiffs’ survey evidence, the court thought, showed only that pharmacists are confused about what the inclusion of a drug on a Price List or drug ordering system means about FDA approval. And, the court noted, it wasn’t convinced that having drugs on a Price List or drug ordering system was “commercial advertising or promotion.” Anyway, there was little evidence that defendants “in any way created” pharmacists’ confusion. The product label/insert arguments were even weaker, both because evidence of confusion was weaker and because content disputes fall even more squarely within the FDA’s primary jurisdiction.
At least at this early stage, the court found that unclean hands substantially decreased plaintiffs’ likelihood of success. Almost up to the moment they sued, plaintiffs did the same things they now seek to enjoin.
Plaintiffs also failed to show irreparable harm. This was a dispute about money, and damages would be easily calculable based on defendants’ sales.
Saturday, October 24, 2009
Laura N. Gasaway, Associate Dean for Academic Affairs, University of North Carolina
Topic: How institutions deal with digital content. Digital content is changing research and library services. Libraries always acquired and used digital content, and used digital tech to preserve analog works.
Four issues for libraries: Creating and managing digital content; creating, hosting, and managing user-generated content; using user-generated content; assisting students and others with their creation of UGC.
Many copyright issues, including new ones about curating and preserving websites. Continues to be confusion among librarians and archivists, especially when collections mix public domain and copyrighted works.
Libraries have been creating their own UGC for some time. Library blogs: NYPL etc. More than 3000 staff for 87 branch libraries—they also tweet daily. Also creating and managing social network sites. Especially for younger users, it’s a way to advertise services. UMinn’s UThink, where faculty, students and staff can post content. Why do libraries host blogs/take risks of crap showing up? Promote intellectual freedom, build communities, explore links between blogging and traditional academia, and preserve institutional memory.
Libraries also organize material that’s neither created nor hosted by the library—online search tools, links to relevant articles (cf. Yen’s discussion earlier). Pointers/finding tools. Also evaluate content, similar to what the library does when it decides what to purchase. Big issue for schools: wikipedia—students don’t know its lack of authoritativeness.
Filtering: control/organization issue. Public & school libraries filter, often not because they want to but because state law or federal funding requires it. Content-based and not copyright-based.
Libraries that host UGC could be required to remove it if they receive a notice.
Internal library uses: creating exhibits etc. Display on Harry Potter might use dust jackets, photos of the actors, news clippings—what about including fan fiction and blog postings as well? Same considerations as for other users of UGC, but libraries are public forums and may achieve greater distribution.
Libraries may also have the advantage of exceptions for face-to-face teaching and distance learning. Face-to-face: copy must be lawfully acquired if it’s an audiovisual work. If teacher shows YouTube version of TV show, is that lawfully acquired content? Might not be! Distance learning: all types of works must be lawfully made to use in distance education, not just audiovisual works.
Libraries get asked questions all the time about fair use, because they maintain AV equipment. Undergraduates stand there begging for help, and librarians end up giving them hints; have to tread a fine line. Pointing them to resources is about the best librarians can do—what the librarian might do herself is likely to be broader than what she’d tell a student, so as not to mislead the student.
Traditional library practices can help, but we will never catch up to the newest tech—clarity on fair use seems illusory at best; there will be new cases with new results.
Michael J. Madison, University of Pittsburgh School of Law
He doesn’t much like the phrase “user-generated content” because it marginalizes certain types of content and certain types of people. Participatory culture, indigenous content, community-curated work, free culture—all these include some people/stuff and exclude others. He likes amateur art, primarily because it focuses attention on what he wants to talk about.
Themes: knowledge, in the colloquial sense—copyright as a system for producing knowledge. Amateur art, and its derivation from pro art and vice versa—the line that crudely divides our thinking about amateur/pro. Shepherd Fairey case as example. Not interested in whether the poster is fair use, though he thinks it is; not interested in whether Fairey copied protectable expression, though he thinks Fairey did. More interested in how we go about thinking whether the poster is the kind of expression we want to protect/legitimize.
Obama poster is not the only version Fairey used—this is the Obama Hope Stencil Collage, hanging in the National Portrait Gallery. Fairey is a trained professional artist, and has been commissioned by Warhol Museum to run around putting his posters on walls. This is a work of fine art, but there are websites around that allow you to create a Fairey-style icon out of any image. Obamicon website is an example.
The fact that you can become your own Shepherd Fairey: we don’t need training in art to produce something very much like what the trained artist produced. Can we think about this as a separate category from classic pro art?
Van Gogh: often cited as one real paradigm romantic genius author-figure. Jamie Boyle cites him. Madison was surprised to discover that Van Gogh was a copyist, not simply borrowing technique or tradition; he painted the same subject matter as other artists—admirer of Jean-Francois Millet. Well known for portraying peasants. Extremely explicit about what he was doing as a matter of theory. Copying for practice: learning how to paint.
Van Gogh says in his letters: I’m justified in working on things Millet didn’t have time to paint in oil: translating into another language, that of color, adding light and shade to woodcuts. It’s taken a lot of time and trouble. Why it seems good to me to copy: they are always asking painters to be nothing but composers. But it isn’t like that in music; someone who plays Beethoven adds his own interpretation. It’s not a rule that only a composer should play his own composition. I started copying accidentally and I find that it teaches me and sometimes consoles me. I did it out of profound and sincere admiration, an attempt to make his work more accessible to the general public.
Sounds very much like an adaptation/derivative work. But there’s more to it. Specific disciplinary contexts: music composition v. painting. Teaching: Van Gogh is acutely aware that his art is not about creativity or originality for the sake of originality or creativity. He has something to teach and something to share; in some sense teaching himself skills/insight, and also sharing with the public. Product and process, related. Van Gogh thinks of himself as a professional artist.
Might disagree with Madison’s judgment on the merits that Van Gogh was engaging in what today we’d call fair use. In fact, Millet was successful during his own life; Van Gogh had no impact on Millet’s market; Van Gogh wasn’t generally appreciated until after Millet died.
We could look at Van Gogh’s work in a modern sense as transformative, but that doesn’t explain a lot. Creativity isn’t an effective line. Rather, it’s organized around a professional discipline with commitments to teaching and learning.
Returning to the present: copyright as knowledge producing system. We’re familiar with some rhetorical moves from history: Statute of Anne’s “encouragement of learning,” Constitution’s “promote the progress.” Maps, charts, and books, first subjects of copyright law, was directed to useful/educational material, not towards entertainment as copyright has become. Progress has largely disappeared from meaningful impact on the law—Eldred. Today we talk about creativity and authorship rather than knowledge, but we shouldn’t be so quick to write knowledge off. Copyright is about producing, distributing, stewarding knowledge in material forms, practices, and ideas. For historical/doctrinal reasons, creativity has become the watchword, but it’s not part of the text of the statute. Now, with amateur art, creativity gives us trouble. Hard to apply default standard of creativity, distinguishing between casual authors and pro. Everyone is now an author. The creativity construct doesn’t give us an effective tool for drawing lines about what should be in/out of the copyright system.
Suggestion: knowledge, in broad/metaphoric sense, might be an effective alternative. Patent law has largely avoided this—law talks a lot about “teaching” the public the art.
Question then: can a focus on knowledge do more and better work than creativity? It’s not a panacea, and doesn’t have built-in concrete solutions; a conceptual matrix.
Analog cases like the hypo Millet v. Van Gogh come out about the same on his theory. Intermediate case: Rogers v. Koons—an early “digital” case because it was a kind of mechanical artistic creation, conceptual art; defendant didn’t physically create the sculpture with his own hands. If you apply creativity, it’s easy to understand how the court judged Koons uncreative, but if you apply a knowledge-based, discipline-based theory, it might favor Koons. UGC: Fairey. In the context of non-professional uses, we should separate our roles as consumers of other creativity from our roles as creators/recreators/remixers. How we access that work is a matter of being a consumer, but when we’re teachers and learners we’re doing new stuff. How explicit must a disciplinary context be? Van Gogh explains himself in his letters, but that evidence isn’t often available so we’d need implicit senses of what was going on.
He thinks his approach is soundly based in copyright history and policy. Knowledge also has an explictly social basis: teaching has its basis in the community. Modern copyright law is uncomfortably individualistic by contrast.
If this is worth pursuing, then there’s a lot of work to do—build a syntax and vocabulary to translate the concept to specific legal rules, including evidentiary considerations.
Rebecca Tushnet, Georgetown Law
I don’t know how well this worked, because my medication was hitting me hard, but I gave a talk about the EFF’s proposed DMCA §1201 exception for noncommercial, fair use remix video, tracking my testimony before the Copyright Office.
Moderator: Steven Hetcher, Vanderbilt University Law School, for Madison: does knowledge fit into “purpose and character”? Fairey: the intuition is that the photo has very little originality: it’s just Obama, and anyone with a camera could have taken that photo. Andy Warhol today would be sued out of existence, when he took photos of Mao/Monroe.
Madison: he thinks knowledge is a structural change; purpose and character would be a logical place to integrate knowledge into the fair use apparatus. On Fairey, Madison’s intuition is different—it’s not the photo anyone would have taken; part of the AP’s case is “respect for photography” and photographers as artists. Some of this is rhetorical overkill. But we tend to undervalue the character of the art that goes into something like photojournalism. There is discipline, training, professionalism. He wants to focus more on what Fairey did than on what Garcia did.
For Gasaway: How many DMCA notices does a university library get?
Gasaway: Many libraries are part of institutions. The notices come through the university. Public libraries are often ISPs. Hasn’t asked how often they get notices, but as libraries host more stuff, they will get more.
McKenna: operationalizing this construct of Madison’s. He’s not sure how to apply the Van Gogh argument elsewhere, such as timeshifting on a VCR. How does knowledge play out there? Knowledge = nature of the user in the community, and there’s a time aspect to that—if we decide something isn’t fair use, then no community develops, or develops sub rosa, but fairness means the community structures itself around the practice.
Madison: Skeptical of whether timeshifting fits into his scheme. SCt might have been better off coming out the other way. Vidding community is an example of a practice coming into existence.
McKenna: What about Napster users?
Madison: Millions jumped into the Napster pool overnight. Is that plausibly characterized as a community of practice/disciplinary community? Probably not: a bunch of people trying to get copies for themselves. There might be learning-oriented subcommunities, but that’s about it.
Bohannon: Copyright generally shies away from process and tries to look at works on their face—PGS work analysis. Those things are manipulable—people can always tell a story that’s about learning. Does this resurrect the idea of a new work, doing something different? If we found a new work of authorship, we used to say there couldn’t be infringement. (Is this true?) Was Van Gogh free riding? Maybe a tiny bit; he could have picked a number of things on which to practice. Free riding was relatively little given new material. Is knowledge about the new work and what it shows us about the old work, or is it about process/intention?
Also, what if there were harm? Absent harm there’s no reason to find infringement. But if we find harm, can we reject a copyright claim anyway?
Madison: Suppose Millet had a robust market and Van Gogh takes the legs out from under it. He says that’s still not enough. Reviving older ways of thinking about infringing. In terms of process/product: to some extent, he’s willing to look at product. But focusing on work-to-work comparison (except for PGS works) is overstated—there’s a fair amount of process consideration implicit in Feist, as well as Bleistein. Doesn’t want to displace product with process, but blend them.
Cotter: Terry Fisher proposed a definition of transformative use that he describes as more process-oriented. A use of copyrighted material that either constitutes or facilitates creative engagement with intellectual products. Wouldn’t limit it by discipline.
Madison: He’s talking about a conceptually similar line of argument. But Fisher thinks the legal system could never actually implement a scheme at his level of generality. Madison ratchets down and looks at disciplinary/community context.
Friday, October 23, 2009
Christina Bohannan, University of Iowa College of Law
The problem of First Amendment overbreadth/vagueness at the intersection of the derivative works right and fair use. Recent dogfighting case before the Supreme Court provides a point of departure. Prohibits certain video depictions of animal cruelty, with exceptions for those with political etc. value. 3d Circuit’s dicta: might be overbroad, covering speech that doesn’t pose the same problems of harm.
Seems likely that the statute will be held unconstitutional for want of narrow tailoring/overbreadth/vagueness. People won’t know what conduct falls within he exemption.
Relationship between derivative works right and fair use is similar. Both the animal cruelty statute and the Copyright Act regulate speech. In Eldred, the Court applied ordinary rational basis review. Copying isn’t speaking, and the CTEA merely extended duration, not scope of protection, and thus didn’t alter copyright’s traditional contours, including idea/expression and fair use. Derivative works: not merely copying, so kicks over into speech even under Eldred.
Derivative works right does alter the scope of copyright. It’s true that, today, the derivative works right overlaps with the reproduction right, but historically the derivative works right has an effect on the scope of protection. Uncle Tom’s Cabin translation case—existence of the right affects what we think counts as sufficient reproduction to constitute infringement. See also the Dr. Seuss case.
Even if we don’t have heightened scrutiny, overbreadth and vagueness concerns still apply to derivative works. In the dogfighting case, the prohibited conduct was defined very broadly. So too with derivative works. The new work must only be “based upon” the older work—doesn’t say that the new work must copy the expression of the old one. So when Dan Brown goes to Holy Blood, Holy Grail for facts, he might appear to violate the derivative works right on its face. Courts have helped by saying that the user must incorporate protected expression, but the language is still very broad. “Any other form” in the definition is also problematic—Congress said it was codifying fair use in 1976, but at the same time the derivative works right allocates a lot to copyright owners—including uses that they might never have anticipated, or satires.
Dogfighting statute: exceptions for serious literary, artistic, etc. value. Copyright’s fair use works as a similar exception for the derivative works right. At oral argument, Justices pointed out that these were vague terms, and this is also true in fair use. In fact, courts also routinely find that there’s no fair use even when the use falls within the listed categories (criticism, research, etc.). If courts and scholars can’t draw the lines consistently between derivative works and fair uses, how is a layperson supposed to?
Modest proposal to save statute from overbreadth/vagueness. Canon of noscitur a sociis—interpret general language in light of specific language, often used in First Amendment cases. Look seriously at harm issue as well as another constraint.
Alina Ng, Mississippi College School of Law, When Users Are Authors: Authorship in the Age of Digital Media
Authorship is her main concern. Web 2.0 facilitates collaboration, sharing of information, etc. Communal owernship of creative works—wikis, community radio, etc. Plus, once content is produced, distribution networks are vast.
Is there any purpose in using the word property for literary/artistic works?
In the pre-digital environment, the production and distribution functions of the copyright system are separate. A user who made a copy would get an inferior result compared to a publisher-produced copy. Author-publisher-user model assumed an efficient market would transmit works to the market. If we move away from patronage, we need to allow rights to transfer from author to publisher. Need exclusive rights to allow author to recover costs of publishing. Efficient market = learning and progress.
Now, the production and dissemination functions of copyright have merged. The user can be publisher and distributor. Authorship isn’t necessarily commercially motivated; baseline right to exclude isn’t necessary and may be waived; economic efficiency is not obviously a normative criterion for protecting works as property.
Authors always borrow from others, so she’s not insisting on the romantic author as a feature of the analog world—but we can identify borrowing, as when Milton borrowed from the Bible. In the digital world, authorship is more communal; the focus is on sharing and not on borrowing. (I’m reminded of Severine Dussollier’s critique of Creative Commons—by this logic, online collaboration could be seen as a matter of grace and not obligation.) Dissemination has moved towards the author and away from the publisher.
Analog world: exclusive rights are fundamentally utilitarian. Digitally: authorship self-perpetuates through desire to share. Rewards are not economic but personal. People say that propertization and removal from the public domain stifles creativity.
Her view: there is still a need for a strong property rights system, not for the traditional reasons of production and distribution. Sustainability: Property rights serve to identify the original author, recognizing individuality. Should only abridge rights if none of the ideals intended (encouraging authorship, providing a connection between authors and readers) when rights were first recognized will be at stake; if a competing right would be undermined otherwise; or if there will be an unanticipated social cost going beyond the cost anticipated when the rights were first granted.
Alfred C. Yen, Boston College Law School
A preliminary First Amendment analysis of a law that would treat content aggregation of news as copyright infringement. Some have suggested that, if we don’t act, there won’t be correspondents in Moscow, or Afghanistan, etc. One way to deal with this: legislation.
Constitutionality is his concern, not good policy, because (1) legislation has been proposed, and (2) discussing the constitutionality of legislation also tells us something about possible judicial solutions.
Problem: Papers’ information is being reused by others for advertising purposes, and newspapers used to have exclusive access to that ad revenue. Papers sold eyeballs to advertisers. News aggregator: someone who posts links to stories. The paper used to make money because you scanned the headlines as you paged through; now you are scanning your aggregator, and advertisers are paying the aggregator. Papers want that money back.
Possibilities: treat copying headline/lead sentence as copyright infringement, or treat linking as infringement. Does it possess the minimal creativity required for copyright? Currently, not every use of a work is infringing—quotes, classically; and there’s no copyright in facts. Plus, merger: when there’s only a limited number of ways to express something, no copyright.
Conclusion: in most cases, aggregators do not commit infringement. Caselaw: linking doesn’t involve a copyright violation, because a link is not a copy. As for copying: most headlines are not copyrightable, though maybe some clever ones are. Even if a court were to find a headline original, merger would defeat claims, because in the limited space headlines have, there are relatively few ways to express the news. Also reaches the same result with lead sentences. Finally, fair use would still shield a number of uses. Purpose: changed, not from reading the work but presenting the work for purposes of deciding whether you want to investigate further. (Again there’s the question of whether that’s a purpose the copyright owner should control. Many newspaper sites put similar teasers on their home pages.) Amount: small. Nature: factual. Harm: not a substitute for reading the article. But if the real market is the market for advertisers, then perhaps the activity of news aggregators could be found infringing.
What if we changed the law? Eldred. A news case wouldn’t justify deference to Congress—copyright doesn’t normally present First Amendment problems, but that’s because the traditional limits on copyright prevent serious speech impacts. Legislation to protect newspapers would change the traditional contours, both fact/expression (subject matter) and fair use. Intermediate scrutiny of the sort applied in Golan, using Turner v. FCC standard. Turner involved must-carry of TV signals, done to sustain over-the-air broadcasters from losing economic viability to cable.
In Turner, the cable operators alleged violation of First Amendment. The Court said the gov’t purpose was constitutionally acceptable: preserving the benefits of free over-the-air TV; promoting a multiplicity of sources; promoting fair competition. Then asked whether it burdened substantially more speech than necessary. Must-carry was ok: only affected a few channels; increased public access to information. We can imagine the same arguments being made on behalf of regulating news aggregation.
Distinctions between news legislation and Turner: (1) None of the cable operators was prohibited from speaking, which isn’t true in news aggregation—a limit on a certain type of activity. (2) Wouldn’t increase public access to news. (Compared to what? If you really believe that this is necessary to save the news, then maybe it would increase access compared to the no-law world.)
Can Congress do other things? Micropayment liability rule only.
Q for Ng: Are you arguing for moral rights of attribution?
Ng: She hasn’t gone hat far—still wants to call them property rights, allowing the author to exclude users of whom he doesn’t approve. She’s not just concerned with economic incentives.
For Yen: What about a hot news doctrine by statute?
Yen: Hot news could be constitutional, if short term. But it wouldn’t help.
Lolly Gasaway, for Ng: There are also digital works with the same attributes as analog works—no serial authorship, etc.
Gasaway for Yen: How do you differentiate aggregation from what libraries do?
What libraries do is fair use too in a large number of situations. They don’t affect profits; the real question is these new business models. Raises the point that new legislation would have to be limited to figure out what it covered.
Gasaway: If headlines are protected in Europe, even directing to the paper’s site would infringe.
Madison: Don’t you have to show a causal connection between aggregation and papers’ economic woes? Reuse may not be affecting the papers—it’s really about a dramatic decline in classified ads, which has very little to do with content. Doug Baird’s chapter on INS in IP stories. Maybe we need to let newspapers sponsor gambling, or something else that allows them to subsidize news.
Yen: On the general Q of why newspapers are suffering, we don’t really know. The business model is broken, and copyright may be irrelevant. But the decline of newspapers is a problem of social importance. Regardless of reasons, gov’t might want to do something about that. But legislation may not be necessary—a lot of aggregators commit flat-out infringement, taking copies of articles/photos. The papers just need to get off their keisters and sue!
McKenna: On causation, if you could say that the real issue was lost eyeballs, the flipside is that people read articles in papers they’d never otherwise encounter. One might think that a tech fix would be useful: if robots.txt says don’t aggregate, then don’t. But most papers are likely to want aggregation (and payment!).
Yen: Google News does respect robots.txt. Papers want it both ways: payment plus aggregation. That’s a policy question. (Compare to Tasini: the freelancers wanted to stay in the database and get paid; this is a matter of economic leverage, as the NYT was able to tell them that it wasn’t going to do that going forward, and contract prices didn’t change when freelancers started giving up all electronic rights. Even with a legal right in the papers, we might expect aggregators to drive the license price down to pretty low/nonexistent.) Google-AP deal: Google may get extras from AP, similar to the Book Search deal.
McKenna for Bohannon: Does Eldred define the traditional contours? If not, what other things count? We’ve had a derivative right for some time.
Bohannon: Depends on how we look at the time period. Abridgements and translations can be viewed as new forms of the same story. But now there are some unforeseeable extensions, which is where the problem lies. Vagueness/overbreadth don’t necessarily concern traditional contours.
Yen: How broadly would Bohannon’s framework apply? If aspects of the Act are unconstitutional, every argument you make could be applied to substantial similarity and fair use.
Bohannon: Certainly would need to think about substantial similarity. She’s trying to follow up on what courts care about with overbreadth; for her, this is really a problem about the intersection of fair use with the exclusive rights. People find it difficult to rely on fair use in any meaningful way. Publishers/copy shops just won’t rely on fair use.
Yen: Perhaps you should look at libel law, not just overbreadth/vagueness. The truth/falsity line was supposed to govern. But the SCt said judges had to interpret that law in certain ways/certain burdens of proof. Perhaps to avoid a chilling effect we need to say that the thumb should be on the defendant’s side of the scale.
Bohannon: Generally, we say that you can’t condemn speech without finding harm that the gov’t has an interest in avoiding.
Lunney for Bohannon: Never quite gets how you draw the line between constitutional and unconstitutional: making a film from a book v. making a mash-up. Economic analysis has a toolbox for drawing lines, given uncertain empirics. But what does the First Amendment have to say about drawing the line?
Bohannon: notice is key—wherever the line is, it must be intelligible. We don’t know how far the entitlement should go. We need to take harm more seriously, though; in libel, we don’t say that failure to benefit a person is a harm, but in copyright we do say “you could have paid a license fee, therefore you’ve harmed the copyright owner.” Amount/kind of harm usually required for other types of legislation to pass First Amendment muster is greater—we tolerate a lot more speculation with copyright.
Lunney for Ng: Your argument: digital property rights play a coordination role just as they do in the real world. Why do we need coordination when there’s no physical rivalousness?
Ng: Attribution to creator is key.
Lunney: But we don’t let farmers control the subsequent use of their tomatoes; why give authors such control?
Ng: Greater connection between author and work than between farmer and tomato.
My thoughts on Yen: without causation, the argument that there’s no less restrictive alternative to restricting aggregators seems to fail—if the problem is the papers need more money, find some way to give them more money. Fred Schauer has a paper on defamation and papers’ willingness to absorb costs that might be of interest here. Frederick Schauer, Uncoupling Free Speech, 92 Colum. L. Rev. 1321 (1992).
Q: On a practical level, is a focus on harm likely to make it difficult to resolve cases quickly, for example where a defendant wants to release a movie on schedule? (Expelled and its use of Lennon’s Imagine.)
Bohannon: We ought to have a pretty clear showing of harm; First Amendment cases pose a higher barrier in terms of what harm needs to be shown. Copyright law tends to accept might/could arguments more readily. Too speculative.
Cotter: cognizability is part of the fourth factor. Critiquing Lennon’s Imagine is not cognizable.
Gervais: The derivative work is something like dilution in TM: you’re not sure you can prove confusion, but you really want to give the owner control, so you extend the right with a slightly different rationale.
Bohannon: She thinks that derivative works has influenced our perception of the reproduction right; we could get rid of derivative works now and see no diminution of scope, but that might not always have been the case. She thinks of it as a clarifying amendment to the reproduction right—look at substantial similarity of expression, not at whether form changed. She hesitates to say that the entire Act is overbroad, but maybe this is really about fair use.
Vanderbilt Journal of Entertainment and Technology Law symposium
Panel 1: Treading the Line: Fair and Derivative Uses
(Note: I’m a bit under the weather; I may be missing parts of the presentations.)
Thomas F. Cotter, University of Minnesota Law, Transformative Use and Cognizable Harm
Transformative use is often more of a conclusion than an analytical tool for determining fair use. Undue emphasis on transformative use obscures the underlying policy issues and creates unnecessary doctrinal knots. Transformative use should be subsidiary to the overarching question of whether a use causes cognizable harm to the copyright owner.
Transformativeness tends to be underinclusive (citing me!). Efforts to define transformativeness have proven elusive. Possible definitions: (1) Any use likely to generate more social benefits than costs. But that’s too vague/unworkable. (2) Any use that would result in the creation of a derivative work. Problem: nullifies derivative works right. (3) Opposite result: any use that doesn’t result in the creation of a derivative work—approach taken by Posner in Ty v. Publications Int’l. Transformative uses are complementary to the original and derivative works are not. Problem: commonplace derivative works are complementary, at least for some users; some fair uses are not. Harry Potter Lexicon case also tried this move: derivative works transform the original into another medium, mode, language, or revision while represneting the original, while transformative fair use transforms the purpose of the original. Thus, the Lexicon was not a derivative work, though it did violate the reproduction right. It had a somewhat transformative purpose, but inconsistently so. Problem: this framework has assumptions about whether a use “represents” the original or what a new “purpose” is—we can define those things many ways. (4) As an empirical matter, transformation of content is less likely to be fair use, but transformation of purpose is more likely—Tony Reese’s article. Cites Kelly v. Arriba Soft & Perfect 10 v. Amazon, where thumbnail images were found to have different purposes from the original images. But see Gaylord v. US—transformation of content, but arguably not of purpose, held to be fair use. Application of these criteria is unpredictable.
We should do something else: assess whether unauthorized use threatens cognizable harm. Emerging scholarship by Balganesh, Bohannan, Sprigman, Stadler. Is this harm of the type the copyright system is intended to prevent? Must consider copyright’s purposes and the purpose of the fair use doctrine and the derivative works right in particular.
An unauthorized use harms the owner if her utility is lower in a world in which the unauthorized use occurs than it would be in a world in which the law forbids the user from engaging in the use absent the owner’s authorization and the user complies with the law. No harm, for example, where the owner would agree to use but for presence of high transaction costs.
Easy cases: I make 100 copies of a best-seller and hand them out to friends. Probably substitutes for authorized sales; a rule allowing this would cause systemic harm. Contrast: I quote portions of the original in a negative book review. Harms the copyright owner by diverting demand, but that’s not cognizable harm. What if my copying whets others’ appetites for authorized copies? This is harm under Cotter’s definition, because in the alternative universe in which the owner could demand compensation and the user would comply, the owner would be even better off. This is cognizable harm under current law, but he’s not sure it should be.
Harder cases: reproduction of works of visual art for purposes ancillary to news reporting/historical commentary: Bill Graham; LA News; Nuñez. Satires: Dr. Seuss. Works that critique the culture that a work symbolizes: SunTrust; Blanch v. Koons.
Relevant considerations: does the work create a conventional type of derivative work? Is the use a type foreseeable at the time of creation, or likely to pose a threat to the copyright incentive scheme? If not, should that be dispositive? If it is dispositive, do copyrights become unenforceable at some point prior to term expiration? Is the use the most effective way for the user to express her viewpoint (a Cohen v. California consideration)? Does the use cause psychological/moral rights harm to the owner, and do we care? Is the author likely to enter the niche served by the use or likely to prevent anyone from entering that niche?
Applications: a finding of cognizable harm may enable copyright owners to inhibit freedom of expression: Salinger v. Colting, wrongly decided unless we accept a moral rights theory.
Glynn S. Lunney, Jr., Tulane Law
The derivative work right proved largely unimportant under the Act. It was originally derived from the translation right as a response to Stowe v. Thomas, in which a German language publisher serialized/translated Uncle Tom’s Cabin. Held: a translation is not a copy. Expanded 1909 to encompass dramatizations; 1976 Act expanded further to any work. In application, hasn’t proved all that important because the reproduction right covers almost everything. Harry Potter Lexicon case is another example: you’d think it’s clearly a derivative work, but the district court finds that instead the Lexicon is a copy.
What role could the right play? A distinction between substitutes (reduce demand for the original) and complements (increase demand). We have a really good story about why copyright law should ban making substitutes without the author’s permission. We don’t have a good theory about why copyright owners should control production of complements. In other economic sectors, we don’t give control over complements: if you make a car, aftermarket sellers can sell tires, lights, little things to hang off the dashboard, etc.
Instances: hardware/software interactions. DVD player is required to play the DVD. You need both: they’re complementary. Better music player = you are more likely to buy more music. Second, uses of the work that are complementary to the work: radio airplay, where it’s well established that radio play drives sales of records. Third: reworking that’s complementary, increasing demand for original work—the film made from a novel. Film release always increases sales of novel.
What are the economics of complements? Copyright law is all over the map with treatments of complements. What is the proper treatment?
Simplistic example: does a monopolist in one market need to control other markets to get the highest return? Answer: no. Assume you are the only seller of shoes, which sell for $10 and cost $2 to produce a pair. You price at $10. What if you only control the left shoe? No, you sell it at $9 and get the same monopoly profit, and the market drives down the price of the right shoe. So often a monopolist earns just as much without controlling complementary markets.
Sony safe harbor, limited by Grokster’s inducement theory. Generally copyright owners don’t have the right to control complementary hardware.
What about radio airplay? There is a performance right, but there are various exemptions like the homestyle equipment limitation for businesses.
Other unclear applications: selling extra levels of Duke Nukem.
When should copyright owners control the complement? The economics of the complement are the same, so our rules should be consistent, but they aren’t.
Incentive argument. But we don’t give copyright owners control over all complements; also, often the complements need incentives too. We want new tech, new ways to experience works. Incentives are required not just for tech but for reworkings—it took an awful lot of effort to create the Harry Potter Lexicon.
Public performance right for radio stations—should we transfer their profits, which can be monopoly profits, to songwriters? Well, do we want more songwriters or do we want more radio stations? Recurring payola scams suggest that, under existing rights, the songwriters are making too much money—they transfer their wealth to the radio stations so that they get more exposure.
Questions: are the complements likely to be natural monopolies? A film version of a novel is expensive, and there’s only likely to be one of them. Good idea to recognize copyright owner’s right where there will be a natural monopoly. If we’re only going to get one film, at least for a given time period, we probably want the author-authorized version rather than unauthorized.
What are the benefits/risks of requiring a license? If the transactions costs are too high, then it’s not worth it. That accounts for a lot of the complementary hardware rule—too hard to ask every copyright owner if it’s ok to make the iPod.
Version of Arrow’s paradox. If you have valuable information, how do you convince someone to pay for it when they don’t know what the information is? A person with a really great adaptation idea may not be able to negotiate with the author without losing the benefits of her idea.
Daniel J. Gervais, Vanderbilt Law, A Principled Approach to Copyright Exceptions and Limitations
We have no framework for limitations/exceptions; can we get one? History of international copyright: a one way rights elevator. Began with basic text and kept amending it to add rights. Derivative work right began as translation right; expanded—US version is probably the broadest in the world.
There’s only one mandatory exception to copyright in Berne/TRIPs: the right to quote. Then there’s a series of exceptions a nation is allowed to make, in most cases responding to countries that insisted that they wanted to maintain certain existing exemptions. In 1967, the negotiators added a general test for exemptions—the 3 step test. Not that far from the “cognizable harm” test—a cap on exemptions/limitations. But copyright rights are not capped. A
Irony: that’s not what the authors who drafted the first version of Berne wanted. Victor Hugo headed the committee that first drafted Berne. He did not want unlimited rights. In other countries as well: The Statute of Anne was an act to “promote learning,” not to give authors every possible right. In the US: “promote the progress of science and the useful arts.” The public interest was the dominant consideration, and in some cases the only relevant consideration.
We could in theory determine the ideal level of copyright protection to match our objective. But we are actually talking about protection levels—depends on type of work, type of user, type of use. That’s codified in fair use and its four-factor test.
Three things copyright can do: which functions should be within author’s control; which should be remunerated (compulsory licensing); which should be free. (The US has more compulsory licenses in its Act of any country.) Recognize need for two equilibria: internal balance of copyright; external balance with things like privacy (in filesharing measures, for example). Exceptions can limitations can’t take care of all these things, but they can help.
Human rights concerns: we see a balance in human rights instruments. They mention both protecting the moral/material interests of authors, and also a right to participate in the cultural life of the community, to enjoy the arts, and to share in scientific advancement and its benefits. A right to development, for example. Conflicts must be managed—someone has to decide when privacy ends and copyright control begins.
What if we reject human rights and say it’s all about money? This is reinforced by the move from Berne, standalone agreement, to TRIPS in the trade arena. Damage to copyright owners is lost revenue. According to two dispute resolution panels, the 3-step test for interference with normal commercial exploitation and the author’s “legitimate interests” must be interpreted in financial terms—are rightsholders losing money? If they are, then the exception violates the test. This could be used either to cabin the right or craft a broad exception in national law—we could enact a three-step test directly into national law and have courts assess it. He thinks this is a cop-out though, a denial of legislative responsibility—the test should be a safety valve.
We cannot assume that all new uses are covered by existing rights or need new ones. Tech can defeat a “right” to control, as P2P suggests. Privacy is going to be a big barrier for copyright. They used to be in harmony; copyright didn’t care what books you read in the privacy of your own home. When the end user became a potential disseminator and re-creator, things changed. No thought given at the international level to exceptions or limitations that would react to these developments.
Principles: copyright rights should not prohibit use in the private sphere of users (remuneration remains a possibility). They should not prohibit access in countries or by groups of users who would not otherwise have reasonable means of access to the content. They should not prevent educational uses that cannot be reasonably licensed. They should not prohibit access by institutions whose purpose is to document and preserve culture. (Notes that filesharing services represent the greatest collection of music the world has ever known.) They should not prevent use and reuse that serve a publi interest in free expression, including the creation and dissemination of culture and information—quotation, parody, caricature, pastiche, research, criticism, and review. Courts should have latitude not to apply exclusive rights, especially where they interfere unreasonably with the right of information or free press—injunctions should not be mandatory. Copyright should not prevent government use in the public interest, though internal/commercial uses by the government should remain subject to exclusive rights. Copyright shouldn’t prevent access to and at least noncommercial use of governmental publications of a general nature—the US already recognizes this, but Crown copyright and the like elsewhere are a big problem. Finally, given the 3-step test, exceptions should consider compensation as a substitute for control.
Q: Nimmer calls the derivative work right superfluous. So is there any optimal level of the derivative work right? Do we need it at all?
Cotter: In the US, there are marginal situations where it matters. Gilliam—taking the original and modifying it, where for technical reasons there’s no reproduction. (Seems unlikely to me that there was no reproduction in Gilliam.) Could tweak the law to capture those instances. May not be necessary as a matter of policy. Moral rights might also justify the concept of a derivative work rights.
Lunney: He doesn’t think it’s necessary; and in fact he wouldn’t extend the reproduction right to works that were very different from the original—he would require a showing of substitution.
Gervais: Within the current rights framework, the first step is to have a principled discussion of what we want out of our exceptions. So for example there is a current dispute over what, internationally, we should do about exceptions for people with disabilities, who can’t read plain text. We keep giving rights and it doesn’t seem to be doing that much.
Mike Madison: Emerging dialogue over the concept of harm. Cotter defines harm for his purposes, but doesn’t define the causal relationship—should we use a concept of causation, and if so how would that work?
Cotter: Ancillary to the overall question of whether certain types of harm should be compensated.
Steven Hetcher: Isn’t cognizable harm circular? Google Books: you’ll sell more books because of our scanning; but the owners say we’d do better if Google had to pay us for scanning.
Cotter: Trying to avoid circularity with his inquiry.
Q: Salinger: do you read it to hold that intent to enter the market is irrelevant?
Cotter: Yes, that’s where the judge ends up—also considers that author may have been motivated to create the work in part because no one could write an unauthorized sequel—weighs the author’s right to veto. Questions whether that’s consistent with copyright as means to promote progress. The harm asserted here would be psychological.
My Q: Cotter & Lunney seem to conflict on complementary uses. Go at it?
Cotter: He thinks his cognizable harm factors are similar to Lunney’s proposal.
Lunney: There are consistencies, but the two start from different places. He focuses on social utility, not whether the author is better off.
Me: but conceptually you seem to disagree about whether the definition of “complementary” is useful.
Cotter: Yes, he thinks that the movie example is helpful—the book isn’t really a movie substitute.
Lunney: He’s not saying complement/substitute maps directly onto the derivative works right. He’s troubled when a court concludes that a use increases the value of the copyrighted work, but it’s still not a fair use because Congress decided to give these rights to the copyright owner.
Mark McKenna: A plaintiff in a TM case can construct a story of confusion in practically any circumstance. Harm in copyright too—there’s always a way to say that potential money was lost. We need to decide whether we want copyright owners to control ancillary markets. Cotter is inviting courts to engage in fine-grained analysis about harm; Lunney seems to want more categorical judgments about kinds of uses where we will ignore harm stories.
Cotter: Asking about harm can screen out some things, for example the high transaction cost case in which no deal would occur.
McKenna: But then the owner says, “give me the right and a market for clearance will develop”—CCC, etc.
Cotter: Still, clarifies the nature of the harm. We can decide whether we care about the kind of harm identified.
Christina Bohannon: She’d add some categories of no-harm situation. For example, when the defendant was never going to pay for the use (the counterfactual world just doesn’t exist). Copy shop cases: these professors wouldn’t assign those works if they had to pay the fees. (There’s very interesting literature on counterfactual worlds that might be of use to Cotter there.) Or when the new use increases sales.