Problem: you get an invention by specifying it. That implies that an invention is a limited series of things/steps. Then you have to claim it, and the claims have to cover the things. You can’t really do that, because if your claim is just a set of things, someone else could easily change a few things—color, nuts and bolts—and get the benefit of the invention. So you claim more broadly: principle or idea.
Example: discover a cure for AIDS using a particular type of radiation and a particular machine. How to claim? No problem of nonobviousness or utility. Levels of abtraction: a single embodiment can translate to many claims: (1) A cure for AIDS, (2) A cure for AIDS using radiation, (3) A cure for AIDS using X radiation, (4) a cure for AIDS using the exact same machine using radiation. Every level of abstraction is accurate, new, useful, and nonobvious. Which level of abstraction will a court let you protect your invention at?
Enablement doctrine. But suffers from two contradictory rules: first, one embodiment: you only need one method of making the invention to satisfy the enablement requirement, but a single specification embodiment enables any claim that describes the embodiment, including the cure for AIDS. Second, the full scope rule—you need to enable the full scope of what you’ve claimed. The specification must teach every embodiment that will be covered—at least teach others of ordinary skill to make it. But then it becomes outdated from the moment the patent issues—suppose you patent a chair and make it out of wood; two years later somebody invents steel. At the time of filing, there’d be no way to describe/enable a steel chair. So full scope rule limits embodiments to extant technologies. Variant: no need to enable after-arising variants, only what’s practicable at the time. But that’s the same as the one-embodiment rule—everything is either pre-existing or after-arising.
Have the same problem in copyright: can describe Romeo and Juliet as “boy meets girl” or down to the last word, and the level will determine how many works read onto that description. This has been recognized as requiring a policy choice of level of abstraction.
Solution: only question is whether we put policy questions explicitly on the table or ask them sub rosa. If levels of abstraction is about an economic quid pro quo, then litigants can supply relevant information: the economic effects/incentives of patent.
Gordon: What kind of evidence would you want litigants to supply?
A: Evidence is very expensive, but one question would be the extent to which a subsequent product gets its benefit from the patent disclosure: if you wouldn’t be able to make it without the patent disclosure, that’s a good sign that the economic benefit of granting a patent with enough scope to cover the product is a good sign. Collateral benefits of innovation are also worthwhile, regardless of whether this particular accused infringer derived benefits from copying—look at whether other people did. Also, with AIDS cure, ask how many lives saved before someone else’s product has come on the market to compete.
Lemley: Is there a right level of abstraction or multiple correct levels? Hand has this problem in copyright; he doesn’t tell us whether there’s a right level of generality at which to assess an infringement claim.
A: Multiple nested claim approach still picks a level—the broadest claim that is protected. At some level, there has to be a right level that balances incentives and monopoly costs. Problem is costs of calculating that, which probably outweigh minor improvements in accuracy in a particular case.
Lemley: and is the same true of copyright? In copyright, I can show similarity at the text level, or at the story level.
A: Yes, but text is concrete/easier to find copying of expression if it’s verbatim. It’s when you have similarity across structure/plot that you’re pushing the envelope of levels of abstraction. Process is more explicit in patents because of the claim structure.
Jeanne C. Fromer – The Psychology of Intellectual Property
Patent and copyright have different standards for protectability. Can psychology tell us anything? One common explanation: copyright has a low standard because rights are limited, because we protect only expression; patent has a high degree of protection. That doesn’t explain how things got to be the way they are—why does patent have a high degree of protection?
Goldstein suggests results theory: copyright wants to create a vast library, whereas patent cares about a smaller set of efficient information units. This is intuition, but she thinks it’s right, with a deeper theoretical framework under it. Clarisa Long and others offer a judgment theory: characteristics of copyrighted works are in the eye of the beholder, so we set up a standard where we don’t have to judge; scientific characteristics are ascertainable, so we do judge. Empirical evidence is the opposite: people are equally likely to agree on the merits of artistic works than they are on the merits of scientific works. (Note that merits aren’t equivalent to meaning.) Another theory: too hard to figure out prior art in copyright in every case. Patent—not so. She disagrees, in previous work. John Wiley’s learning theory: we don’t want artists to care so much about what’s come before, but patent is more cumulative; we want to force scientists to go out and find out stuff to build on. She thinks that’s wrong too, based on the way creativity works.
No matter what IP theory you believe in, creativity should matter. A utilitarian wants to give incentives to act creatively. A Lockean might think hard creative work should be rewarded. A Hegelian would think that creativity is about putting personality into one’s work. Psychology of creativity gets around problems in studying whether IP regimes are good for creation—hard to control variables. Psychology allows us to ask on the front end when we ask what incentives we want to give.
Overview: psychologists define creativity as something that’s new and appropriate: some community recognizes it as socially valuable. People create in context. (I’m thinking of the Professor on Gilligan’s Island.) Also, it’s hard work. Not a thunderclap. You need to understand the domain in which you’re working, which usually takes a decade of work before you can create valuable works. You might need a lot of school. Then creative process often takes months/years/decades. Samuel Taylor Coleridge, Xanadu supposedly came to him in an opium dream: that’s false. Pieces of the poem predate his supposed dream, and prior writings he’d read seep into the poem. Creativity is also deliberately cultivated.
Stages of creativity: (1) Preparation. Identifying a problem and gathering resources. Gutenberg wanted to put images on paper. (2) Incubation. Unconscious processing of information. People are making random/remote connections. Gutenberg made a connection between what he wanted and a wine press. (3) Illumination. A-ha! (4) Verification. Test the idea, develop it, tweak it.
Two key parts: problem-finding, and everything after—problem-solving.
Is artistic production different from scientific production? Two key characteristics of artistic creativity: (1) We really value problem-finding. What is a “problem” in the arts? Taking one’s personal emotions, experiences, themes, and translating them. Experiment on art students: given 27 objects and plenty of time to draw. Followed careers over time; the most successful were the ones who spent, in the studio, the most time on problem-finding. (2) We want some newness in artistic creativity, but not too much. After a certain level of newness, audience reaction turns negative fast.
By contrast, science is more concerned with problem solving and tolerates greater degrees of newness. Scientists good at problem solving have better performance by many metrics than other scientists. Problem finding is important still too. Society is willing to accept inventions that flout accepted conventions—e.g., the telephone; people are very willing to use nanotech. Vacuum tube replaced by transistor = huge leap, doesn’t require immediate change in behavior, so people are happy to do it.
Implications: copyright’s orginality standard is directed at problem finding. Independent creation as a defense: if you’ve created a work independently, you’ve injected something of yourself, which is problem-finding. Holmes recognized this in Bleistein: personality shows itself in expression, which always has something irreducible. Creativity is ill-defined in the law, but the stages of creativity give us a handle on what it is—the phonebook in Feist doesn’t pass through the stages the way the photo of Oscar Wilde did. Also, newness standard is low because we don’t want authors to (have to) push too far.
Patent, by contrast, wants newness. When something has no specific utility, no problem was found, so there can’t be a solution.
But there are a number of non-prototypical works. Some of our problems come because the works don’t fit. In copyright, software: protecting it but it seems to be about problem-solving, not about problem-finding/expression. Jazz music: that’s about flouting convention, encouraging people to be more new (and so the low standard of originality doesn’t give incentives as we want them to be in jazz? I think there’s also something there about performance and its poor fit with copyright theory). Theoretical science doesn’t fit patent well—it’s more about problem finding.
Hughes: just because there’s a low threshold doesn’t mean copyright is intolerant to radical works. Doesn’t see a difference between a pioneer patent and a revolutionary artistic work like Picasso’s. Both systems provide protection to incremental innovation as well as radical innovation, and we allow market responses to reward or misunderstand pioneer patents or radical artistic innovation.
A: Sees differences at lower end of the scale. Patent doesn’t protect a whole host of things, because they reinvent what’s been done; Hand would protect such things if they were within copyright’s subject matter. Thin copyright protection v. thick—some post-Feist cases say you get thicker protection with more creativity, which would suggest an incentive for greater novelty. But the caselaw hasn’t gone that far.
Q: Creativity differs across industries in patent and copyright. Superimpose the institution—large firm, small firm, inventor. How do you cover such different contexts?
A: Can’t expect everything to fit into one story; we should look at whether there are too many counterexamples and then we’d tell a different story.
Ariel Katz – Beyond “Essential Facilities”: Innovation, Intellectual Property and Competition Policy across the Atlantic
Background questions: When does a refusal to license IP violate the antitrust law? When does the essential facilities doctrine apply? His question: What can we learn from this debate? The assumption is that the US and the EU diverge sharply.
2007, Court of First Instance affirmed a finding that Microsoft had violated competition law by refusing to license interoperability—huge fine plus forced disclosure, even if the information was protected by IP. Continued a line of cases finding refusals to license IP may amount to abuse of dominant position. Duty to license may be an appropriate remedy. Many critics say this is compulsory licensing undercutting the very basis of IP, right to exclude—even a disrespect for IP/unwarranted interventionism. Some say the remedy is short-sighted, ignoring benefits from innovation.
US supposedly protects IP rights more vigorously, doesn’t regulate them as much, promoting innovation. Even commentators who approve the EU approach share this view.
Valid point: there is a divergence between the jurisdictions. But there are other dimensions of the interface between antitrust and IP. Current focus is incomplete/mistaken. The underlying IP regimes contain rules and doctrines that are internal and promote competition or otherwise limit exclusionary power. Also, antitrust in process and substance shapes the scope of IP rights. A different picture emerges: the US has tended to circumscribe IP rights and subject them to more aggressive antitrust scrutiny than the EU. The US has a less is more approach to copyright.
Key cases: Magill (1995)—TV listings published without a license; refusal to license was abuse of dominant position. There was consumer demand for this new product; the refusal prevented competition in a new product market; there was no objective justification for the refusal.
Contrast to Feist case in the US. When Rural sued Feist, Feist defended with a list of copyright defenses and also filed an antitrust counterclaim, invoking the essential facilities doctrine. The DCt ruled in favor of Rural on the copyright issue, but it’s less known that the DCt ruled in favor of Feist on the antitrust counterclaim (though the CtApp reversed). Prior to reversal, there was a long line of cases holding phone directories copyrightable. The starting point in both cases, Magill and Feist, was quite similar: competitor seeks license to compilation and is denied a license, but goes ahead anyway. Europe adopts an antitrust solution, but US adopts a more radical copyright solution. US SCt was aware of the antitrust solution but declined, acknowledging it was motivated by competition concerns.
What good does your copyright in TV listings in the EU do you, then? They’re allowed to charge reasonable license fees, unlike in the US.
Feist isn’t singular: merger doctrine, fair use, and copyright misuse work to do work that antitrust may be asked to do in the EU. Strong similarity between fair use and the Magill test: purpose and character = new product; nature of work = de facto, all the European cases involve marginal works, not at core of copyright; effect on value of work = new product/incentive balancing test used in Microsoft; some US courts consider defendant’s alternatives to the use = indispensability. Major differences: fair use potentially applies to all works, while EU only applies to exceptional circumstances/works of monopolists. And fair use, again, means no license fees.
Institutional point about process: in most jurisdictions, antitrust is predominantly the business of the government. In the US, most antitrust litigation is private—9:1 ratio of private to government claimants. This makes it a tool for many IP defendants. In Europe, the ratio is different—very few private cases. This changes the dynamic of litigation, even if the chance of success of a counterclaim is low. Wide availability of antitrust constrains IP owners’ power in ways not available in EU. Even if public enforcers are more aggressive in Europe, antitrust may be more powerful in US. (Which would say something about weakness in Europe, because it’s pretty weak here.)
Hypothesis: to the extent the US has been more innovative, and to the extent the law matters, the less is more approach of the US might be one of the reasons.
Q: Maybe that applies to copyright, but not patent. Several patent cases in the EU; US patent has no fair use; refusal to license doesn’t count as misuse in patent; etc. Also, plaintiffs here lose all their antitrust cases. As a substantive matter, there’s a divergence—the EU is allowing merger/monopoly cases that fail in the US.
A: True, he’s more familiar with copyright. But all the European cases, the language refers to IP generally but they were about copyright or whatever rights Microsoft was claiming, which is unclear.