Intellectual Property – Interpreting the Scope of IP Rights
Moderator: Zahr Said, University of Washington School of Law
Margaret-Jane Radin, The University of Michigan Law School: Patent scope. The problem of describing innovation (thing in the world) in words. Philosophical problem of extensive lineage; judges often don’t know it’s a problem. Related: problem of big picture economic efficiency. If that’s what we want, what are patent rationales? Incentivizing/ex post—coordination of future innovation, signaling market. Recent discussions about notice: notice to public & competitors is important, but patent claims are in words and thus tempting analogy b/t notice in words and fences is flawed. Longstanding puzzle; can’t be solved with analogies or advice to be clearer in claim drafting.
Costs/Benefit balance of large rights may change over time, and that flexibility may be a good idea. Dilemma b/t calibrating rights properly and trying to make rights clear, and both are parts of the efficiency calculation. Some judges gravitate intuitively to one pole or the other. Even with philosophy the dilemma would still be there.
Philips v. AWH: Fed. Cir. en banc. Interpretation focused on entire text: claim, specification, and prosecution history. They arrive at unified guideline, but interpretive guidelines are only useful for notice if they generate more predictable outcomes, and it didn’t. Judges don’t all get the same result—the majority agreed on the interpretive standard, but disagreed on the mandated result in this particular case. Suppose something was indescribable at the time of the patent, but not at the time of the litigation—is it covered in the patent? Can be. But who is doing the describing? Usually the attorney, not the inventor. But that’s an issue the SCt doesn’t go near; we don’t want anyone to think that drafters are co-inventors. Festo introduced the word of describability, but older cases often understood the idea of emergent inventions for which language develops later. The idea of the essential nature of the invention persists; a survival of central claiming, which is supposedly gone.
Mark A. Lemley, Stanford Law School: co-authored paper with Mark McKenna. We generally divide IP into validity doctrines, infringement doctrines, defenses. Result: We apply different rules at different times and sometimes different actors: judge, jury, PTO. Fact that we’ve divided IP into different pieces creates the “nose of wax” problem: you say at T1 that a patent is really broad when it benefits you to do so and at T2 that it is really narrow when it benefits you. It produces bad result because courts don’t at any time take account of the scope of an IP right. We ask “is this IP right invalid b/c it’s too broad” and “is the thing D is doing sufficiently similar to what P is doing” but generally not in an integrated proceeding “is the thing D is doing that P is doing the thing that can be controlled under this right?”
Part of this: (1) we generally allow fragmented infringement. We can point to some sub-piece of product or book that’s sufficiently similar. (2) We’ve also expanded subject matter to cover things that are supposed to be protectable only in part—utilitarian articles for ©; product configuration for TM. But (3) jury has taken on increasing role in resolving IP disputes. Often delegate legal decisions about proper scope of right to the jury in the guise of fact questions about similarity, and often with no guidance to the jury about that.
Example: design patents now allowed to cover prior art and functional aspects. We used to look at prior art, design patent, and accused infringer together, but now we’ve separated them. Said functionality is extremely narrow; limited/narrowed prior art. But then allow the patents upheld under these standards to go to jury under ordinary observer standard. That’s fine if the thing that makes the thing that makes it sufficiently similar is the thing we want to protect. But if the thing that makes it sufficiently similar is prior art or functional, that’s bad. Most recent example: Apple/Samsung design patent case, where the key similarities are also in prior art, but jury doesn’t see them at the same time. Same problem occurs in trade dress cases—Reynolds v. Handifoil case. They are different in many respects, but court finds striking similarity: both boxes say heavy duty, nonstick, Made in USA, and list square footage in the box. Those are similarities, but not similarities in what the law is supposed to protect.
© has similar problems, even though it at least purports to filter out unprotectable elements. Right now we filter for actual copying, but allow jury to load them back in to determine whether copying is so great as to be unlawful, even though that risks jury relying on those unprotectable elements. “I need somebody to love” as lyric whose reappearance is enough to go to the jury in Bieber case; also infects jury in Blurred Lines case where jury hears songs even though it’s supposed to focus on the musical work and ignore intentional but unprotectable stylistic similarities in performance.
Court plays blackjack: IP owners want to get as much as possible, but if they claim too much the right may be invalidated. This is good for gaming, but not good for IP, where we really want to get scope right. Markman has problems, but it’s good b/c it makes parties go into a room and come up with a scope—what the patent covers and what it doesn’t—and that limits the nose of wax problem. We could do more explicit scope proceedings. That would cause the court to focus on what we should be thinking about when we think about IP: how much protection does the law intend to provide, and are you seeking more than that?
Eva E. Subotnik, St. John’s University School of Law: Derivative works: what are they? For a time, appellate courts have avoided interpretations bearing on scope that ask for anything quantitative, as opposed to qualitative. A sea of discarded tests. Examine the roadkill to see the path we’re on. (I have questions about this metaphor.)
How much of the plagiarist’s work the plagiarist didn’t pirate: that would have suggested a brick by brick method.
Feist: Downplays both sweat and binary copied/not copied distinction in favor of a more general approach. Feist decides validity but is still relevant; QP was about scope, and also case sent the tone that clear-cut metrics were out and qualitative metrics were in.
Similar issues about what’s required for a derivative work. Gracen: 7th Cir. required sufficiently gross difference b/t derivative and underlying work to render the derivative work protectable to avoid entangling subsequent artists in problems. “Gross” signals a quantitative method. But more recently, in Schrock, the 7th Circuit capitulated and said it would apply a unitary standard. Now we look for sufficient nontrivial expressive variation to make it distinguishable from the underlying work in some meaningful way. Reaffirms qualitative analysis. Puts a lot of pressure on infringement analysis to ensure the scope of rights in underlying work isn’t unduly curtailed.
Fair use: Cambridge U. Press v. Georgia: DCt used a rule of thumb about amount used: 10% or one chapter where book had more than ten chapters was presumptively fair use. Could wrongly signal to authors that use of more than 10% would always be protected, but that wasn’t the 11th Circuit’s problem—held that improper even as a starting point, b/c case by case/work by work approach was required under Campbell. Cariou v. Prince also downplayed quantitative approach/intention of artist and said transformativeness had to be judged by reasonable observer. But the basis for remand of remaining 5 works was incredibly murky and provided no guidance to judge.
We’re seeing some pushback in quantitative thinking about scope. Garcia v. Google en banc: Fleeting performance on film, which would bear on filmmakers’ © scope. En banc majority referred multiple times to brevity of performance, but also referred to smallness of claims in its policy analysis—cast of thousands would become © of thousands. Cariou also did refer to intent in terms of Prince’s “drastically different approach,” and also did remand on 5 works which did seem to be influenced by how much he “took” and didn’t change. 7th Circuit’s retreat from transformativeness in Kienitz also may represent a retreat from qualitative considerations.
This retreat is a good thing: more transparent. [Not sure the decided cases bear out this transparency. E.g., the Georgia State district court approach was much more predictable as a rule. The Garcia approach is unpredictable if considered as a rule about amount rather than as a rule about performers versus filmmakers. Cariou is, as Subotnik rightly notes, “murky” in its rationale for drawing a line between different prints (a) at all and (b) where it does. And Kienitz, ugh.]
Amy M. Adler, New York University School of Law: Why transformativeness has proved so disastrous in the realm of contemporary art, even though hailed as a savior elsewhere. Requires a court to adjudicate new meaning/purpose, which is a failed enterprise. Three different ways to find meaning: intent; aesthetics; the “reasonable” viewer—each is deeply problematic in assumptions about contemporary art, b/c the assumptions are rejected by contemporary art itself. Copying is now a basic tool of art.
Against intent: freeing art from the artist. Prince (for whom she consulted) was a perfect case where an artist disclaimed any intent to transform the work. Intent rose in importance out of the Koons cases; Jeff Koons learned how to testify in a way that courts liked about his intent to offer “new insights.” Intent is bad, among other things, because of the difficulty of describing images in words—familiar in First Amendment, cultural theory. WJT Mitchell: “Whatever images are, ideas are something else.” Richard Serra, Tilted Arc case—one reason he lost was his inability to describe the meaning of the work in a digestible way. Courts like words b/c it’s familiar, old-fashioned, romantic idea of art. SCt cited Jackson Pollock as artist—but we understand the work as an outpouring of the artist’s soul in a moment of expression, but that’s an old-fashioned way of thinking about the relationship b/t intent and meaning. Many artists reject it. Andy Warhol, asked about meaning: “Why don’t you ask my assistant Gerry some questions? He did a lot of my paintings.” Embracing artist’s lack of control over his own meaning. Technology may have co-authorship.
It’s very hard to figure out intent given that authorship is quite multiple: documentary as example. Famous image won Pulitzer Prize of naked girl on fire from napalm; it was cropped carefully from the image he actually took (cutting off a photographer whose presence in the frame raised uncomfortable questions). Photographer didn’t know this was a key photo; an unnamed editor picked it out of the roll of film and cropped it. He still doesn’t know why this image was picked. Picking images is now a key skill in our digital culture. Shepard Fairey: say what you like, but he knew what image to steal.
Also has similar arguments against aesthetics and against the reasonable observer in the broader paper. Has begun to believe in Kienitz: look strictly at the market, which would be more protective of contemporary art. [The racial and gender politics of this move trouble me, given who is more likely to get recognized as an artist by the market.]
Kevin Emerson Collins, Washington University in St. Louis School of Law: Patent scope should be both/and not either/or: scope is a philosophical issue, and understanding that can allow us to attune patent scope to the world of commerce.
Philosophy of language: what is meaning of “meaning”? Distinction between denotational meaning/reference and “sense.” Meaning found within word-to-world relationships is denotational/referential meaning. Look at all the things referred to when we say “dog” and the word “dog.” Ideational meaning/sense: words gain meaning through another mechanism, the concept of the mind in anyone who understands the expression. Meaning is found in word-to-word relationships (dogs have four legs, sense of smell, member of class mammals). You don’t actually have to connect words to things in the world to get meaning. Identification requires a two-step process: determine meaning of descriptive language, then determine whether the thing in the world meets that description.
Everyday usage: these meanings are interdependent; don’t need to clarify. But that’s not true in claim construction. In patent we fix meaning on a particular date: as PHOSITA would understand it on the date of filing (or invention, there’s confusion). This is an artifice that doesn’t exist in everyday meaning. So are we fixing the denotational or ideational meaning? That affects the ability of later-arising tech to fit in. To fix denotational meaning, you’d idnetify the set of possible objects/actions that PHOSITA can imagine on that date; then if new tech arose, it wouldn’t be within the set; that would change the “meaning” of the fixed-time term. But if we choose ideational meaning, you stabilize the network of linguistic constructs. You don’t have to fix the set of things to which the words refer—ideational meaning remains rigidly fixed as scope of things in the world expands.
Courts usually use ideational meaning, but they switch to denotational form time to time, when they believe that the after developed tech shouldn’t infringe. Should this be a policy lever for the courts? He thinks yes, at least if it’s explicit.
Temporal paradox: enablement requires full scope of claim to be enabled by specification as of th etime of filing. But Merges argues that meaning that determines infringement isn’t fixed until time of infringement. Fully enabled claims thus grow in literal scope over time. Thus claim meaning isn’t fixed at time of filing. Difficulties: Creates instability, lack of ex ante notice. Contra black letter law of claim construction. If we understand ideational/denotational meaning, we can resolve this paradox more simply. Enablement requires disclosure commensurate w/denotational meaning as of time of filing, but we allow ideational meaning to control infringement. Claims remain fully enabled even as they grow b/c denotational meaning didn’t change over time.
Said: what’s the expert’s role in these various approaches? If it’s a policy lever as Collins says, who determines that?
Adler notes dramatic difference in what expert would think of contemporary art and what ordinary viewer would think. Two images that are visually identical have dramatically different meaning according to art experts, but “reasonable” viewer might be different (reasonable compared to what?). How much do we want to defer to standards of the art world? Issues of elitism—over and underinclusiveness. People who aren’t famous: how do we deal with them? But she prefers experts b/c art world is pretty unreasonable.
Lemley: didn’t recognize in either description by Collins the way he makes meaning—if he sees a new animal, he uses an expansive or relational view closer to denotational meaning—does this look enough like what I know to be a dog? Central claiming-ish: how far is this from lodestar? Rather than is it within the universe of things that are already known in some ways to be dogs.
Collins: There is discussion about whether the set is determined by prototypes or full descriptions. But it’s unclear what we do with prototype/archetype theory of meaning when we’re trying to fix meaning. The only way to do that is to figure out what the criteria were that we were using at that time to identify members of the category. (Which may be a complex probabalistic assessment where high conformance with some criteria can be enough even when other criteria are not satisfied, e.g., what is a “game”?)
Lemley: or give up on the idea of fixing meaning.
Lemley expresses concern that under Adler’s market standard, the winners win and the losers lose—if I want to win, I’d better be recognized in the art market before I get sued.
Adler: the art market is brand-driven. Lemley won’t sell b/c he’s Lemley, not Prince. [I think that’s his point.]
Lemley: if filtration a la Altai were applied across the board instead of just to computer programs, the world would be a much better place.
Fred Yen: Isn’t the decision to look at the market value of a work an aesthetic decision? And in any event, don’t market participants have to use aesthetic concepts to set a market value on the work, so we’re just going back to their aesthetic theories? [I agree.]
Adler: the intertwining of economics and aesthetics is a definitional feature of the current art market. That’s a very deep issue; probably right.