IPSC Breakout 4 IP and Property/Contract Theory

Shyam Balganesh, The Eunomics of Intellectual Property

Lon Fuller’s institutional design theory: understand institutional design as a way of understanding institutions on the understanding that they’re not infinitely mutable. One shouldn’t begin with an unilateral approach on their ends. Ordinary assumption: institution is trying to maximize some goal, so tweak construction based on that goal. Fuller says this is problematic. We need to understand institutions as means-based. The means are not fully contingent, but themselves add important values and are not infinitely mutable. Eunomics is that project.

Don’t understand contracting, mediation, adjudication as ways to maximize other variables solely. They emerge from ways humans desire to interact with each other in society—conventional morality. That interjects not just path dependence but some kind of immutability b/c the normative goals of that kind of ordering interact are sticky. People desire that contract be in some part about human relations. Means/forms themselves contribute important normative goals.

Second: inner morality. Often described as having to do with morality & general jurisprudence, but Fuller had a simpler understanding: forms of ordering have their own internal integrity. There are elements in a form which remain immutable such that if you take it away, it ceases to be that form. Take away the impartial judge + adversarial system = not a judicial system. Take away reasoning = ceases to be adjudication. A commitment to integrity in form.

Third: The forms of ordering in society around resources and interactions are necessarily limited, not infinite. Limited set allows humans to channel and recognize how they want to order their lives. The mechanism of adding to the set is slow and evolutionary. There is a strong resonance to numerus clausus.

We can apply this to IP. There are three or four major common forms of ordering IP around which different regimes come to be structure. Newer kinds of resources may hybridize, but they’re defaults. Copyright, patent, trademark.

(1)   Appropriation form. Focuses on nature of wrong, interpersonal relationship of appropriation that doesn’t begin w/identification of resource but w/interaction deemed to be a wrong. That generates right/duty obligation. © is paradigmatic, but expanded to ROP where an intermediary doesn’t have to define a right in advance.

(2)   Grant-based: patents, property, prospects. The resource mediates the interpersonal relation. An impartial third party is supposed to delineate the resource.

(3)   Use perfection form: focus on behavior of entity that can develop over time a right that is perfected through registration or recordation—trademark.

The list is closed, but not closed at all times. It evolves and hybridization is possible. The forms themselves, the defaults, are conventionally determined b/c they come with their own underlying normativity. The means of institutional forms are not necessarily contingent and bring their own values.

Zachary L. Catanzaro, Algorithmic Deadhands: What is Dead May Never Die

Idea of AI trustees that enforce rules after individual’s death. These aren’t conscious systems. Try to get “bounded optimality,” not exactly replicate testator’s intent. Cognitive processes can be extended after death through technological artifacts. We could train a system to monitor for death records and make corresponding dispensations. We could try for active monitoring of beneficiary behavior: A to B so long as B never smokes. That requires more surveillance (Zuboff’s Surveillance Capitalism). What about “as long as B lives a healthy lifestyle”? We could train AI to predict probable intent by looking at data about O’s behaviors. We could exercise more control, in ways that are often bad—genetic or political profiling. Coercion of behavior.
In property law today we teach there’s a balance b/t free alienability of property and honoring testator’s last wishes; there’s conflict in the case law, but there’s a pendulum between formalism and not.

Allowing AI systems to predict and control the present creates new digital mortmains; emerging risk of reinforcing this type of property control. Convoluted way of circumventing abolition of fee tail/restraints on alienation/longstanding rules of property law that prohibit conditions on who a person can marry, what religion they can be, etc. Too much control. Risk of model collapse as systems self-iterate.

Possible solutions: reinforced privacy protection, anti data scraping laws, temporal limits on behavioral data use, etc. Further research: deeper into law of perpetuities/trust law. Philosophy of mind and causality for automated systems. Transhumanist applications. Any hard rule programmed in systems to limit them will necessarily be formalistic in nature.

James Grimmelmann: what would the legal system do in response to one of these dead hand systems? The traditional legal rule would be to void an unenforceable condition. Is that sufficient for a system like this?

A: We do need another form of intervention. The problem w/ relying solely on property law is that most states have weakened dead hand control to the point of meaninglessness—a Fla trust can last 1000 years.

Jennifer Rothman:

[Along with all these other problems, I just think about “my abortion is different”—O’s behaviors and O’s beliefs about what others should do may well differ.]

TJ Chiang, IP Ownership and Penalty Default Rules

Patent v. copyright ownership defaults: employer ownership of employee creations is WFH in © but patent, employee owns absent assignment. Every contributor to conception is automatically a joint inventor; © makes primary author sole author absent agreement. Patent = no duty to account, every coowner is an owner who can practice without accounting; © is different and every coowner must account to others (absent agreement).

Penalty default concept can explain the differences. Patent prosecution is expensive and legally complex, so patents tend to involve at least one legally sophisticated party and higher stakes, making it suitable for penalty default rules. © is automatic on fixation and easy, not suitable for penalty defaults. Criticism of “penalty default”: these aren’t penalties, just majoritarian defaults and conditions for contract formation. Rules that force parties to reveal info to courts are very different from rules that force one party to reveal information to the other; they function quite differently.

His theory: penalty defaults incentivize disclosure of info to courts. Info asymmetry between parties is not his focus. To be efficient, a penalty default requires: a party being penalized to be legally sophisticated enough to know about the rule. Benefit to courts in saved costs must outweigh transaction costs in contracting around the rule; social loss of bad outcome if/when parties fail, and cost to devise and apply the penalty default rule. There must be enough privately at stake to give sufficient incentive to contract around the penalty default. Should create a high private loss but low social loss if bad outcome eventuates.

Jeanne Fromer, First Ideas

First is polysemic. Interweaving of earliest in time and best. We mix these up; it’s important to understand the normativity of choosing “first,” and think through what we’re trying to accomplish with different forms of IP to allocate rights.

General rule in real property is first possession. Pierson v. Post as most famous American case (or McCulloch!). Pursuit isn’t enough; possession is the key. In patent that is reduction to practice, which can be actual or constructive. This creates pressure to file early, especially given statutory time bars after disclosure.

When we think about patent: you have to have actually created every embodiment you claim w/in the scope of invention—not necessarily physically but constructively. If you say “I invented all light bulbs with filament starting with letter T” you have to have constructively enabled tomatoes, thistles, etc. First is fictional in many ways, and this is a theme. You might have done some subset but not a full scope and we’ll deem you to have done that.

Things are also carved out of first—law of nature, abstract idea, natural phenomenon.

Some rules are about not claiming first too quickly: utility rules—there has to be a specific benefit for the invention. Novelty/nonobviousness; getting beyond prior art—if not, no one is first. Continuations and doctrine of equivalents also affect firstness. And people who are prior to first—prior use—are carved out as well.

Copyright is more generous in allocating firstness b/c standard is so low. In some ways © rejects firstness because it doesn’t matter if others are doing the same thing.

TM: use and registration differ in source of firstness; legal fictions like national use also extend firstness. Lack of rights in gross mean that firstness is cabined in other ways.

Real property critiques of firstness—appropriation/expropriation of prior peoples—can also influence IP. Expressive incentives matter as well: firstness as incentive? What about first losers? We don’t have violence in the same way as in real property but we do have power dynamics, like first to file being easier for big entities.

Grimmelmann: there are a lot of similar ambiguities in real property—disputes over the scope of what people have acquired title to. Biggest modern survival is adverse possession; always a question about whether you possessed the whole or only part of the parcel. More continuity than contrast. (I also think about the work of people like Claire Priest and Gregory Ablatsky’s The Rise of Federal Title, which emphasize that constructive possession/allocation from regional capitals was actually a huge driver of US ownership in contradiction both to native title and to small colonizers.)

Mark McKenna: key to emphasize that first is not found in nature. Many disputes: Not trying to decide who’s first; which of two people beat the other to it. Should that make a difference? (Relativity of title!) Also, consider reverse confusion, which isn’t really a confusion doctrine but a priority doctrine. Sometimes we have to pretend there’s confusion to honor the first use or it will be wiped out.

A: We’re often allocating rights and making claims w/o having to think about other parties in dispute. Maybe we shouldn’t as much—wait more until there’s an actual dispute. Reverse confusion was one of the things that got her into this project.

Matthew Murrell, Cyborgs and the Law of the Human Body

Dick Cheney had the wifi for his pacemaker disabled because of fear of cyberattacks; implants have been with us for a while, especially cochlear—eye implants are on the way. 3 million Americans have implants now; this will grow exponentially due to brain/computer interfaces.

Common law: wooden leg is property not person, and no compensation can be awarded for injury. Some states have updated worker’s comp but still treat the artificial body parts as property, not part of the body.

Neuroscience: artificial parts of body are unquestionably part of the body: our brains map our bodies and are constantly updating that map. Phantom limb syndrome: body schema includes lost arm. If a prosthetic is used, the brain adapts—tooth fillings, artificial hips, artificial limbs. Interesting things at edges—in virtual spaces, people can map onto 3 foot long arms, but not lobster claws—but the brain is adaptable. Our body is our body schema in the brain. It is wrong to treat them as personal property only.

One path forward: default rule is that cyborg bodies should be treated the same as corporeal bodies. Privileged lived experience of individuals with atypical bodies and individuals with modified bodies. Individuals with modified bodies are going to use cyborg bodies to express themselves—an important site of political expression. Cyborgness will be immaterial in the normal run of cases—worker’s comp, tort.

Exceptions: capability of a particular body part. Human ability=normal vision. Superhuman ability=hawk vision 20/2. This may be material for, e.g., intrusion on seclusion. Extra-human capability=ability to see infrared light; police officer with this would still have to follow Kelo.

Corporeality: the further a cyborg body drifts from the human form, the more it might warrant departure from the default rule—people may be able to operate spaceships.

Severability: wearables are getting more sophisticated; we may need to distinguish wearables from body parts.

Andrew Gilden: Haraway’s Cyborg Manifesto is the first thing he thinks of—we’re all sort of cyborgs. Worry a little bit about overly focusing on neuroscience and the nervous system as defining what’s a legitimate body. How does this apply in the context of gender-affirming surgery? Advanced wheelchairs?

A: The full paper addresses some of this—social theory and empirical work on virtual spaces—when white subjects are given Black avatars, their scores on implicit bias tests go down for a while. Struggled with how to constrain the definition of cyborg, but trying to clamp down on problems—more expansive you get with the word cyborg, the more it includes wearables. Optional additions will be a part of this explosion.

RT: Peggy Radin’s concept of property—why shouldn’t removable property be part of our personhood? Maybe different treatments in tort and accessibility law are appropriate? How much does it matter if our minds map onto the body part? High heels can be really important to a person, but not worn all the time.

A: countervailing: in tort, the idea of the body is hugely important and expensive—you may be committing battery if you strike a horse and its owner is riding it.

Q: also consider product liability regimes. Strict liability/design defect.

Jeremy Sheff, A Heap of IP: Vagueness in the Delineation of Intellectual Property Rights

Vagueness and notice: a pervasive concern in IP. Vaguely delineated IP rights are worrisome b/c public won’t know what it can do. Patent: specifics vital and a goal of the system is to provide specificity. Copyright=exemplar of work is the work makes the claim inherently vague; TM—vagueness attends both registered and unregistered marks, especially trade dress.

Can philosophical literature on vagueness help?

Sources of vagueness: at least two dimensions. First: qualitative. IP rights are not really in things, they more resemble rules. The property in IP is inevitably a category of possible things.

Second: composite nature of IP rights. IP rights/claims have elements: limitations of patent claim; features, particularly protectable features, of © work, and components of a trade dress. Although they might not be explicit in a claim, in an infringement case the first thing we do is ID the elements and look for them in the accused embodiment.

Qualitative vagueness turns on imprecision of natural language: “no vehicles in the park” doesn’t necessarily tell us what’s inside and outside the rule. More broadly, the idea going to philosophy of language that we have cause to be skeptical that any rule can define in advance its application to all situations. Even with nonlinguistic claims like © and TM, those types of claims often get translated into linguistic terms for purposes of analysis b/c that’s what legal systems do.

Composite nature: there’s another dimension of vagueness, the problem of the heap. The sorites paradox/the Ship of Theseus. How many elements can be removed/replaced before it is no longer the same thing? With patents, all elements are required—unless the doctrine of equivalents applies! Figuring out how many elements we can take away before we’re not dealing with the same thing is a central challenge. How many hairs can be taken away before a person is bald? People disagree, and the predicate operates on a continuum rather than hair/no hair.

Vagueness can be multidimensional—qualitative and quantitative vagueness can occur at the same time and any given element can also be vague, for infinite regress. Best Cellars TM case: what is the trade dress and how many things don’t have to be there before it’s not infringing? If the circles for the wine were hexagons instead, would it be infringing? What if the wood was dark instead of blonde?

Philosophical approaches may not be available to legal systems. One response: epistemicism: there is a sharp boundary, but it is unknowable. Or we could reject bivalence: there is a range of cases for which the predicate is neither true nor false—or both true and false. Legal systems can’t do that—we need an answer, infringement or not.

Responses to qualitative vagueness, one approach is precisification: change the natural language meaning of the term to a quantifiable technical meaning. But multiple precisifications of a vague predicate may be plausible; which should we choose? One approach is asocial: we categorize the range of precisifications and try to determine truth values in certain situations. Narrows the zone of borderline cases but does not eliminate them.

Social approaches: associated with crude Wittgensteinian reading: meaning of vague term is only determined by community’s use. That’s what IP does: allocates authority to community to figure out whether this embodiment does or does not infringe a right.

Interesting things about IP’s solution: (1) community differs across the range of IP. Patent: PHOSITA. Copyright: public. TM: ordinary consumer. (2) Even though the law creates privileged communities, decisionmakers who resolve vague rights claims often come from outside the privileged interpretive community (except in ©). Choices are not about eliminating vagueness but about defining who gets to do it. That tells us a lot about who the system is for.

McKenna: how does the overall scope question get allocated? When a court decides something is a valid trade dress, the court is deciding what the elements are, including thinking about functional elements, then throwing the rest of the decision to the jury. Even in ©, idea/expression or merger has also done definitional work.