Sunday, May 11, 2008

STS and IP: Patent narratives

Session 5: Papers by Henry Huang (“Reevaluating Science in Patent Law: Lessons From the Historiography of Science”) and Dan Burk (“Patent Law’s Black Box: How Intellectual Property Aligns Creative Networks”)

The usual patent-related caveats apply: not my field; I may have missed rather a lot.

Huang: His project is to look at the insights history of science has for patent doctrines and possible reform. What happened after Kuhn's The Structure of Scientific Revolutions? One article proposes that patent law validates inventions that themselves validate theories—patent systems force scientists to create paradigm shifts by theorizing around prior art/patents. But how that interacts with other kinds of scientific advances/publication is unclear.

Other legal literature uses Kuhn to analyze evidentiary reliability—from Frye to Daubert. The courts’ emphasis on falsifiability tracks Kuhn, but that’s no longer the prevailing view in history of science on how changes in science come about.

Patent law turns often on the moment of invention—e.g., priority disputes; reduction to practice. But you can’t sometimes isolate a particular time or person to whom you can attribute an invention. In a single paper from a major lab, there may be hundreds of scientists who contributed to the work. How can we say one person/one small group is responsible? One possibility: create a legal concept of an inventive entity—that would insulate them from some challenges to inventorship. By allowing people to sign on and subsume themselves in a legal entity, we would recognize lots of people as inventors and provide them some insulation of people who’d been left out.

Prior art: teaching, suggestion or motivation as the test for obviousness. KSR represents a big change—before, two references that together suggested the invention wouldn’t have made it obvious; now it might be obvious, in a sort of common-sense test. If there’s no discrete moment of invention, then KSR is a better to fit to actual practice; it’s certainly better than requiring some sort of flash of genius.

Burk: Some not-so-serious advice: The way to succeed in law is to find something from another field that was hot 30 years ago and teach law professors about it. He’s interested in what Bruno Latour has called black-boxing, how it fits into patent law, and why that might be a good thing.

The dominant rhetoric of IP has been economic. But the theories leave some puzzles. Moribund patents: people spend lots of money to get patents, then just sit on them. Why? Perhaps used to attract venture capital and financing, rather than to license. But then why do VCs like patents? Clarissa Long suggests they might signal savvy management. But if that’s a signal of business ability, it can’t be a very good one.

Jessica Silbey has started writing about narrative theory and patents (which I thought of when Huang was presenting)—to win a patent case, you need to tell a story of genius that justifies a plaintiff’s verdict.

Some patents get people very upset—tax patents. But people don’t get upset about patents on new weapons that kill, or on technologies that will disrupt lives.

So can actor-network theory (ANT) tell us about the social work patents do? There’s a period of dispute before scientific results are accepted; how do results become uncontested? There’s a process of negotiation and recruitment of people/resources. ANT doesn’t distinguish between human and nonhuman actors. Lots of strategies are used to corral unruly actors (the equipment that breaks down/produces weird results). Many institutions are mechanisms to align the unruly—peer review; citation practices.

Black-boxing: when things (human or nonhuman) are lined up and behaving, controversy shuts down and we stop thinking about an issue. It becomes a black box. Once the double-helix structure of DNA is accepted, it becomes an uncontroversial conceptual object that’s used to create other concepts, but we ignore all the support/inputs that went into creating that concept. Burk’s view: this is a sociocultural way of quieting title. The statute of limitations has run and we aren’t going to dispute this claim any more.

For patents, there might be a lot of prior references to think about in deciding who was an inventor. Patent has a lot of doctrines for navigating questions of identifying an invention and an inventor. We use the date of invention, the date of going on sale, the location of things overseas, the types of references, and so on to cut off certain inquiries. We also have occasional mechanisms for opening the black box—reexamination; challenges to whether people were left off the patent.

Why do we do this? ANT talks about stabilizing particular cultural artifacts, and that’s what patent law can do.

Possible case study: The Edison light bulb, which ANT has already examined. The 1880 and 1881 versions differ a lot; they’re both patented. Prior art is really quite crowded. Some very similar versions were excluded from consideration because it was from overseas. Moreover, Edison relied very heavily on his staff. There were also unruly materials involved—Joule’s law made copper too expensive to use, so he needed to find another way. Etc.

The patents themselves turned out not to be very helpful. Edison announced the invention a year before he had it; then he was engaged in a lot of litigation over the patents. Why patents then? Latour argues that certain props are necessary for certain roles—the scientist and the test tube. Perhaps you can’t have inventors without patents. Patents validated Edison in the public mind.

Steve Barley: The problem seems to be that if STS is right, then the concepts courts are using to resolve patent disputes are wrong: the fiction of the moment of invention; the issue that people tend to be working on the same problem simultaneously so there may not be a single inventor.

But it’s worse than Huang and Burk say. STS focuses on science, rather than technology. Science and engineering are not the same thing; they have different cultures, social structures, and modes of operating. STS has favored science over engineering because the move to concern with epistemological issues in practice first occurred as sociologists studied scientific labs. Studies of technology tend to look at how designs come into being, not the practices of engineers. Only a tiny fraction of patents pertain to the output of scientific activity; many more pertain to the output of engineering activity (though both may be small fractions of patents as a whole).

Differences in social structure: Citations in a science paper are skewed to publications within the last 5 years; citations in an engineering paper are essentially random as to time. Scientists get ideas from literature and colleagues, who are organized into invisible colleges. Engineers get ideas from vendors, who have access to what’s going on with competitors; from customers; from coworkers who’ve come from other organizations. It's a different idea-spreading system. Science is reading-based; engineering is an oral culture.

Science tends to be physically dispersed, while engineering is physically concentrated. Most engineering patenting is done by organizations, even if they’re forced by the law to put names on it.

He’s not a big fan of ANT. As an American pragmatist, he thinks some STS literature gives actants/nonhumans too much agency. And black-boxing is not necessarily a bad thing. (Biagioli intervenes to say STS doesn’t think black-boxing is bad!) Black-boxing is useful for doing certain things.

Mike Madison : For patent litigation, there are all kinds of avenues to open the black box. The artifact’s stability ebbs and flows. It can be settled for some purposes and contested for others.

Huang’s examples of doctrines that might be changed all involve ways to wrap up a complex history and move forward, because it’s simpler to do so for administrative purposes. Applying STS doesn’t give you a specific policy outcome; it might even push you towards more black-boxing.

Burk: Only lawyers would ask, “why do we care?” Other disciplines would say that more knowledge is good, even if we don’t end up wanting to change the policy. Among the things he’s interested in is whether a patent assists in the black-boxing process. STS generally treats black-boxing as a neutral social phenomenon, but some of it does suggest that black-boxing is intellectually dishonest—denying Rosalind Franklin credit for DNA, for example.

Huang: It’s important to remember that the dominant discussion about patents is economic language—incentives, prospecting, etc. STS provides a new way in.

Burk: There’s an interesting discussion to be had about Daubert and Markman—the opinions read together.

Lemley: In his role as the oppressor/hegemon, it seems to him that the question of whether black-boxing is good is necessarily bound up with the question of why we’re bothering to ask. Why care about who the inventor is? The economic story is one that incentives ought not to be misaligned. Outside economics, we need a story of why it matters to patent law that one person is named the inventor and the other isn’t.

Burk: Economics is like the Borg, attempting to assimilate any discussion into incentives. He thinks that’s a cheat, but it can be done. If you don’t want to do it, ask: If there are gaps/lacunae in the economic story, what’s going on underneath? The economic story coincides with reality sometimes, but it’s a myth. Sometimes myths are good; Schliemann found a city where Homer said it was, but that doesn’t make the rest of the Iliad true.

Samuelson: She’s interested in the notion of an inventive entity and the move away from the romantic inventor. Why doesn’t patent law have a work for hire rule, like copyright? Patents are increasingly corporate. A work for hire rule would allow giving credit without a connection to initial ownership, which really isn’t relevant anyway. The work for hire concept is its own black box, but a different one.

Strandburg: The romantic inventor might work better for scientific papers than patents; maybe the problem is applying the same rule across different endeavors. Also, we could maybe think of patent problems with software, biotech, etc. as a failure of black-boxing: various components refuse to go in the box and let it close behind them.

Burk: The citation strategy in a scientific paper is very different from a citation strategy in a patent application. The former situates and extends, while the latter attempts to differentiate.

Cohen: Patents are at the intersection of science, engineering, law, and finance. Black-boxing in STS was about science, but each discipline has its own ways of doing black-boxing, and patent law is unstable because these ways collide. Businesspeople have to value patents and disclose them to the SEC; it’s uncomfortable to put a value on incomplete assets, so they just call them complete. If the theory is to be powerful in patents, you must excavate the different layers.

Gillespie: There's an analogy to work on “art worlds”: there are intersecting circles of people forming around objects, and boundaries are always being drawn strategically.

Another point: the issue isn’t just different black boxes, but that they open and close at different points and for different reasons across fields. So should patent law attempt to be more like science/engineering in its black boxing?

Lemley: We may be perfectly happy to have finance people treating patents as certain and closed objects for valuation purposes, but the problem is crossover—if people start treating the patent grant as a property entitlement for injunction purposes, you may get pernicious consequences.

Cohen: No, finance behaviors matter, because it comes back into legal reasoning when people start arguing about why they obtained patents.

Biagioli: Latour’s black box is different than what you’ve been saying. The standard story: claims are unstable; they are stabilized by piling social resources on them; the box closes and is padlocked; it can be opened with more resources. Latour doesn’t believe in this. His black box is more like an electromagnetic field; as soon as you turn the electricity off, the thing collapses. He stresses fragility: as soon as the alignment gets disturbed, it falls apart. His argument is about techniques of alignment. It says that nothing is safe.

Latour would be interested in techniques: how you standardize technical drawings so that they can be shared. How specific patent doctrines emerged—patent law as a technology.

1 comment:

Michael F. Martin said...

A patent law without patent lawyers:

I'm glad that the discussion sussed out how the distinction between inventing and artistry is not as wide as some would have it.