Patent Solipsism

I’ve been thinking a fair amount about software patents the last couple of weeks. I recently attended a Brookings Institution conference that focused pretty heavily on software patents, and since then I’ve interviewed several sharp patent scholars in preparation for a forthcoming article. In those conversations, I noticed the same cultural gulf I blogged about on Techdirt last week. You might say that on the subject of software patents, lawyers are from Mars and programmers are from Venus.

I think there’s a universal human tendency to over-estimate the importance of whatever you happen to be an expert on. I know lots of geeks who believe everyone and their grandmother should use Ubuntu, vi, git, RAID, and so forth. A lot of economists believe that the rest of the social sciences would be better off they all started using the methods of economists to do their jobs. When we develop human capital in some particular field, we tend to get a corresponding emotional investment in that field.

So when a programmer thinks about software patents, he’s interested in improving the software industry. Given how screwed up software patents are, a lot of us go straight for the most direct and elegant way to accomplish that objective: excluding software from patentability. In contrast, when a patent lawyers thinks about software patents, he’s interested in fixing the patent system. From that perspective, abolishing software patents looks like a horrible hack, because the underlying problems that caused software patents to be such a mess are probably responsible for problems in other industries too.

Patent scholars tend to view the underlying industries that operate under the patent system as homogenous and interchangeable. From the perspective of the patent theorist, it doesn’t really matter if a patent covers an impotence drug, a one-click shopping website, or a method for managing the consumption risk costs of a commodity. They’re all “inventions” or “technologies” that a patent “protects.” And obviously, if your starting assumption is that “technologies” can be bought and sold like chunks of pig iron, then it’s going to seem weird and arbitrary when people suggest that certain categories of “technology” should be excluded from that process.

The problem is that in the real world doesn’t look very much like the patent theorist’s model. Actual patent-eligible industries are vastly different from one another, and it’s crazy to just assume that patent rules that work well for the pharmaceutical industry will also work well for the software industry. But of course, to appreciate this point it’s helpful to know a fair amount about the underlying industries. And if you spend a lot of time learning about particular industries, then you run the risk of looking like a business or computer science professor, rather than a law professor.

When you’re wearing your law professor hat, you fix the patent system by suggesting tweaks in patent doctrine. I’ve heard a dizzying array of proposals for fixing the patent system, ostensibly to fix the problems that have cropped up in the IT industry. These include changing the rules for obviousness, definiteness, and enablement, having more specialization in the courts, having less specialization in the courts, hiring more examiners at the patent office, giving judges more training, raising fees at the patent office, reducing damages for infringement, switching to a “first to file” system, limiting forum shopping, making injunctions harder to get, establishing an independent invention defense, and probably a lot of others. Each of these proposals are supported by lengthy, erudite law review articles laying out exactly how and why their preferred reform would improve the patent system’s performance.

I don’t doubt that many of these proposals would improve the patent system on the margin. Indeed, I support several of them. But they strike me as bizarrely circuitous ways to deal with what is really a pretty straightforward problem: patents aren’t well-suited to the software industry. Of course, patent scholars tend to object that defining software patents is difficult, and indeed it is. But patent law is chock full of hard-to-define concepts. No serious patent scholar would say that because obviousness is hard to fine, we should just stop trying to invalidate obvious patents. And a “no software patents” rule at least has the virtue of having an obvious and direct causal link to the software patent problem, which is more than can be said for most of the reform proposals now before Congress.

SHARE: 11 comments