Patents

I should probably say a bit more about the substance of Ben’s property rights post, which wasn’t primarily a critique of people with tacky property rights websites. Ben focuses on the conceptually sound idea that property rights are really bundles of rights to dispose of particular things in particular ways. He emphasizes the diversity of these rights:

Text and designs are obviously different from household objects, but we all knew that. In fact, automobiles are also different from household objects. So is land, and the houses built on that land, which differ from each other. Your kidney is also significantly different. Oh, and commodities such as corn, which differ from corn futures. Correspondingly, the rights associated with all of these things are different.

He lists a variety of different rights we can have in some but not all of these things: transfer, sale, use, modification, exclusion, etc. What I think doesn’t get enough emphasis, however, is the “things” part. That is, an automobile, a kidney, and a bushel of corn are all discrete, well-defined things with respect to which which one can have rights. Everyone understands what it means to say that I have the exclusive right to use, transfer, or sell my car. By their nature, cars have only one person driving them at any given time, and so there’s an obvious need for some way of deciding who gets to drive which cars.

Now, compare that to someone who “owns” a “method for managing the consumption risk costs of a commodity.” Ben’s right that we’ve got a bundle of sticks here just as we do with the car. Had Bilski’s patent been upheld, he could have kept the patent for himself or licensed, given, or sold it to others. The problem is that it’s totally unclear what the “thing” is to which the sticks in this bundle relate. Reasonable people can parse the language of the patent and come to radically different conclusions about the scope of the patent right, all of them plausible. A patent is not the ownership of a pre-existing thing that needs an owner. Rather, the things over which the patent system gives people ownership are the creations of patent law. And in many cases, it makes little sense to talk about them as “things” at all.

Unfortunately, the patent bar has developed terminology that papers over these difficulties. We talk of owning a “technology” or an “invention” in the same way we talk about owning a car or a bushel of corn. But in reality, a “technology” is not a discrete thing in anything like the same sense that a car or a bushel of corn is. If I build a website that allows my customers to purchase things with one click, I may technically be infringing Amanzon.com’s patent, but it strikes me as an abuse of language to say that I’m “stealing Amazon’s technology.” What many patents claim is not a specific “technology” or “invention” so much as a broad category of machines or processes that may be only loosely related to one another.

When we describe “inventions” or “technologies” as things that can be owned as property, that carries the implication that inadvertently writing software that’s similar to another company’s software is in the same moral category as stealing a car. I think that’s self-evidently absurd, and that adopting such morally loaded terminology impedes clear thinking. It’s important to remember that the “technologies” and “inventions” patents protect are more often loose categories of related machines or processes rather than sharply-defined entities. The language of property rights obscures, rather than illuminates, this point, and I think we’d all be better off if people stopped employing it.

SHARE: 8 comments

One of the most fundamental disagreements in the debate over software patents concerns the Supreme Court. Some software patent supporters like to cite the case of Diamond v. Diehr as the decision that legalized software patents. Many others argue that the Supreme Court’s classic trilogy of software patent decisions from the 1970s and early 1980s just weren’t clear enough to be of much use in the modern world.

In a new article for Ars Technica, I take a close look at these claims and talk to a couple of prominent patent scholars who make them. I find that like most Supreme Court decisions, the Benson, Flook, and Diehr decisions are hardly models of clarity. It’s possible to find passages in those decisions that could be cited in support of either side of the software patent debate.

However, I think that it’s hard to escape the conclusion that, at a minimum, the Supreme Court intended software patents to be far more limited than they are today. For the last decade, the Federal Circuit has been allowing so-called Beauregard claims, which claim software printed on a machine-readable media such as a CD. As I explain in the article, it’s hard to see a plausible interpretation of the Supreme Court’s precedents that would include these kinds of “pure” software patents. And it seems to me that the most reasonable interpretation of the high court’s decisions is the one Ben Klemens has articulated: that software by itself cannot be patented, and that “insignificant post-solution activity” (to quote the Diehr majority) such as displaying the results of a calculation on a computer screen, cannot transform an unpatentable algorithm into patentable machine.

There’s a lot more detail in my article, so I hope you’ll check it out.

SHARE: 5 comments