Law and Tech A weblog on Law and Technology


The Case for Software Patents

The Supreme Court recently granted cert in Bilski v. Doll, which most believe will change the test for patentable subject matter. Supreme Court precedent has carved out three exceptions to patentability under 35 U.S.C. § 101, these are laws of nature, physical phenomena, and abstract ideas. The quintessential law of nature is Einstein’s famous formula: e=mc^2, which is unpatentable subject matter. A patent cannot be granted that merely states this equation, as it would preempt all uses of the formula. What can -and has- been patented is the application and use of equations, for instance the Arrhenius equation in Diamond v. Diehr.

Computers and their applications have changed the world we live in. Many problems are solved by the use of the computer, from communication to commerce, and from filing your taxes to putting a man on the moon. Computers have taken over many tasks such as running your car to running voting booths. Since their creation there has been a melding of hardware and software to accomplish these goals. Many problems that computers can solve can be handled on both the hardware and software levels. Much of the power in today’s general purpose computers is that we do not need specialized hardware to perform specialized tasks. Methods that can be used to solve difficult problems thus can be solved similarly at both the hardware and software levels and so protection for one without protection for the other is meaningless. Especially since the use of these methods on a general purpose computer could require no additional hardware to perform the needed functionality.

I do not want to discuss the anti-competitive effects of patents generally, or to software patents in particular. I have not studied these in depth but will note that since software patents have been granted the United States has continued to be a hotbed for innovation in software industries having many of the largest software firms around the world. But, I also understand that software is special, there exists a vibrant open source community unlike any other field which is willing to innovate without the need for financial incentive and exclusive rights.

A problem that some have with software is that it is merely an algorithm for performing some task. And, that any task performed on a computer can be performed by a person with a pen and paper, given enough time. I both agree and disagree with that argument, as fundamentally it is correct but I believe that it is an oversimplification of what software does. Simple algorithms like converting a number from decimal to binary and back (as seen in Gottschalk v. Benson) can easily be done using either means. But, what about all those applications that are for computer specific purposes that doing them by hand does not solve the actual problem?

For instance, public key encryption such as RSA was developed as a way to securely encrypt data without first requiring the parties to have created a key to decrypt their communications. Performing the operations by hand or in our mind does not actually solve the problem where it exists, for performing operations on the internet and with computer systems. By limiting the technology to a computer (in a patent’s claims), it does not preempt uses of the algorithm in the mind or on paper, just the method’s use on the machine. This is akin to patenting a new use for a drug. Even though the drug itself may be in the public domain, this new use may be patented. Additionally, there are algorithms that make no sense without a computer system. Programs such as sending email or instant messages have no purpose without interfacing with additional hardware. For this technology to make sense, you must use additional hardware and practicing these algorithms without the use of a computer is pointless.

Software merely makes use of public domain formulas and applies them, just like traditional mechanical inventions. Take a chair for instance. It works by applying the standard laws of physics. Using a computer, or a pen and paper, the chair could be “proven” to work as well, just as a piece of software can be and just as large scale buildings and bridges are. All inventions, mechanical or software based draw upon fundamental principals that themselves are not patentable, but by applying them to solve a problem we move beyond the abstract and into the spectrum of patentability.

Novelty and obviousness also cloud the picture. When we look at the wide spectrum of patentability we see things that immediately make us cringe. Most of the things that make us cringe are because the inventions that we see on this spectrum have been done before. While difficult, it is imperative to separate patentable subject matter from fulfilling all the requirements of patentability, including §§ 102 and 103. Many of the problems that people have with the patenting of software is that the inventions are in the prior art. In this regard, the piece of technology should not be patented, but the class of technology may very fall within the prior art. I see this frequently when discussing patents like the infamous Amazon “1-Click” patent. Most who find that this patent should be invalid think that the technology is obvious.

The patent system is our way of protecting innovation. As we do not know where the next class of innovation is, or how expensive it will be to develop, we should not look to limiting the scope of patentability. Expansive scope has been a cornerstone of our system of patents and enabled the United States to have many technological advances. While not perfect, the patent system is a good trade off of exclusive rights for innovation.