The US Supreme Court decided unanimously this week that many patents have been granted for inventions that are really obvious. To long-time observers of patent law, the conclusion, which absolved the KSR company of infringing Teleflex's patent on a gas pedal with an electronic sensor, was itself obvious. The revised criteria outlined by the court will undoubtedly nullify many patents, and a lawyer for Microsoft, a frequent target of infringement lawsuits, estimates that it will reduce the software company's exposure to liability by as much as 60 percent. Not long ago, I had a discussion with a professor, an expert in semantics, about the word "obvious." People often disagree whether something is true, but is it really meaningful to argue whether something is "obviously true" or whether on the contrary it is indeed true, but not obvious? He mentioned a famous folk story about a math professor who, when asked a question, screwed up his face, furiously scribbled equations on the blackboard for a half-hour and then replied, "It's obvious." Only a few days later, I saw an article in the paper reminding me that arguments about obviousness are not only possible, they are big money. A nine-figure lawsuit was being argued over a patent that the infringer claimed was "obvious," and thus invalid. The patent holder averred that the invention was "not obvious," and the court was obliged to arbitrate. For over a century, courts have agreed that in order to be eligible for a patent, an invention must meet three criteria: it must be novel, useful and non-obvious. If it is not novel, then the applicant is not the inventor and so can not obtain rights. "Useful" is not the opposite of useless, for surely there is no harm in granting patents to useless inventions. It means that the invention is of immediate practical use. The idea is to prevent someone from patenting laws of nature, mathematical theorems and the like. And since patents are meant to stimulate innovation, there is no need to reward inventions that are obvious, meaning that if this inventor hadn't stumbled upon the idea someone else surely would have. (In Europe this requirement is called an "inventive step.") But how exactly do we define obvious? The classical test is whether it would be discovered by someone of "ordinary skill in the art." But that begs the question of who has "ordinary skill in the art" and how much effort that person would have to invest to make the discovery. In order to give a more objective definition of non-obvious, the US Circuit Court, which adjudicates most patent cases, has relied in recent years on the so-called "suggestion test." They try to see if some public document indicates a "suggestion or incentive" to combine existing ideas in a new way. One reason for adopting this criteria rather than the "ordinary skill in the art" was the problem of "hindsight bias" - any invention may seem obvious after you see it, just as the math professor in the story found the result obvious after intensive study. A problem with this test, as many commentators have pointed out, is that some ideas are so obvious that people don't even bother to publish them. In this way, the most obvious ideas may be the easiest to patent! Computer scientist and expert witness Jeffrey Ullman mentions a case of a patent granted in the 1980s for a technique he learned from his high school math teacher in the 1950s. His math teacher, and perhaps hundreds of others, never thought that this technique was worthy of publication, beyond demonstrating it as a neat trick to students. (Eventually, the patent was disqualified for an unrelated reason.) The Court wrote: "In many fields, there may be little discussion of obvious techniques or combinations." Another problem is that something non-obvious later may become quite obvious due to processes unrelated to the patent itself. In a research study I undertook on gene patents, I learned that isolating a new gene required at one time immense skill and ingenuity, and courts concluded that genes could be patented. However, as time went on the process of isolating genes became far more automated, almost routine. Yet patents continued to be granted and enforced. The US Supreme Court noted this process in the landmark case. As science and technology progress, ideas that at one time were innovative later become routine. Decades ago, electronic sensors were arcane devices unfamiliar to the average auto engineer, but today they are an off-the-shelf product familiar even to tinkerers. The judgment states that "as progress beginning from higher levels of achievement is expected in the normal course, the results of ordinary innovation are not the subject of exclusive rights under patent laws." I've been watching the patents scene for some time, and my sense is that virtually all experts agree that it has been far too easy to obtain patents in recent years and that the patent system has been obstructing technological progress as much as it has been stimulating it. Companies are prevented from taking advantage of essentially obvious ideas that have been patented. The fault is seldom with the patent examiners. While there are occasional arcane inventions which stump their expertise, most examiners are highly professional and familiar with the relevant professional literature. The problem is that the criteria, including the "suggestion test," were just far too lenient. One reason for the explosion of patenting in recent years is that there truly has been an immense increase in the degree of innovation and progress. The men and women who devote their 1% of inspiration and 99% of perspiration to creating innovative products and processes deserve legal protection to enable them to profit from their efforts. But a parallel process has been a dumbing-down of the criteria for obtaining patents in the US, and the latest decision will help correct this distortion. firstname.lastname@example.org The writer is research director at the Business Ethics Center of Jerusalem (www.besr.org), an independent institute in the Jerusalem College of Technology.