Software Patent

I think almost everyone I know in the software development field has a deep hatred for patents and the way they've been used in our field. I've had a post on my todo list for ages about this and have finally been moved to write about it after a particularly good piece of investigative journalism by This American Life. The short form of my post is that while patents (even software patents) are a good idea in principle, in practice they have turned into an unmitigated disaster and would be better scrapped.

To begin, however, I'll say why patents are a good idea in principle, indeed they may be one of the most valuable 'inventions' in human history.

A good argument for patents playing such a central case in human development is made by William Rosen in his book The Most Powerful Idea in the World. Rosen's book looks at the Industrial Revolution which he characterizes as one of the most important events in human history, one that made a step change in human wealth.

A skilled laborer—a weaver, perhaps, or a blacksmith—in seventeenth-century England, France, or China spent roughly the same number of hours a week at his trade, producing about the same number of bolts of cloth, or nails, as his ten-times great-grandfather did during the time of Augustus. He earned the same number of coins and bought the same amount, and variety, of food. His wife, like her ten-times great-grandmother, prepared the food; she might have bought her bread from a village baker, but she made pretty much everything herself. She even made her family’s clothing, which, allowing for the vagaries of weather and fashion, was largely indistinguishable from those of any family for the preceding ten centuries: homespun wool, with some linen if flax were locally available. The laborer and his wife would have perhaps eight or ten live offspring, with a reasonable chance that three might survive to adulthood. If the laborer chose to travel, he would do it on foot or, if he were exceptionally prosperous, by horse-drawn cart or coach, traveling three miles an hour if the former, or seven if the latter—again, the same as his ancestor—which meant that his world was not much larger than the five or six miles surrounding the place he was born.

And then, for the first time in history, things changed. And they changed at the most basic of levels. A skilled fourth-century weaver in the of Constantinople might earn enough by working three hours to purchase a pound of bread; by 1800, it would cost a weaver working in Nottingham at least two. But by 1900, it took less than fifteen minutes to earn enough to buy the loaf; and by 2000, five minutes. It is a cliché, but nonetheless true, to recognize that a middle-class family living in a developed twenty-first-century country enjoys a life filled with luxuries that a king could barely afford two centuries ago.

-- William Rosen

The change wrought by the Industrial Revolution is almost unimaginable. We think our current era is one of constant change, but that phrasing captures what we forget. These days we are used to change, but before the Industrial Revolution human life changed very slowly. The biggest change the Industrial Revolution unleashed was change itself.

Thus I think few students of human history doubt the vital importance of the Industrial Revolution, but this raises a couple of important questions - why did it occur when and where it did? What was so special about late 18th century England that set this steam train off? Rosen's view is that patents are the key enabler, because they provided a financial incentive and platform to support inventors and entrepreneurs. Without patents only wealthy people (or those with wealthy patrons) could afford to innovate, and there was little incentive for them to do so.

I find Rosen's argument persuasive and thus think that patents were not just a Good Thing but one of the Best Things to have happened to our species. So why do I loathe software patents so much?

It boils down to the fact that patents generally have become very debased from the animal that enabled the Industrial Revolution. The primary debasement is that of novelty. The whole point of patents is to grant a (limited-term) monopoly to something that is new. The US patent law says you can't have a patent if “the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains.” This boils back to the English 1623 Statute on Monopolies1 which started the notion that a patentable invention is something that is both novel and useful.

The core problem with software patents is that this key principle has been tossed aside. Everyone in the software field has seen a parade of patents which do nothing but try to claim rights on techniques that have already been in use for years, let alone developments that while new, are are still obvious to those of us with ordinary skills in programming.

Although this debasement is quite enough to ruin the integrity of software patents, there are some other debasements worth mentioning too. Patents were originally created with a limited time in mind - the 1623 law placed them at fourteen years. This, of course, at a period of time when change was much slower than it is now, let alone than it is in our field. Proper software patents should hold for a shorter period than that. Further debasement occurs in lack of specificity - most software patents are ridiculously broad and vague, while patents were originally seen as narrow and specific. Narrow patents encourage innovation by incenting people to think of different ways to solve the same problem, broad patents snuff that innovation out.

The result of all this debasement is a world where patents no longer incentivize and communicate new inventions, but where they are weapons to be used to fight legal battles. For large companies they are, at least on the surface, an annoying distraction and cost. But the real damage they do is to small outfits, that can't afford the time and money to fight a patent lawsuit. Thus we see patents used for shakedowns - stifling innovation.

The tragedy is that patents have become a source of reinforcing existing powers. A big company may find patents a significant inconvenience, but in the end patents are good for perpetuating the current power-holders because they can snuff out the smaller ones. This is why it's hard to change the system, those with the power have no incentive to give it up.

I find it particularly depressing that my fellow programmers are complicit in this tragedy. It's not uncommon for programmers to talk about patents they've been involved in getting and how they know how absurd they are. I know it's easy to get on a high horse here, but I do think that any programmer who cooperates in getting a baseless patent should be ashamed of themselves. It shows the kind of lack of responsibility that undermines any justification we have to be treated as professionals.

In theory, I'm not against software patents if we were able to get back to the core beneficial principles of patents and apply them properly. This would imply developing a process that would ensure that patents were only granted for truly novel ideas. But unless such a process were properly put together, I'd rather see software patents eliminated completely. A world without software patents would be better than the mess we're currently in.

Further Reading

There's a huge amount of material that's been written about patents out there. Tim Bray's piece last year on giving up on patents links to a number of good sources. A couple of these point to evidence that software patents have reduced innovation.

Planet Money did a follow up podcast to the one on This American Life, with links to the studies they summarize.

“At a time when our future affluence depends so heavily on innovation, we have drifted toward a patent regime that not only fails to fulfil its justifying function, to incentivise innovation, but actively impedes innovation.” - W.W. of The Economist