Originally Published: Tuesday, 31 August 1999 Author: Sydney Weidman
Published to: featured_articles/Featured Articles Page: 1/1 - [Printable]

Why the Special Case for Software Patents?

The discussion surrounding software patents always seems to revolve around a couple of fundamental assumptions that I think ought to be challenged. The first is that software is a different kind of intellectual "stuff" than other patentable entities....

   Page 1 of 1  

The discussion surrounding software patents always seems to revolve around a couple of fundamental assumptions that I think ought to be challenged. The first is that software is a different kind of intellectual "stuff" than other patentable entities.

Why does software deserve or require special treatment? If software is categorically different from other types of (potential) intellectual property, then it seems legitimate to build a case against software patents without impugning any other classes of protected entities. If, on the other hand, software development is just like any other invention, then whatever is said against software patents applies to all types of patents and similarly any arguments in favour of patents in general would count in favour of software patents.

Some of the arguments that have been presented by the Free Software Foundation are contained in the 1991 article entitled appropriately enough "Why Patents Are Bad for Software". The enquiry is a straightforward one. Patents were intended to "spur innovation and aid in the dissemination of information about technical advances". The authors ask a simple question: Is patent protection for software achieving this aim?

According to the article, patent protection for software is not achieving the purpose for which it was intended. Firstly, it is argued, the task of analyzing algorithms to determine whether they infringe is massively complex and time-consuming. The sheer volume of information is daunting enough, claim the authors, not to mention that it is difficult to know where to look. After all, algorithms can be used in many different ways. But is this really peculiar to software? Integrated circuits must be equally complex and difficult to study. So would many manufacturing processes. I don't buy this as a distinguishing feature of software. There are examples of non-software patents which are equally complex and yet for which patent orthodoxy would still demand protection. If you're going to argue against software patents on the basis of functional complexity, you'd better be prepared to exclude other types of patents as well.

The second major argument presented in the article is that the cost of patent filing and litigation is high enough to prohibit the entry of small companies into software development. Since most small companies drive innovation in software design, this would be quite a blow to technical progress in programming. Again, it seems that this argument has all of patentdom in its purview. The cost of filing and searching patents and the risk of litigation is present for any patent application or in any field of technology. There is nothing here to distinguish software patents from all others.

The third indictment of software patents is that they do nothing to discourage trade secrecy. Since the source code is still hidden from view and defended by trade secret and licensing restrictions (e.g. no reverse engineering) patents have had no impact on this form of knowledge hoarding. Of the arguments against software patents, I believe that this is the argument (the only one, really) which best distinguishes software from other types of intellectual property. Still, one could find other industries which are harmed by secrecy in spite of patent protection. The software industry isn't alone in this.

Opponents of software patents might be aware that their reasoning could easily be applied to all types of patents, but the realpolitik of the situation prevents them from making the larger case. They might say: "If we advocated getting rid of patents altogether, we might lose the battle against software patents".

If we accept the generality of the argument against software patents, we must ask if patents are necessary at all. The necessity of protecting intellectual property is simply taken for granted and not questioned. This assumption, which forms the constitutional basis of patent and copyright law (in the United States, that is) has come under scrutiny in recent years by people other than just John Perry Barlow and the odd Slashdot reader. Although it seems hard to dispute the benefits of patent protection on the surface, the empirical evidence that property based incentives generate more prolific production of intellectual goods is ambiguous and inconclusive.

Many people, notably Stephen Breyer (before he became a Supreme Court Justice), have argued that intellectual property laws are indeed unnecessary because companies receive enough competitive advantage (and therefore profit) from being first to market with an invention. Moreover, if someone brings out a cheap copy, the originator of the product can still compete on the basis of brand recognition and price. The payback period for investments in R&D would be longer, but this should not prevent anyone from profiting. This is probably even more true today than when Breyer wrote in the 70's. The product life cycles are much shorter than they were, so the benefits of getting it out the door first are even greater.

Would a world without patents stimulate enough inventive effort? The answer to this question depends in large measure on how you define "enough". In America, there's no such thing as too much technology. And there's no form of technology that is off limits. The patentability of living creatures ought to have caused some controversy. Instead, it seemed to pass unnoticed into the annals of science history. The race to control the information contained in the human genome is another example of technology invading areas without even pausing to reflect on consequences.

Even if patents do guarantee a cornucopia of new technology, is this necessarily a good thing? Is it good and right for us to promote technological progress? At the end of the eighteenth century, when the US constitution was being drafted, technological developments were in an embryonic state. New inventions needed patent protection or they would be stillborn. Nowadays, the reverse is true. We are overwhelmed with technological change and progress. The regime of patent protection of 250 years ago doesn't belong in the modern world. Rather than innovation, perhaps we should start rewarding dependability or social value.

Sydney Weidman is a writer and agent provocateur living in Winnipeg, Canada, studying Philosophy and Business Computing at the University of Winnipeg.

   Page 1 of 1