
Logic Patents > Reviews > Digital Dilemma 2000 > Newell 1986: Models Broken
Digital Dilemma 2000	Newell 1986: Models Broken	Chisum 1986	Konno 1995: Karmarkar	Tamai 1998: Abstraction

Newell 1986: The Models are Broken

Paul Newell, professor of computer science, in response to law professor Donald Chisum's proposal to affirm the patentability of algorithms, does not directly say whether algorithms should be patentable, but rather describes how both affirmation and negation of this proposal would clash with the underlying assumptions of the patent system and how this will inevitably challenge the foundations of the patent system. As more and more problems are solved by purely mathematical means, the patent system will either have to become less relevant for innovation, or it will have to completely review its model of what an invention is and how it should be appropriated. In particular, Newell explains some basic concepts of informatics and points out that, with algorithms, there can be no meaningful differentiation between discovery and invention, between application and theory, between abstract and concrete, between numeric and symbolic etc. Moreover he explains by a model of game theory that sharing algorithms, as currently practised by programmers, may lead to more innovation than making them appropriatable, so that a crude application of the patent system to algorithms could very well stifle rather than stimulate innovation.

title:: Newell 1986: The Models are Broken
source:: UPLR University of Pittsburgh Law Review 47 (1986) 1023-1031

Chisum 1986: A leading US scholar of patent law who is said to have exerted an overwhelming influence on patent jurisdiction of major US patent law courts such as the CAFC, explains that it is not consistent to grant patents on practical applications of mathematical algorithms but not on the algorithms themselves. Chisum argues that mathematical methods are useful inventions just like any other methods and should be granted independently of any specific tangible application. Any future tangible application that uses the mathematical method should fall under the claim.
Tamai 1998: Abstraction orientated property of software and its relation to patentability: Prof Tamai of Tokyo University shows how patenting of software clashes with some of the underlying assumptions of the patent system. The patent system relies on requirements such as concreteness and physical substance in order to keep the breadth of claims within reasonable limits. Software innovation however is the art of making processes as general as possible, i.e. the art of abstraction. Tamai quotes a set of patent claims from the SOFTIC symposium of 1993, where patent officials from JP, US and EU judged the patentability of an example algorithm at different levels of concretisation. The European representative was more willing than his colleagues from US and JP to grant patents on abstract claims, but even he shyed back from granting them at the level that really represents the innovative achievement. Tamai shows how this inconsistency leads to a series of other inconsistencies. Tamai sees only two ways out of the inconsistency: (1) acceptance of abstract claims (2) exclusion of software patents.
Donald Knuth and Software Patents: Donald Knuth, pioneer and cult figure of informatics (computer science), author of some definitive monumental classics such as "The Art of Programming", finds that software patents are built on some basic misunderstandings, similar to the misunderstandings of certain provincial american legislators in the 19th century or the medieval catholic church. Computer programs are as abstract as any algorithm can be, Knuth says.
Bessen & M 2000: This article is written by two researcher from MIT and concludes, after giving mathematical models and experimental evidence, that in a dynamic world such as the software industry or consulting industry, firms may have plenty of incentive to innovate without patents and patents may constrict complementary innovation. It concludes that copyright protection for software programs (which has gone through its own evolution over the last decade) may have achieved a better balance than patent protection. This new model suggests another, different rationale for narrow patent breadth than the recent economic literature on this subject.
Konno 1995: Karmarkar: In this book which received great public attention in Japan, Prof. Konno, a famous specialist of operations research (applied mathematics) from Tokyo University, explains what is behind the recent drive to extend the scope of patentability to software and mathematics. The book starts from a few fascinating examples of real advances in mathematical research, namely Narendra Karmarkar's interior point method of linear programming. This method was a truly rare achievement of ingenuity with great value for large-scale industrial planning. If any mathematical achievement could be suitable for patenting, then it was this method. The AT&T laboratories did receive a patent for this method in 1988. However the Karmarkar patent did not create very significant revenues for AT&T while causing significant inconvenience to the industry and destroying much of the positive interaction cycles that had previously characterised the cooperation between linear programming research and industry in the US. The AT&T Laboratories no longer produced any Nobel Prize winners after they became patent-oriented, and Karmarkar himself was isolated from the network of mathematic researchers, who soon left him lagging far behind. Konno captures the reader with a fascinating introduction to linear programming, interwoven with a dramatic account of the battles that accompanied the Karmarkar patent within the mathematics community and the patent system. Konno was on the podium or in the first row during the conferences where the drama around Karmarkar, linear programming and mathematical patents evolved. Moreover it was Konno who filed an opposition to the Karmarkar patent in Japan. He shows that the Japanese Ministery of Trade and Industry (MITI) pushed the Japanese patent office to bend Japanese patent law in an inconsistent direction. Lawyers at the time explained that it was not necessary to be consistent but to go with the tide, which was inevitably pro software patents. The MITI acted under the pressure from these lawyers and from the US government. Konno's vivid examples suggest however that software patents are not here to stay. They do not correspond to any need of the software industry. They have stifled innovation and resulted in a massive blood transfer from the software industry to the litigation business. They created a new market of about 2 billion USD per year for the patent law professionals. 'You needn't be named Hercule Poirot to see who is the driving force behind all this', Konno remarks. But there is hope, because the young generation has not lost sight of the problems.