Under normal circumstances, a patent application should always be rejected if there is prior existing knowledge about the product.

But in the United States, "prior existing knowledge" is only recognised if it is published in a journal - not if it has been passed down through generations of oral and folk traditions.

"We are deeply gratified that through our case the EPO has recognised the intellectual achievements of the South," said Linda Bullard, former president of IFOAM [International Federation of Organic Agriculture Movements].

"We were able to establish that traditional knowledge systems can be a means of establishing 'prior art' and thus used to destroy claims of novelty and inventiveness.

Leading the neem campaign were the EU Parliament's Green Party, the Research Foundation for Science, Technology and Ecology (RFSTE), and IFOAM. This story was my first clue that prior art in the EU is not identical to what constitutes prior art in the US, and it is clearly wider in its definition there in Europe.

FFII has sent some questions [PDF] to the Council Legal Service of the Council of The European Uniion about the way the recent A-item vote was handled. Was it procedurally proper? Here is a flow chart. And here is a scholarly paper [PDF] that explains the EU system and the proposed directive, and reading it helped me to get a much clearer picture of how we ended up where we are.

The abstract reads like this:

SOFTWARE PATENTS: A CLOSER LOOK AT THE EUROPEAN COMMISSION’S PROPOSAL

MARIA ALESSANDRA ROSSI
Doctoral School of Law and Economics, University of Siena

Abstract

This paper attempts an evaluation of the merits of the European Commission’s “Proposal for a Directive on the Patentability of Computer-implemented Inventions”. The general thrust of the argument will suggest that it is indeed not surprising that the Proposed Directive has generated such an outburst of reactions. Indeed, while the Commission’s Proposal has been presented as a confirmation of the status quo, it is difficult to question that it will have deep consequences for the design of the regime of legal protection for computer-implemented inventions. The paper addresses three main questions, namely: (a) does the Proposed Directive address worthwhile issues?; (b) is a European Directive necessary?; and (c) is the Proposed Directive likely to achieve its stated objectives? The search for an answer to these questions focuses on the likely impact of the Proposed Directive on legal certainty and innovation.

This all matters, because if the directive is passed, we very likely will want to start invalidating patents right and left, if they interfere with FOSS, so knowing what constitutes prior art in the EU is important. The case involving India has stretched the definition of prior art so that now it is even wider than it was before, including now "common knowledge" that hasn't been set down in written form anywhere. I am seeing possibilities, if worst comes to worst.

The paper also clarified for me, on page 6, how patents have been granted in the EU when supposedly there are no software patents:

2.2. The European Situation

Although as early as 1975 attempts were effected to create Community-wide patents through the signing of the Community Patent Convention, the granting of patents at the European level is still governed by the 1973 European Patent Convention (EPC) signed by 15 member countries and five neighbouring countries. The EPC created the European Patent Organization and the European Patent Office, which are therefore not institutions of the European Union, and endowed the latter with the power to issue patents that take effect in the member states of the EPC selected by the applicant. Patents granted by the EPO represent a “bundle” of national patents and are thus subject to national legislation.

To be patentable under the EPC, inventions must be “susceptible of industrial application”, must be “new” and must involve an “inventive step” (art. 52 EPC). The mentioned requirements set out in art. 52 EPC parallel to some extent those contained in section 35 U.S.C. The novelty requirement is satisfied as long as the invention does not form a part of the state of the art (art. 54 EPC). The “inventive step” requirement bears close resemblance to the U.S.C. “non-obviousness” requirement. Indeed, art. 56 EPC states that “an invention shall be considered as involving an inventive step if [...] it is not obvious to a person skilled in the art”. However, some difference between the U.S. and European legal frameworks is apparent in regard to the “industrial applicability” requirement. Fulfillment of the latter implies, per art. 57 EPC, that an invention “can be made or used in any kind of industry, including agriculture”. Note that the letter of art. 57 should not be interpreted, by analogy with 101 U.S.C. 35, as imposing a requirement analogous to the U.S. usefulness requirement. Indeed, the European Patent Office Guidelines (C-IV, 1.3) specify that “[t]he EPC does not require explicitly or implicitly that an invention, to be patentable, must entail some technical progress or even any useful effect”.

Further differences between the EPC and U.S.C. Title 35 lie in the fact that the EPC does not provide a categorization of patentable inventions, while U.S.C. Title 35 at 101 explicitly indicates statutory subject matter, and in the fact that art.52 EPC explicitly identifies subject matter that cannot constitute the object of an invention, while in the U.S. the only exceptions to patentability are judicially created. In particular, art.52(2) EPC lists (a)discoveries, scientific theories and mathematical methods; (b) aesthetic creations; (c) schemes, rules and methods for performing mental acts, playing games or doing business, and programs for computers; and (d) presentation of information as excluded from patentability. Art. 52(3) modifies, however, the extent of such proscription, by requiring it to be applied only in so far as the mentioned subject matter is claimed “as such”.

The most relevant aspect of the divergence between the U.S. and the European legal frameworks in relation to the patentability of computer-implemented inventions resides in the fact that in Europe, in addition to satisfying the statutory requirements contained in art.52 EPC, an invention must be technical in order to be patentable. This requirement has been generally derived from Rules 27(1) and 29(1) EPC and has recently found an explicit recognition in the European Patent Office Guidelines that read: “the invention must be of ”technical character“, to the extent that it must relate to a technical field (Rule 27(1)(a)), must be concerned with a technical problem (Rule 27(1)(c)), and must have technical features in terms of which the matter for which protection is sought can be defined in the claim (Rule 29(1))”. In Japan, the statutory definition of invention as “a highly advanced creation of technical ideas by which a law of nature is utilized” has traditionally led to interpretations close to prevailing EPO practice. By contrast, the U.S. Patent Office Guidelines for Computer-related Inventions only require the utility of an invention to be within the “technological arts”. As it will be clarified in the following paragraphs, the requirement of “technical character”, while being crucial to most decisions of the EPO Technical Board of Appeal, has been the object of a protracted debate that has not, however, led to an agreed upon and uncontroversial legal definition.

It is to the perceived lack of technical character of computer programs that the exclusion from patentability contained in subsections (2) and (3) of art.52 EPC should be attributed. Computer programs “as such” are assimilated to abstract creations such as discoveries and scientific theories (art.52(2)(a)) or non-technical objects such as presentations of information (art.52(2)(d)) and therefore considered to be non-inventions within the meaning of art.52(2). This reflects the historical development of the concept of invention in most European nations and is confirmed by both the EPC “travaux preparatoires” and by a number of decisions of the EPO Technical Board of Appeal dealing with the issue of patentability of computer-related inventions.

What does it all add up to? If you keep reading, as I did, I think you will see that in actuality there is no barrier that can't be bypassed. I was stunned to learn that business methods patents have already been granted in the EU, despite all the promises that nothing like that would ever be allowed there:

As a final note on the state of software patentability in Europe it is worth commenting on the issue of business method patents. Contrary to a widespread perception, and in spite of the explicit exception contained in art. 52(2) and (3) EPC, methods of doing business have been granted patent protection in a number of instances under case law of the EPO Technical Board of Appeal. The SOHEI case mentioned above constitutes one case in point. The patent at issue in this case covered a general-purpose business management system comprising a screen display with peculiar attributes that were found to satisfy the “inventive step” requirement. The case constitutes by no means an isolated event. Indeed, many more patents relating to the field of business models and e-commerce have been already granted and for many others the application is currently pending. ³⁷

The key test applied to determine the admissibility of claims to business methods under EPO case law seems to be, as in the case of other computer-implemented inventions, whether the invention involves a technical effect of some kind. Thus, provided the invention can be claimed in terms of the technical structures of a computer program and some technical aspect of the invention can be said to provide a contribution to prior art (e.g. involve an “inventive step”), it should be expected to be granted patent protection. The need for fulfillment of a “technical character” requirement distinguishes EPO practice from the prevailing practice in the US, where patents have been granted even for business models claimed independently of computer systems or any other technology. However, it appears to be the case that the requirement of “technical character” could be stretched enough to allow the granting of patents for inventions of the kind involved in the State Street decision mentioned in 2.1, although the corresponding European application for the latter happens to have been first rejected by the examining division and then abandoned by the applicant. On balance, it could be argued that patenting of business methods presents some additional difficulties with respect to the US, but in many instances such difficulties do not seem to be unsurmountable

³⁷Examples of business management and financial systems patents (reported by Beresford [2001]) include European patent 407,026B (Reuters Ltd.), concerning a computer trading system involving a reduced amount of data transmission with respect to prior art; EP 701,717B (Shepherd), protecting a system for setting up insurance contracts by computer; and EP 838,063B (Realkredit Denmark A/S), covering a computer system and the relative software apt to perform the calculations required by specific financial instruments.

I therefore see that Bill Gates' recent suggestion that the US, EU, and Japanese patent systems be harmonized is a practical possibility, with nothing significant standing in the way of its implementation. It's just a matter of stretching things until they fit. And from this paper, that is what has already been happening. Personally, I now believe that any promises that the EU system will be kept separate and distinct from the US system are not to be believed, and any decision on whether or not to adopt software patents in the EU must be based on that understanding, that a yes means that the EU will at some point be indistinguishable from the US system.

But read the paper yourself, and feel free to draw your own conclusions. You might like to read beginning on page 19 the section on questions to ask about the directive:

3. Some basic questions to be asked about the Proposed Directive: preliminary comments

The aim of this section is to provide some preliminary comments on three basic questions concerning the Proposed Directive and worth of some reflection. First, does the proposal presented by the European Commission address the appropriate issues? Second, is a European directive on computer-implemented inventions necessary or could the issues underlying it be addressed in a different way, as for instance by leaving matters to the EPO and national courts or through a direct modification of the EPC? Finally, is the Proposed Directive likely to achieve its stated objectives?

Beginning on page 30, there is a section on whether passing the directive would stimulate innovation, which highlights some issues involving software that make analysis of its patentability unique, and the paper cites some studies on the question of innovation that you may wish to look up and read:

Another piece of empirical evidence concerning the software industry has been recently offered by Bessen and Hunt. The two authors examine a comprehensive database of U.S. patents on inventions that use software granted between 1976 and 1999. The results of their analysis are taken to suggest not only that software patents are particularly “cheap” to obtain and that they are generally acquired for the purpose of building patent thickets, but also that greater use of software patents is associated to lower R&D intensity. Bessen and Hunt’s study has been criticized on various grounds. Of particular interest in this respect is the survey conducted in 2004 by Ronald Mann. Mann forcefully rejects the “myth” of a software patent thicket on the basis of interviews conducted among 50 industry executives (including software developers, venture capitalists, lenders etc.). Indeed, respondents to Mann’s survey explicitly exclude the existence of a patent thicket when asked a direct question. However, the interviews reported by no means indicate an overall positive role for patents as Mann’s argumentation would suggest. In short, Mann demonstrates that small firms value patents in a world in which they are already available to large companies, but that does not demonstrate that small firms would value them if the rest of industry were not already endowed with large patent portfolios. In other words, rather than showing that possession of patents has some value for small firms, the analysis of the interviews gathered by Mann suggests that the lack of patents constitutes a danger for small firms and that this danger is becoming more and more prominent. Note also that participants to the U.S. Federal Trade Commission’s Hearings on Competition and Intellectual Property Law have, for the most part, given responses of a totally different tone than Mann’s responders, confirming the existence of a software patent thicket.

In sum, it seems reasonable to conclude that while it could be the case that “anticommons” and “patent thickets” are not currently posing as much problems as some scholars would suggest, continued and widespread reliance on patents for software may go in the direction pointed at by Cassandras. The question is thus one of balance of benefits and costs. It is certainly true that patent holders’ initiatives such as patent pools and cross-licensing may reduce the adverse consequences of patent-related transactional difficulties, but the question remains as to whether the overall benefits of software patents outweigh their costs, including the costs of devising these private solutions. . . .

As for the objective of increasing innovation in the software sector, our negative conclusion on the ability of the Commission’s Proposal to achieve it comes from a survey of recent economic literature. The evidence surveyed, while not conclusive, can by no means be interpreted as a proof of the existence of a link between extension of patentability and increased innovation, which the Commission seems to assume. The evidence gathered in relation to the recent experience of the United States with the removal of the obstacles to the patentability of software-related inventions suggests the only clear conclusion that we should expect the Directive in its present form to bring about an increase in patent applications and a correspondent increase in litigation.

Of course, what is totally absent from this, and so much other analysis, is the impact of patents on Linux and other Free and Open Source software, particularly on software released under the GPL. Is that not a central issue here? I believe that issue must be stressed and explained clearly in any discussion of software patents in the EU, or anywhere else. Here are some questions: What is Microsoft's leading competition currently? Shall we deliberately hand them an anticompetitive weapon tailor-made to destroy its chief competition? And should the millions and millions of individuals and businesses and governmental agencies that deliberately and consciously chose to use GNU/Linux not be allowed to do so because the former business model champions would like to monopolize the field? Is it to anyone's advantage to allow such a monopoly?

The New Microsoft XML Patent

Before you answer, take a look at this new patent Microsoft has been granted (it is open to challenge now) in New Zealand:

Patent 525484, accepted by the office and now open for objections until the end of May, says Microsoft invented and owns the process whereby a word-processing document stored in a single XML file may be manipulated by applications that understand XML.

It is one of a raft of patent applications Microsoft has dumped on the overworked staff of the office, and on patent offices worldwide.

Some of them might have more merit than this particular piece of junk, but they are part of a strategic effort by Microsoft to control another generation of technology, just when its grip on the personal computer is being undermined by the Open Source movement.

So, a process whereby a word-processing document stored in a single XML file may be manipulated by applications that understand XML. Yoo hoo. OpenOffice. Are you guys paying attention? Note this patent is open for objections until the end of May. The article goes on to say that there are objections that will be filed:

The internet works because the protocols and software behind it were developed by peer-production and left free as a software commons.

Patent 525484 says if a document is created in Extensible Markup Language or XML, it may contain a schema or set of instructions for interpreting the document, which will be used when you want to use another application to make changes to the document. There is no novelty there. The whole idea of XML was to allow interoperability.

Interoperability standards have been around for as long as people realised it would be a good idea for programs to talk to each other.

The patent should be rejected for obviousness, if not for other criteria such as prior art.

New Zealand Open Source Society president Peter Harrison said the society would object.

What does the patent mean, if it is allowed?

If the patent is granted, other word-processing programs will not be able to interact with Microsoft XML word processing documents without Microsoft’s permission.

"It means if you write a document in Word and save it in an XML format, you have to have Microsoft’s permission to read it or change it," he said.

Earlier versions of Microsoft Word had no such patent protection, and although it has taken years to achieve, programs like Open Office can now interact directly with Word.

That, obviously, will become just a brief interlude of interaction. That appears to be the plan. And it's important that everyone understand exactly what is quietly being put in place all around the world on the patent front, and exactly what is at stake. Well, well, here it pops up in Canada.

So, I'll contact Mr. Harrison and find out what would be useful in the way of prior art, and then I'll let you know and we can get started finding it, or at least we can give it a try. Sound like a plan?