Ce site est un chantier à ciel ouvert habité par les éditeurs, lecteurs, auteurs, techniciens, designers de Sens public. Il s'agence et s'aménage au fil de l'eau. Explorez et prenez vos marques (mode d'emploi ici) !

Patents and Liberties

  • Résumé
  • Mots-clés (5)

President, ladies and gentlemen, as an economist turned party activist and senior politician, (a curious turn of events), it’s only right that I should be invited to speak at the Collège International de Philosophie. I have long wondered, and continue to do so, if there hasn’t been a casting error in the story of my life - no directorship is fail-safe. Indeed, I’m not a philosopher. Subjacent culture and the manipulation of philosophical reasoning are not routine subjects for me. Nor am I a computer specialist. When I happened to stumble into the field of software patenting I knew nothing at all. In my office we work with the internet a lot, naturally. But my secretary is the one who deals with that; I’m completely incapable of checking my emails myself and using a computer, because I was born in 1930 and I’m now in that time of life where learning something new frazzles the brain, which is infinitely regrettable as it happens.

I was invited, not in view of my talents, but in view of my involvement in a huge brawl concerning software patenting. President Mathias outlined the subject very well: What are we doing about authors’ rights and patenting? As regards the general theme of this day and a half of study, and the theme of Joëlle Zask’s lecture, software and the internet are not simply one and the same. You need many different programmes inside the internet, but the concept of software relates to a particular an area of activity, a world and a universe that goes way beyond the web, and concerns all current economic activities.

What is a computer program?

I’m going to risk using formulas I’ve learnt by heart. Again, I’m not at all a practitioner in this discipline. Software is a useful manual that makes a computer work and consolidates its calculation functions. So it’s a group of inter-connected mathematical formulae that facilitate the use of a computer in a logical fashion. As such it is a product of human nature. The grand statement was said by Albert Einstein in a moment of anger and conflict: "A mathematical statement is not patentable". This sentence dates back to the 1930s or 1940s, a bit before the war, I think, and it thus stands to reason that a group of inter-connected mathematical formulae is not patentable either.

In any case, this is the conclusion reached by negotiators at the European Patent Convention that was signed in Munich in 1973; the result of lengthy negotiations that defined the European framework for patenting. The Convention defined how registration and verification of patents are practised in Europe. A European definition was needed in order to create uniformity because there was legislation at a national level yet we form a common market. There is not yet a European patent; I’ll come back to that in the conclusion. But there was still the matter of giving common elements to these practises and concepts. The 1973 Munich Convention included Article 52, which plainly specifies in the second paragraph that software isn’t patentable. End of story. Everything was fine, and moreover, software was, as is logical, covered under copyright laws, that is to say, authors’ rights. There are some legislative differences between Anglo-Saxon copyrighting and authors’ rights, but they are roughly the same version of the concept that guarantees the author’s right to financial remuneration and gives the author a certain capacity to protect the integrity of his/her work during his/her lifetime, but not posthumously. It is in the author’s nature to try to fight against the denaturing and deterioration of his/her intellectual production; the author’s desire is for full public access to the work or object, in return for financial remuneration which isn’t subject to the equivalent of taxation.

For the next part of my reasoning, I’m obliged to leave behind the comparison between authors’ rights and patenting.

Authors’ rights. This is a system whereby the authors receive financial remuneration and are able to preserve their right to control their work for a certain time period. Prescribed in it’s concept and definition is the idea that there aren’t any production costs besides the human brain and that what is needed for production is limited to just a sheet of paper and a pen or pencil, possibly a paintbrush and some tubes of paint, but not much else. Things began to change during the 17th and 18th Centuries when the human mind started to invent not only artistic concepts and works but also specific types of material objects and services. In order to create and implement these objects and services, energy, materials and powerful equipment that had been borrowed had to be paid for, if it was to be worth it in any case. Therefore, in terms of the production of human mind this second category causes us to enter a world of expensive production and one which requires financial remuneration beyond that of just the author. This is where the concept of "patents" comes from, which, contrary to authors’ rights, prevent people from using a service without paying royalties. Therefore, by definition, patents constitute a temporary monopoly and essentially call for the implementation of a system that is able to police usage. You have to pay royalties in order to be able to use the patent.

That’s how things stood during that momentous time just over half a century ago when the computer science industry expanded rapidly as did the internet along with it. It all began with authors’ rights, and in thinking of software as a product of the human mind, which only required paper, pen and a good mathematical brain. Silicon Valley started in this way, and remained under the system of copyright for twenty years; it’s a well-known example to conclude my point. What I take from this example is that the regime of authors’ rights gives rise to great software producers and the rapid expansion of computer science.

Some of the larger companies began to think that getting money from people and charging people for the use of their software was better after all, and that it was also probably a good way to protect the secrets of their software from potential competition. And that’s how we came to use patents in this field. Patents first existed in the USA, a country with relatively short legislation. There’s no legislation relating to software patents. This changed the behaviour of service providers slightly and led those who had seen the implementation of such techniques in the USA to suggest that different European states change the 1973 Munich Convention that I mentioned earlier, so as to allow software patenting. At this point, a new conference was convened by Signatory States. This conference, held in November 2000 failed to do so. Disagreement ensued. Because of various details, but I’ll expand on this later, I’ll remind you that the Munich Convention is signed by all the member States of the European Union., including the members that gain accession today and some other countries also: Ireland, Switzerland, Norway, Lichtenstein and I think even the Ukraine and probably even Russia too. Involving countries outside the European Union means that Europe’s, the European Commission’s and the European Parliament’s political institutional authority is not being exerted over the European Patent Office. Written summaries destined for participating states come exclusively in the form of annual reports given to each of the governments, who generally speaking, have better things to do and have no reason to work together. The European Patent Office is therefore a free electron. It must be good to work there, under the control of absolutely no-one and making money selling patents, with little interest in increasing distribution.

In light of the failure of the November 2000 Conference, lawyers defending patents, that is to say all the big ones (Microsoft, Apple, Honeywell, Thalès, Alcatel, etc.), suggested to the European Union in February 2002 that they undertake the redaction of a proposal for a directive. It took ages and was very difficult; we talked a lot about it and the institutions were forceful in their handling of proceedings. The Internal Market Directorate practically abused its role, took charge of proceedings and wrote the Directive without consulting the Research Directorate, the Competition Policy Directorate and the Information Society Directorate (there were three of them that complained a lot about it) and that helps to explain the failures that followed. It’s very interesting. We are in a merciless field.

It has been said that the difference between software patents recognized in the European Union and those that aren’t, correlates to potential money transfers in the region of fifty billion dollars a year, taking into account all potential royalties. I’ve never gone and done the maths myself, but someone told me that never in its short institutional existence has the European Parliament had to handle one directive with such a huge economic impact. If you like, it equates to four or five times the value of the merger between Boeing and Douglas when it was proposed, or of the merger between Honeywell and General Electric. It was very significant.

In the end, this Directive came under the aegis of the Directorate General for Internal Market. The Bureau of the European Parliament saw that it was like giving strawberries to a donkey (the image is a traditional one). The Bureau of the European Parliament decided to entrust the Committee on Legal Affairs with reviewing the Directive, which could be disputed seeing as the Committee on Legal Affairs is there to suggest changes to the law and to the state of affairs, because law is never destined to be permanent. What are destined to be permanent are the principles linked to civilization, culture and the spirit of life, and what have you. The Committee on Legal Affairs has an assisting role and assuming responsibility for the Directive is not at all part of its original mission. We could just as soon have given overall jurisdiction for this Directive to the Committee on Industry, whose permanent core principles include scientism, continuity of progress and competition. We could have also given the Directive to Culture, since it relates to the state of one of the forms of human knowledge. But the Directive was therefore entrusted to the Committee on Legal Affairs, but two Committees are also asked to contribute their points of view: the Committee on Industry and Research, and the Committee on Culture.

At this point in time, I found myself in the bizarre role of President of the Committee of Culture. For the first time in my life I was responsible for cultural affairs, but in politics, more or less by definition, you are involved in many different areas, as everybody knows. When a Committee receives a document, the first thing they do is find a rapporteur for it. Drawing lots or rather, haggling over the negotiation meant that the social democratic party found itself responsible for the draft for the Committee on Culture. And why not? No-one knew what was going to happen, or really understood what was the real issue was. I can promise you here under oath that none of the eminent parliamentary members for the Committee on Culture in the European Parliament understood anything about it or had a clear idea of "software patenting". We had some work to do. This is how the situation arose whereby everything fell to the social democrats, my political party: the President of my political party was impartial, I wasn’t President of the social democrats for the Committee; I coordinated the work for the Committee. The political party was presided over by a charming English woman who was outstanding in the field of theatre, cinema, and "production of culture" to cite today’s media systems. She couldn’t find anyone keen to look into a subject that seemed uninviting, unapproachable, almost incomprehensible and, fundamentally, fairly uninteresting on first examination. This wasn’t the case for me, however. It was either 2000 or 2001 at this point. And there’s another European Parliament tradition, that when there’s work to be done that no-one in the relevant parliamentary Committee wants to do, it’s given to the President who has to deal with it along with the Committee Secretariat. The President and the civil servants of the Committee, write what they can. And that’s how I came to be in charge of writing the draft on software patenting. It was going to take some time, it was a first reading, and there was no time limit. So I learnt. Moreover, taking into account my father and the issue of family dynasty, and because all politicians are driven by the desire to protect their reputation, I couldn’t allow myself to fail, nor to let an anonymous Committee secretary write nonsense. I did my research; I led in-depth discussions within the Committee, I moved around a lot and I met numerous people, etc. And I began to find something quite astonishing, the immense and appalling arrogance in the financial world, a world I had turned my back on; my very own world. As a result, I was quite used to it. But world of finance became more savage when we asked those involved to give their opinion on the incorporation of the effectively juridical or philosophical concept of software patents from a legal perspective. And that’s annoying.

I won’t tell you all the ins and outs of it. We spent eight or nine months learning about it. A whole Committee, that’s quite something. And so we found some things out. Firstly, we discovered that although the Munich Convention explicitly says that software patenting is forbidden and that it is not possible to patent software, the European Patent Office case law has produced some thirty thousand of them, with great numerical uncertainty as regards the quantification of all these patents. We then discovered that there was no law on the matter in the United States and that the field was uniquely governed by case law, and courts which had had somewhere in the region of two hundred, two hundred and ten, two hundred and twenty thousand patents incorporating software to validate and quite quickly (we’re not really sure about these figures either). I’m being very careful with the terminology I use, perhaps you’ll see why later. However, we also learnt along the way that five complaints had been filed in the United States’ Supreme Court. I think these complaints were class actions that were filed against software patents. We don’t have the same vocabulary anyway; here we’re talking about "patenting computer-implemented inventions" and the complaints filed in the Supreme Court relate to "violation of the Constitution" as regards violation of the free circulation of ideas. This is a penal offense and would mean that Bill Gates, if he was ever convicted by the Supreme Court, would receive not only a huge fine but also a prison sentence, and he would take a dozen or so upstanding citizens with him. Therefore, we discovered that it’s all quite serious. We were in a complete haze; we heard a rumour that the Supreme Court of the United States really wanted Europe to introduce legislation to lead the way, and to police the tricky areas of the subject before they set about tackling the matter themselves, and then we tried to understand what had happened. What happened was that the European Patent Office and the American courts allowed themselves to deviate from the definition of what constitutes a patentable invention. In order to have invention, there must also be innovation and production through industrial techniques, indeed use of technology and public applicability, which are not exact legal terms, but come from the mind.

The key criterion is still "the use of technological resources". And that’s something that was specified previously in the definition. But what is technology? Originally, it was the use of resources other than the human mind, and resources that in fact used either material or energy, in short, forces of nature. As soon as we turn to the forces of nature there is a technological element. And that very efficiently clarified case law concerning the authors’ rights/patent split as regards the slightly ambiguous creations of the human mind, and the matter of software. Because, when a product is made with the help of a computer program, and by machines that use material or energy, you notice that the consumption of material and energy is minimal and that the material affects relatively few and limited procedures. In any case, we tried to distinguish the programme that was created by a purely cerebral act and that used no material other than the few thousand millionths of a watt that it takes to make a computer work, and nothing more. But case law, little by little, came to define technological resources by implementing a more rigorous terminology. And we slowly came to patenting; that’s how it started, not with the physical part of an invention, which wouldn’t be able to function without software (something that is not patentable), but with the entirely of this invention. Because most importantly, the software is an essential element of it and the invention itself produces effects, or rather it uses forces and produces effects, similar to forces of nature. To explain I’ll say that a computer program that is a teaching resource for a maths teacher, or a technological resource in the hands of a surgeon in the middle of carrying out heart surgery, is just a concentration of knowledge to aid rapid reproduction, there’s nothing in there, no energy, or anything else apart from intellectual production which serves as a manual for the operation. But at the point when the program in question is used to compress musical signals to create a CD-Rom, or some disk where you have four or five hours of music instead of twenty-five minutes, there is an effect on space and material. We used that a lot as an archetypal example in this business of ABS car braking. You know that the great contemporary invention in braking is that detectors can be put on all the wheels and their axles to see if the ground is wet, humid, slippery or firm, or what it is made of, and can be put on the angle of the bend to transmit information about the location of each wheel to the central computer when the driver put his/foot on the brake pedal. And the software allows you to change the braking effects on each wheel, which greatly improves car safety, and prevents skidding... basically, it’s all completely amazing. It’s obvious that this action presupposes that you have sensory detectors - we’re in the physical domain here, they are patentable without a doubt - and supposes the presence of effectors that are just as physical, they are also patentable without a doubt. But the heart of the system is a program that is intellectually defined to do this. It is of no interest as a piece of software because it relies on a whole physical system, and it is in view of this that we were able to imagine patenting it. But, quite clearly, we went about patenting a product of the human mind. To be very legally precise we had to say that the whole system was patentable, since each of it’s elements, including the software, was unique and inextricably linked to the other elements, and that it provided an innovative solution to a technological problem, but that the software itself could not belong to the claim of the patent since it did not produce the same effects in the material world. And that’s what we discovered when we tried to understand what was happening.

But we also discovered that the extraordinary vehemence of those involved in the debate. We’ll now take a look at the three-year dispute in the European Parliament over this matter. As far as I can remember, no-one had ever experienced such a forceful or strong barrage of emails in the European Parliament in terms of quantity and determination, from those on both sides of the argument. The Free Software Foundation conducted itself badly; it rallied some leftists who either behaved like dock-workers during angry protests or like the lefties in May 1968. As for the terminology of those in favour of patenting, it clearly indicated that whoever didn’t believe that patenting was useful for humanity’s advancement was either behind the times, stupid, or delusional. I was even on the receiving end of insults throughout the debate. It was all very fascinating, and really thrilled us of course. It was an intangible issue in terms of knowing what the potential economic repercussions of patenting were. We didn’t hesitate for a moment as to the fact that we were in the public space. The public/private space argument was dealt with somewhat beforehand. From our perspective, we were plainly in the context of the public space and in its mandatory rules and regulations.

First element: What effect on competition?

Second element: What effect on the production of software?

We also started to discover that the legal situation wasn’t clear in French and European law, and moreover it wasn’t clear in American law either. Because of this patents were somewhat under threat, but at the same time, the system of registering of patents wasn’t disastrous. Around thirty or forty thousand patents of this nature were created in the Euro Zone. This refers to patents that are linked to industrial resources and can therefore be patented by means of the industrial resource. As a result the software is clearly defined as patented. And that’s all very delicate. At the same time, let’s look at France: of the tens of thousands of programs that are invented each year, 95 or 96 percent are produced by young researchers, good mathematicians graduating from university who have quickly launched a start-up around a great idea. The mental fertility of those aged 25-30 has generated the largest number of great ideas. But by nature, a computer program is a chain of logical tools that follow on from one another. It’s a concentration of applied mathematics. Consequently, for Mr X, aged 24, creator of a good computer program which he hopes to use to create a small start-up, patenting presumes that the antepenultimate program he has used is potentially patented by Microsoft, and that because of this he will owe ten thousand dollars. This is preventative.

This could be preventative for serious reasons. An Association of the Federations for Large European Business Software Users, and large international companies, are fighting for the patenting of their software to ensure the possibility of keeping secrets and to protect themselves from the competition. And the terminology, here, proves to be completely extraordinary. If you want to put pay to a computer program - we were told this everywhere, we should have made an anthology or out-takes - you go and betray the French industry to the American industry or to the Chinese industry; basically you take precautions for the future against scientific progress. We could have put all these sentences on one program and have kept the end, "you’re going to condemn capitalism", you’re going to "condemn progress", "denounce us to the competition", etc. We could have then changed the starting sentences to no longer speak about software patenting but about banning children under ten from working, limiting the working day to fourteen hours, all the things that posed the same threat fifty years ago. We found exactly the same terminology to our obviously infinite disappointment. All of this drama didn’t mean that that they were wrong.

As philosophers, this calls you to reflect on the fact that we are in a confrontation of mental worlds that don’t communicate. And I must say that the defenders of free software: a huge federation, have themselves created several programs including a group of programs called Linux that haven’t been patented and that are available to use and copy free of charge. Curiously, I think the French Interior Ministry or one of our Ministries, and some large cities, (Paris for its part is looking on divided about the matter) have rallied together around free software to see what happens. It’s not really with any true conviction because France was virtually silent in the great final dispute because it didn’t know who to side with. The matter of competition is obviously a very significant one. We resolved to look into it in depth. It was greatly discussed. The first examination took two and a half years of work in the European Parliament. We all started out knowing nothing and having nothing but good faith and a real willingness to work, so that we might at last take a particular stance. Naturally, the matter of competition required the most work over all. Faced with the United States, China and also India (a gigantic power in this field), it was out of the question to lessen our chances of our industry in the contemporary world; completely out of the question. But we had the impression that we were being lied to in the details, I really should tell you what made the researchers uncomfortable: the argument was relatively defunct.

It seemed as every bit as important to block every invention of new correlative software that had been around for some time as to defend oneself against the competition’s inventions. This is so firstly because patenting made it possible to do so, but also because imposing a system of royalties made it very difficult for other people to use this knowledge.

We thought on the contrary; patenting served to slow down the use of competition, since it banned the immediate use of a product of human knowledge to benefit humanity. On the other hand, the software industry was born and had existed for fifty years under authors’ rights.

Earlier, I brought up the famous learning tools that are now symbolic of issues central to this matter. When small computer and a clever program can help a maths teacher to conduct a maths class more quickly and capably, the immediate result is that everyone thinks it is urgently needed (if it’s available free of charge) to teach maths in Africa. But the software costs ten thousand dollars. It’s patented. This is where the scandal starts. But in the strictly industrial world, it works more or less in the same way too, and accessing other peoples’ software is a way of developing competition. It’s very curious that a small part of European rightwing ascribes to this idea, because the European rightwing is striving to defend market competition and freedoms, much more so than the left; these are principles that relate only to non-patenting. Patenting is the authorised construction of a monopoly. And from the moment it is created, you are likely to see the majority of all software owned by a few very large companies. There’s also one more detail: a company can only allow itself to enter the world of software patents if it is enormous. Indeed, the legal services required in order to monitor worldwide industrial property in terms of competition, and simply protecting your own patents from the competition, are very expensive.

Basically, somewhat instinctively, we preferred to freely create new programs and prevent any monopoly in this domain. That’s why after a long battle and with around two thirds of the votes, the European Parliament amended the Directive after one examination. This Directive initially banned the patenting of computer programs, and as such upheld what was written in the past and required us to respect what had been written in the Munich Convention by the wise negotiators of the time: "software is not patentable". There is no reason to change the status of knowledge in European civilisation.

I want to add this: a program is an element of human knowledge. In the six thousand years that humanity has existed, we can not say that it has progressed significantly in terms of morals - that’s a subject for philosophy - we still fight as ferociously and kill each other; there’s no perceptible progression and I don’t think that in terms of aesthetic materials and artistic creation that humanity has made advancements either. All the beauty of the world already exists in Lascaux and Grotte Chauvet; we’re now into variety and differentiation, into multiplication and diversity, but definitely not into improvement. So humanity has only really progressed in the area of technology, and in using knowledge. We looked into the matter a lot. All the scientists say so. Human knowledge progresses through copying and transforming. Under the pretext that the computer and the internet are fabulous inventions of the turn of the century, this is essentially so that what has allowed for human progress during the preceding six millennia does not come to an end. We have remained true to the spirit of Albert Einstein and the 1973 Munich Convention. The vote in December 2003 was a thunderbolt in a blue sky. The Commission took it as an insult; there were comments that were very troublesome for the respectability of the elected members of universal suffrage and for the legitimacy of their election. But basically we voted. The Commission decided to take the matter to hand again. And once again, strange things happened.

The process evolved during this time, if I dare say so, and we noticed major changes in the field between the 2003 vote and the December 2005 vote two years later. The first change is that the Indian government decided against software patenting, India is not a small country as regards this matter; it is a major player. The Chinese, who had enough experts, trained two million computer specialists a year, generally to a high standard. They understood everything; have fun and benefit from patenting as long as the Americans allow it, but they are also willing to do anything, because on occasion their manner of reasoning gave the impression that they would side with Linux and free software. In any case, national rights are territorial rights, and it took us some time to realise that the demands of Microsoft, Thalès, Apple, Nokia, Alcatel, etc, call for patenting in Europe so that they could protect themselves from Chinese patents were irrelevant. We could ask for Chinese patents, there is a Chinese Patent Office that works a lot on a large scale and works quite efficiently it seems. There is an American Patent Office, all our companies ask for American patents. The world is therefore heading towards free space in patenting and at the same time towards restricted space, so we’ll see.

So the second element of work in this field: the American Federal Trade Commission. In light of this dispute, and maybe at the request of the Supreme Court, soon to process its complaints, the American Federal Trade Commission launched an in-depth study into the problem: patenting of software on its own. This means that this goes way beyond the internet, but the internet is still involved, and it’s a part of the subject. The results of this study were pretty significant and surprising. First of all, in the turnover of a large company, from the moment they started to patent, the cost of creating, purchasing, inventing and protecting software rose substantially. The budget assigned to creating new programs, which included no legal department costs, well let’s say a very small amount, under copyrighting, now reached 25-30% in a staggering rise that suffocated everything else. The second point of the Federal Trade Commission’s diagnosis: they think that the production rate of new software has significantly decreased, and that it could be on the way to drying up completely, because with patented material, it is only possible for large, established companies to produce new software. It’s practically a quasi-ban for small researchers.

During this time, the large software companies got annoyed, and the Commission that was no longer in a good shape, set about rewriting a new directive with several new characteristics. Firstly, the Commission would not consult the European Parliament or any of the experts auditioned by it. No consultation. Working in secret. From a Parliamentary perspective, to then make legislation is bizarre to say the least. Secondly, it was to be confirmed that the Commission has done none of the work itself, and that it had called upon external consultants, for the first and second examinations, to write the texts, including two or three that came either from Microsoft or Nokia. This means that the consultants had the double role of judge and defendant. And in these conditions, the Commission strangely came to produce an even stricter directive. It almost went as far as recommending the creation of general patent for computer programs "on their own", these words form part of the legal debate, which is surprising. And then we understood that the dates made sense and in the presidencies that immediately followed there were two countries whose governments found it hard to be neutral. One, the Netherlands, where there are a lot of start-ups of this nature and a lot of software companies, and where Microsoft has an enormous influence. The other was Ireland, who went from being last place to first place in terms of income per inhabitant in Europe, by means of extraordinary fiscal dumping. The results of this were that they succeeded in attracting some large companies of this kind to Ireland; for the Irish government it wasn’t possible to act freely as regards all of this. And so they wanted to con us; they submitted their new directive in April 2004 to a Counsel of Ministers (it was then a class B issue, under debate...). The Counsel very rapidly expressed their agreement in principle but it was subject to confirmation (an incomplete discussion). With just one government that refused; the Polish government bizarrely. Then the end of 2003 happened, and on two occasions, under the Dutch and Irish presidencies, the Commission again tried to pass this problem on to the Counsel of Ministers, in making it a class A issue, which means you simply note that everyone agrees on the matter, without opening a debate. It’s a roundtable discussion that notes general agreement without opening a debate but, if there is disagreement on the matter then the process is stopped. Both times, the Polish government was the only one to commit to disagreeing. It was quite astonishing all the same. We had a hard time understanding why. But we quickly bowed to the obvious: patents that incorporate computer programs weren’t found in Poland. Quite simply. And then, in the end, the European elections came around, a change of parliamentary employees and most importantly, of Commission members. The new Commission decided to raise the issue and the commissioner in charge, who happened to be Irish, said that he wanted to remove the directive project from the Council of Ministers, to amend it and take it to the European Parliament for it to be debated and for us to move on from it, whatever the outcome.

We all got ready for battle with great uncertainty. Everywhere, the barrage of media attention had increased and naturally so; the media coverage had become proportional to the financial means of those involved. I won’t give you the whole picture of the mass of documents and emails, or the canvassing or the invitations to dinner which flooded in. On the one hand, the symmetry had disappeared, while during the intimate battle of first examination there was relative symmetry. We enjoyed ourselves. I wouldn’t want to downplay the seriousness of the assembly, but all the same these small pleasures strengthened our convictions.

When the day of the vote arrived we were very worried. This stemmed from the fact that the majority was undecided. It’s clear that some of the members who voted for the freedom of the non-patenting of software at the first examination (two-thirds of the votes) had been turned. That was obvious. But we didn’t know how many and we didn’t know where the majority lay. On the other hand, it was the type of subject that if a single amendment disrupted the balance of the directive, all of its impact of could have been destroyed: indeed it would only have taken one clause in one paragraph to lean in support of future case law or even just to support national patent offices and the European Patent Office to change everything. Moreover, both sides desperately needed to reach the conclusion they wanted and would have preferred not to have a directive than one that favoured the other position. Situations like this don’t often arise. Technically speaking it was a situation that prevented all compromise.

On the morning of the vote there was a great joie de vivre because the Parliament was descended upon (they weren’t as brutal as the dock-workers, there were a few less of them, but it’s the same idea!). We had a good thousand uncustomary visitors who were more aggressive than amicable towards the European Parliament in both cases. So the morning of the vote, the Patenting Federation chartered a beautiful 22-23m (72-75 ft) river boat on the Rhine. We were at the European Parliament, which consists of two buildings straddling one of the canals. And the boat had two masts and a large blue banner, and written in yellow on it (for the colours of Europe) was "Yes to software patenting", "Vote for the directive", which immediately those who were in favour of free software sought to sabotage. They rushed to all the sports-shops in the area and bought five small inflatable rubber kayaks and on the backs of their t-shirts they wrote "No to patenting!", and such like. We witnessed the software naval battle on the Rhine in front of three hundred and fifty beaming ministers, packed onto the footbridge where the little kayaks were preventing the huge boat from sailing over to us.

You see at that point, each side was probably fifty votes from the majority; there are seven hundred and thirty ministers. And the majority was impossible to predict. Are you familiar with Condorcet’s paradox? It is possible for each unique brain or each unique machine to have transitivity of preference and therefore logic. But Condorcet showed that, as soon as a collective voted individually and in absolute freedom, transitivity of preference was no longer necessary or inevitable after voting. Besides, that is also one of the beauties of working in parliament and explains a lot of the foolishness in law-making everywhere. In light of this principle we feared that we would never manage to produce a totally homogeneous and coherent directive. This was obviously a danger. As rapporteur, I therefore suggested that the European Parliament (in fact it was said two days before) reject the proposal for the directive, and the patenting pressure group who felt they were in deep water also adopted this perspective, so much so that with 684 votes to 14 and 18 abstentions, the European Parliament rejected the proposal for the directive.

So what does that mean? Firstly it means that the Collective of large organisations for patenting are obliged to ask themselves some questions. They have to understand the principle whereby those who are elected by the people vote on laws and put pay to their mistaken attitude of us: not only were we capable of understanding the issues but in the end we were able to understand them very well... and were even able to read between the lines!

But on the other hand, the rejection serves to perpetuate the status quo. And the status quo is the European Patent Office and its internal legal system. They felt at liberty to patent integrated computer programs - I gave you the example of ABS, the software that assists with braking and this was its only function. As a computer program it shouldn’t be patentable, but as this is its only function, and is linked to this, why not think of it as included within the ABS braking patent that describes the sensors, the detectors, the effectors, etc.? All the same, it needed clarifying somewhat. In my opinion, the European Office will have to be very careful because it is clear that the parliamentary majority will be in favour of preventing any divergences observed in case law and in patent distribution. I think that we’ve arrived at a place of self-monitoring.

I would like to conclude this slightly lengthy speech by telling you about a mouthpiece, a small press group called European Voice. They are a private group that did the advertising for Europe where Microsoft and a few others were the main sponsors. Via the internet, European Voice created an annual voting system: a free vote to elect the European Commissioner of the Year, the European Manufacturer of the Year and the European MP of the Year. It was therefore under the sponsorship of Microsoft and probably using patented software that I was chosen by the voters as European MP of the Year 2005, we’ve not stopped smiling about that since.

I will finish by saying that I think we must retain the conclusions we’ve reached about this matter in order to find our bearings later on.

There’s one conclusion that is not philosophical but which is completely astonishing. It’s something that if you began a global discussion on this subject it would be worth elaborating on because it is directly linked to some things that you have mentioned. Every government administration has civil servants who are responsible for defining what is in the best interest of either the nation or of Europe. They are expected to do so impartially, under orders and with monitoring from politicians. This involves strategies for combating bird flu, measures taken in Iraq and regulations for dealing with chemical pollution, indeed everything, except industrial property and software patents. This is the only sector of all immensely broad government involvement where the United States government and European states have entrusted the definition of the nation’s position and the defence of its stance to external parties, and have entrusted a paid pressure group with speaking on behalf of the government. There is no other example of this. Since the national offices (the French national office is called the National Intellectual Property Institute (INPI)), earn a living from royalties of patents that they sell, it is difficult for them to be have be impartial enough to hand over such a huge domain. This situation created imbalance in accessing and processing information; an imbalance which was a stumbling block for half of all European governments at the end of the debate. When they all realised this the Council of Ministers and the Irish-Dutch quibblers came up against a courageous Poland. I forgot to quote another piece of information to you: both the Dutch and German parliaments voted on majority resolutions throughout the battle. During the second examination in May 2004, the national parliaments asked their government to withdraw their signatures of agreement for the proposal for the directive, in order to rally together for the prevention of patenting. And it’s not over yet; the two governments put enormous pressure on their parliamentary representation until the end of the debate. As far as the Dutch parliament is concerned scores aren’t settled everywhere.

So the first logical conclusion: on this one subject, governments don’t have an impartial organisation that seeks to find objectivity.

The second logical conclusion, which seemed serious to us, is that we are incapable of understanding ourselves. For the defenders of free software, the large manufacturers in favour of patenting are, for the most part, just a bunch of predators displaying the cruelty of the early stages of capitalism. The matter of competition is not taken into account or respected by this group of people, even though it was the matter of competition that made us begin this evaluation; as it happens it is because it seemed fragile to us that everything has changed. And conversely, for the industrial and financial aristocracy of the software world, and in the legal world, it’s very difficult to advance human progress when we’re faced with a bunch of incompetent fools who are irrelevant and who understand nothing. Words like "free circulation of ideas" and "access to human knowledge for all" are words that don’t penetrate this environment. These are words that evoke a general irresponsibility.

My third conclusion on the subject is that it’s not quite over yet because we still have to clarify the law. Furthermore, the imminent expiration of certain patents is a threat in many ways; the fact that tens of millions of Euros are at stake and the existence of changeable law courts will bring the debate to the foreground. At the moment, the attempt to patent software is currently an administrative work in progress, so low key stuff, but it will become quite public in a year or two when in comes to defining the European patent. I took the precaution of reminding you that there is a convention on how to manage the matter of patents in the framework of national rights; it’s through the Munich Convention that we are trying to define a European patent. It’s jurisdiction as concerns computer programs will have to be defined.

At the moment, the rejection of the directive means that the European Office can carry on as it was, but the Indian decision is a massive one and we are seriously in the process of wondering whether the United States won’t follow us. Indeed, when the Federal Trade Commission requests a study, it is a big deal if it is negative. Furthermore, the Supreme Court hasn’t yet spoken - and the Supreme Court could wreak havoc in the world of patenting, and the subject could change as a result.

And of course, you know that I started out in this matter for reasons that are philosophical at heart, and I dared to do so even though I’m not a philosopher. I wanted to maintain a freedom, and it also seemed to us that some principles, in their banal and archaic makeup, should be respected in their literalism: "free circulation of ideas".

So philosophers (that means you more so than me) in essence, was I right? What do you think about it? I’ll listen to what follows with humility.

Rocard Michel
Cottey Esther
Wormser Gérard masculin
Patents and Liberties
Rocard Michel
Département des littératures de langue française
Sens public 2009-09-14

Conference of Michel Rocard, Collège International de philosophie, Paris, 01/2006. The original french version is available in Les Cahiers Sens Public, "L'Internet entre savoirs, espaces publics et monopoles", n°7/8, octobre 2008.

Monde numérique
Politique et société
Condorcet, Jean-Antoine-Nicolas de Caritat (1743-1794 ; marquis de)
Monde numérique