segurança computacional > english texts > http://www.cic.unb.br/~rezende/sd.htm > english: FOSS Manifesto: Addenda

Best and Worst Practices in Top-Down & Bottom up Legislation

Talk presented at Brazil's Ministry of Foreign Relations - Itamarati

August 23, 2004

Maureen O'Sullivan *


Introduction

All over the world, migrations are being made from proprietary software to free software by companies, individuals and governments, whether at local, regional, national or trans-national levels. Laws relating to free software can be loosely divided into two categories: procurement law and intellectual property (IP) law. Laws promoting the use of free software have been introduced, generally not successfully, in many Latin American countries, including Brazil and in several other parts of the world. This has been happening since about 1999 or 2000 and I believe that France was the first country to try to get the state to favour free software over proprietary software. Other policies elsewhere may be aimed at getting free software an equal consideration with proprietary software, and this is the subject of many discussions in Europe at present. Arguments for using procurement law to favour free software include its security, stability, transparency, competitiveness and cost effectiveness, among others, whereas proprietary software is insecure, opaque, it ensures lock-in and is overpriced. Free software, however, tends not to be so user-friendly at a superficial level, although this is changing as time goes by.

My talk today does not focus principally on the aspect of policy and law relating to procurement because the deployment of free software throughout the world is already a fact – whether governments choose to legislate in order to promote it or not is generally a political matter. If governments have already decided to adopt free software over proprietary software, their roadmap for doing so can be based on practice rather than legislation. Sometimes the use of procurement law to favour free software use can have a negative effect of drawing too much attention to a particular region or country, which may cause a backlash from multinationals with a lot to lose. On the other hand, the use of such laws can have a positive result in that it advertises to the population in general the struggles which face governments when they have to juggle their citizens’ interests along with international trade obligations and bilateral trade negotiation pressure. An alternative route to legislation, not subject to lobbying to the same extent, is a declaration or a simple statement of policy, or other imaginative approaches.

I attended a conference in Ireland last April where a German politician said that in his region they would pay local firms or authorities €200 to switch over. In Extremadura, Spain, the regional government has provided one computer per every two children in public schools, starting last year. Technicians went into every school in the region one Friday evening and the entire installation had been completed by the following Monday morning. Training had already been provided so the transition was quite smooth. So, there are many ways of implementing a decided policy and those are just a few examples. That really is all I have to say about procurement law and practice during this talk.

My talk is divided into a number of parts and focuses on the phenomena of software, proprietary software and free software. First, I deal with theoretical issues relating to intellectual property rights (IPRs) because I believe that unless you understand at a profound level the justifications for a law or a group of inter-related laws, it is very difficult to examine them critically and assess the veracity of what they claim to do. Second, I discuss free software “laws” – referring to how the community has regulated itself and I analyse how intellectual property (IP) laws interact with this type of informal regulation. Third, and finally, I discuss options, including a legislative project of mine which, I believe, is an important step towards the protection of the free software resource and informal regulation throughout the world.

1) Origins of the Modern Construction of Intellectual Property Rights: Moral Rights, Consequentialism and Legislation

Introduction

Modern day intellectual property rights (IPRs) are an invention of Anglo-Saxon or Anglo-American law, although intellectual property (IP) laws in some form have been found in ancient China and medieval Italy. There are various theories which attempt to justify IP laws, most derived from moral rights and arguments about incentives that human beings need to increase their creativity. There is also the question of top-down legislation which does not have a strong apparent justification: it just is.


A) Moral Rights

The English philosopher, John Locke, is the most frequently cited thinker by IP lawyers in the common law tradition, to justify a moral rights based notion of IPRs. It is important to start here because Anglo-Saxon ideology dominates world legislative practice in the realm of IPRs. By moral rights, I mean rights that are inborn or innate: we possess them because we are human. Locke’s work is used a lot by IP academics who like to use the phrase “Lockean-style IPRs” without any further insight. It’s very obvious that anyone who uses this phrase has not read Locke’s work on property – incidentally, he did not write on intellectual property and the whole conflation of these two very different resources is objected to a lot in the free software community. It is a very powerful tactic to appeal to moral rights in order to justify ownership and exclusion because moral rights are essentially human rights and our instinctive reaction is to be protective bout them. In order to do so, with any credibility, however, one must be faithful to Locke’s writings.

Locke based his philosophy of property rights on the Book of Genesis which said that God gave the Earth to mankind – not womankind - in common, meaning that there was no private ownership. Locke believed that God’s intention was that we should make the best use of God-given resources and he claimed that enclosed land was ten times more profitable than communally-owned land. Men could enclose land under two conditions: first, land of the same quality and amount should be left over for everyone else, which arguably is impossible and second, goods could not be allowed to spoil or go to waste. The typical IPR lawyers’ approach, when relying on Locke’s philosophy to justify IPRs, is to excise these two provisos and thereby use Locke to justify an unlimited right to enclose land and by extension, the subjects of intellectual property.

This portrayal of his work is a misrepresentation and unlikely to be a misunderstanding, as Locke only wrote 6 pages on property. Clearly, there is an agenda in using Locke’s work when applied to IPRs because it facilitates the construction of ideas of ownership and piracy. Interestingly, Locke did not say that the encloser of land could keep all the profits for himself. If we are to follow his line of argument logically, as land was given to mankind in common, the profit from enclosed land should be divided between everyone. Locke went further: he said that in cases of extreme necessity, the poor could take from the surplus of the rich and so the occupation of underused private land or, indeed, software by the poor is justifiable, on an unmodified reading of Locke. Yet, those who quote Locke to justify their conservative perspectives like to obscure the socialist undertones and overtones of this work just like colonizers have often re-written history in order to justify their actions.

However, Locke was talking about a scarce resource and subjects of IP laws are not in short supply. For instance, ideas abound and expression is a seemingly limitless possibility. Arguably, therefore, whereas you cannot sensibly apply Locke’s theory to justify private property, if you are seeking to establish a moral rights-based theory of exclusive rights, this may be done in the case of intellectual property, if the resource is not perceived as being scarce.

Locke’s thesis was that as humans, we should make the best possible use of God-given resources: the subjects of intellectual property protection, such as expression and ideas are not limitless but, rather, are unquantifiable. IPRs create artificial scarcity among these subjects, limiting the ideas accessible or the means of expression of those ideas. In the case of software, this may be particularly unjustifiable, given that sometimes there is only one way to express an idea but judges who are “techno-neophytes” may not always appreciate these situations. The moral right to software access and use should be seen as universal because in the same way in which land was granted to us by God for all our use, according to the Book of Genesis, software is a human-made resource, created by some and given to all. Privatisation in the case of free software has no moral place – it would be nothing short of piracy because private companies should not be able to profit exclusively from the efforts of so many.

C) Consequential or Utilitarian Rights

If property or intellectual property rights are not justifiable from a moral rights perspective, another line of argument is often used and, indeed, widely accepted. If human beings are not able to profit economically from their endeavours, they will not create, so goes the argument. Once again, this line of reasoning is based on a land or real property law analogy: when you can enclose and exclude, this will awaken a creativity in some individuals and, in the case of IPRs, without the right to exclude there will be no public domain because there will be no creation. This is nonsense, of course, the vast majority of writers and musicians don’t earn anything from their work and, indeed, much of the Internet was constructed by hackers working in their free time.

Another side to the consequentialist argument is that regulation of individuals is necessary, both in terms of assigning rights and keeping people out. The portrayal of unregulated society as inherently disorderly and wasteful is nowhere better highlighted than in an essay written by Garret Hardin, entitled “The Tragedy of the Commons”, which was published in Science in the 1960s. In this essay, Hardin argued that when land is held communally, it invariably becomes depleted as everyone uses it without putting anything back. Applied abstractly to intellectual resources, this would probably mean that if someone goes to the time and trouble of creating something and is not able to exclude others, those others will copy the resource, pay nothing to the creator and try to profit from it themselves. This, in turn, will kill the creator’s incentive to go on creating. One of Hardin’s solutions is to privatise the resource, to institute strong property rights that encourage productivity whilst preventing free riding from taking place. The history of the creation of the Internet would seem to discount the privatisation aspect of Hardin’s thesis without further discussion – most of the software there is free. His other alternative, somewhat Stalinist, is to regulate externally because in his view, individuals, or groups of individuals are incapable of doing so themselves.

As a matter of fact, this view of communal resources as subject to underproduction or total disorder is a distortion of reality. The failure of Hardin’s thesis and of those who argue along similar lines is that they view the world in black and white, in terms of resources being owned or unowned. There are no middle paths in his perception and yet, intermediate property rights among indigenous groups - I classify hackers or programmers of free software in this way - are widespread, well-established and respected. An intermediate property right is where a resource is held communally among a particular group of people, subject to certain negotiated rules but this resource is not open to the world at large. Such resources are known as “limited access commons”. A good example of how external regulation can interfere with intermediate property rights can be seen in the case of Canadian fisheries and cod where there are certain lessons to be learned for anyone involved in managing free software policy.

At one point in time, about 500 years ago, cod were so plentiful that they were almost jumping out of the sea. Along the Canadian coast where this fish abounded, local fishermen in more recent times managed their area by sharing fish in a particular cove amongst themselves but residents from other areas who “trespassed” were prevented from coming in by the use of loud speakers. This was a means of shaming errant individuals and is an effective enforcement mechanism for locally devised, informal rules. When the Canadian government decided to regulate this resource, they adopted Hardin’s two-pronged approach: privatise or regulate. They disregarded the first option and instead chose the second one. They assessed the Canadian coast as an open access commons, meaning that there were no rights there at all, no ownership or possession or management of any sort and they effectively converted a well-run, limited access resource, in which local actors develop intermediate property rights, into an unsupervised free-for-all. Their actions meant that huge trawlers were able to come into the coast and by 1992 they had cleaned the waters out because fishing techniques are now so sophisticated. The fisheries were temporarily closed in 1992 for a maximum of three years it was claimed, with a loss of 40,000 jobs. Twelve years later, the fisheries still have not reopened and cod is now a scarce resource. While locals were involved in managing the fisheries, they thrived: once excluded, with their rights ignored, the fisheries became totally depleted. Ironically, one of Hardin’s solutions was the cause of the decline, not its saviour.

IPRs can be seen as a type of external regulation and have been introduced with a number of justifications: that the rights are inherent, although this is difficult to establish as none of us are true inventors, we merely stand on the shoulders of giants; that the rights are necessary because without them humankind will not create, however, it can be seen that external regulation can often destroy the very thing it is intended to protect; and privatisation has only really served to enrich a few at the expense of many.

D) Legislation

One other way in which introducing IPRs can be explained is that they are there because the legislature has approved them. This is a very formalistic approach to the law and seeks to explain the existence and validity of the law by its presence on the statute books or in case law, without delving deeper into more complex questions of legitimacy and morality. Before doing so myself, I wish to talk a bit about free software “law”, which is a species of “ground up” law in order to compare and contrast it with the theoretical approaches to and examples of legislation from the top down.


2) Free Software “Law” in Contrast with Intellectual Property “Law”

There’s always a big discussion about what law is: whether it gains its legitimacy from being instituted by a legislature, from being recognised or invented by a court, whether international customary law is really law and whether grassroots or ground-up regulation possesses any qualities that we normally attribute to what we call “law”. Regulation of the people by the people is tinged by the media with the label of “mob rule” and in the case of passionate reactions by a community to a horrendous crime this may be the case but it cannot be applied across the board. In civil matters, for instance, and especially where the management of communal resources is concerned, studies show a surprising level of order emanating from local management of these same resources. Such is the case of free software and I now propose to look at some of the regulatory practices within free software “communities” to highlight these examples.

When software was first developed, copyright law was not applied because value was to be found in the hardware, not the software. Each machine needed its own operating system but this changed when Unix was rewritten in C which made it portable. The network externalities attached to portable software vastly increased its potential and this was seen in economic terms by the US legislature which in 1980, after a four year study, decided to bring software under their copyright laws. One of the Commissioners later commented that instead of creating new bottles for new wine, they protected it under existing ones, although the decision was not universal. First, this was a pragmatic decision for the US because by protecting software as a literary work under copyright law, this was effective internationally under the Berne Convention. The US did not tolerate any dissent from their policy: Australia, for instance, considered protecting software under a separate category in their copyright law but was promptly threatened with a trade war by the US and their compliance was thereby achieved.

Second, this decision was a lazy one: it meant that no new law had to be invented. This should have been the case, of course, but that point is of historical interest only. Copyright protects expression, not ideas and, in this way, it claims to encourage competition. In fact, copyright has facilitated the growth of a monopoly which has dominated the software industry ever since. Free Unixes continued in existence after the introduction of copyright law and the licence style they used, the BSD licence, is very liberal: it allows access to the code, contributions are not encouraged and proprietary derivatives can be made, provided that authorship is correctly attributed. BSD is low maintenance and demands little of its users and would not have facilitated the massive growth in free software in programs such as Linux or GNU/Linux.

Richard Stallman, who started the free software movement, decided that this licensing style was not satisfactory and that if someone benefited from using code created communally, if they wanted the benefits, they would also have to share the code of any of their published modifications. (They can create private forks but these will not benefit from community peer review and maintenance).

The GNU GPL, the licence now used in up to 80% of free software projects worldwide was written by Stallman in order to protect free software in the way he deemed appropriate. It was, in a sense, a “sui generis” or a specially tailored law to protect free software, an informal law, a type of recognition of intermediate property rights and a local regulation by local people with informal but highly effective enforcement mechanisms. At first, the GNU GPL was used principally among other free software users and violations of the licence were uncommon. To this day, they have remained uncommon as participants in the free software community, like fishermen in Canadian fisheries, have strong social bonds and these provide incentives to comply with the licence. Unlike proprietary software licences which may be violated in 100% of cases in some countries, free software licences are almost never breached and are only recently making their court debuts.

External actors who participate in some way in the free software community also comply even if not required to do so by any licence: A good example is IBM’s support of Linux, which uses the GNU GPL and also of the Apache Web Server, which uses a BSD style licence. However, IBM makes many voluntary contributions to Apache, on which Apache’s growth has partly relied.

Non-participating actors have no incentives to comply with any free software community norms and if threatened by the success of free software, will not try to upset these community norms because this will be ineffective but what they can do is to try and disturb the licence through the legal system. This is because the GNU GPL in addition to being a statement or an informal codification of community norms, also operates like an ordinary software licence, with a few exceptions.

All software is automatically copyrighted, which restricts users’ access. Licences grant permission to use the software under certain conditions. Proprietary software licences are very restrictive whereas free software licences are permissive, albeit in different ways as I’ve already explained in the case of BSD licences and the GNU GPL which are at opposite ends of the spectrum in the free software world.


3) Options

As a species of legal protection, the combination of copyright and contract law is not the best way to protect any type of software. Copyright varies throughout the world, especially in terms of authors’ rights, and it does not guarantee that the same legal recognition will be given to a piece of software from one place to the next. The licences, as contracts, are also legally vulnerable in that they depend on contract law, which also differs from country to country and which, in any case, only binds parties to the contract. Where there has been no agreement, copyright, as it binds third parties, can be relied upon but, as already explained, this is not totally secure.

Obviously, there is a great difference, however, in terms of the bargaining powers of the parties and who can or who is likely to sue for what in free or proprietary software realms. Community pressure would need to be extremely strong to have any effect on multinationals to whom the concept of society or community may be alien. There are a number of approaches, however, that may be adopted in order to address this issue.

1. Do nothing. The watch and wait approach - wait no longer – the SCO vs. IBM case could be quite threatening for free software development.

2. Test the GNU GPL in court – the FSF’s general counsel’s approach – a highly risky strategy, in my view. This guarantees little as a positive decision in one court will not have universal effect.

3. Creative Commons-GNU GPL – this is a good idea because at least it offers security within a particular jurisdiction. It will be less effective in Europe, however, where the Union has 25 member states, a lot of inter-state collaboration and different copyright laws in every country, although there has been some level of harmonization at EU level.

4. Introduce an international, harmonized law specifically tailored to free software. This would solve many problems and could clarify a lot of doubts. It could be achieve through the UN, through international customary law or on a more piecemeal basis at EU level, for instance and in trading blocks such as Mercosul. This would not have to be a piece of legislation: it could be a declaration.

5. Take this law and pass it at regional, municipal and national level. If this is not possible, use it as a declaration for political purposes.

6. Make sure that the law is specially tailored to the resource it is governing or protecting. This would mean that the creators of the resource should not be excluded from the drafting process.

7. Tap into my democratically elected Free Software Act for some ideas. This codifies the terms of the major licences but also addresses issues such as creating exemptions from patent laws, changing the penalties from economic damages to a reasonable time to rewrite infringing code in the case of inadvertent copyright infringement.

A final brief point: the democratic deficit in today’s society is about as broad as the digital divide. Free software is a resource which belongs to its creators and to those with whom they want to share it. There are many potential legal threats to its continuance. From a democratic perspective, it demands appropriate protective legislation and a government which respects local laws and protects them rather than serving the interests of multinationals who seek to privatise and profit from resources such as these. Brazil has taken an extremely brave step in opting for free software. You have also been extremely wise in quickly adopting the GNU GPL to your national laws. You should now go a step further and deal with issues such as creating exemptions from patents for free software; indemnifying in some way free software developers from vexatious law suits; and also, perhaps, harmonizing other free software licences with your national laws.


* Maureen O'Sullivan is Lecturer of Law at the University of West of England, in Bristol, UK. She presented this talk at the ministry of Foreign Relations in Brasilia, Brasil, on August 23, 2004, by invitation of the National Institute of Information Technology, and prof. Pedro Antonio Dourado de Rezende on behalf of Projeto Software Livre Brasil.