Virtues and Vices of Open Source
Transcript of a talk by Eben Moglen

Listen to Eben Moglen's Remarks

I've been confused about this title, "The Virtues and Vices of Open Source" for quite some while. I'm sort of accustomed to titling my own talks and I couldn't exactly tell where this one had come from. I'm glad to see what it's about now.

I think I find the matter a great deal more clear. I'm particularly pleased that there are a collection of unavoidable ethical questions before the house. Three of them are eminently avoidable but they're all useful and interesting and it seems to me might be good to give some answers to them.

To begin with then let's take this phrase "open source" and do away with it; it's most unhelpful. We don't worry too much about the difference between open source mathematics and closed source mathematics because none of us has the vaguest idea what closed source mathematics would be. We don't worry very much about the difference between open source chemistry and closed source chemistry because despite the unfortunate tendency of the pharmaceuticals industry to believe that chemistry can be owned, molecules are themselves still the property of everybody. And at least the bad system of injustice about what to do with molecules says only that certain uses of molecules can be the property of parties in possession of a government monopoly. So we can avoid this question of open source if we recognize that what we're really talking about is whether knowledge about how computers can be used to help people's lives should be free in the way that mathematics is free, in the way that biology or chemistry or history or economics are free. We're really talking about whether knowledge ought to be ownable and people ought to be excluded from it.

This is, I admit, a very series ethical problem. And one of the great problems that I had with the 20th century was that though it's a great and obvious problem, people spend a lot of time trying not to answer it. The 20th century was fecund in the production of ways for the ownership of property and ways for the exclusion of people from knowledge. The 20th century did more than any period of equivalent history in the recent past of the human race at institutionalizing ignorance.

Fortunately it is over now. What we have instead is a society in which all elements of useful human knowledge, all things of beauty, all forms of technical understanding, culture, history, philosophy, literature, arts, all pieces of useful information, maps, train schedules, timetables and water diagrams and all the rest can be made freely available to everyone for the cost of making the first copy. As a couple of coauthors of mine, Karl Marx and Friedrich Engels with whom I cooperated a couple of years back on the thing called "The dotCommunist Manifesto" said, "All that is solid melts into air." And at the end of the 20th century, it mostly did. That is to say, the things that we had built since Thomas Edison and Henry Ford into physical, analog, objects for the transmission of knowledge dematerialized themselves into bitstreams capable of indefinite replication and more or less infinitely easy transportation.
And we therefore came face to face with a moral question, not just an ethical question: that is to say, what is it right to do in a particular situation?
But in truth a moral question: what is the nature of our obligation to people as people? And the moral question is this: If you can give everybody something of value or beauty or importance at the price that you can give it to anybody, what is the moral case for excluding anyone from it? If you can make all the bread that the world needs by baking one loaf and pressing a button, what is the moral case for ever charging more for bread than a starving person can afford to pay?
If bread were digital, hunger would be preventable.

Bread isn't but knowledge is. We now live in a world in which ignorance is curable, in which cultural deprivation for economic or other social reasons can be bridged by the use of digital technology. A project on which I am spending a good deal of my time of late, One Laptop Per Child, is an attempt to provide a platform for the universalization of knowledge for the human race outside the grid of electricity and beyond the reach of telecommunications, out past where those twisted copper wires come to an end and measures for the digital transportation of knowledge have to get a little more clever to become complete.
That world is a world enabled by software, which is as crucial to the 21st century economy as steel was to the economy of the 20th.
Software is what the 21st century is made of as steel was with the 20th century economy was made of.

In the making of software, the question of the morality of unfreedom and deprivation arises in a way that it did not arise in the century of steel. Backyard steel furnaces is a lovely idea, but it doesn't work because steel has high marginal costs and every ingot means energy and knowledge and capital equipment. But in the world of software, every copy of a program that makes your life better makes it possible for you to talk to people for free or makes it possible for you to grasp your own ideas or make your music or paint your pictures or make your drawings. Every instance of a program is nothing more than an act of sharing between people, as every sentence I am speaking is nothing more than an act of sharing between people right here right now.

So the question of the ethics of free software (pardon me for sliding that phrase in there; that's really what we're talking about is free software-free as in freedom) so the ethics of free software are very simple really-fundamentally so simple that it's hard to think of how I'm going to spend the rest of the time if I only talk about the unavoidable ethical questions because there's only really one: Why would you deprive people of that which you could give them for nothing?

Now three of the avoidable ethical questions on the sheets you have in front of you arise from the question, Would it be wrong to force people to free their knowledge? And the answer is yes it would be wrong. "Die Gedanken Sind Frei" is a song that's been around since the 13th century. Thoughts are free, and free means free also to keep to ourselves. No one has ever suggested as a part of the free software idea that anybody should be compelled to make free any idea that he wishes to keep to himself. No one has ever suggested that people who have based their "business models" on the withholding of ideas from free exchange in the market should be compelled to exchange them freely. The free world's success is based not on compulsion but on the freedom to offer: "Here I made this. You might like it. Take it it's free." Nobody has ever attempted to free anybody's proprietary program among my clients. I've never thought of going to court to ask somebody else to free an idea he wanted to keep secret. My only goal has been to prevent the owners of ideas from inhibiting free exchange which drove down the value of their hidden stuff.

So what's the virtue of this free software? It restores people to the act of programming. When I was a kid (a time so long ago that there are only two people in the room who remember it) when I was a kid I wrote programs. And I was allowed to write programs because I was allowed to read programs, because I lived and worked in a place where I was allowed to make the programs that I wanted to make out of the things other people had been making first. I lived in that world as a skilled artisan operating on material bequeathed to me by prior generations of artisans in and out of the companies for which I worked. We exchanged ideas freely and we grew up to be programmers.

For contingent reasons which I have analyzed in some of my published writing, between 1980 and 1995 or so that basically ceased in the West, and programming became an activity appropriately described by Craig Mundie of Microsoft as "Never give anybody the source code to anything." In such a world it is hard to learn to be a programmer. In such a world it is hard to turn computers to creative use. You need not just skill or daring or the desire to pursue what's neat, you need permission, you need a license, you need to be allowed.
The effect is to concentrate technological knowledge in a few very powerful hands.

The other effect is to decrease the quality of the technology itself. Two forces, then, have given themselves awareness on the basis of those changes: a set of forces which said, "Gee, we could make better software if we all shared," and a set of forces which said, "It is immoral not to share under the conditions into which we as a technological society are now moving." That division of impression-we are here fighting for the right to make better programs and be better programmers and use our knowledge more effectively and those who said, we are fighting to maintain freedom in the face of ownership gone mad-have been often defined by the distance between two phrases, "open source" and "free software." But in truth, the premises of that debate are based on a kind of historical foreshortening. Galileo is not an open source guy, but he's noticeably a free software person. His concern is with the free exchange of scientific ideas and with the prohibition of the prohibition on exchange of those ideas. Should we think of him as demonstrating the vices of open source?
Well, we had better because he's a big part of why we're standing here right now, and if there are any vices in open source he brought them in....

There are facts about the world that are everybody's right to know and that it's nobody's right to deprive anybody from knowing. But the way to deal with that isn't to demand that the church release everything it knows about astronomy.
The way to deal with that is to figure out what the truth of astronomy is and publish it and demand the right to be read, which is all we've ever done. The practical consequences of 20 years of that insistence are that we have changed the global software industry irretrievably and it will never be the same again.
We have prevented permanently by legal and technical means the vertical ntegration of the software industry in the world. There will never be another Microsoft, never, because the corpus of free knowledge now available is so strong, so powerful, so deep and so rich that it will never be possible for any party to gain again the kind of temporary monopolistic control over the exchange of technological ideas that briefly characterized the Microsoft era.

We are never again going to see a world in which the computer next to your eyeball is primarily engaged in spying on you. It's true, of course, that you can communicate with computers that spy on you; you do it every single day-I guarantee it-every time you run a search. But the one closest to you, the one that lives in your pocket, the one that your life is in, you can make that free as in freedom for yourself. You can be sure that what it does is what you want it to do and not what anybody else wants it to do. You can be sure that powers, limitations are contained in the machines that run your life.

Without that, freedom in a technologically advanced society is very, very perilous in its condition all the time. Imagine for yourself a world in which government has decided that computers will do certain things with word processing documents, summarize their content then email the secret police or just contain a little token which allows every document to be traced back to its source. Oh that wouldn't be fantasy; I'm sorry that would be reality in the Microsoft world. So that everybody is traceable, so that everything that anybody thinks or writes or exchanges has become an item in the indictment of them for inappropriate thought. Surely the first step you'd want to take would be to rip those computers out by the roots, haul up the technical civilization that made unfreedom possible. But you'll never have to do that because we did it first for you. We prevented it from happening.

The political consequences of changes in the global software industry are easy to see if you're not managing quarter to quarter or worrying what your stock value is going to be six months from now. But that doesn't mean that that's the only reason for being interested in what this stuff does. The act of preventing the vertical integration of the software industry, the creation of the commons on which everything now rests has changed lots of other facts of industrial life too. The European Commission reported last month that more than 20 percent of the annual investments in software in the European Union are now investments in free software amounting to some 22 billion Euro per year. There is not a major user of computer technology in any of the globe-girdling industries, not in finance, not in media, not in pharmaceuticals, not in construction, not in banking who does not use free software to run its business. Sure, there are businesses which are more Microsoft less Linux and businesses which are more Linux and less Microsoft but heterogeneity is a fact of life and so is choice.

The structure towards which we are moving in the world of software is a structure which corresponds to larger changes in 21st century economies
as products become services, as things that were non-commodities acts as specialized manufacturing in the 20th century become commodity technologies in the 21st where the value lies in the services that surround the product.
Twentieth century commoditized the road for the car to drive on and built an economy based around the public utility service of cheap ground transportation.
There were, of course, hidden uninternalized costs in that activity and as the icecaps melt we begin to see what those costs were. But what we did in the socialization of ground transportation through the creation of an immense socialist road network in the United States was to enable what we were pleased to call the market economy to flower.

The same fundamental process now goes on across a whole range of activities in the 21st century economy built around the commoditization of what were products and the use of those commodities to enable enhanced application of services for profit. This model is no stranger to the thinking of the executive suites in IBM or Sun Microsystems or Hewlett Packard or any of the other richest and most intelligent corporations in the world. They too now perceive that the growing economy of the socialization of R&D in the world of free exchange of ideas offers them enormous opportunities. They now take shareholders' assets and put those assets in commons in order to increase the value of the commons to the society as a whole.

I need point out only the enormous degree of patent promising, assuring, half-donating and even actual pooling going on in the IT industry. One of the two greatest pillars of the patent system at the end of the 20th century as a basic agonizing reappraisal of the value of the 20-year ownership of an idea goes on.
We are, in other words, watching as this little experiment in treating software ideas as free like math becomes a spotlight cast on the actual mechanisms of change in the economy of the 21st century.

And we begin to address some basic questions of global fairness. When the Massachusetts Institute of Technology puts its entire curriculum on the Web it is striking a blow for social liberation. Ask yourself this: Before the beginning of the 21st century, what percentage of the Einsteins who ever lived learned any physics? My guess is .00001 percent maybe, and that's probably way too high. Radical intellectual deprivation has been the basic principle of human life since before there was the phrase to utter. That only a few had access to knowledge is as characteristic of the shape of human society before now as any other single property of social organization and we're about to change it.

Now I suppose that somebody somewhere may still be willing to argue that there will be more shareholder value in locking up knowledge than there will be social value in letting everybody learn. But it isn't going to be the opinion that rules the 21st century because it isn't an opinion which can be generalized into any conception of global human justice. So this open source-this stuff that we are supposed to consider as having virtues here and vices there in the world of commercial IT-this is not this little phenomenon of how do we share our code?
This is not primarily a confined story about, Is it better to have one Microsoft or a million little consultancies? This is not primarily a story even about whether socialism is dead. The sad truth is that the story's answers are already very clear. The reason that this is such an important story, this ethics of open source, this freedom of free software, is that it's the beginning of a fundamental alteration in how we think about intellectual accessibility in cultural justice. It's the beginning of the end of the campaign against ignorance. It's the beginning of the recognition that all the brains that the human race produces are brains not to be thrown away. It's an attempt to make the motility of ideas as great as the need in human civilization. It is, I regret to tell you because I know how deeply, deeply immoral a thought this is, it's an attempt to get to a world characterized by the idea of, "From each according to his ability and to each according to his need."

And on the question of what the virtues and the vices of that approach may be, I leave it to the commentators to decide.
Thank you very much.

Professor Eben Moglen of Columbia University is the chairman of the Software Freedom Law Center and general council of the Free Software Foundation. In addition to the Free Software Foundation, Moglen has represented many of the world's leading free software developers. In 2003 he was given the Electronic Frontier Foundation's pioneer award for efforts on behalf of freedom in the electronic society. He gave this talk March 26 as part of a series on IT, Ethics, Law, and Society sponsored by the Markkula Center for Applied Ethics, the Center for Science, Technology, and Society, and the High Tech Law Institute.

Redistribution, commercial and non-commercial, of this article is permissible as long as it is passed along unchanged and in whole, and credited to Eben Moglen.