The field of software is a fairly new one. About twenty years ago, people started doing lots of programming. There was no clear basis in tradition for whether software should be considered property or not. There were several possible traditions you could consider extending by analogy to the area of software, no one of which was inevitable, or necessarily more right than the others. One of them was the area of mathematics. Some people believe that a program is like a mathematical formula, and the tradition for mathematical formulas is that they couldn't be property. Another thing that programs are like is recipes. In fact, of all the things that people encounter in daily life, the recipe is the one that's most like a program. It's directions for getting something useful done. And recipes, also, by our traditions cannot be property. There was a third thing, and that was literature. As it happens, literature traditionally could be property. Programs have something in common with all three of these things. They don't have more, particularly, in common with literature than they do with mathematics or recipes, but the decision was made to treat them like literature because the people who had the most to gain from that particular decision were allowed to decide.

But this decision was not good for society as a whole. It's incredibly destructive, and it feeds an overall destructive tendency in this society which we can see in every area. And that's why, having observed the results, I came to the conclusion that it was unacceptable, that software cannot be property, and that I would dedicate my career to bringing that about.

If we look at the traditional forms of intellectual property, we can see they were all designed to fit into niches where they could exist without doing great harm to society. These niches were often created by the technology that existed at the time. For example, consider the area of books, for which copyright was invented. At the time, books could only be copied by mass production. Earlier, books were copied one at a time, by hand, and at that time there was no idea of copyright. In fact, people didn't even have the concept of strictly distinguishing their original book from a non-original one. It was a continuum. They recognized that some writers wrote more new things into their books, while they copied passages [from other books], or they might just be copying entire books. All of these things were legitimate, useful activities, though useful in different ways. But when the printing press was invented, it suddenly became for all intents and purposes impossible to copy a book by hand. Not that it was any harder than it had been before, but by comparison with making a copy on a printing press, it was so ridiculously impractical that no one would want to do it. The situation where books were made by mass production lent itself to copyright in the sense that copyright became a restriction on one particular kind of industry---a form of industrial regulation. It didn't take away the rights of the reading public in any way, because members of the reading public couldn't copy books anyway. They didn't own printing presses, so it was effectively impossible for them to copy books. But this niche was created by a particular technology, and changing technology can make it go away again. The technology of computer programs has never been like that.

[I have since learned that people did copy books by hand to a significant extent after the printing press was developed. What I said was true in recent times, but in England in the 1500's (and perhaps later) government-enforced featherbedding among printers made printed books fairly expensive; as a result, poor people might more easily afford the time to copy a book than the price of the book. But these people were not prosecuted for copying or told it was wrong. Copyright was applied only to authors and publishers. It was treated as an industrial regulation.]

In the time that I've worked as a programmer, I've watched things change from a situation where people generally shared work and generally used work that had been done in whatever way was useful, to one in which is nearly impossible to utilize work that other people have done, where people feel uncomfortably cynical. Constantly, a person will describe a great piece he's doing that's so exciting, you ask him will I be able to use it?'', and his face falls, and he says, no, you probably won't be, and he realizes that he's doing something wrong, but he perhaps doesn't have the commitment to doing right that would be required to decide that he wasn't going to do it any more. But that kind of thing [doing what they know is wrong and being too scared to stop] demoralizes people.

If we look at the two arguments that are advanced most often toward intellectual property in software, they are one, the emotional argument, and two, the economic argument. The emotional argument goes something like,

I put my sweat, and my heart, into this program. It's my soul; it comes from me, it's mine!.''

Well, it's sort of fishy that most of these people would then go sell it to a company. And what ever happened to egoless programming, anyway? So I don't really think that that argument is worth refuting any harder than that. It's a feeling that people can cultivate when it suits them, not one that is inevitable. By comparison, consider all the great artists and artisans of medieval times, who usually didn't even sign their names. They didn't consider that it was relevant who had done the work, only that it existed; and this was the way people thought for hundreds of years.

The economic argument goes something like,

I want to get rich (except that I'll call that make a living'), and if you don't allow me to get rich by programming, then I won't program. And then you'll be screwed---so there! ''

Well, I'll explain later why I think we don't have to go along with that kind of blackmail. But first I want to address the claim [I should have said, the implicit assumption] that society actually benefits from proprietary software development.

The first problem with that argument is that it's comparing the wrong two things [owned software vs no software]. Certainly, software development is useful. But to link the development of the software with subsequent ownership and prevention of use is to beg the question. The question is, Should software development be linked with some ability to stop people from using it?'', and in order to decide this, we have to judge the effects of each of those two activities independently. What I intend to show is that the effects of ownership, by itself, are very destructive, and therefore, if you want to encourage the development which is useful, we should not connect it with something destructive [like ownership]. And I'll go on to explain other ways in which this [encouragement] can easily be done.

Let's look at the overall effects of software ownership. Consider that we have a program that has been developed, and now it either can be free, or it can be proprietary. There are three different levels of material harm that come from making it proprietary. And with each of these, there's a concomitant form of spiritual harm. By spiritual'' I mean the effect on people's feelings, attitudes, and thoughts of the actions that they take. These changes in people's ways of thinking will then have a further effect on their actions in the future.

The first level is the level of simple use of a program. If a program is successfully made proprietary, that will cause fewer people to use it. This doesn't make it any less work, however. Here is a crucial difference between programs and cars, chairs, or sandwiches. That is, once there's one of them, anyone can essentially snap his fingers and have as many more copies of the thing as are wanted. This isn't true for material objects, which are conserved, and which have to be built in the same way that the first one was built. With material objects, it's not necessarily destructive to have a price on them because that means if people decide to buy fewer of them, you can make fewer of them and use fewer resources and spend less of your time. You might say, Well, these people only want to pay for ten of these, so I'll build ten and I'll spend the rest of the day going fishing.'' But that's not relevant when the question is whether they should build some more themselves, which is what users will do with free programs. So we find that there's no overall benefit which comes---no non-selfish, non-obstructive reason to want to put a price on the use---why have a disincentive to use something which really costs nothing to make? A great spiritual harm goes along with this waste, because of the way it's worked.

The spiritual harm comes from the nondisclosure agreements and licenses that people sign, because each buyer is asked to betray his neighbors. If you want to use a program and your neighbor wants to use the program, the Golden Rule says that you should want both of you to get that program. You shouldn't aim for a solution where you get it and the other people don't, if you want to be a good citizen, that is. And the nondisclosure agreement essentially says I promise to be a bad citizen---I promise to say To hell with my neighbors!' To hell with everyone! Just give me a copy!''. And you can see the spiritual effect that has on the community you live in.

A lot of people unconsciously recognize this and that's why the decide to share programs anyway. But they feel somewhat guilty because they haven't reasoned it out carefully and decided that the nondisclosure agreements are illegitimate morally. They know that they must break those agreements in order to be still friends with their friends, but they still think that the agreements are valid also, and so they feel ashamed of being friends with their friends. And that's also a kind of spiritual harm. I say that people shouldn't have to explain in a slightly embarrassed tone of voice that they're pirates''. They shouldn't think of themselves as pirates. They shouldn't use that particular propoganda term.

The second level of material harm is at the level of adapting programs. Most commercially available software isn't fully available, even commercially. It's available for you to take it or leave it, but it's not available for you to change. But changing programs is an incredibly useful thing, and if that is prevented, it's a great harm. I have a friend who told me that she worked as a programmer in a bank for about six months, writing a program that was pretty close to something that was commercially available, and she thought that if she could have gotten source code for that commercially available program, it would have taken much less than six months to adapt it to their needs. But they couldn't get the source code, so she had to do a lot of make-work, a lot of work that you would include in the GNP but was really totally useless. And I'm sure you've all had the experience as programmers of being frustrated because there was a program you wanted to change and couldn't.

We can consider what it would be like if recipes were hoarded in the same fashion as software. You'd say, How do I change this recipe to put in less salt?'', and the great chef says, How dare you insult my recipe, the child of my brain, by trying to tamper with it? You don't have the judgement to tamper with my recipe and make it work right!''. But my doctor says I'm not supposed to eat salt! What do I do?''.

The spiritual harm that comes from this is to the spirit of self-reliance, which used to be greatly valued in this country. If you can't be self-reliant, then the landscape that you live in is frozen under someone else's control and you can't change it. That leads to a demoralized attitude like Well, yeah, this system isn't really what we want, but it's hopeless to think of trying to change it, so we'll just have to live with it and suffer and suffer and suffer.'' And nothing can kill morale of programmers like that.

I remember even at MIT we had an example like that, and I remember how it demoralized people. Our first graphics printer was the XGP, which many of you are probably lucky enough not to have heard of. But that was available ten years ago or more. And all the software for that was written at MIT, and we could change it. We put in lots of nice features, such as it would send you a message when your document had actually been printed; and it would a send you a message if you had anything queued and there was a paper jam; and many other nice features. Then later we got something called the Dover, which was another gift from Xerox. But there, the software was proprietary, and it ran in a separate computer that served as the spooler, and we couldn't add any of those features. So when you got a message that said your document has been sent'', you don't know if it's been printed, you won't find out when it's been printed, and no one would get informed when there was a paper jam, so it might sit there with no one fixing it for an hour. And we all knew that we were capable of fixing these things, probably as well as the original authors were, but somebody found it useful to sabotage us. That was a constant source of frustration.

The third level of material harm is the level of software development. Software development used to go on by a sort of evolutionary process, where a person would take a program and rewrite parts of it for one new feature, and then another person would rewrite parts to put another new feature in; this could go on over twenty years. The final program might have many features that the original one didn't have, but you might still see a few lines of code that you could recognize by their style as having been there from ten years before without ever being changed. And one of the main goals of intellectual property is to prevent this kind of evolution, to try to make it impossible for you to throw together a program quickly and easily by cannibalizing things that are already written. In any kind of intellectual field, progress is built by standing on the shoulders of others. That's what's no longer allowed---you can only stand on the shoulders of the other people in your company nowadays.

And the spirit that's harmed by this material harm is the spirit of scientific cooperation, which used to be so strong that scientists would cooperate with each other even when they were in countries that were at war. Nationalism wasn't able to destroy scientific cooperation, but greed has. Nowadays, people don't publish enough in their papers to be able to replicate anything. They publish only enough for you to be able to gaze in amazement at how much they had been able to do.

[Tape counter 213-227: someone describes the example of Karmarkar's algorithm, an efficient algorithm for linear programming that AT&T is keeping secret.]

So now that I have shown all these reasons why ownership of software is bad for society, I think the conclusion that we shouldn't allow or recognize in any way a concept such as ownership of programs. We should say that programs are not property, and if we want to encourage software development, we should encourage it in another way.

There are many fields in which no one sees a way to get rich by owning something and no one expects to, but there are still lots of people eager to work in the field. They compete bitterly, though more sadly than bitterly, I guess, for the few funded positions available, none of which is funded very much.

Suddenly, however, when it becomes possible by hoarding information to get rich, you find people saying that no one will work in the field if they can't get rich. Well, it may be true at that moment, but it's a consequence of the expectation that you can get rich. Therefore, we shouldn't assume that it has to be true in any particular field. It's essentially a coincidence that that field makes it possible to get rich by hoarding, and if we were to take away the possibility of getting rich by hoarding, then after a while, when the people had intellectually readjusted, they would once again be eager to work in the field for the joy of working in it, and they would ask us only to provide them a pittance to live on. And these pittances can be provided at universities which right now are proprietary software sellers. They can be supported by hardware manufacturers which have to do software development anyway, and twenty years ago did free software development. They can be supported by government contracts and grants, which actually support a lot of proprietary software development today. A large amount of software development is supported one way or another by the military, some of which may have to do with building weapons, but not all of it; but it's proprietary anyway, most of the time.

Some people I know right now have a contract to write a proprietary, improved version of a free program that supports TCP/IP on IBM PC's. And with this contract, they could make a living just as well writing free software, but of course they're not going to; they want to be able to clean up afterwards. And you all know of people who get research grants to develop something at a university, and then they make the unfinished version free, and go on to start companies to sell something with the last few finishing touches, so that you'd actually want to use it.

Our whole world, everywhere we look, all forms of good will to your neighbor are being undermined by the search for ways to obstruct for profit. Fifty years ago, various people in the Mafia would go around to stores in the neighborhood and say You have a nice building there; it would be a shame, wouldn't it, if it burned down, and there have been a lot of fires around here lately. I think you need some protection''. And this was called the protection racket''. Now we have people going to computer users and saying, You've got some nice software there, but I paid those congressmen to say it's illegal for you to have it. A lot of people have been going to jail around here lately for that. You wouldn't want to go to jail, would you? I think you need to pay me some money.'' And this is the software protection racket.

They say that they're providing a useful service to the public, that they are software publishers, or software stores. But these functions only appear to be necessary because of the restrictions they have managed to place on us. You wouldn't need a software store if software were free. You'd only need the public library, which would have a computer with copies that you could download, or copy onto floppies or tapes, of one copy of all the programs that you'd want. And it wouldn't cost very much---nothing, compared with what software stores cost to run, to make this thing go.

But the people in the Mafia fifty years ago didn't want to admit to themselves---they wanted to pretend to themselves that there was something legitimate about what they were doing, and that's why they phrased it in terms of protection. They didn't bluntly say I'll burn your store down if you don't pay me.'' So, too, the software publishers like to pretend, and speak as if they were doing something useful, instead of simply saying if you use that program, I'll send some cops after you if you don't pay me.''

[Unintelligible question, apparently about if there are sufficient programmers to write software w/o monopolies. Ans:]

Well how do you know they're not sufficient? So much of software work today is wasted by duplication that's done only for proprietary reasons, we could probably get as much software delivered with half the programmers if we didn't have this form of featherbedding. In fact, it's sort of funny to look at the quest for software productivity. There are two meanings you could have for software productivity. You can look at the software productivity of the industry as a whole, or you could ask about the productivity of a particular company. There's a lot of work being put into increasing the software productivity of particular companies one by one, but the best way to increase the productivity of the whole industry is to eliminate the artificial obstruction. And that would increase productivity far more than anything people are trying to do now. But they don't want that, because what they want is productivity of my company, not productivity that actually benefits society. But they use that language to try make people overlook the difference.

It's an interesting thing about how language is used for propoganda. By choosing particular words, you don't force people to think a certain way, but you can strongly suggest certain avenues of thought. By using a certain word for two things, you can bias people without their knowing it towards seeing the similarities and underestimating the differences. And this is one of the reasons why software hoarders use the terms property rights'' and ownership'' and theft'' while I use the term software hoarding''. Software hoarding is my name for the crime of trying to prevent someone else from sharing a program with a third party. And that's the same thing that's going on with "software productivity". Software productivity has two meanings, about which you would say different things, but it takes a critical eye to realize that there are two meanings. If you don't look critically, you might say Ah! Improving productivity---that's got to be what we need! Obviously, what we want to do to make society work better is improve productivity!''. You wouldn't guess that they're talking about a very narrow kind of productivity, and excluding something that comes into the overall productivity, an area where they themselves are deliberately making things worse.

What can we say, I guess, about the overall effect of technology on these kinds of restrictions? I've explained how copyright pretty much made sense. It was only a small restriction on people's freedom in the technological situation in which it was invented. But now we've seen a change in technology which makes copying things one at a time a feasible way to copy, and anyone can do it as well as anyone else, which changes the fundamental situation that made copyright legitimate and caused it to become illegitimate. Right now we've seen half of that transformation in the copying of books. You can xerox a book now, but only for very expensive books would you want to, because what you get is not as nice as a bound book, and it takes you more work to do it.

It fundamentally does cost more (now) to make a book by xeroxing than it does on a printing press. And it costs the individual more, so we've only seen a few situations in which people wanted to make any number of copies of a book by xeroxing. But that's just a transient state, because soon we're going to have wonderful computer printers, and we're going to have digitizers---scanners, and it will be possible to scan any book quickly, you'll have a robot to turn the pages, and have it on your disk, and send it across a network to be printed out into a beautifully bound copy, and at that time, book publishers will be just as useless to society as software publishers are. And at that time, the move that's already afoot to eliminate the public libraries or tax them with royalties, may essentially complete itself. You may be unable to get a copy of a book without signing a license of some sort.

In general, the entire thrust of information technology is that information is completely interconvertible. Any form can be changed into any form. Nothing is ever lost. And also, the whole idea of the computer is that it's a universal machine. It's a machine that can be programmed to simulate any other machine. Why is this better than building a specific machine? It's because you can change the program, because you can copy the program, load in new programs. The ability to copy and the ability to change a program are the entire basis for the usefulness of the computer. And those are precisely what the software owners don't want the public to be able to do. Software owners want to reserve all the benefits of computer technology to themselves, with only a trickle down'' to the public. I say the public should not allow this. If we call the bluff of the programmers or the programming investors, they won't have as many Porsches and they won't eat as much sushi. But once they are willing to go along with that, they'll find that there are other ways of organizing businesses that are less profitable, but enough for a good number of programmers to make a living nonetheless. And we'll take a great step forward toward a world in which it's much easier to make a living, because there won't be as obstruction going on, making us do unnecessary work. And we'll see all sorts of indirect effects, because society is pretty well linked.

Software hoarding is one form that the tendency to obstruct society for profit takes; one that pertains to our field, but that actually exists in some other fields as well, to greater or lesser extents. And the more people see this going on, the more are going to be inclined to do so themselves when they get a chance, and the more they're going to be demoralized and willing to put up with it, and not fight against it when they encounter it in others. But conversely, any time we take a stand against some of ownership because it makes the whole world poorer, we'll help other people stand up in the future, and we'll help spread the idea that you should ask not what your country can do for you, but what you can do for your country.

[QUESTIONS]

Q: Your talk is very impressive, it's almost like a sermon...
RMS: Yes, it is.

Q: and it's difficult to put all the pieces together and argue with it, but I find it especially difficult to see how your alternative system would work, or even exactly how you imagine it to function. Let me you this: you didn't mention music. You mentioned recipes and other things.
RMS: yes.

Q: But music is an interesting analog also, with the score; because the composers have a union, and everything, and they not only---
RMS: It supports them very poorly. Copyright as a way to support composers or musicians is a flop. We should throw it away and get something new. We're the only advanced country that still relies on that.

Q: Well, the question I wanted to ask you---I realize it doesn't work very well, but they hope that they can get some profit from the publication of the printed score, but also if a producer of say, a Broadway show, or a movie maker, makes a big profit by performing the composer's work, then the composer would like to have a piece of that action too. And you can make an analogy between that and a user of software.
RMS: You can, and I think in both cases you get problems further down the line. See, the nature of knowledge, and information in general, is to make use of it, you transform it, and mix it in with other things. And to try to keep track of how much of a big program was developed by this author, and how much by this person and so on, it fundamentally can't be done unless you forbid them to work together, which is what the current system effectively does. It's perfectly possible to keep straight which company developed something when nothing can ever go from one company to another. And that's a tremendous harm. The whole idea of encouraging the development of information seems plausible because the information is useful. But in order for it to be fully useful, you have to abandon the idea of trying to keep track of who created it, because any component could be created in many stages. Now it's true it doesn't do very much harm to anyone if this Broadway producer pays some money to the composer, assuming that there was only one composer (of course, there's such a thing as the folk process, and there's the process of arrangement as well), but it just happened, I guess, that in music there wouldn't normally be so many stages of transformation of one work that it was hard to keep track of. And part of the reason is that fundamentally, the value of music is aesthetic, and therefore originality wasn't imposed simply for the sake of intellectual property, it was valued for its own sake because something that was not new was not aesthetically up to contribution. A piece that was made by cannibalizing parts of a few other pieces with a little bit that was new wouldn't turn music listeners on very much; they'd just say oh, yeah! Well, I remember that, I've heard that before in so-and-so's work.'' And in an area where originality has an intrinsic usefulness, you won't get things being tacked on by one person, and another, and gradually changed around so much, although in some areas of music you do---in folk music, that's what happens [originality is not the main concern in folk music]. And when there are fewer people changing a work, it becomes less of an odious burden to try to keep track of who did it, and if you're talking about in particular, the Broadway producer, well, he's running a big business anyway, and he can arrange to make a license and so on, so in that situation it doesn't do very much harm. There's a danger when somebody says that something is wrong with a significant part of society---there's a certain problem that you run into, which is that on one side, there are people who call you a fanatic, and on the other side there are people who call you inconsistent. If you look for every ramification of the problem you see and demand that they all be completely excised from society, people call you a fanatic. And if you don't look for every ramification and demand that it be excised from society, people call you inconsistent. What I believe is that no one coordinate patch will map the whole world very effectively. There are many phenomena; each phenomenon is the controlling phenomenon in some areas and insignificant in others. I'm concerned with eliminating intellectual property mainly in the field of software, because there I see it really uglifying society and causing tremendous waste. If it doesn't cause so much waste in the area of music, then I don't feel it's so important to change it there. Although, there are certain areas where I object to it, and you can see the battle is shaping up again because of the ability of individuals to copy. You may have heard of the proposed laws to tax recorders and tax tape, and pay the money to record companies, or to musicians, in proportion to their sales. Well, I'd be happy to see a tax on tape or tape recorders to support musicians, but not in proportion to their sales! If your goal is to help the struggling musicians (and most musicians are really struggling, because our current copyright system doesn't support them at all), I'd say that that tax should go to the musicians who make a few sales, but not very many. Because the ones who make a lot are rich, and they don't need a handout.

Q: I'm not sure of the legal changes you're proposing. There's at least three ways in which software can be proprietary. One is copyright laws, which you're obviously against.
RMS: yes.

Q: Another is by entering into a contract: I promise not to copy, I promise not to disassemble, I promise to be a bad citizen. The third is you can use encryption methods, which perhaps our copyright laws, hardware manufacturers support in the processor ID [???]. My question is this: are you proposing to make entering into certain kinds of contracts illegal, and are you proposing to make encrypting in order to hoard software illegal?
RMS: Well, I haven't seen hardware copy protection schemes that manage to win. Encryption, I agree, can prevent you from changing a program, but I don't see how it can prevent you from copying it, because you could copy the encrypted thing.

Q: but you're still avoiding the question... [not true; that was one half of the question]
A: I would say that a nondisclosure agreement like that should be called unenforceable, like gambling debts. Clearly to try to stop people from keeping secrets would be tyrannical and I wouldn't be in favor of trying to do that. The other thing to realize is that trade secrecy is not fully built on those contracts. Those contracts are one step you go through to establish something as a trade secret. But there's an entire body of trade secret law that imposes things on people who have not signed those contracts. So in that area, just as in the other areas of copyright and patents, there's explicit government intervention which has nothing to do with voluntary contracts, to give power to software hoarders, and if that were taken so that we had nothing but the voluntary contracts, we'd have a lot less power in trade secrecy than we do now.

Q: One other question. Sometimes you seem very concerned about the [??] of software. But when it comes to, for instance, supporting musicians, you seem to be saying [????]
RMS: The public is already supporting those musicians, and they will make a lot of money anyway, from their concerts and so on. And what I'm saying is that obviously, we should only have a tax like that if people think it's desirable. Maybe people think it's desirable to support starting musicians, and musicians whose lives are marginal but who some people like. And I think you could set that up in such a way that only a person who was yearning to be a musician would consider it attractive, and therefore you wouldn't have to worry about someone sponging off of it in order to get out of doing any work.

Q: While I sympathize with the position you've laid out, it reminds of the debate between capitalism and communism.
RMS: Oh, no, it's very different. I'll explain to you the difference.

Q: OK, well, the specific point I'd like you to explain then is that while communism is appealing, it removes some basic human incentives that are based on self-interest.
RMS: There are lots of kinds of self-interest, and what they are depends on the social situation. There are also lots of kinds of competition. For example, there was competition among hackers in the old days, for who could do a better program, or could make a program shorter and faster. That competition, though, didn't involve obstruction; yet it motivated people very powerfully, and you can see in many areas of life where people are motivated by a wish to excel their rivals, to do better than anyone else had done. And they might even spend gigantic amounts trying to do this. Have you read about the people who were competing to visit all the countries in the world first? They spent incredible amounts of money trying to do that. But as for the fundamental question, communism and capitalism are similar in that they're very concerned with how much money should each person get?''. Capitalism the position of a person who says I want to have the power to get all the money that my might can bring me'', whereas communism says everyone should have the same amount of money, or everybody should have an amount of money that's totally determined by what that person needs in some sense. But what I'm saying is, we all have to pay less attention to how much money each of us has, because we're impoverishing each other by paying too much attention to it. [TAPE ENDS.]

>> *END* <<