Law in the Internet Society

Patent Law at a Crossroads: Bilski and Beyond

A Speech given by Eben Moglen at Cardoza Law School on Nov 2, 2009

Thank you it’s a pleasure to be here. I want to thank Michael and Ryan and their colleagues for making this possible. I’m going to talk about patent law, not intellectual property in general or copyright in particular. I’m not going to talk primarily about why free software is or how it works in the world, but I am going to try to talk about freedom in the 21st century because patent law is a major problem if you want to have freedom in the 21st century. I’m going to try to talk about where we are now which is at a very interesting moment in the history of the patent system.

I need to begin by pointing out that the founders of the American constitutional republic that we live in saw the possibility of monopolies granted by government with extreme suspicion. The statutory monopoly is no 18th century British North American lawyer’s friend. The long common law history beginning in the late 16th century had already taught them the enormous danger of political corruption which arises from grants of statutory monopolies. In fact, the route of the statutory monopoly in English law, from there or any other English British N American lawyer’s point of view, the origin of the statutory monopoly is the crown’s desire to raise unconstitutional revenue. That is, revenue not appropriated by Parliament.

Elizabeth I learned, and her successors on the English throne imitated her success in learning, that a grant of a statutory monopoly is a way to get money out of those who buy the monopoly from you at the net present value of the royalties they can extract. That is, the rents they can extract from civil society, a form of tax farming. Because it is a form of tax farming and a form of tax farming which raises a revenue aside from Parliament it was considered from the very beginning in the late reign of Elizabeth I to be dangerous. That’s why the case of monopolies, so called, as reported by Edward Cook said that there could be no grant of a monopoly, in that particular instance a statutory monopoly on playing cards or an extra statutory monopoly, a prerogative monopoly on playing cards. There could be no monopoly on ordinary product by royal decree, by writ patent.

In 1620 after an outburst of patenting to raise extra Parliamentary revenue by James I, Parliament passed the Statute of Monopolies which enshrined in the statutory law of England the common law’s newly generated hostility to monopolies and stated that only Parliament might grant monopolies in statutory form and that only for new inventions. This principle that the only valid or legitimate use of the otherwise dangerous mechanism of the statutory monopoly - this principle was further endorsed by the first code of law written by British North American settlers in Massachusetts in 1648 - the so called laws and liberties of the Massachusetts Bay Colony, an alphabetical arrangement of all their law under subject headings which stated under monopolies “no monopolies are to be granted among us, save for new discoveries and that for a limited time.”

This of course you will recognize as the basic grammatical structure of the entitlement to the Imperial Congress made under the constitution of 1787 to grant statutory monopolies in the US. But the real purpose as in Massachusetts in 1648, the reason there was such interest in authorizing specifically the grant of statutory monopolies by Congress at the end of the 18th century the reasoning behind that impulse has been long lost. Patent law in the late 18th century setting of the foundation of the US is an incentive to skilled immigration.

I must ask you to remember that nowhere in the western world at the end of the 18th century and nowhere in the world at all was there any regular systematic structure for publicizing and writing down in reference form the technologies of ordinary industrial production in daily life.

The French encyclopédie of the late 1780s, almost coeval with the US Constitution, is the first systematic attempt to record how trades and industries perform their work. 17th century political economy had regarded such matters as national security secrets. How to make silk cloth or drinking glasses in marketable quantities cheaply or any other significant skilled industrial process was regarded as the property of the country in which is was being done.

The US benefited before it was the US from the broad emigration of oppressed Protestants from different parts of Catholic or near Catholic Europe and the first Congress and the Constitutional Convention that preceded it understood that one of the ways of developing the economy of a society with plentiful free natural resources but a very small skilled workforce was to encourage the immigration of skilled people on a deal which required the disclosure of their inventions in a publicly consult-able reference form.

The makers of the patent clause, then, are fishing to bring men to the US who know how. They are also, no doubt about it, attempting to reward those having come here continue to invent. But the goal is quite evident from the conversation, discussion, and correspondence of the period to make a society which benefits by disclosure and that disclosure to supplement the very weak technical literature of the world in which the founders lived.

We must remember that their assumption is that an invention is a thing with a working model. That a physical object constituting a reverse engineer-able example of the invention can be physically deposited in a place where anyone else can go and learn from the thing itself and the mechanical drawings accompanying the thing how it is made. This is the bargain they are thinking about - a limited term commercial monopoly for a new entrant to the country.

In effect, being subsidized for the burden of removing to the US in return for a full and complete teaching of the nature of the invention that produces the thing. And if we ask ourselves about the things that were the antebellum patent law of the US we would find ourselves talking about a cotton gin, an electric telegraph, a sewing machine, objects deposit-able in the patent office in working form from which anyone during the period of exclusivity could learn everything they had to teach and make any necessary modifications to be used elsewhere.

The patent law of the 19th century - that patent law of products - becomes a law which takes onto itself the ordinary procedural characteristics of 19th century administrative law. 19th century American Administrative law is a lost subject. We changed the nature of our administrative system fundamentally in the 20th century, in the mid 20th century when we created the administrative Procedure Act. 19th century patent law looks rather odd in the 21st century, because we still use 19th century patent law as far as process goes in the 21st century, but it looks entirely reasonable under 19th century administrative conditions.

19th century administrative law is the administrative law of the same country we were just talking about - one with a paucity of skilled people. 19th century administrative law doesn’t assume a permanent marriage between government and expertise as 20th century administrative law does. 19th century administrative law assumes that government resources for making deliberate calculations or evaluations of anything are few and far between. Rules to be administered by systems of government in the 19th century must be comparatively simple. They have bright lines and don’t require a great deal of expensive investigation in order to be made good. Patent law, therefore, is as sophisticated an administrative process as 19th century American society can manage. Heavily skilled clerks with technical knowledges are required to make a fine determinations by 19th century administrative standards. Is this invention novel? Is it non-obvious?

I would point out that the patent system of the 19th century therefore becomes expert at answering questions based on its own records. The assumption of the 19th century patent system is that everything that should be patented already is patented already; therefore, when you bring a physical invention to the office and deposit its working model the correct way to determine whether it is novel is to ask whether it is novel in light of the art already in the office and when asking whether it is obvious you are essentially asking whether the very semi-skilled form of art knowledge possessed by the patent examiner would have regarded this as a natural extension of the existing art deposited in the office. In other words, searching is about searching what the office knows. Like other 19th century administrative structures, the patent office assumes that all it can be expected to do is to look in its own files and to make rough binary determination - 0 or 1, novel or obvious or not - on the basis of the consultation of its own administrative record.

We still live there though the entire procedural structure of the administrative system we use for everything else has changed.

The US economy benefited, nobody should doubt that it benefited from the extraordinary technological inventiveness of Americans understood throughout the world by the middle of the 19th century to be the distinguishing characteristic of American civilization in the world - its technological inventiveness. And also understood throughout the world to be an adaptation to the American political economy condition. Lots of land, few skilled workers. Americans invented from the 19th century point of view, ways to use fewer skilled workers to achieve larger outputs. The most important such invention is a business model. It is manufacturing interchangeable parts. It is therefore not patentable or patented.

But the idea, the American system, the breaking down of physical object into parts that could be interchanged among units because they are manufactured in standard gages, sizes, and performances - that’s the central invention of the American seen from the world economy in the middle of the 19th century and that plus railroads, cheap iron and coal and the presence of an enormous land for agricultural and industrial production is the nature of the 19th century industrial superiority of the US.

But the very success of that economy, the American system pursuing industrial production in a new and more productive - that is, more output per hour of skilled work mode. The very success of that instrumentation altered the nature of industrial economy. By the second decade of the 20th century the economy that we actually had depended for its wealth and its power on industrial processes as much as products.

Geochemical discovery and development, the process of manufacturing chemical and electric goods out of the combination of material and skilled knowledge. Steinmetz’s great electric motors for the general electric company. You remember the one that was running Henry Ford’s assembly line in Deerborne and it broke and Steinmetz was brought from Schenectady to Deerborne to fix it and in the end, he added one coil of wire and the general electric company sent a 10,000 dollar bill to Ford and Ford asked for a specification of the items in the bill and Steinmetz wrote back “Wire. $1. Knowing where to use it. $9,999.”

The industrial processes of the 20th century did to time what the industrial processes of the 19th century did to matter. They broke everything down into slices of time. Frederick Winslow Taylor and Henry Ford produced the understanding of what it means to make something in American industrial life. It meant to organize activity in time. That was a process. A series of finely constructed steps precisely indicating how a thing could be manufactured at least cost in materials and labor in a finite way reproducible indefinitely. Whether it’s a motor car or a tire for a motor car or paint or a watch the proposition is the same one - the time necessary to use the machinery to make the ideal object is sliced into steps and those steps can be taught to a worker who can then be joined to the machinery under the supervision of someone with a brain that has been trained to think not about the job slice but about the process as a whole. This is not patentable.

These would be processes and the statute speaks in terms of products until 1953. Then, Congress adds the words under which we now struggle, “or process.” Let us be clear that the Congress that used those words had, in the usual synthetic sense, an intention.

We are reading a statute in light of what it is that the legislature that made it intended. And in the usual way we must reconstruct their intention from the dictionary and/or the committee reports and/or some common sense about the world, but no matter how you do it the process that they have in mind is the Fordist industrial process. A series of steps using specialized machines to produce outputs which are still in their essence products made by employing matter in new forms through processes that can be described precisely in relation to the work being done.

Shortly thereafter, that is shortly after 1953, the US patent office responds both to changes in statutory language and to the changing nature of the inventions being made by ending as an administrative matter the requirement for the deposit of a working model leaving behind, as we were taught in our childhoods if you have as much gray hair as I, that the only things that the patent office required working models of were the machines that violated the second law of thermodynamics. And that was how it was described in patent office propaganda in the early 1960s. From now on the only thing you need to bring us a working model of is anything you claim that reverses entropy. Perpetual motion machines you have to demonstrate. All other things the claims and drawings are sufficient.

Of course, this change, a change away from the idea that there is something you have to give the patent office to prove that somebody else can learn, that reduction to practice actually means giving a thing. That’s about the only change being made in the political economy of the patent office. What is not being changed is the basic administrative law structure which we now need to think about a little bit because by the time the 1953 Act said you could patent processes, we had changed the administrative law of the US completely.

The administrative Procedure Act of 1946 revolutionizes the way we think about government in any number of fashions but two that are crucial here are that it makes a permanent marriage between expertise and government. Government is not supposed any longer to be intervening in the economy or taking other administrative actions that affect private rights without consulting more than its own file cabinet. The theory of the APA is that through rule-making on a record construction of expertise, judicial review of factual bases for regulation, the theory of the APA is that government is responsible for having all the information that expertise can provide, not merely what happens to be in the files from last time around. Everything becomes subject to the rule of cost/benefit analysis, which is regarded as the fundamental technology of government. Understanding whether proposed government actions are more helpful than they are harmful and making a systematic and careful evaluation of the relationship between benefits and cost before deciding to intervene in the economy. This over the subsequent generation becomes the rule that everybody must follow when making marginal interventions in the market. Do you want to make an occupational safety and health rule that will change the way deli workers cut meat or machine shop operators use a drill press? Do you want to set standards for workplace exposures to chemicals of industrial significance but potentially also significant dangers to workers’ health? Do you want to make rules about nutritional labeling or advertising?

In all of those cases you must engage in cost/benefit analysis to demonstrate that the intervention is more helpful in the achievement in public purposes than it is harmful and that cost/benefit analysis is subject to judicial review. That’s the system of government we made in the 20th century and just about the only thing we didn’t change to conform with it was patent law. Patent law continues not just to elide the idea of cost/benefit analysis. Patent law flies in the face of cost/benefit analysis.

The patent law assumption is that the value of every invention is infinite. Show only that it is novel and un-obvious and you are entitled to a 20 year statutory monopoly without any consideration of the possibility that there are offsetting harms.

No need to consult the public interest of any kind. If the private party shows that he meets the definitions contained in the statute then there is no public interest to consult. Everything is assumed to be benefited purely by his taking rents on the basis of his invention.

This is increasingly undefensible. This is increasingly difficult to understand except that nobody is understanding or seeking that it be defended. It’s simply, like so many other things about the law, the consequence of the rule that you don’t ever have to do anything just because you’re obsolete.

Somebody has to come and do something about your obsoleteness. And nobody wants to do anything about the obsoleteness of the procedure - the administrative procedure of patent law because the change in the words process from product to product or process allowed so much new rent seeking. Remember, as I said Congress must have something in mind when it talks about processes and it is talking about processes that move natural materials around and make new things out of them. But within a generation in the change of a statute, the processes being generated are no longer processes that do anything in the real world. The processes involved are now things computers can be made to do and things that businesses can do using computers where they used to use people.

These are after all, the basic processes now at stake as we consider the scope of 21st century patent law. Computer programs and business methods based on the use of computer programs to do what skilled human beings used to do.

In the patent law before 1953, there would have been no issue of such things. They were unpatentable. That was everybody’s plain understanding. When the Supreme Court began to see such claims it made very clear in its own description of the meaning of patent law, that these were unpatentable inventions save in relation to their continued association with industrial processes. This is the great lesson of Diamond against Deer is it not? There’s rubber being vulcanized. Had there been no rubber being vulcanized the court’s completely clear. You could not have patented a computer program to vulcanize rubber. You had to be vulcanizing rubber.

The Court understood wherever it faced the question between 1953 and 1990, the Court understood that the situation had not changed. Patent law could not be used to convey government ownership of facts of nature, mathematical principles, mental steps purely mental in character, or algoriths. All of those represented categories evidently unsuitable for patenting under all previous understandings and began to trench upon the most dangerous aspect of the problem of the statutory monopoly because the statutory monopoly when applied to a pure idea that does nothing in the physical world comes into conflict with the very concept of the freedom of speech and thought. As the Supreme Court has directly shown in recent cases about copyright, there is an inherent potential tension between Article I Section 8 and the First Amendment and as the Supreme Court has said in Harper and Row, in Feist against Rural Telephone, in Eldred, in a line of copyright cases. The more or less simultaneous adoption of Article I Section 8 in the convention’s constitution and the First Amendment by the First Congress suggests that in the minds of those making the instruments, to the extent that you care about initial intention, there is no contradiction. Which means that one ought to interpret the statutes so as to avoid any possibility of the unintended contradiction between the law of statutory monopolies and the law of free expression.

Once you begin to present the possibility of a long-term statutory monopoly on something you can do altogether in your head, or using a language of technological communication between people, you’re trenching on the very distinction that makes patent law secure from First Amendment scrutiny in the first place. So this is the difficulty presented to the Supreme Court currently though it should not have any difficulty because the Supreme Court never created the difficulty because the Supreme Court never said you could patent purely mental operations or ideas.

What happened was that the US Congress in 1982 passed the single most misnamed statute in the history of the US - the Federal Courts Improvement Act, which massively dis-improved the federal courts. You remember that the difficulty that Congress was attending to is the terrible problem of butterflies in the tummy that overcomes district judges when they have to decide patent cases. There are few things that responsible district judges hate more than patent cases, they always did and for a simple reason. A lawyer has spent her entire career learning law and becoming a generalist district judge and she sits in a position of great responsibility and some honor from day to day making difficult decisions of every kind. And now in a trial in which she is going to be required to find facts and make decisions with significant consequences she’s going to be asked to learn about how to manufacture paint or how to make chips in large quantity by doping them with something a little mysterious every once in awhile in the middle of the gallium arsenide or well you understand of course. The problem is pregnant with the difficulty of humiliation. She or he is going to say something that knowledgeable people are going to regard as whacked out. Most likely because she’s been fooled by some fast talking patent lawyer who knew all about how to fool judges because that’s basically his job. Judges don’t like to be fooled by lawyers and most of the time they figure out how to avoid being fooled by lawyers but what can you do when some tassel-loafered guy is working all of his magic on you in regard to stuff that could as well have been written in Klingon for all you can do about it.

Congress, therefore solves the great problem of the tummy ruffling of the district judges by leaving the district judges alone to do the same job in exactly the same way they’ve always done it but to give them a court of appeal full of patent lawyers to do it for. This is insane. It doesn’t improve the quality of patent fact finding in the least. It doesn’t change the difficulty that the federal courts really have in conducting anything like judicial review of patent granting. It doesn’t do anything to assist the system titrating public interest questions in the way patents are actually employed, enforced, used for intimidation, or threat, or as leverage in other litigation. It does nothing beneficial at all but it provides a specialist court of appeals which will be guaranteed to like patents more than anybody else.

The results were deplorable in every other way. Had Judge Giles Rich not been apparently immortal and not gone on from year to year as Chief Judge of the CAFC inventing more and more ways to use patent law to get rents on everything. Had he not found the business method patent all by himself apparently on Mount Sinai some weekend.

Had he not discovered the broad gage patent availability of software, somebody else would have done it instead. But the CAFC was remarkably fertile in ways to take advantage of the fundamental understanding of the industrial world of patents in the post-industrial economy because we live now not in the industrial world where products are made by the jointure of skilled knowledge and high technology to material stratum. We live in the post-industrial information economy where services are enabled by the combination of skilled minds and high technology. Pharmaceutical development and discovery. Financial services. Information technology itself. Those are now the areas in which the patent system is employed. The industrial process users for whom in theory the statute was updated in 1953 have fundamentally abandoned the patent system. They are trade secret protectors of their products now because a 20 year statutory monopoly is useless when the product cycle is a year or two.

And where everybody can see what you are doing and invent around unless you keep it so secret that they can’t see what you’re doing.

So patent migrates to a world in which process now means mental steps. And in the areas where patenting has not traditionally been pursued before, like all the areas of post-industrial life where the CAFC afforded patent scope but there was no history of patenting, the performance of the patent office at reviewing the art was just purer.

Everything in our world, the world we practice in the world of computer software manufacture pure and simple, everything in that world has been patented 3 times. That’s our rule of thumb. Once in the mainframe era. Once in the mini computer era. And once in the micro computer era. But of course in the mainframe area you couldn’t actually patent it. So you defensively disclosed it. In the mini computer era you weren’t sure but you might be able to get system claims out of it so you begin to get a trickle of patents and then under the CAFC all hell breaks loose and the patent office begins to reward rent seekers everything they could possibly want for things that had been invented and disclosed already. But, they’re not in the files of the patent office because they weren’t patentable the first time they were discovered.

So searching the patent office itself is a poor way to find out whether this or that form of computer programming technology has been seen before. It is a poor way to understand whether inventions have been made before.

And of course, it makes very little sense to say about a business method well now it is patentable because it uses a computer when before it wasn’t patentable because it used a man’s mind.

So we have come to a position in which the Supreme Court must deal with a crowd of internal contradictions that the loose and lively patent system of the boom times made possible. Like so many other matters, this is a bubble that has burst. In the last half decade, the Supreme Court has, in a not so gently way, intimated to the CAFC and the patent office that they thought the whole thing had gone a little too far. Whether it was a solid march back towards some meaningful obviousness criterion in KSR or it was the intimation in ATT against Microsoft that software might not be altogether obviously inside patent scope and that the question might be open. The court has offered significant reasons for the parties which had scrambled off on their own in reconnaissance and rent seeking a good reason to come back and get on side.

In their last 2 to 2.5 years the CAFC and the PTO have behaved as though they got the message. Or at any rate, as though they got a message. That it might be a good idea to come a little closer to the mother-ship again.

Which leaves us with our friends Mr. Bilski and Mr. Warsaw. There’s nothing special about them, you know. They’re like a lot of other investors who didn’t have a chair when the music stopped on the bubble music of our time. Oops. A little too late for the party. They had been caught by the descending gates of something like a desire to come to Jesus in the patent office. Suddenly, it has been discovered that merely having a way to use a computer to hedge risk and commodities trading is not actually a patentable invention given that guys with electromechanical adding machines and before that, guys with slide rules and before that, guys with abacuses and checker boards, have been doing pretty much the same thing since the beginning of commodities trade. The “and a computer” just actually doesn’t cut it anymore. It didn’t really ever cut it right? It was an artifact of the combination of a statute meant to do something else with an administrative procedure left from before the black lagoon was drained.

It was only an artifact of bad procedure plus wonky statutory language meant for other purposes that anybody fooled themselves into thinking this could be an opportunity for rent seeking in the first place, and that only by dint of disregarding the obvious role of the First Amendment in making sure that ideas cannot be owned. So Mr. Warsaw and Mr. Bilski now find themselves in a lawsuit in the US Supreme Court in which, as a patent lawyer with whom I agree about absolutely noting else in this lawsuit reasonably said, what everybody knows for sure is this is a patent that shouldn’t be granted. It’s not their victory one way or another which is at stake. It’s whether the Supreme Court believes that the delicate and appropriate retreat of the PTO and the CAFC goes far enough.

What the CAFC has said is “Oh, let us find ourselves in our blacksmith shop a test which restores some similarity between our behavior and that of the statute as read by the Supreme Court.” What the patent office has said is more mysterious. The patent office may have indeed retreated all the way back to Diamond against Deer. It should do. But you know how it is. The logic of capitalism is remorseless. Our friends, who a year or two ago were sure that software patents were more trouble than their worth, have begun to realize that they have tens of thousands of patents on software and that each of those patents, worthless though they may appear to be to the skilled eye, is real estate carried on the balance sheet. It belongs to the shareholders. It is very difficult to go to the Supreme Court and say “Judge, you were always right. We knew you were always right, your Honor. We actually were pretty much on your side, but we have 95,000 building lots and we would not want to have to explain to our shareholders why we decided to agree with you that they should all be vaporized.”

So what has happened as we have proceeded closer to this crossroads is that enthusiasm for passing through them to the other side has begun to flag a little bit among those who have been our allies the past half decade. They were being murdered by the stupidity of the patent system - the large IT companies were.

It was in every way harming them. The trolls were taking random bites. The overhead of the constant dodging, ducking, and weaving was becoming unendurable. And the free software world was valuable to industry in part because it created, through its remorseless determination to have a system of goodwill and free knowledge exchange, it created a demilitarized zone to some extent, where almost all the IT firms in the world with one profound exception came to discover that they were better off agreeing not to engage in patent aggression within the team. The community became more important than the atomized idea. The patenting inventor after all, is no longer some skilled artisan in Moravia who might be persuaded to remove to Pennsylvania. The inventor is an employee of a company that works for a company with 7 other companies to create protocols to improve the NET which also happens to be protocols that are very heavily loved and admired and worked on and framed by our guys who are sitting in their bedrooms on their leave from high school or who are the very paid employees of those companies of themselves. We are making ideas in commons and the ownership of exclusionary rights hurts everybody equally.

So we stand now at a place in Bilski and beyond where patent law gets to decide how much it needs to change for the 21st century and it speaks through the mouths of the Supreme Court Justices. Are we going back to Diamond against Deer? At a minimum we surely should. That’s the law declared by the relevant 5 votes on the relevant court.

Might we beyond that glimpse the possibility that patent law needs a more thorough overhaul than merely a withdraw to the limits set by the First Amendment? No patenting of abstract ideas? No patenting of purely mental processes without specialized machinery? Or material being moved? Transformations of matter as the CAFC puts it?

To go beyond that, to withdraw from 1953 and say that you could only patent products rather than processes, would surely not be right. But to constrain processes within an understanding of the patent law that’s not inconsistent with the requirements of the First Amendment would make some sense and more than everything else some understanding that 20th century, let alone 21st century administrative law, offers a structured method for the consultation of the public interest, requires some structured method for consultation of the public interest. Surely, we ought to be thinking that far about patent law. The administrative procedure of 1850 has no justification in the world of 2015. There isn’t any reason for using 19th century approaches to administration of legal processes. This is not about the question of the substantive law. This is about the question of the improvement of the technology of government. We don’t use government power, we don’t use coercion to produce massive interventions in the private market without asking what the public interest is.

This is the libertarian case against patent law, copyright, and other statutory monopoly. Government is acting coercively in the economy. It’s not even asking what’s good for people. Perhaps from the libertarian point of view it shouldn’t ask what’s good for people. It simply shouldn’t interfere in the economy. It always puzzled me that as acute and longstanding devotee of Ayn Rand as our immortal central banker Alan Greenspan could say with apparent unconcern that government should constrict itself to its proper roles, like protecting intellectual property. I always thought that he must have been absent on the one day when she talked about that during their 30 year association.

How could that not be government intervention in the economy? The handing out of 20 year statutory monopolies right left and upside down? So both the libertarian case and the communitarian case about the problem of patents rest primarily on the process. How come we don’t use a contemporary process for deciding whether to coerce the economy in big long swatches.

If the OCHA wanted to set a rule for a 20 year period then the court of appeals for the DC circuit would say what’s the cost/benefit analysis on your 20 year period? Why 20? Would 5 do 95% as much good for 50% less harm? What are those years there for?

Even the question of term in other words should be subject to some inquiry of a 20th century rational kind in the middle of the 21st rather than being relegated to some 19th century idea that all sizes must be the same because the agency wouldn’t know how to tell the difference.

We have a great deal of work we might do on the patent law in the aftermath of a retreat to the Supreme Court’s own intended lines.

And there really isn’t any case for underwriting the law of the bubble anymore. Everybody knows where we’ve just been. Everybody understands that the very financial industries which gamed the hell out of every other system gamed the hell out of this one. Everybody understands that parties have been maneuvering to use the power of the state to enrich themselves at the public expense. That’s where we’ve been. The religion of the time was greed and everybody sees it. So there isn’t any reason to underwrite that anymore. Nobody is benefited by a continuance of the fast and flashy fluzy-ism that was patent law between 1990 and the day before yesterday. Something will have to give.

Of course, there’s a lot of weight on the other side too. The modern pharmaceutical industry is entirely dependent upon this structure of patent law.

The deepest and best-funded monopoly in the history of the world, Microsoft, is still fundamentally of the belief that it can use arbitrary software patents based on “I have a rule for doing this and if you’re going to do this to interoperate with anybody else you’re going to have to pay me” kinds of threats against every form of the free exchange of technological information regarding software for computers commercially useful.

There’s no doubt about the level of the impacted stakes of those who have what they consider to be an equity interest in the rent stream. This is going to be a big political problem either way and maybe the Supreme Court isn’t the last stop which is why I would urge you to think particularly carefully about the role of the First Amendment here.

Whatever happens in Bilski somebody will go and ask the Congress to change it. Whatever happens in Bilski the assumption will be that if you have 60 votes in the US Senate, you can make any kind of rent seeking statute you want. But that’s not quite right. The First Amendment limits on what the patent law may do are un-described but visible for the same reasons that they are now described and visible with respect to copyright. Bilski may or may not be the case that speaks to that. The US Supreme Court tends not to get to such questions until it hasn’t any choice and perhaps it doesn’t think it hasn’t any choice right here.

But come there we will, because unless we lift this idea that the 21st century will permit the ownership of ideas, we will be living in a 21st century where some very fundamental ideas are owned. Ideas without which you cannot live and which you would have assumed to be merely the common property of human kind will become the property of a few and the rules of exclusion will begin to work. The world food supply, the death and life of every human being, once the genetic revolution comes to fruition, will either be patentable or not patentable and upon that basis, real meaning of political power and 21st century society will be determined. So something will happen. Patent law will come to play an enormous role - either oppressive or facilitative in the life of the US and global society in the 21st century as it came to play a crucial role in the 19th century society of the expansive US. Justice will be at stake, which means that those of you who care about justice, and I hope that’s every single one of you, will find that there is a conflict going on in which your view of justice - one side or the other will prevail.

I know which side I’m on. I know which side my colleagues are on. We come to this not as neutral scholars investigating but as people who believe that the civil rights of human beings depend upon the answers and we don’t mean to compromise. So what I have said comes, I would agree, under the umbrella of advocacy. But I urge you to review the case for yourselves, to check its weak places, and to press on it.

What lies below the surface of this crossroads in patent law is urgently important and how you think about it and what conclusions you come to and how you act on those conclusions will contain much of importance about whether there is justice or not in the 21st century. I’m happy to take your questions. Thank you very much.

Is somebody calling on people. Let me do it if I might. Who would like to ask a question?

Question: You mentioned one size fits all term for patents. Do you believe that if they were to change that for software would there be a one size fits all for software even?

Moglen: No, the proper way of thinking about this, Ryan, is that each intervention in the economy, like each rule made by a regulatory agency, requires a cost/benefit analysis appropriate to the rule not to some class of rules. The proper question would be where’s the crossover point between the good done by providing a monopoly here and the harm done by providing a monopoly here. That requires a little bit of inspection of the political economy of the change. The pharmaceutical patent system hands out 20 year monopolies on the use of molecules to treat disease. This is obviously a bad way to do things.

It’s why we have 3—count them 3—patented pharmaceuticals for the treatment of Alzheimer’s disease in dogs. And it’s why we have in the US and elsewhere in the world, no real interest in the Malaria treatments or trypanosomiasis.

The problem with a 20 year patent in software is what’s 20 years in software? This is not about some notion of the way things ought to be or how to incentivize. This is a rent seekers paradise. The correct answer with respect to term is what’s the product cycle in the business you’re in. What’s the nature of the recovery you’re trying to get for what degree of disclosure. How good are the disclosures in providing for real opportunities for others to grow the technology, and in relation to the quality of the disclosures and the nature of the costs you’re trying to recover from bringing about the invention, what is the period of time economically justifiable for the issuance of a monopoly? The patent office could, under the conditions that are imposed on every other administrative agency under the APA, make those determinations subject to judicial review. The consequences would be - I am anticipating your answer - fewer patents coming out more slowly. Only the things that really need to be patented would be patented. That would be a positive outcome. And it would correct the incentives of American universities- a very important secondary subject I haven’t had a chance to talk about here because time is short. An entire hour should be devoted, or at least an hour plus a lengthy discussion, in every university patent law class in the US about the consequences of the patent system for higher education.

All of which is about winning the 20 year monopoly and then winning it again once you figure out which infinitesimal portion of your university’s patents actually make money for you. At which point, like a pharma company you rush right out and try to re-patent that invention to get more and more time.

When the public patent foundation, headed by Mr. Daniel Ravicher who many of you know, went after the third patent on the world’s most profitable pharmaceutical, Lipitor, which resulted in $84 billion dollars in rent readjustment back from pharma to patients and insurers because 7 years of patent lifetime got lopped off the drug which now earns $12 billion with a “b” dollars a year for Pfizer. The patent we were attacking was a patent which said you know that chemical Litovistatan twice patented for use in clinical administration daily to reduce blood cholesterol. You know that one? The one that you’ve patented twice already? Well we’ve crystallized it. Can we have another 20 years please? And the patent office says, at least until asked under re-exam to consider its judgment, “oh you crystallized it? Really? Wow. We never thought of doing that to any chemical.” More shelf life, better stability. That’s worth another 20 years sure. And the reason we did it wasn’t just Lipitor, though that was plenty of good reason to do it. The reason we did it also was to say “You know how many tens of thousands of ‘we crystallized it’ patents there are?” Each of them renewing a 20 year monopoly on a molecule? A fact of nature? Added to the knowledge about where to use it and how? Wire. $1. Knowing where to use it: rent that makes your eyes bubble. Right?

The proper answer to the term is what are you recouping and why should we give it to you that way instead of some other way? Why shouldn’t you have a prize? Why shouldn’t you have a contract? Why shouldn’t you have some other form of subsidization which leaves the public interest in control? Why is private rent seeking the correct answer to how you should be paid for what you’ve made. Those questions are the questions judicial review of patenting should be about. If the patent system were just another branch of administrative law, that’s what they would be about.

Question:I agree with what you’ve said thus far. My name is Chris and I run a software company in the city. What are the things that software companies can do to protect their intellectual operating barring patents? Answer: But what is the intellectual property of the firm before we decide how to protect it? We better decide what it is. What is it? Other voice:Untangible goods that are considered assets to the public interest.

Moglen: Sure, but is it the secrecy of the source code? So keep the source code secret. We think that’s a really bad idea. We think you’d be way better off with 100 million people finding and fixing the bugs all by themselves. But if you’re convinced that the right way to do it is to bear all of the costs of maintenance internally and reduce your market drastically to the people you can actually provide the code to, then you’re really in a trade secret business and software turns out to be very much like most of the world’s industries in which trade secrecy is really what’s at stake.

If there’s something beyond the trade secrecy of your code that you think of as yours, as opposed to that of the human race in general, what is it? Is it knowing that you can exchange the content of two registers without an intermediate location using 3 consecutive exclusive “OR’s”? Is it the discovery that you can do linear programming problems by manipulating an idealized polyhedron in your head or in your machine? Is it the discovery that you could automate the accounting box chart into a spreadsheet? In other words, is it really the case that what the software industry needs in order to function well is ownership of ideas at all? You can of course copyright your code and give it to people on restrictive terms. Let them read it and think about it but not use it or copy it without your permission. Most people tend to think that’s sort of naked copyright isn’t much use to them. They’re really trying to keep secret what they’ve done. We really think that just under 21st century conditions produces less good software which is why we work with people to try to increase their productivity, their output and their wealth without requiring them to keep generally useful technical information secret.

But I’ll go along with you that the question you ought to be asking is what’s good for my business. The answer is what’s bad for your business is a troll with a patent. That should be your gravest concern - somebody who doesn’t make anything. He doesn’t want to work with you. He just wants you to work for him and the reason he wants you to work for him is computer programs have a lot of features and he owns some. That’s your real big problem and what we want to do is lift the ceiling on you for that. So why don’t we first do that and then we’ll figure out what else you need.

Question: I have a question about the First Amendment argument. I just think it’s really interesting wondering how far you can push it if I can easily see a completely abstract idea is that First Amendment argument. Did you really think you could push that into the realm of software?

Moglen: Well, here’s the question. If I have a patent that can be infringed entirely by some code without anything else - any machine or anything in nature- then what I’ve said is is you can infringe the patent by describing its claims, because computer program source code is a way we communicate ideas to one another. Most computer program source code has a lot more stuff in it meant to communicate with human beings than it has meant to communicate with computers. We use source code to express ideas. You could, for example, say - maybe you want to say - that the best mode of practicing any software invention is source code that describes the practicing of the claims directly in an executable form.

But once you’ve gone to that conclusion, you’ve concluded that you could infringe the patent by talking about the invention. Now you’ve got a problem. That’s the problem posed in Metamune. You go around and you say to doctors “Here, you could correlate this observed situation in the blood count with vitamin deficiency. You know, doc, you could use that to diagnose deficiencies in dietary necessary nutrients in your patients.” Oops, you’re infringing my patent on doing that. What did I do? I talked to a doctor and told him he could do a thing? My communication which is essentially the same thing as the disclosures, so tell me why if a computer program all by itself can infringe a patent it’s ok for me to circulate a copy of the patent but it’s not ok for me to circulate a copy of the program source code. Now, we’re beginning to get to why I do not think that it’s such a stretch to wonder whether the First Amendment has some role to play here. You ought not to grant patents for things that can be infringed solely by communicating an idea.

You never could before. Before 1953 you couldn’t because it wouldn’t have been a product and you wouldn’t have gotten it in the statute. Now the statute says process, meaning vulcanizing rubber and the Supreme Court says no algorithms unless there’s actually rubber vulcanization going on and some subordinate judges and administrators decide to ignore that very important point for 15 or 16 years building up a whole array of rents that are being taken somewhere in the economy creating a whole bunch of guys who’ve got an in-built stake in the existing system and now we’re going to have to do a little heavy lifting and clean it up. But the truth is there wouldn’t be anything so hard to imagine about why we shouldn’t have gotten ourselves into it. That’s why the Supreme Court didn’t go into it in the first place. The problem we now have is they bitched up for a long time, nobody did anything about it, now you’ve got a lot of guys with an equity stake and the wrong answer. What do we do from here? That’s what the Supreme Court faces in Bilski.

Question: My name’s Adam. I’m going to try to see how to get some clarification on what you actually think the Supreme Court should do..

Moglen: Well, the Supreme Court is 9 judges and what 5 say wins. If what you’re asking me is how each one of them ought to vote, there’s never any reason why it shouldn’t be 9–0 for whatever you think the world is really best at, but you know that’s not the way it really works right?

Adam: I guess my point is more to do you think that the Supreme Court should find the 1953 amendment…

Moglen: No, it’s plainly not unconstitutional. There’s nothing wrong with saying you can patent processes in which what goes on is materials are re-manufactured into something else using machinery and human wit. That’s a good patent - vulcanizing rubber however you do it, including with a computer. Digging iron out of the ground and making steel out of it, however you do it, no matter how many computers or how much software is involved, that’s fine. What the CAFC says the Supreme Court’s rule is may not be exactly the Supreme Court’s rule and it may not be exactly an emanation of the statute but what it is is an attempt to get back to a safe neighborhood and the Court ought to say “We meant by our prior cases what we meant by them.” They ought to say “We had a view of patent law and that was right. We understood the statute and we read it carefully and we read it in light of all the things that statute readers are supposed to do and our cases are correct.” Parker against Fluke is correct. Diamond against Deer is correct. There’s nothing wrong with those case. Now, unfortunately, we’ve got a bad situation cause those cases didn’t get followed for a long period of time. The specialist court that was supposed to do it didn’t do it. Now they’re trying to clean up and they’re right. That would be enough. Of course, that would be a cataclysm but there’s nothing complicated about it and it doesn’t require any heavy lifting from the Supreme Court. It requires nothing more than a decision by the Court to vindicate the correctness of its own prior judgments and to adhere to its own precedence.

Question: So the Court thinks that the CAFC has gone beyond what it should have done at State St. why didn’t they deal with this at State St.?

Moglen: The Supreme Court is not required to do anything. That’s its beauty and its problem. If what you mean is had the justices known what was coming might they have done something else in State St.? Maybe. But it is always the Supreme Court’s view that the development of the law is primarily for other courts. And the Supreme Court either wanted to or had no choice but to or decided without deciding to allow the horse to run and it ran.

Any other questions?

Question [Paraphrased]Is the US worried that international treaties would constrain it to support the patent policies of the last 15 years, regardless of the earlier precedents?

Moglen: No. We’re not. We are first of all the US. Would that we were more constrained by international agreements than we are. That somebody will take us to the WTO for not having done what we ordered the rest of the world to do is not how the game is usually played right? The US retreats from the high water mark of its oppressive behavior in the international law system recurrently. Nearly 20 years ago I was involved in fighting the US view that encryption could not be exported in idea form. It took a lot of work to get the US government to recognize that we had as much right to export source code implementing RSA as we had right to export source code implementing the Caesar cipher or the Vigeniere cipher. In 1995 at the high water mark of export controls over encryption, the US secured an agreement from all the other western European union that they were going to adopt the American approach to barring the exportation of encryption technologies. Three years later, the US government conceded that the whole thing had been wrong from the beginning and gave up. Wassenaar is still there. But once the US walks away from an absurd position, the rest of the world is usually content to allow them to do so. Under present circumstances I do not think that the US would find itself in any harm if it restored something like the previous balance in the patent law. We’ll see.

Question: Should the Supreme Court uphold Bilski what will happen next?

Moglen: Patent reform in the US Congress will happen next. What that will consist of remains to be seen. On the other hand, if the Supreme Court decides Bilski the other way, patent reform in the US Congress will happen too. The question will simply be which side has the fire underneath their feet when they get to Capitol Hill.

Question: What changes do you see in legal education if Bilski is

Moglen: Nothing. Law professors don’t ever change. You know that. Thanks very much.


Webs Webs

r1 - 11 Nov 2009 - 22:19:45 - IanSullivan
This site is powered by the TWiki collaboration platform.
All material on this collaboration platform is the property of the contributing authors.
All material marked as authored by Eben Moglen is available under the license terms CC-BY-SA version 4.
Syndicate this site RSSATOM