pageok
pageok
pageok
[Adam Mossoff, guest-blogging, May 6, 2009 at 11:56pm] Trackbacks
Patent Thickets, Bad Patents, and Costly Patent Litigation:

Some might conclude that "patent trolls" are a more serious problem today because of larger problems in patent litigation generally. We often hear reports of an explosion in patent litigation, problems with determining whether patents are valid, problems with determinig patent boundaries, and commercial products being held hostage to the threat of infringement litigation. Such complaints are particularly loud in the computer industry, as supported by academic commentators.

Here, we must be especially cautious in assuming that modern problems are necessarily different from those experienced in yesteryear. Of course, it goes without saying that there are differences between the 1850s and today; in fact, there are differences between 1995 and today. There are always differences between two distinct points in time. So it's important to always figure out if these are differences without a distinction, and, even more important, if there are any similarities.

In this respect, recall that many facets of the Sewing Machine War reflected many of the purportedly new problems with patent litigation today, such as massive and costly litigation and the threat of injunctions. And it's not just cases from the Sewing Machine War that reflected these and other so-called modern concerns. As someone who has been immersed in historical patent jurisprudence for much of his academic career, including having read, among other things, all patent decisions in the Federal Cases reporter (approximately 1,460), I often hear today the echoes of long-forgotten patent disputes.

In 1862, for instance, a judge expressed his frustration at inventors being "frightened off the course by threats of ruinous litigation," and that in the particular case before him, the "astute counsel and experts have been employed to surround this machine or invention ... with a fog of nebulous rhetoric, and to make this concrete machine appear a transcendental abstraction ...."

In 1855, in the midst of the Sewing Machine War, a judge instructed a jury in a patent trial unrelated to the swirling disputes over the sewing machine:

For the maintenance of his right [an inventor] is subjected to legal controversies, which, not infrequently involve him in an expenditure beyond the amount of his profits. Inventors and discoverers are proverbially poor. It is said that the man, by the operations of whose genius the streets of the city of London were first lighted, was a wanderer and a beggar in the streets.

In 1877, another judge complained in a patent decision that "litigation in regard to patents has been found so expensive and so wearisome to the courts." Such problems arose from how each "contest involves an immense sum in value, and where the question turns mainly upon opinions of experts, there is no difficulty in introducing any amount of them on either side and yet this class of cases is one in which there is value to be attached to experts." He further noted that it "is a sad thing to say that perhaps no class of cases coming before the courts have as much fraud, perjury, and wicked conduct, as patent cases. ... there is a large amount of false swearing and corruption in them."

If anyone has a tendency to commit the anachronism of thinking that the inventions of yesteryear were simple compared to today, they need look no further than an 1855 decision from Justice Grier, riding circuit, in which he observed:

It is no reflection on juries or trial by jury to say that many disputes about the originality and infringement of patents depending upon complex mathematical calculations, upon a knowledge of the principles of chemical science, and of mechanical philosophy, cannot be satisfactorily decided by the verdict of twelve men, a majority, if not all of whom, have no knowledge or experience on the subjects they are called to decide on.

And, in 1841, Justice Story, one of the principal architects of American patent law, observed how "Patents and copyrights approach nearer than any other class of cases ... to what may be called the metaphysics of law, where the distinctions are, or at least may be very subtle and refined, and sometimes, almost evanescent."

These are just a few of the many long-forgotten patent cases I have uncovered in my research (some of which I discuss in a previously published article) and which warn against several anachronisms in patent law today. These nineteenth-century complaints remind us that the inherent technical and legal problems in assessing complex innovation — such as distinguishing between different complementary contributions to a commercial product, determining imprecise contours of a patented invention, and litigating a case in a court system that moves at a plodding pace — are long-standing features of the American patent system.

The American patent system, as economist B. Zorina Khan has explained, was tremendously successful in the nineteenth century — as compared to older and more established patent systems in England, Germany, and France. As Professor Khan has shown, the American patent system excelled precisely because it did what none of these other patent systems would do: It secured inventions as property rights within an institutional framework governed by the rule of law.

It was within this framework that the first American patent thicket arose from the incremental invention of the sewing machine. It was this framework that also provided for the resolution of this patent thicket by the sewing machine patentees — exercising their rights of use and disposition in their property by contracting to their mutual benefit.

In my next (and final) post on the sewing machine patent thicket, I will raise an issue that is not yet discussed in my paper — antitrust. The impact of antitrust doctrine on how patent-owners contract with other patent-owners may create significant variances between the nineteenth century and today on how patent-owners may resolve patent thickets. I am still researching the relationship between patent pools and antitrust, and so I am especially keen on receiving feedback from the readers of this series.

_quodlibet_:
I would contend that software patents *are*, in practice, fundamentally different from anything seen in the past. If executable object-code is analogized to a patented machine, then the source code would be analogous to a very detailed blueprint for the machine. However, the cost of building the executable code from the source is essentially zero, and furthermore, the cost of duplicating and redistributing the object code is also essentially zero. So, if one developed a open-source program that used a certain patented algorithm (e.g., an MP3 codec algorithm), then one could distribute the software "blueprints" (i.e., the source code) but not the object code, and users would be legally forbidden from compiling without a license from the patent holder. Given all this, the notion of a software patent basically amounts to the notion of someone owning an idea (an algorithm) and being able to forcibly prevent others from using the idea (i.e., executing a program that implements the algorithm).
5.7.2009 1:01am
BeachBumBill (mail):
Tangential to the Patent thicket idea is the reason for Patent law in the first place. About 15 years ago, there was a very informative article in the Harvard Business Review comparing the Japanese patent system and our own. This article not only delineated the differences between the two systems, but the reasons for thier existence.

For example, the primary reason for the US patent system is to protect the inventor, but in Japan, the motivation is to quickly disperse information to help the country recover from the war.

Patent trolls in Japan quickly reverse-engineer products with just enough differences to allow a challenge to a US patent in Japanese courts, thus freeing them to quickly have a competing product on the shelves and grabbing market share before the US manufacturer has a chance to respond.
5.7.2009 1:11am
Bill Sommerfeld (www):

the primary reason for the US patent system is to protect the inventor

According to the Constitution, the primary reason for the US Patent System is to "To promote the progress of science and useful arts"; the granting of patents is merely a means to that end.
5.7.2009 8:06am
Bruce Hayden (mail):
I would contend that software patents *are*, in practice, fundamentally different from anything seen in the past. If executable object-code is analogized to a patented machine, then the source code would be analogous to a very detailed blueprint for the machine. However, the cost of building the executable code from the source is essentially zero, and furthermore, the cost of duplicating and redistributing the object code is also essentially zero. So, if one developed a open-source program that used a certain patented algorithm (e.g., an MP3 codec algorithm), then one could distribute the software "blueprints" (i.e., the source code) but not the object code, and users would be legally forbidden from compiling without a license from the patent holder. Given all this, the notion of a software patent basically amounts to the notion of someone owning an idea (an algorithm) and being able to forcibly prevent others from using the idea (i.e., executing a program that implements the algorithm.
Sorry to sound snarky, but have you ever made your living writing software? And, in particular, worked on large software projects?

I would argue that your suggestion here is similar to the understanding of software by Supreme Court justices born late in the 19th and early in the 20th centuries.

Hundreds of billions of dollars are spent every year around the world on developing software. Much of the software patenting that I have seen in my practice is the result of a lot more development time, effort, and money than most of the non-software and non-electronics patents I have dealt with.

And then you get into the interface problem - where do you draw the line? The same algorithm that may start in software can start with a computer implementation, and then move into various levels of hardware, including being implemented as custom ICs, all within the life cycle of a product. That means that (intentionally) the system containing the algorithm, etc. can be seen as a black box, with the implementation in software, hardware, or some combination thereof, as irrelevant, and, indeed, (almost) impossible to discern. Banning software patents would create a situation where the inside of that black box would be critical, and allow potential infringers to infringe by merely moving what was in circuitry, into code.
5.7.2009 1:58pm
einhverfr (mail) (www):
Hmmm... You might be misunderstanding my bit about the expenses of patent litigation.

First, I am not saying this is new. In fact, it is old. Patents have always been extremely expensive to litigate on both sides.

Secondly, I am not saying this is a part of the problem. Actually, it is one of the things that keeps this from becoming a huge mess. If I run a small software business and bring home $100,000 per year, and the business makes $20000 a year on top of that in profit, there is simply no way for a litigant to make back money through patent litigation. Covering court costs is hard enough on the defence side. It might force me out of business, but this is only a consideration where the barrier to entry in the market is high enough that patent suits are effective in controlling that. While the automobile and the sewing machine required factories with expensive machines to commercialize, a software shop could be started by one person with pre-existing personal assets and no additional purchases or capital. Furthermore, the barriers to entry for much of this work is actually decreasing.

Thus we really have to look at three scenarios regarding patents on computer software as to how they play out:

1) Submarine patents (Unisys compression patents, for example). Company threatens to sue. Everyone runs away from the patents. Usually the patent holder is forced to then stop pushing for licenses.

2) Little player sues big player for patent infringement (NTP vs Blackberry, for example). Sure, this is a problem but it is the cost of doing business. I don't see it having a strong anti-invention impact on the industry.

3) Big player starts trying to remove competition from small players through patent claims. Microsoft's threats regarding patent infringement in Linux is a good example. This is generally shown to be remarkably ineffective, but it is the one that folks tend to think is the most scary.

I think at this point there is real question as to whether patents are actually useful in the software industry, or whether they are simply something that lack a great deal of the utility that they possess elsewhere.

Now on to one area where patents pose genuinely new problems: Genetically modified crops. These are not fundamental problems with patents but rather how they apply to genetically modified crops.

Plant patents are noting new and are not overly problematic. A normal plant patent protects the discoverer or breeder of a specific plant from infringement of clones of that plant (clones being products of cuttings, graftings, etc. from the patented plant). Plant patents only cover asexual reproduction, so you can take two patented roses, breed them, and patent the result without having to license patents regarding the parent plants. Plant patents therefore are only really applicable to a small subset crops and are kept in check by the fact that a large percentage of competitors are in the public domain. Furthermore, if a cutting shows a sport mutation it is suddenly unencumbered and could be grafted, etc. royalty-free or even patented in its own right.

With GMO patents, however, products of sexual reproduction are covered under the patents. This means that a farmer may not gather the seed from one crop to plant the next year because those plants may contain the patented genetic sequences. The industry is dominated by a couple of big players, and there is an attempt to use this to gain CONTROL over industries which are currently unencumbered by patents, such as rice farming in Indonesia. The control that is offered under these patents is unprecedented. The solution IMO is to oppose attempts to ensure that free trade agreeements require nondiscrimination in GMO imports to developing countries and allow governments to restrict such imports.
5.7.2009 7:05pm
einhverfr (mail) (www):
One unfortunate impact of software patents is that work on things like open source MP3 players occurs primarily in countries that don't protect software with patents.

As to the question of the interface line, I would be inclined to draw that line at the hardwiring of the integrated circuit. THe main reason here is that a logical progression of the rough steps you take to accomplish something mathematically might not be patentable in an ideal world, but the specific invention you create to do it might be. Dedicated hardware poses unique opportunities and challenges and these are better candidates for patent protection. Hence even in that case, a competitor could take the same algorithm and implement it radically differently on a chip and thus fail to infringe.

The other problem is this: Where do you draw the line between patentable algorithm and abstract concept which is not subject to protection? Edgar Codd invented Relational Algebra, which is roughly a sequence of innovations based on set theory for mathematically representing and manipulating information in a specific way. He also invented Relational Calculus which includes some advanced uses of this sort of thing. Why would a Left Join as a mathematical operations be unpatentable but a Left Join operation in software be patentable? Why should there be this difference? In reality they are basically identical and most relational query languages are (imperfect) representations of Codd's abstract inventions. In most cases, software represents a set of abstract mathematical instructions.
5.7.2009 7:19pm

Post as: [Register] [Log In]

Account:
Password:
Remember info?

If you have a comment about spelling, typos, or format errors, please e-mail the poster directly rather than posting a comment.

Comment Policy: We reserve the right to edit or delete comments, and in extreme cases to ban commenters, at our discretion. Comments must be relevant and civil (and, especially, free of name-calling). We think of comment threads like dinner parties at our homes. If you make the party unpleasant for us or for others, we'd rather you went elsewhere. We're happy to see a wide range of viewpoints, but we want all of them to be expressed as politely as possible.

We realize that such a comment policy can never be evenly enforced, because we can't possibly monitor every comment equally well. Hundreds of comments are posted every day here, and we don't read them all. Those we read, we read with different degrees of attention, and in different moods. We try to be fair, but we make no promises.

And remember, it's a big Internet. If you think we were mistaken in removing your post (or, in extreme cases, in removing you) -- or if you prefer a more free-for-all approach -- there are surely plenty of ways you can still get your views out.