USS Clueless              Voyages of a restless mind

Main: normal long no graphics Contact Log archives Best log entries Other articles Site Search

Patents and Intellectual Property

There are now a lot of people who are objecting to patent law, using slogans like "Information wants to be free". (Which is silly: information shouldn't be anthropomorphized, and information doesn't want anything.)

What a lot of these people don't understand is that patent law is the great success story in opening up proprietary technology. That is its purpose, and nothing has been found which does it better. Patent law is extremely subtle, and its real purpose isn't obvious to anyone who looks only at the short run.

Consider for the moment the position of an inventor without any patent law. Why should he reveal what he has created? If he does, then his competitors can duplicate it, and he gets nothing for his brilliance and hard work (and usually considerable investment). In this situation, he wants to keep as much of his invention secret as he can, for as long as he can.

But even then, he's not out of the woods; because simply by analyzing copies of his invention, his competitors may be able to duplicate it.

Patent law offers him something important, but at a high price. If granted a patent, then he gets exclusive ownership over the technology for a limited duration, and has the privilege of preventing others from making things which work the same way even if they are created independently. But in exchange for this, he pays the following prices: he agrees that the technology will enter the public domain in 20 years, and he has to explain in public exactly how it works so that when the patent expires others can duplicate it. And this was deliberate; it was the purpose of patents from the very beginning. The goal of patents is to encourage inventors to reveal how their inventions work and to release that intellectual property into the public domain.

But to convince inventors to do this, it's necessary to give them something in return. The 20 years of exclusive rights to the invention are the prize which makes it worthwhile for an inventor to reveal how the invention works. 20 years is not a long time, and in the long run this causes technology to expand faster.

Patent law is enshrined in the US Constitution for precisely this reason. As short as that great document is, the founders felt it important to include granting of patents as one of the capabilities of the US Government, because they had observed how valuable patents were toward economic and technological progress. Even 200 years ago it had become apparent that countries with strong patent laws tended to flourish relative to the ones which didn't.

Hence the paradox: I support patent law because I want as much technology in the public domain as possible. And nothing has ever been found which is more effective at moving proprietary technology into the public domain.

 

Which is not to say that a patent prevents innovation even while it's in force. It's possible to challenge a patent on numerous grounds, or to demonstrate that your device doesn't infringe a patent. A patent consists of two substantive portions amidst all the boilerplate: the description of the invention and the claims.

The description of the invention is the reason that patents open technology, because it must be sufficiently detailed to permit a person "skilled in the art" to reproduce the invention without further information. Thus when the patent runs out, everyone can and will make their own. Indeed, one way a patent can be challenged is by demonstrating that the description of the invention is incomplete or not sufficiently detailed so that it can't be reproduced. But that doesn't happen often.

The legal protections offered by a patent are embodied in the claims. These are a sequence of statements which describe the aspects of the invention which are to be protected by the patent, along with reasonable applications and extrapolations of that invention to other reasonable applications.

Usually this consists of a series of base claims, each followed by progressively more expansive "derivative claims" based the base claim. This is done for legal reasons; if one of the derivative claims is challenged successfully then the base claims and some of the lesser derivative claims may still stand. So the claims are sort of like the skins of an onion, or successive walls of a fortress, providing a series of fallback positions. On the other hand, if you can challenge one of the base claims, all the derivative claims fall. But this is harder because the base claims are quite restricted and are scrutinized very carefully. Here's a totally fictional example:

1. We invented a pushbutton.

2. We think we can put a plastic cap on it.

3. If we do this a hundred times with different caps, we can make a keyboard.

4. The keyboard can be used to enter data into a computer.

5. This facilitates making a convenient way of producing nearly any user interface for a computer.

So, from the hypothetical invention of a pushbutton, they try to claim control over nearly all computer user interfaces. This is a little over the top, but I've seen worse in patents. Another reason for using a series of claims like this is to see how much you can sneak past the patent examiner. He might grant 1-3 but not grant 4 or 5.

So, we've worked our fannies off and created a device and started shipping it, and then one of our competitors points out a patent they have. What next?

This happened once at a company I worked at, and I was handed the patent and told "Do your best." So I read it in detail, analyzed it carefully, and then composed an analysis of each of the 42 claims in it and showed a plausible case for why we weren't in infringement of that claim. I was able to make a successful case; I never was asked about it but I later saw something written by our lawyer (who never felt the need to talk to me) which echoed my conclusion: we were not in infringement of 41 of the claims, and the 42nd was "overly broad". I guess the other company was convinced as well, because it wasn't necessary for us to go to court and we did continue to ship.

I've been involved with patent law as long as I've been an engineer, both in working with lawyers while writing patents of my own and in challenging patents from other companies. I'm not a lawyer but I now know enough about the fundamental law to discuss it intelligently. Here are the main ways of demonstrating that you're not infringing a patent:

1. Doesn't apply. Simply, that what we did doesn't fall within the claim. This comes down to analyzing the details of the claim and trying to show that it doesn't describe us. It's a matter of fact and by far the strongest defense against a claim of patent infringement because if true it's usually pretty easy to prove in court. Of course, it might not be possible.

2. Contract law. If we've negotiated a license to the patent and paid royalties, then we can't be sued for infringing the patent.

All the other approaches come down to trying to demonstrate that the patent is invalid and was issued by mistake. That's more difficult; the patent holder has more to lose and will defend it more vigorously. "Doesn't apply" means that we're not infringing but it leaves the patent in place. Any successful challenge takes away the patent claim entirely, not merely as it applies to us but as it applies to everyone. Obviously the patent holder takes a dim view of this.

3. Prior art. Patent law doesn't permit anyone to protect anything. You have to have created it, and it has to be new. If someone else did it first, even in secret, even if you didn't know about it, then the patent claim is invalid. The "keyboard" patent above couldn't be issued now because switches and keyboards have existed a long time. This is a very powerful claim because it is also a matter of fact and usually easy to prove. Indeed, there used to be (and I think still is) a time window of 1 year after you first reveal your technology within which you had to file your patent, and if you missed that window then you lost your right to file for a patent. (And of course, no-one else could file a patent because of your prior art.) This is by far the most common way of challenging a patent claim.

4. Obvious. This is more difficult to prove, but it's a valid defense nonetheless. Patent law requires true innovation. If a patent is issued for something that nearly anyone "skilled in the art" could think of without any trouble, then the patent claim is invalid. Unfortunately, "obvious" is in the eye of the beholder, and you have to prove it to a judge. Generally the standard of proof for this is very high. Usually this is difficult to prove in court, so this is not a favored defense except when the claim is egregious. Also, the exact wording of the claim is important. For instance, IBM created and patented a way of using copper on ICs. Shortly thereafter several other companies revealed that they'd also been working on copper technology. Previously, aluminum had been used and the electrical properties of copper are known to be superior. (Silver is even better, the best known for any pure element, and it wouldn't surprise me if someone out there is working on use of silver in ICs.) IBM was granted patents on the specifics of their technology, but if they were granted a patent on the concept of using copper on ICs, it would have been challenged on grounds of being "obvious".

5. Overly broad. The "nearly all user interface" claim above deriving from the invention of a push button would be overly broad. Patent law permits reasonable applications of the invention to be protected, but not unreasonable ones. Again, this is a judgement call but it's usually easier to prove than "obvious" is. Judges experienced in patent law will usually understand this without too much difficulty. In this case there's no evidence needed beyond the patent itself; it comes down to convincing a judge.

6. Significant improvement. This one is also hard to prove. If what we're doing falls within the claims of the patent but is actually dramatically better, then we're not in infringement. Indeed, in such a case we might even be granted a patent of our own. The difficulty is that we have to be a lot better. The standard of proof for this is very high and this is not a favored defense.

7. Dirty hands. This is a rather strange defense; it means that the owner of the patent did something underhanded. In one case I can think of, one company was suing another over a patent and it turned out that the patent holder had stolen something (a prototype) from the other company as part of the process of investigating it. (The patent holder was lucky that criminal charges weren't filed against the employee who did it.) This is luck; you can't rely on the patent holder being so stupid. But if they are and hand you a weapon like this, then it's an extremely powerful defense indeed. Judges have a tendency to punish "dirty hands".

8. Contract law deux. This is a relatively new form of challenge and derives from contract law. Sometimes members of an industry will form a consortium to create an industry standard for something. Not always, but sometimes such an industry standard body will require all participants to give up the rights to any intellectual property which might cover that standard, so that the standard will be open and products compliant with it won't require any license. If a company participates in such a body and makes that agreement, and then later pops up with and tries to enforce a patent on a critical part of the standard, their rights to enforce that patent can be taken away from them. (In essence, it will be argued that they already granted everyone perpetual non-royalty-bearing agreements to the patent.) Dell Computer lost rights to a patent this way, and right now (3/2001) Infineon and Hyundai and Micron are trying to invalidate certain patents belonging to RamBUS on grounds of participation by RamBUS in JEDEC which required such an agreement from its members.

It's not always the case that an industry standard is "open". Indeed, it's very common for an industry standard to require a license to intellectual property. Ethernet was a standard which required a license from Xerox. GPIB (IEEE-488) was a standard which required a license from HP. IS-95 (the standard governing CDMA cell phones) requires a license from Qualcomm. The mere fact that something is an industry standard doesn't mean it's open. However, it does if participation in the standards process required an up front agreement by all participants that the resulting standard should be open, such as was the case in the development by JEDEC of the standard for SDRAM. It remains to be seen whether RamBUS will lose its rights to the disputed patents.

In the case of the patent I analyzed that one time, I was able to directly challenge all but one of the base claims on the basis of "Doesn't apply", and that last one on the basis of it being "overly broad". In each of the former cases what I wrote was a detailed explanation of what the claim actually meant, and then show how what we had done was outside of it. My "overly broad" claim was more a matter of judgement and I simply suggested it as grounds. My write-up was intended for our lawyer. He apparently thought it made sense because he said the same thing in the letter we sent to the other company. They evidently didn't feel like pushing the case; my technical arguments on all the points except one would have stood in court, and they risked losing the last one if we had successfully convinced a judge it was "overly broad". Better to simply let it go.

 

Even during the period of protection, patents usually don't lock up technology. It's rare (although completely legal) for a company to hold its patents closely and refuse to license them, the most common example of this being the Polaroid corporation's patents governing instant film. Far more common is for patent holders to license those patents freely in exchange for royalties or rights to patents owned by others. And they can do this without fear because they retain their rights under the patent. If they were not protected by patent, they'd be much more fearful of doing this. Therefore, even in the short run, the protections offered by a patent facilitate the sharing of technology.

So patent law encourages the creation and distribution of technology by encouraging inventions and by encouraging the inventors to reveal how those inventions work, and best of all by encouraging those inventors to release their inventions into the public domain. What's not to like?

This page has been viewed 1538 times since 20010726.