Thursday, October 30, 2008
Bilski and Continued Uncertainty in Patent Subject Matter
Hot on the the heels of my latest post, which discusses algorithms and business methods patents, the Federal Circuit has issued an en banc opinion shaking the tree in In re Bilski. Alas, the opinion does not follow my suggestions in Everything is Patentable, but the decision isn't as far from my views as it could be.
Let me be clear - I think this is a patent application that is non-inventive, and that should not issue for a variety of reasons. I don't think, though, that subject matter is one of them.
The inventors claimed a method for hedging on consumable commodities. The method involved an intermediary selling commodities at a fixed price and buying commodities at a fixed price. Hedging, in general, is well known, though the inventors here claim to have come up with a better way to do it.
The patent application was rejected by the PTO as not embodying patentable subject matter and not being directed to the "technological arts." The patentee argued that his process, however, achieved a useful, concrete result as was contemplated in Federal Circuit precedent.
The Federal Circuit held, 9-3, that the patent was not directed to patentable subject matter. That's not really a surprise. Even among the three dissents only one judge (Newman) would have found this to be patentable subject matter. Further, the specific holding effectively wiped out the last 10 years of patentable subject matter jurisprudence - it's like Bobby stepped out of the shower on Dallas.
The old rule - State Street Bank and its progeny - basically held that a "useful, concrete and tangible result" was sufficient for a process to be eligible subject matter. So, methods of managing money by computer or measuring heart rhythms, etc. were useful, concrete and tangible. Bilski argued that hedging yielded such benefits. Not so, says the court now. To the extent all of those cases relied explicitly on that rule, they should not be relied on now. And here I just finished a class where I drilled that black letter law into my students' heads.
Here is the new rule - any such process must either be tied to a machine or be a transformation of something physical. However, the transformation can also be a transformation of data representing something physical. But, insignificant "post-solution" machines or transformations don't count. So, because Bilski is not tied to a machine (though no human could do the calculations in the patent), and because the transformations involved are of legal obligations and not anything physical, and because anything physical is "post-solution," the patent is not a "process" and thus not patentable subject matter. How's that for clarifying things.
Here's how the court gets there:
1. We know that natural principles aren't patentable.
2. We know it is difficult to determine what is and what is not a natural principle.
3. We know for sure that there are two ways to tell if something is not a natural principle:
A. If it is tied to a machine or
B. If it transforms the subject matter
4. Therefore, the only way for a process to be patent eligible is to be tied to a machine or tranform subject matter.
5. In order to make sure all principles are excluded, insignificant post-solution machines or transformations don't count.
Anyone else see the logical jump from 3 to 4? I put this leap on the level of (i) if you weigh the same as a duck, you much be a witch (for you Monty Python fans) and (ii) Descartes' proof of the existence of god (for you philosophy fans). To make the leap, the court considers court precedent in what I consider to be a completely unsupportable way (and Judge Newman's dissent takes the majority to task on this point). I'll discuss that in another post, as this one is already pretty long.
So what's the problem?
So, what's the problem with the Court's formulation, aside from the major logical flaw? After all, the concession that the transformation can be of data relating to physical matter is a step forward and something we even argued in our amicus brief.
Well, let's start with unsettled expectations. Is State Street now patentable? It is a transformation of data relating to money. Is money physical? What about transformation of data relating to weather patterns? That's historical data, but doesn't necessarily represent physical data. What about communications? And by the way, if I am transforming data, aren't I tied to a machine? Why do we need the transformation part at all?
Second, in its effort to deal with the high technology, the court has abandoned low-technology. What about all those processes out there that have nothing to do with machines or transformations - new methods for threading a needle more quickly (which can have great cost savings in manufacturing), methods for harvesting fruit, methods for manufacturing products by hand (for example forming wrought iron). One would think that these were what the original process patents in the 1800's looked like, and by this test they are now barred. At the very least one would have to ask whether molded wrought iron is a transformed substance - not a model of clarity. I think the ruling can have a real effect on innovation - and not just "soft" innovation, but all innovation. On a side note, Judge Newman points to English patents from the 1700's that appear to quite clearly be human performed business methods.
Third, just what are insignificant post-solution elements? How might one ever figure that out?
Fourth, the analsysis of tying something to a machine is less than satisfying. The court cites Mackay Radio as a prime example. To me, Mackay is a poster child for why this test fails. There, a well known equation covered the optimal wire lengths for receiving radio signals. The patentee claimed an antenna using these lengths. The Supreme Court said that this was eligible because it was an application of the principle. It seems to me, though, that there is no principled way to separate the antenna from the principle of nature under the test elucidated here. In fact, one could easily argue that the antenna was simply an insignificant "post-solution" part of the claim because the "real" solution was the mathematical formula.
Fifth, let's not lose sight of the actual holding here - the machine-or-tranformation test is supposed to tell us whether a claimed process is more than a "fundamental principle." By saying there is no transformation, the court ruled that the process must, therefore, be preempting a fundamental principle of nature. But can we really call the process of hedging through fixed price contracts a principle of nature? Is it really part of the scientific landscape for all to apply in whatever "applied" way they choose? Does this particular process really preempt a fundamental principle? It seems like the test is trying to shoehorn an otherwise square unpalatable claim into round subject matter rejection.
To be fair, the court is trying to stick with Supreme Court precedent. Of course, as I discuss just above and in prior posts, that precedent is contradictory, unhelpful, and unanalytical. Indeed, in order to attempt to make sense of it the Federal Circuit had to jump through some of the big hoops of logic. In other words, sticking with precedent doomed the effort from the start.
So what's the good news?
The court at least gives lip service to the notion that this must be a statutory interpretation of the term "process." This is something I argue in my article. However, in sticking to Supreme Court precedent - which performs no statutory interpretation - the court doesn't have a firm foundation.
Perhaps more perplexing and dissapointing is the quotation of (and then discarding of) the actual statutory language. Section 100(b) of the Patent Act says:
The term "process" means process, art, or method, and includes a new use of a known process, machine, manufacture, composition of matter, or material.
The court says tht this doesn't work, because the definition of the word process includes the word process. Well, that may be, but what about the rest? A new use of a known process, machine, manufacture, etc. Doesn't this very statutory language preclued the "tied to a machine" part of the test? Why can't the process be tied to a composition of matter? A manufacture? A material? Even in a world where you want to limit subject matter, to claim that you are interpreting the statute but to then exclude a whole variety of claims that fall directly within the statute does not seem to fit even the most basic of statutory interpretation principles.
So, the good news of focusing on statutory interpretation is greatly tempered byt the bad news of no actual statutory interpretation. I think there was a way to get to the court's definition of process (or something like it) such that the interpretation actually had foundation in the statute, and it is unfortunate that the court did not take up that opportunity. Indeed, the Comiskey case, which holds that a method that involved only mental steps is closer to the mark - at least that might fit into the statute because there is no use of any of the 100(b) items.
There is other good news, though:
- The case does reaffirm that business methods and any other method that meets the test is patentable
- The case does reaffirm that software is patentable
- The case makes clear that patents need not cover "technological arts" as the PTO wanted
- The case makes clear that novelty and non-obviousness have no place in subject matter considerations (the PTO guidelines on subject matter inexplicably have examiners doing prior art searches to determine subject matter eligiblity)
- The case does make clear that "transformation" is not limited to physical items (though it doesn't go far enough)
Quick notes about the dissents:
- Judge Mayer says the court doesn't go far enough, and would bar business methods
- Judge Rader says the court doesn't look hard enough (or at all) about why we might want these limitations, and also says that the limitations are too stifling. The dissent says that this is a natural principle and thus invalid even under the broad definition, so no subtest is necessary here to further confuse things.
- Judge Newman takes the court to task on many of the issues discussed here and in my other posts. Judge Newman looks at Morse the same way I do here - but agreeing with the dissent usually means a failure to win.
Finally, particularly gratifying was this line in Judge Rader's dissent:
Much of the court’s difficulty lies in its reliance on dicta taken out of context from numerous Supreme Court opinions dealing with the technology of the past. In other words, as innovators seek the path to the next techno-revolution, this court ties our patent system to dicta from an industrial age decades removed from the bleeding edge.
If that sounds familiar, you need look no further than to my earlier post here called Parroting Dicta. I wrote:
A by-product of the history is what I call "parroting dicta" - the Court makes a statement that is not necessary (or even related to) the holding, and that statement gets repeated in every case, even though it is not necessary for that case either. Eventually, lower courts and even the Supreme Court start believing the dicta despite the fact that it has never actually been helpful in resolving a case.
Unfortunately, my article didn't get to press on time, so now it will look like Judge Rader's comment justifies mine, instead of the other way around!
And truly finally, some other interesting posts on the topic:
Patently-O - Dennis Crouch
And a hat tip to Howard Bashman at How Appealing
TrackBack URL for this entry:
Listed below are links to weblogs that reference Bilski and Continued Uncertainty in Patent Subject Matter:
What about all those processes out there that have nothing to do with machines or transformations - new methods for threading a needle more quickly (which can have great cost savings in manufacturing), methods for harvesting fruit, methods for manufacturing products by hand (for example forming wrought iron).
Don't all of those things "transform a particular article into a different state"? The needle becomes a threaded needle, the fruit goes from fruit on the vine to fruit in a bucket, the iron goes from a bar to some sort of shape, etc.
Posted by: Bruce Boyden | Oct 31, 2008 12:00:47 PM
It's a different "state or thing" - I don't think your examples would cut it. The Supreme Court addressed this in Brogdex - dipping an orange in borax does not constitute a "manufacture." I think under Brogdex and similar cases, fruit off the tree is not a different state as fruit on the tree.
Posted by: Michael Risch | Oct 31, 2008 12:36:41 PM
Interesting. Is your reading of Bilski is that a process must be tied to, or produce, one the other 3 types of patentable subject matters (manufacture, composition of matter, or machine)? That appears to be what Dyk's concurrence is saying, and the quote from Comiskey appears to suggest that. But the discussion of Abele, that no transformation of a physical object is necessary, seems to go the other way. And my own uninformed reading of the word "state" would be broader than just compositions of matter or manufactures. The first dictionary definition is: "1 a: mode or condition of being ." Or, a state of being harvested. New manufactures or compositions would appear to me to be new "things," not new "states."
Posted by: Bruce Boyden | Oct 31, 2008 4:06:41 PM
OK, the last couple of lines are supposed to be:
The first dictionary definition is: "1 a: mode or condition of being (a state of readiness)." Or, a state of being harvested. New manufactures or compositions would appear to me to be new "things," not new "states."
Posted by: Bruce Boyden | Oct 31, 2008 4:08:52 PM
I think your interpretation is reasonable - I just have no confidence that it would actually be used... That said, there is some support for it in a re-reading of Brogdex. Here is the exact quote:
"Manufacture implies a change, but every change is not manufacture, and yet every change in an article is the result of treatment, labor, and manipulation. But something more is necessary, as set forth and illustrated in Hartranft v. Wiegmann, 121 U.S. 609. There must be transformation; a new and different article must emerge 'having a distinctive name, character, or use.'
If it be assumed that the process claims under consideration cover an invention, we think this lacked novelty when application was made for the patent August 13, 1923."
In other words, the product claims are invalid for not being a manufacturer, and the court assumes that the process claim could be valid (but that the process is invalid in any event). Based on this, there is an argument that the process makes the orange change state, but the court never expressly rules.
Posted by: Michael Risch | Nov 1, 2008 9:40:00 AM