By Professor Chris Cotropia
Well, we have gotten early indications from the major patent players as to how the substantive part of KSR will play out. TSM still has a place in the post-KSR world, at least according to Chief Judge Michel of the Federal Circuit and Deputy Commissioner Focarino of the USPTO. These are preliminary statements. We will not know what the Federal Circuit or USPTO really think until the first precedential opinion or interim guidelines issue. But these early comments make one wonder if KSR changes anything, particularly those like myself who witnessed the Federal Circuit's acceptance of, and reliance on, implicit TSM well before In re Kahn, DyStar, and Alza.
While the substantive impact may turn out to be minimal, KSR could have a significant procedural effect. The Court appears to have shifted the line between factual and legal parts of the nonobviousness analysis, moving what was (is?) the TSM test from being a question of fact to a question of law.
The Supreme Court in Graham established that nonobviousness was ultimately a question of law, but there were recognized underlying factual issues—content of the prior art, scope of the claims, and the level of the PHOSITA. When the CCPA, and then the Federal Circuit, started employing the TSM test, the test became another underlying factual component of the inquiry. As a result, there was not much left at the question of law level, and the Federal Circuit affirmed most appeals from rejections by the USPTO and or trials before district courts on the nonobviousness issue.
The Court in KSR introduces a procedural change, folding the TSM-like inquiry into the question of law level of the analysis. This move is most clearly witnessed in Part IV of the opinion, where Court rejects the existence of a dispute over an issue of material fact. Slip op. at 23. The Court goes out of its way to note that "[t]he ultimate judgment of obviousness is a legal determination," rejects the ability of a "conclusory affidavit[s] addressing the question of nonobviousness" to create a fact issue, and shows concern only to whether the first three Graham factors are "in material dispute." Id. In contrast, the final step, determining whether a PHOSITA would modify the prior art, is identified as a "legal question." Slip op. at 21. The earlier, general discussion of the substantive standard also supports this interpretation. The opinion focuses on what "a court" should analyze when determining whether there is "an apparent reason to combine." Slip op. at 14.
The implications of this procedural change are many. Making the TSM-like analysis a question of law facilitates more summary judgments on the issue of nonobviousness (although not as much as one would think, as I determined here). It also gives the Federal Circuit more freedom to reverse both USPTO and district court judgments on nonobviousness. This may lead us down a similar road as Markman has, giving the Federal Circuit so much flexibility that it injects even greater uncertainty into an area that is inherently uncertain, particularly after KSR. We might even end up with KSR hearings.
Admittedly, the Court does not explicitly say it is making this change. And, as the Federal Circuit has done in the past, TSM-like discussions can be framed as part of one of the first three Graham factors that have always been factual issues. However, the main focus of the opinion is on the Federal Circuit's "application" of § 103, and the language and application in KSR strongly suggest a change in the law/fact line in nonobviousness analysis (or maybe this is all just a "common sense" reading).
Chris Cotropia is an Associate Professor of Law at the University of Richmond School of Law and is part of the School's Intellectual Property Institute. Professor Cotropia was counsel of record and a co-author on the Brief of Business and Law Professors as Amici Curiae in Support of the Respondents in KSR.
Nice recap. Thanks.
One other point, someone made a very salient comment about the “equivalence between software and hardware.”
I’m no computer engineer, but I did stay at a Holiday Inn last night and I wrote my first computer program about 30 years ago, so I know a little about the subject. Anyway, software and hardware are interchangeable. A computer program (i.e. software) can be hardcoded into a processor. The result is a unique physical structure within the processor that performs the program. I don’t think that many people would argue that such a structure isn’t non-statutory subject matter.
As such, why should it matter, with regard to a determination as to statutory subject matter, whether or not the computer program is soft-coded or hard-coded into a processor? Moreover, since the hard-coded program would certainly be statutory subject matter, then the soft-coded program should also be statutory subject matter.
Yet another aside, for those who ware looking for the Supreme Court to abolish software/business method patents, I think they are barking up the wrong tree. They should be looking to Congress, not the Supreme Court. The language of 35 USC 101 is pretty expansive, and I very much doubt that the Supreme Court has much to hang its hat on.
As to the person who believes “software” should by copyright protected, you really have no idea as to the difference between a patent directed to software and the copyright to software. Unless one is trying to protect the look and feel of the output of a software program, the copyright to a computer program only protects the code. However, most software patents don’t include (or don’t care about) the code used to implement the software program. Instead, what a software-related patent tries to protect is the over-arching concept(s) behind the code.
MM free
After reading a few comments in this thread, it appears that a few people don’t know how a “software patent” is actually claimed.
The term “recipe” has been thrown out there. However, this term is, in itself, ambiguous since it can takes on different meanings. A recipe, as a series of steps to be performed by a chef or someone making Dipping Dots, is patentable since processes/methods are statutory subject matter, as expressly set forth in 35 USC 101. However, a recipe in its printed form (i.e., in a cookbook) is likely considered to be functional descriptive material. Functional descriptive material, when claimed as descriptive material, per se, is not statutory (as an aside, I can say with confidence that less than 1 in 10 examiners know what the term “per se” means). A discussion of the case law regarding descriptive material (both function and non-functional) is found in MPEP 2106.01.
Claims that can be considered “software-related” are NEVER claimed as functional descriptive material, per se. To even insinuate that software is claimed in such a manner is to evidence a clear lack of understanding of this particular art. Instead, the (3) most common ways of claiming software is: (i) as a method; (ii) as a system; and (iii) as a series of steps (i.e., the software program) recorded on some computer-readable medium.
The method claims are just the series of steps performed (mind you, these steps can be performed by separate and distinct entities). The system claims vary in form but typically recite specific elements performing a particular function (such as performing the steps recited in the method claims). Working with the cooking recipe analogy, some of the system elements may include a blender, an oven, a cutting knife, etc. as part of the structural elements that perform a particular function.
As another aside, food products are patentable subject matter as well as the method of making the food products (i.e., the recipe). Quite some time ago I had the opportunity to prepare an application directed to a food item. One of the good parts was that I got a sample of the food item (quite tasty) to try out during a trip to the manufacturing facility. Food items are difficult to claim because the extent of prior art and the level of creativity of one having ordinary skill in the art is quite high. For this particular food item, the process of how the food item was manufactured left a tell-tale structure in the food item itself. If I recall correctly, one set of claims may have been product-by-process claims.
MM free
“Be honest and say what you REALLY mean – i.e. that NOTHING from the electronic arts deserves patent protection. Then you would be honest and would not be betraying your utter ignorance of computers.”
You would think that after Microsoft the promoters of software patentability would realize that saying “You’re dumb! You’re dumb!” convinces nobody.
“What about hardware? Does hardware deserve patent protection?”
A given piece of hardware deserves patent protection if there is something novel and non-obvious about it BEYOND any information (i.e., “code”) that is inscribed on its surface that is “read” by another part of the machine.
TO “me”,
What about hardware? Does hardware deserve patent protection?
If yes, so why should code ‘hard-wired’ or ‘burned’ into computer hardware (for example, using FPGAs) deserve patent protection but computer code that resides in volatile memory (i.e. software) NOT deserve patent protection?
It seems that you have not heard about equivalence between software and hardware.
So at least be honest and say that NEITHER hardware NOR software deserve patent protection.
Be honest and say what you REALLY mean – i.e. that NOTHING from the electronic arts deserves patent protection. Then you would be honest and would not be betraying your utter ignorance of computers.
“If someone doesn’t know the difference between software and a law of nature, then they are not competent to adjudicate a patent case involving software technology.”
That is my point, software should be copyright protected, like a recipe – and not patented.
Also, regarding the hoped-for eventuality that the PTO will “blitz the backlog”, I point out that KSR is the tri-consonontal afro-semitic root for “break.” According to Juland bin Jaifer Al-Said: “From the root k-s-r, the form I verb is kasar, “he broke”; form II is kassar, “he smashed to bits”; and form VII is inkasar, “it was broken up.”
Smashed to bits indeed!
Re the debate about software patentability:
CAFC, 2005: “Without question, software code alone qualifies as an invention eligible for patenting.”
Mr Waxman, for AT&T: “We all agree that software code in and of itself, removed from a physical structure, cannot be patentable and when software—when some—when an invention that is practiced with software is patented, at the most what you will see is preferred embodiments of the source code which is language that humans understand and which computers do not.”
Is “obvious” fact or law. Let’s use our common sense. Grounds of invalidity are points of law. It is for the court to say whether a claim meets the provisions of 35USC103. But the court needs the facts on which to make that determination. Whether the PHOSITA would have found it obvious, on date A, to have modified teaching D1 by the introduction into it of technical feature X, to create something falling within the scope of the claim in view, is a matter of fact, isn’t it?. Armed with that fact, a court, or a PTO Exr, or an opining attorney, can decide on 35USC103. I hate secondary consideration tests. They rely on things that happened AFTER date A. They muddy the water. There ought to be a workable test for obviousness, that does so without recourse to an examination of after-arising facts. There is. It’s called EPO-PSA. In another thread on this blog, Mark Perdue asserts that EPO Problem and Solution is no different from what the USPTO is already doing. In that case, stand by folks: the USPTO, armed with KSR, is about to blitz the backlog.
“But recipes as such (a distinction I clearly made in my comment) are not patented because (hold your breath) they aren’t patentable subject matter (a feature that recipes share in common with — get read for this!!! — laws of nature).”
I’m not sure the reason why recipes are not patentable is because they’re not patentable subject matter, for recipes are laws of nature?
I thought it’s because the effect of a recipe cannot be quantitatively measured. How do you measure deliciousness of food?
“Second, recipes are patentable. They are usually patented in the form of processes”
Um, yeah, processes are usually patented in the form of processes.
But recipes as such (a distinction I clearly made in my comment) are not patented because (hold your breath) they aren’t patentable subject matter (a feature that recipes share in common with — get read for this!!! — laws of nature).
The CAFC has held that printed instructions to a novel method in an old kit will not render that old kit patentable.
The claims in Dipping Dots did not cover “recipes”, i.e., printed instructions describing how to make food. To the extent there were valid claims in Dipping Dots, those claims were process claims or composition claims.
MM wrote:
If someone doesn’t know the difference between software and a law of nature, then they are not competent to adjudicate a patent case involving software technology.
Computers can read text and control robots.
What is the difference between software and a recipe that is capable of being read and followed by a computer-controlled robot chef?
There must be a difference. Otherwise, the recipe (i.e., the ingredients and instructions, written on a piece of paper) would be patentable and everyone agrees that recipes are not patentable.
Response:
First, a recipe is not a law of nature. Therefore, I have no idea how you are tying my point with yours.
Second, recipes are patentable. They are usually patented in the form of processes (a process for making concrete is the same as a recipe for making concrete). Dippin Dots had a recipe patent, for example.
If someone doesn’t know the difference between software and a law of nature, then they are not competent to adjudicate a patent case involving software technology.
Computers can read text and control robots.
What is the difference between software and a recipe that is capable of being read and followed by a computer-controlled robot chef?
There must be a difference. Otherwise, the recipe (i.e., the ingredients and instructions, written on a piece of paper) would be patentable and everyone agrees that recipes are not patentable.
Eric Wilson’s five good defenses – good list.
Unfortunately, (iii) will be nearly impossible to establish without a TMS requirement.
Also, iv and v involve information not readily available during prosecution.
I disagree with Mr. Sweenze that the defenses are only good if the Examiner cited prior art for a motivation to combine. Those defenses should be good no matter what the Examiner’s “reasons for obviousness” are.
I agree with “me” that Patent Trolls have acquired a disproportionate amount of software patents. Thus, trolls have hurt the software industry more than other industries.
However, is that a reason to punish the software industry by denying patents for legitimate inventions? If the trolls shifted focus to the mechanical arts, would you want to exclude mechanical inventions from patent protection?
TJ said:
Someone forgets that, once upon a time, the debate over TSM was considered dead over three decades ago, except in certain academic backwaters. It takes decades for academic ideas to reach the courts. But the patentability of software (or, at least, mathematical algorithms and laws of nature) will reach the Supreme Court again; and this time it might not take three decades.
Response:
In the case of the patentability of software, the problem is not getting an idea from the academics to the courts. The idea has gone to the courts and been decided. After quite a bit of floundering, the court finally came to several decisions that made a lot of sense. (Allapat is still one of the most sensible decisions I ever recall reading)
Nevertheless, it appears that a few professors out there still like to talk about the issue as if it hasn’t been resolved. I just have to roll my eyes when a law student questions me about the patentability of software because some out-of-touch professor wasted the student’s time by having them study a line of old badly-reasoned already-overruled cases.
If someone doesn’t know the difference between software and a law of nature, then they are not competent to adjudicate a patent case involving software technology.
“Now, when the examiner says “common sense says its obvious”,”
They will have to say more than that to establish a prima facie case. But maybe you were just speaking figuratively, in which case we can only guess as to what you’re driving at.
While I don’t disagree that the software horse is injured, I do not yet believe he has yet died.
Where do we get the most patent trolls? Software and biz methods – remove them and most of the problems immediately vanish!
Re Eric Wilson’s five good defenses: If the examiner does not cite the prior art for the 103 motivation to combine, then your five defenses are useless at the USPTO level.
KSR shifts the 103 burden from the examiner having to prove obviousness to the applicant having to prove non-obviousness. KSR gave legal weight to the examiner’s subject knowledge. Now, when the examiner says “common sense says its obvious”, they need only cite KSR as support for their subjective opinion. The applicant then will have to establish what was in the mind of the examiner when he/she said “common sense says its obvious” and then the applicant will have to show why that is not true. The examiner need only respond “That’s not what I was thinking” and give a final rejection, where the matter is off to extra innings, which only the rich can pursue.
“The debate over the patentability of software ended over a decade ago, except in certain academic backwaters.”
Someone forgets that, once upon a time, the debate over TSM was considered dead over three decades ago, except in certain academic backwaters. It takes decades for academic ideas to reach the courts. But the patentability of software (or, at least, mathematical algorithms and laws of nature) will reach the Supreme Court again; and this time it might not take three decades.
“Why didn’t the SC take the opportunity in Microsoft to do away with software patents”
The debate over the patentability of software ended over a decade ago, except in certain academic backwaters. It’s hard to believe that some people are still whacking this dead horse.
Anyway, good article but not particularly relevant to patent prosecutors. The distinction between law and fact is not something that Examiners care about. To them, KSR simply means that they no longer have to provide any evidentiary support for the reasons they give for something being obvious.
“Making the TSM-like analysis a question of law facilitates more summary judgments on the issue of nonobviousness (although not as much as one would think, as I determined here). It also gives the Federal Circuit more freedom to reverse both USPTO and district court judgments on nonobviousness.”
Does the author mean obviousness here (as opposed to non-obviousness) or am I missing something? It is a rare case when a plaintiff wins (or even requests) summary judgment/adjudication on non-obviousness. Rarer still would be the case when the the USPTO finds non-obviousness and someone requests the Fed. Cir. to reverse that. Presumably he maens that the ruling facilitates summary judgment on the issue of obviousness and reversals of same.
The guidance in KSR is all in the direction of making patents “advance the useful arts and sciences.” I think the court has been made aware of economic arguments that carving out exclusive territories that did not take much work to define has had the opposite effect on US competitiveness. Any Mech. Eng. grad (defined in this case as the relevant benchmark) would know that instrumenting the pivot of a hinged part is the appropriate thing to do. By saying that you can’t carve out accelerator pedals as your exclusive domain the court has reiterated that you need real insightful innovation to get a patent.
Where prior art is said to combine into a development characterising a claimed invention, the KSR judgement spells out four or five good defences:
i) That the prior art ‘teaches away’ from the invention (page 12, paragraph 2, also page 18 item A, and page 22 end of paragraph 2).
ii) That the prior art cannot be upgraded to the invention (Page 22, paragraph 2).
iii) That the prior art requires elements of the invention to be read in using hindsight to be relevant to the invention (Page 17 paragraph 3).
iv) That the prior art has secondary factors which may ‘dislodge’ obviousness – “long felt and unresolved needs”, “the failure of others”, “commercial success” (page 2 paragraph 3).
v) Anticipatory “design need and market pressures” were found in KSR against the patentee (Page 17 paragraph 2) – but will often point the other way concerning genuine inventions, speaking against obviousness.
I believe the re-statement and discussion of these, in a single authoritative judgement concerning easily understood subject matter, helps to somewhat clarify the process for inventors like me
Nice article by the way, and nice brief!
Wasn’t shifting the TSM to a fact analysis the whole point the Fed Cir was attempting when they created and relied on the “old” TSM? They did this to create certainty and a test that could be applied evenly by many. To think Patent Examiners, District Court Judges, and all the rest of the folks who want certainty in the patent system can apply a fuzzy “motivation” standard is a bit silly. This standard will do away with some patents, but create a huge amount uncertainty along the way. Is it really worth it?
Why didn’t the SC take the opportunity in Microsoft to do away with software patents, and then find a case to overturn State Street Bank (biz method patents). With these two areas as non-patentable subject matter under 101 alot of the patent troll issues go away and we don’t need sweeping changes to the system.