“Benefit of the Doubt?” - Redux
I first posted the below item in November 2006. I’m reposting it now in response to a comment made to a recent post.
So … here’s what I wrote in Nov 2006:
You file a patent application at the Patent & Trademark Office. Are you entitled to receive the patent unless the PTO can show you’re not entitled to it? Put another way, who gets the benefit of the doubt on patentability or not … you, or the PTO? Structural rules like this play a large role in driving outcomes, as any lawyer will tell you.
The case law says that, at least so far as novelty (sec 102) and nonobviousness (sec 103) are concerned, the applicant is entitled to the claim unless the PTO proves the contrary. The current approach is stated in, e.g., In re Piasecki, 745 F.2d 1468, 1472 (Fed. Cir. 1984). But the current disregard for any textual difference between sections 102 and 103 traces to In re Warner, 379 F.2d 1011, 1016-17 (CCPA 1967). Warner makes no effort to root in the text of section 103 the burden it places on the PTO to show obviousness (rather than placing on the applicant the burden to show nonobviousness).
Is the case law right? The issue comes to the fore now because, as PFF’s IPcentral has noted, petitioner KSR has filed its Reply Brief at the Supreme Court. (You can pull the brief at IPcentral.) Oral argument is Tuesday, November 28.
There’s a strong argument from statutory text that the case law is wrong. Here it is:
Section 102 begins, “A person shall be entitled to a patent unless … ,” and continues with a variety of scenarios that, if proved, would negate the applicant’s entitlement. This mandate, “shall be entitled … unless,” is quite clear.
How does section 103(a), the core nonobviousness requirement, compare? Well, it’s phrased quite differently: “A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if … .” Unlike section 102, which expressly mandates the applicant’s entitlement unless a negating condition is proved, section 103 commands that a patent can’t be obtained if a particular condition - obviousness - exists. Given that Congress knows how to put the burden on the PTO to show lack of entitlement (as in section 102), the fair reading of section 103’s quite different text is that it puts the burden on the applicant to show nonobviousness.
Petitioner KSR points out that this structural question plays an important role in its case. In doing so, it urges a rejection of the common, undefended (and indefensible?) (mis)reading of the statute. According to KSR, which calls the common reading “an upside down view of sec 103(a),”
[t]he key to understanding [Teleflex’s] argument is the radical, unstated premise on which it is based: that sec 103 purportedly does not state a “condition for patentability” that a patent applicant must satisfy before a patent can properly be granted, but rather purportedly states a condition for challenges to patentability that the PTO or an accused infringer must satisfy before a claim to patent protection can properly be denied.”
KSR Reply Brief at 10 (emphases in original). Strong stuff.
So, we have another issue to watch for in the KSR case … will the Supreme Court state that an applicant has the burden to show nonobviousness, or that the PTO has the burden to show obviousness?
Joe,
I am sympathetic to an argument that the applicant should have the burden for both 102 and 103, or neither. I have difficulty, however, for how they could have the burden for 103 but not 102, from a purely logical standpoint. I may have raised this before.
The logical problem is that the PTO proving anticipation necessarily proves obviousness. On the same coin, an applicant proving nonobviousness necessarily proves novelty. So if you allocate the burden of proof inconsisently, the dispositive question will be who goes first. Stated another way, what if both parties introduce no evidence and thus fail their burden of proof? The PTO fails to prove anticipation, the applicant fails to prove non-obviousness. Do you grant the patent or not?
On another note, I have a quibble with you about the incremental-innovation-is-not-patentable view. Say we are increasing car mileage, and someone invents a 30 mpg car when the prior art has 29 mpg. If 29 to 30 is obvious, then 28 to 29 is obvious, and so on. But if we take away patent incentives in this fashion, the prior art might never have had the 29 mpg car (which was developed with patent incentives), or the 28 mpg car, but instead only the very lousy 10 mpg car that would be developed without any patent incentives.
I have an article about this in more detail. Email me if interested.
Comment by TJ — January 17, 2008 @ 2:52 pm
TJ,
On the 102/103 point, we’re agreed that the logical issues need to be sorted out. I’m simply stating that you can’t make the argument from statutory text alone.
On the incrementalism point, I don’t think the hypo makes sense in the abstract. I have no idea whether getting from 29 to 30 is technologically trivial or technologically gargantuan. My point is that if it IS technologically trivial (i.e., another small tweak here or small tweak there), it’s not the appropriate subject of patent protection. Put another way, the fact that KSR may lead to a noticeable reduction in the amount of patent apps granted (or filed) is a feature, not a bug.
Best,
Joe
Comment by Joe — January 17, 2008 @ 3:46 pm
Professor Joe,
When it comes to patent prosecution, the shifting of burdens is fairly straight forward. We all agree that 102 places the initial burden for making out a prima facie case of anticipation on the PTO.
There is plenty of case law to support the proposition that “anticipation is the epitome of obviousness.” In other words there is a continuum of levels of obviousness that max out at the anticipation level. So if the PTO has the initial burden under 102, it also by extension of the continuum has the same burden under 103. The PTO has always acted that way and has thus recognized it has the prima facie burden with regard to 103.
Another reason why the PTO has the initial burden under 103 comes by way of administrative agency law. The PTO cannot reject a claim without giving a “reason”. Otherwise it would be arbitrary and capricious agency action. Note also use of the word “reasons” in 35 USC 132 (”the Director shall notify the applicant thereof, stating the reasons for such rejection,”).
Once a prima facie case is established, the burden shifts to the applicant to overcome by means of argument and/or rebuttal evidence. Attorney argument is not “evidence”.
After the applicant submits his rebuttal, the prima facie case “dissolves” and the case is decided on the basis of the preponderance of evidence present in the totality of the administrative record.
KSR did not change this well established law. KSR changed that aspect which said obviousness *must always* be proved in the first instance by the show-me-the-TSM method. Examiners must still first supply a “reason” under 35 USC 132. It does not have to be a TSM-based reason.
Comment by step back — January 17, 2008 @ 4:29 pm
Step,
I agree, as I have said from the beginning, that the established practice is as you describe.
I don’t agree, however, that the textual difference between sections 102 and 103 can be disregarded on the ground that courts have referred to anticipation as the epitome of obviousness (which, of course, I acknowledge they have done).
As for grounding the PTO’s initial burden under section 103 on basic agency principles, I don’t follow your argument. It’s true that the agency must give a reason. However, were the initial burden of proving nonobviousness placed on the applicant, the reason would be a properly explained “not proved.” There’s nothing inherent to giving a reason that says one must also bear the ultimate burden of proof.
It’s true that KSR didn’t change the established practice. I suppose it would take a case on appeal from the PTO for that question to be squarely presented.
Best,
Joe
Comment by Joe — January 17, 2008 @ 5:43 pm
“[T]he reason would be a properly explained [by simply saying] “not proved.—
That’s not a reason. That’s circular logic. We reject it because we reject it.
If you want to go deep metaphysical, the inventor does have a first burden, namely to file a specification and claims that meet the requirements of 35 USC 112 and to thereby “prove” entitlement to a patent. The ball then bounces into the PTO’ side of the tennis court to either allow the claims or reject them based on the inventor’s first volley. Since the inventor files his application under oath, statements made in the application as to alleged facts are evidence and thus the inventor has submitted a first set of proofs.
Comment by step back — January 17, 2008 @ 6:10 pm
If “not proved” is not a reason, and (according to your earlier comment) the statement of a reason is what prevents an agency’s action from being arbitrary, are all jury verdicts arbitrary? Are all judicial pre-verdict grants of JMOL arbitrary? I think “you failed to prove what you’re obliged to prove, and here’s how it fell short” is plenty of reason, and not circular.
Comment by Joe — January 17, 2008 @ 6:20 pm
Joe, you write: “are all jury verdicts arbitrary?”
Whoa. Hold up there my good professor. The top of your post starts with “So … here’s what I wrote in Nov 2006: You file a patent application **at the Patent & Trademark Office.** … Put another way, who gets the benefit of the doubt on patentability or not … you, or **the PTO?**”
So I thought we were restricting our discussion to things that happen only inside **the PTO.** There are no juries inside the PTO. There are no JMOL’s inside the PTO. How did we leap from one forum into an entirely different one?
Also, now you’re proposing a case where the PTO says: “You failed to prove what you’re obliged to prove, and here’s how it fell short ….†where presumably the articulation of a reason for rejection of the claim does follow. That’s a horse of an entirely different color from the curt statement: “You failed to prove your case. (Period)” The latter is not a “reason”. It is merely a conclusory statement. The former ostensibly does provide a reason as is required by 35 USC 132. Call it what you want, the reason that the PTO must provide puts the burden first on the PTO.
So essentially, you have taken a circular route to come around to agreeing with my initial assertion, namely, that 35 USC 132 forces the PTO to give a “reason” for a 103 rejection and that is why the burden is first on the PTO’s side.
Comment by step back — January 17, 2008 @ 9:40 pm
Step,
I didn’t suggest what you say I suggest. I didn’t say “You failed to prove your case. (Period).” Look at my next to last comment before this one: I referred to “a properly explained ‘not proved.’” I didn’t suggest an unexplained “not proved.” We agree that wouldn’t be a reason.
As for juries, I realize they aren’t used at the PTO! What I’m suggesting is that your argument about what does, and doesn’t constitute arbitrariness seems tailor-made to get to the outcome you like in the patent context, but hard to apply to similar contexts.
Let’s really step back, Step Back, and look at the bigger picture here: We agree the PTO has to give reasons for its action. We also appear to agree that an explained “not proved” would constitute a proper reason. Can the Patent Office also require, consistent with the properly interpreted text of section 103, that an applicant provide an affirmative demonstration of nonobviousness as part of the initial application? That’s the question that, for me, is not an easy one to answer. (And, just to repeat, I know what the traditional, settled answer is.)
Finally, I don’t agree that the burden is first on the PTO’s side because the PTO must give a reason for its action. The burden is first on the PTO’s side because of tradition (namely, the In re Warner case (1967)). The contrast between sections 102 and 103 makes that tradition at least open to question.
Best,
Joe
Comment by Joe — January 17, 2008 @ 10:14 pm
Joe,
There are two reasons I think the burden is on the PTO; one textual and one policy.
1) There’s no need for Congress to say “and the applicant must prove nonobviousness” in 103 to distinguish 102. 103 is best seen as a continuation of 102. For example, you could easily replace the first seven words of 103 with “Even though” and have the same meaning, so I’m not convinced the lead-in makes that much of a difference in the absence of a more-explicit command.
2) How does the applicant prove the invention is nonobvious in the absence of a PTO action? The universe of patents is, of course, not limitless. However, the burden on the patent applicant would be so great in trying to prove nonobvious ab initio that Congress couldn’t possibly promote the progress of the useful arts by placing that burden on the applicant. Even with my best search, I’m going to miss some relevant art and potentially fail in my quest to prove my invention nonobvious; or worse, erroneously prove my invention nonobvious. The best way to promote this process, and improve its accuracy, is to have the applicant provide as much known art as possible (the IDS) and have the PTO perform its own search, supplementing the art and applying it. Those burdens are properly based on the PTO, and not the applicant, because of resource and knowledge differences.
Comment by Bryan — January 18, 2008 @ 9:02 am
Bryan,
1) I think your textual point is a reasonable one. Of course, the opposing one is, too. So, one must make a judgment.
2) I’m not sure I’m persuaded by the policy point, because I think it may prove too much. It’s unusual to ask a party to prove a negative, I know, but it’s not unheard of. More than that, why would it be so hard, for example, for an applicant to include in the application an explanation of how the claimed invention solves a long-felt, unmet need in the art? Or does something from which the art taught away? Or, at the very least, accomplishes something unexpected, given the teachings of the prior art? These are powerful forms of nonobviousness evidence, and the applicant will have some insight into them, no? They are certainly things for which the examiner will be looking, as the new guidelines instruct.
3) You say “Congress couldn’t possibly promote the progress of the useful arts by placing the burden on the applicant” in this instance. Do you mean to say that it would be unconstitutional to interpret section 103 differently from the way it’s been interpreted on this burden of proof point? That’s an awfully big gun to pull out - “any other reading is unconstitutional” - and it certainly got Justice Douglas into hot water with the patent prosecutor types in Cuno Eng’g and Great A&P Tea (on the basic nonobviousness standard, then a common law matter). You may recall that a serious issue in Graham was whether Congress even had the constitutional authority to pass section 103. The Court’s conclusion in Graham was that Congress had extremely wide latitude to implement the details of the broad constitutional policy.
Best,
Joe
Comment by Joe — January 18, 2008 @ 9:19 am
Bryan,
A few more thoughts about your second argument …
The argument seems to be, “the applicant could not achieve this task with perfection, so we can’t ask the applicant to do it.” I hope this isn’t the standard we’re going to use because, of course, the Patent Office can’t achieve the task perfectly either. A key question, then, is who will suffer the downside of the fact that things aren’t (and won’t be) perfect. As between the applicant, who wants to obtain a long-term, nationally-binding, competition-precluding regulation (otherwise known as a patent claim), and the general public (whose interest the Patent Office should protect), shouldn’t we cast the risks from imperfection on the applicant? Competition is the norm in our economy, after all, not state-conferred bans on competition. (The Supreme Court has said this repeatedly in i.p. cases.)
Best,
Joe
Comment by Joe — January 18, 2008 @ 9:26 am
Joe,
I have only a moment, so I’ll get back to this later, but I wanted to point a couple things out:
I didn’t mean to make a Constitutional argument, so I’ll explain a little more clearly.
As to the policy point, yes you could put the burdens of those statements on the applicant, but that would be awfully unfair given the strong effect of prosecution estoppel in later proceedings. You might say something inadvertent that could really hurt your client down the road; it might even be inaccurate in light of later-discovered art, but it’ll still kill you.
Comment by Bryan — January 18, 2008 @ 11:46 am
Bryan,
The prosecution history estoppel point is, I agree, quite a serious one. And you may be right about the way to resolve it. I would only point out that there’s also an unfairness, if we spare the applicant having to make these statements - namely, an increased risk of an erroneously granted patent. That’s a great unfairness to the public, including other patent owners. Which unfairness is more important to prevent - a narrower scope for a more accurately granted patent, or a more likely error in granting the patent at all?
Thanks,
Joe
Comment by Joe — January 18, 2008 @ 12:13 pm
The other reason I, as a practitioner, don’t want to go making unnecessary arguments is the very real possibility of an inequitable conduct finding on the part of the courts or, as is becoming more and more easy, a disciplinary investigation by the PTO. At this point, inequitable conduct defenses are raised constantly, so anything you say in prosecution can and will be used against you in a court of law. With the PTO’s disciplinary proceedings, the practitioner doesn’t want to be in the position of having previously characterized an invention’s nonobviousness without the benefits of the PTO’s search and rejections, for fear of having later to prove that it was not a knowing misrepresentation of a material fact.
Anyway, so those all go to the policy aspects. There are all kinds of problems with placing the initial burden on the applicant to prove nonobviousness, but that’s a whole paper rather than a blog comment.
As to the textual argument, that would require some more thought. Perhaps a consideration is necessary of whether Congress really needed to affirmatively state the burden of proof as being on the PTO if they were simply codifying what was a preexisting judicial rule.
Finally, as to my “Constitutional” argument, what I mean to say is that Congress couldn’t “promote the progress” in the sense that such a heavy burden on the applicant would retard applications and have (perhaps; a law and economics question) the paradoxical effect of making prospective applicants rely on trade secret protection for their inventions and never disclosing them to the public for fear of the financial cost of doing so.
I believe this also goes to your last comment. If the applicant can’t do a perfect job, then the applicant could, based on many fears glossed over above, keep the inventive discovery to him or herself, with a net loss to society. It’s certainly appropriate to cast the risks of imperfection on the applicant, but if we ask who has the greatest ability to avoid imperfection, then I’ve got to suggest that it’s the PTO; not every inventor is Intel.
Comment by Bryan — January 18, 2008 @ 1:14 pm
Bryan,
All excellent points!
Thank you for participating; the comments for this post have been terrific!!
Best,
Joe
Comment by Joe — January 18, 2008 @ 1:31 pm
Joe,
Your policy point about the imperfection and the competition-precluding patent claim would suggest allocating the burden on 102 to applicants as well.
I’m not sure about the relative merits of the concern over inequitable conduct and prosecution history estoppel. Litigators are careful with their representations to court because they will get slapped hard (and personally, unlike patent prosecutors) if they make a material misrepresentation with intent to deceive. Similarly, judicial estoppel operates very much like prosecution history estoppel–you take a position to win case 1, and in case 2 the statement comes back to bite you.
Now, the difference may well be that in patent prosecution the comes-back-to-bite-you aspect occurs much more frequently. Or the difference may be that inequitable conduct has a lower standard in practice from that required to incur judicial sanctions. But to say that applicants should be permitted to say as little as possible to maintain their strategic flexibility later on–instead of stepping up and recording contemporaneously what they invented–seems strange to me. I can surely see why applicants like it. I don’t see why society does.
Comment by TJ — January 18, 2008 @ 3:28 pm
Looking in on this interesting thread from primitive feudal civil law continental Europe, one can see how the English common law system, with its 700 years of increasing sophistication, is a world away from planet Earth’s default legal system, civil law. Consider for example what possibilities an Applicant at the EPO has to appeal, when the EPO Examiners (all properly educated in the technical field) reason out(relying on their technical education)that the claim pending embraces within its scope something that was already obvious to the person skilled in the art at the filing date of the claim. Applicant appeals. To the Technical Board of Appeal (three members, two with a degree in a relevant science or engineering subject). Is the PTO a party to that appeal? No, it’s ex Parte. At the Appeal Board Hearing, there’s nobody thereto put the Exr viewpoint to the Board. If the Board agrees with the Examiners’ written Decision, the subject of Applicant’s appeal, and rejects the appeal, where does Applicant go next? Answer: nowhere. That was the end of the line. Application dead (hopefully you have filed a divisional). Due process? You tell me. Adversarial relationship between prosecutors and Exrs? Well no, not really, because the Examiner isn’t there at the Board of Appeal Hearing, disputing everything Applicant puts forward. What usually happens, of course, is that Applicant gets his brain properly into gear for the first time when writing his appeal brief, and finds a claim that commends itself to the Board, which then sends the case back down to the Examiners with an order to get on and issue the patent. It’s certainly not the US way, right Joe?. Questions of “Who’s got the burden of proof” get very fuzzy. BUT, it works well. Is this EPO approach to the burden of proof for issues of obviousness applicable to common law-burdened USA? Obviously not. What chance of harmonising the patent laws of the world? About zero, I would say.
Comment by MaxDrei — January 19, 2008 @ 9:27 am