Obviousness, hindsight, KSR

Training the human neural network: by Novasdid (CC BY-SA 4.0)

The 2007 KSR v. Teleflex Supreme Court (SCOTUS) decision is why the patent examiner, ignoring hindsight issues, just used your own teaching against you to reject your patent application claims as being “unpatentable” (obvious) under 35 USC 103.  Unfortunately, the US patent “obviousness” rules and regulations still have some “bugs”. 

The patent examiner has just reviewed your patent application, and has sent you a response. What are all these “rejected under 35 USC 103 as being unpatentable over (various citations)” statements? It almost looks like the examiner just copied your claim, interspersed it with various citations matching some of the claim words, and concluded with “therefore it would have been obvious to one of ordinary skill in the art…

This might even look to you like a standard formula that could be used to reject almost anything. Why does the USPTO work this way?

Some background: Without obviousness rejections, your patent could soon be swamped by many other competitor patents that claim the smallest, most trivial changes to your work. To keep the patent system healthy, there needs to be some “shielding”, some sort of “force field” that keeps competitors from getting too close to your work. In the US, the depth of the “shielding” or “force field” is set by trying to legally determine, often years later, what a person having ordinary skill in the art (PHOSITA) would think was obvious.

The big problem is “hindsight bias”. Lots of non-obvious things look obvious in hindsight. Here the legal system is attempting to cope, with varying success, with a very complex underlying problem of pattern recognition. Once you see the solution to a puzzle, it is hard to see anything else.

Prior to 2007, the USPTO used anti-hindsight rules in an attempt to minimize hindsight problems. However, in the 2007 KSR v. Teleflex case, SCOTUS made what, in my opinion, was a key error. Dictionaries define hindsight somewhat incompletely as: “understanding of a situation or event only after it has happened or developed”. SCOTUS ran this incomplete dictionary definition into the ground. They argued that the earlier anti-hindsight rules were too “rigid”, and that “common sense” should be used. They held that hindsight could be avoided by just considering if the invention would be obvious “at the time of the invention”.

This is an almost meaningless statement.  Who would file a patent application if it could be invalidated by later filed patent applications?

In reality, the patent applicant has just shown the examiner the solution to a puzzle, thus “training the examiner’s neural net” to subsequently view this solution as “obvious”.  However, the examiner is told to examine with 100% hindsight bias.  The examiner can also dismiss “hindsight” rebuttals by merely stating that under the newer, post-KSR, USPTO rules (MPEP 2141.01 III): “Content of the prior art is determined at the time the invention was made to avoid hindsight.”

In other words, thanks to the KSR ruling, the present USPTO rules can be paraphrased as Don’t bother us about “hindsight”, we’re not listening, and SCOTUS says that we don’t have to!

Fortunately, there are other ways to rebut obviousness rejections. Examiners often misquote the citations, have gaps in their reasoning, and their proposed combination is often a Frankenstein monster that differs significantly from the claim.  Secondary considerations can also be raised. So things can be done, but this hindsight “bug” (or feature) in US obviousness patent law is annoying.

Cleaning up obsolete case law: time to revisit In re Gorman


35 USC 103 rejections: Ever have something rejected as “obvious” in view of a combination of 5+ references?  The re Gorman case is how such absurd rejections are rationalized.

According to MPEP 707 section 7.37.07, arguments that the examiner used an excessive number of citations are presently found to be “unpersuasive”.  Section 7.37.07 is based upon an old case In re Gorman, 933 F.2d 982, 18 USPQ2d 1885 (Fed. Cir. 1991).  However Gorman uses reasoning that is now obsolete and unsupported due to patent law developments since 1991.

Some key aspects of the re Gorman decision were:

When it is necessary to select elements of various teachings in order to form the claimed invention, we ascertain whether there is any suggestion or motivation in the prior art to make the selection made by the applicant. Interconnect Planning Corp. v. Feil, 774 F.2d 1132, 1143, 227 USPQ 543, 551 (Fed.Cir.1985). ” ‘Obviousness can not be established by combining the teachings of the prior art to produce the claimed invention, absent some teaching, suggestion or incentive supporting the combination.’ ” In re Bond, 910 F.2d 831, 834, 15 USPQ2d 1566, 1568 (Fed.Cir.1990) (quoting Carella v. Starlight Archery and Pro Line Co., 804 F.2d 135, 140, 231 USPQ 644, 647 (Fed.Cir.1986)).

The extent to which such suggestion must be explicit in, or may be fairly inferred from, the references, is decided on the facts of each case, in light of the prior art and its relationship to the applicant’s invention. As in all determinations under 35 U.S.C. Sec. 103, the decisionmaker must bring judgment to bear. It is impermissible, however, simply to engage in a hindsight reconstruction of the claimed invention, using the applicant’s structure as a template and selecting elements from references to fill the gaps. Interconnect Planning, 774 F.2d at 1143, 227 USPQ at 551. The references themselves must provide some teaching whereby the applicant’s combination would have been obvious

However since 2007 KSR Int’l Co. v. Teleflex, Inc., 550 U.S. 398, the “teaching, suggestion or incentive” requirement for combining citations is no longer required (although still allowed).  When this part of the 1991 Gorman court’s reasoning, shown in the first paragraph above, is removed, it becomes clear that the legal and logical underpinning of their 1991 decision no longer apply!

However under present rules, MPEP 707 section 7.37.07, based upon the now obsolete 1991 Gorman legal reasoning, still acts as a roadblock for any applicant who attempts to argue that the examiner did:  “engage in a hindsight reconstruction of the claimed invention, using the applicant’s structure as a template and selecting elements from references to fill the gaps.”

Given that the Gorman court’s assumptions no longer hold, isn’t it time to revisit this case?  At present, there is no upper limit to the number of citations that an examiner can apply, and indeed I personally have seen combinations of eight citations used in obviousness rejections.