« Watchdog | Main | Sieg Heil! »

July 13, 2008

Not Invented Here

35 U.S.C. §102 - "A person shall be entitled to a patent unless - ... (f) he did not himself invent the subject matter sought to be patented..." §102(f) dates to the 1836 Patent Act, and U.S. case law predates statute. Does §102(f) mean what it says? - That a patent is invalid if not invented by a patent applicant. If so, why is it not applied as such? If not, why not?

Section 15 of the 1836 Patent Act provided for defenses against patent infringement, including what became § 102(f): "that the patentee was not the original and first inventor or discoverer of the thing patented, or of a substantial and material part thereof claimed as new." Clearly, the intent of the clause was that a patent grant in the U.S. required that the patent holder must be the first inventor.

It is not unusual for a U.S. patent to be anticipated by another, foreign patent, but that the foreign filing date fails to qualify under other §102 clauses, except possibly (f). §102 (a) is positively nationalistic: "(a) the invention was known or used by others in this country," as if foreigners don't count for invention, unless published first.

If a foreign filer does file for a U.S. patent, and invokes an interference under §102 (g), for first to invent, then its unoriginal contender application would be extinguished. But, if not challenged during prosecution, the unoriginal issues as a patent, possibly with validity assured, save invocation of §102(f) as originally intended.

There's the rub. Nowadays, no one invokes §102(f) in the context of invalidation. One patent text even proclaims that "subsection 102(f) is neither an anticipation nor a loss of right section."

§102(f) most commonly arises with regard to inventorship, particularly derivation and inclusion. Derivation is that another person than the patent applicant conveyed the invention to the applicant, who took it and ran. The result: loss of right.

Inclusion refers to putting in non-inventors, such as supervisors, and leaving out contributory inventors. §256, incorporated into the 1952 Act, lessened the risk of invalidation for inventor misnaming, by providing for error correction without penalty.

§116, amended in 1984, notes that joint inventorship is appropriate even though an individual may have contributed to only some claims.

U.S. law is peculiar in that it requires that applications are filed in the inventor's name, however its assignment to another, most often a corporation. Many foreign countries permit corporations to file patents.

§102(f) does require that applicants claiming foreign priority under §119 must have the same inventor roster for both the foreign and U.S. application.

Moy's Walker on Patents indicates that §102(f) should serve for invalidation, or, at least, that was the original intent, and long held:

The originality requirement in United States patent law is both stable and extremely long standing.

[T]o be patentable, the invention must be original to the patent applicant. 3

3 See, e.g., Reutgen v. Kanowrs, 20 F. Cas. 555, 556, No. 11710 (C.C.D. Pa. 1804) ("[I]f it appears that the plaintiff was not the original inventor . . . he is not entitled to a patent."); Dawson v. Follen, 7 F. Cas. 216, No. 3670 (C.C.D. Pa. 1808) ("[T]o entitle the plaintiff to recover, they must be satisfied that he was the original inventor...."); Gayler v. Wilder, 51 U;S. 477, 494, 10 How. 477, 13 L. Ed. 504 (1850) ("[The patent act] provides that, if it appears on the trial of an action brought for the infringement of a patent that the patentee 'was not the original and first inventor or discoverer of the thing patented,' the verdict shall be for the defendant."); Odiorne v. Winkley, 18 F. Cas. 581, 582, No. 10432 (C.C.D. Mass. 1814) (StorY J. "The original inventor of a machine is exclusively entitled to a patent for it."); Thomas v. Weeks, 23 F. Cas. 978, 981, No. 13914 (C.C.S.D. N.Y. 1827) ("[T]he patentee can sustain his patent only on the ground of his being the original inventor, ..."); Dawson v. Follen, 7 F. Cas. 216, No. 3670 (C.C.D. Pa. 1808) ("[T]o entitle the plaintiff to recover, [the jury] must be satisfied that he was the original inventor....").

For cases distinguishing "first" from "original," see, e.g., Lowell v. Lewis, 15 F. Cas. 1018, 1021, No. 8568 (C.C.D. Mass. 1817) ("Mr. Perkins, being clearly the first inventor, is entitled exclusively to the patent right, although Mr. Baker may have been also an original inventor; for the law gives the right, as among inventors, to him, who is first in time."); Hayden v. Suffolk Mfg Co, 11 F. Cas. 900, 901, No. 6261 W.C.D. Mass. 1862), aff'd, 70 U.S. 315, 18 L. Ed. 76 (1865) ("[A]lthough [the defendant] may have been an inventor, an original inventor, yet, if he was not the first inventor, [the invention] would not give him any exclusive privilege.").

See also Willard Phillips, The Law of Patents For Inventions, 65 (1837) (''To entitle a party to a patent he must not only be an inventor, but also the original inventor...."); 1 Robinson, The Law of Patents, § 78 (1890) ("An invention is the product of original thought.").

It is long past time to resurrect §102(f) for its original intent, in respect to a patent grant representing sanctity of invention, to insist that a United States patent must issue from the original inventor, a "product of original thought."

Posted by Patent Hawk at July 13, 2008 2:25 PM | Prior Art


I'm a little confused. I thought that what 102(f) says is that you can't have a patent if you didn't actually conceive it yourself. It doesn't say that the first to invent shouldn't get the patent. Remember the default in 102 is that you get a patent ("A person shall be entitled to a patent UNLESS...") 102(f) is not an anticipation or loss of rights provision because those concepts apply to inventions, which by definition you haven't made if you're deriving your work from an inventor.

So rather than reinterpreting 102(f), I think you're advocating that we go to a first to invent worldwide rule -- something that would require a more substantial revision to the statute.

But I think I agree with what I understand to be your premise -- that ownership rules should follow attribution rules. The inventor who was first to solve the problem should have priority in ownership disputes. Is there a principled reason for insisting that the inventor be in the U.S. anymore? If you say yes, then why should China be able to say yes too.

But since we're worrying about fairness to inventors we should recognize that many first inventors are ignored by society. Is it fair to deny a later inventor who solves the same problem an exclusive right when she is able to do what the first inventor was not -- namely, to convince an industry that her solution would actually work? Before you answer, remember that we've known the Dvorak keyboard is better than the QWERTY keyboard for over a hundred years, but I'm still typing on a QWERTY keyboard as I write this.

Posted by: Michael F. Martin at July 13, 2008 6:29 PM

Hi Michael:

To restate the obvious (my first sentence):

35 U.S.C. §102 - "A person shall be entitled to a patent unless - ... (f) he did not himself invent the subject matter sought to be patented..."

What does that say?

Posted by: Patent Hawk at July 13, 2008 10:46 PM

"we've known the Dvorak keyboard is better than the QWERTY keyboard for over a hundred years"

Why is that?

Posted by: JD at July 14, 2008 6:28 AM

@Patent Hawk

Are you saying that you cannot be an inventor unless you were the first to invent? If so, then I think (again) that this would require a radical reinterpretation of the entire statute, not just 102(f).


Why is what? Are you asking why the Dvorak keyboard is better or why we still have the QWERTY keyboard?

I'll assume the latter since you can google for the former.

The conventional answer is "network effects," which overtime increase the "switching costs" to a scale beyond which it is infeasible to adopt a new standard without causing major disruptions to the market. I think I agree with this answer, but it usually does not make explicit the key variable in the switching costs, which is time. The costs of actually retooling the equipment that makes the keyboards are relatively low compared to the social costs of a dramatically less efficient keyboard. So the retooling is not what's preventing the adoption of the better standard -- it's the inconvenience and disruption to the flow of typing for billions of typists, each of whom would be forced to relearn typing, that's the issue.

...which in turn suggests that unstable equilibriums could be pushed into more stable equilibriums by minimizing the amount of TIME that it takes to make the transition. Most discussions tend to emphasize the cost; the real issue is the time. You can say time is money but that just doesn't do justice to either.

Posted by: Michael F. Martin at July 14, 2008 8:57 AM

I guess if you build a better keyboard, the world will not beat a path to your door.

Posted by: JD at July 14, 2008 9:40 AM

Well, very fascinating, but could it be that the intent of the US legislator all along is that no American should be able to profit by taking the invention belonging to another American, and securing an exclusive American patent right from it? European law holds, in Art 138 EPC, that if the patent issues to a party other than the rightful owner, it is invalid. Is the US legislator's intent any more complex than that?

Posted by: MaxDrei at July 14, 2008 9:41 AM