Statutory Bar Prior Art in the Nonobviousness Analysis

I am continually troubled with disconnects between the text of 35 U.S.C. § 103(a) and its application by the courts. 

Section 103(a) defines the doctrine of nonobviousness and reads (in part) as follows:

(a) A patent may not be obtained . . . if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. . . .

A plain reading of the statute that considers the obviousness of an invention “at the time the invention was made.”  In its 1965 Foster decision, the Court of Customs and Patent Appeals (CCPA) made clear that “the existence of unobviousness under that section, as a necessary prerequisite to patentability, we reiterate, must be determined as of 'the time the invention was made' without utilizing after-acquired knowledge.”  343 F.2d 980 (CCPA 1965).

Despite that seeming clarity, in its obviousness analyses, the Foster Court (and subsequently, the CAFC) has given full 103(a) consideration to post-invention activities that qualify as prior art under Section 102(b).  The justification for this interpretation appears to come in the form of an estoppel that blocks inventors who “sleep on [their] rights more than a year after the invention has become entirely obvious to the public.”  Foster indicated that any other interpretation of the statute would be an interpretation that “Congress could not possibly have intended in view of its express indication that section 102(b) is merely a continuation of the prior law.” 

The leading opinion on-point is likely In re Corcoran, 640 F.2d 1331 (CCPA 1981) which was written by Judge Rich.  In that decision, the court held that the application of 102(b) prior art in 103(a) rejections “finds its statutory base in both of those sections.”  See also Lough v. Brunswick Corp., 86 F.3d 1113, 1121 n.5 (Fed. Cir. 1996); Baker Oil Tools v. Geo Vann, Inc., 828 F.2d 1558, 156 (Fed. Cir. 1987) (what was offered for sale before the critical date becomes a "reference under section 103 against the claimed invention").

Although the weight of precedent is on the side of the current rule, the current statutory-language-focused Supreme Court just might have a point of view.  Let me know if you get a test-case going.

See also, Rethinking the Scope of Prior Art in Obviousness Cases.

44 thoughts on “Statutory Bar Prior Art in the Nonobviousness Analysis

  1. 44

    I know what background knowledge the English patents court imputes to its notional PHOSITA. I know what background knowledge the German Federal Patents Court imputes to its notional PHOSITA. I do not know how the USA goes about settling what background knowledge to place in the brain of the PHOSITA, the PHOSITA that is, who is (notionally) going to study the references, and then decide whether any given claim includes within its ambit something that is obvious.

    Is there a case I could read, that would enlighten me?

  2. 42

    We just need to be honest with ourselves that the “person skilled in the art” is something more than a technician with all the basic knowledge and skills of his art, with all the publications of his art as at the relevant date papering his walls.

    “more than…”

    What exactly are ya thinking the criteria to be, my man? all honest to yourself and all?

    Honest to ya, this smacks of an elitist “inventive spark” type O mindset reminiscnet of Justice Douglas that was suppose dto have been excised inthe ’52 statutory changes.

    Just an observation, honest to ourselves and all.

  3. 41

    Is everything that is known to those of skill in an art found in some reference?

    Pretty much everything probably, especially in arts that have textbooks. Some of it might be so basic that you don’t need to cite a reference, or there are dozens of references that will do the trick. But that’s not the point. Sometimes you need a particular reference to teach a particular feature, e.g. if that feature was just invented and it’s only “disclosed” in the one unpublished patent application. It’s a funny sort of public domain that includes as obvious a combination of two references that, while technically both prior art, were not known to a person skilled in the art and therefore could not have taught said person to modify one in view of the other.

    I don’t mind if that’s the rule. We just need to be honest with ourselves that the “person skilled in the art” is something more than a technician with all the basic knowledge and skills of his art, with all the publications of his art as at the relevant date papering his walls.

  4. 36

    Ned sed: “Babel Boy, Congress wrote that the date of invention is the date the prior art is to be assessed for determining obviousness. And you think that is unclear?”

    I don’t think your statement of what you think Congress said is unclear, but I know that what Congress actually said is unclear. Am I clear?

    103, if that’s what you’re referring to, doesn’t say anything about what date the prior art is to be assessed. It says that on the date of invention the subject matter taken as a whole would be obvious.

    That statute is a complete pile of rubbish. It makes less sense when you read it than the Second Amendment.

    At any rate, Mr. Boundy, provided a more substantive response to your point. It’s grammatical.

    Have the courts messed up the 103 mess even more? Ned, I would never disagree with you on that point. You know the law far better than I ever hope to.

    IANAE: “Just to play devil’s advocate here, how can the invention as a whole have been obvious at a time when the 102 prior art was not known to persons of skill in the art?”

    Is everything that is known to those of skill in an art found in some reference? This is a rhetorical question.

    Isn’t that what all the guffaw was about regarding the standing swing method and trimming the crust off the sandwich? Those patents were issued because the examiners couldn’t cite information that is clearly in the public domain but not written down. (I guess they could have cited their own knowledge and given an affidavit, but I have never seen an examiner do that even when demanded.)

    I’m going to patent Mom’s five second rule: “If it’s been on the floor for more than 5 seconds, don’t eat it.”

    But now I can’t because 6 will cite this post against me.

  5. 35

    “ones”, ping? Plural? How can that be, when there is in your world only one “good” jurisdiction.

  6. 33

    Oh, ping, I should have added that any “European” benchmark should be expressed in metric units. The British on their island still use what they call “Imperial” units. They like to be different. Quaint, no?

    I said “benchmark” because it is the European patent law model that Asia has now adopted. Remind me, which jurisdictions are those that have found US patent law something worth copying.

  7. 32

    The benchmark is the European way.

    sure, um, yeah, let’s go with that (wink wink).

    Weza know who wields the gold standard, don’t we Maxie?

    …because I do not myself understand it.

    Never stopped ya before.

  8. 31

    To respond playfully, IANAE, my answer is that the PHOSITA is a mythical being, so he can know, or not know, whatever you want to fix him with.

    But respect for the patent system depends on having a law of novelty, and a law of obviousness, that is objective but, nevertheless, makes some kind of sense to people (like engineers and CEO’s) who are not patent professionals.

    The benchmark is the European way. It goes like this:

    Rule Number 1: You can’t have a patent if your claim embraces within its ambit something that, at the filing date of the claim, was already obvious to a PHOSITA who has at his disposal everything that was public knowledge one day before that filing date.

    Rule Number 2: You can’t have a patent for an invention which is not novel over the enabling disclosure content of earlier filed, later published patent applications at that same Patent Office (it’s “First to File an Enabling Disclosure gets the Patent”, you see).

    Which part of that do readers not understand?

    I can’t explain US patent law on novelty and obviousness to engineers and CEO’s, because I do not myself understand it.

  9. 30

    “The prior art” is defined in § 102, and the ordinary grammatical parse of § 103 doesn’t change it.

    Just to play devil’s advocate here, how can the invention as a whole have been obvious at a time when the 102 prior art was not known to persons of skill in the art?

    I guess section 103 describes a person of ordinary skill at time of the invention but also knowing all of the 102 art. But then, what of combining two 102 references that were not actually known in the art at the time of invention? Surely it could not have been within such person’s ordinary skill to modify one thing he’d never heard of in view of another thing he’d never heard of.

  10. 29

    The current rule comes out of the sentence disagramming you studied in eighth grade.

    “at the time the invention was made” modifies “would have been obvious,” the legal inquiry.

    “at the time the invention was made” does not modify “the prior art.”

    “The prior art” is defined in § 102, and the ordinary grammatical parse of § 103 doesn’t change it.

  11. 28

    John Marshall’s post brings to mind the debate between Scalia and Dworkin in __A Matter of Interpretation__ regarding the proper role of determining the “intent” of a statute vs. construing the “meaning” of its words or phrases, among other things, from its context. John Marshall, akin to Dworkin, presumes to know what is “clearly not intented” by the statute and then implicitly includes the presumed known intent in the “context” from which to determine the meaning of its terms. Scalia eschews this approach. Since Dennis is trying to determine how the current statutory-language-focused Supreme Court just might construe the statute, there are currently four Dworkin-minded justices and four Scalia-minded justices on the court; so, it boils down to which side can succesfully convince Justice Kennedy to think their way.

  12. 27

    Babel Boy, Congress wrote that the date of invention is the date the prior art is to be assessed for determining obviousness. And you think that is unclear?

    Obviously, it was the courts, and not Congress, that went off the tracks in stating that obviousness could be determined against prior art at a date different than the date specified in the statute.

    When Congress wants to make an exception to a general rule, it is known to do so. See for example, §363 which states that international applications have the same effect as a US application except as provided in §102(e).

    Again, what we are talking about here is obviousness. If an invention is fully disclosed in a reference that is a statutory bar, one could not swear behind that kind of reference. What we are talking about is where the §102(b) reference does not describe the invention, but something else which is alleged to make the invention obvious.

    The only “problem” we would have if one could swear behind a §102(b) reference in the context of a 103 rejection is that there would be a few more patents that arguably would be obvious over subject matter that went into the public domain after the date of invention but more than a year prior to the filing date.

    Again, these statues have ancient heritage and go back to the original Supreme Court cases. Unless one understands those cases, one would not understand what Congress meant in §103. The one-year period of §102(b) is a grace period. It was originally 2 years and then changed to one. It was a grace period placed into law by Congress to allow the inventor time to fully develop his invention before he had to file a patent application – because otherwise under the doctrine of Pennock v. Dialogue, any public disclosure would result in abandonment of his invention. But the public disclosure itself by the inventor, or the public disclosure of some other inventor, is not “prior” art to the inventor. This is easy to understand. And it is easy to understand what Congress intended in §103.

  13. 26

    Apologies if this is inarticulate, but one shouldn’t really be troubled by a purposive approach to statutory interpretation. It’s not new, and appears to be in vogue in the present legal era.

    That not all elements must be present for a 103a reference is based on the even earlier finding that not all elements must be present for even a 102b reference

    There is a case earlier than Foster that articulates the issue—Dix-Seal Corp (D.C. Conn. 1964) that is cited in, among other cases, Tri-Wall Containers (Court of Claims, 1969), for the premise that “It should be noted that Section 102(b) is literally applicable only if the public use or sale is of the invention on which a patent application is later filed. However, complete identity of (1) a product sold or publicly used and (2) a product as claimed in an application filed more than 1 year after the sale or public use, is not a prerequisite for invalidation of the patent on the basis of the prior sale or use. Connecticut Valley Enterprises, Inc. v. United States, 348 F.2d 949, 953, 172 Ct.Cl. 468, 474 (1965); Dix-Seal Corp. v. New Haven Trap Rock Co., 236 F.Supp. 914, 917-920 (D.C.Conn.1964).” The Dix-Seal court found that complete, or “literal” identity was not required under 102b, as that would make the section a “paper defense”.

    The logic was extended to 103a in Foster: ““First, as to principle, . . . we cannot see that it makes any difference how [the public] came into … possession [of the information], whether by a public use, a sale, a single patent or publication, or by combinations of one or more of the foregoing. In considering this principle we assume, of course, that by these means the invention has become obvious to that segment of the “public” having ordinary skill in the art.. . . .” (Foster, CCPA, 1965).

    I wouldn’t hold my breath for a “test case” aimed specifically at this issue. The crux as I see it is whether or not the public was in possession of the invention, and therefore disclosure sufficient to fill in the gaps in a PHOSITA’s knowledge will suffice. “Patented”, “described in a printed publication”, “in public use”, and “on sale”, are all ways of saying that the public was in possession, but that is not an exhaustive list thereof.

    Going further still, determining whether or not the public was in possession is just a way to get at a deeper issue: whether the inventor had exclusivity of possession to an extent sufficient to warrant continuation thereof via the grant of patent rights. A sufficient break in the exclusivity of possession, in the absence of evidence to the contrary, is interpreted as evidence of the inventor’s lack of desire to maintain said exclusivity. Since filing for patent rights is an entirely volitional activity, it is evidence of the inventor’s desire at that time to maintain, or to REGAIN, said exclusivity. The most basic issue is, of course, “should we grant patent rights to this inventor?”, and this is the mechanism whereby we answer this basic question.

    There are many public policy and economic reasons for not allowing an inventor to regain exclusivity of possession only when they have realized that conditions are favorable for them to do so. The calculus amounts to an allocation of risk to the inventor, of filing an application when the economic future of future patent rights is uncertain, or less certain than it would be at some future time when the value of such rights is more knowable.

    Given a relatively low cost of filing an application, I think that allocation of risk is reasonable. As the filing and prosecution costs rise relative to potential patent value, what will happen over time will be that those inventions perceived to be less valuable, or to have less potential value, will not be patented. Some of those may see the light of day, some will never be disclosed to anyone, and will effectively disappear until such time as somebody else may invent the same thing.

    This is an argument for keeping filing and prosecution costs low, and to me is an argument favoring a registration system over the current examination system.

  14. 25

    Doogle: “One way to read Section 103 is noting it’s the difference between the prior art of record and the subject matter sought to be patented that you consider at the time of invention.”

    I like Doogle’s analysis, but it seems to me to be off by a hair.

    What is relevant as of the date of invention, according to 103(a), is not the differences but the subject matter as a whole viewed in light of those differences.

    The differences, and, hence, the prior art that gives rise to those differences being publicly articulated, can come at any time.

    The question then becomes: Is it logically possible that contemplation of a reference could render obvious subject matter invented earlier than the reference’s publication date?

    I can’t see why not. Whether subject matter is obvious does not depend on whether or not there is a publication in existence at the time the invention was made. It’s just that from a 103 point of view that obviousness is inchoate until a reference appears.

    What about after a patent grant? Is it possible for someone to publish a reference after the patent issues that would render the subject matter obvious at the time it was invented?

    Let’s face it folks, the 82nd Congress of 1952 was a bunch of semantic bone-heads. Any of us could write a cleaner and clearer Patent Act at a wee fraction of what the public paid Congress for the present one, even in 1952 dollars.

    What we need is a wiki to do just that.

  15. 24

    Patentability shall not be negatived by the manner in which the invention was made.

    Sadly, though, patentability is frequently negatived by the time at which the invention was made.

  16. 22

    Worth clicking on the link from Dennis back to October 2008, if only out of nostalgia for all those pseudonyms that we no longer see on this blog.

  17. 21

    I am continually troubled with disconnects between the text of 35 U.S.C. § 103(a) and its application by the courts.

    Hello Mr. Elephant, welcome to the room:

    Patentability shall not be negatived by the manner in which the invention was made.

  18. 19

    Then that wouldn’t be a statutory bar, now would it?

    It would still be statutory (because it’s in the statute), but it would no longer be a bar.

  19. 18

    “..obvious at the time the invention was made” has to be read in the context that the date of any invention is presumed to be only its filing date, at best, unless proven to be a completed invention before then under the difficult-to-prove standards of 35 USC 102(g). Furthermore, 102(b) prior art may itself have an earlier 102(g) prior art date in some cases. So it would be a relatively rare situation in which the invention date of the subject application could be validly proven to be a year or more prior to its filing date, even if the statutory bar of 102(b)did not prevent it.
    An additional good reason for the 102(b) bar is limiting Rule 131 declarations to swearing back of other peoples much earlier filed patents to only one year before filing dates. Otherwise, even more of these Rule 131 ex parte declarations would be factually inadequate yet erroniously accepted unchallenged by 102(g) unsophisticated examiners, to issue even more overlapping U.S. patents in the same technology.

  20. 17

    Thank you Dennis for raising this point. I anticipate a long thread.

    Ned, on obviousness, I think the difference between FtF and FtI is more fundamental than anybody in this thread has identified.

    In a FtF jurisdiction, the specification is written before the date on which obviousness falls to be decided. What it contains sets out the “contribution to the art” whose obviousness has to be decided upon. In a FtI jurisdiction, however, the specification is written after the date on which the invention was conceived. Looking at that specification for disclosure of technical effects with which to test obviousness is the only way in Europe but (as far as I can determine) a big ex post facto No No in the USA. Scope here, I would say, for loads of trans-Atlantic (and trans-Pacific) misunderstanding and confusion.

  21. 16


    I applaud you on your concern. However, I haven’t noted a similar concern with regard to the disconnects between the text of 35 U.S.C. § 101 and 35 U.S.C. § 100(b) and its application by the courts.

    “Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.”


    There is no indication there that processes are a second class category. In fact, they are listed first.


    When used in this title unless the context otherwise indicates –

    (a) The term “invention” means invention or discovery.

    (b) The term “process” means process, art, or method, and includes a new use of a known process, machine, manufacture, composition of matter, or material.

    No indication of second class status there either. Nor is there any indication that processes must be tied to another category, that mental steps or algorithms generally are precluded.

    Why no concern about that?

  22. 15

    Overlooked is an equally fundamental issue: is the opinion testimony of a person who clearly was not one having OSITA admissible?

    The typical case involves dueling [1] EXPERTS and–forgive me if I’m wrong– [2]never a person who, when the invention was made, actually was then a PHOSITA.

    I’ve succeeded in beating back latter-day experts in PTO proceedings, but await the court case in which one side has a real PHOSITA and asks the judge to disqualify the expert.

  23. 14

    Your interpretation is plausible, but under that logic, why stop with 102(b) prior art? For example, one could argue that post-filing/pre-publication advances, because of the circumstances of their creation, might evidence that the invention would have been obvious. (I find myself explaining frequently that invalidation is about what people actually did prior to the invention, not what someone might have done.)

    I think Dennis proposes a sensible rule, although I doubt he’ll find much support, not on this forum, but amongst those that actually have the power to change it.

  24. 13

    be possible to “swear behind” a statutory bar

    Then that wouldn’t be a statutory bar, now would it?

  25. 12

    Another point is whether Dennis’s approach makes sense in light of how Europe and other foreign jurisdictions handle prior art.

    Um, no.

    US law does not depend on foreign jurisprudence. If there is any similarity, that is fine for as far as that goes (which is to say, there is no causality, it merely is as it is).

  26. 11

    Queen Elisabeth, on the last point you made, you have it backwards. Since secret prior art can be used in the US, but not in Europe, for obviousness purposes, more can be granted in Europe than in the US.

  27. 10

    Max, if I understand European law correctly, using only “publicly” available information for obviousness/inventive step purposes would bring US law back into line with Europe.

  28. 9

    So – properly construed, the prior art that has to be considered for for obviousness purposes is art that was “publicly known, used (in the United States) or described in a printed publication.”

    For years I have been hearing complaints of the logical inconsistency of using secret prior art for obviousness purposes. For many people, that made no sense whatsoever. But low and behold, §103 by its very terms seems to exclude secret prior art, because some, nay most, of what is in section 102 is not prior art, but is other things, such as abandonment, as in §102(c), or legal abandonment such as in §102(d).

    But now thanks to Dennis and this post, I think we also have to include as not being section 103 prior art, anything other than §102(a) art.

  29. 8

    On the other hand, Dennis is right. If one looks to history and origins of the statutory bar, one sees that it actually is a form of estoppel – of abandonment. See Pennock v. Dialogue and the subsequent enactment of a two-year grace period. The facts in Pennock are that the inventor put his invention, a firehose invention, into public use for a number of years and only sought a patent after the secret of the firehose invention was discovered by others. The Supreme Court held that one placing his invention into public use prior to filing a patent application abandoned his invention to the public.

    Therefore, §102(b) “art” is not actually prior art, because it is not necessarily prior to the date of invention.

    In patent prosecution, therefore, it should be possible to “swear behind” a statutory bar reference by showing prior invention of the subject matter disclosed in the statutory bar reference if the reference is use only for obviousness purposes.

  30. 7

    Another point is whether Dennis’s approach makes sense in light of how Europe and other foreign jurisdictions handle prior art.

    It seems that, if you follow the language in a vaccuum approach, you’ll end up with a lot of patents granted in the US that couldn’t grant in Europe.

  31. 6

    The problem with Dennis’s approach is that you get absurd results.

    Inventor makes invention in Jan 2005. Never discloses to the public.

    In 2007-2010, numerous references are published/disclosed that render the invention plainly obvious but do not anticipate the invention.

    Inventor files his patent application Jan 2010.

    Under Dennis’s approach, the invention is nonobvious because all the prior art was after the time the invention was made. This result is clearly not what was intended by the statute. You would be granting patent protection for an invention that is obvious in view of what is in the public domain.

    As another commenter suggested, the language was to be read in context, not in a vacuum.

  32. 5

    And then there are two Transactions that I would call very shady. I always wondered why they were handled the way they were. Not a month goes by that I don’t play those out in my mind, always wondering. The Fingerprints on those should tell the story. I know those will be very very important in getting to the bottom of this whole story. Loose ends? I believe I have a box full of laces that are not tied. Give me a few moments to unload my two Pea Brain.

  33. 4

    I signed a Power of Atty. on May 15, 1996. I mailed it to him as instructed. I also faxed Morrie to tell him I had in fact signed and sent it to PL.
    I was then told by PL that he was very new at this and I needed to fill everything out and send it in myself. And Morris knows the WHOLE STORY!.You can see on the back of my file that I even stated he was my Lawyer.
    All my copies were stolen. And when I asked him for a copy he sent me a piece of paper that was cut in half. with my name on it as. PRO SE. I have evidence in my USPTO Fake Files showing this to be the facts.
    Now Hal if you need to tell me just how you messed up you need to step up.
    Morrie W and S W, and Hal know the whole story. This was done to protect the first thief that claimed my idea was his.
    I get it now. You won’t get away with it. Why did my husband and I pay you on May 2, 1996. REMEMBER MY HUSBAND WAS THERE. The Drafts Person Michael Little will tell you. He was doing the Drawings for the second Application.
    The fees he charged me, what do you think they were for, Ice cream?

  34. 3

    My understanding is that outside of determining who invented first for interference purposes, the “time of the invention” is for all practical purposes equal to or less than 1 year prior to the earliest priority claim. Even in the interference context, if the invention was disclosed by anyone one year prior to the earliest priority claim, the interference is moot.

  35. 2

    One error in a fed circuit opinion that still exists to this day is the statement regarding claim terms in the Phillips decision. The court stated “We have made clear, moreover, that the ordinary and customary meaning of a claim term is the meaning that the term would have to a person of ordinary skill in the art in question at the time of the invention, i.e., as of the effective filing date of the patent application.” Phillips v. AWH Corp. 415 F.3d 1303,1313. However, the time of the invention is not the effective filing date of the patent application. I have run into this in a case where an expert stated that he was interpreting the claim terms as per the date of the invention, which was several years before the patents were filed for. The deposing attorney jumped all over him that he was not interpreting them as of the filing date of the application. For better or worse, this issue was not teed up during Markman.

  36. 1

    Dennis –
    One way to read Section 103 is noting it’s the difference between the prior art of record and the subject matter sought to be patented that you consider at the time of invention.

    The prior art itself can sprout from any part of Section 102, including 102(b), even if later than the time of invention – so that, under Graham, the scope and content of the prior art is as of the date it is effective under Section 102.

    Then, you could look at the differences (using some other Graham factors, including what’s missing from the prior art, possibly certain secondary considerations, etc.) as of the time of invention.

    Here’s the relevant language:
    …if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made

    One interesting question is how often has the time of invention been so early as to make this an issue?

Comments are closed.