Patent Prosecution: 35 U.S.C. § 112(a) Must Be Raised Before a § 102 or § 103

The Patent Office should be estopped from raising a § 112 (a) rejection after raising a § 102 or § 103 rejection in an earlier office action.  35 U.S.C. § 112 (a) states:

The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention

Logically, if the application does not describe an invention in terms that allows one skilled in the art to make and use it, then the Patent Office should not have sufficient information to suggest that the application is not novel or obvious.  In order to determine something is not novel or obvious you first have to know what it is.  I have no objection to the Patent Office putting a 35 U.S.C. § 112 (a) and novelty/obviousness rejection in the same Office Action, where the PTO explains that to the best of their understanding of the invention it would not be novel or obvious for the following reasons.


However, when the Patent Office in a previous office action does not raise a 35 U.S.C. § 112 (a) rejection but does raise either § 102 or § 103 rejection of all the claims and then in a latter Office Action raises a 35 U.S.C. § 112 (a) this is illogical and the Patent Office should either: 1) be estopped from raising the 35 U.S.C. § 112 (a) or 2) provide an affidavit that the examiner is “one skilled in the art” and that they do not have the information necessary to make and use the invention.  If the PTO does submit an affidavit, this should be able to be overcome by an affidavit provided by the applicant by “one skilled in the art.”  The only grounds for the PTO to rebut this is to suggest the applicant’s affidavit does not show that the person is “one skilled in the art”, which should be a very high burden on the part of the PTO>

The particular circumstances that bring this to mind is a case that has been pending over a decade.  The applicant has filed multiple responses, filed an appeal and won, was told that modifying the claims in a certain manner that would make the claims allowable, the applicant made the suggested changes and received new rejections, negotiated new amendments to the claims that they were told would make the claims allowable.  They again submitted a response along the lines negotiated and then the Patent Office, after a decade of prosecution, raised a 35 U.S.C. § 112 (a) issue.  Note all the amendments were based on incorporating dependent claims.

I would appreciate feedback on my proposed rule that the PTO be estopped from raising 35 U.S.C. § 112 (a) rejection after previous Office Action that had no 35 U.S.C. § 112 (a) rejection but did have a novelty/obviousness rejection of all the claims.


Warning & Disclaimer: The pages, articles and comments on do not constitute legal advice, nor do they create any attorney-client relationship. The articles published express the personal opinion and views of the author as of the time of publication and should not be attributed to the author’s employer, clients or the sponsors of

Join the Discussion

20 comments so far.

  • [Avatar for Old Examiner]
    Old Examiner
    March 16, 2013 02:42 pm

    First, with respect to Gene’s comment in #2 above, I take exception to your statement that examiners never read the application. In my lengthy career, I have always read the specification before I even looked at the claims. I agree some examiners may not but don’t make the generalization that it is the practice of all examiners.

    Next, when I make a 112 rejection relating to enablement, I don’t make a 102/103 rejection. It is simply doesn’t make sense. I haven’t found many applications that aren’t sufficiently disclosed. Sometimes an amendment will add a limitation not supported by the original disclosure which requires a 112 rejection; but to go through a decade of examination with the issue of enablement never having raised and then to bring it up is suspect.

  • [Avatar for Dale Halling]
    Dale Halling
    March 15, 2013 02:44 pm

    Old Examiner,

    Thank you very much for your suggestions. By the way this is third examiner on the case and not the one that lost the appeal.

  • [Avatar for Old Examiner]
    Old Examiner
    March 15, 2013 12:39 pm

    In the particular fact situation you describe, you should file a complaint with the Director of the technology center of the examiner. At the least, the Director should transfer the application to another examiner. Clearly, the examiner you are dealing with is adverse to allowing applications, particularly after losing on appeal.

  • [Avatar for Dale Halling]
    Dale Halling
    March 12, 2013 11:25 pm

    Examiner David,

    1) It is extremely rare for the applicant to broaden the claims during prosecution. If this is done, it is usually done in a continuation.
    In addition, the question is not really about the claims, it is about the specification. If it is about the claims, then ,the appropriate rejection would be 112 (b). I do not believe that the applicant should have to file such an affidavit. (see point 3 below for a further explanation)

    2) No. If the Examiner disagrees with the interpretation by the applicant, they can (and do) maintain that the claims/specification does not claim/teach what the applicant is suggesting.

    3) Why should the Examiner not be required to determine if the specification meets 112(a) first? I think your examples all are the result of the PTO examining the claims and not examining the specification. This has been discussed above. The PTO should first read the specification. If it is not clear, then, it should issue a 112 (a) rejection. It may simultaneously issue a 102/103 rejection based on their best understanding of the specification in the interests of compact prosecution. Next, the PTO should undertake a search that is not just of the claims, but the specification – at least that part of the specification that is reasonably related to the claims (This should clarify the case of an omnibus specification that incorporates multiple inventions). The PTO’s argument that amending the claims (narrowing the claims) requires a new search would rarely be necessary if this were the case.

    You bring up some interesting points; but in my opinion, they all result from improper examination procedures on the part of the PTO.

  • [Avatar for examiner David]
    examiner David
    March 12, 2013 09:52 pm

    First, what about the instance if Applicant were to broaden the claims during prosecution, and thus requiring the Examiner to LATER make a 112, first. Should the affidavit at least provide that Applicant will also be estopped from broadening the claims after the first action?

    Second, what about the instance if Applicant were to provide a claim/specification interpretation, and thus requiring the Examiner to LATER make a 112, first? Should the affidavit at least provide that Applicant will also be estopped from questioning Examiner’s initial claim/specification interpretations?

    Third, what about the instance if Applicant cannot end up providing a copy of the reference incorporated by reference that is relied upon for enabling specific claimed subject matter, and thus requiring the Examiner to LATER make a 112, first? Should the affidavit at least provide that Applicant will also be estopped from providing copies of references not already disclosed?

  • [Avatar for Dale Halling]
    Dale Halling
    March 12, 2013 11:34 am


    Excellent points that you should summarize and have Gene put in a new post. Will it make any difference? Did the PTO get back to you?

    When reading anything you have to put it into context. If I say “window”, I could mean a pane of glass with a frame, I could mean a computer screen, I could mean, a means of access, a section of time, a range of frequencies, etc. Window is not a special case either, any word in isolation is subject to numerous interpretations. The rule that words have to be read in context is not specific to patents – it is logic.

    I see two main reasons for the problems you points out:
    1) There is no consequence to examiners giving out spurious rejections – and for many procedural issues, such as, incorrect finals, incorrect divisionals, no recourse for the applicant.
    2) The lack of understand that a patent is a property right. This is not a game. These same examiners would treat a $10k car with car and respect. But the attitude is this is all a game and the government is not protecting the inventor’s rights but granting them a special favor.

  • [Avatar for Dale Halling]
    Dale Halling
    March 12, 2013 10:49 am


    There was an appeal and the applicant won. Then the PTO found new art for the next rejection. Then there was an OA that said certain claims were allowable. The applicant made the changes and new art was cited – Repeat – repeat and then bring up a 112(a).

  • [Avatar for Dale Halling]
    Dale Halling
    March 12, 2013 10:47 am

    Examiner XYZ

    “but that could/should be cleared up with an examiner’s amendment” not a 112(a), which is my point.

    Also,specifications are objected to and claims are rejected, I don’t think the article made that distinction clear.” Good point

  • [Avatar for Ken Fagin]
    Ken Fagin
    March 12, 2013 10:27 am

    Some of this practice is mentioned in the response to thePTO’s request for comments in re the RCE outreach, which response I submitted yesterday. It can be accessed here:


  • [Avatar for Paul F. Morgan]
    Paul F. Morgan
    March 12, 2013 10:24 am

    The indicated problem of 10 years of prosecution delays [presumably without even an appeal] needs better solutions than this. Specifically, better PTO docket tracking, supervision and management that will cause such long-delayed applications to be promply processed, instead of allowing their amendments, RCE’s or continuations to sit unprocessed for many months on some examiner’s docket while very much newer applications are worked on instead. If not improved, then transfering such “submarine” applications to other examiners that WILL process them very promptly, and rewarding them for doing so. The latter is particularly important for applicants that make extensive claim changes and claim additions that make responsive office actions much more burdensome, a tactic for some applicants doing so to delay applications.

  • [Avatar for Anon]
    March 12, 2013 09:57 am

    but that does not allow him/her to change the common meaning of words

    Yes it does – but with a caveat. The ability to change the meaning of common words is limited by art field consistency. You cannot change recognized terms of art to be something they are not. But precisely in my example of Peanut Butter, I am explicitly outside of the food arts – and I have every right to redefinie Peanut Butter as I see fit in that instance.

    That actually makes my point as to the critical nature of the specification.

  • [Avatar for examiner-xyz]
    March 12, 2013 08:30 am

    112’s after a 102/103 does not comply with compact prosecution. See MPEP 2173.06. If the spec or claims are unclear and confusing the examiner should point this out and then explain how he/she is interpreting the spec/claims (an important part that is sometimes left out) and move forward with a rejection if appropriate. Reducing pendency time is one of the goals of the Office. If however the spec and/or claims are so confusing that a proper interpretation can not be made the examiner may merely put forth the 112.only. This should be rare. Under compact prosecution all rejections and objections should be made in the initial office action. I suppose there maybe a situation where a 112 is missed and then made prior to issuing and after a rejection but that could/should be cleared up with an examiner’s amendment, under the same principle of compact prosecution. Also,specifications are objected to and claims are rejected, I don’t think the article made that distinction clear. Also, as to the “peanut butter” issue, while applicant can be his own lexicographer but that does not allow him/her to change the common meaning of words (i.e. should not be allowed to redefine “nut” = “bolt”). That said, the examiner would still have to review the spec for any special definitions. If the spec was confusing about what “peanut butter” is, that should be stated and a rejection should follow with what the examiner understands peanut butter to be.

  • [Avatar for Gene Quinn]
    Gene Quinn
    March 11, 2013 04:56 pm


    How do you know that the claim “Peanut Butter” should be rejected under 102 without looking at the specification?

    What if the specification said this:

    “By ‘peanut butter’ I mean: ‘A sandwich comprising a first piece of bread with a first side, a second side and a perimeter…”

    I think the point is that there is so much nonsense that goes on in examination and it is frequently because the examiners don’t even look at the specification. If that is the case (which it is) then why require a specification at all? If it is required then it ought to be read.

    What about all the expense caused by examiners who simply take the attitude — “well if you don’t like it just file a Notice of Appeal and appeal it.” Do they understand that takes time and costs money? Sometimes I really wonder.

    I would also be supportive of a rule that the USPTO cannot charge fees other than a basic filing fee prior to the issue fee. Then calculate the issue fee based on the necessary actions of the patent examiner, not charging for the nonsense required to address inappropriate examiner rejections/objections.


  • [Avatar for Anon]
    March 11, 2013 04:34 pm


    I agree with Dale, even (especially) for your Peanut Butter example.

    Since the applicant can be their own lexicographer, what you read as “Peanut Butter” in total isolation of a specification you cannot understand MAY NOT be the Peanut Butter that you are thinking of as a food substance. Peanut may be a critical size reference and Butter may be a device to butt.

    I daresay that your example is one mere step away from the (in)famous just-ignore-the-specification-and-examine-on-the-first-go-around-with-a-keyword-search-from-the-claims-only mantra that I have actually heard from examiners, and not only that, but the examiners tell me that such is permissible per their SPEs.

  • [Avatar for Michael Risch]
    Michael Risch
    March 11, 2013 04:15 pm

    I suppose if 112 is meant to ONLY cover stuff that we just cannot understand, then this works. But how about this patent?

    laskdfja;lsd asdl;kfjalsdf adsfja;sldf adskfjasldf adskfljladsk dsafkjasdl;f asdfkj adsf adsfj
    asdf;l adslfj al;dskfj

    Peanut Butter

    Seems like I could reject on 102 without much problem. Now, maybe I should ALSO reject on 112, but the assumption that you can’t understand the claim because the specification doesn’t tell you how to make and use it just doesn’t fly with me. I’ve read hundreds (thousands?) of software patents, and I can tell what is being claimed in most of them (sometimes with frightening clarity and breadth), with only the most cursory reading of the specification, even though I might have no idea whether someone back at that time could have actually implemented it based on the disclosure.

    I understand that in your case this might not have held true, but my entire career of scholarship has been devoted to arguing that we should make general rules based on anecdotes.

  • [Avatar for Dale Halling]
    Dale Halling
    March 11, 2013 04:06 pm


    I understand you example, but you clearly understand what a widget is with 1 vs 2 vs 3 holes. So I do not believe the analogy applies. I also not sure the BRC applies, because the BRC states that it is the broadest reasonable interpretation of the claims in “light of the specification” (something that is constantly ignored by the PTO). If you cannot understand the specification enough to build and use the invention, then you cannot determine the BRC in light of the claims – widget may have a special meaning. You also cannot perform a proper search for the same reason. Logically, dealing with 112(a) issues must come first.

  • [Avatar for Dale Halling]
    Dale Halling
    March 11, 2013 03:56 pm


    What I am suggesting is that it is logically impossible to say that you know an invention is not novel, but you do not understand the invention well enough to practice it. Note if the Examiner, who must likely is not skilled in the art, can understand the invention well enough to reject it then it is likely that one skilled in the art does understand the invention.

    The PTO is quick to hand out finals, but never is compelled to live up to compact prosecution. As to the issue being raised in court – that is separate issue about how the presumption of validity has been eroded to the point that courts treat the process of getting a patent as no better than a registration system. But yes I would rather the PTO be estopped and these issues be raised in court. Courts show no deference to the PTO, so what is the point and no I don’t think examiners are likely to have raised a legitimate issue that was not spotted by the client or myself and save us from embarrassment in court.

  • [Avatar for Michael Risch]
    Michael Risch
    March 11, 2013 03:19 pm

    “Logically, if the application does not describe an invention in terms that allows one skilled in the art to make and use it, then the Patent Office should not have sufficient information to suggest that the application is not novel or obvious.”

    I disagree, and here’s why.

    1. The claims are based on broadest reasonable construction. It’s a bad rule, and it’s not my rule. But it is the rule. Based on BRC, you can often wind up with overbroad, obvious claims wholly divorced from the specification. That is, after all, the point of it.

    2. Both enablement and description are about scope of invention, not scope of claim. It is quite possible – indeed, quite likely, that a patent can fail one without the other. An example:

    I developed a widget. It has 2 holes in it to allow water to flow.

    I claim: a widget with any number of holes.

    Now, let’s say I have a widget with 3 holes in the prior art. As an examiner, I can knock your claim out in 30 seconds, no fight about enablement, no fight about description, no nothing. And I’m going to do it. And I should do it, because you are over claiming.

    Then, you come back and say:

    I claim: a widget with 1 or 2 holes.

    NOW we can talk about 112. Would you really estop the examiner in this example?

  • [Avatar for Gene Quinn]
    Gene Quinn
    March 11, 2013 02:57 pm

    I’m not sure Dale is suggesting that 112 rejections not be made. I think he is suggesting that the 112 rejections can’t be made if the examiner is going on to 102 and 103. I know exactly what Dale is saying. I have a situation now where the examiner says that he is ignoring claim language because he deems it to be new matter in violation of 112. Of course, if the examiner actually read the application (which I know never happens) or the examiner read our amendment explaining where support for the claim limitations is found in the original filing, the examiner would know that the limitations are well supported.

    So how can an examiner actually evaluate 102 and 103 when they choose to ignore what is in the claims based on some bogus 112? The same examiner in this case said that he was ignoring other claim language because it wasn’t clear to him that it was intended to be added… despite being underlined and arguments made relative to those limitations being inserted.

    I’m with Dale. Examiners who do this are wasting their time and our time, as well as the money of clients.


  • [Avatar for Sam]
    March 11, 2013 02:16 pm

    Would you rather have the examiner bring up a legitimate 112 issue when you can still amend? Or the alleged infringer at trial? Estopping 112 rejections sounds like a pretty overbearing way to solve the problem of dealing with bad examiners.