A Brief Defense of the Written Description Requirement

by Michael Risch

The Federal Circuit’s upcoming consideration of Ariad v. Lilly has generated a fair amount of buzz among those who follow patent policy. The dispute arises from the interpretation of 35 U.S.C. 112 ¶ 1, which states in relevant part:

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….

All agree that this language includes an “enablement” rule, which requires that the specification enable a person having ordinary skill in the art (the PHOSITA) to make and use the invention. More controversial is the phrase, “written description of the invention,” and whether that phrase entails a separate requirement apart from enabling the PHOSITA to make and use the invention. It appears that academics are split on the question, and most practitioners appear to disfavor a separate requirement.

This essay briefly describes the dispute, and then raises an important but under-theorized argument in favor of a separate written description requirement. The essay accepts the persuasive grammatical reading of the statute proposed by opponents of a separate written description requirement. This reading suggests that a patent disclosure is sufficient so long as the PHOSITA can practice the invention.

However, while enablement is necessary to satisfy Section 112, it is not sufficient. The statutory language certainly requires that an applicant identify the invention sufficiently to allow a PHOSITA to make and use it. However, even if the specification enables a PHOSITA to practice the invention, the applicant is not relieved of the obligation to identify the invention. Instead, the specification must contain a “description of the invention” even if such description would only serve to reinforce other parts of the disclosure that enable one to make and use the invention.

This reading of the statute is consistent with the grammatical breakdown of Section 112. It is also the preferred interpretation, because a description of the invention fulfills an important purpose in the patent system. The requirement ensures that the applicant actually invented the claimed subject matter. Reading description out of the statute would allow patent applicants to claim subject matter they did not invent, and would effectively rewrite nearly 120 years of precedent about the conception of inventions.

This is certainly not the only argument in favor of a written description requirement; a strong written description requirement provides policy benefits in clarifying claim scope and policing patentable subject matter. This essay leaves such benefits to the side, and focuses only on the statutory basis for the rule.

Finally, the essay considers the Ariad case and concludes, perhaps surprisingly, that under the vision of written description presented in this essay the claims at issue may well be described.

The remainder of the essay is available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1504631.

80 thoughts on “A Brief Defense of the Written Description Requirement

  1. 80

    9 wrote:
    “This is a rather classic line: Where an original claim read in light of the specification fails to provide an adequate written description of the invention because, for example, it claims the invention by what it does rather than what it is, then the original claim is invalid.

    I wish we could get the Fed Circ to go ahead and throw that into some caselaw. I’d lol my behind off. ”

    As an interesting point, this would knock out nearly all the bogus patents on software, which mostly say “A machine which does thus-and-such.”

  2. 79

    “Conceptually, the patent attorney who took over the Wright Brothers’ patent was able to draft claims which elevated the invention, in the eyes of the courts, to covering the mid-wing aileron. Artfully drafted claims can characterize configurations in such a way that the claims cover things never described in the disclosure. Is this fair?”

    Honestly? Probably not. Whether or not it’s fair, it’s undesirable from a public policy standpoint, and it’s also contrary to the intention of Congress and the Framers of the Constitution. People are supposed to get exclusive rights to *their* invention (should Congress decide that such rights promote the progress of Science — if not, they don’t, and Congress can withhold such rights at a whim, as well). They are not supposed to get exclusive rights to other people’s inventions.

    “The CAFC seems to be heading in the direction of requiring that claims be enabled across the full breadth of their scope of coverage. With this requirement in place, how could a patent claim ever cover an unanticipated future improvement.”

    We can only hope that that will be the result.

  3. 78

    Ned, we can call it “intermediate limitation”. But we can put it another way. The written description requirement stops the belated claiming of selection inventions that were not conceived till after the app was filed. It seems that no other legal tool is available, to perform this necessary blocking function.

  4. 77

    Peanut and MaxDrei, interesting example. His original claim, Al with some of three different compounds; but he discloses only one example. The obvious problem is that he never had enough to claim the genus in the first place — the claim is far broader than his example(s).

    It might be better to talk about a case that does not have such obvious disclosure problems because the issue is really not the prior art.

    Try a slightly different example. The disclosure is to Al with three different compounds in ranges. The claim is directed to that, and is allowed. Then, in a continuation, the applicant files a claim to each compound mixed with aluminum in its specific range.

    What is wrong with that? In a sense it is disclosed.

    What if after the original filing date of the parent, a different party found a different and unexpected utility for Al with one of the compounds three compounds in the range. Who gets the patent?

    We all know the answer to these questions do we not? Just because people can make an alloy of Al and each of the tree compounds in isolation, it is the first discoverer of a utility for such isolated alloys that is the first inventor.

    Going back to my range example, 1-10, with examples at 1, 2, 5, 7 and 10; let us suppose another patent is filed with the a disclosure of examples 2 and 5 and a demonstration of unexpected results for that subrange. Would it be apropos for the first applicant to file a claim to the subrange 2-5 even if he could show that he was aware of the utility of the subrange prior to the filing date of the other application? I think not, and the reason is the written description requirement and this has nothing to do with enablement of how to make and use.

  5. 76

    MaxDrei’s scenario:

    “Peanut take the case of an inventor who discloses an aluminium alloy that makes better car body panels. He doesn’t know the art so he claims an aluminium based alloy with some silicon in it, some magnesium and some manganese.

    He files. The PTO cites 20 patents on car body panel alloys of aluminium, all with Si, Mg and Mn but none discloses exactly the alloy of his worked Example. What can he amend down to. Exactly the Example, or some undisclosed intermediate generalisation. Is the intermediate generalisation new matter? Is there a written description of it in the app as filed?”

    What the inventor should do in this case is determine what it is about the alloy that is novel and non-obvious relative to those 20 references. Maybe his worked example is obvious based on those references, so nothing is patentable. On the other hand, it may be the specific percentage of Si, or the specific way he made his alloy. Whatever the significant difference is, that is what the inventor should put in the claim. In most cases, that amendment won’t require the inventor to recite the exact worked example. Further, that amendment cannot introduce new matter, so whatever the difference is must have been expressed somewhere in the original specification.

  6. 75

    “Or are you going to apply the broadest UNreasonable interpretation rule here as well?”

    I might do that for kicks. Nevertheless I don’t see how you think the spec is going to magically cure the WD of an invention because, for example, it claims the invention by what it does rather than what it is. At least in the majority of the cases that I would take issue with over this using this reasoning.

  7. 74

    “This is a rather classic line: Where an original claim read in light of the specification fails to provide an adequate written description of the invention because, for example, it claims the invention by what it does rather than what it is, then the original claim is invalid.”

    That is a nice line, and it might even be accurate, but you have to read all of the words, not just the ones you like. Don’t forget the phrase “read in light of the specification.” Or are you going to apply the broadest UNreasonable interpretation rule here as well?

  8. 73

    Noise we always knew who would win on that front, there was no need for me to bludgeon you and yours in a battle that never existed.

    I still agree with them on the merits of the language of the statute in any event:

    “[t]he statute itself says so, requiring a ********description****** of the invention and of the manner of making and using it, and no amount of parsing can read that conjunction out of the statute.”

    This is a rather classic line: Where an original claim read in light of the specification fails to provide an adequate written description of the invention because, for example, it claims the invention by what it does rather than what it is, then the original claim is invalid.

    I wish we could get the Fed Circ to go ahead and throw that into some caselaw. I’d lol my behind off.

  9. 72

    Patent Docs has what I have seen as the first substantive counter to the “plain language” argument in the 112 paragraph 1 debate.

    link to patentdocs.org

    Take note 6 – rather than engage in a hopeless and futile effort on grammer (not your strong skill), this is how one can bludgeon with case law where legislative law falls short.

  10. 71

    Regarding the EPO case, the Fed. Cir. has recently held en banc that in order to infringe a product by process claim one has to practice the process as well.

    So, if this case stands – and I think it flies in the face of Supreme Court precedent, there is no EPO problem in the US. But, as you might guess, I strongly disagree with the Fed. Cir. case. I agree with the dissent.

  11. 70

    fish bones said, “Jules: I believe that within the past few years there was a FedCir case whose holding discouraged applicants from stating the problem in their applications.

    Can anyone refresh my recollection as to the case and its holding???”

    I don’t remember the exact case, but there are a whole series of cases involving claim construction where if one says something described in the specification IS the invention or IS required in order to make it work, the claims are construed to include that something even if they are nominally broader. As a result, US practitioners routinely stop talking about “the invention” in the specification, no longer say an embodiment is preferred, and in some cases, just, as I said before, describe a host of embodiments with no information as to what is new.

    Circling back to an earlier discussion about genus and species and subranges, I think we all can agree here that the disclosure of a species does not entitle one to claim a genus without more and that the disclosure of a genus does not entitle one to claim a subrange or individual species unless there is some description of such a subrange and species. In the first case, the species->genus case, the problem is both of enablement and written description. O’Reilly v. Morse is seminal case on this issue. But, when there is a disclosure of a broader range but no information as to the criticality of a subrange or species, the major failure is not one of enablement, but of written description. If one were to actually add that description, it would be new matter. This illustrates why a claim to what would be new matter if added to the specification should not be allowed under a written description requirement separate from enablement.

  12. 69

    hindsight — no, my vague memory says that there was a case where you didn’t want to state in your application the problems in the prior art because that gave the examiner his reason for modifying the primary reference with his secondary.

    Any idea what the case was, anyone?

    Thanks

  13. 68

    KSR?
    any reason given is a reason for the examiner to show why it would be obvious to follow the applicant’s direction

  14. 67

    Jules: I believe that within the past few years there was a FedCir case whose holding discouraged applicants from stating the problem in their applications.

    Can anyone refresh my recollection as to the case and its holding???

  15. 64

    Remember I said that barebones/nothing substantial can be interpreted as a POS. Think of it like this. The examiner says, “hmm, yes, I see. It is a POS. It looks slightly different than other POS’s, but yes indeed it is a POS!”

  16. 63

    fishy – there is not one that I am aware of, that was why I went with snark as opposed to yes/no. And I thought I did explain – you get a less comprehensive examination because the examiner does not fully understand or appreciate what you are claiming. So if the examiner allows your claims with a high level of uncertainty, then there may be a chance that the granted claims will be found invalid down the road.

  17. 62

    Luke says: “Fish bones wrote: “Is it possible to CLAIM an invention without identifying what the invention is?””

    Did I really say that?

  18. 61

    Jules — If there is a requirement to describe the “problem” in the specification, please identify the statutory section or case law support for such.

    You said that failure to identify the problem could result in invalidity, please explain.

  19. 60

    PS – when it comes to patents “describing” inventions, if you are a patent professional, even the specification reads like gobbledeegook, to say nothing of the claims.

    One of my best friends who is still in research, unlike me, once came to me coz he ha been told to look at a Japanese patent to heck out a certain reaction. He was still at a university at the time, and his basic attitude was why the hell would want to look a a patent. They are useless as sources of information”. Now he’s in the private sector, he actually has to write patent apps himself as part of the commercialization process of his company’s research; but his attitude hasn’t changed much.

    For many people, researachers, that is, i.e. people who promote the progress of science, patents are valueless, and contribute nothing – they teach nothing because they are so opaque.

    And I would posit having no clearly articulated separate written description requirement is one of the primary reasons. (Making the abstract a pretty well much useless appendix to the rest of the spcification is the other)

    PPS -sorry about all the typos in the previous post. No doubt there are loads here too…

    Cheers, Luke

  20. 59

    Fish bones wrote:

    “Is it possible to CLAIM an invention without identifying what the invention is?”

    Absolutely. Claims don’t identify (as in describe)
    inventions, they identify (as in define) the metes and bounds of a monopoly.

    You have to get your head around the fact that the word “invetion” is used in several different ways in patent law. Only one of them equates to “claim”. Recall “claiming” is a rather more recent concept that patenting inventions. No actual inventor thinks “my invention is…” and concludes with something hat sounds like a claim. When they sign off to having invented something, they are no signing off to having invented a number of claims, but a “thing”. But what is that “thing”. It is much more like the “inventive concept” than it is like “Claim 16”.

    More fro Fish bones:

    “Ned says “not one word of which describes the problem and the solution,” We need to describe the “problem” in our specification?”

    To which I say yes. (taking care that “problem” need not necessarily be a pre-identified problem. Perhaps the greatest inventions solve problems nobody even knew existed)

    Why? Because that is part and parcel of promoting the progress of science and the useful arts.

    OK, I’ll try and describe an actual example from UK litigation. Kirin Amgen didn’t discover Human EPO, but they did sequence it. Using this information, they could do some genetic engineering which enabled them to produce commercial quantities of the stuff – hitherto impossible. Nobody denies this was very inventive work. The resulting patent related to making a polypeptide possessing the important bit of the EPO structure. EPO, in other words, plus or minus unimportant variations.

    The problem was that although the various claims were enabled – you could produce perfectly good commercial quantities of EPO by following the patent, there was a problem in determining if any given sample of EPO, potentially produced by another method not covered by some of the claims, was in fact embraced by those claims or not.

    This is because another person, having generated their own EPO, had problems understanding if their EPO read on the patent’s EPO (product-by-process claims) because the test required comparing it with some natural EPO – but which natural EPO? Not all natural EPO is the same – and this was not understood at the time. Thus the claims (“invention” if you like) were insufficiently clear, even though they were sufficiently enabled.

    Now this was a problem, because lack of clarity in the claims was a grounds for invalidation in the patnt office *but not in the courts* under the UK patent act. In the courts, the only “insufficiency” attack allowed was lack of enablement, not clarity (i.e. written description).

    (This litigation went all the way to the House of Lords, on other grounds however)

    A further aspect of this case you might like to ponder is the advance in knowledge since filing. A the Appeal court stage, the presiding judge said it would be silly if a patent was sufficiently enabled in 1984, but not in 1990. In 1984, people only knew of one form of EPO, so the comparison test was less of a problem. In 1990, two different forms of EPO were known that would fit the comparison test bill, and yet give different results regarding infringement, but the patent didn’t specify which – this wasn’t taught.

    Should the claim be trashed on account of that later knowledge? And if so, on what ground(s) (in say US law). If not, why not?

    Cheers, Luke

  21. 58

    “I don’t see why it would be new matter. I just don’t.”

    While I understand how you could think genus disclosure inherently includes all species, and while this is true in some limited examples, it frequently is insufficient for proving conception of the invention to include a particular species prior to the filing date.

    In the example you took issue with, this would be new matter when the original disclosure lacks either an explicit disclosure of the claimed species or an articulated reasoning that would lead a person reading the original application to inherently arrive at a particular member of the species instead of another.

    Given that the disclosure only calls for “a biocompatible metal”, there is a fairly large number of suitable candidate metals and alloys which could potentially be used. Claiming a particular species after the filing date implies that some other consideration went into picking among the known species. Without such an articulated reasoning in the original disclosure that leads a person to stainless steel, i.e. “in one embodiment the biocompatible metal should be inexpensive, reasonably conductive, mechanically sturdy, weldable, etc.” then there is no reason why an applicant wouldn’t pick titanium, vitalium, gold or platinum, or any of the candidate biocompatible species. Absent such a disclosure, the applicant can’t show that he/she didn’t make additional considerations after the filing date to arrive at a particular solution, even when such considerations are simple. Furthermore, someone reading the application wouldn’t have any reason to arrive at a particular species without making considerations outside of those listed in the application.

    This usage of the written disclosure requirement also prevents an application from amending around anticipatory prior art using new considerations made after the filing date. For this example, let’s say the prior art discloses “[the device] made from a biocompatible metal, for example titanium for it’s durability and conductive properties.” Prior art that discloses a genus isn’t considered to anticipate every claim species, so the amendment limiting to stainless steel or iridium disqualifies the reference from use in a 102 consideration. Should a previously non-disclosed species be considered to overcome the anticipatory reference in this example?

    (On the other hand, if the specification DOES articulate reasoning that would lead a person of ordinary skill to conclude that stainless steel is an inherent embodiment, even without explicitly stating it, then the applicant might be able to appeal a new matter rejection.)

    Anyway, this example and the previous negative limitation example still provide relevant reasons why the written description requirement is useful and protects the public.

  22. 57

    Dear Mr. Heller:
    Ahh In re Koury and Weymouth. One of my all time favorite cases. From my understanding of claimed ranges, which by the way I haven’t done since my days at Townsend, one need not include evidence of the unexpected results. (Although I prefer putting the same in the application). The evidence of the unexpected results may be provied via declaration during prosecution. There is no new matter rejection, because they are not part of the written description or the enablement requirement. It is merely evidence to rebut the presumption of obviousness. The real issue for me is that I think the unexpected results should be in existence before the filing date of the application.

    In my earlier post, I was referring to a self-enabling claim. A patent application may be drafted that has self-enabling claims, but still fails the written description requirement. Additionally, it may very well be that there is inadequate disclosure to allow amending the application to satisfy the written descritpion requirement without running afoul of 35 U.S.C. 132.

  23. 56

    “But the problem comes in with later-filed claims — If they now claim previously disclosed but unclaimed subject matter, what then?”

    “Is it possible to CLAIM an invention without identifying what the invention is?”

    —–
    Even the originally filed claims can fail the written description requirement. The problem can arise when inventors try to prepare and file applications themselves. I’ve seen plenty of inventors go on and on in their disclosures about how great their invention is, describing all of its potential benefits and uses, even going so far as telling what it “does”, without ever actually stating what the invention “is”. Some inventors attempts at claims are the same — pure function — and effectively single-means claims. It sometimes takes a great deal of questioning to get them to disclose the actual invention. If they were to try to file this themselves in such cases, such inventors would likely fail to satisfy the written description requirement.

  24. 55

    “Are you really that opposed to the concept of private property?”

    not at all. infact, im willing to make the secret an actual property for a limited time. if is stays a secret, then it is not private property and if the secret gets out, the owner has nothing. im willing to share for some tangible benefit for the inventor for the betterment of all mankind. im just that kinda guy.

  25. 54

    Broje, could you be so kind as to give us your definition of New Matter? This is of interest since some other people seem to perhaps share this “alternative” definition which most of us here that didn’t play hooky on new matter day have a hard time seeing and fathoming.

  26. 53

    ******Lets say a specification discloses a genus without naming all the species of this genus, i.e. “wherein the device is made from a biocompatible metal.” In response to an office action, the applicant limits the claim to some previously undisclosed, but recognizably obvious species of this genus, “wherein the biocompatible metal comprises stainless steel or iridium.” While these species would arguably be enabled by the written discription, given that they are well-known metals of the disclosed genus, there is no way to confirm that the inventor originally anticipated these species of the genus.**********

    I don’t see why it would be new matter. I just don’t.

  27. 52

    shouldnt there be an act of congress or something to get those secret and valuable business methods out into the public?

    Are you really that opposed to the concept of private property?

  28. 51

    (Sorry if this is a repost, original didn’t appear within an hour)

    Broje,

    “I just don’t follow that logic. If the Specification and drawings enable the claimed invention, and it is a new and non-obvious invention, how is it possible to say that the inventor did not invent the claimed invention? Maybe someone can give me a persuasive example.”

    2 points in defense of the written description requirement, both relating to new matter situations.

    Lets say a specification discloses a genus without naming all the species of this genus, i.e. “wherein the device is made from a biocompatible metal.” In response to an office action, the applicant limits the claim to some previously undisclosed, but recognizably obvious species of this genus, “wherein the biocompatible metal comprises stainless steel or iridium.” While these species would arguably be enabled by the written discription, given that they are well-known metals of the disclosed genus, there is no way to confirm that the inventor originally anticipated these species of the genus.

    In a related example, there is also the problem of previously undisclosed negative limitations. For example, lets say a specification discloses a genus and several acceptable species of this genus, i.e. “suitable materials for the device include biocompatible metals including any of stainless steel, aluminum, iridium,…” In order to amend around an office action which cites prior art that discloses only a species of a claimed genus without disclosing the entire genus, the applicant amends a broad genus claim to state something along the lines of, “wherein the material of the device includes a biocompatible metal except stainless steel.” Again, while such an amendment might be enabled by the disclosure (the disclosure enables fabrication of the device from non-steel metals), there is no written description supporting such an exclusion, and again, no way of knowing that the inventor had anticipated this limitation from the written description requirement.

    Additionally, you can look to Ned’s examples about numerical ranges for another example of the same concepts. The main point is that the written description requirement provides a better protection to the public from new matter amendments.

  29. 50

    Peanut take the case of an inventor who discloses an aluminium alloy that makes better car body panels. He doesn’t know the art so he claims an aluminium based alloy with some silicon in it, some magnesium and some manganese.

    He files. The PTO cites 20 patents on car body panel alloys of aluminium, all with Si, Mg and Mn but none discloses exactly the alloy of his worked Example. What can he amend down to. Exactly the Example, or some undisclosed intermediate generalisation. Is the intermediate generalisation new matter? Is there a written description of it in the app as filed?

  30. 49

    Ken, how would you deal with the range example in my post above? If the applicant claims the subrange that has no disclosure of unexpected results or that the subrange was an invention, but otherwise was fully described and enabled, what is your rejection?

  31. 48

    “However, even if the specification enables a PHOSITA to practice the invention, the applicant is not relieved of the obligation to identify the invention.”

    I agree with several of the comments above – the CLAIMS are the appropriate place for the applicant to identify what exactly the invention is.

    The fact that the claims can be changed after filing does not render the claims unsuited for this purpose. Quite the opposite.

    Patent law has always recognized that the applicant’s belief, at the time of filing, about what the applicant’s exact contribution to the art is may be incorrect. If the applicant had to be correct, at the time of filing, about the applicant’s contribution, then amending to distinguish the art would not be permissible.

    The fears that the applicant will claim somthing that he/she has not invented are unfounded, because whatever claims the applicant gets must recite something that (1) is enabled by the original specification, (2) is novel, (3) is not obvious, and (4) does not introduce new matter into the application.

    On the other hand, there is a real problem with the notion that the “invention” is set forth in both the detailed description and the claims. We have all seen the cases where the courts determine what the invention should be based on the specification, and cause problems when that isn’t identical with what the invention actually is (as defined by the claims).

  32. 47

    Ok I know that we are in a depression, but come on where is the argument here? I just don’t see it. The written description is a different requirement from enablement. What it prevents is an applicant from merely providing a model of the invention to the USPTO and then a set of claims and says, hey, look at my model, a person of ordinary skill in the art would be enabled to make and use the invention. That would be a situation whereby enablment would be satisfied, but the written description would not. OK so they don’t accept patent models any more. That doesn’t mean that they kept the language in the statute just to ensure that applicants knew you had to describe the invention and just not have a set of claims. Now if we talk about the abstract or brief description of drawings and background section . . . .those probably are not necessary. But come on, you need a written description and it is different from enabling claims. Remember claims are part of the written description. However, they may merely define the invention and may fall short as to how to make and use the same.

    In ending, I really don’t see any controversy here. Am I missing something?

  33. 46

    shouldnt there be an act of congress or something to get those secret and valuable business methods out into the public? it seems that the world would be a better place if such secrets were made available to the public.

    jk

  34. 45

    Yes. Are you aware that the USPTO is an imperfect place and sometimes things happen that aren’t exactly 100% in line with the every rule in the big rulebook?”

    Painfully.

    I’m sorry to hear that. But there are ways to make the “human” aspect of the PTO work in your favor. I’ll not be divulging all of my secrets here (or anywhere for that matter). Like nearly all business methods, they are far more valuable when kept secret, as opposed to sharing with the competition. I’m sure you understand.

  35. 44

    “Yes. Are you aware that the USPTO is an imperfect place and sometimes things happen that aren’t exactly 100% in line with the every rule in the big rulebook?”

    Painfully. But what does that have to do with the written description requirement?

  36. 43

    But you are aware that 37 CFR 1.57 expressly addresses the issue of whether the incorporation-by-reference of non-patent literature is sufficient to meet the written description requirement, right?

    Yes. Are you aware that the USPTO is an imperfect place and sometimes things happen that aren’t exactly 100% in line with the every rule in the big rulebook?

  37. 42

    “We need to describe the “problem” in our specification?”

    I know you weren’t asking this for a yes/no answer, so here’s a perspective that might shine a light on things for you. When you describe the problem that the invention is solving, that gives the examiner a fuller understanding and enables them to more comprehensively examine the application. If you don’t give the examiner an understanding, they will not appreciate your “solution” and will have a harder time searching. However, if uncertainty and possible invalidity is what you desire, then a barebones/nothing substantial (read: POS) application maybe be suitable for your needs.

  38. 41

    Michael Risch | Nov 16, 2009 at 09:23 AM

    “Ned Heller –

    Thanks for the thoughtful comments. I left Muncie Gear out for three reasons:
    1. It is a pretty complicated fact pattern to discuss in a short essay
    2. It is about new matter, which I assume in the essay to be non-controversial
    3. The end holding does not support a separate written description requirement, as it is 102(b) type bar that kills the patent for undescribed claims. This implies that if there had been no public use two years before the new claims, then the patent might have been valid. This is not a terribly strong statement in support of a separate requirement.”

    Are you sure new matter was the problem with all the claims? Claims 11 and 13 did not have the apparent new matter problems that claims 12 and 14 had. Claims 11 and 13 were directed to the anti-cavitation plate which was, I believe, fully described in the original specification. Claims 12 and 14 were directed to the new matter — the housing with the “smooth” walls.

    The Supreme Court was careful to state the differences between the claims. Yet it held all four claims barred, even the claims that had a full written description of how to make and use in the original application. The problem with these claims according to the Court was there was no hint in the application as filed that the anti-cavitation plate was the invention of the applicant. (Which is exactly the point you make in your article, btw.)

    Now, I have pointed out there is the statutory requirement of an oath directed at claimed subject matter. The oath is required for an effective filing date. The rules provide for a late claim; but consider the situation where one is never made and the application is abandoned or patented. Legally, that unclaimed subject matter is now abandoned. In the case of a patent, one has a right to file for a reissue, but the issue of a “right to claim” is still presented.

    What of continuations? They themselves can make the oath or claim, but that will be in the continuation. So what is the effective filing date of subject matter unclaimed in any manner in a parent and for which an oath was never filed? Under Section 115, recall, a oath is required for a filing date of an application. Can the effective filing date of subject matter first claimed in a continuation be the filing date of a parent where by statute that subject matter is not entitled to a filing date?

    Now we circle back to Muncie Gear. The Supreme Court held that the amendment with the new matter was effectively a new application. The claim to the anti-cavitation plate was first made then, but it was too late — there was a statutory bar. What this implies is that the effective filing date of disclosed but unclaimed subject matter first claimed in a continuation is the filing date of the continuation. That, I suggest, is the actual holding of Muncie Gear regarding claims 11 and 13.

    Your article makes a lot of very effective points about how a patent application can fully enable an invention but still not provide a date of conception unless the application actually claims the invention in some fashion — either in the claims or in the description. An a good example might involve a range. If the claimed range is between 1-10, with examples at 1, 2, 5, 7 and 10, any example in the prior art of any of these would anticipate the claimed range. One could not subdivide the range into subranges without disclosing some unexpected result in the subrange. That is why one cannot claim a subrange even if the subrange was fully enabled from a make and use point of view. Any description added to support a subrange claim would be “new matter.”

    But is this not generally true of any disclosed but unclaimed subject matter? This is the direct we are heading with the Federal Circuit.

  39. 39

    David French writes:

    I haven’t commented here for a long time but I think the above question is of considerable interest: “Is it possible to CLAIM an invention without identifying what the invention is?”

    The interesting reality is that claims can reach an apparatus that hasn’t been disclosed in the specification. Consider the Wright Brothers’ patent. They disclosed wingtips that were so soft that they could be warped by pulling on cables. The patent was applied against Glenn Curtiss who installed a rigid baby wing intermediate the upper and lower wings, called an “aileron”. The Wright Brothers never thought of an aileron. But the patent was upheld as covering the Glenn Curtiss variant, at least up to a Court of Appeal.

    Conceptually, the patent attorney who took over the Wright Brothers’ patent was able to draft claims which elevated the invention, in the eyes of the courts, to covering the mid-wing aileron. Artfully drafted claims can characterize configurations in such a way that the claims cover things never described in the disclosure. Is this fair?

    The CAFC seems to be heading in the direction of requiring that claims be enabled across the full breadth of their scope of coverage. With this requirement in place, how could a patent claim ever cover an unanticipated future improvement.

  40. 38

    “I could file an application comprising ten of the latest treatises in a particular field and incorporating a few hundred more by reference.”

    Yes, you could. But you are aware that 37 CFR 1.57 expressly addresses the issue of whether the incorporation-by-reference of non-patent literature is sufficient to meet the written description requirement, right?

  41. 37

    Hutz: It’s not hard to imagine situations where an application might disclose all the inventive pieces of a claimed invention, but not describe the invention itself anywhere.

    Indeed, that is the whole point of the written description requirement. I could file an application comprising ten of the latest treatises in a particular field and incorporating a few hundred more by reference. Toss in some boilerplate about combining any these various things to achieve “a device or method that accomplishes more than any of the individual components or steps by themselves.” Then just “amend around the prior art” during prosecution because “there’s an invention in there somewhere.”

    Is this really ‘inventing’? I don’t think so. But it’s an occasionally pursued method of obtaining a patent.

  42. 36

    “6, you sure don’t waste any time do you? You must get Patentlyo on Twitter, right? What a gadfly.”

    No I just check it when I’m bored or watching the news or etc.

  43. 35

    Ned says “not one word of which describes the problem and the solution,”

    We need to describe the “problem” in our specification?

  44. 34

    Luke:

    Re:
    “No. It is certainly possible to “describe” an invention sufficiently well to practice it, without describing it, that is “identifying” it, as to what it “is”.

    The word “description” has many shades of meaning.

    There is a celebrated UK case (or rather, series of cases) on this point regarding EPO (the compound, not the office). The patent described how to make the thing, but didn’t actually identify the thing (as was found out much later) – because, IIRC, a single peptide had inadvertently been lost in the sequencing. ”

    Is it possible to CLAIM an invention without identifying what the invention is?

    I don’t think so. So, assuming the claims identify the invention, and since the claims are part of the specification, why is a redundant “identification” of the invention necessary?

    This is especially, tedious when one considers that inventors are not sure of what they “invented” relative to the prior art. What was “invented” is only determined after or during prosecution. To require that the specification be amended each time a claim is amended, is just so much busy work, wasteful and inefficient. (see how silly redundancy is?)

  45. 33

    MaxDrei, I gather you’re not a US attorney. Your comments about the “best mode” requirement seem a little “off”.
    I’m not sure you’re misunderstanding the requirement, but your comments don’t show that you do understand it.
    The BM requirement requires the inventor to include the best mode of practicing the invention of which he is aware. That is, the patent application is invalid if it leaves out any detail (known to the inventor) of practicing the invention that is superior to the details that are disclosed.
    There is never any duty to label or identify any part of the disclosure as being the “best mode.” The duty is that the inventor must not conceal a better mode than what is disclosed.
    Best regards,
    BM

  46. 31

    breadcrumbs,

    I do believe that your question is quite the succinct counter to the brief defense of the written description requirement.

  47. 30

    “It is unreasonable to expect applicants to anticiapte every possible combination of prior art and every (un)reasonable interpretation thereof before prosecution ensues, and make sure that the claim language that will need to be added by amendment is present in the specification.”

    Is it unreasonable to require them to describe what their invention is at time of filing? Or is that simply not necessary?

  48. 29

    “I just don’t follow that logic. If the Specification and drawings enable the claimed invention, and it is a new and non-obvious invention, how is it possible to say that the inventor did not invent the claimed invention? Maybe someone can give me a persuasive example.”

    I will do so later. I run into this a few times a year. The problem is that the art itself enables a lot of things, therefor, the applicant need not enable those things for them to be “enabled” in their app. So as long as they have some tangential connection with the subject matter they often have a tendency to want to draft a claim that brings in things that the prior art cited against them has in it (and has enabled for them, before they filed) but that they can argue were enabled in their app. True, those features were enabled … by the prior art, not by them.

  49. 28

    Ditto Breadcrumbs.

    This reminds me of a 112 “metes and bounds” rejection I received years ago.

    I argued persuasively that the metes and bounds were clearly defined in the claims, and that the claims formed part of the specification, and that 112 was therefore satisfied.

    For me, the claims answer the question “what” (written description); the rest of the spec answers the questions “how” (enablement) and “why” (utility).

    101, 102, and 103 proceed from the written description, and are not required to be independently declared or proven in the spec.

    Best mode–whatever.

  50. 27

    broje: “From my perspective, it seems like another “gotcha” tool for Examiners and Defendants to get over on applicants and patentees because the precise claim language added or allowed is not present, word for word, in the specification.”

    This sort of thing could be reduced, at least during prosecution, by *explaining* how the spec supports the amendment, rather than including nothing more than a mere sweeping statement about how “support for all amendments can be found in the specification at, e.g., pages 3-8”. Of course, I consider it a good day when a response bothers to even say that much.

  51. 26

    The statute on its face couldn’t be any clearer: enablement is the purpose, and the only purpose, of the written description.

    Consequently, enablement is both necessary and sufficient for written description.

    If you go back to your introductory logic course notes, you’ll recall that that means that if enablement is present, and only if enablement is present, written description is present.

    All of the tautology doesn’t get us anywhere. Never does.

  52. 25

    ******a description of the invention fulfills an important purpose in the patent system. The requirement ensures that the applicant actually invented the claimed subject matter. Reading description out of the statute would allow patent applicants to claim subject matter they did not invent******

    I just don’t follow that logic. If the Specification and drawings enable the claimed invention, and it is a new and non-obvious invention, how is it possible to say that the inventor did not invent the claimed invention? Maybe someone can give me a persuasive example.

    Meanwhile, I think an inventor should be entitled to protection that is valuable in full measure with the value of what they enabled by their disclosure. Otherwise, they are cheated out of something. And the separate written description requirement just seems like a mechanism for cheating inventors out of something of value that they enabled, purely because they did not laboriourlsy lay out every possible variation on their invention, or obvious application thereof.

    From my perspective, it seems like another “gotcha” tool for Examiners and Defendants to get over on applicants and patentees because the precise claim language added or allowed is not present, word for word, in the specification. It is unreasonable to expect applicants to anticiapte every possible combination of prior art and every (un)reasonable interpretation thereof before prosecution ensues, and make sure that the claim language that will need to be added by amendment is present in the specification. If we’re going to be like that, why allow amendment at all, except to place dependent claims in independent form?

  53. 24

    Crackingly good contributions from L Hutz. So, not only does the optimally drafted app refrain from saying what the invention is. It also deliberately obscures what is its Best Mode.

    No wonder the PTO is frustrated. No wonder the worm is turning. Come on judges: do your duty here.

  54. 23

    I believe this issue arises during prosecution. There are cases where the claims that issue do not appear to recite any invention described in the application, but are still fully enabled by the description. These usually crop up during prosecution as art is found and the claims are amended to cover what is not disclosed.

    It’s not hard to imagine situations where an application might disclose all the inventive pieces of a claimed invention, but not describe the invention itself anywhere.

  55. 22

    Max,

    No drafter of any skill writes 50 pages on the best mode. The US does not require an applicant to describe any embodiment AS the best mode and depending on the invention, it would be foolish to do so. I would imagine in cases where it seems that way, it was probably because (1) that was the only mode the inventor thought of, and (2) the patent agent/attorney did not know of any way to broaden the description.

    Generally, al the elements of the best mode should be included, but a good drafter avoids the metaphorical neon sign saying “Best mode” when including those elements.

  56. 21

    Oh I hope so, EG. I do hope so. What fun for lawyers that would be. And what pointless complexity. And isn’t that the same thing?

  57. 20

    Michael,

    A question: why is paragraph 1 of 35 USC 112 judged against a PHOSITA? Paragraph 1 says “one skilled in the art. Also, I know that paragraph 2 of 35 USC 112 is judged against a PHOSITA, even thought it doesn’t use the PHOSITA language. Should there be any difference here between “one skilled in the art” and a PHOSITA?

  58. 19

    MD, good point. I hadn’t considered that.

    “embrace – include in scope; include as part of something broader”

    That definition appears to support your position. This is interesting, but I’ll have to come back to it later.

  59. 18

    Jules what does “embraced” mean? Is this why the originally filed independent claims are invariably drafted wide enough to embrace every second thing under the sun made by man? Whatever amendment you make during prosecution, you are always going to be able to say “Where’s the problem? My original claim “embraced” it, and my amended claim still “embraces” the 50 page Best Mode stuff”. This is best practice benchmark stuff today, isn’t it? Mr Slusky?

  60. 17

    Ned Heller –

    Thanks for the thoughtful comments. I left Muncie Gear out for three reasons:
    1. It is a pretty complicated fact pattern to discuss in a short essay
    2. It is about new matter, which I assume in the essay to be non-controversial
    3. The end holding does not support a separate written description requirement, as it is 102(b) type bar that kills the patent for undescribed claims. This implies that if there had been no public use two years before the new claims, then the patent might have been valid. This is not a terribly strong statement in support of a separate requirement.

  61. 16

    Ned: “But the problem comes in with later-filed claims — If they now claim previously disclosed but unclaimed subject matter, what then? What about the requirement of an Oath in the statute? Are we to just ignore that?”

    Are you referring to the requirement for a supplemental oath when claims are filed that were disclosed but not originally claimed? 37 CFR 1.67(b).

    “(b) A supplemental oath or declaration meeting the requirements of § 1.63 must be filed when a claim is presented for matter originally shown or described but not substantially embraced in the statement of invention …”

    Not sure why the office doesn’t enforce that one, but something tells me it is because “statement of invention” is not clear. The word “must” is pretty clear though.

  62. 15

    MaxDrei, I think you disagree with Dennis.

    I like the article. If you have 50 pages of description, not one word of which describes the problem and the solution, but just describes a number of embodiments, the novel features of which we are left to identify later as the prosecution develops, just what the Hades is this?

    But that is the practice today in so many applications. But it does seem to fly in the face of Section 112, p. 1, and of Section 115.

  63. 14

    Dennis S writes:

    “I think any standard for 112 that declares subject matter to be enabled but not described is problematic.”

    Too right. There ought to be more to good drafting than writing 50 pages on the Best Mode, capped off by a ludicrously broad independent claim. Neither of these deliberate elements of a current, optimally-drafted, US-style app even begins to state what the inventor’s contribution, or inventive concept, “is”. No wonder the USPTO can’t do a half decent search. Never mind though, the courts are bent on putting the situation to rights, using the “written description” stick to beat recalcitrant drafters with.

  64. 12

    Posted by: breadcrumbs | Nov 16, 2009 at 08:04 AM “Don’t claims “identify” the invention?”

    Yes they do. If the application as filed claims an invention enabled but not identified as being an invention of the applicants, I think that is enough.

    But the problem comes in with later-filed claims — If they now claim previously disclosed but unclaimed subject matter, what then? What about the requirement of an Oath in the statute? Are we to just ignore that?

  65. 11

    Very little analysis of the Supreme Court cases. No mention of Muncie Gear, which was based upon a lack of written description for two of the four claims held invalid that claimed previously disclosed subject matter but where there was NO indication that is was part of the invention of the applicant until was claimed years later. This is very consistent with the argument of the author here. It is remarkable that he failed to even mention the case.

    The way written description is emerging at the Fed. Cir. is places it four squares with the Supreme Court. But, until some paper actually looks at the Supreme Court cases, it will remain unconvincing even if I would otherwise agree with it.

  66. 10

    “The [separate written description] requirement ensures that the applicant actually invented the claimed subject matter. Reading description out of the statute would allow patent applicants to claim subject matter they did not invent, and would effectively rewrite nearly 120 years of precedent about the conception of inventions.” “…the law has never required inventors to discover each and every embodiment of the broad principles they claim. The question should be whether Ariad described the broad principles of its invention sufficiently to show that its conception was complete at the time of filing.”

    Although the debate is being defined by most as an issue with the written description requirement, I think an alternative way to address the author’s concerns is to call into question the present interpretation and application of the enablement requirement. If at present the enablement requirement can be satisfied for subject matter that is not adequately described (i.e., identified) as long as ‘enabling’ experimentation that allows a POSITA to then identify the subject matter is disclosed, then maybe the enablement standard needs revisited. If the enablement requirement were interpreted more narrowly, the process of experimentation that is enabled could be patented, but not the subject matter that is identified by using the process of experimentation.

    From another angle, if rather than claiming in a first application some subject matter that was enabled but arguably not described, the inventor instead later files a second new (non-continuation) application which adds no more enablement but does add further description directed to the subject matter, wouldn’t the inventor surely be met by an obviousness rejection of the subject matter claimed in the second application over the subject matter ‘enabled’ by the first application? And if so, is that a desirable result?

    I think any standard for 112 that declares subject matter to be enabled but not described is problematic.

  67. 8

    Luke, is it the EPO cases in which Lord Hoffmann emphasises the distinction between what a document “discloses” and what it “enables”, in the context of whether that document deprives a later-dated claim of novelty?

    First you have to decide whether the earlier doc discloses any subject matter within the ambit of the claim. Then you must ascertain whether that disclosure is an enabling disclosure.

    At least in Europe, the concept of “disclosure” is the same, when deciding whether, in the app as filed, there is a written description of the claimed subject matter.

    But I haven’t yet read Risch. Must do so now.

  68. 7

    Steve M wrote:

    “Curious. So some believe it’s possible for the PHOSITA to be shown how to practice an invention…w/out concurrently identifying what the invention actually is?

    At the very least, it would seem–at least implicitly even if not explicitly stated–that an enabled invention is also one which has revealed by its very enablement…what it is.

    No?”

    No. It is certainly possible to “describe” an invention sufficiently well to practice it, without describing it, that is “identifying” it, as to what it “is”.

    The word “description” has many shades of meaning.

    There is a celebrated UK case (or rather, series of cases) on this point regarding EPO (the compound, not the office). The patent described how to make the thing, but didn’t actually identify the thing (as was found out much later) – because, IIRC, a single peptide had inadvertently been lost in the sequencing.

    Cheers, Luke

  69. 6

    “…even if the specification enables a PHOSITA to practice the invention, the applicant is not relieved of the obligation to identify the invention.”

    Curious. So some believe it’s possible for the PHOSITA to be shown how to practice an invention…w/out concurrently identifying what the invention actually is?

    At the very least, it would seem–at least implicitly even if not explicitly stated–that an enabled invention is also one which has revealed by its very enablement…what it is.

    No?

  70. 5

    When did a court of appeals first recognize the written description requirement as being separate from the enablement requirement?

    What was the case?

    Thanks

  71. 4

    “The requirement ensures that the applicant actually invented the claimed subject matter.”

    Does the application declaration do that?

  72. 2

    “Finally, the essay considers the Ariad case and concludes, perhaps surprisingly, that under the vision of written description presented in this essay the claims at issue may well be described.”

    Actually that is no surprise at all. If there is merely a WD requirement, (a “small” one let’s say) then there patent is in the clear iirc. It’s only if the Fed Circ goes all out and requires them to go above and beyond, like the Fed Circ has, in cases of large genus, that their case is doomed.

  73. 1

    “However, while enablement is necessary to satisfy Section 112, it is not sufficient. The statutory language certainly requires that an applicant identify the invention sufficiently to allow a PHOSITA to make and use it. However, even if the specification enables a PHOSITA to practice the invention, the applicant is not relieved of the obligation to identify the invention. Instead, the specification must contain a “description of the invention” even if such description would only serve to reinforce other parts of the disclosure that enable one to make and use the invention.”

    Sounds like someone has been reading some of my comments.

    Also D I think that there is a chance you are misinterpreting the debate amongst the parties in the case. It is my belief that both sides accept that there is a WD requirement in addition to the enablement requirement. However, where the party disagrees is as to whether or not there is a “separate” WD requirement, which I believe they are taking to mean an additional WD requirement piled on top of the WD requirement that they acknowledge to exist. That “separate” WD requirement is the one that requires you to go above and beyond in cases like the hugely broad genus cases.

    The reason I mostly believe this is the case is because of their first few pages of their brief. Also, they’d be nuts to try to get rid of WD as a whole. There is a 0% chance for En banc Fed circuit to remove the basis for new matter rejections, and there is a 0% chance that they’ll be able to switch the basis for new matter rejections off to the enablement requirement.

    Gl to them.

Comments are closed.