An Empirical Study of the Role of The Written Description Requirement in Patent Prosecution

Table 3[Download the Draft Essay]

Essay Overview: In the pending case of Ariad v. Eli Lilly, an en banc Federal Circuit is considering whether Section 112 of the Patent Act as properly interpreted includes a written description requirement that is separate and distinct from the enablement requirement. Although the USPTO has no direct role in the infringement dispute, the government submitted an amicus curiae brief arguing that a separate written description requirement is “necessary to permit the USPTO to perform its basic examination function.” However, when pressed during oral arguments the government could not point to any direct evidence supporting its contention.

This essay presents the results of a retrospective empirical study of the role of the written description requirement in patent office practice. It is narrowly focused on rebutting the USPTO’s claim that the separate written description requirement serves an important role in the patent examination process. To the contrary, my results support the conclusion suggested by Chief Judge Michel during oral arguments that it is indeed “exceedingly rare that the patent office hangs its case on written description.”

For the study, I analyzed 2858 Board of Patent Appeals and Interference (BPAI) patent opinions decided January-June 2009. Written description issues were decided in 123 (4.3%) of the decisions in my sample. Perhaps surprisingly, I found that none of the outcomes of those decisions would have been impacted by a legal change that entirely eliminated the written description requirement of Section 112 so long as the USPTO would still be allowed to reject claims based on the addition of “new matter” (perhaps under 35 U.S.C. Section 132). New-matter style written description requirement rejections were outcome-determinative in 20 of the 2858 cases – about 1.0% of the cases in my sample. (I am very confident that the PTO will retain its ability to make new matter rejections even if the separate written description requirement is eliminated.)

Although there may be valid reasons for retaining a separate written description requirement, this study safely leads to the conclusion that the government’s conclusory statements regarding the doctrine’s critical importance for patent examination lack a factual basis.

Continue reading the essay. [PDF]

These results fit well with those of UMKC professor Christopher Holman that he reported in his 2007 article, Is Lilly Written Description a Paper Tiger?: A Comprehensive Assessment of the Impact of Eli Lilly and its Progeny in the Courts and PTO , 17 ALB. L.J. SCI. & TECH. 1, 62 (2007).

About Dennis Crouch

Law Professor at the University of Missouri School of Law. Co-director of the Center for Intellectual Property and Entrepreneurship.

68 thoughts on “An Empirical Study of the Role of The Written Description Requirement in Patent Prosecution

  1. NWPA,

    “Each amendment if not literally supported goes to the patentability of the invention.”

    Yes – but you are caught up in a point that screams – so what?

    As I have posited, most amendments go to the patentability of the invention and most do not add new matter. This is true of amendments not literally supported just as it is true of those that are literally supported. Yes it is more difficult to establish antecedent basis for a non-literal support. The crux is still – does the application as filed support the amendment or not. This question applies whether patentability is at issue or not.

    The only procedure to follow is to understand the application as filed.

    As to your examiner proclaiming something new matter when it is not, this is yet another (possible) symptom of poor examination (to give some small leeway to IANAE, there is a very real possibility that the prosecutor does not understand the application as filed). The proper tools are in place – there is a rule that covers new matter – anything more must fall to the particulars of a case by case analysis, and no rule will help you there.

    IANAE,

    I do not doubt that in the current “professional” culture (as witnessed by several of the vocal posters here at the Trainwreck), that if you give the examiner all day, he will still make the rejection the MPEP says is appropriate for the problem he (thinks) he has found. You, once again, drift off into the ideal world that you like to inhabit and ignore the very real problem of poor examination. It does exist. Ignoring the problem will not make it go away.

    You also, once again, fail to fully read and comprehend what I posted. “Further, just because the Office has a habit of not calling out 132 and using 112 in its place neither makes that practice correct nor justified. – Posted by: Noise above Law | Feb 23, 2010 at 08:08 AM”

    Just because the MPEP tells the examiner to do it, doesn’t make the problem go away, nor does it make the process correct unto the law. Also, you are ignoring the fact that the problem NWPA and I are discussing centers around poor examination and the lack of understanding of that application – so the answer to your question of why do I have to drag the discussion back to my pet issue is because the two are causally related. Of course – in your “ideal examiner” world, this does not happen. WAKE UP.

    You also may have missed my earlier post concerning the quasi-widget solution. I have alluded to this for you once before, but you have again glibly passed over it to spout on something I have not said (I never indicated that unlimited or non-limited time is a solution). The possible solution that I shared was to have each art unit’s brightest and best, those most knowledgable about the art form a review committee and review the applications and appropriately rate them for various levels of time permitted. In this manner, “easy” applications gete less time and more “difficult” applications get more time – each not sacrificing to the mythical quasi-average universal widget.

    get it. got it? good.

  2. Let me repeat, then, that the quasi-average widget treatment of applications needs to be eradicated.

    Why do you have to drag the discussion back to your pet issue? This is clearly nothing to do with the time allocated to examiners for examining. It doesn’t take any longer to type “132” than it does to type “112”. It’s right in the MPEP – new matter in the spec is to be rejected under 132 (new matter introduced to the disclosure), new matter in the claims is to be rejected under 112 (not writtenly described in original specification).

    Give the examiner all day, and he’ll still make the rejection the MPEP says is appropriate for the problem he found.

  3. Well, NAL, I don’t disagree with most of what you say, but I don’t think new matter is easy to police. Each amendment if not literally supported goes to the patentability of the invention.

    I think with new matter the problem is the fact there really isn’t a good procedure for the examiners to follow.

    And, I don’t think there is agreement on what constitutes new matter. This cuts both ways where I have examiners refuse to enter my claim amendments when a PHOSITA would implement a function in the specification with the structure of my claim amendment. It really is a difficult problem with no good set of rules for an examiner to follow.

    So, here, we have a problem that needs a solution, but as you say it is probably eclipsed by the greater problems of examination.

  4. NWPA,

    If I can paraphrase what I think you are saying, and then add my reply:

    Amendments to claims can contain new matter for patentability reasons.

    My point on this was that this alone was not necessarily a cause for concern – most amendments are made for patentability reasons and most amendments do not introduce new matter.

    Given that a certain number of claim amendments contain new matter, how is the Office best equipped to police this abrogation of the rules?

    The abrogation is specifically 132 – NOT 112.

    The answer: Face the difficulty head on. The problem is not whether a claim is amended with new matter – per se. The problem is that the examiner does not understand the application enough to distinguish what is new matter from what is not new matter. Thus, when an amendment IS made that contains new matter, it may escape the examiner’s grasp of the invention. I believe that you are focusing on yet another symptom of poor examination – not a different disease. A rejection under 132 would be the proper policing of new matter – the tool is there and it is sufficent for the task. It is in the implementation of that tool that we have the breakdown. This is not something that should be fixed by introducing another rule. It should be fixed by a more arduous and carfeul examination.

    Let me repeat, then, that the quasi-average widget treatment of applications needs to be eradicated. Each application should be examined on its own merits, not force-fed into some predetermined generic time slot. Fix the root cause of the problem, and the symptoms will take care of themselves.

  5. And, one may ask if a claim amendment is not literally supported in the claim can one then argue that amendment as to the patentability of the claim?

  6. I guess a point I am making here is that I have seen many claims that have had what I think are new matter added to them. In litigation these claims are then challenged based 112 written description. Basically, what you see is a patent attorney claim a new invention that is usually better and broader than the invention described in the application.

  7. >>not calling out 132 and using 112

    Yes, I agree that the written description requirement has been fabricated, but the reason it has been fabricated is that in U.S. practice there isn’t currently a good way to police new matter.

    I think if anyone simple imagines the arguments that they can make (and I have made) to argue that something isn’t new matter, then you see that the examiner is really arguing whether the amendment adds anything to the patentability of the claim. (except for of course ridiculous amendments that have nothing to do with the application.)

  8. “…if the 112 objections don’t come from US law, and they don’t come from USPTO training as you know it, what’s your explanation of where all those PTO Examiners picked up their rash?”

    The same way most rashes are spread – from the examiner in the next cubicle, or in the gym, or via the water cooler. The origin is a little harder to pin down – my best guess is an over-creative examiner (6?) who thinks he understands caselaw, and who was told by an SPE to figure out a way to reject that claim, as it’s clearly too short to be allowable.

    I’m no epidemiologist, but I think it might be interesting to study these things. Even from the vantage point of a small firm, we can see the diseases (such as the grossly incorrect paraphrasings of Bilski – there have been two or three varieties) spread, reach a peak, and then fade away. I suspect the rash of spurious 112 rejections described above is no different. But, on the other hand, it could represent the guerilla tactics of a small cadre of EPC-inspired USPTO trainers.

  9. NWPA,

    I think that you are onto something with both the difficulty and the connection to patentability.

    However, it seems we have a bit of the forest and trees going on. You can clearly add patentability to a claim with an amendment without adding new matter. That is one reason why examiners MUST examine both claims and the specification.

    The question, then, is whether the amendment can find antecedent basis in the specification as filed. The tricky part is that this basis can be explicit or inherent – it need not be a word-for-word exacting replication.

    Regardless of the difficulty, and as I have gone into in much greater detail on other threads, forcing even a “good rule” into 112 paragraph one is still misreading the plain language of that rule. There is indeed another section that deals with added matter. Further, just because the Office has a habit of not calling out 132 and using 112 in its place neither makes that practice correct nor justified.

  10. >>between “explicit literal support” and “very >>difficult to answer”.

    That is a good answer and it strikes to the heart of the problem. That when one is enforcing new matter that it really becomes a question of patentability under 102 and 103. If there isn’t literal support then the question has to be asked what does this add to the claim? If it adds some measure of patentability, then it is new matter, if not, then it is not. So, as claims are considered as a whole, the problem is that then the examiner is in a position of allowing in new matter or not based on a judgement of patentability.

    I think the test that is used now isn’t a bad one. Were they in possession of this invention? And if so, then is this the way one would implement that invention.

    I hope everyone can see that this is really a very difficult problem that goes to the heart of patentability. Consider fiber optic cable. A claim would have been enough to enable a PHOSITA to make the cable. And there is nothing novel about any of the structure independently. It is only the combination of the structure and combining the structure is within the ability of a PHOSITA.

    So, imagine a claim that starts off as a simple cable that the patent attorney slowly amends into a fiber optic cable. Which of those amendments adds new matter?

    I think anyone who thinks this is easy to evaluate is kidding themselves. Everyone should be able to see that new matter that goes beyond literalism goes right to the heart of patentability.

    Examiners and SPEs often allow in amendemnts that are highly questionable and others that are clearly new matter.

  11. “what’s your explanation of where all those PTO Examiners picked up their rash?”

    Already asked and answered MaxDrei – incompetence.

    As to my reminding you, we will have to disagree. If it wasn’t actually needed, you wouldn’t keep making the same mistake. Let’s make a deal – you stop conflating law regimes, and I will stop reminding you that you are conflating law regimes – ok?

    Engage your critical faculties, please.

  12. Well, Hobbes, do forgive me if I’m falsely attributing stuff to you, but if the 112 objections don’t come from US law, and they don’t come from USPTO training as you know it, what’s your explanation of where all those PTO Examiners picked up their rash?

    Fight fire with fire, I say. If those Exrs are using tools from the EPO toolbox, learn how to use those same tools, but use them better than the Examiners are doing.

    Or does that not work either?

  13. My dear Max, you’re putting words in my mouth. I’d be quite happy to import the EPO’s training regime (and overall professionalism, for that matter) to the US, with the necessary adaptations for US law. I just think it’s funny that you spot EPO influence in a rash of Section 112 rejections that are based on neither USPTO training nor US law.

  14. Noise, I do understand that the law under which the USPTO and the EPO operate is different, and I’m sure that other readers perceive that I do. Your persistently “reminding” me that US patent law is different from that in Europe might be a comfort for you but it’s not needed, thank you.

    But now to another thing: your prediction of “unmitigated disaster”. Which do you think would create the bigger disaster, passing on to USPTO newbies the benefit of EPO experience in managing prosecution expeditiously, or teaching EPO newbies how to do it, USPTO-style? Hobbes (he cites “incompetence”) seems to think it would be the latter. But you would disagree, right?

  15. “…but that the USPTO trainers are greedily eying the EPO ability to manage its workload, pondering how that is done, and ending up borrowing tools from the EPO examination toolbox (the EPO’s Guidelines book).”

    That’s pretty funny, Max. You really think there are many USPTO trainers that have actually read the EPO’s Guidelines book? I bet there’s no more than, say, two. And that’s a stretch.

    Respectfully, I’m going to stick with incompetence as the most plausible explanation. But please let us know if you spot any more reflections of the EPO in the systematic disfunction of an overloaded, underfunded, US government bureaucracy!

  16. …very interesting concept MaxDrei.

    Of course, it could be a possible source of failure is the difference in Law. And since I remind you of this fairly often, it perhaps would be an unmitigated disaster to let the EPO trainers loose without a full appreciation of this very fact.

  17. Noise, Hobbes, your thoughts are welcome. What I am postulating is NOT that the USPTO trainees are avidly following EPO practice but that the USPTO trainers are greedily eying the EPO ability to manage its workload, pondering how that is done, and ending up borrowing tools from the EPO examination toolbox (the EPO’s Guidelines book).

    I say this because of the feeling of deja vu which assails me, whenever I follow threads in which US attorneys complain about new forms of objection launched by PTO Examiners.

    Only problem is, when the USPTO trainers lack any real understanding how those EPO tools really work, their teaching of USPTO newbies how to use them is a failure.

    Now if the EPO trainers were themselves to be given direct access, to teach USPTO newbies……

  18. Hobbes you set up a false distinction. I suspect that the answer is both of yours, namely that new USPTO Examiners, in their inadequate training, are subjected to EPO-originating thinking.

    Interesting. Incompetence in training in the USPTO leads to EPO-like thinking. Is that supposed to say positive things about the EPO approach? Perhaps you should dismantle your training over there, since you’ll get the same results.

  19. I think once you go outside of explicit literal support that the questions becomes very difficult to answer

    I think there’s a lot of leeway between “explicit literal support” and “very difficult to answer”. Don’t forget, we’re already in the business of determining whether claimed features are taught in a specification that doesn’t use the exact same terminology.

  20. I think once you go outside of explicit literal support that the questions becomes very difficult to answer and the current solution is probably the right one.

  21. Well, the point is what test is used for what is traced to the specification.

    Add amendment x to claim A. How does one tell if it is new matter or not?

  22. NWPA,

    Wrong answer – that would be a new matter rejection – independent of WD.

    Here’s a clue – new matter is that which cannot be traced to something in the specification as filed (your “or not” phrase cannot follow).

    MaxDrei,

    Mr. Hobbes has not set up a false distinction. While you may be in a process of augmenting your worldly knowledge, you will have to keep in mind that the ugly American more or less thinks the world revovles around the US and really does not care about EP prosecution tactics. It is very much in doubt that the US examiner would be ardently following EP philosophies.

  23. Hobbes you set up a false distinction. I suspect that the answer is both of yours, namely that new USPTO Examiners, in their inadequate training, are subjected to EPO-originating thinking.

  24. >>Why is enablement not enough (given that there >>is an explicit requirement that outlaws new >>matter already – and that is outside of WD)?

    Because I can write a claim amendment that would enable a PHOSITA to build my invention without anything from the specification whether it was in the specification or not(outside of the claims.)

  25. Ned,

    Why is enablement not enough (given that there is an explicit requirement that outlaws new matter already – and that is outside of WD)?

  26. Night Writer, I think you have this “right.” If I claim matter that is not literally described in the spec., and which would be new matter if amended into the spec., the issue is “written description.”

    The Euro’s always have the “amendment” issue as they require the broad claims to be set forth in the spec.

    Enablement is not enough to police this issue.

  27. I think the problem is how do you police new matter without a second written description? It isn’t an easy question. And I don’t think you really can. The only real way to police new matter is to say that if it isn’t described in the application then it is new matter.

  28. “Coke, it looks to me like over-enthusiastic application of EPO examining guidelines.”

    Or it could be an examiner with poor training in the proper application of US examining guidelines.

    Max, which of these do you think is the more likely explanation, given that there are probably no more than a few examiners in the USPTO who know the first thing about EPO examining guidelines?

  29. “species b is a separately patentable selection from the known genus.” – I assume you mean from the newly invented and now known genus…

    that about gives you your answer – no?

    btw – I like the “laterer”.

    However, your hypo still does not have enough details. For example, is the “A” application still “in process”? If so, then Applicant A can amend to include the separately patentable selection (there may be restriction issues – hard to tell without more details). Is the “A” application a patent within the two year limit?

    I will state that you are incorrect when you posit that “Applicant A did not invent the selection of species b until after Applicant B’s filing” – if the selection is in the “A” application and is, as you put it, disclosed by A, then without further detail to the contrary, A has invented the species (this also assumes the non-excessive experimentation rule applies). Applicant B would be rightly rejected by the material from Application A (note that if Applicant A does not claim – directly as would be needed in your hypo since you state that “B” is separably patentable – either within the life of his application, or within two years of patent – with a broadening re-issue – the separable invention “B” is dedicated to the public).

    In no case, though did B file before A invented, as the initial filing was prior to B. In no case would B get the patent.

  30. While my posts can be too long – your missive above is too short. I cannot figure out what you are trying to say.

    Okay, let’s try again.

    Suppose arguendo that the genus is a patentable invention, and species b is a separately patentable selection from the known genus.

    Applicant A files for the genus, claims the genus, discloses a long list of species including species b.

    Applicant B later files for species b.

    Applicant A laterer claims species b.

    I would reject (because everybody here knows I’m secretly an examiner) based on the theory that Applicant A did not invent the selection of species b until after Applicant B’s filing (and probably until after B’s publication). Under the same theory that entitles Applicant B to his patent, Applicant A’s disclosure of species b does not show possession of the invention of species b. Put another way, A’s date of invention is later than his filing date.

    Apply 101 (A is not the inventor) or 102(a) (B filed before A invented), for example.

  31. NAL, I was merely thinking that the test whether a prosecution amendment breaches the “written description” requirement is a novelty rather than an obviousness test. But never mind. It’s not important is it?

  32. IANAE,

    While my posts can be too long – your missive above is too short. I cannot figure out what you are trying to say.

    Diet Coke,

    I see your frustration – the examiner is plainly wrong. Exact wording has never been a requirement in the US. Have you tried calling his SPE or proceeding to Appeal?

    MaxDrei,

    “does the generic ground the specific?…” I am going to give you the very hated (but very correct) answer of “it depends”.

    There is no universal answer that covers every generic/specific and specific/generic possibility. Herein lies the unintended consequence of the Supreme’s KSR decision.

    Granting PHOSITA near omniscient power of recall, not only in the direct art unit, but in ANY art unit that may provide an answer likewise empowers PHOSITA in the realm of what an applicant must provide in order to meet 112.

    That same near omniscient power can be used by the applicant, just as it can be used by the examiner.

    Keep in mind that my comments relate to US law.

    The requirement of what must be presented in an application excludes that which the PHOSITA may know – and exclusion includes that which may be known through any experimentation that is not excessive. While the holding in KSR grants the Office greater power to reject claims that are obvious under this rationale (one side of the sword), this power also broadens extensively that which falls to an inventor in presenting a generic and capturing specifics (the other edge of the sword). The tool necessarily takes and gives. You gotta love the law of unintended consequences.

  33. Coke, it looks to me like over-enthusiastic application of EPO examining guidelines.

    It is well-understood in EPO practice that you can’t cherry pick during prosecution, to get to an originally untaught intermediate level of generality. The claimed amended combination of technical features has to be there, in the app as filed, and furthermore at that level of generality. But, in your case, it seems that you could indeed satisfy an EPO Exr on that issue.

  34. The “First to File” is a red herring

    The selection counts as an invention, right? What if the first filer didn’t “invent” the selection until after the second filer’s application was published?

    Presumably an item appearing in a laundry list isn’t enough to establish possession of an invention when that invention is essentially picking one thing out of a laundry list.

  35. Max and Noise–I’m having trouble explaining this, but to clarify–the rejections I’m seeing don’t have to do with new matter (although I can easily seeing the Examiner saying that they do). The support is there in the application; the claimed invention is clearly described so to meet all the 112 tests, it’s just in different phrasing than in the claims. Often this is because different portions of the specification describe different aspects of the invention–for example, a method has several steps and each step is discussed in detail in successive portions of the specification. Then when we present a claim to the method using terminology from these portions of the specification, the Examiner says that the different portions of the specification do not relate to each other despite our setting out in elaborate detail the language that does in fact explain that they do (going all the way back to the earliest-filed priority application). So there’s no new matter problem, or possession problem, or anything else–except that the Examiner maintains the rejections. Very frustrating.

  36. Does “grounded” mean “directly and unambiguously derivable”? If so, Noise, then there is no issue between us. But if it doesn’t, there is.

    I think I know what “directly and unambiguously derivable (by the PHOSITA)” means. I don’t know though what “grounded” means. For example, does the generic ground the specific? Does the specific give a ground for the generic? Or must the grounding in the earlier document be at the “same” level of generality as the statement in the later document that one asserts to have been “grounded” in the earlier document?

    At least in FtF country, these things matter a lot. I suspect they matter just as much (if not more) in FtI country.

  37. “most of the outstanding rejections are written description based on the *exact* language in the claim (i.e., word-for-word as in the claim in its entirety) not being set forth verbatim in the spec.”

    Unlike, perhaps, the EPO, does the US actually require a verbatim antecedent basis for amended claims to the specification?

    I’m pretty sure the answer is no.

    If the amended claim cannot be grounded in the specification as filed, you have new matter – plain and simple. If the amended claims can be grounded in the specification, then the amended claim (even word for word from a later filed application) rightly belongs to the earlier inventor.

    The “First to File” is a red herring – MaxDrei, you will confuse yourself and others by trying to conflate different legal settings into this argument.

  38. Dennis, does your study include Hyatt v. Kappos (2010), Hyatt v. Doll (2009), Hyatt v. Doll (2008), or Hyatt v. Dudas (2007), all of which deal exclusively with written description (new matter) issues? BTW, the Chief Judge wrote the opinion in Hyatt v. Dudas (published at 492 F3d 1365, 83 USPQ2d 1373).

    It seems me that the biggest practical effect of the written description requirement (from the PTO’s point of view) is that an applicant cannot compel the Office to perform a never-ending series of prior art searches in a single application. The applicant can only compel the Office to devote its (vast but limited) resources to performing prior art searches of inventions the applicant had “possession” of at the time he filed that particular application.

  39. “Inviting Body Punches” — I would like to discuss your administrative law issue from earlier in the week. Please contact me. David Boundy DBoundy [at] Cantor [dot] com.

  40. Why the “exact” laguage requirement at the EPO? Because the app is in English but 90% of EPO Examiners have as their first language one that is not English. Imagine you are examining at the JPO. Is Applicant using his prosecution amendment to add matter? Can you tell? If you’re not sure, insist on verbatim support in the app as filed.

    Then there’s this pesky thing called “First to File”. Would you allow an earlier filer to write into his app, during prosecution, claims directed to the valuable selection invention which is the property of another with a later filing (and A publication) date, albeit the FIRST in respect of that precise, and commercially valuable, inventive selection? A TAXOL-coated stent, for example.

  41. Night Writer Patent Attorney said “In many ways I think that literalism of the EPO is better.” My apps are all biotech, and I think we’re there (at EPO literalism)–most of the outstanding rejections are written description based on the *exact* language in the claim (i.e., word-for-word as in the claim in its entirety) not being set forth verbatim in the spec. And if they get rid of the “written description” requirement, I fully expect to see the same rejections but cast as “new matter” rejections, on exactly the same basis.

  42. I do not see how a written description requirement separate from a requirement for “written description so as to enable” aids or is essential to the Patent Office’s examination function. Of course I agree that the “new matter” or “language support” aspect of written description is critical to examination, but this is not the same as Eli Lilly written description. So to the extent that “new matter” issues for claim language are called or fall under 112, first paragraph written description, then that can be a separate requirement.

    But I do not see how the Eli Lilly written description requirement is important to the Patent Office’s examination function. First, the enablement requirement provides all of the tools the Patent Office needs to prevent overclaiming. If the applicant has enabled a certain scope of protection, then there is no need for the Patent Office to separately inquire whether there is some alleged deficit in the scope of the “description” of the invention. What important examination purpose is served (and, for that matter, what important policy is served) by such a separate requirement, other than to deny an applicant patent protection of scope of an invention that has been fully enabled (and thus placed in the hands of the public)?

    Second, as to the significance of Patent Office use of a separate, Eli Lilly written description requirement, it is clearly being used by the Office as a “super-enablement” requirement for biotechnology inventions. Claims that are acknowledged to be fully enabled are being rejected as lacking adequate written description (usually in biotechnology and usually because of an alleged lack of sufficient description of the structure of biological molecules). However, this does not establish that this use of the Eli Lilly written description requirement is important for the Patent Office’s examination function. It is at best circular argument to say that because the Patent Office makes Eli Lilly written description rejections it is important to the Patent Office’s examination function. On the contrary, all it really says is that if Eli Lilly written description is in fact a valid legal requirement for patentability, then it is important to the Patent Office’s examination function to make rejections based on it. But if the Eli Lilly written description requirement is not a valid legal requirement for patentability, then the Eli Lilly written description requirement not only is not important to the Patent Office’s examination function, it is irrelevant to examination altogether.

    Thus, the Patent Office’s examination function is irrelevant to the issue of a separate written description requirement raised in the Ariad case. Rather, the outcome of the Ariad case will determine whether the Patent Office must continue or abandon application of an Eli Lilly written description requirement in examining patent applications.

  43. NWPA – keep in mind that the claims are part of the specification and can be self-enabling (and self-described).

    Another thing to keep in mind is that PHOSITA is a double edged sword. KSR may have bulked up the power of PHOSITA (for example, ANY field of art can be combined if related to a solution) – but that same greater power applies to what applicants are required to present – Applicants are expressly told NOT to include that which a PHOSITA may know. Self-enabling claims that meet the other requirements of the Law simply are patentable – this may indeed be a change from the “old school” where the order of the day was a 75 page specification, but that’s where we are.

  44. In many ways I think that literalism of the EPO is better. It would help promote the progress of technology if there were more certainty in claim construction and I think that the written description requirement helps to nail down the claims. I think it might even be better to have to put in references to the structure of each claim element during prosecution.

  45. I still haven’t completely made up my mind about the third part of the written description requirement. I can say that in litigation that it can be helpful in that in construing claims the other side has to be able to make an argument that the invention they are trying to sell to the court is described in the application. The problem is that one can construe claims very broadly have them still enabled because of the wealth of knowledge one skilled in the art is assumed to pocess, and yet the invention is not really described anywhere in the application but in the construction of the claims.

  46. I’m not sure that I see what that selection bias has to do with anything. Sorry, I just really do not.

    If Linn’s point was that this particular nuance of stat construction might only affect an exceedingly rare case then I agree, but that depends on how the court tailors its decision.

    The fact of the matter is that there are some out there that want the interpretation of 112 to swing wildly in the opposite direction to where it has been for quite some time, further than just this little “nuance” would dictate. There appeared to be a company called Ariad among those people before they did an about face right before their court date.

    I personally don’t care about the specific nuance, aka the “super-enablement” requirement, except that it does make sense to me and does seem reasonable to me. And this isn’t because I want to screw anyone over either. Your claim should be fully enabled either at the time of filing, or by the app itself. Seriously, that isn’t really too much to ask.

  47. From the Holman Abstract: To this day, the conventional wisdom is that the so-called Lilly written description requirement (LWD) exists as a biotechnology-specific super-enablement requirement, substantially more stringent than the enablement requirement (the conventional standard for patentability), and standing as an impediment to effective patent protection for biotechnology inventions.

    Say wha? Most chem/bio biotech prosecutors I know don’t see it that way. It’s not an impediment at all to effectively protecting the embodiments which were reduced to practice because we describe those embodiments in detail regardless. It is an “impediment” to obtaining protection for obtaining scope beyond that, but only insofar as it forces applicants to bulk up their specifications with express support for that additional scope. More of an inconvenience than an impediment.

    The enablement requirement remains, by far, the major impediment for obtaining claim scope that reaches beyond the embodiments reduced to practice.

  48. IANAEI think you two are overlooking the strong selection bias in the data.

    I did try to address this in my 12:12 comment (esp. last paragraph). Not sure if I’d call it a “selection bias” or just a flaw in the approach.

  49. Mooney: I don’t believe 1% of cases is “exceedingly rare.”

    6: And let’s be clear, 1% of 400000 cases per year is still a lot of cases. 4000 to be exact.

    I think you two are overlooking the strong selection bias in the data. This isn’t 1% of all cases, it’s 1% of all cases in which the examiner has found some defect that the applicant is unwilling or unable to correct, and the defect is marginal enough or valuable enough that fighting it is worth the cost of appeal.

    In your typical case that doesn’t make it to the BPAI, either the applicant and examiner agree on 112-compliant claims, or the application goes abandoned because it was so flawed that a nuance of statutory interpretation would not have saved it.

  50. Rob asks

    “how did this become a question of “separate”?

    Rob, I’d recommend looking up the previous discussions of the Ariad case where the written description issue has been beaten to death. From what I’ve seen here, many posters are too tired of the issue to rehash the old arguments.

  51. so goes another round of stimulating reasoning

    Yes, that’s where I present my conclusion AND my reasoning and you put your fingers in your ears and pretend you didn’t hear me.

    Nice try. Once again: whether it’s a “separate” requirement or an essential part of the analysis “under enablement”, it’s still a requirement.

    If you disagree, explain why.

  52. “A time bomb exists in ¶2nd: “what the APPLICANT regards as HIS invention.” Early question in inventor deposition: Mr. Inventor, what do you regard as your invention?”

    I don’t see the problem. He can just say “I regard them all as my invention. The question is which one of my inventions are you willing to give me a patent on?”

  53. “No, it’s not”

    Um, yes, it is.

    And so goes another round of stimulating reasoning posited by the master of debate, Mooney.

  54. “exceedingly rare that the patent office hangs its case on written description”

    I’m not so sure I’d put “exceedingly” before the rare in that sentence but yes, it is otherwise true. What of it? If we had no requirement as we do now then examination would become a crp shoot for many cases. And let’s be clear, 1% of 400000 cases per year is still a lot of cases. 4000 to be exact.

    Even that is assuming that cases that rest on 112 are not more common before they go to the board. They are. Why? Because when an applicant sees nothing but a 112, he usually just amends/RCE’s. It’s a rare few amongst the rare cases that rest on 112 that end up going to the board or further.

    “Although there may be valid reasons for retaining a separate written description requirement, this study safely leads to the conclusion that the government’s conclusory statements regarding the doctrine’s critical importance for patent examination lack a factual basis.”

    Actually you just came to the wrong conclusion, this study supports the USPTO’s statements. 4000 cases per year is not negligible, and your assumption that we’d get to make new matter rejections was in err. That turns your conclusion on its head. D you can do better, I expect it from you.

  55. the separateness is just one more opportunity for finding that a patent is “crap” for want of written description.

    No, it’s not.

    Whether it’s a “separate” requirement or an essential part of the analysis “under enablement”, it’s still a requirement.

  56. “Red Herring”

    Not so fast Mooney, step away from the red sequined pumps, the separateness is just one more opportunity for finding that a patent is “crap” for want of written description. Why would you care?

    Answer: because you are not consistent with anything, much less your anti-patent views.

  57. As a non-attorney, I am interested in the history of this question. Since there plainly is a statutory requirement for a written description of the invention, how did this become a question of “separate”?

  58. Is there more to this essay available? I only see the abstract at the link.

    I remain convinced that all of the concern about the “separateness” of the “written description” requirement is a red herring.

    Also, it seems to me that it doesn’t really matter whether you call it a prohibition against new matter or not. It’s still a written description requirement.

    I don’t believe 1% of cases is “exceedingly rare.” My impression is that in some art units (e.g., chem/bio) the written description requirement is more important than in others. I’d be curious about how the numbers break down according to art unit.

    Also, determining how many cases “hung on” written description ignores the fact that most applicants tend to deal with the issue early on in prosecution as it is usually a straightforward fix. In most cases, either you have written description support, or you don’t.

  59. Again, unbelievable.

    I’ve been practicing for “only” 10 years.

    Why, oh why, is the patent community JUST NOW getting around to asking questions about the BEDROCK of patent law?

    When I passed the patent exam, I KNEW that I had no idea what I was talking about with respect to things like 112, DOE, standard of review on appeal from BPAI–but I knew that nobody else really knew, either. I did the best I could with the information available at the time.

    I continue to do so–and as an attorney, being able to argue both ways is a plus. But uncertainty costs everybody money, and since I am considering filing apps on some of my own inventions soon, I find myself coming down in favor of the side of more certainty, more predictability, and lowered costs.

  60. What is missing from all the discussion regarding the requirements of §112, ¶1st (written description, enablement, etc.) is any serious discussion about what the term “invention” AS USED IN THE PATENT STATUTES means.

    I would submit that you must first look to §112, ¶2nd which specifies where in the application the “invention” is set out: “The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.”

    I would submit that this means that each claim is an “invention.”

    So you read claim 1. Assuming claim 1 makes grammatical sense and is not vague and indefinite (so as to satisfy ¶2nd), then you look to see if the “invention” defined in Claim 1 complies with ¶1st.

    Then you determine if the “invention” defined in Claim 1 is anticipated (¶102) or would have been obvious… (¶103).

    Then you go on to Claim 2 and perform the same analysis and continue claim by claim until you have examined all of the claims submitted.

    My argument is supported by MPEP 2171 (assuming the PTO’s interpretation of the statutes has any credibility left these days).

    A time bomb exists in ¶2nd: “what the APPLICANT regards as HIS invention.” Early question in inventor deposition: Mr. Inventor, what do you regard as your invention?

Comments are closed.