Guest Essay: An Initial Comment on Ariad: Written Description and the Baseline of Patent Protection for After-Arising Technology

Following on the heels of the Federal Circuit en banc opinion in Ariad Pharmaceuticals, Inc. v. Eli Lilly & Co., Professor Kevin Collins asks "What next?"  

Before Ariad, the uncertainty channeled much of the scholarly conversation about the written description doctrine into a binary debate, pitting those who it applied to original claims versus those who believed that it did not. After Ariad, the stage is now set for greater participation in more nuanced conversations about the role that the written description requirement does and should play in curtailing patent protection. Ariad may end one facet of the debate over written description, but it would be a lost opportunity if Ariad were interpreted to end the debate over written description more broadly and cut off these more nuanced conversations.

In this short essay, Professor Collins suggests that patent law scholarship should adopt a "trans-doctrinal approach" directed to the broader purposes and limitations of patent law rather than being confined to doctrinal silos.  The essay indicates that this holistic approach may also be helpful in explaining a perception that patent law varies greatly according to the area of technology. [Download Collins.Ariad]

Read: Kevin Collins, An Initial Comment on Ariad: Written Description and the Baseline of Patent Protection for After-Arising Technology, 2010 Patently-O Patent L.J. 24.

36 thoughts on “Guest Essay: An Initial Comment on Ariad: Written Description and the Baseline of Patent Protection for After-Arising Technology

  1. 36


    Nice thought.

    I am sure you know how I would respond. In inventing a Written Description requirement in a way that makes them feel comfortable, the Federal Circuit has, once again, thrown their hat into the ring and chosen a peculiar theory about the way language works which doesn’t make sense. Specifically, the Court has said that meaning is somehow derived from a one to one correlation between word or concept (or, in other words, something in the head) and some physical reality. On page 21 of the opinion, meaning, under the Written Description requirement, according to the Court, comes somehow as a result of a likeness between a) what one of skill in the art would “…visualize or recognize…” and b) structural (ie, physical) features or a representative number of species falling within the scope of the genus. In other words, something in the head is isomorphic with the physical outside world (the structural limitation and/or species). And, somehow one of skill in the art recognizes that isomorphism and derives ‘meaning’ from it. As if one skilled in the art gets a special feeling when the outside physical object fits, like a hand in a glove, in the inside mental concept.

    According to the Court, meaning sufficient to satisfy the Written Description requirement can NOT come from a functional limitation. Somehow, functional limitations would lack that special ‘feeling’ that structural limitations have.

    Imposing the strange limitation that words on the inside, in the head, have to “…fit like a glove…” around designated physical objects on the outside exalts a peculiar metaphysical theory and has some really weird implications. What would one do with numbers (ie, how can something in the head match something that is ‘5’ or has fiveness in the outside world)? So, what exactly, is fiveness? Certainly, numbers can be recited in claims and they have meaning. And what about colors (how would something have ‘redness’)? So if these would presumably pass, why would functional language fail the test? Is it because the Court can’t “…visualize or recognize…” functional language? Well, how do they visualize or recognize structural language in a way that’s different from functional language, is there really a ‘something’ in the Court’s head for structural language that isn’t there for functional language? And, is that how the Court gets meaning, when the ‘something’ in the head matches a physical ‘something’ in the outside world? When, as a result of that match, they get a special feeling?

    I do not know how meaning works, but I do know that it doesn’t at all work in the weird way that the Court has hypothesized in Ariad. Meaning does not come from isomorphism, a sort of pairing of things on the inside (the head) with only conceivable physical things on the outside. It’s ridiculous to say that one of skill in the art gets a strange or special feeling, that they can ‘visualize’ or ‘recognize’ when they’ve stumbled across a proper written description, a proper meaning, as the Court suggests. Without visualization or recognition, functional language also has meaning. Doesn’t it? That is, one skilled in the art could certainly recognize an invention from so-called functional language alone.

    If the Court really has problems with ambiguity, there is an entirely separate section for that, namely 112, 2. Or, if they don’t feel a claimed genus has the correct number of representative species, there is a section (and test) for that too! Reading some weird theory of meaning into the Written Description requirement shows that the Court has no idea how language operates and is willing to draw some odd conclusions to accomodate their special feelings of recognition.

  2. 35

    T, this is an age old problem. If one claims exactly what is disclosed, it might be easy to circumvent the claims by making insubstantial changes. The Supremes recognized the problem and provided the DOE. See, Winans v. Denmead.

    If one claims the invention in terms of what it does, even to a degree, then the scope of the claim literally will cover later, independently developed technology that does exactly the same thing functionally, but which did not draw on the invention at all. The Supreme again recognized this problem and provided a solution: functionally defined claims covered the corresponding structure, materials and acts described in the specification and equivalents. See, Westinghouse v. Boyden Power Brake.

    Now, it is one thing to broaden a claim by claiming a genus that is fully enabled, and quite another to claim an invention beyond the scope of enablement by functional language, either by using generic terms or by defining the invention in terms of what it does or enables. In the later case, the claims are rightly limited to the disclosed embodiments and equivalents.

    Now, the Fed. Cir. has held that MPF equivalents extend only to known equivalents. With this limitation, such equivalents should be liberally construed to mean any known means one of ordinary skill would know to work. Future inventions are not involved.

    Next, the Fed. Circ has also held, in Philips, that generic terms such as “baffle” are not to be construed to include unnecessary structure from the specification. While this view is laudable, I also think it is inherently dangerous to the extent that broad generic nouns can be used to functionally define an invention.

    The Federal Circuit is aware of the tension between broad claiming and narrow disclosure. I think the author here is right that they are using the written description requirement as one more arrow in their quiver to limit functional claims that would otherwise literally cover future, independent (and underline, independent) inventions of other that achieve the same result.

  3. 34

    I do not feel very smart, because I don’t accept the premise of the article. The article is premised on an analysis of “after-arising technology.” If it is “after-arising,” then that means that the technology came forth after the patent application was filed. But my guess is that any inventor of a patent in the cases cited would have taken the position that their invention *provided* the technology, that it is not “later-arising”, and that the accused is a rip off. So which is it?

    If a claim is generic to an embodiment in a specification and broader than it, then if that is a valid claim, then doesn’t the author of the paper make an unfounded assumption that the accused is “later-arising technology” (assuming of course that the technology does infringe)?

    If I take on the mindset that the accused is “after-arising” then won’t I be biased to arrive at the conclusion that the specification is somehow lacking?

    I am sure I am missing something here. Comments welcome.

  4. 33

    so easily damaged or broken

    Ah, RWA, you mistake my kind and gentle ways for a lack of toughness and ruthlessness. many are lulled by that perception, much to their detriment. The beauty and delicacy of my posts are there to delight the sensitive nature of the rainbow of readers.

    Besides, I have my sharks with laser beams to do my dirty work.

  5. 32

    At least Mooney had a little something going for him in the toughness department. Your posts are like little porcelain butterfly wings, so easily damaged or broken, with your sensitive song lyrics and stuff… so clearly in need of a friend or, better yet, a dog.

  6. 30

    But beyond that, proof of prior invention does not require a full Section 112, p. 1 description. Never has.

    You have not shown this in law. Or is this IMHO-Ned law?

    Once again,

    proof of prior invention…

    what invention? Why the only thing possible (Duh) – the invention as claimed. Claims need to be supported. What level of support? There is only one.

  7. 29

    ping, you still conflate the requirements for an effective filing date and the requirements of proof of prior invention to antedate prior art. Every single case cited involves whether the claim is supported in the specification for the purposes of an effective filing date. I have consistently agreed with this throughout. For an effective filing date, one must have full section 112, p. 1 support for the claim.

    In interferences, full Section 112, p.1 support for the claim in interference is required because claims are required to have full Section 112, p.1 support by statute. Second, the right to an an effective filing date in an earlier application also is essential to establish who is the senior party. They normally win interferences. But beyond that, proof of prior invention does not require a full Section 112, p. 1 description. Never has.

    Circling back, if a reference not claiming the same invention and not a statutory bar, shows so much as the claimed invention as disclosed in an earlier application, such as a provisional, including obvious variations, reference to the disclosure of the provisional (with some oath by the inventors) is sufficient to “swear behind” the reference.

    ping, this is not rocket science. The distinction between when an effective filing date is required and when it is not is not that hard. You are beginning to remind me of Judge Radar.

  8. 28

    Spin me round round baby – right round.

    DiLeone does not say that the test is different in each case, but rather, only the facts of each case need to be looked at on a case by case basis. This supports your contention about a quick and dirty provisional being sufficient NOT AT ALL.

    In In re Smith you are wandering back to the fishing well of red herrings once again with a genus/species discussion. Not only that, but the court does not support your contention, but rather (in words that dance before and after your selected quote:

    In re Risse, supra, relied upon by appellant, involved, inter alia, the question of “support” for a claimed subgenus in the disclosure of a parent application. The court did not speak in terms of the description requirements of § 112, and it is since Risse that this court has focused on the express language of the statute. The recent cases suggest a more stringent requirement for a description of the claimed invention than may have been previously applied in cases wherein the issue was framed in terms of “support” for claimed subject matter.

    Doubt as to the continuing vitality of Risse in light of the subsequent “description requirement” cases has already been cast in Lukach and Fields v. Conover.

    However, the rule extracted from Risse is much too broad… Unfortunately, Risse has been interpreted as so saying, and to put the proposition to rest, we overrule Risse to the extent that it provides the aforementioned “rule” for the satisfaction of the description requirement of the first paragraph of § 112.

    So Ned, your claim of “Now, it is clear that the court in Smith and Hubin was not saying that Section 112, p.1 applied in proving prior invention.” is false. To be certain, not only was Section 112, p.1 applied, it was applied even more strictly.

    sayest the court.

    RWA – don’t you have a Malcolm to chase?

  9. 27

    This is from a recent WSJ Interview with Chief Justice Roberts:

    “Roberts said he doesn’t pay much attention to academic legal writing. Law review articles are “more abstract” than practical, and aren’t “particularly helpful for practitioners and judges.”

  10. 25


    Your weaving is making me dizzy.

    Now, it is clear that the court in Smith and Hubin was not saying that Section 112, p.1 applied in proving prior invention.
    It rather said that there has to be enough written description to support the claimed invention.

    How is “enough written description to support the claimed invention” anything but applying Section 112, p.1?

    There is only one level of “enough” when it comes to claimed inventions.

  11. 24

    ping, I have agreed that if the reference shows all the elements, your proofs have to show all the elements of the claim. In an interference, this is manifestly the case.

    Now, the quote from the concurrence you quoted above actually came from 935 F.2d 1555 link to
    That court went on to explain:

    The CCPA’s “written description” cases often stressed the fact-specificity of the issue. See, e.g., In re Wertheim, 541 F.2d 257, 262, 191 USPQ 90, 96 (CCPA 1976) (“The primary consideration is factual and depends on the nature of the invention and the amount of knowledge imparted to those skilled in the art by the disclosure”) (emphasis in original); In re Smith, 458 F.2d 1389, 1395, 173 USPQ 679, 683 (CCPA 1972) (“Precisely how close the description must come to comply with Sec. 112 must be left to case-by-case development”); DiLeone, 436 F.2d at 1405, 168 USPQ at 593 (“What is needed to meet the description requirement will necessarily vary depending on the nature of the invention claimed”). The court even went so far as to state:

    [I]t should be readily apparent from recent decisions of this court involving the question of compliance with the description requirement of Sec. 112 that each case must be decided on its own facts. Thus, the precedential value of cases in this area is extremely limited.

    Now, it is clear that the court in Smith and Hubin was not saying that Section 112, p.1 applied in proving prior invention. It rather said that there has to be enough written description to support the claimed invention. The amount varies from case to case. But what has to be shown is “possession” in the interference context, which in many cases shows that one actually was aware of what one invented because there was some description of it as being an invention. The discovery of a new compound does not necessarily, for example, imply that one knows this compound has a specific utility not disclosed, or that a genus of such compounds would work for the claimed purpose.

    But this is not to say that what may be sufficient in Pfaff v. Wells to enable one of ordinary skill to make and use the invention would not be sufficient to prove prior invention even without a full written description of the drawings.

    Case by case basis, sayest the court.

  12. 23


    Perhaps you can explain the distinction you continually want to make with a statutory bar or without (as if this somehow excuses the link and baggage to the claimed invention) and a quick and dirty provisional (which I maintain is a dangerous and fool-hardy undertaking).

    From the concurrence:
    Satisfaction of the description requirement insures that subject matter presented in the form of a claim subsequent to the filing date of the application was sufficiently disclosed at the time of filing so that the prima facie date of invention can fairly be held to be the filing date of the application. This concept applies whether the case factually arises out of an assertion of entitlement to the filing date of a previously filed application under § 120 . . . or arises in the interference context wherein the issue is support for a count in the specification of one or more of the parties . . . or arises in an ex parte case involving a single application, but where the claim at issue was filed subsequent to the filing of the application . . . .
    In re Smith and Hubin, 481 F.2d 910, 914 (CCPA 1973); see also In re Wright, 866 F.2d 422, 424 (Fed. Cir. 1989) (“When the scope of a claim has been changed by amendment in such a way as to justify an assertion that it is directed to a different invention than was the original claim, it is proper to inquire whether the newly claimed subject matter was described in the patent application when filed as the invention of the applicant. That is the essence of the so-called ‘description requirement’ of § 112, first paragraph.”).

  13. 22

    I’m sorry, ping, but I just searched the opinion for “prior” and found noting about proving prior invention. The references at issue were both statutory bars. This requires full support for all claim elements in the parent regardless of what the reference shows.

    Perhaps if you can call may attention to exactly what you are talking about in the opinions, perhaps we can have a meaningful conversation, as it always seems we are not talking about the same thing.

  14. 21


    For someone who says “Duh?”, you fight long and hard to not acknowledge the baggage that must be carried with the claims. Even for “Assume Anascapae had been about proving prior invention instead” as this is mentioned in the opinion.

    Why is that? (and I see you don’t understand what it is in showing prior invention…
    …invention of what?)

  15. 20

    Ping, as RWA said, Duh? Assume Anascapae had been about proving prior invention instead of support in a parent, the proof would still have had to show what the references showed: multi-inputs. Since the ‘525 parent did no show this from a written description point of view, neither could a 131 declaration.

    I see ping, you still do not understand what it is you have to show in proving prior invention.

  16. 19

    “consistent reference to the invention as claimed

    What else matters? A true “duh” moment for ping. One of many I predict.

  17. 18

    Not to dredge up an old topic, or to derail the current topic (well, OK, but just a little), my pal Hal shared a snippet of Judge Gajarsa waxing eloquent on WD (Anascape, Ltd. v. Nintendo of America, Inc., __ F.3d __, __ )(Fed. Cir. 2010)(Gajarsa, J.).

    Of note, especially to Ned, is the careful and consistent reference to the invention as claimed.

    Now back to your regularly scheduled blog.

  18. 17

    Paul, IANAE was being obtuse. However, if a claim that recites novel structure is to cover future devices, it will have to do so under the DOE. The Supremes and the Feds have historically been hostile to any claim structure or format that will literally cover future technology.

    That said, if a claim to a new combination broadly recites the elements of the combination with the understanding that they cover the corresponding structure and KNOWN equivalents (meaning substitutions, primarily) then there is no real vice in having a liberal construction of equivalents as the novelty does NOT lie in the corresponding structure, but in the combination.

    The vice of Donaldson is that it fundamentally destroyed the fabric of Section 112, p. 6 and Halliburton. In doing so, we have the worst of both worlds, claims that literally read on the prior art and which have narrowly construed equivalents.

  19. 16

    Re: “How can a claim that recites only corresponding structure described in the specification literally cover future inventions?”
    That depends on what you call or consider with words like “future inventions” [not statutory words] “corresponding” or “equivalents” and what kind of claim it is.
    For example, if one element of a claim recites a “fastener” and the spec shows a conventional fastener, are you going to exclude any claim coverage of any later invented fastener? When a spec shows a then-conventional incandecent lamp as an example of a claimed “illumination source” should that prevent claim coverage of the same apparatus using LED’s or other later invented illumination sources? Should a generic chemical claim supported by 12 species examples in the specification be avoidable by someone “inventing” a 13th species? Or should “pioneer” patent generic claims be avoidable by later inventing by tweaking one of the disclosed and claimed species to get better results and thus get a valid improvement patent on it?

  20. 15

    claims that literally cover future inventions

    How can a claim that recites only corresponding structure described in the specification literally cover future inventions?

  21. 14

    Malcolm, if the scope of protection is to be limited to the disclosed embodiments and equivalents, the claims should also. That is the principle since O’Reilly v. Morse and Halliburton.

    Section 112, p. 6, authorizes MPF elements in combination claiming. In the CCPA case of In re Lundberg, which was the law for 40 years, the CCPA held that the combination itself had to be new regardless of Section 112, p. 6. Lundberg was indubitably right and consistent with Halliburton, but that case was overruled by Donaldson throwing the whole doctrine of MPF claiming into chaos.

    I find it interest how the Fed. Cir. is now retreating from Donaldson, and seems to again understand the vice of claims that literally cover future inventions without corresponding enablement. These claims retard, rather than advance, the useful Arts.

  22. 13

    IANAE, huh?

    MPF claim elements are only authorized in combination claims. If a claim is functional at the point of novelty and is not a combination claim, it is not saved by Section 112, p. 6.

    The claims in Ariad are not combination claims.

    The claims in Ariad were held invalid under written description. But the author speculates they were really held invalid for the same reason the Supremes held the claims in Halliburton and O’Reilly v. Morse invalid: they literally covered future technology without describing or enabling it.

  23. 12

    “Oh my, how horrible!”

    Gratuitous political correctness IS horrible. Horribly silly.

    Let’s ask Mooney what she really thinks about it…

  24. 11


    I have no disagreement with your conclusions, but dispute your premise. Much scholarship (including articles by you and me) addresses the “baseline” problem, in the trans-doctrinal sense you propose. Practitioners and judges might not read those articles, but it is inaccurate to say that they don’t exist. Moreover, there is necessarily some tension between a high-level, pan-doctrine analysis of patent scope and the kind of scholarship that would have ready, direct application to real-life cases. Real-life cases, of course, resemble Ariad in that they turn on specific doctrines.

  25. 10

    Ned That case was one of a number of Supreme Court cases that essentially held that functional claiming at the point of novelty made the claims indefinite in that they literally covered future inventions, inventions neither disclosed nor enabled

    I think it’s the lack of a convincing non-enablement argument in Ariad (and cases like it) that make the written description issue so compelling. Yes, chemistry and biotech are “unpredictable” arts. But that doesn’t mean that extremely accurate predictions can be made about very generalized results. For example, if I know that protein A needs to bind to molecule B in order for a pathway to operate at maximum efficiency, then I’ve got a lot of art-recognized options for reducing the efficiency of the pathway. It’s not a question of which of these options will work — they all will, to some extent, and that’s perfectly predictable — it’s a question of which option works *best* and whether it can be turned into a product that can be sold for a profit.

    The question is: is it better to grant a patent to an applicant who files claims covering all of those options along with a research plan; or should some additional sweat be required, with the claims limited to the disclosed embodiments. The latter approach would seem to result in more patents being granted to individuals who are actually in a position to develop the products and corresponding methods being claimed. The former approach would seem to lead to more patents for NPEs.

  26. 9


    Please explain how

    the Feds. cannot use Section 112, p. 6 in claims such as found in Ariad

    which implies that they had no choice, is consistent with

    they would have to either restrict the claims in some way to the disclosed examples, or declare the claims invalid under Section 112. They decided to declare them invalid.

    which implies that they made a choice, and the equally available alternative was to do exactly what 112 6th says to do.

  27. 8

    I think a obvious reason the Feds. cannot use Section 112, p. 6 in claims such as found in Ariad v. Lilly is that they are not “combination” claims. Here is the sole “step” in one of the litigated claims:

    “reducing NF-κB activity in the cells such that expression of said genes is reduced]”

    Now, if the author is right that the Feds. are concerned that future technology neither disclosed nor enabled is literally covered by these type of claims, then they would have to either restrict the claims in some way to the disclosed examples, or declare the claims invalid under Section 112. They decided to declare them invalid. While they chose Section 112, p. 1, they could equally as well have chosen p. 2. Let me explain.

    While the author cites Halliburton, he does not go far enough. That case was one of a number of Supreme Court cases that essentially held that functional claiming at the point of novelty made the claims indefinite in that they literally covered future inventions, inventions neither disclosed nor enabled, even if the examples that were disclosed are sufficiently enabled. It therefore strikes me that the claims in Ariad might also suffer from invalidity under Section 112, p. 2.

  28. 7

    Does prosecution history estoppel apply?

    I’m not convinced it does.

    Provisionals don’t really have claims. Even if they happen to have an optional section that looks like claims and is handy for supporting the identical text in the claims of a later regular, that’s really just part of a big amorphous ball of text that is never examined and never turns into a patent.

    Besides which, the amendment would have to be both narrowing and for reasons relating to patentability, the latter being tough to make out when the provisional never got examined, any number of other reasons for changing the claims can come up in the course of a year, and it’s more or less standard practice to perform major reconstructive surgery on provisionals to make them regular.

    (assuming that the regular application incorporates the provisional by reference).

    How is incorporation by reference at all relevant? Surely the priority claim to the provisional would be enough.

    Oh wait, “unrelated answer”. I see what you did there.

  29. 6

    unrelated question – think “magic words”.
    Do not ever use “I claim in a provisional (unless for a hugely broad single claim).
    Redraft your actual claims (if available) in a “an embodiment, a further embodiment…” style.

  30. 5

    Yes (assuming that the regular application incorporates the provisional by reference).

  31. 4

    a provisional application with claims is filed, followed by a regular application whose claims differ from the provisional application claims.

    Does prosecution history estoppel apply?

  32. 3

    stand by for politically correct use of the pronoun “her” throughout the article even though most of the Ariad (MIT) inventors were men.

    Oh my, how horrible! Thank you for the alert.

  33. 2

    As the author notes, broad patent claims can be narrowed or even held invalid by judges by citing different sections of 112 or by “claim interpretation” to get to the same end.
    But the real problem seems to me to be that the above is extensively subject to underlying subjective differing views between different CAFC judges as to how broadly claims SHOULD be read beyond more than mere specification examples. Especially with electrical or mechanical inventions clearly outside of the traditional “unpredictable arts” (traditionally requiring more species examples for broader claims).
    The issue is not just a matter of covering “after-invented technology” or not. It also involves a major issue of how easily inventors should be denied appropriate compensation or protection for their inventions and R&D costs by how easily and cheaply the patents can be “designed around” with obvious variants or equivalents (that are NOT later “inventions”) without hugely burdening inventing and patenting costs and time delays with publishing reams of such variants or equivalents?

  34. 1

    I was not aware of “disagreement” regarding claims being part of the original disclosure. Claims ARE considered part of the original disclosure.

    Note 1: as far as I could determine from a review of the Attorneys and Agents link at the USPTO website, Professor Collins does not appear to be listed as a registered patent attorney. Not that there is anything necessarily wrong with that…

    Note 2: stand by for politically correct use of the pronoun “her” throughout the article even though most of the Ariad (MIT) inventors were men.

Comments are closed.