Did the Federal Circuit Revive Halliburton (1946)?

by Dennis Crouch

BASF v. Johnson Matthey (Fed. Cir. 2017)

BASF’s catalytic conversion system patent claims includes the functional limitations that the system is “effective for catalyzing”/”effective to catalyze” NH3 oxidation.

In the infringement litigation, the district court sided with the accused infringer – holding that the functional language was unduly indefinite – rendering the claim invalid under 35 U.S.C. 112 as interpreted by Nautilus. On appeal, the Federal Circuit has reversed — finding that the record  lacks “evidence that would support” the indefiniteness finding.

The Patent Act (section 112) requires “one or more claims particularly pointing out and distinctly claiming” the invention.  This provision serves as the basis for doctrine of (in)definiteness.  In Nautilus (2014), the Supreme Court held that the claims as drafted must define the scope of the invention (from the perspective of PHOSITA) with “reasonable certainty.”   Reasonable certainty is a traditional high evidentiary burden used in several areas of practice — including lost-profit calculations — and is thought to be parallel with the “clear and convincing” evidence standard (i.e., just below “beyond a reasonable doubt.”).  Despite the seeming high standard placed on patentees by Nautilus, the Federal Circuit has continued to give patentees the benefit of the doubt in indefiniteness cases (with the one exception being 112(f) cases).  This case is no exception.

Effective To: The claims here are not limited to particular materials or level of effectiveness, but rather only those that fit within the broad “effective to catalyze” language.  The district court identified this as problematic:

Each claim fails to limit the “material composition A” … to any specific materials. Rather than explicitly defining the material compositions, the claims utilize functional language, specifically “effective,” to purportedly define them.  In other words, the claims recite a performance property the composition must display, rather than its actual composition. Moreover, none of the claims recite a minimum level of function needed to meet this “effective” limitation nor a particular measurement method to determine whether a composition is “effective” enough to fall within the claims.

[FN: The court additionally notes that “a practically limitless number of materials” exist that would “catalyze SCR of NOx, even within the normal operating conditions of an exhaust after treatment system,” indicating that the claims, as written, fail to sufficiently identify the material compositions.]

Without such information, a person of ordinary skill in the art could not determine which materials are within the “material composition A” or “material composition B” limitation, and which are not.

On appeal, the Federal Circuit rejected the lower court’s analysis – first repeating again that functional claim limitations are perfectly legitimate.  According to the appellate court, the basic problem with the district court’s analysis was that it identified the broad claims but then failed to actually consider whether the claims provide “reasonable certainty” to PHOSITA as to the claim scope.

What is needed is a context-specific inquiry into whether particular functional language actually provides the required reasonable certainty. . . . [T]he inference of indefiniteness simply from the scope finding is legally incorrect: “breadth is not indefiniteness.” (Quoting SKB v. Apotex (Fed. Cir. 2005)).

After rejecting the district court’s analysis, the appellate panel went-on to hold that – as a matter of law – the claim is definite. Drawing its own conclusions as to what PHOSITA would think of these claims.

At the point of Novelty: An interesting bit of the opinion harkens back to the Supreme Court’s 1946 Halliburton decision – holding that functional claim language is particularly problematic when done at the point of novelty. Here, the court does not cite Halliburton but does note that the “asserted advance over the prior art” is the particular layered arrangement of the device, not “the choices of materials to perform each of the required catalytic processes.”  Rather, those materials were expected to be ones already well known in the industry.

Here, of course, the Federal Circuit is not suggesting that Halliburton is good law, but the court does implicitly conclude that there are important distinctions in the indefiniteness analysis when considering terms directed to well-known versus novel features.

76 thoughts on “Did the Federal Circuit Revive Halliburton (1946)?

  1. 10

    [I]t’s worth noting again that “TOTALLY functional” is not the correct standard…

    I agree with this, but it is also worth noting that if there is one claim to which 112(f) does not apply, it is the TOTALLY functional claim. By its own terms, 112(f) explicitly applies only to “a claim for a combination.” If you disclose darunavir and claim “a means for slowing the progression of HIV infection,” then your claim will fail, and 112(f) simply cannot save it.

    1. 10.2

      You are confusing what a combination claim and a “TOTALLY functional” are.

      Not sure why this would be a source of confusion for you.

      1. 10.2.1

        The only TOTALLY functional claim is a single means claim. If I claim “a machine having a sweeping means and a suction means” that is not a single means claim (it has two “means”), but it is also not TOTALLY functional. The mere fact that it recites two separate “means” limitations conveys that the machine has at least two parts, and “two” is a structural limitation. The only way that one can get TOTALLY functional is with a single means claim (“a machine comprising a carpet cleaning means”), which is also the only sort of claim that is explicitly not permitted by §112(f).

  2. 9

    I have never seen a person who has invented (both enabled and possesses the full scope of) a novel functional scope, and I doubt I ever will.

    A functional scope does not limit the means by which the function is achieved, which results in*all* means being covered. I have never seen a specification which proves that the disclosed means constitute the entirety of the scope, and in virtually all of the specifications I see (I work in computer software) I doubt such a thing COULD be proved.

    Functional and intentionally vague language is almost always an attempt by attorneys to steal from the public what their inventor didn’t invent.

    When I was at Willkie Farr (one of the AmLaw 25 firms in NYC — Willkie was Wendell Willkie who lost to FDR) just down the hall was a senior partner…

    Language like this almost always precedes someone who either isn’t very good or is about to do something not very smart.

    He told me “This is from the days when you had to draft claims without functional language. This is what I wrote to claim a common household object. Can you guess what it is?” It was a full-page sheet of paper, it must have been 300 or 400 words.

    Yep, not very good.

    I studied it for several minutes before I figured out — it’s a toilet paper roll, the kind that has the two tubes sliding over each other, with shoulders at each end, and a spring in the middle urging the two tube members apart.

    Oh look, you described it yourself in two lines. If you change the language to put the second tube with a greater inner diameter than the outer diameter of the first tube you can remove the word “sliding” and be done with it. (I’d also consider changing “urging apart” to simply “compressed spring”). I’ll split the senior partner’s salary with you?

    Without functional language, it’s nearly impossible to write claims that ordinary mortals can figure out.

    The question isn’t whether a function can be used, the question is what the scope of the claim is. He could have put those 400 words in the spec (in fact, they already have to exist in the spec regardless) and used means-plus to put that structure into the claim with a minimal amount of words.

    Nobody is arguing that means-plus doesn’t exist. Means-plus is functional language that creates a structural limitation. The question is if non-means-plus language (such that there is no structural limitation) in the novelty of the invention is alright. That was settled as far back as General Electric v. Wabash if not earlier, and 1952 didn’t change that.

    Sometimes I have to actually do this. I list the solutions and try to figure out the language that captures them. The reality is that there is hundreds and thousands of variations. This is the way real people work.

    Listing solutions is the only manner of properly claiming non-equivalent solutions. The fact that your inventor may posit five or distinct means of achieving a result does not mean he has posited all means of achieving a result. He posits all means of achieving a result when you can prove that no other means exist other than those equivalent to disclosed ones. That means you have to 1) disclose means and 2) disprove the existence of non-equivalent means. (2) rarely happens.

    All of these problems go away if scope of enablement is used.

    For the reason above, scope of enablement is virtually never met, but regardless the purpose isn’t only to protect an inventor, it’s to protect the public by making clear what the scope is. If the scope can only be determined after having a battle of the experts about what was enabled at the time of the filing rather than what the claim language is, the public has no notice. It does not promote the progress of the art to chill the implementation of possibly non-discovered subject matter simply because the inventor wants to use vague language to put the onus on the public to disprove his scope.

    There already is a mechanism to collect the maximum scope enabled by a disclosure and it’s called means-plus claiming. The very same specification claimed outside of means-plus is an attempt to claim things which are technically enabled by the disclosure but not thought of by the applicant. The question of the validity of that scope was settled in Morse.

    As Ned already noted above, Halliburton never was intended to reach to claims in which the novelty resides in the novel arrangement of old structures. It is perfectly fair, in those circumstances, to identify some (or perhaps all) of the elements by what they do rather than what they are.

    True inasmuch as the structure is not dependent upon the selection of elements that perform the function. i.e. if the structure of the arrangement would be the same regardless of whether Species A or Species B that perform known function F is selected, that is okay. It is generally okay to express known things by function.

    1. 9.1

      Hypothetical to express a point. Suppose one wants to use a computer to do what Google Maps does – receive a first location input, a second location input and then estimate a driving time from the first to the second location.

      Let’s even place this at the point in history where it would have been novel.

      I can certainly posit an algorithm for achieving this: Take A and B, calculate the distance and assume that travel by car averages out to 20 miles an hour, thus you just divide distance by 20 and convert to minutes.

      That’s a structure. It’s a pretty bad structure but it’s a structure. I can easily think of a dozen completely different (i.e. non-equivalent) means of achieving that estimation function.

      Guess what, I have to claim those dozen different ways. What I can’t do is state “here’s a useful end result” point out that I have *some* means of reaching the result, and summarily claim other means I haven’t posited. That prevents Google from developing google maps even though I never posit Google’s algorithm.

      Have I enabled Google’s algorithm? Probably. In as much as I presented a logic problem and people are enabled to solve logic. In the same manner that if I know one of skill knows how to write I have enabled all writing. But if I told people “write a story about a wizard” and then went around claiming I “invented” both Harry Potter and Lord of the Rings, people would laugh at me (even if my disclosure predated both works).

      My “divide by 20” algorithm doesn’t mean I’ve discovered Google’s significantly better and more complicated algorithm, any more than it means that I have written Potter or Rings.

      A computer program which receives an input A and B and then (to use conventional functional-style claiming) includes a processor running code where the code causes the processor to “estimate the amount of time it will take to drive from A to B” is an effort to aggrandize my mundane invention into something greater than what it is. What I have invented is my algorithm (and other algorithms, to the extent I have disclosed them). I haven’t invented the function, because I haven’t foreclosed other non-equivalent algorithms from existing, and in fact Google’s later algorithm is not proof of infringement but proof of invalidity. It evidences that either 1) my claim is directed to an abstract goal rather than a concrete solution or similarly in language this group with better enjoy, is fails to be described and possessed over the breadth of its scope and therefore fails both 101 and 112a, or

      2) the vague claim language that is used when my specification discloses that more particular language exists (indeed, I was capable of actually articulating the algorithm in the spec, so it could be articulated in the claims) is failing to particularly point out my invention with reasonable certainty.

      The same applies to virtually all claims that are functional at the point of novelty – either the spec disclosed a means and the functional language is not an accurate description of that scope, or the *intent* is to claim more than what was disclosed (i.e. I disclosed this manner, and now I have enabled the scope of the function, a Morse-style argument), which renders it abstract and unpossessed. That is why, prior to 1952, there was no written description because overbroad scopes of which some subset were posited always created either an indefiniteness (Wabash, Perkin’s Glue) or abstraction (Morse) problem.

      There exist a very small number of cases where other limitations would so cabin the functional novelty that the specification actually describes all of the equivalent means that fill the scope. There exist an even smaller number of cases where the nature of the world simply prevent a more particular structural disclosure (e.g. application of the uncertainty principle). Beyond that novelty expressed by functional result (or other vague language) outside of means-plus is an attempt to use verbal vagueries to claim more than what the inventor posited. Positing a means is a requirement for invention.

    2. 9.2

      RandomGuy —

      (a) You didn’t read the anecdote very carefully — you’re responding to points I didn’t raise and wasn’t trying to illustrate.

      (b) It looks like you have some misimpression about how the “enablement” requirement works. All I can do is refer you to one of the standard treatises.

      No need to engage this discussion.

      David

      1. 9.2.1

        (a) You didn’t read the anecdote very carefully — you’re responding to points I didn’t raise and wasn’t trying to illustrate.

        I took your point to be that without functional language claims would be in- or less comprehensible (“Without functional language, it’s nearly impossible to write claims that ordinary mortals can figure out.”) and further that to the extent there are problems, scope of enablement is a better test (“I think that the betteer analytical path for objectionable functional language is more likely “scope of enablement” rather than “definiteness””). The former is belied by the fact that your senior partner practiced at such a time, and further belied by the fact that you were capable of summarizing an alleged invention in a reasonable manner.

        Conversely, the general acceptance of functional language at the novelty creates the host of issues above and, contrary to your point at 7.3.3.1, the reasonable certainty analysis is much quicker, cheaper and intrinsic than an undue experimentation analysis to solve them. Moreover, undue experimentation simply gets you the wrong result a lot of the time, and utilizing it to the exclusion of indefiniteness (which I am unsure if you argue for, but Night certainly has and does argue for) results in wrong patentability answers.

        (b) It looks like you have some misimpression about how the “enablement” requirement works. All I can do is refer you to one of the standard treatises.

        For things such as writing computer code, enablement is generally always met. Especially at the office, the view is that once someone expresses a function, the act of translating that command to a computer is mere scrivener’s work. Barring a situation like McRO, where translating the artist’s gut instinct required an inventive act on its own, pretty much any logic you state will be considered enabled.

        Once I command someone to perform some logic (“Code to estimate trip time”) one of skill is enabled to perform the full scope, because there is simply no experimentation required (and thus it can’t be undue) to “estimate” (without a qualifier of the goodness of fit of the estimation) something and the translation to a computer command is non-inventive.

        Let me put it like this – If I put forward a logic problem and then claim a computer with code to solve that problem, the Office will not make an enablement rejection regardless of what (or even whether I include) disclosed particular means are so long as the original disclosure includes a phrase akin to “code to [perform the function].” That alone is reason enough to toss “scope of enablement” as the primary/total standard.

        Even outside of office procedure, so long as one example algorithm is given, together with a boilerplate paragraph that states that the example is not limiting and one of skill would be enabled to achieve other means, that scope will be considered enabled because there is no reasonable manner that one can show undue experimentation; and there simply is *no* manner at the office level, where the lack of independent expert testimony simply makes it a swearing contest after running through Wands.

        The fundamental issue is Night’s statement that:

        There is nothing wrong with functionally claiming the “novel” element. The problem is that all the solutions that would be obvious to a POSITA needs to be included in the scope of the claim. And the DOE is not enough. That forces functional language.

        This argument suggests that if I prove I posit one means of achieving something, but in fact my disclosure happens to enable completely distinct means that I haven’t posited (“all the solutions that would be obvious to a POSITA” though I wonder why he says that when that’s not the standard for an enablement rejection) I am entitled to claim those unposited means by substituting language intentionally created to be vague (functional language) instead of the means I was entirely capable (and in fact did) express in the specification. That simply is not the law.

        Night complains that “these arguments are always discussed in the abstract” so I gave a concrete example – code which estimates a trip time. My disclosure of taking inputs, calculating the distance, dividing it by 20mph, and arguing that other algorithms could be achieved (to use Night’s language here) because they would be obvious or (to use a better standard) without undue experimentation is a situation that 1) the office would not reject under a scope of enablement theory and 2) likely would never be invalidating under that theory. That doesn’t change the fact that its pretty clear the law as the Supreme Court reads it would never allow me to recover from Google, so the suggestion that “all of the problems go away” if you consider only scope of enablement can only be true if you don’t consider “Coming to the wrong conclusion about validity” to not be a problem.

        But no need to engage this discussion.

  3. 8

    Note that this not a final decision, only a preliminary skirmish solely on claim indefiniteness being remanded back for trial on any other issue.
    Also,with a broad claim to using any two different catalysts in the same catalytic converter it would be surprising if an IPR has not been filed?

  4. 7

    All of these problems go away if scope of enablement is used.

    There is nothing wrong with functionally claiming the “novel” element.

    The problem is that all the solutions that would be obvious to a POSITA needs to be included in the scope of the claim. And the DOE is not enough.

    That forces functional language.

    What is always so bothersome about these arguments is that they are discussed in the abstract. Any actual invention and actual language with the set of actual solutions leads immediately to the need for functional language.

    1. 7.1

      Sometimes I have to actually do this. I list the solutions and try to figure out the language that captures them.

      The reality is that there is hundreds and thousands of variations. This is the way real people work.

    2. 7.2

      All of these problems go away if scope of enablement is used…

      … and a whole new set of problems leap in, worse than the ones just solved. If the claim is to be construed as broadly as to be coextensive with whatever the patentee has enabled, then why have claims at all. The point of a claim is to put the public on notice as to what is off-limits (and then, by implication, to give them certainty about what is permitted). If the claim’s scope is simply “whatever is enabled,” that is really no notice or certainty at all.

      Your objection boils down to “it is hard for a patent attorney to write a claim that covers all of the invention and nothing else.” This is no doubt true but what of it? Yes, it is hard to use words to state what the invention is, and much easier to use words to say what the invention does. Congress, however, has rightly chosen to run the patent system for the convenience of the nation as a whole, and not for the convenience of the patent attorney profession.

      Yes, it is difficult to satisfy the requirements of §112 as Congress has written it, but that is why people are willing to pay us to do so.

    3. 7.3

      I share Night Writer’s exasperation at the continued unease over functional language. It’s a practical necessity.

      When I was at Willkie Farr (one of the AmLaw 25 firms in NYC — Willkie was Wendell Willkie who lost to FDR) just down the hall was a senior partner who had tried a patent jury trials in the late 1970s early 1980s. He gave me a page he had written 30 years ago. It was a claim he had drafted as a “doodle” during a slow day of trial.

      He told me “This is from the days when you had to draft claims without functional language. This is what I wrote to claim a common household object. Can you guess what it is?” It was a full-page sheet of paper, it must have been 300 or 400 words.

      I studied it for several minutes before I figured out — it’s a toilet paper roll, the kind that has the two tubes sliding over each other, with shoulders at each end, and a spring in the middle urging the two tube members apart.

      Without functional language, it’s nearly impossible to write claims that ordinary mortals can figure out.

      Courts and know-nothing academics (whose last name begins with L) and know-nothing APJs (see Ex parte McAward) need to just get over it. If functional language gets excluded, the whole patent system will bog down in a burdensome but pointless boggle of word salad.

      1. 7.3.1

        Nobody is talking about excluding functional language entirely. The point is about functional language at the exact point of novelty. I will concede that it is harder to write a claim that is purely structural at the exact point of novelty, but it is not impossible, and it is necessary to a well functioning system.

        Alternatively, one can simply avail oneself of §112(f) as Congress intended, and save oneself a lot of bother.

      2. 7.3.3

        David, Night, can it be, that anybody seriously doubts the necessity of some functional language in a claim, from time to time? Can anybody seriously doubt that there is something between “no functional language in the claim at all” and “functional language in the claim being used to define the contribution to the art”. Here, for example is an extract from the EPO’s MPEP, at F.IV. 4.10

        The area defined by the claim must be as precise as the invention allows. As a general rule, claims which attempt to define the invention by a result to be achieved should not be allowed, in particular if they only amount to claiming the underlying technical problem. However, they may be allowed if the invention either can only be defined in such terms or cannot otherwise be defined more precisely without unduly restricting the scope…….”

        I stop there, even though what follows is vitally important, because it’s enough to catch the drift.

        If you want to reqad more, here the Link:

        link to epo.org

        People all over the world these days take inspiration from the compact White Book of “The Established Caselaw of the Boards of Appeal of the EPO”. I rather suspect that it’s not unknown, at appellate level, even inside the USA.

        1. 7.3.3.1

          thanks Max —

          Yes, unfortunately, there are benighted individuals that have the antediluvian view that all functional language is bad. Michelle Lee mentioned it in several speeches. Mark Lemley has been writing on this for years.

          I largely agree with EPO’s call for judgment rather than per se rules (though I think that the betteer analytical path for objectionable functional language is more likely “scope of enablement” rather than “definiteness”).

          The link to EPO is broken. Can you try again?

          Thanks

          David

          1. 7.3.3.1.1

            Gladly David. Perhaps now:

            link to epo.org

            I agree that “scope of enablement” is a powerful concept. I doubt the EPO jurisprudence fights against that notion. It clings to the notion that scope of protection should correspond to the scope of the “contribution to the art”. The enabled contribution, of course.

      3. 7.3.4

        David, do you think that at least part of the controversy over “functional” claims is confusing functional claiming of a claim element with the claiming of the desired end result [irrespective of how that end result is accomplished]?

        1. 7.3.4.1

          Paul, might I add that it is really important to here distinguish between claiming old elements in a combination or background, and claiming the novel elements. I have been arguing this for years here and before that. Claiming old elements in terms of what they achieve is quite acceptable because skilled artisans know that the claim covers what is discloses and any other known means to achieve the function.

          But when claiming the novel elements, the claim cannot simply claim a result. That is classic claiming in the abstract, and it appears to be forbidden even in the EPO.

          What is new here about this case is that the Federal Circuit (or some litigators) seems to have been listening to me and have now woken up to the distinction.

        2. 7.3.4.2

          Yeah, I think that’s EXACTLY the problem.

          Two examples —

          In re Fuller, 35 F.2d 62, 3 USPQ 51 (CCPA 1930) — the claim was “a fabric tending to wear rough rather than smooth.” The CCPA disallowed this claim as indefinite — I think the better analytical ground is overbroad because full scope is not enabled.

          O’Reilly v Morse, 56 U.S. 62 (1853) — the claim was “I do not propose to limit myself to the specific machinery, or parts of machinery, described in the foregoing specifications and claims; the essence of my invention being the use of the motive power of the electric or galvanic current, which I call electro-magnetism, however developed, for making or printing intelligible characters, letters, or signs, at any distances.” Remember this is only 30 years after Oersted discovered the magnetic current around a wire, and Faraday invented the electromagnet. Electricity was still an “unpredictable art.” So Morse’s claim was too broad as not enabled to its full scope. It’s really too bad that our not-so-patent-savvy nine have turned this into a subject matter case — it isn’t.

          To slightly oversimplify, Morse’s invention works by varying current. Bell’s telephone works off the variation in field (one is the first derivative of the other). I am perfectly happy that Morse should not be allowed to patent Bell’s invention.

          1. 7.3.4.2.1

            David, ditto the famous Neilson case(s). They are reported at WPC 295, 331. link to books.google.com

            The main objection was that the specification only described a heated box between the bellows and the furnace and stated that the form and shape of the box could be varied according to the application. However, it turned out, that a heated box but would not work – at least not to any degree of effectiveness. One had to pass the air from the bellows to the blast furnace through a number of metal tubes that were heated.

            Thus it was left to the skilled artisan to figure out how to implement the invention. At this time, English law required that the specification actually teach one of ordinary skill in the art how to make and use the invention from the specification itself without requiring further invention. (I think that is law today in the United States as well.) Obviously, the Neilson specification was deficient. The invention was simply not enabled.

            Yet in the end, the Appeals Court allowed the the claim because applying heated air to a furnace greatly improved its operation. In effect, Neilson was granted a patent to a method that required heating of air between the bellows and the furnace. The claim covered every apparatus for carrying out the claimed “method.”

            It is said that Cochran v. Diener basically followed Neilson when it held that the details of the apparatus for implementing a method are substantially irrelevant.

            But in the sense, when the essence of the invention is a method, the apparatus might be claimed functionally.

          2. 7.3.4.2.2

            Maybe Morse’s claim would have been OK without the “any.” Then it would have been just a scope of enablement argument. But, by adding “any”, he was likely claiming new non-obvious inventions after his.

  5. 6

    [C]laims with some elements in terms sounding in function are fully legitimate OUTSIDE OF the optional realm of 112(f).

    Sure. No one doubts it. As Ned already noted above, Halliburton never was intended to reach to claims in which the novelty resides in the novel arrangement of old structures. It is perfectly fair, in those circumstances, to identify some (or perhaps all) of the elements by what they do rather than what they are. In fact, where the element is acknowledged to be conventional, pure functional recitation is just fine.

    The problem has only ever been (and continues to be) with functional recitations at the point of novelty. The 1952 act abrogated Halliburton only in the sense that the 1952 Act made it possible (albeit not inevitable) that one might use a functional recitation at the point of novelty, and instead of getting nothing (because of invalidity), your functional recitation is magically changed (by operation of law) into a de facto structural limitation, so that you get the structure you disclosed (and equivalents).

    To this day, however, if you claim by functional recitation at the point of novelty, and you do not invoke §112(f), you just end up with an invalid claim.

    1. 6.2

      Haliburton was abrogated.

      Period.

      Full stop.

      Deal with it.

      That “Point of Novelty” that you want to depend on is just not there in the law.

      1. 6.2.1

        Er, o.k. Care to point to the provision in the 1952 act that achieves that abrogation? I rather doubt that you can.

        1. 6.2.1.1

          Greg,

          I ask for my edification rather than as a challenge…

          Can you point to any post-1952 case law that indicates that Haliburton was not abrogated? I do not think this BASF case counts.

          1. 6.2.1.1.1

            Two responses:

            (1) I can cite you plenty of post-1952 case law that continues to treat Halliburton as if it has continued validity, even in view of the 1952 addition of §112(3). See, e.g.:

            * Technical Tape Corp. v. Minnesota Mining & Mfg. Co., 143 F. Supp. 429, 438 n.9 (S.D.N.Y. 1956)
            * Rockwell v. Midland-Ross Corp., 438 F.2d 645, 653 (7th Cir. 1971)
            * Amdocs Ltd. v. Openet Telecom Inc., 841 F.3d 1288, 1291 (Fed. Cir. 2016)

            That last one is probably the most informative to my point. Consider the following quote

            [Y]ears ago the Supreme Court outlawed such broad ‘ends’ or function claiming as inconsistent with the purposes of the Patent Statute. Congress, however, a few years later softened the rule. Patentees could write claim language to broadly describe the purpose or function of their invention, and when they did the claim would not cover the bare function or goal, however performed, but only as limited to the particular means (and equivalents) for implementing that function or goal as described by the patentee in the patent’s “specification.”

            N.B. two points:

            (a) The CAFC describes the 1952 as “softening” the Halliburton rule. Not as eliminating it entirely.
            (b) The reason that the 1952 act can be understood to have softened the Halliburton rule is that “when [patentees write claim language to broadly describe the purpose or function of their invention] the claim would not cover the bare function or goal, however performed, but only… the particular means (and equivalents) for implementing that function or goal.”

            In other words, the way that the 1952 act undid the Halliburton rule is by turning functional claim language—by operation of law—into de facto structural claim language. Of course, Halliburton is fine with structural language, so the 1952 act does not really reverse Halliburton, it just changes the operation of laws governing claim construction in such a way that fewer claims will fail under Halliburton. That is how Federico explained it (75 J. Pat. & Trademark Off. Soc’y 161, 186) when he wrote that “[t]he last paragraph of section 112 relating to [N.B.] so-called functional claims… [authorizes] some measure of greater liberality in the use of functional expressions… than had been permitted by some court decisions… Halliburton Oil Well Cementing Co. v. Walker, 67 S. Ct. 6, 329 U.S. 1, 91 L. Ed. 3 (1946) [is] modified or rendered obsolete, but the exact limits of the enlargement remain to be determined,” (emphasis added). I call attention to that “so-called” because Federico is acknowledging that §112(3) claims are not really functional—they are structural, it is just that the structural limitation is disguised in functional language. Note also that Federico acknowledges taht “the exact limits of the enlargement remain to be determined,” which conveys that he envisions a continued vitality to Halliburton, just in a more limited range of circumstances.

            (2) I think that it gets the question backwards to ask for case law indicating that a supreme court case was not abrogated. Once the Court hands down a decision, it is presumed to keep on functioning until it is cut off. Therefore, the right question to ask is: where in the statute or subsequent case law do we see the Halliburton holding to be abrogated? Once we can identify that particular statutory or case law text, we can then analyze the extent of the abrogation.

            I think of a lot of lazy lawyers are easily susceptible to the motivated reasoning of thinking ‘I read somewhere that the 1952 abrogated Halliburton, and it suits my convenience for Halliburton to be gone from the books, so I am not going to look too deeply into that point.” In fact, however, one cannot identify any language in the 1952 that clearly abrogates Halliburton in toto, and Federico’s authorized commentary on the 1952 act never says that the act completely reverses Halliburton. Rather, as the Court explained in Warner-Jenkinson Co. v. Hilton Davis Chem. Co., 520 U.S. 17, 28 (1997), “Congress enacted § 112, P6 in response to Halliburton… as a targeted cure to a specific problem” (emphasis added). N.B. “targeted cure,” not “sweeping reversal.”

            1. 6.2.1.1.1.1

              You AGAIN make the mistake that somehow “abrogated” was “abrogated in some particular limited but not defined way”

              That is not how abrogation works.

              The case was thrown out.

              By Congress.

              Period.

              Deal with it.

              1. 6.2.1.1.1.1.1

                You AGAIN make the mistake that somehow “abrogated” was “abrogated in some particular limited but not defined way”

                I do not see anything ill defined in the explanation that I offer. The way that the 1952 Act abrogates Halliburton (as Federico makes explicit) is by the addition of 112(3) (now 112(f)) to the statutory text. Therefore, a claim that refuses to make use of 112(f) cannot avoid the consequences of Halliburton.

                In other words the extent of the abrogation is exactly conterminous with the extent of 112(f). Nothing vague about that in the least.

                1. You are of course completely incorrect.

                  The case was abrogated in toto.

                  Then Congress provided an option for TOTALLY functional claiming. Use of terms sounding in function outside of 112(f) – think Vast Middle Ground.

                2. O.k., baby steps.

                  We both agree, I believe, that Congress abrogated Cuno Engineering in the 1952 Act. We all know the exact language that Congress used to abrogate Cuno: viz, the second sentence of clause (a) of §103, “[p]atentability shall not be negatived by the manner in which the invention was made.”

                  In other words, when Congress wants to abrogate a case holding, Congress cannot merely think ugly thoughts about the case, or scribble graffiti against it in the Capitol bathrooms. They actually have to pass legislative text that is incompatible with the holding that they wish to abrogate.

                  So, what is the statutory language by which Congress abrogated Halliburton? Certainly Federico thought that §112(3) abrogated Halliburton (in whole or in part), and I agree with Federico. Of course, if one understands §112(3) in this way, then it follows that when a claim does not make use of §112(f), it is still subject to Halliburton.

                  I gather that you do not regard §112(3) as especially critical to the abrogation of Halliburton. If not, then where in the 1952 Act do you locate Halliburton‘s abrogation?

                3. Certainly Federico thought that §112(3) abrogated Halliburton (in whole or in part),

                  In whole.

                  Think Greg: Vast Middle Ground

                  (ps: you can leave out the straw man of single means claims – NO ONE is discussing that item)

                4. “In whole.”

                  Federico himself wrote “decisions such as Halliburton Oil Well Cementing Co. V. Walker are modified or rendered obsolete, but the exact limits of the enlargement remain to be determined”.

                  His own word explicitly acknowledge that:

                  1. Halliburton might have been modified.
                  2. He did not know the extent of how the 1952 act changed the law on this issue.

                5. Yes Ben – that was Federico’s opinion (read his whole treatise, as he mentions that some of what he wrote was merely his opinion).

                  Then try to understand what it means for a case to be abrogated.

                  Sometimes you need to do more than merely read what someone writes. Sometimes you need to understand the context of that writing and sometimes it helps to understand the larger context of law.

                6. In whole.

                  Once again, Congress cannot abrogate a Supreme Court holding merely by thinking strong thoughts or sending chain letters with jinxes to the justices. They actually need to enact statutory language that is inconsistent with the holding that they wish to abrogate. What is the language to which you can point in the statute that allegedly abrogates Halliburton completely?

            2. 6.2.1.1.1.2

              I understand your point regarding (2), but I was hoping that a truly clear indication (i.e. a court invalidating a patent under Halliburton) in the case law would make such an order-of-operations type analysis moot.

              I’m not entirely persauded, but your position seems stronger. Especially as anon’s beloved Federico was not certain whether Halliburton was modified or rendered obsolete.

              Thanks.

        2. 6.2.1.2

          Lol – Greg, you go from “partial abrogation” to “prove it was abrogated at all”….

          I merely take your initial position as an admission against interests in your second position (you hoisted yourself on your own petard).

          1. 6.2.1.2.1

            I am not asking you to “prove it was abrogated at all.” I acknowledge that it was partially abrogated. See my response to 1.1.1.1.1.1 below.

            1. 6.2.1.2.1.1

              Your response on “partially” is not sufficient.

              You are still makin things up whole cloth as to what abrogation actually means.

  6. 5

    An interesting bit of the opinion harkens back to the Supreme Court’s 1946 Halliburton decision – holding that functional claim language is particularly problematic when done at the point of novelty.

    Um no – that was merely Ned Heller’s over-reading into the decision his very own hobby horse.

  7. 4

    Greg, I think you missed a nuance. The Supreme Court in Halliburton was concerned only with functionally claiming the novel element in an old combination. (See the follow-on case, Faulkner v. Gibbs, that clarified that Halliburton did not apply to the situation where novelty was in a combination.) Thus, when the prior art was being claimed in a combination claim, i.e., the old elements of a combination, each of those old elements could be functionally claimed and that claim element would be construed broadly to cover virtually any known structure that skilled workers would know to use to carry out the claimed function – otherwise known as known interchangeability. One was not limited to the disclosed structure and mechanical equivalents – something that “adds nothing of significance to the structure.” Valmont Industries Inc. v. Reinke Mfg. Co., Inc., 983 F.2d 1039, 1043, 25 U.S.P.Q.2d 1451 (Fed. Cir. 1993) (discussing the scope of means plus function equivalents).

    The innovation of this case is that the court took the Supreme Court in Halliburton and Faulkner at its word.

    1. 4.1

      Ned,

      As previously pointed out to you (with extreme clarity) both the case you want to present AND the follow on case of Faulkner v Gibbs merely concerned PRE-1952 law.

      You continue to disregard what Congress actually did in 1952.

      1. 4.1.1

        The problem is that all the solutions that would be obvious to a POSITA needs to be included in the scope of the claim. And the DOE is not enough.

        That forces functional language.

      1. 4.2.1

        As ever, Ned is wrong (and you should know better than to try to use as a “back-up case” another case that was PRE-1952.

        Greg – sorry bud, what you want is just not there.

    2. 4.3

      All of these problems go away if scope of enablement is used.

      There is nothing wrong with functionally claiming the “novel” element.

  8. 3

    Here, the court does not cite Halliburton but does note that the “asserted advance over the prior art” is the particular layered arrangement of the device, not “the choices of materials to perform each of the required catalytic processes.” Rather, those materials were expected to be ones already well known in the industry.

    Zing.

    Also worth nothing that both parties’ experts agreed that one skilled in the art would recognize specific, objectively described materials that fell within the scope of the disputed language. In addition, the specification disclosed both specific materials and tests for the adequacy of those materials.

    Contrast this with the reams of claims out there reciting functional language at the point of novelty where there is nothing more in the specification except ineligible vapor, e.g., a “module” described in functional terms. a “correlation”, a bit of logic, a list of rules, or some math.

    1. 3.1

      MM, do you mean that I cannot claim the new transporter (which I might actually have working in my garage) by describing in detail everything we saw in the Star Trek series, and then by drawing a black box and describing that black box as being” “configured to carry out the function of transporting?”

  9. 2

    Nice summary but my one complaint is the your synopsis of the case focuses on what a PHOSITA (i.e., a 103 person) would know as opposed to what a a person skilled in the art (i.e., a 112 person) would know.
    I agree the experts evoke PHOSITA in their declarations but the court on the whole largely defines the question to what a “relevant skilled artisan” would know.
    I recommend not confusing the 103 person with the 112 person.

    1. 2.1

      That is an interesting point, as to whether the §103 “ordinary artisan” is or is not the same person as the §112 “ordinary artisan.” It would be nice if the case law were a little more explicit (or consistent) on that question.

      1. 2.1.1

        If there is a 103 person and a different 112 person, is there also a different person through whose eyes it is decided whether the claim gets priority under 119 or 120, and a different person again through whose eyes it is decided what the claim covers?

        Or is it the 103 person?

        Or is it the 112 person?

        The point is not trivial. In England, for example, the 112 person can indeed be special, and different from the skilled person of the other Sections of the Act.

    2. 2.2

      I recommend not confusing the 103 person with the 112 person.

      I too am very interested in your view as to these being different persons.

      I was taught (way back when) that they are NOT different persons (in the US sovereign, there is only one legal fictional person that is represented by the Person Having Ordinary Skill In The Art, and this single person is in play for all of the patent doctrines calling on this legal fiction.

      “relevant skilled” is no more than that which determines the level of ordinary skill.

  10. 1

    Here, of course, the Federal Circuit is not suggesting that Halliburton is good law…

    I am not sure what to make of that “of course.” I am also not sure what to make of the “revive” in the title. Halliburton is good law. All that 112(3) did was to provide a sort of band-aid savings clause, so that if you had disclosed adequate structure in your spec to account for the function recited at the point of novelty, instead of getting nothing, you get the disclosed structure (plus equivalents). If you neglect to avail yourself of 112(f) (as it has since been reconfigured) and claim in a manner function at the point of novelty, you still get nothing, as per Halliburton. Nothing about 112(f) (or any other provision of the 1952 act or its sequelae) has changed that aspect of Halliburton

      1. 1.1.1

        Halliburton was abrogated with the Act of 1952.

        Well, yes, certain aspects of Halliburton were abrogated. Others were not. If you really think that the whole matter can be neatly summarized as “the 1952 act completely over-turned Halliburton,” you might want to increase your malpractice coverage.

        1. 1.1.1.1

          The CASE was abroagted – not “some portions” of the case.

          Who exactly do you think gets to decide which portions? You? Ned? Malcolm?

          Give.
          It.
          A.
          Rest.

          1. 1.1.1.1.1

            Who exactly do you think gets to decide which portions?

            The courts, including the Supreme Court (for better or worse) in the final analysis. Cite me the Supreme Court case that says that Halliburton was entirely abrogated by the 1952 act.

            1. 1.1.1.1.1.1

              Same with you: give me a case that says it was partially abrogated (and which portion).

              You have already volunteered an admission that the case was at least partially abrogated.

              Further, the courts DO NOT get the say as to making a zombie out of a case that Congress has abrogated with a change in the statutory law.

              You need to go back to basics as to when common law may be applied (especially in light of the direct Constitutional allocation of authority when it comes to the statutory law that is patent law).

              1. 1.1.1.1.1.1.1

                [G]ive me a case that says it was partially abrogated (and which portion).

                Warner-Jenkinson Co. v. Hilton Davis Chem. Co., 520 U.S. 17, 28 (1997)

                Congress enacted § 112, P6 in response to Halliburton…, which rejected claims that “do not describe the invention but use ‘conveniently functional language at the exact point of novelty,'[“]… Section 112, P6 now expressly allows so-called “means” claims, with the proviso that application of the broad literal language of such claims must be limited to only those means that are “equivalent” to the actual means shown in the patent specification… § 112, P6 was enacted as a targeted cure to a specific problem…

                N.B. two points. First, the Court describes §112(3) as a “targeted cure,” not a “sweeping reversal.” Second, the reason §112(3) constitutes any sort of “cure” to the Halliburton “problem” is that “the broad literal language of such claims must be limited to only those means that are ‘equivalent’ to the actual means shown in the patent specification,” which is as much as to say that §112(3) “solves” the problem by turning what might appear to be functional language into de facto structural language. Where the language remains functional at the point of novelty, however, nothing in the 1952 act gives one any reason to expect a different outcome than the one that obtained in Halliburton.

    1. 1.2

      ps: see the Prof. Crouch coined term of Vast Midle Ground, and please realize that the option of 112(f) was for claims TOTALLY functional (and that claims with some elements in terms sounding in function are fully legitimate OUTSIDE OF the optional realm of 112(f).

        1. 1.2.1.1

          I know you’re being sarcastic, but it’s worth noting again that “TOTALLY functional” is not the correct standard. If the structure is not sufficiently definite for performing the claimed function, the claim element is not within the “Middle Ground”. So “3 inch bladed means for cutting hair” could be a 112f element despite not being “TOTALLY functional”.

          1. 1.2.1.2.1

            “TOTALLY functional” vs “sufficiently definite for performing the function” is not a matter of form. If the former was the standard, “processor configured to X” would be a per se non
            -112f limitation. Under the correct standard, whether that limitation invokes 112f is ripe for clarification in the post Williamson world.

Comments are closed.