Computer Implemented Means-Plus-Function Element Must be Supported by Specific Algorithm in Specification

ScreenShot014Aristocrat Technologies Australia (ATA) v. International Gaming Technology (IGT) (2007-1419) (Fed. Cir. 2008)

ATA and IGT have two parallel cases pending before the CAFC. This appeal focuses on whether ATA’s asserted claims are invalid as indefinite. (U.S. Pat. No. 6,093,102). The other pending case questions whether the PTO properly revived ATA’s unintentionally abandoned national stage application. (U.S. Pat. Nos. 7,056,215 and 7,108,603) [See Patently-O discussion of revival case].

This case focuses on a patent covering an electronic slot machine patent that allows players to pre-select which combination of symbol locations will be used to determine a winner.

Claims Indefinite: The district court held the claims invalid as indefinite under 35 U.S.C. § 112 ¶ 2. Its opinion focused on the claimed means-plus-function element “game control means” — finding the term indefinite because the specification did not provide “structure” to perform the claimed functions.  The specification did include the statement that the control means could be a microprocessor-based gaming machine with “appropriate programming.” However, the lower court focused on the absence of any “specific algorithm” or “step-by-step process for performing the claimed functions.”

On appeal, the CAFC affirmed, holding that computer implemented means-plus-function claims must include “more than simply a general purpose computer or microprocessor.”  The purpose of the requirement, according to Judge Bryson’s opinion, is to avoid overbroad “pure functional” claims.  Because of the ubiquity of general purpose computers, a bare microprocessor cannot be considered a sufficiently specific structural disclosure:

For a patentee to claim a means for performing a particular function and then to disclose only a general purpose computer as the structure designed to perform that function amounts to pure functional claiming.

Within PHOSITA’s Ability: The algorithm missing from ATA’s disclosure is almost certainly within the ability of an ordinary game programmer.  In fact, I believe that it would only take me (a law professor) only a few hours to create a rough algorithm showing how the ATA software program could operate.  Judge Bryson shot-down that argument: Structure for a means-plus-function element must be provided in the specification regardless of whether the structure is already well known in the art.

What is Required: “Source code” is not required to be disclosed, nor is a “highly detailed description of the algorithm.”  However, a computer oriented application using means-plus-function claims must “at least disclose the algorithm that transforms the general purpose microprocessor to a “special purpose computer programmed to perform the disclosed algorithm.” (Quoting WMS Gaming)

Notes:

59 thoughts on “Computer Implemented Means-Plus-Function Element Must be Supported by Specific Algorithm in Specification

  1. What if the claim recited “a microprocessor configured to perform function X”?

    Would the court hold that claim invalid for indefiniteness?

    After all, if one of ordinary skill can’t figure out whether their product (1) performs a function and (2) is a microprocessor or an equivalent thereof for purposes of 112p6, how could they do the same if the microprocessor was affirmatively recited in non mpf form?

  2. Thanks Paul. Page 143 of the book. Somebody once said “Two wrongs don’t make a right”. A “Statutory Mandate” on how MPF wording in a claim “must” be construed strikes me as an unfortunate reaction to Walker’s claim, because it imports into US patent law one more needless rigidity. There I go again, writing about how the law might be, instead of how it is. It’s so much more fun though.

  3. If you read the chapter on functional claims in “Fundamentals of Patent Drafting” published by the Chartered Institute of Patent Attorneys in the UK you will find a chapter on functional claims in which the provisions of 35 USC 112(6) are described in detail. Max Drei, incidentally, has a copy.

    It explains the origin of the sub-section with reference to the US Supreme Court decision in Halliburton Oil Well in 1946 [we Englishmen have a respect for history].

    Nowadays we must all draft defensively. That means, inter alia, checking for the existence of structure corresponding to each of the integers we introduce, describing them in appropriate detail and considering and mentioning all foreseeable altenatives because as we are repeatedly warned by the CAFC if a variant is not mentioned in the written description it may not be covered.

    Including a means for or step for limitation without an adequately supporting flowchart or the like is simply courting trouble. Even if your language is not explicitly caught by 35 USC 112(6) it might be held to be within the section so it is best to be on the safe side. Furthermore, compliance with 35 USC 112(6) for all elements of a structure or method is simply good drafting.

    None of this is rocket science!

  4. “Where does this statement come from?”

    From section 112, paragraph 6 itself. The spec has to disclose the structure.

    Bryson, however, seems to think that structure must be a more detailed algorithm, but not source code. If someone can provide a rationale based on the statute for that distinction, I’d love to hear it.

  5. Judge Bryson shot-down that argument: “Structure for a means-plus-function element must be provided in the specification regardless of whether the structure is already well known in the art.”

    I don’t find this statement in the CAFC opinion. Where does this statement come from?

  6. “In other words, you had no factual basis whatsoever in making this statement. Since it is suggested above that you are a law professor, you should know better than to make such unsupported, stereotypical assertions! :-)”

    You all are killing me! At least you had a smiley face… Perhaps I should have said “1L’s.” I had plenty of facts: the strong sales of hornbooks, outlines, and bar review courses (of which I was a customer at one time) implies the law student hankering for black letter law.

  7. Mr. Risch wrote above: “Ouch! I wasn’t talking about MY school – my patent law students did just fine. I was speaking more generally.”

    In other words, you had no factual basis whatsoever in making this statement. Since it is suggested above that you are a law professor, you should know better than to make such unsupported, stereotypical assertions! :-)

  8. “A broad claim, by definition, needs to leave a few things out.”

    The difference between a claim reciting “an antibody that specifically recognizes enzyme X” (CAFC wholeheartedly approved in the absence of data) and a “drug that specifically blocks the activity of enzyme X” (CAFC wholeheartedly disapproved in the absence of data) is quantitative, not qualitative.

    It’s interesting, is all I’m sayin’.

  9. we can all go home, Mooney officially has all the answers…

    And I know cavemen; someone who doesn’t know much about what they invented is no caveman.

    By the way Mooney you seem to be falling into a trap of thinking that a claim needs to be some sort of specification. It doesn’t. A broad claim, by definition, needs to leave a few things out.

  10. “who cares if they get a PBP claim as well?”

    Who cares? Potential infringers who can not determine if their compound falls within the scope of the claim. Also, people who wonder how the PTO can possibly determine whether the resultant compound previously existed without a concrete, description of its properties.

    I might add, also, that a similar problem also arises in the context of purely functional claims, e.g., a claim such as this one:

    1. An isolated antibody that specifically binds to the polypeptide of SEQ ID NO:464.

    (referred to in a recent Patent Docs article link to patentdocs.net)

    Note that the specification does not describe a single example of such an antibody, but merely describes how one would go about making the antibody. As the Pat Docs article notes, the CAFC has expressly approved this practice in Noelle v. Lederman (2004). The key word in the claim is, of course, “specifically”. Assuming that the term is *not* given its most extreme definition (i.e., capable of unambiguously distinguishing the protein of SEQ ID 464 from any other protein on the planet under routinely stringent conditions) the PTO clearly has no means of determining whether such an antibody has already been isolated. Is that a problem? Not now. It’s potentially a policy question.

  11. “A PBP claim could only be asserted against a defendant who used the process to make the product. It would not stop anyone from making the product in a completely different manner.”

    First of all, I think I missed the en banc Federal Circuit case standing for that proposition. Can you remind me what case that is?

    Second, as a defendant, if I think my product differs from “the product” referred to in the preamble, do I infringe regardless just because I’m practicing the recited steps? What if the claim uses “comprising” language and I use additional steps?

    The way you describe product-by-process claim, they sound an awful lot like method claims to me. What’s the difference, Lionel, in your mind?

  12. Malcolm,

    A PBP claim could only be asserted against a defendant who used the process to make the product. It would not stop anyone from making the product in a completely different manner. If the dumb infringer were using the exact process the original inventor, then he can be enjoined just like any other innocent infringer.

    If your actual point is that many of the entities obtaining such claims are much more sophisticated and actually know what is in the product, they probably should just claim the compound and move on. However, who cares if they get a PBP claim as well?

  13. “Instead this case is a straight forward application of WMS Gaming.”

    There is no application of WMS Gaming here. WMS Gaming never held that the an algorithm must be disclosed in order to have sufficient corresponding structure. At most, WMS Gaming held that, when an algorithm is disclosed, it’s part of the corresponding structure.

  14. “Hey, don’t dog on the law students. We’re not all as unaware as those in your WVU classes professor. I bet Dennis’ students know the difference between black and gray patent law.”

    Ouch! I wasn’t talking about MY school – my patent law students did just fine. I was speaking more generally.

  15. For those agonizing over what this case means, I’d recommend looking at the patent including all *three* columns of detailed description to see how “disclosure” there was to work with.

    This case doesn’t really speak against describing structural elements in terms of function either completely or partially. Instead this case is a straight forward application of WMS Gaming.

  16. “Actually, it is possible that people have neither the expertise or the inclination to determine the exact composition and that’s as it should be.”

    Really? Huh. What happens to the alleged infringer who makes that argument in his/her defense, i.e., “I’m just a caveman, I just make the stuff and sell it, I don’t know much else about it.”

    Think that approach will work as a non-infringement defense? If not, why not? Why should it not work? Doesn’t it seem unfair that only the first person who is incapable of describing their composition gets exclusive rights to it? After all, it’s even more challenging for later folks to remain incapable of describing the composition after it’s been publically identified as impossible to describe, the gauntlet having been thrown down, so to speak.

  17. I would say Hutz is right, pretty much the only time they’re appropriate is when you’re too dumb (either by the state of science or whatever) to know what’s actually going on. Funny enough, I have a case right now that might fit just that bill.

  18. Malcom Mooney wrote “Yes I have. And I am not aware of any composition that has been conceived and reduced to practice that can not be claimed as a composition, i.e., by describing its physical properties.”

    Actually, it is possible that people have neither the expertise or the inclination to determine the exact composition and that’s as it should be.

    I believe one of the quintessential PBP claims was to the guy who invented superglue or some other superstrong adhesive. I believe the stiry was that he was working with chemicals related to developing photographs and accidentally produced this adhesive. He was not a chemist, but he was able to say that you can mix substance A with substance B to get adhesive C.

  19. “an unfortunate fact that most law students have not yet figured out”

    Hey, don’t dog on the law students. We’re not all as unaware as those in your WVU classes professor. I bet Dennis’ students know the difference between black and gray patent law.

  20. “In other words SF believes M+F claims should cost 10x what they do today.”

    No, but nice try. The patent office ignores the corresponding structure as a matter of policy and searches, and rejects, based on any structure that performs the function, regardless of equivalency. Opinion counsel does not have that luxury.

  21. “If you “claim a component in terms of its functionality” the examiners these days are likely to construe the “functionality” portion of the claim as an “intended use” and therefore accord it no patentable weight.”

    Don’t claim an intended use and you’re in the clear. There is a difference between a “barely decent” functional claim and in rubbish intended use claims.

  22. “”There is zilcho advantage to M+F claims other than keeping the claims themsleves shorter than they otherwise would be.”

    When included with non M+F claims, M+F claims help to provide a stronger patent thicket because the M+F claims require more analysis for non-infringment/invalidity opinions. It’s probably worth the couple of hundred bucks for an extra independent claim that adds 10 to 25 times that in your competitor’s opinion counsel’s fees.”

    In other words SF believes M+F claims should cost 10x what they do today.

  23. If you “claim a component in terms of its functionality” the examiners these days are likely to construe the “functionality” portion of the claim as an “intended use” and therefore accord it no patentable weight.

    Any comments, anyone?

  24. “When included with non M+F claims, M+F claims help to provide a stronger patent thicket because the M+F claims require more analysis for non-infringment/invalidity opinions.”

    Ha! Well, yes, that is surely true. Long live uncertainty! ;P

  25. “I tried to explain it to him the other day, and I’m sure he’s not so technically deficient as to not get what we’re saying but he hasn’t quite looked at the issue through and through.”

    Yes I have. And I am not aware of any composition that has been conceived and reduced to practice that can not be claimed as a composition, i.e., by describing its physical properties.

    “That said, he’s right that it’s rare that a situation arises where using that format is really called for”

    There is no situation where it’s called for. This argument is what I refer to as Newman’s Fallacy. Granted, she didn’t make up the argument. Some clueless dudes in the 19th century made up the argument. But I believe it’s a form of question begging. No composition has ever been conceived and reduced to practice that can not be described by its properties. Conversely, if you can not describe something that you “made”, all you have “invented” is a method.

    At the end of the day, product by process claims are simply method claims. Construing them to cover anything but the method steps (and ONLY the recited method steps — no step can be deleted or added) is absurd, unfair, and hypocritical. After all, if you can’t characterize the product/composition then there is no way to determine if an infringer who is not following the recited steps is making the product.

    This is quite straightforward, actually. You do need to shrug off a more than a hundred years worth of hogwash about those alleged “otherwise impossible to claim” products, however.

  26. “There is zilcho advantage to M+F claims other than keeping the claims themsleves shorter than they otherwise would be.”

    When included with non M+F claims, M+F claims help to provide a stronger patent thicket because the M+F claims require more analysis for non-infringment/invalidity opinions. It’s probably worth the couple of hundred bucks for an extra independent claim that adds 10 to 25 times that in your competitor’s opinion counsel’s fees.

  27. “But to me the really legally interesting aspect is that it de facto imposes a far stricter and higher specification disclosure requirement for “means-function” claim elements than the law does for 112 “enablement” of NON “means-function” claim elements!”

    That’s correct, and it’s a good thing. M+F claims are a claiming *shortcut*, i.e., a special gift for patent claim drafters who, for whatever reason, like to keep their patent claims short.

    There is zilcho advantage to M+F claims other than keeping the claims themsleves shorter than they otherwise would be. Why should it be otherwise?

  28. “There may be varying degrees of gray but I don’t think that there is anything anymore which can be considered “black letter law” in patent law.”

    Fair enough – and true for almost every area of law – an unfortunate fact that most law students have not yet figured out.

    That said, this one seems to come right out of the statute, and I’ve yet to see a case to the contrary – not that there are a lot of cases on this issue. If you have a means but don’t give the structure, then the claim is indefinite. Of course there are lots of shades, which is why I put it on my exam – the court here could have said this structure was enough, for example.

  29. “But to me the really legally interesting aspect is that it de facto imposes a far stricter and higher specification disclosure requirement for “means-function” claim elements than the law does for 112 “enablement” of NON “means-function” claim elements! ”

    Paul,

    I think the court addressed that. They warned against “conflating” the enablement requirement of 112, 1st with the requirement to disclose the structure that performs the claimed function under 112, 6th. See the paragraph spanning pp. 13 – 14. The court did not say you had to do anything special, or extraordinary, to “enable” the mpf, it just said you had to disclose the structure that performs the claimed function.

    “There is considerable case law saying that enablement can be met by showing that a POSITA can do it “without undue experimentation.” But apparently that will NOT work for “means-function” claim elements.”

    The court addressed this on page 14. There is no change in the law of enablement in this case.

    It’s still entirely possible to: 1) satisfy the requirements of 112, 6th while failing to enable the invention under 112, 1st or 2) enable the invention under 112, 1st while failing to satisfy the requirements of 112, 6th.

  30. “Do you really want to get rid of all functional language in claims?? How about terms such as radiator, light, coater, shield, igniter, and mixer?”

    Ahem, Purely Functional, not structurally indicative functional, those are fine obviously.

    Yes MM doesn’t seem to be a fan of pbp, I tried to explain it to him the other day, and I’m sure he’s not so technically deficient as to not get what we’re saying but he hasn’t quite looked at the issue through and through. That said, he’s right that it’s rare that a situation arises where using that format is really called for, even though the process of making might affect the structure (micro or even macro) being made. They only need to be used as a last resort, and also they need to be used in proper format, specifically stating it is a product made by a process. That also means not introducing the claim as a process and just magically including structural elements rather than steps, or trying to rely on non-process manipulative structural limitations for patentability.

  31. From the decision: “Because general purpose computers can be programmed to perform very different tasks in very different ways, simply disclosing a computer as the structure designated to perform a particular function does not limit the scope of the claim to “the corresponding structure, material, or acts” that perform the function, as required by section 112 paragraph 6.”

    What if the applicant said “adhesive” was a corresponding structure in a claim? Aren’t there an infinite number of very different adhesives that rely on very different combinations of materials with very different molecular structures?

  32. My God … it must be Monday … not only a Monday, but a Fifth Monday in a month as I’m agreeing with both MM and 6K (not as to everything, but as to means plus function claims). Personally, I never use them, and if I run across claims that includes them, I amend the claims (if I’m not worried about PHE) to remove them. There are just too many headaches associated with them for them to be useful to me.

    However, I cannot let 6K and MM go without a little disagreement ….

    6K writes “In other news, we should get rid of means plus function, and also functional, language in claims because people just do not understand what they are.”

    Just another example of where 6K doesn’t think through his proposal. Do you really want to get rid of all functional language in claims?? How about terms such as radiator, light, coater, shield, igniter, and mixer? All of these terms are essentially functional language. Many features are named for what they do (i.e., functional language).

    As for MM, your dislike of product-by-process claims reflects a lack of understanding as to how processes affect structure. This isn’t surprising since most people really don’t understand what happens to the microstructure of many “objects” depending upon what process is used. I think this reflects MM’s chemical/biotech experience. I imagine there are many instances in which a particular isolated product from one reaction may not differ from the same product isolated from a different reaction. However, a device produced from one process may differ significantly from a device produced from another process despite the devices prepared by the different processes looking identical.

  33. if you are amending a claim element to overcome prior art you might consider replacing it with a MPF element so you will at least get the “equivalence” to which MPF entitled you

  34. Lorenzo stated “I rarely use means-plus-function format anymore. I would simply have recited “a game controller” in the claims, and supported it in the spec with a block labeled “game controller” and a simple flow diagram–maybe two or three steps of receiving input, producing a response that controls ________, etc.– and explained that the controller can be a microprocessor, ASIC, hard-wired logic, combination thereof, etc. Is this still a reasonable practice?”

    Absolutely. I cannot believe anyone today still uses MPF claims. It is so easy to claim a component in terms of its functionality with the type of detail Lorenzo proposed.

  35. Perhaps a good flowchart might save patents like this, in spite of the many CAFC decisions for many years now demonstrating the dangers of using “means-function” claims. But there are still large numbers of existing patents being asserted with “means for” claim elements and a mere general indication in the specification that the claimed function can be be performed by programming a general purpose computer, WITHOUT any flowchart, algorithms, or the like. This now seems to be potentially fatal, at least with CAFC panels like this one.
    But to me the really legally interesting aspect is that it de facto imposes a far stricter and higher specification disclosure requirement for “means-function” claim elements than the law does for 112 “enablement” of NON “means-function” claim elements! There is considerable case law saying that enablement can be met by showing that a POSITA can do it “without undue experimentation.” But apparently that will NOT work for “means-function” claim elements. Although also not discussed by the Court, as one commentator above notes, this very different spec support requirement for “means for” claims can be argued to be justified by the need for meeting the special and different infringement scope definition for U.S. “means for” claims in 35 USC 112, which is limited to what is shown in the spec and ITS equivalents, not equivalents in general. I.e., with no structure in the spec corresponding to the “means” [other than maybe an empty black box labeled "computer" or "processor" in some cases], how can one determine this statutory-defined scope of a “means-function” claim?
    {You don’t have to love this situation, but you apparently have to live with it.}

  36. “, you never know how the CAFC will be interpretting claims a decade from now and whether there may be another pendulum swing that will bring measn-plus-function claims back into favor. ”

    True. The issue boils down to competence: if you are going to rely on means-plus-fxn language, don’t do it in a half-axxed manner. Do it right. If you find yourself asking “I wonder if this m+f claim will stand up in litigation,” then assume that it won’t and try again.

  37. Perhaps Dennis could generate a plot of the rise and fall over the last 20+ years of whether means-plus-function claims are in vogue or not! Certainly, sounds like we should be shorting them for now. Still, you never know how the CAFC will be interpretting claims a decade from now and whether there may be another pendulum swing that will bring measn-plus-function claims back into favor. All of which argues for a higher number of independent claims/application as the safe bet is always to take the time to write multiple variations of claim sets from different perspectives

  38. Prof. Risch

    There may be varying degrees of gray but I don’t think that there is anything anymore which can be considered “black letter law” in patent law. You just have to follow the bouncing FedCir panel and which cases the SupCt decides to take :-)

  39. “I rarely use means-plus-function format anymore. I would simply have recited “a game controller” in the claims”

    Ding! Ding! Ding! We have a winnah!

    As with product-by-process claims, the means-plus-fxn claim is a risk not worth taking.

  40. Mr Risch thanks but, in the words of the song, “You’ve got to be joking, right?”.

    - – - -

    Well, I shouldn’t have said “you don’t want…” Of course the patentee wants something that broad – only public policy doesn’t.

    I was also making no real statement about foreign rules, and I frankly know little about the other requirements. I do know, however, if the structural limitation requirement went away in the U.S. it would eviscerate written description even more than that it already is for computer software.

  41. Sorry 6K you lost me yet again. You state as a fact: “Functional language are nowhere near as simple as they may seem on their face, and that is their malice”. But they are, inherently, as simple as they seem, and they are, inherently, free of malice. I’m sorry you feel otherwise. Brought on by having too much to do with prosecuting US attorneys, and reading too much US caselaw, I shouldn’t wonder.

  42. “#6K invokes a public notice notion which strikes me as a circular argument. Because the Statute has 112,6 in it, one needs special rules for claim interpretation. My question was rather: why 112,6? Other countries don’t have it, yet get by alright, on clarity and public notice. 6K makes the extraordinary assertion that functional recitations in claims are beyond the understanding of readers. #6K, from where I’m standing, they’re often the easiest part of the claim to grasp (but of course only when you read a claim free from any “magic words” superaddition)”

    Hardly circular, it goes in two distinctly divergent paths depending on which language you propose to use. I believe you mean to ask “why have 112 6th para?”. You’re right, there is no purpose to having it other than to cause trouble. Some lawmaker somewhere thought it’d make for good chuckles around the lawyer table for years to come so he chucked it in (this is what happened as near as I can tell anyway). You’re right, I say take away functional language, sure it’s easy to understand “hur their claim encompasses things that do that functionality”, but that’s not really as far as the inquiry goes to the informed reader and you know this. Thus, the avg. joe is left “out of the loop” on this extra inquiry(s). Functional language are nowhere near as simple as they may seem on their face and that is their malice. A judge somewhere stated it quite nicely when he said something along the lines of “it is a fault? of funtional language when the applicant claims the prior art right up to where the point of novelty resides only in the functional limitations of a claim” <<<< paraphrased, it’s late and I’m headed home.

    Functional limitations are a devil, and I hope they bite anyone that uses them, there’s simply no need in the vast majority of cases. Other than “to get people”. Oh, that’s just great, no really, it is.

  43. I rarely use means-plus-function format anymore. I would simply have recited “a game controller” in the claims, and supported it in the spec with a block labeled “game controller” and a simple flow diagram–maybe two or three steps of receiving input, producing a response that controls ________, etc.– and explained that the controller can be a microprocessor, ASIC, hard-wired logic, combination thereof, etc. Is this still a reasonable practice?

  44. MPEP § 2181(1) states:

    ——————————
    A claim limitation will be presumed to invoke 35 U.S.C. 112, sixth paragraph, if it meets the following
    3-prong analysis:
    (A)the claim limitations must use the phrase “means for” or “step for;”
    (B)the “means for” or “step for” must be modified
    by functional language; and
    (C)the phrase “means for” or “step for” must not be modified by sufficient structure, material, or acts for achieving the specified function.
    ——————————-

    However, prong (A) is rebutted shortly afterwards:

    ——————————-
    Seal-Flex, Inc. v. Athletic Track and Court Construction, 172 F.3d 836, 850, 50 USPQ2d 1225, 1234 (Fed. Cir. 1999) (Radar, J., concurring) (“claim elements without express step-plus-function language may nevertheless fall within 112 6 if they merely claim the underlying function without recitation of acts for performing that function… In general terms, the underlying function’ of a method claim element corresponds to what that element ultimately accomplishes in relationship to what the other elements of the claim and the claim as a whole accomplish. Acts,’ on the other hand, correspond to how the function is accomplished…If the claim element uses the phrase step for,’ then § 112, 6 is presumed to apply…On the other hand, the term step’ alone and the phrase steps of’ tend to show that § 112, 6 does not govern that limitation.”)

    Personalized Media Communications LLC v. ITC, 161 F.3d 696, 703– 04, 48 USPQ2d 1880, 1886– 87 (Fed. Cir. 1998); Mas-Hamilton Group v. LaGard Inc., 156 F.3d 1206, 1213, 48 USPQ2d 1010, 1016 (Fed. Cir. 1998) (“lever moving element for moving the lever” and “movable link member for holding the lever…and for releasing the lever” were construed as means-plus-function limitations invoking 35 U.S.C. 112, sixth paragraph since the claimed limitations were described in terms of their function not their mechanical structure)
    ——————————-

    Yet the MPEP nevertheless concludes, in the face of established court precedent to the contrary, that: “[t]hus, absent an express recitation of ‘means for’ or ‘step for’ in the limitation, the broadest reasonable interpretation will not be limited to ‘corresponding structure…and equivalents thereof.’”

    I think Mr. Bloom’s prediction/fear is a logical result based on this case and the cases cited above (particularly Judge Rader’s opinion in Seal-Flex). If we consider the Seal-Flex opinion and prong (C) given above by the MPEP, it is clear that 112 6th is invoked if the claim does not provide ***sufficient*** structure/step to perform the claimed function. The critical holding in this case is that, due to the nature of the art, generic microprocessors ***are not sufficient recitation of structure***. When taken together, the logical conclusion is that claims to generic microprocessors for performing XYZ function are in means-plus-function form that lack sufficient structure and therefore presumptively invoke 112 6th.

    It appears to me that the public policy objective is actually one directed to breadth (which, as we know, is typically treated under 102/103 during prosecution). Specifically, claims directed to “a microprocessor configured to X” fairly read on all microprocessors and all algorithms that can achieve X – which, in my view, draws close to the 101 analysis of pre-empting all practical applications of an abstract idea. While the claim certainly narrows the scope to at least computer-implemented algorithms and thus may escape 101, the claim still reads on all algorithms past, present, and future that could possibly achieve the claimed function. Given the nature of software, which is already somewhat resistant to patent protection, it may be good policy to restrict claims of this breadth as much as possible.

    Further points to consider:
    - compare the scope of “microprocessor for performing X”-type claims with, for instance, claims to mechanical inventions.
    - that a PHOSITA would likely be able to develop a program for performing the function without explicit recitation is telling of the nature of software in general and how patent policy should be handled with regards to the software industry. In particular, given a particular result, most programmers would be able to code something to achieve that result without “undue experimentation”; remembering, of course, that the “undue” bar is quite high for software due to the nature of their work and the ordinary level of “experimentation” undertaken in the natural course of their work. Thus, the actual point of novelty is not the software itself (since any PHOSITA could code it), but the result – i.e., the idea. But claiming the idea per se runs into problems with 101 – we aren’t allowed to have monopolies on ideas, only specific implementations of that idea. From this, it should be clear where the logical inconsistencies are, and the reason for why (I think) we’re currently stuck resorting to the entire “computer-readable medium” quagmire and treating software as “functionally and structurally interrelated components” of otherwise well-known computer systems.

    But what do I know, I’m just an examiner.

  45. Mr Risch thanks but, in the words of the song, “You’ve got to be joking, right?”. You “don’t want” a claim that covers an alternative cutting means. As John McInroe once said “You cannot be serious”. ROW does have a “sufficiency” requirement, you know, and its equivalent to Liebel-Flarsheim. Might I ask: which jurisdiction is it that you have in mind when you write that it “doesn’t matter” that the reader can’t build the claimed widget? Do other readers agree with Michael Risch? The gulf between the USA and the ROW is bigger than I imagined. Dennis, something else here to add to your list of “subsidies” that US patent law bestows on ROW.

  46. [quote]
    Thus “means for cutting” for example means “everything under the sun capable of cutting” which is quite a lot when EPO Exrs give that claim term its widest credible meaning.
    [/quote]

    Here’s a reason why: When I develop a new cheeseboard with a “means for cutting” that contemplates a knife and equivalents, I don’t want that claim to read on (or invalidate) a later invention of a cheeseboard that uses a laser as the “means for cutting.”

    A secondary principle is that allowing for such broad claims without support winds up eviscerating the written description requirement and allowing inventions on things that the patentee is actually unable to build. All the patent need say is “any PHOSITA can build Widget X, it doesn’t matter that I haven’t figured out how to do so”

  47. Thanks everybody, esp Leopold Bloom, who seems to be the only one of you able to step back and look at the Statute on a “fitness for purpose” standard. Leopold used the magic words “magic words”. When it comes to words in claims in the USA, some words are more magic than others. That’s the puzzle, to outsiders. Why not just give every word in every claim the meaning that an ordinary PHOSITA would give it. That’s what we do in Europe. Thus “means for cutting” for example means “everything under the sun capable of cutting” which is quite a lot when EPO Exrs give that claim term its widest credible meaning. Consequence: Applicant might have to render his claim clearer to achieve novelty. You would say “more definite”. But then we have Art 69 EPC to help patent owners, when it comes to scope of protection.

    #6K invokes a public notice notion which strikes me as a circular argument. Because the Statute has 112,6 in it, one needs special rules for claim interpretation. My question was rather: why 112,6? Other countries don’t have it, yet get by alright, on clarity and public notice. 6K makes the extraordinary assertion that functional recitations in claims are beyond the understanding of readers. #6K, from where I’m standing, they’re often the easiest part of the claim to grasp (but of course only when you read a claim free from any “magic words” superaddition)

    Somebody asked whether Europe would have come to the same result as the CAFC. I think you know the answer already: NO. I don’t feel like attempting to answer what might have been the intended question: what would Europe have decided? However, I can imagine the claim at the EPO meeting all requirements for allowance. Why shouldn’t this inventor get his patent? Only because his patent attorney used the wrong magic word? Seems disproportionate to me, but of course I’m not qualified to pass any opinion on that.

  48. I don’t know any that are “automatically” treated as means plus function or that are presumptively treated as such, other than the actual “mean for” words themselves.

    However, words such as “unit”, “module”, “section”, etc. can be a bit problematic, particularly when followed by a function and/or the word “for”. As you suggest, these type of problems usually arise in litigation.

  49. To me the issue is how narrowly/broadly you define the “structure.” Claiming “a generic processor to perform function XYZ” effectively equates to claiming “a structural means for performing function XYZ” in an electro/mechanical device. My question was: are there specifically known structural buzzwords (like maybe “processor”) that courts/examiners presumptively treat as “means for” elements? Or, must the claims be litigated every time to determine–in light of the facts, disclosure, etc.–that this claimed element is actually purely “functional.” To me, it’s unclear when a processor, ‘drawn as a square,’ represents a “structural” or “functional” item.

    So, does anybody know any buzzwords that are automatically treated like “means for” consistently, and without regard to the particular claims/patent?

    My guess was that they don’t exist.

  50. An element in a claim for a combination may be expressed as a means or step for
    performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.

    112 6th^

    Now, let’s be clear, the CAFC is NOT construing a processor as not being a definite structure, it is a very definite structure, it is only when you begin to start spouting off functional language that more needs to be specified (as in the case Dennis shows us) in terms of the algorithm used to perform that function. That algorithm doesn’t even have to be in the claim btw, just somewhere in the spec.

    Besides that a processor IS being considered a definite structure we notice a few other things in your example claims. In the first instance you have a claim that basically states “A microprocessor (and maybe it doing a functionlol, examiner please don’t cite caselaw about a structure that can perform a function)” (<<< citing a structure that gets 102′d) and in the second instance you have a claim that states “A processor in my specification that is configured to perform function A just like I have shown you how to configure it in my specification” and definitively invoking 112 6th to tie your claim to the specific structure/method of/algorithm for performing the function in the spec. and probably not getting 102′d quite as easily. Considering what I just told you about how your claims are construed, yes, there is a difference.

    Oh, and the public policy behind this is because when you invoke 112 6th, the “educated reader” (not the layperson, who knows nothing of 112 6th sadly :() you specifically limit your claim to the structures, materials and acts in the specification and when you DO NOT invoke 112 6th, as in first of your latest bunch of examples, you are not so limiting your claim. Thus, the public noticelol function is served by making it clear that you are not limiting your claim to the structures in the detailed desc. you are intending them to be generic parts if you do not invoke 112 6th, and if you do then you are limiting your claims thus.

    In other news, we should get rid of means plus function, and also functional, language in claims because people just do not understand what they are, and what they are there for, nor how to use them, they just throw them in whenever they feel like it. It was so funny to see how the applicant’s own arguments were the most damning evidence against him in the case above.

    Oh, and btw, yes a lot of magic invoking comes from putting those “means or step for” magic words in the claim. And unless the courts find that your claims MUST have been meant as a means or step for or is otherwise completely nonsense then it’s my understanding that you shall not be presumed to have invoked 112 6th. So no, there’s not going to be any hammer dropping with righteous judgment on your bs structure doing a function claims just yet, nor likely anytime soon.

  51. This is pretty well established black letter law, I think. I asked this EXACT question on my patent law exam. One or maybe two students caught the issue and dealt with it well. If there is no structure, then it is impossible to define the claim and its equivalents – it’s not an enablement issue.

  52. “After reading the apparently completely unthoughtout comments above I offer the following commentary.”

    After reading your commentary, it’s apparent that you didn’t you didn’t actually read the “unthoughtout [sic] comments.” Or at least you didn’t comprehend them.

    “That is not a means plus function, and never will be construed as such.”

    Nobody said that it was. But the last part is a pretty bold assertion. Care to explain why? A simple explanation would be that it doesn’t include the magic words “means” or “for,” but that would be wrong.

    How about this: “A microprocessor for performing function A.” Is this really any different than “A processing means for performing function A”? Or “A microprocessor configured to perform function A”? Especially since the CAFC apparently seems to believe that a processor by itself is not a definite structure? But this leads to Max’s real question (which you apparently didn’t comprehend either) – what’s the public policy behind making the validity of a claim depend on whether you used the magic words or not, when there’s apparently no problem with a PHOSITA understanding the scope of the claim?

  53. Also, after reading the apparently completely unthoughtout comments above I offer the following commentary.

    Link to very helpful judgment: link to cafc.uscourts.gov (You might want to link that in your post Dennis so people aren’t confuzzled come monday)

    “PHOSITA able to put the claimed subject matter into effect, to a level of generality proportionate with the scope of the claim. It is just that the presence of a “means plus function” element in the claim triggers this specific “indefiniteness” ground of invalidity.”

    No Max, it was the fact that there was no algorithm recited whatsoever. It was a purely function claim performed by a generic computer with no algorithm stated. Thus, yes, since it “lacked sufficient disclosure of the structure under 112 6th, it is therefore indefinite under 112 2nd”. See page 16, or read the whole thign.

    “”A processing unit configured to A, B, and C” is really a means-plus-function element”

    That is not a means plus function, and never will be construed as such.

    Law student, N O, just N O. If you say a processor then you mean a generic processor presumptively. If you want to make absolute sure you’ll never have a problem with being limited to a specific embodiment, draw it as a square in the drawings and define it as “a processor” in the written desc.

  54. “Judge Bryson shot-down that argument: “Structure for a means-plus-function element must be provided in the specification regardless of whether the structure is already well known in the art.””

    Judge B mah homie.

  55. Question: are there currently structural element names/terms (i.e., broadly generic identifiers) that are routinely deemed to trigger the “means for” approach for determining claim scope? For example, when (if) I write claims with elements like “a computer for controlling. . .” are the claims in danger of *automatic* limitation to the structures disclosed and their structural equivalents? Or, do these claims always need to be litigated (“a computer for” in patent A is different “than a computer for” in patent B)?

    My intuitive guess is that this would be fact specific, and things like “A processing unit configured to A, B, and C” would need to be litigated every time. But, I’m just a novice.

  56. “Maybe the comments to come will reveal all.”

    Not likely.

    Given WMS Gaming, this outcome is not too surprising. I’m waiting for the other shoe to drop – when the CAFC decides that a claim element like “A processing unit configured to A, B, and C” is really a means-plus-function element.

  57. Do I understand this case aright? No problem with 101, 102, 103. Claim clear. Claim scope commensurate with contribution to the art. PHOSITA able to put the claimed subject matter into effect, to a level of generality proportionate with the scope of the claim. It is just that the presence of a “means plus function” element in the claim triggers this specific “indefiniteness” ground of invalidity.

    I know we should accept the law as it is, not whinge about what it should be. All the same, I should just like to know what is the public policy objective, for this provision of 35 USC and this line of CAFC authority. Maybe the comments to come will reveal all.

Comments are closed.