Ex Parte Bilski: On the Briefs:

In Ex Parte Bilski, an en banc Federal Circuit plans to reconsider the scope of patentable subject matter as it relates to business methods and so called mental methods. Perhaps more importantly to the patent system as a whole, the court is considering the proper procedures going forward for determining whether a particular invention falls within the scope of 35 USC 101.

In its en banc decision, the CAFC invited non-party amici briefs, which were due April 7. (Scroll down to find the briefs).  In reading through the briefs, the first aspect that caught my attention was a common theme that institutional strengths and weaknesses of the PTO and Courts should help dictate the ultimate subject matter rule.  

  • Prof Morris: Through its examiners, the PTO has expertise in determining the technical questions of novelty, nonobviousness, and indefiniteness. On the other hand, examiners do not have the expertise to decide “philosophical and abstract” issues of statutory subject matter.
  • Prof Lemley*: Arbitrary subject matter boundaries have generally been difficult to enforce and usually result in patent attorneys using “magic words” to avoid the limits. (*NOTE: I signed Prof Lemley’s brief along with 21 other law professors.)
  • Prof Collins: A test that excludes “human cognition” elements is administrable.
  • EFF: Proposed three-step process provides a more “efficient and meaningful” way to administer the Section 101 threshold.
  • AIPLA: Section 112 should guard claim scope rather than Section 101.

Narrow or Expansive: The main thrust of the Bilski arguments, however, focus on whether patentable subject matter should be narrow or expansive. I have categorized the briefs on this axis:

Expansive Subject Matter:

  • Prof Lemley: We cannot predict the next area of innovation, and arbitrary limits on patent scope reduces incentives in those potential areas. “Bad patents” should be dealt with using the true tools of the Patent Act: Sections 102, 103, and 112. “Mental methods” should be allowed if they fall within the other requirements of patentability.
  • Regulatory Data Corp (Prof Duffy): Even under a narrow definition, applied economics is now part of the “useful arts.”  Statutory subject matter should only limit claims directed to abstract ideas, physical phenomena, or principles of nature.
  • AIPLA: We should continue to follow Diehr, State Street, & AT&T.
  • RMC: Business methods should be patentable.
  • American Express: Patenting of business and information management processes encourages the development of those useful societal tools.
  • Accenture: Business methods should be patentable regardless of any physicality limitations. Congress has deemed that business methods should be patentable via 35 USC 273.
  • Greg Aharonian: The Supreme Court’s 1876 Cochraine test does not exclusively define “process.” Rather, a patent eligible process should be broadly defined to include any process or method that yields a “useful concrete and tangible result.”
  • Koninklijke Philips: The court should be wary of relying upon precedent that focused on traditional manufacturing methods. Rather, the court should look at the broad definition of process required by Congress in 35 U.S.C. 100(b).

Narrow Subject Matter:

  • Prof Sarnoff: The court should return to the precedent of Flook. The inventive concept of a patent cannot be an abstract idea (such as hedging risk). Likewise, an information processing method must include significant post-solution activity. State Street is unconstitutionally over-broad.
  • End Software patents: Software should not be patentable even when loaded on a computer. Rather, to be patentable, there must be significant additional (non-information processing) physical activity.
  • American Civil Liberties Union: Patents mental processes would violate the first amendment.
  • EFF: There must be a technological component of a patentable invention.
  • Computer & Communications Industry Association: The CAFC should shed its “patentee-centric approach” and insted try to meet the needs of the modern world. In particular, the court should consider the systemic policy implications of its decisions. The policy implications of broader patent coverage is more litigation & rent seeking.
  • IBM: There is no sound policy for allowing business method patents.
  • American Institute of CPAs: Tax methods should not be patentable because they preempt free use of the tax laws. (Of course, the same could be said of CPAs charging corporations for their service).
  • SAP: A process should both (1) have a concrete, useful, and tangible result and (2) be “sufficiently machine-like” in order to avoid preempting work-arounds. However, software processes should be patentable.
  • Prof Collins: The court should add a “human cognition” exception to Section 101. Steps involving human cognition should receive no consideration in judging patentability.
  • Red Hat: Software patents put a huge kink in the open source software movement.
  • Financial Services Industry: State Street and its progeny are unduly broad. A token inclusion of a ‘machine’ in a claim would not render that claim patentable subject matter.
  • Dell & Microsoft: A patentable invention must operate on “something physical.” To be patentable, software should be tied to a computer and cause some physical transformation (such as movement of electrons). And, following Comiskey, a patent should not be granted under 103 if the inventor merely combined well known computer hardware with inventive but otherwise unpatentable software.

Other?

  • Yahoo! and Prof Merges: A strict “technology” requirement is too inflexible. State Street taught us that such a strict requirement does not fit well with “onrushing technology.” The Yahoo!/Merges test: a patent eligible process must itself be “stable, predictable, and reproducible” and its result must be “useful, concrete, and tangible.”  Bilski’s claims would not be eligible because they do not define a “stable” process.
  • Intellectual Property Owners Association: A process that is either implemented by a machine or that transforms matter into another state is patentable subject matter.  IPO favorably cites the Flook limitations on on information processing.
  • Business Software Alliance: Courts should err on the side of patentable subject matter because Sections 102, 103, and 112 make-up any slack. Software should be patentable. However, Bilski’s invention is not patentable because it is an abstract idea.
  • Washington State IP Law Assn: The CAFC should re-write State Street to be consistent with Supreme Court precedent.
  • Prof Morris: Subject matter questions should be avoided. Rather the PTO and courts should look to the substantive rules of 102, 103, and 112 to decide the issue. Section 101 jurisprudence has been both haphazard and unfair.

The elephant in the room is the recent Comiskey decision. There, it appeared that the court refused to give any patentable weight to the portion of the invention directed to non-statutory subject matter. In its brief, the Boston Patent Law Assn asks the court to clarify the following statement from Comiskey:

“The routine addition of modern electronics to an otherwise unpatentable invention typically creates a prima facie case of obviousness. Moreover, there is no pertinent evidence of secondary considerations because the only evidence offered is of long-felt need for the unpatentable mental process itself, not long-felt need for the combination of the mental process and a modern communication device or computer.”

Notes:

  • I signed Professor Lemley’s brief along with twenty-one other law professors. The theory behind the brief follows the IPO brief that I helped draft in the Metabolite case.

Amici Briefs:

 

 

 

 

66 thoughts on “Ex Parte Bilski: On the Briefs:

  1. The side-stepping of the other issues aside, I believe Landis proves my point: “a process that solely manipulates abstract ideas or concepts”. If maniuplation of ideas or concepts has real utility, the ideas or concepts are not abstract. As an example, a new (software) process sorts data in a manner that makes all computers 3 times faster. The act of perfoming this sorting modifies the physical properties of the computer system. The utility is clear and the results are not abstract.

    Let me put it another way, you could write any software claim in for form: a circuit for performing steps A, B and C. A processor is, afterall, nothing more than a type of circuit. This claim type has been widely accepted by both the USPTO and the courts. What is the logic behind preventing a patent for a computer-implemented method that implements steps A, B and C?

  2. “Singling out such a standard for software suggests either a misunderstanding of the technology or that there is some other motive.”

    So far as I’m aware, all method claims have this requirement.

    Keep in mind the difference between a method and a composition claim. They are quite different. Attorneys routinely forget this when practicing before me. Routinely. As in nearly every single day.

    For your benefit, Landis:

    link to djstein.com

    See Chapter 4. Also, see the portion about Computer method claims for a discussion about how badly SS was interpreted at the time of that writing. Landis feels as though SS made an exception to the rules of method claims just so that they could be patentable. He’s right that it would have had to. But he’s wrong that the courts would do such a thing.

  3. “Alternatively it could provide a tangible effect. “A computer product … on a computer accessible storage device … which causes the computer to perform the steps: compare a to b, wherein comparing a to b comprises 56 transistors turning off and then back on in a sequence of 101010101101011010101011010… with a time of 1 nanosecond or less between each transition”

    The ones and zeros represent a physical property. Would you be happier if it was 2 nanoseconds, or 100, or 10 minutes? If the turning on and off has utility, then I am confused as to the point of your example. Your logic seems to suggest that if you cannot see the change and the change happens really fast, it must not be patentable.

    Arbitrary distinctions are arbitrary.

    “This is inherently impossible to do in any reasonable way with a modern +computer method claim. Wouldn’t you agree? You said so yourself, we require these “abstractions” to understand what is going on at the transistor/atomic level in the chip. But these abstractions don’t even show up in modern claims. Abstractions of abstractions show up.”

    If one can show two levels of abstraction, a patent claim is invalid? Interesting logic. I am unaware of any support for such a novel view. Claims are by definition an abstraction of the invention (they are afterall just words). Please apply the actual rules.

    Fabricated rules are fabricated.

  4. “Suggesting that a “software” claim must recite an “output,” is, quite frankly, laughable. ”

    Alternatively it could provide a tangible effect. “A computer product … on a computer accessible storage device … which causes the computer to perform the steps: compare a to b, wherein comparing a to b comprises 56 transistors turning off and then back on in a sequence of 101010101101011010101011010… with a time of 1 nanosecond or less between each transition”

    ALLOWED!

  5. “Moreover,recitation of steps carried out by a computer is an abstraction of physical changes in the computer system”

    And that is the problem with software claims. It’s an abstraction of physical changes that you couldn’t possibly describe in any meaningful way (at least meaningful enough for a claim). And the last time I checked, “steps” being carried out indicate a method claim. And also the last time I checked, a method claim is to transform an object and set forth what the transformation involves. This is inherently impossible to do in any reasonable way with a modern +computer method claim. Wouldn’t you agree? You said so yourself, we require these “abstractions” to understand what is going on at the transistor/atomic level in the chip. But these abstractions don’t even show up in modern claims. Abstractions of abstractions show up.

    Ridiculous practice is ridiculous.

  6. Some of the earliest processing circuits required that the circuit by physically changed (redesigned) to perform a different processing step. As the technology evolved, configurable inputs (such as switches) were added to allow for a user to implement multiple functions in response to the inputs. Feedback mechanisms were not uncommon and allowed the circuit to, in some sense, configure itself. As the complexity of the configurable inputs and feedback grew, it became increasingly difficult to conceptualize all of the different configurations of the circuit. To compensate for our shortcomings as human beings, we developed abstractions to represent the function of complex configurations. Software is such an abstraction. Software code is merely a representation of vast numbers of different configuration settings, yet these settings represent the physical state of the computer system just as much as the original switches did.

    Simply put, written code is just an abstraction of physical properties of the computer system that helps us to understand. Should it be denied patentable protection because we simplify the complexity using abstractions?

    Whether or not you are able to understand that a computer system configured with software is no different than specifically tailored circuits in hardware (some basic physics lessons along with a class in computer hardware are suggested if you do not), it would seem that all policies reasons for allowing patents on specifically designed hardware circuits (and indeed most, if not all, technical areas) also apply to software.

    Suggesting that a “software” claim must recite an “output,” is, quite frankly, laughable. Any claim could add a clause for output something, making the requirement superflous. Requiring that it be explicitly recited serves no meaningful purpose. Moreover,recitation of steps carried out by a computer is an abstraction of physical changes in the computer system, making an output redundant. Whether or not the changes are detectable by your five senses is not a requirement of the patent law and indeed is a rather arbitrary distinction.

    The proper question is not whether the claim positively recites the utility, but whether the claimed subject matter has utility. There is no question that a claim to a novel chemical compound does not need to positively recite a use. Singling out such a standard for software suggests either a misunderstanding of the technology or that there is some other motive.

    As as been suggested by others, any problem lies not with the utility of “software” patents, but with the failure to reject applications that are obvious. This is a problem, however, in every technical area.

  7. “Red Hat: Software patents put a huge kink in the open source software movement.”

    Being a software engineer, I understand the importance of the open source movement. However, I also respect an engineer’s right to her intellectual property.

    Software, similar to an electronic device, a machine, or a pharmaceutical, may be new and useful. Moreover, software does not exist unless encoded on a physical medium.
    When a patentee claims a software product or business method implemented by software, the courts and the USPTO should interpret the claims to include encoding of the software on a physical medium.

    For more information about my technical background visit link to windycityip.com

  8. Mark Nowotarski | Apr 30, 2008 at 06:56 AM wrote:

    I read the Financial Services brief. The motivation for their brief, although not stated, is pretty clear. It is just a second front on their attack on the DataTreasury patents, which would certainly be ruled invalid under 101 if the positions in the brief were taken.
    —–
    Mark, I am no an expert on the DataTreasury patents, however I can say with certainty that none of the briefs advocating a narrowing of 101 and/or elimination of business methods have any effect for the simple fact the CAFC cant over rule the Supreme Courts precedent in Diehr that the claims must be evaluated as a whole.

    As long as this case is good law all business method inventors have to do is argue their claims based on Diehr.

    Also business method is just a word and to the best of my knowledge has never been defined by any court. So if the CAFC should rule that business methods are no loner patentable per se, then all inventors have to do is change or amend their applications to use the term industrial processes instead of business methods. The Supreme Court has upheld industrial process as statutory subject matter for the last 150 years.

    CASE CLOSED!

  9. e#6k | Apr 29, 2008 at 09:17 PM wrote:

    Pro- Are you out of your mind?

    ————-

    No. I am quite in command of my faculties and the case law for 101 statutory subject matter.

    e#6k wrote:

    The courts making their exceptions are not there to
    protect people from what came before, that’s what
    102 and 103 are there for. They’re
    there to protect people from gaining a patent on
    things that are directed towards the things excepted.
    ———

    e#6k I specifically addressed the fact that a person may not receive a patent
    that preempts one of the judicial exceptions of laws of nature, natural
    phenomenon and abstract ideas, or preempts all application of one of these
    judicial exceptions. Traversely, if a person discovers an application for a law of
    nature, natural phenomenon, or abstract idea, the Supreme Court has ruled such
    discoveries/inventions are patent eligible subject matter. ( See Diehr ) Of
    course the application may fail at 112, 102, or 103, but none the less is
    required by law to pass the bar at 101 and be examined.

    Understand now?

    I am not sure what you are referring to when you write “gaining a patent on
    things that are directed towards the things excepted.”

    None the less, please cite the legislative history, Federal Statue, or
    Supreme Court case law that dictates, a person may not receive a patent on an
    invention/thing “directed” towards a judicial exception?

    ::silence::

  10. I read the Financial Services brief. The motivation for their brief, although not stated, is pretty clear. It is just a second front on their attack on the DataTreasury patents, which would certainly be ruled invalid under 101 if the positions in the brief were taken. Their brief didn’t simply attack “mental process” patents, their brief went after all computer-implemented inventions. It looks like the group of financial services companies signing onto this brief want to take to the concept of technology back to the 1950s.

    Needless to say, I like the Regulatory Data Corp (Prof Duffy) brief.

  11. SS by itself I don’t believe needs overturning, it needs restricting to what is a proper interpretation of it rather than the hog wild approach so many apps/reps have foisted upon the office trying to game the system. But, if they feel like it’s just too far gone then sure, overturn it, it’s a bit open to over-interpretation and might still be even after clean up measures are taken.

  12. The best brief for upholding State Street was the brief from Regulatory Data Corp. The best brief for over-turning State Street was the brief from the Financial Services Industry.

    It makes sense that these two amici were invited to oral argument.

    Particularly since Bilski’s brief weakly defended State Street and the USPTO did not argue that State Street should be over-turned.

  13. I’m going to have to agree with Big Guy for the first time ever.

    Pro- Are you out of your mind? The courts making their exceptions are not there to protect people from what came before, that’s what 102 and 103 are there for. They’re there to protect people from gaining a patent on things that are directed towards the things excepted.

    Case closed, dismissed, and we’re ready for the next case. Stick to being a pro se let me know how it goes when all of your stuff is properly rejected through several appeals.

  14. A method of concluding a blog posting in the most obnoxious manner possible, the method comprising closing the blog posting with with a phrase selected from the following:
    ‘CASE CLOSED!';
    ‘Class Dismissed!'; and
    ‘Next?”

    Patentable subject matter?

  15. Posted by: Sue D. Nimmer | Apr 10, 2008 at 09:23 AM wrote:

    I read the human cognition brief and I don’t see the problems
    mentioned. Human
    cognition doctrine seems different than mental
    steps doctrine (narrower). There is no
    overlap with computer programs because it’s
    limited to “human” cognition. Second,
    infringement is determined as a matter of fact.
    And, juries are great at deciding such
    matters. For example, mens rea. Actually, this
    point is doubly-important, because how
    can examiners find prior art of thinking steps?

    The answer is, they are do not. What they are supposed to do according to the
    law is to evaluate and find prior art for the claims as a whole, not for individual
    elements or steps. That’s the problem we have now, is Patent Examiners totally
    disregarding the law and forcing inventors to go to court to get patents for
    inventions the laws were designed to protect.
    Posted by: Sue D. Nimmer | Apr 10, 2008 at 09:23 AM also wrote:

    However, I’m not sure this human cognition doctrine is best placed in 101. It
    seems
    more of a 103 approach. Also, there was something
    in the brief about distinguishing
    between different types of thinking steps
    (cognitive versus motor skills control or
    something), which I would imagine would cause
    line drawing issues.

    —–
    Actually it does not matter how the brief defines cognitive vs. mental steps
    because it’s all been made irrelevant by the new scientific breakthrough in
    Brain Plasticity. Whenever we apply knowledge, as every person making and using
    a process invention must do, we change the physical shape, anatomy of the
    brain. This scientific fact can be proven with MRI and Positron Emission Tomograpy
    scanners ( PET).

    In fact, if need be, every process claim that requires an application of the
    process whether it be curing rubber, or making a sandwich, can have the brain
    plasticity verified with an MRI or PET scan and presented in court as objective
    scientific evidence.

    So all applicants of process patents need to do is argue that the so called
    mental/human cognition/behavior step of the claim is statutory because it
    causes a definite physical transformation of an anatomical part of the body.

    Next?

  16. Posted by: Law Student | Apr 09, 2008 at 06:01 PM wrote: I agree with e#6k on this one. Prof. Collins’
    brief is right on. I hope the Fed. Cir. gives
    it the time of day.

    You didn’t mention on your blog how Collins’
    human cognition approach stems from
    the printed matter doctrine. I think that’s
    significant, because PMD has proven
    administrability and limited scope (i.e., it
    won’t shake things up too much).

    Collins presents some interesting legal gymnastics in an attempt to use the
    PMD as a spring board to empower the CAFC to run an end game around the SCOTUS
    precedent to examine the claims as a whole and in effect reinstate Flooks Point
    of Novelty test, at least for mental steps.

    However what Collins has failed to address is that printed matter is not
    patentable in and of itself because it generally does not effect the operation of
    the invention. Whereas mental processes and human cognition are essential to
    the manufacture , operation and use of every invention. To deny inventions
    because they involve a mental step would to be eliminate an entire class of
    inventions ( process) which congress and the supreme court have said the patent
    laws were designed to protect.

    Class Dismissed!

  17. Posted by: e#6k | Apr 09, 2008 at 05:52 PM wrote:

    A claim like:

    A method of creating a sandwich comprising:
    Providing bread
    Providing cheeses
    Providing meats
    Determining which meats are which
    Determining which cheeses are which
    Comparing cheeses to other cheeses, and comparing meats to other meats
    and choosing a cheese and a meat to stack on the bread and cover up with the bread (fold it over to make a fold over sandwich)

    The parts about determining, comparing etc. etc. just shouldn’t recieve attention. People do that stuff naturally, they’re natural concepts. You can replace meats breads and cheeses with signals or data etc. and the outcome is still the same. Yet I see these types of claims every so often.

    What you fail to understand e#6k is that the determining and comparing steps
    in the claim do not preempt the mental process or natural phenomena of making
    determinations and comparisons. Nor do the claims preempt the application of
    making determinations or comparisons to all fields of technology and
    invention. The claims are only limited to the process of making a sandwhich. Therfore
    the phenomena of determining and comparing are still within the store house of
    human knowledge.

    The judicial exception of laws of nature, natural phenomena and abstract
    ideas are designed to prevent society from being excluded in using what has
    already existed. Since no such exclusion is apparent in the above clams you MUST
    allow the claims to be examined as statutory subject matter under 101. To do
    other wise would be a flagrant disregard of the law.

    CASE CLOSED!

  18. “there is a huge difference between software and abstract mathematics. Software is goal-oriented applied mathematics.”

    That’s a huge difference?

  19. Re the Lemley brief, I agree with the point that subject matter rejections can usually be justified on other grounds, but I’m not persuaded by the brief’s explanation of Gottschalk. But it has always seemed to me that the algorithm in Gottschalk was obvious. BCD stores 10s and 1s digits separately in binary form. The least significant (1s) digit is already in binary, so it doesn’t have to be altered (so claim 13(5) says don’t process the least significant digit). The 10s digit represents 10*N where N is represented in binary, e..g N = 6 = [0110]. To convert to binary you have to take the tens digit and multiply it by 10, then add the result to the 1s digit. To multiply by ten you take each binary digit individually and multiply that by ten e.g. 10*[0110] = 10*[0010]+10*[0100]. Looking at claim 13, each repetition specified in (5) treats a separate binary digit. The machine / assembler language I am familiar with from the late 70s /early 80s didn’t have a ‘multiply’ command. (I don’t know if they do now). Instead you have to use a ‘shift’ command. Shifting one digit left corresponds to multiplying by 2; shift by 2 digits multiplies by 4; shift by 3 digits multiplies by 8. So to multiply by 10, you shift by 1 and independently shift by 3 and add the two results together. This is specified expressly as shifts in claim 5, and in claim 13(3) the same process is described as adding a binary 1 at the (i + 1)th and (i + 3)th positions. So the claim as a whole says “take each binary digit in the 10s column, multiply by 10 and add to the 1s column.” Given that machine language constrains you to the shift function instead of direct multiplication, this seems like a very obvious algorithm. Of course hindsight is 20/20, but I can’t really think of a non-trivial alternative algorithm using only assembler commands. (Of course there are trivial alternatives, like starting from most rather than least significant binary digit position.) Does anyone else know of an alternative algorithm for BCD to binary conversion?

  20. Hutz, I’m aware they’re nothing new, I deal with some of them currently. I staun chly believe that there should be some physical transformation in order to qualify, and that the physical transformation should be explicitly spelled out in the claim where appropriate. The “I cured cancer using a heart medicine” may be a rare type of exception where functional language might need to be called for to distinguish the “inventive step?” (lol what’s that in america?). Usually however that sort of language is a ba d thing to have clogging up our system.

    Step back- I believe the SC merely recognized what many simpletons cannot seem to, the patent laws, in all their various shapes and forms over the years, were meant to apply to physical manifestations of technology and not to ever preempt any types of thought etc.

  21. Big Hairy,
    Good point.
    At the end of the day this debate is a socio-economic one about who is “deserving” of compensation in the patent sense and who is not. The Supreme Court, in its beyond infinity wisdom, has decided that mathematicians are not deserving of compensation for having come up with some new mathematical process. Accordingly they are not among the “Whosoever” of 35USC 101.

  22. “there is a huge difference between software and abstract mathematics. Software is goal-oriented applied mathematics”.

    The difference is not all that great. Granted that theorems are not equivalent to software but all software is equivalent to a logical (ie mathematical) construct. The idea in any software that might warrant patenting is that part of it which can be expressed as an abstract logical construction – the particular implementation (in Fortran, C, C++, C#, Java or what have you) of that abstract construct (usually expressed in mathematical notation or meta code) is protected by copyright. Most mathematics is goal oriented.

    Should the Black-Scholes formula have been patentable? Should the Simplex method have been patentable? Should conjugate gradient or finite element methods have been patentable?

  23. Brad- didn’t you hear? Physics theories are just that, theories. In addition, I heard even Japan gave up AI. I disagree with your statement that our resources are better spent on 102 103 112 IF applicant’s persist in trying to control the thoughts of their fellows. Of course, that’s not going to change, so we’re going to have to keep on the 101.

    Maybe the patent act needs a little redrafting to clarify just what CONGRESS would like to stimulate, rather than what everyone and their grandmother thinks should be “stimubackwardslolated”. Problem is, our leaders are vote whipped wusses.

  24. How much easier life would be if 101 was limited in application to why it appears in the first place…that an invention must be “useful”. 102, 103, 112, and other relevant provisions of Title 35 are more than up to the task of taking it from there and determining if the other necessary requirements are met.

  25. E#6K

    Sorry to be the bearer of bad news, but new use patents have been around forever. They are method patents. If I discover a century-old leather tanning chemical can be used to treat cancer, I should be able to obtain a patent on the use of that chemical for the treatment of cancer. Even if a valid patent on the underlying chemical is still in force, if I discover a use that has not been suggested by any previous literature or data, I should be able to obtain a patent on that use.

  26. I am sorry, but there is a huge difference between software and abstract mathematics. Software is goal-oriented applied mathematics. I can understand the requirement to have a “use of the result” type limitation.

    I am with the camp that says 102, 103, and 112 should be used to reject inventions, and not 101. Rejecting applications simply because they claim a particular category of subject matter is too arbitrary to be just, and will run into problems where people will simply change their claim language to avoid nonpatentable subject matter rejections.

  27. I am left wondering if Lemley knows anything about math or software. The distinction he seeks to draw between software and abstract mathematics is precisely the type of distinction that patent attorneys are already “gaming”.

    His basic point is that 101 should not be used to weed out the garbage at the beginning for fear that there might be an ounce of gold hidden in the tons of dross.

    In saying that there are other remedies for bad patents Lemley basically assumes that litigation is free and that there is little or no cost through externalities from bad patents. After assuming no cost from bad patents he compounds the error by simply assuming benefits for innovation from patents for business methods.

  28. “The simple fact is that NO ONE ever ACTUALLY thought these involved mental steps.”

    No one other than the inventors, the patent attorney(s), the Examiner and now everyone who reads the patent, the case or PatentlyO!

    The law has long struggled with the challenge of mental processes in many areas other than patents, like whether crimes are premeditated, whether there is a conspiracy, or whether there is willfulness or inequitable conduct. And just like in all of those areas, we are nowhereclose to understanding and defining black and white lines about mental processes that can be easily defined and defended. Even if we were somehow able to measure mental processes, just like in Minority Report, there will be no certainty in those measures.

    Mix in the complexity of trying to define something that, by virtue of Sections 102 and 103, must be new and and not obvious and the challenges of defining or measuring mental processes become even more daunting. Add in the further layers of advances in information theory and quantum physics, and the supposed bedrock of testing for transformation of something “physical” melts away. Throw in advances in neuroscience and artifical intelligence, and the future futility of trying to define “mental processes” in terms of “man vs. machine” may make all of these arguments nothing more than good humor, possibly even within the lifetime of patents granted today.

    In my opinion, the Supreme Court got it right in Charkbarty (anything under the sun) and wrong in Flook (getting lost down the rabbit hole of algorithms).

    Humans (and even patent attorneys) are and will be creative and, regardless of the decision, clients will be willing to pay for creative attempts to craft around any decision by the CAFC (or even Supreme Court) in Bilski.

    From a pragmatic view, the resources of the PTO, attorneys and clients are probably better spent on 102, 103 and 112 issues, particularly given the new KSR yardstick of “ordinary creativity” that can be used to attack many claims that might otherwise be directed to mental steps, business methods, algorithms or printed matter. From a public policy view, this issue may be more properly the purview of the legistlature than the courts; but if the current experience with the Patent Reform Act of 2005-06-07-08 is any guidance, that option may be little better than tilting at the windmills.

    Brad Pedersen

  29. someone above hit the nail on the head – shame these folks did not provide these briefs “in lieu” of testimony on the patent reform act – not really sure the judiciary can determine “business questions”

    … that being said … what is a “business method”? what is a “patent? what is “software”? more particularly, what is a “computer”? there are ONLY “magic words” when search and examination is directed at words not concepts …

    fwiw, greg aharonian has the most consistent viewpoint (another way to paraphrase – if you *knew* the next “good” “patent” it is unlikely to be patentable (disclosure makes something so objective it begs obviousness – but then we can create “magic words”) or good (you know so why share?) >>> what is more tangible than manna? MONEY

    aclu the most bizarre … would prefer they argued the balance between “privacy” and “piracy” (inlcuding anything to do with mca v sony and recent efforts to undermine codification of fair use rights for consumers) …

    messaging matters … keep calling those senators …. got nothing to do with appointees … but sen lott did get out at the right time just not the right cause (my opinion)

  30. “The overall filing rate would probably go down 25% if we start interpreting 101 properly”

    And revenue would drop 25% too. I’d be sad to see you work for any less, 6k.

  31. Upon perusal Lemley is even worse than I had thought from the outset. See page 17, “Take, for example,” making an old product new again through a new process of its use, thus denying the original inventor of the product exclusive rights. <<< is what it should have said. Take further for example the discovery of a function of a certain DNA sequence, under Flook this fails, yet all of a sudden it’s an amazing invention since someone did … nothing but look and find it (i.e. no inventing, just discovery of something old). Truely, great arguments Lemley makes here. “The same is true of software:” the value of a new program lies in the fact that information is processed differently or that the program generates new information as results since the “inventor” decided to use an old product in a “new” process. Again, wonderful points you’re making there Lem, except for the fact that we are not discussing what the value of a new program is, we are discussing if it is patentable subject matter, and so far as I know, “utility” would be the only requirement that “value” even comes close to, but even that isn’t the same thing. So in essence, Lem’s argument is, software is valuable, and yet it fails under Flook. Hmmm, good point Lem. Irrelevant though as far as I can see in interpreting statutes properly. “If patentably subject matter is restriced to physical embodiments then some things that are not physical embodiments and are as-yet-unforseeen fields of “not inventing physical embodiments” [will] not qualify for patent protection”. Again, I wish the rest of us had the intellectual prowess to have figured that one out Lem and I’m sure glad you’re here to point that out to us. If we (they?) wanted patent laws to cover ideas, and all functions etc. Lem then Congress would EXTEND 101 to include them, they do not have to cut back on what they already have put in to express their distaste of them. I’d comment on the rest, but it’s all together too funny. I will say this though, it’s Crp like this that’s helping to overwhelm the office. The overall filing rate would probably go down 25% if we start interpreting 101 properly, and not only properly in relation to Lem’s situations, I mean distinguishing methods in terms of steps, products etc in terms of structure.

    I will say this though Lem, D, etc. sure, making functional language, software etc patentable might be a good thing. Might. But it’s improper under current law.

  32. Why not do this the old fashioned, and actually decide if things are patentable subject mattter or not? They got it right in Chakrabarty, they got it wrong in State Street and seem to be correcting the error in their ways. No need for the sweeping academic brief of Lemley and friends that has its head in the clouds, and it nose in the air.

    BTW – also no need for patent reform, the system works!

  33. big hair rat writes “Surely a process has to be a complete recipe (algorithm?) to take you from point A to point B with a full description of all of the steps in the middle or at least an invocation of standard existing sub-processes for those steps.”

    If the claim is obvious but for the algorithm, then that is surely the case. If nothing else, one would expect in such circumstances that such a detailed recitation of the algorithm would be needed to forestall an enablement rejection. If the application argues that no detailed recitation is needed, then the claim should be deemed obvious.

  34. CaveBoy flings his usual thoughtless poo and writes: “trademark law protects the ability of a mark to form a proper perception held in someone’s head and riding on the good will”

    Only you, CaveBoy, would “forget” to expressly mention “consumer confusion” in your definition. Remember those little people, CaveBoy? The people that don’t own trademarks but need to buy things occasionally to survive? Preventing consumer confusion is the purpose of trademarks and it’s the only one worth mentioning. “Riding on good will”? Whatever. As long as there is no confusion, it’s a free country, baby. I’ll ride your good will to Abilene then shoot it in the street just to watch it die.

    The obvious question is: what if I don’t want that allegedly “proper” perception in my head? The answer is: I should be permitted to alter that perception by, e.g., creating marks that destroy the perception. As long as consumers are not confused, this shouldn’t be a problem. In fact, it isn’t a problem when a mark “at issue” can be characterized as parody. It shouldn’t stop there, however. My point (again) is that “tarnishment” is a crap doctrine.

  35. PDS, you’re hitting this one out of the park.

    I think some common sense should also be applied.

    Regarding ATT, the claims says “for use in a telecommuniations system” and “automatically”. The simple fact is that NO ONE ever ACTUALLY thought these involved mental steps. In anywhere other than that court, people would laugh at you if you said a method for telecom switching would be performed mentally.

    In contrast, Ex6k’s ridiculous claim is CLEARLY something that a human mind could or would do in a beneficial manner.

    There should be some demarcation based on common sense.

    To this end, Lemley’s brief is spot on. You could never prove that someone was *thinking* the way you claimed. However, it’s fairly easy to look at code or get testimony as to how someone has programmed their system.

    I think it’s all a bit unnecessary.

    I don’t care that signals aren’t patentable (Bilski). It’s easy enough to recite a computer-readable STORAGE media, or claim “a transceiver”.

  36. ” Many of the hairs trying to be split in some of the briefs can easily be cured by clever claim drafting. As such, this leads to the determination of patentable subject matter resting, not on the invention itself, but instead, on whether or not the right magic words were included in the claims. ”

    Amen to that, and may those hairs continue to be split – that’s why our clients pay us!

    But, with more sincerity, “the right magic words” are would not just be *magic*. They would definitively limit claim scope to 101-acceptable subject matter (whatever that turns out to be), giving up some of the claim in order to attain allowance over the statutory grounds for rejection. What’s wrong with that? Seems just like the sort of words added to overcome 112, 102, 103 …

  37. I don’t have time to read the links but it seems to me that one of the problems with the Bilski claim is that it is not “new” and does not define a “process” since it leaves out completely any method of doing key steps. Surely a process has to be a complete recipe (algorithm?) to take you from point A to point B with a full description of all of the steps in the middle or at least an invocation of standard existing sub-processes for those steps.

  38. David French writes:

    While the comments are short on this subject, I would like to make the following observation. Patently-O is a very successful and popular blog. This is due to:

    1. Dennis’s insightful selection of topics and observation;

    2. The extremely valuable hyperlinks that Dennis provides as particularly demonstrated by this posting, and

    3. The extensive commentary inspired by 1. and 2. above.

    Without 1. and 2., 3 could not exist.

    Thank you Dennis.

  39. One of the things I noticed in perusing the briefs was just how badly many of them cited AT&T Corp. (the case I spent a lot of time reading yesterday).

    AT&T Corp. doesn’t discuss *physical* transformation. Instead, AT&T Corp. refers to transformation. Also, the transformation can involve data. Many briefs describe AT&T Corp. a dealing with a process that *required* a machine. This was certainly not the case.

    Claim 1 of U.S. Patent No. 5,333,184 (the patent at issue in AT&T Corp. is reproduced below:
    1. A method for use in a telecommunications system in which interexchange calls initiated by each subscriber are automatically routed over the facilities of a particular one of a plurality of interexchange carriers associated with that subscriber, said method comprising the steps of:
    generating a message record for an interexchange call between an originating subscriber and a terminating subscriber, and
    including, in said message record, a primary interexchange carrier (PIC) indicator having a value which is a function of whether or not the interexchange carrier associated with said terminating subscriber is a predetermined one of said interexchange carriers.

    The only findings of what the process required was made by the district court and not the Federal Circuit. The Federal Circuit stated the following:

    “The district court concluded that the method claims of the ‘184 patent implicitly recite a mathematical algorithm. See AT&T, 1998 WL 175878, at *6. The court was of the view that the only physical step in the claims involves data-gathering for the algorithm. See id. Though the court recognized that the claims require the use of switches and computers, it nevertheless concluded that use of such facilities to perform a non-substantive change in the data’s format could not serve to convert non-patentable subject matter into patentable subject matter. See id. at *6-7. Thus the trial court, on summary judgment, held all of the method claims at issue invalid for failure to qualify as statutory subject matter. See id. at *7″

    Both the steps of “generating a message record …” and “including … (PIC) indicator” could be considered mental steps. Although, in reality, these steps would always be performed on a computer, these steps could be performed mentally.

    Although AT&T Corp. is chock full of interesting comments, two of my favorites are:

    the mere fact that a claimed invention involves inputting numbers, calculating numbers, outputting numbers, and storing numbers, in and of itself, would not render it nonstatutory subject matter, unless, of course, its operation does not produce a ‘useful, concrete and tangible result.’

    and,

    Since the claims at issue in this case are directed to a process in the first instance, a structural inquiry is unnecessary.

  40. If the court addresses human cognition, the court should address feline cognition:

    “A method of inducing aerobic exercise in an unrestrained cat comprising the steps of: (a) directing an intense coherent beam of invisible light produced by a hand-held laser apparatus to produce a bright highly-focused pattern of light at the intersection of the beam and an opaque surface, ***said pattern being of visual interest to a cat*** …”

    (from Claim 1 of U.S. Patent No. 5,443,036)

  41. I read the human cognition brief and I don’t see the problems mentioned. Human cognition doctrine seems different than mental steps doctrine (narrower). There is no overlap with computer programs because it’s limited to “human” cognition. Second, infringement is determined as a matter of fact. And, juries are great at deciding such matters. For example, mens rea. Actually, this point is doubly-important, because how can examiners find prior art of thinking steps? These steps should not even be considered.
    Third, even examiners can handle disregarding a “determining step” their analysis (sorry examiners).

    However, I’m not sure this human cognition doctrine is best placed in 101. It seems more of a 103 approach. Also, there was something in the brief about distinguishing between different types of thinking steps (cognitive versus motor skills control or something), which I would imagine would cause line drawing issues.

  42. Lemley has a perfect record – he is not a patent attorney and it shows! He is wrong about EVERYTHING either in whole in part, and not just this brief.

  43. I agree with pds in that I see E#6K’s representative claim as an example of extremely poor drafting (not to say that this doesn’t occur) that can be tinkered with to provide a better claim directed to a process that cannot be purely mental.

    Good luck with the the PTO listening to the CAFC on 101 decisions. I have found that even after quoting the appropriate part of recent CAFC decisions directly contradicting the examiner’s point (on various 103, 112 issues), most examiners ignore the citation entirely in his/her response.

    It is also unfortunate that the elevation of form over substance (in the form of “magic words”) has become so prevalent in the business methods/software area.

  44. I was going to say what PDS already said.

    This is much ado about nothing.

    You can’t prove infringement by the human mind of a thinking step.
    In the practical world, the “determining” step will be one carried out by a machine.

    The real intent behind some of these amicci briefs is simply to block innovators in the software arts (i.e. samll inventors) from being rewarded for their inventions even though innovators in other “useful” arts are.

  45. I searched “determining” in “claims” in pto database and got 319,036 hits — it is a very common claim term

  46. I spent all day combing through AT&T Corp., so this thread is very topical to me.

    Anyway, in reading the briefs, one thing I took out of it was the many different (and inconsistent) meanings for the concept of “transformation.” Some meanings involved having the transformation be physical, some did not. Also, certain writers said that a physical transformation encompassed the transformation of data, while others did not.

    Also, many of the briefs that wanted a narrower view of statutory subject matter still want software to be included as statutory subject matter. The briefs by IBM, Microsoft/Dell, SAP are examples. These briefs state that software (when claimed properly?) are tied to machines, and are thus patentable.

    Personally, I like the brief that Dennis signed. Many of the hairs trying to be split in some of the briefs can easily be cured by clever claim drafting. As such, this leads to the determination of patentable subject matter resting, not on the invention itself, but instead, on whether or not the right magic words were included in the claims.

    Personally, I have no use for the policy arguments being made in some of the briefs. Neither the Federal Circuit nor the Supreme Court should be involved in making public policy. That job should be left to Congress, which leads to the biggest hurdle in narrowing statutory subject matter. Specifically, the old “anything under the sun that is made by man” statement from Chakrabarty with regard to Congress’ intent. This is a fairly high hurdle to overcome.

    On a philosophical note, with regard to the question as whether certain steps which could be performed by the human mind should be patentable, I have always wondered, how do you establish infringment? Unless somebody actually states “I was thinking x, y, and z,” you cannot establish infringement. Personally, I would rather this issue decided upon lack of proof of infringement than under 35 USC 101. The problem I have with the “human thinking” step is that in many instances, the same step (as claimed) could also be performed by a machine. Knocking out a step, just because it could be performed by a human mind, also excludes steps that involve a machine, which should be statutory under 35 USC 101.

    In going through the briefs I found that a few of the writers struggled with how to make this distinction (i.e., exclude step performed in human mind versus step performed by machine). My hope is that the Federal Circuit takes this opportunity to draw the approrpriate line and make it clear how one can draft claims that fall on the (right) side of the line. Getting back to AT&T Corp., I spent a good portion going through this case because the claim I’m dealing is spot on identical to the claim at issue in AT&T in terms of statutory subject matter yet it still has been rejected.

    My sincerest hope is that with a clear decision, the Federal Circuit will drastically reduce the number of 101 rejections coming out of the USPTO. I successfully traverse 95%+ of the 101 rejections I receive, and the ones I don’t are because it is the only thing keeping me away from a notice of allowance. Responding to 101 rejections is one of the biggest wastes of time I have. Considering that the examiners have to both write the rejections (and respond to my arguments), they are also a big waste of time at the USPTO.

  47. I reviewed one of the patents cited in the EFF brief – “Character assessment method” … while it may or may not be patentable subject matter, it definitely contributed to my knowledge of a useful art …

  48. “Tradmark is the worst offender…” Offender of what exactly? Trademark law protects the ability of a mark to form a proper perception held in someone’s head regarding origin of goods and prevent a third party from free riding on the good will established by that perception. Assault allows a criminal cause to be raised for an idea formed in someone’s head regarding a perceived ability of a perpetrator to carry out a threat of harm.

    You’re not too good at this law stuff are you Mooney.

  49. I had planned to make a bunch of ill-informed comments responding to discussion of the summary of the ACLU brief (which is the only one that currently comes up in response to clicking the above links-to-briefs).

    Then I actually read the brief.

    It presents a cogent argument for cabining the scope of patentable subject matter.

  50. “The important issue is that copyright law is limited by the 1st amendment, and, consequently, so should patent law.”

    The first amendment does not bar copyright protection to art, music, and abstract ideas due to their subject matter. Why just patents?

  51. ex6k writes :

    “A method of creating a sandwich comprising:
    Providing bread
    Providing cheeses
    Providing meats
    Determining which meats are which
    Determining which cheeses are which
    Comparing cheeses to other cheeses, and comparing meats to other meats
    and choosing a cheese and a meat to stack on the bread and cover up with the bread (fold it over to make a fold over sandwich)
    The parts about determining, comparing etc. etc. just shouldn’t recieve attention.”

    If the “determining” parts are the new contribution of this method, then they should certainly receive attention because the applicant is attempting to patent a way of THINKING about a problem. You can’t put an idea in someone’s head and then sue them for thinking about it or making a decision based on facts known to them. People have a right to THINK about ANYTHING.

    By the way: trademark law is arguably the worst offender in this regard, with the so-called “tarnishment” doctrine alleged to protect the sanctity of an idea about a corporation or its products as the idea exists in a third party’s head.

  52. “This is my problem with the ACLU brief: they argue that things like art, music, and abstract ideas can’t be patented because such protection impedes 1st amendment rights”

    I didn’t get that from the ACLU brief. My understanding is that the ACLU merely observed the fact that art, music, and abstract ideas aren’t patentable. The important issue is that copyright law is limited by the 1st amendment, and, consequently, so should patent law.

  53. “I thought the ACLU brief was an interesting take.”

    This is my problem with the ACLU brief: they argue that things like art, music, and abstract ideas can’t be patented because such protection impedes 1st amendment rights, which are paramount. Yet art, music and literature are elligible for copyright, which in many ways is a stronger protection than patent rights (at least in length). How can patents interfere with 1st amendment protections if copyrights do not?

  54. “‘Bad patents’ should be dealt with using the true tools of the Patent Act: Sections 102, 103, and 112.”

    Right on. 103s got some serious teeth against human cognition steps.

  55. FYI, PMD is a hybrid 101/103 type analysis, where claimed elements with non-functional printed matter (outside of 101) are given zero “patentable weight” with regard to 103. (Exam ‘s in two weeks!)

  56. I agree with e#6k on this one. Prof. Collins’ brief is right on. I hope the Fed. Cir. gives it the time of day.

    You didn’t mention on your blog how Collins’ human cognition approach stems from the printed matter doctrine. I think that’s significant, because PMD has proven administrability and limited scope (i.e., it won’t shake things up too much).

    (my two cents)

  57. I disagree with Lemley, and agree with Sarnoff, IBM, Collins especially, and off hand Microsoft as well. Lemley’s ideas, which I presume you adhere to D, create way too much confusion. Patent laws were never meant to preempt thinking, no matter if the thinking was done during a process. Allowing there to be law that allows patents to capture thought processes just because they’re tied to a tangible process (whether the tangible process is novel or not doesn’t matter).

    A claim like:

    A method of creating a sandwich comprising:
    Providing bread
    Providing cheeses
    Providing meats
    Determining which meats are which
    Determining which cheeses are which
    Comparing cheeses to other cheeses, and comparing meats to other meats
    and choosing a cheese and a meat to stack on the bread and cover up with the bread (fold it over to make a fold over sandwich)

    The parts about determining, comparing etc. etc. just shouldn’t recieve attention. People do that stuff naturally, they’re natural concepts. You can replace meats breads and cheeses with signals or data etc. and the outcome is still the same. Yet I see these types of claims every so often.

  58. I agree with this:

    –Prof Lemley: We cannot predict the next area of innovation, and arbitrary limits on patent scope reduces incentives in those potential areas. “Bad patents” should be dealt with using the true tools of the Patent Act: Sections 102, 103, and 112. “Mental methods” should be allowed if they fall within the other requirements of patentability.–

Comments are closed.