Microsoft v. AT&T, now pending in the Supreme Court, involves important but recondite issues of the extraterritorial application of patent law. Most observers expect Microsoft to win, especially because it has the support of the U.S. DOJ.

Two of the amicus briefs filed raise more fundamental issues, though -- they argue that software should be unpatentable, root and branch. One of these comes from the Software Freedom Law Center, a central pillar of the Free Culture Movement, and is thus no surprise. The other is from Eli Lilly of big pharma, which is a surprise. (This one is available only from Lexis.) One would think that pharma, whose own patent-based business model is under sharp assault, would be careful about spitting in the soup -- where will it end? With Microsoft supporting the anti-pharma campaign?

Patent expert Greg Aharonian has serious objections to both the Lilly and SFLC briefs. His unexpurgated comments to his newsletter list follow:

COMMENT ON SFLC BRIEF

In a case before the Supreme Court, Microsoft v. ATT, the main issue is if supplying a (software) component from the United States to an overseas operations triggers infringement depending on how that component is used overseas. I won't be discussing this issue (partly because Microsoft can workaround any adverse decision, given the fluidity of software - indeed, why is any company in this era of high bandwidth, multiple error-corrected coding, physically shipping "golden masters" anywhere?).

Rather, a secondary issue raised (to attract attention) reduces to "Is software patentable?". The issue probably won't - and shouldn't - be decided by the Court, though maybe commented on by one of the more scientifically ignorant Justices.

Anarchist Eben Moglen, along with big pharma as represented by Eli Lilly, last month filed amicus briefs arguing that software (Moglen) and business methods (Lilly) should not be patentable. More on Moglen's anarchist beliefs at: firstmonday.org/issues/issue4_8/moglen, where he talks about his wish for the death of software copyright. In his Microsoft brief, he argues for the death of software patents. Since the logic of both arguments applies equally well for hardware, it is easy to conclude that Moglen doesn't want any IP protection in the digital world, software or hardware (an anarchist's dream he shares with Richard Stallman and Brian Kahin). Lilly as well, not that Lilly lawyers think that deep, so in revenge, how about eliminating IP for drugs? Something Lilly should think about before playing these legal word games.

What follows is a critique of the Moglen brief (Lilly to follow). Both briefs make use what I think are of outright lies, repeated use of many undefined terms, and selective use of caselaw and law review articles.

If so, the legal ethics and scientific competence of Moglen should be questioned.

====================

The brief was filed by Eben Moglen, Daniel Ravicher and Richard Fontana of the Software Freedom Law Center, which supports open source efforts.

Moglen's argument is that the Supreme Court was right in Benson/Flook/Diehr, while the CAFC got it wrong with its cases Alappat/[Lowry]/State Street.

The problem for Moglen is that Benson was a politically motivated decision based on false science, as is its continuation Flook - both decisions repeatedly criticized by lawyers who actually understand patent law and science. Worse for Moglen, he argues that Diehr continues this chain, without realizing that Diehr was a face-saving way for the Supremes to correct its mistakes in Benson.

Moglen, near the end of his brief, concludes:

"The Court's decision in Diehr upheld the holdings in Benson and
Flook, ..."

the holdings being that software, no matter what the industrial use, was just an unpatentable mathematical algorithm

"... and merely found that the claimed invention in that case
was not substantially directed to just software, but instead was
- in totality - directed towards an 'industrial process ....."

So the Benson/Diehr test reduces to checking to see if the software patent claims are for a process in "totality" or just a process "substantiality", whatever the hell "totality" and "substantial" mean. "Totality" is the loophole that the Supremes used to dispose of the bad science of Benson.

To rely on Benson/Diehr as a basis for an attack on software patents thus seems self-defeating, especially since many Benson/Flook/Diehr arguments are more strongly used against hardware patents. Why did the Supreme Court need this loophole for Benson? I quote from Chisum's 1986 article that reviews Benson, "The future of software protection: the patentability of algorithms", where he writes in the intro:

Second, the Benson decision, which held that mathematical
algorithms could not be patented, was poorly reasoned and
stemmed from an antipatent judicial bias that cannot be
reconciled with the basic elements of the patent system
established by Congress. Third, the awkward distinctions
and seemingly irreconcilable results of the case law since
Benson, including the Supreme Court's decisions in Parker
v. Flook and Diamond v. Diehr, are a product of the analytical
and normative weakness of Benson itself.

and to be more clear, one main section of Chisum's article is a detailed critique of Benson. Why?, he writes:

The reasoning in Benson is monstrously bad.

Such a charge should and can be backed up by paragraph-by-paragraph,
line-by-line examination of the opinion, authorities upon
which it relies, and the policy justifications that it recites

But Moglen addresses none of these problems with Benson, which he uses to attack software patents (and indirectly attack hardware patents), as we shall see below.
====
LIE #1

The first lie of Moglen's brief is a big lie of omission. Nowhere in his brief does there appear the word "hardware". It is unethical to talk about the patentability of software without simultaneously talking about the patentability of hardware, especially in light of hardware/software codesigns tools. And even using the word "hardware" is pointless unless you provide rigorous definitions of "hardware" and "software". Moglen doesn't. So when Moglen bases his software patent hatred on Benson:

"The holding of Benson is properly applicable to all software,
because a computer program, no matter what its function, is
nothing more or less than the representation of an algorithm."

as well, he is arguing hardware patent hatred:

"The holding of Benson is properly applicable to all hardware,
because a digital circuit, no matter what its function, is
nothing more or less than the representation of an [Boolean]
algorithm."

So all of the hardware companies giving money to Moglen, stop doing so
- it betrays your stockholders' interests (admittedly in a minor way).

====
LIE #2

With this lie of omission (not mentioning "hardware"), what better way to start a Supreme Court brief than with a lie of fact:

"Much of the world's most important and most significant software
is distributed under terms that give recipients freedom to copy,
modify and redistribute the software ("Free and Open Source
Software").

Use of "much" is a lie. Much of the world is NOT the Internet, and much of the world's software has nothing to do with the Internet. From the January 2007 edition of the Dr. Dobb's Journal: "Q: How many lines of COBOL source code currently are in use?" "A: ... the answer is 250 billion (BILLION) lines of COBOL source code.", much of this COBOL code being significant to its users. And Moglen knows nothing about the history of software. In the 1970s and 1980s, long before software patents became numerous enough to be socially relevant, and long before open source's delusions, science and professors were routinely completely large numbers of significant computer programs. But why bother sharing these truths with the Supreme Court, when a nice tight lie will do?

Indeed, if Moglen actually researched the issue (like asking me for my statistics), he would know that there is little overlap between software patents (which reflect current significant applications) and open source software. Most software patents are for embedded systems and specific industrial uses, not the ten-billionth open source packet sniffer or small Linux kernel. Indeed, one important piece of software for the original growth of the Internet was the proprietary SS7 switching software. And most software that governs the physical infrastructure of the United States (on which the Internet relies) - power industry, telecom industry, etc., is not open source.

LIE #3

Worse, open source lawyers like Moglen refuse to address the deep legal problems of open source copyrights, such as: what type of linking creates a derivative, if at all; what aspects of open source software illustrate, depict and embody the processes in the open source software to then be uncopyrightable under 17 USC 102b; what aspects of open source source codes survive Altai's abstraction and filtration? Until Moglen cleans up these copyright problems, he should shut up about patent problems.
Maybe some Justices reading Moglen's brief might think - "Moglen is right - let's get of software patents for software copyright", with such Justices probably not realizing Moglen wants to get rid of software copyright, leaving no protection.
----

The next two pages are a brief history of open source software and their Law Center in which none of these problems are discussed.

Moglen then turns to question I: "Software cannot be a component .. because software is not patentable subject matter under 101", because:

"... the Federal Circuit's holding in Eolas that software
is patentable subject matter conflicts with long-standing
precedents of this Court."

Precedents that Moglen doesn't cite until some paragraphs later:

"Confronted with the rise of new technologies, this Court has
addressed the issue of patentable subject matter several times.
[Benson (1972), Flook(1978), Chakrabarty(1980), Diehr(1981)]"

So this must be the anti-software patent basis for Moglen. Something in either Benson, Flook and/or Diehr (Chakrabarty doesn't talk about software).

"Since before the Civil War, this Court has consistently made
it clear that subject matter which would have the practical
effect of preempting laws of nature, abstract ideas or
mathematical algorithms is ineligble for patent protection.
O'Reilly v. Morse, Benson."

Whoopee. Software is not a law of nature, it is not abstract [hey Mogly, define "abstract", huh?], and as long as the software provides specific, useful results, "in totality" (Diehr), even the Supreme Court agrees that software algorithms are patentable.

Moglen basically argues Flook and Diehr.

"This Court stated in Flook that to be eligible for patent
protection, '[a] process itself, not merely the mathematical
algorithm, must be new and useful.'"

Well, with this logic, here is how to argue against hardware patents:

"This Court stated in Flookie that to be eligible for patent
protection, '[a] device itself, not merely the underlying
Boolean algorithms, must be new and useful."

After all, there is no physical difference between a digital processing circuit where the circuit is fixed in nature at the factory (hardware) and a digital processing circuit where the circuit is fixed in the field (software), especially for processors reconfigurable in realtime by the software, making any hardware/software boundaries irrelevant. You can't talk about patenting math algorithms in software without talking about patenting math algorithms in hardware.

"In alignment with Benson and Flook, this Court's decision in
Diehr held that structures or processes must, when considered
as a whole, perform functions intended to be covered by patent
law in order to be eligible for patent protection."

"Structures" allows all of Moglen's arguments to be applied to hardware, and I suspect that as an icing on the cake, Moglen wants to get rid of both software AND hardware patents. (Put it this way, if Moglen succeeds in getting software patents banned, I will file lawsuits to use the same legal reasoning to get hardware patents banned - just think of silicon generics companies turning out Intel and AMD chips). To use Moglen's silliness, software and hardware are different "forms" of the same "substance", and since he argues that software "form" is not patentable, so too then is hardware "form" not patentable.

"This Court requires that one look, not simply at the language
of the patent claim to see if it recites a structure of multiple
steps or components, but also at the practical effect of the
claim to see if it in fact covers - or otherwise would restrict
the public's access to - a principle, law of nature, abstract
idea, mathematical formula, mental process, algorithm, or
other abstract intellectual concept."

LIE #4

Meaningless in the context of software. Moglen repeatedly uses "abstract"
without defining it. This is important to do in the software world because computer scientists routinely use the phrase "abstract data type"
to describe data structures that are NOT abstract in the Lowry/State Street sense (concrete, tangible, useful), for example, in languages such as Standard ML.

To confuse the issue of which category he wants to apply to software, Moglen lists multiple categories. And it is mathematically ignorant to use both "mathematical formula" and "algorithm" in the same sentence without defining what you mean by the terms, since the terms are generally the same thing (depending on whether or not you include a "result" as part of the "algorithm"). The vast majority of mathematical formulas correspond to first order predicate logic statements, each of which is mappable into an algorithm in a programming language such as C
- heck there is even software to do this. And without accurately defining these terms, you can't assess whether the type of software being patented falls into one of these JUDICIALLY-CREATED categories.
----

Moglen then starts attacking the history of CAFC software patent caselaw as a betrayal of the Supreme Court. While interesting, the discussion doesn't talk about the underlying science.

"Many scholars have noted that the creation of the Federal Circuit
'did away as a practical matter with Supreme Court jurisdiction
in patent case'.

"Initially, the Federal Circuit used the opinions of legal
commentators to justify straying from Benson and Flook."

LIE #5

Moglen thus complains about the CAFC playing politics. But what Moglen doesn't mention is that three key paragraphs of Benson come from a 1960s presidential panel (legal commentators) dominated by IBM, and that Benson was decided as much politically as it what on science grounds. IBM didn't want software patented in the 1970s because it interfered with sales of IBM hardware, and what IBM wanted in the early 1970s was easily given by the Supreme Court. When IBM started making lots of money, it abandoned this position, to the extent of forcing the PTO to allow computer program copyrightable expression to be patentable as a useful article of manufacture (the contradiction Beauregard causes for 17 USC 101/102). And now that IBM is making money from services, it doesn't mind undermining software and business method patents and helping open source. I wonder how much money IBM is giving Moglen. Benson was much more of a political decision that Alappat.

And there is a good scientific reason (science as in the constitutional command to progress science) why the CAFC strayed. For example, Benson is riddled with scientific errors (as documented by patent law professsor Donald Chisum in his classic law review article that Moglen doesn't cite).
I doubt if Moglen has ever read Chisum's article, and to not do so and not address Chisum's criticisms in his brief is unethical. Moglen then continues

"Roughly two years later, the Federal Circuit said that this
Court's precedent on patentable subject matter was too unclear
to follow. In re Alappat, ..."

Not a surprising thing for the CAFC to say, since many lawyers (starting with Chisum) have documented the confusions in Benson.

"Contrary to the Federal Circuit's characterizations, however,
this Court's precedent on subject matter is plainly clear;
the analysis is one of substance, not form, ...

But what does Moglen mean by "substance" and "form"? Otherwise this phrase is meaningless.

"... and asks whether a patent claim is substantially directed
to a law of nature, natural phenomenon, abstract idea or
mathematical algorithm."

So in which category is Moglen placing software? And note the caselaw distortion - the Court's precedent was the phrase "law of nature, natural phenomenon, abstract idea" separate from the phrase "mathematical algorithm" in other caselaw.

"As just one example, it ignores the firm statement in Diehr
that '[a] mathematical formula [in hardware] does not suddenly
become patentable subject matter simply by having the applicant
acquiesce to limiting the reach of the patent for the formula
to a particular technological use [in hardware]."

I added "[in hardware]" to the Supreme Court quote, and ask, is my addition inconsistent with the original quote? If not, this quote is equally applied to most hardware and many analog patents, especially signal processor patents, if you understand the science. And Diehr's statement is not firm, since if the "acquiescence" is "in totality", Diehr says it is patentable.

"Since Alappat, the Federal Circuit has continued its
expansion of patentable subject matter through the
implementation of its formalists approach. State Street,
..."

I agree with Moglen here - once you make the scientific assumptions of Alappat, much of the following caselaw logically follows. But Moglen doesn't attack the science, but rather just the CAFC's behavior. So 9 pages into his 12 page brief, Moglen has been attacking the CAFC's legal behavior in "betraying" the Supreme Court, but not the science that the CAFC defends, with Moglen using lots of undefined scientific terms. A good way of sucking up to the Supreme Court, maybe, with the hope that they stop at this point, where Moglen now addresses the science.

"C. SINCE SOFTWARE DOES NOTHING OTHER THAN EXECUTE MATHEMATICAL
ALGORITHMS, IT IS NOT PATENTABLE SUBJECT MATTER AND, THUS, CAN
NOT BE A "COMPENENT[] OF A PATENT INVENTION UNDER 271(E)."

Fine. Here's my lawsuit heading:

"C. SINCE HARDWARE DOES NOTHING OTHER THAN EXECUTE MATHEMATICAL
ALGORITHMS, IT IS NOT PATENTABLE SUBJECT MATTER AND, THUS, CAN
NOT BE A "COMPONENT[] OF A PATENT INVENTION UNDER 271(E)."

Moglen implicitly argues against hardware patents because in relying on Benson/Flook/Diehr, Moglen writes ".. is eligible for patenting is one of substance, not form." Well, the only difference between hardware and software is one of "form". And again, unethically, Moglen doesn't define what he means by "form" and "substance". In fact he can't (or is planting a time bomb), since any scientifically-base definition of "form" for software to be used in patent law can be used with 17 USC 102(b) (which uses "form") to argue that software is not copyrightable (another Moglen goal). (Indeed, for a long time Moglen has refused to address and educate the open source community about the many legal problems with
17 USC 102.

"This Court rejected in Benson the patentability of a software
patent directed to a specific application of a generic
formulation because 'the mathematical formula involved here
has no substantial practical application except in connection
with a digital computer'."

Here's the scientific nonsense inherent in Benson. If I have a new, useful, non-trivial mathematical algorithm (and many software patents involve complex specific hierarchical data flows and structures, no mere
formulas) written in C that I blast into a field programmable gate array to be shipped to your computer and plugged into a chip socket, that is patentable as hardware. But if instead, I compile the C program and write it onto a USB drive to be shipped to your computer and plugged into a USB socket, that is NOT patentable as software? The only difference is Moglen's "form", whatever he means by that, not his "substance", whatever he means by that.

"The holding of Benson is properly applicable to all software,
because a computer program, no matter what its function, is
nothing more or less than the representation of an algorithm."

For my future lawsuit:

"The holding of Benson is properly applicable to all hardware,
because a digital circuit, no matter what its function, is
nothing more or less than the representation of an algorithm."

For Eli Lilly:

"The holding of Benson is properly applicable to all drugs,
because a drug, no matter what its function, is nothing more
or less than the representation of a unpatentable mathematical
quantum wavefunction."

LIE #6

The lie of omission Moglen makes here is in not stating that his arguments apply equally to hardware. The substance (say the underlying mathematics) is the same for both hardware and software - the Church-Turing thesis - all that differs is the form (say the physical embodiment). This principle is built into hardware/software codesign tools - a human designs the "substance" of what he or she wants in an application, and the computer generates the "form" in hardware, software or a combination of the two.
Moglen is implicitly arguing against both hardware and software patents if he understands anything about automata theory.

"[Software] is not conceptually different from a list of steps
written down with pencil and paper for execution by a human
being."

Nonsense for two reasons. First, it is true for hardware:

"Hardware is not conceptually different from a list of steps
written down with pencil and paper for execution by a human
being."

LIE #7

Second, the process of hardware and/or software executing something is very physically different from a list of steps executed by a human. First, much software involves large numbers of instructions, executing a large number of times - it is why we need gigahertz processors. The reason why we don't have humans do the same calculations is an issue of energy and physical resource engineering efficiencies. In the end, information processing is VERY VERY physical. It is impossible for humans by hand to execute most software.

"In no uncertain terms, the Court in Benson held that software,
which contains and upon command executes algorithms that
solve mathematical problems through the use of a computer,
was not patentable under 101."

To the contrary, because of the inaccurate scientific definitions used by the Supreme Court in Benson, the conclusions of Benson are uncertain.
Moglen, please I beg you, read Chisum's critique of Benson.

Moglen then quotes Flook:

"The notion that post-solution activity, no matter how conventional
or obvious in itself, can transform an unpatentable principle
into a process exalts form over substance."

Maybe. Maybe not. What are your mathematical and engineering definitions of "principle", "form", "substance"? Without such definitions, this Flook statement is meaningless (will someone buy Moglen a dictionary!). As are arguments based on Flook, such as what Moglen argues:

"Thus, claims to implement some method or accomplish some
process substantially through the use of software, which
does nothing more than encode [in binary form configuration
patterns for a general processor] and execute upon command
an algorithm to solve a mathematical problem, are no more
patentable than direct claims to software that solves
such a problem itself."

the comment in "[...]" my addition, an argument which I will use in my hardware lawsuit:

Thus, claims to implement some method or accomplish some
process substantially through the use of software, which
does nothing more than encode [as field programmable gate
array patterns for an embedded device] and execute upon command
an algorithm to solve a mathematical problem, are no more
patentable than direct claims to software that solves
such a problem itself.

The only difference between these two statements is that of "form", not "substance". Thus, Moglen opposes hardware patents as well.

And it is here that Moglen implicitly concedes his entire argument is
bulllsheet:

"The Court's decision in Diehr upheld the holdings in Benson and
Flook, and merely found that the claimed invention in that case
was not substantially directed to just software, but instead was
- in totality - directed towards an "industrial process for the
molding of rubber products", which is undeniably included within
the realm of patentable subject matter. ... Had the applicant
sought to claim the software used in that process, by itself,
however, this Court would have most assuredly found it to be
unpatentable subject matter just as it had in Benson and Flook."

LIE #8

So Moglen actually supports software patents (because he approves of Diehr, and Diehr approves of software patenting in general), as long as the result is not "substantially" but rather "in totality", is directed towards some "industrial" use, no matter what the "form" of the software is - its "substance" is patentable as software. WHATEVER THESE UNDEFINED TERMS MEAN. Look, Moglen's whole argument is meaningless, as long as he keeps on using and introducing multiple scientific terms without defining them. What does "in totality" mean, if that is the threshold for patentability?

Worse for Moglen, in accepting Diehr's approval of software patents under some conditions, Moglen agreee with the CAFC, the same CAFC that Moglen insults in his brief. For in the case In re Johnson (589 F.2d 1070 [1978]), the CCPA comments: "Additionally, the [Supreme] Court in both Benson and Flook REFUSED to hold computer programs non-statutory subject matter per so."

Diehr was the Supreme Court's facesaving way (using yet a new undefined concept - "in totality") to correct the science/engineering mistakes it made in Benson and Flook. There is an inconsistency across these decisions (as Chisum points out), because the Benson Supreme Court didn't fully understand computing. You might not like Alappat/Lowry/StateStreet, but at least they are consistent with each other (as well as consistent with science and engineering).

And if Moglen bothered to research, he would see that many, many hardware patents, especially signal processing patents (think of FFT chips) are directed towards no specific industrial use. Many hardware claims are litle more than pure mathematical algorithms, which definitely conflict with Benson and Flook, let alone Diehr. Moglen doesn't address this issue.

He partially concludes with:

"Thus, this Court's precedent repeatedly sets out that software,
which is nothing more than a set of instructions - an algorithm -
to be performed by a computer in order to solve some mathematical
problem, is subject matter than [sic] is not patentable under 101."

No, this Court's precedent is not that consistent. Benson is based on incorrect science, Flook continues that incorrectness, and Diehr plays a bit of word games to get everyone to ignore its mistakes in Benson. As opposed to the accurate science and engineering of Alappat - that software turns a general processor into a specific machine - that software is more efficient for shipping than shipping the same "substance" on a motherboard pluggable gate array.

Think of three different "forms" of the same "substance":

- an application specific integrated circuit (ASIC)

- a general processor with a socket for a programmable
gate array (FPGA)s.

- a general processor with a socket for an executable-carrying
USB disk drive

I write a program in the C language. With one compiler, I can generate instructions to have the ASIC manufactured. With a second compiler, I can manufacture the gate array by burning connections. And with a third compiler, I can manufacture an executable to be transported on the USB.
One "substance" - three "forms" - a trinity. In fact, with only one hardware/software codesign compiler, I can generate all three.

No one disputes the patentability of hardware-y ASICs, and barely dispute the patentability of shipable FPGAs. Why then dispute the patentability of the executable? Indeed, 13,000+ Beauregard patents argue to the contrary.

So these unscientific attacks on software patents need to find stronger ammunition than Benson. Because if this issue is directly addressed in a future case before the Supreme Court, all of the science and engineering problems with Benson will be formally addressed. And it won't be pretty.

Which is why the Supreme Court dissenters should not encourage this bad science. I think there are times when the Supreme Court deludes itself about its' scientific competence (at least Scalia admits that he doesn't understand science).

And Moglen - learn science, learn engineering, learn mathematics, learn about sematic definitions, learn about the political history of IP law, learn about Due Process, learn about critical academic inquiry (which includes reading Chisum) - just go back to collge and learn something.

Anytime, anywhere, I challenge Moglen to a debate. Otherwise, shut up.

And law professors should adopt a code of scientific ethics with regards to IP law briefs. Or stop filing them.

Greg Aharonian
Internet Patent News Service

COMMENT ON LILLY BRIEF

The electronics/software industries are in a mini-battle with bio/pharma industries over patent policy. In particular, bio/pharma is upset with attacks on injunctions. In a counterattack, bio/pharma, as championed by Robert Armitage of Eli Lilly, has filed a purely political amicus brief with the Supreme Court in the Microsoft/ATT case, asking the Supremes to eleminate all software patents.

There is little scientific and legal logic in the brief, so it is obvious that the point of the brief has nothing to do with law, but politics, a warning from bio/pharma to software. I understand and agree with the tactics of a warning, the fun of battle. But not in the form of a legal brief that is so badly written as to make it unethical to submit it to the Supremes as a legal brief. Eli Lilly should apologize.

What follows is a critique of the physics and computer science mistakes made by Eli Lilly in their amicus brief to the Supreme Court. The brief was written by Robert Armitage and James Kelley of Eli Lilly. The brief mostly ignores the main issue (whether any infringement is triggered under 271(f) when a component is shipped and use overseas), and instead focuses on whether or not software is patentable-enough to then be considered to be a component under 271(f).

Lilly's basic argument is that software is just information, information is not physical and doesn't fall under any of the Section 101 categories of patentability, and therefore software shouldn't be patentable.

First Lie of Omission

Armitage has little experience in physics and computer science. His brief cites NOTHING from the literature of physics and computer science, where issues such as what it means to be "physical" are heavily discussed. It is obvious he consulted no one with experience in such matters to help him combine such science with IP law. It is as if he holds the science he talks about in contempt. Not citing the science literature is a betrayal of the Constitutional imperative to progress science. What he also fails to mention is that out of Lilly's 3000+ patents, few if any deal with software. This is one area where Lilly has little to no expertise.

He starts out his whining early in his brief:

"The courts have overreached in determining those acts that infringe a patent. While perhaps not self-evident from the record below, ..."

Then why not talk about that issue in the brief, which Lilly has much experienced with, as opposed to software patents?

"... this overreaching is a byproduct of the failure of the courts
below to rigorously apply the prime statutory requirement for
patenting - that every claimed invention in a patent must be
limited to patent-eligible subject matter."

His arguments focuses on the categories of Section 101, and immediately he focuses on "physicalness":

"Machines, manufactures, and compositions of matter are clearly
tangible, physical things."

Maybe. Maybe not. It depends on what you mean by "physical" (a very complex concept in modern theoretical physics), and Armitage doesn't offer a definition of "physical". Indeed, it is easy to argue from definitions that information is physical by using dictionaries in the same way that Bob uses a dictionary in his brief (see my argument below).
Also, scientists now view that "physical" is a property of "information", as opposed to "information" being a property of the "physical". It all depends on your definitions, which Armitage doesn't provide.

He continues:

"Similarly, patent-eligible processes have historically been
defined as consisting of one or more tangible, physical steps,
rather than intangible or mental ones."

Again what do you mean by "physical" and by "tangible"? Otherwise, this sentence is meaningless. And to be honest, intellectually, he should have written this sentence:

"Similarly, patent-eligible processes have historically, BASED
ON THE SCIENCES OF EARLIER TIMES, been defined as consisting of
one or more tangible, physical steps, rather than intangible or
mental ones."

What Armitage is referring to is the 1877 decision, YES, 1877 decision, Cochrane v. Deener. Well, science has changed incredibly in 130 years, and it is also well known caselaw that laws based on science need to be reexamined periodically as the science changes. I quote from In re Bergy, which was reaffirmed in Diamond v. Chakrabarty:

"To insist on ... Congressional foresight in construing [patentable
subject matter standards] would be the very antithesis of the
Consitutional and Congressional purpose of stimulating the creation
of new technologies - by their nature unforeseeable - and their
progressive development. In re Bergy, 596 F.2d 952, 973 (CCPA 1979),
affirmed sub nom., Diamond v. Chakrabarty, 447 U.S. 303 (1980).

This applies to Supreme Court decision such as Cochrane as well - the Cochrane court could not foresee the next 100 years of science. What's pathetic about this "historial" argument that is contradicted by Bergy and Chakrabarty is that if anyone on the planet should be familiar with Chakrabarty, it is Armitage, since he was one of the leading proponents of expanding patentability in the Bergy and Chakrabarty era.

He continues on with history for awhile:

"The requirements for physicality and tangibility of what is being
claimed, even if claimed as a 'process', have been part of the
patent statute from the very beginning of the U.S. law on patenting.
In the 1790 Patent Act, Congress used ....."

Again Bob, reread Bergy/Chakrabarty which you have already read. When Congress drafted the 1790 patent act, it could not draft the laws in light of Maxwell's field equations for signals, the stastical physics of information and entropy, quantum field theories and general relativity theories on the nature of the physical, computers, gene sequences, etc., ...

- BECAUSE NONE OF THESE SCIENCES EXISTED AT THE TIME!!!

The Founding Father would be outraged as scientists to see us limiting our laws of science to the science assumptions of their times 200 years ago.

Much like Moglen in his lies to the Supreme Court, Armitage shifts to bashing the Federal Circuit:

"The Federal Circuit has taken views that appear on their face
to be inconsistent with the requirement that patented inventions
themselves must be physical and tangible."

Again, a meaningless sentence unless you define "physical" and "tangible".
(not that the Federal Circuit is brilliant in its handling of semantics).
He discusses the Federal Circuit doing this across State Street and Eolas:

"The Federal Circuit's State Street Bank opinion provided the
foundation for its subsequent decision in Eolas Techs. v.
Microsoft ... The Eolas court held that computer software code
by itself represented patent-eligible subject matter."

Why shouldn't it? This issue was hypocriticially papered over in 1994 when the PTO overruled its own Board of Appeals, and accepted Beauregard claims as patent-eligible subject matter, due to the insistence of IBM and the software industry. And despite the many lies to the contrary, Beauregard claims are claims on pure computer source code expression as useful articles of manufacture, with over 13,000 patents with such claims (claims that, if you are honest, render software all uncopyrightable).

So why shouldn't the CAFC, in State Street and Eolas, support the position of the entire computing industry (all of whom use Beauregard claims)?

What is bizarre is that Armitage can legitimately argue to the Supreme Court about software patents by just arguing Beauregard, without having to make use any "physical" arguments where he has little expertise.

At this point in the brief, Armitage starts with his science arguments:

"Software, whether machine-recognized 'object code' or human
intelligible 'source code', is - in and of itself - simply
information. It is instructions or directions. It is clearly
not a 'process, machine, manufacture or composition of matter'
under section 101 any more than it would have been an 'art,
manufacture, engine, machine, or device' under the 1790 Patent
Act."

Again, it depends on what you mean in the 21st centry by "manufacture" or "machine". I mean, how could the 1790 Patent Act reflect anything about software, when computers didn't even exist.

"The Federal Circuit, therefore, contradicted the patent statute
when it reached its Eolas conclustion that software code alone
could be subject matter eligible for patenting."

Maybe. Maybe not. It is a real shame here that Armitage doesn't focus on the real argument here - whether or not Beauregard claims are valid. What the Federal Circuit so concluded in Eolas completely logically follows from Beauregard claims. Sadly, other than making some assumptions of physics that are false, Armitage offers no legal or scientific reasoning.
He continues on with this physical argument:

"Section 101 completely refutes the notion that Congress permitted
an invention to be patented simply because the inventor could
point to a useful, concrete and tangible result produced by an
invention. It is the invention itself as set out in the claim in
the patent that must be tangible and physical, as must - in the
case of a combination - any of its discrete elements."

Not having any experience with software patents, Armitage is unfamiliar with software patent claiming practices, especially involving Beauregard claims (which qualifies him to be a district court judge in Northern California). The software industry, led by IBM and accepted by the PTO, have pushed boundaries here, by defining as a patentable computer program product a piece of software that is downloadable over a network (which presumably they view as a signal). So Lilly's complaint really isn't with the Federal Circuit, but yet another chapter in the patent policy wars between the pharma/bio industries and the electronics/computing industries.

"The jurisprudential foundations of the present case, which include
the errorneous State Street Bank framework for deciding issues of
subject matter eligibility for patent and, ...."

Technically, the Supreme Court upheld State Street when it didn't grant cert.

"... more especially, the Eolas holding that software code alone
qualifies as an invention eligible for patenting, clearly conflict
with section 101 and led to the wrong decision in this case."

Since Armitage uses multiple undefined terms, and incorrect statements of physics (re information), there is no basis for his "clearly conflict"
conclusion.

"To resolve this case, therefore, this Court should specifically
disavow the analytical framework in State Street Bank, which can
be wrongly construed to support patent eligibility for software
code and other intangible subject matter so long as it produces
'concrete results'."

Maybe the Supreme Court should do this, but not in this case. There was little discussion in the lower courts about this issue, and little real science provided in the filings and briefs about this issue. This is the absolutely wrong case to decide this issue, and Armitage knows this. Let's hope the Supreme Court knows this as well (and frankly, given reports about how Chief Justice Roberts wants to decide issues narrowly, I suspect the Supremes will wait for an honest case on this issue).

"This Court should hold that Congress, while making eligible for
patenting anything under the sun made by man through expansive
and inclusive language, has for more than 200 years consistently
limited what can be patented to physical and tangible results,
not intangibles that might produce some 'useful, tangible and
concrete results'."

Illogical. The argument rests on many undefined terms, it improperly restricts patent policy to rest on archaic scientific assumptions, it gives Congress more credit for IP law consistency than it deserves
(17 USC 102 is totally inconsistent undefinedness), and it ignores current trends in physics. The brief then starts focusing a bit more on the 271 issues:

"Thus, even if section 101 did not so require, every invention
expressed as a combination of elements must be physical and
tangible because section 112 commands that each of its
constituent elements must be structures, materials or acts."

Again, he is repeating meaningless statements that rest on undefined terms
- "physical" and "tangible".

"The plain meaning of section 112, sixth paragraph, is that a
discrete element of a claim to a combination must either be
set out as something tangible and physical (i.e., a structure,
material or act), or, if it is expressed as a means or step
for performing a specific function, then it will nonetheless
be interpreted as something tangible and specified function,
then it will nonetheless be interpreted as something tangible
and physical, i.e., the structure, material or acts described
in the specification or equivalents thereof. Either way, each
and every element in a claim to a combination can only be
subject matter that is tangible and physical."

Again, what do you mean by "TANGIBLE" and "PHYSICAL"? Define these terms, or else these sentences are meaningless.

"The reason [the CAFC] would have made such a conclusion is
that, as discussed above, software code by itself is neither
a structure, a material, nor an act, as each discrete element
of the claims to a combination must be."

Again, it depends on what your definitions of all of these terms are.
No such definitions are provided. What's hilarious is that Bob does have a discussion on semantics in the brief, but for lesser confusing
terms:

"Furthermore, the terms 'element' and 'component' are linguistically
synonymous. An 'element' is commonly understood as a 'fundamental,
essential, or irreducible constituent of a composite entity', or
'a constituent part' or 'a distinct part of a composite device'. A
'component' is either the same (e.g., a 'constituent element, as
of a system' or a 'constituent part') or very nearly so (e.g., a
'part of a mechanical or electrical complex'."

The Merriam-Webster online dictionary used for these definitions.

"....... It necessarily follows that each component of that
combination must likewise be tangible and physical."

Fine Bob, but where's your linguistic analysis of the much more complex terms "physical" and "tangible"? Let me prepare one for you. I use the much more authoritative Webster's Encyclopedic Unabridged Dictionary of the English Language, 1996 edition. In arguing that information and software is not physical, Bob is making use of the following definitions:

physical. adj. 2. of ... that which is material

that is, something is physical if it is material as thought of in the touchy-feely sense where Pamela Anderson is very physical. But so relying on this definition is illogical, since the same authoritative dictionary's definition of "material" is "material. adj. 8. formed or consisting of matter; physical" - that is the dictionary definitions of "physical" are meaningless because they are circular definitions.

Additionally, even if the defintitions weren't circular, the definitions have a second meaning:

physical. adj. 2. of OR PERTAINING TO that which is material;
3. NOTING OR PERTAINING TO THE PROPERTIES OF MATTER AND ENERGY
other than those peculiar to living matter

with the related definitions:

pertain - to relate; to reference
relate - to reference
reference - to refer to
refer - to direct for INFORMATION

which one can then use to argue, linguistically, that the definition of "physical" should be:

physical. adj. 2. of OR INFORMATION ABOUT that which is material;

which makes "information about" something "physical" in the patent world sense of the word "physical", including as a component. Everyone, which includes Lilly scientists, view temperature as something physical, but temperature is purely a statistical information measurement, which is why this second definition appears in the dictionary (and elementary particles with spin don't "spin"). Funny that Bob doesn't cite this part of the dictionary. Worse for Bob, my definition is consistent with current physics trends on the nature of the physical. I quote from a PATNEWS of March 2006:

Let's start with a new book that has just been published: "Decoding
the Universe: How the New Science of Information is Explaining
Everything in the Cosmos, from Our Brains to Black Holes", by Charles
Seife (Viking Press). The book addresses the issue so well that I
will be buying copies and mailing them to the policy people at the
PTO. Let me quote from just the introduction to his book (page 2):

"The laws of thermodynamics - the rules that govern the motion
of atoms in a chunk of matter - are, underneath it all, laws
about information. The theory of relativity, which describes
how objects behave at extreme speeds and under the strong
influence of gravity, is actually a theory of information.
Quantum theory, which governs the realm of the very small, is
a theory of information as well. The concept of information,
which is far broader than the mere content of a hard drive,
ties together all these theories into one incredibly potent
idea."

How you ask?

"Information theory is so powerful because INFORMATION IS PHYSICAL.
Information is not just an abstract concept, and it is not just
facts or figures [i.e., data], dates or names. It is a concrete
property of matter and energy that is quantifiable and
measurable. It is every bit [greg note: actually, it is bits -
see below] as real as the weight of a chunk of lead or the energy
stored in an atomic warhead, and just like mass and energy,
information is subject to a set of physical laws that dictate
how it can behave - how information can be manipulated, transferred,
duplicated, erased or destroyed. And everything in the universe
must obey the laws of information, because everything in the universe
is shaped by the information it contains."

"Information appears, quite literally, to shape our universe. The
motion of information may well determine the physical structure
of the cosmos. And information seems to be at the heart of the
deepest paradoxes in science - the mysteries of relativity and
quantum mechanics, the origin and fate of like in the universe,
the nature of the ultimate destructive power of the black hole,
and the hidden order in a seemingly random cosmos."

So if the physical is information, as modern physics as arguing, then wherever "physical" appears in the statutes, explicitly or implicitly, the statutes encompass information as well. Making Bob's argument meaningless.

Bob concludes in part with:

"Amicus Eli Lilly and Company, therefore, asks that this Court
declare the following: (1) While Congress intended that
anything under the sun made by man is eligible for patenting,
it has limited patent-eligible subject matter to what is
physical and tangible. This rule excludes patenting software
and software code alone, but does not bar from patent-eligibility
tangible inventions expressed as novel computer machines."

I am sorry, but because of multiple lies of omission, faulty definitions, and a lack of references to modern physics, the Supreme Court should totally ignore this request of Lilly. Lilly's brief is an insult to modern physics, semantics and computer, and it is a disgrace that crap like this is being filed to the Supreme Court.

Greg Aharonian
Internet Patent News Service

posted by James DeLong @ 1:22 PM | Patents

Link to this Entry | Printer-Friendly | Email a Comment | Post a Comment(0)