More on Patenting Computer Programs
1. State Street was the decision that opened the floodgates.
After it was decided, the number of computer-program-related patents
that the PTO issued rose to thousands per year. Therefore it merits
special scrutiny, both for the breadth of its holdings and for its consistency
with precedent and basic principles.
2. The State Street panel addressed two basic questions: whether
the judge-made exception for "abstract ideas" or the putative "business
method" exception to patentability rendered the accounting "system" at issue
unpatentable. In both cases, the panel answered in the negative, in
the process obliterating a the presumed business-method exception entirely.
3. Again, it may be helpful to consider what the claimed "invention"
was. The "system" was a machine and process for automating the accounting
operations of a so-called "hub and spoke" investment partnership. Among
the accounting items to be calculated was the daily net asset value of investments
allocable to each "spoke" of the investment partnership. For example,
suppose the pooled "hub" of the partnership held 1000 shares of X Company,
each of which, after the close of trading on a particular day, had a closing
market value of $20 per share. If a particular "spoke" of the partnership
held five percent of the investment pool, the net asset value of X Company
allocable to that spoke would be 1000 x $20 x .05 = $1,000.
This sort of calculation is representative of the kinds of simple arithmetic
the "system" in State Street was supposed to automate. Moreover,
there was little, if any, creativity involved in setting up the calculations
to be done. Virtually all were specified and required by rules and
regulations of the SEC, the tax authorities, and the accounting profession,
including the Financial Accounting Standards Board.
The "inventor" here certainly had not invented arithmetic. Nor had
he invented the accounting rules of the SEC, tax authorities, or the accounting
profession. Nothing in the district court's or appellate opinion contains
the slightest suggestion that he had invented any computer hardware or programming
language. So what did he invent?
4. The patent claimed the "invention" using so-called "means-plus-function"
claims. These claims describes an invention as a series of particular functions
to be peformed, each by unspecified means. Section 112, paragraph
6, of the Patent Act allows this sort of claiming if the patent specification
properly discloses means that will perform each of the specified functions.
See generally, Jay Dratler, Jr., Intellectual Property Law: Commercial,
Creative, and Industrial Property § 2.05[c] (Law Journal Press
1991 & Supps.), available on LEXIS: Secondary Legal: Law Journal Press.
On pages 258 and 259, the Federal Circuit has replaced, in brackets, the
nonspecific words "means" in each of the claim elements with descriptions
of the corresponding means stated in the patent specification, making it
seem as if the invention were a specialized set of hardware specially designed
to perform each arithmetic function as described. But was the invention
at issue really a special-purpose computer system with special hardware?
Wouldn't that type of claim be easy to circumvent? Couldn't
a competitor avoid the claims of the patent by performing the arithmetic
functions differently? "[D]uring licensing negotiations, [the patentee
had] informed State Street that any data processing system designed to perform
book accounting for a multi-tiered fund based on a partnership portfolio
configuration would infringe the . . . Patent." State Street Bank
& Trust Co. v. Signature Financial Group, Inc., 927 F. Supp. 502, 516 (D.
Mass. 1996) (district court's opinion), rev'd on other grounds, 149
F.3d 1368, 47 U.S.P.Q.2d (BNA) 1596 (Fed. Cir. 1998). The district
court also found as a matter of fact, not challenged on appeal, that:
"If Signature's invention were patentable, any financial institution desirous
of implementing a multi-tiered funding complex modelled on a Hub and Spoke
configuration would be required to seek Signature's permission before
embarking on such a project. This is so because the . . . Patent
is claimed sufficiently broadly to foreclose virtually any computer-implemented
accounting method necessary to manage this type of financial structure."
Id. When the true breadth of the claims is considered, isn't
the claimed "invention" in State Street any general- or
special-purpose digital computer system programmed to perform the stated
accounting operations, regardless of how it is programmed?
Back to Top
5. If the answer to that question is affirmative (if not, isn't
the patent rather useless?), then the basic issue of State Street
appears in stark relief. The inventor there did not claim to have
invented arithmetic, any particular accounting procedure used, any computer
hardware, or any programming language. For all that appears in the
opinion, the alleged inventor simply wrote a computer program, using a
programming language invented by someone else, to run on a computer invented
by someone else, to perform arithmetic calculations required by rules
and regulations devleoped by someone else, using arithmetic operations
known for millennia. By virtue of writing such a program, should
an "inventor" deserve a patent that can control a particular type of business
by monopolizing the accounting procedures necessary to engage in that
form of business? Wouldn't such a patent be a state-granted monopoly
on a type of business, which has been outlawed in Anglo-American society
for about four centuries, since Parliament adopted the Statute of Monopolies?
If one characterizes the effect of the patent in this manner, isn't the
court's extinction of the putative "business-method exception" merely
an afterthought? Could such a patent, so construed, coexist with
a rule precluding patents that provide monopolies on ways of doing business?
6. Are there any limits to the panel's holding in State Street?
The panel repeats the phrase "a useful, concrete and tangible result"
like a mantra, but what does it mean? In accepting the printout
or display of accounting numbers as such a result, isn't the court in
fact accepting any system or process that produces "useful" numbers? Aren't
most computed numbers useful to somebody, in some context? Aren't
specific numbers "concrete," and isn't a paper printout of them "tangible"?
Is there any perceptible limitation here? Also, does the court's
emphasis on usefulness confuse the requirement for statutory subject matter
with the separate requirement that an invention be "useful" under Section
7. More fundamentally, does the abstract/concrete distinction discussed
in Diehr, Alappat, and State Street make economic
sense? The three judge-made exceptions to patentability mentioned
in Diehr"laws of nature, natural phenomena, and abstract
ideas"seem to be based on the notion that no one should monopolize
the basic tools of invention and discovery, i.e., the fundamentals of
math, science and engineering that every inventor needs to do his or her
work. But does the abstract/concrete distinction properly capture
this policy? Is is clear and administrable? Isn't the question
whether claimed subject matter is an invention or part of the tools for
making inventions and discoveries a matter of discrerning judgment that
no formula can capture?
8. In any event, does the abstract/concrete distinction properly
draw a line between specific inventions which need and can benefit from
patent protection, on the one hand, and, on the other, ordinary business
ventures of the type that have been subject to the general rule of free
competition for about four centuries? Shouldn't the dividing line
be based on economic substance, rather than an evanescent distinction
that is itself an abstraction, such as the difference between abstract
and concrete results?
9. In 1966, in articulating the economic policy basis for the "nonobviousness"
criterion for patentable invention under what is now Section 103(a), the
Supreme Court said, "The inherent problem was to develop some means of
weeding out those inventions which would not be disclosed or devised but
for the inducement of a patent." Graham v. John Deere Co., 383 U.S.
1, 11, 86 S.Ct. 684, 15 L.Ed.2d 545, 148 U.S.P.Q. (BNA) 459 (1966). Isn't
this the proper policy basis for determining what "inventions" merit patents
and what do not? If this basic criterion were applied, would the
patent in State Street stand? Would Signature Financial Group
have declined to write a computer program to manage its hub-and-spoke
investment partnerships without the incentive of a patent? Would
Alappat have refused to use his cute mathematical trick to smooth curves
on computer screens? In contrast, would a pharmaceutical company
embark on a project of basic reserach, followed by exhaustive and risky
clinical trials (which together can cost hundreds of millions of dollars)
without the incentive of patent protection to prevent piracy of the new
drug once discovered and proven to be safe and effective?
Doesn't the "but for" test of Deere provide the proper balance
between the general rule against monopoly and the need for patent protection
where progress requires financial incentives? Yet isn't the "but
for" test itself rather vague and difficult to administer? How could
it be refined, without relying on equally slippery abstractions like the
abstract/concrete distinction? See Jay Dratler, Jr., "Does Lord
Darcy Yet Live? The Case against Software and Business-Method Patents,
43 Santa Clara L. Rev. 823, 840-853 (2003).
10. Some judges on the Federal Circuit appear to believe, as did
the late Judge Rich, that enforcing the subject matter line is not important.
The separate criteria of novelty and nonobviousness will weed out
"bad" patents, they seem to think, without having to draw lines between
patentable and unpatentable subject matter generally.
But is this so? The Patent and Trademark Office apparently granted
the patent in State Street after a thorough examination of the
alleged "invention's" novelty and nonobviousness. While the patent
might have been invalidated in litigation for failure to meet the requirements
that patentable inventions be new and nonobvious, wouldn't that litigation
be costly and dilatory, impairing competition by putting up a barrier
to entry into the "hub and spoke" investment partnership business?
11. To see how costly and dilatory full-blown patent litigation
can be, consider the case of Amazon.com, Inc. v. Barnesandnoble.com, Inc.,
239 F.3d 1343, 57 U.S.P.Q.2d (BNA) 1747 (Fed. Cir. 2001). There
Amazon had gotten a patent on "one-click" shopping on the Web. See
id., 239 F.3d at 1357-1348. The claims were quite broad,
not limited to any specific programming techniques; indeed they were broad
enough to encompass "single-action" shopping by means of technology other
than the ubiquitous mouse. See id. at 1348-1350 (reciting
and discusing claims). Apparently on the basis of cases such as
State Street, no one had challenged the alleged "invention" as
improper subject matter. The district court refused to invalidate
the patent for obviousness and granted a preliminary injunction. See
id. at 1347-1350. About two years later, the Federal Circuit,
although affirming the finding that infringement was likely if the patent
were valid, see id. at 1355-1358, vacated the injunction and remanded
for reconsideration of obviousness in light of new and improperly considered
prior art. See id., at 1366.
At this point, the alleged infringer, Barnes and Noble, had paid for a
complete trial and an appeal to Federal Circuit, with two years having
passed and no end to the litigation in sight. Had the defendant
been a small, innovative start-up company, rather than a substantial competitor
like Barnes and Noble, could it have managed to challenge such a substantial
barrier to entry into the "one-click" Web shopping business?
12. Patent doctrine also suggests that preliminary litigation over
subject matter is likely to be far less expensive and dilatory than full-blown
patent litigation. Full-blown patent litigation requires judicial
construction of the patent claims, which may entail a special "Markman"
hearing and an appeal of claim construction to the Federal Circuit. Only
after the patent claims have been construed can the district court address
their validity, exhaustively comparing the claimed invention with "prior
art," both that cited by the patent examiner and that introduced by the
infringement defendant in litigation. Then, if the patent is valid,
the court must compare the properly construed and validated patent claims
with the accused device, in detail, in order to assess infringement, perhaps
with the aid of a jury. Finally, in assessing infringement, the
court may have to address the doctrines of equivalents and prosecution
history estoppel, which may require an exhaustive review of statements
and claim amendments made by the patentee during patent prosecution. See
generally, Jay Dratler, Jr., Intellectual Property Law: Commercial,
Creative, and Industrial Property § 2.05 (Law Journal Press 1991
& Supps.), available on LEXIS: Secondary Legal: View More Sources: Law
Journal Press (discussing claim construction and infringement assessments).
In the end, full-blown patent ligitation may require as many as
three hearings (preliminary injunction, Markman, and full trial),
and three appeals to the Federal Circuit.
Should an infringement defendant who believes that an alleged "invention"
is clearly not of the kind for which the Constitution and the patent statute
authorize patent protection have to go through all this? Does the
daunting prospect of such prolonged and expensive litigation provide a
barrier to entry into competition nearly as high as a that of a valid
patent itself? Woudn't it be far simpler and more efficient economically
for the courts first to decide whether one-click shopping is a patentable
technology or an ordinary business venture subject to the age-old rule
of free competition in free markets? Under these circumstances,
does the Federal Circuit's refusal to draw a line on subject-mattter ground
make economic sense?