|
04.09
What should be
patentable?
By Lee Hollaar
What should be patentable? It
seems like that question is answered by the
patent statute, which since 1793 has indicated
that “statutory subject matter” — what can be
patented — is “any new and useful process,
machine, manufacture, or composition of matter,
or any new and useful improvement thereof.” (In
1956, the original term “art” was replaced by
its up-to-date synonym “process”.) That
definition seems
particularly broad, especially when you consider
a "process" includes methods or ways of doing
things expressed in a number of steps —
essentially anything done by a person or
machine, and that Congress noted that “machine” and
“manufacture” encompass “anything under the sun
that is made by man.”
But statutory subject matter
must be expressed in broad terms. Because
patents are granted for new technology, it is
not possible for Congress to enumerate specific
categories (such as software-based inventions)
in the patent statutes. Such enumeration would
mean that truly pioneering inventions that
define a new field could not be patented, since
their field would be added by Congress only
after it was recognized. This outcome would clearly be
an unfair and undesirable result, and is
prevented by the broad sweep of the statute. The
same language has served for more than 200
years, despite dramatic changes in technology.
Court decisions hold that laws
of nature, natural phenomena and
naturally-occurring products are not patentable,
because they have no “inventor,” and patenting
them would be removing them from the public
domain. Abstract ideas also aren’t patentable,
because they lack “utility” by having no
particular use. And courts have found mental
processes unpatentable, but without a clear
explanation what a mental process is and why it
is unpatentable. One way of viewing mental
steps is that a process that involves steps that
require a subjective or aesthetic judgment on
the part of the person performing the process is
not patentable. This is not because the process
is somehow any less of a process, but because
the disclosure requirement for patents has not
been met. All patent applications must teach how
to make and use the claimed invention. If
judgments, rather than procedures that can be
taught, are necessary to carry out the process,
then there hasn’t been the required disclosure
and the application is faulty.
There are good reasons why not
everything made or done by man should be
patentable. Unlike copyrights and trade secrets,
a person can be an infringer even if he or she
has never seen the patented invention, heard of
it or its inventor, or even inadvertently
performed the steps of a patented method.
Independent creation is never a defense to a
patent infringement suit, just a justification
for not enhancing the damages. And the
independent creator who has invested millions of
dollars can be left with the choice of paying
royalties to somebody who did not contribute to
the product or be shut down by an injunction and
lose his or her investment.
Bilski’s method
Applicants commonly test the
boundaries of what may be patentable. In its
recent Bilski decision, the Court of
Appeals for the Federal Circuit (which was
established in 1982 to handle all patent
appeals) was faced with the following process
claim:
A method for managing the
consumption risk costs of a commodity sold by a
commodity provider at a fixed price comprising
the steps of:
-
initiating a series of
transactions between said commodity provider
and consumers of said commodity wherein said
consumers purchase said commodity at a fixed
rate based upon historical averages, said
fixed rate corresponding to a risk position
of said consumer;
-
identifying market
participants for said commodity having a
counter-risk position to said consumers; and
-
initiating a series of
transactions between said commodity provider
and said market participants at a second
fixed rate such that said series of market
participant transactions balances the risk
position of said series of consumer
transactions.
While the most likely infringer
would be a commodity provider, if the wording of
the claim were slightly different, it might also
be infringed by “market participants” or even
consumers. One can easily see how it might be
impossible to know whether a process claim like
Bilski’s, which requires a particular reason for
selecting a market participant, is being
infringed without filing suit and then asking
the defendant in a burdensome deposition what he
or she was thinking when he or she performed a
claimed step, if that step indicated a reason
for doing it.
That’s the problem with allowing
process claims not requiring the transformation
of an article or producing a special-purpose
computer. Anyone can be a patent infringer,
whereas before you had to make or sell things or
write computer programs. Before, only product or
software developers needed to be aware of
patents and clear their activities for possible
infringement. Now, everybody has to because they
may accidentally perform all the steps of some
patented process.
For example, if the process were
a new way of putting (and there are such patents
[1]), golfers could be liable for patent
infringement if they inadvertently putted the same way as the claimed method. While
there is no indication that these patent owners
eagerly watch golf on television, hoping to see
Tiger Woods putt using their claimed process so
they can collect royalties from him, it is
unlikely that they paid several thousand dollars
to apply for and get a United States patent just
to frame it and hang it on the wall.
The Federal Circuit addressed
this by saying that to be patentable, a process
had to be what traditionally was regarded as a
patentable process: “it transforms a particular
article into a different state or thing.” But
the Federal Circuit also said that a process was
patentable if “it is tied to a particular
machine or apparatus.”
Further confusion
While this resolved the issue of
the patentability of Bilski’s method claim (it
isn’t patentable), for methods that are claimed
as programming a general purpose computer, the
Federal Circuit essentially “kicked the can down
the street.” Now, instead of the question, “When
is a series of steps patentable?”, there are at
least two new questions: “What is a particular
article?” and “What is a particular machine or
apparatus?” While in a previous case [2] the
Federal Circuit held that “transient electric or
electromagnetic transmission” — signals — were
not “articles” as that term is used in patent
law, or any other class of statutory subject
matter, like Bilski that tells us that
something isn’t patentable, not what
is patentable.
Whether a computer was a
“particular machine” seemed to have been
answered by the Federal Circuit when it said
that programming a general purpose computer
“creates a new machine, because a general
purpose computer in effect becomes a special
purpose computer once it is programmed to
perform particular functions pursuant to
instructions from program software” [3]. A
recent decision of the Board of Patent Appeals
and Interferences, though, held that the claimed method
was “not tied to ‘a particular machine,’ but
rather is tied only to a general purpose
computer” [4] It is likely that decision, or one
like it, will be appealed to the Federal
Circuit, giving another data point on when a
computer-implemented process is patentable.
It should be clear that trying
to determine what aspects of computer-based
inventions should be patentable on a
case-by-case basis is not the best way to do it.
The old saying “hard cases make bad law” is
particularly apt here, where a person trying to
push the boundaries of method patents to include
things not tied to a computer made the validity
of perhaps thousands of issued patents for
methods that did not specifically mention a
particular machine, but rather said (or implied)
that the method was for a general purpose
digital computer, questionable.
IEEE-USA’s positions
IEEE-USA, in its
new position statement on patent reform [5],
calls on Congress, which specifies overall
patent policy, to “draw clear lines regarding
software patentability and infringement
liability that promote innovation” rather than
just resolve a dispute in light of existing law
like the courts.
A recent book [6] posits that a
major problem with patents today is that it is
difficult for people to determine what is
covered by a patent, and this lack of a
predictable property right produces uncertainty
for developers and costly disputes that may
outweigh the positive incentives of the patent
system. The authors’ research found that only in
some sectors of technology, such as the
pharmaceutical industry, do patents act as
advertised, with their benefits outweighing
their costs, while for software, the lack of
clear claiming has a definite negative effect.
With the recent uncertainly regarding methods
implemented using digital computers caused by
the Bilski opinion and decisions of the
Board of Patent Appeals and Interferences, this
is not only true for software developers who are
concerned with infringing a patent, but software
innovators who are not sure how to protect their
creative efforts with a patent.
In another
position statement [7], IEEE-USA noted that
“Because software-based inventions require
claiming in a variety of ways to get full
protection, such patent applications need to
have four or five times the number of claims as
other technologies” and that this can cause
problems when the patent office proposes
exponentially-increasing fees for claims in an
application, or requires burdensome “examination
support documents” when there are more than a
small number of claims. As part of its overall
approach to patents for computer-based
inventions, “Congress should amend the patent
statutes so that intentional distribution of a
computer program that uses a patented method is
an infringement of that patent, so that multiple
claim forms are no longer necessary in these
cases.”
Related to that, in
another position statement [8] IEEE-USA also
asks Congress to address the problems that may
result as people try to expand copyrights to
cover artistic expression. It notes that “Claims
to the media storing the program as an ‘article
of manufacture’ — so that those producing and
marketing computer programs that use a patented
method can be sued as direct infringers — are
now common.” But this results in “unrealistic
distinctions now used to explain why a medium
such as a CD containing a computer program is a
patentable ‘article of manufacture,’ while a CD
containing digitized music isn't patentable,
even though it has the same physical structure.”
The problem comes from the
patent statute’s requirement that the claim be
considered "as a whole" and not ignoring any
element of the claim. (This is supposed to keep
an examiner from rejecting a claim even if the
prior art does not teach a part of the claim.)
A proposed “media test”
What IEEE-USA’s position
statement proposes to Congress is that the
patent statute’s obviousness requirement should
be amended to say that patentable weight will
not be given solely to information stored in a
medium unless there is a novel and non-obvious
relationship between the information and the
medium. In addition, patentable weight should
not be given solely to the use of a computer in
business method claims where a prior art method
is implemented on a computer in a
straightforward way.
The second part just says that
implementing a well-known business technique on
a computer won't get you a patent, even though
there is no prior art that shows the claimed
combination of the technique and a computer.
To understand how the proposed
"media" test works, consider a new access
control technique that works only because of a
quirk in how a CD is formatted and processed
when it is read. The information stored to make
the technique work would be considered as a
claim element because of its novel and
nonobvious relationship to the medium.
Now consider a CD storing a song
as information, and the following hypothetical
claim:
We claim an article of
manufacture comprising a medium storing a song
with the lyrics: "Who put the bomp in the
bomp-a-bomp-a-bomp."
It would have been easy, if such
claims were really considered as a whole, for
Barry Mann and Gerry Goffin to get such a
patent. The claimed invention is likely both
novel and nonobvious from all the prior art
songs. But there is no special tie between the
lyrical claim element and the medium, so under
the proposed approach, the examiner gets to
exclude the lyrics (give them no patentable
weight) when considering the novelty and
nonobviousness of the claim. So, what's left is
just a "medium storing a song," which certainly
isn't novel.
Notice that in applying the
test, whether something was "technology" or
"artistic" wasn't even considered, and isn't
even mentioned in the test, so there is no need
to define those terms.
References
[1] United States Patents
7,261,652 (28 August 2007), 6,296,577 (2 October
2001), 6,019,689 (1 February 2000), 5,776,016 (7
July 1998), 5,616,089 (1 April 1997). The first
claim for the patent 7,261,652 is:
A method of a player holding a
golf club with a target side hand and a second
hand, such golf club having a shaft, a shaft top
having an end face, a club head attached to said
shaft, and at least one location on said shaft
for gripping the club, said method comprising:
pressing said end face of said shaft top against
the player's armpit or area at the front of
shoulder of the arm of the target side hand;
with the target side hand, grasping one of said
at least one location for gripping; and grasping
with a second hand, or resting all or a portion
of said second hand against, one of said at
least one location for gripping.
[2] In re Nuijten, 500
F.3d 1346, 84 U.S.PQ2d 1495 (Fed. Cir. 2007).
[3] In re Alappat, 33
F.3d 1526, 1545, 31 U.S.PQ2d 1545, 1558 (Fed.
Cir. 1994).
[4] Ex parte Langemyr,
Appeal 2008-1495 at 24 (BPAI, 2008).
[5]
http://www.ieeeusa.org/policy/positions/patentreform.pdf.
[6] James Bessen and Michael J.
Meurer, Patent Failure: How Judges,
Bureaucrats, and Lawyers Put Innovation at Risk,
Princeton University Press, 2008.
[7]
http://www.ieeeusa.org/policy/positions/patentclaimforms.pdf.
[8]
http://www.ieeeusa.org/policy/positions/PatentsLimited.pdf.
Lee Hollaar is a member and
past chair of IEEE-USA's Intellectual Property
Committee. He is a professor in the School of
Computing at the University of Utah, where he
teaches intellectual and computer law in
addition to technical courses. A senior member
of IEEE, he joined in 1966 as an electrical
engineering student at Illinois Institute of
Technology. He received his Ph.D. in computer
science from the University of Illinois at
Urbana-Champaign in 1975. Prof. Hollaar is a
registered patent agent, and the author of
Legal Protection of Digital Information,
available online at
http://digital-law-online.info.
Comments on this article may
be submitted to todaysengineer@ieee.org.
|
E-mail
this page to a friend
