[Patents] More on why business methods are technical (fwd)

[phm@a2e.de (PILCH Hartmut)]

Greg wrote:

> I agree with Hartmut that there is one key essential argument behind all
> of these discussions on the patenting of software, hardware, business
> methods, genetics, etc.  The key issue is in Hartmut's words, this idea
> of "controllable physical forces", or my preference, "patenting
> non-physical processes".  This issue is what is underneath the partially
> thought out semantics of the EPO's Pension Patent decision - that
> business methods don't achieve a technical effect - with one view of
> technical effect being a physical result in the claimed process.

Well said.
The main logical problem with attempts such as that of the Pension Benefit
System decision is its failure to distinguish between the technical
teaching, in return for whose disclosure a patent is granted, and the
predictable "technical effects" occuring during execution of that teaching
in the material world.  In fact the Pension Benefit System contributed to
further blurring that distinction.

> Thus the question - "Should non-physical processes be patented"?  In
> America, while it has taking time to clarify the thinking, the answer
> for over 100 years has been YES, starting with the words of our Supreme
> Court in the late 1800s cases involving the early telephone and telegraph.
> The State Street and ATT/Excel cases pretty much settle the issue -
> business methods (whatever they are) are patentable.

It is true that patents have been granted for purely logical teachings for
a long time.  Maybe you know Rick Feinberg's book "peculiar patents".
These contain e.g. the idea of using music on a child toilet as an
incentive for a child to get done more quickly: as soon as the shit hits a
sensor, music playing is activated.  See patent US4849782 of 1989.
Another US patent from 1989 lets parents make children participate in
chosing an appropriate punishment by activating a card shuffling machine.
In all these and many older cases, we are taught nothing new about how
controllable physical forces achieve an unforeseen effect, although what
is claimed is a physical process.

> Thus the question - "Should non-physical processes be patented?" is of
> interest for two reasons.  First, how will Europe answer the question
> (since Asia [Japan, Australia] seems to be answering the question YES
> following the US logic)?  Second, if Europe answers NO (i.e. retaining
> the "computer program" and "business method" exceptions in EPC 52.2),
> which answer, YES or NO, will prevail when someone files a protest under
> TRIPS.
>
> How you answer the question depends on your view of the evolution of
> society and laws.  Hartmut prefers to stay in the past, when most
> inventions and case law was grounded in "physical" inventions, such as
> those traditionally thought of as mechanical.  This was a time when such
> inventions made up a large part of the economy, and therefore not
> surprisingly were the basis for much case law for patenting.  For
> example, about 20 years ago in the United States, half of all patents
> were mechanical, 35% chemical and 15% electrical.  Going back further I
> suspect even more patents were mechanical and less electrical (the
> electrical being where the boundary blurs between physical and
> non-physical).

As we can see from Feinberg's example collection, the boundary question
exists also in mechanical patents.  German courts were faced with many
patents of Feinberg's type and thought out the technicity doctrine in
order to deal with this problem.

If they had not found a solution, I suspect that we would already in the
60-80s have had a large percentage of non-physical innovations patented.
There have always been more non-physical than physical innovations, and
the

        Disposistionsprogramm decision of 1976
        http://swpat.ffii.org/vreji/papri/bgh-dispo76/

says clearly that the patent system is a specialised system limited to one
specific area of human creative endeavor, namely that of technical
inventions = teachings of direct application of controllable physical
forces, leaving the larger field of logical combinations to copyright,
competition law and other mechanisms.

> However, presently about 30% of US patents are mechanical, 35% chemical
> and 35% electrical (and I suspect similar percentages hold up around
> the world).  We see clearly that the role of the mechanical, the
> physical, is diminishing in importance to the economy.  Indeed, daily
> we are flooded with talk of the global information economy, the global
> service economy, etc.  I wouldn't be surprised one day to learn that the
> farms of America, farming being very mechanical and physical, were being
> run by one human and a large number of computerized machines.
>
> So the dilemma is this - to we allow the patent system to evolve to handle
> the types of processes, in this case "non-physical", that are becoming the
> predominate part of the economy, or argue as a fundamentalist - that the
> patent system was spawned in a mechanical/physical world, and should never
> evolve?

Patent law was not designed to regulate the "predominant part of the
economy" but a specific part, that of innovative use of physical forces.
For other parts, other regimes exist.

It is good when the law has some fundamentals to rest on as long as the
case for change has not been argued convincingly, both from the side of
legal systematics and that of macro-economic needs.  So far the arguments
I have heard, even the legal ones at the heart of US court cases (such as
those of Donald Chisum), fail to solve even the questions of legal
systematics, let alone macro-economics, and they do not come from those
who are working in the concerned arts (such as programming or finance) but
only from parties interested in patent law.

> This is the fundamental question, once you filter out the rhetoric.  And
> it is here that Hartmut and I disagree.  I say YES, the patent system
> should be allowed to evolve to handle non-physical processes such as
> software and business methods, because the patent system is the best way
> to describe such processes for public knowledge and setting boundaries.
> Hartmut says NO, either suggesting that software and business methods
> get no protection, or that we rely on copyright, and repeatedly citing
> old German case law grounded in the physical/mechanical.

Not quite right:  the German caselaw (BPatG/17 until today, BGH before
1992) explicitely deals with computer programs and is grounded in a
tradition of legal dogmatics that dates back 100 years but nevertheless
reached its climax of refinement in decisions and legal literature of the
70s and 80s.  It has no special relations to mechanics.  On the contrary,
the main decisions revolve around the question of why biological
applications can be considered technical while computer programs can not.
Theoretically clean answers were given, while at the same time Donald
Chisum and others rightly attacked the US courts for failing to give such
answers.

Thus my impression is simply that the German courts and scholars did
better work than most others, and that US law behaved according to the old
prejudice that people in continental Europe have concerning anglosaxon
law:  excessive pragmaticism and dogmatic shallowness, lack of systematic
considerations.

> I doubt anything Hartmut says will convince me to abandon my position,
> and I doubt anything I way will convince Hartmut to abandon his position.
> The question for the rest of Europe is which position makes most sense
> for the European economy.

Again well said.

I am glad to find that we did make some progress at clarifying a few
questions on this list after all.

--
Hartmut Pilch                                         http://phm.ffii.org/
Protecting Innovation against Patent Inflation      http://swpat.ffii.org/
90000 signatures against software patents       http://www.noepatents.org/