[Patents] The ignored default option: Reassert the underlying rules of Art 52!

[PILCH Hartmut phm@a2e.de]

Dear Mr. Mueller, Dear Mr. Bolkestein, Dear European Commission,

Under

        
http://europa.eu.int/comm/internal_market/en/intprop/indprop/studyintro.htm

the European Commission's General Directorate for the Internal Market (DG 15)
has published a study titled


    The Economic Impact of Patentability of Computer Programs

which was prepared at the order of DG 15 by

      Robert Hart (Independent Consultant), 
      Peter Holmes (School of European Studies, University of Sussex) and 
      John Reid (IP Institute) 
      on behalf of Intellectual Property Institute, London

We have discussed this study at FFII and Eurolinux. 

It is to be welcomed that DG 15 has found it necessary to start investigating
this subject.

However, we find that this study does not address those questions that should
form the subject of the current consultation.  Moreover, some of the responses
from the study are not based on thoughrough investigations of reality but
rather on widespread prejudices of the patent law community.

Let me explain this by citing some pieces of text:

The IPI study says:

  Summary of and conclusions from the Study  

  [...]

  We return to the more general aims of the study.  

  I To provide information on a number of key elements to enable the 
  Commission to assess whether the envisaged Directive should be 
  confined to harmonising the relevant laws of the Member States on the 
  basis of the status quo as defined by the jurisprudence, or whether 
  it should extend the scope of application of the Directive.  

This is not the question that we should be discussing now.
We are now facing the following options

  (a) harmonise the relevant laws of the member states on the basis of
      the recent jurisprudence of the European Patent Office (EPO)
  (b) harmonise the relevant laws of the member states on the basis of
      Art 52 of the European Patent Convention and its underlying theory
      of technicity, as it was developped by national courts such as the
      Court de Paris and the German Federal High Court (BGH) in a series 
      of decisions in 1973-81 and as it has been upheld by some of the 
      national courts until this day

This can be further simplified to mean

  (a) adaption of the legal rules to the caselaw of the EPO 
  (b) adaption of European caselaw to reformulated legal rules

or

  (a) patentability for anything man-made under the sun, subject to
      occasional arbitrary restrictions based on EPO caselaw      
  (b) patentability only for inventions, i.e.
      technical products and processes with an industrial application,
      subject to clear criteria to be written into the EC Directive and
      perhaps into the European Patent Convention

Option (b) is really the default option that should first of all be considered
when an EC Directive is written and when the EPC is revised.  We are here
talking about a basic rule that lies at the foundation of information society.
Such rules need careful formulation by the legislator.  As the practise of the
courts evolves and deviates from the original rule, it is normal that the
legislator will reassert the original rule, unless there is an overall
consensus in society (not in patent-law circles) that this rule should be
changed.

The original rule of Art 52 EPC was at the time of its formulation in the 70s
carefully worked out on the basis of epistemological and economic theories
that went far beyond a simple discussion of caselaw, see

     http://swpat.ffii.org/vreji/papri/grur-kolle77de.html

The sophistication of these theoretical considerations of the 70s has hardly
ever again been reached in the discussions of the EPO and its friends.  There
can be no reason to attribute greater weight to the jurisprudence of the EPO
than to the underlying theories formed at the time where today's law was
formulated.

The IPI Study concludes

  On II and III: There may be other ways of formulating the options for 
  possible more wide ranging harmonisation beyond current levels but we 
  suggest the following three:  

  Option 1: we could stay with the status quo, subject to removal of 
  the exclusion of "computer programs" "as such". This would, we 
  consider, have no consequence except the important one that SMEs and 
  independent software developers will be less likely to consider 
  computer program related inventions unpatentable.  

This remark seems erroneous in several respects.

IPI option 1 ERROR 1:

What IPI calls "the status quo" is in reality only the jurisprudence of the
EPO, which is not fully enforcable in Europe.  Once this "status quo" is made
enforcable by a harmonisation directive, we arrive a different reality.  While
currently the German Federal Patent Court rejects claims to computer programs

          http://swpat.ffii.org/vreji/papri/bpatg17w6998de.html

on the grounds that they are excluded under the EPC, any move to legalise EPO
jurisprudence would force the same court to grant such claims.  The economic
consequences of this could be quite significant.  While currently software
patent claims are never enforced in courts, we could after such a change
expect a wave of infringement litigation.

IPI option 1 ERROR 2:

Under current law, not "computerprograms as such" but "computer programs" are
excluded from patentability.  Likewise, we are not currently discussing about
a possible deletion of the "as such" formula from Art 52.3 but only about a
possible deletion of the "computer programs" item from Art 52.2c.  It is as
silly to construct a difference between "computer programs as such" and
"computer programs not as such" as it is to construct a difference between
"games as such" and "games not as such".  Art 52.3 simply reminds the reader
that a patentable invention can be connected to a game or a computer program,
as long as it is a distinct entity and not the game or the program as such are
claimed by the patentee.

IPI option 1 ERROR 3:

To learn about the legal rules of patentability, any SME will always have
to consult secondary literature.  Just reading the raw law text is not
advisable, nor is it common practise.  On the other hand, if Europe adapts
the original clear rules of the EPC to the unclear practise of the EPO, SMEs
will have great difficulty in understanding the rules.  Currently, even
patent lawyers and national patent offices have great difficulty in
understanding and following the EPO's caselaw.  Dr.iur. Thomas Winnischhofer
concludes in his study at

          http://swpat.ffii.org/vreji/prina/drtw.pdf
          Patentrecht und Computersoftware

that there is no coherent system in this caselaw.  At the UNION conference in
1997, a large round of the patent lawyers and patent office representatives
complained that they don't understand the rules for deciding which computer
programs contain a "further technical effect".   If patent experts can't
understand the EPO's ever-changing criteria, how can SMEs be expected to
understand them.

The above-stated 3 errors committed by the IPI study represent common
prejudices of the patent law community.  The fallacies in these arguments are
quite easy to point out.  Some have been pointed out already in the 1970s, see
article of Gert Kolle quoted above.
                
The IPI Study goes on:

  Option 2: European law could ensure that the mere use of a computer 
  program/computer to implement an invention brings an invention within 
  technology, as appears to be the case in the USA. This would be a 
  substantial change from basic principles of Eiropean patent law. It 
  would be highly controversial. But it would bring European law into 
  alignment with U.S. law on patentability of business methods.  

This is a fake option.  Even in the USA attempts to patent the mere use of a
computer program to implement a known process are regularly rejected by the
patent office and the courts.

  Option 3: European law could be altered to have no requirement that 
  patents be limited to technology. If it were accepted that business 
  methods should be patentable simpliciter then this is the logical 
  consequence. But any attempt to make such a change would cause great 
  controversy. This would interfere in achievement of option 1 and in 
  achievement of adequate understanding by SMEs and independent 
  software developers of the opportunities and risks from the 
  patentability of computer program related inventions under option 1 
  or under the status quo.  

The IPI correctly recognises that, if computer programs are to be patentable,
it would be consistent to go even further, as has been done in the United
States by the State Street decision.  However, the IPI advises European
lawmakers to avoid the controversy that such a consistent approach could stir
and rather just legalise the current inconsistent EPO approach as a first
step.

This may be a smart political advice, but it does not serve the public's 
interest.

Fortunately, the IPI authors do not leave the latter completely unmentioned:

  Very relevant to options 2 and 3, as to any extension of patents, is 
  the caution amongst economists specialising in IP rights. As shown in 
  our economic study of the literature (Section III of our report), 
  most economists have doubts whether economic efficiency, i.e. 
  increased overall welfare, is achieved by having or making computer 
  program related inventions patentable. This caution is supported by 
  the continuing, indeed growing, concern in the USA on the issues 
  surrounding patents on computer program related inventions. The 
  debate in the States is not finished.  

  ...

Evidently, the IPI study was ordered at a time where the Directorate for the
Internal Market was not yet considering the option of reasserting a more
consistent and narrow definition of technicity as the one underlying the
currently still valid EPC.

We wonder why the IPI study wasn't published earlier.  Judging from its
content, it must have been known to the Directorate for quite a while.

Could it be that the IPI's scepticism about the macroeconomic effects of
software patents, cited above, were not in line with what DG 15 was planning
at the time?

Maybe we should stop speculating and forget this study altogether, since it
doesn't answer our questions anyway.

Within the Eurolinux Alliance there is a consensus for option (b), i.e. the
return to a narrow and consistent definition of technicity.

We propose to define the concept of invention as

     solution to a problem of manipulating physical objects,
     consisting of

       (1a) physical apparatuses
       (1b) knowledge of physical reality
       (2) calculus / algorithm / instruction to the human mind

     with the requirement that the innovative part must lie in 1.
     It can be discussed whether both 1a and 1b or only one of them
     need to be innovative.

The following texts contain detailed thoughts on how our technicity defining a
rule can be best formulated today and how it builds on existing legal
concepts:

     http://swpat.ffii.org/stidi/epue52/
     http://www.pro-innovation.org/     ;(French report, chapter 5)
           
We hope that the European Commission will choose one variant of this concept
as the basis of its Directive.

Moreover, we hope that the EPO and its board of directors can be persuaded to
leave Art 52 untouched in November.  It is neither acceptable to delete
computer programs from 52.2c nor to introduce the TRIPS universality clause
into 52.1.  This universality clause changes the delimiting nature of Art 52,
and it has repeatedly been used as a pretext to extend patentability.  It is
justified within the context of WTO free trade harmonisation, but not within
the context of definining the limits of patentability.

We can not have "some pregnancy".  Either we find a reliable way of performing
the traditional cut between matter and mind aka physics and logics aka world 1
and worlds 2 and 3, or we give up using the criterion of technicity and move
in the direction of unlimited patentability as practised by the United States.

We believe that we have found a way to formulate a clear rule, and that this
rule is helpful in separating real patentworthy inventions from the those
types of innovations, for whose promotion the patent system can in general not
serve well.

There are of course other systems that can be used to promote progress in the
realms of software and intellectual methods.  Please consider our concept for

       http://swpat.ffii.org/stidi/basti/
       Promoting Abstract-Logical Innovation: Logilege and Logidend

which also proposes some new rules that could be applied to the traditional
patent system in order to fix many of its problems.  Some people indeed
propose that the problems of the patent system must be solved mainly in the
areas of novelty and inventivity and in procedural areas such as that of
public scrutiny and compulsory licensing.  These difficult issues should be
addressed very soon.

For the moment however, we are facing the problem of finding clear definitions
and rules for determining technicity.  Such a definition is not only of
central importance to society, it is also surprisingly easy to find and to put
into practise.

Your work is of central importance to the development of the Information
Society in Europe and the world.  We need the most extensive and intensive
possible level of communication, involving expertise from science, philosophy,
epistemology, economics etc much more than from patent jurisprudence.  No step
that you might take in order to assemble such expertise can be to bold.

Sincerely wishing you courage, strength and wisdom.

--
Hartmut Pilch
FFII workgroup on software patents
http://swpat.ffii.org