From phm@a2e.de  Thu Jan  3 12:43:18 2002
From: phm@a2e.de (PILCH Hartmut)
Date: Thu, 3 Jan 2002 13:43:18 +0100 (CET)
Subject: [Patents] industrial application and swpat
In-Reply-To: <200201030029.g030TR708213@flem.fambjerke.dk>
Message-ID: <Pine.LNX.4.33.0201031254130.9779-100000@mulix.oas.a2e.de>

> According to the concise oxford dictionary, "industry" has several meanin=
gs:

> >From this it is clear the meaning of "industry" is much broader than jus=
t
> manufacture, and 'software industry' is correct and fully meaningfull. In
> Danish, it has a much more narrow meaning, i.e. manufacture whereas 'bran=
che' is more close to english
> 'industry'.

Indeed the english "industry" corresponds to something like "branche" or
"sector" etc in many other european languages, whereas "industrxx" in
these languages is more narrow, referring to serial production of material
goods.

It is rather striking to me that the english language seems to lack clear
distinctions between the material and the immaterial throughout several
worldfields.  I suspect that this is connected to a lack of sense for this
kind of distinction and what has been called "pragmatism" or
"utilitarianism" in british intellectual traditions.  It seems also that
both at the EPO and at the European Commission the British experts have
played a major role in blurring the related patent law concepts.  I have
heard this also from technicity experts within the EPO:  their previous
concepts started crumbling under the strong influence of "pragmatic"
but superficial englishmen who found these concepts particularly difficult
to understand and were supported by the tide of the time in not
understanding them.

> It is right that "industry" is not far form the meaning of technique
> (skill in art, method of achieving ones purpose, manner of
> performance, Oxford ...)  and technology (the study of mechanical arts
> and applied sciences, Oxford ..). Although industry has a broader
> meaning. The word technique stems from Greek techne (art, handicraft).
> But, it is not constricted to material production (let alone serial
> production) it is used about arts as well.

The greek "techne" actually comes from the opposite side: the fine arts.
Yet that is not more than etymology.

In German and other languages, words like "Technik"  have been firmly
established as equivalents of "engineering" (using controllable physical
forces) and antonyms of those arts that are based on pure reasoning.
A "Technische Hochschule" (polytechnical university) has these physical
engineering disciplines as its core subject matter.

There are also diverging uses of the word, especially in composites such
as "Klavierspieltechnik" (piano playing technique), but when standing for
itself, "Technik" (such as in Audi's slogan "Vorsprung durch Technik" --
cutting edge engineering) clearly evokes engineering with physical forces,
just as "Industrie" evoques those industries whose knowhow is in the
mastery of physical forces (not those who use a book-press for
commercialising a knowhow that is in the books but not in the press)
Just as the recent terms "Musikindustrie" or "Softwareindustrie" are
unnatural americanisms introduced by lazy journalists, so is the
term "Technologie" (lazy word for "technology"), which in real German
means "Techn[ik] - ologie" =3D "theory of engineering".  Again, seen
from the German (and other continental european) semantic perspective,
it appears that the english speaking people have a habit of blurring
the borders of material phenomena and mental theories about such
phenomena.   The english language seems to be exerting a bad influence
on thinking habits in this area, which is probably one of the reasons
why all anglo-saxon countries (US, UK, AU) were, according to Kolle's
articles, strongholds of software patents in the 70s, while the
european continent resisted them.

> So, looking at the meaning of the words, we can't conclude that technique=
 or
> industry just means manufacture, nor serial production. It has a much bro=
ader
> meaning - in English. The same is to some extent the case in French where
> 'industrie' also has a somewhat broader meaning than manufacture (e.g. vi=
vre
> d'industrie), and technique is used both concerning art and manufacture.

See the example of "Klavierspieltechnik".  Diverging uses don't change
the narrow basic meaning of the alone-standing word.

> Incidentally, I thought that "gewerblich" means manufacture as well as
> handicraft and trade.

> Of course the big question is if technique and industry is tightly
> associated with manufacture, i.e. serial production, in legal
> terminology. According to Hansen there are tendences to associate it
> with manufacture, but it is not that simple. "Industrial
> applicability" has also a broader meaning, "applicable in trade", and
> it is not confined to "serial production". And this must also be the
> case with EPC since "industrial application" means "made or used in
> ANY industry, including agriculture."

The latter could be a confirmation: agriculture is, in part, an industry
with serial production of material goods, even though the physical forces
involved are usually of biological nature.  The example also shows that
some of the readers apparently doubt whether agriculture is an industry.
They seem to have a narrower view of industry.

> Now, my point is: if EPO had interpreted EPC in that way, it would have b=
een
> fairly simple to decide that software cannot be patented. Any industrial
> applicable invention can be patented except a specified range of activiti=
es,
> i.e. programs for computer, presentation of information, business methods=
,
> etc.

The latter are, according to patent doctrine, not inventions.  So it is
inconsistent to say "any i.a. invention except for ...".

> The whole dirty trick made of EPO is avoiding taking point of departure
> in what  programs, presentation of information, etc. are. Instead, they
> defines "industrial applicability" and patentablity as being a physical
> technique, and everything which has physical effect or solve physical
> problems, must then be industrial and patentable. From this they conclude=
,
> that software which has physical effect or solve physical problems must b=
e
> industrial applicable, i.e. patentable. This has a groteque consequence w=
hen
> the appeal chamber of EPO in T110/90, IBM, Editable Document says that it=
 is
> not sufficient that a patent has industrial applicability for its
> patentablility - the technical character is decisive!!!! (Hansen, p 136) =
Art.
> 52.1 is not valid any more!!!! Keine Hexerei, nur Beh=E4ndigkeit.

The argument is of course quite right:  before industrial applicability
(together with novelty and non-obviousness) can be examined, it must be
established that there is an invention.  What is an invention is defined
negatively by art 52(2), but such a list of negative examples is not
enough for reaching a decision: it must be supplemented by an auxiliary
theory, that of technical character.  So far the EPO reasoning is
perfectly valid.

The trouble comes when the EPO allows its auxiliary theory to yield
results that contradict art 52(2) and then chooses its theory instead of
the law.  The law should supersede the theory and force its correction,
not vice versa.  The law is the authority, the theory is only auxiliary.

This becomes quite clear in

  Moses, the 10 Exclusions from Patentability and "Stealing with a
  further Righteous Effect"
  http://swpat.ffii.org/stidi/epc52/moses/

> So, what puzzles me is how the legal system would react to following
> argument. We do not want to discuss technical character or physical effec=
t
> because that is not mention in EPC at all. What is mentioned is industria=
l
> applicability. That is, patents can be taken in any trade, but there are =
a
> number of important exceptions, and software is one of them. Therefore al=
l
> software patents is illegal.

You can discard the auxiliary theory of technical character if you want.
But you still need a theory to explain the invention concept in art 52(2).
A simple negative list of examples is not enough.  You will at least need
some theory about each of the items listed therein and probably also some
theory about how these items are interrelated.  This follows from the
normal methodology of law interpretation.  And I doubt that you would find
any reasonable theory that allows you to patent computer programs as the
EPO has been doing since 1986 and, more bluntly, since 1998. Only the
pre-1986 theory (as found in the Guidelines of 1978) allows a clean
interpretation of the law, and from a constitutional view it is hardly
permissible for a lawcourt to change such basic rules which touch on
constitutional freedoms without a mandate from the legislator.  At least
German jurisprudence has explicitely seen it this way up to the 90s.

Once you do tweak Art 52(2) into the opposite of its original meaning,
"industrial applicability" will hardly help you.  This concept has been
devoided of its original meaning and left dangling there for a very long
time already.  Also, business methods and the Karmarkar algorithm for
linear programming are of course applicable to industries, even oil and
steel producing industries.  Of course this application is only indirect,
but as soon as you start getting into that argument, you are back in the
midts of the invention (technical character) discussion.

Another option would be to say that only an industrial application (in the
non-anglosaxon narrow sense) can constitute a patent infringement.  But
then you are not talking about whether a patent should be granted but
where it should be enforced.  That is an entirely different question which
has nothing to do with art 52 or 57 (industrial application) and is
currently not even regulated at the EPC level.  In Germany this is
regulated by Art 11 of the PatG (patent act).

--=20
Hartmut Pilch                                         http://phm.ffii.org/
Protecting Innovation against Patent Inflation=09    http://swpat.ffii.org/
10000 signatures against software patents       http://www.noepatents.org/





