[Patents] Re: On claim language. (Was: could you amend the statistics if they

[halber@timsi.com.pl (Jozef Halbersztadt)]

From: Axel H Horns <horns@ipjur.com>

>
> Now let us proceed to "abstract functionalities" and to the problem-
> / solution approach. There is a truth that the subject-matter of a
> patent claim always should discriminate some inventive "solution"
> relative to some posed "problem". However, there is also a myth
> continuously fed by ffii and other anti-swpat activists (and
> regrettably not even properly recognized by some patent examiners,
> for example Mr. Kiesewetter-Koeblinger of DPMA recently interviewed
> by the German computer magazine c't), namely that functional claim
> language does not represent a "solution" because of the expert has
> to find the details of the implementation when attempting to work 
> the claimed subject-matter.
>
> What do you think of the claim language as set out below?
>
> "I claim a monopoly right to make (commercial) use of a certain
> technical teaching, the said technical teaching represented by
> making a device for the conversion of in each case a given first
> binary digital block of a first length into an associated 
> digital block of a equal length with the use of at least one
> freely second binary selectable binary control block, comprising  
> transistor and a second transistor, the gate of the second transistor 
> a first being electrically connected with the collector of the first 
> transistor ..."
>
> This would be a claim with "non-functional" but "structural
> features" loved in earlier decades by patent examiners e.g.  
> of the German PTO.
>
> Would such claim language be ok?
>
> Today, we prefer "functional features", e.g. the original claim
> language of claim 1 of EP 0 482 154 B1.
>
> Are "functional claims" a bad practice?
>
> I don't think so. Moreover, on a more basical level there is no
> difference at all.
>
> What _all_ patent claims do is to provide a _recursive_
specification
> of how to separate a first problem (the problem "solved" by the
> claimed subject-matter) into a set of sub-problems (for which
> solutions are known to the expert skilled in the art). If all of the
> said sub-problems are solvable without further inventive activity
> (e.g. due to some information generally available to the expert or
> by specific information set out in the patent description) then
> what the claim makes is to specify how in a bulding-block like manner 
> the solutions to the sub-problems can be combined such that the first
> problem is solved. In practice, the sub-problems are never
> unambigously solvable. Each sub-problem has to be treated by the
> expert separately until a semantical level is reached where the
> expert working the claimed invention can use e.g. trade mechanisms
> in order to obtain missing building blocks, e.g. by going into a shop
> and simply buying such building block components. There the recursion
> stops in practise. This is also true with regard to computer-
> implementable inventions. The expert recognises that at least one
> feature of the claim relates to a sub-problem that can be
> transformed into other sub-sub-problems by buying a processor and 
> coding some software text.
>
> If a sub-problem cannot be solved without further inventive activity
> the claim will be deemed invalid due to lack of workable disclosure.
>
> For example, a certain claim feature language refers to an "electric
> motor" or to a "pnp-transistor". What does "electric motor" or "pnp-
> transistor" mean? Both concepts are nothing more than a narrow
> "verbal dressing" for a functional specification, namely
>
> "electric motor" = any "means consuming electrical energy and
> outputting torque"
>
> "pnp-transistor" = any "pnp semiconductor means capable of
> controlling a first current by a second current"
>
> or the like. The real difference between "structural" claiming and
> "functional" claiming is this "verbal dressing". And, as a practical
> consequence, the concepts used for purposes of this verbal dressing
> are in general much more narrow with regard to their semantics than
> functional specifications. This is the reason why functional claim
> language is preferable.
>
> Strucural claim language is nothing else than an "under-cover"
> variant of abstract functional claim language but in general with
> very narrow scope.
>
> For example, assume there is the first and true inventor of an Anti-
> Blocking-System (ABS) for automotive brakes. The basic idea is to
> control the brakes of a car in a way so that the brakes never cause
> blocking of the wheels. Let us assume the inventor has implemented a
> laboratory prototype device entirely based on pneumatic technology.
> So, if no claims would be allowed openly showing functional language
> he might be directed to draft a claim entirely relying on pneumatic
> concepts, in fact being a narrow "under-cover" functional language.
> Such patent would, of course, be worthless in practice.
>
> If the inventor recognizes that there are other ways to implement
> his invention e.g. by electrohydraulics, he might wish to draft
> another additional claim covering electrohydraulical ABS. After 
> having slept a night and thinking further he might have another
> dozend ideas how to implement an ABS, resulting in another dozend
> of independent claims.
>
> Of course it would be extremely unhealthy if the regulations of
> patent law would force applicants to draft immense numbers of
> independent claims in order to maximize patent claim coverage of
> their invention. Under industrial conditions, inventions made by
> engineers oftenly are not of a somewhat narrow and particular nature
> but in the shape of a more general technical teaching. It would be
> unjust if the patent system would be made inelastic so that the full
> scope of the invention cannot be covered by a claim. Today's typical
> inventor is a scientific engineer, not a Daniel Duesentrieb-like
> individual (What is the name of Daniel Duesentrieb in U.S. editions
> of Disney's comics?; German word "Tueftler"). Therefore, functional
> claim language is absolutely essential for today's patent practise.
>

Polish examiners are even stronger belivers in structural claims. As
once I has been told by a head of the patent department IBM-Germany we
are 20 years behind Germans. It is a paradox, but Poles, Czechs as well,
kept some good German habits which in Germany withered away. Still,
compromises are inevitable. I do not deny admissibility of functional
claims. They allow to present the scope of protection in more general
way compare to traditional structural claims. They are inevitable
because modern technology is so complicated that sometimes only
"compressed" description is realisable on paper. But everybody should be
conscious where it is leading. When scope of protection is definited
loosely a second-comer can easily cut off a piece of land and take into
own possession. It is one of the reasons why we have nowadays
cross-licensing and inflation of patents (inflation of claims in one
appliciation too).

But I do feel it is unjustified to mix the set of problems and
phenomena from the paragraph above with drawing a line between
tangible and intangible innovative solutions. It is difficult to drawn
a line between physics and mathematics. None will deny that there is a
grey area, and a spectrum of views is justifiable. But those who deny
rationality of differentiation, who pretend that the wise men of the
past have not deeply researched the problem, and their findings are not
applicable today, have another reason to be deaf as an ignorance. I
susupect that for them the patent system is achievement of humanity of
such magnitude that they would like to offer its service to those who
unhappy remain beyond the civilisation. Behind this expansion is
similar drive as it once was behind colonial idea.

As we can differentiate between physics and mathematics we can in
future differentiate between tangible and intangible innovations. We
can also elaborate a procedure to deal with intangible innovations
that remain protected by granted patents after the legislative
bodies amend the law. Let's apply the rules from Article 31 WTO-TRIPS
Agreement that permits member states to issue compulsory licenses on
patented technology. The internal US law is accepting them as well - to
help government in accomplishing some political objectives. It is
interesting to find out when the United States interest would be
compromised if the patented invention were not adequately licensed and
when patents are not granted at all. In industrial field such special
treatment is given to atomic energy (nuclear arsenal) and air pollution
(to force a transfer of inventions from small inventors to Ford, GM and
alike ?). In Europe computer programs can be excluded from patentability
on general terms as backbones of the information society. To make it,
nothing more is needed as the political will.

The sui generis rights for software utility model would have been very
helpful in conducting such policy because it can simplify a procedure to
extract algorithms from the patented invention and to remunerate the
rightholder for the compulsory license.
1. Any person may at any time make application to the court for a
compulsory non-exclusive license to use the computer program embedded
in the patented invention and the court shall grant such license
unless it would be contrary to public interest.
2. The application should point out claimed elements of each algorithm
the compulsory license is sought and shall set forth the steps taken
by the applicant to obtain a patent license from the owner of the
patent.
3. The court is determining the scope of the granted license, the
owner of the patent shall be entitled to royalty payments determined
by automatic mechanism from the software utility model law.

Jozef Halbersztadt