…well not really, but the German Federal Court of Justice has recently issued a decision (Kreuzgestänge, X ZR 103/13) that may expose Germany’s “Bifurcation System” to even more questions and criticism than in the past.

Bifurcation is a term probably originating from geography and generally means “splitting of a main body into two parts”. An example is a river bifurcation, meaning that a river flowing in a single stream separates into two or more separate streams (called distributaries) which continue downstream. See wikipedia for a couple of interesting examples. If one tries to transfer this term to the German patent enforcement system, it is readily apparent that this term does not fit well. Namely, wikipedia defines Bifurcation in law as “a judge’s ability in law to divide a trial into two parts so as to render a judgment on a set of legal issues without looking at all aspects”. This is not what happens in Germany when a patentee seeks to enforce his patent. The issue in Germany is that the defendant cannot successfully argue before the infringement court that patentee’s action should be dismissed because the patent is invalid. What defendant must do is to raise his objections against the validity of the patent at stake before a separate tribunal, i.e. the Federal Patent Court (Bundespatentgericht) from the outset. In the end, this means that there are usually two separate proceedings ongoing in parallel in Germany: infringement proceedings before the civil courts and invalidity proceedings before a specialised court, i.e. the Federal Patent Court. The “Bifurcation System” should therefore better be called “Two-Track System”.

Why does the Federal Patent Court not also hear infringement cases? Or why do the civil courts not also properly examine invalidity objections? These are good and legitimate questions to which cynics might answer that this is because history sometimes stands in the way of reason. Indeed, historically there was no Patent Court when the first German Patent Act was established. Thus, in the absence of an alternative, actions for infringement were heard before the ordinary (civil) courts. At the same time, however, it was possible to attack the validity of a patent by way of an opposition or a nullity action before the Patent Office. Under these circumstances the civil courts found it convenient to leave a thorough examination of what were perceived as “technical questions”, such as novelty and inventive step, to the Patent Office that had plenty of technically qualified staff, whereas the judges at the civil court were mostly “normal” German jurists with no special technical training. This remained so until after the Second World War. In 1959, the Federal Court of Justice decided in Moped-Kupplung (I ZR 93/57) that a patent that survived (pre-grant) opposition proceedings had to be accepted by the infringement court in the form in which it was finally granted.

When the Federal Patent Court was established in 1961 so as to comply with the constitutional requirement that review by a court must be possible against decisions of an authority, such as the German Patent Office, the legislators missed a great opportunity to fundamentally review the German patent enforcement system and only awarded the Federal Patent Court the same rights as the previous “Appeal Panels” of the German Patent Office had held. The main reason for this is that Germany is a federal state with considerable rights being awarded to the Länder (States), including the setting up of the civil courts. And the Länder liked the competence of “their” civil courts for patent infringement matters, because the patent boards of the civil courts were unusually profitable, compared to other branches of German civil courts. Therefore the Länder did not want to assign these competences to the “Bund”, i.e. the Federal Administration.

There we are, still today, and will remain until the Unified Patent Court redeems us from the “Bifurcation System”. Indeed I assume that the UPC will typically deal with both infringement and validity at the same time and by the same panel of judges, even though the local divisions will have the discretion under Art. 33(3) UPCA to bifurcate the nullity case to the central division.

History has shown that the German two-track-system is quite patentee-friendly and attorney-friendly :). To put it very briefly, this is because the infringement courts used to decide on their actions much faster than the Federal Patent Court came to a decision on validity. This opened up opportunities for owners of patents of questionable validity, but with a good case on infringement, to proceed against a defendant and obtain a provisionally enforceable injunction before the Federal Patent Court (or EPO Opposition Division) began scrutinizing the validity of this patent.

Aside from the fact that Germany is home to many patentees and that this blogger is a patent attorney, there are also other arguments that speak for a two-track system, such as an increased efficiency and a greater deterrence of infringers or even “product pirates”. But there are also good arguments against it, and the best of them are probably (i) that such a system is at risk of inherently causing injustice to the defendant who may receive an injunction before he has had a proper chance to present his objections to the validity of the patent asserted against him, and (ii) that there is an inherent risk that the patent at stake is given a different construction in invalidity proceedings than in infringement proceedings. The latter scenario is well known under the term “angora cat strategy”, meaning that the patentee might feel tempted to give the patent a narrow construction in invalidation proceedings so as to save it from prior art objections, but then forget about this construction in infringement proceedings when trying to capture the accused infringing embodiment under the wording of the claim.

It goes without saying that judges in a two-track system should be aware of these two risks and act responsibly to mitigate them. In regard to risk (ii), an obvious solution is that the infringement court accepts the claim construction of the nullity court or vice versa so as to avoid conflicting and unfair decisions. Moreover, both the infringement case and the nullity case may reach the Federal Court of Justice (FCJ) as the final instance in the end and at least at that stage one should assume that the FCJ will adopt the same claim construction in both cases. Anything else is, as one might argue, a bit schizophrenic and runs the risk of seriously undermining the foundations of the Two-Track System.

And then came the Kreuzgestänge decision of the FCJ in regard to a collapsible stroller (a buggy) (see X ZR 103/13). The court’s first headnote of this decision reads in my translation as follows:

The infringement court is called upon to interpret the patent in suit independently and is neither legally nor in fact tied to the interpretation by the Federal Court of Justice in patent invalidity proceedings concerning the patent in suit.

And in fact, the Federal Court of Justice then decided that the interpretation it had attributed to claim 1 in its earlier decision on validity and that was followed by the (infringement) court of appeal is rechtsfehlerhaft, i.e. wrong under the law. Whow! It does not happen every day that a Supreme Court tells the world that it had erred recently and now knows better. Here is how the Court argued its case:

The determination of the meaning of a claim is a matter of law and must be carried out by the infringement court as well as by any other court concerned therewith under its own responsibility (FCJ, Judgment dated 31 March 2009, X ZR 95/05, BGHZ 180, 215 para. 16 Straßenbaumaschine; FCJ, Decision dated 29 June 2010, X ZR 193/03, BGHZ 186, 90 para. 15 Crimpwerkzeug III). This includes the possibility that the infringement court will arrive at an interpretation that differs from the one that the Federal Court of Justice has reached in patent nullity proceedings concerning the same patent. It is true that such divergence, if it is relevant to the decision, justifies leave for judicial revision (BGHZ 186, 90 para. 11 et seq. Crimpwerkzeug III), but this is not different from other cases of a judgment on a question of law by a court of appeal that deviates from a decision of the Federal Court of Justice and hence justifies leave to judicial revision according to Sec. 543(2) no. 2 Code of Civil Procedure. In this as well as in those cases, the court of revision has to consider whether it stands by its previous ruling or whether the better reasons argue for the judgment of the appeal court. Such better reasons may also result from facts found by the court of appeal in patent litigation proceedings that were not found in the nullity proceedings proceedings but have an effect on the interpretation of the patent (FCJ, Judgment of 11 October 2005, X ZR 76/04, BGHZ 164, 261 para. 19 Seitenspiegel; Judgment of 12 February 2008, X ZR 153/05, GRUR 2008, 779 para. 31 Mehrgangnabe).

What does this mean? It seems to this blogger that the Federal Court of Justice had to weigh two fundamental principles here: Legal certainty and consistency of the Two-Track-System on the one hand against judicial autonomy and substantive justice on the other hand. It clearly went for the latter. At least the Federal Court of Justice may always correct itself, and even lower instance courts have the right to come to a different claim construction than the supreme court, provided only that they grant leave to judicial revision. What drove the court towards this decision can perhaps best be taken from a further citation from the decision: It was the fact that the FCJ’s earlier interpretation had excluded several of the working examples from the scope of the claim, and the court felt that this is generally undesirable.

Accordingly, an interpretation of the patent claim resulting in that none of the working examples described in the patent would be covered by the subject-matter of the patent can only be considered if other possibilities for interpretation that result in the inclusion of at least a part of the working examples can be compellingly ruled out or if sufficiently clear indications can be taken from the claim that something is indeed claimed that differs so largely from the description (BGH, Judgment of 14 October 2014, X ZR 35/11, GRUR 2015, 159 Rn. 26 Zugriffsrechte). If several working examples are presented in the specification as belonging to the invention, then the terms used in the claim are to be understood in case of doubt such that all working examples can be used to illustrate the invention.

As far as the risk of incompatible decisions on infringement and validity is concerned, the FCJ satisfied itself in the Kreuzgestänge decision (para. 28) that its change in claim interpretation in regard to the two- or three-dimensional implementation of the stroller’s cross bar (Kreuzgestänge) would have had no impact on the evaluation of the nullity grounds, i.e. the result on validity. Unfortunately, however, the decision did not consider the (not unrealistic) scenario where the claim interpretation does have an impact on validity. In such a case, judicial creativity might be required so as to avoid an unfair result for the defendant, i.e. that he loses the invalidation action due to a narrow understanding of the claim and then the infringement action due to a broad understanding of the same claim. The Kreuzgestänge decision is thus not really helpful to appease the critics of the German Two-Track System who had always been arguing that this system inherently involves risks of inconsistencies and potential unfairness to defendants.


Leave a Reply

Your email address will not be published. Required fields are marked *