Friday, 26 February 2016

UNION-IP Roundtable on Patent Priority

Today I attended a 'roundtable' meeting of UNION-IP, which was held at the German Patent Office in Munich. I had been invited by Reinier Wijnstra (WijnstraWise B.V) to do a presentation on poisonous priority, following interest in what Reinier had called "Tufty's Law" in an IPKat post from June 2013, referring to my earlier post here, and in light of the recent referral to the Enlarged Board in G 1/15 (which I have written about here).

The meeting, titled "Patent Priority: prosperous or poisonous" involved presentations from the EPO (Heli Pihlajamaa and Tilman Quarch from the Directorate of Patent Law), me, Mr Justice Birss of the England & Wales High Court, Judge Klaus Bacher of the German Federal Court and Judge Edger Brinkman of the Netherlands District Court at The Hague, with the meeting being chaired by Reinier Wijnstra. Each of us talked about issues relating to priority, with the main focus being on how partial priority should to be assessed, particularly in light of the problem around poisonous priority and poisonous divisionals.

The EPO's presentation went into a lot of detail about what the issues were, mentioning a lot of case law along the way. The most useful slide I found was the one shown here, which presents a list of the cases from the EPO Boards of Appeal on the two different sides of how partial priority is viewed. On the left side are those cases where a "literal" view of G 2/98 has been taken, where partial priority has been refused for a generic claim encompassing a narrower priority disclosure, while on the right are cases where a broader "conceptual" approach has been taken, allowing partial priority based on a theoretical division of a generic claim. The broader approach is the one I have been arguing for since I wrote my post on the subject back in 2013. The situation at the moment is that there is a clear inconsistency in the case law, which the referral in T 557/13 aims to resolve by referring questions to the Enlarged Board.

My presentation was about the same case law, with emphasis on the copper/metal example I have used a few times before on this blog. For a bit of fun I got several references to Alice's Adventures in Wonderland into the presentation, which might have made some sense to at least some of the audience. The key point, which I have expressed in this blog a few times, is that the whole issue of poisonous priority and poisonous divisionals goes away if you consider the generic claim to be theoretically divided into two parts, one of which has valid priority and the other of which has a later priority but is inevitably novel. This is essentially the conclusion reached in T 571/10.

The other presentations from judges in the UK, Netherlands and Germany did not concentrate exclusively on partial priority, but also discussed how priority in general was considered in the courts in these countries. There was a general view that these courts tended to be more lenient than the EPO on added matter, and also on priority in general, with one particular example from the German courts (discussed here) resulting in some surprise, not to say disbelief, from members of the audience. There have apparently been no cases from the courts in the Netherlands and Germany relating to partial priority, but this could be a feature of the more relaxed approach taken to allowing priority for a generic claim from a specific disclosed example. In the UK though, there has been a case specifically involving partial priority (Nestec v Dualit, which I have written about here). Mr Justice Birss expressed a clear view that the narrow approach taken in the Nestec case, and the various cases at the EPO, was not the right one, and the idea of poisonous priority was "obviously stupid" (which raised a spontaneous round of applause). In his view, the Enlarged Board would simply have to find a way to make the problem go away, and the conceptual division of claims was clearly the correct approach. He did, however, raise some interesting points about priority in general, and referred to the case of HTC v Gemalto, where the issue was whether a generic claim could have a valid priority right (not a partial one). In his view, it was more a question of support in the application rather than a test based on added matter. If the priority document would justify, in light of its disclosure as viewed by the skilled person, a broader claim then a later application having this broader claim could have a valid right to priority. In many cases this would make the poisonous priority issue go away, but would not work in all cases, one example being the ribonucleotides case of T 1127/00. In those cases though, since refusing priority was contradictory to the examples in the FICPI memorandum, the narrow literal approach was clearly just wrong.

At the end of the presentations, a couple of straw polls were held to see what the audience thought of two issues raised. The first was whether it was possible to have valid priority for a generic claim based on specific examples, where the skilled person and common general knowledge were used to justify the broader scope of the later claim. The audience was roughly split on that one, probably because it would be difficult to define a general rule and each case would depend on the facts. The second poll was about the question of whether the literal or conceptual approach was the correct one to take when assessing partial priority. The audience in that case was quite emphatic that the conceptual approach was correct, with not a single vote for the literal approach. Clearly the presentations on the day had got the point across to the audience. Let's now hope that the Enlarged Board reaches the same conclusion.


  1. Dear David,

    Many thanks for this interesting report. I have two comments on this topic.

    First, I do not think it is correct that there has been no case of partial priority in Germany. A few years ago I had the opportunity to work on a multinational litigation brought by Novartis against Johnson & Johnson on contact lenses. Claim 1 of the patent contained a functional definition of contact lenses. Three priorities were claimed. The priority documents contained examples of contact lenses but not the general functional definition. The Bundespatentgericht held that the priorities were invalid and that the patent claims were not novel over the priority applications. The decision (in German) can be found here:

    Second, when you look at T 1222/11, where the conceptual approach was used, you realize that the application was ultimately refused, due to a priority issue. More specifically, the Board found that the priority document was not the "first application" for the invention. So, in this case, the conceptual approach is what killed the application. My prediction is that if the Enlarged Board embraces the conceptual approach, there will be no more toxic divisionals, but there will be many more priority challenges based on the "first application" requirement. I have commented on this topic here:


  2. I am sorry for message butting in.

    I agree to your prediction that there will be many priority challenges based on the "first application" requirement. The EPC aside, however, the Paris convention stipulates this situation naturally.

    Namely, BODENHAUSEN explains as follows:

    “[I]f an even earlier application as a whole has already specifically disclosed these elements, that application will be considered the first application and priority cannot be recognized on the basis of the application mentioned earlier.”
    Please see Section H (d) on page 58.

    With regard to case of partial priority in Germany, in “Prioritätsdisclaimer” case (BpatG 30.4.2003; GRUR 2003,953), BpatG stipulates as follows:

    “If in a patent the priority of an earlier application has not been validly claimed, because the patent claim includes a feature which extends beyond the content of the earlier application, and if moreover the subject matter of the claim is not patentable if the filing date with the German Patent and Trademark Office is considered the date relevant for the state of the art relating to the patent, the patent may still be maintained in a limited form if a) the patent claim comprises a sub-combination disclosed in the priority document and b) a declaration is introduced into the patent claim stating that the feature in question extends beyond the content of the earlier application and that the feature cannot support patentability of the patent claim based on the priority of the earlier application.”

    I think that this logic is consistent with the explanation of Section 4F (d) on page 54 of BODENHAUSEN. However, the logic was denied by EBoA in G3/93.

    Patent practitioner in some Asian Country

  3. "Literal" approach vs. "conceptual" approach

    A better approach would be for the Enlarged Board to say that "a limited number of clearly defined alternative subject-matters" is not the correct test. Question 1 of the referral invites this - all they need to do is say "no".

    There would then be no need to argue whether the alternative subject-matters should be viewed literally or conceptually.

    1. I couldn’t agree with you more.

      In point 4 of the Reasons in 2/98, it was explained that “The possibility of claiming multiple priorities was introduced into the Paris Convention in order to avoid improvements of the original invention having to be prosecuted in applications for patents of addition.”

      In Australia, a Patent of Addition can be granted for a non-inventive “improvement or modification” notwithstanding the earlier publication of a “main invention” by the same inventor(s). Publication of the “main invention” between the priority date of the application for the main invention and the priority date of the application for the “improvement or modification” is excluded from the prior art base for the purpose of assessing the obviousness of the “improvement or modification. ”
      (Please see page 4.)

      Section 26(7) of UK Patents Act 1949 stipulated validity for patents of addition in the same manner as the current Australian Patent Act.

      If the effect of claiming multiple priorities is same as the effect of filing patents of addition, even publication of the original invention will not constitute prior art citable under Article 54(2).

      Article 54(3) goes without saying.