Comments on Mr Shibata's article in EPI information 2/2017

The article " Reconsideration of G3/93 “Priority Interval” from the Perspective of G1/15 “Partial Priority”" of Mr Shibata in EPI Information 2/2017 deserves some comments.


The article points to paragraph 23 of the FICPI Memorandum about partial priority. Here the case is considered wherein (simplified) the priority document discloses chlorine, the application claims halogen, and public prior use of the chlorine has occurred in the priority interval. Now, with partial priority, the claim is said to be "valid in its entirety". However, the public prior use of chlorine renders the alternative with bromine obvious (that this alternative is obvious, follows from paragraph 8 of the Memo). Furthermore, the Memorandum indicates that if the public prior use was also of the bromine embodiment, then the claim would be invalid for the bromine embodiment (paragraph 24).

See by the way this comment of kaz-xmas on another blog.

Hence, the FICPI memorandum seems to assume the shield effect of the priority for chlorine for publication of that embodiment in the priority interval, i.e. such publications can not be cited, also not for the embodiments with bromine. On the other hand, any publication of the bromine embodiment can be cited. Hence, the priority of the chlorine embodiment does not extend to the bromine embodiment, renders the chlorine publication uncitable even for the bromine embodiment lacking priority.

Mr Shibata then proposes (as I understand it) that in case of partial priority, the part of the claim not covered by the partial priority, should benefit from the "shield effect" of the priority document. 

Mr Shibata's argument is interesting, but in my opinion not fully convincing. Article 89 EPC quite simply states that the effect of priority is the date of priority "shall count as the date of filing".  The FICPI Memorandum provides the legislative intent of Article 88(2) and (3), which are about the requirements (and limits) for claiming priority. In particular paragraphs 1-32 of the Memorandum are specifically about Article 88(2), second sentence "multiple priorities may be claimed for any one claim". The Memorandum does not engage with the quite clear statement of the effect of priority, even though the "shall count as the date of filing" wording was used from the outset of the drafting, see e.g. IV/5569/1/61-D in the travaux and the Avant-Project draft version (VE 1965; edit: possible link; page 17). 

Therefore, I do not think that it can be said that the FICPI Memorandum is clear enough to establish that it was the legislative intent to have a shield effect in case of OR claims.

In particular, for the shield effect, it should not matter if "halogen"  was claimed or simply "bromine" alone. There is also no logical ground why the shield effect would depend on whether an equivalent is claimed (such as bromine), or "chlorine + obvious feature B). However, then priority would provide protection for AND claims, and that is (unambiguously) rejected in the Memorandum (page 2, top). 

Finally, Mr Shibata is not entirely accurate where he says that "scope of the referral in G3/93 was "AND"-claim". The scope of the referral (by the President) was an application that  "comprises subject-matter not disclosed in the priority document". G 3/93 uses the example of an AND claim in point 3, but the following analysis is not so restricted and refers to priority that is invalid because the application "claims subject-matter not disclosed in the priority document". That clearly encompasses not only AND claims, but also equivalent (chlorine vs. bromine), and irrespective of whether a claim is entirely directed to such undisclosed subject-matter or only in part (a claim for chlorine OR bromine, or a claim for halogen).

Of course, one can maintain that G 3/93 is wrong in the sense that the shield effect of priority ought to be recognized by the EPO. Of course, priority needs to give not only the shield effect, but also the redate effect to protect against publications of equivalent embodiments during the priority interval), and for the redate effect there needs to be a limit on the protected subject-matter. Furthermore, the redate effect alone is sufficient to implement Article 4B of the Paris Convention, if one assumes that the effect of priority extends only to the same subject-matter in the later application. Mr Shibata challenges also that - citing Wieczorek, page 156. However, I do not find the remark of Wieczorek very convincing: the fact that some subject-matter does not form a second invention (but an obvious equivalent), does not mean that it is the same invention. Hence, Article 89 EPC and the redate approach are not irreconcilable with the Paris Convention.

Addendum 04.04.2020: see also T680/08, T665/00 and T1443/05 mentioned by Bremi in Singer/Stauder/Luginbuhl 8th ed. (2019) Art.89 rdn. 9 fn.11 and Mitt. 2006 p.490 (these decisions appear to actually deal with partial priority, G1/15).