Following the issuance of G 2/21 last year, we asked whether the plausibility elephant had left the room. Our Kluwer colleague and friend Miquel Montañá discussed this issue more recently here. Several decisions have meanwhile been issued applying the new “test” in G2/21, the lucidity of which may have reminded readers of the oracle of Delphi. Among these decisions, T 116/18 is by far the most valiant attempt to show the elephant the door. Yet comparison of this case with T 681/21 begs the question of whether we are already starting to see divergence on this important topic.

“Jumbo’s pitiful refusal to leave London Zoo tugged at the nation’s heartstrings”, Public domain, via Wikimedia Commons


T 116/18 – 3.3.02’s interpretation of G 2/21

To briefly recap, the relevant test in G2/21 is:

A patent applicant or proprietor may rely upon a technical effect for inventive step if the skilled person, having the common general knowledge in mind, and based on the application as originally filed, would derive said effect as being

i) encompassed by the technical teaching and

ii) embodied by the same originally disclosed invention.

TBA 3.3.02 interpreted this as follows:

  • i) means that the purported technical effect together with the claimed subject-matter need only be conceptually comprised by the broadest technical teaching of the application as filed – no literal disclosure of the effect is required, it is enough that the skilled person recognises the effect is necessarily relevant to the claimed subject matter (reason 11.10);
  • ii) requires asking: would the skilled person, having the common general knowledge on the filing date in mind, and based on the application as filed, have legitimate reason to doubt that the purported technical effect together with the claimed subject-matter, is an embodiment of the originally disclosed invention, i.e. the broadest technical teaching of the application as filed (reason 11.11)? If the answer is no, the effect can be relied upon and again no experimental proof or positive verbal statement is required in the application as filed (reasons 11.12 and 11.13).

Much ink has already been spilled analysing this decision. For us the key point is that this is by far and away the clearest and most detailed interpretation of G 2/21. If adopted by other Boards, it could provide much needed legal certainty for parties concerning the reliance on post-published data to support technical effects for inventive step. However, as the following analysis of T 681/21 shows, the early indications that T116/18 will become the generally accepted interpretation of G 2/21 are not good.


T 681/21 – background

To give a brief overview of the facts behind this case:

  1. The main request related to a fabric treatment composition comprising (inter alia) silicone and a specific polymer “CPP”;
  2. The applicant relied on the technical effect of improved softness and in particular a synergy due to the combination of silicone and CPP (reason 1.2.1);
  3. The description did not contain any examples;
  4. The description generally stated that the invention related to a composition including silicone which displays improved softening due to being a separate composition to the detergent;
  5. A polymer was essential, and CPP was the preferred polymer, though no reason for this preference was given.

The question before the Board was whether the effect 2. could be relied upon in assessing inventive step.


T 681/21 main request – the deadly side of G 2/21?

Topsy, the elephant who killed a man, Public domain, via Wikimedia Commons


Unlike in T 116/18, no detailed analysis of G 2/21 was given. Instead, at reason 1.2.3 the Board held that:

In the present case it is not in dispute that the application as filed does not relate to a synergistic effect arising from the combination of a silicone with CPP or any other component. Also the fact that the application as filed (page 2, lines 7-8) indicates the CPP to be a preferred cationic polymer without explaining the reason for this preference cannot foreshadow that the claimed combination would provide any type of synergism. The respondents did also not file any evidence that it was common general knowledge that silicone and cationic polymers may provide a synergism in terms of improved softness.

Therefore, it follows from the above reasons that the alleged synergistic effect would not have been considered by the skilled person as being encompassed by the technical teaching of the application as filed and has to be disregarded.

Although the underlying situations are not the same, it is difficult to reconcile this approach with that of T 116/18. After all, for this same element of the test i) “encompassed by the technical teaching”, T 116/18 held that the purported technical effect together with the claimed subject-matter “need only be conceptually comprised by the broadest technical teaching” – no “literal disclosure” is required, and the effect must instead merely be “necessarily relevant”. Arguably the broadest teaching in T 681/21 (i.e. that the invention related to a composition including silicone and a polymer which displays improved softening) encompasses the specific technical effect of improved softness and in particular a synergy with the claimed combination of silicone and the preferred polymer CPP. While the effect was not explicitly disclosed as in T 116/18, surely it would at least be understood to be necessarily relevant?

The objection in T 681/21 that patentee didn’t file evidence of the common general knowledge about synergism and improved softness is reminiscent of the ab initio plausibility approach, which was not endorsed by the EBA (but also not explicitly rejected – remember that G 2/21 rather resembles an oracle than a clear-cut opinion). It doesn’t sit too easily with T 116/18 either, which if anything adopted the opposite ab initio implausibility approach in holding that the opponent must demonstrate that there is a legitimate reason to doubt that the technical effect applies to the claimed subject-matter.


T 681/21 auxiliary request 4 –  the cuddly side of G 2/21?

Auckland Museum, CC BY 4.0 <>, via Wikimedia Commons


Things get even more complicated when we turn to section 5.2.1 of T 681/21 discussing auxiliary request 4, in which the silicone has been further specified as a preferred anionic type. For this request, patentee relied on the effect that the combination of anionic silicone with CPP provided unexpectedly better softening than a similar combination comprising a cationic polymer which was not a CPP. Once again, this effect is not supported by any data in the application as filed, and nor is it expressly identified. Nevertheless, the Board found that this effect could be relied upon under G 2/21:

In the board’s view, a skilled person reading the application as originally filed and having the common general knowledge in mind would derive therefrom as a technical teaching that the addressed improved silicone softness is especially obtained by using a combination with the components indicated as preferred, such as a CPP and an anionic silicone (page 2, lines 7-8 and 16). Therefore, said alleged technical effect can be considered to be encompassed by the technical teaching and embodied by the same originally disclosed invention and may thus be considered in view of G 2/21.

This more patentee-friendly outcome is much easier to reconcile with T 116/18. Concerning the different outcomes for the main and auxiliary requests in T 681/21, the Board in the passage above emphasised the preferred nature of the CPP and anionic silicone. But it is difficult to understand why the preferred nature of the anionic silicone in the auxiliary request would be a game changer allowing for consideration of post-published evidence in the absence of any data in the application. At the very least, T 681/21 seems to take a different approach to T 116/18 on this point, where it was sufficient that the purported technical effect together with the claimed subject-matter be conceptually comprised by the broadest technical teaching of the application as filed.

Another difference was the specific reference to “synergy” for the main request. If this was the reason that the effect was disregarded for the main request, it wouldn’t fit well with T 116/18 either. After all, synergy is a specific type of improvement, so one might expect that synergistically improved softness would be “encompassed by the technical teaching” of improved softness. Certainly 17.4.3 of T 116/18 suggests that synergistic effects do not require a special approach:

A synergistic effect, however, does not deserve a special position compared with other effects on which patent applicants or proprietors regularly rely for inventive step. 



Unfortunately, our borrowed African proverb

When the EBA fights the plausibility elephant, it is legal certainty that suffers

holds as true at the time of writing as it did when G 2/21 was issued last year. While the cases reviewed above relate to different underlying facts and cannot be directly compared, the fact remains that applying the approach of T 116/18 to T 621/21 would point to a different conclusion to that reached in the latter decision for the main request. It is difficult to see how this issue will be resolved until more Boards follow T 116/18 in clearly explaining what they think this test means.


To make sure you do not miss out on regular updates from the Kluwer Patent Blog, please subscribe here.

Kluwer IP Law

The 2022 Future Ready Lawyer survey showed that 79% of lawyers think that the importance of legal technology will increase for next year. With Kluwer IP Law you can navigate the increasingly global practice of IP law with specialized, local and cross-border information and tools from every preferred location. Are you, as an IP professional, ready for the future?

Learn how Kluwer IP Law can support you.

Kluwer IP Law
This page as PDF


  1. Loved the tenor of the article. But do tell: where in the world is the rock face which concludes the piece? Looks a bit like somewhere in the British Isles. Lundy Island perhaps? Googling “Elephant Rock” leads to many examples but not the one pictured. Perhaps the writer’s choice was to demonstrate how difficult it is to discern an elephant even when you are told that it is there?

    I’m mindful of that old saying: if it ain’t broke, don’t fix it. I had not realised that the plausibility test was broken and so hadn’t seen any pressing need to “fix” it.

    Then there’s Hilaire Belloc’s poem about Jim who ran away from his nurse and was eaten by a lion. It contains the line, perhaps having a degree of relevance here:

    “always keep a-hold of Nurse, for fear of finding something worse”.

  2. “Don’t make a mountain out of a molehill” –
    in German = Don’t make an elephant out of a mosquito

    To my understanding, the BoA in T681/21 did allow the post-filed data. However, the data of pthe patenteee were not convincing nor proving any kind of synergism. In sum, data only showing that A+B is better than A does not match the definition of “synergism” nor is it surprising (in particular if you consider that B would be also A) . This was also correctly found by the BoA. In addition, the opponent could establish doubt on the patentee’s data by providing conclusive counter data. In T116/18 such counter data was not provided by the opponent, thus the case hang on the one and only post-filed data point of the patentee proving synergism (where A+B was better than A and B separately according to Colby’s formula).
    All in all, the case in T 681/21 relied on very specific technical data and not on any plausibility elephant.

  3. Excuse me, but since when do applications have to include technical data to prove an effect? In the applications I deal with they are never provided and it would be impossible to provide them, but the effect is always plausible and even logical, it is up to the EPO to prove that the effect is not achievable: where is it written that there is no inventive step unless it is experimentally proven that there is an improvement? You just have to show that the person skilled in the art could have thought of it but would not have done so, all the rest is just academic games that should be played in a university, a technology centre, a research centre, etc. but not to recognize to someone an idea that they can then decide to use industrially or not depending on the practical improvements in reality. A patent office should not go beyond certain technical industrial assessments that are not within its competence.

    1. I agree that there is no requirement that applications include technical data to prove an effect. I don’t know which field you work in, but my understanding is that this issue arises most with complex cases in chemistry and biotech. In these cases one often cannot rely on logic to derive an effect from the claim features due to the complexity – you can speculate but will not know the answer until tests have been carried out.

      1. On the US American blogs people draw a distinction between the “predictable” sciences (broadly, the engineering sciences) and those sciences where the technical effects are not predictable (broadly chem-bio). In the USA, one has to demonstrate in the application as filed that one is “in possession of” the claimed subject matter. At the EPO, in those areas where results are not predictable, the objection to patentability varies, depending on whether the result is presented, in the claim, as a feature of the invention, or only in the description. But, no matter which jurisdiction, if you haven’t shown in your petition for exclusive rights that you are in possession of the effect you don’t qualify for a patent.

        What’s not to like, about any of that?

      2. Thanks, I definitely work in “predictable” sciences, I know that chemistry and biotech are always a bit special, I have the impression that a lot of regulatory effort is only put into these areas, which are however only a small part of the whole, and I find it a pity when someone tries to apply rules to other areas that only make sense in these areas, often creating a lot of confusion and uncertainty.

  4. Thanks for your comments Max – I’ll ask my co-author about the picture. As for Belloc’s poem, this is new to me and seems something like “better the devil you know”. But for this you need to know the devil – something which doesn’t seem to be true for G 2/21.

    And Thorsten, I agree that the data was not particularly convincing in T 681/21. But to my understanding this is not a factor in deciding whether the effect can be relied upon under G 2/21. In addition, while criticising the data, the board does initially discuss G 2/21 and hold that the synergistic effect does not meet the test for the main request in reasons 1.2.2 and 1.2.3. As for whether this is a mountain, molehill, elephant or mosquito, time will tell. But at least at present it remains difficult to predict when post-published data can be relied upon at the EPO, and these 2 decisions seemed to us to illustrate this.

    1. Adam, thanks for that reply.

      I’m a lot older than you so perhaps that’s why I remember the Belloc line from my childhood.

      BTW, I think that in those days, long ago, what Belloc called a “nurse” we would these days call a “nanny”, that is, a species of child-minder, rather than a medically-qualified healthcare professional.

      I’m not in chem-bio but it seems to me that every case will inevitably turn on its unique technical facts (that’s what makes our profession so rerwarding) and that therefore one is never going to be free of the “difficulty” of “predicting” whether, and if so how much, post-published evidence will win the case. We need a methodology for “plausibility” that approaches in intellectual rigour the problem-solution methodology for the obviousness enquiry. The evolution of such rigour takes at least a generation, i would think. My hunch is that the EBA is trying to stimulate that intellectual enquiry.

  5. Dear Thorsten,

    You have looked at two decisions following G 2/21. In the meantime, there are a few more. I will mention them here, but for a more detailed analysis, I refer to another blog.

    The “naughty” word “plausibility” might have disappeared, but now I have seen the word “credibility” creeping up. Where is the gain?

    I am still not convinced by T 116/18, as the synergy was disclosed for a different type of moth and I have difficulties in following the reasoning of the board. I might have missed something as not being a chemist.

    T 681/21 is much more convincing. The application as filed did not say anything about a synergistic effect, and it was right not to accept the pot-published filed evidence proving an alleged synergistic effect.

    In T 1989/19, the original application only spoke about stability during the manufacture of the medicine and did not say a single word about storage stability after manufacture. And yet, the board admitted post-published evidence showing a good storage stability. I am inclined to say that it is pretty obvious for a medicine to require some storage stability. Knowing that the product at stake has a natural tendency to clump, it does not seem really inventive to limit the water content during manufacture.

    When looking at T 116/18 and T 1989/19, the original specification contained a buzzword, i.e. synergy. So it was easy for the proprietor to claim a synergy which was not even thought of or envisaged at filing.

    In this respect, there is a danger that proprietors may latch on a buzzword mentioned in the originally filed application to later come with post-published evidence in order to nevertheless justify inventive step.

    In T 887/21, it was made clear that a mere hypothesis in the application as filed, without any experimental data supporting said hypothesis, is not enough to admit post-published evidence in the meaning of G 2/21.

    In T 2716/19, the proprietor filed post-published evidence he had been in possession for 18 years. Do you think this is reasonable? I do not.

    I do agree with Adam Lacy that it remains difficult to predict when post-published data can be relied upon at the EPO.

    G 2/21 is not one of the most pungent decisions of the EBA. It will have to be seen whether it will bring more problems that it might actually solve.

Comments are closed.