EP 2 960 746 B1 relates to a control system and method for controlling the movement of an aerial apparatus, i.e. an aerial firefighting ladder controlled by a joystick.
Brief outline of the case
The patent was revoked due to lack of N of most of the requests over a public prior use, i.e. a real firefighting ladder inspected by the OD, and some supporting documents as well as a witness hearing. Two AR were not admitted by the OD.
The board confirmed the revocation.
The case is interesting in that the board gave a very clear interpretation of G 1/24.
The proprietor’s point of view
According to the proprietor, “consulting” the description and the drawings means that the definitions of a claim feature provided in the description, and the drawings, should be read into a claim.
The proprietor argued that the present board deviated from the interpretation of the EPC contained in G 1/24 within the meaning of Art 21 RPBA by not considering the description and the drawings of the opposed patent when interpreting the claims of the patent. It did, however, not submit any specific question to be referred to the EBA.
The board’s decision
With respect of the manner in which G 1/24 should be applied, the board stated the following:
- First, the wording of the claims forms the “basis” for their interpretation and, for the purpose of interpreting the claims, the description and the drawings should always be “consulted” or “referred to”. G 1/24 is, however, silent as to the very purpose of such “consultation” or “reference”.
- In particular, it cannot be inferred from G 1/24 whether the description and the drawings should be “consulted” for the purpose of, for example, deriving “explanatory aids” to claim interpretation, see e.g. UPC_CoA_335/2023, second headnote, confirming the ordinary meaning of a claim feature, determining the function or the purpose of a claim feature, or simply defining the technically skilled reader of a claim see e.g. T 1924/20, Reasons 2.7, etc.
- Second, discrepancies in claim interpretation practice between national courts, the UPC and the BA are undesirable and should thus be avoided.
- Third, it is up to a patentee to remedy discrepancies between the description and the claims. In other words, patentees are the masters of their fate.
In reply to the proprietor’s point of view, that G 1/24 was to be understood such that a feature of the claim must be interpreted in a more limited manner based on features set out in the embodiments of the invention as disclosed in the description, the board found no authority for interpreting a claim more narrowly than the wording of the claim as understood by the skilled person.
First, because “consulting the description” does not imply any specific result of such consultation.
A consultation of two conflicting authorities does not a priori determine who is master. G 1/24 does not, as the proprietor seems to suggest, require the boards to give the description preference over the wording of the claim. To the contrary, according to the Order of G 1/24, “the claims are the starting point and the basis for assessing the patentability of an invention”, and the present board added that, in line with UPC_CoA_335/2023, second headnote, presently not available, the claims are the “decisive basis”.
Second, such interpretation would contradict the well-established jurisprudence of the BA, which was essentially confirmed in decision G 1/24, Reasons 10, see e.g. also T 1999/23, Reasons 5.6.
Third, such an interpretation would directly contradict the second takeaway from decision G 1/24, namely a harmonised claim interpretation between national courts or the UPC and the BA. For a long time, it has been considered unsatisfactory that patentees could get away with different claim interpretations for validity, on the one hand, and for infringement on the other.
The board quoted Sir Robin Jacob in decision UK Court of Appeal, European Central Bank v DSS [2008] EWCA Civ 192 at n 81, in which “Professor Franzosi likens a patentee to an Angora cat…”.
- For the board, the problem Professor Franzosi described would become relevant if the board were to interpret a claim below the wording of the claim as understood by the skilled person skilled in order to find it patentable, while an infringement court would interpret the claim literally when determining infringement.
The board also noted some further judicial precedents:
UPC-CD Paris-30.04.2025-UPC_CFI_230/2024, first headnote, temporary not available
- “a broad, general term used in the main claim is not to be limited to an understanding derived from the more specific or narrower features disclosed in a dependant claim or in the description.”
LD-Hamburg-30.04.2025- UPC_CFI_278/2023 (Agfa/Gucci)
- a claim feature was interpreted in a restricted manner on the basis of an explicit definition set out in the description.
German Federal Court of Justice-07.09.2024- X ZR 255/01– Bodenseitige Vereinzelungseinrichtung-
- “An example embodiment does not as a general rule permit a restrictive interpretation of a patent claim that specifies the invention in general”
Given the above precedents, the board concluded that, in order to do justice to the second takeaway from G 1/24 and align the judicial practice of the BA with that of the national courts and the UPC, a claim should not be interpreted, based on features set out in embodiments of an invention, as having a meaning narrower than the wording of the claim as understood by the skilled person skilled.
Only as an aside, the board notes that the above findings also correspond to court practice outside Europe, e.g. in the US and in Japan.
The board has “consulted” and “referred to” the present description and drawings to define the technically skilled reader from whose perspective or view point a claim is to be interpreted.
After careful analysis of the features at stake, the board confirmed the lack of N of claim 1 as granted in view of the PPU.
In its catchword, the board specified that
- In cases of discrepancy between the claim language and the description, it falls upon the patentee to remedy this incongruence by amending the claim. It is not the task of the BA to reach such alignment by way of interpretative somersaults.
Needless to say that the board never considered a referral to the EBA.
Comments
The present interpretation of G 1/24 makes it very clear that the boards will apply it in a way which will not suit most applicants/representatives. To me, the notion that the patent is its own dictionary is dead.
The catchword is very clear in this respect. Any incongruency, not just inconsistency, has to be removed by claim amendment.
If the applicant/proprietor wants to give a claimed feature, clear as such for the skilled person, a more limited or a broader interpretation, it is up to the applicant/proprietor to amend its claim in order to give the claim the interpretation it wishes. The interpretation cannot be hidden somewhere in the description.
To sum it up, interpretative somersaults are not to the liking of the boards.
As the present decision has been taken by an enlarged composition ot three TQM and two LQM, as well as distributed to all chairpersons and members of the OD, it bears even more weight.
In view of the way boards interpret G 1/24, it can be guessed how the Guidelines might be amended.
If the differing interpretation from the description is incorporated in the claims, alignment between claims might become quasi automatic. It could well be that only embodiments not falling under the claims will have to be marked as such.
Comments
5 replies on “T 2027/23-A further interpretation of G 1/24-No interpretative somersaults for the boards”
T2027/23 was discussed todau on the IPKAT blog. There, the blog author, Dr Rose Hughes, stated:
‘By stating that patentees themselves must fix inconsistencies, and that the Boards of Appeal will not perform “interpretative somersaults” to do it for them, T 2027/23 might be seen to indirectly support the case for requiring such amendments to ensure legal certainty. On the other hand, making it up to the patentee to align the description and the claims could be read as supporting the case that mandatory description amendments should not be forced on patentees, given that legal certainty is provided by an established approach to claim interpretation that focuses on the primacy of the claims and does not allow inconsistencies in the description to override the claim language. If this is how claims are to be interpreted, what need is there to amend the description? ‘
I thought this summarized the issue quite well, that’s why I copied it.
I support the second of these two interpretations, whereas you appear to be in the first camp.
On the issue of the patent being its own dictionary, I still do not agree with you. There are at least two cases in which the patent can be its own dicstionary:
1) If a certain term is defined differently in the description than the common understanding, it follows from T2027/23 that the applicant/proprietor should amend the claims. If it is a limitation, usually the applicant/proprietor can do this, and in such a case the patent IS, at least before the amendment, its own dictionary.
2) Also, T2027/23 deals, like any T decision, with a specific case. T2027/23 did not concern an explicit definition, in which the description says ‘in this document, feature A is defined as ….’
In such a case with an explicit definition, the applicant should have been forced to place this definition in the claims before grant, but if, somehow, this was missed, I find it hard to believe that that an OD or BoA would then ignore such an explicit statement and consider that an interpretative somersault, instead of a straightforward interpretative act. So, also in the case of an explicit definition in the description, I still believe, notwithstanding T2027/23, that the patent can well be its own dictionary.
I agree that the patent will not very easily be taken to be its own dictionary, but it is definitely not excluded either.
Dear BJ,
Thanks for your comments and drawing my attention to the IPKat dealing with T2027/23.
I have great respect for Dr Hughes opinion, but like your comment I consider it similar to the glass being half full for some, and half empty for others.
I do agree that “T 2027/23 might be seen to indirectly support the case for requiring such amendments to ensure legal certainty”. I would go one step further: T 2027/23 directly and not indirectly, supports the case for requiring such amendments to ensure legal certainty.
It is your good right to disagree with me on the issue of the patent being its own dictionary. I still beg to differ.
What good is it to decide that the patent is its own dictionary if the so-called notion of a dictionary disappears when the claim is amended in order remove the incongruence between the claims and the description? A patent could only be its own dictionary if the dictionary definition stays in the patent, or have you seen definitions in a dictonay vanish?
When an intermediate generalisation is corrected by adding the missing features present in the description, which allow to overcome the objection of added matter, would you consider the patent being its own dictionary? I do not, as it boils down to re-establish what has been originally disclosed.
I can agree with you that T2027/23, like any T decision, deals with a specific case. T2027/23 and did not concern an explicit definition. If a different meaning is given in the description with respect to that of the skilled person reading the claimed feature, the applicant/proprietor will not herald t so bluntly.
I agree with you that in the case of an explicit definition, the applicant should have been forced to place this definition in the claims before grant. If, somehow, this was missed, it is correct in my view that an OD or BoA will then ignore such an explicit statement and consider that an interpretative somersault, and not accept the difference as a straightforward interpretative act.
Until the interpretation of G 1/24 will have diffused in applications, we will often be confronted with such difference of interpretation between claims and description. The lesson to be taken from board decisions following G 1/24, is that the applicants/proprietors cannot bank on the fact that a different interpretation of claimed features in the description will be accepted. The board made clear in T 2073/23 that patentees are the masters of their fate.
After all, in T2027/23, the board has purportedly included a reference to the famous Angora cat. It should not be permitted to the proprietor to have one interpretation in case of validity and one in case of infringement.
The implementation of G 1/24 in T 2027/23 is welcome. The Board’s statement that « a claim should not be interpreted, based on features set out in embodiments of an invention, as having a meaning narrower than the wording of the claim as understood by the skilled person skilled » is fully aligned with the settled case law of the BOAs, i.e. the rule of the broadest reasonable interpretation.
The same principle is stated in UPC decisions, such as CD Paris [230/2024] Kinexon v. Ballinno cited by Mr Thomas, dealing with a revocation request. The decision is actually available on the UPC website, it was published on 21 May 2025. Generally speaking, the primacy of the claims is the leading principle in the case law of the UPC. See also the recent decision of LD Paris [363/2024] dealing only with equivalents.
A clear conclusion can be drawn from T 202723 : if the applicant or patent owner wishes to rely on a narrow interpretation based on information in the description or in dependent claims, in order to distinguish over the prior art, it is up to them to amend the independent claims accordingly.
T 697/22, the referring decision to G 1/25, provides in Reason 15.1.2 a potent policy argument supporting what it calls the « second line of case law « finding there is no basis in the EPC for the requirement to adapt the description to the amended claims. The Board reasons that the decisions under the second line of case law share the underlying idea that the inconsistent description and the possible consequences are the applicant’s responsibility.
This « underlying idea » is implemented at the 71(3) stage in GL C-V 1.1 which states that the text the ED intends to use for the grant « may include amendments and corrections that the examining division made on its own initiative and can reasonably expect the applicant to accept ». If in doubt as to the applicant’s agreement, the ED should contact them, and the applicant’s agreement should be recorded.
The applicant’s responsibility also implies that the EPO has no responsibility, hence no liability, even if a patent is revoked or held invalid as a result of its own action, e.g. if an amendment entered by the ED at the 71(3) stage is in violation of Art 123(2).
The « underlying idea » emphasised by T 797/22 can justify the applicant’s disagreement with description amendments which delete or disclaim embodiments no longer covered by the amended claims. Nonetheless, it is not applicable to all kinds of inconsistencies and all stages of examination. If an inconsistency casts doubt as to the meaning of a claim term (see 2024 version of GL F-IV 4.3, in which « casts doubt » in the present tense has replaced the conditional « would cast doubt » of the 2021 version), this violates the clarity condition of Art 84 and would warrant a refusal if it is not addressed by the applicant.
In respect of the « patent being its own dictionary », I agree that this is acceptable but subject to restrictive conditions. First, the « underlying idea » of applicant’s responsibility implies the applicant’s discretion as to the drafting of the original application. The applicant must be free to include definitions in the description which depart from the ordinary meaning of claim terms, much like a list of definitions in an agreement. But the definitions should be worded so as to express the applicant’s clear intent that the claim terms be interpreted in accordance with the definitions of the description, and be consistent with the description as a whole. On this topic, see Rose Hughes’ post of 10 June 2025 on IPKat and her comments on the CAFC decision Alnylam v. Moderna dated 4 June 2025 setting out the restrictive conditions for a patent to be its own « lexicographer ».
Dear Thomas,
Thank you for your excellent write-up of T 2027/23.
While I agree with you on most points, on the point of the patent being its own dictionary I still have to disagree with you while agreeing with BJ. Although, I consider that we have reached an consensus to the point where the disagreement in mainly on choice of words rather than on what “the patent can/cant be its own dictionary” actually means.
You bring up the point of “A patent could only be its own dictionary if the dictionary definition stays in the patent, or have you seen definitions in a dictionary vanish?” to which I would like to reply: As you are well aware in the days of printed dictionaries definitions and/or terms constantly “vanished” once a certain term fades common use in its language. Similarly, definitions of terms regularly change over time once science progresses and better descriptions are available.
To be more clear, you have so far not been able to convince me that a patent can be its own dictionary for terms that have no commonly accepted and clear definition for the skilled person.
As a practical example, let’s consider a claim directed to a technology platform for increasing efficacy of proteins. Let’s consider the claim is directed to a protein covalently conjugated to a specific signal molecule moiety providing for an inventive effect. Now, let us further consider that a competitor wants to sell an enfuvirtide (HIV medication comprising 36 amino acids) conjugate comprising a moiety falling under the signal moiety as claimed.
To determine whether the competitor falls under the scope of the claim the skilled person first has to determine whether enfuvirtide is a protein. While everybody knows – in principle – what a “protein” is, this term has no clear-cut limits regarding size length or the presence of glycosylation. It appears that there is a technical consensus that a peptide with more than 20-50 amino acids is called a “protein”, while anything below that is referred to as “peptide” or “polypeptide” (further problems relating to whether a “protein” can be glycosylated will be omitted for simplicity). Therefore, it is not unambiguously clear from the common technical understanding of the skilled person whether enfuvirtide is still considered a peptide or falls under the term “protein”, i.e. whether the competitor would fall under the scope of the claim or not (disregarding for now the doctrine of equivalent infringement).
In cases such as this, which are more common in the fields of chemistry and biochemistry as one would expect, it can be vital if an applicant has used the description to supply the skilled person consulting it with an explanatory aid that provides clear-cut borders, e.g. a “protein” as referred to herein is a polypeptide compound comprising more than 50 amino acid units”. If present, such a definitions provides clarity to the public as to the actual scope of the claim. In such a case, the patent can be its own dictionary.
To underscore my point, I point to the much-discussed Agfa/Gucci case, also cited in T 2027/23. To the best of my knowledge, there was no commonly accepted and clear definition for the disputed term “achromatic” base coat. While the proprietor tried to argue for infringement of an “ivory” base coat based on a cobbled together technical definition of the term “white” as “achromatic”, the board – correctly – cut through the proprietors nonsense and bound the proprietor to a definition of “achromatic” the proprietor himself provided in the application as originally filed. Therein, the CFI used the patent as its own dictionary for a term without a clear and common understanding in the art and denied infringement, but at the same time had to uphold patent validity (as clear borders between “ivory” prior art and the “achromatic” claimed matter had been established).
In my mind, would the CFI have disregarded the patent as its own dictionary for a term with no clear-cut technical borders or understanding in the art, it would have acted contrary to legal certainty for the public, as it was not clear from the claim wording alone where the borders of “achromatic” had to be drawn. As every claim requires interpretation and the description has to be consulted with interpreting, the public is not faced with an additional burden in considering definitions within the patent. It is another matter whether such a definition should have been included within the term from the onset or at the latest prior to grant. In my personal opinion, including every definition for every “blurred” technical term leads to bloated and incomprehensible claims in many technical fields. I find their inclusion as “explanatory aids” in the description much more suited in yielding clear and concise claims.
Where we all appear to agree is that a patent cannot be its own dictionary for terms that do have a commonly understood technical meaning or for implying a limitation to terms used in a claim based on the context within the application or the problem to be solved. Any one of T 1561/23, T 2027/23, T 1069/23, T 1465/23, T 1846/23, T 400/23, T 1164/23 and T 1999/23 released after G 1/24 made that clear.
Lastly, I remain undecided on the position whether “unfit” definitions which either try to define terms with their own clear-cut and established technical meaning, or worse, try to limit or broaden technical terms in the definition creating “Angora cats”, should be removed from the patent or not.
On the one hand, the primacy of the claims has clearly been confirmed. It is the position of the vast majority of patent offices around the world, that the public is more than capable of determining whether a part of a patent can be ignored in construing a claim or not.
On the other hand, I do tend to agree with Thomas that in order to increase legal certainty and avoid any potential mistakes in infringement courts the patent should be adapted to the claimed matter. However, I feel that when one accepts that is the explicit goal of the EPO to avoid any possible mistakes and/or confusion in construing granted and claimed matter, the current practice of including a “Further disclosed herein is” feels like a sorry bandaid. If taking this position and being consequent I feel that any matter described in the description that does not fall within the scope of the claim or serves to explain such matter should – like any definition conflicting with the skilled persons understanding of a term – be excised from the granted patent.
Personally, I have had to analyze 1000+ page patents wherein a good 800-900 pages referred to irrelevant and/or only adjacent matter. I feel conflicted with calling such patent descriptions “clear and concise” even though the granted and claimed matter is linguistically correctly marked. Infringement courts should be more than capable of establishing equivalent infringement by referring to the application as filed and their common sense. As such, I am still torn whether to adopt the world-wide practice of leaving a granted patent as-is, or whether to subject any granted patent to the “red pencil” of cutting it down to the granted scope. What I can say is that the current EPO practice of merely linguistically excising non-claimed matter does – in my opinion – fails to achieve what it sets out to do. I do see, though, that the second option of forcing excision of non-claimed matter will have both patentees and the representatives up in arms.
I look forward to hearing your further opinion on the matter.
Best regards, L.T.
L.T., you wrote ‘On the other hand, I do tend to agree with Thomas that in order to increase legal certainty and avoid any potential mistakes in infringement courts the patent should be adapted to the claimed matter. ‘
This is a noble principle, but I wonder whether it is neccessary. Are there really that many cases where the proprietor succeeded in getting a wider scope accepted than the plain (to the skilled person) meaning ? This is an honest question, because I don’t know of any (except the pemetrexed case of course).
And similarly, are there that many cases where the proprietor managed to get a narrower scope accepted in order to avoid N/IS objections?
I know that proprietors TRY to broaden or narrow the scope based on the description, but I see that the courts (and ODs/BoA) generally are clever enough not to let them succeed.
Further, you wrote: ‘I do see, though, that the second option of forcing excision of non-claimed matter will have both patentees and the representatives up in arms.’
‘ Up in arms’ is, in my case, a bit exaggerated. I am opposed to it, basically for reasons of a ridiculous cost/benefit ratio (ALL patents have to be conformed, just so that the risk of a judicial error for 1 in 10000 or 1 in 100000 is reduced), not for disingeneous Angora car reasons.
Also, I hate to state that an example is not according to the invention. Firstly, because an examiner in another country (eg US) might read it and attach a meaning to such a statement that it doesn’t have.
Secondly because it may not be my true opinion. It may well be that I accept certain claims because I realise that they are the best that the examiner will allow, but that doesn’t mean that I agree with the examiner, I have just given up fighting.
In other words: An applicant accepting a certain allowable claims set is often the result of practical/financial considerations, a business deal so to speak, whereas an applicant stating that something is not according to the invention is a fundamental matter touching on the fundamental freedom of the applicant to accept the outcome but nevertheless disagree with the position of the other side.
Does the EPO then need to humiliate the applicant by forcing them to state something that they honestly don’t agree to, when the claims already make abundantly clear to the whole world what the scope of protection is?