Saturday, 12 June 2010

Interpretative spotlight : “comprising”

[29] The meaning of the word “comprising” is generally interpreted as encompassing all the specifically mentioned features as well as optional, additional unspecified ones, whereas the term “consisting of” only includes those features as specified in the claim.

Therefore, “comprising” includes as a limiting case the composition specified by “consisting of”. The board is satisfied, therefore, that the amendment from the former into the latter does not extend beyond the content of the application as originally filed (see also e.g. decisions T 457/98 [2.1.1] and T 425/98 [3.1]).

Admittedly, this finding is not surprising, but it is always good to have a decision where it is clearly stated.

To read the whole decision (T 997/06), please click here.


Anonymous said...

This is surprising to me. Obviously "comprising" includes "consisting of", but that does not mean that the amendment from "comprising" to "consisting of" does not add matter. The term "metal" includes "iron", and everybody will agree that an amendment from "metal" to "iron" adds matter.

The question to be answered is whether the application as filed specifically disclosed "consisting of", either explicitly or implicitly. I do not believe that such a question can be answered purely linguistically; the Board should have examined the technical disclosure. The Kinkeldey Board appears to attribute to the skilled person the abilities of a patent attorney rather than the abilities of a technical expert. (Another example is T 1107/06, in which the Kinkeldey Board reasons that "[generic], e.g. [specific]" can ALWAYS be amended to "[generic] not being [specific]".)

T 812/08, r. 1.1.1 seeks actual support in the application as filed for an amendment of "consisting essentially of" to "consisting of" and for an amendment of "comprising" to "consisting of". Idem for T 197/06, r. 2.1, for T 1082/05, r. 2.1, for T 985/03, r. 2.1, for T 272/98, r., and for T 916/95, r. 2.1.

In T 1063/07, r. 2.1-2.4, the Board concludes that the amendment from "comprising" to "consisting of" is NOT allowable. The Board mentions the argument of "implicit support", but dismisses it as a mere assertion.

Oliver G. Randl said...

Thank you for this very interesting comment.

It is true that from a hardliner point of view (and the EPO certainly qualifies as hardliner when it comes to A 123) one could find an amendment from 'comprising' to 'consisting' unallowable if not explicitly or at least implicitly disclosed.

Having said this, I think your analogy (metal/iron) is not quite to the point. There is some sort of unwritten convention in the patent world (I guess it originated in the U.S. case law) that when you write "comprising" in a claim this is a shortcut for writing "including the following elements but not necessarily excluding others", whereas "consisting" is to be understood as "including the following elements to the exclusion of additional unspecified elements". If this convention is accepted, then the amendment from "comprising" to "consisting" should be acceptable, because then the term "comprising" implicitly also discloses what is referred to by "consisting".

As you have cited decisions that do not accept this convention, it might be wise to add a sentence on the meaning of "comprising" in the specification.

Anonymous said...

I am aware of and accept this convention, but imo it only implies that "comprising" covers "consisting of". It is also beyond question that "metal" covers "iron". I don't really see a difference here.

In my view "a composition comprising ..." does not imply a collection of separately claimed alternative compositions, but a single claimed collection of compositions. A claim to "composition A or B" can be amended to "composition A", but a claim to "composition with property X" claims the whole family of compositions with property X and cannot be amended to a particular element of that family (without a basis in the application as filed).

In terms of the novelty test: "composition comprising A and B" does not take away novelty of "composition consisting of A and B".

Anonymous said...

The novelty test fails for A123(2) when you amend into a more restricted scope, ie from "comprising" to "consisting".
The ultimate test for amendments is whether they are "directly and unambiguously" derivable from the original disclosure and this depends solely on the facts of the case. It is therefore not surprising to come to different conclusions depending on how one assesses the interpretation by the skilled person of the original disclosure.

Anonymous said...

I know that the novelty test is not a conclusive test for 123(2), but it works perfectly fine in one way. IF the novelty test fails (i.e. the amended claim is new over the application as filed), THEN the amendment adds matter.

Since "composition comprising A and B" does not take away the novelty of "composition consisting of A and B", an amendment from "comprising" to "consisting of" needs further support in the application as filed.

"It is therefore not surprising to come to different conclusions depending on how one assesses the interpretation by the skilled person of the original disclosure."
I agree, but my point is that there needs to be an assessment of the original disclosure. T 997/06 does not assess the original disclosure.

If the linguistic argument used in T 997/06 were correct, it would ALWAYS be allowable to amend "comprising" into "consisting of", no matter the technical disclosure of the application as filed.

Anonymous said...

Hmmm, if the anonymous of 15:53 meant that an amendment from "comprising" to "consisting of" fails the novelty test and is therefore not allowable under Art. 123(2) (unless a basis can be found in some other part of the application as filed, which of course fully depends on the facts of the case), then we are actually in agreement.

Anonymous said...

My reliance on the "directly and unambiguosly derivable" test was to give more weight to the original disclosure.
Often the invention is described as "consisting" of some features while the claim is drafted as "comprising" such features to allow for a larger scope of protection. In such cases the amendment to consisting would be permissible. In some other circumstances it may well be that the invention has been described as comprising the features and a restriction to consisting may be an extension. It depends on the facts and a general linguistic rule is hard to discern.
On the other hand, I would interpret "comprising" as meaning "consisting of at least". This could be interpreted as meaning, in some kind of weird "feature space", the point defined by the "consisting" features and a region surrounding it. However, I don't know if this picture helps in solving this issue.

Anonymous said...

We agree that more weight should be given to the original disclosure.

Regarding your last paragraph, I suppose what you mean is that a claim with "comprising" could be interpreted as an OR-claim: "consisting of" OR "consisting of plus extra features". This might be what T 997/06 means by "limiting case". If "comprising" should be interpreted as specifically disclosing the limiting case (i.e. consisting of) as one alternative, then the amendment to "consisting of" simply limits the claim to one disclosed alternative, hence does not add matter. I personally don't like this approach, though.

What about the claim: "table having four legs". I think we can agree that this claim covers all tables having at least four legs. Is it possible to amend this claim to "table having exactly four legs" (without any extra support in the application as filed)?

I say this is not possible. The amended claim is more specific than the original claim. It has an extra feature not present in the original claim: exactly four legs.

The Board of T 997/06 would probably say this is possible. A table having exactly four legs is a "limiting case" of a table having four legs, hence specifically disclosed by the original claim. However, by that reasoning, the original claim is equivalent to "table having exactly four legs OR a table having more than four legs", and this is equivalent to "table having exactly four legs OR a table having five legs". The original claim can therefore be amended to "table having five legs". Now, by the same reasoning, we can amend to "table having exactly five legs". By induction, we can amend to "table having exactly N legs" for any integer value of N >= 4. This would mean that "four legs" or the equivalent "at least four legs" specifically disclose all possibilities "(exactly) N legs" for N >= 4. In my view, that is nonsense. The "limiting case" reasoning is therefore wrong.

Anonymous said...

I think that you made a quntum leap when going from "a table having more than four legs" to "a table having five legs". Both statements are not equivalent, since more than four does not directly and unambiguosly disclose five. However, the statement "a table having exactly four legs" is directly and unambiguosly derivable from "a table comprising four legs". I think in this point we do indeed differ :-)

A similar issue arises with parameter ranges. Eg the statement "a concentration of an element A greater than or equal to x%" can it be amended to state "a concentration of an element A greater than x%" or to the statement "a concentration of an element A equal to x%"? Is "greater than or equal to" an alternative of two different embodiments or is it a single property that cannot be divided? Is there any difference if x is zero (ie element A is present or not)? I would greatly appreciate your views on this issue.

Anonymous said...

If I understand you correctly, then your position is that "a table having five legs" discloses both "a table having exactly five legs" and "a table having more than five legs", but that "a table having more than four legs" does NOT specifically disclose "a table having exactly five legs", so that it cannot be amended to "a table having five legs" (because that WOULD specifically disclose "exactly five legs").

I agree that your position also avoids the nonsensical conclusion I referred to above. So your position makes sense.

However, I still prefer my own position ;).
The skilled person knows that a table with more than four legs is a table with at least five legs and vice versa. It seems undesirable if the following claims had subtle differences in meaning:
1. a table having more than four legs.
2. a table having at least five legs.
3. a table having five legs.
4. a table having five or more legs.

If the applicant should wish to have an OR-claim, he could phrase the claim as:
a table having exactly five legs or having more than five legs.

I think this is also reasonable for the following reasons. The applicant will only want to amend from "five legs" to "exactly five legs" if he has something to gain by that. The amendment would probably be intended to overcome an Art. 56 or 83 objection. The embodiment "exactly five legs" will somehow be technically different from "five legs". If that is the case, it seems only fair to require the application as filed to contain a clear indication that the applicant knew that "exactly five" was somehow special.

In my view "a plurality of legs", "two legs", "more than one leg", "two or more legs" are interchangeable. This implies that "two or more" cannot be amended to either "more than two" or "three".

So in my view, "greater than or equal to" is a single property that cannot be divided. It is the expression ">=" which, put into words, simply happens to have the word "or" in it. Otherwise, the expression ">=" could not be amended to "greater than or equal to" without subtly adding matter.

Of course this is only my view and everybody is free to differ ;).

Anonymous said...

Btw, I just realised that I'm a bit quick in concluding that if "table having four legs" can be amended into "table having exactly four legs", this would imply that "table having four legs" can also be amended into "table having more than four legs". Although this would follow from T 1107/06, point 46, I don't agree with that decision.

Since the Boards in T 997/06 and T 1107/06 are the same, or at least have the same chair(wo)man, the last paragraph of my post of 16 june 2010, 02:52 still applies, though.

Anonymous said...

Thank your for your comments :-)

I suppose the only way to get a final (and hopefully convincing)answer would be that the question be referred to the EBA. However, I don't see this as very probable.

Anonymous said...

I just noted point IV of decision T 2017/07:
"On 18 August 2008, the Board issued a communication indicating (...) that no basis could be found in the application as filed for replacing the original feature "comprising" defining the claimed composition by the fresh feature "consisting of" in the then pending auxiliary requests 1 and 2 (Article 123(2) EPC)."