By Pierre Véron
is a hot topic in the world of patents at the beginning of the 21st
century. In Europe, the Protocol on the Interpretation of Article 69 E.P.C.
was revised in November 2001 to include an Article 2 making an express
reference to the doctrine of equivalents as follows: “For the purposes of
determining the extent of the protection conferred by a European patent, due
account shall be taken of any element which is equivalent to an element
specified in the claims”.
However, the Member States have failed to agree on a definition of
equivalence. As regards the
The doctrine of equivalents, as it results from French case law, may be summarised as follows: a means whose structure does not literally reproduce the claims of the patent is considered equivalent to the patented means when it performs the same function (in the same way ?) as the patented means and achieves similar results2.
This definition has some flavour of that given by the famous Graver Tank3 judgment of the US Supreme Court: indeed, the French doctrine of equivalents has much in common with what is generally accepted in other countries. However, a closer analysis of French case law shows a much wider interpretation of equivalence.
The French Courts have repeatedly decided that a means having a different structure is equivalent when it performs the same function, to achieve the same result, as the patented means. It is however questionable whether the “way” the means works must also be taken into consideration.
Two means performing the same function
The identity of the functions is the key element of the French doctrine of equivalents. Two means perform the same function when they produce the same primary technical effect4 The “primary technical effect” can be defined as the effect directly and immediately produced by the implementation of the means. It must not be confused with the result, which is the immaterial advantage provided by the means5.
It is not always easy to catch the difference between function and result. The French supreme court6 (Cour de Cassation) decided upon this question in a case involving a patent for a process of obtention of doxycycline through hydrogenation of methacycline, with a modified catalyst. The alleged infringer argued that the function of the process was to reduce, through hydrogenation, methacycline into doxycycline, which was not novel. The patentee argued that the function of the process was rather to obtain selectivity and stereospecificity at almost 100 % of the required epimer (α), which was novel. The Cour de Cassation
affirmed the judgment of the Court of Appeal which had decided that the obtention of a very high yield of epimer α was a novel function, not a result. As the process carried out by the defendant involved not only the known result of reduction through hydrogenation of methacycline into doxycycline, but also the novel function of selectivity and stereospecificity of epimer α, the defendant was found infringing the patent.
The result obtained by the means at issue must be similar, but not necessarily identical. As a consequence, the result may be of a different quality, may it be higher or poorer than the result obtained by the patented means7. In the same way, the result may be of a different degree than the one obtained by the patented means8.
This does not mean that protection can be granted to a result. If French law admits the protection of the function itself9, it does not admit, though, the protection of the result10 per se. Thus, a means which enables to obtain the same result than a patented means, but which differs not only in its structure but also in its function, will not be held equivalent11.
It is not perfectly clear whether the way in which the means performs the same function should be taken into consideration. Several authors do not mention the “way” test and the readers of their books could think that the French doctrine of equivalents is only twofold (function, result) as opposed to the threefold U.S. Graver Tank test
Patent World November 2001
(function, way, result). However, several judgments of the 1990s do mention the way in which the accused device achieves the same result12. And when reading many other judgments, one can realize that, although not explicitly mentioned, the way the means works is also taken into consideration. It may well be that the French courts give more weight to this factor in a near future.
A wider interpretation of equivalence
Five main differences exist between equivalence in
· The theory of the so-called “general means”
· “Improving is infringing”
· The intention of the patentee
· The prosecution history file
The theory of the so-called “general means”
According to the wording of French case law, the doctrine of equivalents is applied only when the patentee can claim for a wide scope for his patent.
When the scope of the claim appears to be limited to a specific embodiment, because the function of the claimed means is not novel, the Courts do not resort to the doctrine of equivalents. They treat minor or immaterial variants as literal infringements without applying the doctrine of equivalents.
The situation is different when the scope of the claim is broader. French law has the peculiarity to grant protection to the function, even if it is not claimed, through the theory of the so-called “general means” (“moyen général”). According to this theory, when a claim is directed to a specific means (“moyen particulier”), the scope of this claim is determined in a different way when the function of this means is known from the prior art, on the one hand, and when the function is novel, on the other hand. When the means function is known, the scope will be limited to the claimed structure (and to minor variants). On the contrary, when the means function is
novel, the scope of the claim will extend to any other structure performing the same function: the French Courts say, in such a case, that the claim covers a “general means”, which is nothing else than the function performed by the specific means.
Thus, if a function, although not claimed, is novel, it will be
protected even though the claim is directed to a particular means which
performs this function. Therefore, a second means, of a different structure
from the one of the claimed means, will be held infringing by equivalence if
it performs the function of the patented means. The Court of Appeals of
Obviously, if the function is novel and claimed, any means which performs this function will be held infringing, whatever its structure may be: in this case, the infringement is literal, and not carried out by equivalence, since the patent thus covers the function itself.
On the contrary, if only the structure is novel, the function it performs being known, then the doctrine of equivalents does not apply14. In such a case, the essential part of the invention is only its particular structure, not its function. Therefore, only a means reproducing the patented structure exactly or with minor variants may be held infringing15.
In Germany and in the United Kingdom, the test of equivalence, as applied by the current case law, includes a condition which tends to check whether the accused variant was
obvious for the person skilled in the art, when reading the claim. Under French law, such a condition is not required for a means to be held equivalent to a patented means. Obviousness is not a criterion of infringement, but only a grounds for invalidity: it is taken into account only when the inventive step is assessed. Therefore, a means will not be held equivalent on the grounds of its obviousness, nor held not equivalent because it was not obvious: this is so undisputed that, as of today, such defence has never been raised and accordingly no case has ever been decided in this regard16.
As a consequence, the question
of when the obviousness of the alleged infringement is appreciated is not of
topical interest in
One of the peculiarities of the French doctrine of equivalents is to consider that improving a patented means may be an infringement. This solution is summed up in the saying: “Improving is infringing” 17.
Under French law, a means improving a prior patented means infringes it when it reproduces the essential structure and the function of the prior invention. Of course, any improvement does not constitute an infringement. An improvement which would implement different means, with a different function, in order to obtain a better result, would not constitute an infringement. But an improvement may be held equivalent, although its structure is modified compared to the patented invention, if:
· it reproduces the function of the patented invention,
· it allows to obtain a similar result.
In this respect, reference can
be made to the decision of the Court of first instance of
Patent World November 2001
Such an improvement, even if it is patentable, will thus constitute an infringement of the prior invention. This is a consequence of the fact that obviousness from the viewpoint of the person skilled in the art is not required for equivalence to be admitted.
Under English law, the intention of the patentee seems particularly important for the appreciation of infringement. In particular, the Courts check whether the patentee intended to cover the variant reproduced by the allegedly infringing means or if he intended to strictly limit the claim to its wording19.
As regards French law, the possible intentions of the patentee are not taken into account20. Equivalence is appreciated in a purely objective manner, without any search for the intention of the patentee. This is not to say that the variants excluded by the patent or going against the teachings of the patent can be held infringing. When the patent explicitly and unambiguously excludes a variant from its scope, this variant, if exploited by a third party, will not be held equivalent to the patented invention21. In the same way, when the variant directly and unambiguously goes against the teachings of the patent, it will not be held equivalent.
Generally speaking, estoppel does not exist as such under French law. Thus, a party may theoretically contradict itself at different stages of the proceedings, without running any risk from a strictly legal point of view22. However, some French decisions do take into account the foreign proceedings which had the same subject matter than the current proceedings23, and a party often seeks to use as an argument the positions taken by the other party at this period of time. If the judges cannot base their decision only on the prior declarations of a party, they do take into account the declarations made during the prosecution of the patent at issue, as well as, if appropriate, at the prior stages of the judicial proceedings24.
However French case law is far from the strictness of the “Festo” decision: estoppel is only a general procedure principle, the application conditions of which are not yet exactly defined, and which remains largely unfamiliar to French law.
Therefore, it is not sure that French Courts will always apply this doctrine so broadly. Changes could occur in the near future.
LAMY, VÉRON, RIBEYRE
The author wishes to thank Isabelle Romet and Olivier Moussa for assisting with the preparation of this article.
2. Cour de Cassation, Chambre Commerciale, October 26, 1993, PIBD 1994, No. 558,III,21 - Cour d’Appel of Paris, October 30, 1996, PIBD 1997, No. 626,III,78
Graver Tank & Mfg. Co. v. Linde Air Products
4. Cour d’Appel of Paris, September 11, 1996, PIBD 1996, No. 621, III, 577
5. "Le nouveau droit français des brevets d'invention", P. Mathély, p.60
6. Cour de Cassation, Chambre Commerciale, January 26, 1993, Dossiers Brevets 1993, I, 2
7. "Droit de la propriété industrielle", A. Chavanne and J.-J. Burst, Dalloz, 1998, No. 406
8. Cour d'Appel of Paris, March 15, 1996, PIBD 1996, No. 613 , III, 337
9. Cour d'Appel of Paris, June 16, 1987, Annales de la Propriété Industrielle, 1987, 96 – Tribunal de Grande Instance of Paris, June 28, 1996, PIBD 1996, No. 620, III, 547
10. Cour d'Appel of Paris, June 16, 1987, Annales de la Propriété Industrielle, 1987, 96 – Tribunal de Grande Instance of Paris, June 28, 1996, PIBD 1996, No. 620, III, 547
11. Cour d’Appel of
12. Tribunal de Grande Instance of Paris, April 8, 1998, Alcan v. Schuco, JURINPI No. 98B0105 - Cour d'Appel of Paris, May 28, 1999, PIBD 1999, No. 687, III, 501
13. Cour d'Appel of
14. Cour d'Appel of
15. Cour de Cassation, Chambre Commerciale, June 18, 1996, PIBD 1996, No. 619, III, 511- Cour d'Appel of Paris, October 30, 1996, PIBD 1997, No. 626, III, 78
16. The Cour de Cassation has however decided twice (July 4, 1984, Annales de la Propriété Industrielle, 1985, 210 – October 26, 1993, Annales de la Propriété Industrielle, 1993, 94) that, when an alleged infringer owns a patent for the accused device, the lower Courts must check, before holding that this device infringes the prior patent of the claimant, whether the content of the defendant's patent is identical or different from the content of the plaintiff's patent. Obviously, if the second patent is identical to the plaintiff's patent, it is invalid and the defendant is held infringing. On the other hand, when the second patent, which covers the accused device, is different from the plaintiff's patent, the lower Courts must check whether the defendant's device constitute an infringement by equivalence. The proper reading of these judgments of the French supreme court is therefore to oblige lower Courts to check whether the accused device differs substantially from or is equivalent to the plaintiff's patent; by no way the French supreme court has invited lower Courts to check whether the defendant's patent regarding the accused device was or not obvious over the plaintiff's patent.
17. Cour d'Appel of Paris, March 16, 1977, Dossiers Brevets 1978, II, 2
18. Tribunal de Grande Instance of Paris, December 20, 1990, PIBD 1991, No. 502, III, 367
19. Catnic Components v. Hill & Smith,  RPC 183
20. Tribunal de Grande Instance of the Seine, March 29, 1965, cited by M. de Haas, in Aspects actuels de la contrefaçon, Litec, 1975, p.76 – Tribunal de Grande Instance of Paris, October 15, 1997, PIBD 1998, No. 649, III, 133
21. Tribunal de Grande Instance of the Seine, March 29, 1965, cited by M. de Haas, in Aspects actuels de la contrefaçon, Litec, 1975, p.76 – Tribunal de Grande Instance of Paris, October 15, 1997, PIBD 1998, No. 649, III, 133
Patent World November 2001
22. Cour d'Appel of Paris, April 24, 1998, Revue du Droit de la Propriété Intellectuelle, 1998, No. 86, 38
23. Cour d'Appel of Paris, November 19, 1999, PIBD 2000, No. 700, III, 305
24. Cour de Cassation, 3° Ch. Civ., December 10, 1985, Bull. civ. III, No. 166 - Cour de Cassation, 2° Ch. Civ., April 16, 1986, Bull. civ. II, No. 53 - Cour de Cassation, 1° Ch. Civ., March 6, 1990, Bull. civ. 1990, No. 59 - Cour de Cassation, Chambre Criminelle, June 11, 1996, Dalloz 1997, Jurisprudence, 576
Patent World November 2001