Jonathan DC Turner




Intellectual Property
Competition Law
IT Law
Curriculum Vitae


Three Stone Chambers

3 Stone Buildings

Lincoln's Inn 

London WC2A 3XL, UK

Tel: +44 (0)20 7242 4937

Important: This is a copy of an article published in the European Intellectual Property Review, NOVEMBER 1999 and the CIPA JOURNAL, AUGUST 1999. It is not guaranteed to be up-to-date when downloaded. Copyright is reserved by the author and moral rights have been asserted. The article may be copied and distributed provided this done free of charge and in relation in the whole article, including details of authorship and this notice, without any amendments.

Recent judgments of English courts have continued to apply the “purposive construction” approach propounded by Lord Diplock in Catnic v Hill & Smith[1] and elucidated by Hoffman J[2] in Improver v Remington (the depilator case)[3] to determine the extent of protection conferred by patents[4].  In doing so, they have rejected the view of the Court of Appeal in PLG Research v Ardon[5] that discussion of Catnic should now be left to legal historians, since it was decided under the law prior to the implementation of the European Patent Convention in the United Kingdom. Despite the impressive show of unity in recent judgments, the point remains acutely controversial[6]. This article contends that the Catnic/ Improver approach is wrong and incompatible with the European Patent Convention.

1. Purposive construction of patent claims requires knowledge of their purpose.

Lord Diplock’s central premise is that patent claims must be interpreted purposively through the eyes of the person skilled in the art to whom the patent is addressed[7].  However, the person skilled in the art does not know what is the purpose of the claims until he is informed by a person with knowledge of patent law.  For the person skilled in the art but untutored in patent law[8], there is merely a heading “claims” followed by a shorter or longer list which appears to be a rather repetitive and often convoluted summary of what has just been described[9]. This list does not help him carry out the invention; it appears to him to be there for some legal or other non-technical reason, as of course it is.

Thus the skilled person can only interpret the claims purposively when he has been told what is their purpose in patent law; and when he has been told, his interpretation will depend on what he has been told.  If he has been told (in the British tradition) that the claims are there to define the outer limits of the patentee’s exclusive rights in the invention so that competitors know where they stand, he will interpret them relatively strictly.  If, on the other hand, he has been told (in the German tradition) that the claims are guidelines identifying the principal features of the invention which have been described in more detail in the description and drawings, he will interpret the claims relatively liberally.

It is circular for patent law to refer the interpretation of patent claims to the person skilled in the art, as was rightly recognised by the Court of Appeal of The Hague in the Dutch Improver case[10]. The skilled person has to refer back to the lawyer to find out what the claims are for.  This loop must be cut.  The first question is: “What is the purpose of the claims?” 

    2. Catnic is internally inconsistent with regard to the purpose of the claims.

Lord Diplock’s analysis starts with the observation that a patent specification is a statement by the patentee by which he informs those skilled in the art what he claims to be the essential features of the new product or process for which the patent grants him a monopoly[11]. This confuses the description with the claims and conflates two different functions of the specification: on the one hand, that of disclosing the invention to those skilled in the art; on the other, that of informing lawyers, managers assisted by lawyers and the courts what monopoly has been granted.  However, to the extent that it refers to the claims, Lord Diplock’s starting point seems to treat them as a “signpost” identifying the features of the invention believed to be novel at the time (as in the German tradition), rather than as a “fence” defining the outer boundary of the monopoly (as in the English tradition).

Lord Diplock then sets out an algorithm for determining whether a claim should be interpreted as extending to a variant from its strict wording[12]. The full implications of this algorithm are unfolded by Hoffman J in Improver[13] in terms which presuppose that the purpose of the claims is to define the outer boundary of the monopoly.  According to this algorithm, a variant is “outside” the monopoly claimed unless the variant makes no material difference to the way the invention works, it is obvious that the variant makes no material difference to the way the invention works, and the skilled person would understand that the patentee intended the claim to extend to the variant. 

The presence of the assumption that the claim is there to define the boundary of the monopoly is borne out by Hoffman J’s reasoning that the person skilled in the art but ignorant of the law (to whom patent claims are erroneously supposed to be addressed – see 1 above) would be entitled to suppose that there was some reason to do with the patent office for the patentee limiting its claim by reference to a helical spring, and that accordingly it was not apparent that the claim was intended to extend to the use of an equivalent of a helical spring[14].  This reasoning would not be pertinent if the claims were there merely to identify the essential features of the invention for the information of the person skilled in the art.

The inconsistency between the starting point and the algorithm reflects the disparate foundations of Lord Diplock’s approach.  Before Catnic English law treated the claims as defining the limits of the monopoly, but acknowledged the possibility of infringement of the “pith and marrow” by a colourable imitation which evaded the words of the claims[15]. Lord Reid observed that in this regard “there is some illogicality, but our law has always preferred good sense to strict logic”[16].  In combining these two conflicting principles into a single approach of “purposive construction”, Lord Diplock glossed over the differences between them and obscured the fundamental question of the purpose of the claims.

           3. The purpose of the claims presupposed in the Catnic/Improver algorithm is inconsistent with the Protocol on the interpretation Article 69 of the European Patent Convention. 

Article 69 of the European Patent Convention provides that the extent of protection of a European patent is to be determined by the terms (or “inhalt” or “teneur” in the equally authentic[17] German and French texts) of the claims, but that the description and drawings shall be used to interpret the claims. The Protocol (an integral part of the Convention[18]) requires article 69 to be interpreted neither in the sense that the extent of protection is determined by the strict, literal meaning of the wording used in the claims, nor in the sense that the claims serve only as a guideline, but in an intermediate way which combines fair protection for the patentee with a reasonable degree of certainty for third parties.

However, according to the Catnic/Improver algorithm, the claim is only infringed if it is obvious that the variant makes no material difference and that the patentee intended to cover it.  If this approach were followed rigorously, it would achieve complete certainty at the expense of fair protection for the patentee, since a patent would only be infringed by a variant from a strict meaning of the words of the claim in the rare case where it was clear that the patentee intended to claim the variant. 

If, on the other hand, the Catnic/Improver algorithm is not followed rigorously and, where the intention is not clear, the court neither assumes that equivalents are included nor that they are excluded[19], then the court has to make its best estimate of the skilled person’s best estimate of what the patentee intended in relation to a situation which it never contemplated. The exercise becomes a hypothetical speculation which multiplies the degree of uncertainty, instead of providing the “reasonable degree of certainty” sought by the Protocol[20]. 

The hypothetical nature of the exercise is demonstrated by the case of an alleged infringement which is itself inventive.  According to Improver the court has to consider what the skilled person would suppose the patentee had intended if it had known of the inventive infringement[21].  But if this is right, in considering validity (including sufficiency) the court should consider what the skilled person would suppose the patentee would have intended had it known, not just of the alleged infringement, but of all possible variants (including those not yet invented even at the time of consideration), and then what alternative generalisations the patentee would have adopted to define the scope of the monopoly instead of the words actually used in the claims[22].

The central problem is that the Catnic/Improver algorithm presupposes that the purpose of the claims is to define the boundary of the monopoly. This is inherently inconsistent with a degree of flexibility to cater for a patentee’s failure to anticipate all of the ways in which its invention may be used. Yet under the European Patent Convention, interpretation of the claims is the only way in which such flexibility can be provided. Article 69 precludes a doctrine of equivalents along the lines of US law or the pre-Catnic English approach, according to which a claim means X but can be infringed in certain circumstances by an equivalent or colourable imitation of X[23].

It is in this context that the Protocol on the interpretation of Article 69 requires the courts to find a way between the traditional German approach of treating the claims as guidelines and the traditional British approach of treating them as defining the boundary of the monopoly claimed.  It is true that the Protocol does not include the Catnic algorithm in the approaches which it expressly rejects.  The draftsmen of the Protocol were not gifted with foresight of Lord Diplock’s reinterpretation of the traditional British approach[24].  They referred instead to the way in which the difference between the traditional German and British conceptions of the functions of the claims had previously manifested itself, namely the relationship between the claims on the one hand and the description and drawings on the other.  Nevertheless the purpose of the Protocol was clearly to obtain a reasonable degree of flexibility by requiring the courts to find a way between the traditional alternative conceptions of the claims, namely boundaries or guidelines.

Accordingly, patent claims should be interpreted on the basis that their purpose is neither to define the boundary of the monopoly, nor to act merely as guidelines for ascertaining the scope of the protection from the description and drawings, but rather to state the invention described and exemplified in the description and drawings, with a view to combining fair protection for the patentee with a reasonable degree of certainty for third parties.

Third parties may fairly be taken to have the knowledge of both the technical and legal background reasonably available to them through in-house or external advisers.  Patent claims should thus be interpreted through the eyes of a person who is aware of the requirements of the Protocol and the reasons for it.  This includes the knowledge that claims should define the matter for which protection is sought in terms of the technical features of the invention[25], but that it is recognised that patentees and their advisers may not foresee when the patent is granted all of the ways in which the invention may be used, and also that they cannot be expected to incur disproportionate expense in the formulation of every patent[26].  It also includes the knowledge that there is no principle of colourable infringement under the European Patent Convention, and hence that the balancing of the requirements of reasonable certainty and fair protection has to be achieved by a degree of flexibility in the interpretation of the claims.

In some cases, further legal knowledge is required to interpret a claim sensibly.  For example, “Swiss-form” claims (use of a specified substance for the manufacture of a medicament for a specified use) cannot realistically be interpreted without a knowledge of the law on novelty of new uses or new effects and on the exclusion from patentability of methods of treatment of the human or animal body, as expounded by the European Patent Office[27] and accepted by national courts[28].  As Jacob J observed of a quasi-Swiss-style claim, “I daresay that an ordinary skilled man (to whom it is notionally addressed) would find it puzzling, unless he had been initiated in some of the Byzantine logic of patent law and jurisprudence”[29].

The above analysis only identifies the basic principles which should be followed in interpreting patent claims in Europe.  The European Patent Convention goes no further.  Detailed principles have to be developed by the courts building on the right foundations. 

          4. Invocation of Catnic is contrary to the objective of the Protocol of achieving consistency in Europe. 

Even if the Catnic approach could be regarded as a possible way of combining reasonable certainty with fair protection, it differs conceptually from the approaches generally followed in other European countries, which tend to focus on what the claims convey to the skilled person rather than what the patentee intended[30].  It also differs in practice, in that patent claims tend to be interpreted more narrowly in the United Kingdom than on the Continent[31].

The travaux préparatoires for the European Patent Convention show that the Protocol was adopted in response to concerns expressed by representatives of European commerce regarding the difficulties and expense caused by the differences between European countries[32]. It was pointed out that these difficulties would be accentuated where a single European patent application covered countries with different approaches to the interpretation of claims.  A claim drafted with a view to relatively strict interpretation in the United Kingdom might be too broad in Germany, while a narrower claim suitable for Germany might not provide fair protection in the United Kingdom.

The disparities between the different countries create further complexity and expense for industry in relation to enforcement and commercial transactions. The fact that the result of infringement proceedings may vary from country to country gives rise to multiple litigation and jurisdictional disputes.  Differences in the scope of corresponding patents also increase the difficulty and expense of considering their value in connection with commercial and corporate transactions.

These and other disparities between different countries contribute to the overall complexity of the patent system.  This complexity presents a formidable barrier to individual inventors and small companies, and often diminishes the attention paid to patents by the management of bigger concerns.  It thus undermines the objectives of the patent system of promoting innovation and rewarding inventors.  While conformity in the principles of interpretation of patent claims in Europe would not eliminate complexity from the system, it would at least be a significant step in the right direction.

           5. Invocation of Catnic reduces the influence of British case-law on the Continent. 

Since British decisions on this issue are apparently based on pre-EPC principles, they are readily dismissed as of no relevance, as in the final decision of the Dusseldorf Court of Appeal in the German Improver case.  That court found in favour of the patentee, rejecting the conclusion of Hoffman J in the English Improver case on the ground that it was based on Catnic[33].  Unfortunately this was not drawn to the attention of the English court in the later case of Beloit v Valmet[34] where Jacob J made observations on the footing that the last word of the Dusseldorf Court of Appeal had been its interlocutory decision in favour of the defendant.

The continuation of a purely national line of authority in the United Kingdom is not conducive to a reciprocal exchange of principles between the United Kingdom and other countries.  It reduces the prospect of a convergence of European case-law influenced by British judgments.  This is regrettable because the reasoning of decisions on the interpretation of patent claims in other European countries frequently leaves a lot to be desired[35].  British lawyers and judges could contribute a great deal to the developing case-law if only they proceeded from a basis shared by other European countries.

           6. Continued reliance on the Catnic/Improver approach is contrary to any rational doctrine of precedent. 

The only case in which it has been seriously argued that the Catnic/Improver approach is inconsistent with the Protocol was PLG Research v Ardon[36].  After hearing full argument[37], the Court of Appeal concluded that Catnic should be consigned to the history books and that more attention should be paid to the developing case law on the Continent.  Yet recent judgments of English courts have preferred to follow other cases where the Catnic/Improver approach was uncritically accepted without argument and hence without exposure of its incompatibility with the Protocol. 

English judges have wrongly supposed that they are bound by judgments in cases where the compatibility of Catnic/Improver with the Protocol was not disputed.  This view does not accord with the English doctrine of precedent.  On the contrary, a proposition in a judgment does not rank as a decision for the purpose of the English doctrine of precedent unless the point was argued[38].

At the same time the recent judgments have treated the observations of the Court of Appeal in PLG on this point as not binding on the ground that they were obiter dicta. However, in PLG the Court of Appeal concluded that the German courts applied a test, which differed from Catnic, of considering whether the embodiment in issue is deducible by the skilled person from the wording of the claim in the light of the description and drawings[39].  They then applied this test in determining the issues of construction and infringement in that case. A proposition used in this way is a ratio decidendi.

Indeed, the Court of Appeal added an important qualification to the construction which had been adopted by Aldous J on the basis of Catnic, namely the use of a functional test to assess the substantiality of departures from uniplanarity[40].  Applying this test, they then found infringement in relation to the defendants’ uniaxially orientated products, reversing the decision of Aldous J[41].  Finally, they held that the defendants’ biaxially orientated products did not infringe because “while the starting material may be said to be a functional equivalent of a uniplanar starting material, this would not have been obvious to the skilled man and could not be deduced from the language of the patents in suit.”[42]  This conclusion clearly recalls the test distilled earlier in the judgment from the German case-law.

The Court of Appeal did say that it was “unnecessary to consider whether Lord Diplock’s purposive construction was an accurate if proleptic application of the Protocol”[43].  They gave two reasons for this.  Firstly, in PLG the relevant phrase was “substantially uniplanar”.  Since the claim itself used the word “substantially” this precluded an exact and literal construction.  Secondly, they concluded that “If the two approaches are the same, reference to Lord Diplock’s formulation is unnecessary, while if they are different it is dangerous.”  Accordingly Catnic should be ignored whether it accords with the Protocol or not.

Thus in PLG the Court of Appeal applied a test based on the Protocol and Continental authorities after hearing full argument and giving reasons for ignoring Catnic both in the instant case and generally.  This should not be dismissed as mere obiter dicta.

It would now take a brave judge at first instance to follow PLG v Ardon and defy the views expressed in more recent judgments of the Court of Appeal.  However, it would be helpful for High Court judgments to analyse issues of claim interpretation applying the Protocol directly, having regard to the points made above and the Continental case-law, in addition to, and separately from, any analysis under Catnic/Improver.  This may mitigate some of the problems identified at 4 and 5 above and 7 below.  It would also provide useful material for further review by the appellate courts.

           7. Continued reliance on the Catnic/Improver approach is delaying the development of the case-law which is needed to elucidate the Protocol. 

As indicated above, the European Patent Convention gives only limited guidance as to how to proceed and needs to be developed by case-law.  However, British courts are not contributing to this development when they decide cases on the basis of Catnic.

It has been suggested that abandoning Catnic now would throw away the benefits of years of case-law[44].  This argument might have some merit if Catnic were a valid basis for future development.  But since it is not, the very same argument is a powerful reason for starting to build a body of case-law on the right foundations without further delay. 

© Jonathan DC Turner 1999

[1] [1982] RPC 183, EPIC (Marks & Clerk’s European Patent Infringement Cases edited by Alastair Wilson QC, Jonathan DC Turner and Rory Sullivan, Trenton Publishing, 1999) 341

[2] Now Lord Hoffman

[3] [1990] FSR 181, EPIC 354

[4] See e.g. Henry Brothers (Magherafelt) v Ministry of Defence [1999] RPC 442 (CA - Butler-Sloss, Ward, Robert Walker LJJ), Auchinloss v Agricultural & Veterinary Supplies [1997] RPC 649 (Prescott QC), [1999] RPC 397 (CA – Simon Brown, Aldous, Clarke LJJ), Union Carbide v BP Chemical [1999] RPC 409 (CA - Simon Brown, Aldous, Mantell LJJ), Hoechst Celanese v BP [1999] FSR 319 (CA – Nourse, Aldous, May LJJ), Kastner v Rizla [1995] RPC 585, EPIC 445 (CA – Balcombe, Otton, Aldous LJJ), Minnesota Mining & Manufacturing’s (Suspension Aerosol Formulation) Patent [1999] RPC 135 (Pumfrey J), Consafe Engineering v Emtunga [1999] RPC 154 (Pumfrey J), Visx v Nidex [1999] FSR 405 (Neuberger J), Uni-Continental Holdings v Eurobond Adhesives [1999] FSR 263 (Pumfrey J), Brugger v Medic-Aid [1996] RPC 635 (Laddie J), Strix v Otter [1995] RPC 607 (Ferris J), Beloit v Valmet [1995] RPC 705, EPIC 425 (Jacob J), Assidoman Multipack v The Mead Corp [1995] RPC 321, EPIC 396 (Aldous J)

[5] [1995] RPC 287, EPIC 370. Corresponding proceedings in Baltimore (where a jury found in favour of the patentee on all points) are reported sub nom Tenax v Tensar at USPQ

[6] An after-dinner debate organised by The Intellectual Property Organisation on 8 June 1999, in which Lord Hoffman, Judge Randall R Rader, and Jacob, Neuberger, Pumfrey, Laddie JJ participated, revealed considerable doubts as to the desirability of continuing to apply Catnic

[7] [1982] RPC at 242 l.44, EPIC at 350G. 

[8] The hypothetical skilled man is not assumed to be skilled in patent law, per Hoffman J in Improver v Remington [1990] FSR 181 at 197, EPIC 354 at 368A

[9] See the observations of Jacob J in Bristol Myers Squibb v Baker Norton [1999] RPC 253 at para 43, discussed below

[10] EPIC 301, [1992] BIE 286

[11] [1982] RPC at 242 l.44, EPIC at 350G

[12] [1982] RPC at 243 l.5-20, EPIC at 351A-D

[13] [1990] FSR at 189, EPIC at 361B-D

[14] [1990] FSR at 197, EPIC at 368A

[15] See Beecham Group v Bristol Laboratories [1977] FSR 215

[16] Van der Lely v Bamfords [1963] RPC 61 at 75 l.40

[17] EPC art 177.  The German and perhaps the French have a slightly broader connotation that the English word.

[18] EPC art 164

[19] As suggested by Aldous J in Assidoman Multipack v The Mead Corp [1995] RPC 321 at 332 l.19, 336 l.35-47, EPIC 396 at 402F, 407A

[20] Lord Diplock’s speech was forcibly criticised on this ground by Dr Bruchhausen, President of the Patents Division of the German Supreme Court: paper of May 1981, published in Kemp, Patent Claim Drafting and Interpretation (1983).  This part of Judge Bruchhausen’s paper appears to have been overlooked by Aldous J in his comments on it in Assidoman Multipack v The Mead Corp [1995] RPC 321 at 336, EPIC 396 at 406C, 407B

[21] [1990] FSR at 192, EPIC at 364

[22] See Minnesota Mining & Manufacturing’s (Suspension Aerosol Formulation) Patent [1999] RPC 135  at 143 l.41, Consafe Engineering v Emtunga [1999] RPC 154 at para 9

[23] See the decision of the Zurich Commercial Court to this effect in Tool Retainer Spindles EPIC 323 at 327A-D

[24] Although decided under the English law in force prior to implementation of the European Patent Convention, the Catnic case concluded in the House of Lords in 1980, some 7 years after the Convention was concluded.

[25] EPC art. 84, EPC rule 29(1)

[26] This is a point which should not be distorted by the perspective of judges and litigators accustomed to investigations on a scale which could not be practicably applied to every patent application at the time of drafting

[27] Eisai/Second Medical Indication [1979-85] EPOR B241, Mobil/Friction reducing additive [1990] EPOR 73

[28] In the case of the UK, see Wyeth’s application [1985] RPC 545, Merrell Dow v Norton [1996] RPC 76

[29] Bristol Myers Squibb v Baker Norton [1999] RPC 253 at para 43

[30] A number of the more significant cases are collected and translated into English in Marks & Clerk’s European Patent Infringement Cases edited by Alastair Wilson QC, Jonathan DC Turner and Rory Sullivan (Trenton Publishing, 1999)

[31] As indicated by the Improver litigation.  See generally European Patent Infringement Cases (above); Bodenheimer and Beton [1993] EIPR 83

[32] See Marks & Clerk’s European Patent Infringement Cases (details above), p. 21 sq

[33] EPIC 216 at 221B

[34] [1995] RPC 705 at 719-721, EPIC 425 at 433-5

[35] As can be seen from the cases collected in Marks & Clerk’s European Patent Infringement Cases (details above)

[36] [1995] RPC 287, EPIC 370

[37] Including citation of cases decided in Germany, The Netherlands, France, Austria and Italy, cf. the comment of Aldous J in Assidoman Multipack v The Mead Corp [1995] RPC 321 and 337 l.30-35, EPIC 396 at 407E. 

[38] Re Hetherington [1990] Ch 1

[39] [1995] RPC at 309 l.25, EPIC at 385D

[40] [1995] RPC at 310 l.44 – 311 l.10, EPIC at 386F-387A

[41] [1995] RPC at 319 l.23-40, EPIC at 394E-G

[42] [1995] RPC at 320 l.7-15, EPIC at 395B-C

[43] [1995] RPC at 309 l.35-52, EPIC at 385E-G

[44] Assidoman Multipack v The Mead Corp [1995] RPC 321 and 337 l.24-35, EPIC 396 at 407E (Aldous J), Kastner v Rizla [1995] RPC 585 at 605, EPIC 445 at 462E (Balcombe LJJ)


Home ] Intellectual Property ] Competition Law ] IT Law ] Curriculum Vitae ] Cases ] Publications ] Instructions ] Links ]