Guidelines November 2019:

Purposive selection no longer needed for novelty of sub-ranges?

R. van Woudenberg (NL)

In the (pre-published) 2019 edition of the Guidelines for Examination in the European Patent Office, the paragraph relating to the novelty test of a sub-range was amended by deleting the purposive-selection criterion. For many years, the test followed established case law on the novelty of selection inventions as developed in particular in T 198/84 and summarised briefly in T 279/89, reason 4.1, according to which a selection of a sub-range of numerical values from a broader range is new when each of the following criteria is satisfied: (a) the selected sub-range should be narrow, (b) the selected sub-range should be sufficiently far removed from the prior-art range illustrated by means of examples (and, in the Guidelines version, from the end-points of the known range), and (c) the selected area should not provide an arbitrary specimen from the prior art, i.e. not a mere embodiment of the prior description, but another invention (purposive selection); i.e., just (arbitrarily) different values do no confer novelty. However, in recent years some board decisions, e.g., T 261/15, deviated from this well-established test, and considered the purposive-selection criterion an issue for inventive step rather than novelty. This different line of reasoning was not followed by the Guidelines over various years, (implicitly) indicating that the line of T 198/84 and T 279/89, including the purposive-selection criterion, was to be considered as established case law. However, in the (pre-published) November 2019 of the Guidelines, the test for novelty of a sub-range was suddenly change to the line of T 261/15. In this article, it is challenged whether the line of T 261/15 is correct, and whether the amendments to the Guidelines by deleting the purposive selection criterion - and thereby giving a different definition of the prior art then before - is legally correct, also in view of recent Enlarged Board decisions, notably G 3/89, G 11/91 and G 2/10, emphasizing the relevance of what the skilled person derives directly and unambiguously, using common general knowledge and also taking into account any features implicit to a person skilled in the art in what is expressly mentioned in the document, from the disclosure of the application as filed rather than what is literally disclosed.

1. Novelty of a sub-range – Guidelines G-VI, 8(ii)

In section G-VI, 8(ii) of the (pre-published) Guidelines 2019 (GL/EPO), the paragraph relating to the test for the novelty of sub-range has been amended to:

A sub-range selected from a broader numerical range of the prior art is considered novel if both of the following two three criteria are satisfied (see T 198/84 T 261/15 and T 279/89):

  1. the selected sub-range is narrow compared to the known range; 
  2. the selected sub-range is sufficiently far removed from any specific examples disclosed in the prior art and from the end-points[1] of the known range;
  3. the selected range is not an arbitrary specimen of the prior art, i.e. not a mere embodiment of the prior art, but another invention (purposive selection, new technical teaching).

An effect occurring only in the claimed sub-range cannot in itself confer novelty on that sub-range[2]. However, such a technical effect occurring in the selected sub-range, but not in the whole of the known range, can confirm that criterion (c) is met, i.e. that the invention is novel and not merely a specimen of the prior art. The meaning of "narrow" and "sufficiently far removed" has to be decided on a case-by-case basis. The new technical effect occurring within the selected range may also be the same effect as that attained with the broader known range, but to a greater extent.

i.e., the paragraph was amended to:

  1. replace the reference to T 198/84 by a reference to T 261/15,
  2. while keeping the reference to T 279/89,
  3. delete third criterion (c), which is the purposive-selection criterion,
  4. while keeping the text that related to the third criterion (c) (“An effect … greater extent”)

In my view, the amendment may create unclarity as to what the tests now comprises. Firstly, maintaining the reference to T 279/89 results in referencing to two decisions which are conflicting: T 261/15 deletes the purposive-selection criterion, while T 279/89 embraces T 198/84 and the purposive-selection criterion[3]. Secondly, maintaining the paragraph indicated above relating to the effect (“An effect …. greater extent”) results in an inconsistent part (ii) of GL/EPO G-VI, 8. The first line of that paragraph originated from -the now deleted decision- T 198/84. Also, the presence of that paragraph suggests that there is still a (now hidden) criterion as to a new effect. This makes the paragraph unclear. If a reader would, when confronted with this unclarity, check the cited case law, he would not find a clarification as cited T 261/15 and T 279/89 give different and conflicting tests. Also, there is no decision from the Enlarged Board that resolves the matter.

Further, it is doubted whether moving away from T 198/84 towards T 261/15 is correct in view of the arguments presented below.[4]

2. Established case law, conflicting case law

Purposive selection
In T 261/15, the Board argued that the purposive-selection criterion should not be part of novelty, but part of inventive step. T 1233/05 (Reasons 4.4), T 230/07 (Reasons 4.1.6), T 1130/09 (Reasons, 3.2), T 1948/10 (Reasons 3.6) and T 378/12 (Reasons 4.8 to 4.9), T 305/16 (Reason 2.1.3) argued similarly. This conflicting case law was however not followed in the Guidelines until and including the edition of November 2018.

T 198/84 as well as T 279/89 however argued the contrary. T 198/84 as well as T 279/89 have always been followed in the Guidelines as well as in the Case Law Book[5]. The decisions imposed the “purposive selection” criterion for a sub-range selected from a known continuous numerical range. T 198/84, reason 7 provides hereto: “It remains to be established whether this view of novelty really entails more than just a formal delimitation of the process concerned vis-à-vis the state of the art. It would be delimited only in respect of the wording of the definition of the invention, but not in respect of its content, if the selection were arbitrary, i.e. if the selected range only had the same properties and capabilities as the whole range, so that what had been selected was only an arbitrary specimen from the prior art. This is not the case, since the effect of the substantial improvement in yield may be believed to occur only within the selected range, but not over the whole known range ("purposive selection").”[6] A sub-range which is arbitrarily selected has no new technical effect and no new technical teaching with respect to the prior art broader range, and is thus not novel according to the test of T 198/84 and T 279/89.

There has not been any referral to the Enlarged Board on the topic, so it is difficult to understand why the Guidelines have now deleted criterion (c).

“Gold” standard
Decision G 2/10 from the Enlarged Board of Appeal[7] emphasizes that the “gold” standard[8] of what an application (or the prior art, as the same tests are to be applied according to G 2/98) discloses is that the skilled person directly and unambiguously derives from the disclosure as a whole[9]; not what he reads there literally (this was confirmed in G 1/16). The reasoning of T 198 and T 279/89 are fully consistent with that approach: it recognizes that the skilled person does not see some numerical values, but also an understanding of what happens in the range / what he can expect in the range based on the overall disclosure. T 261/15 and the other T-decisions that deny the criterion of purposive selection seem to put only little weight as to what the skilled person directly and unambiguously understands from the prior art range in its context (i.e., that the range reflects a certain physical, mechanical or chemical characteristic and that the prior art range reflects a certain purpose or effect upon the prior art subject-matter that is in part described by a range of a certain parameter). It may be doubted whether denying the relevance of a purpose/effect of the prior art on its disclosure and, consequently, denying the relevance of a new purpose/effect is consistent with the “gold” standard and the reasoning behind that.

Exact values of continues parameters are arbitrary
Further, T 198/84 and T 279/89 recognize that the exact values used to define a numerical range and the exact values of examples are arbitrary in the sense that they could be slightly different while substantially not changing any technicalities. Rather, the disclosure of the specific end points and examples just serves to proof that the claimed range is novel and enabled, over the full range with the technical characteristics as understood by the skilled person from the application as a whole. Only if something differently than what the skilled person derives from the disclosure as a whole would occur in a sub-range, can a sub-range be considered to relate to something technically different. This seems to be ignored in T 261/15.

End points
Also, other parts of decision T 261/15 (cited in Case Law Book (2019) I.C.6.3.1) are in my view difficult to understand and go against the common understanding, practice and the principle of legitimate expectations (see below). E.g., T 261/15, reason respect 2.3.2 comprises

“In the view of this Board, the limit values of a known range, although explicitly disclosed, are not be treated in the same way as the examples. The person skilled in the art would not, in the absence of further teaching in this direction, necessarily contemplate working in the region of the end-points of the prior art range, which are normally not representative of the gist of the prior art teaching.”

However, according to section I.C.6.3.2 “Overlapping ranges” of Case Law Book (2019), the Board held in T 240/95 that, “in accordance with established case law, disclosure of a range was considered to be an explicit disclosure of the end values”, followed by. e.g., T 1115/09 as mentioned in the same section.

So, reason 2.3.2 of T 261/15 seems to go against the common approach for sub-ranges where end points of a prior-art range are considered to be explicitly disclosed and where an end point of a prior-art range destroys the novelty of a later range that comprises that end point and extends on either side. T 261/15 seems to be incorrect (at least as a general principle; it may be correct under possible special circumstances of a specific case, e.g., where the teaching of the prior art document as a whole indicates that the mentioned end points are speculative).

This usual approach is also explicitly used elsewhere in the Guidelines, e.g. in GL/EPO G-VI, 8(iii):

“As to overlapping ranges or numerical ranges of physical parameters, novelty is destroyed by an explicitly mentioned end-point of the known range, explicitly mentioned intermediate values or a specific example of the prior art in the overlap.”

Citing T 261/15 for sub-ranges while the same decision goes against established practice in relation to other aspects of ranges and against the Guidelines with respect to the interpretation of end points seems inconsistent. In my view, a decision relied on in the Guidelines should be followed with respect to all of the aspects addressed in that decision, as it reflects the underlying argumentation. Further, the novelty-test for sub-ranges shall be consistent with G 2/10. See also my arguments above.

Authority of T 194/84
In contrast, the Case Law Book attributes a high degree of authority to decision T 198/84. Firstly, the opening paragraph of the Novelty chapter I.C.1 “General” refers to this decision as one out of only two when describing the purpose of the novelty requirements as:

“An invention can be patented only if it is new. An invention is considered to be new if it does not form part of the state of the art. The purpose of Art. 54(1) EPC is to prevent the state of the art being patented again (T 12/81, OJ 1982, 296; T 198/84, OJ 1985, 209).” [emphasis added]

Secondly, when discussing distinguishing features in I.C.5.2.1 “Difference in wording”, T 198/84 is cited in support of the later T 114/86:

“In T 114/86 (OJ 1987, 485) the board held that a mere difference in wording was insufficient to establish novelty (see T 12/81, OJ 1982, 296; T 198/84, OJ 1985, 209; T 248/85, OJ 1986, 261). In T 565/90 the appellant submitted that only preferred ranges or examples amounted to a technical disclosure destructive of novelty, and that generic ones could not anticipate the more specific teaching of the patent in dispute. The board did not agree and confirmed earlier case law that the definition of an invention which differed from the prior art only in its wording was insufficient to establish novelty. The board stated that what had to be established was whether or not the state of the art made the subject-matter of the invention available to the skilled person in the form of a technical teaching.” (emphasis added]

In my view, where a decision has such a high level of authority in the Case Law Book, and not just established case law but also leading as established case law, extra care must be taken to deviate from the decision. Also, a difference in approach between the first instance search, examining and opposition divisions, who follow the Guidelines, and second instance Boards, is to be prevented.

Narrow and far-removed criteria cannot be seen independent from technical meaning
The narrow and far-removed criteria can in my view not be seen as a purely numerical comparison of end points of claimed and known ranges and end points of claimed ranges and known examples. When comparing the values, the skilled person will compare the technical differences, i.e., compare the technical teachings. Whether two values are close or far will be considered taking into account the difference between the complete technical embodiments at the two points. E.g., comparing embodiments comprising water at 85 and 95⁰C may be considered close to each other, whereas embodiments comprising water at 95 and 105⁰C will surely be far removed from each other in view of the - to the skilled person well-known - phase change occurring at 100⁰C. The values 85, 95 and 100 as such have no technical meaning; their technical relevance, as understood by the skilled person, incorporates - and may even be considered to be fully determined by - the effects that they provide. Similarly, a claimed sub-range of 45-65% of a certain parameter compared to a prior art range of 40-70% may render novelty to the claimed sub-range if there would be some kind of new physics in the sub-range of which the existence is not known from the known end points 45 and 65%.

In view of this, the three criteria of T 198/84 should not be seen as independent, isolated tests, but the purposive-selection criterion may also play a role also for the narrow and far-removed criteria, as it gives technical meaning to the subject-matter associated with the respective values.

Principle of legitimate expectations
A change of the novelty-test for sub-ranges has a major impact on various existing and new cases: it changes the definition and the effect of the prior art and it also changes the definition of the “same subject-matter” in the context of amendments and of the “same invention” and the “first invention” in the context of priority (Art. 87 EPC). In my view, such an important change cannot occur silently, but would at least require an explanation in the Official Journal, either from the Enlarged Board or from the EPO. In view of the good faith requirement (principle of legitimate expectations), such a change would need to be announced far in advance such that the applicant can take it into account when drafting an application, during the substantive examination of his application, as well as during any subsequent opposition and/or appeal proceedings. In my view, making a new interpretation that goes against a long-established practice also applicable for already filed applications goes against this principle.

The Guidelines should be a reliable source of legitimate information. Established practice can -by definition- not change instantly by amending the Guidelines if it is not due to a change of the legal provisions, a T/J-decision on a topic for which there is no established case law yet, or a G-decision; none of which apply in this case.

Established case law can - by definition - not change instantly, unless there is a G-decision that overturns the earlier main way of reasoning; which there is not in that case. Also see Case Law Book (9th edition, 2019) III.A.1.1:

“1.1 Sources of legitimate expectations
Sources of legitimate expectations include information provided by the EPO in individual cases (e.g. in the form of communications to the party), information contained in official statements of general applicability and published by the EPO (e.g. the Guidelines and the Official Journal), established practice of departments of the EPO, and decisions taken by the Enlarged Board because of its special role (see J 25/95 and the decisions given on the same day, namely J 14/95, J 15/95, J 16/95, J 17/95 and J 24/95; see also T 905/90, OJ 1994, 306, Corr. 556). The case law of the boards of appeal may also be a source of legitimate expectation, in particular, if it is established case law which has become enshrined in the consistent practice of the department of first instance (see J 27/94, OJ 1995, 831; see also in this chapter III.A.5). Courtesy services provided by the EPO may also be a source of legitimate expectation (see J 1/89, OJ 1992, 17; see also in this chapter III.A.2.2)” [emphasis added]

Even though the recent editions of the Case Law Book indicates that there is also some conflicting case law, the Case law Book (2019) takes a clear position as to what is the established case law, in I.C.6.3.1 “Selection from a broad range”:

“The principles applied by the boards of appeal as part of their established case law on the novelty of selection inventions were developed in particular in T 198/84 (OJ 1985, 209). They are summarised briefly in T 279/89, according to which a selection of a sub-range of numerical values from a broader range is new when each of the following criteria is satisfied:
The Guidelines recall the three criteria (see G‑VI, 8 (ii) – November 2018 version).
The three postulates for the novelty of a selected sub-range are based on the premise that novelty is an absolute concept. It is therefore not sufficient merely for the wording of the definition of an invention to be different. What has to be established in the examination as to novelty is whether the state of the art is such as to make the subject-matter of the invention available to the skilled person in a technical teaching (T 198/84, OJ 1985, 209; see also T 12/81, OJ 1982, 296; T 181/82, OJ 1984, 401; T 17/85, OJ 1986, 406).” [emphasis added]

Except for the addition to the reference to the Guidelines, the text has been unamended in this wording in the Case Law Book since the first edition of 1993, including the explicit indication that “The principles applied by the boards of appeal as part of their established case law on the novelty of selection inventions were developed in particular in T 198/84 (OJ 1985, 209)”.

It violates the principle of legitimate expectations to deviate from what the Boards have consistently referred to as established case law for over 25 years. Such a change could in my opinion only be appropriate if there would be a decision of the Enlarged Board of Appeal that overturns the established case law. In the absence of such a decision, it is difficult to see a justification for the amendment made to the Guidelines with the 2019 edition.

3. Consistency with the criteria for overlapping ranges

Guidelines G-VI, 8(iii) addresses overlapping ranges; this section is not amended in the 2019 edition. In my view, deleting the “purposive selection” criterion from the sub-range test is not consistent with the criteria, arguments and clarification given in the Guidelines for novelty of overlapping ranges.

In my view, the tests for novelty of overlapping ranges are consistent with G 2/10 and maintaining them unamended in the Guidelines is correct and appropriate. Guidelines G-VI, 8(iii) provides:

“It has to be decided which subject-matter has been made available to the public by a prior-art disclosure and thus forms part of the state of the art. In this context, it is not only examples, but the whole content of the prior-art document which has to be taken into consideration.” [emphasis added]


“It is not sufficient to exclude specific novelty destroying values known from the prior-art range, it must also be considered whether the skilled person, in the light of the technical facts and taking into account the general knowledge in the field to be expected from him, would seriously contemplate applying the technical teaching of the prior-art document in the range of overlap. If it can be fairly assumed that he would do so, it must be concluded that no novelty exists. In T 26/85, the skilled person could not seriously contemplate working in the area of overlap, since the prior art surprisingly contained a reasoned statement clearly dissuading him from choosing said range, although the latter was claimed in said prior art.” [emphasis added],

i.e., G-VI, 8(iii) recognizes, in the first two cited passages, the concepts of G 2/10 in that prior art is not just numeric values, it comprises a technical teaching. It emphasizes that it is not just the numerical values, but the what is understood as well as seriously contemplated from the prior art as a whole.

GL C-VI, 8(iii) also provides:

“The concept of "seriously contemplating" is fundamentally different from the concept used for assessing inventive step, namely whether the skilled person "would have tried, with reasonable expectation of success", to bridge the technical gap between a particular piece of prior art and a claim whose inventiveness is in question (see G‑VII, 5.3), because in order to establish anticipation, there cannot be such a gap (T 666/89).” [emphasis added]

I.e., G-VI, 8(iii) recognizes that novelty is about lack of technical gap and inventive step is about bridging the technical gap. As G-VI, 8(iii) also provides that

“In the case of overlapping ranges (e.g. numerical ranges, chemical formulae) of claimed subject-matter and the prior art the same principles apply for the assessment of novelty as in other cases, e.g. selection inventions.” [emphasis added]

Hence, in my view purposive selection needs to be maintained as part of the novelty test for sub-ranges: without purposive selection, there is no technical gap between a prior art range and a claimed sub-range. Thus, in my view, maintaining the test for overlapping ranges while amending the test for sub-ranges results in embracing conflicting principles for the two situations (see discussion above), while the current tests for the novelty of sub-ranges (i.e., including purposive selection) is fully consistent with the test for overlapping ranges and, at the same time, both embrace the principles of G 2/10.


In this article, it is argued that the test for novelty of a claimed sub-range of a known range should include the purposive selection criterion, as it did in the Guidelines until the revision of November 2019. Even though a decision of the Enlarged Board would be strongly preferred to clarify the situation, the purposive-selection criterion seems to be consistent with the principles of G 2/10 (“gold” standard), whereas a novelty test that excludes the purposive-selection criterion (i.e., is only based on comparing numerical values) seems to fail to recognize what the skilled person would directly and unambiguously derives from the prior art disclosure as a whole (i.e., the broader range and the technical meaning of the prior art subject-matter within its broad range). Further, it is argued that the purposive-selection criterion for sub-ranges is based on the same principles as the seriously contemplating concept for overlapping ranges: the whole prior art disclosure is to be taken into account and, for novelty, there must be a technical gap between what the skilled person would directly and unambiguously derive from the prior art disclosure as a whole and the claimed subject-matter with its sub-range. As first instance practice will suddenly change per 1 November 2019 due to the change to the Guidelines and the situation at second instance is also uncertain in view of conflicting Case Law, there is a strong and urgent need for alignment at first instance (via consistent Guidelines) as well as at second instance (via a referral to the Enlarged Board of Appeal) to provide for clarity as well as legal certainty for applicants and third parties.

The author wants to thank Cees Mulder, Derk Visser and Jessica Kroeze for discussions on the topics addressed in this article and for reviewing its draft versions.

  1. Note that the Case Law Book (2019) indicates in III.W.3 that: “In T 261/15, in respect of the argument that a selected sub-range has inter alia to be sufficiently far removed from the end-points of the known range, the board pointed out that it was not aware of any jurisprudence stating this condition in such a general way. It was true that the Guidelines for Examination, G‑VI, 8, recited under point (ii)(b) this criterion as a condition for acknowledging novelty of a numerical selection. However, neither decision T 198/84 nor T 279/89, which are cited in this passage of the Guidelines, stipulates this condition.”
  2. This sentence is taken from the headnote of T 198/84, which provides “The sub-range is novel not by virtue of an effect which occurs only within it; but this effect permits the inference that what is involved is not an arbitrarily chosen specimen from the prior art but another invention (purposive selection).”. Thus, the headnote clarifies that a new effect permits to conclude that the selection is not arbitrary, but rather a purposive selection. It is addressed in some more detail in reason 7.
  3. The wording of the three criteria actually stems from T 279/89, reason 4.1, wherein the principles applied by the boards of appeal as part of their established case law on the novelty of selection inventions, as developed in particular in T 198/84, were summarized. Also see Case Law Book (2019) I.C.6.3.1.
  4. Reference is also made to Derk Visser’s article elsewhere in this issue of epi Information, expressing a different point of view as well as some other aspects.
  5. Case Law Book (2019) I.C.6.3.1 “Selection from a broad range” as well as corresponding sections in all earlier editions
  6. Reason 7 of T 198/84 continues with: “To prevent misunderstanding, it should be expressly emphasised that when examining so-called selection inventions as to novelty the Board adheres to the principle that the sub-range singled out of a larger range is new not by virtue of a newly discovered effect occurring within it, but must be new per se (cf. T 12/81 "Diastereomers/BAYER", OJ of the EPO 8/1982, 296, 303). An effect of this kind is not therefore a prerequisite for novelty; in view of the technical disparity, however, it permits the inference that what is involved is not an arbitrarily chosen specimen from the prior art, i.e. not a mere embodiment of the prior description, but another invention (purposive selection).” Thus, reason 7 emphasises that the third requirement is “what is involved is not an arbitrarily chosen specimen from the prior art, i.e. not a mere embodiment of the prior description, but another invention (purposive selection)” and that the latter may be interfered from a newly discovered effect. 
  7. G 2/10 is cited in the Guidelines (2019) in, e.g., H-IV, 2.1, last paragraph and H-V, 3.1.
  8. G 2/10 did not develop the “gold” standard itself, but G 2/10 introduced the term in reason 4.3: “[…] the general definition of the requirements of Article 123(2) EPC established in opinion G 3/89 and decision G 11/91, which definition has become the generally accepted, one could also say the "gold" standard, for assessing any amendment for its compliance with Article 123(2) EPC.”
  9. For the definition of the “gold” standard, G 2/10, reason 4.3 refers to G 3/89 and G 11/91: “4.3 The basic principle underlying Article 123(2) EPC, in the jurisprudence of the Enlarged Board The importance and the applicability, without exception, of Article 123(2) EPC was underlined in the jurisprudence of the Enlarged Board of Appeal as early as in its opinion G 3/89 and decision G 11/91 (OJ EPO 1993, 117 and 125, relating to amendments by way of correction). From these rulings it follows that any amendment to the parts of a European patent application or of a European patent relating to the disclosure (the description, claims and drawings) is subject to the mandatory prohibition on extension laid down in Article 123(2) EPC and can therefore, irrespective of the context of the amendment made, only be made within the limits of what a skilled person would derive directly and unambiguously, using common general knowledge, and seen objectively and relative to the date of filing, from the whole of these documents as filed, points 1., 1.3 and 3. of the Reasons.”