Digipos Store Solutions Group Ltd v Digi International Inc

This is an appeal from a Decision of Mr George Salthouse, the Hearing Officer for the Registrar, the Comptroller-General, dated 14^th May 2007 whereby he upheld the opposition to registration of the Appellant's marks set out in the Decision at paragraph 1. The grounds of opposition were based primarily on the Respondent's prior registrations, which are set out in paragraph 2 of the Decision. The Hearing Officer held that the opposition succeeded under section 5(2)(b) of the Trade Marks Act 1994 ("the Act") against all of the Appellant's applications in all classes.

Although all of the marks identified in the Annexes to this judgment are relied on as a basis for opposition it is convenient to undertake the analysis by reference to two of the Respondent's marks namely

The central analysis on this appeal can be undertaken by reference to one of the Appellant's marks namely UK application 2358343

I deal with the other marks at the end of this judgment.

It is important at the outset to bear in mind the nature of appeals of this kind. It is clear from Reef Trade Mark [2003] RPC 5 ("Reef") and BUD Trade Mark [2003] RPC 25 ("BUD") that neither surprise at a Hearing Officer's conclusion nor a belief that he has reached the wrong decision suffice to justify interference by this court. Before that is warranted, it is necessary for this court to be satisfied that there is a distinct and material error of principle in the decision in question or that the Hearing Officer was clearly wrong (Reef). As Robert Walker LJ (as he then was) said:

This was reinforced in BUD, where the Court of Appeal made it clear that it preferred the approach of the appellate judge but nonetheless held that there was no error of principle justifying departure from the Hearing Officer's decision. As Lord Hoffmann said in Biogen v. Medeva [1997] RPC 1 at 45, appellate review of nuanced assessments requires an appellate court to be very cautious in differing from a judge's evaluation. In the context of appeals from the Registrar relating to section 5(2)(b) of the Act, alleged errors that consist of wrongly assessing similarities between marks, attributing too much or too little discernment to the average consumer or giving too much or too little weight to certain factors in the multi-factorial global assessment are not errors of principle warranting interference. I approach this appeal with that in mind.

The Appellant contends that the Hearing Officer made ten errors. These may be summarised as follows.

It is convenient first to address the issue of law relating to the assessment of distinctiveness of the respective marks and the factual assessment of distinctiveness in the light of the relevant law. That involves consideration mainly of points (3), (6), (7) and (8) together. In that context, it is convenient also to deal with the issue of what account should be taken of the representations in the file which are relied upon (point (1)) since it is said that these should have featured more prominently in the assessment of distinctiveness of the mark and the scope of protection of the earlier Respondent's marks. I deal separately with the other points: (4), (5), (9) and (10).

Of the arguments, it seems to me that point (3) lies at the heart of this appeal. Before evaluating the arguments relating to it, it is convenient to summarise the Appellant's and Respondent's positions on this point more fully.

The Appellant (represented by Mr Richard Arnold QC) contends that the Registrar is obliged, as a result of the judgment of the ECJ in Canon Kabushiki Kaisha v. Metro-Goldwyn-Mayer Inc. (Formerly Pathe Communications Corporation) Case C-39/97 [1998] ECR I-5507, [1999] RPC 117 ("Canon") and other cases, to assess not only whether the earlier mark enjoys enhanced distinctiveness, and therefore enhanced protection, as a result of its use but, conversely, whether there are features about the prior mark which diminish its distinctive character and therefore reduce its scope of protection.

In a case where the strength of the earlier registered mark is in issue, the Appellant contends that the Hearing Officer is obliged to consider whether the prior mark consists of, or comprises any elements which are, descriptive or which are or may be used descriptively. It is said that the Hearing Officer did not perform this assessment and that his failure to do so led to an erroneous approach to comparison of the respective marks in making the global assessment. The Appellant contends that the Hearing Officer should have compared the marks on the basis that the Respondent's mark afforded only narrow protection.

The Appellant develops the argument as follows. Once it is appreciated that the Hearing Officer should have assessed the mark's potential weaknesses, the necessity of evaluating matters which, it is said, were not considered properly or at all becomes apparent. If the Hearing Officer had been alive to the need specifically to assess the distinctive character of the marks for "weakness", he would not have so readily discounted the factors and the evidence which went to that issue. Assessment of whether the mark had a conceptual meaning, what third party use there had been, as shown by actual use and by registration of marks, the assessment made by third parties of the meaning of DIGI-, as a prefix, by OHIM and the respondent's agent's own submissions as to weakness of the DIGI mark should have been analysed in greater detail and not brushed aside. In this context, the Appellant says that the Hearing Officer should have given weight to statements by the Respondent's agents to the effect that the prefix DIGI- was weak and that no one company could claim to have a monopoly in it. Had this analysis been undertaken, the Appellant submits, the Hearing Officer should have and would have concluded that the mark "DIGI" was weak in the relevant sense in the context in question. That finding should then have been put together with the relatively high level of discernment likely to be employed by the average consumer for the goods in question, the nature of the goods and services for which the comparison was made and should, accordingly, have led to a conclusion that overall, there was no likelihood of confusion. If the analysis is correctly done, the Appellant contends that a result is reached which, if not proved by, is at least consistent with the evidence that parallel trading has not revealed any confusion and is consistent with the representations on the file.

In response, the Respondent (represented by Mr James Mellor QC and Ms Fiona Clark) argues as follows.

First, the Respondent submits that there is no principle that requires weakness to be assessed in the way suggested by the Appellant, or that it is a requirement of Community jurisprudence derivable from the judgment of the ECJ in Canon . The Respondent accepts that the ECJ contemplates that, in applying Article 5(2)(b), the Registrar must examine earlier marks to determine whether they enjoy enhanced reputation arising from use but there is no converse requirement to examine them for diminished reputation, once it has been established that they have sufficient distinctive character to be registrable.

Second, it is contended that, even if the Registrar is required to take "weakness" into account, this does not have to be done as a separate exercise. Rather, it forms part of the global assessment. Moreover, it is submitted that the contrary approach would be artificial and inconsistent with the judgment of the ECJ in L'Oreal (cited below): the Respondent submits that the global assessment does not mandate separate assessment of the strength or weakness of a prior mark.

Third, the Respondent submits that, even if such were required, the Hearing Officer in fact undertook the required assessment in this case. He considered the submission of alleged weakness as much as he was required to do and did not find it persuasive. There is no room for this court to interfere with that assessment.

Fourth, as to the other factors, the Respondent submits that the weight given to them was a matter for the Hearing Officer's judgment. He took them all into account, giving them the weight he thought appropriate, made the global assessment on the basis of all of them and there has been no error of principle.

Fifth, the Respondent submits that, with rare exceptions, no account should be taken of the trade mark prosecution file in principle and that there are strong reasons against taking any account of it in this case. Accordingly, the Hearing Officer was right to ignore it.

Finally, the Respondent contends that the Appellant's points are a sophisticated but impermissible way of inviting this court to make a different assessment and to identify an error of principle where none exists. The Respondent contends that the Hearing Officer was right and even if this court would have taken a different view as to the likelihood of confusion, that affords no basis for interfering with his decision. The warnings from Reef, Bud and other cases concerning unwarranted appellate interference with the assessments of the Registrar are emphasized.

I deal first with the central issue of law arising on this appeal, namely what is required by way of assessment of distinctive character in a case in which it is alleged that the common element of the marks has a lesser distinctive character, or in the common shorthand, is "weak".

Section 5(2) of the Trade Marks Act 1994 which reflects the Trade Marks Directive 89/104, Article 4, provides:

It is also convenient to have regard to Section 3(1) of the Act. This provides:

The Community Trade Marks Regulation 40/94 contains similar provisions.

Three preliminary observations arise. First, neither in Community nor in UK trade mark law is there an express legislative requirement that the Registrar or OHIM should assess the degree of distinctiveness of the earlier registered trade mark in determining whether or not there is a likelihood of confusion. Second, an earlier mark may perfectly validly contain signs or indications which may serve in trade to designate the kind, quality, quantity (and so forth) of goods or services and is registrable without proof of use, unless it consists "exclusively" of such signs. Third, even where an earlier mark consists exclusively of such signs, it may nonetheless be registrable as a result of the distinctive character that has resulted from the use made of it. So the legislation contemplates that signs which are to some extent descriptive may be registered and protected from use and from the registration by others of similar marks.

Nonetheless, both Community jurisprudence and the courts in this country have recognized that the nature of the earlier mark and, in particular, its distinctiveness may be relevant to the question of whether there is a likelihood of confusion. Such issues have arisen both in the context of disputes over registrability of a later mark and in the context of infringement proceedings, as the authorities show.

It is convenient to begin with Canon , since the Appellant bases its argument that it is necessary to consider the weakness of a prior mark primarily upon that decision of the ECJ.

In Canon , a case on reference from the Bundesgerichtshof, the ECJ was invited to consider whether account could be taken, in assessing the similarity of the goods and services covered by two marks, of the distinctive character and, in particular, the reputation of the earlier mark. The reference arose because, under earlier German law, no account was taken of the reputation of the earlier mark. Underlying the issue was the question of whether consideration of reputation could make it easier to demonstrate a likelihood of confusion.

The point at the heart of Canon was whether the greater reputation of the earlier mark could make it easier to prevent registration or use of a later mark for a broader range of goods or services. The reference in Canon did not directly raise the question of the extent to which the reputation of the earlier mark was of relevance in assessing the likelihood of confusion in situations where there was no dispute that the goods or services were identical or similar. The ECJ nonetheless made observations in general terms:

The Appellant relies on this as indicating the need to assess the distinctive character of the earlier mark, including determining whether it has a lesser degree of distinctiveness. The Respondent relies on it as demonstrating that the ECJ did not in that case contemplate that there should be an assessment of reduced distinctiveness, only of whether the distinctiveness was enhanced through use.

In my judgment, there is merit in both these submissions. Canon states that it is necessary for the competent authorities to take account of the distinctive character. That proposition was stated in unqualified terms and there is no reason in principle why, if account must be taken of enhanced distinctive character, account should not be taken of reduced distinctive character. It would appear that the ECJ may have regarded Canon as a basis for the proposition that such should be done in its judgment in L'Oreal at least by analogy (see paragraph 12 of the judgment of the ECJ in L'Oreal ). On the other hand, Canon is not direct or explicit authority for it. The issue did not arise expressly in the reference, the Court was not required to make any observations on that issue and did not directly do so. In my judgment, while Canon is consistent with the proposition for which the Appellant contends and afford a basis for an argument by analogy or by implication, it is not direct authority for it. To that extent, aspects of both the Appellant's and the Respondent's submissions on Canon are justified.

However, more powerful support for the Appellant's argument that it is necessary to assess whether the earlier mark enjoys reduced distinctive character may be found in the cases following Canon cited during argument on this appeal of which the following are the leading examples.

In Lloyd Schuhfabrik Meyer & Co. GmbH v. Klijsen Handel B.V. Case C-342/97 [1999] ECR I-3819 (" Lloyd" ), the ECJ said (emphases added):

Both in Windsurfing in a different context and in Lloyd , more directly, the ECJ contemplated that the national court should take account of the extent to which the mark does or does not contain an element descriptive of the goods, in performing the global assessment required under Article 5(2)(b) and in particular should take the lesser distinctive character into account.

In Reed Executive plc v. Reed Business Information Ltd [2004] EWCA Civ 159 [2004] RPC 40 (" Reed "), Jacob LJ said, in a judgment with which Auld and Rix LJJ agreed:

This latter proposition is consistent with the approach taken by the CFI summarized in paragraph 74 of its judgment in Devinlec Developpement Innovation Leclerc SA v. OHIM Case T-147/03 where the distinctive and dominant components of the marks in issue were QUANTUM (as compared with QUANTIEME) for watches and similar goods:

The CFI overturned the assessment by OHIM that the marks were not sufficiently similar to cause confusion and one can readily understand why on the facts. That decision was upheld by the ECJ on appeal (Case C-171/06P [2007] ETMR 38).

More recently, in L'Oreal SA v. OHIM Case C-235/05P, Judgment of 27 ^th April 2006, (" L'Oreal" ), the ECJ summarized the approach to the global assessment, where a prior mark was said to be weak. To understand what the ECJ was saying and the submissions relating to it, I set out the passage in full.

In L'Oreal , the applicant for registration of FLEXI AIR, L'Oreal, had argued that, in the case of an earlier mark with weak distinctive character there would only be a likelihood of confusion if the mark was completely reproduced: see paragraph 11 of the Judgment. The ECJ rejected that argument: see paragraphs 45 and 47 of the Judgment. But, in so doing, it did not suggest that the weakness of the earlier mark should not be taken into account. Although it is not particularly easy to understand paragraphs 42 and 43 of the judgment, taken in context, the ECJ was not in my view, there saying that the court should take no account of the weakness of the earlier mark in assessing the likelihood of confusion: that would have been to depart from Lloyd . Rather, that, in that case, notwithstanding the finding of weakness of FLEX, the CFI rightly concluded that there was a likelihood of confusion. The conclusion on the facts may be unsurprising: the marks there related to products in class 3 of the Nice Classification more likely to be purchased in a high street environment.

A similar approach was taken in the ECJ's judgment in the appeal from the CFI in Devinlec Developpement Innovation Leclerc SA v. OHIM ( T.I.M.E Art Uluslararasi Saat Ticareti ve Dis Ticaret AS v.OHIM , Case C-171/06P [2007] ETMR 38) where it was held that the CFI was entitled to treat the fact that the earlier mark was only of weak distinctive character as not being of overriding importance: see paragraphs 38-42. Nonetheless the weak distinctive character was treated as a relevant factor which should be taken into account.

The following propositions may be drawn from the relevant authorities as to the approach that the Registrar should take to the assessment under Article 5(2)(b) of the Act, in the case of a mark where the element alleged to be similar is also alleged to be of weak distinctive character. (I do not set out the aspects which are common to all assessments such as the requirement that use of the earlier mark is assumed, the general approach to assessment on the basis of Lloyd and so on).

First, in undertaking the global assessment of likelihood of confusion (as to which see Lloyd passim) where the earlier mark, or an element of it, is said to be descriptive of goods or services or otherwise weak, the Registrar must take into account the extent to which the sign has a lesser capacity to distinguish as well as whether it has a greater capacity to distinguish ( Canon , Windsurfing , Lloyd, Reed and L'Oreal ). That involves evaluation of the evidence relating to that issue. Because the examination of capacity to distinguish is undertaken in the course of comparing the marks to see whether confusion is likely, for the purpose of a section 5(2)(b) determination, it is necessary to look at the characteristics of both the earlier trade mark and the mark applied for. In my judgment, although not expressly stated in the cases, it follows from the general principles that there are cases in which an element may be distinctive in the context in which it appears in the earlier mark, but not in the different context in the mark applied for. If, in that latter context, it is unlikely to be taken to denote trade origin, that must be taken into account in making the global assessment, since it is one aspect of the lesser capacity to distinguish.

Second, the fact that the element has a lesser capacity to distinguish should be taken into account in the global assessment by recognizing that the average consumer is more likely to consider that a descriptive element is not being used, in the mark applied for, to identify the origin of the goods or services in question and therefore is less likely to be confused as to origin. Although it is convenient shorthand to talk of the scope of protection of a weak earlier mark being reduced, that is neither a proposition of law nor a blueprint for assessment. A weak mark is, in practice, likely to have a reduced scope because there is a lesser likelihood that other marks will be confused with it, if all that the respective marks have in common is the descriptive element. That, in turn, is because, as a matter of fact, the common element would not be thought by the average consumer to signal that the goods in relation to which the respective marks are used come from the same trade source, not because of any special approach under section 5(2)(b) to marks composed of descriptive elements ( Reed , L'Oreal ).

Third, although the lesser capacity to distinguish (the "weakness") of an earlier mark or element of the earlier mark should be taken into account in the global assessment, weakness of the earlier mark is not conclusive of whether there is likelihood of confusion. It is one factor which goes into the global assessment. In some cases, an earlier mark may be descriptive in whole or in part but nonetheless there may be a likelihood of confusion ( L'Oreal ).

In the light of that analysis, I turn to consider the Hearing Officer's approach to the assessment of the greater or lesser capacity of the DIGI mark to distinguish which is, in the main, contained in paragraphs 35 to 38 of the Decision.

Paragraph 35 of the Decision summarises the effect of the ECJ authorities, Sabel Lloyd, and Canon. Although that summary refers to Lloyd, it does not specifically pick up on the point in the judgment of the ECJ referred to above, namely the importance of assessing the lesser capacity of the mark to distinguish. Paragraph 36 of the Decision summarises the Hearing Officer's approach. However, in paragraph 37, the Hearing Officer refers to the decision of Mr David Kitchin QC sitting as the Appointed Person in Steelco Trade Mark (BL O/268/04) where he referred to the passage from Reed set out above, which is to the same effect as Lloyd. The Decision then considers whether the DIGI mark was entitled to an enhanced reputation and concludes that it is not. In considering the question of reputation, he said, at paragraph 38:

Pausing there, the Hearing Officer did not, at this point, undertake any (or any meaningful) analysis of the descriptiveness of DIGI-, when used as a prefix in contexts other than the earlier registered marks and, in particular, when used as a prefix in the Appellant's marks used in relation to digital or computer-related goods or services. His findings were there confined to the distinctiveness of the Respondent's mark DIGI and whether it was entitled to enhanced reputation. However, he returned to the question of the meaning of DIGI in the respective marks later in the Decision. At paragraphs 46 and 47, the Hearing Officer said:

In my judgment, although expressed briefly, paragraph 46 amounts to a determination that DIGI- used as a prefix in the Appellant's mark would not be taken to be descriptive by the average consumer. Only the POS suffix would be so taken.

The Appellant contends that this amounted to a finding by the Hearing Officer that the average consumer would not regard the prefix DIGI as a reference to digital and that, as such, it is plainly wrong. It is possible to argue that these passages, taken in the context of the judgment as a whole do not amount to such a finding as regards the meaning of the prefix DIGI- in the context of the Appellant's marks and that might be said to accord better with paragraphs 35-38 and paragraph 47. However, I do not think that is right. In my judgment, in paragraph 46, the Hearing Officer was there seeking to evaluate how the average consumer would regard the DigiPos mark as a whole and in particular the DIGI- element of it. Moreover, he had, only a few paragraphs earlier, referred to the Steelco decision and, in particular, observations of Jacob LJ in Reed there cited. This refers to the need to consider the descriptiveness of marks or elements of them. It would be surprising if, having referred to the need to undertake the analysis, the Hearing Officer wholly failed to perform it or have it in mind in what he said later.

It is therefore not possible to say that the Decision, taken as a whole, does not undertake the assessment which I have held should be done. Neither L'Oreal nor Reed require the analysis of the descriptiveness of the respective marks to be addressed in a formulaic way, so long as adequate account is taken of it. The Hearing Officer did not analyse the material referred to in detail nor did he make any specific findings about it (see paragraph 47). But it is reasonably clear that he at least had the points in mind. He had referred to the evidence going to this issue earlier in the Decision and must be taken to have rejected the argument that this was a case where the lesser degree of distinctiveness of both sides' marks was established (or was decisive, having regard to all the other circumstances).

Accordingly, the Hearing Officer did make the required evaluation of the descriptiveness of the marks in question.

Had the Hearing Officer not made the evaluation of "lesser" distinctiveness at all there would, in my judgment, have been an error of principle. Although the boundary between errors of principle and other errors is not always easy to trace and the court must not be too ready to categorise an error as one of principle, a failure to analyse the way in which the average consumer would view elements alleged to be descriptive in the Respondent's marks and in the Appellant's marks would have been an error of principle. It is fundamental to performing the global comparison required by the Act and the Directive that the right factors are taken into account. Descriptiveness (or the greater or lesser degree of distinctiveness) of the marks and the relevant elements of the marks in question is potentially an important factor. Moreover, the Court of Appeal in Reed emphasised that it could be important to take into account factors such as whether the marks contained an element descriptive of the goods in making the comparison (see per Jacob LJ, paragraph 86) and, when the Court of Appeal came to apply section 5(1)(b) of the Act, it specifically took into account the fact that "small differences may suffice to differentiate" (paragraph 96). Accordingly, an error, if committed, which consists in a failure adequately to evaluate such a factor and take it properly into account is an error which merits appellate re-evaluation.

However, for the reasons given above, I am of the view that the Hearing Officer did evaluate the arguments on this matter and came to a conclusion relevant to the global assessment, namely that DIGI in the Appellant's and the Respondent's marks had no conceptual meaning, was not descriptive but distinctive, albeit allusive to characteristics of the goods and services in question. Accordingly, he compared the marks on the basis that both were distinctive (albeit allusive to "digital") as regards their use of the DIGI element.

Was the Hearing Officer's assessment in this respect plainly wrong? Answering that question requires consideration of the evidence in some detail and involves addressing a number of other points in issue on the appeal as to the matters which it is said that the Hearing Officer wrongly assessed.

The material relating to the meaning of the term DIGI, particularly as a prefix in the context of a composite mark for digital or computer-related goods or services may be summarised under five heads.

The Hearing Officer correctly recorded that the term DIGI is defined by the web-site www.acronymfinder.com as "digital": Decision paragraph 38. He also said that the term was not defined in any dictionary but this does not appear to accord with the approach previously taken by the Registrar who noted, during the course of prosecution of one of the marks, that DIGI was defined in De Sola's Abbreviations Dictionary as "Digital": see Examination Report for 1577042, dated 8 December 1994. The Registrar's initial reaction was that the term DIGI was not registrable either in Part A or in Part B of the Register under the 1938 Act because DIGI was descriptive and non-distinctive for digital computers. There is no evidence that DIGI features in any other dictionary and the initial objection was overcome. In my judgment, the Hearing Officer was entitled to conclude that this material at least provides a starting point but not a conclusion, for the argument that, when used as a prefix or abbreviation, in a term relating to digital or computer-related goods or service, "DIGI" denotes "digital". Attributing the weight he did to it was not plainly wrong.

More compelling, to my mind than the web-site definitions, is the evidence of use by third parties of a range of marks with DIGI- as a prefix for what appear to be digital or computer-related goods or services. In paragraph 12 of the Decision, the Hearing Officer referred to the evidence of Mr Charlton showing web-sites of undertakings using marks with DIGI- as a prefix. He also referred at paragraph 21 to the responsive evidence from Mr Tibber on behalf of the Respondent. Mr Tibber's evidence shows that it is not possible to draw firm conclusions as to whether the marks revealed by the search are in use, are in use in the United Kingdom or were in use at any relevant date, but there is, nonetheless, a significant number of undertakings which are either using or at least appear to wish to use the prefix DIGI- to denote digital in a number of contexts.

Summarising the position and incorporating the more important of Mr Tibber's observations (from his Statement, at paragraph 13), internet searching disclosed the following uses or potential uses:

Two matters deserve comment. First, all of these uses, or possible uses, appear to relate to goods and services which are likely to be digital or computer-related. Second, all use the prefix DIGI-.

It is, in my judgment, to be inferred from this range of uses and potential uses that numerous undertakings, commercially unrelated to the Appellant or Respondent, either use or have used DIGI- as a prefix as part of a composite term so as to denote that the goods or services have something to do with digital or computer technology. It is not possible to discern from the evidence that there has, in each case, been use in the United Kingdom and none of the uses appear to be specifically in relation to computer connectivity or point of sale hardware, although there are uses broadly in the retail sector and which may be regarded as permitting communications. But that does not detract from the overall position shown by these uses namely that the prefix DIGI- as such is a term which, when used as a prefix to another term in relation to digital or computer-related goods is unlikely to be apt uniquely to identify the goods or services of a given undertaking.

The Hearing Officer touched upon the use by third parties of the DIGI- prefix lightly, in his analysis of the evidence and he did not revisit it in the latter part of his decision (see paragraphs 38-50) or explain why he found this material unpersuasive. In my judgment, notwithstanding imperfection, this material provides real evidence that the prefix DIGI-, used in the context of another sign of the kind described above has a "lesser capacity" to identify digital or computer-related goods or services as coming from a particular undertaking (in the words of Lloyd). However, the question is not whether this court would have found the material compelling but whether the Hearing Officer was entitled to make the determination that he did, in the face of this material (as to which, see further below).

I should say that I am unpersuaded by the Appellant's argument that there is no likelihood of confusion merely because other marks with a DIGI- prefix were used and co-existed. Co-existence does not, of itself, negate the possibility of confusion. Rather, the existence of the other marks rather shows that the common element relied upon (DIGI-as a prefix) is unlikely to be viewed, in the context of the Appellant's marks, as denoting a specific trade origin. There is also reference to another undertaking using the mark DIGI as such which may have pre-dated the Respondent's use but, for present purposes, I regard that material as of lesser relevance. What it does show is that the term DIGI, even otherwise than as a prefix, is not unique as a mark but that does not seem to me to be directly relevant to the analysis required under Article 5(2)(b).

The evidence of use is re-inforced by what has been described as "state of the Register" evidence. Before turning to that material, it is necessary to consider what reference it is appropriate to make to the state of the register. In British Sugar [1996] RPC 281 at 305 Jacob J said:

That was said in the context of a case where absolute grounds were relied on as the basis for refusing registration. However, in the present case, this material is not relied on in support of an argument that a mark which shares characteristics with those already registered should, for that reason, also be registered: the register is not relied on for its precedent value, in my judgment, rightly so. The register is not in this case deployed to show actual use of the marks recorded there either: for that purpose too, it would be of limited, if any, evidential value. Instead, the register is relied upon to show that, as a matter of fact, a significant number of traders have expressed the intention of using (and may be using) the prefix DIGI- as part of a mark in relation to class 9 goods which (one is invited to infer, in the absence of detailed specifications) involve a digital or computer-related element. It does not seem to me to be illegitimate, as a matter of principle, to deploy material of this kind for that limited purpose and, for that purpose, it is of some value as part of a larger body of material.

The material relied on is at exhibit PJC2 to Mr Charlton's statement. It shows a large number of marks registered or applied for in class 9 by a large number of different undertakings, which have DIGI- as a prefix. Some, it is true, are for DIGITAL or variants. But a large number are not. They are for marks employing a DIGI- prefix. They include marks such as DIGIANSWER, DIGIBAR, DIGIBET, DIGIBLADE, DIGIBOOK, DIGIBOX, DIGIUS, DIGICALL, DIGICAM, DIGICARD, DIGICARE, DIGICASH, DIGICAST, DIGICAT and so on. The number of such marks is striking.

It is, in my judgment, to be inferred that the reason that the prefix DIGI- is intended to be used (as these registrations or applications demonstrate) in at least a significant proportion of those marks is so as to indicate that the product or service is or involves digital apparatus, computers or software and that, to a large number of traders, that is what is means. That seems the most logical explanation for (a) the number and (b) the diversity of such marks. This provides further support for the proposition that DIGI- used as a prefix has a limited capacity as such to distinguish goods or services of one undertaking from those of another when used in relation to goods or services involving digital apparatus, computers or software and is common to this general trade.

That said, it does not seem to me that it is an error of principle for the Hearing Officer to attribute only limited weight to this material as, in effect, he did in paragraph 47 of the Decision. To do so is not, in my judgment, plainly wrong.

Next, the Appellant relies on three OHIM decisions.

In CeWe Color AG & Co OHG v. OHIM Joined Cases T-178/03 and T-179/03 (Judgment of 8 September 2005), the issue was whether the word marks "DigiFilm" and "DigiFilmMaker" were registrable, inter alia, for a range of goods in class 9 relating to storage of photograph images in digital form, having regard to Article 7(1)(b) and 7(1)(c) of the CTM Regulation. The examiner rejected the applications. The Third Board of Appeal dismissed the appeals. The applicant appealed further to the CFI. The judgment of the CFI recorded that the applicant for registration "acknowledges that "digi" is a frequent abbreviation of "digital" (para. 15). The CFI said in paragraph 31,

The CFI does not appear to have been in doubt about this decision because it added:

So, the undertaking applying for the mark, the examiner, the members of the Board of Appeal and the members of the CFI were all of the same view as to the meaning of the DIGI- prefix. A similar view was expressed on behalf of the Respondent before the Hearing Officer, in my judgment, correctly: "You can readily see that "Digi" in the context of DigiFilm to the average consumer is going to mean digital film and, accordingly, to be a descriptor." (Transcript, p24).

The Opposition Division of OHIM rejected an opposition to registration of the CTM DIGICOM for a range of goods and services in class 9, based on the earlier German word mark registration DigitOne (Decision No. 1738/2000, 11^th August 2000). The Opposition Division said:

The Opposition Division, referred to Sabel. It analysed the similarities and differences between the marks and concluded that the marks were sufficiently different. In reaching that decision, the Opposition Division took into account that the marks "…enjoy a narrow protection because of their descriptive beginnings."

In Decision 1705/2004 Medion v. Sony KK (26 May 2004), the Opposition Division of OHIM also rejected an opposition to register digipad on the basis of the earlier mark DIGI FLAT. It held that the prefix "DIGI" was descriptive and devoid of distinctive character for the goods involved, referring inter alia to the definition of DIGI in www.acronymfinder.com .

The common theme of these decisions is recognition on the part of the tribunals concerned that the prefix DIGI- was likely to denote digital goods or services related to them. The Hearing Officer treated these decisions shortly saying that they were "not on all fours to the instant case and determined upon their individual circumstances". It is true that they are not. He did not take account of their value as showing that the prefix DIGI- would be regarded as an abbreviation for digital in the context of a composite sign. The Hearing Officer rightly disregarded their precedent effect as such but took no account of the support they provided for the case on the meaning of DIGI- as a prefix that the Appellant was advancing.

The Hearing Officer dismissed the evidence of definition of DIGI saying that "the best that could be found" was an acronym website. But, it might be said, what better evidence is there of the meaning of a sign than actual evidence of the way in which a particular sign is being used and proposed to be used? That is first hand, "real world" material. The Appellant located numerous instances of both actual and intended use of the term in question as an apparently descriptive prefix (or at least common to the trade to use historical phraseology) in computer-related products and services. The Hearing Officer could be said to have discounted that body of evidence of use and intention to use too readily. However, again here, I do not believe that in so doing, he can be said plainly to have gone wrong.

Pausing there, before turning to the prosecution file material, the combination of the definition in the www.acronymfinder.com website, the actual use, the intended used as evidenced by the register and the recognition by the decisions of OHIM taken as a whole provide to my mind persuasive evidence that the prefix DIGI-, when used as part of a composite term in relation to digital or computer related goods or services is of lesser distinctive character and is more likely to be taken to be a reference to the nature of the goods or services and not as an indication of origin. But the fact that I would have found this material more persuasive had I been determining the matter from scratch, is not enough. The question is whether it was open to the Hearing Officer to find that it did not persuade him. In my judgment, it was.

There is, finally, a further category of evidence (or alleged evidence) as to which a fundamental dispute arises over whether it may be taken into account at all. The Appellant contends that the Hearing Officer erred in failing to take account of what it describes as "admissions against interest" made on behalf of the Respondent. The statements in correspondence with the Registrar particularly relied upon are as follows (emphases added):

I should also refer to an earlier letter, dated 3 August 1995, from the Respondent's agents, to which the Appellant does not draw attention and upon which it does not suggest that reliance should be placed, which states (inter alia):

Following that letter, the Registrar waived its objections to Mark No. 1577044 which were based on sections 9 and 10 of the Trade Marks Act 1994 (see letter of 30 August 1995). The Respondent contends that the Hearing Officer was entitled to have little, if any, regard to the correspondence relied on by the Appellant and that, in general, with very limited exceptions, the Registrar should pay no regard to such material.

Whether the court should refer to the genesis of a document in the context of a dispute over its meaning and effect is an issue that arises in numerous areas of law, from legislative interpretation to patent construction. Each area has developed its own particular approaches, influenced by numerous factors, including historical practice.

As regards patents, in general, where a party has sought to rely on the file in the context of a patent dispute, it has been in support of an argument about construction. The judgment of Laddie J in Russell Finex v. Telsonic [2004] EWHC 474, provided a helpful summary of the uncertain state of English law with respect to file wrapper estoppel at least as it applies to patent construction as follows:

Later that year, giving the leading speech of the House of Lords in Kirin-Amgen Inc. v. Hoechst Marion Roussel Ltd [2005] RPC 9 ("Amgen"), Lord Hoffmann expressed the approach taken by the English courts in the context of patent disputes as follows:

This provides a starting point for analysis of the position with respect to the trade mark prosecution file, although different issues of scope of protection arise in patent and trade mark law. The Patents Act 1977 and the EPC on which it is based contain specific provisions relating to the interpretation of patents. Neither the Trade Marks Act 1994 nor the CTM Regulation contain any such provisions. Although that absence may be thought to leave more open the question of the materials that may be considered in a trade mark context than in patents, it would not in my judgment be right to draw that conclusion from the legislative silence in this respect. In patents, the question is largely one of meaning: whether a term used in a claim can properly be said to cover the allegedly infringing product. In trade mark law, the scope of protection only infrequently depends on the precise meaning of technical terms in the specification of goods or services. It more frequently depends on more general matters such as the reputation of the mark. Indeed, that is plain from Canon.

There are other differences as to the kind of matters for which the file may be relevant. However, in my judgment, this court should not adopt an approach to consideration of the prosecution history of a trade mark which differs markedly from that applicable to patents. The approach described by Lord Hoffmann in Amgen articulates a general stance of hostility to arguments made on the basis of material in the file, coupled with a reminder that the assistance which may be gained from consideration of this material is likely to be very limited. That approach is no less appropriate in trade marks than in patents. Without attempting to lay down a hard and fast rule, reference to the prosecution file in a trade mark case should be discouraged, if not absolutely prohibited. There may be exceptional cases in which it provides genuine assistance but those will be rare.

It is also important to bear in mind the difference between admissions and submissions. Admissions may be of specific primary facts, or, to use the terminology of CPR 14.1(1), admissions of the truth of the whole or any part of another party's case. Different weight may be placed on different kinds statements by or on behalf of a proprietor in the file in so far as reference is made to it. At one end of a scale may be clear and unambiguous statements of primary fact. The mere fact that such statements are made in documents included in the file should not, of itself, preclude the tribunal from having regard to them in appropriate cases, although they may well not be evidentially conclusive of an issue. At the other end of the scale are submissions advanced during the course of prosecution of a mark by agents on an applicant's behalf and a fortiori, those which may not have been accepted by the Registrar. In my judgment, reference to submissions of that kind is particularly unlikely to be of assistance in the resolution of disputes which later arise in different contexts. One of the reasons for reluctance to refer to the prosecution file is that, quite often, as in this case, the file will not speak with one voice. The file and individual statements made in it are likely to be inconclusive and to give rise to wasteful collateral arguments about the meaning, effect and correctness of the submissions made.

Ultimately, the Registrar must act on evidence not historical submission and, to echo the alleged words of the Caliph Omar in giving instructions for the burning of the famous library at Alexandria (see Amgen, paragraph 46): if the documents in the prosecution file agree with the evidence, they are not required; if they disagree, they are not desired.

In this case, the Appellant contends that the Respondent represented to the Registrar that:

In my judgment, those representations can best be characterised as submissions or conclusions of legal argument than admissions against interest. They were made in a different context some years ago, it is not clear whether they were accepted by the Registry (and there is at least some reason to suppose that they were not in every respect) they were not specifically directed to the points in issue in this case and the file contains other matter which could be regarded as contrary to the overall position taken in the correspondence in question.

The Hearing Officer said in paragraph 27 of the Decision:

I therefore return to the overall conclusion on the basis of the material referred to above, with a view to assessing whether the Hearing Officer was plainly wrong in the conclusion that he reached in paragraph 46 of the Decision.

In my judgment, there is evidence which would entitle a tribunal to conclude that DIGI- when used as a prefix in the context of a composite mark used in relation to digital or computer-related goods or services is likely to be taken to mean, "digital", even if, taken alone, DIGI was not descriptive. In the context of such a composite mark (as are the Appellant's), there is no real reason to think that it is any less likely to be taken as a "descriptor" than the prefix DIGI- in DigiFilm which was accepted on behalf of the Respondent before the Hearing Officer to denote "digital" in that composite mark. It does not matter that such use has not been specifically in the field of connectivity or indeed point of sale equipment. Moreover, it is not easy to reconcile a finding that the prefix DIGI- has no conceptual meaning in the mark DigiPos with the facts that (i) DIGI is defined as digital in at least one electronic dictionary and was originally regarded as descriptive by the Registry coupled with (ii) evidence of actual use and intention to use shows that DIGI- prefixes have been used or are proposed to be used to convey the concept of digital in a large range of marks for computer-related goods or services by a considerable number of different undertakings and (iii) the meaning of DIGI- has been accepted without question by OHIM and the CFI as being such as to denote "digital".

On the basis of all of the material, I would have concluded that the DIGI- in the Appellant's DigiPos marks did have a conceptual meaning as indicated above. However, with inevitable hesitation, I have come to the conclusion that the approach of the Hearing Officer in this respect cannot be described as plainly wrong. In addition, it might be said that the Hearing Officer's approach to the assessment of descriptiveness (or lesser capacity to distinguish) was supported by a number of matters. First, the absence of evidence of any established definition of DIGI in mainstream dictionaries. Second, the absence of real evidence as to how the mark DigiPos (and in particular the individual elements of it) was actually understood by members of the trade and public although short statements were put in on other matters. Third, it could be said that the Appellant's marks are presented, at least in some of the marks in question, as composed of two elements, DIGI- and –POS, and that the average consumer could regard the first element as a distinctive element, and POS as a descriptor. Fourth, the evidence from the trade as to use and from the file in different contexts provided imperfect proof of how the element would be seen in the marks in question for the goods in question and was, at best, relevant by analogy.

It follows, that the Hearing Officer cannot be said to have taken no account of the potential weakness of the DIGI element or, in consequence, failed to take account of the factors which need to be taken into account. Nor can it be said that he failed entirely to evaluate them, or did so in a way that was plainly erroneous. The Hearing Officer was also entitled to treat the question of weakness as a less important factor (as did the ECJ in L'Oreal).

However, in case this matter goes further and it is concluded either that the Hearing Officer failed to make the evaluation required by the relevant authorities or that, in so doing, he was plainly wrong and that the approach to assessment of plain error above is, notwithstanding Reef and Bud, unduly cautious, I should state, briefly, how I would have approached the section 5(2)(b) assessment had it been appropriate to consider it afresh. In so doing, I address the other main errors alleged to have been made.

The starting point for the assessment is the average consumer. In determining the approach that the average consumer would take, it is necessary to bear in mind the characteristics of the goods and services in question and the circumstances in which they would be selected. It is well established that "the average consumer's level of attention is likely to vary according to the category of goods or services in question" (Lloyd, paragraph 26) and the assessment must be approached with that in mind.

The Hearing Officer held that the purchasers would be reasonably well informed and reasonably circumspect and observant and that the goods and services would not be purchased without some consideration (Decision, paragraph 49). Goods and services of the kind in question, which are, broadly speaking, specialist IT equipment, software and services, are likely, in my judgment, to be selected with considerable care by knowledgeable and discriminating consumers. The warning in Reed against an overly "nanny" approach to comparison of marks has particular force, to my mind, in the context of specialist products which are not selected in the hurly burly of the high street. In that latter context, one would expect the average consumer, even if notionally well informed and reasonably observant, to be more susceptible to confusion, and therefore warrants a more precautionary approach, than that which would be appropriate with business purchasers of IT goods and services.

In paragraph 43, the Hearing Officer referred to the "high degree of consideration" and he is criticised by the Appellant for later adopting an inconsistent approach to the level of discrimination of the average consumer (see Appellant's point (9)). I am not persuaded that his approach in this respect was inconsistent or wrong. He clearly had the relevant evidence in mind in making his assessment.

However, if the Hearing Officer was wrong not to feed into his analysis that the average consumer would be likely to regard the prefix DIGI as having a conceptual meaning, at least in the Appellant's marks, and, when used as part of a composite mark in relation to digital or computer-related goods, would indicate that the goods were digital or computer related, a different view of the approach of the average consumer would have emerged. If that is fed into the assessment, it is harder to see why there should be a likelihood of confusion or that the mark should be taken by the average consumer to be anything other than what prima facie it appears to be: namely a combination of two elements in which the DIGI- prefix refers to the digital nature of the goods or services rather than trade origin. This is a case in which the assumed alertness to small differences set out in Reed is potentially of importance, having regard to the nature of the marks in question and the goods and services for which they are registered and proposed to be registered. But it only comes into play if the element in question is descriptive.

If the lesser capacity of DIGI- as a prefix to distinguish in the context of the Applicant's marks should have been taken into account to a greater extent than the Hearing Officer did and if the assessment is approached on the basis of the notional use of the prior marks (cf. Compass, cited below), having regard to the notional average consumer for the goods and services in question, the better conclusion seems to me to be that there is not a real likelihood of confusion between the respective marks within the meaning of section 5(2)(b). That leaves out of account what was said about the sign DIGI on behalf of the Respondent during prosecution.

I would therefore probably have reached a different overall conclusion to that of the Hearing Officer as to whether section 5(2)(b) was satisfied, had I been considering the matter afresh. Ultimately, the difference arises largely from the more robust (or perhaps optimistic) view I would have taken of the abilities of the average consumer in this particular field to distinguish the marks, having regard to their nature (which should mean that small differences are enough to obviate confusion) and having regard to the nature of the goods and services. But a difference of this kind from a Hearing Officer whose experience at making determinations under section 5(2)(b) far exceeds mine is an unpromising hunting ground for errors, whether plain or of principle.

I turn to the other points made by the Appellant which do not focus so much on the way in which the marks were compared.

It is said that the Hearing Officer erred in concluding that the Appellant's services were for the same specialised purpose as the Respondent's goods (Appellant's Point (5)). The evidence shows that equipment of the kind to which the Respondent's specifications relate have extensive potential and actual retail applications. The specifications of the marks applied for may be said to relate more specifically to point of sale equipment. But the reason for the overlap is, as Mr Vesper's evidence explains, that microcomputer hardware and software which effects communication between digital electronic devices is used in or with point of sale equipment. It is, perhaps, more accurate to say that, although the goods were not necessarily for the same specialized purpose, there was at least a degree of overlap in the specialist purposes.

There is one aspect, however, which, although not an error of principle taken alone, is relevant in the re-evaluation above. The specifications of the Respondent's goods and services are focused on communication between digital electronic devices. This is a general function of use in many different computing applications. The specifications for the Appellant's goods are focused on the specialist retail and store management equipment, software and services. That is relevant both to the global assessment and to the nature of the average consumer. The specialised field of application of the Appellant's goods and services means that the notional assessment does not fall to be conducted notionally across the whole range of potential customers for the Respondent's equipment for communication between digital computers. Rather, the assessment is across the whole field of specialist retail store management equipment and services to which the Appellant's marks relate. These are particularly specialised goods and services involving installations, software and servicing for retail applications. Ultimately, however, I do not see this as an issue of principle in the global assessment or one in which the Hearing Officer fell into error, still less, clear error. The overlap in uses entitled the Hearing Officer to deal with this matter broadly as he did.

It is also said that the Hearing Officer paid too little attention to the fact that there had been parallel trade without confusion. In my judgment, that does not provide an independent basis for criticising his approach or an error of principle.

The weight to be given to evidence of this kind is for the primary tribunal of fact. Moreover, mere absence of confusion in side-by-side trade is sometimes of limited evidential value. In Compass Publishing BV v. Compass Logistics Ltd. [2004] RPC 1, Laddie J pointed out that the comparison required under Article 9(1)(b) of Regulation 40/94 was between notional use on the part of the proprietor and the use in question. Laddie J did not find the evidence of absence of confusion of assistance in undertaking that analysis. Moreover, it is possible for there to be relevant mark-for-sign confusion under section 5(1)(b) but for any actual confusion in the marketplace to be dispelled before it is operative by other circumstances of trade. If that were the case, it would be unlikely for any such confusion to come to light.

It is said by the Appellant that he disregarded the evidence on a false basis: namely because there was inadequate evidence of use by the Respondent of its mark. Although it may fairly be said that the Decision does not explain in detail upon what basis this evidence was disregarded, given that it was common ground that both marks had been in use, upon reading the transcript of the hearing below, that part of his Decision is likely to have been expressed in that way because the Respondent did not press with great vigour a suggestion that it had a significant reputation as a result of use of its marks in the United Kingdom in 2004 in the context of the section 5(4) case which was advanced before the Hearing Officer.

This is not, in my judgment, a case like Fiorelli [2007] RPC 18, upon which the Appellant relies, in which the court concluded that the Hearing Officer had made an error of principle in failing to consider evidence of this kind. To the contrary, in the present case, the Hearing Officer did consider it, but found it unpersuasive. Although he might have expressed himself more fully in discounting it, he was entitled to find that it did not assist the Appellant. Very briefly summarised, the Respondent's evidence shows that the DIGI brand has been established in the UK since 1985 (with DIGI INTERNATIONAL established in the UK since 2001). The Respondent has stated that it has had a significant reputation in the UK for at least 15 years for serial connectivity products. The annual turnover of the Applicant's DigiPos products sold in the UK for the years 2000 to 2004 are said to be as follows (rounded to the nearest £500,000): 2000: £2m; 2001: £3m; 2002: £3.5m; 2003: £6m; 2004: £7m and there has been substantial marketing expenditure. Doubt was sought to be cast on the length of time for which DigiPos has been in use in the responsive evidence. Although, at face value, the figures are such that the absence of confusion coming to light may be considered surprising, if it is really likely that retailers would be confused by the similarity of the trade marks used by Digi and DigiPos (cf. Mr Vesper's statement) other explanations are possible and, for example, the evidence as to precisely when the Appellant commenced significant use of DigiPos as opposed to other marks is not particularly clear.

In my judgment, the Hearing Officer was entitled to conclude, as in effect he did in paragraph 50 of the Decision, that parallel use without confusion was merely one factor in the global assessment and was not, in the present circumstances, a factor of great weight.

Finally, I am not satisfied that the Hearing Officer misconstrued the Respondent's specification of goods or that, even if he did, it mattered at all to the overall decision (Appellant's point (4)). Even if the goods were not identical, they were in my judgment similar and, to that extent, the point advanced in the Respondent's Notice (point (iii)) is justified. The real issue in this case is the approach to assessment of the similarity of the marks, not the similarity of the goods or services.

That addresses (albeit not in the same order as advanced and, in some cases, taking related points together) the main arguments advanced on appeal and in the Respondent's Notice.

I said at the outset that this appeal could be determined by reference to two of the Respondent's marks and one of the Appellant's applications. In the light of the conclusions above, the other marks may be treated briefly.

First, the Hearing Officer was in my judgment entitled to treat the applications for the marks for related services in classes 42 and 37 in the same way as the class 9 applications, given the nature of these goods and services. Second, the Hearing Officer was also correct not to distinguish between the work mark DigiPos (2358343) and the word plus device (2358344). The word DigiPos is the distinctive and dominant feature of the latter mark although the "swish" element does add something. Third, the Respondent's case is no stronger on the basis of any of the other marks, in particular 2118197 even leaving aside the issue of expiry of that mark. That was not suggested either before the Hearing Officer or on this appeal. The word "PORTSERVER" in small letters next to the DIGI in logo form does not, in my judgment, give rise to an increased likelihood of confusion.

I see no basis for a concluding that the case as to confusing similarity is any stronger (or weaker) having regard to DIGI INTERNATIONAL (UK 1577044 and CTM 3477023). This judgment proceeds on the basis that there was genuine use of that mark before 9 July 2004. Taking account of the DIGI INTERNATIONAL CTM makes no real difference to the global assessment: the more distinctive and dominant component of DIGI INTERNATIONAL is the word DIGI and, for practical purposes, the case can be decided on the basis that the earlier marks are for DIGI tout court. Finally, in my judgment, the Hearing Officer was entitled to conclude that the result reached in this judgment is no less applicable to the marks as compared for the respective services in class 37 and 42 and no different case arises on the basis of the other marks.

I will hear argument as to whether any aspect of this decision should be made conditional upon registration at OHIM and if so what steps should be taken in that regard. It is conceivable that, having regard to the overall outcome, it may not matter.

For the reasons given, while I have real doubts about the ultimate outcome of the global assessment in this case, I am unable to conclude that the Hearing Officer was plainly wrong in his approach or in his conclusion or that he made an error of principle. Despite the intelligence and power of Mr Arnold's arguments, I have concluded that the Decision does not fail to take account of or assess the relevant factors. The weight to be given to each of them in the global assessment was a matter for the Hearing Officer in deciding what is, effectively, a "jury" question.

Accordingly, having regard to the very cautious approach which this court must take before overturning a decision of an experienced Hearing Officer concerning the likelihood of confusion under section 5(2)(b) of the Act, this appeal will be dismissed.