Publications > Torres must drown their sorrows over Velasco case

Torres must drown their sorrows over Velasco Case


 More case law on Compound marks – indivisible logical conceptual units


MANSO DE VELASCO  T-259/06 (Spanish and French). Torres v OHIM - Navisa Industrial Vinícola Española, SA, 16 December 2008
December 2008 was not a good month for Miguel Torres SA, the Spanish family-run winery. Six separate CFI judgments involving the company were delivered - every single one against them. At least there would have been no shortage of bottles of wine to reach out to for comfort…

This decision was an ultimate confirmation of rejection by both the Opposition Division of OHIM and by the Board of Appeal of an application by Miguel Torres SA to register the name ‘Manso de Velasco’ in class 33: alcoholic beverages (except beers). The name is that of one of the applicant’s own vineyards in Spain where Cabernet Sauvignon is made. The initial opposition was based on a Spanish word mark VELASCO
(no. 214212) for goods that included vinegar in class 30, beer and must (non-fermented pressed grapes) in class 32, and wines, vermouth and quinquina  (a fortified aperitif containing quinine)  in class 33. The opponent was Navisa Industrial Vinícola Española, SA. The Opposition Division, based on their global appreciation of similarity between goods and between signs, considered that there was a risk of confusion and a risk of association between the marks, leading Spanish consumers to the belief that MANSO DE VELASCO was a particular type of VELASCO wine coming from the same enterprise. The Board of Appeal, on their part, noted that there was moderate visual, phonetic and conceptual similarity between the marks, but given the principle of interdependence by virtue of which a greater degree of similarity between goods (such as in this case) can offset a lesser degree of similarity between goods (and vice-versa), the risk of confusion was significant.

  Applicant’s arguments

On appeal to the CFI the applicant argued that, with the exception of wines, there was only a weak similarity between the goods (and consequently there was a misapplication of the principle of interdependence). With regard to the signs, their argument was that established case-law states that a compound trade mark cannot be regarded as being similar to another trade mark which is identical or similar to one of the components of the compound mark, unless that component forms the dominant element within the overall impression created by the compound mark. In this case, there was a wrong assumption that VELASCO was the dominant part of the applicant’s mark. On the contrary, this name was a common surname (as evidenced by the frequency of appearance in telephone listings, there being about 3000 in Madrid), and was also found often in association with alcoholic brands. The term MANSO, coming at the beginning of the mark, would be what first catches the consumer’s eye and would also be what the mark would be abbreviated to. Visually, phonetically, and even conceptually (Manso de Velasco evoking a family of nobility, Velasco, that of common people) there are significant differences. Taken together with the weak similarity between most of the goods, overall there was no risk of confusion.
CFI’s riposte
Torres’ arguments did not convince the CFI. After having established that the relevant consumer was the average Spanish consumer who was reasonably attentive and informed, it first tackled the question of similarity of goods, reminding the applicant in the process, that class divisions have an administrative function only and that goods could not be considered different merely because they happened to fall into different classes. Thus, and applying established case law, beginning with CANON, whereby in order to assess the similarity of the goods concerned, all the relevant factors which characterise the relationship between those goods should be taken into account, those factors including, inter alia, their nature, their intended purpose, their use and whether they are in competition with each other or are complementary (cf. a later iteration in Case T-85/02 Pedro Díaz v OHIM – Granjas Castelló (CASTILLO) [2003], par. 32), the Court ruled , as one example, that beer and wine are only weakly similar goods, but that vermouth and quinquina were identical to the applicant’s goods in respect of wines, and similar in respect to other alcoholic drinks.
With regard to the comparison of signs, the court referred to Nestlé (C-193/06) – reported in the November 2007 issue of the ITMA Review and still only available in French - which emphasised, echoing MATRATZEN (T-6/01), that although a mark is generally seen by the average consumer as a whole without examining its individual details, in certain circumstances the impression formed by a compound mark may be dominated by one or more of its components. Similarity would be found with another trade mark which is identical or similar to one of its components where ‘that component is likely to dominate by itself the image of that mark which the relevant public keeps in mind with the result that all the other components of the mark are negligible within the overall impression created by it’. (par.37-38)
The court dismissed the applicant’s argument that the Board of Appeal had implicitly considered VELASCO to be the dominant part of the compound mark, and wrongly so, whilst giving MANSO an undeservedly lesser role. MATRAZEN, on which the applicant had based this part of its appeal, had indeed stated (in paragraph 35) that:
“with regard to the assessment of the dominant character of one or more given components of a complex trade mark, account must be taken, in particular, of the intrinsic qualities of each of those components by comparing them with those of other components. In addition, and accessorily, account may be taken of the relative position of the various components within the arrangement of the complex mark.”
Nonetheless, repeating the words of TORRE MUGA (T-247/03), par. 50 (in French and Spanish), of which more below:
“the relative position does not, in all circumstances, make one element of a mark dominant while removing all significance, in the overall impression, from the other elements which are components of that mark.”
In other words, in a compound mark, just because a word comes first, that does not always mean that it is the more significant or dominant for that reason alone. Regarding the commonality of VELASCO, even if this is indeed a popular patronymic denomination, as long as the mark is able to fulfil the function of a trademark, this fact alone of being a common name is irrelevant. As NICHOLS, C-404/02, par 30, puts it:
“In the same way as a term used in everyday language, a common surname may serve the trade mark function of indicating origin and therefore distinguish the products or services concerned where it is not subject to a ground of refusal of registration other than the one referred to in Article 3(1)(b) of Directive 89/104, such as, for example, the generic or descriptive character of the mark or the existence of an earlier right.”
 The result of this is that there is no ground for insisting that MANSO is the distinctive and dominant part of the compound mark.
The CFI continued to analyse the comparison of the marks made by the Board of Appeal, taking the visual, phonetic and conceptual examinations in turn.
The Board had incisively concluded that a visual comparison between MANSO DE VELASCO and VELASCO had revealed a ‘50%’ degree of similarity [as it happens, 7 letters as against 14], the junior mark being entirely reproduced within the applicant’s mark. The court, with its further insight, concludes (in my informal translation) that:
“from the visual point of view, the conflicting marks present at the same time elements of similarity and elements which differ. Indeed, if it is true that the applicant’s mark has a number of distinct elements, namely ‘manso’, ‘de’ and ‘velasco’, and is longer than the earlier national mark, it has to be emphasised that the ‘velasco’ element is identical in the two marks in conflict. This verbal element is not only common to the two marks, but equally constitutes the only element of the earlier national mark and, furthermore, the earlier national mark as such. Also, it is necessary to note that the elements ‘manso’, and ‘de’ in the application do not necessarily confer visual distinctiveness such that the relevant public seeing this mark would have an overall impression that is totally different from that which it would have from the earlier national mark.”(par.13)
The conclusion: contrary to what the applicant asserts, the Board of Appeal committed no error in its analysis of the visual similarity between the marks.
Unfortunately for the applicant, exactly the same conclusion was reached in terms of the phonetic comparison, the marks, unsurprisingly, being deemed to be ‘50%’ similar. However, the fact that the identical part is not at the front of the compound mark does not mean, for that reason, that the marks cannot be deemed similar.
“…the fact that the common element “velasco” is not situated at the beginning of the applicant’s mark is not decisive to the extent that, if it is true that the first part of word marks are able to attract the attention of consumers more than the parts of the marks that follow, this consideration does not hold good in all cases” (par.58)
Finally, regarding the conceptual comparison, again the court held that the conclusion of the Board of Appeal was not mistaken. If the word MANSO is taken as a generic word, its etymological meaning in Spanish is that of a plot or parcel of land. Under this hypothesis, the surname VELASCO would be conceptually related to MANSO, as land belonging to the family of Velasco. If, on the other hand ‘Manso de Velasco’ was taken to indicate, as the applicant argued, a more noble form of the Velasco family, this in itself would be insufficient to differentiate the terms such as to avoid confusion. Either way, there was a conceptual link between the marks.
The overall conclusion was that the applicant’s mark viewed globally suggested to the average Spanish consumer an extension of the range of Velasco wines or indicated an economic link. The court dismissed the action.
The TORRES opposition cases
TORRE DE FRIAS T-285/06 Torres v OHIM – Bodegas Cándido
TORRE ALBÉNIZ T-287/06 Torres v OHIM – Bodegas Peñalba López
TORRE DE GAZATE T-286/06 Torres v OHIM – Vinícola de Tomelloso
TORRE DE BENÍTEZ T-16/07. Torres v OHIM – Sociedad Cooperatíva del Campo San Ginés
TORRE GALATEA T-8/07. Torres v OHIM – Gala-Salvador Dalí
All cases are in French and Spanish, although Albéniz is also in English. All decisions were issued on 18 December 2008.
Unlike the case just considered, Manso de Velasco, where the appellant (applicant of the action before the court) was the applicant, in these cases the appellant, Miguel Torres SA, was the opponent.
The first opposition, Torre de Frias, was filed in November 1998, the last of the series – Torre Galatea – in August 2003. In two cases, Albéniz and Galatea, the opponent was successful at first instance, the decisions of which were appealed by the other parties. The Second Board of Appeal adjudicated in each of the five cases (where all original applicants appear to be independent wineries), but the decisions were deferred pending the outcome of another parallel case that was to be heard by the Court of First Instance, the case of TORRE MUGA T-247/03 (mentioned above). The judgement was issued on 11 July 2006 (which was unfavourable towards Miguel Torres) and because of the similarity of the cases where an opposition had been filed against applications each beginning with the word TORRE, followed by either one or two more words (and with an additional device or figurative element in the case of Albéniz and Galatea), the reasoning of the CFI would also be relevant and applicable to the cases now being considered by the Board of Appeal.
Unsurprisingly the Board came to the same conclusions, the decisions in each case being delivered either days or at most a few months following the TORRE MUGA decision, differing but in a few details. Appeals to the Court of First Instance were filed in all instances within a short period of each other. The judgement of the court in all cases, which appeared on the same day in December of 2008, were expressed in an almost identical way. Furthermore, they were almost identical to the judgement of Torre Muga. All found in favour of the original applicants of the opposed marks.    
As this is not a review of Torre Muga as such, the analysis that follows will take Torre de Frias as the model case, with variations according to the other appeals noted.

 TORRE DE FRIAS T-285/06 Torres v OHIM – Bodegas Cándido 

The application for the mark ‘Torre de Frias’ in class 33 ‘wine’ was made on 4th July 1996 (this particular prosecution having run therefore for almost 12½ years!), and was opposed by Miguel Torres based on several of its earlier national TORRES and LAS TORRES word marks in class 33. The opposed mark already existed as a Spanish registration (this too had been opposed by Torres in the Spanish courts, but unsuccessfully).
The opponent’s arguments that run in this case are repeated in all the cases under review since the reasons for rejecting the oppositions given by the Second Board of Appeal were practically verbatim the same (except for the particular visual and phonetic analysis of the different marks).
In its oppositions, Miguel Torres SA relied on various arguments: the notoriety of its TORRES mark to indicate greater distinctiveness; the similarity taken overall of the marks; the dominance within each opposed mark of the word TORRE; and (in three cases) the fact that their earlier marks constituted a family or series, thereby increasing the likelihood of confusion.
The Board of Appeal, on the other hand, found for the applicants of the marks. It stated: (quoting from Torre de Frias but editing (in square brackets) to be inclusive of all cases under review): 
“the fact that the word ‘torre’ is placed first in the mark applied for did not make that word dominant as compared with the other elements of that mark [rendering the other elements negligible in the overall impression produced by the mark]. The Board of Appeal also held that while the fact that an earlier mark is well known may strengthen its distinctiveness, that fact does not, in all circumstances, make more distinctive an identical or similar element in a compound mark to the extent that that element becomes the dominant component in the overall impression produced by the compound mark, where that element forms with the other components an indivisible logical and conceptual unit which can be entirely distinguished from and cannot be associated with the earlier mark which is well known. Next, the Board of Appeal held that the overall impression produced by the marks at issue was different, that the phonetic coincidence of the element ‘torre’ was offset by the other elements of the sign [namely, ‘de frias’, ‘albéniz’, ‘de azate’, ‘de benítez’ & ‘TG + galatea’) and that, conceptually, the perception of the relevant public will vary according to whether or not it is familiar with the meaning of the word ‘torre’. For those who understand the meaning of that word [esp. Spaniards, Portuguese and Italians since the word means ‘tower’ in their languages] there will be a conceptual difference, while for those who do not perceive any meaning in the word ‘torre’, the conceptual similarity is of little relevance. [Finally, to the extent that the word ‘torre’ is not the dominant element in the overall impression made by the applicant’s mark, the fact of having this word in the mark could not lead to the conclusion that it would be perceived as being part of a family of marks composed of the term ‘torres’ accompanied by other elements].” (par 13)
Therefore, it was held that as the condition of similarity between the marks at issue was not satisfied, there was no likelihood of confusion between them.
Arguments of the applicant
In support of its action the applicant before the court relied on a single plea in law, alleging infringement of article 8(1)(b) of regulation No 40/94
It claimed that the Board of Appeal made two errors when applying the rules set out by the case-law relating to the interpretation of that provision. First, the Board took no account of the fact that the earlier marks were well known when it determined which element in the mark applied for was dominant and when it examined whether there was a likelihood of confusion. Second, the Board did not base that examination on the perception of the relevant public (the average consumer in the European Union) but solely on the perception of a minority group of consumers (Spanish, Italian and Portuguese consumers).
A third error was claimed in the cases of ‘Albéniz’, ‘de Benítez’ & ‘TG Torre Galatea’, this being that the Board took no account of the fact that the applicant’s marks were a ‘family of marks’ or a ‘series of marks’.
The plea in law was based on three arguments.
  1. The fact that the TORRES sign is well known throughout the European Union adds to the distinctiveness of the element which is common to the signs at issue and which is thereby dominant.
  2. Having regard to their distinctive and dominant elements, the signs at issue can be seen to be phonetically and visually similar.
  3. The fact that the TORRES sign is well known, added to the fact that the goods are identical, strengthens the ius prohibendi of [right to prohibit] the earlier marks in the assessment of the likelihood of confusion.
In the ‘Albéniz’, ‘de Benítez’ & ‘TG Torre Galatea’ decisions a fourth argument was added: the likelihood of confusion is increased by the existence of a ‘family of marks’ containing the word ‘torres’ (and the graphic element of one or more towers in Albéniz).
Regarding argument one, what the applicant was in effect claiming was that there is a sort of ‘spill-over’ of distinctiveness and dominance from TORRES, through its status of being well-known (having first traded in 1800, exporting to 85 countries of the world, winning many international prizes etc), to the word TORRE of the compound opposed mark. This doctrine appears to be current in Spanish jurisprudence and academic writing.
Miguel Torres also argued that the Board of Appeal’s assertion that the mark applied for is ‘an indivisible logical and conceptual unit’ is valid solely in Spain, Portugal and Italy, countries where the meaning of the marks at issue is understood or may be understood. In other countries where the TORRES marks are also registered the word ‘torre’ and the other elements of the compound marks have no specific meaning and only with difficulty therefore can they be regarded as forming a unique logical and conceptual unit.
In addition, they pointed out that Torre was but the singular form of Torres, thus the marks were effectively identical, meaning tower(s).
The positioning of the word Torre, though not decisive, will in this case lead consumers to pay particular attention to it and remember the mark in abbreviated form by it.
Regarding similarity, the applicant argued that, as according to the Court’s case-law, where one of the two words which alone constitute a word mark is identical, both visually and aurally, to the single word which constitutes an earlier word mark, and where those words, taken together or in isolation, have no conceptual meaning for the public concerned, the marks at issue, each considered as a whole, are normally to be regarded as similar within the meaning of Article 8(1)(b) of Regulation No 40/94. Add to this the extra distinctiveness of the mark due to it being well-known, then there is even greater likelihood of confusion.
Regarding the argument of its marks constituting a family of marks, the applicant claimed that it fulfilled the conditions necessary to permit recognition of the effect of a family of marks on the likelihood of confusion as per BAINBRIDGE (T-194/03 Il Ponte Finanziaria v OHIM), namely, first, the use and presence on the market of all the marks belonging to the family on which the opposition is based and, second, the existence in the mark applied for of the fundamental element which identifies the family of marks.
 Findings of the Court 

As occurred with the first case reviewed in this article, the Court referred constantly to the TORRE MUGA case, as well as the usual recitations of general principles found in established case law.
Is Torre the dominant component of the compound mark Torre de Frias?
Regarding the issue of dominance, and repeating the point already made above, the Court states:
“according to established case-law, a compound trade mark cannot be regarded as being similar to another trade mark which is identical or similar to one of the components of the compound mark, unless that component forms the dominant element within the overall impression created by the compound mark. That is the case where that component is likely to dominate, by itself, the image of that mark which the relevant public keeps in mind, with the result that all the other components of the mark are negligible within the overall impression created by it. (see Torre Muga, paragraph 46 and case-law cited).  However, that approach does not amount to taking into consideration only one component of a complex trade mark and comparing it with another mark. On the contrary, such a comparison must be made by examining the marks in question, each considered as a whole (see Torre Muga, paragraph 47 and case-law cited).” (par. 45-46)  Although account may be taken of the relative position of the various components within the arrangement of a compound mark to assess the dominance of one particular component, in this case the Court ruled:
“the fact that the word ‘torre’ is placed first in the mark applied for does not mean that the word placed second, namely ‘de frias’ [‘albéniz’, ‘de azate’, ‘de benítez’ & ‘TG + galatea’], is insignificant in the overall impression produced by that mark, in particular phonetically and conceptually. On the contrary, the decisive factor in the distinctiveness of the mark applied for is the combination of the words ‘torre’ and ‘de frias’ [et al], which form, together, a unique logical and conceptual unit. Consequently, even if the earlier mark was well known, the existence of the other components of the mark applied for cannot be ignored.”
It added (par. 53-54):
“There is no basis for the proposition that the average consumer who is reasonably well informed and reasonably observant and circumspect will systematically disregard the second part of the word element of a mark to the point of remembering only the first part, as claimed by the applicant.
 That is more particularly true in the alcoholic beverages sector, where consumers are accustomed to the practice of goods being designated by marks which include several word elements (Torre Muga, paragraphs 52 and 53).”
Thus, concluding the analysis on the question of dominance:
“it must be held, for the purposes of the comparison of the signs at issue, that the Board of Appeal did not err in law by declining to hold that the word ‘torre’ was the dominant element in the impression produced by the mark applied for.” (par. 57)
On visual similarity, the Court concludes, after noting that the element ‘torre’ is not identical to ‘torres’, and that visually the marks give a different overall impression, the mark applied for being longer, with (in the case of ‘de frias’), three individual components, the Board of Appeal did not err in determining that the differences as perceived by the consumer are greater than the similarities.
Phonetically, there are likewise differences in the number of syllables to be pronounced and in the number of words, these differences outweighing the similarities when considered overall. The Board of appeal did not err in coming to this view.
Conceptually, the court held that a distinction must be made between the impression produced by the signs at issue in countries where consumers know the meaning of the word ‘torre’, namely Spain, Italy and Portugal, and the impression produced in other countries. It then states:
“As regards countries where consumers know the meaning of the word ‘torre’, the Board of Appeal has not erred by holding that the ideas suggested by the marks in conflict are different. Whereas the word ‘torres’ evokes, at least for the Spanish public, the idea of towers, in the plural, the second part of the mark applied for evokes the idea of a specific tower. While there is some degree of similarity, the frequent use of the word ‘torre’ to designate the goods in question in Spain, Italy and Portugal will none the less cause consumers in those countries not to overlook the element ‘de frias’ [et al]  in the mark applied for and, consequently, to be more alert to the conceptual difference between the signs at issue.” (par. 65)
 On the other hand, in countries where the meaning of the word ‘torre’ is not known, the conceptual similarity is of limited relevance, as is pointed out by both the applicant and OHIM… the fact that the conceptual comparison of the signs at issue is of limited relevance in the majority of European Union countries does not preclude account being taken of the conceptual differences between the signs at issue in countries where the meaning of that word is known by the public.” (par 66-67)
The overall conclusion on the comparison of signs:
“In light of the foregoing, it must be held that the Board of Appeal was right to conclude that the differences between the signs at issue prevailed, visually and phonetically, for all the consumers concerned, and, conceptually, for the Spanish, Italian and Portuguese public. It is clear from the comparison of the signs at issue that the overall impression produced by them is different. The mere coincidence of the first five letters of the earlier sign and the element ‘torre’ in the mark applied for does not alter the finding that, in the overall impression, the differences between the signs are greater than the weak points of similarity.” (par. 68)
Coming, then to the overall assessment of likelihood of confusion, the Court held that in light of the differences between the signs at issue, the Board of Appeal was correct to hold that there was no likelihood of confusion between them, notwithstanding the fact that the goods concerned were identical and, importantly, answering the argument concerning the well-known status of the opponents mark:
“The fact that the TORRES sign is well known in all the relevant countries cannot call in question that conclusion. While there is established case-law that the more distinctive the earlier mark, the greater will be the likelihood of confusion, it must be emphasised that a likelihood of confusion presupposes that the signs are identical or similar. The fact that a mark is well known is one factor which, far from giving rise, in itself, to a likelihood of confusion, must be taken into account when determining whether the similarity between the signs or between the goods and services is sufficient to give rise to a likelihood of confusion (see Torre Muga, paragraph 72 and case-law cited).” (par. 70)
  Family of marks

Finally, regarding the argument based on the alleged existence of a ‘family of marks’, found in Albéniz’, ‘de Benítez’ & ‘TG Torre Galatea’, the court again upheld the Board of Appeal’s rejection of its application, giving as its reason the fact that there was no similarity between the marks at issue.
It may be useful to remind ourselves of the case law on family marks as summarised in par. 79-81 of Albéniz:
“As regards the applicant’s argument that its earlier marks constitute a ‘family of marks’ or a ‘series of marks’, which can increase the likelihood of confusion with the mark applied for, such a possibility was recognised in BAINBRIDGE and confirmed in Case C‑234/06 P Il Ponte Finanziaria [2007] ECR I‑7333.
According to that case-law, there can be said to be a ‘series or a ‘family’ of marks when either those earlier marks reproduce in full the same distinctive element with the addition of a graphic or word element differentiating them from one another, or when they are characterised by the repetition of the same prefix or suffix taken from an original mark (BAINBRIDGE, paragraph 123). In such circumstances, a likelihood of confusion may be created by the possibility of association between the trade mark applied for and the earlier marks forming part of the series where the trade mark applied for displays such similarities to those marks as might lead the consumer to believe that it forms part of that same series and therefore that the goods covered by it have the same commercial origin as those covered by the earlier marks, or a related origin. Such a likelihood of association between the trade mark applied for and the earlier marks in a series, which could give rise to confusion as to the commercial origin of the goods identified by the signs at issue, may exist even where the comparison between the trade mark applied for and the earlier marks, each taken individually, does not prove the existence of a likelihood of direct confusion (BAINBRIDGE, paragraph 124). When there is a ‘family’ or a ‘series’ of trade marks, the likelihood of confusion results more specifically from the possibility that the consumer may be mistaken as to the provenance or origin of goods or services covered by the trade mark applied for and considers erroneously that that trade mark is part of that family or series of marks (Il Ponte Finanziaria, paragraph 63).
However, according to the above case-law, the likelihood of confusion attaching to the existence of a family of earlier marks can be pleaded only if both of two conditions are satisfied. First, the earlier marks forming part of the ‘family’ or ‘series’ must be present on the market. Secondly, the trade mark applied for must not only be similar to the marks belonging to the series, but also display characteristics capable of associating it with the series. That might not be the case, for example, where the element common to the earlier serial marks is used in the trade mark applied for either in a different position from that in which it usually appears in the marks belonging to the series or with a different semantic content (BAINBRIDGE, paragraphs 125 to 127).”
The court dismissed each of the five TORRES actions (as well as the MANSO DE VELASCO action), and ordered Miguel Torres SA to pay the costs (ouch…)

If the five TORRES cases reviewed here were indeed decided correctly, one wonders why MANSO DE VELASCO could not also have been considered ‘an indivisible logical and conceptual unit’ with the opposite outcome. It is indeed curious how, given the identity of goods, MANSO DE VELASCO and VELASCO were found to be confusingly similar, yet the highly distinctive TORRES and TORRE DE FRIAS (and others) were found to be quite different, given especially the CANON principle of interdependence, recently reaffirmed in CFI T-357/07 FOCUS MILENIUM, par. 42, that where the goods and services in question are identical, even a low degree of similarity between the signs at issue may be sufficient to give rise to a likelihood of confusion.
This article first appeared in the ITMA Review, March 2009