Advertising in a Foreign Language or the Consumers' Native Language?

ABSTRACT

Research on foreign languages in advertising stresses the role of the match between product and language. Ads with a match are more effective; however, their effectiveness has not been compared to that of ads in the consumers’ native language, which are not restricted to products with a match, but are used for a range of products. An experiment tested whether native ads were more effective than those utilizing foreign languages. We also took consumers’ foreign language proficiency into account. Results suggest that foreign language is not the only key to success in advertising and effectiveness might also depend on the foreign language used.

KEYWORDS:

• Attitude
• congruence
• effectiveness
• foreign language
• native advertisement
• print advertising

Introduction

Foreign languages are frequently used in international advertising. English, in particular, is very popular in non-English speaking countries. However, other languages are also frequently used, mainly in advertisements in which the product advertised is congruent with the language, or the country in which the language is spoken. An advertisement for perfume that uses French as a foreign language is an example of an advertisement with a congruent relation between the product and the foreign language used. If the same advertisement would use Russian as foreign language, the relation between foreign language and product is incongruent because Russia is not usually associated with perfume. There are quite a few studies (see below) on the effectiveness of the congruent vs. incongruent relation between the product advertised and the foreign language used. The overall conclusion these studies draw is that congruent advertisements are more effective than those in which language or country and the product are incongruent. However, despite the frequent use of foreign languages other than English in advertising, a large part of advertisements in one country is in the consumer's native language. Naturally, those advertisements are often incongruent with the product they advertise, especially if the product is of foreign origin. Few studies actually compared ads using foreign languages (other than English) to ads in the consumers’ native language to see whether the use of foreign languages is actually more effective than the use of consumers’ native language in advertising. In the present study, we compare the two types of advertisements.

Background

A number of studies have shown that advertisements that make use of foreign languages are very common in countries all over the world. Griffin (1997, 13) found that 88% of the ads in a variety of Polish magazines he analyzed contained English words. Gerritsen et al. (2000) showed that 33% of the TV commercials shown in a week's time on one of the Dutch nationwide TV stations contained English. Piller (2000) even found that 70% of the German TV commercials she analyzed were multi-lingual. Gerritsen et al. (2007) showed that in the Netherlands 81%, in Germany 57%, in France 63%, in Belgium 73%, and in Spain 75% of the ads in glossy women's magazine were (partly) in English. Though the figures of the different studies vary, depending on the definition of what is considered English and the part of the ads analyzed, English is by far the most frequently used foreign language (Bhatia 1992; Piller 2003). With an increasing globalization, these figures will probably have increased in the past years. Though more sparingly, other foreign languages are also used in advertising, 8% of the advertisements in Piller's (2000) study contained French and 6% Italian.

In the literature, there seems to be a consensus that the foremost reason for the use of foreign languages in advertising is to attract consumers’ attention (cf. Petrof 1990; Domzal, Hunt, and Kernan 1995; Gerritsen et al. 2000; Piller 2001; Hornikx and Starren 2008). Domzal, Hunt, and Kernan (1995, 100) argued that the use of foreign languages in advertising functions as “attention getting device” because the use of a foreign language deviates from the norm and is unexpected. This makes the advertisement stand out. Domzal, Hunt, and Kernan (1995) and Piller (2001) also claimed that pro-cessing a foreign language takes more cognitive effort and as a consequence the consumer spends more time on the foreign language elements in order to be able to understand them. Processing a foreign language requires more cognitive effort to understand the message and a longer processing time probably affects the ability to remember such an ad. For the attention getting function of a foreign language in advertising, it does not seem to matter which foreign language is used. However, if English is frequently used in a country where it is not a first language the attention getting function might decrease because habituation might occur. On the other hand, English is most frequently used because it is assumed to be a language that is spoken or at least understood by consumers in large parts of the world.

However, Kelly-Holmes (2000) argued that an understanding of the foreign words might not be necessary for an advertisement to be effective because foreign languages are not merely used to directly communicate about a particular product via the actual meaning of the individual words or utterance, but to employ associations individual consumers have with a particular language. In that case, all that is needed is the recognition of a language as the language spoken in a country X. This means that consumers do not need to understand that Vorsprung durch Technik, a slogan Audi has been using for more than 30 years (Rice-Oxley 2012), means “progress through technology,” but simply have to recognize that the slogan is German. If recognized, German triggers all kinds of associations (intended are usually only the positive ones) that are transferred to the product advertised. Kelly-Holmes (2000) therefore talks about the symbolic function of foreign language use in advertising. She stated (Kelly-Holmes 2000, 71) that “[i]n intercultural advertising, however, [a foreign] language now seems to be used primarily for its symbolic value, while the communicative or utility value of the particular words has come to be obscured or mystified through the process of fetishization to the point where it becomes irrelevant.” In this respect, the use of English differs from that of other languages because the use of the English language is in the majority of cases (there are exceptions, e.g., for advertisements for tea) not associated with the country in which English is spoken as a native language, such as the United Kingdom, the United States, India, or Australia. English typically evokes ideas, such as modernity, youthfulness, and global mindedness (Piller 2003). This means that the English language can not only be used for products with an origin in an English speaking country, but also for a range of products that are not typically associated with a particular country of origin (COO). This means that the English language is suited for standardized advertising campaigns for a large variety of products from many different countries. In addition, English is very popular because it is assumed to be widely known and spoken all over the world. This provides internationally operating companies with an economic advantage because they do not have to adapt their campaigns to the languages of their different markets, but use an English version of their campaign. Whether these so-called standardized campaigns are successful or not is a matter of much debate (for an overview see Melewar and Vemmervik 2004).

The use of other foreign languages, such as French, German, or Spanish, is to a lesser extent a matter of standardization (this is probably why they are less frequently used) because languages other than English are typically used to evoke associations that go hand in hand with the respective language or the country it is spoken (Haarmann 1989; Ray, Ryder, and Scott 1991; Gerritsen et al. 2000; Piller 2003; Hornikx, van Meurs, and Hof 2013). The associations related to foreign languages other than English are more specific and more closely related to the stereotypes associated with a particular country and the COO of a particular product. This means that products advertised using a foreign language other than English are typically products that come from a particular region or products whose characteristics are somehow related to a stereotype held by the intended target group about a particular country. So, the slogan in the example above (Vorsprung durch Technik) functions as COO marker in such an advertisement. Hornikx and van Meurs (2017) were able to show that the use of a foreign language in advertising has the same effect as that of other COO markers (such as “Made in Germany”). However, since the associations are based on stereotypes, they might be different for individual cultures since stereotypes are not universal. Gürhan-Canli and Maheswaran (2000) showed that associations with a particular country can vary for different cultures. Despite the fact that the associations might not be the same for all countries, what they usually have in common is that the intended associations utilized in advertising are positive, and are expected to be mirrored onto the advertised product. Given that these associations are related to a particular country, only products that are related to this country can be advertised making use of these associations. Research (Verlegh and Steenkamp 1999; Verlegh, Steenkamp, and Meulenberg 2005; Usunier and Cestre 2007; Hornikx, van Meurs, and Hof 2013) has also shown that the relation between the product and the COO (or the foreign language spoken in this particular country which functions as COO marker (see Hornikx and van Meurs 2017)) used for such campaigns is crucial. Usunier and Cestre (2007) looked at the congruency relation between the product, on the one hand, and the COO and the associated language, on the other. In an experiment with products that are typically associated with a particular country (e.g., cosmetics and France, or vodka and Russia), they manipulated the COO of those different products. They were able to show that consumers were more willing to buy a particular product after having seen a congruent advertisement than after having seen an incongruent version (cosmetics and Mexico). Usunier and Cestre (2007) concluded that congruency between the product and the country of origin is very important for the success of an advertising campaign utilizing a foreign language. Hornikx, van Meurs, and Hof (2013) came to similar conclusions in their study of print advertisements in the Netherlands. They compared the effectiveness of three advertisements for products with a clear congruency between product – COO (France – wine, German – sausages, Spain – oranges) to three advertisements with an incongruent product – COO relation (France – beer, Germany – olive oil, Spain – washing machine) in which the slogans of the advertisements were in the respective foreign language. For all three languages, the advertisements with a congruent relation between product and foreign language used were more effective than their incongruent counter parts. Hornikx, van Meurs, and Hof (2013) concluded that congruency between product and country of origin is vital for the effective use of foreign languages (other than English) in advertising.

What is typical of studies like the ones discussed above is that they compare the use of a congruent foreign language in an advertisement to the use of an incongruent, but also foreign language, but they do not compare the use of these foreign languages to advertisements in the consumers’ native language. However, looking at advertisements in different countries, advertisements in the native languages of the consumer are still common even for products that are typically associated with a foreign COO. In the Netherlands, pasta and espresso are typically associated with Italy, but nonetheless Honig, a Dutch brand that belongs to the Heinz corporation, uses all Dutch advertisements to promote their pasta. Douwe Egberts, one of the Sara Lee coffee brands, uses an all Dutch advertisement to promote their L' OR espresso in the Netherlands. In those advertisements, the relation between product and COO is either not stated, or assumed to be Dutch, since the Dutch language is used. In the latter case, the COO is incongruent with the products that are commonly assumed to originate from Italy. What seems to be more important here is that the consumer is addressed in his or her native language. Previous research (Verlegh and Steenkamp 1999; Verlegh, Steenkamp, and Meulenberg 2005; Usunier and Cestre 2007; Hornikx, van Meurs, and Hof 2013) has shown that the congruency relation is important for the effectiveness of advertisements employing foreign languages other than English, but is this also the case when these ads are compared to advertisements in the consumers’ native language? Despite the positive effect of a foreign language and the associations foreign languages evoke, consumers might prefer to be addressed in their native language. There are a few studies that compare the use of English to that of the consumer's native language (Krishna and Ahluwalia 2008; Hornikx, van Meurs, and de Boer 2010; Micu and Coulter 2010; Ahn, Ferle, and Lee 2017). Krishna and Ahluwalia (2008) compared the effectiveness of advertisements in Hindi, English, and bilingual English–Hindi advertisements for different types of products (necessity vs. luxury) from local and multinational companies. However, in their study participants were native speakers of both of these languages and therefore the results are difficult to compare to those of studies that look at advertisements that utilize foreign languages because those Indian consumers are used to communicate in both languages and the associations evoked by the two languages are not related to the COO, but much rather related to social stereotypes (Krishna and Ahluwalia 2008, 693). Comparing advertisements using English to those using Romanian with respect to the attitude towards the advertisement among Romanian consumers, Micu and Coulter (2010) were able to show that English advertisements evoke a more positive attitude towards the advertisement than ads in the native language if the product advertised is from a multinational company. For products from local companies they found no difference between ads that made use of a foreign language and those in a native language. Again, the associations that play a role in this study are not related to the COO of the products advertised, Micu and Coulter (2010) assumed that positive associations that are related to the English language “such as globalness, prestige, and being cosmopolitan” (2010:79) are responsible for the effect of English on consumer attitudes. In a study by Ahn, Ferle, and Lee (2017) who compared Korean–English, English–Korean and transliterated Korean–English advertisements, associations with a COO did not play a role either. Their study was mainly focused on issues related to the fact that Korean and English use different alphabets that makes code-switching between these languages more complex than for languages using the same alphabet. Hornikx, van Meurs, and de Boer's (2010) study compared the use of a foreign language to the use of the native language of the consumer in advertising. In an experiment, they compared English slogans in car advertisements to Dutch native slogans. Again COO association did not play a role in their study. However, the English slogans they used in their study varied in their degree of complexity. They not only measured the appreciation of the slogans and the preference for the native or English slogan, but also tested the effect of comprehension of the slogan used in the advertisements. Their results showed that the more complex the English slogan the higher the preference for the native Dutch slogan. More complex English slogans were less well understood than simpler ones. These results led Hornikx, van Meurs, and de Boer (2010) to assume that the preference for a native slogan is related to the comprehension of the English slogan or a lack of understanding. If respondents did not understand the English slogan they preferred the Dutch one. This suggests that comprehension of the foreign language influences the effectiveness of foreign languages used in advertising. These results are interesting in the context of the present study under the assumption that consumers’ proficiency in foreign languages other than English is lower (see European-Commission 2012). Consumers might prefer an advertisement in their native language if they do not understand the foreign language slogan in the advertisement. In that case, the understanding of the advertisement seems to be more important than simply triggering the association related to the COO as suggested by Kelly-Holmes (2000). Apart from an abundant sample of print advertisements and TV commercials illustrating the use of foreign languages described in Kelly-Holmes's (2005) review study, there seems to be no empirical evidence for the effectiveness of the symbolic function of foreign languages used in advertising compared to the use of the consumer's native language. We would expect advertisements that make use of a foreign language to be more effective than the native ones irrespective of whether consumers understand the foreign language or not because of the positive associations triggered by the foreign languages.

The aim of the present study was twofold. The literature has shown that foreign language use in advertising is most effective if product and foreign language (or COO) used is congruent (see above). Our first aim is to establish whether advertisements with product-language (COO) match are also more effective than advertisements for the same products in Dutch consumers' native language. So the question is whether the native language of the consumer (which is not congruent with the product) is more effective than an ad in a foreign language with a product language congruency. The second aim was to find out whether the understanding of the foreign language has a positive influence on the effectiveness of the use of a foreign language. Given that German advertisements are understood by the majority of Dutch native speakers (given that according to the European-Commission (2012) 71% of the Dutch speak German well enough to participate in a German conversation), whereas advertisements in Spanish are less well understood (Spanish is rarely learned at school and Spanish does not appear in the Eurobarometer 386 report as one of the foreign languages most frequently spoken in the Netherlands) consequently, we assumed that Dutch consumers are able to understand German, but not Spanish. In order to control for this assumption, our respondents provided a self-assessed measure of proficiency for the two languages used.

Method

In order to test whether print advertisements that make use of foreign languages and a language-product match are more effective than Dutch print advertisements, we carried out an experiment. Participants were shown different print advertisements and had to fill in a questionnaire.

Participants

In total 191 participants took part in this experiment. Fifteen participants were excluded from the analyses because their mother tongue was not Dutch. Of the remaining 176 participants, 128 (72.7%) were women and 48 (27.3%) were men. Their mean age was 23.51 with a minimum age of 18 and a maximum age of 61. Most of the participants were students at the Radboud University Nijmegen, but there were also participants with various other backgrounds (mostly highly educated). Participants indicated their proficiency in German and Spanish on a 5-point Likert scale (1 = poor, 5 = very good). Their mean proficiency for German was M = 3.24 (SD = 0.78) and M = 1.78 (SD = 1.05) for Spanish. A repeated measure analysis with language as factor showed a significant difference between the self-assessed proficiency of German and Spanish F (1,175) = 214.82, p < .001, η² = 0.55. Proficiency in German was significantly higher than that in Spanish.

Material

In order to determine the match between product and country, a pre-test was carried out. This test consisted of two separate parts. In part one, participants were asked to name five products they typically associate with Germany, the same was done for Spain. In part two, participants were presented with a list of products and had to indicate on 7-point Likert scale in how far they considered the match between a product and a country a good one. All products were matched with Germany and Spain to make sure that a product was uniquely matched to one country.

The four products that were rated highest for a particular country and that were most frequently mentioned in the open questions of the pre-test were selected for the experiment. Although sangria was not among the products with the highest association scores, sangria was selected rather than fish because it was frequently mention in the open questions and olives, olive oil and fish could also be considered Italian or Greek (Table 1 displays the mean association scores for the pre-tested products). The following eight products were selected for the respective countries beer, sausage, car, and schnitzel for Germany, and paella, sangria, oranges, and tapas for Spain. Eleven native speakers of Dutch evaluated the selected products with respect to COO for the Netherlands. None of the products was associated with the Netherlands.

For the above products, we selected pictures from the Internet to create print advertisements. These pictures did not show a brand name to avoid interference of pre-conceived attitudes towards certain brands. The textual message was limited to product names (e.g., sangria on the bottle of sangria) for some of the products to be able to pinpoint the effect of a foreign slogan without interferences from other textual elements. This led to simplified advertisements that limit the generalizability of our results, but minimized the influence of other aspects of an advertisement. For each of the products, a slogan was created in Dutch. This slogan was translated into the matching foreign language. Those slogans were back translated into Dutch and in addition checked by native speakers of the two foreign languages. Table 2 shows the slogans that were used for the different products.

Table 1. Typical mean associations score for pre-tested products (1 = not very typical, 7 = very typical), n = 10.

	Germany (M)	Spain (M)
Product	Sausage (6,5)	Tapas (Spanish appetizer/snack) (6,8)
	Beer (6,5)	Paella (Spanish fish and rice dish) (6,6)
	Car (6,4)	Oranges (6,3)
	Schnitzel (6,3)	Olives (6,1)
	Mulled wine (Glühwein) (6,1)	Fish (5,7)
	Kaiser roll (Kaiserbrötchen) (6,0)	Olive oil (5)
	Wine (3,6)	Sangria (Spanish punch) (4,5)
	Skies (3,8)	Wine (4,3)

Table 2. Overview of slogans used in the print advertisements.

Country	Product	Dutch slogan	Foreign language slogan	English translation
Germany	Sausage	De beste worst	Die beste Wurst	The best sausage
	Beer	Heerlijk verfrissend	Herrlich erfrischend	Deliciously refreshing
	Car	Laat geen wensen open	Lässt keine Wünsche offen	Does not leave any wishes
	Schnitzel	Mals van binnen, krokant van buiten	Innen zart, aussen knusprig	Tender inside, crispy outside
Spain	Tapas	Geniet van tapas! …mmmhhh	¡a disfrutar de las tapas! …mmmhhh	Enjoy tapas! …mmmhhh
	Paella	Het beste uit de zee.	Lo major del mar.	The best from the sea
	Sangria	Elke dag mag gevierd worden	Cada día merece una fiesta	Every day can be celebrated
	Oranges	Blijf gezond, eet meer fruit	Mantente sano, com más fruta	Stay healthy, eat more fruit

Two versions were made for each print advertisement, one with a Dutch slogan and one with a slogan in the respective foreign language. See Figure 1 for an example.

Figure 1. An example of the print advertisements used in the experiment.

Design

For this experiment, a within and between subject design was used. The participants only saw one version of the print advertisements (either with the Dutch slogan or with the foreign language slogan), but saw advertisements in all conditions (within and between). This resulted in two questionnaires with each two German, two Spanish, and four Dutch print advertisements. Each participant was randomly assigned to one of the online questionnaires by means of a computer script.

Instrumentation

The effectiveness of the individual advertisements was measured in terms of the attitude towards the product, attitude towards the advertisement, and the intention to buy the product. All questions were posed in Dutch.

The attitude towards the product was measured by seven semantic differentials on a scale of 1 to 7. The statement This product is… was followed by not nice-nice, bad quality-good quality, not attractive- attractive, not useful- useful, not of additional value-of additional value, not appealing-appealing and cheap-expensive. Cronbach's alpha for these seven items was very high (α = 0.92).

The attitude towards the advertisement was measured by eight semantic differentials on a scale of 1 to 7. The statement This advertisement is…. was followed by not original-original, not interesting – interesting, boring – fascinating, easy – difficult, not nice-nice, unprofessional – professional and not understandable – understandable. Cronbach's alpha was again very high (α = 0.91).

The intention to buy the product was measured by six semantic differentials on a scale of 1 to 7. The statement This product… was followed by does not appeal to me-appeals to me, is nothing for me- is really for me. I would…. followed by not want to have this product-want to have this product, not recommend this product to someone else-recommend this product to someone else, not buy this product-buy this product and not seek information about this product-seek information about this product. Cronbach's alpha for these items was very high (α = 0.90).

The alphas for the individual questions for the three dependent variables were very high, therefore the separate questions were combined into the three variables.

The self-assessed proficiency in the respective foreign language was measured using one 5-point Likert scale item, with 1 = non-existent and 5 = fluent, following the statement My Spanish (German respectively) is …. following Hornikx, van Meurs, and de Boer (2010).

Procedure

Participants were approached via e-mail and social media. They received an URL with a link to start the questionnaire. They were randomly assigned to one of the two versions of the questionnaire. Once they had started the questionnaire they saw an introduction explaining that they would see a number of concept advertisements designed for the Dutch market and that they would be asked to give their opinion on them. They were also told that the experiment would last about 15 min. Once they clicked the start button, they saw the first advertisement. The advertisements were followed by 21 questions per advertisement. After having seen all the advertisements and the questions related to the advertisements, they were asked to provide some information about their personal background (e.g., age, gender, level of education, etc.).

Results

Repeated Measures analyses were used to assess whether the differences between the means were significantly different from each other with a critical value of 0.05. Table 3 shows the mean ratings for the three dependent variables.

Table 3. Overall mean ratings (scale 1–7, 1 = negative and 7 = positive) for the different conditions, SD between brackets.

Language of slogan	Advertisement attitude	Product attitude	Buying intention
Dutch	4.13 (0.69)	4.32 (0.74)	3.95 (0.79)
German	4.23 (0.84)	4.08 (1.21)	3.58 (1.30)
Spanish	4.14 (0.95)	4.54 (1.19)	4.21 (1.31)

Advertisement attitude

Repeated measures analysis for the attitude towards advertisement with as factor Language of Slogan (Dutch, German or Spanish) was carried out. Mauchly's test of sphericity was significant, therefore Huynh–Feldt F-statistics are reported. The analysis did not show a significant main effect of Language of Slogan F(1.731, 302.953) <1. The different languages of the slogan did not seem to have an influence on the attitude towards the advertisement.

The influence of the proficiency in the individual languages on the advertisement attitude

German

A linear regression analysis for the attitude towards the advertisement with the variable self-assessed German proficiency showed that the model was non-significant F (1,174) < 1. This means the proficiency in German did not predict the attitude towards the German advertisements.

Spanish

A linear regression analysis showed that the attitude towards the Spanish advertisement could not be explained by the variable self-assessed proficiency in Spanish F (1,174) <1. This means proficiency in Spanish did not predict the attitude towards the Spanish advertisement either. So, for neither of the two foreign languages an influence of language proficiency was found for the attitude towards the advertisement.

Product attitude

The results of a repeated measures analysis for the attitude towards the product with as factor Language of Slogan (Dutch, German or Spanish) showed a significant main effect of Language of Slogan, F(2,350) = 10.66, p < .001, η² = 0.06. Pairwise Comparison with Bonferroni correction showed that the German slogan differed significantly from the other two languages as can be seen in Figure 2 (p. 8). The products in the advertisements with the German slogan (M = 4.09, SD = 1.21) were appreciated least, the advertisements using a Dutch slogan (M = 4.32, SD = 0.74) and the products in the advertisements with a Spanish slogan (M = 4.53, SD = 1.19) were appreciated equally well (p = .078)

Figure 2. Mean ratings (semantic differentials with scale 1–7, 1 being negative and 7 positive) for Product Attitude and Buying Intention for all three language slogans.

The influence of the proficiency of the individual languages on the product attitude

German

A linear regression analysis for the attitude towards the product with the variable self-assessed German proficiency showed that the model was non-significant F (1,174) < 1. This means the proficiency in German did not predict the attitude towards the product in German advertisements.

Spanish

A linear regression analysis showed that the attitude towards the product in Spanish advertisements could be explained for 9% by the variable self-assessed proficiency in Spanish F(1,174) = 17.31, p < .001. The effect of proficiency in Spanish was significant for the attitude towards the product (B = 0.343, p < .001). If the level of proficiency increases with one score on the scale, the attitude towards the product in Spanish advertisements increases about one-third on the scale.

For the attitude towards the product in German advertisements, the proficiency in German did not seem to have an influence either, but the proficiency in Spanish influenced the attitude towards the product in Spanish advertisements. Therefore, another repeated measure analysis with proficiency in Spanish was carried out in order to see whether high proficient second language speaker of Spanish had a higher attitude towards the product for advertisements with a Spanish slogan than for those with a native Dutch slogan. The respondents were divided into two groups: The group of low proficient speakers of Spanish (n = 160) consisted of participants with a score of 1, 2, or 3 on the self-assessed proficiency score for Spanish, the high proficient speakers (n = 16) had a score of 4 or 5 on the same measure. The mean scores can be seen in Table 4.

Table 4. Overall mean ratings (scale 1–7, 1 = negative and 7 = positive) for the product attitude for high and low proficient learners of Spanish for Dutch and Spanish slogans, SD between brackets.

	Product attitude
	High proficiency in Spanish (n = 16)	Low proficiency in Spanish (n = 160)
Dutch slogan	4.4 (0.68)	4.3 (0.75)
Spanish slogan	5.6 (0.82)	4.4 (1.17)

A repeated measure analysis for the attitude towards the product with as factors Language of the Slogan (Dutch, Spanish) and Level of Spanish Proficiency (high, low) revealed significant main effects for Language of the Slogan F (1,174) = 16.97, p < .001, η² = 0.09 and Level of Spanish Proficiency F (1,174) = 9.83, p = .002, η² = 0.05. These main effects were qualified by a significant interaction effect F (1,174) = 11.62, p = .001, η² = 0.06. Separate analyses of the high and low proficient second language speakers of Spanish showed for the low proficient group that there was no significant main effect of Language of the Slogan F (1,159) = 1.31, p = .255, η² = 0.01. The results for the high proficient group showed a significant main effect of Language of the Slogan F (1,15) = 48.91, p < .001, η² = 0.75. High proficient learners of Spanish had a higher rating for the Spanish slogans (M = 5.59, SD = 0.82) than for the native Dutch slogan (M = 4.37, SD = 0.68). The interaction was due to the fact that the high proficient group showed a significant difference for the attitude towards the product for advertisement with a Spanish slogan than for the one with a native Dutch slogan, whereas the low proficient group did not.

Buying intention

Repeated measures analysis for Intention to Buy with as factor Language of the Slogan (Dutch, German or Spanish) also showed a significant main effect of Language of Slogan, Mauchly's test of sphericity was violated therefore Huynh–Feldt F-statistics are reported, F(1.869, 327.022) = 16.99, p < .001, η² = 0.09. Pairwise comparison with Bonferroni correction showed that the three languages of the slogan differed significantly from each other (p < .05). Figure 2 reveals that products advertised with a German slogan were least likely to be bought (M = 3.58, SD = 1.30), followed by products advertised with a Dutch slogan (M = 3.95, SD = 0.79). Products advertised with a Spanish slogan were the most likely to be bought (M = 4.21, SD = 1.31).

The influence of the proficiency in the individual languages on the buying intention

German

A linear regression analysis for the intention to buy the product with the variable self-assessed German proficiency showed that the model was non-significant F (1,174) < 1. This means the self-assessed proficiency did not predict the intention to buy the product.

Spanish

A linear regression analysis showed that the intention to buy the product could be explained for 8% by the variable self-assessed proficiency in Spanish F (1,174) = 15.4, p < .001. The effect of proficiency in Spanish was significant for the intention to buy the product (B = 0.36, p < .001). If the level of proficiency increased with one score on the scale, the intention to buy the product increases about one-third on the scale.

For the intention to buy the product, the proficiency in German did not seem to have an influence either, but again the proficiency in Spanish did. Therefore another repeated measure analysis with proficiency in Spanish was carried out in order to see whether high proficient second language speaker of Spanish had a higher buying intention after having seen the advertisement with the Spanish slogan than the one in their native language. The division of groups was the same as above. The mean scores can be seen in Table 5.

Table 5. Overall mean ratings (scale 1–7, 1 = negative and 7 = positive) for the buying intention for high and low proficient learners of Spanish for Dutch and Spanish slogans, SD between brackets.

	Buying intention
	High proficiency in Spanish (n = 16)	Low proficiency in Spanish (n = 160)
Dutch slogan	4.14 (0.56)	3.94 (0.81)
Spanish slogan	5.37 (0.80)	4.09 (1.29)

The repeated measure analysis for intention to buy the product with as factors Language of Slogan (Dutch, Spanish) and Level of Spanish Proficiency (high, low) revealed a significant main effect for Language of Slogan F (1,174) = 17.22, p < .001, η² = 0.09 and Level of Spanish Proficiency F (1,174) = 11.31, p < .001, η² = 0.06. These main effects were qualified by a significant interaction effect F (1,174) = 10.15, p = .002, η² = 0.06. Separate analyses of the high and low proficient second language speakers of Spanish showed that there was no significant main effect of the Language of the Slogan F (1,159) = 2.41, p = .123, η² = 0.02 for the low proficient group. The results for the high proficient group showed a significant main effect of Language of Slogan F (1,15) = 40.71, p < .001, η² = 0.73. High proficient learners of Spanish had a higher buying intention after having seen the Spanish slogans (M = 5.37, SD = 0.80) than after having seen the native Dutch slogans (M = 4.14, SD = 0.56). The interaction was due to the fact that the high proficient group showed a significant higher intention to buy the product for advertisement with a Spanish slogan than for the one with a native Dutch slogan, whereas the low proficient group did not.

Conclusion and discussion

The aim of this study was to investigate whether advertisements with a congruent product – country of origin (or language) relation were more effective than advertisements for the same product in the native language of the consumer. Another aim was to establish whether the consumers’ foreign language proficiency had an influence on the effectiveness of the foreign language advertisement compared to the same advertisement in consumers’ native language. Participants were shown a number of ads. The effectiveness of these ads in the different conditions was measured in terms of buying intention, attitude towards the advertisement and the product. Rather than having participants with different levels of proficiency in one foreign language participate in this study, we selected two foreign languages, one assumed to be widely spoken in the Netherlands (German) and one that is less frequently taught and consequently less frequently spoken (Spanish). By means of self-assessment, we controlled whether the assumption about the proficiency in the two languages was correct. The results of the experiment among Dutch consumers showed that foreign language use as such did not necessarily make an advertisement more effective than the same advertisement in the native language of the consumer. Overall, the advertisements with the German slogans were less effective than the native Dutch ones, the advertisements with the Spanish slogans, on the other hand, were more effective than the Dutch native slogans and the German slogans with respect to the attitude towards the product and the intention to buy the product. For the attitude towards the advertisement, we did not find any influence of proficiency in a foreign language.

Looking at the role of proficiency in the foreign language, the results of this study are not as straight forward as expected. In terms of the overall proficiency in the two language (reasonable proficiency in German, low proficiency in Spanish), the effect is contrary to what we anticipated, Dutch consumers rated the slogans in the language in which they were least proficient (Spanish) better than the slogans in their native language, but also better than the ones in the language they were more proficient in (German). In fact, German slogans got the lowest ratings. These results do not seem to coincide with the results by Hornikx, van Meurs, and de Boer (2010) who found that low proficient Dutch speakers of English preferred native slogans to the English equivalent when the English slogans were more complex. The present results are even more surprising when we take a closer look at the slogans in the three languages (see Table 2). The slogans in Dutch and German are very similar, the wording in the two languages was almost parallel. This means that German slogans could possibly be understood without any knowledge of German because the Dutch slogans were almost the same. In contrast, the Spanish slogans are very different from the Dutch ones, which meant that they were difficult to understand without any knowledge of Spanish. In the present study, the aim was to keep the slogans the same as possible, the original Dutch slogans were translated into German and Spanish and back translated. Given the linguistic relatedness of Dutch and German, this resulted in very similar slogans for both languages and thus to relatively easy slogans for the German advertisements and more difficult ones (in terms of vocabulary) for Spanish. The difference in difficulty could have influenced the results of this study. However, if that were the case, we would have expected that the advertisements with a Spanish slogan received lower ratings for the three measures of effectiveness, but this was not the case. Follow-up studies should control for the difficulty of the slogans in the different languages to determine the role complexity of the slogan plays for languages other than English.

When looking at the self-assessed proficiency of individual participants in both languages, the picture is even more complex. Consumers’ proficiency in German had no effect on attitude towards the advertisement, the product, or the buying intention. Advertisements with German slogans were rated lowest on all three measures. This might again be due to the fact that the German slogans were very similar to the Dutch ones, as well as the fact the participants in this study were relatively high proficient in German (M = 3.25, on a scale from 1 to 5). Proximity to the German boarder might be another factor that influenced the results. The participants in this study came mainly from the east of the Netherlands, the parts closest to the German border, they have access to German media and are more familiar with German than consumers elsewhere in the country. Consequently, the use of German no longer functions as an “attention getting device” as proposed by Domzal, Hunt, and Kernan (1995) because the use of German is no longer unexpected or deviant from the norm and given the familiarity with the language, processing German probably does not take much longer than processing the ad in the mother tongue and consequently this did not influence the attitudes measured in this study.

For Spanish, the participants’ proficiency in Spanish had no influence on the attitude towards the advertisement either. Given the average proficiency in Spanish was low (1.78 on a 5-point scale) and it was even lower (1.55) when the participants were divided in a high and low proficient group, we would have expected, in line with the results by Hornikx, van Meurs, and de Boer (2010), that participants would prefer a Dutch slogan over a Spanish one, but this was not the case. Little can be said about the group of participants that was high proficient in Spanish because their number was too small.

For the attitude towards the product and the buying intention, the role of proficiency is more transparent. Despite the low number of high proficient speakers, proficiency in Spanish could predict the attitude towards the product as well as the buying intention. For low proficient speakers of Spanish, we did not find differences in their attitude towards the product and the buying intention for advertisements with a Spanish or native Dutch slogan. High proficient speakers of Spanish, on the other hand, had better scores for both measures when they rated advertisements with Spanish slogans. Given the small number of participants that were high proficient in Spanish and the variation within this group, the results for the comparison of low and high proficient speakers have to be treated with caution. However, the results for the low proficient speakers did not coincide with results by Hornikx, van Meurs, and de Boer (2010) because consumers that were low proficient in Spanish did not have a better attitude towards the advertisements with the Dutch slogans than towards the advertisements with the Spanish slogans, in fact there was no difference between Spanish and Dutch advertisements. Under the assumption that the participants of this study did not understand the actual meaning of the Spanish slogans, the understanding does not seem to be a prerequisite for the effectiveness of the Spanish slogans. Consequently, something else must be responsible for the higher effectiveness of the Spanish advertisements, and this could be the symbolic function of foreign language use as proposed by Kelly-Holmes (2000). The fact that Spanish slogans were even more effective for more proficient speakers of Spanish could be explained by the fact, that this group, while learning the language had more contact with the language and the country and consequently had more (positive) associations with both. These associations were triggered when reading the Spanish slogans and led to higher ratings for the attitude towards the product and the buying intention.

Another possible explanation for the results of the present study could be that Spanish is more popular than German among the student population that participated in this study. However, this is difficult to verify because German is part of the curriculum in Dutch secondary school, whereas Spanish is only offered as an elective and only at some schools. A study by Du Bois-Reymond et al. (1999) which was based on two surveys and qualitative interviews found that Dutch children and young people have a more negative attitude towards the Germans than towards other European countries. According to Jaffe and Nebenzahl (2001) in advertising, a negative image country is very difficult to counter-balance. This could in part be responsible for the results of this study; however, a study concerning students’ attitude towards Spanish and Spaniards to support this line of reasoning is lacking. Further research will not only have to take more foreign languages into account, but will also have to look at the attitudes towards these language to determine factors that influence the effectiveness of foreign languages in advertising in comparison to the native language of consumers.