Different studies show that English is widely used in television advertising messages targeting consumers whose native language is not English (Gerritsen et al., 2000; Hsu, 2008; Kuppens, 2010; Lee, 2006; Martin, 2002, 2007; Pétery, 2011; Piller, 2000; Riaz, 2019; Ustinova & Bhatia, 2005). Raedts et al. (2015) found that almost two out of three prime-time television commercials in Belgium, France, Italy, the Netherlands, and Spain contain English songs and/or English words in spoken and/or written form.

Why is the use of English in (television) advertising so widespread? The first reason can be found in multinationals’ belief that English product names and brand slogans, and by extension standardized English-only advertising messages, help them to globalize the image of their products (Akaka & Alden, 2010; Hornikx & van Meurs, 2020; Martin, 2002; Nelson & Paek, 2007). Second, advertisers also believe that the use of English increases the brand’s competitiveness and/or evokes the idea that the brand behind the advertising message excels when it comes to the use of the latest technology (Hornikx & van Meurs, 2020; Martin, 2002). Third, a lot of advertising agencies think that the target audience has no problems understanding the English words and phrases in their advertising messages (de Mooij, 2004; Gerritsen et al., 2000).

However, a large-scale survey by the European Commission (2012) showed that Europeans’ English proficiency level is not as high as advertising agencies might assume. Only 27 % of the 15- to 24-year-olds in the EU member states rate their English skills as very good; among the over-55s that percentage drops to 15 %. Gerritsen et al. (2010) found that a substantial part of university students in Belgium, Germany, France, the Netherlands, and Spain was not able to provide an accurate translation of English phrases used in print ads targeting their age group. A study by Gerritsen et al. (2000) highlighted that non-native English viewers overestimate how much they understand of the English words and phrases in television commercials. The Dutch research team showed fifteen- to eighteen-year-olds and people in their fifties six prime-time television commercials. All ads contained at least one spoken English sentence (e.g. ‘The spirit of freshness’ or ‘One break is never enough’). Over 80 % of participants claimed to understand these English sentences, and 76 % thought they were able to translate the English phrases correctly. However, when they were asked to write down the English sentences correctly, and to subsequently translate these English sentences into their native language (Dutch), the majority of participants failed to do so. Gerritsen et al. (2000) found striking age differences. Only 10 % of the older participants (50- to 57-year-olds) were able to give a correct transcription of the English phrases that did not appear on the screen, and only 9 % of the older participants were able to correctly describe the meaning of these English sentences in Dutch. The younger generation performed significantly better on both comprehension tasks: 37 % of the teenagers correctly transcribed the English sentences, and 35 % correctly translated or paraphrased the meaning of the English sentences in their mother tongue. Not only participants’ age but also their educational level was a decisive factor. More than four out of ten (42 %) participants who (had) attended an academic high school were able to explain what was meant by the spoken English sentences, whereas only 7 percent of the participants who (had) attended a vocational secondary school were able to paraphrase in Dutch what was said in English.

Researchers on the use of foreign languages in advertising do not agree on the role of comprehension in the persuasion process (Hornikx & van Meurs, 2020). According to sociolinguists (e.g. Haarmann, 1989; Kelly-Holmes, 2005; Piller, 2003), it is of minor importance that readers or viewers understand the foreign language words in advertising messages because these words primarily have a symbolic function (Kelly-Holmes, 2005). Consequently, it is sufficient that they recognize the foreign language in an advertisement. The moment consumers recognize the foreign language, they activate positive associations about the (inhabitants of the) country where that language is spoken (Hornikx & van Meurs, 2017). These positive associations are subsequently transferred to the advertised product. Hence, foreign languages in advertising messages serve as an implicit country-of-origin cue (Hornikx & van Meurs, 2017). English is a special case, as this language is seldom used to evoke specific ethno-cultural associations (Hornikx et al., 2020). Instead, English is primarily used as a ‘marker of globalness’: “Ads with English are interpreted as projecting that the company operates internationally as a global player, and that the product is used by consumers worldwide” (Hornikx & van Meurs, 2020, p. 99).

From a psycholinguistic perspective, foreign language elements make an advertising message more difficult to process and understand than an advertising message in the consumer’s own language (Domzal et al., 1995). The foreign language mental lexicon is smaller, less rich, and less extensive than the first language mental lexicon (Skehan, 2018). Advertisers seem to take this fact into account: they use relatively simple foreign language utterances (words belonging to basic vocabulary) and/or cognates (Hornikx & van Meurs, 2020, p. 79).

Several experimental studies have shown that understanding foreign language utterances in advertising messages matters for the effectiveness of the advertisement (Hornikx & van Meurs, 2020). Raedts et al. (2016) found that consumers valued English slogans in a standardized advertising campaign more when they understood the English catchphrases. Participants who were able to translate the English slogans in their mother tongue also had higher purchase intentions compared to participants who did not or only partially understood the English catchphrases. Hornikx et al. (2010) and Hendriks et al. (2017) found that Dutch consumers appreciate easy to understand English language slogans more than difficult to understand English language slogans. Hendriks et al. (2017) also concluded that easy to understand English slogans resulted in a better attitude towards the ads and the advertised products, and in higher purchase intentions.

Gerritsen et al. (2000, p. 28) showed that viewers’ understanding of the written and especially the spoken English words in television ads “is not as rosy as the makers of the commercials suppose”. Their findings raise the question of whether the language of standardized cross-national television advertising campaigns should be translated into the language of the local viewer.

Many European countries have either a subtitling or a dubbing tradition. In the Netherlands and in the Dutch-speaking part of Belgium, subtitling was and still is the standard (European Commission, 2011; Koolstra et al., 2002). Yet, even in typical dubbing countries (e.g. France, Germany and Spain), the trend is shifting towards subtitling (Riggio, 2010; Perego, 2018).

Research on the effectiveness of local-language subtitles in English-only commercials on European consumers is scarce. Moreover, the research results contradict each other. In addition, most of the participants were viewers from dubbing areas. Scarpi et al. (2020) found that dubbing works better than subtitling to increase Germans’ and Italians’ attitudes toward the ad, but only when the viewers find that English is very different from their own language. Pagani et al. (2015) observed that viewers of standardized English-only TV ads in Italy, Germany and Spain (all typical dubbing countries) had less positive attitudes towards the ad and towards the brand compared to viewers who saw the subtitled or dubbed version of the ads. No significant differences in attitude scores emerged between participants in the dubbing and in the subtitling condition. Cuykx and Smits (2017) replicated Pagani et al.’s (2015) experimental study in Belgium, a country with two different major language communities, and two different media landscapes. Flanders (where Dutch is the official language) is a typical subtitling region, whereas Wallonia (where French is the official language) has a dubbing tradition. Dubbing yielded lower attitude scores towards the ad and the brand within the group of Dutch-speaking participants. Dutch- and French-speaking participants had similar attitude scores for both the standardized and the subtitled versions of the ads. An experimental study by Raedts et al. (2019) among pupils in Flanders and Wallonia revealed that subtitles boost the communicative effectiveness of English-only ads, especially when the dialogues in the commercials contain English words that exceed the viewer’s English language proficiency level. Effects for marketing effectiveness, however, were only found for ads with English words above the level of basic English learners.

In the above-mentioned studies, participants were young and usually highly educated. Experimental studies on the added value of subtitles for viewers of Generation X, baby boomers and people over 75 of different educational levels are lacking (Raedts et al., 2015, 2019). The primary purpose of our study was to fill this research gap, all the more because Gerritsen et al.’s (2000) study showed that mostly elder and low-educated Dutch-speaking viewers overestimated their understanding of spoken English in TV ads. Gerritsen et al.’s (2000) study, however, is more than twenty years old. It is likely that the English proficiency level among viewers of Generation X, baby boomers and people of the silent generation has increased in the past two decades. Viewers of Generation X are born between 1965 and 1980, baby boomers are born between 1946 and 1964, and people of the silent generation are born between 1928 and 1945 (Dimock, 2019). In 21st century Europe, many people are frequently exposed to the English language, for instance when they surf on the Internet, or watch English-spoken movies and series on their smartphones and tablets (Seidlhofer, 2020).

Following on from the studies mentioned above, we formulated the following two research questions:

Raedts et al. (2019) found that local-language subtitles have a positive effect on young consumers’ attitudes towards the ad, provided that the complexity of the English used in the television commercial exceeds the knowledge level of the viewer. Based on the results of Gerritsen et al. (2000) and the results of the language survey of the European Commission (2012), we assumed that people aged 40 and over, i.e. people of Generation X (born 1965–1980), baby boomers (born 1946–1964) and people of the silent generation (born 1928–1945), are less proficient English language users than the millennials (born 1981–1996) and Generation Z (born 1997–2012). Therefore, we expected to find a positive effect of subtitles on these viewers’ attitudes towards English-only ads. This leads us to our third research question:

Our study complements previous research into the effects of subtitles in television commercials in three important ways. First, we conducted an experiment with older viewers of different educational levels. In addition, we examined the added value of subtitles for the communication effectiveness of English-only TV ads. Except for Raedts et al. (2019), previous researchers only measured the effects of subtitles on the marketing effectiveness of English-only commercials (i.e., consumers’ attitude towards the ad and the brand). Third, previous researchers used commercials of different brands. As research has shown that consumers’ prior brand attitudes are related to their attitudes towards advertisements of the brand (Lee, 2010; Muehling & McCann, 1993), we chose commercials from the same brand. Hence, we kept prior brand attitude equal and avoided brand bias in our research results. In addition, using commercials from the same brand allows clustered analyses. Previous studies analyzed the effects of subtitles on a commercial-by-commercial basis.

For our experiment, we chose four cross-national car commercials, since the use of English in advertising messages of multinationals manufacturing luxurious and sophisticated high-tech culture-free products is widespread (Hornikx & van Meurs, 2020, p. 111). The market share of this luxury car brand (less than 1 %) is very low in Belgium (Marklines, 2020), which made it unlikely that a large part of our participants would own a car of the brand. Hence, we could exclude brand familiarity as a possible interference variable in our experiment.

All four commercials lasted 30 seconds and contained spoken English in the form of a monologue or dialogue between actors. The first two commercials (‘Copilot’ and ‘Toy ride’) belong to a pan-European campaign called Performance obsessed. In the first commercial (Copilot), a young girl wearing a red racing cap, is giving directions to a woman (probably her mother): “Turning into left, two. Turning, turning. Come on. Left, three. Do not cut.” The images have been edited in such a way that the viewers get the impression they are watching a street rally. At the end of the commercial, however, it becomes clear that mother and daughter are in an outdoor garden center. In the second commercial, called ‘Toy ride’, a father picks up his son from a children’s party. When they leave the house, the father hides a yellow robot in the vegetation next to the house. In the next scene, the father receives a phone call. He replies to the caller: “The yellow robot? Oh, we’ll be right over.” In the following scene, we see the father, his son and the yellow robot inside a SUV. In both commercials, the sounds of a racing car can be heard as a Jaguar driving in a residential area disappears from the screen. The third and fourth commercials, featuring the French actress Eva Green, were part of a global advertising campaign launched in 2018. The ‘Electrical storm’ commercial starts with images of a severe thunderstorm in a city late at night. In the next scene, Eva Green descends a giant staircase of a hotel. A life-size (computer generated) jaguar walks by her side. When the night porter notices her, he says: “Evening ma’am, it’s a terrible night”. Eva replies: “It’s just a bit of electricity. Nothing to be afraid of”. Together with her pet, she walks out of the hotel, into the storm. The fourth ad is called ‘Popular choice’. In this commercial, Eva Green enters a hotel lobby, again accompanied by her unusual pet. When the assistant at the hotel counter sees her, he confirms he got her message and shows her a tablet, saying “This is our most popular choice”. The actress replies: “Thank you, but I don’t do popular choices”. Next, she confidently walks away from the hotel desk. The four commercials are published by Jaguar Belux on YouTube (https://www.youtube.com/c/JaguarBELUX/videos).

Both campaigns use different types of narratives. The ‘Copilot’ and ‘Toy ride’ commercials are so-called slice-of-life commercials in which the product is embedded in an everyday situation (Wahl, 2014). The ‘Popular choice’ and ‘Electric storm’ commercials, however, are so-called arthouse commercials, commercials in which the aesthetic quality seems more important than the advertised product itself (Wahl, 2014). We decided to run separate analyses for both campaigns because of their different narratives, but also because the participants in the study by Verlinden (2018) felt that the ‘Popular choice’ and ‘Electric storm’ commercials better represented the brand Jaguar than both slice-of-life commercials (‘Copilot’ and ‘Toy ride’).

We used a between-subjects design. Each participant saw either the subtitled or the non-subtitled version of the four car commercials. Participants were randomly assigned to one of the experimental conditions. To avoid order effects, the four commercials were shown to the participants in random order.

We created a General Data Protection Regulation (GDPR) compliant online questionnaire in Qualtrics (an online survey tool). At the beginning of the questionnaire, participants were informed that they would see four commercials of the brand Jaguar. They were informed that their participation was voluntary, that their answers were anonymous, and that they could end their participation at any point. After they had given their consent, the questionnaire started.

We first checked participants’ familiarity with the car brand through three yes-no-questions: ‘Do you know the brand Jaguar?’ (n = 184 ‘yes’), ‘Do you own a Jaguar? (n = 2 ‘yes’)’, ‘Was one of your previous cars a Jaguar?’ (n = 4 ‘yes’). So, most participants knew the car brand, and only a very small proportion (had) owned a Jaguar. Therefore, we decided not to remove participants from the sample based on their familiarity with the car brand. These questions were followed by a seven-points four-item Likert scale measuring participants’ attitudes towards the brand: ‘I like Jaguar’, ‘Jaguar is a brand to look out for’, ‘Jaguar fits in my life’, and ‘I would love to be seen driving a Jaguar’. The items were based on Aggarwal and McGill’s scale (2012) for brand attitude. The scale’s reliability score was .79.

Then, we measured participants’ perceived English proficiency level with the following question: ‘If the subtitles are missing when watching an English-language movie or series, how well can you continue to follow the story?’. Participants had to fill in a number from 0 (I cannot follow the film or episode at all) to 10 (I can follow the film or episode as good as a native speaker of English).

Next, participants saw the commercials one by one. Participants were instructed as follows: “In the rest of the questionnaire, we ask you to watch four commercials. You will be shown one clip each time. After each commercial we will ask you some questions about the commercial. It is the intention that you watch each commercial only once. Therefore, we have set a timer. The questionnaire automatically switches to the next page if you have not clicked through yourself.”

Previous exposure to the ad was elicited by the following question: ‘Have you seen this commercial prior to this viewing?’. Participants could check one of the following three response options: no, yes, or don’t know. Subsequently, we measured participants’ attitudes towards the ad with five seven-point semantic differentials: The commercial is (not) fun, boring/exciting, (not) original, (not) attractive, and (not) interesting (cf. Hendriks et al., 2017). The internal reliability of the scales was good: Cronbach’s alpha was above .94 for all four commercials.

On the next page of the questionnaire, two multiple choice questions verified participants’ comprehension of spoken English in the commercial. The comprehension questions were all recognition questions with five answer possibilities (see Table 1 for examples). For each correct answer, participants received one point. For each campaign, they could therefore score four points.

Next, we asked participants about their year of birth, gender and educational level. We also requested participants to indicate whether their data were suitable for the study, i.e. whether they had watched the commercials carefully, and had not looked up the answers to the comprehension questions on the internet. On the next page of the questionnaire, we showed a debriefing text in which we explained the design and the purpose of the study. At the end of this page, participants were given the opportunity to withdraw from the study. On the last page of the questionnaire, we thanked the respondents once again for their cooperation in the study.

All participants were residents of the Dutch-speaking part of Belgium. Our final sample comprised 188 participants (61.7 % women) between 41 and 88 years old (M = 56.10, SD = 9.01); 58 % of participants belonged to Generation X (born between 1965 and 1980), 38.3 % were baby boomers (born between 1946 and 1964), the remaining 7 participants (3.7 %) were born between 1933 and 1945 (silent generation). More than one out of four (26.6 %) participants held a university degree (bachelor or master or Ph.D.), 36.7 % held a degree of a professionally oriented bachelor programme, and 36.7 % did not start or complete higher education. Participants’ average score on the self-assessment question regarding their English listening skills was 6.99 out of 10 (SD = 2.12). More than half of the participants (52 %) estimated their English listening skills to be high: they rated themselves with an 8, 9 or 10. Respondents did not receive any incentive for their participation.

Only a small group of participants indicated that they had already seen the ‘Copilot’ commercial (1.1 %), the ‘Toy ride’ commercial (3.7 %), the ‘Popular choice’ commercial (16 %) and the ‘Electric storm’ commercial (9 %). Because of the small numbers, we decided not to exclude participants’ data from further data analyses based on their familiarity with one or more commercials.

Participants were recruited through social media messages (Facebook, Twitter, LinkedIn) and e-mail messages containing an anonymous link to the questionnaire. It took participants about 12 minutes on average to complete the survey.