Can You Make a Difference? The Use of (In)Formal Address Pronouns in Advertisement Slogans

Abstract

Directly addressing the consumer in advertisements has a positive impact on their perception of the product. Dutch, however, makes a distinction between a formal and an informal second person pronoun. If and how this distinction impacts consumers’ appreciation in and beyond advertising has not received much attention. Quantifying consumers’ appreciation in terms of the attitude toward the advertisement and toward the product, purchase intention, and price estimation, we show experimentally that customers’ attitude toward the advertisement receives higher scores when the informal pronoun is used than when its formal counterpart is.

Keywords:

• Advertising
• second person pronouns
• experiment
• persuasion
• involvement

Introduction

It is safe to say that one of the most famous slogans was printed in 1917: J.M. Flagg’s “I want YOU (for U.S. army)” propaganda poster. The slogan is accompanied by a depiction of Uncle Sam – a personification of the United States – who menacingly looks and points at the reader. The poster and the slogan make clear use of a persuasion strategy in which the target audience is directly addressed, in this case in an effort to enlist new recruits for military service. This rhetorical strategy has been widely adopted for over a century. Christopher (2012) investigates a sample of 500 advertising slogans from the past century, which have been used by thirteen international companies from various branches. She observes that a prominent persuasion technique in these slogans is personalization, a strategy with the aim of connecting the advertiser and the consumer. Although first person pronouns such as I and we referring to the advertiser are also used for this purpose, most slogans in this study used a second person pronoun you referring to the consumer. Christopher (2012, 520) gives the 1932 British American Tobacco slogan “Do you inhale? Of course you do!” as an example of this strategy.

The use of second person pronouns as a persuasion strategy is also evident from Hradilova’s (2014) corpus study of one-page advertisements targeted at women. She reports that 50 out of 65 (76.9%) advertisements contained person deixis − 43 of which (66.2%) included the second person pronoun. Sušinskienė (2013) shares a handful examples of slogans with a second person pronoun from the sector of fragrances and argues that the second person pronoun is not only the most direct form of address, it also places the consumer in the center of attention and thereby underlines their individuality. This may be one of the many reasons why the use of the second person pronoun seems to be a growing trend across genres in and beyond advertising (Sorlin 2017).

That second person pronouns can increase the feeling of engagement with written text forms has also been found in experimental research (Brunyé et al. 2009, 2011, 2016; Ditman et al. 2010; Andeweg et al. 2013; Hartung et al. 2016). In narrative texts, second person pronouns facilitate taking the (internal) perspective of the agent; third person pronouns by contrast facilitate taking the (external) perspective of the observer. The power of second person pronouns to trigger self-ascription (Wechsler 2010) makes them a powerful tool not only in narration, but in (charity) fundraising (Macrae 2015), political discourse (Kuo 2006), TV shows (Sorlin 2015), and tweets from broadsheet publications, as compared to news headlines (Moncomble 2017).

It is crucial to note here that most studies discussed so far investigate the influence of second person pronouns in English. Many languages, however, employ a system of pronouns with a distinction between formal and informal second person pronouns, i.e., V and T forms, respectively, after Latin vos and tu (Brown and Gilman 1960; Helmbrecht 2013). Dutch is one such language. The slogan “I want you” therefore has two possible translations in Dutch: one with the formal address pronoun ik wil u ‘I want you[V]’ and one with the informal address pronoun ik wil jou ‘I want you[T]’. And while researchers have studied preferences for T or V forms in advertising (Weyers 2011, 2022) and translations of advertisements with English you (Smith 2004), the question remains what the impact of different second person pronouns in advertising is. The current paper investigates the impact of T and V pronouns of address in Dutch advertising (through an image of a product accompanied by a slogan) experimentally, by means of a questionnaire. We tested participants’ attitudes toward the advertisement and the product, their estimation of the price, and their intention to buy. The results indicate that T is generally better appreciated than V by our participant group in the sense that the advertisements were rated as more appealing in T than in V.

Theoretical background

Persuasion in advertising

There is a shift toward more second person pronouns in and beyond advertising. Sorlin (2017) cites a sense of trust and attracting attention as possible reasons for this shift toward more second person pronouns, as it simulates a more personal relationship between persuader and addressee. She quotes Trush (2012, 20), who indicates that “marketing involves initiating and developing relationships similar to the one you have with a spouse”, with trust being the most important component. This may be achieved through the use of second person, because “what matters is to get inside the prospect’s head, find out what she is thinking, what her desires, fears, and beliefs are and provide a tagline that is a direct answer to them” (Sorlin 2017, 138). The shift toward more direct address forms in advertisements then correlates with a shift from highlighting facts to highlighting benefit, with the consumer essentially placed in the center of attention. That is, the advertisements no longer present (only) facts about the product, but insist on its potential benefit for the client.

Thus, advertisements commonly relate the content of the entity (product, service) to the addressee and one strategy to do so is by using personal pronouns. Second person pronouns in particular have the ability to change the conception of the addressee from a generic audience, or a “disembodied abstraction” (Moncomble 2017), to the consumer – an actual human being reading the text. This strategy is also known as self-referencing and has been shown to lead to a closer reading of the arguments contained in the message (Burnkrant and Unnava 1989, 1995; Debevec and Romeo 1992; Mick 1992; Meyers-Levy and Peracchio 1996; Escalas 2004, 2007; Cruz, Leonhardt, and Pezzuti 2017). These studies show that the use of second person pronouns instead of third person pronouns in English advertisements can lead to stronger persuasion, in that participants report higher attitudes toward and higher credibility of the advertisement and the product, as well as higher intentions to try or purchase the product. Participants also show improved recall of the message content (Burnkrant and Unnava 1989; Mick 1992; Meyers-Levy and Peracchio 1996). Yu, Hudders, and Cauberghe (2017) examined the effectiveness of a personalized email advertisement for a fictional luxury watch brand in Chinese and Dutch. They manipulated personalization by using the second person pronoun you in the body of the personalized email and no pronouns in the non-personalized email. Similarly, the slogan in the non-personalization condition was “This is elegance”, while in the personalized condition the second person pronoun was used: “Elegance especially designed for YOU.” A Dutch and a Chinese version of the ad were designed, because the authors were interested in the differences in influence of personalization on purchase intention between a highly individualistic culture (The Netherlands) and a lowly individualistic culture (China). For the online experiment they recruited Dutch and Chinese middle- and high-class working participants living in The Netherlands and China. They found that personalization had a positive effect on purchase intention, but only in Dutch. These findings suggest that the use of second person pronouns motivates Dutch consumers to imagine that the message content applies to their own personal situation (mental simulation in Escalas 2004, 2007) and to relate this to the advertisement’s content. This leads us to our first hypothesis:

H1: Consumers will respond more positively to advertisements featuring a pronoun of address (T or V) than to advertisements without a pronoun of address.

Yu, Hudders, and Cauberghe (2017) did not provide information about which second person pronoun they used in their Dutch ad, the informal T pronoun or the formal V pronoun. Shubin Yu (p.c.), the first author of the article, kindly informed us that the researchers used the informal second person pronoun in Dutch, because that is the most common one used in Dutch ads, as will be discussed below. However, since it was an ad for a luxury product (8,000 Euro), they recruited participants with a monthly income higher than 3,000 Euro. The use of V pronouns in the Netherlands is often associated with older age, and older people tend to be in a better financial position (Vismans 2013b). Therefore, it would have been interesting to compare the Dutch ad in Yu, Hudders, and Cauberghe’s (2017) study with an ad that contained a V pronoun instead of a T pronoun, to see if it would have a different effect.

Cross-linguistic differences in formal and informal second person pronouns

As pointed out above, various (European) languages distinguish between a formal and an informal second person pronoun (V and T, respectively, see Brown and Gilman 1960; Helmbrecht 2013). Research on pronoun usage has shown that languages vary widely in terms of their pronoun inventories and the factors that determine the speaker’s choice of one pronoun over another, see, for example, Vismans (2019) on Dutch, Kretzenbacher and Schüpbach (2015) on German, Foster, Aalberse, and Stoop (2019) on (Colombian) Spanish, and Coveney (2010) on French. In Spanish, pronouns with the grammatical function of subject are often not expressed, so that only the conjugation of the verb reveals whether T or V is used. Interestingly, the pronoun V is expressed overtly more often than T in Spanish (Serrano 2017). Serrano (2017) links this variation to differences in inherent subjectivity and objectivity of T and V, respectively. Posio (2011) notes that the overt subject pronoun V is typically used in contexts where personal involvement of the addressee is at stake, adding that in such cases the V pronoun is not used to avoid ambiguity, but as a marker of politeness toward the addressee.

Levshina (2017) examined translations of utterances with second person pronouns from a set of original English movies to subtitles in ten European languages. The advantage of her quantitative and qualitative parallel corpus study is that the situations in which T and V are used were kept completely identical across the different languages. Levshina found the lowest number of T pronouns (more V than T) in the French and German subtitles, but the opposite pattern (more T than V) in two related languages, Spanish and Dutch.

Even within languages, there are striking differences. For example, unlike the Netherlandic Dutch pronominal system, the endogenous Flemish system does not distinguish between informal and formal pronouns for the second person singular (Vandekerckhove 2005; Plevoets, Speelman, and Geeraerts 2008). An important observation in this regard is the fact that the object pronouns and possessive pronouns of the Flemish system correspond to the V pronouns of the Standard Dutch system, without being perceived as formal or informal by Flemish speakers of Dutch, who mostly use the pronouns in interactions with friends and family. Moreover, the informal Netherlandic Dutch T pronouns are perceived by Flemish speakers as exogenously Dutch and thereby as rather formal, since they are associated with the Standard Dutch language. These T-forms are hardly used by adult Flemish people, except in interactions with young children (De Houwer 2003; Van de Mieroop, Zenner, and Marzo 2016).

Research has shown that certain tendencies exist in how and when the T and V pronouns are used in advertising across languages. Smith (2004) reports on a parallel corpus study of 45 English-to-Russian translated advertisements, primarily from glossy women’s magazines (1997–2001), with a diverse set of non-essential products. She further reports on a monolingual corpus with 23 similar advertisements in Russian. While Smith (2004, 289) does not share the exact numbers of her research, she does report to have found a general preference for the formal variant in both corpora, with a “high percentage of adverts using V in the Russian parallel corpus.”

Weyers (2022) demonstrates that Colombian Spanish speakers from Medellín have a preference for the informal imperative T verb form tuteo in advertisements of national brands, but for the formal imperative V verb form ustedeo in advertisements of international brands. A third form voseo was never the top choice for any item, but overall, men preferred ustedeo forms and women tuteo forms. Weyers (2022) assumes that men’s preference for ustedeo is in fact a way for them to avoid tuteo, because voseo is a widespread alternative informal T form used mainly by men, functioning as an identity marker. However, this alternative T form is considered less prestigious in written language, hence their switch to the formal verb form ustedo. Female participants would not use voseo instead of tuteo in their everyday speech and thus they preferred tuteo. Weyers (2011) shows that 88% of the Mexican Spanish commercial advertisements in his sample used tu ‘you[T]’, whereas only 54% of the noncommercial advertisements did. Weyers suggests that these findings show the persuasion strategy of using informal pronouns of address to engage with the consumer on a more personal level.

Van Compernolle (2008) found in a corpus of 200 on-line dating ads from heterosexual men and women residing in Quebec that an address pronoun was present in about three quarters of the personal ads, which highlights the interactive communication that takes place in romantic relationship initiation processes and “demonstrates some awareness of a definite other” (van Compernolle 2008, 2067). Van Compernolle finds that almost 63% of personals contain a singular second person pronoun (T or V), with a preference for the T (81%) over the V pronoun (19%). This finding, van Compernolle argues, reflects the general preference in Quebecois to use T to address strangers and is arguably generalizable to environments of non-computer mediated communication (cf. Pires 2004). Taken together, these studies show that preferences for T or V forms differ between languages.

Dutch is a language with a strong preference for the use of T forms in various contexts (Vismans 2013b, 2018; Levshina 2017; den Hartog, van Hoften, and Schoenmakers 2022), albeit only in the Netherlands – not in Belgium (Vandekerckhove 2005). However, the T/V distinction has been little explored in the world of advertising. Moreover, the impact of T and V second person pronouns on the addressee remains largely unexplored, even though both pronouns address consumers directly in advertisements. In the following, we discuss several studies that have investigated the impact of T and V pronouns on Netherlandic Dutch readers in a written persuasive text advertising a holiday (van Zalk and Jansen 2004), government-issued texts (Jansen and Janssen 2005), HR communication (de Hoop, Levshina, and Segers 2023), marketing communication (Leung et al. 2022), and product advertisements (de Hoop et al. 2023).

The impact of T and V pronouns in Netherlandic Dutch

Van Zalk and Jansen (2004) investigated the attractiveness of a text from an online travel brochure about three hiking areas in Ireland. One variant of the text used the Dutch formal pronoun of address u and the other variant used the informal form of address je. They found a positive effect of the formal variant: the formal V-text led to both greater interest in the content of the text and to more positive opinions about the topic. A post-hoc analysis revealed an interaction with age. While younger people rated the formal V-text as equally high or higher than the informal T-text, older people actually valued the informal T-text more. This result was the opposite of what van Zalk and Jansen had expected.

In a follow-up study by Jansen and Janssen (2005), participants rated formal and informal versions of two government texts. One of the texts (on violence on the streets) could count on high involvement, the other (on the Dutch food inspection service) much less so. The results showed that if readers of texts were less involved, the pronoun of address did not matter. However, readers who were more involved appreciated the text more if they were addressed with the formal pronoun of address u than with the informal one je. The age of the readers played no role in this. Notoriously, consumers who are more involved with a product have a more positive attitude toward an advertisement than consumers who are lowly involved (Zaichkowsky 1985, 1994). Personal involvement can be defined as “[a] person’s perceived relevance of the object based on inherent needs, values and interests” (Zaichkowsky 1985, 342). It can be measured by scales such as “[object to be judged] is of no concern – is of concern to me.”

Michaelidou and Dibb (2008) review the literature on involvement as an important concept of marketing and consumer behavior. They note that Zaichkowsky’s (1985) measures capture both enduring involvement, representing an individual’s attachment to a product or advertisement, and situational involvement, which represents an individual’s involvement in the purchase of a product. Several factors, such as the type of purchase, level of motivation, and brand choice, determine an individual’s temporal purchase involvement with a particular product. However, rather than manipulating consumers’ situational involvement, as is often done in advertising research, Jansen and Janssen (2005) measured their personal persistent involvement by asking them about their interest in the topic of the text and their agreement with the advice given in the text. They found that the degree of agreement with the advice given in the text appeared to influence the rating of the T and V variants of the text. If readers did not find the advice relevant or only moderately relevant, it did not matter whether they were addressed with V or with T. But the more involved they were with the text and in particular the more they agreed with the advice given in the text, the use of the formal V pronoun of address contributed to a higher rating than the use of the informal T pronoun. We predict the following:

H2: Consumers who are highly involved will respond more positively to advertisements than consumers who are lowly involved.

The type of involvement we will measure in the experiment is not the situational, but the enduring type (Zaichkowsky 1985; Michaelidou and Dibb 2008), which represents consumers’ attachment to a product, but independent of the advertisement for the product, and therefore independent of the pronoun used in the advertisement.

In an online experiment among native speakers of Dutch, de Hoop, Levshina, and Segers (2023) measured addressees’ responses to emails in HR communication. They examined whether ratings for a fictitious company and a fictitious recruiter differed depending on whether recipients were addressed with an informal or formal pronoun. Because the reduced informal pronoun je is considered more neutral and weaker than jij and jou(w), they used the stronger unreduced versions of the T pronoun as much as possible. The results showed that emails with the formal V pronoun were generally rated slightly higher than emails with the informal pronoun T. De Hoop et al. suggest that in formal situations such as a job interview, the formal pronoun may still be considered more appropriate. They note that the strong unreduced forms of the informal T pronoun may be particularly salient and out of place in a formal context.

Nevertheless, the informal T pronoun seems to be the preferred pronoun today in generic recruitment advertisements in the Netherlands (den Hartog, van Hoften, and Schoenmakers 2022) as well as in product advertising (Leung et al. 2022). In a series of five studies, Leung et al. (2022) established that consumers in general have a greater preference for advertisements featuring an informal T pronoun of address. In one of their studies, Dutch participants had to choose between two versions of slogans for twenty real brands, which were selected on the basis of warmth and competence ratings. One version of a slogan used the formal V pronoun u (e.g., Altijd dicht bij u ‘Always close to you’) and the other the informal T pronoun jij or jou(w) (e.g., Altijd dicht bij jou ‘Always close to you’). The results showed a clear effect of brand warmth and competence on consumers’ preferences for either slogan. Participants were more likely to prefer T-slogans for brands high in warmth, and to prefer V-slogans for brands high in competence. However, when both warmth and competence were high or both were low, consumers preferred the T-slogan.

In one other experiment, Leung et al. examined the effect of using T or V in Dutch on participants’ preference for one of two restaurants, describing one as competent with words such as excellent, capable, and professional, and the other as warm with words such as warm, friendly, and attentive. Next, subjects read one slogan with the name of the restaurant and your choice with either a T or a V possessive pronoun. When asked which restaurant they preferred, the majority of participants indicated a preference for the warm restaurant, although the number of participants who preferred the competent restaurant went up significantly when they read the V version of the slogan. De Hoop et al. (2023) suggest that the overall preference for the warm restaurant could be a result of the difference in descriptions, where the competent restaurant may have been perceived as more expensive by terms such as excellent, and therefore less appealing to the young subjects (all undergraduate students). Hence, the use of V and T in a slogan could not only be associated with a brand’s competence and warmth (Leung et al. 2022), but also with a higher or lower estimation of the brand’s product price.

Indeed, de Hoop et al. (2023) found that the price estimate of a product was higher for ads with the formal V pronoun. A possible explanation could be that V is generally used for older people (Levshina 2017), who are usually in a better financial position (Vismans 2013a). The use of V in the ads led to a larger effect on the estimated product price among older participants. In two experiments, de Hoop et al. (2023) examined the effects of using a formal V or informal T pronoun of address in product advertisements, not only on consumers’ ratings of the ads and the advertised products, but also on their price estimates. In one experiment, they found that the informal T pronoun in a Facebook advertisement text (150 words) for a robotic lawn mower positively affected participants’ rating of the attractiveness and persuasiveness of the advertisement, but only among older participants. Also, the informal T pronoun led to a lower product price estimation among older participants. De Hoop et al. (2023) suppose that the two results may be related, in the sense that a higher rating of the advertisement goes hand in hand with a lower price estimation. The second experiment made use of advertisement slogans for different products, and here advertisements with both T and V pronouns led to a significantly higher attitude toward the advertisement, the advertised product, and purchase intention than advertisements without pronouns, but advertisements with T and V pronouns did not significantly differ from each other in these respects. However, the use of a formal V pronoun led to a higher price estimate of the product, and this effect was stronger among older participants.

H3: Consumers will estimate the price of a product advertised with a V pronoun higher than the price of a product advertised with a T pronoun.

The above results raise the question what can explain the difference in preference between van Zalk and Jansen (2004), Jansen and Janssen (2005), and de Hoop, Levshina, and Segers (2023) on the one hand, and Leung et al. (2022) and de Hoop et al. (2023) on the other. The first three studies find a small but significant higher rating for a text that makes use of the formal V pronoun, while the latter two studies show consumers’ preference for the use of an informal T pronoun in Dutch. De Hoop, Levshina, and Segers (2023) suggest that the difference may be caused by the difference in type of text, as also suggested by Jansen and Janssen (2005). Leung et al. (2022, 11) also point to the communicative context as a potential moderator, when they assume that “the impact of brand personality on address preference might be stronger in a branding context (e.g., mass media advertising) than in contexts which are less focused on emphasizing the brand’s personality (e.g., communicating with customers by email about some service issue).” This might indeed explain the difference that was found between a preference for the V pronoun in the longer persuasive texts and recruiter emails versus the preference for the T pronoun in shorter product ads (Leung et al. 2022; de Hoop et al. 2023). Using the unexpected V pronoun in an advertising slogan could then elicit less positive reactions in consumers. Based on the above literature review we put forward our fourth hypothesis:

H4: Consumers will respond more positively to advertisements featuring an informal pronoun of address T than to advertisements with a formal pronoun of address V.

The experiment

In order to test the four hypotheses we conducted a questionnaire in which we investigated the impact of pronouns of address on Dutch consumers’ evaluation of advertisements and advertised products. The experiment was approved by the Ethics Assessment Committee Humanities of the Radboud University Nijmegen (ETC-GW number 2021-9598).

Participants

393 Dutch speakers were recruited through network/snowball sampling and completed the full questionnaire. Data from one participant were removed, as they did not meet the minimum age requirement of 16. Data from five more participants were removed because they correctly identified the purpose of the experiment (i.e., by explicitly mentioning the way in which consumers are addressed). Data from 387 participants were thus entered into statistical analysis (263 female, 122 male, 1 non-binary, 1 “prefer not to say”), whose ages ranged between 16 and 75 years old (M = 37.58, SD = 16.22).

Materials

The stimulus material consisted of a diverse set of fictitious advertisements with a slogan and an image. The products advertised were designed within three main categories and subjected to a pretest to create a diverse but balanced item set. Participants in the pretest (n = 32) were presented with 34 fictitious advertisements and were asked the questions below and instructed to respond to these questions on a 100-point scale.

• Is the product depicted above more essential or more luxurious?

• Is the product depicted above more masculine or more feminine?

• How cheap or expensive is the product depicted above?

For each of the three questions the two highest and two lowest rated responses were included in the main experiment. The experiment thus contained twelve experimental items in total, which included a museum card,¹ a weekly meal box subscription, health insurance, an energy supplier, lipstick, mascara, beard oil, aftershave, wine, a toothbrush, a laptop, and a coffee machine. The latter four advertisements also included the names of two cheap and two expensive brands (respectively: Lidl, Action, Apple, DeLonghi) to further diversify the item set. Four filler advertisements without a form of address were added to the item set, in an attempt to distract participants from the purpose of the experiment (shampoo, hair gel, cleaning service, and car insurance). An overview of all advertisement slogans can be found in Table 1.

Table 1. Overview of the slogans used in the advertisements.

Product	T/V slogan	Control slogan
mascara	Jij/u moet deze mascara echt proberen ‘You really need to try this mascara’	Deze mascara is echt het proberen waard ‘This mascara is really worth trying’
lipstick	Zo’n goede lippenstift heb/heeft jij/u nog nooit gebruikt ‘You’ve never used a lipstick this good before’	Een betere lippenstift is er niet ‘There is no better lipstick’
aftershave	Zo’n lekkere aftershave heb/heeft jij/u nog nooit gekocht ‘You have never bought such a tasty aftershave before’	Dit is de lekkerste aftershave ‘This is the tastiest aftershave’
beard oil	Jij/u vindt deze baardolie vast geweldig ‘You will love this beard oil’	Deze baardolie is geweldig ‘This beard oil is awesome’
bottle of wine	Geniet jij/u ook van een glaasje? ‘Do you enjoy a glass too?’	Ook genieten van een glaasje? ‘Enjoying a glass too?’
tooth brush	Wist jij/u dat het verstandig is om elke drie maanden een nieuwe tandenborstel te gebruiken? ‘Did you know that it is wise to use a new toothbrush every three months?’	Elke drie maanden een nieuwe tandenborstel gebruiken is verstandig ‘It is wise to use a new toothbrush every three months’
laptop	Thuiswerken alsof jij/u op kantoor bent? ‘Working from home as if you were in the office?’	Thuiswerken net als op kantoor ‘Working from home just like in the office’
coffee machine	Wil/wilt jij/u thuis ook de lekkerste koffie zetten? ‘Would you like to make coffee as delicious at home?’	Thuis ook de lekkerste koffie zetten? ‘Making equally delicious coffee at home?’
museum card	Wil/wilt jij/u onbeperkt musea bezoeken? Dan kan/kunt jij/u een museumpas nemen ‘Visit museums whenever you want? Then you should get a museum card’	Onbeperkt musea bezoeken? Neem een museumpas ‘Unlimited museum visits? Get a museum card’
weekly meal box	Jouw/uw wekelijkse maaltijdboxen. Download onze app en jij/u krijgt 50% korting op eerste box ‘Your weekly meal boxes. Download our app and you get 50% discount on first box’	Wekelijkse maaltijdboxen. Download onze app en krijg 50% korting op eerste box ‘Weekly meal boxes. Download our app and get 50% discount on first box’
energy supplier	Ben/bent jij/u op zoek naar een passende energiemaatschappij? ‘Are you looking for a suitable energy supplier?’	Op zoek naar een passende energiemaatschappij? ‘Looking for a suitable energy supplier?’
health insurance	Wil/wilt jij/u zonder zorgen genieten van het leven? Dan neem/neemt jij/u een uitgebreide zorgverzekering ‘Do you want to enjoy life carefree? You can with comprehensive health insurance’	Genieten van het leven zonder zorgen? Neem dan een uitgebreide zorgverzekering ‘Enjoy life carefree? Get comprehensive health insurance’
shampoo	n.a.	Benieuwd naar deze vernieuwde shampoo? ‘Curious about this renewed shampoo?’
styling gel	n.a.	Test nu de beste stylinggel ‘Test the best styling gel now’
cleaning service	n.a.	Op zoek naar een persoonlijke schoonmaker voor in huis? ‘Looking for a personal cleaner?’
car insurance	n.a.	Ook goed verzekerd op weg? ‘Well insured on the road?’

Slogans in the experimental advertisements were written in T or V, or without a pronoun of address (control condition C). Three versions of a sample advertisement can be found in Figure 1. The four filler items did not contain a pronoun of address but, with regards to statistical analysis, were not included in the control condition C either. These items were the same for each participant. The experiment was conducted in Qualtrics.

Figure 1. Three versions of an advertisement for mascara. Condition C: “This mascara is really worth trying”; Condition T and V: “You (T/V) really need to try this mascara”.

Design and procedure

A within-subject design was used. Each participant saw all sixteen items and all conditions of the factor pronoun (T, V, C). That is, they saw the twelve experimental advertisements in one version only, and for each participant four ads were presented in T, four in V, and four in C. The four filler advertisements were the same for all participants.

Participants first read an information and consent form and agreed to the terms and conditions of participation. This form addressed the participants in T, because this seems nowadays the most natural form to use. Participants then read the general instructions of the experiment, in which no form of address was used: Several ads will be shown with a number of questions per ad. Please answer as honestly as possible. There are no wrong answers. Read the ads and the questions carefully as it is not possible to go back to a previous page. Completing the questionnaire will take about 15 min.

We used the statements below to quantify four dependent variables, namely Attitude toward the advertisement (1), Attitude toward the product (2), Purchase intention (3), and Price estimation (4). The first three statements were based on Hornikx and Hof (2008), the last scale (semantic differential) on de Hoop et al. (2023).

1. I find this advertisement appealing.

2. I find this product attractive.

3. I want to purchase this product.

4. My estimation of the price of this product is [cheap – expensive].

We chose to select one statement to measure attitudes (1–2) and intention (3), instead of using five or three separate semantic differentials as Hornikx and Hof (2008) did, because the experiment would otherwise take too long. The choice for the specific statements was based on the likeliness of their results to be affected by the choice of address pronoun (T vs. V). Participants were instructed to indicate on 100-point slider scales to what extent they agreed with the statements (1–3) and whether they estimated the product as cheap or expensive (4). The slider bar was initially set to the middle of the scales and participants were forced to move the slider bar in order to continue to the next item, which was presented on a new page. The experiment did not allow participants to return to earlier advertisements. We forced participants to move the slider bar so as to eliminate the issue of non-response and/or overestimation of the initial value (see Funke 2015). Note that the use of different scales (7-point Likert scales or 100-point slider scales) generally does not change the outcome of experiments much (see e.g., Cook et al. 2001; cf. Funke and Reips 2012). Here, we chose for a 100-point scale to provide participants with a way to nuance their answers, because in our experience the reported T/V-attitudes are subtle at best. Participants did not see the numerical value corresponding to their response and were thus invited to rely on their intuition.

After the main part of the experiment, participants were asked to indicate on another 100-point slider scale to what extent they were involved with each product. For this, we used the statement “A [product] is really for me.” Participants were asked to what extent they agreed with the statement, providing us with an index of by-participant product involvement (cf. Jansen and Janssen 2005). Participants were then asked to provide their gender (male, female, non-binary, prefer not to say), age, and education level (secondary, vocational, higher professional, university). Finally, participants were given the opportunity to give qualitative comments about what they thought was the purpose of the experiment.

Analysis

We used the software R (version 4.2.1; R Core Team 2022) and the package lme4 (Bates et al. 2015) to perform a series of linear mixed effects regression analyses with Attitude toward the product, Attitude toward the advertisement, Purchase intention, and Price estimation as the dependent variables. Each base model included the fixed effects of pronoun (C, T, or V), which was Helmert-coded (C vs. T and V; T vs. V). The random structure of the base models included by-participant and by-item varying intercepts. By-participant varying slopes for the effects of pronoun were added to the models when this improved the model fit (as assessed by Likelihood Ratio tests). This was the case for the two Attitude models and for the Price estimation model, but not for the Purchase intention model.

After having established the base models, other potentially relevant factors (covariates) and their interactions with pronoun were added to the model one by one: age (mean-centered), involvement (measured by-item on a 100-point scale), and education level (university, higher professional, vocational, secondary; with secondary education set as the reference level). The factor or interaction was included in the model if it improved the model fit. We used Likelihood Ratio Tests to determine p values for main and interaction effects using the package afex (Singmann et al. 2022), which were then Bonferroni-corrected to counteract putative problems with inferences due to multiple comparisons using the package emmeans (Lenth 2022). That is, the p values printed in the tables below represent Bonferroni-corrected values whenever applicable. We further used the package MuMIn (Barton 2022) to calculate the (pseudo) R² of each model as a measure of the model fit.

Results

The results of the questionnaire are visually presented for each of the judgment scales in Figure 2. Notice that the figure does not fully represent the scale available to participants, which were 100-point scales. Differences between the mean scores of the conditions are only small. Visual inspection suggests that the T-forms are generally better appreciated than V-forms on all four scales, with the control condition rated somewhere in between. Further, the mean scores are on the lower end of the scales.

Figure 2. Mean scores per scale per condition (error bars represent 95% confidence intervals).

The fixed effects included in the final model were the same for the dependent variables Attitude toward the product, Attitude toward the advertisement, and Purchase intention: pronoun, age, involvement, and education level – without the interactions between them. The final model for Price estimation only included a fixed effect for pronoun.

The estimates of the final model with Attitude toward the advertisement as dependent variable are presented in Table 2. The analysis indicates that pronoun has a significant effect in the model (χ² (2) = 6.55, p = .038). The control condition C did not significantly differ from T and V (p = .999), but the advertisements with T and V were rated significantly differently (p = .032). Differences between the control condition C and T (p = .458) and between C and V (p = .495) were not significant. We further find significant effects of education (χ² (3) = 11.38, p = .010), involvement (χ² (1) = 94.55, p < .001), and age (χ² (1) = 7.76, p = .005). These findings indicate that scores increase the more involved participants were with the product, but decrease for older and higher educated participants.

Table 2. Model specifications Attitude toward the advertisement.

Fixed effects	β	SE	t	p
Intercept	43.66	3.82	11.433		<.001
C vs. TV	−0.02	0.86	−0.02		.999
T vs. V	−2.92	1.14	−2.56		.032
Education (HBO)	−7.79	3.29	−2.37		.110
Education (MBO)	−2.80	3.42	−0.82		.999
Education (WO)	−5.55	3.33	−1.67		.578
Involvement	0.11	0.01	9.82		<.001
Age	−0.12	0.04	−2.81		.005
Random effects	Variance	SD
By-participant intercept	125.5	11.20
By-participant slope (C vs. TV)	83.3	9.13
By-participant slope (T vs. V)	227.9	15.10
By-item intercept	58.3	7.63
Residual	542.8	23.30
	R^2marginal: .034	R^2conditional: 0.331

The estimates of the final model with Attitude toward the product as dependent variable are presented in Table 3. The analysis shows that pronoun did not have a significant effect in the model (χ² (2) = 4.25, p = .119). The control condition C did not significantly differ from T and V (p = .707), nor did T significantly differ from V (p = .208), C from T (p = .174), or C from V (p = .999). The model yielded significant effects of education (χ² (3) = 11.82, p = .008), involvement (χ² (1) = 324.67, p < .001), and age (χ² (1) = 12.09, p < .001). The findings indicate that scores increase with increased involvement and with higher education, but decrease with age.

Table 3. Model specifications Attitude toward the product.

Fixed effects	β	SE	t	p
Intercept	38.49	3.78	10.18		<.001
C vs. TV	0.46	0.50	0.93		.707
T vs. V	−1.64	0.90	−1.82		.208
Education (HBO)	6.89	2.90	2.37		.109
Education (MBO)	2.55	3.02	0.85		.999
Education (WO)	4.06	2.94	1.38		.999
Involvement	0.22	0.01	18.47		<.001
Age	−0.13	0.04	−3.53		<.001
Random effects	Variance	SD
By-participant intercept	85.7	9.26
By-participant slope (C vs. TV)	23.3	4.83
By-participant slope (T vs. V)	92.6	9.62
By-item intercept	80.0	8.94
Residual	569.7	23.87
	R^2marginal: .085	R^2conditional: 0.311

The estimates of the final model with Purchase intention as dependent variable are presented in Table 4. The analysis shows that pronoun did not have a significant effect in the model (χ² (2) = 3.24, p = .198). The control condition C did not significantly differ from T and V (p = .630), nor did T significantly differ from V (p = .407), C from T (p = .312), or C from V (p = .999). The model yielded significant effects of education (χ² (3) = 11.77, p = .008), involvement (χ² (1) = 385.41, p < .001), and age (χ² (1) = 16.21, p < .001). This indicates that scores increase with increased involvement and higher education, but decrease with age.

Table 4. Model specifications Purchase intent.

Fixed effects	β	SE	t	p
Intercept	26.36	3.38	7.80		<.001
C vs. TV	0.50	0.50	1.01		.630
T vs. V	1.291	.865	1.49		.407
Education (HBO)	3.19	2.95	1.08		.999
Education (MBO)	−2.17	3.06	−0.71		.999
Education (WO)	1.08	2.98	0.36		.999
Involvement	0.23	0.01	20.16		<.001
Age	−0.15	0.04	−4.07		<.001
Random effects	Variance	SD
By-participant intercept	90.1	9.49
By-item intercept	42.0	6.48
Residual	578.0	24.04
	R^2marginal: .104	R^2conditional: .270

The estimates of the final model with Price estimation as dependent variable are presented in Table 5. The analysis shows that pronoun did not have a significant effect in the model (χ² (2) = 2.55, p = .279). The control condition C did not significantly differ from T and V (p = .507), nor did T significantly differ from V (p = .800), C from T (p = .378), or C from V (p = .999).

Table 5. Model specifications Price estimation.

Fixed effects	β	SE	t	p
Intercept	51.58	5.30	9.73		<.001
C vs. TV	0.58	0.51	1.14		.507
T vs. V	0.90	0.91	1.11		.800
Random effects	Variance	SD
By-participant intercept	52.0	7.21
By-participant slope (C vs. TV)	92.7	9.63
By-participant slope (T vs. V)	74.2	8.61
By-item intercept	334.3	18.28
Residual	351.9	18.76
	R^2marginal: < .001	R^2conditional: .544

Conclusion and discussion

The aim of this study was to investigate the impact of using an informal T or a formal V pronoun of address in Netherlandic Dutch advertisement slogans. Based on the literature, in particular Yu, Hudders, and Cauberghe (2017), we hypothesized that consumers would respond more positively to an ad with a pronoun of address (T or V) than to an ad without a pronoun of address (H1). On the basis of Leung et al. (2022) and de Hoop et al. (2023), we expected that participants would respond more positively to an informal T pronoun of address than to a formal V pronoun of address (H4). Moreover, based on the findings of de Hoop et al. (2023) we expected that consumers would estimate the price of a product advertised with a T slogan lower, and the price of a product advertised with a V pronoun higher (H3). A second aim was to establish whether consumers’ involvement with the advertised products had an influence on the impact of T and V pronouns of address. We hypothesized that highly involved consumers would respond more positively to the advertisement than lowly involved consumers (H2), where involvement was measured independently for each product.

Participants were shown 16 diverse product ads each consisting of a picture of the product and a corresponding slogan. The impact of the presence of T and V pronouns compared to the control condition (H1) and the difference between T and V pronouns (H4, H3) was measured through attitude toward the advertisement and the product, purchase intention, and price estimation. After the main part of the experiment, participants were asked about their gender, age, and level of education. They were also asked to indicate to what extent they were involved with each type of product, by means of the statement “A [product] is really for me” in order to test H2.

As for the influence of involvement, results provide support for our second hypothesis. Indeed, we found that more involved participants appreciated the advertisement as well as the product more, and they had a higher purchase intention, compared to less involved participants. As pointed out above, the experiment did not manipulate participants’ involvement to see the effects of product involvement on the appreciation of the advertisement, the advertised product, and purchase intention. Rather, we measured product involvement independently after the main part of the questionnaire. In general, we found lower appreciations of the advertisements and products for older participants, as well as a lower purchase intention.

No main effect of pronouns was found for three out of four dependent variables, i.e., attitude toward product, purchase intention, and product price estimation. This is somewhat surprising because de Hoop et al. (2023) found a significant difference between V and T with respect to price estimation. The difference between de Hoop et al.’s and our study is that they found the difference between V and T in price estimation in one experiment in which only one product ad was evaluated and in another experiment in which three product ads were evaluated, viz. a laptop, a coffee machine, and an exercise bike. In our experiment, the products varied much more in terms of price range: a bottle of wine from a cheap brand (Aldi) and a toothbrush are obviously many times cheaper than a health insurance or a laptop. The differences in price range between the experimental items in our study are so large on a scale from cheap to expensive that they may have overshadowed other effects. It is, then, not surprising that the difference between the pronouns T and V did not play a significant role in price estimation. However, our results did show a significant effect of the use of T or V on attitude toward the advertisement, in the sense that consumers found the advertisement more appealing when the advertisement slogan featured a T pronoun compared to a V pronoun.

Thus, while van Zalk and Jansen (2004), Jansen and Janssen (2005), and de Hoop, Levshina, and Segers (2023) found a slight preference for the formal V pronoun in Dutch, Leung et al. (2022), de Hoop et al. (2023), and the present study found a general preference for the informal T pronoun in advertising slogans. One suggestion that Jansen and Janssen (2005) made, but did not explore further, is that text genre could make a difference as well. They suggested that readers of an advertising text may want to keep more distance and therefore would value the formal V pronoun more. Clearly, our finding that consumers value the T pronoun more than the V pronoun in advertising slogans does not corroborate this prediction. It is not clear, however, that Dutch consumers feel more distance, i.e., less addressed, when they read the formal V pronoun in comparison to when they read the informal T pronoun. Both van Zalk and Jansen (2004) and Jansen and Janssen (2005) made use of the informal T pronoun je in Dutch, which is in fact a reduced form of the informal T pronoun jij (subject), jou (object), and jouw (possessive pronoun) used in our study. Notoriously, this reduced pronoun is very often used generically (de Hoop and Tarenskeen 2015), while this is less often the case for the formal V pronoun, and also less often for the unreduced T pronouns (Tarenskeen 2010). We suspect that in short advertising slogans the second person pronouns T and V get a generic reading, irrespective of their exact form, and therefore do not differ in their level of personalization, i.e., the degree to which customers feel personally addressed when they encounter the pronoun. Using the informal T pronouns in advertising slogans could have contributed to the positive evaluation of the advertisement, because they are the default pronouns used in advertisements (Leung et al. 2022).

Strikingly, while the results support our fourth hypothesis, they do not support our first hypothesis. That is to say, consumers did not value being addressed with a second person pronoun more than being not addressed with a second person pronoun in an advertisement slogan. Clearly, this goes against our first hypothesis, which was based on research that found a significant positive effect of the use of a second person pronoun in advertisements in English (e.g., Cruz, Leonhardt, and Pezzuti 2017). The question is how to explain this.

The results of our study show a general trend in which T leads to a higher evaluation of the ad than V in Netherlandic Dutch, while the control condition C with no pronoun is in the middle. This suggests that the T pronoun not only leads to a higher rating of the ad, but also, rather unexpectedly, that V actually leads to a lower rating of the ad, not only lower than T, but also lower than the control condition without a pronoun. Therefore, the presence of a second person pronoun (T and V taken together) does not lead to a higher appreciation of the ad than the absence of a second person pronoun. So it seems that it truly matters whether consumers are addressed by T or V in an ad slogan. An effective ad, at least one that is better rated, does not simply have a second person pronoun as a pronoun of address, but specifically the informal T pronoun. Moreover, the formal V pronoun is better avoided because it leads to a lower rating of the ad.

Acknowledgements

This publication is part of the project ‘Processing pronouns of address: The impact of being addressed with a polite or an informal pronoun’ with project number 406.20.TW.011 of the research programme Open Competition SSH, financed by the Dutch Research Council (NWO). We are grateful to an anonymous reviewer and the editor for their valuable comments on an earlier version of this article. We thank Charlotte van den Broek and Fieke Litjens for designing part of the stimuli and collecting part of the data.

Disclosure statement

The authors report there are no competing interests to declare.

Data availability statement

The data that support the findings of this study are openly available in a Radboud Data Repository at https://doi.org/10.34973/jwwq-p517.

Additional information

Funding

This work was supported by the Dutch Research Council (NWO) under grant number 406.20.TW.011.

Notes

1 Note that four experimental items and two filler items advertise services rather than products. For the purposes of this paper we abstract away from this difference and treat services as products.