Abstract
Consumers often resist marketing information about death-related products and services (DRPS), let alone purchase them. Yet, there is a massive market and universal necessity for DRPS offerings like life insurance and funeral services. Proactively considering DRPS in advance allows consumers to better prepare for death and its aftermath. The current research builds on terror management theory and the literature on language use and psychological distance to propose a language-based sales communication strategy for DRPS. The findings from eight studies, including two field experiments and an eye-tracking study, indicate that DRPS sales messages that appear in consumers’ second (vs. native) language (e.g., using muerte, Spanish for death, when talking to a native English speaker whose second language is Spanish) decrease consumers’ fear of death by creating greater psychological distance to death, which in turn induces more consumption (e.g., actual purchases). However, this effect becomes attenuated when consumers perceive high control over death or when the DRPS feature transcendence after death. In contrast, the effect is amplified for DRPS that require an intermediate level of customer participation. These findings offer novel insights into how marketers and policymakers can motivate consumers to consider DRPS earlier and engage in more proactive decision-making.
Keywords
In many cultures, death remains a taboo and “unmentionable” topic (Crespo-Fernández 2023). Although people acknowledge that death—whether their own or that of loved ones—is inevitable and that advance planning is beneficial, they often avoid death-related information, potentially due to an innate self-preservation instinct (Kemp and Kopp 2011). For example, less than 25% of people over age 50 plan for their funeral (Levy 2015), and only about 1% of the Chinese population prepares a will in advance (Gardner 2017), leaving bereaved family members in a passive, vulnerable, and difficult position (Gabel, Mansfield, and Westbrook 1996; Gentry et al. 1995). Yet, proactively considering death-related products and services (DRPS)—defined as products or services that help individuals prepare for death and its aftermath—offers instrumental benefits, such as securing loved ones’ financial futures through life insurance, ensuring that personal wishes are honored via advance directives, and alleviating emotional distress for grieving families.
Moreover, the market for DRPS is vast, driven by over 100,000 global deaths daily (Worldometer 2025). This industry includes life insurance, funeral products, and innovative offerings like memorial diamonds made from cremated remains (around $24,000 per carat; Algordanza 2025). With the global death care services market valued at $147.1 billion in 2024 and projected to reach $209.6 billion by 2030 (Deep Market Insights 2025), the economic stakes are high. In the United States alone, for instance, funeral expenses reached up to $12,000 in 2025 (Lincoln Heritage Life Insurance Company 2025). Overall, this critical industry could offer even greater economic and social value if marketers helped consumers overcome their mortality avoidance and engage with DRPS earlier.
Drawing on terror management theory (TMT; Pyszczynski, Greenberg, and Solomon 1999) and extant research on language use as a strategy for increasing psychological distance (see Table 1), we propose that presenting DRPS sales information in consumers’ second language (L2), rather than their native language (L1), increases consumption by psychologically distancing them from death-related content and subsequently reducing fear of death (i.e., fears of mortality and nonexistence). For instance, using “muerte” (Spanish for death) rather than “death” (in English) could create psychological distance to death and reduce fear for English native speakers who know Spanish as L2. However, we expect this effect to be attenuated when consumers perceive higher control over their own death or when the DRPS highlight a sense of transcendence after death (i.e., a disembodied existence beyond physical death; Hood and Morris 1983) and stronger if the product or service requires an intermediate level of customer participation. Across eight studies, we tested and found support for our predictions using various L2–L1 pairs (e.g., Chinese–Japanese, French–English, English–German) and DRPS (e.g., organ donation, cancer insurance, biodegradable urn) via multiple methods (eye tracking, field, and online studies). We captured actual behavioral outcomes (e.g., organ donation registration, insurance purchases) and explored the boundary conditions of purchases made for close others and non-DRPS purchases.
Summary of Literature of Second Language Use on Psychological Distance.
Our research contributes to the literature on bilingualism and psychological distance by applying the foreign language effect to the substantive domain of DRPS sales communication. We not only offer practical insights into how L2 use can attenuate fear of death and enhance DRPS purchases but also advance the field by demonstrating that the psychological distancing effect of L2 extends to contexts where existential concerns and emotional defenses are particularly salient. Furthermore, we provide a nuanced empirical account of when and how the language effect operates by identifying critical boundary conditions. Our findings suggest that language is not merely a vehicle for information but a psychological lever that can fundamentally alter how consumers cope with mortality in the marketplace, providing actionable guidance for marketers and policymakers to encourage consumers to foster earlier and more proactive engagement with DRPS.
Theoretical Background and Hypothesis Development
Death-Related Products and Services (DRPS) and Fear of Death
The instrumental benefits of early engagement with DRPS are substantial. For example, early purchase of life insurance mitigates financial risks, protecting family members by securing their future (Quinn 2008). End-of-life preplanning also alleviates emotional distress by ensuring that personal wishes are honored (Detering et al. 2010; Kemp and Kopp 2010) and reducing logistical burdens for the bereaved—ultimately fostering security and peace of mind. Providers of DRPS frequently urge consumers to preplan end-of-life affairs, as it is the right thing to do (SupportiveCare 2025). Despite these benefits, research on death-related consumption remains limited and is primarily exploratory or qualitative. For instance, scholars interviewed funeral industry professionals (Korai and Souiden 2017) and consumers about death rituals (Bonsu and Belk 2003), and bereavement (Gentry et al. 1995).
In practice, consumers exhibit substantial resistance to DRPS due to profound psychological barriers (Huang, Du, and Qu 2021; Klackl, Jonas, and Kronbichler 2013). According to TMT, people adopt distal defenses when thoughts of death are unconscious (e.g., unintentionally thinking of mortality when reading news about a plane crash; Huang, Huang, and Jiang 2018). However, when death-related thoughts are conscious (e.g., deliberately thinking of the risk of death when experiencing turbulence on a flight), people engage with proximal defenses by seeking distractions or denying their vulnerability to death (Hayes et al. 2010). DRPS promotions inevitably make death salient and conscious in consumers’ mind, thereby sparking proximal defenses. In such circumstances, consumers experience an intense fear of death (Lambert et al. 2014) and attempt to suppress these thoughts from consciousness. They tend to ignore DRPS promotion information and avoid communicating with DRPS sales, which ultimately reduces the likelihood of DRPS consumption. Therefore, mitigating the fear of death during DRPS sales interactions can lower consumer resistance and increase purchase intent.
Some may counterargue that a reduced fear of death will diminish interest in DRPS consumption and, consequently, purchase behavior. We suggest, however, that consumer decision-making operates via a dual-process framework involving both an immediate, affective response (System 1), and a deliberative, cognitive evaluation of instrumental benefits (System 2) (Rottenstreich, Sood, and Brenner 2007; Shiv and Fedorikhin 1999). Consumers’ decisions regarding DRPS consumption could result from the interplay between these two systems. Specifically, when confronted with DRPS-related information, fear of death functions as an automatic, intensive emotional barrier (System 1) that overrides slower and more effortful evaluations of the products’ benefits (System 2), thereby inducing avoidance behaviors (Huang, Du, and Qu 2021; Klackl, Jonas, and Kronbichler 2013). This aligns with reinforcement sensitivity theory, which similarly predicts that fear triggers avoidance rather than approach behaviors (Perkins, Kemp, and Corr 2007). Importantly, reducing fear does not eliminate the underlying need for DRPS; rather, it “unlocks” the cognitive pathway. When the overwhelming emotional response is alleviated, consumers are better able to recognize and appreciate the instrumental, rational benefits of DRPS—such as providing financial security for one's family, minimizing logistical burdens for survivors, and attaining peace of mind (Detering et al. 2010). Therefore, lowering the fear response allows cognitive reasoning to become the primary determinant, facilitating DRPS consumption through a more objective assessment of its practical value. Taken together, because DRPS evoke the idea of death, making it proximal, consumers are likely to avoid information of DRPS. Such instinctive resistance creates considerable challenges for DRPS marketing, and we propose that using consumers’ L2 in DRPS sales communication might help alleviate this issue by creating psychological distance to death, reducing fear, and prompting consumption (Figure 1).
Conceptual Model.
Language Use, Psychological Distance, and Consumers’ Fear of Death
Multilingualism is widespread, with bilingual people constituting nearly half of the global population (Newsdle 2023). Fluency in two languages is increasingly common due to globalization and immigration trends (Maher 2017). Duolingo, a leading language-learning app, witnessed a 51% growth among daily active users in 2024 (D’Souza 2025). Using L2 is also common in mediated communications (e.g., email; Marcella and Davies 2004).
Language is a basic yet crucial component of marketing communication (Packard and Berger 2023). Existing research mainly focuses on linguistic phenomena within a specific language, such as the meaning conveyed by linguistic structures (e.g., phonemes; Lowrey and Shrum 2007; Yorkston and Menon 2004) or styles of marketing messages (e.g., concreteness; Packard and Berger 2021). Meanwhile, emergent research has centered on choices among different languages, particularly among native and second languages (L1 vs. L2; see Web Appendix A for a review): Some suggest that L2 use leads to inferior marketing outcomes (e.g., increased consumer dishonesty; Gai and Puntoni 2021), while others point to positive effects (e.g., healthier food selections; Klesse, Levav, and Goukens 2015). A few studies further suggest that the effectiveness of L2 depends on the specific foreign language used (Noriega and Blair 2008) and its congruence with the product (Hornikx, Van Meurs, and Hof 2013). Notably, psychology research also addresses the effect of L2 use on moral decision-making and gambling behaviors (Gao et al. 2015; Hayakawa et al. 2017). Framing moral dilemmas or risks in L2 reduces emotional or irrational decision-making (Costa et al. 2014), such that individuals would choose to kill an innocent person to save more lives (Corey et al. 2017) and display less risk aversion in betting (Keysar, Hayakawa, and An 2012).
Collectively, prior research has established that processing information in L2 (vs. L1) tempers emotional reactions (Pavlenko 2017). As Ivaz, Griffin, and Duñabeitia (2019) put it, L2 serves as an emotional buffer that absorbs the stark emotional impact of affective states by creating psychological distance. Psychological distance is a subjective experience of something as near to or distant from the self, here, and now, reflecting social, spatial, or temporal dimensions (Trope and Liberman 2010). Distancing the self along any dimension enables one to transcend the here and now and move beyond direct experience (e.g., plan ahead, take others’ perspectives; Trope and Liberman 2010) by allowing people to form abstract representations and be less affected by emotions (Moran and Eyal 2022). Prior research found that using L2 can increase psychological distance through representations at more abstract construal levels (e.g., Shin and Kim 2017; see Table 1 for a review). In our research context, L1 use brings thoughts of death psychologically closer to individuals’ egocentric experience, thus triggering proximal defenses and intense fear of death (Moran and Eyal 2022). In contrast, when processing death-related information in L2, people construe death through an abstract, high-level lens (Caldwell-Harris 2015) and thus perceive greater distance to death. Moreover, using L2 also decreases self-diagnosticity, such that consumers might feel like “someone else” (Gai and Puntoni 2021). Because language encodes psychological distance, and psychological distance reduces negative emotions (e.g., fear), we propose that using consumers’ L2 (vs. L1) in sales communications reduces their fear of death by creating psychological distance, leading them to be more receptive to DRPS sales information and to engage in more consumption. Formally,
Moderators of the Language Effect on DRPS Consumption
The effectiveness of marketing communication tools hinges on both what the product or service is and who the targeted consumers are (Kemp and Kopp 2011). Prior research suggests that consumer characteristics affect their reactions to terror (Mandel and Smeesters 2008). Thus, to identify potential moderators of the effect of L2 use, we consider both consumer characteristics and product features and posit that consumers’ perceived control over death and DRPS transcendence features might weaken the proposed language effect, whereas an intermediate level of customer participation in the service production process may amplify it.
Consumer perceived control over death
In line with prior literature (Han and Gershoff 2018; Skinner 1996), we define perceived control in our context as the extent to which people believe they can avoid undesirable events associated with death. We propose that the effect of L2 use on psychological distance to death depends on consumers’ perceived control over death. Most people feel little or no control over mortality and thus automatically resist reminders of death (e.g., talking about or contemplating it; Gorer 1955; Yalom 2008). Prior research has established that people with a lower sense of control are more reluctant to take action and more susceptible to negative emotions (Ross and Mirowsky 2013). Therefore, consumers who perceive lower control over death will be less willing to consider DRPS information.
Nevertheless, perceived control is malleable and can be manipulated (e.g., Cutright and Samper 2014; Han and Gershoff 2018). We propose that the effect of sales language use on psychological distance to death could be alleviated when consumers perceive more control over death. Gaining control gives people a sense of resourcefulness and makes it easier for them to face difficulties (Skinner and Zimmer-Gembeck 2010). They are less likely to experience helplessness and more likely to act to confront challenges (Rothbaum, Weisz, and Snyder 1982). Moreover, greater perceived control allows people to adopt more effective coping mechanisms to manage undesirable situations, such as pain (Bandura et al. 1987) and distress (Thompson et al. 1993). Thus, when consumers perceive higher control over death, the benefit of L2 use is attenuated, because the perceived control removes the defensive tendency against DRPS sales information when consumers use L1. In addition, we expect that when consumers feel in control over death, they are inclined to think about death in a less overwhelming, more objective, and more distant way even when using L1. Formally stated,
DRPS transcendence features
According to the TMT literature, fear of death derives from the existential terror of “not being” anymore (Thorson and Powell 2015). The recognition that their existence will cease fills many people with dread (Bradley 2015), prompting them to seek comfort from a sense of transcendence (e.g., religion, Jonas and Fischer 2006). Marketers could also leverage this appeal by highlighting the transcendence possibilities in DRPS, such as biodegradable burial urns that enable consumers to return mineral-rich remains to the earth and nurture a tree after they die (Capsula Mundi Team 2020). This product feature evokes connections between dead bodies and something lasting (e.g., trees). When consumers gain a sense of transcendence, they think of their lives continuing in another form after death. By transcending the here and now and going beyond the present, they might also perceive an altered psychological distance (Moran and Eyal 2022). That is, consumers might distance themselves from their actual, unavoidable death by transcending it through a form of continuous existence (Vandecreek and Nye 1993). Overall, regardless of the language use, if transcendence increases the psychological distance between consumers and their death, it should mitigate their fear of death.
DRPS customer participation
Consumers are value cocreators (Payne, Storbacka, and Frow 2008) who devote effort and time to service production processes (Chan, Yim, and Lam 2010). Such customer participation is common for DRPS. For example, funeral providers in Ghana ask families to wash and dress the deceased family member (Korai and Souiden 2017), and some end-of-life planning services require people to become fully immersed in envisioning their funeral and make plans (Winn 2019). Prior research has categorized three levels of customer participation, positing low, intermediate, and high levels in the service design and delivery process (Dong et al. 2015), and has found that these different levels could yield nonlinear effects on customer satisfaction and loyalty (Stokburger-Sauer et al. 2016; Wang, Wu, and Wei 2021). Building on prior research, we propose that as the level of customer participation in DRPS varies from low to intermediate to high, the effect of L2 use may differ.
First, we propose that the language effect may become more pronounced as customer participation increases from a low to an intermediate level. When consumers provide information or feedback, they shift from passive recipients to active participants, which enhances their involvement and emotional arousal (Auh et al. 2019). One way of providing such feedback is through making choices (Dong et al. 2015), which activates more episodic memories, as individuals frequently rely on past experiences to evaluate options and make informed choices (Bornstein and Norman 2017). In the DRPS context, making end-of-life medical decisions by actively choosing (vs. passively reading) options demands greater cognitive effort, thereby heightening emotional responses and psychological reactance (Botti, Orfali, and Iyengar 2009). Although consumers generally prefer the autonomy of making their own decisions, this preference weakens in highly emotional, high-stakes situations such as end-of-life decisions, where making choice intensifies feelings of causality and responsibility (Botti, Orfali, and Iyengar 2009). Accordingly, using L2, which is less associated with death-related memories than L1, may more effectively help consumers distance themselves from thoughts of death. In other words, the beneficial effect of L2 is amplified when the level of customer participation increases from a low level (e.g., reading about end-of-life plans) to an intermediate level (e.g., making choices for end-of-life plans).
However, when the customer participation level is high, the language effect may diminish. High participation requires consumers to draw more extensively on their own knowledge and skills to complete increasingly complex tasks (Yim, Chan, and Lam 2012). In this context, deeper involvement may shift cognitive processing and lead consumers to become more deeply immersed in death-related memories. According to language-specific episodic trace theory, words presented in L2 can activate corresponding L1 counterparts through an indirect route (i.e., L2 → L1 → L2), facilitating access to L1-coded memory traces before reproducing the lexical representation in L2 (Puntoni, De Langhe, and Van Osselaer 2009). Thus, when tasks require high participation—such as writing a funeral plan rather than simply reading or selecting options in L2—consumers are prompted to initially generate death-related ideas in L1 (Göpferich and Nelezen 2014). This process then probes L1-coded episodic memory traces, infusing the L2 stimuli with some of the emotional weight associated with L1. Consequently, this indirect route diminishes the psychological distance to death, as the L2 stimuli become more emotionally charged, reducing the buffering effect of L2 use against fear of death and thus alleviating the proposed language effect. In contrast, when participation ranges from low to intermediate, the direct route (L2 → L2) prevails, meaning that there is little activation of the lexical representation of death in L1. Overall, DRPS with high levels of customer participation lower the effectiveness of L2 use in creating psychological distance and reducing fear of death. Hence, we propose that:
Overview of Studies
We report on eight main studies to examine the proposed effects of L2 (vs. L1) use on DRPS consumption. We also test the underlying mechanisms and proposed moderators. Briefly, we offer field evidence for the main effect of L2 use on consumers’ pamphlet-taking and organ donation registration (Study 1a) and their actual purchase of cancer insurance (Study 1b). Study 2 uses eye-tracking techniques to affirm the key mediating role of fear of death with physiological measures (e.g., pupil dilation). Studies 3a and 3b investigate psychological distance and fear of death as the serial mediators for our proposed language effect, replicate our results when consumers purchase DRPS for close others (Study 3a), and explore non-DRPS contexts as a boundary condition (Study 3b). Studies 4–6 test the proposed moderators: consumer perceived control over death, DRPS transcendence features, and DRPS customer participation, respectively. Across these studies, we use different DRPS manifestations (e.g., organ donation, cancer insurance, biodegradable urn), diverse L2–L1 pairs, and varied questionnaire languages (i.e., the same language as the stimuli in Studies 1a, 2, and 3b; participants’ L2 in Studies 4 and 5; and participants’ L1 in Studies 1b, 3a, and 6). 1 All participants’ bilingual eligibility was filtered by translation tasks and an attention check. Participants who failed these filter questions were excluded. We also include comprehension checks (e.g., true/false and multiple-choice questions). Web Appendices B–D contain summaries of the studies, measurements, and attrition rates. All studies received ethical approval from an institutional review board. All questionnaires and data can be found on an open-source platform (https://researchbox.org/3489). Web Appendix H provides a sample informed consent statement.
Study 1: Field Experiments
Study 1a: Organ Donation
Studies 1a and 1b are field experiments that test H1, the proposed main effect of language use on DRPS consumption, with actual behavioral outcomes. In Study 1a, we examine organ donation, a critical DRPS that is still new and not widely accepted in Asia but contributes significantly to societal welfare (Johnson and Goldstein 2003).
Method
This field experiment used a two-cell, between-subjects design (language use: English as L2 vs. Chinese as L1). English is the professional language in Hong Kong, where local professionals use English as L2 with international colleagues and clients. In this study, a bilingual confederate, unaware of our hypotheses, approached native Chinese professionals to promote organ donation for the local government. The confederate initially approached 379 professionals who work for international institutions in various industries (e.g., education, finance, business, law) through email invitations. The confederate ultimately was able to present the information to 120 of these professionals, either face-to-face or through Zoom (55.8% female, 44.2% male; 70.8% between 30 and 60 years old). 2 Using a randomly assigned language (English or Chinese) throughout the whole procedure, the confederate emphasized the importance of organ donation, described which organs and tissues can be used (e.g., liver, outer layer of eyes, heart), and outlined the organ donation process and benefits, using a script adopted from the official promotional brochure provided by the local government (see Web Appendix E1, Figure WA-1). Next, the confederate gave each participant a pamphlet (in English or Chinese, according to the condition) and asked if they wished to keep it for more information (see Web Appendix E1, Figure WA-2). Six months later, 3 the confederate followed up with each participant to determine whether they had registered.
Pamphlet-taking
We found that 68.3% of participants took the pamphlet after listening to the promotional message in Chinese (L1), but significantly more, 86.7%, took the pamphlet after listening to it in English (L2: χ2(1) = 5.78, p = .016, odds ratio [OR] = 3.01).
Organ donation registration
After six months, 8 participants were not reachable, leaving a final sample of 112 (57.1% female, 42.9% male; 72.3% between 30 and 60 years old). In the L1 condition, only 10.7% of participants registered as organ donors after six months. Significantly more, 26.8%, did so in the L2 condition (χ2(1) = 4.75, p = .029, OR = 3.05). The results remained significant when we included industry and interaction mode (face-to-face or Zoom) as covariates (see Web Appendix G1). Thus, promotional information presented in L2 (vs. L1) prompted more pamphlet-taking and more organ donation registration, supporting H1. Moreover, we conducted additional analyses on how pamphlet-taking might influence registration rates. Overall, there was no significant relationship between pamphlet-taking and sign-up rates in general (25.9% for takers vs. 16.5% for nontakers, χ2(1) = 1.20, p = .273, OR = .56) without considering language use. Importantly, we found a higher sign-up rate in the English (L2) condition among those who took the pamphlet (L1 = 5.4%, L2 = 25.0%, χ2(1) = 5.83, p = .016, OR = 5.83), but this effect was absent among the nontakers (χ2(1) = .79, p = .373, OR = 2.25). We speculate that the nontakers had already made up their minds, whereas the takers were still considering registration, making the L2 use effective for the takers and warranting further research on motivations behind pamphlet-taking.
Study 1b: Cancer Insurance
The global insurance market is projected to grow from $8.3 trillion in 2025 to $11 trillion in 2030 (The Business Research Company 2026). Yet, in China, insurance penetration has reached only 3.9%, far behind the global average of 6.8% (Rudden 2023). This shows a significant market potential for insurance among Chinese consumers. Given the high cost of cancer care, we adopt cancer insurance promotion in Hong Kong as our research context.
Method
We collaborated with two senior insurance sales managers, equally experienced and bilingual in English and Chinese, from a leading international insurance firm based in Hong Kong. 4 Over two weeks, we recruited 147 valid postgraduate Chinese mainland students 5 (76.9% female, 23.1% male; Mage = 24.57 years) and randomly assigned them to one of two identical rooms on campus, where we manipulated the language conditions, such that the cancer insurance plan was introduced in English in one room and in Chinese in the other.
The insurance introduction was around 45 minutes and consisted of three parts: (1) statistics about cancer in Hong Kong; (2) information about cancer treatment expenses; and (3) the plan overview, benefits, and premium (Web Appendix E2). Next, participants completed a short survey about their understanding of the key insurance message as a comprehension check. They also indicated their age and mode of L2 acquisition and basic demographic information (age, gender). Two months after this sales encounter, we contacted the insurance agents to find out which participants had purchased the cancer insurance plan. 6
Purchase
In the Chinese (L1) condition, only 8% of participants purchased the cancer insurance, whereas significantly more participants (19.4%) did so in the English (L2) condition (χ2(1) = 4.09, p = .043, OR = 2.78). Thus, presenting cancer insurance in the participants’ L2 (vs. L1) prompted their purchases.
Discussion
These field experiments offer convergent evidence that L2 (vs. L1) use increases actual consumption of DRPS (H1), represented by pamphlet-taking, organ donation registration, and cancer insurance purchases. However, we acknowledge that both studies involved confederates whose behaviors or beliefs could have influenced participants’ responses, raising the possibility that social influence or conformity effects impacted our results (Kuhlen and Brennan 2013). To address this, we designed subsequent studies to eliminate the use of confederates and thereby minimize these potential sources of bias.
Study 2: Sea Burial
In a controlled lab setting with eye tracking, we test fear of death as a key emotional driver of DRPS consumption. We also examine sea burial, an ecofriendly corpse disposal service, to test H1's robustness. In Hong Kong, green burials comprised only 15.4% of funerals in 2022, highlighting a need for more promotion (Government of HKSAR 2023).
Method
With a two-cell (language use: English as L2 vs. Chinese as L1) between-subjects design, Study 2 collected valid responses from 112 administrative employees at a large Hong Kong university (88.4% female, 11.6% male; Mage = 34.60 years), all native Chinese speakers proficient in English as a L2. In a lab, participants were randomly assigned to answer a survey in English or Chinese, on a computer with an eye tracker (Tobbi Pro X3-120). The survey contained promotional information about government sea burial service, highlighting its benefits like ecofriendliness, and contained a gloomy picture of typical burial niches in Hong Kong, noting that sea burials help “ease the shortage of niches and graves” (Web Appendix E3, Figure WA-4, Panel A). The eye tracker captured fear through two physiological indicators: the pupil dilation and the first fixation duration 7 (Hopkins et al. 2015; Sharpe et al. 2017). Participants who feel greater fear tend to exhibit larger pupil sizes and shorter first fixation durations when viewing the frightening picture (i.e., avoid looking at it).
After reading the material, participants indicated their likelihood of service adoption (“When in need, how likely are you to apply for the sea burial service for yourself?”; 1 = “very unlikely,” and 9 = “very likely”) as an indicator of DRPS consumption. We also measured the extent to which participants found the material easy to understand (1 = “not easy to understand at all,” and 9 = “very easy to understand”), their L2 proficiency (“In general, your language proficiency of English is …”; 1 = “novice-like,” and 7 = “native-like”), their chronic worries about death due to fatal illnesses (“I think the chance that I will get [an] incurable disease is …”; 1 = “very low,” and 7 = “very high”) or accidents (“I think the chance that I will accidentally die is …”; 1 = “very low,” and 7 = “very high”; Lehto 2011), and age and gender.
Results
Fear of death
To analyze pupil dilation, we set the average pupil size of participants when they read the introductory information, prior to encountering the frightening picture, as a baseline (L2 = 3.04 ± .34 mm; L1 = 3.08 ± .45 mm; t(110) = .54, p = .587, d = .10; Hopkins et al. 2015). Then we identified their maximum pupil size during the presentation of the frightening picture, after removing blinks (L2 = 3.50 ± .36 mm; L1 = 3.69 ± .67 mm; t(110) = 1.86, p = .066, d = .35; Kafkas and Montaldi 2012). The pupil dilation response equaled the percentage change from the baseline pupil size to the maximum pupil size, such that we controlled for individual differences in pupil sizes (Hopkins et al. 2015). According to a one-way analysis of variance (ANOVA), participants in the L1 condition exhibited greater pupil dilation (M = 119.88%, SD = 11.55%) than participants in the L2 condition (M = 115.46%, SD = 7.05%; t(110) = 2.44, p = .016, d = .46), indicating that the former were more aroused when they saw the frightening picture with information presented in L1.
With regard to the duration of the first fixation, we created an area of interest (AOI; see Web Appendix E3, Figure WA-4, Panel B) over the image stimuli. In the AOI, we deliberately included the frightening picture only, without any text, because a first fixation duration on text might reflect the time or effort required to process L1 or L2 texts, rather than fear. A one-way ANOVA revealed that participants’ first fixation on the image in the L1 condition was significantly shorter (M = .24 s, SD = .11 s) than that in the L2 condition (M = .30 s, SD = .16 s; t(110) = 2.39, p = .018, d = .45), indicating that participants directed their attention away from the death-related stimuli in the L1 condition more quickly than the L2 condition. As such, we extrapolated that consumers felt more fear when looking at death-related information alongside the DRPS promotional message in their L1 compared with their L2.
DRPS consumption
The results of another one-way ANOVA indicated that participants were significantly less likely to adopt the service for themselves in L1 (M = 4.71, SD = 2.31) than in L2 (M = 5.75, SD = 2.28; t(110) = 2.39, p = .019, d = .45). This difference remained significant after we incorporated the control variables (see Web Appendix G3).
Mediation analysis
A mediation test with PROCESS Model 4 (Hayes 2017) indicated a significant mediating effect of fear of death, whether operationalized by pupil dilation (b = .371, SE = .157, 95% confidence interval [CI]: [.0887, .6992]) or first fixation duration (b = .313, SE = .160, 95% CI: [.0474, .6682]), of the effect of providing information in L2 (vs. L1) on consumers’ willingness to adopt the DRPS.
Discussion
With eye-tracking data, we obtain objective evidence of a mediating effect of fear of death (manifested by pupil dilation and first fixation duration) on the effect of L2 (vs. L1) use on consumers’ adoption of a green burial option.
Study 3: Boundary Conditions
Study 3a: Cancer Screening Service for Self or Close Others
In addition to the focal mediator, fear of death, in Studies 3a and 3b, we measure psychological distance to death to address its potential mediating role (H2). Purchases of DRPS might be for oneself or for close others (e.g., parents). We suspect that the language effect observed in purchases for the self also appears for purchases for close others, as close others represent an extended self (Aron et al. 1991; Belk 2014).
Method
This study featured a 2 (language use: Chinese as L2 vs. Japanese as L1) × 2 (beneficiary: self vs. close others) between-subjects design. Four hundred participants received a cancer screening service introduction from a research data collection agent. 8 All participants are native Japanese speakers who also know Chinese as a foreign language (31.0% female, 69.0% male; Mage = 52.45 years). Initially, the survey revealed to participants (in Chinese or Japanese) that the researchers sought to learn their opinions about cancer screening services and whether they would consider purchasing for themselves (self condition) or a family member who is closest to them and whom they care about most (close others condition). They were asked to keep their own initials or the initials of the close family member in mind while reading about the cancer screening service (Web Appendix E4). Participants then indicated their psychological distance to death (e.g., “How close to or distant from your death do you feel at this moment?”; 1 = “very close,” and 9 = “very distant”; r = .76). They also noted their fear of death with three items (e.g., “I have a scary idea that I would disappear forever with my death”; 1 = “strongly disagree,” and 9 = “strongly agree”; α = .92; Huang and Wyer 2015) and their consumption intentions (e.g., “The probability that I would consider buying the cancer screening service is …”; 1 = “very low,” and 9 = “very high”; α = .96). Moreover, we asked a comprehension check question (i.e., true or false question: “Only people aged 70 and above die from cancer”). We also collected participants’ demographic information (e.g., age, gender), efforts to understand, L2 proficiency, age and mode of L2 acquisition, perceptions of the formality of the stimuli, and chronic worries about death by fatal illness or accidents as control variables (Web Appendix C). In all subsequent studies, we included this same set of control variables as covariates.
Psychological distance to death
A 2 × 2 ANOVA with the psychological distance to death as the dependent variable indicated a significant main effect of language (F(1, 396) = 14.32, p < .001,
Fear of death
A 2 × 2 ANOVA with the participant's fear of death as the dependent variable indicated a significant main effect of language (F(1, 396) = 14.83, p < .001,
Consumption intentions
A 2 × 2 ANOVA with consumption intentions as the dependent variable revealed a significant main effect of language (F(1, 396) = 19.01, p < .001,
Serial mediation
Noting the lack of interaction between language use and beneficiary, we separated the data according to the beneficiary and ran PROCESS Model 6 twice, to test the serial mediation effects. In the self condition, the serial mediating effect was significant (b = .025, SE = .017, 95% CI: [.0012, .0651]), suggesting that psychological distance and fear of death sequentially mediate the effect of language use on consumption intentions. When the beneficiary was close others, the serial mediation was also significant (b = .038, SE = .025, 95% CI: [.0025, .0980]).
Discussion
With Study 3a, we reaffirm the serial mediating influences through psychological distance to death and fear of death, in support of H2. We further reveal that the proposed foreign language effect appears regardless of whether the beneficiary is self or close others, demonstrating the robustness of our findings.
Study 3b: Life Insurance (DRPS) and Property Insurance (Non-DRPS)
Study 3b tests a boundary condition of the language effect, by including a non-DRPS offering. Specifically, we focus on non-DRPS with emotional connotations that are less likely to trigger mortality salience. For clarity in comparison, we use property insurance as a representative of non-DRPS and life insurance to exemplify DRPS.
Method
We adopted a 2 (language use: Spanish as L2 vs. English as L1) × 2 (service type: DRPS vs. non-DRPS) between-subjects design. On Prolific, we recruited 400 native English speakers who know Spanish as a L2. However, 13 participants were excluded due to preregistered exclusion criteria, 9 resulting in 387 qualified participants (50.9% female, 48.1% male, 1.0% others; Mage = 35.45 years). In the DRPS condition, the life insurance, named Life Power Cover, covered common causes of death, including critical illnesses and accidents. In the non-DRPS condition, the property insurance, Property Power Cover, covered common causes of property loss, such as climate-related and human causes. We then briefly introduced the payment scheme, which was the same across both conditions (Web Appendix E5).
After reading the insurance information, participants indicated their psychological distance to death on a diagram, with a column of two dots with varying distances, such that one represented the participant and the other represented their death (adapted from Wang, Touré-Tillery, and McGill [2023]), as well as their fear of death (α = .94) and consumption intentions (α = .96). A manipulation check referred to the service type (i.e., “To what extent do you think the life insurance is related to death/associated with loss of life?”; 1 = “not at all,” 9 = “very much”; r = .89).
Manipulation check
A 2 × 2 ANOVA with the service type perceptions as the dependent variable yielded a significant main effect of service type (MDRPS = 7.06, SD = 1.71; Mnon-DRPS = 3.80, SD = 2.28; F(1, 383) = 243.44, p < .001,
Psychological distance to death
A 2 × 2 ANOVA with the psychological distance to death as the dependent variable indicated a marginally significant main effect of language (F(1, 383) = 2.84, p = .093,
Fear of death
A 2 × 2 ANOVA with participant's fear of death as the dependent variable indicated a significant main effect of language (F(1, 383) = 4.96, p = .027,
Consumption intentions
A 2 × 2 ANOVA with consumption intentions as the dependent variable revealed nonsignificant main effects of language (F(1, 383) = 1.55, p = .213,
Effects of Language Use on Consumption Intentions for DRPS (Life Insurance) vs. Non-DRPS (Property Insurance) (Study 3b).
Serial mediation
We used PROCESS Model 85 (Hayes 2018) to test the moderated serial mediation, with language use as the predictor, psychological distance and fear of death as serial mediators, service type as the moderator, and consumption intentions as the outcome variable. We found a significant moderated mediation (index = .028, SE = .018, 95% CI: [.0019, .0695]). The serial mediating effect was significant (b = −.023, SE = .014, 95% CI: [−.0572, −.0022]) in DRPS conditions: Psychological distance and fear of death serially mediated the language effect on consumption intentions. We also used PROCESS Model 6 to get the path coefficients of significant serial mediations for this and following studies (see Figure 3). However, for non-DRPS, the serial mediating effect was not significant (b = .005, SE = .009, 95% CI: [−.0115, .0239]). 10
Path Coefficients for Conditions with Significant Serial Mediations (Studies 3a–6).
Discussion
The results confirm the serial mediation of psychological distance to death and fear of death underlying the effect of L2 (vs. L1) use for DRPS (e.g., life insurance) consumption intentions, but not for emotionally charged non-DRPS offerings such as property insurance. We further conducted a posttest (N = 240) to identify key factors influencing property insurance purchase decisions. The results replicated our findings such that the effect of L2 (vs. L1) use on consumption intentions and fear was again observed in life insurance decisions, but not in property insurance decisions. For property insurance, practical factors like insurance premium (β = .35, p < .001) and coverage (β = .55, p < .001) outweigh emotional influences of fear (β = .10, p = .176) in determining the property insurance purchase decisions (Web Appendix F3). While these findings are robust, we acknowledge that employing fear of death as a mediator for both DRPS (i.e., life insurance) and non-DRPS (i.e., property insurance) conditions may not be entirely comparable. Notably, the study did not specifically assess fear related to property damage, which might have yielded different results had the target of fear been differentiated. This limitation highlights an important avenue for future research.
Study 4: Sea Burial with Consumer Perceived Control
In Study 4, we test our prediction that consumer perceived control over death moderates the language effect, such that the proposed effect disappears if the consumers perceive high control over death (H3).
Method
We adopted a 2 (language use: French as L2 vs. English as L1) × 2 (consumer perceived control over death: low vs. high) between-subjects design. On Prolific, we collected 400 completed responses. After excluding those who have failed the filter questions, we retained 384 valid responses (66.9% female, 30.2% male, 2.9% others; Mage = 37.94 years). In either English or French, participants read an article that indicated people's low (high) control over their death, suggesting that nothing was more important than a person's genetic makeup (diet and lifestyle choices) (Web Appendix E6). Then they answered manipulation check questions (e.g., “How much do you agree that you have control over your own death?”; 1 = “strongly disagree,” and 9 = “strongly agree”; α = .80). Next, they read about sea burial services and indicated psychological distance (r = .78), fear of death (α = .90), and consumption intentions (α = .94).
Results
Manipulation check
A 2 × 2 ANOVA with the manipulation check measures as the dependent variable yielded a significant main effect of the control manipulation (Mhigh = 5.49, SD = 1.52; Mlow = 3.92, SD = 1.59; F(1, 380) = 97.94, p < .001,
Psychological distance to death
A 2 × 2 ANOVA with the participant's psychological distance to death as the dependent variable indicated a marginally significant main effect of language (F(1, 380) = 2.92, p = .088,
Fear of death
In a 2 × 2 ANOVA with fear of death as the dependent variable, we found a nonsignificant main effect of language (F(1, 380) = 1.96, p = .162,
Consumption intentions
With a 2 × 2 ANOVA with consumption intentions as the dependent variable, we found nonsignificant main effects of language (F(1, 380) = .60, p = .438,
Effects of Language Use on Consumption Intentions for DRPS When Consumers Have Low Versus High Perceived Control over Death (Study 4).
Serial mediation
With PROCESS Model 85 (Hayes 2018), we tested the moderated sequential mediation with language use as the predictor, psychological distance and fear of death as two serial mediators, consumer perceived control as the moderator, and consumption intentions as the outcome variable. With 5,000 bootstraps, we identified a significant moderated mediation (index = −.113, SE = .056, 95% CI: [−.2387, −.0128]). When perceived control over death is low, the serial mediating effect was significant (b = .098, SE = .040, 95% CI: [.0276, .1872]). We also used PROCESS Model 6 to get the path coefficients for these conditions (Figure 3). However, when perceived control over death is high, the serial mediating effect was not significant (b = −.015, SE = .035, 95% CI: [−.0893, .0512]), supporting H3.
Discussion
With Study 4, we confirm H3 that when consumers have high perceived control over death, the effect of salespeople's use of L2 (vs. L1) on DRPS consumption, through psychological distance to death and fear of death, becomes attenuated.
Study 5: Biodegradable Urn and Transcendence
With another DRPS, biodegradable urns, we test whether DRPS transcendence features moderate the language effect (H4). When DRPS enable consumers to distance themselves psychologically from their unavoidable death by transcending it, consumers’ fear of death should be reduced, such that the language effect also likely is attenuated.
Method
We adopt a 2 (language use: English as L2 vs. German as L1) × 2 (DRPS transcendence features: no vs. yes) between-subjects design, using a biodegradable urn (similar to https://www.capsulamundi.it/shop/). On Prolific, we recruited 520 native German speakers who also know English as a L2. Seven participants were excluded due to preregistered exclusion criteria, 11 resulting in 513 eligible participants (43.9% female, 55.9% male, .2% others; Mage = 39.62 years).
We created four versions of a video embedded with audio and subtitles (around 2 minutes, similar in length to actual sales videos), each of which introduced a biodegradable urn, with or without transcendence features, in English or German (Web Appendix E7). The video noted the limited location choices in German public cemeteries and their expensive maintenance. Then, it introduced a nonprofit company, Final Farewell, and its new burial service, which would put remains in a biodegradable urn and bury it in a forest. In the nontranscendence condition, the service was called “Rest in Forest,” and the biodegradable urn in the video appeared next to a tree, with the caption “Your remains will decompose into nutrition in the soil and return to nature … and the deceased can quietly rest in the forest.” In the transcendence condition, the service was called “Become a Tree,” and the biodegradable urn was placed under a tree, with the caption “Your remains will decompose into nutrition in the soil and nurture your tree… the deceased can continue to live as a tree.”
After watching the advertisement, participants indicated their psychological distance to death using the diagram item from Study 3b, fear of death (α = .81), DRPS consumption intentions (α = .92), and a manipulation check of transcendence (e.g., “I feel like I will live on past my own death,” 1 = “not at all,” and 9 = “very much”; α = .91; Dunn, White, and Dahl 2020).
Results
Manipulation check
A 2 × 2 ANOVA with the manipulation check as the dependent variable yielded a significant main effect of this product feature (Mtranscendence = 5.22, SD = 2.30; Mnontranscendence = 4.43, SD = 2.30; F(1, 509) = 14.88, p < .001,
Psychological distance to death
A 2 × 2 ANOVA with psychological distance to death as the dependent variable indicated a significant main effect of language (F(1, 509) = 4.57, p = .033,
Fear of death
A 2 × 2 ANOVA with fear of death as the dependent variable yielded nonsignificant main effects of language (F(1, 509) = 1.56, p = .212,
Consumption intentions
With a 2 × 2 ANOVA using consumption intentions as the dependent variable, we found a significant main effect of language (F(1, 509) = 7.08, p = .008,
Effects of Language Use on Consumption Intentions for DRPS Featuring Nontranscendence Versus Transcendence (Study 5).
Serial mediation
We used PROCESS Model 85 (Hayes 2018) to test the moderated serial mediation effect, including language use as the predictor, psychological distance and fear of death as two serial mediators, transcendence features as the moderator, and consumption intentions as the outcome variable. We found a significant moderated mediation (index = −.021, SE = .013, 95% CI: [−.0504, −.0019]). When the DRPS did not feature transcendence, the serial mediating effect of psychological distance and fear of death was significant (b = .020, SE = .011, 95% CI: [.0031, .0444]). We also used PROCESS Model 6 to get the path coefficients for this condition (see Figure 3). Yet, when the DRPS featured transcendence, the serial mediating effect was not significant (b = −.001, SE = .006, 95% CI: [−.0151, .0119]), in support of H4.
Discussion
In Study 5, we confirm that when a DRPS does not feature transcendence, psychological distance to death and fear of death serially mediate the effect of language use on DRPS consumption intentions, which is attenuated when DRPS features transcendence.
Study 6: Participating in Funeral Planning
In Study 6, we investigate another DRPS feature, required customer participation, which is common for funeral planning services (Winn 2019). To test the potential nonlinear moderating role of DRPS customer participation (H5), we used a 2 (language use: Spanish/French/German as L2 vs. English as L1) × 3 (DRPS customer participation: low vs. intermediate vs. high) between-subjects design.
Method
From Prolific, we recruited 610 participants who speak English as L1 and know either Spanish, French, or German as L2 12 ; 17 participants were excluded due to preregistered exclusion criteria, leaving us 593 qualified respondents (58.7% female, 40.5% male, .8% others; Mage = 45.08 years). In both language conditions, we manipulated customer participation as follows: In low-participation conditions, participants read statements listing the end-of-life service options (e.g., “The way your body will be treated after you die includes: Option 1: Traditional burial; …; Option 4: Others”). In intermediate-participation conditions, participants chose among given options (e.g., “How would you wish your body to be treated after you die? Traditional burial; …; or Others, please specify”). In high-participation conditions, participants wrote down their own preferred funeral plan details (Web Appendix E8). Then, we adopted the Becker–DeGroot–Marschak method as an incentive-compatible measure to elicit their true willingness to pay (WTP; O’Donnell and Evers 2019). Specifically, we asked all participants to indicate their WTP for a service trial on a £0 to £2 scale and informed them that if their WTP was higher than a randomly generated price, they might be selected to purchase the trial via a lottery. Selected participants could be compensated with a one-month premium membership from a (now defunct) end-of-life planning company, Cake, minus the amount they indicated. Following that, they indicated fear of death (α = .89) and psychological distance to death as in Study 3b. They also answered three manipulation check questions about customer participation (e.g., “The funeral planning process was very engaging”; α = .93), filter questions and control variables as in Study 3a, and a measure of perceived complexity of the funeral planning service (1 = “simple/easy,” and 9 = “complex/difficult”; r = .87). This perception arguably might interact with our manipulation of DRPS customer participation.
Results
Manipulation check
A 2 × 3 ANOVA with the manipulation check of customer participation as the dependent variable yielded only a significant main effect of customer participation (F(2, 587) = 13.35, p < .001,
Psychological distance to death
A 2 × 3 ANOVA with the participant's psychological distance to death as the dependent variable indicated a significant main effect of language (F(1, 587) = 12.31, p < .001,
Fear of death
A 2 × 3 ANOVA with fear of death as the dependent variable indicated significant main effects of language (F(1, 587) = 22.89, p < .001,
WTP
In a 2 × 3 ANOVA with WTP as dependent variable, we identified significant main effects of language (F(1, 587) = 18.86, p < .001,
Effects of Language Use on WTP for DRPS Featuring Low, Intermediate, and High Customer Participation (Study 6).
Serial mediation
We used PROCESS Model 85 (Hayes 2018) to test the moderated serial mediation and included language use as the predictor, psychological distance to death and fear of death as two serial mediators, customer participation as the moderator, and WTP as the outcome variable. The significant moderated mediation results indicated the differences of the indirect effects among participation conditions (index = −.006, SE = .004, 95% CI: [−.0148, −.0004]). Psychological distance to death and fear of death serially mediated the language effect on WTP in both low- (b = .015, SE = .007, 95% CI: [.0043, .0305]) and intermediate- (b = .009, SE = .004, 95% CI: [.0024, .0182]) participation conditions, but not the high-participation condition (b = .003, SE = .004, 95% CI: [−.0048, .0121]). We also used PROCESS Model 6 to get the path coefficients for both low- and intermediate-participation conditions (see Figure 3).
Discussion
In this study, we reveal a nonlinear moderating effect of DRPS customer participation, as predicted by H5. Specifically, we find that when the customer participation level increases from low to intermediate, the L2 effect gets stronger. Nevertheless, when the required customer participation reaches a high level, the language effect becomes attenuated. To verify our theorization of the dominant indirect route (L2 → L1 → L2) activated at high-participation levels, we conducted a posttest using a three-level (customer participation: low vs. intermediate vs. high) between-subjects design with 187 native Chinese speakers whose L2 is English (see Web Appendix F4). Participants completed a preplanning funeral service trial in L2, and then verbalized and recorded their thoughts during the trial following the retrospective think-aloud protocol (De Larios, Murphy, and Marín 2002) and uploaded the audio recordings. A three-item scale (e.g., “I frequently thought in Cantonese (L1) during the service trial”; 1 = “not at all,” and 7 = “very much”; α = .89, Sasaki and Hirose 1996) assessed the extent of thought processing in L1 during the service trial. Results showed that compared with low or intermediate levels of participation, participants in the high-participation condition (1) used fewer L2 words (log-transformed means: Mlow = 1.66, Mintermediate = 1.61, Mhigh = .75; high vs. low: p = .002, d ≥ .55; high vs. intermediate: p = .002, d ≥ .55; low vs. intermediate: p = .854, d = .03), (2) showed a higher percentage of L1 use (Mlow = 71.8%, Mintermediate = 79.0%, Mhigh = 96.3%; high vs. low: p < .001, d ≥ .51; high vs. intermediate: p = .004, d ≥ .51; low vs. intermediate: p = .249, d = .21), and (3) claimed a greater extent of L1 thought processing (Mlow = 4.98, Mintermediate = 5.26, Mhigh = 5.96; high vs. low: p < .001, d ≥ .49; high vs. intermediate: p = .005, d ≥ .49; low vs. intermediate: p = .271, d = .20), supporting the dominance of the indirect route when participation is high.
General Discussion
Despite death being an inevitable aspect of life, people commonly refrain from discussing or contemplating their own mortality. This tendency leads many to avoid marketing information related to end-of-life matters—even though advance planning can yield important benefits for both individuals and their loved ones (Quinn 2008; SupportiveCare 2025). We propose a viable communication tool to encourage consumers to foster earlier and more proactive engagement with DRPS: using consumers’ second language (L2) (vs. native language, L1). Introducing DRPS in consumers’ L2 (vs. L1) increases consumers’ psychological distance to death and reduces fear of death, which enhances their consumption. This effect also applies to DRPS purchases for close others but not to purchases of non-DRPS (e.g., property insurance). Furthermore, consumer characteristics and DRPS features can influence the effect of L2 use: When consumers perceive greater control over death, or when the DRPS promise transcendence after death, the language effect is attenuated. However, the language effect amplifies when DRPS require an intermediate (vs. low or high) level of customer participation. These findings remain robust across various language pairs and types of DRPS.
Theoretical Contributions
Our research makes several contributions to the extant literature. First, this study extends research on language use and psychological distance into the sensitive context of DRPS consumption. While prior work has primarily investigated the effect of L2 use in varied domains such as moral decision-making (Gai and Puntoni 2021), memory recall (Baumeister et al. 2017), and emotional responses (Caldwell-Harris 2015), we broaden the scope by applying the language effect to the more substantive domain of DRPS sales communication. Specifically, our findings reveal that the L2 use can alleviate fear of death by increasing consumers’ psychological distance to death, thereby increasing their DRPS consumption. This effort contributes to a deeper understanding of how linguistic choices can shape consumer behavior in contexts characterized by high emotional sensitivity.
Second, this study adds to the emerging marketing literature on death-related consumption by advancing beyond qualitative approaches (e.g., Bonsu and Belk 2003) and offering a more comprehensive and nuanced explanation of when and how L2 use influences consumers’ DRPS purchases with robust empirical evidence. Importantly, we identify key boundary conditions of the L2 effect (e.g., close others as beneficiary, non-DRPS product type), providing a more detailed understanding of this linguistic phenomenon.
Third, this research extends TMT and the literature on bilingualism to marketing contexts. Marketing research has addressed how distal defenses affect consumers (Mandel and Smeesters 2008), such that they increase spending to defend their worldview or reinforce connections to brands to bolster their self-esteem (Rindfleisch, Burroughs, and Wong 2009). However, the implications of proximal defenses for consumer behaviors, such as avoidance or suppression of the idea of death (Greenberg et al. 2000), are understudied. This research provides initial evidence that using L2 (vs. L1) to communicate about DRPS can reduce consumers’ proximal defenses by increasing their psychological distance to death. Moreover, our investigation offers a more nuanced view of the language effect by examining moderators, including customer characteristics and DRPS features such as customer participation, with a nonlinear moderating effect, highlighting the complex dynamics and intricacies of L2 usage.
Practical Implications
Our research offers implications for various stakeholders in DRPS consumption. First, considering the communication barrier of DRPS, marketers need to find effective ways to convey death-related sales information. We establish that using consumers’ L2 represents a viable option to boost their interest in DRPS. Thus, we recommend that marketers recruit salespersons proficient in multiple languages and train them to assess consumers’ language skills and speak in consumers’ L2 to reduce their resistance to sales information about DRPS.
Second, language use influences consumers through psychological distance, and firms can thus leverage it to help regulate consumer emotions. Previous studies suggest that distancing techniques require less effort than some other tactics (e.g., suppression; Powers and LaBar 2019). In addition to L2 use, for example, marketers can present more abstract information about DRPS (e.g., intangible benefits and rationale rather than specific features) in their initial appeals to increase consumers’ psychological distance to death (Williams, Stein, and Galguera 2014). Furthermore, marketers should be cautious when describing the probability of certain death-related events, as high perceived probabilities make death seem too close, which could increase consumers’ fear and hinder their DRPS consumption (Rim and Summerville 2014).
Third, firms can also intervene by leveraging DRPS product features. For example, marketers could emphasize transcendence features of DRPS when L2 use is not applicable. To further amplify the effectiveness of L2 use, marketers could foster intermediate levels of customer participation, thereby encouraging customers to express preferences and share information in L2 but not activating L1 processing. Leveraging the nonlinear moderating effect of customer participation, structured yet casual interactions (e.g., simplified L2 preference surveys or conversational prompts) could strike this balance. Staff training can also emphasize subtle techniques that gently steer conversations toward L2 while monitoring signs of discomfort, ensuring that participation remains within the optimal participation level.
Marketers also need to note other consumer characteristics beyond bilingualism, such as consumers’ perceived control over death. To determine if L2 use is likely to be influential, marketers could use surveys, for example, to learn about consumers’ sense of control over death or manipulate their perceived control. We note that marketers often portray end-of-life events as unpredictable and beyond consumers’ control to highlight the need for DRPS (e.g., Lambert 2022). However, our findings suggest the opposite: To encourage DRPS consumption, marketers should refrain from highlighting the inevitability of death and instead increase consumers’ perceived control by emphasizing the actions they can take.
Our findings offer important implications for consumer welfare and public policy. Presenting information in L2 can help consumers overcome their fear of death when confronted with death-related topics about themselves or close others. For their well-being, consumers are advised to browse relevant websites or discuss related topics in their L2. However, if they are not fluent in another language, consumers still can address their fear of death by focusing on the transcendent potential of DRPS or by cultivating perceived control over death. Policymakers might also use L2 to promote DRPS for positive societal outcomes. For example, they could develop promotional information in multiple languages and default to displaying the information in consumers’ L2 to nudge them to review it. For DRPS that are readily linked to transcendence notions (e.g., organ donation), policymakers might incorporate transcendence as a key element of their promotional practices. Such efforts may be particularly practical when the service process involves intermediate levels of customer participation. Finally, policymakers could promote messages that foster a sense of control over death, emphasizing that while it is ultimately unavoidable, it can still be managed.
Limitations and Directions for Further Research
Several limitations of the present study offer opportunities for continued research. First, DRPS are not the only offerings that elicit fear. Further research could examine whether the language effect extends to fear in other types of consumption, such as risky financial investments (Visser-Keizer et al. 2016). In addition, while we did not observe the language effect for property insurance, future research could employ more context-specific measures of fear to better examine the potential effect in other important product categories. Further investigation into the nuanced impact of L2 use on other non-DRPS products associated with different emotions (e.g., condom purchases related to embarrassment) may also help clarify the complexities of the language effect across diverse marketing contexts.
Second, based on the TMT literature, we examined the fear of one's own death due to existential terror or concerns of “not being” (Thorson and Powell 2015), but fear of death could also relate to the loss of family members. Our supplementary study (N = 200; Web Appendix F6) found that consumers’ fear of losing their own lives, rather than the fear of losing family members, reduced DRPS consumption. Interestingly, fear of losing family members is positively related to DRPS consumption; thus, this type of fear might drive consumers to approach, rather than avoid, death-related information. Moreover, it is important to note that our research centers on promoting preplanning for one's death, where the threat of death is not imminent. In other words, our proposed effects of psychological distance and fear of death on DRPS consumption might not hold when the threat of death is imminent (e.g., terminal illness). According to the mere urgency effect, limited time before a deadline heightens attentional focus and induces discomfort and tension, prompting immediate action (Zhu, Yang, and Hsee 2018). This “fight” (vs. “flight”) response contrasts with our proposed mechanisms of psychological distance and fear of death, which typically influence consumer decisions under normal circumstances. While these perspectives are compelling, further research could investigate contexts where our proposed effects may diminish or even reverse, such as situations where psychological distance might hinder DRPS consumption.
Third, our research focuses on the initial introduction stage, with the aim of reducing resistance and encouraging consumption. Future studies should examine whether L2 benefits extend to later stages, such as after sales or service recovery. Additional analyses indicate that L2 use may be more effective in offline (vs. online) and less formal settings (marginally significant, ps < .10; see Web Appendix G1 and G4). The results suggest that offline settings may amplify the L2 effect due to sensory cues and physical presence, whereas formal contexts may diminish it due to greater emphasis on professionalism. Future research can explore these dynamics.
Fourth, our findings reveal that several control variables varied across language conditions with mixed patterns. For L2-related variables, age of acquisition (AOA) differed marginally, with L2 conditions showing earlier AOA in Study 3a and later AOA in Study 3b. Mode of acquisition varied significantly: In L2 conditions, Studies 3a and 6 had more participants learning in mixed settings. In addition, L2 proficiency was higher in the L2 condition for Study 6. We speculate that the earlier the AOA, the more immersive mode of acquisition, and higher proficiency might give L2 stronger access to emotional memories (Wu, Shi, and Zhang 2023), potentially attenuating the language effect. For death-related controls, participants’ chronic worries about accidental death in Study 3b were marginally higher in L2 conditions, while their chronic worries about illness in Study 4 were marginally higher in L1 conditions. Since individuals often deny their own vulnerability to death when confronted with mortality reminders (Pyszczynski, Greenberg, and Solomon 1999), heightened chronic worries may diminish optimism bias and increase attention to unpredictable fatal events, thereby increasing consumers’ intentions to consume DRPS. Consequently, the language effect might be amplified (attenuated) in Study 3b (Study 4). For demographic-related controls, Study 5 had more male participants, potentially amplifying the language effect due to testosterone's inhibition of fear responses (Lebron-Milad and Milad 2012). While our results remain robust with these variables included as covariates (see Web Appendix G), the mixed patterns warrant further investigation. Moreover, the sensitive topic of DRPS could potentially lead to attrition, which might affect our results. However, our field (Studies 1a and 1b) and lab (Study 2) studies showed negligible attrition, and our online studies (Studies 3a–6) showed attrition below the 20% threshold of concern (Zhou and Fishbach 2016). Attrition patterns also varied between language conditions across studies: Attrition was higher in L2 for Studies 3b and 6, and in L1 for Study 5, while Study 4 showed no significant difference (Web Appendix D). We reason that higher L2 attrition may reflect increased cognitive load, while the higher L1 attrition in Study 5 could result from stronger negative emotional responses to video stimuli. Overall, the low and inconsistent attrition rates are unlikely to bias our findings. Future research could use strategies such as prewarnings to further reduce dropout in online studies.
Finally, future research on language use could gain from utilizing advanced monitoring technologies or software to detect and prevent real-time translation tool usage during online surveys. In addition, although we have gathered actual outcomes from our field experiments, the self-reported aspect of organ donation registration may introduce limitations. We encourage further studies to corroborate our data with official records or evidence when feasible.
Supplemental Material
sj-pdf-1-mrj-10.1177_00222437261440816 - Supplemental material for Death or Muerte? Effects of Sales Language Use on the Consumption of Death-Related Products and Services
Supplemental material, sj-pdf-1-mrj-10.1177_00222437261440816 for Death or Muerte? Effects of Sales Language Use on the Consumption of Death-Related Products and Services by Lexie Lan Huang, Xueni (Shirley) Li and Kimmy Wa Chan in Journal of Marketing Research
Footnotes
Acknowledgments
The authors thank a number of colleagues and seminar participants for feedback and helpful advice. They gratefully acknowledge the anonymous companies for sponsoring the field experiments. All errors and omissions remain the responsibility of the authors.
Authors Contributions
The three authors contributed equally and shared the first authorship.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is supported by grants from the Hong Kong SAR Research Grants Council, awarded to the second (HKBU22500521 and HKBU12500723) and third (HKBU12501222 and HKBU12501325) authors.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability
Supplemental Material
Supplemental material for this article is available online.
Notes
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
