Introduction

One basic function of language is information transmission, and this transmission can influence various aspects of our cognitive processes, including decision making. Since the last century, researchers have analyzed how language impacts on our decision-making, that is framing effects. Tversky and Kahneman [1] recruited two groups of participants to make choices regarding a common situation (The Asian Disease Paradigm). These choices were described as either "gain-framed" or “loss-framed” policies. The authors observed that when a case was phrased in terms of potential gains, people tended to choose the less risky option. Conversely, when the participants confronted a potential loss scenario, they gravitated towards the option that required greater risk. This experiment confirmed that language-framed statements (or problems) can clearly influence people’s decision-making processes. Over the decades, this theory has been replicated and accepted well by specialists from various fields, including but not limited to behavioral economists, linguists, and psychologists [2]. Although the study of framing effects generated by descriptive language has been investigated many times, it was only in the past few years that scholars started to analyze how the ‘nativeness’ of a language influences decisions [3].

More than half of the world’s population nowadays is capable of communicating in a foreign language [4], and this number has increased steadily in recent years. Along with the trend of globalization, English serves as the lingua franca that builds connections among speakers with different native tongues and diverse cultural backgrounds [5]. Early in the 21^st century, three quarters of English users were non-native speakers who learned it as a foreign language in classroom settings [6]. Due to this large population of foreign language users, researchers have recognized that it is vital to learn how and why the use of a foreign language influences an individual’s choices in different situations [7].

Numerous studies have shown that thinking in a foreign language can lead to both positive and negative decision-making processes and outcomes. On one hand, some scholars have stated that using a foreign language can reduce biases in decision-making, because this process enables people to rely on systematic processes that are more remote from the immediate intuitive system than they are when one is thinking in one’s native language [8]. For example, when using their native language, one study’s participants were less likely to overestimate the likelihood of positive outcomes after a series of previous successes than when the situations were presented in their second language [8]. In addition, Keysar, Hayakawa and An [9] reported that, when the Asian Disease Paradigm was introduced in a foreign language, asymmetry in framing effects was lessened in relation to both losses and gains. On the other hand, some researchers have also suggested that a foreign language may interfere with one’s train of thoughts from multiple perspectives. At first, performance may be limited by foreign language anxiety, especially during listening and speaking sessions. The anxious states of foreign language users can reduce the accuracy of both information input and output [10]. Also, generating messages in a non-native tongue may induce a greater burden on the cognitive load and hence reduce the efficiency of decision-making [11]. De-contextualization can also be triggered because each language requires a specific cultural background that is quite challenging for language learners to understand fully [12].

Various aspects of the effects of thinking in a foreign language have been investigated. These have included role inferences [13], moral judgements [14, 15], and affective processing [14]. These studies focused on the outcomes of high-level cognitive processes like reasoning, perception and thinking. Hence, the procedures used in these studies have always consisted of continuous responses from participants in multiple trials with scenarios that are relatively rare in daily life.

To investigate further how the use of a foreign language influences regular human performance in day-to-day living, an analysis of other daily higher-order cognitive processes may enrich understanding about the underlying psychological mechanisms and the linkage between foreign languages and human cognition [16]. This line of research should use simple designs and concrete contexts as the testing materials [17]. Game Theory may serve as one compatible and appropriate tool for studying the fundamental processes involved in strategic decision-making process. Therefore, the present study used several simple Game Theory scenarios that usually occur in real-life business settings to look into the relationship between the effects of foreign language and human cognition (in particular, our decision-making processes).

In this research area, previous experiments mostly investigated cases within Anglophone languages (e.g., English), Latin European Languages (e.g., Spanish, Italian and French) or Germanic European languages (e.g., Dutch and German). These languages belong to the Indo-European Language systems, and they share a certain degree of intercultural similarities [18]. Studies of foreign language effects on individual responses have rarely been conducted between Chinese and English languages [19]. Compared to all the languages used in past research endeavors, Chinese and English languages share comparatively few intercultural similarities. Therefore, the present study aimed to investigate the foreign language effects on decision-making from this cross-cultural perspective.

Based on Boroditsky’s studies is the view that thinking in one language leads to culture-based decisions [20, 21]. According to Boroditsky [21], different languages build up various meaning-generating frameworks due both to the rules of expression (grammar, words) and language experiences (patterns and values). Cultural accommodation has been applied widely to identify and explain the underlying mechanisms of foreign language effects [22]. When processing occurs in a certain language, cultural constructs of that language are activated. Harzing [23] invited participants from 24 different countries to respond to a questionnaire. Two experimental groups of non-native English speakers were established. In one of these groups, the participants were asked to provide responses in English, while those in the other group were required to respond in their native tongue, while the control group consisted of native English speakers. The results showed that cultural differences among regions were reduced when the participants responded to the questions in the language in which they were presented. Previous literature suggested that cultural accommodation may start from incidental acculturation [24]. Studying a new language may guide the new language learners to know about the corresponding culture. Norms and values of that culture eventually influence a learner’s perceptions. This results in modified behaviors influenced by language [25].

How do we classify the underlying cultures of different regions? Researchers have suggested numerous systems to induce unique characteristics. Cattell [26] concluded that patterns of culture emerged from the syntax dimensions. Hofstede [27] highlighted espoused values of individuals from various nationalities, and designed a classification system by assigning countries into cultural clusters based on four dimensions of variation: Power Distance, Uncertainty Avoidance, Individualism and Masculinity [28]. One of the most globalized classifications used currently was developed by the GLOBE project, which included 61 nations [29]. In this project, the researchers enhanced the original four dimensions proposed by Hofestede [30] by separating social practices and social values. This permitted the independent analysis of the influences of social values on individuals [31]. Here, because we aimed to study foreign language effects on individuals’ responses to strategic games (Cooperative vs. Competing, Volunteering vs. Selfish), the classification model proposed by the GLOBE project was most appropriate.

According to the GLOBE Project’s Societal Cluster classification model [29], the Chinese language belongs to the Confucian Asia Cluster, while the English language belongs to the Anglo Cultural Cluster. Compared to the Anglo Cultural Cluster, Confucian Asians are characterized by higher values in terms of both institutional and in-group collectivism, higher levels of uncertainty avoidance and more support towards masculinity in their mindsets [32]. People from the Anglo Cultures Cluster tend to be more agreeable to individualism and gender egalitarianism, but they are less likely to exhibit uncertainty avoidance [33]. When applying culture patterns to the analysis of individuals’ strategic behaviors, several assumptions emerge. First, because of the higher tendency towards collectivism, people immersed in Confucian Asian cultures may tend to play cooperative rather than competing roles. Second, due to their higher uncertainty avoidance, Confucian Asians may be less likely to cooperate when the costs of the cooperation are not clear. Therefore, when language contexts activate cultural priming, behavioral tendencies occur [34]. Related to the idea of cultural accommodation, the first hypothesis (H1) was about participant tendencies displayed when playing cooperative roles in the games in different languages. We predicted that in the Prisoner’s Dilemmas, Chinese natives would be less likely to cooperate when the contexts were displayed in English; while in the Volunteer’s Dilemmas, Chinese native speakers would be less likely to cooperate when the contexts were described in their native language.

In addition, past studies have indicated that foreign language processing does not influence people in the same ways, even if they share the same native tongue. These differences could be explained by several language-related factors. First, Volk and his colleagues [11] indicated that the increased cognitive loads in foreign language processing bounded the rationality of a human’s decisions. However, this was less likely to occur in the more proficient foreign language users [35]. Also, language users’ self-evaluations of their proficiency may be negatively related to the level of foreign language anxiety, which might lead to biases and misjudgments in decision making. Since foreign language proficiency can improve cognitive fluency, it could be a moderator of foreign language effects.

Second, a foreign language can increase the psychological distance, which can influence people to rely more on their general and fundamental attitudes [36]. More frequent use of the foreign language could reduce the psychological distance by involving more direct experiences in the foreign language processing [37]. Hence, frequency of use of the foreign language might influence the effects of foreign language processing.

Furthermore, as language priming leads people to accommodate the norms and values corresponding to that language, personal connection and agreeableness with the culture are related to the tendency to display cultural-based behaviors [34]. Hence, we assumed that cultural identity with the corresponding language could have an impact on the outcomes of foreign language priming. Therefore, to investigate the influences of the three factors discussed above, the second hypothesis (H2) focused on the correlations between the scale of foreign language effects and these three factors. We predicted that foreign language proficiency, frequency of use of the foreign language and cultural identity with the corresponding language could be correlated to the strength of the foreign language effects.

Method

Results

The main dependent variable in each trial was the binary choice between cooperating and competing. We coded the choice of “cooperating” as value 1 while the choice of “competing” as value 0. First, we conducted Chi-square tests of independence (χ) in each round to examine the association between the types of language and likelihood of cooperating (H1). Second, we used correlational analyses of the participant responses and other variables, namely the participants’ foreign language proficiency, frequency of foreign language use and cultural identity with the corresponding foreign language to examine the other hypothesis (H2). The percentages (%) of cooperation for the four games in each language group are presented in Fig 1. The correlation matrix of those language-related variables is presented in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. The percentages (%) of cooperation for the four games in each language group.

a: Percentage of cooperation in Prisoner’s Dilemmas; b: Percentage of cooperation in Volunteer’s Dilemmas.

https://doi.org/10.1371/journal.pone.0277556.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Correlations of the language-related factors when the contexts were presented in the foreign language.

https://doi.org/10.1371/journal.pone.0277556.t001

Discussion

Previous studies have demonstrated clearly that language affects the way we think, and recent studies have also reported that our (cognitive) processing in a foreign language not only alters our attitudes and perceptions, but also impacts actual behaviors [13]. Here, we extended these observations to another important cognitive process, decision making. In the present study, we used Chinese and English as the focus languages to investigate the foreign language effects on human decision making while playing strategic games. Past relevant research on how foreign language affects cognitive processing focused on a few higher-order thinking processes, such as moral judgment [49], reasoning [50] or time-perception [51]. This was one of the first experimental studies to use Chinese-English bilinguals in research of this nature. In general, the overall patterns of results were in line with the cultural accommodation hypothesis [22] as well as the risk-aversion preference when using one’s native tongue [9].

Our results from games 1 and 4 supported the main hypothesis about foreign language effects. In the Prisoner’s Dilemma Game, when the contexts were described in Chinese (their native tongue), the participants were more likely to provide cooperative responses than when the contexts were presented in the foreign language. First, the participants’ behavior patterns were in accord with the norms and values of the culture backgrounds corresponding to each language [52]. This supported the assumption of cultural accommodation. Language is a valid source of cultural priming since each language is associated with a specific background. Thus, when contexts presented in different languages activate cultural priming, behavioral tendencies will occur [53]. The negative correlation between cultural identity with the corresponding foreign language and the likelihood of cooperation could also be explained by the Cultural Accommodation Hypothesis, as the agreeableness and connection to cultures corresponding to the foreign language enlarged the performance of language-specific responses. Second, regarding the reduction in emotional resonance of processing in a foreign language [54], when the game was presented in English, the participants (or players) might have perceived a lower degree of risks in competing. On the opposite side, those who went through this game in their native language might have felt a relatively higher level of risks for defecting [35]. Therefore, the risk-aversion tendency in the native language group increased the possibility of cooperative behaviors.

From the outcomes of each choice, cooperating was the main strategy that maximized the respondent’s personal gains in this game [39]. Lower rates of cooperation in the English language group showed that decision-making bias was increased by foreign language processing. Volk et al. [11] suggested that foreign language processing could induce heuristic bias in decision making, which is consistent with the present results. However, the foreign language effects are not always negative. Keysar, Hayakawa and An [9] argued that if decisions require less emotional reactions, foreign language(s) could depend more on systematic processes, thereby reducing bias. Therefore, the consequences of foreign language effects are influenced by the requirements of each task.

Similarly, in the Volunteer’s Dilemma Games, when the cost of cooperating (volunteering) involved higher levels of uncertainty (in Game 4), the participants were more likely to give cooperative responses if the contexts were described in English (the foreign language). The explanations for this are similar to those for the PD (Prisoner’s Dilemma) games. According to the Cultural Accommodation theory, the Chinese language primes the values of the Confucian Asian cultures. As Confucian Asians are typically less like to cooperate when a situation involves uncertainty, the rate of volunteering was relatively lower in this study when the contexts were presented in the Chinese language. Similarly, the higher acceptance of uncertainty triggered by the English language increased the possibilities of volunteering in this case [19]. The positive correlation between cultural identity with the responding foreign language and likelihood of cooperating could also be explained by the Cultural Accommodation Hypothesis, as the agreeableness and connection to the culture corresponding to the foreign language enlarged the performance of language-specific responses [55].

Second, regarding the increase of psychological distance in processing in a foreign language [37], when the game was presented in English, the players might have perceived lower degrees of loss in volunteering. On the opposite side, those who went through this game in their native language might have felt relatively higher levels of loss for being the volunteers [56]. Therefore, the loss-aversion tendency in the native language group increased the possibility of cooperating.

However, unlike the result of Game 4, the differences between players’ choices for each language treatment in Game 3 were not significant statistically. Inconsistent patterns in these two conditions of Volunteer’s Dilemmas may have been due to the probability judgements. Probability judgement is the essential component in making decisions about events with uncertainty. Decision makers mostly start from generating their understandings about the whole event, and then estimating the possibilities of alternative pathways before coming up with their strategies [57]. In these two cases, compared with the plane crash in Game 4, power failures might happen more frequently and a lot of the participants can experience these events directly. According to Volk et al. [11], situations which are easy to remember but occur rarely can result in greater levels of heuristic bias caused by foreign language effects. This could explain the larger differences between participants’ choices from each language treatment in Game 4.

Conclusively, assumptions about the foreign language effects on human decision making are still immature and the underlying mechanisms are unclear [58]. Therefore, to further evaluate the rationality of the Cultural Accommodation theory, or the underlying psychological mechanism governing how foreign languages impact on our daily decisions, experiments are being designed in our laboratory to test the foreign language effects of pairs of languages from the same cultural clusters (e.g., Chinese and Japanese; English and Dutch) or different cultural clusters (e.g., Chinese and English). Certainly, the nature of playing games involves the player’s cooperation and the goals. There are several factors affecting the processes: for example, the group size of the game, risk of the outcomes and individual involvement, etc. [59, 60]. These factors are all interacting with (or moderated by) the language effects (native and foreign) during decision making, and hence this is another important issue in dire need of investigation in future research. It is hoped that a more comprehensive picture of this topic can be revealed through solid, scientific evidence.

Supporting information