Abstract
Business and demographic trends are conflating to bring language issues at work to the forefront. Although language has an inherent capacity for creating interpersonal bonds, it can also serve as a means of exclusion. The construct of linguistic ostracism encapsulates this phenomenon. Drawing on ethnolinguistic identity theory, we identify how linguistic ostracism influences two interpersonal work behaviors: interpersonal citizenship and interpersonal deviance. We conduct a set of studies that uses multisource data, data across time, and data from three countries. Our results reveal that linguistic ostracism was associated with the enactment of lower interpersonal citizenship behaviors and higher interpersonal deviance behaviors. We find that disidentification served as a mechanism to explain why linguistic ostracism resulted in interpersonal citizenship behaviors and interpersonal deviance behaviors. Furthermore, linguistically ostracized employees with low (vs. high) social self-efficacy engage in fewer interpersonal citizenship behaviors and greater interpersonal deviance behaviors. We discuss theoretical implications associated with the phenomenon of linguistic ostracism and the implications for managers working in linguistically diverse organizations.
Keywords
Imagine that you enter a meeting room where some of your colleagues are already seated. Your colleagues greet you and proceed to continue their conversation in Danish—a language that you do not understand. It bothers you that they don’t switch to French—the language in which you usually converse at your workplace. You are not sure whether your colleagues are talking about you or about some work issue that you should know about. In that instance you feel a lack of connection with your work colleagues. It turns out that they were only talking about having to dig themselves out of their driveways after a recent snowstorm, but you didn’t realize that was what they were talking about.
One key demographic trend is that there is an increase in the linguistic diversity of the population across many countries. For instance, in the United States and Canada, currently over 20% of citizens speak more than one language (e.g., Spanish, French), and these percentages are only expected to rise (Finaccord, 2014; Statistics Canada, 2017). In the European Union, more than 54% of the population speaks more than one language (Special Eurobarometer 386, 2012). In China, an estimated 416 million citizens received foreign language instruction, 390 million of whom learned English (Wei & Su, 2012). With this increase in the linguistic diversity of the population globally comes a similar increase in the number of languages spoken in the workplace (Piekkari, Welch, & Welch, 2014), and for this reason, many organizations have crafted workplace language strategies (Harzing & Feely, 2008). For instance, Microsoft has developed a strategy to manage 80 languages at work (Spolsky, 2009). Similarly, IBM has identified 8 languages, in addition to English, that complement its global strategic objectives, and it hires and trains staff based on these linguistic competencies (Neeley & Kaplan, 2014). These trends, however, also signal the possibility that employees could experience exclusion based on language.
In the anecdote above, the focal employee’s perception of feeling excluded is real. But that employee’s work colleagues were not intentionally excluding this person—they were unlikely to recognize that the focal employee felt excluded in the first place. Unlike other forms of ostracism where a single employee can ignore or avoid another, exclusion based on language necessarily involves at least two or more people forming an emergent group from which the focal employee perceives exclusion. In other words, there are two unique attributes of the type of ostracism discussed in the anecdote: it is generally nonpurposeful, and it occurs in the context of a workgroup (Ferris, Chen, & Lim, 2017; Robinson, O’Reilly, & Wang, 2013). More formally, such linguistic ostracism reflects instances where focal employees perceive that others at work have rejected and/or excluded them by using a language that they do not comprehend (Dotan-Eliaz, Sommer, & Rubin, 2009; Kulkarni, 2015; Kulkarni & Sommer, 2015).
The potential for language to inadvertently exclude employees has begun to garner increased research attention. Early research on language in the workplace primarily focused on macrolevel factors, such as language policies and mandates. These include specific events that trigger linguistic tensions, such as international firm mergers or the implementation of English-only language policies (e.g., Neeley, 2013). Other work in this domain has examined how organizations conceptualize language as a strategic competency when operating globally (e.g., Harzing & Feely, 2008; Harzing & Pudelko, 2013). Recently, however, scholars have begun to explore more microlevel language processes that prevail in daily organizational life (e.g., Lecomte, Tenzer, & Zhang, 2018). For instance, members of multinational teams switched to their mother tongue to converse with colleagues from their country—rather than speaking in the corporate-mandated language—because it streamlined communication and because it was easier to express emotions (Tenzer, Pudelko, & Harzing, 2014). Such “code-switching” (i.e., changing from one language to another; Harzing & Feely, 2008; Neeley, 2013) could inadvertently place informational boundaries between those who understand what is being said and those who do not (Tenzer et al., 2014). In an increasingly multilingual workforce (Piekkari et al., 2014), language could thus unintentionally be a conduit for exclusion. As such, understanding the impact of linguistic ostracism is critical.
To gain a better understanding of the potential effects of linguistic ostracism, we develop an integrated model based on ethnolinguistic identity theory—a theory that highlights the centrality of language use in triggering social categorization processes (Giles & Johnson, 1981, 1987). Specifically, we propose that in a work environment with increasing linguistic diversity, contact with employees who speak a language that the focal employee does not understand is likely to activate a cognitive process by which ingroups and outgroups could be formed (Allport, 1954; Giles & Johnson, 1981, 1987; Neeley, Hinds, & Cramton, 2012). In brief, employees who are linguistically ostracized will perceive disidentification from their workgroup because they view themselves to be members of a linguistic outgroup (Kulkarni, 2015; Voss, Albert, & Ferring, 2014). This is because language acts as a primary method of disseminating information at work, and being unable to understand what is being communicated places a significant strain on focal employees’ interpersonal relationships (Kulkarni, 2015; Lauring, 2008; Zhang & Peltokorpi, 2016).
In turn, this disidentification will influence the enactment of interpersonal actions such that linguistically ostracized employees will perform fewer interpersonal citizenship behaviors (Sluss & Ashforth, 2007) and greater interpersonal deviance behaviors (Vadera & Pratt, 2013). Interpersonal citizenship behaviors are those actions that benefit workgroup members and contribute indirectly to effective organizational functioning (e.g., passing along important information, helping others who have heavy workloads; Lee & Allen, 2002). Interpersonal deviance behaviors, on the other hand, encompass actions that are harmful to other workgroup members and, thus, detrimental to organizational functioning (e.g., acting rudely, making hurtful remarks; Bennett & Robinson, 2000). This part of our model examines why linguistic ostracism influences the enactment of interpersonal work behaviors.
Clearly, however, employees will differ in the extent that language-based social identification processes affect them. In this regard, we propose that social self-efficacy (SSE)—a person’s ability to successfully navigate interpersonal interactions—will be an important moderator of the effects of linguistic ostracism on interpersonal work behaviors (Anderson & Betz, 2001). After experiencing linguistic ostracism, employees who have had limited prior success developing and maintaining relationships with others—those with low SSE—will lack confidence in their capacity to reestablish relational bonds (Dotan-Eliaz et al., 2009). We posit that low SSE employees will disidentify from their coworkers and attempt to assert their sense of self through the enactment of fewer interpersonal citizenship behaviors and greater interpersonal deviance behaviors (Kreiner & Ashforth, 2004; Vadera & Pratt, 2013). This part of our model examines when linguistic ostracism influences the enactment of interpersonal work behaviors.
We present our theoretical model in Figure 1. In Study 1, we test the direct effect of linguistic ostracism on interpersonal citizenship behaviors and on interpersonal deviance behaviors. We also test the moderating role that SSE plays in these two relationships. In Study 2, we test the integrated model where the indirect effect of linguistic ostracism on interpersonal citizenship behaviors and on interpersonal deviance behaviors (via disidentification) is moderated by SSE. Collectively, through these studies we clarify the conceptualization, operationalization, and effects of linguistic ostracism in the workplace.
Theoretical Model
Conceptualizing Linguistic Ostracism
Linguistic ostracism is embedded within the broader construct of workplace ostracism. Workplace ostracism occurs “when an individual or group omits to take actions that engage another organizational member when it is socially appropriate to do so” (Robinson et al., 2013: 206). Although workplace ostracism is not as explicit as other harmful work behaviors, such as aggression or harassment, its effects are pernicious because it involves a “loss of social engagement” (Robinson et al., 2013: 207). Such social distancing occurs because there is no interaction between perpetrators and targets of ostracism (Ferris et al., 2017). That is, ostracism is an act of omission that violates social norms (Robinson et al., 2013). For instance, perpetrators may purposefully shun or ignore their coworkers by giving them “the silent treatment” to intentionally punish or retaliate against targets (Quade, Greenbaum, & Petrenko, 2017; Robinson et al., 2013). In this regard, however, linguistic ostracism is unique because it is a generally nonpurposeful form of ostracism (Robinson et al., 2013). Linguistic ostracism functions by removing verbal social contact, specifically, through a nonmutually understood language (Neeley, Hinds, & Cramton, 2009). The difference in this medium of exclusion is relevant, and, as such, linguistic ostracism “may not be attributed to ill will” (Ferris et al., 2017: 333).
Notwithstanding linguistic ostracism’s generally nonpurposeful characteristic, the nonmutually understood language spoken in workgroups could still be perceived as an ostracizing act (Chrobot-Mason, Ruderman, Weber, & Ernst, 2009; Kulkarni, 2015). Furthermore, linguistic ostracism, by its very nature, occurs in many-to-one settings (Dotan-Eliaz et al., 2009). That is, at least two or more people are required to interact in a language unfamiliar to a focal employee in order to linguistically ostracize him or her. Workplace ostracism, on the other hand, can occur in one-to-one settings such that only one employee could ignore or avoid his or her coworker and, thus, ostracize the coworker (Ferris, Brown, Berry, & Lian, 2008).
As discussed previously, with the increase in the number of languages spoken in workplaces, there is a heightened prospect of linguistic ostracism. The generally nonpurposeful characteristic of linguistic ostracism coupled with the workgroup-based setting in which it takes place makes it a unique form of ostracism. For these reasons, investigating the effects associated with this form of ostracism is opportune and important (see Ferris et al., 2017). 1
Linguistic Ostracism: The Social Identity Perspective and Interpersonal Work Behaviors
A key tenet of the social identity perspective is that people derive their self-concept—their sense of who they are—from their connection to relevant social categories and that they are motivated to possess a positive self-concept (Tajfel, 1978; Tajfel & Turner, 1979, 1986). Ethnolinguistic identity theory, a conceptual descendent of the social identity perspective, highlights that language is a relevant social category that shapes people’s self-concept (Giles & Johnson, 1981, 1987). Because language is an essential mechanism for workgroups to plan, communicate, and execute shared objectives (Charles, 2007), it is a salient social category in the workplace (see Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). Put simply, language plays a fundamental role in forming a person’s social identity (Bordia & Bordia, 2015). According to ethnolinguistic identity theory, when a particular language is heard by focal employees, it activates their awareness of their own social category and signals their membership status (i.e., ingroup or outgroup based on the language spoken). That is, language, as a social category, activates ingroup/outgroup formation and shapes employees’ self-concept. Employees feel a stronger connection to those who are a part of their linguistic ingroup and a weaker one with those of their linguistic outgroup (Giles & Johnson, 1981).
Therefore, when coworkers speak a nonmutually understood language, focal employees could perceive themselves as members of a linguistic outgroup. Being a member of a linguistic outgroup will adversely affect focal employees’ self-concept, and they are likely to adopt actions to disidentify with their workgroup in order to enable them to reassert a more positive self-concept (Hogg & Abrams, 1988; Matschke & Sassenberg, 2010). Following this line of reasoning, which we elaborate below, we propose that two interpersonal actions that linguistically ostracized employees will enact are decreases in interpersonal citizenship behaviors and increases in interpersonal deviance behaviors.
According to ethnolinguistic identity theory (Giles & Johnson, 1981, 1987), there are two explanations for why linguistically ostracized employees are more likely to reduce their interpersonal citizenship behaviors and increase their interpersonal deviance behaviors (Kulkarni, 2015; Kulkarni & Sommer, 2015). First, employees decide on the level of effort that they are willing to contribute to their workgroup on the basis of their membership (ingroup/outgroup) status (Tyler & Blader, 2003). In instances where employees identify strongly with their coworkers, they are more likely to exhibit interpersonal citizenship behaviors because they consider their personal success and that of their coworkers as inextricably linked (e.g., Bartel, 2001). As employees continue to experience the benefits of ingroup membership over time, this reinforces their desire to remain and contribute to the collective (Kulkarni & Sommer, 2015). When this relationship is strained, however, such as in the case of linguistic ostracism, focal employees no longer experience the benefits of belonging to the ingroup. As such, they begin to disidentify with the workgroup and make fewer beneficial—interpersonal citizenship—contributions to their workgroup (Lauring, 2008; Sleebos, Ellemers, & de Gilder, 2006).
Second, the more that employees perceive socially distancing linguistic ostracism behaviors from a particular group, the lower the value they attach to identifying themselves as a member of that group (Giles & Johnson, 1981, 1987; Tenzer et al., 2014). As the value associated with membership in such a group erodes, so does the level of effort placed in attempting to maintain relational bonds (Hinds, Neeley, & Cramton, 2014; Marschan, Welch, & Welch, 1997). Focal employees will engage in fewer helping behaviors because they are affronted by their treatment (Thau, Aquino, & Poortvliet, 2007; Twenge, Baumeister, DeWall, Ciarocco, & Bartels, 2007) and feel more disconnected from the ingroup (O’Reilly, Robinson, Berdahl, & Banki, 2014). Thus, we propose:
Hypothesis 1: Linguistic ostracism is negatively related to interpersonal citizenship behaviors.
Consistent with the preceding reasoning, when employees are linguistically ostracized, they view themselves as members of a linguistic outgroup. Being a member of a linguistic outgroup will adversely affect focal employees’ self-concept because of the associated status loss (Neeley, 2013; Ridgeway & Correll, 2006; Tajfel & Turner, 1979). For this reason, focal employees are likely to adopt actions that enable them to reassert a more positive self-concept (Hogg & Abrams, 1988; Matschke & Sassenberg, 2010). In that regard, employees could differentiate themselves from the outgroup by enacting interpersonal deviance behaviors (Kreiner & Ashforth, 2004) because engaging in such behaviors represents “an individual’s way of establishing what he/she is not” (Giacalone & Rosenfeld, 1987: 369). Thus, we propose:
Hypothesis 2: Linguistic ostracism is positively related to interpersonal deviance behaviors.
Linguistic Ostracism–Interpersonal Work Behaviors Relationships: The Moderating Role of SSE
Being ostracized in the workplace is generally an aversive experience for employees and frays interpersonal relational bonds (Robinson et al., 2013). Research, however, suggests that employees’ personal belief that they can restore relational bonds may significantly influence their reaction to linguistic ostracism (Dotan-Eliaz et al., 2009; Robinson et al., 2013). When employees perceive linguistic ostracism, one particular trait that may influence their beliefs of reinclusion to the ingroup is SSE. SSE is a person’s “self-rated ability to deal effectively with others” (Sherer, Maddux, Mercadante, Prentice-Dunn, Jacobs, & Rogers, 1982: 670) and refers to a personal conviction that previously developed friendships are the result of one’s interpersonal skill (Watson & Nesdale, 2012).
Employees low in SSE lack the confidence in their social abilities to navigate tenuous interpersonal situations because of their limited prior success in developing and reestablishing relationships with others (Anderson & Betz, 2001; Smith & Betz, 2002). When low SSE employees perceive linguistic ostracism, because of their pessimistic belief in their ability to form social bonds, they will perceive a greater sense of disidentification with their workgroup and act in a socially unproductive way that is dysfunctional to their work colleagues (Giles & Johnson, 1981; Vadera & Pratt, 2013). On the other hand, if high SSE employees (i.e., those who have confidence in their capacity to handle delicate interpersonal situations) perceive linguistic ostracism, they are less likely to disidentify with their workgroup and perform actions that are dysfunctional to their workgroup (Syna Desivilya & Eizen, 2005). Therefore, we reason that linguistically ostracized employees with low levels of SSE will enact fewer interpersonal citizenship behaviors and more interpersonal deviance behaviors relative to their high SSE counterparts.
Hypothesis 3a: SSE moderates the relationship between linguistic ostracism and interpersonal citizenship behaviors such that the relationship is stronger when SSE is low than when SSE is high.
Hypothesis 3b: SSE moderates the relationship between linguistic ostracism and interpersonal deviance behaviors such that the relationship is stronger when SSE is low than when SSE is high.
Study 1
Procedure
Undergraduate students at a Singaporean university recruited participants for research participation credit. We informed students that the participants had to be work colleagues who had to interact with one another for their work. Students provided the contact information, specifically, the work e-mail addresses, and organizational affiliation of focal employees and their work colleagues. This method for collecting matched data has been used in prior research and yielded data of comparable quality to those obtained through other sampling methods (e.g., Bonner, Greenbaum, & Quade, 2017; Grant & Mayer, 2009; Greguras & Diefendorff, 2010).
We sent out two surveys: a focal employee survey and a work colleague survey. Focal employees completed items on linguistic ostracism, SSE, and the control variables (discussed below). Work colleagues completed items related to the focal employee’s interpersonal citizenship behaviors and interpersonal deviance behaviors (i.e., peer-reports).
We recognize that there are two possible limitations of utilizing peer-reports of interpersonal work behaviors. First, it is possible that peers, other than those who were included in the sample, could rate focal employees’ interpersonal work behaviors differently (Brannick, Chan, Conway, Lance, & Spector, 2010). In this regard, however, work colleagues reported the extent to which focal employees engaged in interpersonal citizenship behaviors and interpersonal deviance behaviors in general and not only directed toward themselves (e.g., Brebels, De Cremer, & van Dijke, 2014; Holtz & Harold, 2013). A second limitation is that it is conceivable that peers are limited in their observational capacity and may not possess adequate information on focal employees’ interpersonal work behaviors (Fox, Spector, Goh, & Bruursema, 2007). In this regard, however, research suggests that peers are considered to be reliable observers of interpersonal citizenship behaviors and interpersonal deviance behaviors because focal employees publicly exhibit such behaviors; as such, peer-reports of interpersonal citizenship behaviors and interpersonal deviance behaviors also converge with self-reports (Berry, Carpenter, & Barratt, 2012; Carpenter, Berry, & Houston, 2014). In addition, using peer-reports also mitigates concerns related to common method bias (Podsakoff, MacKenzie, & Podsakoff, 2012).
Participants
We sent out surveys to 265 employees who met the prescreening requirements of being a full-time employee. Of these, 244 employees (142 female) completed the focal employee survey (92% response rate). Employees had an average age of 38.57 years (SD = 13.24), worked an average of 43.48 (SD = 13.42) hours per week, and reported working for their organization for 6.63 years (SD = 8.16), on average. The majority of participants worked in groups of more than 2 to 10 employees (66%) and in organizations with 100 to 500 employees (57%).
A total of 241 participants (125 female) completed the work colleague survey. Work colleagues had an average age of 35.79 years (SD = 11.41), worked an average of 41.47 (SD = 13.01) hours per week, and reported working for their organization for 5.83 years (SD = 7.67), on average. Because of missing data, the final matched sample of employees and their work colleagues was 222.
Measures
Linguistic ostracism (employee-report)
By drawing on content domain experts and utilizing data from three samples of American workers, we developed a five-item measure of linguistic ostracism (please see the online supplemental material for additional details regarding this scale validation procedure). Employees rated the linguistic ostracism they perceived as a member of their workgroup on a 5-point scale (1 = never to 5 = all of the time). The coefficient alpha for linguistic ostracism was .95.
Interpersonal citizenship behaviors (peer-report)
We used Lee and Allen’s (2002; α = .87) eight-item measure to assess interpersonal citizenship behaviors. Coworkers indicated the extent to which the focal employee engaged in interpersonal citizenship behaviors (e.g., “my work colleague helps others who have been absent”) on a 5-point scale (1 = strongly disagree to 5 = strongly agree).
Interpersonal deviance behaviors (peer-report)
We used Bennett and Robinson’s (2000; α = .85) seven-item measure to assess interpersonal deviance behaviors. Coworkers indicated the extent to which the focal employee engaged in interpersonal deviance behaviors (e.g., “my work colleague cursed at someone at work”) using a 7-point scale (1 = never to 7 = daily).
SSE (employee-report)
We assessed SSE using the six-item SSE scale (Sherer et al., 1982; α = .75). For each item, employees indicated the extent to which they agreed with a number of statements pertaining to their social lives (e.g., “I have acquired my new friends through my personal abilities at making friends”) on a 4-point scale (1 = strongly disagree to 4 = strongly agree).
Control variables
We controlled for employees’ tenure, their language ability (i.e., the number of languages spoken by the employee), demographic dissimilarity with the workgroup, and workplace ostracism. We chose these control variables on the basis of recommendations for control variable usage (Bernerth & Aguinis, 2016). Specifically, employees’ tenure can influence their interpersonal work behaviors because employees with shorter tenure could be motivated to work harder to establish new relationships (Ng & Feldman, 2011). On the basis of ethnolinguistic identity theory, we also controlled for employees’ demographic dissimilarity and their language ability because both variables may influence ingroup and outgroup formation within the workgroup (Giles & Johnson, 1981, 1987). As regards demographic dissimilarity, perceived demographic differences are associated with reductions in extrarole behaviors (e.g., Chattopadhyay, 1999) and increases in deviance behaviors (Liao, Joshi, & Chuang, 2004). Employees reported their perceived demographic dissimilarity with members of their workgroup on the basis of six diversity dimensions: age, education, lifestyle, ethnic background, religion, and language (we added the dimension of language to the original measure; Pelled, Ledford, & Mohrman, 1999; α = .76).
Furthermore, we included employees’ language ability as a control because multilingualism has been linked to greater cultural and language acceptance, which may enable employees to connect more easily with a diverse set of work colleagues (Gunesch, 2003). We assessed employees’ language ability by asking them about the number of languages that they could converse in fluently. Finally, we included workplace ostracism as an additional control variable, which, as a broader form of ostracism, influences work behaviors (e.g., Ferris et al., 2008). We measured perceptions of workplace ostracism in the workgroup using the 10-item Workplace Ostracism Scale (Ferris et al., 2008; α = .92).
Results
Table 1 presents the descriptive statistics, reliability estimates, and bivariate correlations. We first performed a series of confirmatory factor analyses (CFA) using Mplus 8.0 to assess the distinctiveness of the measures completed in the employee survey. The default four-factor model was composed of linguistic ostracism, workplace ostracism, demographic dissimilarity, and SSE, with each construct loaded on separate factors.
Study 1 Variable Intercorrelations, Descriptive Statistics, and Reliability Estimates
Note: N = 222. Coefficient alpha values are shown on the diagonal in boldface.
p < .05.
p < .01.
This model provided a good fit to the data: χ2(318) = 559.54, p < .01, comparative fit index (CFI) = .95, Tucker-Lewis index (TLI) = .95, root mean square error of approximation (RMSEA) = .06, standardized root mean square residual (SRMR) = .04. We compared this model to alternative models through chi-square difference tests. A three-factor model in which demographic dissimilarity and SSE were collapsed onto one factor yielded a significantly worse fit to the data: Δχ2(3) = 270.59, p < .01, χ2(321) = 830.13, p < .01, CFI = .84, TLI = .83, RMSEA = .09, SRMR = .10. A two-factor model in which both linguistic ostracism and workplace ostracism were collapsed onto one factor and demographic dissimilarity and SSE were collapsed onto a second factor yielded a worse fit to the data: Δχ2(2) = 882.87, p < .01, χ2(323) = 1,713.00, p < .01, CFI = .57, TLI = .54, RMSEA = .14, SRMR = .14. Finally, a model in which all constructs were collapsed onto one factor also yielded a weak fit to the data: Δχ2(1) = 286.57, p < .01, χ2(324) = 1,999.57, p < .01, CFI = .50, TLI = .46, RMSEA = .15, SRMR = .14. 2 These results provided evidence of the distinctiveness of the constructs.
We proposed that linguistic ostracism is related to lower interpersonal citizenship behaviors (Hypothesis 1) and higher interpersonal deviance behaviors (Hypothesis 2). The bivariate correlations between linguistic ostracism and interpersonal citizenship behaviors (r = −.25, p < .01) and interpersonal deviance behaviors (r = .38, p < .01) provided preliminary support for Hypotheses 1 and 2 (see Table 1). We formally tested these hypotheses using hierarchical regression analysis (see Table 2). Results indicated that after we controlled for employees’ tenure, language ability, demographic dissimilarity, and workplace ostracism, linguistic ostracism was negatively related to interpersonal citizenship behaviors (b = −0.13, SE = 0.06, p < .05; Model 2, Table 2) and positively related to interpersonal deviance behaviors (b = 0.44, SE = 0.09, p < .01; Model 5, Table 2; results were similar without including any control variables, Bernerth & Aguinis, 2016). Hypotheses 1 and 2 thus received support.
Study 1 Hierarchical Regression: Interpersonal Citizenship Behaviors and Interpersonal Deviance Behaviors
Note: N = 222. Coefficients are unstandardized. Standard errors are shown in parenthesis.
p < .05.
p < .01.
In Hypotheses 3a and 3b, we proposed that SSE moderates the effect of linguistic ostracism on interpersonal citizenship behaviors and interpersonal deviance behaviors such that compared to high SSE employees, those with low SSE will engage in fewer interpersonal citizenship behaviors and greater interpersonal deviance behaviors. The interaction term of linguistic ostracism and SSE was statistically significant for interpersonal citizenship behaviors (b = 0.07, SE = 0.03, p < .05; Model 3, Table 2) and for interpersonal deviance behaviors (b = −0.11, SE = 0.05, p < .05; Model 6, Table 2).
To better clarify the pattern of these interactions, we followed the recommendations of Aiken and West (1991) and conducted simple slopes tests. For interpersonal citizenship behaviors, results indicated that there was a stronger negative relationship between linguistic ostracism and interpersonal citizenship behaviors when SSE was low (b = −0.20, SE = 0.07, p < .01) than when SSE was high (b = −0.02, SE = 0.08, n.s.; see Figure 2a). For interpersonal deviance behaviors, results indicated that there was a stronger positive relationship between linguistic ostracism and interpersonal deviance behaviors when SSE was low (b = 0.53, SE = 0.11, p < .01) than when SSE was high (b = 0.23, SE = 0.14, n.s.; see Figure 2b). Supporting Hypotheses 3a and 3b, results indicate that linguistically ostracized employees were less likely to engage in interpersonal citizenship behaviors and more likely to engage in interpersonal deviance behaviors when SSE was low compared to when SSE was high.
The Effects of Linguistic Ostracism on Interpersonal Work Behaviors as a Function of Social Self-Efficacy (Study 1)
Discussion
Consistent with the literature on social identity formation, the results from Study 1 suggest that language in the workplace is a relevant social category that could spark ingroups and outgroups (Bordia & Bordia, 2015; Harzing & Feely, 2008; Tajfel & Turner, 1986). If workgroup members do not converse in a mutually understood language, focal employees could view themselves as members of a linguistic outgroup. In turn, this will adversely affect focal employees’ self-concept, and they would seek to disidentify with their workgroup so as to reestablish a more positive self-concept. This reasoning, which draws on ethnolinguistic identity theory (Giles & Johnson, 1981, 1987), highlights that disidentification serves as an underlying mechanism that explains why linguistic ostracism influences interpersonal work behaviors. We now further develop the rationale regarding this theoretical mechanism of disidentification. 3
Linguistic Ostracism–Interpersonal Work Behaviors Relationships: The Mediating Role of Disidentification
Disidentification is considered as “an active separation from a group and thus a negative self-defining relation to a relevant group” (Matschke & Sassenberg, 2010: 892). Put simply, disidentification acts as a means of distancing oneself from an undesired social group (Elsbach & Bhattacharya, 2001; Kreiner & Ashforth, 2004). Because disidentification invokes an unfavorable categorization between the self and the group, it is strongly associated with negative feelings, such as anger, that reinforce the separation from the outgroup (Elsbach, 1999; Pratt, 2000). As such, linguistically ostracized employees will perceive a loss of connection with their workgroup and differentiate themselves by engaging in actions that run counter to outgroup norms in an effort to improve their self-concept (Elsbach & Bhattacharya, 2001; Giles & Johnson, 1981; Tajfel & Turner, 1979). Such actions could involve limiting help to workgroup members or being discourteous to them (Sluss & Ashforth, 2007; Tyler & Blader, 2003; Vadera & Pratt, 2013). As this line of reasoning suggests, linguistic ostracism constitutes a threat to focal employees’ self-concept and activates cognitive processes of disidentification from the workgroup (Giles & Johnson, 1981, 1987). That is, disidentification serves as a key explanatory mechanism for understanding why linguistic ostracism will result in lower enactment of interpersonal citizenship behaviors and higher enactment of interpersonal deviance behaviors. Thus, we propose:
Hypothesis 4a: Disidentification mediates the relationship between linguistic ostracism and interpersonal citizenship behaviors.
Hypothesis 4b: Disidentification mediates the relationship between linguistic ostracism and interpersonal deviance behaviors.
In light of the preceding hypotheses where we posit the mediating role of disidentification, coupled with our earlier discussion of the moderating role of SSE, we propose that SSE will moderate the indirect effect of linguistic ostracism on interpersonal citizenship behaviors and interpersonal deviance behaviors (via disidentification). That is, we posit that for linguistically ostracized employees, those with low SSE (i.e., employees who have lower confidence in their ability to connect with their work colleagues) compared to those with high SSE, are more likely to perceive disidentification from their workgroup and engage in fewer interpersonal citizenship behaviors and more interpersonal deviance behaviors. Thus, we propose:
Hypothesis 5a: SSE moderates the indirect relationship between linguistic ostracism and interpersonal citizenship behaviors via disidentification such that the indirect effect is stronger when SSE is low than when SSE is high.
Hypothesis 5b: SSE moderates the indirect relationship between linguistic ostracism and interpersonal deviance behaviors via disidentification such that the indirect effect is stronger when SSE is low than when SSE is high.
Study 2
In Study 2, we endeavored to extend the findings from Study 1 by assessing the mediating role of disidentification and the associated conditional indirect effect of linguistic ostracism on interpersonal citizenship behaviors and interpersonal deviance behaviors via disidentification at low levels of SSE. In so doing, we also incorporated changes to our research design guided by the objective of a constructive replication (Lykken, 1968). Our first goal was to identify whether our findings replicated in a different national and linguistic context and using a different research design. Because Study 1 was set in an Eastern culture (Singapore), we conducted Study 2 in a Western culture (Canada). Second, in Study 1 we asked work colleagues to provide reports on focal employees’ interpersonal citizenship behaviors and interpersonal deviance behaviors. To complement these peer-reports, in Study 2 we asked employees to provide self-reports of their interpersonal citizenship behaviors and interpersonal deviance behaviors. We did so because employee self-reports are “not only justifiable but probably necessary” (Chan, 2009: 326) as a result of the high level of information that employees have about their own behaviors. Third, in Study 1 we used multisource data to minimize common method bias concerns (Podsakoff et al., 2012). For similar methodological considerations of minimizing common method bias, in Study 2 we temporally separated the assessment of the independent and dependent variables. Finally, given these changes to the research design, to ensure comparability across the two studies, we retained the same set of variables in both studies (with the exception of including the mediator of disidentification in Study 2).
Procedure
Master of business administration students at a Canadian university recruited full-time employees for this study (e.g., Grant & Mayer, 2009; Morgeson & Humphrey, 2006) and received a gift card for their efforts. Prospective participants received study packets, which included an invitation letter with information about the study and an online link to access the Time 1 survey. When participants accessed the link, they received an invitation to complete two surveys separated by 10 days. Participants received a $10 gift card for their involvement in the study. In the Time 1 survey, participants provided responses to measures of linguistic ostracism, SSE, disidentification, and the control variables. In the Time 2 survey, participants provided responses to measures of interpersonal citizenship behaviors and interpersonal deviance behaviors.
Participants
In total, 267 participants met the prescreening requirements of being a full-time employee and completed the Time 1 survey. We sent the second survey after 10 days and received complete responses from 206 participants (77%). Employees in our sample had an average age of 29.91 years (SD = 8.81) and an average organization tenure of 4.31 years (SD = 4.52), and 53% of the sample was female. The majority of participants worked in groups of more than 2 to 10 employees (56%); participants worked in organizations of varying size (50–100 employees, 31%; 100–500 employees, 22%).
Measures
Linguistic ostracism (Time 1)
We measured the extent to which participants perceived linguistic ostracism in their workgroup using the same five-item measure from Study 1. The coefficient alpha for this scale was .94.
Disidentification (Time 1)
We used Kreiner and Ashforth’s (2004; α = .93) six-item measure (e.g., “I am embarrassed to be a member of this workgroup”) to assess participants’ perceived disidentification with their workgroup on a 5-point scale (1 = strongly disagree to 5 = strongly agree).
SSE (Time 1)
We assessed SSE using the same six-item SSE scale from Study 1 (Sherer et al., 1982; α = .76).
Interpersonal citizenship behaviors (Time 2)
We used the same eight-item measure from Study 1 to assess interpersonal citizenship behaviors (Lee & Allen, 2002; α = .88). On the basis of self-referent items, we asked participants to report the extent of interpersonal citizenship behaviors they enacted in their workgroup.
Interpersonal deviance behaviors (Time 2)
We used the same seven-item measure from Study 1 to assess interpersonal deviance behaviors (Bennett & Robinson, 2000; α = .82). On the basis of self-referent items, we asked participants to report the extent of interpersonal deviance behaviors they enacted in their workgroup.
Control variables
Similar to Study 1 we controlled for tenure, language ability, demographic dissimilarity (Pelled et al., 1999; α = .74), and workplace ostracism (Ferris et al., 2008; α = .93).
Results
Table 3 presents the descriptive statistics, reliability estimates, and bivariate correlations. Similar to Study 1, we first performed a series of CFAs to assess the distinctiveness of the constructs. The default five-factor model was composed of linguistic ostracism, workplace ostracism, demographic dissimilarity, SSE, and disidentification, with each construct loaded on separate factors. This model provided a good fit to the data: χ2(485) = 797.24, p < .01, CFI = .94, TLI = .94, RMSEA = .06, SRMR = .05. We compared this model to alternative models through chi-square difference tests. A four-factor model in which SSE and disidentification were collapsed onto one factor yielded a significantly worse fit to the data: Δχ2(4) = 272.20, p < .01, χ2(489) = 1,069.44, p < .01, CFI = .84, TLI = .83, RMSEA = .09, SRMR = .08. A three-factor model in which workplace ostracism and linguistic ostracism were collapsed onto one factor, SSE and disidentification were collapsed onto a second factor, and demographic dissimilarity was kept independent yielded a significantly worse fit to the data: Δχ2(4) = 954.38, p < .01, χ2(493) = 2,023.82, p < .01, CFI = .58, TLI = .55, RMSEA = .12, SRMR = .12. A two-factor model in which linguistic ostracism and workplace ostracism were collapsed onto one factor and demographic dissimilarity, SSE, and disidentification were collapsed onto a second factor yielded a weak fit to the data: Δχ2(2) = 132.19, p < .01, χ2(495) = 2,156.01, p < .01, CFI = .55, TLI = .52, RMSEA = .12, SRMR = .12. Finally, a model in which all five constructs were collapsed onto a single factor also yielded a weak fit to the data: Δχ2(1) = 275.78, p < .01, χ2(496) = 2,431.79, p < .01, CFI = .47, TLI = .44, RMSEA = .14, SRMR = .13. 4
Study 2 Variable Intercorrelations, Descriptive Statistics, and Reliability Estimates
Note: N = 206. Coefficient alpha values are shown on the diagonal in boldface.
p < .05.
p < .01.
We proposed that disidentification mediates the relationship between linguistic ostracism and interpersonal citizenship behaviors (Hypothesis 4a) and interpersonal deviance behaviors (Hypothesis 4b). We assessed mediation based on 95% bias-corrected confidence intervals (CIs) constructed through 5,000 bootstrapped samples (PROCESS, Model 4; Hayes, 2013). The indirect effect of linguistic ostracism (via disidentification) on interpersonal citizenship behaviors (indirect effect = −.09, SE = .04, 95% CI = [−.193, −.036]) and on interpersonal deviance behaviors (indirect effect = .14, SE = .06, 95% CI = [.062, .288]) was statistically significant. These results provide support for Hypotheses 4a and 4b.
Using Hayes’s (2013) PROCESS macro (Model 7), we assessed our complete model, which is a first-stage moderation model (Edwards & Lambert, 2007; for a recent example, see Bonner et al., 2017; see Table 4). In Hypotheses 5a and 5b, we proposed that the indirect effect of linguistic ostracism on interpersonal citizenship behaviors and interpersonal deviance behaviors, respectively, via disidentification, is stronger when SSE is low than when SSE is high. First, the interaction term of linguistic ostracism and SSE was statistically significant (b = −0.45, SE = 0.14, p < .01; see Table 4). Additionally, a simple slopes test revealed that the linguistic ostracism to disidentification relationship was stronger when SSE was low (b = 0.49, SE = 0.09, p < .01) than when SSE was high (b = 0.09, SE = 0.11, n.s.; see Figure 3).
Study 2 Moderated-Mediation Regression: Interpersonal Citizenship Behaviors and Interpersonal Deviance Behaviors
Note: N = 206. Bootstrap sample size = 5,000. Coefficients are unstandardized. Standard errors are shown in parentheses.
p < .01.
The Effects of Linguistic Ostracism on Disidentification as a Function of Social Self-Efficacy (Study 2)
Second, for interpersonal citizenship behaviors, the conditional indirect effect of linguistic ostracism on interpersonal citizenship behaviors via disidentification was statistically significant when SSE was low (−1 SD, indirect effect = −.13, SE = .04, 95% CI = [−.233, −.054]) compared with when SSE was high (+1 SD, indirect effect = −.01, SE = .05, 95% CI = [−.142, .052]). For interpersonal deviance behaviors, the conditional indirect effect of linguistic ostracism on interpersonal deviance behaviors via disidentification was statistically significant when SSE was low (−1 SD, indirect effect = .20, SE = .07, 95% CI = [.085, .372]) compared with when SSE was high (+1 SD, indirect effect = .02, SE = .07, 95% CI = [−.090, .193]). Hypotheses 5a and 5b thus received support (see Table 5).
Study 2 Estimates and Bias-Corrected Bootstrapped 95% Confidence Intervals for the Conditional Indirect Effect of Linguistic Ostracism on Interpersonal Citizenship Behaviors and Interpersonal Deviance Behaviors (via Disidentification) at ±1 SD of Social Self-Efficacy
Note: N = 206. Bootstrap sample size = 5,000. Bootstrapped estimates for standard errors are presented in parentheses.
As an additional check, we consulted the index of moderated mediation, which assesses whether there is a statistically significant difference between two values of the conditional indirect effect at different levels of the moderator (Hayes, 2015). The index of moderated mediation provided supportive evidence for both interpersonal citizenship behaviors (Δ indirect effect = .10, SE = .05, 95% CI = [.012, .216]) and interpersonal deviance behaviors (Δ indirect effect = −.16, SE = .08, 95% CI = [−.360, −.027]). Collectively, results of Study 2 provide a constructive replication (Lykken, 1968) of Study 1’s findings and a more comprehensive test of the theoretical model.
General Discussion
Managing workplace diversity is essential to ensure organizational effectiveness (Welch & Welch, 2018). Ample research has identified the effects of surface-level characteristics such as gender and ethnicity and their associated effects on work outcomes (e.g., Bhave, Kramer, & Glomb, 2010; Joshi & Roh, 2009). Increasingly, scholars have advocated the importance of considering other attributes that prevail in contemporary workplaces that could also affect work outcomes (see Roberson, Ryan, & Ragins, 2017). In light of this, and concomitantly spurred by demographic and business trends, organizational scholars have considered the use of multiple languages in workplaces as a diversity characteristic (Piekkari et al., 2014; Tenzer, Terjesen, & Harzing, 2017). Although language has an inherent capacity to connect people, it could also serve as a way to exclude others. We draw on ethnolinguistic identity theory (Giles & Johnson, 1981, 1987), and a growing body of research on microlevel language processes (Lecomte et al., 2018), to clarify the construct of linguistic ostracism. Through a set of two studies, we test an integrative model to identify linguistic ostracism’s connection with relevant interpersonal outcomes for employees and organizations.
In line with the tenets of ethnolinguistic identity theory, we observe that linguistic ostracism triggers ingroup and outgroup formation and results in two interpersonal behavioral consequences: lower enactment of interpersonal citizenship behaviors and higher enactment of interpersonal deviance behaviors. An underlying reason for these interpersonal behaviors is that focal employees view work colleagues who speak a language they do not understand in their presence as outgroup members and disidentify with such colleagues. As a result, they are less likely to initiate interpersonal citizenship behaviors and more likely to initiate interpersonal deviance behaviors. We observe that an individual differences variable—SSE—shapes this underlying pattern such that for employees with pessimistic beliefs about their ability to succeed in social situations (i.e., low SSE employees), linguistic ostracism creates a stronger sense of disidentification from the workgroup and subsequently influences their enactment of interpersonal work behaviors.
Theoretical Implications
Our results suggest that workplace language is integral to how employees view themselves in relation to their workgroup (e.g., Hinds et al., 2014), and when employees perceive linguistic ostracism, it can result in feelings of disidentification. In so doing, we provide an important rationale for why employees perceive acts of linguistic exclusion by their workgroup members—even if they are generally nonpurposeful in nature—to be aversive and how this influences their subsequent enactment of interpersonal work behaviors (Kulkarni, 2015; Robinson et al., 2013). By illustrating that disidentification could also occur with one’s workgroup, these findings also contribute to the broader literature on disidentification, which has explored how disidentification occurs with one’s organization (e.g., Elsbach & Bhattacharya, 2001; Vadera & Pratt, 2013).
Our research also contributes to the broader literature on language use in workplace settings (Brannen, Piekkari, & Tietze, 2014; Piekkari et al., 2014; Tenzer et al., 2017). For instance, international business scholars have investigated the impact of corporate language mandates (Brannen et al., 2014; Fredriksson, Barner-Rasmussen, & Piekkari, 2006; Neeley & Dumas, 2016), where organizations institute a common language, such as English, regardless of location or host nation (Neeley, 2013) and in specific contexts (e.g., choosing an official corporate language in an international merger; Vaara, Tienari, Piekkari, & Säntti, 2005). Other work has considered the socialization challenges encountered by members in multilingual international management teams (Henderson, 2005), and the role of expatriates’ language proficiency and their work and nonwork adjustment (Zhang & Peltokorpi, 2016). Our study contributes to this body of work by demonstrating that language could serve as a “cultural replicator” (Pagel, 2009: 405). That is, language transmits culture and shapes a person’s worldview and, thus, plays a fundamental role in identity creation (e.g., Bordia & Bordia, 2015) and group boundary formation (Giles & Johnson, 1981, 1987).
Emerging research on microlevel language processes within organizations (Janssens & Steyaert, 2014) offers another connection to our findings. This body of work suggests a more fluid conceptualization of language from discrete languages to a consideration of how people speak in everyday life on the basis of their linguistic resources (Janssens & Steyaert, 2014; Lecomte et al., 2018). For example, bi/multilingual speakers make quick, natural transitions between discrete languages in everyday life to communicate with each other (Janssens & Steyaert, 2014) and may combine multiple languages and slang to create a hybridized language (Gaibrois, 2018). In so doing, they draw on their entire set of linguistic resources, crisscrossing the boundaries of discrete languages—a phenomenon linguists refer to as translanguaging (Otheguy, García, & Reid, 2015). Such microlevel language processes are likely to operate in increasingly multilingual workplaces. Our findings suggest that these processes could also activate linguistic ingroups and outgroups and inadvertently exclude focal employees.
Finally, we observed a similar pattern of results in Eastern (Singapore) and Western (Canada) cultures. Furthermore, invariance analyses (see the online supplemental material) revealed that participants across three national contexts (Singapore: Study 1; Canada: Study 2; and the United States: scale development study) interpreted the phenomenon of linguistic ostracism in similar terms. These results indicate that linguistic ostracism occurs across national boundaries and draw attention to the effects associated with the increasing linguistic diversity across workforces globally.
Practical Implications
Our results suggest that linguistic ostracism could be costly for organizations because linguistic ostracism influences two important domains of job performance: interpersonal citizenship and interpersonal deviance (Rotundo & Sackett, 2002). Thus, it is important to make organizational members conscious of the potential effects of language use at work. In many cases, organizations may simply need to emphasize the importance of common courtesy and awareness regarding language use in the workplace. Other situations may call for focusing on metacommunication (Tenzer et al., 2014); that is, senior leaders could dispense information about workplace communication practices in an effort to develop a shared understanding regarding the possibility of linguistic ostracism.
The moderating role of SSE highlights that employees’ beliefs in their social capabilities are influential in limiting the adverse effects of linguistic ostracism. As such, we suggest that organizations offer training programs to equip employees with relational and conflict resolution skills so as to bolster their beliefs in their social abilities and enhance their confidence to resolve interpersonally tense situations constructively. In this regard, assessing levels of SSE of workgroup members could serve as a useful diagnostic when there are reports of linguistic ostracism.
Finally, assessing the prevalence of linguistic ostracism in the workplace could help in the design of targeted leadership, language acquisition, and cross-cultural training initiatives (Tenzer & Pudelko, 2015). For instance, expatriate employees or new immigrants in the workforce may be more susceptible to linguistic ostracism (see Zhang & Peltokorpi, 2016). Thus, assessments of employees’ perceptions of linguistic ostracism will provide a helpful metric to evaluate the efficacy of diversity and inclusion programs.
Limitations and Future Directions
Our studies were set in three different national contexts (Singapore, Canada, and the United States) and used different designs: multisource data and time-separated data. Notably, the pattern of our findings is consistent across studies. Despite these methodological features, on the basis of the design of our studies, we are circumspect to make causal inferences. For instance, linguistic ostracism occurs in workgroups where employees speak more than one language. As such, it could be argued that it is multilingualism itself that is associated with the outcomes of interpersonal citizenship behaviors and interpersonal deviance behaviors. 5 In this regard, however, we included two relevant control variables in the analyses: number of languages spoken (multilingualism) and perceptions of demographic dissimilarity. As we noted earlier, multilingualism is associated with greater openness toward others and may favorably influence employees’ connection with their work colleagues (Gunesch, 2003). Conversely, perceptions of demographic dissimilarity (which also included an item on dissimilarity based on language) are associated with lower citizenship behaviors (Chattopadhyay, 1999) and higher deviance behaviors (Liao et al., 2004). Accounting for the effects associated with these two control variables provides greater support for the directionality of the relationships. Furthermore, in Study 2, we measured focal constructs at different time points and, as such, adhere to one essential element to establish causality: temporal precedence (Cook & Campbell, 1979). Nevertheless, employing other designs, such as randomized field experiments, could offer additional insight regarding causal processes (Antonakis, Bendahan, Jacquart, & Lalive, 2014).
Focal employees reported the linguistic ostracism they experienced in their workgroup and not from specific coworkers. Additionally, a limitation of Study 1 is that it is possible that coworkers, other than those who we surveyed, could provide different ratings of focal employees’ work behaviors. Related to this, we did not directly assess perceptions of ingroup and outgroup membership (Giles & Johnson, 1981, 1987). The explanatory mechanism of disidentification with workgroup members, however, does provide supportive evidence of employees’ distancing from undesired social groups. Nevertheless, other designs, such as network survey designs (e.g., Wasserman & Faust, 1994), could help in identifying how the entire workgroup (rather than a single work colleague) views focal employees’ interpersonal work behaviors. Such a network design could also help to assess ties between workgroup members and the role that network centrality plays in employees’ perceptions of linguistic ostracism and their subsequent relationship to interpersonal work behaviors.
Although linguistic ostracism is a generally nonpurposeful form of ostracism, as we mentioned earlier, language at work could also be used in an explicit fashion to exclude others. Recent work has identified how employees could engage in linguistic maneuvering such as language-based communication avoidance and linguistic power plays (e.g., Lauring & Klitmøller, 2015; Tenzer & Pudelko, 2017). Such purposeful actions align with how the broader construct of workplace ostracism is conceptualized (e.g., Robinson et al., 2013). To that end, future work could integrate the role of motives to engage in ostracizing behaviors (e.g., Neuman & Baron, 2005; Williams, 2001) and whether SSE influences the accurate detection of purposeful versus nonpurposeful ostracism (e.g., Dotan-Eliaz et al. 2009; Watson & Nesdale, 2012).
Given the breadth of workplace language and communication, we intentionally narrowed our focus to aspects of formal language that could be perceived as a form of exclusion. Yet there could be other facets of workplace communication (e.g., nonverbal communication; Bonaccio, O’Reilly, O’Sullivan, & Chiocchio, 2016) in addition to the use of slang, company speak, or jargon (Aichhorn & Puck, 2017; Wells, 2018), which could also be perceived to be exclusionary. There is also scope to expand the outcome domain. For instance, a growing body of research has found that deviance can also be constructive (Vadera, Pratt, & Mishra, 2013). In that vein, in addition to disidentification, other mechanisms such as moral disengagement (Lee, Kim, Bhave, & Duffy, 2016) or interpersonal justice (Smart Richman & Leary, 2009) could have a role to play to explain the linguistic ostracism–constructive deviance relationship.
Conclusion
Given the rise in multilingualism in organizations, it is imperative to broaden our understanding of how language influences intragroup relationships at work. Drawing on ethnolinguistic identity theory, we clarify how language could be perceived as a form of exclusion in the workplace. Our results illustrate that if employees perceive linguistic ostracism, they disidentify with their workgroup and enact fewer interpersonal citizenship behaviors and greater interpersonal deviance behaviors, and that this phenomenon is particularly salient for those who are low in SSE. Considering the potential adverse consequences associated with linguistic ostracism, a generally nonpurposeful form of ostracism, our study identifies this phenomenon as an important area for future theoretical and empirical development.
Supplemental Material
JOM_833445_online_supplement – Supplemental material for Mind Your Language: The Effects of Linguistic Ostracism on Interpersonal Work Behaviors
Supplemental material, JOM_833445_online_supplement for Mind Your Language: The Effects of Linguistic Ostracism on Interpersonal Work Behaviors by John Fiset and Devasheesh P. Bhave in Journal of Management
Footnotes
Acknowledgements
We are grateful to the editor, Christopher O. L. H. Porter, and the reviewers for their insights and guidance. We also thank Gary Greguras and Abhijeet Vadera for their helpful comments and feedback on previous iterations of this work. This research was supported by funding from the Government of Canada, Social Sciences and Humanities Research Council of Canada (Grant 430-2018-00321) and from Singapore Management University’s Internal Research Grant (C207/MSS17B012).
Supplemental material for this article is available with the manuscript on the JOM website.
Notes
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.
