Do Chinese subordinates trust their German supervisors? A model of inter-cultural trust development

In this qualitative study based on 95 interviews with Chinese subordinates and their German supervisors, we inductively develop a model which advances theoretical understanding by showing how inter-cultural trust development in hierarchical relationships is the result of six distinct elements: the subordinate trustor’s cultural profile (cosmopolitans, hybrids, culturally bounds), the psychological mechanisms operating within the trustor (role expectations and cultural accommodation), and contextual moderators (e.g., country context, time spent in foreign culture, and third-party influencers), which together influence the trust forms (e.g., presumptive trust, relational trust) and trust dynamics (e.g., trust breakdown and repair) within relationship phases over time (initial contact, trust continuation, trust disillusionment, separation, and acculturation). Our findings challenge the assumption that cultural differences result in low levels of initial trust and highlight the strong role the subordinate’s cultural profile can have on the dynamics and trajectory of trust in hierarchical relationships. Our model highlights that inter-cultural trust development operates as a variform universal, following the combined universalistic-particularistic paradigm in cross-cultural management, with both culturally generalizable etic dynamics, as well as culturally specific etic manifestations.


INTRODUCTION
Trust is recognized as central to effective working relationships (De Jong, Dirks, & Gillespie, 2016;Dirks & Ferrin, 2002;Fulmer & Gelfand, 2012;Luo, 2002). In an increasingly globalized, multicultural and multinational work environment, a key challenge is developing and maintaining trust across national cultural boundaries (Pudelko & Liu, 2020), for example between supervisors and subordinates whose cultural values, beliefs and norms can substantially differ.
Despite the need for practical guidance on developing trust in inter-cultural relationships, models of interpersonal trust development have almost exclusively focused on mono-cultural contexts, particularly in Western nations. These models mostly adhere to an etic perspective, assuming that the development and dynamics of trust are universal and culturally generalizable. Challenging this perspective is much-needed research, which takes an emic view based on evidence that trust has culturally specific elements and is interpreted and developed differently across cultures (Branzei, Vertinsky, & Camp, 2007;Chua, Morris & Ingram, 2009;Pudelko, Tenzer, & Harzing, 2015;Wasti, Tan, & Erdil, 2011). Emic research has largely taken a comparative cross-cultural perspective, examining how trust varies across national cultures (Ferrin & Gillespie, 2010). This literature has theorized how a variety of cultural dimensions influence the bases and forms of trust (e.g., Chen, Chen, & Meindl, 1998;Doney, Cannon, & Mullen, 1998), with empirical work focusing on individualismcollectivism. This work has shown, for example, that collectivists are more heavily influenced by affective and situational trust cues, whereas individualists are more influenced by dispositional and cognitive trust cues (Branzei et al., 2007;Chen, Eberly, Chiang, Farh, & Cheng, 2014;Wasti et al., 2011). Recent studies further suggest that collectivists may respond in a more moderate and adaptable way to a relationship breach compared to individualists (Eckerd, Boyer, Qi, Ecker & Hill, 2016;Vogel, Mitchell, Tepper, Restubog, Hu, Hua, & Huang, 2015).
In contrast to this comparative cross-cultural trust literature, a smaller literature has examined trust within inter-cultural relationships, which is between members of differing cultures (Fulmer, Ferrin, Dennison & Gillespie, forthcoming). This emerging research suggests that trust development is significantly different in inter-cultural compared to mono-cultural relationships (Ferrin & Gillespie, 2010). Due to the lack of shared cultural values and norms, and cultural differences in the trust cues that are expected and relied upon, the encodingand-decoding of trust signals is more complex in inter-cultural relationships (Branzei et al., 2007;Johnson & Cullen, 2002), and often leads to misunderstanding, conflict and difficulties in establishing trust (Ajmal, Helo & Kassem, 2017;Luo, 2002). Given the ubiquity of inter-cultural relationships in modern organizations, combined with evidence that trust is a fundamental prerequisite for inter-cultural collaboration (Child, 2001;Zaheer & Zaheer, 2006), it is surprising that only limited inter-cultural trust research has been conducted so far (Ferrin & Gillespie, 2010;Fulmer & Gelfand, 2012).
There are three key limitations in the emerging literature on inter-cultural trust research in working relationships. First, this research mostly takes a static perspective (Pudelko & Liu, 2020), informing our understanding of what factors either contribute to or inhibit inter-cultural trust in working relationships, rather than how they do so (for a review see Ferrin & Gillespie, 2010). As such, this research overlooks the dynamic and processual nature of trust development (Dietz, Gillespie, & Chao, 2010), which includes the possibility that processes that develop over time, such as cultural adaptation and cross-cultural learning, may play a significant role in inter-cultural trust development (Pudelko & Liu, 2020).
Second, what is missing are models of intercultural trust formation in working relationships, derived from empirical data, that capture the trust development process in specific contexts (Tenzer, Pudelko, & Harzing, 2014). This is important as we should not limit our understanding of inter-cultural trust formation to conceptual reflections only. Empirically grounded models are particularly important given trust formation is at least partially context dependent (Ferrin & Gillespie, 2010).
Third, the inter-cultural literature ignores the fact that working relationships are frequently embedded within a hierarchy of roles. While it is well established that expectations of leaders and subordinates are heavily culturally bound (Cheng, Jiang, Cheng, Riley, & Jen, 2015;Wasti & Tan, 2010;Wasti et al., 2011), we do not currently understand how hierarchy and role expectations influence trust development in inter-cultural work relationships.
In sum, what is required to facilitate and guide inter-cultural working relationships is emic research providing an in-depth understanding of the dynamic processes which influence trust development between supervisors and subordinates of different cultures (Dietz et al., 2010;Pudelko & Liu, 2020;Tenzer et al., 2014).
In this paper, we address these limitations in the literature by examining the question: How does trust develop in inter-cultural relationships between subordinates and their supervisors? We examine this question in an inter-cultural setting comprised of countries with distinctly different cultural values of direct relevance for inter-cultural trust development in hierarchical relationships: China as the nationality of subordinate trustors and Germany as the nationality of supervisor trustees. Illuminating the process of inter-cultural trust development in a Chinese-German context is particularly promising, as trust development between Chinese nationals and Western counterparts is still widely regarded as a ''black box'' (Ang & Tan, 2016: 45). More specifically and given our interest in hierarchical trust relations, China is a high-context, vertical collectivist country in contrast to Germany, which is a low-context, horizontal-individualist country (Hall & Hall, 1990;Hofstede Insights, 2020;Triandis, 1995). Given these oppositions, we treat both countries as a 'pragmatic extreme case', an ideal choice for inductive theory building. Importantly (and as we show in more detail in our study), the trust concept in both countries combines an interesting mix of etic commonalities and emic particularities, suggesting promising opportunities of inter-cultural theory building (Kriz & Keating, 2010). Furthermore, both countries constitute major economies (the second and fourth largest in the world and the largest in their respective continent) with strong economic exchanges between them, constituting an active site for inter-cultural work relationships.
Given the nascent stage of the literature on intercultural trust development, we respond to calls recommending inductive theory building based on a qualitative research design and incorporating an emic perspective to advance inter-cultural trust theory in hierarchical relationships (Dietz et al., 2010;Tenzer et al., 2014). While our empirical material speaks only to the specific Chinese-German inter-cultural context, our aim is to develop mid-range theory that is applicable to wider country contexts. Qualitative research is particularly well suited to investigating the dynamic processes underlying trust developments (the how), as well as the underlying reasons for trust development or decline (the why; see Pratt, 2009).
Our study draws on a rich dataset of 95 interviews with Chinese subordinates and their German supervisors (including 50 from dyads). From these data, we develop a comprehensive model of intercultural trust development in subordinate-supervisor relationships. Key aspects of our model only emerged after multiple rounds of data analysis, as is typical for inductive theory building (Edmondson & McManus, 2007), underscoring the relevance of the qualitative approach for the purpose of our study.
Our study makes a number of important contributions to the literature on inter-cultural trust. First, our model advances theoretical understanding by showing how inter-cultural trust development in hierarchical relationships is the combined outcome of six distinct elements: the trustor's cultural profile (cosmopolitans, hybrids, culturallybounds); the relationship phases (initial contact, trust continuation, trust disillusionment, separation and acculturation) between trustor and trustee; the ensuing trust dynamics (e.g., trust breakdown and repair); the resulting trust forms (e.g., presumptive trust and relational trust); the psychological mechanisms operating within the trustor (e.g., role expectations and cultural accommodation) and various contextual moderators (e.g., country context). To our knowledge, this is the first empirically derived model of inter-cultural trust development. Second, we indicate that some of these elements and, hence, trust development processes in inter-cultural hierarchical relationships in general, are etic and universal (context-free), whereas others are emic and particular (contextspecific) due to cultural differences (Ferrin & Gillespie, 2010;Pudelko, 2006). Third, our findings challenge the dominant assumption in the literature that cultural differences between parties result in low levels of initial trust. In contrast, we find uniformly high levels of presumptive trust early in the relationship, mainly based on subordinates' heavily socialized, culturally-bound role expectations of their supervisors. This finding challenges the assumption of seminal models that intercultural trust starts low and only develops incrementally over time (Mayer, Davis & Schoormann, 1995). Fourth, our model advances inter-cultural trust theory by highlighting that the distinct cultural profiles of subordinates strongly influence the dynamics and trajectory of trust development. We find the subordinates' cultural profiles influence two central psychological mechanisms of inter-cultural trust development, role expectations and cultural accommodation, which in turn result in trust consolidation or loss.

Definitions and Models of Trust Development
In the interdisciplinary management literature, our key concept, trust, is commonly defined as ''…a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another'' (Rousseau, Sitkin, Burt, & Camerer, 1998: 395). Mayer et al. (1995) colleagues' seminal model suggests that the positive expectations that underlie trust are largely captured by three dimensions of perceived trustworthiness: ability, benevolence and integrity. Ability (also known as competence) refers to the ''group of skills, competencies and characteristics that enable a party to have influence within some specific domain'' (Mayer et al., 1995: 717). Benevolence refers to a positive orientation of the trustee toward the trustor, including expressions of genuine care and concern. Integrity refers to adhering ''to a set of principles that the trustor finds acceptable'' (Mayer et al., 1995: 719), including promise fulfilment and word-action alignment. Seminal work by McAllister (1995) popularized the notion of cognitive and affective trust. Cognitive trust refers to beliefs about another's reliability, competence and trustworthiness, and taps closely into the dimensions of ability and integrity, whereas affective trust is based on the reciprocated interpersonal care and concern between the trusting parties, and hence is similar to the benevolence dimension (McAllister, 1995). It is generally assumed in the trust literature that trust develops over time as a function of the frequency, duration, depth, and diversity of experiences between parties (Lewicki, Tomlinson, & Gillespie, 2006). Seminal models of trust development by Shapiro, Sheppard and Cheraskin (1992) and Bunker (1995a, b, 1996) propose trust has different bases and stages. These models propose that trust begins 'at zero' with calculusbased trust, where one party makes a rational calculation of the benefits and costs of trusting. From this low basis, some relationships then develop into knowledge-based trust, which is grounded in sufficient interaction to get to know the other party. A few relationships then develop identification-based trust, which is grounded in the trustor's identification with the other party. Rousseau et al. (1998) integrate and simplify the concepts of knowledge-based trust and identification-based trust from these earlier models into the concept of relational trust, which develops from repeated or complex interaction between parties over time. Relational trust emphasizes the affective element of care and concern for each other that often develops in relationships over time.
In contrast to these models of incremental trust development over time, seminal work by Kramer (1996Kramer ( , 1999 and Meyerson, Weick and Kramer (1996) emphasizes presumptive bases of trust that enable trust to be assumed and swiftly established without an interactional history. Presumptive trust enables a high level of trust at the start of a relationship, however, such trust is fragile and can erode quickly if expectations and norms are violated (Meyerson et al., 1996). Two pertinent bases of presumptive trust identified by Kramer and colleagues are role-based and category-based (Kramer, 1999;Kramer & Lewicki, 2010). Role-based trust is founded on the strong expectation that role incumbents (e.g., doctors, supervisors) possess specific competencies and will fulfil the obligations and responsibilities associated with their role (Barber, 1983). As Kramer (1996) explains, to the extent that people have confidence that the role signals competence and positive intention, they can trust in the absence of individualized information or experience with the specific person fulfilling the role. Category-based trust is founded on the other party's membership of a social or organizational group, and the attribution of positive characteristics of the group onto individual group members. Where this group membership is shared, it can further facilitate trust development, as we tend to attribute positive characteristics such as cooperativeness, honesty and trustworthiness to in-group members, and be more suspicious of out-group members (Brewer, 1981;Kramer, 1999).

Emic Versus Etic Trust Research
The above-summarized insights on the development of trust have largely been won by monocultural empirical studies based on Western samples, particularly from North America (Fulmer & Gelfand, 2012). Most trust models take an etic or universalistic stance, assuming that trust is a universal or culturally generalizable concept (e.g., McAllister, 1995;Schoorman, Mayer, & Davis, 2007; for a review see Fulmer & Gelfand, 2012), and that the Western understanding of trust, as we have outlined above, is applicable for the study of trust irrespective of national cultural contexts (Zaheer & Zaheer, 2006). However, other researchers take an emic or particularistic stance, arguing that trust is a culture sensitive concept (e.g., Chen et al., 2014;Doney et al., 1998;Johnson & Cullen, 2002;Li, 2013;Noorderhaven, 1999;Wasti & Tan, 2010). For these scholars, trust can be interpreted and developed differently, depending on one's cultural imprint. Emic trust studies can be separated into two main categories: comparative cross-cultural studies, which investigate how trust varies across national cultures, and inter-cultural studies, which examine the development of trust between culturally diverse trustors and trustees (Dietz et al., 2010;Ferrin & Gillespie, 2010;Fulmer & Gelfand, 2012;Li, 2013).

Cross-Cultural Research on Trust
Most emic research on trust is cross-cultural, i.e., conducted in a comparative way with a focus on explaining differences between national cultures that influence trust (Ferrin & Gillespie, 2010). Based on the view that the role and nature of trust differ significantly across cultures, Zaheer and Zaheer (2006) call for more trust research which considers the specific cultural contexts. In line with our research setting, we focus on cross-cultural trust studies that examine interpersonal trust in (collectivistic) Chinese and (individualistic) German contexts.
In the Chinese culture, trust is usually translated as xin, which is a key component of Confucianism (Luo, 2000;Zhao, Shi, Xin, & Zhang, 2019). In an interpersonal context, xinren (ren = people) as a verb means 'to trust in or to confide in' (Eye, 2007;Yen, Barnes, & Wang, 2011), whereas xinren as a noun describes the heart-and-mind-confidence that another party will perform according to explicit or implicit expectations (Kriz & Keating, 2010). Xinren often serves as the interpersonal trust component of the Chinese concept guanxi, which takes a more instrumental approach by focusing on mutual obligations based on the continuous exchange of favors or to gain competitive advantage through personal relationships (Chen, 2001;Dunfee & Warren, 2001;Tsang, 1998). These definitions suggest that the concept of trust in China has important similarities to Western notions of trust (e.g., confidence, positive expectations, reliability etc.), while at the same time featuring culturespecific elements (Kriz & Keating, 2010). For example, the Chinese conceptualization of interpersonal trust xinyong (=credit) is different from the Western understanding of interpersonal trust as it is associated with a hierarchical relationship, in which a higher-positioned person will have more xinyong than a lower-positioned one (Leung, Bhagat, et al., 2005;Leung, Lai, et al., 2005). Another Sino-Western difference relates to affect-and cognition-based trust, which tend to be separate concepts in the West but are more interwoven in China (Chua et al., 2009). Therefore, Western conceptualizations of trust are only to a limited degree transferable to the Chinese context (Chen et al., 2014;Kriz & Keating, 2010;Luo, 2005). Doney et al. (1998) proposed a model of how dimensions of national culture (individualism-collectivism, power distance, uncertainty avoidance and masculinity-femininity) influence the form of trust in the relationship (calculative, prediction, intentionality, capability or transference processes). They postulate that trust is more likely to form when the trustor and the trustee share similar cultural values and norms. Similarly, Johnson and Cullen (2002) propose national culture influences the salience of the bases of trust (e.g., dispositional, experience, instrumental-calculus), which in turn influences trust. These authors note a limitation of their models, however, which is that they do not portray the trust formation and development process between members of different cultures. This, by contrast, is a key focus of the current paper.
Most comparative trust research has focused on how individualism and collectivism influence trust processes (Fulmer & Gelfand, 2012). Scholars report that collectivists place greater importance on benevolence and affective trust in their relationships than individualists do (Triandis, 1995;Wasti et al., 2011). Chen et al. (1998) argue that cognition-based trust is a stronger determinant of cooperation in individualist cultures, whereas affectbased trust is a stronger determinant in collectivist cultures. They suggest that cognition-based trust by itself is insufficient to foster cooperation with collectivist members. In contrast, Chua et al. (2009) found that economic exchange and thirdparty ties influence interpersonal trust among Chinese managers, whereas friendship has a stronger impact among American managers. Branzei et al. (2007) find that collectivists focus more on situational information, such as predictability and benevolence, and interpersonal ties, when assessing trustworthiness, whereas individualists base trustworthiness more on dispositional signals, such as ability, integrity, and common membership.
There are also a few studies on how different cultures react to trust violations and breaches (Fulmer & Gelfand, 2015). Members from collectivist societies tend to have a higher tolerance towards psychological contract violations (Westwood, Sparrow, & Leung, 2001) and respond in a more moderate way compared to Americans (Eckerd, et al., 2016). Similarly, Vogel et al. (2015) found Chinese subordinates perceived abusive supervisory behavior as more interpersonally fair than their US-American counterparts. As a result, following a relationship breach, Chinese still expressed higher trust towards their supervisors and continued to be committed to their workplace. In their conceptual article, Ren and Gray (2009) propose specific restoration mechanisms after members from individualistic and collectivistic cultures experience trust violations, showing restoration behaviors vary across cultures and emphasizing the cultural embeddedness of trust, violations and repair.
While comparative cross-cultural trust research informs us about trust behavior in different domestic contexts, these studies are largely static and do not investigate how trust develops over time between members of different cultures. Only inter-cultural trust research can inform us about trust development across cultural boundaries.

Inter-Cultural Research on Trust
Johnson and Cullen view inter-cultural trust building as a process of ''formative mutual realignment '' (2002: 344) in which exchange parties must ''mutually develop and agree on what behavior, activity or gesture in the relationship serves as a trust signal' ' (2002: 358). The comparatively few intercultural trust studies investigating the development of trust between parties of different cultures suggest that inter-cultural trust building is significantly different from intracultural trust building, i.e., amongst members of the same cultural group (e.g., Li, 2013;Newell, David, & Chand, 2007;Zaheer & Zaheer, 2006). In the latter context, signals of trustworthiness are easily understood since actors share the same socialization (Doney et al., 1998). In contrast, in inter-cultural contexts, encoding-and-decoding of trust signals is more complex (Johnson & Cullen, 2002). As Branzei and colleagues explain in their conceptual paper, ''mismatched signs'' hinder trusting choices, whereas ''signs that are aligned with the culturally laden expectations of the trustor hasten trust production' ' (1998: 78). Given these complications, it is hardly surprising that numerous scholars focused on the exploration of negative effects of cultural diversity on trust, suggesting that the greater the cultural differences or distance between parties are, the more difficult it is to establish trust between them (Branzei et al., 2007;Doney et al., 1998;Johnson & Cullen, 2002). For example, Jiang et al. (2011) found in their field survey of senior Chinese senior executives that executives show higher affect-based trust in overseas partners who shared the same cultural ethnicity.
Empirical work on inter-cultural trust between members of the same organization has focused on identifying the antecedents which either contribute to or inhibit trust. For instance, without cultural sensitivity people misinterpret situations and develop negative stereotypes. Therefore, being culturally sensitive is essential to communicate and work with others from different cultural backgrounds in a successful way (Kim, Heo, Lee, Suh & Kim, 2015) and, as such, is an important skill for people working in international contexts (Shapiro, Ozanne, & Saatcioglu, 2008). McCrae and Costa's (1997) concept of openness to experience is another construct which has been described as an antecedent to inter-cultural trust formation. Open individuals live through new experiences in a deeper, more intensive and more positive way than closed individuals: they seek exposure to a broad range of new experiences and ideas, cultivating an ability to critically reflect on accepted values and assumptions, as well as a high tolerance for ambiguity and high degree of behavioral flexibility (McCrae & Costa, 1997), which in turn fosters cross-cultural trust building. Referring to McCrae and Costa (1997), Gelfand, Aycan, Erez and Leung (2017) conclude that cross-cultural research with a Chinese focus has to integrate emic cultural facets (such as the concepts of reng qing, i.e., following norms of interaction, or of mianzi, i.e., face) which go beyond traditional Western constructs and concepts to generate more comprehensive theories across cultures.
Many of the inter-cultural studies on trust focus on relationships between parties of different organizations operating in a competitive frame, such as in negotiations (Lopez-Fresno, Savolainen, & Miranda, 2018) or buyer-seller contexts (Danik, 2014). Only few studies have examined inter-cultural trust between members of the same organization or organizations working collaboratively (e.g., joint ventures and strategic alliances). In a survey study of managers involved in US-Mexican strategic alliances, Rodriguez and Wilson (2002) found that trust for Mexican managers was driven predominately by social and affective cues, whereas for US managers, trust was based mainly on economic and strategic cooperation. Of particular relevance to our empirical context, Muethel and Hoegl (2012) examined conceptualizations of trust by Chinese and German managers in Sino-German joint ventures. They found shared trust expectations across cultural contexts, such as credibility, honesty and dependability, but also unique trust expectations not shared across cultures, such as openness and reliability for the Germans and shared understanding and morality for the Chinese.
A further key limitation of most of the intercultural trust literature is that it treats trust mainly as a static phenomenon. While the above referenced studies shed light on the antecedents of trust in inter-cultural relationships, and/or whether the forms of trust differ across cultural contexts, they overlook conceptually and practically most relevant dynamic trust development processes, which are likely to include cross-cultural learning practices, action-reaction loops, as well as cross-cultural adaptation and identity negotiation dynamics (Molinsky, 2007) which all have the potential to change trust outcomes in significant ways. The model of inter-cultural trust development, proposed by Dietz et al. (2010), integrates some of these considerations into a 'tentative' five-stage process in which (1) contextual factors (each party's cultural preconceptions, cross-cultural capabilities and motivation to adapt) influence (2) the parties' opening trust stance and (3) early interactional encounters, which then lead either to (4a) a trust breakthrough characterized by understanding, acceptance and accommodation to each other and a subsequent (5a) maturation of the relationship, or (4b) a trust breakdown characterized by a failure to understand, accept and accommodate, and hence (5b) dissolution of the relationship. The authors call for inter-cultural research to test and extend their 'tentative' model and deepen understanding of its etic and emic elements.
Furthermore, research on inter-cultural trust formation in hierarchical relationships is particularly rare, failing to account for how hierarchy and power differences may change the way inter-cultural parties navigate cultural differences in subordinate-supervisor trust relationships. This is surprising, given on one side the supreme importance of hierarchical relationships in organizations and on the other side ample evidence indicating strong cultural differences in the way hierarchical differences and power discrepancies are negotiated (House, Hanges, Javidan, Dorfman, Gupta & GLOBE Associates, 2004). Linking trust and leadership in culture-specific contexts, the GLOBE study highlights that some features of supervisors, such as competence, benevolence and integrity are universally endorsed and thus serve as etic trust facilitators (House et al., 2004), while other studies highlight that trust facilitating leadership elements might vary across cultures (Stahl & Sitkin, 2005). Ang and Tan (2016) found that cultural intelligence helps expatriate managers in China to develop trust (particularly benevolence-based trust) by fulfilling their Chinese subordinates' expectations of paternalistic leaders.
To address these conceptually and practically important limitations in inter-cultural research on trust, we respond to calls for more context-sensitive, emic examinations of inter-cultural trust dynamics (Dietz et al., 2010;Ferrin & Gillespie, 2010;Fulmer & Gelfand, 2012;Tenzer et al., 2014), specifically in a hierarchical working relationship, by conducting an in-depth study of trust development between Chinese subordinates and their German supervisors.

Research Design
We adopted a qualitative, inductive research design as the most appropriate to address our research questions for a number of reasons. First, qualitative designs are ideal for gaining an in-depth understanding of phenomena and processes in fields where there is limited theoretical development (Morgan & Smircich, 1980;Suddaby, 2006). Given that trust formation in inter-cultural relationships is still a nascent research area with limited unified theory and given our aim of inductively theory building, a qualitative design was ideal. Qualitative approaches are particularly suitable for studying complex, dynamic phenomena such as trust (Tenzer et al., 2014) and pursuing inductive theory building (Eriksson & Kovalainen, 2008) in the form of robust mid-range theory (Eisenhardt, 1989). Our differentiation of various trustor groups we uncovered only after several rounds of data analyses. Similarly, our discrimination of different relationship phases, and the role of acculturation as a specific construct of relevance for cross-cultural trust development emerged inductively from our data. Only as these specific foci emerged did we return to the literature. This iteration between data and literature review (Gioia, Corley, & Hamilton, 2013), spurred the development of our research questions and became the starting point for our theory development (Edmondson & McManus, 2007). It was only after several iterations between data collection, data analysis, and literature study, that we were able to develop our inter-cultural trust model. Second, our research design, which is based on in-depth, semi-structured interviews, is most suited to investigate the 'how' and 'why' questions (Pratt, 2009) we intended to explore and provided a more holistic picture than could be unveiled by deductive, quantitative research (Rubin & Rubin, 2012). Our interviews allowed us to study in-depth the trust relations between Chinese subordinates and their German supervisors and allowed us to recount vivid and contextualized description of (dynamic) occurrences which were ''close to the informants' experience'' (Gioia et al., 2013: 19) and (possible causal) structuring of the socially constructed ''worldviews of the people under study'' (Lee, 1999: 43, Maguire & Phillips, 2008. We used a ''semi-grounded approach'' (Fox-Wolfgramm, 1997;Strauss & Corbin, 1998), which is inductive and uses core techniques of grounded theory (e.g., theoretical sampling and constant comparison; Glaser & Strauss, 1967) but also allows initial frames of reference from the literature for gathering and interpreting data (Rynes & Gephart, 2004). In our case, this enabled us to build on relevant existing trust concepts and theories advanced in a mono-cultural or comparative context. Third, our qualitative design helped to unveil dynamic processes which constitute inter-cultural trust formation and development. While researchers repeatedly point to the value of process research in the field of international business, it remains heavily underrepresented (Jones & Khanna, 2006;Welch & Paavilainen-Mäntymäki, 2014): ''the process view has always played a minor role in the literature on International Management compared to the static perspective'' (Kutschker, Bäurle, & Schmid, 1997: 102). Our research design enabled us to combine data related to processes with process theorizing (see also Pettigrew, 1992;Welch & Paavilainen-Mäntymäki, 2014).

Research Setting
We chose our countries of investigation, China and Germany, for multiple conceptual and practical reasons. First, conceptually they represent a collectivistic (East Asian) and individualistic (Western) country, respectively (Hofstede Insights, 2020). Second, in China, employees typically show a significant interdependence of their personal and professional life (Zhang, Li, & Harris, 2015), whereas in Germany, employees segregate more strictly between personal and professional life (Bader, Froese, & Kraeh, 2018). Third, they represent different approaches to subordinate-supervisor relationships as China is a country where employees expect personal care by their supervisors (Chen et al., 2014) which is not the case for Germany (Brodbeck, Frese, & Javidan, 2002). Comparative research has shown the relevance of these three dimensions for the relationship between subordinates and supervisors and different trust conceptions (Huff & Kelley, 2003;Wasti & Tan, 2010). Finally, these countries represent the largest economies of Asia and Europe respectively, which have between them major trade and FDI flows and, as such, an active context for inter-cultural work relationships.

Data Collection
In order to investigate the trust of Chinese subordinates towards their German supervisors in all its ramifications, we chose a relatively complex data collection approach. We interviewed not only Chinese subordinates of German supervisors (the trustors) but also German supervisors of Chinese subordinates (the trustees) and in both cases, we collected these data at German headquarters as well as at subsidiaries in China. In total, we conducted 95 interviews. This multitude of different voices across hierarchical levels, nationalities, and locations allowed us to gain an in-depth and balanced understanding of the inter-cultural trust formation processes, which stands in contrast to most prior trust research that focuses on the trustors without considering the trustees. Furthermore, 50 of our interviews involved intact subordinate-supervisor dyads, i.e., Chinese subordinates with their respective immediate German supervisors. Table 1 provides a detailed demographic breakdown of our respondents.
Participants were recruited through China-related trade shows, professional social network platforms as well as through the first author's own professional networks. Respondents were exclusively white-collar employees, representing diverse functional areas (e.g., human resources, research and development, manufacturing, logistics, accounting, marketing and sales) and hierarchical positions (ranging from trainees, regular employees, lower, middle and upper managers, to vice presidents). Most of our interviews were conducted with companies in the manufacturing sector (e.g., high-tech, machinery, automotive, heavy steel). Furthermore, 42 (90) percent of our Chinese (German) respondents were male. The time worked in different cultural settings ranged from 3 months to 15 years, and the duration of the current supervisor-subordinate relationship ranged from 3 months to 4 years, with an average of approximately 2 years. We did not include shorter subordinate-supervisor relationships, given their limited utility for understanding trust development over time.
We chose a semi-structured interview design. This enabled comparability between our interviews by addressing the same core issues and questions in a structured way, whilst also enabling the emergence of new issues by flexibly asking respondentspecific follow-up questions (Myers, 2008) and allowing our respondents to provide us with rich, thick descriptions (Doz, 2011). By employing a narrative interviewing technique, we gave our interviewees the opportunity to describe their trust development processes. In the course of our data collection, the significance of developments over time became apparent. Qualitative, interview-based research is particularly well suited to detect such change processes.
To achieve investigator triangulation, six investigators conducted the semi-structured interviews. This allowed for the comparison and contrasting of findings, thus decreasing potential bias in the findings and interpretation (Denzin, 1970). All six interviewers were familiar with the German and Chinese culture and had knowledge of Mandarin, German and English. To ensure consistency among the interviews, the first author designed the semistructured interview protocols after having tested it in his first interviews, and thoroughly discussed it in group and individual sessions with the interviewers (Sinkovics, Penz, & Ghauri, 2008). Before interviewing the German supervisors, we explicitly asked the Chinese subordinates for permission to do so. Interviews were recorded and transcribed, and lasted on average a little over 1 h, with some taking close to 3 h. English interviews were transcribed in English, and German as well as Mandarin Chinese interviews in German or in English, depending on the researcher, while keeping culturally rooted and difficult to translate idioms and phrases in Chinese. In aggregate, our interviews represented over 95 h. In line with most interview studies, our investigation was not longitudinal in a strict methodological sense as the relationship development with supervisors were assessed retrospectively. While this might be a source of potential bias, retrospective interviews are a highly effective way to examine experiences and dynamics over multiple years and obtain information about long lasting change processes (Gustafsson, Gillespie, Searle, Hope-Hailey & Dietz, 2020). We asked our interviewees to take us through their relationship with their supervisor (respectively subordinate) from the very start of their collaboration up to the moment in which the interview took place, focusing in particular on the implications for trust building during this process. This emphasis on the temporal perspective helped us to understand how trust development varied for different groups of trustors over time, and to identify the particular relationship phases, and how these phases linked to the cultural adaptation processes of our interviewees over time. In interviews held in English we spoke about ''trust'', in Mandarin about xinren and in German about Vertrauen. To ensure that our respondents were also specifically referring to trust, we explicitly asked for trust in delineation to other related concepts, such as help (bangzhu) or reciprocal help (huzhu) (see also Kriz & Keating, 2010). However, we purposefully never asked our respondents to provide their definition of trust or any other abstract conceptualization, as we were more interested in their trust related experiences. From these narratives, we deduced different cultural approaches towards trust. For example, the differentiation of (more Chinese) relationship-based trust versus (more German) competence-based trust was never explained to us as such but resulted from our analysis of the transcripts and subsequent coding.
We asked both the Chinese subordinates and the German supervisors to interpret trust from the other culture's point of view and provide suggestions on how to increase levels of trust in such a cross-cultural, hierarchical setting. The first trustrelated set of questions, which inquired about how our interviewees view, form and develop trust in their own cultural environment, was the same for the Chinese subordinates and for the German supervisors. Subsequently, regarding the Chinese subordinates, we asked them to describe the trust formation process towards their German supervisors over time and to give account on the resulting consequences in their thinking and behavior. In this section, we asked for specific instances and critical incidents that influenced their perception, assessment and reinterpretation of trust towards their German supervisor. These critical incidents helped us to gain access to our interviewees' concrete feelings, thoughts and behaviors (Janssens, Cappellen, & Zanoni, 2006). From these particular occurrences, we subsequently deduced more general patterns. As for the German supervisors, we asked them how they establish trustworthy behavior towards their subordinates, if they adjust their behavior depending on the national (or other) characteristics of their subordinates and if so, what specifically they do. We also asked the German supervisors if they changed their attitudes and behavior towards the Chinese subordinates (with focus on establishing and developing trustworthiness) over time and asked for critical incidents which made them adjust in this respect. The interview protocol we used with Chinese subordinates working in Germany was slightly adapted for Chinese subordinates working in China and for the German supervisors working in China and Germany.
Our interviewing, simultaneous data analysis and literature consultation took place over the course of 4 years. As a result of these constant iterations, the interview questions were also in a process of continuous change. Through this long iterative process, we gained increasing confidence in the transferability of our interview questions across cultural contexts (paralleling internal validity in quantitative studies) and the ecological validity (paralleling external validity) and dependability (paralleling reliability) of our results.

Data Analysis
We coded transcripts using an ''open coding'' technique (Strauss & Corbin, 1998) with the help of the qualitative research software atlas.ti. During this stage, we labeled every passage with an appropriate code. All codes were developed in English to facilitate the data analysis. To assure reliability, all six members of the data collection team contributed to the coding. While there was in general agreement on the meaning, codes often differed. In these cases, the first author integrated similar codes to ensure consistency across the coding structure.
We followed Hollensbe, Khazanchi and Masterson (2008) coding technique and assigned, for example, a ''trust'' code in one of two possible situations: (1) when the passage explicitly included the word ''trust'' or (2) when the passage was a response to a question that has asked about trust. An example of (1) is: ''You give trust very quickly.'' An example of (2) is the response to: ''Describe how trust between you and your supervisor was built?'' Some of our codes were derived from the interviewees' statements (in vivo codes). For example, we assigned the code ''change in face concern'' to any passage describing a change in the importance a Chinese subordinates attached to the concept face (e.g., ''In China I was always concerned not to lose my face, but after I came to Germany, I did not care about it so much anymore.''). Other codes were taken from the literature (Strauss & Corbin, 1998) (e.g., the quote: ''I trust my boss because he is really good at doing his job'' generated the code ''competence-based trust'').
After completing the open coding phase, we compared our coded data, employing the constant comparative method (Glaser & Strauss, 1967), in different ways in order to merge codes into more conceptual categories and finally into theoretical constructs (Lee, 1999). First, we compared different parts of each interview to ensure consistency. Subsequently, we contrasted interviews with the Chinese subordinates (other-perception) with those with the German supervisors (self-perceptions) on the trustworthiness of the Germans, paying particular attention to convergence tendencies. At this stage in analyzing our data, we identified three distinct types of trustors (Chinese subordinates). When we further investigated how these three types distinguish themselves from one another, we lifted our analysis up to a more theoretical level, establishing the concept of 'cultural profile'. Our use of the term cultural profile does not refer to the way culture manifests itself at the individual level (e.g., Maznevski, Gomez, DiStefano, Noorderhaven, & Wu, 2002) but rather denotes a combination of the individual's cross-cultural experiences and identification with their own cultural background, which in turn influence the individual's degree of cultural sensitivity and openness to other cultures. In this vein, a few cross-cultural management scholars have already employed the term cultural profiles for cross-cultural interactions (Caprar, 2011;Yagi & Kleinberg, 2011). Regarding cultural profiles, we distinguish between cosmopolitans, hybrids and culturally-bounds. Furthermore, we also noticed that trust was for most of our Chinese respondents more relation-based and less competence-based. This was never explained to us as such but we became aware of this difference from our coding efforts of the transcripts. Furthermore, we juxtaposed the four groups ''Germans in Germany'', ''Germans in China'', ''Chinese in Germany'' and ''Chinese in China'' to obtain an understanding whether there is a connection between home and host country and the applied cultural adaptation strategies. During these comparative processes, further connections between the codes emerged. We also coded in particular for factors encouraging or inhibiting the adaptation processes to understand the underlying reasons why some interviewees engaged in cultural adaptation while others did not. Throughout our coding process we compared our data in an iterative fashion with existing literature until we reached the point of theoretical saturation (Guest, Bunce, & Johnson, 2006).
Subsequently, we integrated related first-order codes into superordinate categories to move from a primarily descriptive to a more conceptual level (van Laer & Janssens, 2011). For example, the codes ''paternalistic expectations leading to initial trust'' and ''feeling of dependency leading to initial trust'' were consolidated into superordinate category ''presumptive trust based on role-expectations''. During this phase of coding, our data also revealed the processual nature of trust, which was later segmented into several phases of trust formation and development. We came to distinguish, for example, between the initial contact, trust continuation, trust disillusionment and acculturation phase. Appendix 1 (online only) depicts in detail our data structure with the aggregation process from first order codes, to superordinate categories to theoretical constructs.

FINDINGS
From our data analysis, we inductively developed a model of inter-cultural trust development in hierarchical relationships. Next, we describe the findings underlying our model, before presenting our conceptual model.

Trustor Cultural Profiling: Cosmopolitans, Hybrids and Culturally-Bounds
We noticed in our data certain patterns in the trusting behavior of the Chinese subordinates towards their German supervisors. When studying the underlying reasons for these differing trusting behaviors, we found that their demographic background in terms of cross-cultural experiences as well as the extent of their identification with their own (Chinese) cultural background (as opposed to the German host culture) played decisive roles. We further observed that these differences in demographic backgrounds resulted in different degrees of cultural sensitivity and openness to experience among our trustors. Taken together, these aspects led us to distinguish between three (cultural) profiles, which, in turn, resulted in clearly distinguishable paths of trusting behavior. We also considered possible differences in the trustees, i.e., the German supervisors, and identified that the length of cross-cultural experiences was of most relevance. We first describe the three cultural profiles of Chinese subordinate trustors. For quotes, illustrating all three cultural profiles, we refer to Appendix 2 (online only).

Cosmopolitans
Our first group of trustors, which we labeled cosmopolitans, represented about one quarter of our Chinese interviewees. Those cosmopolitans had mainly been raised in one of the first-tier cities in China and belonged to the 'new generation' of workers (xingshengai yuangong), being born after 1980 (Zhu, Xie, Warner, & Guo, 2015). In terms of their parents' upbringing, educational approaches ranged all the way from very traditional to nontraditional approaches with respect to Chinese values, such as obedience and harmony. However, what they had most in common was extensive exposure to a foreign environment during the formative years of their life: they either had studied abroad (entirely or as exchange students) or had significant exposure to foreign exchange students at their Chinese universities. Many also did internships abroad or had prior assignments in foreign countries. Cosmopolitans had a multicultural circle of friends which, according to their own judgment, significantly impacted their mindsets and behaviors. Characteristic for cosmopolitans was a high degree of cultural sensitivity and openness to experience, which can be attributed to their comprehensive exposure to foreign contexts at a formative age. Cultural sensitivity and openness to experience appeared to be highly relevant for the acculturation and trust formation processes of this group of trustors in their current positions. Cosmopolitans predominately worked in Germany as opposed to being based in their home country (as locals in German-owned subsidiaries). As a result, in particular those cosmopolitans working in Germany became strongly acculturated to their (German) host culture and often got significantly detached from their (Chinese) home culture environment.

Hybrids
The majority of our Chinese respondents belonged to the second group which we called hybrids. Members of this group kept (deep-level) characteristics from their (Chinese) home culture but at the same time came to adopt also certain (surface-level) aspects of the (German) host culture. In their own words they are 'a good mix' or 'a blend' between their original and host cultures. Just like cosmopolitans, this group had been exposed to an international environment, however in more limited ways or later in their lives. Most of them were not working in Germany but in their home country (China) as locals for a foreign (German) subsidiary. Their cross-cultural exposure was particularly impactful in cases where their employing organization was pursuing a strong globally standardized strategy with globally uniform policies and processes. They usually worked closely as subordinates of (German) expatriates. Some had studied abroad for a limited period (i.e., one or two semesters) or had gone on business trips to Germany. Several hybrids had recently been assigned to Germany as an inpatriate. However, what distinguishes the hybrids from the cosmopolitans is that they had not fully immersed themselves into a foreign environment and often continued to socialize with those of the same national background, thus separating their private from their professional sphere. As a result, even though they have experienced cross-cultural exposure and were open to Western concepts on the organizational level, we define hybrids as still being traditionally-oriented, especially regarding their interpersonal relationships and particularly role expectations (i.e., towards their supervisors but also towards themselves as subordinates). Compared to the cosmopolitans, hybrids show a moderate degree of cultural sensitivity and openness to experience. However, hybrids self-assessed themselves as being more cross-culturally savvy than their own description of their attitudes and behaviors would account for.

Culturally-bounds
About one quarter of our Chinese respondents can be regarded as traditionally Chinese in all regards. They had only limited exposure to cross-cultural (living or working) environments. Culturallybounds are so deeply engrained in their own culture that, by their own account, they struggled to be open to non-Chinese cultures, and appeared the least culturally sensitive of all cultural profiles. As a result, they do not adjust to any substantial degree to foreign environments. This was particularly evident for their role expectations of supervisors and subordinates. They were often raised in a traditional way in small-or medium-sized cities or in the countryside and moved to a more cosmopolitan Chinese city merely for professional reasons. They appeared to work for a foreign organization for financial purposes only, rather than for intrinsic motives. In the following, we describe for all three trustor cultural profiles, the trust development process across various phases which we inductively established (for additional quotes pertaining to the trust development phases, we refer to the online Appendices 3a-d, 4a&b and 5a&b).

Initial Contact Phase: A High Level of Initial Trust
Our interviews reveal the existence of a distinct initial contact phase during which the Chinese subordinates and German supervisors began to interact and work together for the first time. Hence this phase represented the early formation of the relationship and lasted approximately 1-2 months. During this phase, Chinese subordinates enter the relationship with their German supervisors with very high levels of initial trust: Why should I not trust my supervisor? I think there is no reason to work for a supervisor if you don't have the intention to trust him. (Chinese 30) This is remarkable insofar as the few, initial studies on trust between parties of different cultural background suggest that cultural differences negatively affect the development of trustful work relationships (Doney et al., 1998;Luo, 2002) due to conflicting goals, beliefs or expectations between people from different cultures (Fiske & Ruscher, 1993;Kramer, 1991;Sitkin & Roth, 1993;Williams, 2001). This literature suggests low instead of high trust levels. Given this surprising finding, we were at first concerned that our Chinese interviewees might express trust particularly in Germans, as Germans tend to be highly respected in China. However, we interviewed several Chinese subordinates who had previous experience with foreign supervisors of other nationalities before they were assigned to a German supervisor and probed them on this particular issue, obtaining the same results: Before coming here, I worked for an American and also for an Italian boss.
[…] I trusted all of them in the beginning when I started working with them. Their style was very different, but this one aspect [of trust] was the same. (Chinese 3) In addition, and more importantly, we observed that Chinese subordinates trust their German supervisors at the same level as their Chinese supervisors, so that we are confident that the subordinates relate to their supervisors in terms of their role rather than their nationality: I worked for several supervisors so far. When I started out to work with them, I always trusted them. No matter if they were from China or, as in my current situation, from Germany. (Chinese 17) We were also astonished by the uniformity of our results across the three groups of Chinese trustors. In subsequent interviews, we therefore probed for the reasons for these homogeneous findings. Our data suggest that most of our Chinese interviewees (i.e., the hybrids and culture-bounds) regard their German supervisors in the initial contact phase not primarily as individuals and not even mainly as foreigners, but instead categorize them as supervisors who are expected to provide care and guidance and therefore are to be trusted: In China people are used to get orders by their parents, teachers or their bosses at work. They get specific orders and follow them, because only if you follow, you are seen as a good employee. This way I can trust that the boss teaches me how to do things and this helps me grow. (Chinese 5) Based on our data, we therefore conclude that the key to our unexpected observations lie in the phenomenon of quickly formed presumptive trust (Kramer & Lewicki, 2010). Our data reveal that the trust Chinese subordinates have from the start in their supervisors, stems from their unambiguous and hierarchy-based role expectations regarding obligations and rights that are deeply embedded in the Chinese culture. Following the five clearly defined Confucian dyadic relationships (traditionally: father/son; older brother/younger brother; husband/wife; emperor/subject; friend/friend), subordinates owe their supervisors (parallel to the father or older brother) obedience, while supervisors can be trusted to show in return care for their subordinates. Consequently, the German supervisors are categorized by their Chinese subordinates as paternalistic care providers, who can be presumed to be trustworthy and to demonstrate benevolent behavior: When first working with my German boss, I had this basic trust towards him. It was not difficult for me to trust him because I was open. I saw him as an older brother. Yes, I saw him like an older brother who will take care of me. (Chinese 4) This expected benevolence is therefore met from the start with high levels of presumptive trust. In addition to this element of presumptive trust, we also saw calculative aspects, based on the necessity for Chinese subordinates to get along with their supervisors: Chinese are relationship-based. You have to build your relationships. It all depends on the relationship. For example, our company is managed by people. By leaders, not necessarily by regulations. In other countries, the rule is king, but in China, the rule is made by the people. That's why we need this relationship. And so we trust our manager. We have people here who say they work in this department because of the manager, because of how he influenced the department in many ways. (Chinese 52) This necessity to establish a trusting relationship with the supervisor becomes even more evident when considering the near absolute power Chinese supervisors have over their subordinates' salary, bonuses and career prospects: By contrast, for the cosmopolitans, we found a similar but not entirely identical explanation for their presumptive trust. Here, the role expectations were not about German supervisors presumed to act similar to Chinese supervisors but instead more as foreigners, or more specifically as Westerners, a culture they were very open to engage with:

Journal of International Business Studies
At the beginning I trust especially foreign people. (Chinese  2) To conclude, in the initial contact phase, our interviewees seemed to experience presumptive trust, which was based not on their own experience or personalized knowledge of their supervisors as individuals, but rather entirely on their culturebased role expectations. This suggests that (except for the comparatively few cosmopolitans) supervisor status is far more important than foreignness. Some of our interviewees even went so far as to state that they consider it to be their job to trust their supervisors.

First Bifurcation: Trust Continuation Phase vs. Trust Disillusionment Phase
After the initial contact phase, which showed high trust levels for all three groups of trustors, we observed in the next stage of the trust development process a first bifurcation. This led to what we labeled the trust continuation phase, characterized by continuous high levels of trust, and the trust disillusionment phase, typified by a gradual but steady decline in trust.

Trust Continuation Phase: Trust Remains High
Those Chinese subordinates who reported continuing to trust their German supervisors beyond the initial contact phase and throughout the entire relationship with them were almost exclusively cosmopolitans. We attribute their continued high trust levels to the deep-level cross-cultural adaptation this group already went through and to their profound wish to live and to work in an international environment. Members of this group therefore have fully immersed themselves into the German culture, also by spending spare time with German friends or even by entering a romantic relationship with a German: I have a lot of international experiences. I worked only 4 months with a Chinese boss, but when I started my career at [German company] I only had German bosses. We worked together very well during the entire time, we had a very good relationship all along. I received a lot of information about Germany, learned new working skills and had the chance to go to Germany and connect to Germany. I have an aunt in Berlin and she told me a lot of interesting things, so I wanted to work and live in Germany. It feels like home, I even have a German girlfriend. (Chinese 48) Furthermore, we also noticed that most of these respondents had assumed at work the informal role of boundary spanner, facilitating smooth communication between Chinese and Germans and mitigating conflict, a role which necessitated trusting both sides of their interaction partners. The identification of these cosmopolitans with the German culture went for some even as far as not being able to imagine anymore to work for a Chinese organization or even for a Chinese supervisor: I  We found that many of our cosmopolitan respondents experienced category-based trust towards their German supervisor in both the initial contact as well as the continuation phase, not because they identified with this individual personally, but more with the culture he or she represented. As the cosmopolitan Chinese subordinates' are characterized by a high degree of cultural sensitivity and openness to experience, it is relatively easy for them to adapt to the German culture. Consequently, irrespective of whether the German supervisors demonstrated any culturally sensitive behavior towards their Chinese subordinates or not, most cosmopolitan Chinese subordinates had no difficulty in trusting them. Hence, our results suggest that the presumptive trust, the cosmopolitan Chinese subordinates bestowed upon their German supervisors during the initial contact phase and which was based on role expectations and identification with the German culture, continued over time. Moreover, it matured into a more robust form of relational trust as the cosmopolitans developed greater knowledge and experience with their German supervisors.

Trust Disillusionment Phase: Gradual but Steady Decline of Trust
The significant majority of the Chinese subordinates we interviewed, particularly the large group of hybrids and the smaller group of culturallybounds, did not move from the initial contact phase on to the trust continuation phase. To the contrary, most of them reported experiencing a gradual but steady decline of trust in their German supervisors. Over time they came to recognize that their expectations of their German supervisors, which had been the foundation of presumptive trust in the initial contact phase, were not met as Chinese and Germans have fundamentally different expectations about their roles as subordinates and supervisors. Their German supervisors, particularly those based in Germany, often failed in their view to recognize the needs of their Chinese subordinates, specifically regarding their need for personalized care and benevolence, and were not willing to deviate from their customary way of treating subordinates. This unwillingness or inability to fulfill their role in the way Chinese subordinates expected was met with an erosion of trust: The Germans just don't understand us. They act in ways which makes it impossible to keep up the trust. (Chinese 8) During this phase, the initial presumptive trust eroded in most cases gradually but steadily because the German supervisors did not live up to the aforementioned role expectations their Chinese subordinates had placed in them and the Chinese subordinates were unwilling to reevaluate their own culturally bound role expectations. The failure of the German supervisors to adapt to their Chinese subordinates can be explained by the fact that most of them had little to no previous international working experience and therefore often lacked cultural awareness of these expectations. Many Chinese interviewees complained about their German supervisors showing little respect for culturebased differences. In particular, this was true regarding the relationship-orientation of the Chinese, which continuously clashed with the German task-orientation. This fundamentally violated the Chinese role expectations based on Confucian obligations and benevolence shown by supervisors to their subordinates. As a consequence of their own limited cultural sensitivity and openness to experience, Chinese subordinates were largely unable to reconsider their own cultural role expectations of their supervisor, making it impossible to continue to trust, or develop relational trust: This lack of benevolence in the subordinate-supervisor relationship was very much mirrored by many comments we received from the German interviewees: No, I don't talk about personal things at work. These things are private. I don't even ask employees to go and have lunch with me. (German supervisor of Chinese 9) As the Germans, especially those working in Germany, did not seem to show sufficient interest in the personal and family affairs of their subordinates, this disregard for their subordinates' relational needs inevitably resulted in disillusionment. As a result, the German supervisors were no longer categorized or viewed as a father or older-brother figure (to whom collectivists generally show a high level of trust) resulting in lower levels of trust. Interestingly, we noticed that the decline of trust in the trust disillusionment phase, which took anything between several weeks to several months, was almost as uniform a trend across our respondents (with the exception of the cosmopolitans) as were the high trust levels in the initial contact phase. The only nuance within the trust disillusionment phase we observed was that trust erosion in German supervisors was more pronounced in the subsidiaries (located in China) compared to the headquarters (located in Germany). We attribute this to Chinese subordinates on their ''home turf'' expecting even more to be treated according to Chinese customs. Overall, the Chinese progressively realized in these still relatively early stages of their intercultural interactions that the initial high degree of presumptive trust in their German supervisors had been premature as the Germans did not meet their hierarchical role expectations. This trust disillusionment led to a quasi-uniform trend of gradual but steady decline in trust.

Second Bifurcation: Separation Phase vs. Acculturation Phase
We already have established that the cosmopolitans simply continued to hold high trust levels (trust continuation phase) without undergoing any further change, which is why in the further depiction of the trust development process we do not discuss them further. In contrast, for the hybrids and culturally-bounds, whose trust in their German supervisors had gradually but decisively declined, we observed in the next stage of the trust development process a split trust pattern. This next stage constitutes the third phase in our trust model. This third phase is distinct from the previous ones insofar as the Chinese trustors take at this juncture for the first time a more proactive role in evaluating their specific supervisor-subordinate relationship. Up to this point, trust was very much a function of undeviating behavioral patterns: in the initial contact phase, high trust levels were due to (positive) role expectations in the supervisor; and in the trust disillusionment phase, declining trust levels were an almost automatic (negative) reaction to the unfulfilled role expectations in the supervisors. Only in the third and final phase, which in most cases lasted until the end of the working relation, did Chinese subordinates finally process their experiences and make their own, largely independent and conscious choice about how to react to the previously experienced disillusionment. Depending on this choice, we observed in this third phase another bifurcation of trust development. One group of hybrid and culturebound trustors chose to enter what we called the separation phase, losing trust in their supervisor permanently. However, the (larger) group of hybrid and culture-bound trustors decided to go the opposite path, leading to what we label the acculturation phase.

Separation Phase: Perpetuation of Lost Trust
We identified a set of Chinese subordinates (and German supervisors), who did not take any steps to adapt to the other party's cultural values and behaviors. Subordinates who opted for the separation strategy were mainly culturally-bounds, who continued to be unable or unwilling to reconsider their original cultural expectations, which remained unmet. The supervisors of these subordinates had little extensive international working experience and showed no willingness or sign of cultural adaptation. This resulted in persisting low levels of trust of Chinese subordinates in their German supervisors: I am working here for 3 years now and I worked with several [German] supervisors. But it is not so easy for me to work with them. In China, there is more focus on the relationship, but in Germany it is only about work, work, work. I want to have a good relationship with my supervisors, but I do not know anything about them, except for their work. And they also do not ask questions about myself. How can I trust people who are not interested in me? (Chinese 10) In most cases, Chinese subordinates understood the reasons for their unmet expectations lied in cultural differences, however, they were simply unwilling to adjust their expectations because they described their cultural imprint as too strong and inflexible to adapt: I have the feeling that I am still Chinese, I did not change. When I met my parents again after quite some time, they told me that I did not change…when we have a problem in China we handle it more delicately. Here it is still very difficult for me to say things directly as my [German] colleagues do. My colleagues and my supervisor say it is ok to say if something is wrong or if there is a problem with a project, and that it is not fair for the others not to say it right away, if something is wrong. I know it is not easy for them to understand why it is difficult for me to mention problems. But it is also difficult for me to change, my heart is still Chinese. (Chinese 29) In other cases, Chinese subordinates actually wished to be able to adapt, however their cultural identity was too strong to allow them to do so: From my working experience here in Germany I notice the big differences between China and Germany. I tried to adapt, but after some time I gave up, because somehow I already knew myself, I knew I cannot adapt to this company culture because my personality is so strong. (Chinese 49) The main problem of adaptation lies in the fact that the own cultural identity is seen as being at stake: I am here for 3 years now, but it is difficult to adjust because I don't want to give up my Chinese identity. (Chinese 20) While the trust disillusionment phase was characterized by an emotional reaction (decline of trust) to having one's role expectations blatantly violated, the case for the separation phase is different: here the trustors had enough time and experiences to analyze and evaluate the situation. Overall, we also clearly noticed that the Chinese subordinates reflected more on cultural differences and resulting differing behavioral patterns than did the German supervisors. Consequently, based on these reflections, the Chinese trustors were able to make a personal trust choice. And a substantial number of them, unable or unwilling to adjust, came to lose trust in their German supervisors for good. They withdrew emotionally from the relationship with their supervisor, avoiding interaction with them wherever possible and, in more severe cases, even left their employing organization.

Acculturation Phase: Trust Repair and Trust Consolidation
While a non-trivial amount of Chinese subordinates irrecoverably lost trust in their German supervisors, the clear majority of them, the bulk of hybrids and some culturally-bounds, went the opposite path: they successfully overcame the trust disillusionment phase by adapting to the German cultural context and thus were able to rebuild trust in their German supervisors. Our interview data showed that this acculturation process underwent two distinct consecutive sub-phases: the early acculturation phase during which trust is repaired and the mature acculturation phase during which trust is consolidated.

Early acculturation phase: Trust repair
In this first stage of the acculturation phase, the Chinese subordinates successfully managed to repair their trust relationship with their German supervisors which led to increased trust levels. This trust repair was a consequence of initial adaptation steps taken towards the German culture: I trust him [the supervisor] because he knows very well how to do the job. We don't have private contact. I found out that in Germany teams do not have so much private contact during or after work. This is totally different than in China. But I got used to it. Not immediately, but after some time. (Chinese 15) These initial cultural adaptation steps were not the result of a change of cultural identity on the part of the Chinese, rather they simply learned over time how the Germans operate and the underlying reasons for their conduct. In this process, they came to realize that Germans focus more on competence than relationship building. In addition, they accepted that this behavior, while being different from Chinese conduct, is not untrustworthy but simply following a different logic. Consequently, in their decision whether to trust their German supervisors or not a shift of focus took place: away from an emotional reaction to a violation of one's relationship-based role expectations regarding benevolence (''my supervisor doesn't care about me'') and towards the cognitive understanding and acceptance of cultural differences (''my supervisor cares about my performance and that is fine with me''). This reattribution of violated behavioral expectations helped the Chinese subordinates to adjust the basis of their trust, shifting away from it being relationship-based and towards being competence-based. This shift was decisive for trust repair. As one respondent put it: In the beginning everything is strange and it is very difficult. The first step is to know. After a while it is not so difficult anymore, you just have to know things, you need to say: They [the Germans] are just like this, this is how they behave, you learn about them, you learn to understand them and you can talk about these differences…Now I would say that it was not so difficult to get adjusted [to working with a German supervisor], but I also had a couple of learning processes to tackle. (Chinese 30) Many of our Chinese respondents stated that understanding the counterpart's culture was crucial for triggering the cultural adaptation process. More specifically, two kinds of knowledge were of relevance: they learned what are appropriate role expectations for them to have of their (German) supervisor, and they learned what are the role expectations their supervisor had of them. With this newly gained knowledge, the behavior of the German supervisors became predictable and understandable which led to an increase of knowledgebased trust. The biggest hurdle for them in this process was to acknowledge that they could not expect of their German supervisors much of a relationship-based orientation; instead, they had to accept the task-orientation of the Germans. Once they understood that the trust for Germans is based less on benevolence in the relationship, and more on competent task performance, they were willing to adapt and readjust: In China your manager normally just comes to your desk and asks you how you are doing. This made me feel good because my boss recognized and appreciated me. This created a positive atmosphere. But German managers don't do this. I realized that in Germany you use regular meetings to talk business and not to do small-talk. Now I know. Now I can understand my manager, what he wants and why he wants it this way and this is completely okay. We work very well together. I think this is what also Germans say is a productive and trustful working atmosphere. (Chinese 11) Consequently, most respondents still considered themselves at this stage to be largely Chinese and adapted only for the time being to the cultural context of their supervisors.
During this phase, we noticed that a series of factors improved the chances of trust repair and ensuing increasing trust levels by the Chinese subordinates. The first of these moderating factors is cultural accommodation by the supervisor. We noticed that trust repair was facilitated if also the German supervisors made a step towards the Chinese and acted towards them in a more relationship-oriented fashion. While many Chinese subordinates, particularly those working in Germany, ultimately went a long way to adapt to their German supervisors, most of the German supervisors, in particular those in Germany, did not show the same efforts to adapt to the Chinese value system. However, in those cases when they chose to adapt, this was clearly beneficial for trust repair: I had [a Chinese] in my team who was showing me photos of his family and then he wanted to see photos of my family. This was really strange because we usually don't do something like that. But I liked it and it left a positive impression. When he saw that I am interested in China, he was inviting me to his home, which is also not so normal here in Germany. It is also a little bit difficult because of the hierarchy issue; I don't know what the others might think. But from now on I try also to talk about personal aspects every now and then with him. I have the feeling this means something for him. (German 7) Country context was another factor influencing the likelihood of trust repair. We noticed that cultural accommodation and trust repair occurred more frequently with those Chinese working in Germany as opposed to those being employed as locals in China: I chose to come to Germany. Understandably, I am required to adapt here as I would expect Germans to adapt when they come to my country. (Chinese 23) A third moderating factor was the trusted third party of trustor nationality. Some Chinese subordinates were helped in their trust repair process by other Chinese (often from the group of cosmopolitans) who took on the role of boundary spanners. By explaining why Germans acted the way they did (often related to their preference for task orientation over relationship orientation), they facilitated the cultural adaptation process and subsequently the trust repair of their colleagues, leading to an increase of knowledge-based trust: It helps to learn from other Chinese because we are used to a different leadership style. So, it helps if a Chinese colleague has been here for 3 years to tell me that the expectation of a German manager is that I need to give him proactively feedback and that the German managers do not always proactively approach their employees and appraise their performance. It helps a lot that my Chinese colleagues explain how German leadership works. If I make these experiences by myself, I would be demotivated very quickly. But if I have this information upfront, then I know what to expect. That helps me to change expectations. (Chinese 36)

Mature acculturation phase: Trust consolidation
This phase, which follows on from the early acculturation phase, is characterized by Chinese subordinates moving from surface-level to deeplevel cultural adaptation and the ensuing consolidation of trust in their German supervisors. However, not every Chinese subordinate who went through the early acculturation phase also succeeded at transitioning into the mature acculturation phase. We found that mostly hybrids and less so culturally-bounds crossed the threshold to this phase.
In the early acculturation phase, the Chinese subordinates adapted to a differing cultural context, as a means to better function in this new environment. As such, adaptation was more about behavioral alignment than about changing their own cultural identity. By contrast, in the mature acculturation phase the Chinese subordinates internalized the culture of the German supervisors as a way to integrate or even assimilate themselves into the other culture. In this sense, this phase is similar to the consolidation phase, typical for the cosmopolitans. The main difference is that with the mature acculturation phase, the hybrids working in Germany first went through the trust disillusionment phase before arriving at this stage. At this point, the degree to which the German supervisors also made an effort to adapt to their Chinese subordinates was of no relevance anymore, as the Chinese were now intrinsically motivated to integrate themselves more fully into the German culture. Our Chinese interviewees emphasized that it was their own decision to renegotiate their cultural identity and to move from a surface-level to a deep-level cultural adaptation, internalizing German cultural norms: I found out that Germans are very direct. I like this now. I am not scared of criticism, I also criticize a lot... This helps me because I try to learn as much as possible here and always try to improve…. I also know that Germans take their work seriously and they do not intend to offend me. … Now, I am like this also. I am nice in my private life, but strict at work. (Chinese 47) As a consequence, this internalization of the German culture at times could also lead to an estrangement with the own, Chinese culture: I and my [Chinese] colleagues changed a lot in Germany. Because we had to change. That's good. We take some good things from the West. But whenever I go back home, it is difficult. If you left your friends for years and then you go back, you feel that you don't match anymore. So there is some change. Also with the family, after you have conversations with your parents, they feel you have changed. (Chinese 43) This adaptation process away from the original culture also did not remain unnoticed by others, including members of the deep ingroup, such as family members, colleagues and friends: It is not only me who realizes that I became somewhat German since I have been working for this company. Also my friends and especially my parents realized this. They are a little concerned about how my communication changed because I became more direct and talk back (laughs). (Chinese 54) Our data also clearly show that the just outlined cultural adaptation process has major implications for the trust development process. Feeling now very comfortable with German cultural norms, a deeper level of a trusting relationship is achieved: My boss puts a lot of trust in me. Over time he also gave me a lot of freedom and liberties. He trusts me to think for myself and wants me to try out new things. I develop a concept and he will give his feedback. He respects me and I try not to disappoint him. I do my best to satisfy his expectations and I think that is a very good feeling indeed. He can trust me and I can trust him. (Chinese 47) The repaired trust from the early acculturation phase was still fragile as it was only based on the trustors' newly gained knowledge about cultural differences and the perceived necessity to adapt to differing cultural norms. By contrast, in the mature acculturation phase, trust in the German supervisors was now consolidated on a much stronger foundation: the adoption of cultural norms as a result of a cultural negotiation process.
Our interviewees reveal that during the acculturation phases, new cultural norms were more and more internalized. As a consequence, trust became now firmly consolidated and gradually transformed from knowledge-based to identification-based. Trust levels therefore further increased, substantially improving the supervisor-subordinate relationship: After all these years I feel good here. I would even say that sometimes I think like a German. I have good relationships with my boss, my colleagues and also with my international friends. So the trust that I feel about my boss and my colleagues, my friends, but also the company grew stronger. But to be honest, this is not a surprise. The longer you know someone the more you trust the person...or the more you distrust the person if the experiences are bad. (Chinese 34) During the mature acculturation phase, we identified a set of moderating factors which influenced the chance of trust consolidation. The first is the trusted third party of trustee nationality. While also for the earlier acculturation phase third parties played a role, they were predominantly Chinese nationals who, in their role as boundary spanners, helped their colleagues to better understand the other culture. In this mature stage, however, third parties were friends or colleagues mainly of German nationality who helped the trustors to immerse themselves more fully into the foreign environment and ultimately to achieve much higher trust levels: When I think about it, without my German friends I would have never been able to understand the German culture to the point that I do now. More than just helping me to understand, they also helped me to internalize the culture so that it is completely normal for me now. (Chinese 43) The second moderating influence was time spent in the foreign environment. Deep-level cultural acculturation and ensuing high trust levels in the German supervisors requires a renegotiation of one's own cultural identity, which needs time. Consequently, our data clearly show that only those Chinese subordinates who already had spent several years of intense contact with and socialization in the host country culture, succeeded in deeplevel cultural adaptation. By contrast, local employees of subsidiaries of German companies in China reached this phase less often: Years ago, I was entirely Chinese, but by now not so much anymore. In the past, I took everything personal. In those times, I was Chinese. But now, especially at work, I am not so personal anymore. I don't care.
[…] On the contrary, now I enjoy it that I can have discussions and I don't have to be too careful what I am going to say. (Chinese 13)

A CONCEPTUAL MODEL OF INTER-CULTURAL
TRUST DEVELOPMENT On the basis of our findings, we derived a model of inter-cultural trust development between supervisors and subordinates (see Figure 1).
The main conceptual building blocks of our model are the three distinct cultural profiles of trustors (cosmopolitans, hybrids and culturallybounds), which are each associated with varying trusting behavior across consecutive relationship phases, which we label the initial contact phase, trust continuation phase, trust disillusionment phase, separation phase and acculturation phase (early and mature). Embedded in these relationship phases, are ensuing trust dynamics, with trust either remaining high, declining, being restored or consolidated. Each of these various trust dynamics has a clearly distinguishable trust form and underlying psychological mechanisms (role expectations and cultural accommodation).
In the initial contact phase, trust was high for all three cultural profiles of subordinates based on a presumptive form of trust, specifically category-based and role-based trust. These high levels of presumptive trust are based on the psychological mechanism of positive role expectations. For the cosmopolitans, due to their high degree of cultural sensitivity and openness to experience, their presumptive trust is based on the role expectations of supervisors behaving according to their own cultural values and norms (e.g., for Germans, focusing on task competence). By contrast, due to their lower degree of cultural sensitivity and openness to other cultures, the presumptive trust of hybrids and culturally-bounds is based on their own culture's specific role expectations towards supervisors (e.g., for Chinese, to be benevolent and provide care and guidance). Hence, the psychological mechanism of positive role expectations is key to strong presumptive trust early in the relationship, with supervisor trustees primarily perceived as representatives of collectives, namely either as 'supervisors' or 'representatives of their cultural group' (e.g., Westerners/ Germans).
The extent to which the trustor's expectations are met over time results in a first bifurcation of trust: Cosmopolitans' role expectations of supervisors (e.g., as carriers of the supervisor's cultural norms) were met, which sustained trust into the trust continuation phase. This sustained high trust over time reflects a category-based form of trust as cosmopolitan trustors identified with the trustees' culture. This trust developed into a relational form of trust over time, as the subordinate trustor gained greater knowledge and experience with the supervisor trustee. By contrast, the hybrids' and culturally-bounds' role expectations of supervisors (e.g., to act in line with Chinese norms as a benevolent provider of care) were not met and continued to be violated. As a result, these trustors entered into a disillusionment phase in which trust levels steadily declined. These trustors came to realize that their initial presumptive trust based on role expectations of supervisors within their own cultural frame of reference was misplaced, resulting in low levels of role-based and relational trust.
The second bifurcation of trust occurred for hybrids and culturally-bounds right after the trust disillusionment phase and revolved around the psychological mechanism of cultural accommodation. In the early acculturation phase, the increasing trust levels of (largely) hybrid trustors occurred through a process of trust repair facilitated by cultural accommodation: these trustors gradually came to understand and accommodating cultural differences in role expectations (e.g., reduced their expectations of supervisor benevolence, and instead basing trust on the ability and task-oriented aspects of the relationship). Trustors who progressed from surface-level early acculturation to deep-level mature acculturation, consolidated trust further, by basing trust not only on knowing and understanding the trustee's cultural norms and role expectations, but also on renegotiating their own cultural identity. These trustors went beyond merely accepting and adapting to the supervisor trustees' culture and role expectations and started to identify with and adopt the culture.
In contrast, the (mainly) culturally-bound trustors, who understood that their role expectations of supervisors were not going to be met, were simply unwilling or unable to culturally adapt and reconsider and repair their loss of role-based and relational trust, resulting in a separation phase in which trust levels always remained very low. Hence, we noticed that from the first to the second bifurcation, trustors changed their trusting behavior from an emotional reaction to a violation of their own relationship-based role expectations to a more experience-based and reflected personal trust choice that was intricately entwined with their acculturation.
We identified several contextual moderators that influenced trust during the acculturation phases. In the early acculturation phase, three moderators influenced trust repair: cultural accommodation by the supervisor, i.e., whether the supervisor trustee attempted to adapt towards the trustor's culture; country context, i.e., trust development occurring in the HQ country where subordinate trustors were on foreign turf facilitated trust repair compared to when trustors were local employees in the subsidiary country on their home turf; and finally trusted third party of trustor nationality, i.e., compatriots of the trustor who facilitated the trust repair process. In the mature acculturation phase, the following two moderators influenced trust consolidation: trusted third party of trustee nationality, i.e., in our case, a trusted German who assisted the Chinese trustor to understand more deeply the trustee's culture; and time spent in the foreign environment, i.e., the time period the trustor had spent in the trustee country, which in turn influenced socialization in this culture and renegotiation of one's own cultural identity.

DISCUSSION AND CONCLUSION
In this study, we engaged in inductive theory building to develop the first comprehensive conceptual model of trust development in inter-cultural hierarchical relationships. Our model advances theoretical understanding by showing how trust development in inter-cultural relationships between (Chinese) subordinates and (German) supervisors is the result of a combination of six distinct elements. Specifically, we find the cultural profile of the trustor (e.g., cosmopolitan, culturally-bound, hybrid), the phase of the relationship (e.g., initial contact, trust continuation, trust disillusionment, separation and maturation), the psychological mechanisms operating within the trustor (e.g., role expectations and cultural accommodation), and contextual moderators (e.g., country context, time spent in foreign culture, and thirdparty influencers) combine to influence the trust dynamics (e.g., trust decline and repair) and trust forms (e.g., presumptive trust, relational trust, category-based trust) that develop in the relationship over time, and ultimately whether the relationship endures or separates. cultural trust development, as well as more broadly to inter-cultural management research. First, our model is based on the view -informed by our data -that trust in inter-cultural hierarchical relationships operates as a variform universal (Ferrin & Gillespie, 2010) and follows, in more general terms, the combined universalistic-particularistic paradigm in cross-cultural management (Pudelko, 2006). Specifically, we assume the cultural profiles to be more generalizable etic concepts, in that people can be predominately bound to one culture (culturally-bound), a mix of cultures (hybrid) or fluid across cultures (cosmopolitan). A few cross-cultural management scholars have already identified the importance of cultural profiles for cross-cultural interactions (Caprar, 2011;Yagi & Kleinberg, 2011). While the existence of such cultural profiles might be an etic phenomenon, the 'cultural content', that make up these profiles, will be emic, that is unique to the specific inter-cultural context. Similarly, we view the psychological mechanisms of expectation fulfilment and cultural accommodation as etic. In the trust literature, expectancy fulfillment supports trust (Mayer et al., 1995), whereas expectancy violation is recognized as a fundamental determinant of trust breakdown (Burgoon & Le Poire, 1993;Hornsey, Chapman, Mangan, La Macchia & Gillespie, 2020).
Similarly, cultural accommodation and acculturation is widely understood to facilitate inter-cultural relationships (Berry, 1992). However, we see the strong presumptive trust based on the assumption of role expectation fulfillment as an emic manifestation unique to Chinese culture (and other East Asian Confucian cultures), as well as the content of role expectations, which are unique to the trustor's culture. The same applies to the content of the cultural accommodation (i.e., what values, norms and behavior change), which is specific to the trustors' and trustees' cultures. Similarly, we consider cultural sensitivity and openness to experience as etic antecedents to inter-cultural trust formation, whereas the concrete forms these antecedents take are emic. We further view the contextual moderators identified in our model as etic and hence generalizable across contexts. Regardless of the specific cultures represented in the inter-cultural hierarchical relationship, we expect cultural accommodation by the supervisor, working in the foreign country, trusted third parties, and time spent in foreign country to facilitate the subordinate's cultural accommodation and acculturation to the culture of the supervisor, and hence trust of that supervisor. Hence, our findings and conceptual model support the notion that trust operates as a variform universal (Ferrin & Gillespie, 2010), following the combined universalistic-particularistic paradigm in cross-cultural management (Pudelko, 2006): an etic principle with specific emic manifestations across national boundaries, which need to be identified (Zaheer & Zaheer, 2006). Our model further supports, and substantially extends, the staged model of trust development across cultural boundaries, proposed by Dietz et al. (2010: 28-29) in several ways. Our model differs in its aim and ambition in that it is empirically derived from in-depth data on inter-cultural relationships (rather than based on insights from prior literature) and is focused on inter-cultural trust development in hierarchical relationships (rather than generically). Like Dietz and colleagues, our model highlights accommodation and/or acculturation processes as central for ongoing trust, and a lack of cultural insight or inability to reconcile cultural differences as central to trust breakdown and termination of the relationship. However, our model shows that what is central is acceptance and understanding of the other party's culture, rather than acceptance and knowledge of the other person as an individual. Another key difference is Dietz and colleagues assume expectation violation only occurs once the relationship is established and assume the violator of the expectations needs to lead the trust repair effort. In contrast, our model shows violations of expectations can occur early in the relationship, and it is mainly the trustor (not the trustee) who proactively restores trust through processes of cultural accommodation and acculturation. In this way, our model both supports, challenges and extends prior conceptual work on inter-cultural trust development.
Second, our research more broadly advances the emerging inter-cultural trust literature by showing that the cultural profile of the subordinate trustor powerfully influences the trust dynamics and trajectory in hierarchical relationships. Specifically, we find that the cultural profile of the subordinate influences the extent to which their role expectations of their supervisor are met or conversely violated, as well as their subsequent capacity to culturally accommodate to differing role expectations in the hierarchical relationship, with direct implications for trust and ultimately the continuation or dissolution of the relationship.
The importance of this insight is underscored by the fact that inter-cultural trust research to date has not identified the cultural profile of trustors as important for trust development. Our finding connects to and advances existing literature on cultural profiles in cross-cultural contexts. While crosscultural scholars' approach to cultural profiling focuses on identifying, analyzing and describing sub-group-specific cultural identities (Caprar, 2011;Yagi & Kleinberg, 2011), we extend the concept by showing how cultural profiles influence the process of inter-cultural trust development. With a few exceptions, the inter-cultural trust literature typically assumes that members of the same national culture are homogenous, and hence treat them as an undifferentiated generic group (e.g., Branzei et al., 2007;Muethel & Hoegl, 2012). Our findings challenge this assumption by showing that members of the same national cultural group are far from homogenous in their cultural background, socialization and expectations, but rather represent distinct cultural profiles (cosmopolitans, hybrids and culturally-bounds), which in turn powerfully influence trust development in inter-cultural relationships. However, these cultural profiles are not completely deterministic of trust development: trustors still retain agency in how they respond, and the extent to which they choose to engage in cultural adaptation and acculturation processes. Third, our model challenges an implicit assumption of seminal models of inter-cultural trust development that assume that cultural differences between trusting parties result in low levels of initial trust. Scholars argued that trust in intercultural relationships develops incrementally as a function of interaction and relational understanding between parties over time. Our findings challenge this view, highlighting that subordinates can enter the relationship with strong presumptive trust based on heavy socialization of the role expectations of supervisors. Indeed, we were astonished to observe the uniformity of high presumptive trust in the initial phase of the relationships across most subordinates. These results contradict previous assumptions that 'trust begins at zero' as suggested by previous trust models (Lewicki & Bunker, 1995a, ba, b, a, ba, b, 1996Lewicki et al., 2006;Shapiro et al., 1992).
We explain these results with the cultural particularities in China, particularly Confucianism, role expectation and the need for harmony (Chen, Chen, & Huang, 2013;Tan & Chee, 2005). We thereby do not explicate differing patterns of trust development with generic Western concepts, but instead with concepts which are specific to the Chinese cultural context. In doing so, we addressed Barkema, Chen, George, Luo and Tsui's (2015: 460) critique that ''our knowledge about management and organizations in the East remains relatively limited or colorized with a Western lens''. Our findings confirm that trust building is deeply embedded in the respective cultural context, and hence requires a context sensitive approach (Kriz & Keating, 2010). Our findings suggest the cultural context and background of trustors and trustee is an important boundary condition of Westernbased mono-cultural theories of trust development.
Fourth, we find that trust forms are also culturallybound and heavily intertwined with the aforementioned trust dynamics. To our knowledge, there has not been research on the role of presumptive trust in an inter-cultural context and how presumptive trust based on roles can be so mis-calibrated in hierarchical relationships. Our findings support the concept and dynamics of presumptive trust proposed by Kramer and colleagues (Kramer, 1996(Kramer, , 1999Kramer & Lewicki, 2010) with high initial trust that is fragile and gradually erodes if presumptive expectations and norms are violated (Meyerson et al., 1996). Our findings suggest collectivists (e.g., Chinese) may be more heavily influenced by presumptive trust towards their supervisors than individualist (e.g., Western) supervisors are in their relationship with their subordinates. Relatedly, by evidencing how powerful presumptive trust can be early in the formation of inter-cultural hierarchical relationships, our research challenges the influence of propensity to trust as an individual difference variable early in relationships. Propensity to trust is integral to many dominant and seminal Western-based models of trust development (Mayer et al., 1995;Schoorman et al., 2007), based on the assumption that trust is typically built over time through relational mechanisms. Yet our findings suggest disposition to trust is less significant in Chinese contexts where initial trust is presumptive and based on role and group membership. Future research is warranted to examine how disposition to trust and presumptive trust play out in different collectivist-individualist inter-cultural relationships.
Fifth, our findings extend understanding of the role of the subordinate trustor and the supervisor trustee in inter-cultural hierarchical relationships. The growing body of research on trust breakdown and repair focuses either heavily on the role of the trustee, and what they can do to restore trust (Ren & Gray, 2009;Tomlinson & Mayer, 2009) or theorizes trust repair as a process which ''involves the interaction of both the trustor and trustee as they attempt to resolve discrepancies in their beliefs'' (Kim, Dirks, & Cooper, 2009: 404). Our findings and model directly challenges this focus on the trustee and instead suggests that in the context of inter-cultural relationships, subordinate trustors play the central role in trust restoration through processes of cultural accommodation and acculturation. In contrast, supervisors had only one significant role in the trust development process: whether or not they behaved in line with their subordinate's role-based expectations.
Our model highlights that trust development in inter-cultural hierarchical contexts are rather onesided, as it is mainly subordinate-trustor driven: trust facilitation mainly depends on the successful inter-cultural sense-making process by the subordinates. It is mainly up to the trustors how they psychologically cope with the violation of (culturally-bound) role expectations, and whether they are willing and able to reframe the violated expectations to understand it as normative behavior in the cultural context of their supervisor (rather than a lack of benevolence), and thereby rebuild trust.
Finally, our findings highlight the central role of expectations and cultural accommodation for inter-cultural trust development. We join other trust researchers by highlighting that expectations are fundamental to trust and often are implicit and culturally influenced (Doney et al., 1998;Mayer et al., 1995). This highlights the importance of making role expectations explicit before they are (often unwittingly) being violated. While our model highlights that Chinese subordinates' trust behavior reflected their socialized understandings, it also highlights that over time Chinese subordinates can and do reflect upon the appropriateness of their socialized understandings, reacting to new information and experiences and, often succeeding in establishing a revised understanding and accompanying trust routines. This revision process is based on cultural adaptation processes through phases of acculturation. This stands in contrast to most cross-cultural trust research, which focuses on how different cultural values lead to different expectations of trustworthy behavior, jeopardizing a trusting relationship between people of different cultural background (Doney et al., 1998;Wasti & Tan, 2010). However, cultural values and resulting trusting behaviors are not static or immutable constructs (Leung, Bhagat, et al., 2005;Leung, Lai, et al., 2005). The human mind is sensitive to environmental influences (Yagi & Kleinberg, 2011) and individuals can culturally adapt and reassess their values. Importantly, our model highlights how subordinates are able to culturally retool themselves (Molinsky, 2013) by engaging in a cultural adaptation process that helps them to manage internal conflicts and engage in reflective sense-making to deal with the other culture, supporting intercultural trust. Our model highlights that this cultural accommodation process is supported by third parties of both cultures, the country context of the relationship, and time spent in the foreign context. Each factor helps the trustor to understand the culturally embedded nature of role expectations in the hierarchical relationship and importantly shift attributions of the supervisor's behavior so they are not seen as a violation of trust, but rather normative and appropriate in the context of the supervisor's culture.

Managerial Implications
Our study has significant practical implications for Western-East Asian working relationships. First, given the increasing global mobility and multicultural nature of the workforce, it is important to recognize the influence that the cultural profile of employees have on trust development in working relationships. This is very practical as people can be asked about this early in the relationship -and misaligned expectations can be surfaced and discussed before they become a basis for the loss of trust. Second, our results confirm that whereas Westerners tend to focus more on task-oriented aspects of trust, East Asians tend to rely more on relational cues to assess the trustworthiness of the other party (Chen et al., 2014;Tan & Chee, 2005). To reduce misunderstandings and to increase trustworthiness, Westerners can augment the relational cues in their daily routines when working with East Asians, while East Asians can focus more on taskoriented elements. Third, knowledge about the other culture helps not only to understand and appropriately attribute behavior that runs contrary to one's own role expectations, but also to adapt one's own role behavior, leading to a stronger foundation for trust. Both Westerners and East Asians can foster a common understanding of trustworthy behavior by transparently communicating their needs and expectations. Another way to increase an understanding of the other culture and to align expectations is through cross-cultural training preceding cross-cultural collaboration (Black & Mendenhall, 1990;Caligiuri, Phillips, Lazarova, Tarique, & Burgi, 2001). Our results suggest that when starting a new collaboration, it would be helpful for Western managers to ask about and reflect upon the cultural profile of their East Asian subordinates to be aware of role expectations and either adapt their behavior to meet these expectations, or alternatively explain why they do not do so. Fourth, we found that in particular local German supervisors, i.e., those based at headquarters, are more reluctant to adapt culturally. Therefore, not only expatriates and inpatriates should receive cross-cultural training, but also the often-neglected local staff, as they equally collaborate with culturally diverse colleagues (such as inpatriates).

Limitations, Suggestions for Future Research, and Conclusions
While our study makes important contributions, it is not without limitations, which can serve as a springboard for future research. One aspect is the gender imbalance in our sample of supervisors. However, we note that companies in the manufacturing sector commonly have a high ratio of male supervisors. In addition, by far most (German) expatriates on foreign assignments are male (Brookfield Global Relocation Services, 2016). Nevertheless, while we consider the gender imbalance in our data to mirror corporate reality, we acknowledge this is a potential source of bias, and recommend future studies seek to achieve a more equal gender distribution of supervisors in their sampling. In line with the majority of qualitative studies, we investigated the trust development process on the basis of retrospective interviews. While we acknowledge that this may introduce certain biases (e.g., differential memory recall of recent versus distal events), given the rich, specific details discussed by our interviewees, we are confident that the accounts captured the key events, dynamics and processes of the emotionally and practically salient issue of trust in their own supervisor. Nevertheless, we recommend future research examining and testing our inter-cultural trust development model to employ longitudinal designs.
While this study was unique in integrating data from both the subordinates' and supervisors' perspective, we zoomed in to focus exclusively on subordinates' trust of their supervisors. Our study also focused on the context of German supervisors managing Chinese subordinates. In the future, we expect Chinese supervisors will increasingly manage Western subordinates. Examining the extent to which the six elements identified in our trust development model hold in this hierarchically reversed context, as well as for supervisors' trust in their subordinate, is likely to reveal interesting insights and boundary conditions to the model.
Finally, our conceptual model was derived from inter-cultural relationships between Chinese and German employees working in German companies. Future research is required to understand the extent to which our model holds across other conceptually well-justified country combinations, given significant variation in trust expectations might be expected across countries (Harzing, Pudelko, & Reiche, 2016), as well as how the nationality of the corporation may affect cultural adaptation and trust formation in cross-cultural settings.
Our study has developed the first comprehensive conceptual model of trust development in intercultural hierarchical relationships. We hope that our model will guide future research in this domain and that its application will ultimately increase trust building across cultural boundaries.

FUNDING
Open Access funding enabled and organized by Projekt DEAL.

OPEN ACCESS
This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/ by/4.0/.