Language choice at a Chinese subsidiary of a Spanish bank

This study explored language management in Banco Santander in China, a multilingual workplace where two major languages, Spanish and Chinese, are used along with English as a lingua franca. We collected data through conducting interviews with four senior managers in charge of human resources, strategic development, retailing and commercial banks to understand how languages are used and managed in this Spanish bank. Analyses of data revealed consistencies and inconsistencies between employees’ choices of language and beliefs as mediated by relevant social cultural, corporate and linguistic factors. The findings shed light on the complexity of language management at a Spanish bank in China against the backdrop of the changing scenario of foreign direct investment. They highlight the importance for multinational corporations to manage languages in response to multilingual challenges at workplaces in China.


Introduction
Multinational companies (MNCs) have many subsidiaries or branches in different parts of the world where employees from diverse cultural backgrounds using different languages work for shared corporate goals. As a result, emergent challenges such as language barriers and ineffective communication caused by these language barriers (Aichhorn and Puck, 2017;Harzing, Köster, and Magner, 2011;Kingsley, 2013) have been observed, undermining multinational team morale and MNCs' global business performance. MNCs are confronted with growing pressure related to language choice, faced with a decision between the use of English as global business lingua franca or a selection of other languages than English as their 'corporate language', particularly in the relationships between MNC headquarters and multilingual subsidiaries.
Language choice in multilingual workplaces has attracted increasing attention in research on international business (Aichhorn and Puck, 2017), international business communication (Tenzer, Terjesen, and Harzing, 2017), and language management (Lauring and Selmer, 2012). Relevant studies have mainly focused on language standardization (Marschan- Piekkari, Welch, and Welch, 1999) or language diversity (Aichhorn and Puck, 2017) in both internal and external corporate communication. While language standardization emphasizes "efforts by top management to instill a common corporate language to harmonize internal and external communication through general rules and policies" (Piekkari and Tietze, 2011), such a top-down centralized approach has always been criticized for neglecting the contextual, dynamic, and complex nature of language practices to facilitate or impede communication. Recent studies from a micro perspective have pointed to the complex effect of language diversity on communication and relationship building (Aichhorn and Puck, 2017;Maclean, 2006) as a growing body of literature has identified the nature of language barriers and language choices between headquarters and overseas subsidiaries, especially in European contexts (for example, Kingsley, 2013). Yet only a few studies have been done in MNCs in China, particularly in MNCs originally from non-English speaking countries.
To address the aforementioned gaps, this study sets out to identify language choice in a Chinese subsidiary of a Spanish bank, the language barriers and corresponding strategies, and how the interplay at different levels shapes employees' beliefs about and choices of language. This discursive understanding of implicit language management and cross-border acquisition in a Spanish bank will help us to better understand corporate language management in Spanish multinational companies and global project teams in general.

Language Choices in the Multilingual Workplace
Language choice in MNCs has garnered increasing research interest in the literature surrounding language management, business English as lingua Franca and international business (Ehrenreich, 2010;Feely and Harzing, 2003;Harzing and Feely, 2008;Harzing and Pudelko, 2013). Thomas (2008) classified languages used in MNCs' work settings into three groups, parent company language, common corporate language (usually English), and local (foreign) language, which is similar to Harzing and Pudelko's understanding of home, corporate, and host language (2013). Although Thomas (2008) called for more research on the benefits of using the home language, it is still an impediment to subsidiaries to use the parent company language instead of the local language (Feely and Harzing, 2003). Therefore, English as a Business Lingua Franca was encouraged in many MNCs (Kankaanranta and Louhiala-Salminen, 2010;Poncini, 2004;Rogerson-Revell, 2007, 2008 to eliminate language differences for increased communication efficiency. However, communication barriers may exist between the parent company and subsidiaries due to the insufficient use of English in developing countries and the difficulty experienced by expatriates in learning the local language (Charles and Marschan-Piekkari, 2002). However, more recent studies have recognized the positive effects of language diversity (Goodall and Roberts, 2003;Maclean, 2006;Tietze, 2010). Therefore, language choices from the perspectives of language standardization (e.g. English as Business Lingua Franca, corporate, or host language) or language diversity remain an important area for research.
Research into language choices in the workplace is usually associated with the identification of language barriers and corresponding coping strategies. Previous studies have investigated the causes of language barriers and suggested solutions to overcome hindrances (Buckley et al., 2005;Du-Babcock and Babcock, 2007;Janssens and Steyaert, 2014). Harzing and Feely (2008) drew on social identity theory and elaborated on the eight causes underlying the language barrier, including loss of rhetorical skills, face, group identities, incorrectly and negatively attributed motives and actions, code switching, power-authority distortions, parallel information networks, and cognitive schema. Harzing, Köster, and Magner (2011) identified 12 different solutions, from informal day-to-day ones to more structural types, for example, code-switching, language training, and bilingual employees as linking-pins. However, most previous studies drew on observations without solid empirical evidence. Therefore, more studies are needed to generate context-specific understandings of the types of language barriers, solutions, and underlying beliefs in multilingual workplaces.

Language Beliefs and Management
Another research perspective related to language choices in the workplace is language management. Relevant studies have found that language choice and language practices shaped by underlying language beliefs are likely to affect the development of employees, team morale, corporate communication, and long-term business profit (Neeley, 2013;Neeley, Hinds, and Cramton, 2012;Spolsky, 2004Spolsky, , 2009. Therefore, effective language management at MNCs can not only help eliminate communicative barriers and disputes among employees, but also increase working efficiency and promote collaboration (Dhir and Gòkè-Paríolá, 2002;Louhiala-Salminen and Kankaanranta, 2012).
While language management is considered to play a crucial role by scholars, it has been largely neglected by the top management teams in MNCs (Harzing et al., 2011). As an MNC is understood as a multilingual community, it is believed that a MNC's language regime should be designed to comply with its global strategies (Luo and Shenkar, 2006;Wang and Xu, 2017). However, only a few studies have been done on corporate language management in MNCs. Research is needed to address the relationship between language choices and language beliefs from the meso and micro perspectives. Research into beliefs about and choices for language at both corporate and individual levels will reveal how these beliefs and choices have been shaped by the interplay between various factors in different layers of the organization.
Spolsky's theory of language management (2009) provided an interactive and dynamic perspective on the interrelationship between language practices/choices and language beliefs. He distinguished three components of language policy: language beliefs (what people hold to be appropriate conduct), language practices (what people actually do), and language management (explicit and deliberate efforts with authority to regulate practices and /or modify beliefs). He pointed out that '... language practices, beliefs and management are not necessarily congruent. Each may reveal a different language policy. The way people speak, the way they think they should speak, and the way they think other people should speak may regularly differ… Within social groups, it is common to find conflicting beliefs about the value of various language choices. One is therefore faced regularly with the question of which the real language policy is.' (Spolsky, 2004: 217). Kaplan and Baldauf (1997) suggest that the top-down approach refers to the direction of the planned change, associated with actors with significant power. Bottom-up processes of change are often not exactly planned, but nonetheless very efficient (Mortensen, 2014). The topdown and bottom-up perspectives are often connected with macro-and micro-level processes, which relate to the work of corporations and individuals. Shohamy (2006) adopted the top-down and bottom-up approaches to explore the interaction between language practices, beliefs, and management, and postulated a stronger role for micro-level and bottom-up corporate language policy. Whether top-down or bottomup approaches are adopted in an MNC requires more exploration, as well as their fundamental reasons and influential factors at different levels.
A brief review of relevant studies in language management at MNCs also reveals that the majority of studies were conducted in European contexts (for example, Louhiala-Salminen et al., 2005;Sherman, 2009a, 2009b;Kingsley, 2013;Lønsmann, 2017). This can be partly explained by multilingualism in Europe and the prevailing language policy in European multilingual contexts. However, few empirical studies have investigated current corporate language management at MNCs from non-English speaking countries, for example, against the changing social and economic backdrop of China (Cui, 2013). Although the Chinese context has been considered in some studies, for instance, Chinese headquarters' language management of its subsidiaries abroad (Harzing and Pudelko, 2013) and MNCs' management of expatriates in China (Zhang, Harzing, and Fan, 2018), there is a lack of in-depth explorations into the interactions among language beliefs, language practices/choices, and language management in non-English-speaking MNCs in China.
In addition, previous studies have attempt to examine language practices through quantitative research. For example, Harzing and Pudelko (2013) collected large-scale quantitative data from more than 800 subsidiaries in 13 host countries, related to MNCs with headquarters in 25 home countries. They used two dimensions, the global importance of the parent group's local language and the level of English skills of managerial employees, to determine four country clusters: Anglophone, Asian, Continental European, and Nordic MNCs. For example, Spain is one of the Continental European countries, where English skills are comparatively higher than in some Asian countries (for example, China and Japan) but not as high as in the Nordic countries (like Sweden and Denmark). The global importance of the role of local languages, for example Spanish, gradually declines. China is characterized by a relatively low level of English skills and the increasing importance of the role of the Chinese language worldwide.
Relatively few studies have specifically investigated the interaction between a Spanish MNC and its Chinese subsidiaries. It is important to understand more about the beliefs about language, language choice, and language management, which will be achieved by examining different geographical sites of a Spanish bank. Such understanding will enrich language management theories by providing additional evidence about banking industry in China, as well as providing a useful reference for managing languages at MNCs or in global project teams in China in general.
Therefore, this study sets out to investigate employees' language choices and the contextual factors affecting such choices in a Chinese subsidiary of a Spanish bank, Santander. Specifically, two research questions were raised as follows: 1. What are the employees' choices of language in the Chinese subsidiary? 2. What are the employees' beliefs about language and Santander's language management?

Research Design and Setting
As no single theory can explain the complexity of language choices (Aichhorn and Puck, 2017;Kingsley, 2013), we adopted an explorative case-study approach to address our research questions. Our case study was conducted in a Spanish bank's subsidiary in China. The analysis of the interaction between a Spanish MNC and its Chinese subsidiary provides a reasonable extension of previous studies.
The Santander Group, headquartered in Madrid, Spain, is one of the world's largest retail and commercial banks. Santander has a balanced geographical spread. Its ten core markets are Spain, Germany, Poland, Portugal, the United Kingdom, Brazil, Mexico, Chile, Argentina and the United States (Santander, 2018). It also has operations in China through its wholesale and consumer finance business. Concerning its experience and reputation, Santander is one of the world's top ten banking brands and ranks 77th in the Fortune Global 500 list (Fortune, 2018). Santander is a typical example of MNCs whose headquarters are based in countries where the official language is not English and is different from the language of global operations.
To expand its global commercial network and build its presence in Asia, Santander acquired an 8% stake in the Bank of Shanghai (BoS) for 470 million euros (Santander, 2013). BoS, established in 1995 and headquartered in Shanghai, is now the second largest urban commercial and retail bank in China and has the third largest banking network in Shanghai. Since this acquisition Santander has become the second largest shareholder in BoS and its strategic partner. According to the agreement, Santander provides BoS with a team of professionals to transfer its knowledge and experience in risk management and commercial and retail banking. This cross-border acquisition will allow Santander to develop investment banking activities and strengthen its position in China.
Up to 2018, Chinese, Spanish, and English were the top three most spoken languages ranked by the number of native speakers and the total number of speakers worldwide (Ethnologue, 2018). In mainland China, Putonghua, also known as Standard Chinese, is the official language. Regulated by the National Language Regulating Committee, Putonghua is the language chosen for communication in the overwhelming majority of cases. In terms of foreign languages, English has become the most popular foreign language. According to China's national survey in 2006, 93.8% of respondents had studied English, while only 0.05% reported learning Spanish (Wei and Su, 2012). Therefore, Santander and its Shanghai branch are an ideal choice for the purpose of analysing the language choice in a MNC where the parent company's host language is merely one of the 'minor languages' in the home country.

Participants
The primary data for this research consisted of semi-structured interviews with four respondents. The four case informants, with diversified education backgrounds particularly in language education, had different multilingual potentials and capabilities. They have worked in multiple geographical sites in Spain, mainland China and Hong Kong. As shown in Table 1, Kevin, born in Hong Kong, had the chance to exchange in Europe as a college student and obtained his Master's degree in Britain. Li, born in China, joined a double-major undergraduate program studying Business English and International Trade. She also studied for an MBA degree in Hong Kong. Jian, born in China, studied and lived in Britain for about ten years, from high school to his PhD degree. Lucas, born in Spain, studied as a telecommunication engineer in Spain and obtained his Master's degree in finance in Finland. Having worked in the international banking industry for about ten years, they all had experience in various departments and teams, such as retail and commercial banking, global corporate banking, and central services (for example, risk control and product marketing). We expected them to have rich insights and in-depth understanding of language choices and management, and chose them as participants for this reason.

Male
Spanish English 11 (7.5 years in Spain and 3.5 years in Shanghai) 1. Acronyms in brackets refer to the languages a participant is able to use for work purposes. 2. Li had 5 years' working experience in another international bank before she joined Santander. Table 1. Brief information about the four participants.

Data Collection and Analysis
Jian, Li, and Kevin preferred to be interviewed in Putonghua, of which they are native speakers. We also respected their use of English words and expressions from time to time. The interview with Lucas was conducted in English, which was the only language that could be both understood by interviewers and interviewee. In the interviews, to identify the language choices in Santander sample questions included: 'Which languages do you use most in the Spain headquarters and the Shanghai branch?' 'Do you use different languages for written reports, emails, meetings, and informal talk?' Further questions were asked to understand language barriers and solutions, including: 'Do you meet challenges in languages when getting used to the Santander environment, or not?' To address their beliefs about language management, sample questions included: 'Do you think that language management is taken as an important part of corporate strategy in Santander?' 'Do you find that language capability makes a difference in the process of interviewing and promotion, or not?' 'Does Santander provide language training?' The interviews, which lasted 30-50 minutes, were recorded with the interviewees' permission. We asked two research assistants to transcribe the audio files, which yielded a total of 40 single-spaced pages of transcript.
The qualitative research software MAXQDA was used as an analytical tool for categorizing and coding data. The transcripts were analyzed in their original source language and codes were developed in English only for data analysis. First, we read the transcripts to achieve an overview of the themes. Second, the coding categories were defined to address the research questions, including language choice, barriers and corresponding solutions, beliefs, and factors influencing choices and beliefs. Third, we grouped a list of predefined codes, which was developed based on the literature, into our categories. The codes, for example 'the usage of Spanish/English/Chinese', comply with Harzing and Pudelko's catagorizations of home, corporate, and host language (2013). Language strategies, such as 'code switching' and 'language training', were put into the categories classified by Harzing, Köster, and Magner (2011). During coding we also compared interview responses with the predefined codes and developed more invivo codes, for example 'macro/corporate-level', 'meso/team-level', and 'micro/individual level' factors that affect employees' choices and beliefs.

Findings
Our findings will be presented and discussed in the light of existing literature and will address our research questions point by point. Employees' language practices were found and studied in different locations and contexts. To follow up, we analyzed detailed language barriers and relevant solutions. These daily language practices and strategies effectively reflect employees' beliefs on language use and management. This section also sums up the factors contributing to the formation of beliefs.

Language Choices
It was not surprised to find that local languages (for example, Spanish in Madrid and Chinese in Shanghai) were used in daily work, from internal communication between native speakers to external communication with local clients and organizations. Regardless of location, English is used as the working language; this is required by a top-down corporate policy, because no other single language could connect all employees. Employees also have no doubt about accepting English as a 'common corporate language' (Thomas, 2008) to improve work efficiency. As commented: Kevin (CPE): If you work in an international team, the counterpart located in Spain is also very international. Then everyone in the team would be comfortable to use English as the working language. … Many managerial employees, for example Li and the CEO of Asia Pacific, only use English as the working language. (Interview originally in Putonghua) Although Kingsley (2013) found that languages were used differently in written communication such as reports and emails than they were in spoken communication such as presentations and meetings, our interviewees were indifferent to language use in different genres. Communication objectives and the language competence of different parties played a more decisive role in language choice. It was also agreed that using a common corporate language that can be understood by the majority is important to show respect to other people in the workplace. As commented: Jian (PEC): It is rather difficult to answer this question (which language is used more often The interviewees all agreed that English is a must and another language is a plus. Spanish was especially important when an employee was trying to build relationships and 'personal trust' in the working community and with Santander headquarters, especially if the counterpart was not international or proficient in English. As commented: Jian (PEC): Spanish is not always on a must list. But positions will have different requirements. For example, if you work in a risk department, you would frequently need to contact the headquarters. Therefore, the ability to use Spanish as a working language will be a huge advantage. If your work is about internal communication, speaking a few languages will be a great help. (Interview originally in Putonghua) Similarly, in the process of strategic cooperation with BoS, the relationship plays an important role. Using Chinese shows a non-native speaker's commitment to building relationship with their local counterparts. As commented: Li (PEC): When you speak some Chinese, it is the way to show your commitment. … (The language courses) may also help (expatriates) live better in Shanghai and build and maintain relationships with your (local) counterpart. (Interview originally in Putonghua) As part of language use and practices, we also found that employees encountered language barriers in the work setting. Their solutions for overcoming these barriers were therefore investigated. In Spain, non-Spanish speakers may have found it difficult to understand corporate information from the official website, documents or emails written in Spanish. They tended to be tolerant of this situation as 'Santander is after all a Spanish bank'. The job nature and the counterparts' language competence may decide how much Spanish would be used. For example: Li (PEC): For example, at the headquarters, people from the risk department are almost all Spanish, without overseas education or work experience. To communicate with them efficiently, using Spanish is the only choice. Besides, many internal emails and emails with Latin American counterparts are written in Spanish. (Interview originally in Putonghua) In these occasions, non-Spanish employees would seek help from Spanish-speaking teammates or online translation tools, for example Google translate, as an interviewee comments: Li (PEC): My ex-supervisor was not a native Spanish or English speaker. Every time before he sent out a very important email in Spanish or English, he would ask a native to check his grammar and wording. (Interview originally in Putonghua) In addition, misunderstandings between non-native language users may be due to language competence, characteristics or a different understanding of jargon in the banking sector. Employees may use code-switching to confirm accuracy in communication. For example: Jian (PEC): The word 'billion' means 1,000,000,000 (十亿, shí yì in Chinese pinyin) in English, but 'billon' means 1,000,000,000,000 ( 万亿, wàn yì in Chinese pinyin) in Spanish. Every time this word is mentioned in talk, we may pause and switch language, for example from Spanish to English, to confirm if we mean the same thing. (Interview originally in Putonghua) After acquiring shares in BoS, Santander decided to build and send a team to guarantee cooperation and share management expertise with BoS. This team intended to work on consultancy and knowledge transfer. However, Santander's experience in Asian markets was limited. As commented: Li (PEC): At that time, almost half of the (Santander) team were non-Chinesespeaking experts. However, (in China) it is common to use Chinese when communicating with BoS, local authorities and other corporate clients. Only a limited number of local managers can speak English very well. (Interview originally in Putonghua) Furthermore, the responsibility of the local branch also has an impact on language practices. As commented: Kevin (CPE): Currently, the main responsibility for the Shanghai team is 'transformation', different from traditional international business, which requires integrated understanding of fintech innovation, retailing and best practices. These jobs relate to (reading and speaking) multiple languages. (Interview originally in Putonghua) Therefore, to overcome this communication barrier, Santander decided that every non-Chinese-speaking expert would be supported by one Chinese-speaking assistant, helping with internal coordination and providing translation in external communication. As commented: Li (PEC): I have been working as an assistant to the Deputy CEO of Asia Pacific (who did not speak Chinese), coordinating and guaranteeing cooperation between Santander and BoS, and then corporate development, finding new investment opportunities in China. (Interview originally in Putonghua) In addition, when working in Shanghai, non-Chinese speakers confront language barriers due to cultural divergence and communicate patterns. To overcome these barriers, they need to spend extra time outside work to learn a new language. As commented: Lucas (SE): My major responsibility is to identify business opportunity for Santander here in China. I have to say that at the beginning it was quite difficult because I didn't speak the language and the culture is completely different to the Western world. … I struggled a bit. It's not easy, to be honest. I would recommend to anyone to come to China, but it's not easy. You need to know that people are completely different, the way they act, the way they negotiate, the language skills are completely different, even though some of the guys in BoS can speak English, but their accent, the way they express themselves is different. … So during the first one and a half years I studied a lot of Chinese. So I used to wake up around six to study before coming to work and to study after leaving the office in the night. So … I could communicate a little bit with most of the people, with the driver, with the restaurant, with everyone (if they use Putonghua).
(Interview originally in English) After analysis, in short, various solutions to language barriers were found in communication at different levels. In daily communication, employees as individuals reported using machine translation and code-switching, for example the use of 'billion/billón'. At the team level, expatriates may be sent to a local branch for knowledge transfer, while bilingual/multilingual employees work as linking-pins between expatriates and local counterparts. At the corporate or organizational level, Santander provides extensive language training courses, and English is implicitly suggested by interviewees as a common corporate language-for example, the interviewees felt comfortable to use English as the working language.

Language Beliefs and Language Management
Implicitness is a typical feature of Santander's language management. Although no explicit language policy may be found in areas such as recruitment, training, or promotion, employees have formed their own beliefs in this international working setting.
The interviewees all mentioned Santander's provision of language courses for non-native speakers-for example, Spanish courses in Spain and Putonghua courses in Shanghai. They had all learnt, or tried to learn, a second foreign language for varying amounts of time. Employees have freedom to make their own decision on when and how frequently to take language courses. As commented: Li (PEC): Similar to Spanish courses in Madrid, the bank does not pressure or track employees' improvement in language studies. It would be employees' personal choice to take courses and exams. (Interview originally in Putonghua) Although Santander may have different expectations of employees' improvement in language skills, it is generally agreed that this second language provision constituted encouragement by the bank, not pressure or obligations on learners. As commented: Jian (PEC): At that time we were asked to take exams, as we took language courses frequently and regularly. I achieved level 4, while others achieved level 6, 5 or even 2 (where 1 is the lowest level). But the bank had no fixed criteria for evaluation. It was set as a goal and an encouragement instead of a compulsion. (Interview originally in Putonghua) Currently Santander has no document that explicitly identifies its language management approach. If this was needed, interviewees assume that a top-down approach would work and would be integrated into the corporate culture. For example: Jian (PEC): It would be better for the bank to adopt a top-down approach (than a bottom-up one), as everyone is busy with his/her job and lack of action. So something core, for example languages … can be regarded as a part of corporate culture. Employees' language competence is also a reflection of corporate culture. So language courses can be a useful way to build up and promote corporate culture. … But for sure the company needs to consider costeffectiveness. (Interview originally in Putonghua) In terms of job interviews, employees' career paths, and promotion, there is agreement that the ability to use English is crucial. As commented: Lucas (SE): If you don't speak English, (it's) very difficult to be promoted beyond a certain level, because … one third of our revenues comes from countries where Spanish is not their language. And you need to interact with certain people in different countries. So if you don't speak English or you only speak Spanish, you may have an issue in being promoted. (Interview originally in English) However, it may not be feasible to set other languages, for example Spanish, as a mandate, because this would dramatically narrow down the field of potential candidates. As commented: Kevin (CPE): In job interviews, English is certainly mandatory. It is important for a candidate to communicate naturally in English. International exposure, for example overseas experience, would help a lot. (But) it's difficult for an employer to set (other) foreign languages as a mandatory requirement. If you did so, hardly anyone would apply (for this position). (Interview originally in Putonghua) Accordingly, employees' language beliefs are influenced by the external (outside the corporation) and internal (within the corporation) environment. The external environment is primarily composed of customers and regulatory authorities. Li (PEC): He (Jian) is responsible for corporate clients in our cooperation with BoS, so he has no language barrier in communicating with local and foreign companies in Shanghai. I once contacted Shanghai Financial Authority, banking regulators, news media and other cooperating third parties. (Interview originally in Putonghua) The internal factors can be further categorized into three levels: the macro/corporate level, the meso/team level, and the micro/individual level. Within the bank, macro/corporate-level communication is needed between headquarters, subsidiaries or branches, departments, and teams. Although employees share an understanding of English as a lingua franca in this international bank, other languages are likely to be used in certain settings. For example, different languages may be used when employees contact the headquarters and local colleagues. As commented: Jian (PEC): When I contact headquarters, I use Spanish more (than English), especially if he knows you (I) can speak Spanish. … When I work with Chinese colleagues, it's unnecessary for us to speak English, even if we all can. (Interview originally in Putonghua) Teamwork is encouraged in modern corporations, in order to create and transfer knowledge. Therefore, a team leader's opinion on how to improve team communication shapes team organization and language choice. Meso/team-level factors were as follows: Li (PEC): (In Spain) I was in a global financing team. To nurture his team, our team leader intentionally chose and recruited people from different countries, asked us to speak English in meetings, and made English the working language within the team. (At that time) our counterparts were in non-Spanish speaking countries, like Britain, America, and Asian regions. (Interview originally in Putonghua) As an international company, the language choice in different departments or teams may be determined by the most commonly understood language. As commented: Lucas (SE): (In Spain) I was within asset management … so my responsibility was to elaborate and manage certain funds. … Most of my peers … were Spanish people. And in Europe, even with the United Kingdom unit, we have a huge Spanish committee. So I talked to them in Spanish mostly. (Interview originally in English) Micro/individual level factors also influence language choices and beliefs. During our observations, these four interviewees were open-minded and cooperative with the multilingual workplace. It does not seem that they were perplexed or felt competitive due to language incompetency or barriers. The first reason for this may be their overseas education experience. As they had been exposed to an international education environment over long periods of time, they had become used to multilingual contexts. Instead of feeling pressured or incompetent, they had learned to adapt to and cooperate with the environment.
An individual's learning motive also influences their choice of whether to learn a new foreign language. Although all interviewees believed that the ability to use a second foreign language was a 'plus' in the workplace, they might also choose to give up learning a language due to heavy workload, lack of time, and difficulty in breaking through a bottleneck. As commented: Kevin (

Discussion
This research intended to investigate the language choices in a multilingual work setting and the reasons for these. English is adopted as a common corporate language to improve work efficiency. In terms of the interaction between the parent group and subsidiaries, Santander's expatriates in key positions are accompanied by assistants who are native speakers of the local language. Santander also places emphasis on job candidates' English skills and provides language training courses. The language courses imply that Santander understands both the importance of learning the local language and the difficulty for expatriates of mastering a new foreign language. These measures, in accordance with Harzing, Köster, and Magner (2011), have proved to be pragmatic in dealing with macro-, meso-, and micro-level communication barriers between the parent company and subsidiaries. This study also identifies employees' beliefs about language management and language choice in Santander Bank. From the employees' accounts and our document analysis, there is no explicit language policy at the corporate level. Santander's corporate language management involves emphasizing English as a lingua franca in the work setting, allowing native speakers to use their mother tongue, and encouraging expatriates to learn the local language in non-Englishspeaking countries.
To extend the theory of corporate language management, factors at different levels need further study. Language choice is influenced by external and internal factors. The location of a MNC's branches or subsidiaries defines the language it uses to serve local clients, report to authorities, and contact news agencies, especially in a country where English is not used as an official or functional language. To further identify the interplay in the internal environment, factors can be further divided into three levels: the macro-, meso-, and micro-levels. The corporate environment comprises macro-level factors, including job responsibility and counterparts. As required by modern corporate structure, employees always work in a team or department, which forms a smaller working community influenced by team leaders and context. Micro-level factors relate to individual experience and actionfor example, individual education experience and learning motive.
The importance of efficiency in the global work setting has been mentioned by Luo and Shenkar (2006), Barner-Rasmussen andAarnio (2011), van Mulken andHendriks (2015) and Barfod (2018). However, it has seldom been studied thoroughly or given adequate attention. This research underlines efficiency as an important factor in language choice. Efficiency can be considered as the ratio of language output to the sum of the users' efforts to surmount barriers in a communication context. The higher the language competence the user has, the less effort the user needs to overcome barriers. A language which requires a low ratio of language output to effort made by users is preferred for work-related or informal talk between employees. For example, provided that each employee is able to use English, either Spanish or Chinese may be used between two employees as it will be the mother tongue for at least one of them. Among three or more employees who have various mother tongues, English is quite often more efficient to connect all users. Similarly, formal documents are written in English to guarantee the widest acceptance and understanding within the MNC.
This research shows that language choice does not just facilitate communication at different levels; it also supports knowledge transfer, and the development of respect and trust with local Chinese counterparts and clients. Language choice is more likely to be determined by the nature of the international team, for example, the knowledge and managerial expertise transfer and the nature of the consultancy work. It should be noted that the employees' beliefs about and practices of language choice were greatly influenced by their educational and prior work experiences in multicultural and multilingual contexts, and their career goals for global mobility.
Unlike previous studies focusing on language choices and language barriers, this study, through a micro lens, unravels the contextual and individual factors that shape language choices and beliefs about language choice. The global mindset suggesting the proficient use of English and the additional use of Spanish and/or Chinese, which has been cultivated by multicultural and multilingual education or shaped by implicit language management in Santander, may help employees to embrace a flexible and quite effective approach to coping with language barriers at work. Such findings might be useful for a MNC to take into consideration a similar bottom-up approach when drafting corporate language management policies, as they might initiate individual agency to improve communication in a similar multilingual working environment.

Conclusion
This study aims to provide additional evidence of positive aspects of language diversity, implicit corporate language policy, and language for knowledge transfer in a globalized world. This research contributes to the existing literature in its study of Santander's laissez-faire corporate language management and the underlying factors at different levels. The study identifies employees' beliefs about language choices, and their practical language choice as a result of the interplay of social, cultural, institutional, team, and individual factors. More data from other sources will be valuable to triangulate and develop a more in-depth understanding. A longitudinal study could investigate how corporate language policy might change and be influenced. It is also crucial to address more individual voices, reflecting beliefs and their effects on language management. Investigating more MNCs from different sectors will help to understand the language choice of MNCs in the Chinese work setting.