跨國建築公司翻譯實務:中國子公司資源運用

全文

(1)Advisor: Dr. Tze-wei Chen. Translation in a Multinational Architecture Firm Resource Allocation in a Chinese Subsidiary :. By Yu-ping Wang. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts of Graduate Institute of Translation and Interpretation, National Taiwan Normal University July, 2015.

(2) Acknowledgements First of all, this long overdue thesis would not have been possible without Professor Tze-wei Chen. I’d like to thank him for his words of encouragement, insight, and timely kicks in the behind.. I would also like to thank Gensler for its commitment to research and collaboration, which was translated into incredible support from my colleagues, near and far, throughout this arduous process.. Big thanks to my incredible friends, Clifford Champion, Jan Marek, and Christina Adamidou, and Mark Blankestein, for moral and technical support.. Saving the best for last: I would like to dedicate this thesis to my late father, Haikang “Eric” Wang. He was the ultimate linguistic nerd, and I his best project. Thank you..

(3) Abstract. Language in the international business context is a topic that's been steadily gaining attention, but in-depth understanding of how multinational corporations handle day-to-day translation tasks is still eluding researchers. This thesis aims to conduct an investigation into the allocation of resources for translation to better understand demands and current practices. Through personal interviews and an online survey at a Chinese subsidiary of a multinational architecture firm, the results showed that most employees have translated content themselves and considered translation an unspoken part of their jobs. The majority of the employees are wary of translation quality and review outsourced translations using machine translation tools. Suggested next steps include developing corporate glossary and translation memory to maximize the existing linguistic assets and improve translation speed and accuracy.. Keywords: translation, multinational corporations, outsourcing.

(4)

(5) Table of Contents. Chapter 1 Introduction................................................................................................ 1 1.1 Background of the study ............................................................................... 1 1.2 The Study and Thesis Organization ............................................................. 4 1.2.1 About Gensler ....................................................................................... 4 1.2.2 The Subsidiary ...................................................................................... 7 1.2.3 Localization Efforts .............................................................................. 8 Chapter 2 Literature Review .................................................................................... 11 2.1 Language in Multinational Corporations .................................................. 11 2.4.1 Lingua Franca ..................................................................................... 13 2.4.2 Corporate Language ............................................................................ 15 2.4.3 Multilingualism ................................................................................... 18 2.2 Translation and Language Problems in Management ............................. 19 2.3 Language as Asset ........................................................................................ 20 Chapter 3 Research Methodology ............................................................................ 25 3.1 Research Methods ........................................................................................ 25 3.1.1 Pre-Survey Interview .......................................................................... 25 3.1.2 Online Survey ..................................................................................... 29 3.1.3 Participants .......................................................................................... 33 Chapter 4 Findings .................................................................................................... 34.

(6) 4.1 Survey Results .............................................................................................. 34 4.1.1 Linguistic Assets ................................................................................. 36 4.1.2 Content and Assignment ..................................................................... 42 4.1.3 Resource Allocation ............................................................................ 56 4.1.4 Aspects of Translation ........................................................................ 62 4.2 Summary of Results ..................................................................................... 72 Chapter 5 Discussions ................................................................................................ 75 5.1 Discussion of Survey Results ....................................................................... 75 Chapter 6 Conclusions............................................................................................... 77 6.1 Conclusion .................................................................................................... 77 6.2 Limitations of the Current Study ............................................................... 80 References ................................................................................................................... 82. Appendix ..................................................................................................................... 90 Appendix i: Survey Questions........................................................................... 90 Appendix ii: Survey Results ............................................................................ 123 Appendix iii: Interview Questions .................................................................. 126 Appendix iv: Call to Action Message ............................................................. 128.

(7) List of Tables Table 1. Type of Design Work ................................................................... 6 Table 2.. Interviewees ................................................................................ 27. Table 3. Job Roles ..................................................................................... 35 Table 4. Native Language ......................................................................... 37 Table 5. English Fluency .......................................................................... 39 Table 6. Mandarin Fluency .......................................................................... 40 Table 7. Types of Content Translated - General ....................................... 44 Table 8. Translation Assignment .............................................................. 45 Table 9. Native Language vs. Translation Coordination .......................... 45 Table 10. Types of Content - Outsourced ................................................. 48 Table 11. Reasons for Outsourcing ........................................................... 49 Table 12. Translation Agencies ................................................................ 51 Table 13. Types of Content - Self Translated ........................................... 52 Table 14. Role vs. Translation Demand.................................................... 54 Table 15. Cost - Outsourced Translation ..................................................... 58 Table 16. Cost - Self Translation ................................................................. 60 Table 17. Cost - Peer Translation ............................................................. 61 Table 18. Translation Quality ................................................................... 62 Table 19. Most Useful Translation Resource ........................................... 63 Table 20. Review - Outsourced Translation ............................................. 65 Table 21. Review - Self Translation ......................................................... 66.

(8) Table 22. Review - Peer Translation......................................................... 66 Table 23. Review Tools - Outsourced Translation ................................... 68 Table 24. Tools - Self/Peer Translation .................................................... 69.

(9) Chapter 1 Introduction. 1.1 Background of the study. Globalization and technological advancements have transformed business communication in many ways. Information exchange happens faster, and virtual collaboration has become commonplace (Klitmøller & Lauring, 2013). But while communication can be instantaneous, effectiveness isn’t guaranteed. Language barrier can still pose a problem for anyone intending to communicate to a foreign client, a team member in a remote location, or to an entirely new market (A.-W. Harzing & Feely, 2008; Tenzer, Pudelko, & Harzing, 2013); it can also prevent effective communication within multinational companies (Welch, Welch, & Piekkari, 2005).. The challenges of language barrier and growing linguistic diversity in international business have been gaining visibility in management literature. Most of the focus have been on how multinational corporations manage language barriers, through language training (Charles & Marschan-Piekkari, 2002; Anne Wil Harzing, Köster, & Magner, 2011; Leeds‐Hurwitz, 1990), expatriate assignments (Du-Babcock & Babcock, 1996; Anne W. Harzing, 2001a; Nurmi, 1995), corporate language policies (Dhir & Gòkè-Parí olá, 2002; R Fredriksson, 2006; van den Born & Peltokorpi, 2010), or cross-cultural communication training (Cowley & Hanna, 2005; 1.

(10) Spinks & Wells, 1997). Janssens, Lambert, & Steyaert (2004) invoked the importance of translation studies in helping multinationals establish effective language strategies over ten years ago. They defined language strategy as multi-faceted: the choice of language, or languages spoken and urged translation experts to turn their attention to corporate communication. But few researchers have examined the role of translation within the international business context.. Piekkari, Welch, Welch, Peltonen, and Vesa (2013) explored translation activities within Nordea, a Nordic bank with multilingual translation demands, identifying the various methods employees used to complete these tasks. They found that employees working in this multinational firm encountered growing translation needs on a daily basis. Burdened with the responsibility of ensuring effective communication without the background training or tools, they resorted to their personal network, within and outside the company, to accomplish the job. Piekkari et al. found that meeting translation demand in a timely matter became part of the corporation’s competitive advantage.. This thesis expands on Piekkari et al. (2013) and their theory of language absorptiveness capacity, establishing multilingual staff as linguistic assets. By gathering information on the day-to-day translation activities at Gensler Shanghai, a 2.

(11) subsidiary of a multinational architecture firm, we see if the employees approach translation requests in a similar to those at Nordea, and gain a thorough and realistic understanding of the translation demands within a multilingual workplace. Through interviews with key individuals and an online office-wide online survey, the thesis will assess the linguistic assets within the office, how these resources are allocated, how employees process translation demands, types of content translated, amount of work that occurs, and how translation quality is controlled. This thesis will further propose that translation should be an accepted and expected part of a multilingual workplace, and takes another step towards making the “invisible” translation efforts visible (Bielsa, 2005),. The data collected will provide some insight into the linguistic assets within Gensler’s Chinese subsidiary, help the firm make informed decisions on effective linguistic resource allocation, and identify resources the firm may need to invest in to meet their translation demands (Tsedal Neeley & Kaplan, 2014). Provided with the appropriate tools and resources, the employees can efficiently and accurately perform or oversee the translation tasks, improving communication and delivery quality without jeopardizing their core services.. 3.

(12) In agreement with Piekkari et al. (2013), the researcher suggest that Gensler invest in tools that will capture and protect the language capacity within the office. By setting up centralized corporate glossaries and developing translation memory, the firm will benefit from the local linguistic assets while improving the quality of its service delivery.. 1.2 The Study and Thesis Organization. The following section will introduce the firm and the subsidiary that is the subject of this research, and provide some background on the firm’s localization efforts thus far.. 1.2.1 About Gensler. Gensler is a privately-owned American architecture and design firm, founded in San Francisco in 1965 by Arthur Gensler, Drue Gensler, and James Follett. At the time of writing, Gensler had offices in 46 cities in 16 countries around the world. Nine of these 16 countries were located in the Asia-Pacific region, with Shanghai as the regional hub. The multinational design firm employed over 4,500 people and had more than 3,500 clients from various industries.. 4.

(13) The firm’s culture puts heavy emphasis on collaboration, and operates under the “One-Firm Firm” motto, a management model popular amongst the world’s leading professional service firms, such as McKinsey and Lathem & Watkins (Maister, 1985). In an industry where competition is fierce and developers fight to work with “starchitects,” or celebrity architects, Gensler focuses on building collaborative teams, cultivating long-term client relationships, and growing and training talents internally.. When it comes to global expansions, architecture firms generally have two different approaches: establish sales and marketing offices abroad and produce all design work back in headquarters, or through merging with or acquiring local design studios. In contrast, Gensler’s expansion into China began with an office in Hong Kong in 1991, to serve clients who had expanded into the region, true to the One-Firm Firm management approach. The Shanghai office was officially established in 2003, led by a Shanghai native architect who had been with the Gensler Denver office for seven years prior. This organic expansion approach means that most Gensler offices have leadership who are well-versed in the Gensler culture. By embedding long-time Gensler leaders in its subsidiaries, the Gensler’s corporate culture has been transplanted in the remote offices (Anne W. Harzing, 2001b), with the expected cultural differences from region to region.. 5.

(14) Gensler’s design expertise is categorized by client industry and by project type, commonly referred to as practice areas. As of July 2015, the firm has 31 practice areas (see Table 1 below), with professionals designing everything from desk lamps (Product Design) to entire cities (Master Planning). With a global network of design professionals, clients in different regions could benefit from a wide scope of expertise and talent. The offices are staffed with architects and designers specialized in various practice areas, but if a client approached with a special project that requires a unique expertise, it was common practice for the firm to staff the project with designers from another office who are experts in the project type, so collaboration with offices in other regions are quite common.. Table 1. Type of Design Work. Workplace. Lifestyle. Community. Commercial Office. Brand Design. Arts & Culture. Entertainment. Aviation &. Building Developers Consulting. Transportation Consumer Products. Hospitality. Convention Centers. Corporate Campuses. Mixed-Use. Education. Defense & Aerospace. Residential. Health & Wellness 6.

(15) Energy. Retail. Mission Critical Facilities. Financial Service Firms. Retail Centers. Not-for-Profit. Government. Sports. Planning & Urban Design. Landlord Services. Tall Building. Life Sciences Media Product Design Professional Service Firms Technology. 1.2.2 The Subsidiary. The Gensler Shanghai office had 216 staff members at the time of writing. Fifty of the employees were internationals from a diverse range of nationalities and cultural backgrounds. Most of the international employees are locally hired, although a few had been transferred from other developed Gensler offices. The office employed staff members in all functions, and served as a fully-fledged hub to the other 8 offices in the Asia-Pacific Region: Bangalore, Bangkok, Beijing, Hong Kong, Seoul, Singapore, Sydney, and Tokyo. The administrative staff in Shanghai, including IT, Finance, Accounting, Risk Management, and Marketing provided support to all 9 regional offices.. 7.

(16) Gensler Shanghai is a wholly foreign-owned enterprise, with a license to act as architecture and design consultants in China. Designers at Gensler work with local design institute like Tongji Architectural Design and Research Institute. The local design institutes are responsible for ensuring that the design work is compliant with local building codes, providing necessary construction documentation, and submitting drawings for government approval. Many local design institute members are stationed in the Shanghai office for ease of collaboration.. 1.2.3 Localization Efforts. The firm had taken initial steps to produce localized content, with Simplified Chinese being the front-running language. The Firmwide Communications studio, an in-house editorial and publication team, produced the company annual report and trend forecasting magazine in Simplified Chinese in March 2015, and the company blog had featured posts translated into Simplified Chinese. These blog posts mainly featured the Shanghai Tower, a megatall building and a milestone project for the firm. The company’s official web page has been undergoing evaluation for localization, with Simplified Chinese and Japanese as the two priority languages for translation, Costa Rican Spanish a close third.. 8.

(17) The company’s executive broadcast video, a biannual event where the Co-CEOs report the firm’s achievements and future ambitions to its employees, also featured Simplified Chinese subtitles since 2014.While most translation efforts had been met with positive response, feedback were mixed regarding the subtitled video broadcasts. The main issue was that turnaround time was too slow, as the subtitled video did not reach the Chinese offices until weeks after the initial firmwide broadcast.. Regional offices have made significant efforts in localizing client-facing material, such as contracts and marketing collateral. But there is currently no established firm-wide workflow or standard procedure, and no centralized translation department. The Global Advisory Studio provides support for design teams working across borders on matters such as taxes, currency, legal liability, and local customs. This team has been responsible for assessing translation needs for the firm, comparing and vetting various translation service providers in order to select a suitable firmwide vendor.. A translation management platform service was engaged for a pilot program. The studio had dedicated months to customize the tool and train staff members. The tool would have served as a centralized portal for all the firm’s translation needs, with context-specific translation memory, project tracking, billing, and outsourcing 9.

(18) capabilities. The goal was to track the amount of translation work around the firm, capturing content for translation memory, and streamlining translation workflow.. Asia was elected as a pilot region to beta test the tool, onboard local vendors, and train translation project managers. However, the platform’s implementation proved to be difficult, and the user interface and workflow were unintuitive. The concern was that most employees would not utilize the tool because of the steep learning curve. The firm chose not to extend the pilot program with this translation vendor.. Following an introduction to the multinational architecture firm and the subsidiary that is the subject of this research, the thesis will provide a review of relevant literature on language management and translation in multinationals. The methodology chapter explains the development of the survey questions, including the pre-survey interviews and the previous survey results, and the data collection process. Findings from the survey and interviews are laid out in Chapter 4, followed by discussions in Chapter 5. And finally, a summary of the research is presented in Chapter 6, along with limitations of the current study and suggestions for future research.. 10.

(19) Chapter 2 Literature Review. This chapter reviews previous studies related to the issues of language barrier in business, business English environments, and translation that inform the fundamentals of this research, and explores various approaches multinationals have taken to counter these challenges.. 2.1 Language in Multinational Corporations. Since the advent of trade, businessmen have had to learn the languages of business, and as the world becomes more globalized, the impact of language barrier have extended beyond sales and marketing functions, and can now be felt in all divisions, from I.T. to operations. Businesses have long been coping with this issue, but the field of international management did not truly begin to discuss languages in multinational corporations until the late 1990s, beginning with a series of research by Rebecca Marschan-Piekkari, where language was dubbed “the forgotten factor” (Marschan, Welch, & Welch, 1997), or simply “in the shadow.”. This neglect has been contributed to language being considered as an operational or technical issue, rather than a pervasive, daily occurrence; corporate language thought of as simply a matter of choose one over the other, based on company profile. 11.

(20) (Maclean, 2006); the legacy of Hofstede’s (1980) cultural research, which overlooked language entirely as a factor. Feely and Harzing (2002), on the other hand, contributed this oversight to the predominantly English-centric management literature. The assumption being that English-speaking researchers are less keenly aware of the challenges posed by a lack of translation. Alan J. Feely and Anne-Wil Harzing (2002) performed a quick database search, demonstrating that while any keyword relating to communication, management, or international business returned “tens of thousands of journal articles,” merely adding the word “language” eliminated over 99% of the search results, most of which were targeted at language training. This was back in 2002, and the field has come a long way since then.. Researchers focusing on language in the international business context became more prolific in the early 2000s,. (A. Feely & Harzing, 2002; A.-W. Harzing &. Feely, 2008; A.-W. Harzing & Maznevski, 2002; Henderson, 2005; Janssens & Steyaert, 2014; Peltokorpi & Clausen, 2011; Pudelko & Harzing, 2008) Brannen et al., (2014) declared language as “the heart of international business activities” in a Journal of International Business Studies special issue, which generated 78 submissions and 12 published articles, demonstrative of the increased interests in language as a research domain in the field. Janssens et al. (2004) focused on the. 12.

(21) potential contribution of translation studies in developing interdisciplinary language management strategies by inviting translation scholars to enter the field of international management. Over time, language became an increasingly cross-cultural management problem (Harzing & Christensen, 2004; Harzing, 2001; Noorderhaven & Harzing, 2003; Reiche & Harzing, 2010) and a factor in headquarter-subsidiary relationships. Harzing & Feely (2008) described a vicious communication cycle resulting from language barrier, and illustrated how this can affect management decisions.. 2.4.1 Lingua Franca. One reason for the lack of early academic attention in language barriers may be the popularity of English as the language of international business, commonly known as Business English as Lingua Franca, or BELF (Gerritsen & Nickerson, 2009; Kankaanranta & Louhiala-Salminen, 2010; Kankaanranta & Louhiala-salminen, 2013; Louhiala-Salminen & Kankaanranta, 2012). The factors for the popularity of English are manifold, including: the geographical extension of the British Empire; the economic and political power of the U.S.; the rise of digital communication technologies; and cross-border mergers and acquisitions (House, Javidan, Hanges, & Dorfman, 2002). To further its reign in the global lingua-sphere, English is also the 13.

(22) dominant language of academia, specifically in the professional management field, where most of the research and theories are developed and honed in North America. All of these influences combined to make English the de facto language for global commerce (Kankaanranta & Louhiala-salminen, 2013; Rogerson-Revell, 2007).. Adopting English as the corporate language would theoretically negate the need to explore other possible solutions for language barrier in the workplace, and is a policy adopted by many multinationals to facilitate communication between headquarters and subsidiaries. Rakuten, a Japanese online marketplace on a fast global expansion track, announced that their corporate language would henceforth be English, replaced all Japanese cafeteria menus and signage with English, and declared that employees who did not improve their English fluency within two years would face demotion or termination. While this decision was met with criticism, the policy did contribute to staff diversity. Adopting English as corporate language allowed Rakuten to recruit from a larger pool of international candidates. (T Neeley, 2012). Research also showed that diversity in teams, whether in discipline, language, or culture, can be beneficial to performance (Polzer, Milton, & Swann, 2001).. 14.

(23) 2.4.2 Corporate Language. Companies that have adopted corporate languages include General Electric, Daimler-Chrysler, Airbus, Fast Retailing, Renault, SAP, Technicolor, Nokia, Electrolux, Nordea (Rebecca Piekkari et al., 2013), Kone (Marschan et al., 1997), Microsoft Beijing (T Neeley, 2012), and Siemens (Riikka Fredriksson, BarnerRasmussen, & Piekkari, 2006). A shared corporate language has been shown to promote corporate cohesion, trust, and common vision within multinationals (W Barner-Rasmussen & Björkman, 2005). Electrolux, a multinational appliance manufacturer headquartered in Stockholm, Sweden, refers to their corporate language as “bad English,” a neutralized version of a “Midatlantic” English that allows for the average employees to successfully communicate with each other (Rebecca Piekkari et al., 2013). This can be translated into cross-cultural communication training, where native speakers of the corporate language can learn to simplify their vocabulary and moderate the speed of their speech to improve communication with non-native speaker colleagues (T Neeley, 2012).. Fredriksson, Barner-Rasmussen, and Piekkari, (2006) argued that top-down corporate language policies, particularly when it comes to choosing one common corporate language, may not really be a viable solution for multinational companies. 15.

(24) Their studies found that top management, while aware of the political nature of these decisions, tended to proceed with the decision for the sake of efficiency and cost cutting. Rather than a rigid single-language policy, which Fredriksson et al. (2006) has observed as a difficult policy to implement, the researchers advised multinational companies to employ a more flexible approach to language policy, by leaving the corporate language intentionally undecided. This alternative strategy avoids the highly political decision of declaring any one language “corporate,” while allowing different parts of the company to adopt the languages used as they see fit.. Corporate language is certainly not without its challenges and challengers. Compliance is not guaranteed, as discovered by Fredriksson et al. (2006) at Siemens. Headquartered in Munich, Germany, the company has over 400,000 employees, who are mostly based in or around Germany. While executive communication and training are conducted in English, most of the day-to-day interactions still happen mostly in German. Rebecca Piekkari et al. (2013) conducted a case study on Nordea, a Nordic financial service firm, headquartered in Sweden, operating out of Nordic and Baltic nations. Most of its employees were located in Scandinavian countries, and the company has undergone numerous mergers with Finnish, Swedish, Norwegian, and Danish banks over the years prior to the case study. Their study found that rather than. 16.

(25) being a simple administrative task, translation was crucial to service delivery, and employees will resort to personal social network, external to the company, to ensure that the task is done in a timely manner.. While the company had an official language policy along with an in-house translation department, a significant amount of translation activities was still recorded outside of the management designated language solutions. Employees often relied on their personal internal and external network of contacts, especially when the language requirements were time-sensitive, effectively crowd-sourcing translation from willing and able volunteers (Costales, 2012; Kittur, 2010; Olohan, 2014). Machine translation and outsourcing may provide some relief (Potsus & Deschamps-Potter, 2002), but more research needs to be done on the effects on cost, accuracy, and quality, to fully evaluate the effectiveness of this method of network translation.. Some international companies such as Caterpillar and General Motors establish controlled languages, which limits the vocabulary and syntax usage in communication to ensure all text can be easily accessible for all, without demanding fluency in the corporate language. Content created in controlled languages can also be easily localized into multiple languages through machine translation with greater accuracy. General Motors have rolled-out a Controlled Automotive Service Language (CASL). 17.

(26) And Caterpillar, an American multinational that designs, manufactures, and distributes machinery, launched a controlled language system, dubbed CCE, or Caterpillar Controlled English. The system permits a limited vocabulary of 8,000 words (effectively eliminating over 98% of the English language) and is still in use today (A. J. Feely & Harzing, 2003; Rychtyckyj, 2006). Note that both of these examples are in the manufacturing industry. This may not be a suitable approach for multinationals in the design industry, a similar approach may be applied to content types with consistent, repeated work, such as project bidding documents, or construction drawings.. 2.4.3 Multilingualism. Recent research have begun to suggest language management strategies that acknowledge and address the practical challenges of operating in multiple languages (Angouri, 2013; Badalotti, 2010; Luo & Shenkar, 2006; R Piekkari, Welch, & Welch, 2014; Steyaert, Ostendorp, & Gaibrois, 2011). Riikka Fredriksson et al. (2006) examined multinational corporations as both multilingual and multicultural, and Maclean (2006) in particular suggests that this interest in language management will continue towards more sophisticated, multilingual solutions, stepping away from the monolingual and corporate language policy. Language has gone beyond an 18.

(27) operational decision, with spreading interests in fields such as human resource management, knowledge management, and project management.. Previous research has been conducted on companies in other industries like manufacturing, construction (Ochieng & Price, 2010), and banking. The case of Nordea's banking service (Piekkari et al., 2013) can provide some insight to language solutions for professional service providers, but researchers have yet to explore translation needs within a professional design firm. Architecture and banking are both service providers, but the services are delivered in very different formats. Collecting and analyzing the various types of translated content and understanding the linguistic needs for design service delivery will facilitate future language management research in other professional services firms.. 2.2 Translation and Language Problems in Management. Some researchers have discussed translation in multinational businesses. One frequently cited example was a Pepsi marketing copy, “Come Alive with Pepsi." Through an unfortunate mistranslation and lack of subsequent review, the advertisement ended up suggesting to the brand’s Chinese audience that the beverage would revive their ancestors from the graves (Kaye, 2006; Ricks, 1983). The translation error was largely linguistic in nature, but the lack of overview raises an 19.

(28) important question: how can organizations manage the translation process to avoid these translation blunders when entering and competing in foreign markets (Steyaert & Janssens, 1997). Gensler had a few of its own to add to the collection of translation blunders: one of the project teams had failed to review a contract, which was translated from German into English with Google translate. The contract was then presented to a major global client, with a consultant’s name, “Rösti & Partner,” translated as “hash browns & Partner.”. 2.3 Language as Asset. Marschan-Piekkari et al. observed how language proficiency affected information flows and relationship building, creating an alternative organization structure that empowers linguistically able staff members (Marschan-Piekkari, Welch, & Welch, 1999b). Barner-Rasmussen and Aarnio (2011) argued that the lack of clear directives for future language strategy development is due to lack of understanding in the linguistic makeup of multinational corporations. Their quantitative study confirmed that MNCs are in fact multilingual, with language fluency varying greatly amongst various functions and organizational levels. In response to BarnerRasmussen and Aarnio, Harzing and Pudelko (2013) conducted a large-scale quantitative research study that covered over 800 subsidiaries and provided an 20.

(29) overview of language proficiencies, policies and practices in multinational companies. The results showed regional differences in language management challenges, suggesting that linguistic diversity can be advantageous to communication, and the potential of language as a source of power transcends beyond the official company hierarchy.. The literature has shown that multinational firms are highly reliant on language nodes. A “language node” is often an individual in a subsidiary fluent in the company’s working language. These language nodes, or language intermediaries, will be tasked with translation work, or act as messengers in issues that are not part of the person’s job scope. This bridging role can be fulfilled by an expatriate, inpatriate, or locally hired non-native individuals (Anne Wil Harzing et al., 2011; Nurmi, 1995). These bridge individuals are often tasked with translating content of various nature, well beyond their role, unit or function, and in multiple organizational directions. This may be beneficial in improving horizontal communication, and help the language nodes establish personal networks within the firm. But these translation tasks may also prevent language nodes from focusing on their projects (Marschan et al., 1997; Marschan-Piekkari, Welch, & Welch, 1999a; Marschan-Piekkari et al., 1999b), which. 21.

(30) might prove to be problematic for architects and designers, whose work hours are billable.. Sometimes the language node may not be completely fluent in the host country language, where he or she is stationed. This lack of fluency can prove detrimental to the organization’s information flow, or the individual’s career development. Conversely, the appropriate language skills can significantly improve an employee’s career trajectory, opening up opportunities to assignments that would not have otherwise been available (R Piekkari, 2008). The language node often has the power to influence crucial communication and decision making, but not necessarily in the most constructive way. Often such individuals might engage in gatekeeping behavior, limiting or filtering the information according to personal agenda (Macdonald & Williams, 1994). Poor translation or communication skills can also lead to distortion of the original message which, depending on the direction of the information, can lead to inaccurate information being circulated throughout the organization (Marschan et al., 1997).. A lot of multilingual management research has emerged from Scandinavian countries like Denmark (Tange & Lauring, 2009), Norway (Peltokorpi & Clausen, 2011), Finland (Wilhelm Barner-Rasmussen & Björkman, 2007; Marschan-Piekkari 22.

(31) et al., 1999b), and Sweden (Louhiala-Salminen, Charles, & Kankaanranta, 2005)), where the general population encounter a diversity of languages, albeit in the same linguistic family. Such exposure has evidently led to more awareness of translation issues. One of the most often cited case in the literature is Kone, a Finnish multinational that is a leader in the global elevator industry, had adopted English as its corporate language. But this did not resolve the issue of language barrier. An informal “shadow structure” formed based on language clusters, around individuals with the relevant language proficiencies as language nodes. These language nodes found themselves in a position of power since they can control the flow of information and influence decisions that should be beyond their pay grade (Marschan et al., 1997).. Forbes Insights conducted a survey in collaboration with Rosetta Stone in July 2011, aimed at senior executives at high-grossing U.S. companies. The results indicated the importance of language skills in career advancement, 40% of the respondents stated that they expect their expatriates to possess intermediate proficiency in the host country’s language; 93% surveyed agreed that expatriate managers would be more successful if they were bilingual. This perceived advantage can affect hiring decision, and can further impact overall language policies in multinationals, and echoes the. 23.

(32) exceptional advantages and power of language skills presented by Marschan et al. (1997).. One notable finding was the relationship between language fluency, effective communication, and trust formation (W Barner-Rasmussen & Björkman, 2005; Wilhelm Barner-Rasmussen & Björkman, 2007). This suggests that proficiency in the corporate language and trust in select individuals, whether working in foreign subsidiaries or headquarter, are major factors in effective knowledge sharing (Lauring & Selmer, 2011; Louhiala-Salminen & Kankaanranta, 2012; Moenaert, Caeldries, Lievens, & Wauters, 2000; Tenzer et al., 2013), and successful knowledge transfer and expertise sharing is integral for any multinational to stay competitive in the market (Home-based & Mne, 1995). For a professional service firm such as Gensler, one of the biggest competitive advantages is the wide network of global talent and expertise. Offices often collaborate on projects across borders and time zones. If this accumulative expertise bottlenecks in individual offices due language the whole firm suffers.. 24.

(33) Chapter 3 Research Methodology 3.1 Research Methods. Data for this research was predominantly collected from an online survey, which was developed based on previous internal survey and semi-structured one-on-one interviews.. 3.1.1 Pre-Survey Interview. Questions and responses from a previous survey conducted by Firmwide Global Advisory served as a blueprint for the initial draft of the Shanghai office survey. This previous survey reached 74 respondents: project managers, risk managers, and marketing staff, mostly in the U.S. or Latin American offices. The results were presented to the executive leadership of the firm to demonstrate translation demands. Survey questions ranged from types of content, translated cost, content, roles, and workflow.. A series of pre-survey interviews were conducted with key individuals in the Gensler Shanghai office to ensure the survey questions were comprehensive and inclusive.. 25.

(34) With permission from the Gensler Shanghai office, the researcher conducted one-on-one interviews with individuals: an in-house translator, a studio coordinator, an executive assistant, as well as a two designers. Additional questions were added to the survey based on preliminary interviews with these key individuals. Non-native Mandarin speaker designers were also interviewed to gauge the amount of translation workload, understand how day-to-day communication takes place between local and international team members, and how employees cope with language barrier. The interviews were conducted during office hours, in a small private conference room for added privacy. Each session lasted 30 minutes to an hour. The interviews were documented using an iPhone application, Recording Lite, and later transcribed by the researcher. The interviewees were given a printed copy of the survey question draft to review, then asked about their personal experience dealing with translation in the office: the challenges, the frustrations, the workarounds, and the tools they used. They were then asked to suggest additional questions, answers and other adjustments to the survey draft.. Aside from their involvement in translation work, the individuals were also selected based on their job role, as the employees come from very different functions and backgrounds. These interviewees had also expressed willingness and concrete. 26.

(35) suggestions to improve the office’s translation process, and demonstrated understanding of the challenges and complexity of translation work.. Table 2. Interviewees Role. Language Capacity. Function. Former in-house translator. Native Mandarin speaker, professional English fluency. Marketing/Admin. Executive assistant. Native Mandarin speaker, professional. Admin. English fluency Architect. Working level Mandarin (spoken), native English fluency. Design. Architect. Elementary level Mandarin (spoken), native English fluency. Design. The first interviewee was a former in-house translator for the Gensler Shanghai office. She was brought on board as a full-time translator, supporting marketing and administration and translated content such as marketing copy, project case stories, contracts and other bidding documents. She was also responsible for coordinating with an outside translation agency for any work outsourced. She had since transitioned to a new role in knowledge management, and only occasionally provides translation support.. The second interviewee was an executive assistant to the Asia region managing principals, who had previously worked as a studio coordinator. The researcher had worked with her on a client requested translation prior to the survey. She spoke fluent 27.

(36) Mandarin, English, and Japanese, and translated meeting minutes, project descriptions, and day-to-day communication for the principals.. The third interviewee was an American-born Shanghainese architect, who spoke working level Mandarin and Shanghainese and is a native English speaker. Her team had produced a bilingually narrated and subtitled video for a company research project, and she worked with the researcher on a design survey translation in the past. The last interviewee was an American architect with minimal Mandarin fluency. He provided insight into communication and collaboration with local design institute partners, and how expatriates cope with language barrier in the office and during client meetings. The unilingual and partially bilingual interviewees. The interviewees each approached translation from a different perspective. For the former in-house translator, translation was part of her job description. For the executive assistant, translation was part of facilitating and supporting the expatriate principals, translating everything from an address to a project document. For the architects, translation was a key component in internal collaboration and client communication. The interviewees all provided valuable input that informed the survey question and answer design.. 28.

(37) 3.1.2 Online Survey. The survey was built in an online survey engine called SurveyGizmo (www.surveygizmo.com), for which Gensler had an enterprise account. The survey’s user interface is very friendly and intuitive, and the engine provides a reporting tool, as well as a piping function that allows the questionnaire to draw information from previously answered questions, so further information can be gathered from respondents based on their answers. Through SurveyGizmo’s “display logic” function, which displays questions based on previous answers, respondents are prevented from having to answer questions regarding unfamiliar content types.. For example, a respondent who answered “no” to the question “have you ever outsourced translation to an outside translation agency or translator before?” was not shown the series of follow-up questions regarding translation outsourcing. The survey was set to only allow one response per computer to prevent duplicate responses, as long as cookies were enabled on the respondent’s browser. Since most respondents would be likely to be taking the survey during working hours, the survey was designed to take less than 10 minutes to complete.. 29.

(38) The questions were developed and reviewed based on Fowler’s guidelines (Fowler, 1995). The survey questions were translated into Simplified Chinese by the researcher, and reviewed by a native Chinese speaker from the marketing team for clarity and accuracy. Since the respondents are from different functional roles and with varying levels of fluency in English, the survey used concise, simple and nontechnical language. Translation industry terms such as TM (translation memory) or Target Language were avoided.. For the sake of inclusivity and ease of understanding, native language is defined in the survey question as “the language you've been speaking since birth, or the language that you speak the most fluently.” Respondents were then asked to selfassess their fluency in the two working languages in the Shanghai office: English and Mandarin. The employees were not asked to specify whether this represented written or oral fluency, with the assumption that most non-native Mandarin speakers would have some oral fluency and minimal understanding of written Chinese. The five-level fluency rating is modeled after the Interagency Language Roundtable scale (Lowe Jr, 1987), and slightly modified for simplicity. Language fluency is self-reported based on the following prompts:. . No proficiency – limited to a few words 30.

(39) . Elementary proficiency – basic greetings and taxi Chinese. . Limited working proficiency – basic social conversations, carry on non-technical conversations. . Professional working proficiency – can conduct business and discuss technical, professional topics with ease, making only minor mistakes occasionally. . Native or bilingual proficiency. The various types of content that require translation were collected and refined through the pre-survey interview process, and the respondents were asked to select all of the content types that they have needed translation for. Identifying and understanding the translation requirements of each type of content will help language service providers put forth the best teams and solutions for the jobs, and improve the overall quality and efficiency of translation.. The survey was beta tested by 6 employees, who submitted feedback and edits via e-mail, face-to-face feedback sessions, or directly in the SurveyGizmo testing platform. The beta testers were again, all from different functions. One American PR specialist, who had high working level spoken and written Mandarin fluency; one local Chinese finance personnel with working level English fluency; one American architect who spoke minimal Mandarin, but was fluent in Spanish and Catalan; a 31.

(40) native Vietnamese speaking designer; and the last two were the pre-survey interviewees, the former in-house translator, and Chinese American architect.. The overall feedback was positive, the testers considered survey comprehensive, clear and easy to complete, with the exception of a couple of questions. One particular question that was met with some complaint: respondents were required to rate the importance of turnaround time, accuracy, and cost, with an answer format known as the “Star Rating Grid” in SurveyGizmo. A rating was required for each type of translation content. With a total of 8 content types and 6 quality requirements to rate, the respondents would have to click 48 times before they could finish the question. The beta testers considered this rating system too exhaustive. Since the question appeared at the very end of the survey, it felt even more daunting. The beta testers suggested that this might cause respondents to abandon the survey when they’re just a few questions away from completion. One tester suggested that the question be broken down and distributed throughout the survey. However, since the survey was organized in sections for Self Translation, Outsourcing, and Peer Translation, this approach could confuse the respondents on the type of translation work being rated. Another strategy would have been to minimize the number of translation type categories by merging some of the options based on function. But this would make it. 32.

(41) difficult to compare the data collected. In the end, the questions were considered too exhaustive and were eliminated from the survey entirely.. 3.1.3 Participants. The online survey was launched through the Gensler Shanghai office e-mail list to 216 employees. The call to action e-mail informed potential respondents of the survey’s anonymity. The message was sent to all 216 employees of the Shanghai office. Two reminder e-mails were sent out in the following weeks, which contributed to visible peaks in participation. The researcher also relied on personal network within the office to encourage participation in different studios.. 33.

(42) Chapter 4 Findings. This chapter presents the findings from the data generated from the methods and procedures described in the previous chapter. The results of the quantitative analysis on the shadowing data is first presented, followed by the findings of the qualitative analysis of the interview data. Discussions on the implications of these results can be found in the next chapter.. 4.1 Survey Results. Over the course of two weeks, a total of 77 completed responses were collected for the survey, with another 16 partially completed results disqualified from the final data. Of all respondents, 64.9% of the employees surveyed are designers, followed by project managers 14.3% (see Table 3). Note that for the purpose of this survey, designers refer to all designers, including architectural, interior, branding, signage and wayfinding, and graphics. The designers are organized into “studios,” each with their own project managers, coordinators (similar to assistants), and directors. Administration, IT, finance, accounting, human resources, marketing, communications, and risk management are all considered part of the “Administrative Studio,” and support all of the architecture and interior design studios.. 34.

(43) Table 3. Job Roles. Role. Percent. Count. Architecture / Design. 64.9%. 50. Project Management. 14.3%. 11. Administration / Reception. 1.3%. 1. IT. 1.3%. 1. Finance / Accounting. 2.6%. 2. Management. 6.5%. 5. Human Resources. 3.9%. 3. Marketing/Communications. 3.9%. 3. Risk Management. 1.3%. 1. Total. 100%. 77. 35.

(44) 4.1.1 Linguistic Assets. As an American firm founded in San Francisco, the Gensler working language is English, though it was never officially declared as the “corporate language.” With 30 offices in North America (29 in the United States, 1 in Toronto, Canada), the firm operates chiefly in American English. London, the hub office for the EMEA (Europe, Middle East, and Africa) region is also a native English speaking location. The company intranet is in English, and most of the firm’s external communication channels, such as the official website, blog, and LinkedIn page are English-only. All internal communication between headquarters and subsidiaries, written or spoken, is conducted in English. Working level English fluency is a basic requirement and an inherent part of the company’s human resource strategy. This can be referred to as logical ethnocentrism, which in some cases can lead to some neglect to host country culture, language and practice (Anne Wil Harzing & Pudelko, 2013). Also, a shared second language does not always lead to clear communication, since cultural values can affect communication styles and prevent. Gensler Shanghai also has many expatriate directors, most of whom have little to no fluency in Mandarin. When a project lead is brought to a studio, the project information needs to be translated into English for the directors to evaluate whether or 36.

(45) not to bid. There are also international designers working within the studios, in various functions and levels. These international designers need to collaborate with their local colleagues on projects and work with local design institutes to understand local building codes, project requirements, and client feedback. But all this translation work is rarely documented.. The collaborative and flat nature of Gensler demands effective communication. Employees depend on their network across offices and divisions to facilitate their work and share information across borders and time zones. Developing well-rounded policies and support systems will ensure that the organization remains effectively collaborative in practice. Language training and selective recruitment of linguistically capable staff can also be beneficial, but a thorough language audit (Reeves & Wright, 1996) would help the firm acquire a thorough understanding of its existing linguistic resources and translation demands. Ideally, language will eventually be considered an inherent part of the firm’s overall strategy, and not just translating everything like an afterthought (Marschan-Piekkari et al., 1999b).. Table 4. Native Language. Native Language. Percent. 37. Count.

(46) Mandarin Chinese. 68.8%. 53. English. 24.7%. 19. Cantonese. 2.6%. 2. German. 1.3%. 1. Korean. 1.3%. 1. Vietnamese. 1.3%. 1. Total. 77. Regional and international collaboration and knowledge sharing are highly dependent on effective and efficient communication, and most staff members speak English at working level fluency or above. The Gensler Shanghai office operates in both English and Mandarin Chinese. All staff meetings are conducted in English, with sprinklings of Chinese, but public office announcements are bilingual. All human resources, legal, and financial documents are bilingual. Shanghainese, the local dialect in Shanghai, can be heard in studio meetings and pantry conversations. Externally, the choice of language may vary depending on the client, communication with local consultants, vendors, and local design institutes is generally carried out in Mandarin Chinese. The firm provides a lot of resources for its subsidiaries, such as stock responses to RFPs (requests for proposals), standard Non-Disclosure Agreements, 38.

(47) research data, and marketing materials such as case stories and other project metrics. But these resources are usually produced in English only, and must be localized. Table 5. English Fluency. How would you rate your English. Percent. Count. No proficiency - limited to a few words. 0.0%. 0. Elementary proficiency – basic. 3.9%. 3. 27.3%. 21. 41.6%. 32. 27.3%. 21. fluency?. greetings and taxi Chinese Limited working proficiency – basic social conversations, carry on nontechnical conversations Professional working proficiency – can conduct business and discuss technical, professional topics with ease Native or bilingual proficiency Total. 77. 39.

(48) Only 3 designers claimed “elementary proficiency” in English, with zero respondents claiming “no proficiency,” which was the anticipated result, since employees are not likely to self-report having no fluency in English, the company’s de facto corporate language. All in all, 68.9% of the office speak professional to native level English. 32 out of the 77 surveyed (41.6%) reported professional working proficiency in English, and with 19 native English speakers, 21 reported native to bilingual level fluency in English (see Table 5).. Table 6. Mandarin Fluency. How would you rate your Mandarin. Percent. Count. No proficiency - limited to a few words. 2.6%. 2. Elementary proficiency – basic. 10.4%. 8. 11.7%. 9. Chinese fluency?. greetings and taxi Chinese Limited working proficiency – basic social conversations, carry on nontechnical conversations. 40.

(49) Professional working proficiency – can. 6.5%. 5. 68.8%. 53. conduct business and discuss technical, professional topics with ease Native or bilingual proficiency Total. 77. Reversely, only 5 non-native Mandarin speakers rated themselves as having professional working fluency in Mandarin (8 at elementary; 9 at limited working proficiency, see Table 6), but 91% of the office has at least elementary proficiency in Mandarin, indicating that while few have mastered the local language, everyone has made an effort to communicate with their local team members. Only 2 reported “no proficiency” in Mandarin, and both of these are designers, these along with the 3 designer respondents with elementary English proficiency suggest that architects and designers can perform their work effectively, even if they were only proficient in one of the two working languages in the office.. 41.

(50) 4.1.2 Content and Assignment. The data gathering process for this research also helped identify various types of content that requires translation, crucial to refining internal translation needs. There are many different kinds of clients within the firm’s structure: the legal team requires accuracy and confidentiality, but the content is repetitive and recyclable. The marketing and communications community need content that is localized to the target market, and can afford a lot more creative liberty in the target language. The designers and project teams can benefit from a robust, industry-specific glossary or translation memory, and minimize the time they spend searching through various files and references.. Qualifications, proposals, and presentations are the most common client-facing written communications. All of these publications include standard, boilerplate text introducing the firm’s history, philosophy, and body of work. Qualifications are curated collections from the portfolio, with project images, case stories, and basic metrics to demonstrate design capabilities to clients; proposals are more tailored, specific to individual client requirements, with a proposed scope of work, design team and fees; presentations are for face-to-face client interaction, and can have a variety of content from design concept to site analysis. 42.

(51) Contract and other legal documents include bidding and tender documents, contracts and agreements with clients, vendors, and consultants. Marketing collateral refers to printed publications that are gifted to clients for marketing and business development purposes. This can be a curated portfolio with a specific point of view, or a corporate periodical and annual report that provides an overall positioning for the firm. Project deliverables include models, renderings, sketches, project books, or multimedia materials created for design projects to showcase options to the client.. Internal and external communication include e-mail messages, phone conversations, and meetings. This content type can include written translation or oral interpretation, but the distinction was not made, since the purpose of the questions was to assess frequency and content types, rather than the method of translation.. Drawing and sketches are usually translated to help facilitate internal collaboration, or to communicate designers’ requirements to the local design institute, who is responsible for producing the final construction drawings for government approval. Meeting minutes sometimes need to be translated depending on the language used in the meeting. If a non-Mandarin speaking designer or design director is present at the meeting, the meeting minutes will be translated to help get the designer up to speed. 43.

(52) Table 7. Types of Content Translated - General. Content Types. Percent. Count. Proposals / Qualifications /. 54.7%. 35. Contracts and other legal documents. 35.9%. 23. Marketing collateral. 26.5%. 17. Project deliverables (project book,. 73.4%. 47. 64%. 41. 56.3%. 36. Drawings and sketches. 68.7%. 44. Meeting minutes. 67.2%. 43. Other (required). 6.3%. 4. Presentations. multimedia, etc.) Client / consultant communication (email, phone, conversations, etc.) Internal communication (e-mail, phone, conversations, etc.). Total. 64. The types of content that most often require translation are project deliverables, followed by drawings, sketches, and meeting minutes. External communication with 44.

(53) clients and consultants make up for 64% of the general translation workload (see Table 7).. Table 8. Translation Assignment. Frequency. Outsource. Self Translate. Peer Translate. 24%. 41%. 33%. When faced with a translation task, employees most often translate documents themselves (41%, see Table 8 above). Considering most employees at Gensler Shanghai are moderately proficient in both English and Chinese, the tendency to self translate was not unexpected. However, this question does not specify the type or size of the translation task. Even though later questions asked respondents to identify how each type of content is handled, this question could have been more specific about the time or word count that qualifies a translation task. Since translation was defined loosely at the beginning of the survey, respondents may be reporting that they “self translate” a conversation with an English-speaking colleague or an e-mail message. If the survey is to be duplicated in another office or region, the definition of what constitutes a “translation task” should be more descriptive to yield better results.. Table 9. Native Language vs. Translation Coordination 45.

(54) Mandarin English. Cantonese German. Korean. Other. Total. 64. Chinese Yes. No. 40. 19. 2. 1. 1. 1. 75.5%. 100.0%. 100.0%. 100.0%. 100.0%. 100.0%. 13. 0. 0. 0. 0. 0. 24.5%. 0.0%. 0.0%. 0.0%. 0.0%. 0.0%. 19. 2. 1. 1. 1. 100%. 100%. 100%. 100%. 100%. Total 53. 100%. 13. 77. Translation is most often handled by native Mandarin speakers: 75.5% of the Mandarin-speaking staff have performed translation themselves and have delegated translation work to their colleagues. The native English speaking staff also require a lot of linguistic assistance and reported a lot of peer translation. But since most of them have limited Mandarin fluency, they’re rarely tasked to coordinate with the translation agency or perform translation themselves.. 46.

(55) Unlike Nordea, as of the time of writing, Gensler Shanghai does not have inhouse translation support. Up until May 2015, the office had a full-time translator employed from an external translation agency stationed in the legal department, where the majority of the workload came from. Few people in the office knew about her role, and those who did were not sure if she was allowed to work for other studios. After the head of China legal left the firm, she no longer received translation requests, and the office decided that the amount of translation work in the office did not demand a full-time translator on board. The office’s former in-house translator had been transitioned to a different role a few months ago, but prior to that, her time was chiefly dedicated to the marketing and administration studios. An interviewee had stated that since the in-house translator’s billing rate at the time was very high, it wasn’t economical to insource to her. “It just doesn’t make sense. She’s admin, not billable,” said the interviewee. Nordea’s central translation agency met similar challenges. Employees either didn’t know how to reach out to them, unsure if they were allowed to or thought the process too bureaucratic and inefficient (Rebecca Piekkari et al., 2013).. A solution would be to establish a translation service desk. Gensler’s company intranet already houses an IT service desk, where employees can submit support. 47.

(56) requests, choosing from a drop-down menu of technological challenges, and assigning the requests with levels of priority. A task force would need to be assembled to monitor and manage the translation requests service desk, with access to good vendors and internal linguistic assets. Contact information and rates for pre-vetted translation agencies can also be posted here to direct employees to trusted language service providers.. Table 10. Types of Content - Outsourced. Types of Content Outsourced. Percent. Count. Proposals / Qualifications / Presentations. 42.5%. 17. Contracts and other legal documents. 47.5%. 19. Marketing collateral. 12.5%. 5. Project deliverables (project book,. 42.5%. 17. 22.5%. 9. multimedia, etc.) Client / consultant communication (e-mail, phone, conversations, etc.). 48.

(57) Internal communication (e-mail, phone,. 2.5%. 1. Drawings and sketches. 25%. 10. Meeting minutes. 17.5%. 7. Other (required). 7.5%. 3. conversations, etc.). Total. 40. Contracts and legal documents make up a large portion of the translation work outsourced. 19 out of 23 who have encountered legal translation chose to outsource the task (Table 10). This may well be due to the sheer amount of text in every contract, but it also demonstrates that employees understand that legal translation requires domain knowledge that themselves or their colleague do not possess. Since legal translation for architectural and design services is likely to contain terminology that’s repeated in each document, working with a reputable translation agency with a dependable translation memory can be very beneficial to the firm.. Table 11. Reasons for Outsourcing. Reasons for Outsourcing. Percent. 49. Count.

(58) Not enough time / too much content to. 90%. 36. The content is too technical or complicated. 57.5%. 23. It's an uncommon language and no one in. 5%. 2. 25%. 10. 5%. 2. translate. the office can do it Billing rate for in-house translation is too high Other, please specify (required) Total. 40. Two major factors that lead employees to outsource translation work are the usual suspects: not enough time (85%), or when content is too technical (55%), see Table 11. Time is the obvious reason for outsourcing. Fifty-five percent of the employees would choose to outsource when the content is too technical, which coincides with the high rate of outsourcing for contracts and other legal documents. The majority of the translation work is outsourced to Tongji (82.5%, Table 12), an agency that specialized in architecture and design related translation. The agency also handles a lot of legal translation for the office. The pre-survey interviewees had mixed responses regarding the quality of their work, and all stated that they reviewed 50.

(59) everything translated by this vendor whenever possible, indicating lack of trust in the quality of their work. The agency has also had a few mishaps regarding project confidentiality and have reportedly send translated documents from competitors to Gensler employees.. Table 12. Translation Agencies. Which translation agency have you. Percent. Count. Tongji. 82.5%. 33. Capita. 2.5%. 1. Yeeyan. 2.5%. 1. Linguitronics. 2.5%. 1. Other. 10.0%. 4. Not sure. 15.0%. 6. outsourced to?. Total. 40. Thirty-three out of the 40 respondents who outsource translation work with Tongji. The other three vendors listed have only been engaged by the marketing 51.

(60) studio, with minimal collaboration with the designer. But this survey brought the discovery of another translation vendor previously unknown to the researcher and the pre-survey interviewees, Wutong. All 4 of the respondents who answered “other” listed Wutong as their agency of choice.. The process of surfacing, identifying and vetting translation vendors can be tedious at best, and downright overwhelming for employees who are unfamiliar with the subject. While the information asymmetry (Mishra, Heide, & Cort, 1998) in the market can make it difficult to evaluate translation quality, Gensler Shanghai has many domain knowledge experts in both working languages, and should be able to assess the accuracy and quality of translated content.. Table 13. Types of Content - Self Translated. Types of Content Self Translated. Percent. Count. Proposals / Qualifications / Presentations. 38.7%. 19. Contracts and other legal documents. 20.4%. 10. Marketing collateral. 12.2%. 6. Project deliverables (project book,. 69.4%. 34. multimedia, etc.) 52.

(61) Client / consultant communication (e-mail,. 61.2%. 30. 57.1%. 28. Drawings and sketches. 59.2%. 29. Meeting minutes. 55.1%. 27. Other (required). 2.0%. 1. phone, conversations, etc.) Internal communication (e-mail, phone, conversations, etc.). Total. 49. Employees are most willing to translate project deliverables themselves, followed closely by external communication, internal communication, drawings and sketches, and meeting minutes. Note that project deliverables, drawings and meeting minutes all contain highly industry or project specific content, require extensive design expertise and involvement in client projects. Evidently the employees are quite comfortable translating materials they are familiar with. One of the architect interviewees also stated that outside translators, even though they specialized in architectural translations, often “just translate” rather than fact check. For example, they would create an English name for a Chinese client’s company, rather than identifying the official English company title. 53.

(62) Marketing collateral is amongst one of the least reported types of content that is self-translated, but this is relative to the number of marketing staff participating in the survey. Even though only 4 people self-identified as part of the marketing/communications team, 6 people reported having translated marketing collateral (see Table 7 below).. Only native Mandarin-speakers with elementary fluency in English reported never having encountered translation tasks at all during their time at Gensler Shanghai. Respondents with low proficiency in English also unanimously reported as having no experience in outsourcing translation to agencies, demonstrating that translation work is usually handled by Mandarin-speaking staff, and that nonMandarin speaking staff members are completely reliant on on colleagues to handle anything translation related. Of the respondents who claimed that they’ve never encountered translation needs/tasks in the office, majority (76.9%, see Table 14) selfidentified as designers, suggesting that translation tasks are usually handled by administrative or project management staff, and not by the designers themselves.. Table 14. Role vs. Translation Demand. 54.

(63) What is your role at Gensler?. Yes. No. Total. Architecture / Design. 40. 10. 50. 62.5%. 76.9%. 10. 0. 15.6%. 0.0%. 1. 0. 1.6%. 0.0%. 1. 0. 1.6%. 0.0%. 0. 2. 0.0%. 15.4%. 5. 0. 7.8%. 0.0%. Project Management. Administration / Reception. IT. Finance / Accounting. Management. 55. 10. 1. 1. 2. 5.

(64) Human Resources. Marketing/Communications. Risk Management. Totals. 2. 1. 3.1%. 7.7%. 4. 0. 6.3%. 0.0%. 1. 0. 1.6%. 0.0%. 64. 13. 100%. 100%. 3. 4. 1. 77. 4.1.3 Resource Allocation. One key question is to understand how much time designers and architects spend of handling translation work. As the core service providers of the firm, their time is the firm’s biggest asset. If translation distracts the designers from providing satisfactory services to the clients, how can that be mediated? And if translation is an 56.

(65) inevitable, necessary part of the job, how can we make the most efficient use of their time? What tools or processes can the firm provide them to guarantee the best use of their billable hours?. As professional service providers, architects bill time to projects, much like lawyers charge for their legal services by the hour. Employees at Gensler are required to submit a weekly timesheet, logging their hours to projects they are working on under project numbers, with labor codes specifying the type of service provided. Time is the resource, and the project numbers dictate where the resource is allocated. Client and project work hours are considered “billable,” and will be charged to the client’s project fee. For administrative work, such as internal meetings, the hours are overhead expenses. There is no labor code or project number dedicated to translation. Some individuals log translation work hours to client projects, others simply absorb the hours without ever charging them. This practice can be especially costly if a designer with a high billing rate were tasked with translation work, often because of his or her fluency in a client language. The hours dedicated to translation are therefore invisible.. In general when it comes to translation, there is no standard protocol for what is billed and what is not. Employees make these decisions based on personal understanding of their roles, quick judgment call, or follow instructions from their 57.

(66) project managers, who are responsible for the overall cost control of any client project. It would be worthwhile to set up a labor code, project number, or a similar tracking tool to calculate how much time is actually spent on translation. The cost of insourcing to a designer can then be easily weighed against the cost of outsourcing.. Table 15. Cost - Outsourced Translation. Outsource Translation Billing. Percent. Count. To the related project number. 85.0%. 34. To a general number. 5.0%. 2. Dealt with on case-by-case basis. 10.0 %. 4. Total. 40. When translation is outsourced, it is almost always billed to the project number (85%, see Table 15). Translation cost in Shanghai is relatively low, with Gensler Shanghai’s preferred vendor, Tongji, charging RMB 180/1000 words (NTD 0.9/word) for English into Simplified Chinese, or RMB 200/1000 words (TWD 1.0/word) for Simplified Chinese into English. Tongji translators do not use any machine translation tools, translation memory, or content management system, therefore the same content 58.