Introduction
“Thank you for calling. For English, press or say ‘one.’ Para español, oprima o diga ‘dos’.” With variations, this statement is the initial telephone message used by numerous corporations for services directed to mass consumers in the United States. The statement reflects the relevance of the Hispanic market segment and the Spanish language in reaching corporate objectives. The incorporation of the Spanish language in the business practices of North American companies supports the hypothesis of a shift in favor of Spanish in the United States, especially in the southwest where historically an important segment of the HispanicFootnote 1 population is concentrated (Jenkins Reference Jenkins2009; Martínez Reference Martínez2009). This refers to an unintentional change, termed a language shift reversal, produced by demographic and market forces, as businesses rapidly respond to the increase of Hispanic clients (Hidalgo Reference Hidalgo1995).
Following directly from business decisions to use Spanish to serve the growing Hispanophone market, during the last decade call centers, especially bilingual Spanish-English call centers, have witnessed a notable growth on the US-Mexico border. This involves a significant new source of jobs occupied mainly by Hispanic workers in locations such as El Paso, McAllen-Edinburg-Mission, and Corpus Christi, Texas. These locations are abundantly supplied with bilingual operators, relatively skilled but with low pay rates in comparison to other locations in the United States, as detailed below. Since these linguistic activities are conducted, at least in part, by workers located within United States territory, we can hypothesize that there will be an increase in the labor market value of the Spanish language. As Roy indicates (Reference Roy, Bayley and Schecter2003:269), “the marketplace in the new economy is redefining who is seen as bilingual.” Nevertheless, relatively little is known empirically about what directions management is actually adopting in bilingual production sites. Corporations can opt to handle Spanish as an ethnic attribute, inherently characterizing an individual from a specific group, or as a quality of human capital that can be compensated and subject to economic investment (Heller Reference Heller2003, Reference Heller2010; Grin, Sfreddo, & Vaillancourt Reference Grin, Sfreddo and Vaillancourt2010).
To examine this, we look at accounts of call center worker background (community, educational, and training) and workplace linguisitic practices, focusing on the kinds of Spanish used and policies toward such usage. We analyze the characteristics of the site, on the one hand, and the language policy of the corporations on the other. First, we focus on Spanish-English bilingualism as an attraction factor for businesses in the international division of labor. Second, we examine management practices and beliefs about language at the point of production (Spolsky Reference Spolsky2009:4). Our key finding is that Spanish is more often treated as an uncompensated ethnic attribute than as human capital, directly rewarded as such. Obtaining Spanish as a “free” attribute of labor, and more generally flexible Spanish-English bilingualism, is likely to be important in global locational decisions.
We do not find that call centers seek professional Spanish developed by business and educational institutions. Instead, we find that businesses acquire Spanish as a “heritage language” from the bilingual border community. A heritage language emerges organically from usage in a speech community, rather than through formal training, resulting in a high degree of variability in the personal, cultural, sociodemographic, and linguistic traits (Wiley Reference Wiley, Peyton, Ranard and McGinnis2001). To businesses, obtaining heritage Spanish eliminates the need for labor market premiums or workforce investments. Indeed, it is a sign of inexpensive and flexible labor. What, then, are the consequences of these policies for Hispanic workers and more broadly the Hispanic community?
The consideration in the business world of a physical and human space as a “bilingual location” has different meanings and implications for various agents, from executives to local personnel, as well as representations of firms, products, and workers in the minds of customers. This redefinition is produced in the domain of existing linguistic beliefs and practices, which highlight latent fears and anxieties in the heartland of American society, phenomena the literature notes for off-shore call centers (Heller Reference Heller2010:109). El Paso, our field site, frequently brings up negative feelings about the border community among managers from large firms, owing to its proximity to Mexico and especially Ciudad Juárez, its majority Hispanic composition, the linguistic “quality” of the site, and its relative physical isolation with respect to the home sites of companies. These anxieties are extended to Hispanic operators when they come into contact with Anglo American clients.Footnote 2 If the Spanish language is treated more as an ethnic attribute than as human capital, Hispancity and bilingualism are a comparative advantage of the region in attracting investment but the representation of Hispanicity as a marker of exploitability of the workforce is reproduced.
The international division of labor and the linguistic location of call centers
The growing centrality of information (Reich Reference Reich1991; Castells Reference Castells1996) and the linguistic intensity of products and processes (Harris Reference Harris1998; Cameron Reference Cameron, Erreygers and Jacobs2005) makes the need for linguistic management of organizations more evident and requires more urgent solutions (Spolsky Reference Spolsky2009). According to Heller (Reference Heller2010:104), late capitalism consists of the expansion of markets and their progressive saturation, resulting in an increased importance for language in: (i) the need to manage communication at a world scale where diverse linguistic agents come together; (ii) computerization implying new linguistic skills; (iii) the growth of more communication intensive services; and (iv) the use of symbolic resources, often linguistic, to compete in saturated markets. Language is converted into a selling point and, as Duchêne (Reference Duchêne2009:30) indicates, “the recruitment of multilingual employees is becoming a clear financial advantage for companies as it is cheaper and more practical, for example, to have employees who are able to answer phone calls in more than one language”—assuming, of course, that bilingualism comes at zero or close to zero cost to business.
Linguistic management policies in the new economy are still undergoing definition and are subject to important contradictions (Heller Reference Heller2003, Reference Heller2010). In particular, according to Heller, businesses in the new economy move between the search for Taylorist standardization in the use of language in providing service, exemplified by call center scripts as a system of control over discourse and speech (Cameron Reference Cameron2000), and the value of authenticity in the context of product saturation and increased competition, where distinction through authenticity and affective-identity bonds with clients can be key to business success. Additionally, businesses alternate between viewing language as a technical competence, rewarded as such, versus an innate talent of people (or of ethnic groups), viewed as unworthy of investment and compensation.
Connecting these contradictions with the management literature, Batt, Doellgast, & Kwon (Reference Batt, Doellgast and Kwon2004) identify two principle call center management models. The professional service model is characterized by high value clients, particularly other businesses and professionals, where firms emphasize quality service, adaptability, and loyalty, which imply management practices based on high levels of worker preparation, employee autonomy, and relatively high wages. In turn, the mass market model, focused on simple transactions, such as telemarketing, reservations, or credit card handling, requires relatively low skills, and jobs are likely to be highly routinized with low pay (Heskett, Sasser, & Schlesinger Reference Heskett, Earl Sasser and Schlesinger1997; Batt Reference Batt2002).
These typologies interact with the linguistic availability of sites with respect to the international division of labor within a framework that is often characterized as follows: “call centers involve connecting service-seeking customers from wealthy, English-speaking countries with service-providing workers from more multilingual, less economically powerful states or regions” (King Reference King2009:1). The outcomes range between imposing intensive models of control over national identity in the context of wages that are globally low but locally high (the case of India studied by Poster Reference Poster2007:293) to utilizing authenticity in contexts of wages that are low locally but high globally (the case of Switzerland analyzed by Duchêne Reference Duchêne2009:48), to cite two extremes.
While worker preparation and wages are key locational variables in the literature on call center management, bilingualism and the linguistic characteristics of how it is enacted is less well-studied. A good part of the research on language and call centers has focused on the international division of labor, with the paradigmatic example of call centers located in south Asia that offer services to Anglo-Saxon clients (Taylor & Bain Reference Taylor and Bain2005; Poster Reference Poster2007; Morgan & Ramanathan Reference Morgan and Ramanathan2009; Rahman Reference Rahman2009). As this research shows, customers often respond with linguistic prejudices. Businesses attempt to micromanage linguistic variables, such as accent, and workers in turn respond in various ways—compliance, resistance, agency within hegemony (Mirchandani Reference Mirchandani2004, Reference Mirchandani2005; Cowie Reference Cowie2007; Taylor & Bain Reference Taylor and Bain2005; Rahman Reference Rahman2009; Sonntag Reference Sonntag2009). As a result, the predominant focus of the literature on call centers and language is on how narrowly linguistic variables are managed and disciplined, and also focuses only on varieties of English. This differs considerably from the case we describe.
Possibly, the reason why our case differs is that we are addressing a complex situation brought about by migration, both for the labor force and the target public. The typical case examined in the call center literature addresses nation-state homogeneity in the destination markets. The operator, wherever she or he is found, and independently of the cultural-linguistic variety in that point of origin, normally serves a market conceived of as homogenous by the business's communications policies. This includes the interesting multilingual call center case studied by Duchêne (Reference Duchêne2009), where a Swiss tourist call center serves different countries, and operators are assigned to a language to communicate with the client according to the nation from which the call arrives. This case of a diverse workforce but rigid, nation-based linguistic delineation of customers contrasts with the linguistic diversity of both workers and customers that we document, based on both history and current migrations. For business, this demands flexibility as a valuable component of production, while at the same time the global site selection logic continues to emphasize reduction of costs. Border regions and their language actors are well suited for reconciling these two priorities.
Strikingly, the narrow linguistic features of call center work, such as accent, are generally not tightly managed or rewarded in US-Mexico border bilingual call centers. Rather, the linguistic issue is broader. Call centers seek cheap and flexible labor with specific markets in mind. In this case, the market in the US is predominantly monolingual English-speaking but includes a substantial subpopulation that is either monolingual Spanish-speaking or that code-switches between the two languages. Why would firms employ native speakers of Spanish or bilingual English/Spanish speakers with a Chicano accent for English language calls, which we find predominates? As we show in the following sections, these workers are inexpensive and abundant for serving English calls, but sufficiently flexible to handle the Spanish or code-switching calls that do occur. The language agenda of corporations is less a matter of fine-tuning specifics of language performance than it is recognizing bilingualism as a sign of desired low-cost workers without special premiums for preparation in either language. That labor is produced in the borderlands via Spanish acquired as a heritage language and a mixture of formal education and heritage language learning in English without need for extra compensation for these valued language skills as a factor of production in the information economy. The cultural-linguistic reproduction of the workforce is central to this segment of the new economy.
Bilingual El Paso as a location in the new economy
El Paso has a notable situation of diglossia, with Spanish as the subordinate language and English as the prestige language in public and institutional life, that has its origin in the political and economic control that the Anglo American minority exercised in the region at least until the second half of the twentieth century (Timmons Reference Timmons1990). According to the 2008 American Community Survey (ACS),Footnote 3 73.9% of the population speaks Spanish in the home and 70.7% speak English “very well.” Almost all of the Mexican-born (“Mexicans”) people maintain Spanish in the home (97.3%) but only 30.4% indicate that they have very good knowledge of English. In the case of Mexican Americans (52.3% of the whole population), 84.6% use Spanish at home and 82.9% report very good knowledge of English. Among non-Hispanic whites (12.9% of the whole), 12.5% report using Spanish at home (often with spouses).
The association between speaking Spanish and lack of English language competency and work assignments of low communicative requirements and low position in the labor market is a historical trait even today present in the social and labor structure of El Paso. Before the North American Free Trade Agreement entered into force, the main private industry in this region was assembly, especially of men's slacks. This type of industry was characterized by low compensation and Taylorist-Fordist work regimes where the linguistic skill of the workers, mainly Mexican immigrant women with little or no ability in English and low education, was irrelevant to their role in production (Boutet Reference Boutet2001:56; Spener Reference Spener, Gereffi, Spener and Bair2002; Cohen Reference Cohen2009:26). Even today, significant sectors (e.g. janitorial services) follow this pattern. Currently, 29.8% of Spanish speakers in El Paso are below the poverty line, versus only 14.1% of those who only speak English, according to the 2008 ACS data. Language in El Paso is closely associated with inequality in the prestige of activities and labor market remuneration (Gibson, Olmedo, & Caire Reference Gibson, Olmedo and Caire2008).
The new information economy could hypothetically change this situation. As the main local newspaper optimistically claims, El Paso has become “a critical part of the call center-service center ecosystem” that “has grown to be one of the city's largest private employers, has continued to hire people, even during the recession” and in the context of stagnation in the sector in the United States.Footnote 4 This industry started in El Paso during the final years of the 1980s, with the creation of the first bilingual call centers, but it was at the end of the 1990s and the beginnings of the 2000s when large corporations made a more determined commitment to this location. According to the annual statistics of the Quarterly Census of Employment and Wages, in 2001 telephone call centers (NAICS 6056142) employed 1,434 workers in the private sector, while in 2009, the figure reached 7,338. If we consider a wider definition of the sector, that would include thirty-seven different economic activities that involve work specialties in having telephonic contact with clients (Wirtz Reference Wirtz2001; Moss, Salzman, & Tilly Reference Moss, Salzman and Tilly2008), El Paso call centers employ 13,992 workers in thirty-four work centers (REDCO 2010).
Despite the relative local importance of this sector, these figures remain modest vis-a-vis the entire United States. Comparing data disaggregated into 577 Metropolitan Statistical Areas (MSAs) for Occupational Employment Statistics (May 2009), looking at customer service representatives, the main group of workers in Call Centers, El Paso occupies position 103 for absolute numbers (and position 156 for “telemarketers”). In relative terms, 2% of employment in this MSA are such workers, placing El Paso in position 88.
The thirty-four call centers in El Paso at the time of this study are mainly inbound customer services, that is, call receivers. A minority have as their main activity outbound (call senders), doing questionnaires (market research and analysis). In terms of ownership, we find more in-house than outsourced in numbers of centers, but the former are more limited in numbers of positions (workstations) and employees. We find mainly large corporations with US capital that exclusively serve the US market. El Paso call centers largely provide mass consumer services (e.g. in-house for banks, insurance, telephone services; outsourced for sale of television services, etc.). Only in a few cases can centers (of small size) or services offered be classified under the professional service model typical of business-to-business centers.
Following data provided in the Public Use Microdata Sample (PUMS) for 2006–2008 relating to customer service representatives and telemarketers (n = 292), the majority of workers are women (60.3%). Educationally, 39.7% have no college, 46.2% have some college, and 14.0% have a college degree or more. The average age is 29.98 with a standard deviation of 11.41. In 64.9% of cases, they are Mexican Americans, the second group being Mexican-born (16.2%), and the third being non-Hispanic whites (8.3%), lower than the regional population as a whole but overrepresented, as we have seen in fieldwork, in managerial and intermediary positions. The remaining 10.6% belong to other ethnic groups, where 4.9% are Hispanics not of Mexican origin. Among these workers, 79% indicate that they speak Spanish at home and 11.6% speak English “less than very well,” an interesting finding for a “bilingual” workforce that we will address.
Methods
Our research focuses on language and the management, practices, and beliefs of working languages in bilingual call centers. Following Spolsky (Reference Spolsky2009:4–5), we consider language management as the effort (formal or informal) by someone or some group that has or claims to have authority over the participants in the domain to modify their practices or beliefs. Regarding language practices we focus on the socially agreed rules about use of English and/or Spanish for working activities and also the microlevel questions of usage, such as vocabulary choice, pronunciation, and code-switching (Spanish-English). Correspondence between top-down management of language and actual workplace practices is highlighted in the analysis. In relations to beliefs, we consider the interaction between the linguistic background of the worker in life history (such as family, education, etc.) and language practices and management of the enterprise (with attention to linguistic management of worker selection, training, promotion and remuneration methods, and worker-customer interactions).
To analyze these topics for El Paso call centers, we systematically investigated three fundamental dimensions in a sample of semi-structured interviews: (i) the labor hierarchy, examining the positions of operator, supervisor, technician, intermediate manager, language specialist (e.g. translator), and executive, as well as relations with clients; (ii) ethnolinguistic origin and linguistic competence of the participants in the sector, in particular Mexicans, Mexican Americans, and Anglo Americans; and (iii) the basic organizational types of call centers as reported in the management literature: inbound/outbound and in-house/outsourced. Based on this classification and its empirical correspondence with actual establishments in El Paso, we established three call center types for field research: five in-house inbound; three outsourced-inbound; and one outsourced-outbound, which encompass the main types in the area.
Our method concentrated on accounts of local agents, that is, economic development promoters, executives, and call center workers. We performed, recorded, and transcribed in-depth, semi-structured interviews with thirty-one operators, customer service representatives, and technicians, twelve language professionals related to the centers, and eight executives and intermediate managers. Sample selection was based on key informants' interviews (see below) and snowball procedure. The method of inclusion of the respondents was established to fill up the three dimensions mentioned above, and to ensure diversity of experiences and perspectives, which guards against the potential narrowness of snowball selection.
The interview protocol included the following categories of questions: (i) language skills, oral and written; (ii) basic characteristics of the enterprise; (iii) a detailed discussion of the linguistic characteristics of work, including language and language style used, by self, colleagues, supervisors, and customers; (iv) language preparation and adequacy of language skills for work needs; (v) past occupational language use and anticipated future career language use; (vi) the enterprise's language policies; (vii) personal history in terms of nation, family, and schooling, focusing on language acquisition, as well as residential mobility (especially border crossing); (viii) linguistic ambitions and educational methods for children; (ix) language ideologies, including language politics concerning the place of Spanish in the United States and attitudes toward Spanish usage and Spanish-English code-switching; and (x) basic demographics. The analysis of the interviews' transcripts followed the previous substantive categories, adding in a second stage the theoretical categories of practices, beliefs, and management of language. We crossed these categories according to the three dimensions of our sample (hierarchical position in the companies, ethnolinguistic origin and type of call center).
We also did eight key informant interviews, such as regional economic development staff, public authorities, and university experts, focused on the business situation for call centers. Adding significantly to our knowledge, we socialized with sector employees during the fieldwork period, having numerous conversations and informal get-togethers. A common job of students of the university where our research was based, University of Texas at El Paso, is call center operators, which offered many conversational opportunities. This allowed for spontaneous ethnographic learning, going beyond the sometimes cautious or constrained accounts that actors provide in interviews. We also gathered secondary sources, mainly statistical (especially population-level data about language and occupation contained in the Public Use Microdata Sample and in Occupational Employment Statistics) and journalistic (notably, articles from the business and local press). Given the mass of information, here we provide a selection that addresses our key research questions (discussed in the first and second sections above), being careful to provide quotations and facts that are representative of the larger corpus of material we collected.
Localization at the border: Wages and linguistic quality
To understand language use within El Paso call centers, it is important to consider first why call centers would locate on the US side of the border. Because of inexpensive communications technology, call centers are relatively free to locate around the world except for some legal restrictions (e.g. insurance firms we studied) and security risks that limit off-shoring. However, workforce language and interactional skills are important considerations in locational decisions, along with management comfort, labor costs, and workforce flexibility (Batt et al. Reference Batt, Doellgast and Kwon2004). Locationally, El Paso competes with US on-shore sites, where English-Spanish bilingualism is less common (though certainly it can be found), and off-shore sites in Latin America that provide Spanish and sometimes English-Spanish bilingualism (Micheli Reference Micheli2006; Rojas Reference Rojas2007). The particular character of the El Paso workforce is best understood as a balance point between these two extremes.
According to officers of REDCO, the main El Paso economic development agency (involved in continuous collaboration with call centers), the reason call centers invest in El Paso are a combination of: (i) “bilingual labor force;” (ii) “low wages/cost of living;” and (iii) “do not need to pay extra for the bilingual labor force.”Footnote 5 Analysis of statistical sources allows us to generalize this situation for the border and other locations with a Hispanic majority in the United States. The El Paso average wages according to the Occupational Employment Statistics for May 2009 are $11.23 per hour for customer service representatives and $9.30 for telemarketers, among the lowest in the country. In this, El Paso is 543rd of the 577 Metropolitan Statistical Areas (MSAs) with such data. In the thirty-four MSAs with lower wages in the sector, locations that can be considered “bilingual” stand out. The MSA Puerto Rico Metropolitan Area 2, with $7.76 per hour stands out as the lowest average salary. Likewise, Puerto Rico-Ponce is $9.60. San Juan-Caguas-Guaynabo, Puerto Rico has a wage average twenty cents higher than El Paso, and has the largest concentration of customer service representatives at 11,000, that is, more than double the number in El Paso. On the Mexican border, the MSA of McAllen-Edinburg-Mission, Texas, with the highest percentage of Spanish speakers in the fifty states (slightly higher than El Paso), pays $9.72 an hour and has a slightly higher number of workers in the occupation. Likewise, Corpus Christi, Texas, has 4,668 employees and a salary of $10.80 per hour. Our findings should be generalizable.
Nothing appears to indicate that the large concentration of Hispanic workers with their linguistic capital, on the Mexican border or Puerto Rico, favors better wages in comparison to monolingual call centers in the rest of the United States, although they may operate favorably in attracting investments. These examples indicate that the border, as with the past assembly industry, is attracting “cheap labor” firms in the new economy, just as it did for low-wage manufacturing in the past.
A Mexican-American economic promotion officer placed the representations about the linguistic qualities of El Paso in the higher offices of US corporations in the framework of “negative feelings about the border community.” The location on the US-Mexico border awakens these negative feelings due to its proximity to Mexico and especially to Ciudad Juárez, stigmatized by violence, by the majority Hispanic ethnic composition with relatively high rates of poverty, and the relatively great distance from central corporate offices. These factors bring executives to consider El Paso in some ways “near shore” (call center jargon for sites not in the United States, but close by) with respect to its cultural and geographic relationship to corporate headquarters, even though legally it is “on shore” (within the United States).
Following this informant: “The managers [AngloAmericans] from headquarters think that El Pasoans don't speak English. They are worried that El Pasoans cannot give quality service in English, but they assume that their abilities in Spanish are sufficient or they are not even considering the topic.” The negative values about linguistic competencies in English focus on southern accent, Mexican-American (Chicano) accent, low-class register, and excessive use of English-Spanish code-switching. There are also concerns about uneven ability to convert reading into speaking (literacy problems). Following the 2006–2008 PUMS cited above, 11.6% of operators in the sector indicate speaking English “less than very well,” despite the fact that in our fieldwork we never found any positions devoted exclusively to Spanish service. Nevertheless, these deficits are not impediments to businesses hiring these sorts of workers with limitations in English, since they consider that with practice (learning by doing) they can end up with sufficient linguistic abilities. That is consistent with the minimization of labor costs and reduced concern with linguistic quality of service in the mass market service model.
The promoters of local institutions (Mexican-Americans) put into action the local ideology against subordination to the home company managers. They argue to investors that the bilingualism of the El Paso population indicates that “everybody has an accent in English,” emphasizing that the population is “bilingual fluent in Spanish” (our emphasis). The definition by local promoters of the local population as “bilingual fluent in Spanish,” emphasizes to investors the Americanness of El Pasoans: as Americans, English speakers plus possessors of a heritage language. In no instance does this refer to a population formally linguistically educated in Spanish. The Mexican-American middle management, as well as the workers who have some fluency limitations in either English or in Spanish, argue that complaints about language express more a dissatisfaction with the services offered by the business than with linguistic shortfalls. With respect to the border dialects, the Mexican-American operatives feel completely legitimate in using the local variety of English with clients. Especially in the Chicano case, the imposition of an English accent different from their own would be considered as illegitimate for the team: as was repeated in the interviews, “in the United States, everyone has an accent.” Meanwhile, local Spanish, even with its specific features (discussed below), is regarded as sufficient to serve the Hispanic public of the whole United States. Anglo-American managers and Mexican-American operators share a vision of Spanish as a free, unspecific heritage language.
The relatively few employees who occupy positions as language professionals in the call centers are occupied by young people with high educational levels, largely done in Mexico with a university degree or credential from the United States. They consider themselves as the only “legitimate” Spanish speakers, standing out against the linguistic “poverty” of the area, which leads them to develop a critical discourse about the role of Spanish in the companies. Their duties in the various call centers include the recruitment and training, examination of linguistic abilities during recruitment, translation of scripts, and duties in small departments of translation and interpretation (such as translation into English of conversations between clients and operators with legal implications).
Those Mexicans with university-level studies are very critical of the competence in Spanish of the Mexican-Americans: “Few [Mexican-Americans] can pass a written entry examination.” Particularly, the extensive use of “Spanglish” (borrowed words, code-switching, etc.) on the part of that group is viewed as a speech form associated with a low cultural level. Differing from what occurs among Anglo-American managers and Mexican-American operators, they view Spanish as a technical competency, acquired through the educational system, with effort. As such, it should be paid for. These professionals claim their role at the moments of assuring linguistic quality of service.
Ideally, these language professionals would want the workforce to employ an anonymous or neutral language for the whole Hispanic population by using their Spanish scripts. In some call centers they elaborate lists of prohibited words for use with clients, explicitly prohibiting code-switching and words commonly used among US Hispanics (in place of “llanta, goma, tire o [or] rine, emplea [use] pneumático”).Footnote 6 However, this “verbal hygiene” remains far distant from being an integral style in communication practices (Cameron Reference Cameron1995, Reference Cameron2000:326). The absolute lack of control over the use of such guidelines (in the rare occasions when they exist) occurs within a conflict between Mexican-Americans (heritage language users) and Mexican professionals of language (as guardians of legitimate language). In the following quote, a Mexican employee indicates that she also has adopted these practices, so as not to humiliate the client.
I want the person to whom I am speaking always to come away with the impression that I am a professional. And a professional does not speak Spanglish. I don't know where we got this idea because probably a professional has to speak Spanglish. Now, let me tell you something. When you are an interpreter, let's say talking with a client, and you say to him “Sr. Gonzalez, aquí tenemos su factura. – ¿Qué es eso? – El cobro que le mandan, ¿Qué?, el bill.” [Mr. Gonzalez, here we have your bill (formal Spanish word). What is that? The statement we sent. What? The bill (English borrowing).] And you act as if you do not understand? Or “estaba tanqueando cuando le dieron el golpe. ¿Y qué es tanquear? Pues llenar de combustible.” [I was tanking up (English borrowing) when they hit it. And what is to tank up? Well, to fill up with gas.] Then we need to adapt to the needs of the client because we do not want to humiliate them. (Translator & interpreter, in-house/in-bound; born in Chihuahua, Mexico; thirty-five years old, female)
Mexican workers hold a linguistic ideology about “legitimate” educated Spanish and identify their Chicano colleagues as acculturated. Among Mexican-Americans, we observe an enormous variability in their abilities to communicate in Spanish. Nevertheless, US Spanish speakers, the largest part of whom are working class people of Mexican origin, do not perceive lower register, mainly oral Mexican-American Spanish as bad. A trainer describes the resulting situation below.
Spanish proficiency test. Supposedly someone passes it, but you see the quality of the employee. It is entirely an oral exam (…) Probably 20% would pass a written test, because now we have many employees from Juárez, now that the violence has begun. Those speak almost no English. Rather, they speak Spanish. They pass the Spanish test and more or less the English one, but when they take a call, they cannot speak English (…) but they remain. (Training manager, in-house/in-bound; Juarense origin; thirty-nine years old, female)
This complex statement proposes both that many Chicanos do not know formal written Spanish, and conversely recent migrants from Mexico speak almost no English. Yet her firm does not hesitate to employ people from both groups as “bilingual operators.” Overall, heritage Spanish and English are the dominant forms in the call centers, associated with cheap, flexible labor, speaking to customer markets with somewhat parallel linguistic practices.
Linguistic organization
The explicit linguistic philosophy of including Spanish in these services focuses, according to managers, in the satisfaction of and access to the Hispanic public as a factor of competitivness. Additionally, they adopt a paternalistic vision, indicating that these sorts of linguistic policies contribute to the defense of linguistic rights of consumers. In this regard, what type of consideration does the Hispanic minority in the United States receive in the call center services? Are the services focused on the “Americanness,” “Hispanicity,” or “Mexicanness” of the employees and of the Mexican-origin clients? As we have seen, the predominant approach is Spanish as a heritage language.
Managers adapt to the local ideologies and linguistic conflicts, transforming the technical competence for which the well-educated Mexicans advocate into “authentic” social competences justified in terms of ethnolinguistic politics within the United States. Faced with the practices observed by Poster (Reference Poster2007) with respect to identity management, the explicit advice from managers and supervisors to their employees having communication problems with their clients was to “tell them the truth, tell them that you are in a call center in El Paso, Texas.” The linguistic management of these centers focuses fundamentally on standardization of scripts, but with an absolute permissiveness about local border accents, as much for the Chicano as the Mexican varieties of Spanish and English.
From the first call centers established in the 1980s, service was offered in English and Spanish, especially directed at the final consumer in large consumption markets (telephone, gas, cable television, insurance). We only find as exceptions with monolingual English service or without guidelines for the use of SpanishFootnote 7 some small call centers or specific services within outsourced call centers, oriented toward business-to-business or specialized professions, precisely in keeping with the theory of the professional model of client satisfaction.
This is a case of an industry oriented to the national market, with few exceptions, as is typical in the majority of the sector in the United States (Batt et al. Reference Batt, Doellgast and Kwon2004). As such, the volume of calls handled is heavily weighted to English. Based on the data obtained for the nine centers we researched, only two to eight percent of calls are attended in Spanish. For example, one in-house-inbound center of a large insurance company with around 600 employees indicated that of six million calls attended in 2009, only 120,000 (2%) were in Spanish. An initial evaluation of this low volume of calls would lead us to think that a small Spanish department within the center would be sufficient to attend to these volumes, and there is no justification for why El Paso has a sector with 14,000 bilingual agents, especially since the largest segment is inbound centers where the majority of the calls arrive sorted by information systems and therefore could be directed efficiently to a workstation according the linguistic criteria. This type of strict linguistic division of labor is very far from reality, making the primary consideration linguistic flexibility of the operators.
We find this linguistic flexibility in different forms, as much standardized as dependent on the know-how of workers. In general, the base operations (the first line of attention to calls) can change flexibly between English and Spanish according to the instructions that appear on the monitor with each entering call, but always following a script in each of the languages. At immediately higher levels not all of the service information is in the form of a script, although normally it is for the most recurrent topics. For this, the employee translates freely from/into English, producing a continual code-switching without a pre-established guide, flowing between the two interlocutors, while consulting an order screen, the telephone menu, or a television screen. This is especially common with technical support workers. It is also the case for supervisors, who have the autonomy to “break some rules” and as such have nonstandardized conversations, while they negotiate with clients prices or conditions of products and services with the goal of obtaining a sale, in a way impossible for an operator. As a customer technical service representative points out: “Supervisors have to take calls from agents. For these reasons, they have to be bilingual” (our emphasis). The next section addresses these policies in a more systematic way.
Typologies of linguistic organization
To attend to the distribution of calls in Spanish and English, the linguistic organization of activity in El Paso call centers follows two different patterns: (i) all the workforce is linguistically flexible, or (ii) a part of the positions are monolingual in English and part are bilingual in English and Spanish. These two types suppose different models of linguistic management.
Call centers with 100% bilingual workstations
There are varied situations in which businesses use a 100% bilingual workforce to attend to all calls, from inbound-in-house centers to outbound-outsourced. In outbound centers, where the linguistic flexibility of the employee is key to productivity, a 100% bilingual workforce is essential. A paradigmatic example is survey research firms where the operator can only infer from first and last names the language(s) of the person being called. However, these suppositions can be offensive to the survey target, so that the operator needs to identify rapidly the preferred language of the interlocutor. There are no explicit instructions from management for this: intuition, mental agility, and empathy in redirecting the conversation to the language of the client are the tools the operator uses. It is worth noting that despite all of these required skills, the employees in this subsector are among the worst paid, with the highest turnover, and worst working conditions in the sector. These social skills, as well as knowledge of Spanish, are not recognized as worthy of compensation. The poor labor conditions and low wages make it especially difficult for these centers to recruit personnel with high levels of English, drawing mainly on Mexican-born personnel, Mexican-Americans with low education, and university students who seek temporary work.
Management as much as employees recognize that some of these workers do not have sufficient formal competence in Spanish and low ability to communicate in English. Despite this, they make calls in either one of the two languages. If there are irresolvable communicative problems in this type of firm, it is rare that the call is transferred to another station or to a supervisor, habitually generating conflicts with clients.
Besides the importance of linguistic flexibility in outbound call centers, it is worth noting in the following quote that, despite identity conflicts over accents that resemble those analyzed in the South Asia–United States call center literature, there is no indication of micromanagement of linguistic variables, and indeed a managerial policy of accepting inexpensive recent immigrants from Mexico into these low value centers.
What they did with me… was that they gave me a list of numbers. And I said OK. Here you have 200 numbers. You are going to work from 3 PM to 9 PM. You do questionnaires. I had my telephone, activated it with the headset in place, entered the number, waited for them to answer, and then I began my script. I had two screens. The E for English and the S for Spanish. And it had my script. I could not deviate at all. If the client responded with “Bueno” [standard Spanish telephone greeting], “Haló” [English borrowing, greeting in Spanish], then the S script. If the client speaks in English… Then in English. … I had to give an immediate response, English or Spanish. If I spoke in English and it seemed that they did not understand, I would make a swift change to Spanish. … The majority of Hispanics answered like my father. He answers “hello” but then cannot follow the conversation. Then I have to be very attentive and switch to Spanish. Then there are misunderstandings and angry moments with clients…but it was more because of the type of call I make than the language. Anglos, yes, get angry if I spoke to them in Spanish. They would say to me “you are talking to a New Mexican, why did you speak in Spanish?” I would just turn on a dime, they were very rapid decisions. … No one explained it to me. (Operator, outbound-outsourced; upper class from Mexico; twenty-eight years old, male)
With respect to the inbound call centers where all of the El Paso personnel occupy bilingual positions, we examined in-house operations of large corporations with thousands of employees. In these companies the linguistic division of labor is established between centers located in different parts of the country, with El Paso being the location or one of the locations where any call can be sent, whether in English or Spanish. In this division of labor El Paso does specialize in Spanish calls, but it also is expected to handle both languages flexibly; in reality, these centers deal with no more than 4% of calls in Spanish. Also, in these centers, despite better conditions of work and employment (e.g. health insurance), problems of linguistic deficits in English are observed, and problems derived from this in satisfaction of clients. Especially the centers with greater bilingual emphasis find that there is more and more competition to recruit linguistically qualified workers despite the superior reputation of the corporations and the relatively good working conditions in the centers.
They [workers] speak however they want. [What is the cause?] I think that since they [managers] do not want to pay more, people go to other jobs. And they want to hire personnel with the same pay, and it continues that way. And besides, it has a lot to do with the managers. For me at first they were very strict. To me and everyone who worked there. Before, Spanish was one of our strengths, but now… is terrible to see how they treat the client, don't care about them, don't have the same respect as for those who speak English. [describes scene] There are ten trainees. They are looking for an hour and 45 minutes at one of the best workers, for how to attend to clients… but I had not realized that his Spanish was so horrible. We never allow them to say “tú” [informal second person] to clients. Use “usted” [formal second person] and last name. Then… [in Spanish] “my name is Eric from the 24 hour Friendly service… what is your name [using tú and asking for first name]?” First call, never this “your [informal] first name.” And he said “What do you [informal] want? You [informal] want to speak to the agent [agent, the English word, not the Spanish agente]? Ah, well, look, your policy is not here [several words pronounced in working class style, pos, not pues, and nosta, not no está].” I had to say to the trainees not to follow this example. But it made me angry to see this… and it affects the relation between client and agent a lot. (Training manager, in-house/in-bound; Juarense origin; thirty-nine years old, female)
The previous excerpt with a critique of language usage is not part of a policy of linguistic fine-tuning; it is an internal criticism of actual practices when corporations accept informal, working class (heritage) Spanish as adequate to serve US Hispanic customers, performed by high-turnover, inexpensive “bilingual” labor.
The linguistic division of labor: Monolingual English positions and bilingual positions
In call centers where there is a linguistic division of labor, there are two types of positions: monolingual in English and bilingual. Monolingual Spanish positions are not found. This division of labor is more frequent in inbound-outsourced call centers. These provide diverse services for various business clients with differing degrees of interest in establishing a Spanish language service, although the majority are interested in bilingual service. The positions are customarily divided into four monolingual English positions for each bilingual one. The bilingual positions receive between 10% and 40% of their calls in Spanish, depending on the product (e.g. television services in Spanish) and areas they serve (regions with a Hispanic concentration, e.g. the southwest).
The bilingual workstations are generally occupied by workers with longer time in the organization. That is, they are judged only on experience, and this constitutes a type of compensation for seniority. In this fashion, to occupy a bilingual position is considered a merit within the organization. Historically, these positions have carried a bonus of between $.50 and $1.50 per hour, amounts that have not increased in recent years. However, various operators and managers told us that this policy is changing, and it is less and less common to receive this bonus. Particularly, one of the principal firms in this sector has established bureaucratic requirements for a court-issued official certificate based on an examination of knowledge of Spanish (costing $60) to receive the wage bonus. In other companies, the call centers have used examinations from private companies that accredit Spanish knowledge.
With a certain audacity on the part of management, while workers have not obtained this official certificate, they are still moved into bilingual positions, but without the bonus. The demand for official Spanish certificates eliminates opportunities for a good part of the Mexican-Americans, with important linguistic gaps at the formal level. Mexican-Americans consider the demand for a certificate as part of a process of reMexicanization, toward which they are disinclined through the various generations in the United States. The Mexicans for their part have show in some cases resistance to the idea that a business or US public authority (e.g. a state court) can ratify their knowledge of Spanish while Mexican credentials are not recognized. We found that the existence of wage bonuses for knowledge of Spanish and the establishment of linguistic credentials brings out latent conflicts in this community. One operator indicated that:
When I received a job offer, they told me that because I speak Spanish they would raise my wage from $8.25 to $9.25 an hour. I did an exam in English and Spanish to qualify as bilingual, but the company turned around and said to me, you have to be certified by the state to qualify as bilingual. Then you have to go to the court, pay $60, and do the exam. This I did not do. I don't want to pay to say that I am bilingual. And I did not do it. (Operator, inbound-outsourced; Mexican origin, university student; thirty-two years old, male)
Various operators and managers described the linguistic exams involved in initial entrance to the company. The processes are generally very informal: As an operator narrated: “They had me read a page in English and another in Spanish. To ensure I could read well in both languages. Then, if you can read, you can speak.” The interviews consistently note that central management considers that “if they [El Pasoans] understand Spanish, then they can sell in it.” A local economic development officer indicated that the managers “do not have a very good idea what is Spanish.” Likewise, the training periods are generally minimal, between one and fifteen days, though by the second or third day employees are already assisting clients and making sales. Distinctly linguistic training effectively does not exist, even in large in-house call centers within the most reputable companies.Footnote 8
In my folder I have two pages of etiquette in Spanish in a book of 400 pages. It is the last thing. The only exercise is a question that asks how you say motor in English. Please! They should put in a little etiquette! The client should be listened to! This is a training manual! (Trainer, in-house-inbound; Ciudad Juárez origin; thirty-two years old, female)
Concerning the “priority” of satisfying clients, companies shift Spanish-language calls to monolingual English workstations, depending on the volume of calls and the saturation of bilingual workstations, even though the employee may not necessarily be competent in Spanish. If there are difficulties, the monolingual operator can subsequently try to shift the call back to a bilingual position, although according to the criterion of minimizing time per call, this is something operators should avoid. The key elements behind this practice are reducing the time the client has to wait for a call to be answered. and adding to the flexibility among monolingual English stations during dead times, by contrast with considering the language choice made by the Spanish-speaking client. That is, the priority remains optimization of work time in the call center, not linguistic precision. Possibly the heritage bilingualism of many “monolingual” English operators in El Paso allows this imprecision to persist.
In summary, the philosophy of linguistic management in call centers and the wage bonus for knowledge of Spanish are tied closely to the linguistic division of labor. If there is no linguistic division of labor, there is no reason for a bonus. And these bonuses appear to be in a process of redefinition. Everything points to a new “cheap labor”: the maintenance of the Spanish language conceived as an instrument for access to the labor market, although it is less clear to what extent this affects promotion and better pay.
Conclusions
The study of call centers in El Paso provides important information for understanding the possibilities and limitations of the new information economy in cultivating language as a technical expertise. These large corporations have opted to manage Spanish as an authentic social attribute and, with some exceptions, not worthy of special compensation or development. This managerial conceptualization is consistent, on the one hand, with local ideologies (Achugar & Pessoa Reference Achugar and Pessoa2009; Velázquez Reference Velázquez2009) and, on the other, with neotaylorist management practices typical of mass market call centers, that consist of: (i) standardization (Milroy & Milroy Reference Milroy and Milroy1998; Cameron Reference Cameron2000:324), especially through routines and scripts that permit control (Leidner Reference Leidner1993; Poster Reference Poster2007), rotation of personnel, and low need for investing in training (Del Bono Reference Del Bono2000); (ii) permissiveness with local accents, as an “authentic” way of relating to the client (Heller Reference Heller2003, Reference Heller2010; Duchêne Reference Duchêne2009) that in turn permits access to an abundant, low wage workforce, especially in the case of languages with lower social prestige. The combination of these factors converts the cost of Spanish to these organizations to near zero.
Examining the linguistic practices in call centers permits a greater understanding of the contradictions between explicit managerial philosophies about Spanish (satisfaction and access for the Hispanic consumer; promotion of “Hispanic talent”) and actual business practices, fundamental to competition and productivity in the call center sector. These involve: (i) diverting Spanish language calls to English operators in order to reduce wait times; (ii) reducing the service time to handle calls by using Spanish because of the inefficiency in English of part of the Hispanic population; (iii) greater attention to the diversity implied by a large Hispanic population, but at zero cost, due to treating the Spanish language as an innate talent of local populations rather than as a technical competency; and (iv) in relation, preference for the authentic varieties of Spanish spoken on the border as a way to guarantee an abundant and cheap supply of bilingual workers, as opposed to other possible (trained, formalized) forms of Spanish.
These practices, considered in the theoretical debate about the role of language in the new economy, indicate to us that “authenticity” not only can involve a patrimony that makes local cultures profitable, but also for managers can be an optimal form of linguistic management of “disposable workers,” even with greater communicative intensity of work. The novelty that is diversity is accepted in organizational management, but fundamentally as a “heritage,” enculturated and informal. This coincides with and reinforces Mexican-American community tendencies to consider it unnecessary to formally educate their children in Spanish (Velázquez Reference Velázquez2009). As such, this type of linguistic management in the new economy brings us to question the role of the market in the revaluation of Spanish in the United States. The new situation that comes from the management of Spanish “at zero cost” appear to favor a greater respect and inclusion of the Hispanic population via business recognition of their heritage, but not a social and economic advancement of the group based on Spanish as a rewarded technical expertise. The linguistic intensity of the new information economy does not in all cases promote a focused technical expertise approach to language, in business or community.
