The Use of Languages by Valencian Business-to-Consumer Service Companies in their Telematic Communications

This study analyses the languages used by companies in their telematic means (website, social media, etc.) in a specific geographical area, the Valencian Community (Spain), where two official languages (Catalan and Spanish) coexist and where there is a large influx of foreign tourists (mainly British). More specifically, the aim of the study is to understand the weight each language has in companies’ telematic communication. The theoretical approach used for the analysis is the centre-periphery model, which is used to analyse national identity and language. Different techniques (use of secondary sources, mystery shopping, content analysis and direct observation) have been used in the quantitative empirical study to obtain a statistically disaggregated data matrix. The results strongly emphasize the peripheral and marginal position of Catalan in this region, and, on an international scale, the resistance of Spanish, which clearly maintains its hegemonic position over English in the telematic communications of these companies.

regions (there could be a partial regulation in some companies. See Lüdi et al. 2013). In fact, in previous studies, during which we analyzed the use of languages in telematic communications of the Valencian local governments (Català-Oltra/Penalva-Verdú 2019a, 2019b), there was a regulation in force -Llei 4/1983, de 23 de novembre, d'Us i Ensenyament del Valencià 3 (Valencian Government)-which establishes the use of the two official languages of the Valencian Community (Castilian or Spanish, and Valencian or Catalan) 4 by the local government in those areas within the 'Valencian linguistic domain '. 5 Partly because of this, the use of this dialectal variant of Catalan in the local government of these areas is relatively high. Valencian companies, as we have pointed out, do not have this obligation (as Catalan companies partially do 6 ), but one would expect them to adapt their website and social media content and their linguistic policy accordingly in areas where there is a high percentage of customers who use the native language of the region as their everyday language. The study reported on in this this article has provided us with data on this matter, along with data regarding the use of foreign languages in an area where tourism has a significant impact on the economy, in large part thanks to important European holiday destinations such as Benidorm and the surrounding county. This information is highly useful to outline public policies in terms of linguistics in a context of valorization of multilingualism.
The general aim of this study was to identify which languages are used (especially official ones) by Valencian business to consumer 7 service companies in their telematic communications. This included determining to what degree different languages are used on the websites of these companies, especially on the homepage, identifying the initial response language of the switchboard, determining the response language to a query made in Valencian via email or a website form, analyzing the use of languages on the social media profiles (Twitter, Facebook, etc.) of these companies and, additionally, collating results according to the linguistic domain, territory, number of employees, share capital, environment (rural/urban) and activity, in order to obtain a profile of those companies that choose to express themselves in Valencian or in foreign languages. But, first, we must address two aspects of the reference framework: the use of languages in companies and, specifically, in their telematic means, and the centre-periphery dynamics in relation to languages in a globalized economy.
We are well aware of the fact that, for a while now, language has become a field of interest for business research, among others, because it is an important part of the content of websites (Tzikopoulos et al. 2007; quoted in Singh et al. 2008: 163). Companies and the economic sphere have also become a focus of linguistics studies (Lüdi et al. 2013(Lüdi et al. , 2016Varela 2015: 132-133;Block 2014: 24-26). And even more so considering that corporate websites are more complex than institutional ones when considering minority languages (Kalganova 2014: 1220), as we have found when comparing the results of this study with the previous research already mentioned (Català-Oltra/Penalva-Verdú 2019a, 2019b. The importance of language in company-consumer service relationships has been pointed out by Holmqvis/Grönroos (2012: 431), based on the fact that more than half of the world's consumers speak more than one language. Thus, markets are less nation-state focused and, therefore, it is less only offered information in English, in some cases in their telematic means (Martinez Normand et al. 2014:72). 3 http://www.dogv.gva.es/portal/ficha_disposicion.jsp?id=26&sig=0428/1983&url_lista= (visited on 2/10/2019). 4 'Valencian' is the dialect of Catalan spoken in the Valencian Community and the denomination enshrined by the Statute of Autonomy of this region. We will use this denomination preferentially, although we also refer to the language as 'Catalan' or 'dialect of Catalan'. 5 Valencian linguistic legislation divided the territory into two parts (see Figure 1, Section 3.3), one where Valencian had not been spoken historically or was no longer spoken at all ('Castilian linguistic domain') and another where Valencian was still spoken at the time the law was drafted in 1983 even if it was just by a minority of inhabitants ('Valencian linguistic domain'). The 'Valencian linguistic domain' includes about three quarters of the municipalities and 87% of the population of the Valencian Community. Therefore, in addition to being the geographical area where the legal framework has a more obvious impact, it is widely representative of the whole region in demographic and territorial terms. 6 See https://llengua.gencat.cat/ca/serveis/empresa_/marc-legal/pmf-empreses/ (visited on 3/12/2020). 7 Business-to-Consumer (B2C) is used to name the set of strategies (usually accompanied by technical architecture and software) that allow companies to interact with their potential or actual customers. likely that the consumer will speak the same language as a company. In B2C service companies, this is even more important as the way in which interaction is developed will determine how the consumer perceives the quality of the service and this greatly depends on using a language that the consumer can understand: In studies that connect language use to service perceptions, (...) research has shown that consumers prefer using their native language (Puntoni et al. 2009) and that even consumers who are fluent in more than one language expressly indicate that the use of their native language influences their perceptions of service providers (Holmqvist 2011). These findings reinforce the relevance of language for service encounters (Holmqvist/Grönroos 2012: 431). Therefore, it is relevant to know if companies address the linguistic dimension of their service when in a bilingual environment, such as the one at hand. Even more so when considering that the minority language of the Valencian territory is officially used and there are areas where it is widely spoken.
Existing studies on customer service departments and corporate websites have addressed intonation, nonverbal language, meanings, types of messages, etc. and not so much the actual languages used (Varela 2015: 135;Yanaprasart et al. 2013: 137). Nevertheless, we have come across qualitative studies on the presence of Spanish on corporate websites in the US (Singh et al. 2008) and on the use of working languages in Basque companies (Van der Worp et al. 2017) and legal studies on the use of Tatar over Russian in Tatarstan economic sphere (Kalganova 2014). We have also found quantitative studies along the same lines as our subject at hand which analyse the use of languages on the websites of Ukrainian companies (Ugolkov/Karyy 2017), of multinationals ) and of the customer service departments of Galician companies (Varela 2015).
However, there are very few quantitative studies that systematize data on the use of languages in telematic means with a statistically representative sample of companies from a specific territorial area. The most similar prior study is one that focuses on Catalan, in fact, the Estudi ELAN.cat de 2009 (Marí et al. 2010: 142), which is not a systematic study, however, but a limited survey of companies in Catalonia with between 10 to 249 workers (small and medium-sized) carried out by the Open University of Catalonia, the University of South Wales and Catalan government agencies. This survey addresses the languages used in communication in internal and external markets, including the languages in which the websites are written. A few years earlier, the Secretariat de Política Lingüística (2009) had carried out a survey of service companies that also included the website language (revealing that Spanish was used predominantly over Catalan on the corporate websites of these companies, in contrast to the languages used in their advertising and the signage of their establishments). In any case, obtaining this information via a survey results in an important bias, since it implies an indirect observation that assumes the truthfulness of the verbally transmitted information. In contrast, in our case, the information strictly reflects the reality of the companies at the time of the field work, since we were able to check the languages that each of the websites were written in by going to the URLs.

Central and Peripheral Languages in Global Capitalism: Pride and Profit
With the emergence of the structuralist economic theories developed by the ECLAC (United Nations Economic Commission for Latin America and the Caribbean) in the 1950s, the centreperiphery structure became established as an explanatory model of relations in a world in transition towards a globalized economy (Lampa 2019;Bianchi/Salviano 1999). In summary, this model states that there are central countries that dominate and determine the relations of other peripheral countries, which are dependent on the former, in many cases as a result of colonization. This vision in which hegemonies and subordinations occur worldwide was subsequently transferred to identity studies (Rokkan/Urwin 1983;Letamendia 1997) and also to sociolinguistics (De Swaan 1998;Heller 2013;Morgenthaler 2008: 48-49;Woolard 2018: 115-118). 8 There is essentially some agreement about the presence of a world system which fits this centre-periphery structure and which can be extrapolated to nations, to States and even to languages. Everyone's position is structurally and historically defined in terms of their ability to exercise control over, first, their immediate area of influence and, progressively, over areas of greater scope that culminate in the entire world. Applying this to the case of languages and to our geographical and linguistic unit of analysis, Valencian can be considered a peripheral dialect in the context of Catalan, in which Eastern Catalan, which is the form spoken in Barcelona, would be considered the central language. However, Catalan, including the form spoken in Barcelona, is peripheral in relation to Spanish within the geographical framework of Spain. And, beyond that, in the European and global context, Spanish is peripheral in relation to English (in relation to which Catalan also is peripheral in this larger frame of reference). However, Spanish would still be considered central in the regional context of Latin America, as a result of the colonization that began in the 15th century, which established a relationship of dependence of the American territories on the metropolis of Spain. These types of issues condition the entrenchment of a language in a given territory or environment and its present and future development. A diglossia often exists in these cases in which, over time, the central language or dialect is reinforced to the detriment of the peripheral one, unless its speakers react adopting a discourse of pride and identity (Heller/Duchêne 2012) with subsequent measures to contain its decline (De Swaan 1998).
Today's (globalized) stage of capitalism implies significant changes in different fields (Piqueras 2017;Streeck 2014), including linguistics. Heller/Duchêne (2012: 3-4) point to the interaction between two types of discourse in relation to language, pride and profit. The first was dominant at the start of capitalism, which coincided with the creation of modern nation states and other nations, which employed the (national) language as a tool for consolidating an ethno-national project and its corresponding collective identity. This type of discourse or way of proceeding can also be related to the defensive reactions that are taking place locally in the face of globalization (Kroon/Swanenberg 2019). Meanwhile, the second, the discourse of profit, has emerged in the current era of globalization, 9 as a result of various changes: market saturation (which requires lower prices, differentiation, customization, etc.); partly because of the above, expansion in the search for new markets; also differentiation or added value, to compete in that saturated market; outsourcing (and this involves information and, therefore, language playing a more important role); and the flexibility of companies and workers to adapt to an increasingly competitive and changing environment (Heller/Duchêne 2012: 8-9). In this context, discourse and actions based on profit have gained in importance to compete with the discourse of identity pride. For example, content may be presented in a minority and peripheral language not so much to affirm an identity, but because it can be a way of connecting with a new niche market in a very competitive environment. In other cases, both discourses are interwoven and content may be presented in one language for both reasons (pride and profit).
Our quantitative and observational study will not address the construction of discourses, but it is worth bearing in mind that these may be at the root of the decisions made regarding languages. For example, in the Ukraine prior to the 2014 armed conflict with Russia, Russian was the dominant language on company websites because the discourse of profit prevailed (there are many people who only speak Russian and those who speak Ukrainian also often know Russian). However, since the outbreak of the conflict, identity has become more important and, in the western part, where there are more Ukrainian speakers, the presence of Ukrainian on websites has increased, even more so as a result of the campaign to boycott Russian products (Ugolkov/Karyy 2017: 37-41). This indicates that the discourse of pride is interwoven with that of profit (intertwined discourses can also be seen with the speakers of Irish in Brennan/O'Rourke 2019). In the case of Catalonia, where there is a relatively high percentage of individuals who feel exclusively Catalan and can be said to have a militant attitude (Colomer 2018: 280), 10 it is understandable that identity pride is the motivating factor behind the choices made by business leaders with respect to language, although another factor in play is that, with a broad target of consumers sensitive to the identitylinguistic issue, these same companies see a market opportunity (profit) to respond to the need for information in Catalan. On the other hand, in the Valencian Community, our geographical scope of analysis, where the exclusively Valencian identity is not felt as strongly (the Spanish-Valencian dual identity is clearly dominant and, secondly, the exclusively Spanish identity), 11 one would expect the actions of company leaders, in relation to language, to be driven by the discourse of profit, which may or may not imply offering content and services in Valencian (depending on the expected profit). Whether motivated by profit or pride, one would expect to find a relatively high use of Valencian/Catalan or at least a greater use on the websites of service companies in specific areas where this language is spoken more.
Indeed the clear predominance of the discourse of profit in times of globalization is hampering the advance of minority languages, almost all of them with a limited territorial scope, as is the case of Catalan. The minorization (and, in some cases, extinction) process of languages occurs "in bilingual or multilingual contexts in which a majority language -that is, a language with greater political power, privilege, and social prestige -comes to replace the range and functions of a minority language" (May 2006: 257-258; see also Vila 2014: 59;Léglise/Alby 2006: 67-68). This has intensified with the spread of ICT, which are the subject of our study. This "digital diglossia" (Melero 2018: 154) has prompted Internet and social media users to opt for those languages that provide more information on the net. Ten hegemonic languages 12 make up 80% of the content of the world wide web. Although English has lost ground in the last two decades to the other nine, over half of the information is still in this language. However languages that are not part of this group are more difficult to be established (Melero 2018: 154-157;Marí et al. 2010: 140;Boix-Fuster 2010: 166) and it is likely that companies will use them to a lesser extent.
English is undoubtedly the lingua franca 13 (and central) language in the academic world, but also in the business world (Martínez Normand et al. 2014: 72;Janssens/Steyaert 2014: 623;Gal 2012;Marí et al. 2010: 140). A systematic study carried out by more than 200 experts on thirty European languages and the degree of technological support available in each of them (translators, dictionaries, linguistic corpora, speech recognition technologies, etc.), showed that only English is properly supported, though Catalan is the non-state language that scores highest and also ranks among the highest of the minority languages (though far behind Spanish, French, German, etc.) (Rehm/Uszkoreit 2012; quoted in Melero 2018: 161-162). Already in 1997, shortly before The Oslo Recommendations regarding the Linguistic Rights of National Minorities (OSCE 1998), the High Commissioner on National Minorities of the Organization for Security and Co-operation in Europe-OSCE "was already observing (…) limitations on the choice of language used by private enterprises could affect the ability of minorities to fully participate in economic life, which in turn could increase potential for conflict" (Croft 2018: 53).
Regardless, despite the pervasiveness of English and other hegemonic languages, companies are finding it necessary to adapt linguistically to the territories where they carry out their activity (Brennen et al. 2017: 140), not just by translating content, but by doing so with sufficient quality, taking into account the language's own rules and sociocultural context, so that the  (Melero 2018:157). 13 Although we are aware that English as a lingua franca can be seen as a multilingual phenomenon itself since one can find layers of other languages in it (Lüdi et al. 2013). reader/consumer does not perceive it to be a translation (Nantel/Glaser 2008: 112-114;Martínez Normand et al. 2014: 70;Marí et al. 2010: 140;Yanaprasart et al. 2013: 138). Even more so if the language is understood more from an emotional perspective and, to a lesser extent, as a functional element (Holmqvist et al. 2014). Singh/Pereira (2005;quoted in Yanaprasart et al. 2013: 142) have observed five degrees of localization, referring to the websites of multinationals or internationalized companies: "'standardized' (one website for all countries), 'semi-localized' (one site gives information on many countries), 'localized' (a whole translated site for each country), 'highly localized' (translation plus country-specific adaptations), and 'culturally customized' (a new site completely immersed in the target culture)". Naturally, the customer expects a website to be 'culturally customized' or, at the very least, 'highly localized'. In the case of local companies that operate in the Valencian Community, this is relatively simple, since there is usually no need for any cultural adaptation and the only action required is a competent translation from Spanish, which is closely related to Catalan. But, as we will see, in most of the Valencian service companies these actions have not been carried out, even the undertaking of a simple translation (regardless of quality), hypothetically because of the prevalence of the logic of profit, which considers that there is no significant demand for this among customers and, therefore, it is not worth wasting resources on localization. In some public administrations in the territory (at the autonomous community/ regional level and even in a significant number of town councils), the sense of identity pride prevails and, therefore, the use of Valencian in their telematic means is more widespread. This is also because it is stipulated by law, drafted under that principle and, largely, as a result of pressure from civil society organizations whose focus is on language and identity. This is the case of the current Plataforma per la Llengua in the Catalan-speaking territories. 14 In Europe, the European Charter for Regional or Minority Languages was drafted in 1992 (Kalganova 2014: 1220) signed by practically every country and ratified by most of them, including Spain. Indeed, one of the languages protected by this document is Catalan and its dialects. One of the action areas included in the Charter is 'Economic and social life' and the objective is to eliminate prohibitions on the use of minority languages in the public documents and information of companies and, beyond that, to protect the use of these languages in the economic sphere. A few years later, the Oslo Recommendations (OSCE 1998) also included a section on 'Economic life', with recommendation number 12 which stated that: "All persons, including persons belonging to national minorities, have the right to operate private enterprises in the language or languages of their choice" (quoted in Croft 2018: 52). However, in light of the special difficulties faced by minority languages as a result of the spread of ICT, in 2017, the European Parliament presented a report entitled Language equality in the digital age, which "analyzed the status of technological inequality between languages in Europe and discussed the challenge that this poses for the single digital market and for equality among citizens" (Melero 2018: 162-163).
Legislation on the protection, use and promotion of languages in Spain is the responsibility of the autonomous communities and, in this regard, the Valencian Community is governed by the Law 4/1983 on use and teaching of Valencian. But the use of Catalan in the Valencian Community, which was already in steady decline compared with Catalonia at the beginning of the new democratic era in which many powers were delegated, is now even more diminished due to the inaction of governments prior to 2015, notably in the long period  in which the centre-right was in power (Català-Oltra/Penalva-Verdú 2019b; Pradilla 2017). Catalonia has a tradition of forty years of Catalan nationalist governments, which pursue active policies to promote the language in different fields, while, in the case of the Valencian Community, these lines of action have been implemented with less intensity and more intermittently. Specifically, in the economic field, in recent years under a progressive government supportive of the entrenchment of the language, only actions with a limited scope have been taken to promote the language and there have certainly been no interventions that stipulate, for example, the use of Valencian in the public information of companies.

Sampling Procedure
The sample universe of this study is service companies whose activity usually involves a direct business-consumer relationship (what is known as B2C in marketing terminology). Under these conceptual premises, we went to the Central Companies Directory (DIRCE in its Spanish initials) of the Spanish National Statistics Institute to determine the sample number and the basic structure for obtaining the sample. The first step was to decide, using the 2009 National Classification of Economic Activities (CNAE in Spanish), which activities were part of the universe. After analyzing all the categories we ended up selecting a set of 17 divisions of activities. 15 The DIRCE also allows segmentation by number of employees, since this number is recorded for every company. This allowed us to classify the units into micro (0 to 9 employees), small (10 to 50 employees), medium (50 to 250) and large (more than 250) enterprises.
For the activities indicated above, we found 160,608 companies out of a total of 350,065 business entities in the Valencian Community in 2017. That is, 45.9%. Since we wanted to ensure a maximum margin of error of ± 5% (for a confidence level of 95.5%), this gave us a figure of around 400; 17 activities divided into 4 categories of number of employees imply 68 segments. If we had to distribute 400 interviews across 68 segments, many of them, due to their small sample volume, would not be represented. This fact, and the high fragmentation of the sample, would make compliance with the quotas very difficult. For this reason we decided to group activities together resulting in fewer categories, a total of 7 (in brackets, CNAE-2009 codes of the activities that are part of each grouped category): This meant that there was still a very unbalanced structure, with a sample basically of micro companies (97.7%) and activities related in one way or another to tourism (retail, hospitality, real estate etc., which added up to 68.6% of the total). Although we understand that this is the reality of Valencian companies, it did not make sense to make a strictly proportional sample, because this would not provide us with enough cases to understand what is happening with medium or large companies, nor to analyze certain activities, such as cultural ones. This led us to a nonproportional allocation in which these underrepresented categories were given more weight than they actually have. Therefore, we ensured that the final sample included 50 companies for each category of number of employees and 40 per activity, and we distributed the rest proportionally.
Once the overall results were obtained, the weighting coefficients were applied to ensure the sample was proportional to the reality of Valencian service companies. Table 1 shows the technical sheet of our empirical study.

Data Collection Techniques
The sample was used to apply less reactive techniques compared with a survey. Therefore, different quantitative techniques were combined to obtain all the information that responded to the objectives: • Use of secondary sources: these were mainly the 'explanatory' variables (number of employees, share capital, territory, linguistic domain, etc.) which derived from databases such as SABI or the DIRCE. These databases also allowed us to structure the population and the sample, which is the case with DIRCE, and SABI has worked as a sampling frame. Both databases are the most common in Spain when working with companies as a unit of analysis (see, for example, Larrán et al. 2010: 67 and Camisón-Haba/González-Cruz 2020: 5).
• Content analysis: we visited websites and social media profiles of the companies that use these resources to obtain different indicators (official website home page language, website languages, languages used in social media communications, etc.) in the same way we did previously (Català-Oltra/Penalva-Verdú 2019a, 2019b) and as Callahan/Herring (2012) did with a sample of universities.
• 'Mystery shopping': this participant observation technique used in market research is a technique in which the researcher or field worker adopts the role of a random consumer and simulates a request for information to then record the codified response. This technique was applied to telephone calls and emails, in which we asked a straightforward question common to every company to see what language was used in the interaction. We have assured data protection and the anonymity of the responses (no enterprise is identified with a specific answer), and it was a single and banal question that did not imply an evaluation of workers or enterprises, as occurs with regular mystery shopping. Therefore, we are not faced with ethical problems in the way that Jacob et al. (2018) explain or as seen in the Guidelines for Mystery Shopping of the Mystery Shopping Providers Association. 16 The application of these techniques for gathering information provided data to generate a single matrix with different indicators for the 397 companies.

Zoning
Some of the explanatory variables are territory-specific although there are sociolinguistic, administrative or socio-urban factors that are at the root of the resulting spatial pattern. These variables are 'zone' (based on linguistic criteria 17 and urban dynamics 18 ), 'province' (which is the level that in the Nomenclature of territorial units for statistics corresponds to NUTS-3) 19 and 'linguistic domain' (based on Law 4/1983 on the use and teaching of Valencian). 20 All these delimitations appear on the map below: Figure 1: Marking out of zones, provinces, (based on linguistic and urban dynamic criteria) and linguistic domains according to the regional Law 4/1983.  The differences in the use of Valencian in telematic means between service companies and Valencian town councils are very significant. As a general summary, the results indicate that the use of Valencian by companies is residual whereas it is used by the vast majority of town councils, around two thirds. 21 The only cases in which the percentage of use of Valencian by companies worthy of mention are when the potential customers have expressed themselves in this language and, in these cases (by email and telephone), only just over a third of the companies respond accordingly. In other words, about two thirds of the customers of Valencian service companies that use Valencian to address the companies do not receive a response in their language. Moreover, there is a higher proportion of companies than town councils whose responses in Valencian to emails contain errors (58.9%, compared to 37.6% of town councils), confirming that these companies have little or no interest in training their personnel in the local language of the Valencian Community. With regards to email, we also noted that just under half of the companies (49.5%) did not even answer when they received an email in Valencian. Of those that answered, as we already mentioned, about a third responded in Valencian. However, if we analyse this information against the total number of companies, only one-sixth of them respond to an email expressed in Valencian in that same language. And, what is more, those that answered with no linguistic errors made up only 6.8% of the total. Therefore, the companies that do provide at least a 'highly localized' answer, referring back to the typology described by Singh and Pereira (2005), make up a very small proportion. And this is despite the fact that we are talking about companies with Valencian share capital that operate in a bilingual region and not about multinationals or companies with registered offices located in other areas of Spain. With regards to telephone conversations, it is not just that half the number of companies compared to town councils (34.6% versus 69.4%) respond accordingly when the caller speaks in Valencian, but also that Valencian is rarely the initial language used by businesses when they pick up the telephone (3.6%). In this first step of telephone interaction, the vast majority of town councils use Valencian. Therefore, in town councils there is hardly any increase in the proportion of public employees who use Valencian after hearing the caller speak in that language, because they are already accustomed or committed to do so. This is not the case for customer service employees in companies, since practically 90% of those who end up speaking in Valencian do so in the second instance, in order to adapt to the caller. In town councils, it is the other way around: 87.3% of those who end up speaking Valencian on the phone do so from the first moment.  If Valencian is far less used by companies than by town councils, the opposite occurs with English. In all cases, the figures are very low, always below 15%, but there are more than twice as many companies as town councils whose website content is in English, a few tenths more on Facebook and more than six times more on Twitter (9.8 vs. 1.6%). The Valencian economy's focus on tourism and exports is at the heart of these results, which, in any case, do not reveal a high use of English, even though the level of synchronicity of the media used in communication is low (Varhelahti et al. 2017: 174). This is partly because the vast majority of companies are micro-SMEs.

Bearing in mind that
The website is the communication modality most used by companies (50.7% of the companies have one) and the distribution of languages in this modality is as follows: Adding together the categories for 'All' and 'More than half' of the content written in the language in question, we can summarize the use of the different languages on the websites of companies as follows: 99.2% are in Spanish while only 14.1% are in English, 7.7% in French and 6.7% in German. Russian and other foreign languages register marginal percentages, but, in the case of Valencian, the figures are even lower, because only 1% of websites are in the native language of the region.
In the case of Catalonia, a 2010 sample of companies with 10 to 250 employees, the percentage of websites in Spanish was 85. 3%, 63.5% in Catalan, 42.4% in English, 15.8% in French and 6.6% in German (UOC-GenCat 2010). 22 If we focus on Valencian companies of this size in our sample, the distribution would be 100% Spanish, 4.6% Catalan, 19.8% English, 9% French, and 5.3% German. Therefore, whether in the use of the local language or foreign languages, Valencian companies show little regard for adapting to the language of the potential customer.
Our presentation and descriptive analysis of the results continues with a bivariable reading of the use of Valencian and English based on different explanatory variables, including the size of the company (number of employees), the environment (rural/urban), geographical variables (zone, province and linguistic domain), economic activity, and share capital.
Regarding the use of Valencian based on these variables, the aforementioned overall (very low) percentages prevent us from drawing conclusions worthy of consideration for most telematic means. However, taking into account telephone and email interaction, which are the communication methods used by around a third of the companies that express themselves in Valencian, we can start to draw up a profile of the service companies that use this Catalan dialect in their telematic interaction with customers:  The companies that use more Valencian in their interaction with customers through telematic means are, logically, those that have their registered offices in the Valencian linguistic domain, those which are not located in an urban environment, those from the Valencian Central Counties (south of Valencia and north of Alicante), and the Valencian-speaking zone of Castelló, with medium-low share capital, offering personal services and workshops in the telephone category, and culture/sports/games in the email category. But only the semi-urban environment and workshops in the telephone category, the rural environment and services related to culture/sports/games in the email category, and, in general, the indicated geographical zones register medium percentages worth mentioning (around 50% or more). In any case, and, as observed using the Chi-Squared statistic, only geographical variables are associated with the use of Valencian in companies.  The size of the company and its share capital have a direct correlation with the use of English on the company website. A relatively high percentage of medium and large companies and those that exceed 50,000 euros of share capital (therefore with more resources than those with fewer employees or less share capital) have their websites in English, 50% in the case of large companies. There is not such a clear correlation between these two variables and the two most widespread social networks, Facebook and Twitter. The two zones with the highest regular influx of foreign tourists (mostly British), the Valencian Central Counties (especially Benidorm and its surroundings, the Marina) and Vega Baja, contain the highest proportion of companies with websites and social networks in English, with figures that more than double the average of Valencian companies as a whole. Finally, the activity of a company is also a decisive factor in the use of English in the telematic means of companies and, in this regard, the real estate sector, whose target market is mostly foreigners, stands out especially. The figures for workshops, some of the highest in terms of the use of English on websites and Facebook, can be explained by the links their websites have to the car companies of which they are dealers or partners. On the other hand, the relatively low figures for the hospitality and tourism industry stem from the inclusion of traditional bars throughout the autonomous region that either do not have regular foreign customers or do not have the resources to pay for decent translations (in any case, if we take hotels in isolation, practically all of them use English in their telematic means). Finally, the healthcare sector also stands out in social networks, as a result of the high numbers of elderly tourists who reside in the region for part of the year, especially in coastal areas of the province of Alicante.
On the whole, whereas for Valencian the results are explained by more territorial aspects, in the case of English, the characteristics of the company (its size, capacity, activity, etc.) carry more weight, although some territorial particularities to do with the productive activity of those zones are also a factor. Therefore, the companies that use English the most in the content of their telematic means are those which are largest (only in the case of websites), those with the highest share capital, those in the real estate sector, and those located in coastal areas with a large influx of foreigners, in what can be inferred as following the logic of profit in the terms described at the beginning of this study.

Conclusions
Consequently and adopting the approach by Heller/Duchêne (2012), identity pride is not a motivating factor behind the use of Valencian in telematic means, but neither is profit as evidenced by the marginal number of companies that offer their telematic content in this language. In short, there seems to be no motivation behind the use of the region's native language despite the fact that in certain areas (Central Counties, second ring of metropolitan Valencia, northern Castelló) more than half of the population usually speaks Valencian in their immediate environment, 23 and these are results that runs counter one of our hypothesis. Thus, these companies are not taking into account what different studies (Holmqvist/Grönroos 2012) understand as one of the bases for delivering a good service: adapting to the consumer's native language.
Valencian is only marginally used on websites, to the point that its percentage (1%) is lower than that of Russian. This contrasts with the relatively prominent presence of English (14%), German and French (both about 7%). The case of social media and answering machines is similar. Only for situations where the potential customers are expressing themselves in Valencian (by telephone and email), are the numbers something more than just residual (around 33%), although they are far from most of the cases. Therefore, Valencian companies do not consider it necessary to adapt linguistically to the customer, unlike Catalan companies do (UOC-GenCat 2010). The telematic environment is one area that reflects this situation more clearly, to the extent that the 'digital diglossia' which we referred to at the beginning (Melero 2018) is a prominent aspect of the Valencian linguistic ecosystem.
On the other hand, we have seen that Spanish holds out strongly against English, with its use very much entrenched in the telematic means of companies in an autonomous community with two official languages, while English has barely penetrated the websites and social media of local companies. The centre-periphery model, in this case, is reconfigured by the strength of a language that is central in its national context and that, in addition, has a notable weight in the international context. In short, Spanish manages to impose its central position on Valencian, and manages to contain the expansion of English in the business world.
In any case, the representative sample of Valencian-owned service sector companies that have a direct relationship with the end consumer is very much conditioned by the fact that the vast majority are micro-enterprises. This has a significant impact on the final results because these companies are mostly made up of freelancers and small businesses that do not usually have telematic communications technology and platforms available to them, nor do they have sufficient resources to translate content into different languages. In the case of English, to a certain extent, the results vary in three segments: larger companies (in volume of workers and/or share capital), those whose activities have foreigners as the target market, and companies located in zones that attract most tourists.