Discursive media representations on telecommunications privatization

Abstract

This paper analyzes the discursive representation of the hegemonic written media on the privatization of telecommunications in Brazil. Since the country’s political re-democratization, the privatization of state-owned companies has remained on the agenda of governments from different ideological positions. Assuming the mainstream media as one of the leading influencers of public opinion, we analyzed a corpus of 869 articles from widely circulated newspapers and compared it to 344 articles in pro and anti-privatization media. The quantitative and qualitative analysis showed how the press restricts the lexical field of privatization to companies and businesses, suggesting the discourse that telecommunications is like any other business, which renders its discussion with society unnecessary. The analysis of the main collocations also showed how privatization has been represented as a historic landmark. The research contributes in two ways to public administration and organizational studies: it brings materiality to the discussion about the representations of privatization and presents a methodology of discursive analysis based on extensive collection of documents.

Keywords: privatization; telecommunications; discourse; media; corpus linguistics

Resumo

O artigo analisa a representação discursiva da mídia escrita hegemônica sobre a privatização das telecomunicações no Brasil. Desde a redemocratização política no país, a privatização de empresas estatais permanece na pauta de governos de diferentes posições ideológicas. Assumindo a grande imprensa como uma das principais influenciadoras da opinião pública, analisamos um corpus de 869 artigos de jornais de grande circulação e o comparamos a 344 artigos de mídias pró e antiprivatização. A análise, quantitativa e qualitativa, mostrou como a imprensa restringe o campo lexical das privatizações às empresas e aos negócios, sugerindo o discurso de que as telecomunicações são como qualquer outro negócio e tornando desnecessária sua discussão com a sociedade. A análise das principais colocações também mostrou como a privatização vem sendo representada como um marco histórico. A pesquisa contribui de duas maneiras para a administração pública e para os estudos organizacionais: traz uma materialidade para a discussão acerca das representações da privatização e apresenta uma metodologia de análise discursiva com base em grandes coleções de documentos.

Palavras-chave: privatização; telecomunicações; discurso; mídia; corpus linguistics

Resumen

El artículo analiza la representación discursiva de los medios escritos hegemónicos sobre la privatización de las telecomunicaciones en Brasil. Desde la redemocratización política del país, la privatización de las empresas estatales ha permanecido en la agenda de los gobiernos de diferentes posiciones ideológicas. Asumiendo que la prensa convencional es uno de los principales influenciadores de la opinión pública, analizamos un corpus de 869 artículos de periódicos de amplia circulación y lo comparamos con 344 artículos de medios pro y antiprivatización. El análisis, cuantitativo y cualitativo, mostró cómo la prensa restringe el campo léxico de la privatización a empresas y negocios, sugiriendo el discurso de que las telecomunicaciones son como cualquier otro negocio y haciendo innecesaria su discusión con la sociedad. El análisis de las principales colocaciones también mostró cómo la privatización ha sido representada como un hito histórico. La investigación contribuye de dos maneras a la administración pública y a los estudios organizacionales: aporta materialidad a la discusión sobre las representaciones de la privatización y presenta una metodología de análisis discursivo basada en grandes colecciones de documentos.

Palabras clave: privatización; telecomunicaciones; discurso; medios de comunicación; lingüística de corpus

1. INTRODUCTION

Debates on privatization, defined here as the use of the private sector to provide a public good or service, has remained on the Brazilian government’s agenda since before the country’s re-democratization in 1985, as shown in the main milestones by presidential mandate detailed in Box 1. Since then, the private sector has been managing and operating a variety of public services, whether through the acquisition of former state-owned companies whether through state concessions - in both cases, infused with the efficiency and quality improvement discourse and with the fiscal aim to make cash for the state (Almeida, 2014). In the recent governments of presidents Michel Temer (2016-2018) and Jair Bolsonaro (2019-), privatization is still in vogue as a way of reducing the state and its fiscal deficit (Esposito, 2017; Fernandes & Tomazelli, 2019; Época, 2019), although not without controversy and contestation (Lacerda, 2020; Resende, 2019).

Box 1
Privatizations in Brazil from 1980 to 2015

For Costa and Peci (1999, p. 3), privatization or denationalization (for a discussion of the use of the two terms, see Pachi, 2008a) “is part of the history of the transformation process of the Brazilian State, also dictated by social and political reasons”. This statement refers to the fact that the transfer of public services and industrial activities to the private sector is situated, framed in a specific political and historical context, associated with the crisis of the national-developmentalist state at the national level and the expansion of neoliberal ideology at the international economic level (Bresser-Pereira, 1998; Fadul, 1999). In this context, there was “a liberal political consensus”, based on the “dependent and associated development strategies”; neoliberal economic stabilization strategies; and the dominant administrative strategies in the scenario of market-oriented reforms” (Paula, 2005, p. 38).

At the administrative level, the emergence of a managerialist ideology became pervasive in areas of activity outside the private sector, such as universities, hospitals and public sector agencies (Cherchiglia & Dallari, 2003; Doolin, 2002; Du Gay & Salaman, 1996; Onuma, Zwick & Brito, 2015; Siqueira & Mendes, 2009). Despite some claims that under Lula and Dilma governments a societal public administration replaced the so-called managerial public administration (Paula, 2005), Misoczky, Abdala and Damboriarena (2017, p. 12) move from the view of the 1990s reform as “a historically localized episode” to one that frames it as a socio-political project “related to the neoliberal stage of capitalism and managerialism as its operational arm”, explaining somewhat the permanence of the privatization agenda in governments of different ideological positions, as shown in Box 1.

For management research, approaches to privatization have been predominantly instrumental, highlighting performance gains (or losses) with the shift from public to private ownership (Araújo & Silvestre, 2014; Bachiller, 2017; Cardoso, Maia, Santos & Assis Soares, 2013; Fillardi, Leite & Torres, 2014; Scriptore & Toneto, 2012; Silvestre, Hall, Matos & Figueira, 2010) and the impacts on service quality and customer service (Pina, Torres & Bachiller, 2014; Zilber, Lex & Ades, 2005). Other studies view privatization through an organizational change and human relations perspective, approaching the deterioration of employment conditions (Mccarthy, Reeves & Turner, 2011; Silva, 2002), the new human resource management and the new required management skills (Castro, 2005; Luz, 2002; Oliva, 2002), employee resistance (Palassi, 2002), strategic changes in organizational structure (Faria & Fischer, 2001; Veloso & Trevisan, 2010), culture (Ferreira, Fandiño, Segre & Nascimento, 2010), values (Luz, 2007) and organizational identities (Carrieri, 2002).

Despite the contribution of these studies to the understanding of the phenomenon, only few analyze the media role in the social imaginary construction on privatization, along with the role of the State in the provision of public services. Since history shows us the non-evolutionary, political-ideological character in the transition from public to private sector provision of utility services (Clifton, Lanthier & Schröter, 2011; Fadul, 1999), it is therefore relevant to better understand the ways in which a favorable or unfavorable imaginary on privatization are socially constructed. In this regard, the media, with its power to reaffirm “a social imaginary about the present, past and future, in which individuals of the present time internalize images and references” (Pachi, 2015, p. 32), plays an influential role on public opinion by constructing a possible interpretation of privatizations, justifying, legitimizing and naturalizing them in society.

Regarding mainstream media, the discussion of its role in the discursive construction of organizational phenomena acquires relevance, since “mostly the main journalistic publications in Brazil are privately owned and for-profit [linking] directly with other media and other business sectors” (Smith, 2000, p. 20). Therefore, as organizations of public and private nature constituted from the outset as profit makers, wide circulation and significant political action (Capelato, 2014), newspapers assume “different positionings, influencing and directing people’s daily lives [and publishing] what they chose as appropriate to reach the public” (Santos & Costa, 2019, p. 373).

In the field of linguistics, researchers like Pachi (2008a, p. xi, 2008b, 2015) and Leal (2005) adopt the theoretical-methodological approach of critical discourse analysis (CDA) to understand how media discursive practices frame “privatization as a symbolic landmark of a period of prosperity”, establishing it favorably. In public administration, CDA researchers have studied the expansion of the managerialist ideology in public companies human resources management (Onuma et al., 2015; Siqueira & Mendes, 2009), the use of language as a way of excluding citizens regarding society matters (Rosa, 2011), and the knowledge building within the public administration field (Wallmeier, Helmig & Feeney, 2019). Yet, none of these approach media texts. Conversely, CDA business administration studies have analyzed the media on topics such as mergers and acquisitions (Riad, Vaara & Zhang, 2012; Vaara & Monin, 2010), the closing of organizational units (Vaara & Tienari, 2008) and leadership (Elliott & Stead, 2018; Liu, Cutcher & Grant, 2017), among others.

We align our research with those mentioned above addressing two questions: how the hegemonic and mainstream media have represented the privatization of telecommunications over the years? How do these representations differ or coincide with that of medias that clearly position themselves in favor or against privatization? By addressing these questions, the contributions of the study are threefold: 1) from a methodological point of view (as explored in Baker et al., 2008; Freitas, Biar & Martins, in press), bringing to public administration field the corpus linguistics approach, a set of methods suitable for revealing patterns of media representation that are hardly distinguishable with the naked eye; 2) analyzing, for an extensive period (from 1994 to 2015), the hegemonic media discourse and representations of Brazilian telecommunications sector privatization, assuming the cumulative effect of the media power through the repetition of certain representations and forms of causality and agency (Baker, 2010); and, finally, 3) comparing the hegemonic media representations to those of media positioned as pro and anti-privatization, uncovering how the discourse of the former differs from other possible discourses on privatization.

2. PRIVATIZATIONS IN CONTEXT

The privatization of Brazilian state-owned companies and, more specifically, of the Brazilian telecommunications sector - considered as the “crown jewel” of Brazilian state-owned companies - in the 1990s is usually framed as part of a broader international movement that took a particular form at the national level, having as main feature the rise of neoliberal ideology as a way of thinking about the State, its role, its size and its scope.

The State reform carried out in Brazil in the 1990s, during the government of Fernando Henrique Cardoso (FHC), embraced this ideology through proposing the reduction of the State using privatizations, outsourcing and what was then called publicization (the transfer of public social services to the non-state sector); using deregulation to reduce State intervention and limit it to just the necessary; increasing state governance by means of fiscal adjustment and administrative reform to reach a managerial public administration; and increased governability, with political institutions guaranteeing the intermediation of interests (Bresser-Pereira, 1998, p. 60).

For Bresser-Pereira, the father of the 1995 state reform, the economic crisis that started in the 1970s and that peaked in the 1980s would have been a crisis of the State

partly as a consequence of the inability to recognize the new facts that were occurring at the technological level, partly due to the mistaken view of the role of the state as a social demiurge, and partly, finally, because of the inevitable distortions of any state administration system as time goes by (1998, p. 53).

The author compares this argument to another crisis of the economic system, that of 1930, framed as a crisis of liberal capitalism, having its origin in the market. This crisis would have led to the rise of the national developmentalist state in Brazil and its first state reform called the Estado Novo (“New State”) carried out by the dictatorial government of Getúlio Vargas. The next reform that followed Estado Novo would have occurred in 1967 during the military dictatorship and in the second phase of national developmentalism in the country. The third reform would have been that of 1995, which included the privatization of telecommunications, during the FHC government (Costa, 2008).

Authors such as Gurgel (2017, pp. 164-165), however, reject Bresser-Pereira’s (1998) diagnosis of the origin of the 1980s state crisis arguing that the source would be both the market and the state; to blame the state for the crisis would have been “a skillful way to be able to reduce it, mainly by transferring assets and public services to the private sector”. For the author,

even if it was the current discourse, in the media and in academia, the wasteful action of the State that was not the reason for privatizations and concessions. But the fact that the great operation of the reform of the State consisted in promoting the transfer of assets and services to the private sector aiming at the revaluation of over-accumulated capital in the real economy. Then, in recent years, public-private partnerships have emerged, in which the State provides not only the investment opportunity, but its own financing, through development agencies where the BNDES pontificates.

Thus, neoliberal ideology would have endured in governments of different ideological positions, as in the leftist governments of Luiz Inácio Lula da Silva and Dilma Rousseff the use of public-private partnerships and public concessions stood out. Misoczky et al. (2017, p. 1) reinforce this argument by defining neoliberalism as a “dynamic and resilient political program that organizes the current stage of capitalism, operationalized from managerialism, a rationality that proposes the reproduction of the market logic in all dimensions of associated life”.

Even before the 1995 reform, the government of Fernando Collor (1990-1992) was already inspired by a neoliberal agenda with the National Privatization Plan (PND), which aimed to reduce public debt and promote greater national competition (Almeida, 2013). Even though the National Privatization Program of the government of João Figueiredo (1979-1985) and the sale of companies that had been nationalized due to insolvency under José Sarney (1985-1990) had already occurred, it was in the Collor government that the neoliberal agenda was prominently included in a government economic reform program (Almeida, 2013). However, it was in FHC’s government (1995-2002) that the PND was used as one of the main instruments of State reform, with privatization of state banks, electricity and energy companies (such as Light), railways and highways, Companhia Vale do Rio Doce and the telecommunications sector (Couto & Abrucio, 2003).

At the international level, the neoliberal influence would have started since the end of the 1970s, with the governments of Margareth Thatcher, in the United Kingdom, and Ronald Reagan, in the United States. Under the discourse of efficiency, market orientation and improving service quality, several state-owned companies, especially those of the infrastructure sector, were denationalized or have their markets deregulated, in an international context marked by the rise of the New Right, by the substitution of an industrial economy for a service economy, by financial globalization, and by valuing the individual as opposed to the collective (Clifton et al., 2001; Kroeze & Keulen, 2014).

The difference between the international and national domains was the imposition of bodies such as the International Monetary Fund (IMF) and the World Bank to indebted countries like Brazil - that suffered fiscal crisis and high inflation - to adopt the recommendations of the Washington Consensus, with its neoliberal agenda of economic opening, fiscal adjustment, deregulation of industries, privatization of state-owned companies and reduction of the state in the economy (Kogut & MacPherson, 2011). No wonder, according to Ramamurti (1992), the countries with the highest number of privatizations were those with the largest fiscal deficits, greater external debt, and greater dependence on international financing agencies.

Costa and Peci (1999, pp. 1-2) stress that even if there were at least 3 strategies for fiscal adjustment - asset reform, with the transferring of assets to the private sector mainly through privatizations; tax reform, aimed at increasing revenue, and managerial reform, aimed at increasing the efficiency of public spending - “by ideological affirmation or by pressure from international organizations, most Third World governments have opted to carry out asset reforms”. In particular, the low efficiency of state-owned infrastructure companies, instead of being credited to the economic crisis and to the use of these companies’ profits for public accounts or as an instrument of monetary policy, were ascribed to the lack of competition and the consequent lack of incentives for management (Almeida, 2014; Clifton et al., 2011). In addition, in particular for the telecommunications sector, technological innovations allowed the convergence of information and communication technologies, eliminating barriers to entry in the sector, offering new business opportunities to the private sector, making it possible to de-characterize industry as a natural monopoly (Maculan & Legey, 1996), and forcing other countries besides the United States and the United Kingdom, to liberalize their telecommunications industry for international competitiveness (Kornelakis, 2015).

We must remember that the private sector presence in infrastructure sectors was not unprecedented. These same sectors had originally been made possible by the alliances of entrepreneurs and private banks, given the high investments and the risk associated with these activities, both in Brazil and in parts of Europe. The State’s presence was limited to price regulation and distribution of concessions (Clifton et al., 2011; Fadul, 1999). As recalled by Fadul (1999), various infrastructure services in Brazil, such as public cleaning, urban transport and water supply, were carried out in the cities by private entrepreneurs, and services such as electricity, road transport, postal services and telecommunications were offered by foreign companies. In Brazilian telecommunications, until the 1950s, the private sector was the main responsible for the operation of telephone services in states and municipalities, through concessions (Fascina, 2002).

In this way, we assume a non-evolutionary vision of the State Reform and, consequently, of the public services transferring to the private sector. We therefore agree with Costa’s (2009, p. 80) view of the State Reform as a political issue, of collective choice:

Thus, (State) reform cannot be treated as an eminently technical issue, detached from the commitments of the political forces of society and the interests that present themselves in the state arena, that is to say, without major political consequences. State reform raises issues of a distributional nature and therefore conflicts of interest between different social groups.

If so, how did governments and society reach a consensus on the need for privatization? With regard to governments, for MacPherson (2006), the spread of neoliberal ideology was the result of a planned effort by economists from the Chicago school, as there was no consensus among economists or public administrators in the early 1980s on the scope of the State’s action. The author recalls that, before the privatizations of the 1980s, only Germany and Chile had privatized some public companies. The Chicago school, through a global network of economists associated with sympathetic governments - as in the case of economist Milton Friedman in the Thatcher government -, would have implemented a proactive strategy to disseminate its economic ideology that ended up becoming virtually a consensus among the main economics schools in the world.

And what about society? Vaara and Tienari (2002, p. 276), for example, suggest that organizational phenomena such as mergers and acquisitions “create a need to make sense collectively of what is happening. Much of this collective sensemaking takes place through media texts or is reflected in them. The emerging discourses shape collective understandings of these complex phenomena”. Thus, the media holds a prominent position in the way in which themes that affect society are (re)produced and received by the latter.

Pachi (2008a, p.5) reiterates this position when addressing the hegemonic press “as a place that is socially legitimized and of production of an opinion-forming discourse that expands in the public space”. He also reminds us that “the press discourse is one of those produced in the public space, not being configured as unique nor as a ‘reflection of a reality’”. Thus, we emphasize the media role in the social construction process of organizational and political phenomena such as privatizations, influencing the justification, legitimation and naturalization of these phenomena in society.

Vaara and Tienari (2002), for example, analyzed the discursive construction of mergers and acquisitions in the media in Finland and identified 4 types of discourses about the phenomenon: rationalistic, cultural, societal and individualistic. The first relates to the company value creation, in financial terms; the second, to a process of cultural confrontation between two companies now made one; the third, to the impact of these mergers on society, on employment and on the loss of national sovereignty - in the case of mergers with foreign companies -; and the fourth, to individual personalities, such as CEOs, simplifying the phenomenon for readers. These 4 types of speeches end up justifying, naturalizing, and legitimizing the impacts of these organizational changes in the eyes of public opinion.

Concerning the public sphere, how does the transfer of public services to the private sector take place? Pachi (2008a, 2008b, 2015), for example, analyzed the discourse on privatization of Brazilian telecommunications in the newspapers Folha de S. Paulo and Estado de S. Paulo, for the years 1997 and 1998 (year of privatization). The author argues that the word privatization “is inscribed in a memory and is related to a time when the State reviews its role in economic development” (Pachi, 2008a, p. 22). His analysis uncovers the press emphasis on the rupture between past and present, representing the telecommunications privatization as a founding event. The image of the past that the press has constructed discursively about the state-owned period of these companies reinforces the discourse of inefficiency, of lack of investments, of lack of telephones and of inability to manage the State, driving telecommunications into chaos and negatively affecting citizens’ lives. “It feeds on an imaginary of the past that produced negative effects for society and should therefore be rejected and denied in its continuity. For this reason, a break with this past is promoted in the discourse, establishing privatization as the symbol of the new” (Pachi, 2008b, p. 2). The media, therefore, disqualifies the country’s telecommunications past, centered on the figure of the state-owned company and the role of the state as a manager. Pachi (2008a) also shows that in the press discourse there is a link between the Brazilian telecommunications’ past and its lovers to the military regime, associating this past to a period in the country’s history considered until recently controversial and commonly rejected by contemporaneity. The press also connects the telecommunications sector to communist countries: “The operation of Telebrás is compared to that of a Soviet state-owned company, meaning negatively in this speech. [...]Thus, the communist regime works as a parameter of comparison for what happened in Brazil” (Pachi, 2008a, p. 165). Both associations - military regime and communist countries - aim to give a negative and outdated connotation to the status quo of the Brazilian telecommunications sector.

Thus, through a distinction between the new and the old, the press makes a utopian construction of the future, in which supposed benefits will be realized for the whole society. For Pachi (2008b), the creation of this utopia of the future makes it possible to establish privatization as a symbolic landmark, and the imagined future an argument in favor of privatizations, normalizing the phenomenon among readers. Therefore, the telecommunications past is characterized as chaotic and inefficient while the future is promising and of well-being.

Similarly, Leal (2001) analyzed the press discourse about the privatizations of Usiminas, Companhia Vale do Rio Doce and Telebrás and showed how the media favorably represents these privatizations. The author shows how the images of the state-owned company’s past in the press carry a negative connotation, showing “State as an elephant and a dying father. The “elephant” gives the idea of something slow, not functional, and the “father” image as of someone who gives without receiving anything (Leal, 2001, p. 82). In this way, the image of the State as father and elephant is opposed to the imaginary of modernity that privatizations would bring. On the other hand, the public manifestations against privatizations are represented in the press in a negative way, with the use of words such as stage, old characters and ham actors, representing “the manifestations as enactments, something false. The representation of the protesters as old characters emphasizes the sense that they are outdated and suggests that they are already known, insubordinate individuals. And ham actors bring the sense of poor performance, incompetence” (Leal, 2001, p. 84).

The media allows a collective understanding of events that are relevant to society. Not only can journalists build new meanings for these events but can also reproduce discourses that already exist or are expected to be heard (Vaara, Tienari & Laurila, 2006). Moreover, the effects of the media can be cumulative, through the repetition of representations such as those demonstrated by Pachi (2008b) and Leal (2001). Pachi (2008a), for instance, had as analytical period the very own years of telecommunications privatization. Using a methodology that complements those of Pachi’s (2008a, 2008b) and Leal’s (2001), this research addresses the ways in which the media builds these phenomena cumulatively over time, and pinpoints the variations in the discursive constructions of different types of media.

3. THEORETICAL-METHODOLOGICAL ASPECTS

In this research we analyzed the media discursive representations of the privatization of the telecommunications sector from 1994 to 2018. Since this is a 24-year period, it is appropriate to adopt corpus linguistics methods, capable of generating a non-linear or a distant reading (Freitas, 2017), borrowing the term from the literature (Moretti, 2008).

In non-linear reading, the different point of view made possible by the disembodied text allows to create correlations not available by conventional reading, high frequency or not. In distant reading, distance as a specific mode of knowledge privileges relationships, patterns and forms while forgoing details. The use of large collections of text for distant reading is not considered a single method, but several, used and selected according to the research questions (Baker, 2010). In common, they share analysis procedures that use large collections of electronic texts treated with specific computational tools and that go back to an approach that emerged in the 1990s in linguistic studies (McEnery & Hardie, 2011).

If the type of linguistic work had been customarily subordinated mainly to the lexicographic and pedagogical dimensions, since the mid-2000s, however, work has begun to emerge that articulates corpus and sociolinguistics (Baker, 2010) and corpus and critical discourse analysis (CDA) (Baker et al., 2008; Baker, Gabrielatos & McEnery, 2012; Strom & Alcock, 2017; Freitas et al., in press), and it is to this tradition that we align ourselves.

For sociolinguistics, the use of corpus linguistics can assist in sociolinguistic variation studies, that is, the phenomenon in which the language is used in different ways by different users. This approach can reveal differences and similarities in the use of language by different social groups, in a synchronous way (among social groups that coexist) or diachronic (referring to variations over time), and the way in which variables such as age, region, social class and gender impact these differences (Baker, 2010).

An interesting use of corpus linguistics for public administration and administration in general could be the combination with the interdisciplinary approach of CDA, already used in these fields to understand how the managerialist ideology has expanded to the human resources management of public and private companies (Onuma et al., 2015; Siqueira & Mendes, 2009) or how language can be used as an instrument to exclude individuals with different cultural and symbolic capital than that of text authors (Rosa, 2011). The use of large corpora - ranging from media articles, organizational communication, government speeches and even transcription of interviews -, combined with CDA, may assist in the representativeness of the analyzes, since the CDA works with a reduced number of texts and is therefore less likely to reveal frequent and cumulative patterns that can represent hegemonic discourses.

An important article in the field of linguistics that advocates this combination is that of Baker et al. (2008). For these authors, the CDA, as well as the use of corpora - both relatively recent approaches in the field of linguistics -, is usually not considered a single method. The CDA “adopts any method that is adequate to realize the aims of specific CDA-inspired research” (Baker et al., 2008, pp. 273-274).

Different CDA approaches have in common the analysis of the discourse in addition to the text itself, focusing not only on the use of language and grammatical choices revealing the authors’ ideologies and biases, but also the production and reception of these texts, as well as their insertion in the cultural, social, political and historical contexts of their time. Through these levels of analysis, discourse analysts have shown how language is used for specific purposes, such as a way to represent in a biased manner refugees (Baker et al., 2008), immigrants (Strom & Alcock, 2017) and protesters against privatization (Leal, 2005). At the linguistic level, CDA researchers have commonly unveiled the use of nominalization, which occurs when an author chooses to represent processes that occur in the world as entities. An example is that of former British minister Tony Blair’s quote brought by Fairclough (2003, p. 13): “The modern world is swept by change”. In this case, there is an omission of process agents such as companies and governments, and the change itself becomes an entity. Other linguistic analyses include lexical choices and the use of figures of speech, like euphemism, hyperboles, and metaphors. In studies such as those of Carrieri and Palassi (2002) and Carrieri, Leite-da-Silva and Pimentel (2007), for example, the use of metaphors by employees of newly privatized companies gives clues on changes in organizational identity and on the propensity to join the union movement. In a complementary way, the texts are analyzed in the light of their relationship with society, including the production and receiving contexts - indicating their interdiscursivity and intertextuality - as well as the cultural, political, historical and social contexts in which they are inserted.

Given the complexity of CDA, its approaches are often essentially qualitative, comprising of a small number of texts (Baker, 2010). On the other hand, studies with large corpora are quantitative and carried out with the aid of specific programs; however, they also demand qualitative analyses, such as examining concordance lines. For Baker et al. (2008, p. 283), this is how large corpora can be complementary to CDA, since some patterns of language use are not distinguishable with the naked eye, that is, in a conventional reading: “A small-scale analysis may not be able to identify which linguistic patterns are cumulatively frequent (and therefore likely to represent powerful discourses) and those which are less frequent (and therefore may constitute minority or resistant discourses)”.

In this way, linguistic patterns can be revealed “via a combination of automatic and qualitative forms of analysis” which, in turn, can also be combined with contextual analyzes such as those of the CDA (Baker, 2010, p. 123). Like Baker (2010), we understand that the best use of large corpora is when it improves small-scale qualitative analyses, and not when it replaces them.

In the field of administration, quantitative analysis of texts is often associated with content analysis (Bardin, 1977). However, in the articulation between quantitative methods and CDA, we move away from Bardin’s proposal more for epistemological than methodological reasons, since we share systematic and computer procedures, for example. On the other hand, content analysts trust a certain transparency of the language, in the neutrality of technologies and their role as aides in the introduction of an “additional order revealing an internal structure” (Bardin, 1977, p. 55).

In the perspective that we assume here technologies are not neutral - the simple activity of “counting” words already involves positioning what should be considered a word (Anthony, 2013) -, and the gross dispersion of the data is not thought of as something that should be fixed for the benefit of the norm, but it deserves a perspective that privileges it. With the help of technologies, we ended up introducing some order, neither supplementary nor revealing, but that allows us to create narratives capable of shedding light on our practices. For us, categorization does not reveal, but builds (for a more in-depth discussion of the differences between CDA corpus research and content analysis, see Freitas et al., in press).

In this research, we use procedures familiar to the linguistic exploration of corpora: generation and analysis of list of words, keywords, concordance lines and collocation, detailed below.

The word list indicates the frequency of all words (or terms and phrases) in the corpus. The most frequent words in a corpus (of any language, by the way) are articles, prepositions, auxiliary verbs, and certain pronouns that are considered semantically empty by themselves, being called function words. But we should not eliminate them from the corpus, for when combined with other words they may acquire greater relevance (Pearce, 2014).

For both the word list and the keyword list, two indicators are used: that of tokens, which indicates the total number of words in a list, including repeated ones, and that of types, that counts the number of words without repetition of this corpus.

The second procedure is the generation of a list of keywords, which are the most frequently used words (or group of words) from one corpus in relation to another, termed as control corpus (reference corpus or contrast corpus), used to highlight lexical contrasts between different materials. When comparing one corpus to another, the list of keywords highlights the central concern of the corpus under analysis, which is called the lexical field. The underlying idea is that, by comparing two corpora with different characteristics, what is specific to the material under analysis will emerge. Therefore, the characteristics of the reference corpus will influence what the tools return as a keyword, and the more similar the corpora, more chances of relevant elements being filtered (or made invisible) when generating keywords.

Despite some authors’ advice for a reference corpus 5 times larger than the one under analysis, recent studies show that reference corpora of varying sizes and qualities extract a similar set of keywords, invalidating the previous requirement (Baker, 2010). For the analyses used in this research, except when indicated, the contrast material was a corpus with 45 works by the Brazilian writer Machado de Assis (209,486 tokens and 14,976 types). The choice of this material as a reference corpus was mainly their easy accessibility, for they are in the public domain, enabling comparisons with other data and analyses.

We performed keyword analysis in 3 ways. The first, in the keyword analysis section, included the survey of keywords from the sum of hegemonic media articles from 1994 to 2015. Then, in the keywords longitudinal analysis section, we analyzed the hegemonic media keywords by presidential mandate (approximately): 1994 to 1998, 1999 to 2002, 2003 to 2006, 2007 to 2010, 2011 to 2014 and 2015 to 2018. This division aimed at examining the differences in the central concern of the hegemonic media over time. Finally, in the analysis section among different medias, we made a keyword comparison by media type (hegemonic, pro and anti-privatization) for the years 2007 to 2018, addressing the differences between the central elements of each one of them. The reason for choosing this period was the availability of articles online for the pro and anti-privatization media, as shown in Table 2.

We used the Antconc 3.5.7 (Anthony, 2016) to manipulate the corpus, for it is a corpus linguistics program that calculates keywords taking into account relevance (keyness), based on the statistical test of Log-likelihood, since language data do not have a normal distribution. According to Baker (2010), corpus linguistics researchers usually analyze the first 20, 50 or 100 keywords, which was followed in this research.

The third procedure is the analysis of concordance lines, which allow us for a more qualitative analysis. With the help of the tool, we observed selected words within their contexts (closest texts), and then order them according to research needs. With this procedure, it is possible to identify patterns of terms use not visible by other procedures.

Collocation was the last procedure used in this study and indicates the co-occurrence of two words in a corpus, one being the node - the word under analysis - and the other the collocate, which contributes to the meaning of the node. Collocation is “a lexical relationship better discernable in the analysis of large amounts of data, and, therefore, it is less accessible to introspection or the manual analysis of a small number of texts” (Baker et al., 2008, p. 278). As the collocates do not need to be contiguous to the node to influence the meaning of the last, the program usually brings them in a predefined space of words, allowing for identification of non-sequential patterns (McEnery & Hardie, 2011). For this research, we followed Baker et al. (2008) by defining this space from 5 words to the left and 5 words to the right of the node “privatization”. Likewise, we defined a minimum frequency of 10 times for a word to be considered as collocate. The statistic used was Mutual Information (MI), pre-set by Antconc.

3.1 Corpus Compilation

In order to analyze the media discourses on the telecommunications privatization, 3 corpora were compiled. The first or the main one included online news from mass media, which we call hegemonic media. According to the website Poder360 (2018) with data from the Circulation Verifier Institute (CVI), the three newspapers with the greatest digital circulation, as shown in Table 1, are Folha de S.Paulo (Folha), O Globo e o Estado de S. Paulo (Estadão). These media were considered hegemonic because, in addition to being the main broadsheet newspapers, declare themselves as non-partisan, striving for the plurality of opinions and the search for the truth for readers.

The second corpus comprises online news from Instituto Liberal, Instituto Millenium and Instituto Mises, think tanks with a liberal economic profile, called pro-privatization. The third includes online news from the medias Brasil de Fato and Carta Capital, associated with a more interventionist economic profile, called anti-privatization.

Table 1
Circulation of printed newspapers and digital subscriptions from 2015 to 2017

	average of copies - in thousands (audited by the CVI)				evolution-Dec. 2014 to Dec. 2017
	dez. 14		dez. 17		%
media	print	digital	print	digital	print	digital	total
Folha (SP)	212	159	121	164	-42.9%	3.3%	-23.1%
Globo (RJ)	205	148	130	113	-36.3%	-23.9%	-31.1%
Super Notícia (MG)	284	39	157	48	-44.9%	23.1%	-36.7%
Estado (SP)	163	74	115	89	-29.9%	19.7%	-14.4%
Zero Hora (RS)	164	38	101	80	-38.6%	112.6%	-10.4%
Estado de Minas (MG)	56	53	26	25	-52.7%	-52.1%	-52.4%
Correio Braziliense (DF)	41	10	26	18	-35.2%	87.2%	-11.4%
Valor Econômico (SP)	43	16	29	29	-31.9%	83.5%	-1.0%
Gazeta do Povo (PR)	38	2	*	n.d.	*	n.d.	n.d.
A Tarde (BA)	30	10	17	13	-44.9%	37.9%	-24.9%
O Povo (CE)	19	n.d.	14	n.d.	-28.2%	n.d.	-28.2%
total	1,256	549	736	580	-41.4%	5.8%	-27.0%

• * The “Gazeta do Povo” stopped circulating in daily printed version in 2017. The “Povo” has no data on digital signatures.
• Source: Poder360 (2018), with official CVI (Circulation Verifier Institute) data. Newspapers: average daily circulation + digital signatures.

In this analysis, we build an opportunistic corpus (McEnery & Hardie, 2011), constructed with the material available to meet the research objective. For the automatic compilation of the material, we used the Bootcat tool (Zanchetta, Baroni & Bernardini, 2011), which extracts web pages based on a selected set of words and convert them into a text format corpus. The selected set of words were “privatize*, telecom*”; “liberalizes*”; “telecom*”; “denationalizes*, telecom*”; “regulates*, telecom*”; “privatization*, telecom”; and “regulates*, telecom*”¹. We manually categorized the generated txt files according to the article’s year of publication and the specific media (metadata), “cleaning” the data when it contained advertising information that made data analysis difficult.

Table 2 details the 3 corpora. Types and tokens were calculated for both the word list and the keyword list. The last two columns show the ratio of keywords to total types and tokens. For the media comparative analysis, we used the period from 2007 to 2018, due to pro and anti-privatization media online articles availability.

Table 2
Corpora Details

Media	Website	Download	Positioning	Period	Articles	Types (a)	Tokens (b)	Type/token% (a/b)	Types Keywords (c)	Tokens Keywords (d)	Type/token Keywords % (c/d)	Type Keywords % (c/a)	Token Keywords %(d/b)
Estadão	estadao.com.br	16/10/18	Hegemonic	2000-2018	211	14,854	153,693	9.66
Folha	folha.uol.com.br	16/10/18	Hegemonic	1994-2018	403	17,703	264,599	6.69
Globo	globo.com	16/10/18	Hegemonic	2007-2018	255	14,079	155,243	9.07
Total Hegemonic corpus				1994-2018	869	27,347	570,779	4.79	1,094	314,742	0.35	4.00	55.14
Instituto Liberal	institutoliberalorg.br	14/11/18	Pro-privatization	2010-2018	12	6,242	36,900	16.92
Instituto Millenium	institutomillenium.org.br	14/11/18	Pro-privatization	2010-2018	13	4,005	16,702	23.98
Instituto Mises	mises.org.br	14/11/18	Pro-privatization	2008-2018	63	13,531	148,794	9.09
Total pro-privatization I				2008-2018	88	15,568	185,694	8.38	893	90,807	0.98	5.74	48.90
Brasil de Fato	brasildefato.com.br	23/10/18	Anti-privatization	2010-2018	83	13,742	101,194	13.58
Carta Capital	cartacapital.com.br	23/10/18	Anti-privatization	1999-2018	173	20,481	245,348	8.35
Total anti-privatization corpus				1999-2018	256	25,108	346,531	7.25	1.038	164,260	0.63	4.13	47.40
TOTAL CORPORA				1994-2018	1.213	41,448	1,103,004	3.76	1.101	573,809	0.19	2.66	52.02

• Source: Elaborated by the authors.

4. ANALYSIS AND DISCUSSION

4.1 Keyword Analysis

Table 3 shows the first 50 keywords in the corpus, excluding function words and words like “Folha”, “Paulo” (Folha de S.Paulo), “year”, “second” and “three”, unrelated to privatization. Some of these words were checked for meaning through lines of concordance and cluster analysis, which counts compound words in the corpus, such as “Telebrás System” and “Brasil Telecom”. We illustrate this point with the word “telecom”, which appeared 762 times as part of “Brasil Telecom”, 543 times as “Telecom Italia” and 412 times as “Portugal Telecom”, in addition to “France Telecom” (84) and “British Telecom” (47), indicating that the first word on the list, “Telecom”, refers to companies.

Table 3
Hegemonic media keywords - 1994 to 2018

N.	Frequency	Keyness	Keyword	N.	Frequency	Keyness	Keyword
1	2387	1495,19	Telecom	26	617	385,97	Telebrás
2	2472	1321,16	Brazil	27	606	379,08	Opportunity
3	2208	1273,05	Company	28	646	365,93	Fixed
4	1966	1194,05	Companies	29	582	364,06	Telemar
5	1396	873,79	Telephony	30	558	349,04	Control
6	1346	828,67	Oi	31	570	335,19	Funds
7	1293	809,25	Billions	32	532	332,77	Participation
8	1284	803,61	Telecommunications	33	532	332,77	Phone
9	1643	795,92	Government	34	525	328,39	Service
10	1188	719,18	Marketplace	35	520	325,26	BNDES
11	1220	688,05	President	36	649	323,01	Minister
12	1062	664,56	R (R$)	37	497	310,87	Carriers
13	1062	664,56	Millions	38	496	310,25	Internet
14	1026	642,02	Privatization	39	496	310,25	Tele
15	968	605,7	Country	40	506	304,68	Purchase
16	962	601,94	Shares	41	486	303,99	Agency
17	957	585,71	Sector	42	475	297,11	Investments
18	859	537,45	Cellphone	43	452	282,72	Operator
19	857	536,2	Anatel	44	468	281,06	Band
20	826	493,95	Bank	45	428	267,7	Federal
21	762	476,72	Services	46	565	253,59	Portugal (telecom)
22	904	457,5	Group	47	394	246,43	Shareholders
23	753	424,05	US (US$)	48	582	245,35	Deal
24	651	407,25	Auction	49	391	244,55	Operation
25	621	388,47	(Telecom) Italy	50	390	243,92	Price

• Source: Elaborated by the authors.

Table 3 unveils the predominance of words related to companies and the government, in addition to telecom services. To facilitate the analysis, we categorized the first 100 keywords of the hegemonic corpus into 12 categories, as shown in Table 4. This time we included function words, which represent 4% of total keyword types and 34% of total keyword tokens.

The categorization shows that the lexical field of privatization in the hegemonic media revolves around companies, their services (telephony, telecommunications, cellular), the business size (millions, billions, R$), the privatization process per se (auction, consortium, purchase, sale, merger) and government actors and individuals with decision-making power or influence over the sector (Lula, minister, president).

Table 4
Categories for 100 keywords of the hegemonic media corpus, 1994 to 2018

Category	Type	% total types/ keyword (a)	Tokens	% total tokens/ keyword (a)
Company	22	2%	20,028	6%
Service	9	1%	7,393	2%
Size	7	1%	6,248	2%
Privatization process	9	1%	4,838	2%
Government	5	0%	3,442	1%
Country / Region	4	0%	2,631	1%
Individual	4	0%	2,602	1%
Property	4	0%	2,446	1%
Regulation	2	0%	1,343	0%
Economy / Business	1	0%	1,188	0%
Others	1	0%	346	0%
Function words	32	3%	108,305	34%
Total first 100 keywords	100	9%	160,810	51%
Total keyword list (a)	1,094	4%	314,742	55%
Total word list	27,347		570,779

• Source: Elaborated by the authors.

The hegemonic media definition of the lexical field ends up restricting the topic of privatization to the sphere of business, of companies and their relationship with the government. The individuals in the keyword list are public persons, such as the president of the country, the presidents of the telephone companies and the minister of telecommunications, limiting matters regarding privatization to individuals with decision-making power. This result is quite interesting, considering that one of the main arguments for the privatization of state-owned companies was the service quality improvement for customers and the citizens’ universal access to telephone services. The words “customers”, “consumer”, “workers”, “employees” and “jobs”, however, appear as the 110th, 144th, 171th, 188th and 1,042nd position in the list, with a frequency of 293, 227, 201, 189 and 38, respectively.

Interestingly, the word “universalization” appears 80 times, taking the 498th place. The hegemonic media lexical field for the privatization of telecommunications, therefore, relegates the topic to business, bypassing the broader interests of society.

It must be stressed, however, that the corpus analysis allows visualizing patterns of representation about selected words or themes with respect to quantity. The fact that words like “company” and “government” appear more frequently in the corpus does not mean that there are no articles in it that discuss, for example, universal access to telecommunications services for citizens. On the contrary, these articles exist in the corpus. However, while the word “universalization” appears 80 times in the 24-year period, the Oi company, for example, appears 1,346 times in the same period, suggesting that the central concern of the hegemonic media when representing the theme of privatization is in companies, not in its users.

We also did not find in the corpus in a significant volume the topic of fiscal deficit, considering that one of the arguments for privatization was the need to make cash for public coffers. Although the financial side seems to be very important in the media, as shows the “size” category in Table 3, which includes the words “billions”, “millions”, “R” (for R$) and “US” (for US$), the word “fiscal” appears 78 times in the corpus, and “public debt”, 14 times. A possible explanation is the long period of analysis, making the topic of a one-time sale for fiscal deficit matters disappears over the time.

The next analyses deepen the results above in 3 ways: 1) with a keyword analysis splitting the corpus into subcorpus according to periods roughly corresponding to the presidential terms, raising marked differences in the lexical field of hegemonic media around the theme of privatization over time; 2) with keyword analysis of pro and anti-privatization medias to raise alternative representations of privatization; and 3) with collocation analysis of the word “privatization” in the hegemonic media.

4.2 Keywords Longitudinal Analysis

The longitudinal analysis examines the hegemonic media lexical field on telecommunications privatization from 1994 to 2018, taking into account variations over time. The analysis was divided into 6 periods, approximately according to presidential terms: 1994 to 1998 (year of privatization), 1999 to 2002, 2003 to 2006, 2007 to 2010, 2011 to 2014 and 2015 to 2018.

Figure 1 indicates the frequency of articles per year on the topic of telecommunications privatization, by type of media (hegemonic, pro and anti-privatization). It is important to note that only Folha had online articles for the entire period (1994 to 2018), whereas the coverage period of other media varied considerably, as shown in Table 2, in the section on theoretical and methodological aspects.

Figure 1 shows that the number of articles varied over time, with peaks at certain times. The main peak was clearly around the sector’s privatization, which occurred in July 1998, justifying the large number of articles.

In 2000, two years after the telecom privatization, the Telecommunications Services Universalization Fund (Fust) was regulated and the Fund for the Technological Development of Telecommunications (Funttel) was created. In addition, disputes concerning the acquisitions of Brasil Telecom and Companhia Riograndense de Telecomunicações (CRT) arose among the partners of the winning consortium, formed by Opportunity bank, by Telecom Itália and by Brazilian pension funds.

The next peak of articles was in 2008, when Oi acquired its competitor Brasil Telecom using BNDES loans. The acquisition generated quite a stir for it imposed a change in the 1998 General Plan of Concessions to allow the acquisition of a telephone company competitor- one of the goals of privatization was the end of monopoly and foster competition in the sector.

In 2010, the peak in the number of articles relates to both the Portugal Telecom sale of Vivo shares to Telefónica and the alliance of the former with Oi. Two of the main discussions in the media were the consolidation of the market in a few players and the matter of the foreign capital getting into a market that was previously 100% national, a great concern for the PT left-wing government at the time. The merger between Portugal Telecom and Oi ended up happening only in 2013, explaining the other peak that year. At a later peak, in 2016, Oi’s judicial recovery took place, raising criticism regarding the creation of the then-called national “supertelecom company” during the Lula and Dilma governments.

Figure 1
Telecommunications privatization articles per year

For the keywords analysis of the hegemonic corpus by presidential term, the control corpus used was that of the previous presidential term period, with the exception of the first period, when the Machado de Assis corpus was maintained. Box 2details the control corpus for each period:

Box 2
Control corpus for each temporal subcorpus

Table 5 shows the keywords for the period from 1994 to 1998. Since this period addresses the years before privatization and the very own privatization year, words like “north” and “east” appear, which refer to Tele Norte-Leste (North-East Telecom), old name for Telemar, and “south” for Tele Centro-Sul (Center-South Telecom), then named Brasil Telecom. The Minister of Communications, Luiz Carlos Mendonça de Barros, also shows up in the list, having replaced the then-recently deceased Minister Sérgio Motta during the privatization process. Barros at that time was involved in the privatization scandal of Tele Norte-Leste - his telephone conversations had been tapped and uncovered his preference for the Opportunity consortium, leaded by the banker Daniel Dantas. In general, however, the keywords revolve around the same categories presented in Table 4, which included the period from 1994 to 2018.

Table 5
Hegemonic subcorpus keywords from 1994 to 1998

N.	Freq.	Keyness	Keyword	N.	Freq.	Keyness	Keyword
1	415	1083.69	Companies	26	119	309.12	sector
2	325	874.85	Auction	27	116	292.94	bank
3	314	845.2	Telecom	28	102	274.34	Telephonic
4	310	834.42	Privatization	29	101	271.65	BNDES
5	295	794	Telebrás	30	98	263.58	control
6	314	755.71	company	31	95	255.5	state-owned
7	253	680.85	Telecommunications	32	114	252.4	South
8	227	610.82	Telephony	33	123	248.91	group
9	224	602.74	Cellphone	34	103	245.3	fixed
10	260	536.96	Brazil	35	89	239.36	communications
11	258	512.2	Government	36	89	239.36	consortiums
12	196	506.19	Marketplace	37	118	236.55	system
13	183	492.34	Tele	38	86	231.29	Anatel
14	182	489.65	Billions	39	92	229.28	east
15	172	462.73	consortium	40	113	220.5	Portugal
16	160	430.43	Telesp	41	103	211.32	sale
17	178	401.6	Minister	42	78	209.77	participation
18	144	387.36	Shares	43	77	207.08	Barros
19	159	373.93	North	44	75	201.7	minimum (price)
20	154	372.86	US ($)	45	75	201.7	telephone
21	129	346.99	Millions	46	75	201.7	telephones
22	225	343.27	R ($)	47	74	199.01	Mendonça
23	149	342.43	President	48	70	188.25	Italy
24	119	320.08	Country	49	67	180.18	Embratel
25	115	309.32	Price	50	66	177.49	services

• Source: Elaborated by the authors.

Comparing the subcorpora of each period in relation to the previous one, we are able to identify the differences in group of words statistically more frequent in one period than in another, according to Table 6.

Table 6
Keywords compared to the previous period

	1999 to 2002			2003 to 2006			2007 to 2010			2011 to 2014			2015 to 2018
N.	Freq.	Keyness	Keyword	Freq.	Keyness	Keyword	Freq.	Keyness	Keyword	Freq.	Keyness	Keyword	Freq.	Keyness	Keyword
1	246	79.24	opportunity	73	160.35	citi	480	272.81	Oi	89	146.33	alibaba	525	128.13	Oi
2	47	41	dantas	507	126.32	telecom	41	31.4	Dilma	141	73.22	us	66	70.62	creditors
3	36	37.47	pimenta	56	123	kroll	362	30.35	R	104	73.09	g	73	69.26	recovery
4	33	34.35	veiga	42	92.24	citigroup	110	28.89	Pt	64	71.21	china	62	60.63	tele
5	32	33.31	Tim	432	90.92	Brazil	72	26.58	Crisis	38	62.47	bernardo	51	57.82	temer
6	164	32.91	Funds	41	66.29	telmex	97	26.2	Fusion	97	55.36	customers	65	48.83	judicial
7	29	30.18	Intelig	39	58.55	broad	34	26.04	suicides	69	47.61	points	47	47.27	Jato (wash)
8	29	30.18	Vésper	25	54.9	miro	264	26.02	Billion	70	44.78	high	46	46.06	Lava (car)
9	28	29.14	Debentures	177	50.96	Italy	84	23.53	telefónica	35	44.29	ibovespa	56	45.52	assets
10	33	27.13	(Broad)Band	26	44.31	citibank	46	23.18	consumer	64	43.68	exchange	34	36.54	odebrecht

• Source: Elaborated by the authors.

The differences between periods show each seasons’ remarkable events not only for the sector, but also for the country. From 1999 to 2002, the highlight was the first major telecommunications privatization scandal, led by economist Daniel Dantas, from Opportunity Bank, an episode that included a dispute among partners (one of them was Telecom Italia), the company’s bond issuance to pay the acquisition installments of Tele Centro-Sul (named Brasil Telecom afterwards), and disputes regarding the CRT acquisition. During the period analyzed, the keyword list also brings names of mirror companies like Vésper (mirror of Telemar) and Intelig (mirror of Embratel), both created to compete with the incumbents. In the same period, the government started selling new licenses for C, D and E cellphone bandwidths.

From 2003 to 2006, one more scandal involving Opportunity bank comes into spotlight: the hiring of the spy agency Kroll by Daniel Dantas to investigate Telecom Italia, a partner at Brasil Telecom. The sale of Citigroup’s share in Brasil Telecom also significantly shows up in the media, within the context of partners’ disagreements at Brasil Telecom. During this period, the Mexican company Telmex acquires the long-distance company Embratel from the American MCI / Worldcom, and services like high-speed or broadband internet, expand in the market and constitute an important segment in the services portfolio of telecom companies.

The highlight of the subsequent periods is Telemar company, renamed to Oi afterwards. The company at that time merged with Brasil Telecom, then established an alliance with Portugal Telecom, and finally requested a judicial recovery during the 2015 to 2018 period. Before that, in the 2007 to 2010 period, the highlight was the acquisition of Vivo’s shares by Telefónica.

The highlights of the first term of Dilma government and Temer government were the scandals of their respective mandates - such as the suicides of France Telecom employees after privatization and the so-called Lava-Jato (Car Wash, a criminal investigation led by the Brazilian Federal Police in the state-owned oil company Petrobras). The then-new Minister of Telecommunications, Paulo Bernardo, also makes it to the list from 2011 to 2014, as does China, with the opening of its giant cellphone market and the public offering of shares of the Chinese internet company Alibaba.

In this analysis by period, it is interesting to note the emergence of “consumer” and “customers” keywords in the period from 2007 to 2010 and 2011 to 2014, respectively. The qualitative analysis of the concordance lines for the word “consumer”, for the period from 2007 to 2010, showed that the rise in the frequency of the word was related to the acquisition of Brasil Telecom by Oi. This acquisition, for the media, represented a decrease in competition and a potential harm to the consumer. In addition, there were rising criticism for companies’ noncompliance with the new rules for customer service established at the time.

Regarding the word “customers”, in the 2011 to 2014 period, the qualitative analysis of concordance lines uncovered a greater concern for telephone services quality - in particular, that of cellular phones. The telecom companies’ infrastructure investments had not followed the rapid growth of the customer base at the time, incurring in punishments by the regulatory agency Anatel. The merger of Oi and Portugal Telecom and the acquisition of Vivo by Telefónica also explain the high frequency of the word “customers” since there was a higher concentration of customers in the merged companies.

However, why the words “customers” and “consumers” are more significant in this period than in the others? One explanation may be the maturation of the sector at this point, with the consequent rise of issues that affect services provision to clients, such as the need for network investments with the expansion of the customer base and changes in the telecom usage profile with the rise of data services. Another explanation may be the more active role of regulatory agencies such as Anatel and Procon, as well as the creation of the National Consumer Secretariat in 2012, during the Dilma government.

The longitudinal analysis, in this way, reveals the changes that occurred in the post-privatization telecommunications sector and the emphasis of the hegemonic media in issues that arose in each period. Are there, however, other possible emphases that were not tackled by the hegemonic? The next analysis deals with this question, comparing the lexical field of hegemonic media to that of pro and anti-privatization ones.

4.3 Comparative Media Analysis (Hegemonic, Pro-privatization and Anti-privatization)

The analysis comparing the 3 corpora - hegemonic, pro-privatization and anti-privatization - considered the first 100 keywords of each group for the period from 2007 to 2018. We also categorized the keywords from each corpus to facilitate the analysis, ending up with the same categories (only adding the “Customer” category) of previous analyses, indicating the categories’ adherence to the field.

Table 7
Categorization of the first 100 keywords by media type (2007 to 2018)

Categories	Absolute numbers						Participation percentages (%)
	Anti		Hegemonic		Pro		Anti		Hegemonic		Pro
	Types	Freq.	Types	Freq.	Types	Freq.	Types	Freq.	Types	Freq.	Types	Freq.
Company	4	1,662	19	9,117	5	1,546	0.4	1.1	1.9	5.5	0.6	1.7
Service	19	5,646	12	4,425	7	1,614	1.9	3.7	1.2	2.7	0.8	1.8
Privatization process	3	706	6	1,620	3	628	0.3	0.5	0.6	1.0	0.3	0.7
Government	7	2,783	6	1,971	6	2,056	0.7	1.8	0.6	1.2	0.7	2.3
Size	2	539	5	2,962	2	224	0.2	0.3	0.5	1.8	0.2	0.2
Property	6	1,509	5	1,202	12	1,951	0.6	1.0	0.5	0.7	1.3	2.1
Regulation / legislation	10	2,896	4	1,071	1	95	1.0	1.9	0.4	0.6	0.1	0.1
Individual	3	742	4	1,494			0.3	0.5	0.4	0.9	0.0	0.0
Economy / Business	4	836	3	1,089	19	3,540	0.4	0.5	0.3	0.7	2.1	3.9
Country / Region	5	2,470	2	993	3	713	0.5	1.6	0.2	0.6	0.3	0.8
Client	1	180	2	390	3	425	0.1	0.1	0.2	0.2	0.3	0.5
Others	5	976			8	1,043	0.5	0.6	0.0	0.0	0.9	1.1
Function words	31	63,692	32	56,237	31	26,840	3.1	4.3	3.2	33.9	3.5	29.6
Total 100 Keywords (a)	100	84,637	100	82,571	100	40,675	10	55	10	50	11	45
Total Keywords (b)	1.011	154,218	1.012	166,007	893	90,807	100	100	100	100	100	100

• Source: Elaborated by the authors.

Table 7 shows the distribution of the categories considering the first 100 keywords of each corpus, both in absolute numbers and in percentages of each category participation in the total of words. The percentages indicated in row (a) refers to the participation of these 100 keywords in the sum of all keywords (b).

Regarding the share of each category in the total of keywords, a very simple analysis reveals that the hegemonic media seems to give more attention both to companies (category “Company”) and to the privatization transaction amounts (category “Size”). The anti-privatization media, in turn, seems to place more emphasis on regulatory and legislative issues (category “Regulation/Legislation”), while the pro-privatization media emphasized ownership and economic/business issues. This analysis corroborates those of Scherling’s (2014), who analyzed the ways in which the term “privatization” is linguistically constructed in texts by pro and anti-privatization groups. In his preliminary results, the pro-privatization discourse would include categories related to economics, to quantification, to individuality and to facts and arguments. These categories reveal that the “success” of privatizations is measured regarding the impacts on the economy, and not on the population, privileging individual choice and freedom of choice. In contrast, anti-privatization discourses would emphasize cooperation and collectivity, claiming for individual rights and concerning for the impacts on employment and wage.

To confirm these results, we compared the corpora pro and anti-privatization in relation to the hegemonic, placing the latter as control corpus. Table 8 shows the first 25 keywords statistically more frequent in anti and pro media than in hegemonic media. For this analysis, we removed the function words.

Table 8
Anti and pro-privatization media keywords from 2007 to 2018 - hegemonic media control corpus

N.	ANTI-PRIVATIZATION			PRO-PRIVATIZATION
	Freq	Keyness	Keyword	Freq	Keyness	Keyword
1	580	393,15	communication	330	531,79	property
2	403	217,49	civil	316	481,44	free
3	674	179,5	law	359	326,82	prices
4	235	173,47	means	285	276,23	competition
5	228	162,85	media	531	192,21	state
6	338	140,82	society	90	132,51	theory
7	214	124,28	freedom	158	130,91	production
8	165	119,48	constitution	158	121,44	taxes
9	388	108,75	access	150	119,59	privatizations
10	274	105,85	Milestone	119	118,62	monopoly
11	459	99,1	public	109	118,22	post-office
12	197	97,06	interests	73	115,24	arrangement
13	141	94,04	expression	149	114,49	private
14	156	93,61	debate	127	110,45	freedom
15	151	88,65	people	173	108,33	assets
16	448	88,33	project	238	108,25	state-owned (company)
17	280	88,03	rights	69	107,84	governmental
18	73	85,36	democratization	907	106,48	government
19	148	83,34	construction	56	100,51	regulations
20	299	80,96	public	614	93,03	marketplace
21	119	79,47	popular	206	87,43	state-owned (companies)
22	67	77,77	law (article)	48	85,15	individual
23	707	77,28	internet	66	79,45	regulators
24	183	77,06	regulation	95	76,49	agencies
25	334	73,59	power	51	74,94	owner

• Source: Elaborated by the authors.

Confirming the previous analysis, the words of the anti-privatization media keywords reveal a greater emphasis on regulatory and legislative issues (projected law, civil milestone, constitution, article of law, regulation), on collectivity (society, public, debate, people, democratization, public, popular) and on rights (freedom of expression, rights). Therefore, the concern of anti-privatization media articles seems to foster a debate on the rights of the population regarding the telecom privatization, with emphasis on regulation and legislation in the telecommunications sector.

The keywords for the pro-privatization groups, in turn, revolve around economic terms, such as “property”, “prices”, “free competition”, “taxes”, “monopoly”, “freedom”, “production” and even “theory”, uncovering an educational profile of these media by bringing economic concepts as a defense of a political agenda. The role of the State and of property is widely discussed, as the words “state”, “governmental”, “government” and “private” show. Like the anti-privatization, the pro-privatization discourse also emphasizes regulatory matters and the role of regulatory agencies, albeit not bringing to the forefront the merit of the law and the constitution.

These results relate to the media discourses types identified by Vaara and Tienari (2002) on mergers and acquisitions: the rationalistic, the cultural, the societal and the individualistic. Pro media would fit the rationalistic discourse, bringing the economics theory to rationally explain the need for privatization; the anti-media would emphasize a societal discourse, concerned about the consequences of privatizations on society. We couldn’t find in our analyses any cultural discourse on privatizations, even though it would have been a possible discourse, since several foreign companies entered the Brazilian telecom market, and cultural organizational differences could emerge with the acquisition of state-owned companies by private ones. Individualist discourse, which simplifies the phenomenon by means of telling stories about certain individuals, could slightly be observed in the hegemonic media, which brought to attention the action of individuals with decision-making power on telecommunications - like the banker Daniel Dantas, the Ministers of Communications and the President of Brazil.

4.4 Collocation Analysis

The last corpus analysis was the collocation. Table 9 shows the top 25 collocates for the node “privatization”. This node has in total 114 collocations with a frequency of at least 10 and, in total, these collocates appear 6,530 times in the hegemonic corpus. The frequency column indicates the number of times that the collocate appears next to the node (privatization); the freq column. (L) indicates the number of times the collocate appears to the left of the node, and the freq column. (R), the number of times it appears on the right. The stat column indicates the statistics of each collocate.

Table 9
Collocates for Privatization (5L and 5R)

#Total No. of Collocate Types: 114 - PRIVATIZATION
#Total No. of Collocate Tokens: 6530
N.	Freq	Freq (L)	Freq (R)	Stat (MI)	Collocate
1	19	2	17	772,383	telebras
2	226	9	217	767,081	telebrás
3	108	4	104	706,729	system
4	10	9	1	694,983	post
5	76	70	6	655,711	since
6	64	57	7	642,976	post
7	68	66	2	627,944	process
8	34	30	4	627,648	model
9	40	11	29	622,252	teles (telecom companies)
10	10	9	1	613,791	Auctions
11	25	23	2	607,711	period
12	12	12	0	600,428	participate
13	75	73	2	600,206	auction
14	18	15	3	589,737	program
15	32	28	4	561,593	before
16	10	10	0	560,880	processes
17	22	5	17	560,765	state-owned companies
18	41	37	4	551,153	after
19	18	3	15	547,590	july
20	10	3	7	544,169	started
21	21	5	16	531,732	State-owned company
22	17	15	2	528,836	during
23	55	31	24	508,894	year
24	15	11	4	508,413	from
25	58	17	41	507,536	sector

• Source: Elaborated by the authors.

As one would expect, since it is a corpus on the privatization of Brazilian telecommunications, the most frequent candidates are “Telebrás” (the previous telecom holding company), “system” (for Sistema Telebrás), “process”, “model”, “teles (telecom companies)”, “auctions”, “state-owned”, “July” (month of privatization) and “sector”. What stands out, however, are words that mark time, such as “post”, “since”, “after”, “season”, “before”, “after”, “during”, “years” and “from”.

Analyzing the collocates that mark time (since, after, season, before, after, during) through the generation of 254 concordance lines, we may conclude that privatization has been represented in the hegemonic media as a historical landmark, to which all subsequent events in the sector will be compared. Here we connect with Pachi’s (2008a, 2008b) findings on media discourse in the privatization years, when privatization was then-represented as a watershed event that divided the new and the old, a chaotic situation from a well-being one.

Some representations, in fact, show the “arrival of the new and the modern” with privatization:

Before the telecommunications privatization, in 1998, the Brazilian consumer paid dearly for a telephone line. The telephone, unlike today, was an asset. There was even a parallel market for telephone lines. With the sale of telephone companies to the private sector, prices fell and access to the service significantly increased. At the time, it was estimated that the pent-up demand was 13.3 million handsets. Less than a third of the households had a telephone (Estadão, May 2011).

Only a few articles that take stock of privatizations (5, 10 and 20 years after), however, discuss the improvements and challenges for the sector. Much of the articles that establish privatization as a milestone do so to situate big businesses or to give a temporal notion about existing relationships in the industry, like the examples below:

The incorporation of Vivo is the largest business in the sector since privatization. With this operation, Telefónica is adding to its capital BRL 31.1 billion, market value of Vivo shares. In total, the group will have 77 million customers, of which 62 million will come from Vivo. It is more than Oi, now vice-leader (Folha de S.Paulo, April 2011).

On December 18, the National Telecommunications Agency (Anatel) approved the acquisition of Brasil Telecom (BrT) by Oi, considered the largest business in the sector in the last ten years, after the privatization of the Telebrás System. The supertelecom company that resulted from the merger of Oi and Brasil Telecom was born with annual gross revenue of BRL 41 billion, concentrates 22 million wire line phones and almost 30 million cell phones (Estadão, December 2008).

Daniel Dantas, a 53-year-old born in Bahia, has been in the midst of controversial issues since the privatization of Telebrás in 1998. Months later, wiretaps arose that caught the then Minister of Communications Luiz Carlos Mendonça de Barros articulating benefits to Opportunity in the privatization process- the “BNDES telephone tap” episode (O Globo, July 2008).

The concordance analysis shows that the privatization of telecommunications is still seen as a major milestone in the sector, serving as a measure for all subsequent events. Its use as a founding event varies, sometimes positively and sometimes negatively, although the emphasis is almost always on the businesses generated by it and on the companies involved.

5. CONCLUSIONS

What can a corpus linguistics analysis tell us about media representations on the privatization of the Brazilian telecom sector? By using the linguistic exploration methodology of large corpora to analyze the hegemonic media privatization discourse, we visualize how the media restricts the matter to the field of business, of companies and their relationship with the government, with little participation of matters regarding users, customers and, mainly, the Brazilian population. This finding opposes the alleged “neutrality” and “plurality” of the analyzed media, something relevant especially if we take into account the role of opinion maker these media play.

The ways in which this type of media discursively constructs the theme of privatization end up ratifying the idea that telecommunications are no longer a public service and, therefore, of State responsibility. To equate telecommunications to other services / consumer goods is a discursive and not neutral construction, since privatization has not always been a consensus, as history shows us. At the time of privatization, Fadul (1999, pp. 76-77) raised questions that deserve reflection

about the role and the place that public services should occupy in the economy and contemporary society for which they exist, as well as about the State’s mission to ensure the fulfillment of certain primary functions and the defense of certain social values, together with the promotion of economic and social progress.

Twenty-two years after the Brazilian telecom sector privatization, we can assume that the texts of the hegemonic media played a role in socially constructing the place of services such as telecommunications in the private arena. By restricting the lexical field of privatization to companies and businesses, the hegemonic media transforms and reproduces the discourse that telecommunications is a business like any other, mischaracterizing its categorization as a public service and making its discussion and debate with society unnecessary. This discursive construction is ratified when confronting the discourse of the hegemonic media with that of anti and pro-privatization media. The first, with a well-defined societal discourse, makes us question about types of discourses that can’t be found (or can hardly be found) in the hegemonic media, especially with regards to regulatory issues involving the participation of civil society. In the societal discourse, there is a greater concern on the effects of the telecom services privatization on society. The pro-privatization media, in turn, suggests another way of looking at the phenomenon, clearly with greater emphasis on the individual and his/her freedom (and possibility) of options when choosing a telecom service, and on the way the government regulates the sector so that these choices may exist.

In analyzing these other discourses, we again question the very idea of “plurality” in the hegemonic media, since the interests/points of view of the “consumers/citizens” have been neglected during these 20 years after the privatization of the sector. When building an image of privatizations that privileges the point of view of companies, and not that of the population, the media normalizes the government’s privatization agenda for state-owned companies, limiting the debate and the participation of society. These results corroborate those by Leal (2001), in her analysis of the (favorable) media discourse in relation to privatizations.

It is unquestionable, however, the establishment of privatization as a milestone for the sector. At the time of privatization, it made sense for the media to build an image of the future to legitimize the denationalization of these companies, as argued by Pachi (2008a, 2008b, 2015). However, developments following privatization also reflected political changes in the country, that complicated the relations between the government and businesses and that, consequently, reflected on the representation of the hegemonic media about the sector.

Future research may use this same corpus to identify changes in patterns of representation as government changes. The alternation of governments from different ideological positions may have altered the way in which the media represented privatization, in a more critical or more effusive way, according to their ideological predilections.

One of the main contributions of this study is methodological, bringing corpus linguistics to organizational studies and public administration. Critical studies on organizational phenomena are often qualitative, using methods such as CDA to unveil ideologies behind organizational discourses. The use of large textual collections may show the cumulative way in which these discourses are being propagated and reproduced not only by the media, but also by organizational communication, by government bodies, by political parties, among others. In addition, throughout the study, we used tools (AntConc and BootCat) and resources (the control corpus) of public use and free of charge. In this way, we believe to contribute to the dissemination of this type of approach.

The study also shows how the media must be included as part of the public and organizational arena, in which discourses on phenomena that affect society are legitimized and reproduced within it. If on the one hand the results are not new, on the other hand this same predictability signals confidence in the methodology used, which may be used on other issues at the organizational level. Much of the research in the administration area is based on textual documents, such as reports, speeches, and laws. The possibility of analyzing them from another point of view, both quantitative and qualitative, must be able to raise new questions as well as answer classic ones in novel ways.

References

• Almeida, M. P. (2013). A privatização estruturada: o PND no Governo Collor (1990-1992). Outros Tempos-Pesquisa em Foco-História, 10(16), 65-87.
• Almeida, M. P. (2014). A reestruturação do setor público estatal brasileiro rumo às privatizações dos anos 1990. Anos 90, 21(39), 363-397.
• Anthony, L. (2013). A critical look at software tools in corpus linguistics. Linguistic Research, 30(2), 141-161.
• Anthony, L. (2016). AntConc (Version 3.5.7) [Windows]. Tokyo, Japan: Waseda University. Recuperado de http://www.laurenceanthony.net/software/antconc/
  » http://www.laurenceanthony.net/software/antconc/
• Araújo, J. F. F. E; & Silvestre, H. C. (2014). As parcerias público-privadas para o desenvolvimento de infraestrutura rodoviária: experiência recente em Portugal. Revista de Administração Pública, 48(3), 571-593.
• Bachiller, P. (2017). A meta-analysis of the impact of privatization on firm performance. Management Decision, 55(1), 178-202.
• Baker, P. (2010). Sociolinguistics and corpus linguistics Edinburgh, Scotland: Edinburgh University Press.
• Baker, P; Gabrielatos, C; Khosravinik, M; Krzyżanowski, M; McEnery, T; … Wodak, R. (2008). A useful methodological synergy? Combining critical discourse analysis and corpus linguistics to examine discourses of refugees and asylum seekers in the UK press. Discourse & Society, 19(3), 273-306.
• Baker, P; Gabrielatos, C; & McEnery, T. (2012). Sketching Muslims: A corpus driven analysis of representations around the word ‘Muslim’in the British press 1998-2009. Applied Linguistics, 34(3), 255-278.
• Bardin, L. (1977). Análise de Conteúdo Lisboa, Portugal: Edições 70.
• Banco Nacional de Desenvolvimento Econômico e Social. (1992-2015). Programa Nacional de Desestatização: relatório de atividades Rio de Janeiro, RJ: Autor.
• Banco Nacional de Desenvolvimento Econômico e Social. (2020). Histórico/Processos Encerrados/Desestatização/Transparência Recuperado dehttps://www.bndes.gov.br/wps/portal/site/home/transparencia/desestatizacao/processos-encerrados/Historico
  » https://www.bndes.gov.br/wps/portal/site/home/transparencia/desestatizacao/processos-encerrados/Historico
• Bresser Pereira, L. C. (1998). A reforma do Estado dos anos 90: lógica e mecanismos de controle. Lua Nova: Revista de cultura e política, 45, 49-95.
• Capelato, M. H. (2014). História do tempo presente: a grande imprensa como fonte e objeto de estudo. In L. A. N. Delgado, & M. M. Ferreira (Org.), História do Tempo Presente Rio de Janeiro, RJ: FGV Editora.
• Cardoso, V. I. D. C; Maia, A. B. G. R; Santos, S. M. D; & Assis Soares, F. (2013). O impacto da privatização no desempenho econômico: um estudo em empresas brasileiras de grande porte. Revista Ibero Americana de Estratégia, 12(3), 183-211.
• Carrieri, A. D. P. (2002). A transformação das identidades uma empresa de telecomunicações antes e depois de sua privatização: um estudo de metáforas. Organizações & Sociedade, 9(23), 1-29.
• Castro, C. S. D. (2005). Gestão de recursos humanos no setor de telecomunicações: novas empresas, novas práticas? Revista de Administração de Empresas, 45(SPE), 36-47.
• Cherchiglia, M. L; & Dallari, S. G. (2003). A reforma setorial do Estado brasileiro na década de 1990: eficiência e privatização? Um estudo num hospital público universitário. Revista de Administração Pública, 37(2), 313-334.
• Clifton, J; Lanthier, P; & Schröter, H. (2011). Regulating and deregulating the public utilities 1830-2010. Business History, 53(5), 659-672.
• Costa, F. L. D. (2008). Brasil: 200 anos de Estado; 200 anos de administração pública; 200 anos de reformas. Revista de Administração Pública, 42(5), 829-874.
• Costa, F. L. D. (2009). Bases Teóricas e conceituais da reforma dos anos 1990: crítica do paradigma gerencialista. Revista Brasileira de Administração Política, 2(2), 79-103.
• Costa, F. L. D. & Peci, A. (1999). Desestatização como estratégia de reforma do Estado: Análise preliminar da privatização no Brasil da Nova República. Revista de Administração Pública, 33(3), 191-205.
• Couto, C. G; & Abrucio, F. (2003). O segundo governo FHC: coalizões, agendas e instituições. Tempo Social, 15(2), 269-301.
• Doolin, B. (2002). Enterprise discourse, professional identity and the organizational control of hospital clinicians. Organization studies, 23(3), 369-390.
• Du Gay, P; Salaman, G; & Rees, B. (1996). The conduct of management and the management of conduct: Contemporary managerial discourse and the constitution of the ‘competent’manager. Journal of Management Studies, 33(3), 263-282.
• Elliott, C; & Stead, V. (2018). Constructing women’s leadership representation in the UK press during a time of financial crisis: Gender capitals and dialectical tensions. Organization Studies, 39(1), 19-45.
• Esposito, I. R. (2017, agosto 23). Governo anuncia 57 projetos de concessões e privatização de empresas públicas. Agência Brasil Recuperado de https://agenciabrasil.ebc.com.br/economia/noticia/2017-08/governo-anuncia-57-projetos-de-privatizacao-de-empresas-publicas-e
  » https://agenciabrasil.ebc.com.br/economia/noticia/2017-08/governo-anuncia-57-projetos-de-privatizacao-de-empresas-publicas-e
• Fadul, É. M. C. (1999). Reforma do Estado e serviços públicos: transformação de um modelo ou adaptação a uma nova ordem social? Revista de Administração, 34(1), 70-78.
• Fairclough, N. (2003). Analysing discourse: Textual analysis for social research London, UK: Routledge.
• Faria, L. D. O; & Fischer, T. (2001). Privatização, mudança & evolução da estrutura organizacional em três momentos: Tebasa, Telebahia e Telemar. Organizações & Sociedade, 8(21), 59-75.
• Fascina, A. (2002). De Alexandre Graham Bell à Anatel. Pensamento & Realidade, 5(11), 48-61.
• Fernandes, A; & Tomazelli, I. (2019, novembro 01). Governo busca acelerar processos de privatização. Estadão Recuperado de https://economia.estadao.com.br/noticias/geral,governo-busca-acelerar-processos-de-privatizacao,70003072050
• Ferreira, E. C. D. V; Fandiño, A. M; Segre, L; & Nascimento, R. P. (2010). Privatização e mudanças organizacionais: um estudo de caso sobre transformação cultural e comportamental. Cadernos EBAPE.BR, 8(3), 468-491.
• Filardi, F; Leite, A. L. D. S; & Torres, A. A. G. (2014). Análise de resultados de indicadores de gestão e de regulação após a privatização: estudo de caso da Light Serviços de Eletricidade. Revista de Administração, 49(1), 18-32.
• Freitas, C. (2017). Estudos linguísticos e Humanidades digitais: corpus e descorporificação. Gragoatá, 22(44), 1207-1227.
• Freitas, C; Biar, L; & Martins, F. (No prelo). Construções do feminino na literatura brasileira: discurso, grandes corpora e leitura não-linear. Revista Brasileira de Linguística Aplicada
• Gurgel, C. R. M. (2017). Para Além da Reforma do Estado-crises e limites da razão liberal. Administração Pública e Gestão Social, 9(3), 159-170.
• Kogut, B; & Macpherson, J. M. (2011). The mobility of economists and the diffusion of policy ideas: The influence of economics on national policies. Research Policy, 40(10), 1307-1320.
• Kornelakis, A. (2015). European market integration and the political economy of corporate adjustment: OTE and Telecom Italia, 1949-2009. Business History, 57(6), 885-902.
• Kroeze, R; & Keulen, S. (2014). The managers’ moment in Western politics: The popularization of management and its effects in the 1980s and 1990s. Management & Organizational History, 9(4), 394-413.
• Lacerda, N. (2020, abril 08). Estatal na lista de privatizações de Bolsonaro é essencial para renda emergencial. Brasil de Fato Recuperado de https://www.brasildefato.com.br/2020/04/08/estatal-da-lista-de-privatizacoes-de-bolsonaro-e-essencial-para-renda-emergencial
  » https://www.brasildefato.com.br/2020/04/08/estatal-da-lista-de-privatizacoes-de-bolsonaro-e-essencial-para-renda-emergencial
• Leal, M. C. D. (2005). O discurso jornalístico sobre privatizações e protestos nas ruas. DELTA: Documentação de Estudos em Lingüística Teórica e Aplicada, 21(SPE), 73-92.
• Liu, H; Cutcher, L; & Grant, D. (2017). Authentic leadership in context: An analysis of banking CEO narratives during the global financial crisis. Human Relations, 70(6), 694-724.
• Luz, T. R. (2002). Telemar-Minas: a reinvenção dos gestores. Revista Economia & Gestão, 2(3), 56-73.
• Luz, T. R. (2007). Valores do Livre Mercado e seus Impactos no Trabalho Gerencial: O Caso Telemar-Minas. Gestão e Sociedade, 1(1), 1-22.
• Macpherson, J. M. (2006). Palace wars and privatization: Did Chicago Beat Cambridge in influencing economic policies. European Management Review, 3(3), 190-198.
• Maculan, A. M; & Legey, L. R. (1996). As experiências internacionais de regulação para as telecomunicações e a reestruturação dos serviços no Brasil. Revista de Economia Política, 16(4), 67-86.
• McCarthy, D. C; Reeves, E; & Turner, T. (2011). Changing the Rules of the Game: The impact of privatization on firm industrial relations. Public Management Review, 13(6), 845-860.
• McEnery, T; & Hardie, A. (2011). Corpus linguistics: Method, theory and practice Cambridge, UK: Cambridge University Press.
• Misoczky, M. C; Abdala, P. R. Z; & Damboriarena, L. (2017). A trajetória ininterrupta da reforma do aparelho de Estado no Brasil: continuidades nos marcos do neoliberalismo e do gerencialismo. Administração Pública e Gestão Social, 9(3), 184-193.
• Moretti, F. (2008). Distant reading New York, NY: Verso Books.
• Oliva, E. D. C. (2002). A privatização de empresas siderúrgicas brasileiras: reflexos na gestão de recursos humanos. Revista de Administração Contemporânea, 6(1), 141-161.
• Onuma, F. M. S; Zwick, E; & Brito, M. J. (2015). Ideologia gerencialista, poder e gestão de pessoas na administração pública e privada: uma interpretação sob a ótica da análise crítica do discurso. Revista de Ciências da Administração, 17(42), 106-120.
• Pachi, F. F F. (2008a). Privatizar para tornar” público”: uma análise do discurso sobre a privatização das telecomunicações em jornais (Tese de Doutorado). Universidade Estadual de Campinas, Campinas, Brasil.
• Pachi, F. F F. (2008b). O passado e a necessidade de ruptura no discurso da imprensa sobre a privatização das telecomunicações. Revista de Letras, 10, 1-19.
• Pachi, F. F F. (2015). A construção do futuro no discurso sobre a privatização das telecomunicações na imprensa. Línguas & Letras, 16(32), 30-50.
• Palassi, M. P. (2002). Privatização e enfrentamento: elementos integradores e desintegradores da resistência coletiva. Revista de Administração Pública, 36(3), 433-461.
• Palassi, M. P; Carrieri, A. D. P; Leite-da-Silva, A. R; & Pimentel, T. D. (2007). Os discursos e as metáforas na análise da propensão à participação: um estudo em um banco privatizado. Revista de Administração Pública, 41(3), 443-473.
• Paula, A. P. P. D. (2005). Administração pública brasileira entre o gerencialismo e a gestão social. Revista de Administração de Empresas, 45(1), 36-49.
• Pearce, M. (2014). Key function words in a corpus of UK election manifestos. Linguistik online, 65 (3), 23-44.
• Pina, V; Torres, L; & Bachiller, P. (2014). Service quality in utility industries: the European telecommunications sector. Managing Service Quality: An International Journal, 24(1), 2-22.
• Poder 360. (2018, janeiro 31). Tiragem impressa dos maiores jornais perde 520 mil exemplares em 3 anos Recuperado de https://www.poder360.com.br/midia/tiragem-impressa-dos-maiores-jornais-perde-520-mil-exemplares-em-3-anos/
• Ramamurti, R. (1992). Why are developing countries privatizing?. Journal of International Business Studies, 23(2), 225-249.
• Resende, S. (2019, novembro 06). Alcolumbre diz que resistência à privatização da Eletrobras é ‘muito grande’ no Senado. G1 Recuperado de https://g1.globo.com/politica/noticia/2019/11/06/alcolumbre-diz-que-resistencia-a-privatizacao-da-eletrobras-e-muito-grande-no-senado.ghtml
  » https://g1.globo.com/politica/noticia/2019/11/06/alcolumbre-diz-que-resistencia-a-privatizacao-da-eletrobras-e-muito-grande-no-senado.ghtml
• Revista Época. (2019, novembro 08). Se não privatizar a Eletrobras, será um crime contra o governo brasileiro’, diz secretário Recuperado dehttps://epocanegocios.globo.com/Brasil/noticia/2019/11/se-nao-privatizar-eletrobras-sera-um-crime-contra-o-governo-brasileiro-diz-secretario.html
  » https://epocanegocios.globo.com/Brasil/noticia/2019/11/se-nao-privatizar-eletrobras-sera-um-crime-contra-o-governo-brasileiro-diz-secretario.html
• Riad, S; Vaara, E; & Zhang, N. (2012). The intertextual production of international relations in mergers and acquisitions. Organization Studies, 33(1), 121-148.
• Rosa, D. D. S F. (2011). Poder simbólico no reajuste tarifário do transporte metropolitano de passageiros de Porto Alegre na crise de 2004. Revista de Administração Pública, 45(3), 643-668.
• Santos, C. A. S; & Costa, A. S. M. (2019). Imprensa, Discurso Ideológico e Golpe de Estado: uma Análise Crítica do Discurso. Revista Eletrônica de Ciência Administrativa, 18(3), 371-393.
• Scherling, J. (2014). (Neo)Liberalizing the state: Privatization of core government social competences - A CDA approach. In Anais do CADAAD 2014, ELTE University, Budapest, Hungary.
• Scriptore, J. S; & Toneto, R J. (2012). A estrutura de provisão dos serviços de saneamento básico no Brasil: uma análise comparativa do desempenho dos provedores públicos e privados. Revista de Administração Pública, 46(6), 1479-1504.
• Silva, J. C. (2002). A reestruturação produtiva no setor bancário em Santa Catarina. Revista de Ciências da Administração, 4(8), 1-13.
• Silvestre, B. D. S; Hall, J; Matos, S; & Figueira, L. A. P. D. A. (2010). Privatização: bom ou ruim? Lições do setor de distribuição de energia elétrica do nordeste brasileiro. Revista de Administração de Empresas, 50(1), 94-111.
• Siqueira, M. V. S; & Mendes, A. M. (2009). Gestão de pessoas no setor público e a reprodução do discurso do setor privado. Revista do Serviço Público, 60(3), 241-250.
• Smith, A. M. (2000). Um acordo forçado: o consentimento da imprensa à censura no Brasil Rio de Janeiro, RJ: Editora FGV.
• Strom, M; & Alcock, E. (2017). Floods, waves, and surges: the representation of Latin@ immigrant children in the United States mainstream media. Critical Discourse Studies, 14(4), 440-457.
• Vaara, E; & Monin, P. (2010). A recursive perspective on discursive legitimation and organizational action in mergers and acquisitions. Organization Science, 21(1), 3-22.
• Vaara, E; & Tienari, J. (2002). Justification, legitimization and naturalization of mergers and acquisitions: A critical discourse analysis of media texts. Organization, 9(2), 275-304.
• Vaara, E; & Tienari, J. (2008). A discursive perspective on legitimation strategies in multinational corporations. Academy of Management Review, 33(4), 985-993.
• Vaara, E; Tienari, J; & Laurila, J. (2006). Pulp and paper fiction: On the discursive legitimation of global industrial restructuring. Organization Studies, 27(6), 789-813.
• Veloso, E; & Trevisan, L. (2010). Estratégias Organizacionais e Processos de Fusão: Estudo de Caso Sobre Mudanças de Estrutura em Agências Bancárias. Revista Administração em Diálogo, 12(3), 60-79.
• Wallmeier, F; Helmig, B; & Feeney, M. K. (2019). Knowledge construction in public administration: A discourse analysis of public value. Public Administration Review, 79(4), 488-499.
• Zanchetta, E; Baroni, M; & Bernardini, S. (2011, julho). Corpora for the masses: the BootCaT front-end. In Pecha Kucha Presented at the Corpus Linguistics 2011 Conference, University of Birmingham, Birmingham, UK.
• Zilber, M. A; Lex, S; & Ades, C. (2005). As privatizações e o novo modelo do setor elétrico brasileiro: o impacto sobre o atendimento das necessidades do consumidor. Gestão & Regionalidade, 21(61), 51-63.

• 1
  Asterisks indicate that the word may contain any type of termination. So, for example, “telecom *” includes “telecom”, “telecommunication” and “telecommunications”.

• [Translated version] Note: All quotes in English translated by this article’s translator.

Publication Dates

• Publication in this collection
  09 July 2021
• Date of issue
  May-Jun 2021

History

• Received
  10 Jan 2020
• Accepted
  21 Sept 2020